eLife Assessment

This study examines the role of the fungal pathogen Candida albicans in the progression of colorectal cancer, a relevant and urgent topic given the global incidence of colon cancer. While the findings are useful and provide solid experimental work and insight into how Candida may contribute to tumor progression, the small patient sample size, reliance on in vitro models, and absence of in vivo validation may limit its impact. This work will interest scientists studying cancer progression and the role played by pathogens.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers. The authors have addressed most of the comments raised in the previous round of review.]

Summary:

This study addresses the emerging role of fungal pathogens in colorectal cancer and provides mechanistic insights into how Candida albicans may influence tumor-promoting pathways. While the work is potentially impactful and the experiments are carefully executed, the strength of evidence is limited by reliance on in vitro models, small patient sample size, and the absence of in vivo validation, which reduces the translational significance of the findings.

Strengths:

(1) Comprehensive mechanistic dissection of intracellular signaling pathways.

(2) Broad use of pharmacological inhibitors and cell line models.

(3) Inclusion of patient-derived organoids, which increases relevance to human disease.

(4) Focus on an emerging and underexplored aspect of the tumor microenvironment, namely fungal pathogens.

Reviewer #2 (Public review):

The authors in this manuscript studied the role of Candida albicans in Colorectal cancer progression. The authors have undertaken a thorough investigation and used several methods to investigate the role of Candida albicans in Colorectal cancer progression. The topic is highly relevant, given the increasing burden of colon cancer globally and the urgent need for innovative treatment options.

Strengths:

Authors have undertaken a thorough investigation and used several methods to investigate the role of Candida albicans in Colorectal cancer progression.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study addresses the emerging role of fungal pathogens in colorectal cancer and provides mechanistic insights into how Candida albicans may influence tumor-promoting pathways. While the work is potentially impactful and the experiments are carefully executed, the strength of evidence is limited by reliance on in vitro models, small patient sample size, and the absence of in vivo validation, which reduces the translational significance of the findings.

Strengths:

(1) Comprehensive mechanistic dissection of intracellular signaling pathways.

(2) Broad use of pharmacological inhibitors and cell line models.

(3) Inclusion of patient-derived organoids, which increases relevance to human disease.

(4) Focus on an emerging and underexplored aspect of the tumor microenvironment, namely fungal pathogens.

Weaknesses:

(1) Clinical association data are inconsistent and based on very small sample numbers.

We thank the reviewer for this important comment. We have investigated 4 colorectal tumors (2 in early stage and 2 in late stage), and we observed Candida albicans in the 2 late-stage samples while not in the early-stage ones. This result is consistent with TCGA data (which is large-scale) that Candida albicans mainly detectable in the late-stage colorectal tumors (Fig. 1c) and suggests that Candida albicans contributed to colorectal cancer progression, which is the main research direction of this study.

(2) No in vivo validation, which limits the translational significance.

We appreciate the reviewer’s concern regarding the lack of in vivo validation. While we recognize the value of in vivo models, our current institutional biosafety protocols and animal facility designations do not support the handling of pathogenic microorganisms like Candida albicans in live infection models. Consequently, these experiments were beyond the immediate technical scope of this study. To validate the findings using cell lines, we have performed Candida albicans infection experiments using organoids collected from colorectal cancer patients instead (Fig. 7). We have revised the Discussion section to acknowledge this limitation and clarify that the current work serves as a mechanistic study based on in vitro and ex vivo systems. We have also incorporated references to recent studies demonstrating the in vivo effects of C. albicans in tumor models, which support the biological relevance of our findings.

(3) Species- and cell type-specificity claims are not well supported by the presented controls.

We thank the reviewer for this insightful comment. We agree that our current dataset does not warrant definitive conclusions regarding species- or cell type-specificity. Accordingly, we have tempered our claims throughout the manuscript, describing the observed effects as context-dependent across different epithelial models. Specifically, we observed differential responses among the cell lines and epithelial systems evaluated, suggesting variability rather than strict specificity. Furthermore, the Discussion has been expanded to address potential underlying factors, such as variations in EGFR expression levels and other signaling determinants. We have also added a dedicated section to acknowledge this limitation and emphasize the need for future systematic investigations using a more diverse array of fungal species and cell models.

(4) Reliance on colorectal cancer cell lines alone makes it difficult to judge whether findings are specific or general epithelial responses.

We appreciate the reviewer’s thoughtful concern. Although most of our mechanistic experiments were performed in colorectal cancer cell lines, we also evaluated our finding across a broader range of epithelial models, including normal human colon-derived organoids and the breast epithelial cancer line MCF7 (Fig. 8). Neither model exhibited HIF-1α activation upon C. albicans exposure, supporting that the hypoxia response we observed might not be universal. Interestingly, the observed response in non-colorectal epithelial cancer lines (e.g., HCC1937, NUGC-3) suggests that this mechanism is not strictly confined to CRC. Based on these observations, we propose that the specificity is likely related to EGFR levels but may involve additional epithelial determinants, which we aim to investigate in future work.

Reviewer #2 (Public review):

The authors in this manuscript studied the role of Candida albicans in Colorectal cancer progression. The authors have undertaken a thorough investigation and used several methods to investigate the role of Candida albicans in Colorectal cancer progression. The topic is highly relevant, given the increasing burden of colon cancer globally and the urgent need for innovative treatment options. However, there are some inconsistencies in the figures and some missing details in the figures, including:

(1) The authors should clearly explain in the results section which patient samples are shown in Figure 1B.

We thank the reviewer for pointing out this omission. We apologize for the lack of clarity in the initial submission. The patient samples shown in Figure 1B are from the CRC patients with Stage III. We have revised the manuscript to explicitly state this information in the legend for Figure 1B to ensure better clarity for the reader.

(2) What do a, ab, b, b written above the bars in Figure 1F represent? Maybe authors should consider removing them, because they create confusion. Also, there is no explanation for those letters in the figure legend.

We thank the reviewer for this helpful comment. The letters above the bars represent statistical groupings from post-hoc multiple-comparison tests (a standard convention used after ANOVA or similar analyses): bars sharing the same letter are not significantly different, whereas different letters indicate statistically distinct groups. We chose this letter-based system over asterisks to avoid the visual clutter and potential confusion that often arise from numerous pairwise comparisons; therefore, we will retain the letter-based grouping. In the revised manuscript, we have explicitly defined this notation in the figure legend to be ease of interpretation for the reader.

(3) The authors should submit all the raw images of Western blot with appropriate labels to indicate the bands of protein of interest along with molecular weight markers.

We appreciate the reviewer’s request for raw data. We have now included the raw images of the Western blots in the supplementary materials, with clear annotations of the bands corresponding to the proteins of interest as well as molecular weight markers.

(4) The authors should do the quantification of data in Figure 2d and include it in the figure.

We thank the reviewer for this valuable suggestion. In the revised manuscript, we have quantified the subcellular localization of HIF-1α in PBS-treated versus C. albicans–infected cells shown in Figure 2d. The quantification results are shown in the following figure and provided in Supplementary Figure 3c.

(5) In Figure 2h, the authors should indicate if the quantification represents VEGF expression after 6h or 12h of C. albicans co-culture with cells.

We thank the reviewer for pointing this out. We have updated Figure 2h to specify that the quantification represents VEGF expression after 12 hours of co-culture with Candida albicans.

(6) In Figure 2i, quantification of VEGF should be done and data from three independent experiments should be submitted. The authors should also mention the time point.

We thank the reviewer for this helpful comment. In the revised manuscript, we have quantified VEGFA fluorescence intensity based on three independent experiments (the other 2 replicates were shown in Author response image 1). The corresponding time point (12 hours of co-culture) has been clearly indicated in the figure legend.

Author response image 1.

Recommendations for the authors:

Reviewing Editor Comments:

(1) Some of the statements regarding Candida albicans and CRC progression in Figure 1 may be overstated (since the association with stage/survival may be cross-confounded). That is, analyses of survival ought to be stage-adjusted.

We thank the editor for this important comment. We agree that the association between C. albicans and patient survival may be influenced by tumor stage as a confounding factor. In the revised manuscript, we have moderated our statements regarding the clinical associations and clarified the limitations of these analyses, now presenting these findings as correlative observations rather than causal relationships. We have also noted in the Discussion that stage-adjusted analyses would be required to more rigorously assess the independent contribution of C. albicans to patient outcomes.

(2) Fan et al. (citation 26) is incorrectly referenced. The paper states that Bacteroides fragilis does not affect Candida albicans colonization. Instead, Bacteroides thetaiotamicron was shown to reduce C. albicans colonization, but B. fragilis was used in the current study as a control.

We thank the editor for pointing out this error, and we have corrected the citation accordingly. As noted, the referenced study indicates that Bacteroides thetaiotaomicron, rather than Bacteroides fragilis, reduces C. albicans colonization. We selected B. fragilis as a control in this study because it is a prevalent gut commensal and has been previously implicated in colorectal cancer progression. Although prior reports suggest that B. fragilis does not significantly affect C. albicans growth, we observed that co-culture with B. fragilis led to a noticeable inhibition of C. albicans growth under our experimental conditions. This discrepancy may reflect differences in experimental settings. We believe these findings provide additional context for the complex interactions between gut microbiota and fungal pathogens.

(3) The link between hypoxia signaling is interesting, but for the most part, these experiments were largely done in normoxic conditions, while the colon is generally hypoxic. So I would have encouraged the authors to consider testing the effect of C. albicans presence/absence under low-oxygen conditions, which may be more physiologically relevant.

We thank the editor for this insightful suggestion. We fully agree that evaluating the effects of C. albicans under hypoxic or anaerobic conditions would be highly relevant to the physiological tumor microenvironment. Although we have attempted to assess the impact of C. albicans on cell migration under hypoxic conditions, we observed that tumor cells exhibited markedly accelerated migration and proliferation, resulting in near-complete wound closure within 24 hours in control groups. This limited our ability to reliably detect differences between conditions using standard migration assays. We agree that in vivo models may provide a more physiologically relevant context to address this question, and we will pursue this direction in future studies when appropriate experimental conditions become available.

Reviewer #1 (Recommendations for the authors):

(1) Figure 1 inconsistencies: In Figure 1C, there is no significant difference in C. albicans detection between stage II and stage III CRC patients. In fact, more patients in stage II appear positive, which is inconsistent with Figures 1A and 1B. For Figures 1A and 1B, the sample size (n=2) is too low to support meaningful conclusions. Please also clarify which stage is represented in Figure 1B.

We thank the reviewer for this important comment. In the revised manuscript, we have clarified the sample information and explicitly stated that the samples shown in Figure 1b are derived from stage III CRC patients. We have also moderated our conclusions and described these findings as exploratory observations. Regarding the apparent inconsistency between Figure 1C and Figures 1a-b, we consider that this discrepancy may be partly due to the small number of clinical samples analyzed in our study. In addition, the TCGA-based analysis relies on transcriptomic data, whereas our analysis is based on immunohistochemistry (IHC). These methodological differences may also contribute to the observed variation.

(2) Weak link between clinical and in vitro data: The transition from clinical samples to CRC cell line models feels tenuous. While C. albicans may induce hypoxia signaling, it is unclear whether this is specific to CRC cells or could occur in other epithelial cell types. Some broader testing would help strengthen this link.

We thank the reviewer for this insightful comment. We agree that reinforcing the bridge between clinical observations and in vitro mechanistic findings, as well as clarifying cell type specificity, is important for a comprehensive study. In the revised manuscript, we have clarified that the clinical data provide correlative evidence, while the mechanistic insights are derived from controlled in vitro systems. To address the issue of cell type specificity, we have included additional analyses across multiple epithelial cell models (Figure 8). These results indicate that the response to C. albicans is not restricted to colorectal cancer cells but varies across different epithelial contexts.

(3) Lack of in vivo validation: The mechanistic findings would be substantially strengthened by in vivo data, e.g., murine CRC models. Without this, the translational impact is limited.

We appreciate the reviewer’s concern regarding the lack of in vivo validation. While we recognize the value of animal models, our current institutional biosafety protocols and facility designations do not support the handling of pathogenic microorganisms like Candida albicans in live infection models. Consequently, these experiments were beyond the immediate technical scope of this study, and better be performed in future studies to validate the mechanisms.

(4) Figure 8B interpretation: The authors conclude that C. albicans shows the strongest effect on c-Myc and c-Jun activation. However, from the presented blots, the differences compared to other fungi are not obvious. The claim should be toned down or quantified more rigorously.

We thank the reviewer for this important comment. We agree that the differences in c-Myc and c-Jun activation among fungal species are not sufficiently pronounced to support a strong comparative claim. In the revised manuscript, we have moderated the corresponding statements to avoid overinterpretation.

(5) Cell type specificity: Since the title emphasizes CRC specificity, the cell line comparison shown in Figure 8 should be moved earlier in the results. This would clarify from the start whether the described mechanisms are CRC-specific or more generalizable.

We thank the reviewer for this insightful suggestion. We agree that earlier presentation of cell type comparisons would help clarify the scope of the observed effects. We have revised the Results section accordingly: “To evaluate the cell type specificity of this response, we further analyzed additional epithelial cell models, as shown in Figure 8”.

In this study, we initially identified the effects of C. albicans in colorectal cancer (CRC) cells and therefore focused on establishing the underlying mechanisms in this context. Subsequently, we extended our analysis to additional epithelial cell types to evaluate whether these effects are shared or context-dependent. We believe that this stepwise organization, from detailed mechanistic investigation in CRC cells to broader comparison across cell types, provides a logical and coherent flow for the reader. In the revised manuscript, we have further clarified this rationale in the text to improve readability and interpretation.

(6) It would be good to use a negative fungi control instead of a PBS control for most of the experiment.

We thank the reviewer for this valuable suggestion. We agree that a negative fungal control would further strengthen the conclusions. Unfortunately, we were unable to incorporate additional controls during the revision, while we believe that our comparative analysis across multiple fungal species (Figure 8) partially addresses this issue by demonstrating differential signaling responses. Future studies will incorporate appropriate negative fungal controls to further validate the specificity of these effects.

(7) It is surprising that the Dectin-1 inhibitor shows a smaller effect compared with the TLR2 inhibitor. This result warrants further explanation, as Dectin-1 is a well-known receptor for C. albicans β-glucans. The authors should clarify whether this difference reflects cell type-specific expression (e.g., low Dectin-1 in CRC cells), ligand accessibility, or pathway redundancy, and discuss how it aligns with existing literature.

We thank the reviewer for this insightful comment. We agree that the relatively modest effect of Dectin-1 inhibition compared to TLR2 inhibition warrants further consideration. In the revised manuscript, we have expanded the Discussion to address this observation. We propose several possible explanations: Firstly, the expression level of Dectin-1 is relatively low in these epithelial cancer cells, thereby limiting its functional contribution. Secondly, differences in ligand accessibility, particularly in the context of fungal cell wall architecture, may influence receptor engagement. Finally, redundancy and cross-talk among pattern recognition receptor pathways compensate for Dectin-1 inhibition. These observations highlight the context-dependent nature of host–fungal interactions.

Reviewer #2 (Recommendations for the authors):

All my comments that need to be addressed are given above and below:

(1) What do a and b represent in Figure 2f? They should be removed or clearly explained in the figure legend, as they are creating confusion for the audience.

We thank the reviewer for this comment. The letters indicate statistical groupings from post hoc multiple comparison tests. In the revised manuscript, we have added a clear explanation of this notation to the corresponding figure legends to be ease of interpretation for the reader.

(2) In the figure legend of S3a, the authors mentioned only the Caco2 cell line, whereas in the figure, there are two more cell lines, HCT116 and SW48. The authors should revise the figure legend.

We thank the reviewer for this comment. We have addressed this point and made the necessary corrections in the revised manuscript.

(3) The scale bar information is missing for Figure S3b. It should be included.

We thank the reviewer for this comment. The same scale bar was applied across all images in this panel. We have clarified this in the figure legend.

(4) In Figure 2e, the HIF-1α level in the Caco2 cells at 24 hr time point is a lot higher compared to the level at the 12-hour time point after C. albicans infection. But in the WB quantification in Figure 2f, the level of HIF-1α is not higher when compared to 12hr. Although it is relative data based on control, authors should check this calculation again for any errors.

We thank the reviewer for carefully examining the data. We have re-verified the quantification and confirmed that the values represent relative protein levels normalized to the corresponding controls at each time point.

Because samples from different time points were processed and analyzed separately, direct comparison of absolute protein levels across time points is not appropriate. Therefore, relative quantification within each time point provides a more accurate and representative assessment of HIF-1α changes.

(5) Line 125-127: This sentence should be rephrased.

We thank the reviewer for this comment. We have revised the corresponding section to improve clarity.

(6) PHD-mediated ubiquitination is the primary mechanism regulating HIF-1α protein stabilization. The authors should add an appropriate reference here.

We thank the reviewer for this suggestion. An appropriate reference has been added in the revised manuscript to support this statement.

(7) The authors claim that they observed that although the total level of HIF-1α increased, the ratio of its ubiquitinated form to total HIF-1α decreased. The authors should clearly indicate in the figure which protein band from the WB image was used for quantification from Figure S3c, which resulted in the graph presented in Figure S3d.

We thank the reviewer for this suggestion. We have revised the figure legend to improve clarity.

(8) In Figure 3a, there are some faint grey color lines. These graphs should be reformatted.

We thank the reviewer for this comment. We did not observe obvious faint grey lines in the original figure; however, these artifacts may have arisen during image conversion or file transfer. To ensure optimal image quality, we have provided high-resolution vector files to improve clarity.

(9) What do a and b in the bar graphs shown in Figure 3d,e; S4d,e,f represent?

We thank the reviewer for this comment. The letters indicate statistical groupings from multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend of this notation to the corresponding figure legends.

(10) What do a,b,c in the bar graphs shown in Figure 4c,d,h represent?

We thank the reviewer for this comment. The letters indicate statistical groupings from multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend of this notation to the corresponding figure legends.

(11) There are some faint grey lines in the bar graphs shown in Figure 4g. These lines should be removed.

We thank the reviewer for this comment. We did not observe obvious faint grey lines in the original figure; however, these artifacts may have arisen during image conversion or file transfer. To ensure optimal image quality, we have provided high-resolution vector files to improve clarity.

(12) Grey line below HIF-1α in the graph shown in Figure h should be removed.

We thank the reviewer for this comment. We did not observe obvious faint grey lines in the original figure; however, these artifacts may have arisen during image conversion or file transfer. To ensure optimal image quality, we have provided high-resolution vector files to improve clarity.

(13) The authors wrote - notably, despite treatment with AG1478, the levels of HIF-1α and c-MYC in C.albicans-infected cells remained significantly elevated compared to the uninfected control group (Figure 4b). There is no quantification for c-MYC. Statistics for HIF-1α quantification are missing. These should be added.

We thank the reviewer for this comment. We have quantified HIF-1α levels, and the results are presented in Figure 4d, including statistical analysis.

(14) There is no data for knockdown of MYD88, Dectin-1, and SYK as mentioned in the text lines 222-224. The authors should explain this discrepancy.

We thank the reviewer for this important comment. MYD88, Dectin-1, and SYK are expressed at relatively low levels in HCT116 cells, and our preliminary qPCR analyses indicated that it would be technically challenging to achieve reliable and quantifiable knockdown of these targets. Nevertheless, previous studies have reported that Dectin-1 can be present on the surface of epithelial cells, suggesting that it may still contribute to fungal recognition even at low expression levels. Therefore, given the technical constraints of gene knockdown in this specific context, we reasoned that pharmacological inhibition would provide a more robust approach to suppress this pathway.

(15) In line 227 in the results section it should be Figure S5c-e instead of Figure S5b-e. Figure S5b results do not match the results that are being explained here.

We thank the reviewer for this comment. We have corrected the typos in the revised manuscript.

(16) What do a,b,c in the bar graphs shown in Figure 5 a,b,i represent?

We thank the reviewer for this comment. The letters indicate statistical groupings from multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend of this notation to the corresponding figure legends.

(17) Was the experiment in Figure 5e done in triplicate? If not, it should be done in triplicate and quantified. The scale bar information is missing for IF images shown in Figure 5e. It should be added.

We thank the reviewer for this comment. The experiments were independently repeated for three times, and the quantification shown in Figure 5g represents the combined results from these biological replicates. The same scale bar was applied across all images in this panel. We have clarified this in the figure legend.

(18) Lines 273-274 in the results section: Als3 and Hwp1 are known to be involved in the adhesion of C. albicans to epithelial cells, while Ece1 encodes the virulence factor candidalysin. References should be added.

We thank the reviewer for this suggestion. We have added a reference in the revised manuscript to support this statement.

(19) What do a and b in the bar graphs shown in Figures 6 f,h,r represent? Since these letters are confusing and are present in several figures, they should be either deleted or clearly explained in the figure legends or text.

We thank the reviewer for this comment. The letters indicate statistical groupings from multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend of this notation to the corresponding figure legends to be ease of interpretation for the reader.

(20) What do a,b, and c in the bar graphs shown in Figure S8 b represent?

We thank the reviewer for this comment. The letters indicate statistical groupings from multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend to of this notation to the corresponding figure legends to be ease of interpretation for the reader.

(21) Scale bar should be added in Figure S9.

We thank the reviewer for these helpful comments. We have addressed this point and made the necessary corrections in the revised manuscript.

(22) What do a and b, in the bar graphs shown in Figure S11 represent?

We thank the reviewer for this comment. The letters indicate statistical groupings from post hoc multiple comparison tests. In the revised manuscript, we have added a clear explanation in the figure legend of this notation to the corresponding figure legends to be ease of interpretation for the reader.

(23) Were the organoids used in this paper characterized? If yes, how? Also, it should be mentioned in the appropriate section in the manuscript.

The organoids are not characterized; they are cultured using patients’ samples according to our previous protocols (He et al. Cell Stem Cell 2022).

eLife Assessment

This paper presents a collection of analyses relating structure and function in the whole-brain Drosophila EM connectome and whole-brain calcium imaging data. The linkage of detailed anatomical structure with population activity is of broad interest in circuit neuroscience in light of increasingly detailed brain maps, but the methods used made the evidence inadequate due to a lack of consideration of neurotransmitter identity and technical issues with the network analysis. The conclusions are useful for specific network observations, but a more thorough analysis of the anatomical and functional data is needed to support the overall claims.

Reviewer #1 (Public review):

In this revision the authors address some of the points, but they also make some technical errors. My overall view of the manuscript hasn't changed since the original evaluation.

Previously I noted that SC sparsity presents an issue when comparing to full FC matrices. They authors misinterpreted the Honey et al paper. They resampled ALL entries of the SC matrix (including zeros) from a Gaussian distribution. In effect, this assigns zeros small (but uniform) weights. In Honey et al, the authors resampled only existing edge weights from a gaussian distribution (the rationale at the time was that there might be pushback against the extremely heavy-tailed edge weight distribution). In other words, the zeros are still zeros following this resampling procedure.

That said, I agree that the log transform is likely useful or necessary given edge weight distributions.

In short, I still think that the approach is interesting and meritorious, I just don't think the execution is correct.

Reviewer #3 (Public review):

Summary:

In this manuscript, Okuno et al. re-analyze whole-brain imaging data collected in another paper (Brezovec et al., 2024) in the context of the two currently available Drosophila connectome datasets: the partial "FlyEM" (hemibrain) dataset (Scheffer et al., 2020) and the whole-brain "FlyWire" dataset (Dorkenwald et al., 2024). They apply existing fMRI signal processing algorithms to the fly imaging data and compute function-structure correlations across a variety of post-processing parameters (noise reduction methods, ROI size), demonstrating an inverse relationship between ROI size and FC-SC correlation. The authors go on to look at structural connectivity amongst more polarized or less polarized neurons, and suggest that stronger FC-SC correlations are driven by more polarized neurons.

Strengths:

(1) The result that larger mesoscale ROIs have higher correlation with structural data is interesting. This has been previously discussed in Drosophila in Turner et al., 2021, but here it is quantified more extensively.

(2) The quantification of neuron polarization (PPSSI) as applied to these structural data is a promising approach for quantifying differences in spatial synapse distribution. The revision now uses morphological cable length for some analyses rather than straight-line distance, which improves the realism and interpretability of these results.

Weaknesses:

One should not score noise/nuisance removal methods solely by their impact on FC-SC correlation values, because we do not know a priori that direct structural connections correspond with strong functional correlations. In fact, work in C. elegans, where we have access to both a connectome and neuron-resolution functional data, suggests that this relationship is weak (Yemini et al., 2021; Randi et al., 2023). Similarly, I don't think it's appropriate to tune the confidence scores on the EM datasets using FC-SC correlations as an output metric. While it is likely that some FC-SC relationship does exist at large scales, it does not in my view justify use of this metric for evaluating noise removal methods, since such methods may inadvertently remove real neural correlates. This concern remains unaddressed in the revision.

Any discussion of FC-SC comparisons should include an analysis of excitatory/inhibitory neurotransmitters, which are available in the fly connectome dataset. The authors examine the ratios of input and output neurotransmitters in different defined regions. However, I think it would be more useful to integrate the neurotransmitter information more fully into the assessment of SC, for instance: examining the signed weight (excitatory - inhibitory), or by examining the excitatory and inhibitory networks separately.

Comparisons between fly and human MRI data are also premature here. Firstly, the fly connectomes, which are derived from neuron-scale EM reconstructions, are a qualitatively different kind of data from human connectomes, which are derived from DSI imaging of large-scale tracts. Likewise, calcium data and fMRI data are very different functional data acquisition methods-the fact that similar processing steps can be used on time-series data does not make them surprisingly similar, and does not in my view constitute evidence of "similar design concepts."

The comparison of FlyEM/FlyWire connectomes concludes that differences are more likely a result of data processing than of inter-individual variability. If this is the case, the title should not claim that the manuscript covers individual variability.<br /> The analysis of the wedge-AVLP neuron strikes me as highly speculative, given that the alignment precision between the connectome and the functional data is around 5 microns (Brezovec* et al, PNAS 2024).

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

In this paper, the authors analyze connectome data from Drosophila and compare the physical wiring with functional connectivity estimated from calcium imaging data. They quantify structure-function relationships as a correlation of the two connectivity modalities. They report correlations roughly comparable to what has been described in the literature on sc/fc relationships in mammalian connectome data at the meso-scale. They then repeat their analysis, focusing on segregated versus unsegregated synapses. They derive separate connectomes using one or the other class of synapse. They show differential contributions to the sc/fc relationships by segregated versus unsegregated synapses.

Strengths:

There is nice synthesis of multimodal imaging data (Ca and EM data from flies and meso-scale data from human and marmoset).

Thank you very much for your comments.

Weaknesses:

(1) The paper is written in an unusual way. The introduction intermingles results with background, making it hard to figure out what precisely is being tested.

Thank you for pointing this out. We have revised the introduction to make it more concise.

(2) There are also major methodological gaps. Though the mammalian connectomes are used as a point of reference, no descriptions of their origins or processing are included.

The reanalysis of marmoset data is presented in Ext. Data Figure. However, as pointed out by other reviewers, the data was obtained in [10], and the processing is also described in [10]. Therefore, we have revised the caption and removed the Ethics Declaration.

(3) A major weakness stems from the actual calculation of the sc/fc correlation. In general, SC is sparse. In the case of the EM connectomes, it is *exceptionally* sparse (most neural elements are not connected to one another). The authors calculated sc/fc coupling by correlating the off-diagonal elements of sc (the logarithm of its edge weights) and fc matrices with one another. The logarithmic transformation yields a value of infinity for all zero entries. The authors simply impute these elements with 0. This makes no sense and, depending on whether these zero elements are distributed systematically versus uniformly random, could either inflate or deflate the sc/fc correlations. Care must be taken here.

Thank you for pointing this out. As you mentioned, the SC matrix becomes increasingly sparse as the number of ROIs increases (Ext. Data Fig.2-2b). In contrast, the FC matrix may contain values even when there are no direct connections between ROIs (indirect connections). We conducted an investigation into this issue. To deal with this issue, Honey et al. (2009) [6] resampled the elements of the SC matrix in rank order using a Gaussian distribution and calculated the FC-SC correlation between this resampled SC and FC.

Ext. Data Fig.2-2a shows a comparison between resampled SC (Honey et al.’s method) and log-scaled SC (our method). Up to 200 ROIs, the proportion of SC matrix elements that are zero is less than 10% (Ext. Data Fig.2-2b), and there is little zero replacement of logarithmic elements. In this situation, replacing with Gaussian arithmetic tends to increase the correlation coefficient (Ext. Data Fig.2-2a). On the other hand, with 10,000 ROIs, where sparsity is extremely high, the proportion of SC matrix elements that are zero exceeds 70%. In this situation, 70-80% of the zeros are randomly assigned from the smaller end of the Gaussian distribution, which causes a lowering of the correlation coefficient (Ext. Data Fig.2-2a, c, d). For these reasons, we believe that log-scaled SC has less bias than resampling with a Gaussian distribution, and conclude that using log-scaled SC as is in this paper is reasonable. Log-scaled SC has also been used in previous studies [9, 68] and is considered a simple method for showing the relationship (correlation) between FC and SC. To show that we have considered this issue, Ext. Data Fig.2-2 has been added to the manuscript.

(4) Further, in constructing the segregated versus unsegregated connectomes, they use absolute thresholds for collecting synapses. It is unclear, however, whether similar numbers of synapses were included in both matrices. If the number is different, that might explain the differential relationship with fc; one matrix has more non-zero entries (and as noted earlier, those zero entries are problematic).

Author response image 1.

a, Sparsity rate histogram of SC matrix with cPPSSI (0-0.1) and subsampled null SC matrices corresponding Fig.4e. Red line indicates sparsity rate of SC matrix with cPPSSI (0-0.1). b, Sparsity rate histogram of SC matrix with cPPSSI (0.9-1) and subsampled null SC matrices corresponding Fig.4f. c, Sparsity rate histogram of SC matrix with reciprocal synapse (≤2𝜇𝑚) and subsampled null SC matrices corresponding Fig.4i.

Thank you for pointing this out. The number of synaptic connections in the SC matrix shows a large difference between those extracted from cPPSSI (0-0.1) and cPPSSI (0.9-1) (Fig. 4e, f). However, when null SC matrices (99) were generated for each and compared with the cPPSSI-extracted matrices, the FC-SC correlation was significantly higher or lower. At this point, since the sparsity rates of the null SC matrices differed a lot from that of the SC matrices extracted by cPPSSI, we regenerated the null SC matrices in Fig. 4e and 4i. As shown in Author response image 1, we ensured that the extracted SCs (red lines) fit within the null-generated matrices. This figure was added to Ext. Data Fig.4-5, and the main text was also revised. The sparsity rates are 0.52 for cPPSSI (0-0.1) and 0.123 for cPPSSI (0.9-1). Since both cases involve comparisons with null SC matrices that have closely similar sparsity rates, we believe comparison using log-scaled SC is appropriate.

(5) There was also considerable text (in the results) describing the processing of the Ca data. In this section, the authors frequently refer to some pipelines as "better" or "worse" (more or less effective). But it is not clear what measures they adopted to assess the effectiveness of a pipeline.

Detailed registration flow of Ca data is described in “Preprocessing of D. melanogaster calcium imaging data” in Materials and Methods section (Ext. Data Fig. 1-1a). Then, optimal nuisance factor removal methods and smoothing size were investigated. We used both correlation analysis (FC-SC correlation) and ROC curve analysis (FC-SC detection). Since signals are assumed to be transmitted between regions based on SC, when SC is treated as the ground truth, we considered a pipeline with a FC-SC higher similarity and higher detection to be better. We updated the Results section to include this point.

Reviewer #2 (Public review):

Summary:

Okuno et al. investigate the structure-function relationship in the fruit fly Drosophila melanogaster. To do so, they combine published data from two recent synapse-level connectomes ("hemibrain" and "FlyWire") with a dataset comprising functional whole-brain calcium imaging and behavioural data. First, they investigate the applicability of fMRI pre-processing techniques on data from calcium imaging. They then cross-correlate this pre-processed functional data with structural data extracted from the connectomes, including a comparison to humans. The authors proceed to compare the two connectomes and find significant differences, which they attribute to differences in the accuracy of the synapse detections. Next, they present a novel algorithm to quantify whether neurons are segregated (pre- and postsynapses are spatially separate) or unsegregated (pre- and postsynapses are mixed). Using this approach, they find that unsegregated neurons may contribute more to function than segregated neurons. Applying a general linear model to the functional dataset suggests that activity in two brain areas (Wedge and AVLP) is suppressed during walking. The authors identify a GABAergic neuron in the connectome that could be responsible for this effect and suggest it may provide feedback to the fly's "compass" in the central complex.

Strengths:

The study tackles a relevant question in connectomics by exploring the relationship between structural and functional connectivity in the Drosophila brain. The authors apply a range of established and adapted analytical methods, including fMRI-style preprocessing and a novel synaptic segregation index. The effort to integrate multiple datasets and to compare across species reflects a broad and methodical approach.

Thank you very much for your comments.

Weaknesses:

The manuscript would benefit from a clearer overarching narrative to unify the various analyses, which currently appear somewhat disjointed. While the technical methods are extensive, the writing is often convoluted and lacks crucial details, making it difficult to follow the logic and interpret key findings. Additionally, the conclusions are relatively incremental and lack a compelling conceptual advance, limiting the overall impact of the work.

(1) The introduction currently contains a number of findings and conclusions that would be better placed in the results and discussion to clearly delineate past findings from new results and speculations.

Thank you for pointing this out. We have revised the introduction to make it more concise.

(2) The narrative would benefit greatly from some clear statements along the lines of "we wanted to find out X, therefore we did Y".

Thank you for pointing this out. In many biology papers, the problem is clear, but as you say, this paper starts by comparing the very fine SC and FC of flies, which makes the problem unclear and the results sporadic. We have revised the structure of the introduction.

(3) More concise terminology would be helpful. For example, the connectomes are currently referred to as either "hemibrain", "FlyEM", "whole-brain", or "FlyWire".

Thank you for pointing this out. We revised the manuscript to separate "hemibrain" and "whole-brain" from "connectome." "hemibrain" and "whole-brain" retain their original meanings.

(4) The abstract claims "a new, more robust method to quantify the degree of pre- and post-synaptic segregation". However, the study fails to provide evidence that this method is indeed more robust than existing methods.

We apologize, but this information was not included in the main figures or the Results section. It is presented in the Methods section and Ext. Data Fig. 4-1i, j. We moved related texts from the Methods to the Results section.

(5) The authors define unsegregated neurons as having mixed pre- and postsynapses in the same space. However, this ignores the neurons' topology: a neuron can exhibit a clearly defined dendrite with (mostly) postsynapses and a clearly defined axon with (mostly) presynapses, which then occupy the same space. This is different from genuinely unsegregated neurons with no distinct dendritic and axonal compartments, such as CT1.

Thank you for pointing this out. Regarding this point, we think it is difficult to discuss the neuron’s topology in this paper. We defined PPSSI and demonstrated only that unsegregated neurons with mixed pre- and post-synapses are scattered throughout the brain (Ext. Data Fig. 4-2e). Further research is needed to determine the relationship with morphology in individual neurons.

One possibility is that inhibitory, non-spiking unsegregated neurons, such as CT1 amacrine cell [24, 27, 28] or interneurons in Antennal Lobe [29], may be widely used throughout the brain (WAGN is also a candidate for this). Grimes et al. [34] mentioned “The retina is a beautiful example of a neural network that optimizes signal processing capacity while minimizing cellular cost.” To maintain the signal dynamic range, A17 amacrine cells must optimize the processing units and wiring costs. If one unit equaled one cell, an enormous number of cell bodies would be required, reducing the number of processing units per volume and increasing the energy cost during development. To optimize this, they proposed arranging units capable of parallel processing within a single cell, thereby maximizing the processing units and wiring costs per volume.

Signal bursts might also occur in the central nervous system (CNS), in which case CNS neurons also require dynamic range adjustment. The concept of optimizing processing units per volume is highly compelling and is thought to apply not only to the retina but throughout the entire brain.

(6) It is not entirely clear where the marmoset dataset originates from. Was it generated for this study? If not, why is there a note in the Ethics Declaration?

Marmoset data were reported in [10] and it was not generated for this study. We therefore removed the Ethics Declaration.

(7) On the differences between hemibrain and FlyWire: What is the "18.8 million post-synapses" for FlyWire referring to? The (thresholded) FlyWire synapse table has 130M connections (=postsynapses). Subsetting that synapse cloud to the hemibrain volume still gives ~47M synapses. Further subsetting to only connections between proofread neurons inside the hemibrain volume gives 19.4M - perhaps the authors did something like that? Similarly, the hemibrain synapse table contains 64M postsynapses. Do the 21M "FlyEM" post-synapses refer to proofread neurons only? If the authors indeed used only (post-)synapses from proofread neurons, they need to make that explicit in results and methods, and account for differences in reconstruction status when making any comparisons. For example, the mushroom body in the hemibrain got a lot more attention than in FlyWire, which would explain the differences reported here. For that reason, connection weights are often expressed as, e.g., a fraction of the target's inputs instead of the total number of synapses when comparing connectivity across connectomic datasets. Furthermore, in Figure 3b, it looks like the FlyWire synapse cloud was not trimmed to the exact hemibrain boundaries: for example, the trimmed FlyWire synapse cloud seems to extend further into the optic lobes than the hemibrain volume does.

Thank you for pointing this out. FlyEM connectome data version 1.2 was downloaded and used as described in Data Availability. This data is provided in the format defined by https://neuprint.janelia.org/public/neuprintuserguide.pdf, and we extracted neurons and synapses from it.

The entire segmentation body is 28M segmentations, and there were 99,644 Traced proofread neurons. In addition, there were 73M (pre- or post- alone) synapses, 87M records in synapseSets and 128M records in synapseSet-to-synapse. When we extracted post-synapses between Traced neurons, the total number was 21.4M (i.e., connections from Traced neurons to other body fragments like Orphans were excluded).

The FlyWire dataset (v783) was downloaded from the flywire codex and Zenodo. This dataset contained 139,255 proofread neurons and 54.5M (pair of pre- and post-) synapses, as described in Dorkenwald et al. [13], with 18.8M post-synapses in the regions corresponding to the hemibrain primary ROIs. We have updated the Results and Methods sections by taking into account your comment.

In Fig. 3b, these images were created using a mask that extended the boundaries of the hemibrain primary ROIs, making the boundaries unclear. Therefore, we corrected the images in Fig. 3b by adjusting the mask so that the boundaries were properly aligned.

Reviewer #3 (Public review):

Summary:

In this manuscript, Okuno et al. re-analyze whole-brain imaging data collected in another paper (Brezovec et al., 2024) in the context of the two currently available Drosophila connectome datasets: the partial "FlyEM" (hemibrain) dataset (Scheffer et al., 2020) and the whole-brain "FlyWire" dataset (Dorkenwald et al., 2024). They apply existing fMRI signal processing algorithms to the fly imaging data and compute function-structure correlations across a variety of post-processing parameters (noise reduction methods, ROI size), demonstrating an inverse relationship between ROI size and FC-SC correlation. The authors go on to look at structural connectivity amongst more polarized or less polarized neurons, and suggest that stronger FC-SC correlations are driven by more polarized neurons.

Strengths:

(1) The result that larger mesoscale ROIs have a higher correlation with structural data is interesting. This has been previously discussed in Drosophila in Turner et al., 2021, but here it is quantified more extensively.

(2) The quantification of neuron polarization (PPSSI) as applied to these structural data is a promising approach for quantifying differences in spatial synapse distribution.

Thank you very much for your comments.

Weaknesses:

One should not score noise/nuisance removal methods solely by their impact on FC-SC correlation values, because we do not know a priori that direct structural connections correspond with strong functional correlations. In fact, work in C. elegans, where we have access to both a connectome and neuron-resolution functional data, suggests that this relationship is weak (Yemini et al., 2021; Randi et al., 2023). Similarly, I don't think it's appropriate to tune the confidence scores on the EM datasets using FC-SC correlations as an output metric.

Thank you for pointing this out. We believe that the FC in C. elegans uses cell body dynamics, which is different from the synaptic population dynamics in a region of fly calcium imaging or fMRI data (BOLD [Blood Oxygenation Level Dependent] signal). The BOLD signal in a region is thought to correspond to the neurovascular coupling of synaptic population dynamics. Furthermore, compartmentalization of a neuron has been observed in C. elegans (Hendricks et al., 2012)*, showing different dynamics across neuron compartments. Thus, the dynamics of the cell body and the dynamics of the synaptic population in other regions are different in C. elegans. We speculate that there is some relationship between FC-SC between regions, because the FC-SC correlation in the fly brain reached r=0.87 with 20 ROIs (Fig. 2d). We believe that this result is different from the cell body dynamics in C. elegans.

*Hendricks et al., “Compartmentalized calcium dynamics in a C. elegans interneuron encode head movement,” Nature 487, 99-103 (2012)

Any discussion of FC-SC comparisons should include an analysis of excitatory/inhibitory neurotransmitters, which are available in the fly connectome dataset. However, here the authors do not perform any analyses with neurotransmitter information.

A comparison between FC-SC and neurotransmitter has been written in the Results section. We investigated the ratios of neurotransmitter input (ExtFig.3-2a) and output (Fig. 3f) in each region, and investigated the relationship between this ratio and FC-SC correlation in each neurotransmitter. This revealed significant correlations for acetylcholine (r=0.39, p=0.0013) and GABA (r=-0.25, p=0.046) (Fig. 3g). That is, the higher the percentage of excitatory connections, the higher the FC-SC correlation; conversely, the higher the percentage of inhibitory connections, the lower the FC-SC correlation.

Comparisons between fly and human MRI data are also premature here. Firstly, the fly connectomes, which are derived from neuron-scale EM reconstructions, are a qualitatively different kind of data from human connectomes, which are derived from DSI imaging of large-scale tracts. Likewise, calcium data and fMRI data are very different functional data acquisition methods-the fact that similar processing steps can be used on time-series data does not make them surprisingly similar, and does not in my view, constitute evidence of "similar design concepts."

Thank you for pointing this out. As you say, fiber bundles of DTI and EM connectome are completely different. Nevertheless, the fact remains that the FC-SC correlation is high in both the fly and human brains. As mentioned above, both regional signal from calcium imaging and BOLD signal from fMRI are based on synaptic population dynamics. It was estimated that 43% of the energy consumption in the gray matter is due to synaptic activity of neurons (Harris et al., 2012), and the BOLD signal fluctuates greatly due to this activity. Furthermore, synaptic activity is thought to be much faster than the activity of microglia and astrocytes, so the FC of fMRI is thought to mainly capture the regional correlation of synaptic activity. In other words, in both flies and humans, although the size is different, the pre-synaptic activity in one region and the pre-synaptic activity in another region via neural fibers are being compared in a common manner in the form of FC-SC.

In addition, non-spiking unsegregated neuron exists in mammals as well, such as the amacrine cell of the retina [34], and even pyramidal cells in the neocortex show local mixtures of pre- and post-synapses (Ext. Data Fig.1-2). If a functional unit is realized by local compartment in a neuron as mentioned in [34], the fly will be a powerful model organism for investigating them, and its functional “design concept” may also be useful for mammals.

Harris et al., “The Energetics of CNS White Matter,” J. Neurosci., 2012, 32 (1) 356-371

The comparison of FlyEM/FlyWire connectomes concludes that differences are more likely a result of data processing than of inter-individual variability. If this is the case, the title should not claim that the manuscript covers individual variability.

Thank you for pointing this out. Inter-individual variability is relevant to both SC and FC. Regarding SC, we think the difference in the number of synapses between the two individuals is due to the difference in detection power caused by differences in the resolution of the electron microscope. Regarding FC, as stated in the Results section, “Spatial smoothing is useful for absorbing inter-individual variability and conducting second-level group analysis.” Increasing the smoothing size improves the correlation and AUC between group-averaged FC and SC, indicating the presence of inter-individual variability in FC (Fig. 2b, Ext. Data Fig. 2-1b, especially when the number of ROIs is high). We added this text in the Introduction and Results sections to address your comment.

The analysis of the wedge-AVLP neuron strikes me as highly speculative, given that the alignment precision between the connectome and the functional data is around 5 microns (Brezovec* et al, PNAS 2024).

As you mentioned, functional analysis has limitations in spatial resolution. In particular, the resolution in the Z axis is 4 μm, which is 1,000 times lower than the resolution of electron microscopy data. This makes it difficult to perfectly match synaptic activity to a synapse in the structural data. Furthermore, spatial smoothing is applied to functional images to absorb inter-individual variability, which can only provide blurred results for group analyses. These are considered limitations of the methods used in fMRI analysis. Despite these limitations, we applied GLM analysis to walking behavior and observed clear inactivity region. This region roughly corresponds to the synaptic cloud of a neuron named WAGN (Fig.5b and c). This neuron also connects to WPNb and ANs in the connectome data, suggesting a possibility that it is related to walking behavior. This is merely a screening reference; therefore, further biological experimentation is needed to pursue this topic.

Recommendations for the authors:

Reviewing Editor Comments:

We should emphasize that the reviewers encouraged revision and resubmission. If the reviewers' comments were to be addressed in full in a revision to strengthen the evidence, this would significantly increase the impact of the findings and the relevance of the work to the fly neuroscience community and to the connectomics field more broadly.

Thank you very much for your comments.

Major Issues:

(1) Structural correlation and functional correlation measure very different aspects of network data, yet a simple correlation between the off-diagonal elements of the two is used. It would be expected that this would not be directly proportional, and it's not clear why this would be a sensible measure. The authors need a better solution for dealing with the zero entries in the SC matrix. Replacing the infinities with zeros and then running the linear regression to get an SC/FC relationship is not appropriate. Even with a better metric, given that both intuition and other studies have shown a weak correlation between FC and SC, using FC-SC correlation as a quality descriptor for other properties is not proper. Furthermore, the authors don't account for neurotransmitter identity in the structural data, which would have strong implications for the relationships between FC and SC.

Thank you for pointing this out. To investigate this issue we compared the FC-SC correlation between the Gaussian resampled SC approach used in Honey et al. (2009) [6] and the log-scaled SC used in this study (Ext. Data Fig.2-2a). With a small number of ROIs, the sparsity rate is low (Ext. Data Fig.2-2b), resulting in less zero replacement. Therefore, log-scaled SC is likely to more accurately represent the FC-SC relationship. Furthermore, with a large number of ROIs, the sparsity rate exceeds 70%, and Gaussian resampled SC randomly assigns a large number of zero elements from the smaller end of the distribution. This tends to lower the correlation (Ext. Data Fig.2-2c, d), suggesting that log-scaled SC provides fairer results. Log-scaled SC has been used in previous studies [9, 68] and is considered a simple method for showing the relationship (correlation) between FC and SC. When zero replacement is undesirable, using connection weights (the proportion of connections originating from the target region among all connections) can yield results similar to log-scaled SC (data not shown). It may be possible to compare various methods, but this is outside the scope of this study and requires further research.

The C. elegans studies presented by Reviewer #3 showed a weak correlation between FC and SC. However, C. elegans neurons do not fire and exhibited different calcium fluctuations depending on the region (Hendricks et al., 2012). This suggested that the cell body and various synaptic terminal regions have different FCs, which is consistent with the objective of our study (neuronal compartmentalization). If a functional unit is locally composed of multiple neurons and synapses, it is expected that SC and FC from that region will show a strong relationship. Larger regions would include multiple functional units, and a relationship between SC and FC would also be found, which is consistent with the results of our study. The C. elegans study compared FC of the cell body (a region) with SC of whole cell (not a same region), which would be inconsistent.

(2) Synaptic segregation on neurons can be topologically present even if pre- and post-synaptic synapses are present in similar regions of space, as an axon branch and dendrite branch can overlap in space but remain distinct along the arbor. The authors emphasize a region-based definition that does not reflect cellular anatomy. Moreover, the authors do not make an argument for their claim of better robustness of their new synaptic segregation measures.

Author response image 2.

Distance calculation for DBSCAN. a, Example synapse pair (black dot) of distance calculation. Red line shows the straight-line distance, and green line shows the morphology-based distance. DBSCAN will places two synapses in the same cluster based on straight-line distance, but they will be in different clusters based on the morphology-based distance.

Thank you for pointing this out. We changed from using DBSCAN based on the straight-line distance between synapses to DBSCAN based on the morphology-based distance via the branch nearest to the synapse (Author response image 2a). This resulted in a synaptic segregation measure that incorporates cellular anatomy. We updated all related figures, such as Figure.4, Ext. Data Figure.4-1, 4-2, 4-3, 4-4, Figure.5h. Also, we updated related text in the Results and Methods sections.

(3) Reviewers found the overall structure of the paper is difficult to follow, with sections appearing disjoint and the aims of different sections not well described. This extended to the paper organization as well, with the introduction not clearly setting up the questions and being distinct from the results. The manuscript would benefit from a clearer overarching narrative to unify the various analyses.

Thank you for pointing this out. We have revised the introduction to make it more concise.

(4) Similarly, there are several descriptions of data and analysis that are unclear or lacking, including the source of the marmoset data and how the FlyWire synapse was subsampled.

As pointed out by other reviewers, the marmoset data was obtained in [10], and the processing is also described in [10]. Therefore, we have revised the caption and removed the Ethics Declaration.

We have updated the Results and Methods sections regarding the extraction of "traced" neurons and synapses in FlyEM connectome data, and the extraction of post-synapses in hemibrain primary ROIs in FlyWire connectome data.

(5) Comparisons between FlyWire and Hemibrain have shown many similarities and some clear examples of inter-individual variability. There was concern that technical decisions with handling FlyWire synapse sampling were responsible for some of the differences observed between the datasets.

In response to Reviewer #2's question, we answered that both FlyEM and FlyWire use proofread neurons and their connecting synapses. We also updated Fig. 3b and the Results and Methods sections.

Reviewer #1 (Recommendations for the authors):

The paper is written in an unusual way. It would be helpful if the introduction read more like a standard introduction. Describe the relevant background that the reader needs to understand the results that come later. Frame the experiments in terms of a question or hypothesis. Results should be relegated to the results section (or, if you like, a final paragraph that summarizes the findings). They should not be intermingled throughout the introduction.

Thank you for pointing this out. We have revised the introduction to make it more concise.

The authors must be more attentive in terms of how they construct the segregated/unsegregated connectomes. I suggest exploring various thresholds/bins, but also considering proportionality thresholds that match the number of synapses.

Thank you for pointing this out. As pointed out by other reviewers, we changed from using DBSCAN based on the straight-line distance between synapses to DBSCAN based on the morphology-based distance via the branch nearest to the synapse (Author response image 2a). This resulted in a synaptic segregation measure that incorporates cellular anatomy.

We also considered about the sparsity rates of the SC matrices. Since the sparsity rates of the null SC matrices differed a lot from that of the SC matrices extracted by cPPSSI, we regenerated the null SC matrices, shown in Fig. 4e and 4i. As shown in Author response image 1, we ensured that the extracted SCs fit within the null-generated matrices. This figure was added to Ext. Data Fig.4-5, and the main text was also revised.

The authors need a better solution for dealing with the zero entries in the sc matrix. Replacing the infinities with zeros and then running the linear regression to get an sc/fc relationship is not appropriate.

Thank you for pointing this out. To investigate this issue, as pointed out by other reviewers, we compared the FC-SC correlation between the Gaussian resampled SC approach used in Honey et al. (2009) [6] and the log-scaled SC used in this study (Ext. Data Fig.2-2a). With a small number of ROIs, the sparsity rate was low (Ext. Data Fig.2-2b), resulting in less zero replacement. Therefore, log-scaled SC is likely to more accurately represent the relationship. Furthermore, with a large number of ROIs, the sparsity rate exceeds 70%, and resampled SC randomly assigns a large number of zero elements from the smaller end of the distribution. This tends to lower the correlation (Ext. Data Fig.2-2c, d), suggesting that log-scaled SC provides fairer results. Using connection weights (the proportion of connections originating from the target region among all connections) can yield results similar to log-scaled SC (data not shown), because this matrix can also be very sparse. It may be possible to compare various methods, but this is outside the scope of this study and requires further research.

It would be useful to include a description of where the human/marmoset datasets came from. It would be useful to describe the processing of those datasets and whether they're comparable to how the fly data was processed.

As pointed out by other reviewers, the marmoset data was obtained in [10], and the processing is also described in [10]. Therefore, we have revised the caption and removed the Ethics Declaration.

The pre-processing of fly calcium imaging data is described in the Methods section. Unfortunately, this processing method is not comparable to that used in humans/marmosets as it was highly customized.

The authors report sc/fc correlations for the human/marmoset datasets based on single papers. However, in the human case, especially, the strength of sc/fc correlations is highly variable. Not just based on number/size of parcels, but based on amount of data, processing pipeline, single-subject versus group averaged (incidentally, single-subject sc/fc is ‘much’* lower than group-averaged, which has big implications for this study, where the fly datasets are, in essence, N=1 studies).

Yes, there are numerous FC-SC correlation studies. We think Honey et al. (2009) [6] to be a highly representative study. It showed r = 0.39 to 0.48 for individual participants in 998 ROIs, and r = 0.36 for averaged one, but it increased r = 0.53 excluding absent or inconsistent structural connections. So, single-subject may not be much lower than group-averaged. Since the SC for a fly is an N=1 study, the FC-SC correlation for the same individual cannot be calculated. We think further research will be necessary.

Reviewer #2 (Recommendations for the authors):

Abstract:

Please introduce the term "ROI"

Thank you for pointing this out. We have revised the Abstract.

Introduction:

(1) On a general note: the introduction reads like an extended abstract (i.e., a mix of results and discussion).

Thank you for pointing this out. We have revised the introduction to make it more concise.

(2) Line 43: Does this mean FC-SC correlation is higher in flies but not significantly so? Please clarify.

We performed Mann-Whitney U test and it was not significant (p= 0.2667).

(3) Line 51: The "confidence" score does not indicate the degree of synaptic detection.

In the NeuPrint user guide, https://neuprint.janelia.org/public/neuprintuserguide.pdf it states “confidence - The certainty that an annotated synapse is correct and valid.” Since “degree of synaptic detection” may be difficult to understand, we changed it to “certainty of an annotated synapse.”

(4) Line 59-61: This statement needs refining: post-synapses do not "receive" neurotransmitters, action potentials aren't conducted along nerve fibres.

We changed “receive” to “sense.” About “action potentials,” we changed “conduct an action potential” to “graded potentials”, and removed “along nerve fibers.”

(5) Line 61: calcium activity as detected via GCaMP correlates with (electric) neuronal activity - please cite relevant GCaMP literature here.

We added F. Helmchen and J. Waters, "Ca2+ imaging in the mammalian brain in vivo," Eur J Pharmacol., vol. 447, pp. 119-129, 2002.

(6) Line 76: "interconnected" is rather vague; just say "many Drosophila neurons are reciprocally connected".

Thank you for pointing this out. Lin et al., (2024) showed motif analysis and there are many reciprocal, three-node and rich-club connections. However, introduction was updated and this sentence was removed.

(7) Line 77: comparing unsegregated vs reciprocal synapses is overly simplistic; these are separate features of the same object - i.e., a synapse can be reciprocal and at the same time be segregated in the presynaptic neuron but unsegregated in the postsynaptic neuron.

Thank you for pointing this out. As you say, the relationship is complicated. In this paper, we are concerned with the degree of segregation of pre- and post-synapses, and we are looking at the segregation within a neuron. In this case, nearby reciprocal synapses (<=2 μm) are included in unsegregated synapses. We have made a correction to the sentence.

(8) Line 79: I don't understand how we get from unsegregated synapses to local activity.

Retinal amacrine cells have extensive unsegregated synapses, which provide local feedback inhibition of burst inputs [34]. We changed the text around these descriptions.

(9) Line 80: What does "more essential function" mean?

We removed this sentence.

(10) Line 85: "as shown earlier": Is this based on results in this study or prior work? See also the general above note on mixing results/discussion into the introduction.

Thank you for pointing this out. We have revised the introduction to make it more concise.

(11) Line 85-87: I don't understand how the applicability of certain fMRI analysis methods in turn means that functional activity is locally compartmentalized. Did you mean to say something along the lines of "we applied common fMRI methods which showed functional activity is locally compartmentalized"?

These sentences discuss the commonality between fMRI (BOLD signal) and calcium signal, which both represent presynaptic population dynamics within a local region (voxel). Furthermore, unsegregated synapses are widespread throughout the fly brain (Ext. Data Fig.4-2) and can also be observed in human pyramidal cells (Ext. Data Fig.1-2). Unsegregated synapses suggest local compartment activity [33, 34, 39, 40] and contribute more to functional activity (Fig.4b). Therefore, the similar trend in FC-SC correlation (Fig.2d) between humans and flies suggest that both species exhibit localized compartmental activity via unsegregated synapses throughout the entire brain.

Because these sentences contain many conclusions, they have been moved from the Introduction to the Discussion section.

(12) Line 87: Please provide a reference for "common among various species".

Thank you for pointing this out. Because these sentences contain many conclusions, they have been moved from the Introduction to the Discussion section.

Results:

(1) Line 91-92:

(a) Please explain where the calcium data came from, how it was generated, etc.

We added the data source and a reference (Brezovec et al. [14]).

(b) Please clarify: what registration method?

This is not simple. Please see the Methods section and Ext. Data Fig.1-1. This is also indicated in the text.

(c) "calcium image" → "calcium image data"?

We changed “calcium image” to “calcium imaging data”.

(d) What is the "FDA template"?

This is a brain template created by Brezovec et al. [14]. JRC2018 is a well-known brain template, but it was created by immunostaining postmortem brains and did not fit well with calcium imaging data from living flies. Therefore, we used the FDA template.

(2) Line 93: Please introduce the term "ROI".

We added “(Region of Interest)” in Line 38.

(3) Line 94: Ito et al., Neuron (2014) "A systematic nomenclature for the insect brain" is a better reference for Drosophila neuropils; for the hemibrain, the ROIs were generated to match that original atlas

Thank you for pointing this out. We added a reference.

(4) Line 95/96: It is unclear what was used as the basis for the k-means/distance-based clustering

This was because we wanted to investigate whether nuisance factor removal methods are robust, even for such diverse types of ROI. We added this point to the text.

(5) Line 120ff: I'm not sure how the total number of ROIs is relevant for comparing flies and humans, given (a) the huge difference in brain size and (b) the difference in resolution of the functional data.

Indeed, the fly brain and the human neocortex are completely different. We are investigating whether there are commonalities between them using a metric called FC-SC correlation. As described in our answer for (11), both the fMRI (BOLD signal) and calcium signal represent presynaptic population dynamics within a local region (voxel). FC represents the synchronization of synaptic activity between regions, and SC represents the structural connectivity of neurons. Both flies and humans showed high SC-FC correlation and showed similar trends (Fig. 2d), so we believe it would be interesting to investigate this phenomenon.

(6) Line 123: "by contrast" is misleading here since, as you say, there isn't really a difference.

We changed “by contrast” to “and.”

(7) Line 141: I'm somewhat worried that the differences between FlyWire and hemibrain synapse counts are due to the issues mentioned above.

Thank you for the comment but we are not sure about “the issues mentioned above” is referring to.

(8) Line 148: There is no evidence that any differences in synapse are due to the resolution or anisotropy (as suggested in the introduction).

We apologize that we don’t have direct evidence for it. We changed this to the sentence “This may be caused by differences in detection accuracy resulting from the resolution of EM scanning, but not to inter-individual variability.”

(9) Line 155: References "39,45" have no brackets.

These are not referencing numbers, but brain regions of Brodmann area 39 and 45.

(10) Line 155-157: I don't think we can infer the composition of brain areas in humans based on a tenuous correlation in flies; this is highly speculative and really should be in the discussion.

In humans, there are areas with strong and weak FC-SC correlations [8], which may be due to the E-I (Excitatory-Inhibitory) balance of connections. We investigated this possibility by comparing the correlation between neurotransmitters and FC-SC correlations in the fly brain. We slightly changed this sentence.

(11) Line 159: I find the first 2-3 sentences in this paragraph confusing. Are you saying that you did all these things in the prior results sections, or that you wanted to look at X and therefore you did Y? Maybe there is an issue with the tense here?

We changed the sentences around this description.

(12) Line 161: "whole-brain" = FlyWire?

We changed “whole-brain” to “FlyWire”.

(13) Line 163: Please explain the "PPSSI" acronym.

This is now explained on Line 75.

(14) Line 165: The description of how the cPPSSI was calculated is hard to follow. For example, what's the "fraction of synapse number".

We changed our sentences around this description to be clearer. The cPPSSI is the degree of segregation within a cluster and is also assigned to each synapse. The PPSSI is then the average of the cPPSSI values of all synapses in a neuron.

(15) Line 166: Is there a difference between "cPPSSI" and "PPSSI"?

Yes, there is. Please see our answer for (14).

(16) Line 167: "The result showed a histogram resembling a normal distribution" → I suggest running a normality test.

Thank you for pointing this out. We tested it by Lilliefors test and the result was p=0.001 (significantly not a normal distribution). Since there are numerous values with PPSSI=1, it is not judged to be a normal distribution. We therefore changed this description.

(17) Line 173: I am somewhat worried about a selection bias in your correlation of segregated vs unsegregated synapses. First, it seems like only a small fraction of neurons are in the 0-0.1 and 0.9-1 PPSSI range. I would suggest running a proper correlation between PPSSI and FC-SC correlation instead of looking at just the two extremes. Second, your examples for segregated neurons (APL + CT1) are large neurons that densely innervate spatially close and functionally very similar neuropils. If the sample of unsegregated neurons consists mainly of these large interneurons, I'm not at all surprised that they contributed strongly to FC-SC correlation.

Thank you for pointing this out. For this work we investigated synapses (not neurons), extracting those with cPPSSI of 0-0.1 and 0.9-1, and performed a rank text with the FC-SC correlation of random sub-sampled synapses. We aimed to demonstrate that unsegregated synapses in particular, strongly contribute to FC-SC, and we hope to investigate overall trends in a future study.

(18) Line 185: I don't think the function of reciprocal synapses is "considered to be clear". There are examples of feedback inhibition through reciprocal synapses, in particular in the visual system, but that does not mean that this is true across the board.

We changed “considered to be clear” to “considered to be clearer than unsegregated synapses.” Of course, the function of reciprocal synapses is unknown for the whole brain, but we think it is more well-studied than unsegregated synapses.

(19) Line 188 / Figure 4h: that figure panel does not appear to show transmitter pairs.

Figure 4h (FlyWire) showed transmitter pairs. Ext. Data Fig.4-1g did not, because FlyEM does not have transmitter information.

(20) Line 192: Please clarify "functionally common".

We changed our sentences to clarify this.

(21) Line 199: "ventral nerve code" → "ventral nerve cord".

We fixed this typo.

(22) Line 201: I don't think you can use "conversely" here.

We changed “Conversely” to “Moreover.”

(23) Line 201: How certain are you that the WAGN neuron is the only candidate? Also, it would be nice to provide the neuron IDs so that people can identify them in the connectome.

Thank you for pointing this out. We added Root ID: 720575940644632087 in the text. Actually, we found several GABA neuron candidates, such as 720575940637611365, 720575940644632087, 720575940613552947, 720575940640333109 and 720575940612264817. We investigated whether ER1(L) was present in these downstream connections and found that 720575940644632087 had the strongest connection with the largest number of synapses, so we adopted this.

(24) Line 207: When you say "the left WAGN was strongly connected", are those connections not also present for the right WAGN?

There is a right WAGN (Root ID: 720575940624377224), but it does not have strong interconnections with WPNb tier 2/3 (left) neurons. For the right WAGN, there are few inputs from WPNb tier 2/3 (left). We added “(left)” in the text.

(25) Line 212: I don't think you can use "however" here.

We removed “however.”

(26) Line 214: "well unsegregated" → "very unsegregated"?

This sentence was removed, because we recalculated Fig. 5h.

Ethics Declaration:

It seems the marmoset data were reported on in [10], so why is there a reference to the generation of the dataset?

Yes, marmoset data were reported in [10], so we removed the Ethics Declaration.

Reviewer #3 (Recommendations for the authors):

(1) In my opinion, the title and framing of this manuscript dramatically overstate the results presented here. Also, the results presented in the different figures in this manuscript seem disjointed and are not very related to each other.

Thank you for pointing this out. We have rewritten our manuscript slightly to address this. Inter-individual variability is relevant to both SC and FC. Regarding SC, we think the difference in the number of synapses between the two individuals is due to the difference in detection power caused by differences in the resolution of the electron microscope. Regarding FC, as stated in the Results section, “Spatial smoothing is useful for absorbing inter-individual variability and conducting second-level group analysis.” Increasing the smoothing size improves the correlation and AUC between group-averaged FC and SC, indicating the presence of inter-individual variability in FC (Fig. 2b, Ext. Data Fig. 2-1b, especially when the number of ROIs is high). We added this text in the Introduction and Results sections.

(2) There are multiple ways to compute structural correlation matrices-the methods the authors implemented should be discussed in greater detail in the manuscript.

Thank you for pointing this out. To investigate this issue, as pointed out by other reviewers, we compared the FC-SC correlation between the Gaussian resampled SC approach, used in Honey et al. (2009) [6] and the log-scaled SC approach, used in this study (Ext. Data Fig.2-2a). With a small number of ROIs, the sparsity rate was low (Ext. Data Fig.2-2b), resulting in fewer zero replacement. Therefore, log-scaled SC is likely to more accurately represent the relationship in our study. Furthermore, with a large number of ROIs, the sparsity rate exceeds 70%, and resampled SC randomly assigns a large number of zero elements from the smaller end of the Gaussian distribution. This tends to lower the correlation (Ext. Data Fig.2-2c, d), suggesting that log-scaled SC provides fairer results. Using connection weights (the proportion of connections originating from the target region among all connections) can yield results similar to log-scaled SC (data not shown), because this matrix can be also very sparse. The log-scaled SC aprroach has been used in previous studies [9, 68] and is considered a simple method for showing the relationship (correlation) between FC and SC. It may be possible to compare various methods in-depth, but this is outside the scope of this study and requires further research.

(3) The use of the FC-SC detection score defined by the authors should be discussed and justified more extensively in the text.

Thank you for pointing this out. This has already been discussed in [10]. We defined our own “FC-SC detection score,” but we consider the overall approach to be well established in the literature. For example, Stafford et al. (2014) carried out FC-SC detection for 168 mouse cortical regions, and obtained 78.26% sensitivity and 81.69% specificity for the top 1% of SC. Hori et al. (2020) also investigated FC-SC detection for 55 cortical regions of the marmoset brain left hemisphere, achieving an AUC of 0.72. We think FC-SC detection is an index that evaluates the relationship between FC and SC from a different angle than FC-SC correlation and is worthwhile.

Hori et al., (2020). Comparison of resting-state functional connectivity in marmosets with tracer-based cellular connectivity. NeuroImage, 204, 116241.

Stafford et al., (2014). Large-scale topology and the default mode network in the mouse connectome. Proc. Natl. Acad. Sci. U.S.A., 111(52), 18745-18750.

eLife Assessment

This useful study addresses the interesting question of how immune cells recognise infected erythrocytes in malaria. It proposes the parasite protein PfGBP-130 as an interaction partner of the human cell surface protein LFA 1, which could help explain how NK cells recognize infected erythrocytes. The conclusions are partially supported by pull-down and cell-based activation data. However, the overall evidence of direct interaction at the cell-cell interface and downstream effects is incomplete; stronger evidence is required to demonstrate surface exposure of PfGBP-130, as well as a direct role of this antigen in killing.

Reviewer #1 (Public review):

In this manuscript, the authors aim to determine the ligand on Plasmodium falciparum infected erythrocytes for the NK cell integrin, LFA-1, following up on previous evidence that LFA-1 is important for immune cell-mediated recognition of iRBCs.

They start by incubating LFA-1 with iRBCs and show by flow analysis that a substantial population of these iRBCs binds to the LFA-1 (Fig 1C). They do conduct the control with uninfected RBCs, but put this in the supplementary material. As this is a critical control, I think that it should be moved to Figure 1C as it is essential to allow interpretation of the iRBC data. The authors also do not state which strain of P. falciparum they used (line 144). This is critical information, as different strains have different variant surface antigens and should be included. With these changes, this data seems convincing.

They next incubated LFA-1 with the iRBCs, cross-linked and conducted a pulldown, identifying GP130 as a binding partner. Using cross-linkers is a dangerous strategy as it risks non-specific cross-linking. Did they try without cross-linking and find an interaction?

They raised antibodies to PfGBP and showed IFA, which reveals that these antibodies stain iRBCs (Figure 2Ciii). This experiment lacks a critical control of uninfected RBCs, which needs to be included to show that the staining is specific. Without this, it is not possible to conclude that there is iRBC-specific staining with PfGBP.

They then conduct a pulldown using LFA-Fc, which does show GP130 only in the presence of the LFA-Fc, but not when empty beads are used. This is convincing. BLI measurements are also used to study this interaction (Figure 2Ci). The BLI data is presented in such a way that any association phase is obscured by the y-axis, which makes it impossible to know whether there is binding here. I think that the data needs to be shown with some baseline before the addition of the ligand so that association can be seen. The data is also a bit messy with a downward drift and the curves showing different shapes, for example, with the 1.0uM curve seeming to have a different association rate. As this is the only data which shows a direct interaction between LFA1 and GBP, as pulldowns are done with lysates, which might mean bridging components. I think that it is important to repeat the BLI, or use additional biophysical methods to assess binding, to obtain more convincing data.

The authors next do some modelling of the putative complex. This is done by homology modelling and docking, which is not the most up-to-date method and is overinterpreted. Personally, I would remove this data as I did not find it convincing and it is not important for the story. If the authors wish to include it, then I think that they should validate the modelling by mutagenesis to show that the residues which the models indicate might bind are involved in the interaction.

They next made GP130 and tested the binding of this to THP-1 cells, which are often used as a model for macrophages. They observe greater binding of PfGBP-Fc to these cells when compared with hIgG and show that LFA-1 siRNA reduces this binding. I was a little confused about how the flow plots related to the graph in the bottom right corner of Figure 3Bii. In the flow plots, hIgG control shows 12.8% of cells in the gated region, while the unstained cells has 5.63%, but the MFI data shows a decrease in binding for hIgG vs unstained cells. How is this consistent? Also the siRNA reduces the number of cells in gated region from 66.6% to 25.9%, which is still substantially more that 5.63% in the unstained control. This also doesn't seem quite consistent with the MFI data. Could the authors explain this? Also perhaps an additional experiment would be to add soluble LFA-1 into this assay as an additional control to determine whether this blocks PfGBP binding to the THP-1 cells? It could. Be that there are additional mechanisms of binding which indicate why the siRNA has a partial effect. The same is true for the NK cell experiments in Figure 3Ci in which the siRNA has a partial effect. The authors also test binding to HEK, HepG2 and 'stem' cells and claim 'only background levels of binding', but in each case, there is more binding to these cells by PfGBP-Fc than by hIgG, albeit less than in THP-1 and NK cells. Why have the authors decided that these increases are not significant? All in all, these experiments do indicate a role for the GBP-LFA1 interaction in the binding of immune cells to iRBCs, but perhaps not as absolutely as is suggested.

The authors next produce CHO cells with PfGBP on the surface. These cells bind to LFA-1 specifically. When these cells were incubated with primary NK cells, they did see increases in activation markers, which were reduced by addition of antiCD11a, suggesting these to be specific. They also conduct the same experiment with anti-GBP with iRBCs but this is in a different figure. It would be easier for the reader if Figure 5B were in the same figure as Figure 4B as it is related data using the same method. I found this data convincing, showing that the LFA1:GBP interaction does contribute to immune cell recognition and activation.

The authors next conduct an experiment in which they assess parasite growth in the presence of NK cells and in the presence of anti-GBP. They use Heochst staining as a measure of parasite growth and claim that NK cells reduce the number of parasites, but that anti-GBP abolishes this effect (Figure 5A). I found this experiment very unconvincing as there are small effects and no demonstration of significance. More commonly used approaches to study parasite growth are lactate dehydrogenase GIA assays or calcein-AM labelling. I did not find this experiment convincing and would either remove or supplement with additional data using a more robust assay, with repeats and tests of statistical significance.

In summary, the authors present a set of data which comes together to indicate an interaction between LFA1 and PfGBP on the Plasmodium infected erythrocyte surface. Pulldown studies show convincingly that these two proteins co-precipitate and BLI data suggest that this is direct. Also convincing is that NK cell activation can be reduced using antibodies against either LFA1 or PfGBP, indicating that this interaction does play a role in immune cell recognition of iRBCs.

Comments on revised version:

The authors made some minor changes in response to my review, but did not present any substantial new data to demonstrate a direct interaction between PfGBP and LFA1 or to convincingly show differences in NK cell-mediated killing.

Reviewer #2 (Public review):

Summary:

The authors used an LFA-1 αI-Fc fusion protein to pull down potential ligands and LC-MS/MS, leading to selection of PfGBP-130 as a potential membrane protein on the surface of infected cells. PfGBP-130 antibodies were raised and used to support the surface localization. This putative ligand interacted strongly with LFA-1 (Kd = 15 nM). A presumed PfGBP-130 ectodomain interacts with monocytes and NK cells but not cells that lack LFA-1. PfGBP-130 antibodies also interfered with NK cell-mediated infected cell killing; the effect, although statistically significant, is modest. The authors propose that NK cells recognize infected cells via LFA-1 interaction with PfGBP-130 exposed on the host cell and that this interaction is critical to initiation of NK cell activation and killing of infected cells.

Comments on revised version:

The authors submit a minimally revised manuscript that does not address any of my comments, as itemized here:

(1) This reviewer suggested immunoblotting with hypotonic lysis and alkaline extraction as a simple test of whether PfGBP-130 is a membrane protein as the authors propose despite PEXEL cleavage that removes a signal peptide they originally proposed to be a TM domain. Instead of performing this simple immunoblot, the authors state that it is unnecessary because their LC-MS/MS of membrane-associated proteins recovered PfGBP-130, it must be a membrane protein. Unfortunately, this is insufficient because the high sensitivity of LC-MS/MS leads to detection of many soluble proteins. (For example, it is almost certain that their LC-MS/MS recovered hemoglobin, which is soluble and not a surface-exposed protein on infected cells.)

(2) I also suggested a simple immunoblot using a few different immature-stage cultures to detect the full-length and pre-proteins of PfGBP-130 because their immunoblot detected only a 95 kDa band whereas the PEXEL-processed protein is expected to migrate at 85 kDa. The authors state this is unnecessary because their LC-MS/MS of LFA-1 pulldowns enriched for PfGBP-130 and that a single band was detected in immunoblots. This is insufficient because pulldowns often enrich for more than one protein (e.g. some proteins adsorb onto the immunoprecipitation beads or precipitate with beads in certain buffers); immunoblotting often fails to detect some proteins depending on stringency of blocking and wash buffers. They state that the processed form at 85 kDa "may not be well resolved under our current conditions" as a reason not to perform the simple experiment. This reviewer's original statement that P. falciparum antigens frequently cross-react with nominally specific antibodies (with two examples provided in my original review) remains an important concern that would undermine the authors' main conclusion.

(3) As PfGBP-130 is not essential, a knockout was suggested to more directly test their model given the above concerns. The authors state this cannot be done and that their "multiple orthogonal approaches" suggest it is unnecessary. This reviewer considers this an essential experiment to support a provocative, fundamentally new finding, such as the identification of the NK cell activation ligand.

(4) This reviewer suggested that the authors add some speculation about why PfGBP-130 is retained in parasites if triggers NK cell-mediated killing and is nonessential. Rather than adding relevant hypotheses to the Discussion, the authors appear to dismiss this suggestion by stating that PfEMP1, STEVOR, and RIFIN are retained despite being nonessential. The problem with this response is that each of these other antigens has a clearly defined role on the surface of infected erythrocytes that benefits the parasite. It is not clear that the authors have considered possible advantages the parasite may gain from exposing PfGBP-130 on the red cell surface.

Reviewer #3 (Public review):

Summary:

Malhotra and colleagues present evidence that the integrin LFA-1 on NK cells is a ligand for the Plasmodium falciparum protein GBP130 on the infected erythrocyte surface and that this interaction plays a role in the clearance of infected erythrocytes by NK cells.

The authors first select a subdomain contained within the CD11a subunit of LFA-1 as a probe to discover possible binding proteins on the infected erythrocyte surface. Parasite-infected erythrocytes stained positively with this probe; the level of staining increased as the parasites progressed through the life cycle. Using the LFA-1-based probe in cross-linking pull-down experiments, GBP130 was identified by mass spectrometry as a co-purifying parasite protein. The N-terminal portion of GBP130 was recombinantly expressed and shown to interact with LFA-1 alpha-I by biolayer interferometry experiments. The full-length extracellular domain of GBP130 was then recombinantly expressed and used to stain primary human NK cells and THP-1 cells. Knocking down LFA-1 by siRNA reduced staining by GBP130. To assess the contribution of GBP130 to the activation of NK cells, CHO cells exogenously expressing GBP130 were incubated with primary NK cells. Transfecting CHO cells with GBP130 led to increased activation of co-incubated NK cells compared to mock-transfected and compared to GBP130 transfected cells, with the inclusion of anti-CD11a to block NK cell adhesion. Finally, CHO cells expressing GBP130 led to increased activation of NK cells compared to mock-transfected CHO cells.

Overall, although the authors present data from NK cell killing assays that include appropriate controls, the data suggesting a direct interaction between PfGBP-130 and LFA-1 does not include the same necessary controls, for example, the use of blocking antibodies. Most critically, the biolayer interferometry experiments use a recombinant fragment of PfGBP-130, which does not include the residues predicted to be important for mediating specific interaction with LFA1. The biolayer interferometry data instead suggest non-specific interactions between PfGBP-130 and LFA1, as binding does not reach saturation.

Comments on revised version:

The authors have addressed all minor concerns, however the major point regarding the biophysical data supporting direct interaction between PfGB130 and LFA-1, in my opinion, has not been satisfactorily addressed. Biophysical data supporting the interaction was generated using a fragment of PfGB130, which does not include residues that the authors predict by structural modelling to be important for the interaction. The authors argue that PfGB130 is a repeat containing protein and may have multiple binding sites for LFA-1. If this is the best mechanistic hypothesis given the current data, the authors need to explain this in the results section.

Overall though, I agree with Reviewer#1 that the structural modelling results are not convincing and given that the modelling data do not straightforwardly agree with the experiment, the clarity of the manuscript would benefit from their omission.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

(1) They start by incubating LFA-1 with iRBCs and show by flow analysis that a substantial population of these iRBCs binds to the LFA-1 (Figure 1C). They do conduct the control with uninfected RBCs, but put this in the supplementary material. As this is a critical control, I think that it should be moved to Figure 1C as it is essential to allow interpretation of the iRBC data. The authors also do not state which strain of P. falciparum they used (line 144). This is critical information as different strains have different variant surface antigens and should be included. With these changes, this data seems convincing.

We thank the reviewer for this important suggestion. We agree that the uninfected RBC (uRBC) control is critical for interpreting the specificity of LFA-1 αI-Fc binding. In the revised manuscript, we have ensured that these control data are clearly presented and appropriately referenced in the main text; however, we have retained them in the Supplementary Information (Supplementary Figure S1) to maintain clarity and avoid overcrowding Figure 1, while still ensuring their visibility and accessibility to the reader. Importantly, these data demonstrate negligible binding of LFA-1 αI-Fc to uRBCs compared to iRBCs, supporting specificity. We have explicitly stated the parasite strain used (Plasmodium falciparum 3D7) in the Methods section (line 475).

(2) They next incubated LFA-1 with the iRBCs, cross-linked and conducted a pulldown, identifying GP130 as a binding partner. Using cross-linkers is a dangerous strategy as it risks non-specific cross-linking. Did they try without cross-linking and find an interaction?

We agree that cross-linking can introduce potential artefacts. To mitigate this, we included hIgG control pulldown experiments performed under identical conditions. Proteins identified in the control eluate were excluded as background (summarized in Supplementary Table S1). Importantly, PfGBP-130 was the only protein specifically enriched in the LFA-1 αI-Fc pulldown across all three biological replicates (Fig. 2A, Venn Diagram). While cross-linking was used to stabilize transient interactions, consistent enrichment of PfGBP-130 across the three biological replicates precludes any concerns of non-specificity.

(3) They raised antibodies to PfGBP and showed IFA, which reveals that these antibodies stain iRBCs (Figure 2Ciii). This experiment lacks a critical control of uninfected RBCs, which needs to be included to show that the staining is specific. Without this, it is not possible to conclude that there is iRBC-specific staining with PfGBP.

The question pertains to Fig. 2Biii. The IFA images include both infected and neighboring uninfected erythrocytes within the same field. No PfGBP-130 staining is observed in uninfected cells. PfGARP staining, specifically done to verify parasite-infected cell and surface localisation, shows complete resonance with PfGBP-130 staining. This unequivocally shows that the antibodies raised specifically recognise only infected RBCs.

(4) They then conduct a pulldown using LFA-Fc, which does show GP130 only in the presence of the LFA-Fc, but not when empty beads are used. This is convincing. BLI measurements are also used to study this interaction (Figure 2Ci). The BLI data is presented in such a way that any association phase is obscured by the y-axis, which makes it impossible to know whether there is binding here. I think that the data needs to be shown with some baseline before the addition of the ligand so that the association can be seen. The data is also a bit messy with a downward drift and the curves showing different shapes, for example, with the 1.0uM curve seeming to have a different association rate. Also, is this n=1? I think that this data needs to be repeated and replicated. As this is the only data which shows a direct interaction between LFA1and GBP, as pulldowns are done with lysates, which might mean bridging components. I think that it is important to repeat the BLI or use additional biophysical methods to assess binding, to obtain more convincing data.

We sincerely thank the reviewer for highlighting this important concern regarding the BLI data presentation and interpretation. We would like to clarify that the baseline signal prior to ligand addition was subtracted during data processing; therefore, the plotted curves represent the net response following ligand association. However, we agree that this may have obscured the visualization of the association phase. Accordingly, in the revised manuscript, we have re-plotted the data with adjusted y-axis scaling to better capture the association kinetics. In addition, to ensure robustness and reproducibility, the BLI experiments were performed in multiple independent replicates (n ≥ 3) using independently purified protein batches. The original figure showed a representative dataset; we have now included averaged sensorgrams along with standard deviation in the calculated KD values [K_D = (1.7 ± 0.22) × 10^-8 M] (Figure 2C (i)). These revisions provide a clearer and more accurate representation of the binding interaction.

(5) The authors next do some modelling of the putative complex. This is done by homology modelling and docking, which is not the most up-to-date method and is over-interpreted. Personally, I would remove this data as I did not find it convincing, and it is not important for the story. If the authors wish to include it, then I think that they should validate the modelling by mutagenesis to show that the residues which the models indicate might bind are involved in the interaction.

We thank the reviewer for this thoughtful comment regarding the modelling analysis. We agree that computational docking and homology-based modelling have inherent limitations and should not be over-interpreted. In our study, these analyses were included strictly as supporting evidence to provide a structural framework for the PfGBP-LFA-1 interaction, while the primary conclusions are based on direct biochemical and functional validation, including pull-down, BLI measurements, receptor knockdown, and cellular inhibition assays. Importantly, the use of docking approaches such as ClusPro, followed by interface analysis and MD simulations, is a widely accepted and routinely used strategy to generate testable hypotheses for protein-protein interactions, particularly when experimental structures are unavailable (e.g., Comeau et al., 2004; Weng et al., 2019). We believe that the current modelling serves as a useful complementary analysis that is consistent with, and supportive of, the experimentally validated interactions.

(6) They next made GP130 and tested the binding of this to THP-1 cells, which are often used as a model for macrophages. They observe greater binding of PfGBP-Fc to these cells when compared with hIgG and show that LFA-1 siRNA reduces this binding. I was a little confused about how the flow plots related to the graph in the bottom right corner of Figure 3Bii. In the flow plots, hIgG control shows 12.8% of cells in the gated region, while the unstained cells has 5.63%, but the MFI data shows a decrease in binding for hIgG vs unstained cells. How is this consistent? Also, the siRNA reduces the number of cells in the gated region from 66.6% to 25.9%, which is still substantially more that 5.63% in the unstained control. This also doesn't seem quite consistent with the MFI data. Could the authors explain this? Also, perhaps an additional experiment would be to add soluble LFA-1 into this assay as an additional control to determine whether this blocks PfGBP binding to the THP-1 cells? It could be that there are additional mechanisms of binding which indicate why the siRNA has a partial effect. The same is true for the NK cell experiments in Figure 3Ci, in which the siRNA has a partial effect. The authors also test binding to HEK, HepG2 and 'stem' cells and claim' only background levels of binding', but in each case, there is more binding to these cells by PfGBP-Fc than by hIgG, albeit less than in THP-1 and NK cells. Why have the authors decided that these increases are not significant? All in all, these experiments do indicate a role for the GBP-LFA1 interaction in the binding of immune cells to iRBCs, but perhaps not as absolutely as is suggested.

We thank the reviewer for this insightful comment. The apparent discrepancy arises because the flow plots depict the percentage of cells within a defined positive gate, whereas the graphs quantify mean fluorescence intensity (MFI) across the entire population. We have revised figure legend accordingly to indicate the same. Regarding the partial reduction in binding upon LFA-1 (CD11a) knockdown, we agree that this indicates LFA-1 is a major but not exclusive contributor, which is biologically plausible given incomplete siRNA depletion and the known avidity-dependent nature of integrin interactions. Importantly, our conclusion is supported by multiple orthogonal approaches (αI-domain binding, LC-MS/MS identification, BLI, docking, receptor knockdown, and functional blockade). We also appreciate the suggestion of soluble LFA-1 competition, which we acknowledge as an important future experiment. Finally, we have revised the text regarding HEK293T, HepG2, and stem cells to reflect that PfGBP-Fc binding is minimal but not absent, consistent with low/non-expression of LFA-1 in non-immune cells. Overall, we have moderated our claims to state that PfGBP-LFA-1 interaction is a dominant and functionally relevant mechanism, while not excluding additional low-affinity or accessory interactions.

Figure legend change: Representative flow plots depict the percentage of cells within a predefined positive gate, whereas the accompanying summary graph quantifies fluorescence intensity across the analyzed population. These two metrics report distinct properties of the distribution and are therefore not expected to be numerically identical.

(7) The authors next produce CHO cells with PfGBP on the surface. These cells bind toLFA-1 specifically. When these cells were incubated with primary NK cells, they did see increases in activation markers, which were reduced by the addition of anti-CD11a, suggesting these to be specific. They also conduct the same experiment with anti-GBP with iRBCs, but this is in a different figure. It would be easier for the reader if Figure 5B were in the same figure as Figure 4B, as it is related data using the same method. I found this data convincing, showing that the LFA1:GBP interaction does contribute to immune cell recognition and activation.

We thank the reviewer for this positive assessment and helpful suggestion regarding figure organization. We agree that the CHO-PfGBP and iRBC-based NK cell activation assays represent conceptually related experiments that both address LFA-1-PfGBP dependent activation using similar readouts. We have retained separate panels to distinguish the reductionist CHO-based system from the physiologically relevant iRBC context. We believe that the combined evidence from both systems strengthens the conclusion that PfGBP-LFA-1 interaction is a key contributor to NK cell recognition and activation.

(8) The authors next conduct an experiment in which they assess parasite growth in the presence of NK cells and in the presence of anti-GBP. They use Heochst staining as a measure of parasite growth and claim that NK cells reduce the number of parasites, but that anti-GBP abolishes this effect (Figure 5A). I found this experiment very unconvincing as there are small effects and no demonstration of significance. More commonly used approaches to study parasite growth are lactate dehydrogenase GIA assays or calcein-AM labelling. I did not find this experiment convincing and would either remove or supplement with additional data using a more robust assay, with repeats and tests of statistical significance.

We respectfully disagree that the assay should be removed, because flow-cytometric quantification of P. falciparum parasitemia using DNA dyes such as Hoechst is a widely used, accepted, and high-throughput approach for measuring infected erythrocytes and parasite growth, with clear separation of infected from uninfected RBCs and good reproducibility across malaria studies (Dent et. al., 2009; Jang et. al., 2014). Importantly, closely related immune-cell killing experiments in the malaria field have used the same general strategy, co-culture with effector cells followed by flow-cytometric enumeration of parasitemia to infer parasite control, including the seminal NK-cell study by Chen et. al., 2014, which our assay design follows conceptually, and later work showing reduced parasitemia after co-incubation with cytotoxic lymphocytes measured by nucleic-acid dye flow cytometry. We therefore believe the experiment is methodologically valid and directly relevant to the biological question, namely whether disrupting PfGBP-LFA-1 engagement alters NK-cell-mediated restriction of parasite expansion.

Reviewer #2 (Public review):

(1) PfGBP-130 is proposed to be a membrane protein based on a single predicted transmembrane domain. Figures 2b and 3a show ribbon schematics with this TM domain at residues 51-68, in agreement with TM prediction algorithms such as TMHMM 2.0 and Phobius. However, this predicted TM is upstream of the PEXEL motif (residues 84-88, sequence RILAE), a conserved sequence for parasite protein export to host cytosol that is proteolytically processed at its 4th residue. Thus, residues 1-87are removed from PfGBP-130 prior to export, yielding a mature protein without predicted TMs. Prior studies have determined that the mature PfGBP-130 lacks TMs and is retained as a soluble protein in host cell cytosol (PMID: 19055692, 35420481). Thus, the authors' model of PfGBP-130 as a surface-exposed membrane protein conflicts with both computational analysis of the mature protein and these prior reporter studies. An important simple experiment would be to evaluate PfGBP-130membrane association in immunoblots using the authors' PfGBP-130 antibody after hypotonic lysis (PMID: 19055692) and after alkaline extraction (e.g. 100 mM NaCO3, pH 11 as frequently used, PMID: 33393463). If the prior studies and computational analyses are correct, the protein will be predominantly in the soluble and/or alkaline supernatant fractions.

We thank the reviewer for this important observation regarding PfGBP-130 topology and export. We agree that the presence of a PEXEL motif supports proteolytic processing and that the mature protein may lack a classical transmembrane domain. However, consistent with our model of surface accessibility, we would like to clarify that in an independent proteomic study performed in our laboratory on the membrane-enriched fraction of Plasmodium falciparum-infected erythrocytes, PfGBP-130 was reproducibly identified by LC-MS/MS among membrane-associated proteins (data not shown; can be provided upon request). These findings support the conclusion that, irrespective of the absence of a canonical transmembrane domain, PfGBP-130 is associated with the iRBC membrane compartment, likely via peripheral or protein-complex–mediated interactions, as described for several exported Plasmodium proteins.

(2) Many findings rely on the specificity of antibodies generated against PfGPB-130 or NK cell receptors. Although the authors have included key controls (use of isotype control antibodies, lack of anti-PfGBP-130 binding to uninfected cells), cross-reactivity between P. falciparum antigens is well-recognized and could significantly undermine the interpretation of experiments (PMID: 2654292 and 1730474 provide key examples of antigens recognized by antibodies raised against other proteins). For example, the surface localization in IFA experiments (Figure 2B(iii)) could reflect anti-PfGBP-130binding to an unrelated parasite surface antigen, a possibility not addressed by any of the authors’ controls. As another example, the iRBC lysate immunoblot using this antibody in Fig. 2B(iv) suggests a MW of 95 kDa, which corresponds to the unprocessed pre-protein before export; cleavage in the PEXEL motif yields a processed mature protein of 85 kDa, which should be readily resolved from the pre-protein in immunoblots (PMID: 19055692). A better immunoblot using immature infected cell stages might show both the pre-protein and the mature protein as a doublet band.

We thank the reviewer for raising this important concern regarding antibody specificity. We agree that cross-reactivity among P. falciparum antigens is a known issue and have taken multiple steps to ensure specificity in our study. First, the anti-PfGBP-130 antibodies were generated against a defined recombinant fragment and show no detectable binding to uninfected RBCs and no signal in hIgG control immunoprecipitates, supporting specificity. Importantly, in our LC-MS/MS analysis of LFA-1 αI-domain pull-downs, PfGBP-130 was specifically enriched and consistently identified across replicates, independently validating the target recognized by the antibody. Furthermore, the same antibody detects a single dominant band in both iRBC lysates and αI pull-down fractions, arguing against widespread cross-reactivity. Regarding the apparent molecular weight (~95 kDa), we agree that this likely corresponds to the precursor form, and that a processed form (~85 kDa) may not be well resolved under our current conditions.

(3) PfGBP-130 is not essential for in vitro cultivation (PMID: 18614010 and MIS of 1.0 in the piggyBac mutagenesis screen as tabulated on plasmodb.org, indicating a highly dispensable gene). The authors should use the knockout line as a control in their IFA localization experiments to address antibody specificity. More fundamentally, their model predicts that NK cells should not recognize or kill infected cells from the knockout line when compared to their untransfected parent. Such results with the knockout line would compellingly support the authors' model without reliance on antibodies that may cross-react with other parasite antigens. PMID: 18614010reported that the PfGBP-130 knockout exhibited increased membrane rigidity, suggesting an intracellular scaffolding protein rather than a surface localization and use as a ligand for LFA-1 interaction and NK cell-mediated killing.

We agree that a PfGBP-130 knockout line would provide a powerful genetic validation of both antibody specificity and the proposed functional role of PfGBP-130 in NK cell recognition. At present, such experiments were not included in this study, and we acknowledge this as an important limitation. However, we would like to emphasize that our conclusion does not rely on antibody-based localization alone; rather, it is supported by multiple orthogonal approaches, including LFA-1 αI-domain pull-down coupled to LC-MS/MS, biophysical interaction analysis, receptor knockdown, and functional blocking assays. In addition, in one of our previous proteomic analyses of the membrane-enriched fraction of infected erythrocytes, PfGBP-130 was identified among the proteins present in the membrane fraction, supporting its association with the iRBC membrane compartment despite lacking a classical mature transmembrane domain.

(4) PfGBP-130 non-essentiality raises the question of why the gene would be retained if it triggers NK cell-mediated killing of infected cells in vivo. Presumably, this killing would pose strong selective pressure against retention of PfGBP-130. Some speculation is warranted to support the model.

We thank the reviewer for this thoughtful evolutionary question. We agree that if PfGBP-130 enhances NK-cell recognition, its retention likely reflects a context-dependent fitness trade-off rather than a simple benefit or cost. This situation is not unusual in P. falciparum: several exported or surface-associated proteins are retained despite being immunogenic because they also provide advantages in other settings, such as erythrocyte remodeling, cytoadhesion, niche adaptation, immune modulation, or transmission. The clearest precedent is the PfEMP1/var system, in which highly immunogenic surface antigens are nevertheless strongly maintained because they mediate sequestration and in vivo fitness, while antigenic variation limits continuous immune exposure (Chew et. al., 2022). Similarly, other variant surface antigens such as STEVOR and RIFIN are retained despite immune recognition because they contribute to erythrocyte binding, antigenic diversity, and immune evasion or modulation (Niang et. al., 2009; Sakoguchi et. al., 2025). More broadly, many P. falciparum genes that appear dispensable in standard in vitro culture are nevertheless preserved because culture does not recapitulate the selective pressures present in vivo, including splenic clearance, endothelial interactions, immune attack, and within-host competition.

Reviewer #3 (Public review):

(1) Anti-GBP130 antibodies are used in the cellular assays to block the interaction between GBP130 and LFA1. They should therefore also block interactions betweenGBP130 and LFA1 recombinant proteins in the biolayer interferometry experiment. Do the authors have data to show this? Similarly, the anti-CD11a antibodies used to block the interaction in the cellular assays should also block the in vitro interaction between recombinant LFA1 and GBP130.

We thank the reviewer for this insightful suggestion. We agree that demonstrating antibody-mediated inhibition of the recombinant PfGBP-LFA-1 interaction would provide an additional orthogonal validation of the interface. While such blocking experiments were not included in the original BLI dataset, our current study already establishes the specificity of this interaction through multiple independent approaches, including αI-domain pull-down and LC-MS/MS identification, BLI-derived high-affinity binding (KD ~10^-8 M), structural docking, receptor knockdown, and antibody-mediated inhibition in cellular systems. We note that antibody-mediated blocking in a purified biophysical system is not always directly comparable to cellular assays, as epitope accessibility, orientation on biosensor surfaces, and conformational states of integrins (which are known to undergo activation-dependent structural changes) can influence inhibition efficiency. Nonetheless, we fully agree that this represents an important validation experiment.

(2) The structural modelling analysis of the predicted complex between GBP130 andLFA1 (Figure 2cii) predicts that the majority of the important GBP130 interface residues are located in the region D509-N607. However, the authors present BLI data for the GBP130-LFA1 interaction, which used the N-terminal fragment of GBP (residues 69-270), which does not include the GBP130 residues predicted to be important for the formation of the complex between the two proteins. Could the authors provide an explanation for how an interaction was observed with theGBP130-N fragment, which does not contain the residues predicted to be important for interacting with LFA1?

We thank the reviewer for this important observation. We agree that the structural model predicts a major interaction interface within the D509-N607 region of PfGBP-130; however, this does not preclude the existence of additional or auxiliary binding determinants within the N-terminal region used in our BLI assays (aa 69-270). PfGBP-130 is a multi-domain, repeat-containing protein, and such proteins frequently exhibit distributed or multivalent interaction interfaces, where individual regions can independently engage binding partners with lower affinity while the full-length protein achieves higher avidity through cooperative interactions. In our study, the BLI data using the N-terminal fragment demonstrate that this region is sufficient to mediate direct interaction with the LFA-1 αI domain, whereas the structural model based on full-length predictions likely captures a dominant or higher-affinity interface in the C-terminal region. Importantly, the interaction is supported by multiple orthogonal datasets, including pull-down/LC-MS/MS, cellular binding assays, and functional inhibition, indicating that the observed binding is not an artefact of fragment choice.

Author response image 1.

To further examine this, we performed docking and binding energy analyses comparing the full-length PfGBP-130-LFA-1 complex with the N-terminal domain-LFA-1 complex. Using the PRODIGY server, the predicted binding affinity for the full-length complex was -9.8 kcal/mol, whereas the N-terminal domain complex exhibited a still favorable binding energy of -5.6 kcal/mol. Similarly, HawkDock (v2) analysis yielded binding energies of -22.2 kcal/mol for the full-length complex and -14.1 kcal/mol for the domain-only complex. While reduced relative to the full-length protein, these values remain well within the range of stable protein-protein interactions, supporting the ability of the N-terminal region to independently contribute to binding. These energy calculations take into account all non-covalent interactions. For clarity, hydrogen bonds have been specifically highlighted in the figure to represent key interaction interface.

(3) There is no section in the materials and methods describing how the BLI was performed; this should be added. The highest concentration ofGBP130 used in the interaction measurements is 1.4uM, almost 100x the measured Kd (0.015uM) for the GBP130-LFA1 interaction. At these high concentrations ofGBP130, I would expect to start seeing saturation of binding, but the interferometry curves show that saturation is not close to being reached. This strongly suggests that the binding of GBP130 to LFA1 is non-specific.

We thank the reviewer for raising these important technical points. We have included a detailed description of the biolayer interferometry (BLI) methodology in the Materials and Methods section in the manuscript. Regarding the concern about lack of saturation at higher analyte concentrations, we respectfully disagree that this necessarily indicates non-specific binding. In BLI assays, incomplete saturation can arise from several well-recognized factors, including suboptimal orientation or partial inaccessibility of immobilized ligand on the biosensor, mass transport limitations, or heterogeneous binding populations particularly relevant for integrins such as LFA-1, whose αI domain exists in multiple conformational states with distinct affinities. Importantly, the interaction exhibits clear concentration-dependent association and dissociation kinetics that fit a 1:1 binding model with a KD in the nanomolar range, which is inconsistent with non-specific interactions that typically show poor fitting and minimal dissociation. Furthermore, the specificity of the PfGBP-LFA-1 interaction is supported by multiple independent lines of evidence in our study, including selective enrichment in αI-domain pull-downs, absence in IgG controls, reduction upon CD11a knockdown, and functional inhibition by blocking antibodies in cellular assays. We have now clarified these points in the revised manuscript and tempered the interpretation to acknowledge potential experimental constraints of BLI while maintaining that the cumulative data strongly support a specific interaction.

Minor points:

(1) For the pulldown experiments, can the authors confirm that cross-linking was also performed for the protein A beads + hIgG control?

Yes, DTSSP cross-linking was performed identically in the protein A beads + hIgG control arm. This is consistent with the control design described in the manuscript.

(2) If the recombinant CD11a I subdomain used as a probe is correctly folded and functional, it should bind ICAM1. Do the authors have this data?

We agree that ICAM-1 binding is an important functional validation for the recombinant CD11a αI probe (Hogg et. al., 1998). The isolated αI domain of LFA-1 is well established as the principal ICAM-1-binding module, and soluble αI-domain reagents have previously been shown to bind/block ICAM-1 interactions. We did not include this control in the current version.

(3) Were the authors able to perform the reciprocal pull-down, using pfGBP130-N-Fc to pull down LFA1 from cell surfaces?

We did not perform a reciprocal pull-down with PfGBP130-N-Fc and native cell-surface LFA-1 in the present study; we agree this would be a useful orthogonal experiment.

(4) After identifying GBP130 as a co-purifying protein in the LFA-1 pull-down experiments, the authors select an N-terminal fragment of GBP130 to recombinantly express and use. How did the authors narrow down which region of GBP130interacted with LFA-1?

The N-terminal PfGBP130 fragment (aa 69-270) was selected empirically as a tractable, soluble recombinant segment containing a defined repeat-containing extracellular region, rather than because we had already mapped the full LFA-1-binding interface. We agree with the reviewer that our structural model suggests that additional residues, including a likely dominant interface outside this fragment, may contribute to the full interaction, and we have clarified that the N-terminal fragment should be interpreted as a minimal binding-competent region, not necessarily the sole binding site.

(5) As erythrocytes age, their surface undergoes biochemical changes, most notably a drop in levels of sialylation, decreasing the net repulsive negative charge, and they generally become more adherent. Can the authors exclude the possibility that, rather than binding to a parasite-derived ligand, LFA alpha 1 is instead binding to a marker of older erythrocytes? In the data presented, increased binding of LFA alpha 1 is observed as parasites progress through the life cycle, but the host erythrocytes will be ageing during parasite replication, which could account for the increased levels of LFA alpha 1 binding. To rule out this explanation, data from LFA alpha 1 staining of age-matched uninfected erythrocytes could be provided.

We agree that erythrocyte aging can alter surface sialylation and adhesiveness, and loss of sialic acid is known to reduce erythrocyte surface charge and increase adhesiveness. However, our data argue against aging alone explaining the signal, because LFA-1 αI-Fc binding was compared with uninfected RBC controls and the interaction led to enrichment of a parasite-derived ligand, PfGBP130, in pull-down/MS analyses.

(6) Figure 3b(i) Surface staining of THP1 cells was performed using GBP-130 Fc as a probe, which should detect all LFA1-positive cells. But no accompanying staining data using an anti-LFA1 antibody are shown, so it is not possible to determine whether staining profiles with GBP-130 Fc match staining profiles with anti-LFA1 antibodies. This is important to show what proportion of LFA1-positive cells can recognise parasite-derived GBP-130 Fc.

(7) Figure 3c(i) Surface staining of peripheral NK cells is performed using GBP-130 Fc as a probe, which should detect all LFA1-positive cells. Here, as well, there are no staining data using an anti-LFA1 antibody. This would allow a comparison between cell population LFA1 staining with an anti-LFA1 antibody and cell population LFA1 staining with GBP-130 Fc. The two staining profiles should be similar as both probes bind the same surface marker. However, it appears this might not be the case because the staining data using GBP-130 Fc show that only a minor proportion of NK cells (~20%) stain positive, but the majority of peripheral NK cells usually express CD11a, as it is a key adhesion molecule in the formation of immune synapses with target cells. This suggests that GBP-130 can only bind to a subset of NK cells, and if it is binding LFA1, then it can only play a role in mediating the formation of an immune synapse with this subpopulation of NK cells. Could the authors include a comment in the manuscript making clear that the GBP-130 only assists a small proportion of NK cells in adhering to parasite-infected erythrocytes? Are there any reasonable hypotheses as to whyGBP-130 was only able to stain a small subpopulation of LFA1-expressing NK cells?

For minor comment 6 and 7

We agree that parallel staining with anti-CD11a would help relate PfGBP130-Fc binding to total LFA-1-positive THP-1 and NK-cell populations. Importantly, LFA-1 expression and ligand binding competence are not equivalent, because integrin binding depends strongly on activation/conformation and avidity state; in NK cells, only a subset can display LFA-1 in a partially activated conformation at baseline despite broader CD11a expression. Thus, a smaller PfGBP130-Fc-positive subset than the total CD11a-positive population is biologically plausible and does not imply inconsistency.

eLife Assessment

This manuscript investigates inter-hemispheric interactions in the olfactory system of Xenopus tadpoles. Using a combination of electrophysiology, pharmacology, imaging, and uncaging, the transection of the contralateral nerve is shown to lead to larger odor responses in the un-manipulated hemisphere, and implicates dopamine signaling, likely originating from the lateral pallium, in this process. The study convincingly uses a rich and sophisticated array of tools to investigate olfactory coding, and uncovers valuable mechanisms of signaling likely to be conserved across vertebrates.

Reviewer #1 (Public review):

In this study, the authors investigate responses to methionine in the olfactory system of the Xenopus tadpole. They show that the LFP response is local to the glomerular layer, arises ipsilaterally, and is blocked by pharmacological blockade of AMPA and NMDA receptors, with little modulation during blockade of GABA-A receptors. They then show that this response is translently enlarged following transection of the contralateral olfactory nerve, but not the optic lobe nerve. Measurement of ROS- a marker of inflammation- was not affected by contralateral nerve transection, and LFP expansion was not affected by pharmacological blockade of ROS production. Imaging biased towards presynaptic terminals suggests that the enlargement of the LFP has a presynaptic component. A D2 antagonist increases the LFP size and variability in intact tadpoles, while a GABA-B antagonist does not. Finally, the authors provide anatomical and physiological evidence that the contralateral dopamine signal may arise from the lateral pallium. Overall, I found the array of techniques and approaches applied in this study to be creatively and effectively employed.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

In this study, the authors investigate LFP responses to methionine in the olfactory system of the Xenopus tadpole. They show that this response is local to the glomerular layer, arises ipsilaterally, and is blocked by pharmacological blockade of AMPA and NMDA receptors, with little modulation during blockade of GABA-A receptors. They then show that this response is translently enlarged following transection of the contralateral olfactory nerve, but not the optic lobe nerve. Measurement of ROS- a marker of inflammation- was not affected by contralateral nerve transection, and LFP expansion was not affected by pharmacological blockade of ROS production. Imaging biased towards presynaptic terminals suggests that the enlargement of the LFP has a presynaptic component. A D2 antagonist increases the LFP size and variability in intact tadpoles, while a GABA-B antagonist does not. On this basis, the authors conclude that the increase driven by contralateral nerve transection is due to DA signaling.

Overall, I found the array of techniques and approaches applied in this study to be creatively and effectively employed. However, several of the conclusions made in the Discussion are too strong, given the evidence presented. For example, the authors state that "The observed potentiation was not related to inflammatory mediators associated to inury, because it was caused by a release of the inhibition made by D2 dopamine receptor present in OSN axon terminals." This statement is too strong - the authors have shown that D2 receptors are sufficient to cause an increase in LFP, but not that they are required for the potentiation evoked by nerve transection. The right experiment here would be to get rid of the D2 receptors prior to transection and show that the potentiation is now abolished. In addition, the authors have not shown any data localizing D2 receptors to OSN axon terminals.

Similarly, the authors state, "the onset of LFP changes detected in glomeruli is determined by glutamate release from OSNs." Again, the authors have shown that blockade of AMPA/NMDA receptors decreases the LFP, and that uncaging of glutamate can evoke small negative deflections, but not that the intact signal arises from glutamate release from OSNs. The conclusions about the in vivo contribution of this contralateral pathway are also rather speculative. Acute silencing of one hemisphere would likely provide more insight into the moment-to-moment contributions of bilateral signals to those recorded in one hemisphere.

We thank the reviewer for their positive evaluation of our manuscript. We agree with their opinion about the necessity of including new experimental evidence to back up discussion and conclusions

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

This is a creative and careful study, but I felt that the conclusions in the Discussion were too strong. I think these could either be toned down or additional experiments could be done to support the idea that D2 receptors are required for the nerve transection-evoked potentiation, that the source of glutamatergic input is OSNs, and that contralateral interactions are mediated by DA. In particular, I think anatomical stains showing which neurons are carrying the DA signal and whether there is any potentiation of DA release after nerve transection would greatly strengthen the conclusions.

This new version of the manuscript contains two new figures: 6 and 9.

New figure 6 addresses the suggestion of this reviewer and provides anatomical evidence for the distribution of dopaminergic neurons in the olfactory bulb of X. tropicalis tadpoles using a tyrosine hydroxylase antibody (mouse monoclonal, Immunostar cat. no. 22941, 1:250; RRID:AB_57226). We identified a discrete neuronal population present in the border between the mitral cell layer and the glomerular layer that resembles the type1 TH+ population described in adult frogs (Boyd and Delaney 2002). TH+ neurons send their processes to innervate olfactory glomeruli and we provide evidence that they contact the GFP lateral glomerulus labelled in Dre.mxn1:GFP X. tropicalis tadpoles (Fig. 6C). These results reinforce a modulatory role for dopamine on glomerular neurotransmission. Materials & methods (lines 152-167), results (lines 393-399) and discussion (lines 550-563) have been modified accordingly.

Figure 9 provides new evidence on the interhemispheric connections involved in the potentiation of glomerular responses. We first demonstrate that dorsolateral pallial neurons participate in the processing of olfactory information based on the general consideration that the lateral pallium is an olfactory cortex. We confirmed this possibility by stimulating the olfactory epithelium and recording ipsilateral calcium transients in pallial neurons of tubb2b:GCaMP6s tadpoles. We next injured the dorsolateral pallium and 24-48h afterwards we recorded odor-evoked responses in the GFP labelled glomerulus located contralaterally. We observed a ~70% potentiation of responses, which was comparable to the ~75% potentiation obtained by olfactory nerve transection. These results illustrated the involvement of pallial neurons in the control of glomerular output by likely modifying the activity of TH+ neurons. The results (473-506) and discussion (569-576) now include these new results.

Does the contribution of DA signalling change across development? I think this would be helpful to interpret the results and relatively straightforward to do: apply raclopride at different developmental stages and measure how much potentiation occurs at each stage.

This is indeed an interesting point, but conducting a comprehensive study of dopamine release throughout development would require a substantial amount of work and delay the publication of this paper. To perform these experiments, we should first implement new technical approaches, such as successfully injuring young tadpoles or recording from late premetamorphic stages. We believe that the proposed experiments could define a new line of arguments rather than complement the present work. Nonetheless, we acknowledge the suggestion of this reviewer.

In this new version, we provide strong evidence for dopamine release in the glomerular layer, and a key question that arises is the nature of TH+ positive neurons. Recent findings obtained in mice show that there are five different types of dopaminergic interneurons present in the olfactory bulb (Kosaka, Pignatelli, and Kosaka 2020), and important functional differences exist between axon-bearing and anaxonic neurons (Dorrego-Rivas et al. 2025). This evidence suggests a key role for development. A completely new study based on transgenic X. tropicalis displaying labeled TH+ neurons could bring together development, anatomy, and physiology to gain an understanding of how dopaminergic signaling shapes glomerular function.

In addition, there are several places where showing additional raw data in the figures and carefully quantifying variability would be helpful. For example, in Figure 3B, the authors should show equivalent raw traces from intact and transected tadpoles. In Figure 5D, it would be helpful to show raw traces for LFP equivalent to what is shown for presynaptic imaging in Figure 5E. In Figures 6E-F, it would be helpful to show raw traces.

Thank you for this suggestion. The examples have been added to the figure panels.

I found the last experiment with photobleaching somewhat inconclusive, and I am not sure what it adds to the study as presently written. Line 418: Please quantify how many OSNs remained. Line 423: What is the hypothesis for the source of variability?

The goal of this experiment is to investigate the participation of chemotopy in the potentiation induced by contralateral injury. The elimination of 30-50% of topographically related OSNs did not alter contralateral glomerular responses. This evidence suggests that chemotopy was not relevant to the gain of function observed ; however, we cannot completely rule out a certain topographical contribution, as it was not possible to completely silence all inputs of the studied glomerulus. We now link these findings to the likely innervation of several glomeruli by TH+ neurons, which suggests the absence of a one-to-one glomerulus relationship. LFP amplitudes and their variance are now illustrated in box plots to highlight the absence of significant differences. Lines (457-471).

An increase in the variance among the recordings obtained is a consistent empirical observation. Although it is a hallmark of the potentiation recorded, we cannot provide a mechanistic explanation. Considering that neurotransmitter release from OSN axon terminals is normally inhibited by dopamine, we hypothesize that disinhibition drives an increase in release probability , leading to larger variations in glutamate release. Such variations could be reflected in the amplitude of LFP negativities.

It would be helpful to include a measurement of LFP over time so we have some idea of how stable the odor delivery is.

The amplitude of LFP responses was stable for >30 min. Figure 3B shows recordings obtained during 30 min and new Figure 7F over 42 min. We believe that these examples illustrate that the amplitude, as well as kinetics of the responses obtained were consistent over the period studied.

Line 227: Small upward deflection - could this be an electrical artifact? Can you run the stimulus delivery with no odor (say, with water) to see if you get the same signal?

We do not know the precise source of this upward deflection. It is not an electrical artifact related to stimulation, which is sometimes evident (Fig 7A, methionine application). When present, it occurs after the activation of OSNs. One possibility is that the deflection originates in the layer of nerve fibers reflecting some aspect related to the conduction of APs and the relative position of the electrode. Interestingly, some recordings of LFP responses at the level of glomeruli carried out in rats also show a positive deflection (see Figs. 1B, 2A, 3B in (Lecoq, Tiret, and Charpak 2009), thus suggesting it is an intrinsic characteristic of this type of recordings.

Line 237-239: I wasn't clear from the text whether this was a variation due to development, to transection, or natural variability.

We now indicate that the relationship reflects normal development (lines 261-264).

Line 521: N-type VGCCs: can these be targeted with pharmacology to strengthen the argument?

We acknowledge this suggestion but we have not carried out these experiments as we believe that the interpretation could be complex due to the high density of synapses present in glomeruli and the likely involvement of other types of VGCCs in neurotransmitter release.

Small issues:

(1) Line 190-196: Some of this could potentially be moved to the Discussion section.

These are some arguments to defend the validity of our experimental approach to record the response of the lateral glomerulus labeled by GFP. If we move them to the discussion, the information related to the spatial extent of our recordings would be split between results and discussion. We believe that the current format of the paper allows to focus the discussion on the interpretation of the results obtained.

(2) Line 268: exponential recover phase.

Thanks. Corrected.

(3) Line 278: affected to -> arises from

Thanks. Corrected.

(4) Line 282: affect to -> can affect.

Thanks. Corrected.

(5) Line 403: 2Phatal technique: Please state briefly what this is

It is now indicated: two-photon chemical apoptotic targeted ablation (2Phatal).

NOTE:

During the revision of this manuscript we realized that Figures 3C and 4B indicated mean±SD. The panels have been amended to show mean±s.e.m.

References

Boyd, J. D., and K. R. Delaney. 2002. "Tyrosine hydroxylase-immunoreactive interneurons in the olfactory bulb of the frogs Rana pipiens and Xenopus laevis." J Comp Neurol 454 (1):42-57. doi: 10.1002/cne.10428.

Dorrego-Rivas, A., D. J. Byrne, Y. Liu, M. Cheah, C. Arslan, M. Lipovsek, M. C. Ford, and M. S. Grubb. 2025. "Strikingly different neurotransmitter release strategies in dopaminergic subclasses." Elife 14. doi: 10.7554/eLife.105271.

Kosaka, T., A. Pignatelli, and K. Kosaka. 2020. "Heterogeneity of tyrosine hydroxylase expressing neurons in the main olfactory bulb of the mouse." Neurosci Res 157:15-33. doi: 10.1016/j.neures.2019.10.004.

Lecoq, J., P. Tiret, and S. Charpak. 2009. "Peripheral adaptation codes for high odor concentration in glomeruli." J Neurosci 29 (10):3067-72. doi: 10.1523/JNEUROSCI.6187-08.2009.

eLife Assessment

This important study addresses the unresolved and long-debated question of whether atypical protein kinase C is required for the maintenance of synaptic potentiation and long-term memory. The convincing results confirm previous findings that persistent activity of PKMζ is required for lasting potentiation of hippocampal synapses and spatial memory. The study also adds new genetic evidence to support the earlier suggestion that enhanced expression of PKC iota/lambda compensates for the genetic reduction of PKM zeta to support synaptic potentiation and memory.

Reviewer #1 (Public review):

Summary:

The authors convincingly demonstrate that when PKMzeta is genetically deleted from the hippocampus, the related atypical PKC, PKClambda is upregulated and compensates both neurophysiologically and behaviorally for the missing PKMzeta. Specifically, the upregulatiion of PKClambda supports late-phase hippocampal long-term potentiation (L-LTP) and long-term spatial memory in the PKMzeta knockout mice.

Strengths:

The study uses up-to-date transgenic techniques to alter the expression of the two atypical PKCs. The synaptic and behavioral experiments are well-controlled and appear to have been carefully executed.

Weaknesses:

None

Reviewer #2 (Public review):

Summary:

The authors significantly advance understanding of the role of unconventional PKC's, PKCM𝛇 and PKC𝜄/𝝀 in maintenance of late-phase LTP. Their results help to clarify the interplay between "structural" and "biochemical/enzymatic" mechanisms of LTP and learning in the hippocampus.

Strengths:

A strength is the use of state-of-the-art conditional knock-outs of PKCM𝛇 and PKC𝜄/𝝀 to confirm that PKC𝜄/𝝀 compensates for KO of PKCM𝛇 in the hippocampus to maintain long-term potentiation even when PKCM𝛇 is conditionally knocked out in the adult. The authors use both electrophysiological and behavioral methods to assess the effects of genetic manipulations on late-phase LTP and long-term memory. The authors present an informative discussion of the possible molecular mechanisms that may enable compensation by PKC𝜄/𝝀 for KO of PKCM𝛇 in the hippocampus. They correctly emphasize that the notions of "structural" and "enzymatic" mechanisms for maintenance of LTP are not mutually exclusive. With this publication, the experimental case for a role of PKCM𝛇 in maintenance of late-phase LTP is now quite strong.

Weaknesses:

There are no significant weaknesses.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

An ongoing controversy in the field of learning and memory is the specific neural mechanism that maintains long-term memory (LTM). A prominent hypothesis proposed by Sacktor and Fenton and their colleagues is that LTM is maintained by the ongoing activity of the atypical PKC isoform PKMζ. Early evidence in support of this hypothesis came from experiments showing that an inhibitory peptide, ZIP, whose activity was purported to be specific for PKMζ, blocked late-phase hippocampal LTP (L-LTP) and LTM. However, in 2013, two articles reported that LTM was normal in PKMζ knockout mice and that ZIP erased LTM in the knockout mice, indicating that ZIP lacked specificity for PKMζ. In response, Sacktor and Fenton and colleagues reported in 2016 that in PKMζ null mice, there is an increase in the expression of PKC𝜾/𝛾, a related isoform of atypical PKC, and this increased expression can compensate for PKMζ; their data indicated that the upregulation of PKC 𝜾/𝛾 mediates L-LTP and LTM in the PKMζ. In the present article, the authors provide additional support for this idea. They replicate the finding of an upregulation of PKC 𝜾/𝛾 expression in the hippocampus of PKMζ knockout mice; in addition, they show that the expression of several other PKC isoforms is upregulated in the knockouts. They find that down-regulation of PKC𝜾/𝛾 expression in the hippocampus using the Cre-LoxP technology, the 2016 paper merely used an inhibitor to block the activity of PKC𝜾/𝛾-blocks L-LTP. Finally, the authors demonstrate that, although LTM is preserved in the single PKMζ knockout mouse, it is eliminated in the PKMζ/PKC𝜾/𝛾 double knockout mouse.

Strengths:

The experiments appear to have been carefully executed, the results reliable, and the paper well-written. Overall, the article provides significant additional support for the idea that the activity of PKMζ is critical for the maintenance of hippocampal L-LTP and LTM. The article uses genetic methods, rather than simply pharmacological ones, to demonstrate that when PKMζ is genetically deleted, PKC𝜾/𝛾, compensates for the missing PKCζ.

Weaknesses:

The paper sets up what I believe is probably a false dichotomy between a structural explanation - a change in the number of synaptic connections among neurons - and the persistent kinase activity explanation for memory maintenance. Why are these two explanations necessarily antithetical? It is possible that an increase in synaptic connections and the ongoing activity of PKMζ both contribute substantially to memory maintenance. The authors certainly don't provide any evidence that the number of synapses in the hippocampus remains unchanged after the induction of L-LTP or LTM. Indeed, I see no reason why persistent PKMζ activity could not be a mechanism for the maintenance of an enhanced number of synaptic connections following the induction of LTP/LTM. To the best of my knowledge, this possibility has not yet been explored. Consequently, I don't see why the present results would lead one to favor a biochemical explanation over a structural one for memory maintenance. Given the significant experimental evidence that LTM involves persistent structural changes in neurons, both explanations are equally plausible at present.

As requested, we eliminated the discussion of a dichotomy between structural and biochemical mechanisms of long-term memory in the Abstract and Introduction. We now briefly address the relationship between the two hypotheses, which are not mutually exclusive, in the Discussion.

Reviewer #2 (Public review):

Summary:

The authors are attempting to advance understanding of the role of unconventional PKCs, PKCM𝛇, and PKC𝜄/𝝀 in maintenance of late-phase LTP. Their results help to clarify the interplay between "structural" and "biochemical/enzymatic" mechanisms of LTP and learning in the hippocampus.

Strengths:

A strength is the use of conditional knock-outs of PKCM𝛇 and PKC𝜄/𝝀 to assess the role of these two enzymes in maintaining long-term potentiation and in compensating for each other when one of them is conditionally knocked out in the adult.

Weaknesses:

The paper is extremely difficult to read because the abstract does not clearly state the advances made over earlier studies by the use of conditional KO mutation. For example, in line nine of the abstract, the authors state, "Here, we found PKC𝜄/𝝀 persists in LTP and long-term memory when PKM𝛇 is genetically deleted." This is confusing because it sounds as though the experiments have repeated earlier published experiments in which the gene encoding PKM𝛇 is deleted in the embryo. The authors are not clear throughout the manuscript that they are using conditional KO of the two enzymes in the adult animal, rather than deletion of the gene. The term "genetically deleted" does not mean "conditionally deleted in the adult." The final sentences of the abstract are: "Whereas deleting PKM𝛇 and PKC𝜄/𝝀 individually induces compensation, deleting both aPKCs abolishes hippocampal late-LTP. Hippocampal 𝜄/𝝀-𝛇 -double-knockout eliminates spatial long-term memory but not short-term memory. Thus, in the absence of PKM𝛇 , a second persistent biochemical process compensates to maintain late-LTP and long-term memory." These sentences do not convey a clear logical conclusion. The Discussion does a better job of stating the importance of the experiments.

We have clarified the genotypes of the mice in the abstract and throughout the text.

Reviewer #3 (Public review):

Summary:

The manuscript addresses an important, yet unresolved and long-debated, question: whether atypical protein kinase C is required for the maintenance of late-long-term synaptic potentiation (L-LTP) and long-term memory (LTM). The authors confirm previous findings that persistent activity of PKMζ is required for hippocampal L-LTP and spatial memory. They demonstrate that genetically deleting PKCι/λ and PKMζ individually induces compensatory upregulation, whereas deleting both atypical PKCs abolishes hippocampal L-LTP spatial long-term memory. The study uses an elegant combination of immunoblots, electrophysiology, and behavioral assays. The use of Cre-recombinase to target specific hippocampal regions and neurons adds to the rigor of the findings.

Strengths:

The manuscript addresses an important, yet unresolved and long-debated, question; whether PKMζ is required for the maintenance of L-LTP and LTM. The study demonstrates that PKCι/λ, which was previously shown to be critical for the initial generation of the early phase of LTP and short-term memory, becomes persistently active in L-LTP and LTM in a PKMζ knock-out model, compensating for the loss of PKMζ. Furthermore, when the compensation mechanisms are eliminated by simultaneous deletion of both PKMζ and PKCι/λ, maintenance of LTP and long-term spatial memory, but not of short-term memory, is diminished. The strength of this study is that the authors used a double-knockout strategy to directly address the controversy concerning the roles of PKMζ in memory formation. By showing that PKCι/λ compensates when PKMζ is deleted, the authors provided a compelling explanation for previous contradictory findings.

Weaknesses:

(1) The authors should provide the numerical values for all data.

(2) It appears that blind procedures were only used for the behavioral experiments. Some explanation is warranted.

(3) The description of the immunoblotting procedures lacks sufficient detail. The authors state that immunoblots were stained with multiple antisera to visualize multiple PKCs on the same immunoblot. To conserve antisera, the immunoblots were cut to isolate the relevant proteins based on molecular weight. Isoforms with similar molecular weights were either stained with antisera of different species or on separate blots. Despite this explanation, it is unclear how immunoblotting was performed in practice. For example, in Figure 1B, the authors compared the changes of four conventional PKC isoforms. Because all four antibodies are mouse monoclonal antibodies recognizing proteins of similar molecular weights, each probing should presumably have its own actin loading controls. However, these controls are missing from the figure. Some clarification is warranted.

(4) The statement in the legend to Figure 4B, that the increases of maximum avoidance time from pretraining to trial 1 are not different, indicates both groups of mice successfully established short-term memory, which is not correct. The analysis only reveals that there is no difference between the two groups. No differences could be due to both groups learning the same, as the authors suggest, or alternatively to no learning in either group.

(5) The labeling on some of the illustrations (e.g., Figure 2B) is unreadable.

(6) In Figure 4B, only the single statistical comparison between "pretaining" and "1 trial" is shown. The other comparisons described in the legend should also be illustrated.

(7) There is no documentation to support the statement that "The prevailing textbook mechanism for how memory is retained asserts that stable structural changes at synapses, the result of initial protein synthesis and growth, sustain memory without the need for ongoing biochemical activity dedicated to storing information" or for the statement in the Discussion that the structural model of memory storage is the standard account.

(1) Numerical data used in statistical analyses are now provided for LTP experiments in Figure 4 figure supplement 1. Numerical values for all other experiments are presented in the figures.

(2) Blind procedures were performed for all experiments except for LTP experiments that involved the transfection of eGFP as control, as the eGFP could be detected visually in the hippocampal slice by the experimenter. This is now clarified in the Statistics section of the Methods.

(3) The description of immunoblotting was clarified in the Methods, and actin loading controls presented for all immunoblots in Figure 1 and Figure 1 figure supplements 1 and 2.

(4) Short-term memory (Figure 5B) is now determined by 2 methods. First, we show that for both groups the times to enter the shock zone increase in the first training trial, as compared to the pretraining session with the shock off. The increases are not different between the groups. Second, we show increases of the maximal avoidance time from pretraining to trial 1 for both groups are significant, and that the increases are not different. These data show that short-term memory was present in both groups and not measurably different between the groups.

(5) The fonts of the figure labels were enlarged.

(6) The comparisons between pretraining and training trial 1 and between training trials 1 and 3 for the two groups are now shown in Figure 5B.

(7) We abbreviated our discussion of the structural model, which is now presented at the end of the Discussion (as per Reviewer 1), and removed the comment that it is the prevailing view, stating instead that the hypothesis is “widely held.”

Additional points: As requested, the timing of tamoxifen injections and tissue collection for immunohistochemistry is clarified in the protocol schematic of a new Figure 2A and Figure 2A legend.

eLife Assessment

This valuable study reports that the ALDH-abundant cells display stem cell properties and may play a key role in the endometrial epithelial development in the mouse. The data supporting the main conclusion are solid, although further improvements are needed to strengthen the conclusions. This work will be of great interest to reproductive biologists and biomedical researchers working on women's reproductive health.

Reviewer #1 (Public review):

The manuscript by Tang et al. characterizes the expression dynamics and functional roles of aldehyde dehydrogenase 1 activity in uterine physiology. Using a combination of in vivo lineage tracing and cell ablation coupled with organoid culture, the authors propose that Aldh1a1 lineage-marked cells contribute to uterine gland development and epithelial regeneration. The descriptive data will be of interest to reproductive biologists and clinicians and will build on established hypotheses in the field. The manuscript is well written and scientifically sound; however, several experimental limitations and interpretation caveats should be addressed.

The methods surrounding the passage number and duration of culture following sorting prior to transcriptomic profiling should be clarified in the figure legends. Related to this, the representative images in Figures 1D and 1E do not appear consistent with the quantification presented in Figures 1F-H and should be reconciled.

The conclusion that ALDH1A1+ cells are enriched in populations with stem cell characteristics relies primarily on transcriptomic analysis. Protein-level co-localization should be performed to strengthen this claim.

The overlap of 19 genes between the data set here and AXIN2 HI data is presented as evidence of shared stemness identity, but no statistical assessment of this overlap is provided. A hypergeometric test should be performed to determine whether this overlap is greater than expected by chance.

The impact of tamoxifen injection on Aldh1a1 expression should be characterized in the neonatal uterus, as tamoxifen itself has known estrogenic activity that could confound interpretation of the lineage tracing results at early postnatal timepoints. Related to this, while low-dose tamoxifen is shown to label individual cells within 24 hours of injection, the translation dynamics of the label following Cre-mediated recombination can require up to 72 hours. The presence of only a few labeled clones at PND8 but multiple separate clones per cross-section at later timepoints warrants discussion and may reflect labeling kinetics rather than clonal expansion.

It would strengthen the in vivo ablation data to validate the degree of cell death following diphtheria toxin treatment directly. It is possible that a general decrease in cell number rather than specific loss of a stem cell population is responsible for the observed reduction in gland number and FOXA2 expression (Tongtong et al 2017).

The lineage tracing data in the postpartum endometrium demonstrate that Aldh1a1-marked cells are present during regeneration, but it remains unclear whether these cells are preferentially activated or expanded in response to tissue injury. Coupling these studies with diphtheria toxin-mediated ablation during active regeneration would more directly test the proposed regenerative role of this population.

The contribution of stromal Aldh1a1 lineage-positive cells is underexplored in the discussion, given the lineage tracing data showing stromal labeling across multiple timepoints and its potential relevance to mesenchymal-to-epithelial transition.

Finally, the word 'control' may overstate the functional evidence presented. 'Contribute' may be more accurate given the partial and context-dependent nature of the phenotypes observed.

Reviewer #2 (Public review):

Tang et al. investigated the contribution of Aldh1a1+ cells, as putative stem/progenitor cells, to endometrial development, maintenance during the estrous cycle, and postpartum repair in mouse models. They employed in vitro organoid formation and in vivo lineage tracing models coupled with RNA-seq to test the stem-ness of Aldh1a1+ cells. They found that mouse endometrial cells with high ALDH activity (using the ALDEFLUOR assay) formed more and larger organoids and were enriched for stem/progenitor cell gene signatures. Similar results were shown using endometrial cells from a human patient sample. Epithelial ALDH1A1 expression was shown to be hormonally regulated, becoming more restricted to the glands, a putative epithelial stem cell niche, under estrogen stimulation. Using lineage-tracing initiated postnatally/prepubertally, Aldh1a1+ epithelial cells were shown to expand, contributing to both the luminal and glandular epithelium into adulthood, whereas adult initiation of labeling showed expansion of stromal Aldh1a1+ cells but not epithelial. Postnatal ablation of single-labeled Aldh1a1+ epithelial cells resulted in impaired gland development. Lastly, Aldh1a1-lineage traced cells (adult labeled) were present during postpartum endometrial repair as were epithelial/mesenchymal transitional cells.

This study addresses an important area of research in the field of endometrial stem/progenitor cell biology. The authors are commended for their use of multiple complementary methods, including lineage tracing, DTR-mediated cell ablation, organoid assays, and RNA-seq in mouse and human models to assess the stem-like nature of Aldh1a1+ cells. The data support the stem/progenitor phenotype of Aldh1a1+ epithelial cells during endometrial development; however, there are noted discrepancies between organoid formation assays and lineage tracing experiments regarding the stemness of Aldh1a1+ epithelial cells in adults. Specifically, organoids were generated from adult cells and demonstrated in vitro stem cell activity; however, in vivo lineage-tracing of adult cells either during the estrous cycle or postpartum repair does not show expansion of Aldh1a1+ cells, suggesting they do not have stem/progenitor activity. Additionally, the stem-ness of epithelial vs stromal Aldh1a1+ cells is confounded in the study because epithelial cells were not purified for organoid experiments, epithelial cells were not exclusively lineage-traced as stromal cells were also labeled, and mesenchymal-epithelial transition was suggested to occur during postpartum repair. The following specific comments are presented to detail these concerns:

(1) The statement in the brief summary, "...critical for lifelong endometrial regeneration," is not supported by the data provided.

(2) AlDH1A1 is not restricted to the endometrial epithelium, and epithelial cells were not purified by flow cytometry for experiments in Figure 1. Figure 2 clearly shows the presence of mesenchymal cells, even using the described method for enriching for epithelial cells. Therefore, contaminating mesenchymal cells with high ALDH activity may confound the experimental results in Figure 1, either through promoting epithelial cell growth or through MET. The authors should provide clear evidence of epithelial purity in organoid experiments or that mesenchymal cells are not contained in the ALDHhi population. These comments also apply to the human organoid experiments in Figure 7.

(3) Lines 186-187: Susd2 was increased in EpSC clusters, yet this is a mesenchymal stem/progenitor marker in humans. The authors should discuss the implications of this.

(4) In Figure 5, RFP+ epithelial cells should be quantified as in previous figures to substantiate the statement in lines 279-280, "At PPD5, the proportion of RFP+ epithelial cells had expanded relative to PPD1 and PPD3 (Figure 5E-E')." Especially because in the low mag images (C-E), RFP+ epithelial cells appear to be most abundant at PPD1 and decrease at PPD3 and PPD5, suggesting that they may not be involved in endometrial regeneration/repair (contradicting the interpretation in line 285). Further, if there is in fact a decrease over postpartum repair, then regeneration should be removed from the title of the manuscript. RFP+ stromal cells should also be quantified.

(5) For Figure 7F, it should be clearly stated in the main text that the results are from one patient sample and the data presented are experimental replicates, so as not to be confused with biological replicates (the same for Supplementary Figure S4). Were B and G in Figure 7 also from one patient?

(6) Lines 425-427: "Ovariectomized mice treated with 90-day E2 pellets, on the other hand, showed a complete restriction of ALDH1A1 to the glandular crypts." In Figure 2 S' ALDH1A1+ cells are visible in the LE (the staining is lighter than in the GE but looks real), contradicting this statement.

(7) Lines 466-467: "In cycling mice, we found sporadic cells that expressed both stromal and epithelial markers in the ALDHA1+ cells." These data are not presented.

(8) These data support the role of Aldh1a1+ cells in endometrial epithelial development, but conclusions about their role in repair/regeneration should be tempered as the data are much weaker here.

Reviewer #3 (Public review):

Summary:

Tan et al demonstrated the importance of ALDH-high cells in the epithelial development in the mouse endometrium, and these cells displayed properties of stem cells.

Strengths:

The findings are solid, supported and validated through a combination of technical methods. I appreciated this combined use of mouse and human endometrial cells to strengthen the findings. Genomic results from a single-cell sequencing dataset were informative as they depicted the different stages of the estrus cycle during the regeneration process. Verification with immunostainings with various markers made it convincing for readers to visualize the cell's location, progression, and status at different timepoints. Utilizing human endometrial cells further demonstrated that the phenomenon observed in mice can be translated to humans.

This work will greatly advance the understanding of endometrial regeneration for reproductive biologists.

Weaknesses:

No major weaknesses were identified by this reviewer.

eLife Assessment

This important study examines the evolution of virulence and antibiotic resistance in Staphylococcus aureus under multiple selection pressures, specifically host immune function and antibiotic exposure. The evidence presented is convincing, supported by rigorous phenotypic and genomic data from within-host evolution experiments. The manuscript now provides a nuanced and robust interpretation of how pathogens adapt to complex selective landscapes.

Reviewer #1 (Public review):

Summary:

The authors investigate how methicillin-resistant (MRSA) and sensitive (MSSA) Staphylococcus aureus adapt to a new host (C. elegans) in the presence or absence of a low dose of the antibiotic oxacillin. Using an "Evolve and Resequence" design with 48 independently evolving populations, they track changes in virulence, antibiotic resistance, and other fitness-related traits over 12 passages. Their key finding is that selection from both the host and the antibiotic together, rather than either pressure alone, synergistically results in the evolution of the most virulent pathogens. Genomically, they find that this adaptation repeatedly involves parallel mutations in a small number of key regulatory genes, most notably codY, agr, and saeRS.

Strengths:

The main advantage of the research lies in its strong and thoroughly replicated experimental framework, enabling significant conclusions to be drawn based on the concept of parallel evolution. The study successfully integrates various phenotypic assays (virulence, growth, hemolysis, biofilm formation) with whole-genome sequencing, offering an extensive perspective on the adaptive landscape. The identification of certain regulatory genes as common targets of selection across distinct lineages is an important result that indicates a level of predictability in how pathogens adapt. Furthermore, the detailed mapping of specific parallel mutations provides a highly useful genomic resource for the microbiology community.

Revisions and Re-Appraisal:

In the initial version of the manuscript, a primary limitation was the use of causal language to link specific mutations to phenotypes, despite the evidence from the evolution experiment being correlational. In this revised version, the authors have excellently addressed this limitation. They have meticulously revised the text to accurately reflect these relationships as strong, statistically significant genetic associations rather than confirmed facts. Furthermore, they explicitly acknowledge that future ancestral reconstruction experiments will be required to confirm direct causality. The authors have also appropriately clarified the visual interpretations of their data (such as the PCA clustering) and refined their discussion of mutation rates. With these revisions, the claims made are fully supported by the data presented.

Impact and Context:

The authors successfully achieve their aims, demonstrating that the combined effects of host and antibiotic pressures collaboratively propel the evolution of heightened virulence. While the nematode model does not perfectly mimic human or mammalian infection, the evolutionary principles uncovered here are highly relevant to both evolutionary biology and infectious disease management. The evidence presented is compelling, and the strong correlational hypotheses generated by this study offer a robust and significant basis for upcoming mechanistic research into pathogen adaptation.

Comments on revisions:

I commend the authors for their thorough, thoughtful, and highly constructive revision. You have successfully addressed all of my major and minor comments. The addition of Table S2 and the careful revisions to the causal language have significantly strengthened the manuscript and clarified the data interpretation. I have no further recommendations. Great work!

Reviewer #2 (Public review):

Summary:

The manuscript describes the results of an evolution experiment where Staphylococcus aureus was experimentally evolved via sequential exposure to an antibiotic followed by passaging through C. elegans hosts. Because infecting C. elegans via ingestion results in lysis of gut cells and an immune response upon infection, the S. aureus were exposed separately across generations to antibiotic stress and host immune stress. Interestingly, the dual selection pressure of antibiotic exposure and adaptation to a nematode host resulted in increased virulence of S. aureus towards C. elegans.

Strengths:

The data presented provide strong evidence that in S. aureus traits involved in adaptation to a novel host and those involved in antibiotic resistance evolution are not traded-off. On the contrary, they seem to be correlated, with strains adapted to antibiotics having higher virulence towards the novel host. As increased virulence is also associated with higher rates of haemolysis, these virulence increases are likely to reflect virulence levels in vertebrate hosts.

Weaknesses:

Right now, the results are presented in the context of human infections being treated with antibiotics, which, in my opinion, is inappropriate. This is because

(1) exposure to the host and antibiotics was sequential, not simultaneous, and thus does not reflect the treatment of infection, and

(2) because the site of infection is different in C. elegans and human hosts.

Nevertheless, the results are of interest; I just think the interpretation and framing should be adjusted.

Comments on revisions:

Following the revision, I now think the weakness I initially described has been addressed well by the authors.

Reviewer #3 (Public review):

Summary:

Su et al. sought to understand how the opportunistic pathogen Staphylococcus aureus responds to multiple selection pressures during infection. Specifically, the authors were interested in how the host environment and antibiotic exposure impact the evolution of both virulence and antibiotic resistance in S. aureus. To accomplish this, the authors performed an evolution experiment where S. aureus were fed to Caenorhabditis elegans as a model system to study the host environment and then either subjected to the antibiotic oxacillin or not. Additionally, the authors investigated the difference in evolution between an antibiotic-resistant stain MRSA and an isogenic susceptible strain MSSA. They found that MRSA strains evolved in both antibiotic and host conditions became more virulent and that strains evolved outside these conditions lost virulence. Looking at the strains evolved in just antibiotic conditions, the authors found that S. aureus maintained its ability to lyse blood cells. Mutations in codY, gdpP and pbpA were found to be associated with increased virulence. Additionally, these mutations identified in these experiments were found in S. aureus strains isolated from human infections.

Strengths:

The data are well-presented, thorough, and are an important addition to the understanding of how certain pathogens might adapt to different selective pressures in complex environments.

Comments on revisions:

For the most part, my comments have been addressed. It seems that the authors have not addressed my comments about quantifying population sizes in order to understand mutation supply, particularly in light of which experimental phase exhibits the strongest selection and possible increases in mutation rates. While I think this information would be very useful if they had collected it during the experiment, I don't think it is important enough to require additional experiments. I am therefore satisfied with the current state of the manuscript.

Author response:

The following is the authors’ response to the original reviews.

eLife Assessment

This important study examines the evolution of virulence and antibiotic resistance in Staphylococcus aureus under multiple selection pressures. The evidence presented is convincing, with rigorous data that characterizes the outcomes of the evolution experiments. However, the manuscript's primary weakness is in its presentation, as claims about the causal relationship between genotypes and phenotypes are based on correlational evidence. The manuscript needs to be revised to address these limitations, clarify the implications of the experimental design, and adjust the overall narrative to better reflect the nature of the findings.

Thank you for your feedback. Here, we summarize the major changes made in the revised manuscript:

(1) We did not test causality between mutations and phenotypes in our study. We were intentional about not using causal wording (“mutation X caused/led to/resulted in phenotype Y”), and only discussed these results using the terms “correlation” and “association”, and only when they were statistically significant. We understand that some readers may view these terms as being equivalent to “causation”, thus in the revision, we have modified our wording as suggested (please see below for specific lines).

(2) We agree that experimental evolution in nematodes is not a direct simulation of evolution in humans. The goal of our study was first and foremost, a test of how multiple selective pressures can shape pathogen evolution. This point was presented in the first paragraph, the second to last paragraph of the Introduction (which included our hypotheses), and the last paragraph of the manuscript. References to humans and other mammalian systems were intended to point out similarities between our findings and what had already been found in S. aureus outside the lab. Despite differences between mammals and nematodes, several parallels arose at both the phenotypic and genomic levels, which is interesting from an evolutionary standpoint. We understand that more experiments and tests would be needed before we can make claims about the selective pressures acting on S. aureus outside the lab. We presented some information in the context of humans because a large part of the literature on S. aureus is on its role as a major bacterial pathogen; we did not want to neglect this aspect of its natural life history.

In the revised manuscript, we are more explicit in stating these points, as well as tempering some language regarding human infection, and removing some references to humans. Please see below for specific lines as well as justification for specific references to humans/mammalian systems.

(3) We have including additional details on the experimental design below. We hope this is sufficiently clarifying.

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors investigate how methicillin-resistant (MRSA) and sensitive (MSSA) Staphylococcus aureus adapt to a new host (C. elegans) in the presence or absence of a low dose of the antibiotic oxacillin. Using an "Evolve and Resequence" design with 48 independently evolving populations, they track changes in virulence, antibiotic resistance, and other fitness-related traits over 12 passages. Their key finding is that selection from both the host and the antibiotic together, rather than either pressure alone, results in the evolution of the most virulent pathogens. Genomically, they find that this adaptation repeatedly involves mutations in a small number of key regulatory genes, most notably codY, agr, and saeRS.

Strengths:

The main advantage of the research lies in its strong and thoroughly replicated experimental framework, enabling significant conclusions to be drawn based on the concept of parallel evolution. The study successfully integrates various phenotypic assays (virulence, growth, hemolysis, biofilm formation) with whole-genome sequencing, offering an extensive perspective on the adaptive landscape. The identification of certain regulatory genes as common targets of selection across distinct lineages is an important result that indicates a level of predictability in how pathogens adapt.

Thank you very much.

Weaknesses:

(1) The main limitation of the paper is that its findings on the function of specific genes are based on correlation, not cause-and-effect evidence. While the parallel evolution evidence is strong, the authors have not yet performed the definitive tests (i.e., reconstruction of ancestral genes) to ensure that the mutations identified in isolation are enough to account for the virulence or resistance changes observed. This makes the conclusions more like firm hypotheses, not confirmed facts.

We have replaced instances of “association” and “correlation” with wording similar to that suggested where applicable, including:

L 342 – 344: “The loss of SCCmec and ACME was more often identified in populations exhibiting an increase in total growth from the ancestor outside the host…”

L 371 – 375: “Mutations in three genes were regularly identified in populations exhibiting significant increases in virulence from the ancestor: codY, gdpP, and pbpA. Mutations in agr in general were not associated with changes in overall virulence, but MSSA populations harboring mutations in this gene were more likely to exhibit greater virulence compared to MRSA populations (Wilcoxon rank sum exact test P = 0.045).”

L 377: “Mutations in specific genes were often found in populations able to hemolyze red blood cells…”

L 379 – 381: “There were also significant differences between the mutations regularly identified in oxacillin-resistant populations evolved from the MSSA ancestor...”

L 384 – 385: “By contrast, mutations in agr were often in populations exhibiting loss of hemolytic activity, consistent with previous findings...”

L 409 – 410: “Mutations that arose during experimental evolution are regularly found in strains associated with human systemic infections.”

We have also stated that ancestral reconstruction is needed:

L 553 – 555: “Future experiments may include introducing these mutations into the ancestral background to directly link the mutations in these genes to evolved virulence.”

(2) In some instances, the claims in the text are not fully supported by the visual data from the figures or are reported with vagueness. For example, the display of phenotypic clusters in the PCA (Figure 6A) and the sweeping generalization about the effect of antibiotics on the mutation rates (Figure S5) can be more precise and nuanced. Such small deviations dilute the overall argument somewhat and must be corrected.

In reference to Fig. 6A, we have revised the statement as suggested: “…where populations exposed to host and sub-MIC oxacillin clustered together, largely separating from all other treatments…” Line 442

In reference to Fig. S5, we conducted statistics to include both MRSA and MSSA populations and examined the effect of oxacillin on the number of mutations. While oxacillin had a significant effect on the number of mutations, we agree with the reviewer that this may be driven by the MRSA populations and have clarified: “Sub-MIC oxacillin selection also resulted in more mutations than in its absence ( = 5.92, P = 0.015), although this is likely driven by MRSA populations.” Lines 310 – 311

Reviewer #2 (Public review):

Summary:

The manuscript describes the results of an evolution experiment where Staphylococcus aureus was experimentally evolved via sequential exposure to an antibiotic followed by passaging through C. elegans hosts. Because infecting C. elegans via ingestion results in lysis of gut cells and an immune response upon infection, the S. aureus were exposed separately across generations to antibiotic stress and host immune stress. Interestingly, the dual selection pressure of antibiotic exposure and adaptation to a nematode host resulted in increased virulence of S. aureus towards C. elegans.

Strengths:

The data presented provide strong evidence that in S. aureus, traits involved in adaptation to a novel host and those involved in antibiotic resistance evolution are not traded off. On the contrary, they seem to be correlated, with strains adapted to antibiotics having higher virulence towards the novel host. As increased virulence is also associated with higher rates of haemolysis, these virulence increases are likely to reflect virulence levels in vertebrate hosts.

Weaknesses:

Right now, the results are presented in the context of human infections being treated with antibiotics, which, in my opinion, is inappropriate. This is because

(1) exposure to the host and antibiotics was sequential, not simultaneous, and thus does not reflect the treatment of infection, and

(2) because the site of infection is different in C. elegans and human hosts.

We have removed the two sentences referencing site of infection:

Introduction: “In the host, antibiotic concentrations will gradually decline after administration due to metabolism and excretion.”

Discussion: “…in addition to infection of antibiotic-treated hosts, where there is uneven distribution of drugs across tissues.”

For our rationale for discussing humans in general, please see below.

Nevertheless, the results are of interest; I just think the interpretation and framing should be adjusted.

Thank you very much.

Reviewer #3 (Public review):

Summary:

Su et al. sought to understand how the opportunistic pathogen Staphylococcus aureus responds to multiple selection pressures during infection. Specifically, the authors were interested in how the host environment and antibiotic exposure impact the evolution of both virulence and antibiotic resistance in S. aureus. To accomplish this, the authors performed an evolution experiment where S. aureus was fed to Caenorhabditis elegans as a model system to study the host environment and then either subjected to the antibiotic oxacillin or not. Additionally, the authors investigated the difference in evolution between an antibiotic-resistant strain, MRSA, and an isogenic susceptible strain, MSSA. They found that MRSA strains evolved in both antibiotic and host conditions became more virulent, and that strains evolved outside these conditions lost virulence. Looking at the strains evolved in just antibiotic conditions, the authors found that S. aureus maintained its ability to lyse blood cells. Mutations in codY, gdpP, and pbpA were found to be associated with increased virulence. Additionally, these mutations identified in these experiments were found in S. aureus strains isolated from human infections.

Strengths:

The data are well-presented, thorough, and are an important addition to the understanding of how certain pathogens might adapt to different selective pressures in complex environments.

Thank you very much.

Weaknesses:

There are a few clarifications that could be made to better understand and contextualize the results. Primarily, when comparing the number of mutations and selection across conditions in an evolution experiment, information about population sizes is important to be able to calculate the mutation supply and number of generations throughout the experiment. These calculations can be difficult in vivo, but since several steps in the methodology require plating and regrowth, those population sizes could be determined. There was also no mention of how the authors controlled the inoculation density of bacteria introduced to each host. This would need to be known to calculate the generation time within the host. These caveats should be addressed in the manuscript.

While the population sizes within hosts and generation time could be determined, we would need to conduct additional experiments (e.g., infecting nematodes with S. aureus, then crushing, plating, and counting colony forming units across time intervals) in order to obtain measurements for pathogen growth in hosts across time. For experimental evolution, we crushed a set number of dead nematodes (30) and all bacteria that were released were allowed to grow in liquid media before an aliquot (25%) was used to seed the next passage. Picking and crushing nematodes across 48 populations for one time point was an arduous task. The additional steps of picking, crushing, and plating nematodes across multiple time intervals at the same time experimental evolution was being performed would not be logistically sound.

In terms of the inoculation density of bacteria, all nematodes were placed on abundant lawns of S. aureus. Nematodes were exposed to full lawns the entire infection step; bacteria remained in abundance. While we do not know the exact inoculum each individual nematode was exposed to, we know that they ingested the bacteria because of the high mortality rate. Furthermore, we followed the same procedure for every replicate across every host-associated treatment. Host individuals within and across passages were also genetically identical to one another. Altogether, these factors allowed for more consistency across the experiment, such that relative inoculum size should be similar across individual hosts. Please refer to the evolution experiment diagram (Author response image 1) for more details.

Ultimately, while knowing the absolute population size, inoculum size, and generation time within the host is interesting, the rounds of selection (the number of times each population was exposed to the selective pressures) is also important in addressing our major question. Every treatment, which started out from one ancestral clone (MRSA or MSSA), was exposed to the same number of bouts of selection (passages), yet we see significant divergence in terms of traits and mutations. Future directions would certainly involve determining the number of steps (e.g., number of generations within hosts) required to reach these end points, but not knowing exactly how many steps were required do not detract from addressing the larger question of determining how pathogens respond to multiple selective pressures.

Another concern is the number of generations the populations of S. aureus spent either with relaxed selection in rich media or under antibiotic pressure in between the host exposure periods. It is probable then that the majority of mutations were selected for in these intervening periods between host infection. Again, a more detailed understanding of population sizes would contribute to the understanding of which phase of the experiment contributed to the mutation profile observed.

We conducted every step of the evolution experiment on the same timeline. For example, all replicates across treatments were grown in liquid media at the same time (see Author response image 1.). All populations were exposed to the same selective pressures at this step of the experiment. We can then compare populations that were subsequently exposed to hosts against those that were not. Populations passaged without a host served as the control. Mutations that were solely unique to host-exposed populations would more likely contribute to the traits of interest, compared to mutations that were in common between the host-exposed and no-host treatments. Similar comparisons could be made with the oxacillin-exposed and no-oxacillin populations.

In general, the only differences between treatments would be driven by the treatments themselves. Given that we are interested in treatment-level effects, any differences in population size or generation time between treatments could contribute to the treatment effects we observe, and thus were not something we aimed to hold uniform across our experiment.

Author response image 1.

Schematic of procedural steps involved in one passage of S. aureus through nematodes (+host -ox) compared to without nematodes (-host -ox).

Recommendations for the authors:

Reviewing Editor Comments:

We encourage you to address all other comments raised by the reviewers; however, the review team has identified the following points as the most critical and fundamental to improve your manuscript:

(i) Reframing the narrative: You will need to adjust the narrative so that the study is presented as a "proof of principle" rather than a direct simulation of a human infection.

While we referenced human infection, we believe the study had been presented as a proof of principle. Examples include:

(1) We discussed the gap of knowledge in the first paragraph: “It is unclear how virulence evolves in the face of more than one selective pressure and whether this trait is constrained or facilitated by antibiotic resistance.” Lines 86 – 88

(2) In the second to last paragraph in the Introduction, we presented the main hypotheses: “Adaptation may require resources to be expended toward either virulence or antibiotic resistance, leading to a trade-off between these traits (Ferenci, 2016). Alternatively, weaker selection from sub-MIC antibiotics may interact synergistically with hosts and facilitate the evolution or maintenance of high virulence and antibiotic resistance.” Lines 176 – 179

(3) The last paragraph concluded with “Our findings ultimately emphasize the importance of considering the host context in the evolution of antibiotic resistance. Integrating multiple traits, such as virulence, antibiotic resistance, and fitness may be critical in identifying the factors that facilitate host shifts and persistence of drug-resistant pathogens.” Lines 613 – 616

These paragraphs, which set up the context for our work, did not primarily discuss human infections.

In the revised manuscript, we have further tempered language regarding human infection:

L 169 - 172: “Experimentally evolving S. aureus in C. elegans thus allows us to track the early stages of virulence and antibiotic resistance evolution in novel host populations with the potential to identify conserved genomic regions underlying evolved traits.”

L 595 – 596: “Additional direct tests are needed to evaluate the role of these mutations in adaptation of S. aureus to different infection sites.”

L 610 – 611: “Pathogen evolution in a tractable invertebrate animal model yielded phenotypes and genotypes similar to those identified in mammalian hosts, highlighting the utility of evolution experiments to identify potential ecological and genetic mechanisms that may give rise to pathogen traits conserved across systems.”

And removed some references to humans:

In the Introduction: “In the host, antibiotic concentrations will gradually decline after administration due to metabolism and excretion.”

In the Discussion: “…in addition to infection of antibiotic-treated hosts, where there is uneven distribution of drugs across tissues.”

Otherwise, our rationale for referencing humans/mammalian systems in our Introduction include:

Setting the context of our study system: we discussed humans and clinical significance when we first introduced S. aureus (lines 132 – 151) and experimental evolution (lines 153 – 172). Much of what is known about S. aureus outside the lab is when it is interacting with humans, thus we weaved in relevant information that has been discovered in other organisms.

Hemolysis: This ability is important for S. aureus virulence toward C. elegans (Sifri et al., 2003).

S. aureus genomic database: we intended to leverage this large-scale database of genomes isolated from S. aureus outside the lab to compare patterns emerging from experimental evolution to those in existing isolates. Due to its relevance as a major bacterial pathogen, most of the isolates happen to be from clinical settings.

(ii) Adjusting the causal language: You will need to soften the language so that correlational claims do not appear to be causal.

We have adjusted language as noted above.

(iii) Clarifying methodological aspects: You will need to provide more details on the methodology, such as population sizes, and clarify the implications of these in the conclusions of the work.

We have provided additional explanation of methodology and the role of control (no host) treatments above.

Reviewer #1 (Recommendations for the authors):

The paper is robust, and the study is of great significance. Tackling the subsequent issues would greatly enhance the paper and elucidate its findings.

Major Recommendations:

(1) Revising Causal Language: The main flaw of the manuscript lies in its presentation of correlational data as if it were causal. We highly suggest a thorough review of the text to soften causal language when connecting genotypes to phenotypes. The absence of ancestral reconstruction should be recognized as a constraint. Assertions ought to be presented as robust, evidence-based hypotheses. For instance, rather than saying a mutation "associated with significant increases in virulence," you might say "was regularly identified in groups that developed increased virulence, strongly suggesting this gene's role in the adaptation." This will more precisely clarify the contribution of the work.

We have softened language and stated that ancestral reconstruction is needed as noted above.

(2) Expand on Parallel Mutations: The examination of parallel evolution in Figure 4A is intriguing but would be notably stronger with additional details. I suggest including an additional supplementary figure or table detailing the specific non-synonymous mutations identified in the highly parallel genes (e.g., codY, agr, gdpP). It is essential for the reader to understand whether parallel evolution is happening at the gene level (different mutations in a single gene) or at the nucleotide level (the precise same mutation appearing again). Kindly specify if any of these mutations were nonsense mutations, as this suggests that the loss-of-function is advantageous.

The full table of mutations is in fig share (10.6084/m9.figshare.28745558). We have added a Supplemental Table (Table S2) containing mutations in genes occurring in more than two populations. Many of these mutations were not the same, indicating parallel evolution at the gene level (lines 315 – 317).

Minor Recommendations for Clarity and Accuracy:

(1) Introduction:

Lines 176-177: Please add a citation for the statement describing the function of the SCCmec cassette, as this is established knowledge.

Done.

(2) Results:

Section Title (Line 254): The title "Host and sub-MIC antibiotic promoted growth..." is imprecise. Figure 3B shows that it is the combination of these factors that promotes growth in MRSA, while oxacillin alone is detrimental. Please revise the title to reflect this synergistic effect.

“Synergistically” has been added to the title: “Host and sub-MIC antibiotic synergistically promoted growth of MRSA…” Lines 269 – 270

Lines 261-263: The description of Figure 3B is incomplete. The text should explicitly state that the -host+ox treatment resulted in the lowest growth for MRSA, which provides a critical contrast and suggests a fitness cost.

We have added “By contrast, exposure to sub-MIC oxacillin alone yielded the lowest growth, suggesting a fitness cost.” Lines 277 – 278

Line 294: The claim that "Sub-MIC oxacillin selection also resulted in more mutations" is a generalization not supported for the MSSA genotype, according to Figure S5. Please revise this sentence to specify that this effect was observed in the MRSA populations.

We have clarified: “Sub-MIC oxacillin selection also resulted in more mutations than in its absence ( = 5.92, P = 0.015), although this is likely driven by MRSA populations.” Lines 310 – 311

Lines 419-421: The claim that the +host+ox populations in Figure 6A "formed a distinct cluster" is an overstatement, as there is visible overlap with one other treatment (e.g., host-ox). Please revise this to more accurately describe the visual data (e.g., "clustered together, largely separating...").

We have revised the statement as suggested: “…where populations exposed to host and sub-MIC oxacillin clustered together, largely separating from all other treatments…” Lines 442 – 443

Lines 422-424: The interpretation of the MRSA PCA (Figure 6A) focuses on the correlation between virulence and sub-MIC growth. However, the correlation between "biofilm production" and "growth without oxacillin" appears visually stronger. Please address this correlation as well for a more complete interpretation.

We have added “For MRSA populations, biofilm production and growth without oxacillin also appeared to be positively correlated.” Lines 447 – 448

(3) Discussion:

Lines 469-470: The statement that "exposure to oxacillin resulted in pathogens causing the greatest host mortality" is imprecise. The data in Figure 2A show that it is the combination of host and oxacillin. Please revise this for accuracy and add a direct citation to Figure 2A here.

We have added clarification: “Nonetheless, we observed differing evolutionary trajectories, where exposure to oxacillin in host-associated treatments resulted in pathogens causing the greatest host mortality.” Lines 496 – 498

Reviewer #2 (Recommendations for the authors):

After reviewing the paper and reading the previous reviews from PLoS Biology, my biggest criticism of the paper is the way the story is told. In principle, the results are interesting and relevant, but the analogy to human infection and immune system/ antibiotic treatment strategies does not fit entirely with the experimental design or the results. I think the motivation needs to be reframed. In the study, antibiotic exposure is purely environmental, i.e., not in the host. How does environmental antibiotic use affect in vivo evolution, as this is not tested? As previous reviewers have pointed out, S. aureus is not an enteric pathogen in humans but most often causes skin infections. Furthermore, much of the results and discussion is focused on haemolysis of red blood cells, a cell type that C. elegans does not have. What the paper does present, on the other hand, and something that is interesting and novel, is a test in a model system of how a bacterial pathogen evolves to competing selection pressures. I might have hypothesised a priori that these competing pressures result in trade-offs, something which there is no evidence of, even though growth rate does not appear to be negatively impacted as a consequence of selection for drug resistance and virulence together. Instead, many traits are correlated and seemingly at the mechanistic level. This is cool and is a proof of principle, even if the system does not completely mirror reality, and I think the story should be told as such.

We agree entirely with the reviewer that testing how pathogens respond to multiple selective pressures and the resulting lack of trade-offs are significant and interesting. We presented this question (lines 86 – 88) and our hypothesis about such trade-off in the Introduction (lines 176 – 179). As stated above, we had framed our paper to highlight these points and have removed references to antibiotic concentrations in treated humans.

We measured and discussed hemolysis because it is important for virulence toward C. elegans (lines 195 – 197) (Sifri et al., 2003). We believe our manuscript contained a reasonable discussion of this trait. For example, three panels of the main figures presented the main hemolysis results (Figures 2B, 2C, and 2D), whereas 23 other panels did not at all involve hemolysis. In the Discussion, hemolysis took up half of the shortest paragraph (lines 509 – 519) and an additional sentence (line 589 – 591), out of seven total paragraphs.

Specific comments:

(1) L137-138. Can S. aureus really survive for long periods of time outside of the host? Can you clarify this statement? Do you mean it is an opportunistic pathogen and can also replicate in the environment?

S. aureus can form biofilms and persist for weeks on inert surfaces (Kramer et al., 2024; Tran et al., 2023), indicating that it may replicate in non-host environments. We have included the phrase “opportunistic pathogen” to clarify (line 145).

(2) L187 - to ascertain

Corrected.

(3) Figure 2B - there seems to be a benefit of haemolysis activity to oxacillin resistance, perhaps a crossover in mechanism? In MSSA, without a host, it goes to complete fixation, whereas it is completely lost when antibiotics aren't present. I know this is discussed later, but I would appreciate a more detailed hypothesis of why this could be.

Antibiotics have been found to induce expression of virulence traits, such as in the case of oxacillin and hemolysis. Thus, it is reasonable that exposure to oxacillin during evolution would maintain MSSA’s hemolytic ability. We hypothesize that the loss of hemolysis in the absence of oxacillin may be due to the cost of hemolysis expression without a stimulant (oxacillin), hemolysis may not be expressed as often and be subject to deleterious mutations. Alternatively, the stress that cells were under favored virulence in some way, rather than the direct action of the antibiotic.

(4) L225-228 - As C. elegans do not have red blood cells, why would we expect this? Do you see increased lysis of C. elegans gut cells? Or could it be due to iron accumulation as you are growing the staph on BHI?

We measured and correlated nematode mortality with hemolytic ability because hemolysis had been found to be involved in virulence toward C. elegans (Sifri et al., 2003). The hemolysis phenotype is a surrogate for S. aureus virulence gene expression.

(5) Figure 3A - There seems to be a growth cost of evolving oxacillin resistance in the absence of a host. Why might this be?

MRSA populations exposed to oxacillin without a host during evolution visually exhibited the lowest growth rate. While this is an interesting question, the result was not statistically significant, so we cannot speculate in the manuscript.

Reviewer #3 (Recommendations for the authors):

(1) Some claims in the introduction are either non cited or not correctly stated. The second sentence has a claim about the interplay between antibiotic resistance and virulence with no citation listed. Additionally, there is a claim about S. aureus "evading detection" by attacking the host's immune cells. That is by definition not avoiding detection. Perhaps phrasing it as resisting host immune function would make it clearer.

We have added a citation (lines 80 – 81) and clarified our wording: “Once inside the host, S. aureus resists host immune function by hindering or lysing immune cells.” Lines 140 – 141

(2) Once in the introduction and in the discussion, the authors referred to S. aureus as a novel pathogen for C. elegans, I do not think enough is known to make this statement.

This S. aureus strain is novel because it was isolated from humans, so at least in its recent evolutionary past, it has not interacted with C. elegans. Furthermore, we used a C. elegans isolate (N2) that had been frozen and maintained in the lab on E. coli, and had not been exposed to other microbes in its recent evolutionary past. Finally, S. aureus has not been found to be a native pathogen of C. elegans in nature (Ekroth et al., 2021).

(3) Key suggestion: Change Figure 1C to reflect the design better. So you could have the +OXA before the host and then have an arrow looping back again to show the cycle of each step. So a figure that would have something like: MRSA > +OXA > +host>+OXA --> MRSA .

We have updated the figure as suggested.

(4) Suggest changing "greatest" on line 191, section header to greater.

Done.

(5) Line 258: Rich media can still provide selective pressures that are difficult to quantify - fast growth, cofactor and other nutrient limitations due to that fast growth

We have adjusted our wording: “Importantly, rich media reduced the risk of introducing additional selective pressures than those being tested.” Lines 273 – 274

(6) Why were intergenic mutations routinely ignored? These can often be very important phenotypically.

We had focused on genes because there was a sufficient number of genes to discuss, but we have added a Supplemental Table (Table S2) containing all mutations (including intergenic and synonymous) appearing in more than 2 populations. We have also added information regarding mecA, an accessory gene, highlighting the role non-core genes may have in shaping bacterial evolution:

“Despite evolving in similar environments, MRSA and MSSA populations differing only in the presence of an intact accessory gene (mecA)—proceeded on divergent evolutionary paths…” Lines 66 – 68

“Carriage of Staphylococcal cassette chromosome mec (SCCmec), which encodes mecA, an accessory gene that provides resistance…” Lines 187 – 188

“As MRSA and MSSA only differed in the presence of an intact mecA gene at the start of the experiment, accessory genes may play important roles in shaping bacterial evolution (Jackson et al., 2011).” Lines 472 – 474

(7) Line 294: more mutations than what?

We have clarified the sentence: “Sub-MIC oxacillin selection also resulted in more mutations than in its absence…” Lines 310 – 311

(8) Lines 295-297: wording is pretty confusing. It seems that the discussion is about increased mutation rates, possibly due to hypermutators resulting from mutL or recA mutations, but this isn't well-thought out and much is implied here. Furthermore, see the above comment about comparing mutations across conditions - it's hard to make inferences of mutation rates without knowing the mutation supply as a result of varying population sizes across conditions and through the experiment.

We have clarified the sentence: “…there were only two mutations in DNA and mismatch repair genes (mutL and recA), suggesting repair genes were not the sole mechanism involved.” Lines 313 – 314

Because all populations evolved from one ancestral clone (either MRSA or MSSA), all mutations that are found at the end of the experiment would have arisen de novo from that ancestor. Since all populations experienced the same number of passages/rounds of selection, we determined mutation rate by counting the number of mutations that were found at the last passage for each replicate population. Populations that acquired significantly more mutations had a higher mutation rate in terms of # of mutations/# of selection rounds.

(9) Line 486: typo "Mutations genes".

Corrected.

(10) Line 487: "antibiotics may allow" is awkward; suggest changing to more precise language, possibly relating to pleiotropy if that is what was meant here.

We had intended to mean “adaptation [to antibiotics] may allow”. We have clarified: “Mutations in genes involved in resistance to antibiotics were found more often in populations with increased virulence, suggesting that antibiotic adaptation may also favor evolution of virulence.” Lines 514 – 516

REFERENCES

Ekroth AKE, Gerth M, Stevens EJ, Ford SA, King KC. 2021. Host genotype and genetic diversity shape the evolution of a novel bacterial infection. ISME Journal 15:2146–2157. DOI: https://doi.org/10.1038/s41396-021-00911-3, PMID: 33603148

Kramer A, Lexow F, Bludau A, Köster AM, Misailovski M, Seifert U, Eggers M, Rutala W, Dancer SJ, Scheithauer S. 2024. How long do bacteria, fungi, protozoa, and viruses retain their replication capacity on inanimate surfaces? A systematic review examining environmental resilience versus healthcare-associated infection risk by “fomite-borne risk assessment.” Clinical Microbiology Reviews. PMID: 39388143

Sifri CD, Begun J, Ausubel FM, Calderwood SB. 2003. Caenorhabditis elegans as a model host for Staphylococcus aureus pathogenesis. Infection and Immunity 71:2208–2217. DOI: https://doi.org/10.1128/IAI.71.4.2208-2217.2003, PMID: 12654843

Tran NN, Morrisette T, Jorgensen SCJ, Orench-Benvenutti JM, Kebriaei R. 2023. Current therapies and challenges for the treatment of Staphylococcus aureus biofilm-related infections. Pharmacotherapy 43:816–832. DOI: https://doi.org/10.1002/phar.2806, PMID: 37133439

eLife Assessment

This important study shows that action potentials undergo frequency-dependent failure along the axons of fast-spiking interneurons during sustained high-frequency firing, offering a mechanistic explanation for why inhibition may fail to restrain seizures. The evidence is solid, though additional analyses could further strengthen the mechanistic interpretation. The work will be of broad interest to neuroscientists studying axonal physiology, cortical inhibition, and epilepsy.

Reviewer #1 (Public review):

Summary:

This paper examines whether action potentials (APs) reliably propagate to the distal axon in neocortical parvalbumin-expressing interneurons (PV-Ins) during prolonged high-frequency activity, as occurring during epileptiform activity. The authors use dual soma and axon-attached patch-clamp recordings from mouse and human PV-INs and show that axon AP amplitude declines when the firing frequency exceeds ~200 Hz and fails during seizure-like bursts. Finally, they show that elevation of external K+ to 10 mM also reduces AP amplitude. Taken together, these data strongly suggest that the reduction in transmitter release observed during intense PV-INs activity or during seizure-like events is mainly mediated by the reduction in the presynaptic AP amplitude in PV-INs.

Strengths:

This paper is very interesting, well-written and technically impressive. It provides new and important results. The paper will have a great impact in the field of both axon physiology and epilepsy.

Weaknesses:

I did not find any significant weakness in the methods, data analysis and results.

Reviewer #2 (Public review):

Summary:

The authors demonstrate a frequency-dependent progressive failure of action potential propagation through the axonal arbors in fast-spiking interneurons

Strengths:

The experimental protocols are technically challenging, but the data is of very high quality, and the presentation and writing are very clear.

I congratulate the authors on submitting a really excellent study demonstrating an activity-dependent alteration in the efficacy of axonal propagation of action potentials in fast-spiking interneurons. It is a well-designed project involving technically challenging experiments, and yet the data is of very high quality, the results are compelling, and the presentation is clear.

Weaknesses:

I have some minor suggestions and comments, including those below, but I hope and expect that these could be performed quickly and without difficulty.

Two of the most interesting figures were consigned to the supplementary information, and I would recommend that they are "upgraded" to be in the main document. The two figures are Figure 1 - Figure Supplement 2, showing the inverse correlation of the AP size with recording distance and branch; and Figure 6 - Figure Supplement 1, showing the postsynaptic effect. My rationale for saying this is that I feel that both add useful biological information to the narrative.

I was glad to see that "realistic" firing patterns were used, because I recall an old modelling paper from Mainen and Sejnowski (https://pubmed.ncbi.nlm.nih.gov/7770778/) that is highly relevant to this paper and should be referenced. However, I would like to suggest one further bit of analysis of the data presented in Figure 4, because I think it will support the main story. In Figure 4, the ostensible conclusion is that there is relative preservation of spike amplitude for this natural firing pattern, but that is almost certainly because the average firing rate is substantially below the level where spike amplitude suppression was seen in Figures 2 & 3. Instead, I recommend analysing for each consecutive spike pair, the ratio of the heights of the two spikes with respect to the interspike interval. Viz<br /> t2 - t1 versus spike 2 amplitude / spike 1 amplitude

The data may be a little noisy, but given the very large number of spike pairs, I would expect to see the suppression effect to be fully evident, and that can feed directly into the model.<br /> I think the author's intuition that dissipation of ionic gradients is a key factor is correct, so I was pleased that Na+ was not ignored in the discussion (the results section only talked about K+).

Perhaps the fact that Na gradients may also be depleted could be mentioned in the results section, too. In the discussion, perhaps the authors could mention two other details: that this "fatigue" may reflect ATP depletion, and progressive failure of the Na-K-ATPase in the axons. That could be examined in a follow-up study (I certainly am not suggesting a raft of experiments for this study), but it could be mentioned in the discussion. And second, that the ionic depletion may be greater within the confines of the cell-attached pipette tip, which is why the branching pattern/distance data (F1FS2), the Ca imaging data and the post-synaptic effects (F6FS1) are such important additional supporting data, because together they indicate that the effect is along the whole axon.

Regarding the rise in [K+]o, it would be worth mentioning the fact that this will be greatly exacerbated by the postsynaptic effects of high-frequency PV activity, because the consequent Cl loading of the postsynaptic cell is subsequently cleared by coupling to K+ extrusion. A good reference for this is http://www.ncbi.nlm.nih.gov/pubmed/20211979; a recent review (https://pubmed.ncbi.nlm.nih.gov/39637123/), which argues that this may even be the dominant source of raised [K+]o in the immediate preictal period, larger even than that exiting cells through the Hodgkin-Huxley mechanism.

The referencing needs some attention. In some instances, the citations either do not really illustrate preceding statements or are simply the wrong citation.

Reviewer #3 (Public review):

Summary:

This is an interesting paper which asks a compelling and translationally relevant question: since the firing rate of GABAergic PV+ interneurons (which powerfully control pyramidal cell excitability) increases prior to and during seizures, why doesn't this increase in inhibition do more to prevent seizure propagation? The authors hypothesize that increased PV+ spiking might lead to spike propagation failures in the axon.

To test this hypothesis, the authors conduct paired electrophysiological recordings from PV+ neurons in acute barrel cortex slices of mice and from a handful of human neurosurgical samples. They use patch clamp recordings to measure the membrane potential of PV+ neurons at the soma, while simultaneously measuring spike propagation with a recording electrode in the axon of the same neuron.

After a variety of elegant experiments and modeling, the authors conclude that extracellular K+ accumulation around the axon during high-frequency firing might be causing propagation failures.

Strengths:

Overall, the paper is nicely written, the experiments are technically challenging, and the figures are, for the most par,t well laid out. The topic will be of broad interest for the neuroscience field, given the relevance of PV+ interneurons to cortical circuit function, plasticity across development, and disease.

Weaknesses:

In addition to the strengths here, I feel the need to highlight a few weaknesses which, if rectified, could improve the work.

(1) The key hypothesis in this paper is that extended periods of somatic spiking lead to progressive decreases in the axonal AP amplitude, which eventually lead to failures, potentially (but not necessarily) at branchpoints. Two comments here.

It would be helpful for the authors to show us examples of the axonal spike waveforms at a faster time base (along with the somatic recording) so that we can really understand what's happening to the spike in the axon.

Their data are also compatible with failures of spike initiation at the AIS. Could the authors show us the first derivative of somatic voltage for successes and failures, and maybe show us some phase plots of Vm vs dV/dt for the failures, successes, and attenuated spikes? Effectively, what I'm asking is whether the changes they see in the distal axon are downstream of the initiation zone. It's very possible that extended spiking is simply depolarizing the AIS and inactivating Na+ channels there. In which case, the authors should be able to pull this out in a phase plot.

(2) There's no baseline period for their calcium fluorescence signals, which is necessary to compare their "signal" magnitude to frame-by-frame variance of dG/R. Could the authors correct this issue in Figure 6B?

(3) Some of their stats are a bit unorthodox. Why are they doing two separate Wilcoxon tests in 6D and 6E? Why not throw all that into a one-way ANOVA model followed by appropriate post hoc tests?

(4) Why don't the authors observe washout of their effect after high K+ application? This concerns me that their high K+ application is having secondary and long-lasting effects on PV excitability, which mimic (but are not necessarily identical) to their hypothesized mechanism of axonal failures.

eLife Assessment

This valuable study contributes to the field of neuro-glial biology by establishing a direct causal link between astrocytic metabolism (glycolysis) and the structural wiring of neural circuits. Connecting the metabolic-synaptic mechanism to locomotor reorientation in the dopaminergic circuit offers new insights into how energy metabolism shapes circuit assembly and function. The evidence offers a solid foundation, moving logically from molecular mechanisms to circuit-level anatomy and finally to behavior; however, several central conclusions currently exceed the direct evidence presented. With appropriate calibration of claims and interpretations and/or additional clarifying experiments, the manuscript has the potential to make a significant contribution to our understanding of glial regulation of circuit assembly.

Reviewer #1 (Public review):

This study investigates how astrocyte metabolic state influences astrocyte-synapse interactions and the organization of the dopaminergic circuit in the Drosophila CNS. Using a creative split-GFP-based contact reporter ("PEAPOD"), combined with genetic perturbations of glycolytic enzymes, synaptic labeling, EM, transsynaptic tracing, single-cell transcriptomics, and behavioral assays, the authors propose that disruption of astrocyte glycolysis enhances astrocyte-dopamine neuron contacts, promotes synaptogenesis, and biases dopaminergic-motor circuit connectivity through a mechanism involving altered Neuroligin 2 trafficking.

The work is conceptually ambitious and technically broad. The development and application of a contact reporter for astrocyte-neuronal interfaces is potentially valuable to the field, and the convergence of multiple glycolytic perturbations on similar phenotypes is a notable strength. However, several central conclusions currently extend beyond the direct evidence presented. Clarification and calibration of these claims would substantially strengthen the manuscript.

Major Points:

(1) Astrocyte glycolytic impairment is inferred rather than directly demonstrated

The central premise of the manuscript is that reduced astrocyte glycolysis drives the observed phenotypes. While multiple glycolytic enzymes (e.g., pfk, eno, pyk) are genetically perturbed and produce similar increases in PEAPOD signal, the manuscript does not directly demonstrate altered glycolytic flux or metabolic state in astrocytes under these conditions. Reduced enzyme levels or genetic mutation do not necessarily establish functional metabolic deficiency, particularly given potential compensatory mechanisms.

Because glycolytic impairment is foundational to the proposed mechanism, the conclusions should either be supported by direct metabolic readouts in astrocytes or framed more cautiously as perturbations of glycolytic enzymes rather than confirmed metabolic deficiency.

(2) Interpretation of the PEAPOD signal requires clearer calibration

The PEAPOD system is an innovative tool to detect membrane proximity between astrocytes and dopamine neurons. However, the manuscript frequently interprets increased PEAPOD intensity and volume as increased "ensheathment" or increased synaptic contact. A split-GFP-based reporter measures membrane apposition within a defined spatial range but does not directly quantify structural ensheathment, synapse number, or functional synaptic engagement.

Although the authors show an association of the PEAPOD signal with presynaptic markers in some regions, the distinction between increased membrane contact, altered membrane organization, and true changes in perisynaptic coverage should be more explicitly discussed. Several conclusions would benefit from clearer wording that distinguishes contact proximity from ultrastructural or functional synapse remodeling.

(3) Evidence for biased dopaminergic-motor circuit wiring is indirect

The manuscript proposes that disruption of astrocyte glycolysis biases dopaminergic-motor connectivity. This conclusion relies heavily on trans-Tango labeling intensity and downstream cell-type composition analysis via FACS and single-cell RNA sequencing.

Transsynaptic labeling approaches can be influenced by expression levels, reporter trafficking, labeling efficiency, and differential recovery during dissociation and FACS. Changes in labeled cell abundance or reporter intensity do not necessarily equate to altered synaptic wiring. Given that this conclusion represents a major conceptual advance of the study, the manuscript should either provide additional orthogonal support or temper the claim to reflect that altered labeling efficiency or synaptic engagement, rather than definitive rewiring, has been demonstrated.

(4) Mechanistic claims regarding Neuroligin 2 trafficking are suggestive but not definitive

The authors propose that astrocyte glycolytic disruption alters Neuroligin 2 (Nlg2) trafficking, leading to ER retention and downstream synaptogenic effects. The observation of Nlg2-positive intracellular bodies colocalizing with ER markers is intriguing. However, quantitative analysis, additional compartment markers, and/or biochemical support would be necessary to firmly establish altered ER exit or glycosylation status.

At present, the mechanistic model is plausible but should be presented more explicitly as a working model supported by suggestive evidence rather than a fully established trafficking defect.

(5) Behavioral phenotypes are not yet causally linked to dopaminergic circuit changes

The locomotor phenotypes observed upon astrocyte glycolytic perturbation are clear. However, the manuscript attributes these changes to altered dopaminergic-motor connectivity. A direct causal linkage between astrocyte metabolic state, dopaminergic circuit remodeling, and behavior is not conclusively demonstrated. The discussion should either clarify the inferential nature of this link or provide additional evidence supporting dopamine-specific dependence.

Minor Points:

(1) Statistical analyses across multi-group comparisons should be more clearly justified, particularly where multiple pairwise tests are performed. A clarification of the multiple-comparison correction and the exact comparison strategy would improve rigor.

(2) The temporal interpretation of activity-dependent remodeling experiments would benefit from a clearer explanation of what timescale is being tested.

(3) Developmental compensation versus the acute effects of glycolytic perturbation are not fully distinguished and should be discussed.

(4) The orthology and functional equivalence of Drosophila Nlg2 should be described with greater precision to avoid potential confusion.

Reviewer #2 (Public review):

Summary:

This manuscript presents a significant advance in our understanding of how metabolic states in astrocytes directly influence the structural assembly and functional output of neural circuits. By focusing on the Drosophila larval dopaminergic system, the authors uncover an interesting mechanism: astrocyte glycolysis acts as a negative regulator of PEAPODs, ultimately altering locomotor behavior. Metabolic fluctuations (e.g., due to diet, development, or disease) could fundamentally reshape neural connectivity, with broad implications for neurodevelopmental and metabolic disorders.

Strengths:

The manuscript offers a compelling narrative linking astrocyte metabolism to DA-MN circuit wiring and behavior. For the field, this study serves as an important prompt to investigate how metabolic states might dynamically tune neural connectivity during development and in disease.

Weaknesses:

The definitive acceptance of the proposed linear mechanism depends on future validation through genetic interaction tests and rescue experiments.

Reviewer #3 (Public review):

Summary:

The authors are trying to demonstrate how astrocytes influence the connections within neural circuits that control behavior.

Strengths:

The data presented in the manuscript are thorough and well-executed, using advanced Drosophila approaches (Ca2+ imaging, GRASP, clonal analysis, trans-Tango) in new ways (PEAPODS) and with new tools (pyk mutants, anti-pyk Ab, LexAop2-pykRNAi). Use of two RNAi lines for each of three glycolytic enzymes is strong evidence that perturbation of glycolysis is responsible, though it does not rule out that inappropriate build-up of intermediates, or shunting to alternative pathways, may play a role here. Subsequent focus on Pyk alone is understandable.

Weaknesses:

As strong as the data is, it does not always support some of the stated claims, and this should be addressed in any revision. In addition, there seems to be an oversimplification of the possible effects of Pyk RNAi, and some missing pieces that could fill in gaps and align the proposed mechanism with observed phenotypes.

Where the data does not support stated claims:

(1) The authors claim larvae executed more reorientation actions during locomotion "as a result" of attenuated astrocyte-to-DAergic neuron signaling through neuroligin 2 (astrocyte) and neurexin 1 (DA Neuron). They correlated these, but did not make a direct connection.

(2) There is a claim that "at the circuit level, behavioral alterations were found to arise from increased DAergic neuronal synaptogenesis and DAergic-motor connection" (sic). However, the work does not build a causal relationship between behavior and synaptogenesis or connectivity. At present, the manuscript does not directly address whether increased DA-motor neuron synapses are sufficient to explain the increased orientation reactions observed.

(3) It is asserted that (line 182, and elsewhere) "astrocyte glycolysis deficiency increased PEAPODs and DAergic neuron synaptogenesis". While astrocyte Pyk KD increased PEAPODS (Figure 2), and it also increased endogenous Brp-GFP in DA neurons (via STaR, Figure 3F), the added Brp-GFP was not localized to synapses under these conditions (pyk KD), to unequivocally demonstrate that the increased PEAPODS are at the sites of DAergic synapses. Also seen in 6I-J.

(4) It may be premature to refer to this strictly as synaptogenesis, as alternative explanations (e.g., stabilization or impaired pruning) could also account for the observations.

(5) The use of trans-Tango is an elegant way to support the idea that extra DAergic synapses are formed onto motor neurons, with potential impact on motor circuits. But again, the claim (line 215, and elsewhere) that this "Biased DAergic-motor wiring" is what "alters motor output", would benefit from additional evidence.

(6) Oversimplification of the possible effects of Pyk RNAi: Because Pyk knockdown is likely to alter glycolytic flux rather than abolish glycolysis entirely, it may be clearer to describe the manipulation as 'Pyk loss' rather than 'glycolysis-deficient' in most contexts.

(7) Filling gaps to align the proposed mechanism with observed phenotypes:

a) Figure 6K-M - the ER retention of Nlg2 should also be tested using Pyk-RNAi, in addition to the pyk mutants shown. This would confirm the astrocyte-specific nature of this effect and close the loop to align the phenotypes.

b) From the mechanism proposed (ER retention of Nlg, presumably leading to loss of Nlg function in astrocytes), one might expect that the effects of loss of Nlg2 from astrocytes could phenocopy the behavioral deficits seen in pyk KD (from astrocytes). Ackerman et al (2021) knocked down Nlg2 from astrocytes and examined motor behavior with FIMTrack. They saw increased accumulated distance but did not see the effects seen upon pyk KD in this manuscript (increased pausing, sweeping). The authors could perform this experiment themselves or alternatively should address this inconsistency in the discussion.

eLife Assessment

This important study by Kong et al. systematically and rigorously dissects the gene regulatory network underlying melanoma and breast cancer risk at the multi-cancer 2q33 locus. The authors provide compelling evidence that rs3769823 is a key functional variant that acts through allele-preferential binding of the transcription factors E4F1 and IRF2 to regulate CASP8 and FLACC1 in a cell-type-specific manner. The work makes a significant contribution to understanding the mechanisms operating at multi-cancer risk loci.

Reviewer #1 (Public review):

Summary:

In this manuscript, Kong et al. conduct a systematic analysis of the multi-cancer risk locus at 2q33. The authors start with a careful analysis of co-localization between the melanoma risk SNPs and several other cancers and conclude that a subset of credible causal SNPs is shared among different cancers, including breast cancer. Next, they define a starting list of 27 SNPs as potential credible causal SNPs and analysis of TADs (topologically associating domains) to zoom in on CASP8 and FLA CC1 as potential target genes. They then systematically rule out coding and splicing variants in the set and focus on a smaller set of three SNPs constituting a melanocyte enhancer element. Using a combination of mass spectrometry, reporter assays, and electrophoretic mobility shift assays, the authors define a role for transcription factors IRF2 and E4F1 in the regulatory network driving risk at the locus.

This work represents a high-quality tour de force, using multiple tools, to zoom in on a gene expression regulatory network associated with risk for multiple cancers. It provides a detailed framework for analyses of other multi-cancer risk loci. Limitations of the work, which is rather a current limitation of the field, is the lack of a model to study how the identified network of regulatory elements, transcription factors, and target genes mechanistically drive risk at the organismal level. Advances such as those described in this manuscript contribute significantly to our knowledge of how common risk variants drive risk.

Reviewer #2 (Public review):

Summary:

Kong et al. investigate a well-validated risk locus at chromosome band 2q33.1 adjacent to CASP8, a ubiquitously expressed and central initiator caspase in the extrinsic apoptotic pathway. Importantly, this region is a multi-cancer risk locus harboring multiple highly correlated risk alleles that are confounded by linkage. In addition to protein coding and splicing variants, further evaluation of eQTL and TWAS results for the locus suggests a cis-regulatory effect is present for CASP8 and nearby FLACC1. The authors prioritize variants using orthogonal statistical fine-mapping approaches and triage top candidates for functional assays. Luciferase reporter assays demonstrated convincing allele-specific regulatory activity of rs3769823 variant as well as suggestive evidence for rs3769821 and rs59308963. These three variants lie in close proximity within a melanocyte regulatory element marked by overlapping promoter and enhancer chromatin state signals. The authors employ a haplotype reporter assay, which shows that the combination of risk alleles in the forward direction has additive effects compared to the protective haplotype. These effects are also cell type specific among melanocytes, melanoma, and breast cancer cell states. Utilizing electron mobility shift assays, the authors convincingly show augmented nuclear protein binding of the rs3769823-A risk allele, and mass spectrometry of allele-specific rs3769823 binding proteins revealed specific activity of E4F1 and IRF2, whose motif score is strengthened by the risk allele. Correlation of these transcription factors' expression with CASP8 expression suggested repressive effects of E4F1 and activating effects of IRF2, which were confirmed in siRNA assays across multiple cell types. These data provide important evidence towards the molecular mechanisms governing disease susceptibility at the 2q33.1 risk locus and nominate s3769823 as a causal variant through cis-regulatory activity by E4F1 and IRF2.

Strengths:

Major strengths of the work include the authors' employment of orthogonal fine-mapping approaches and functional assays in multiple cell types. These help to fortify a novel molecular mechanism of rs3769823 and also work together to propose a complicated multi-variant and cell-type-specific effect at this locus, which is worth future investigation.

Weaknesses:

The rs3769823 variant is a protein-coding variant for CASP8. While the authors conclude that this is likely neutral to CASP8 function, their evidence is suggestive at best and does not close the door on a protein-coding function for this variant.

Similarly, another variant, rs10804111, is associated with alternative splicing of CASP8. The authors do well to include the potent rs10804111 sQTL effect on CASP8 and further confirm it by a minigene assay. However, its exclusion from the fine-mapping results may be due to a potent bias towards active chromatin marks. Therefore, rs10804111 still requires further investigation.<br /> Some attention is given to FLACC1, whose promoter may be in contact with multiple variants. However, little is known about FLACC1 function, and the authors don't provide meaningful supporting data to illustrate whether FLACC1 is relevant in the context of melanocyte, melanoma, or other cancer types that share this risk locus (breast, prostate). Showing the absolute expression levels in the eQTL analysis would be helpful towards this.

Phenotypic assays interrogating the rs3769823-E4F1-IRF2 relevance to melanocyte biology and melanoma pathogenesis are not included.

Finally, the segmented figure organization negatively impacts the readability of the paper.

eLife Assessment

This important study introduces an innovative synthetic nanobody approach to probe the function of the bacterial SMC complex. The work is a compelling example of the potential of this approach. The authors generate protein chimeras to provide convincing evidence that their identified nanobodies target the coiled-coil region of the SMC subunit, demonstrating that this region is critical for SMC function in vivo. Overall, the work is significant for the fields of genome organisation, SMC protein biology, synthetic biology, and bacterial cell biology.

[Editors' note: this paper was reviewed by Review Commons.]

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers. The authors have addressed the major comments raised in the previous round of review. Public Reviews below refer to the version submitted to Review Commons.]

Summary:

Gosselin et al., develop a method to target protein activity using synthetic single-domain nanobodies (sybodies). They screen a library of sybodies using ribosome/ phage display generated against bacillus Smc-ScpAB complex. Specifically, they use an ATP hydrolysis deficient mutant of SMC so as to identify sybodies that will potentially disrupt Smc-ScpAB activity. They next screen their library in vivo, using growth defects in rich media as a read-out for Smc activity perturbation. They identify 14 sybodies that mirror smc deletion phenotype including defective growth in fast-growth conditions, as well as chromosome segregation defects. The authors use a clever approach by making chimeras between bacillus and S. pnuemoniae Smc to narrow-down to specific regions within the bacillus Smc coiled-coil that are likely targets of the sybodies. Using ATPase assays, they find that the sybodies either impede DNA-stimulated ATP hydrolysis or hyperactivate ATP hydrolysis (even in the absence of DNA). The authors propose that the sybodies may likely be locking Smc-ScpAB in the "closed" or "open" state via interaction with the specific coiled-coil region on Smc. I have a few comments that the authors should consider:

Major comments:

(1) Lack of direct in vitro binding measurements:

The authors do not provide measurements of sybody affinities, binding/ unbinding kinetics, stoichiometries with respect to Smc-ScpAB. Additionally, do the sybodies preferentially interact with Smc in ATP/ DNA-bound state? And do the sybodies affect the interaction of ScpAB with SMC?

It is understandable that such measurements for 14 sybodies is challenging, and not essential for this study. Nonetheless, it is informative to have biochemical characterization of sybody interaction with the Smc-ScpAB complex for at least 1-2 candidate sybodies described here.

(2) Many modes of sybody binding to Smc are plausible

The authors provide an elaborate discussion of sybodies locking the Smc-ScpAB complex in open/ closed states. However, in the absence of structural support, the mechanistic inferences may need to be tempered. For example, is it also not possible for the sybodies to bind the inner interface of the coiled-coil, resulting in steric hinderance to coiled-coil interactions. It is also possible that sybody interaction disrupts ScpAB interaction (as data ruling this possibility out has not been provided). Thus, other potential mechanisms would be worth considering/ discussing. In this direction, did AlphaFold reveal any potential insights into putative binding locations?

(3) Sybody expression in vivo

Have the authors estimated sybody expression in vivo? Are they all expressed to similar levels?

(4) Sybodies should phenocopy ATP hydrolysis mutant of Smc

The sybodies were screened against an ATP hydrolysis deficient mutant of Smc, with the rationale that these sybodies would interfere this step of the Smc duty cycle. Does the expression of the sybodies in vivo phenocopy the ATP hydrolysis deficient mutant of Smc? Could the authors consider any phenotypic read-outs that can indicate whether the sybody action results in an smc-null effect or specifically an ATP hydrolysis deficient effect?

Significance:

Overall, this is an impressive study that uses an elegant strategy to find inhibitors of protein activity in vivo. The manuscript is clearly written and the experiments are logical and well-designed. The findings from the study will be significant to the broad field of genome biology, synthetic biology and also SMC biology. Specifically, the coiled coil domain of SMC proteins has been proposed to be of high functional value. The authors have elegantly identified key coiled-coil regions that may be important for function, and parallelly exhibited potential of the use of synthetic sybody/designed binders for inhibition of protein activity.

Reviewer #2 (Public review):

Summary:

Structural Maintenance of Chromosome proteins (SMCs), a family of proteins found in almost all organisms, are organizers of DNA. They accomplish this by a process known as loop extrusion, wherein double-stranded DNA is actively reeled in and extruded into loops. Although SMCs are known to have several DNA binding regions, the exact mechanism by which they facilitate loop extrusion is not understood but is believed to entail large conformational changes. There are currently several models for loop extrusion, including one wherein the coiled coil (CC) arms open, but there is a lack of insightful experimentation and analysis to confirm any of these models. The work presented aims to provide much-needed new tools to investigate these questions: conformation-selective sybodies (synthetic nanobodies) that are likely to alter the CC opening and closing reactions.

The authors produced, isolated, and expressed sybodies that specifically bound to Bacillus subtilis Smc-ScpAB. Using chimeric Smc constructs, where the coiled coils were partly replaced with the corresponding sequences from Streptococcus pneumoniae, the authors revealed that the isolated sybodies all targeted the same 4N CC element of the Smc arms. This region is likely disrupted by the sybodies either by stopping the arms from opening (correctly) or forcing them to stay open (enough). Disrupting these functional elements is suggested to cause the Smc-dependent chromosome organization lethal phenotype, implying that arm opening and closing is a key regulatory feature of bacterial Smc-ScpAB.

Significance:

The authors present a new method for trapping bacterial Smc's in certain conformations using synthetic antibodies. Using these antibodies, they have pinpointed the (previously suggested) 4N region of the coiled coils as an essential site for the opening and closing of the Smc coiled coil arms and that hindering these reactions blocks Smc-driven chromosomal organization. The work has important implications for how we might elucidate the mechanism of DNA loop extrusion by SMC complexes.

Reviewer #3 (Public review):

Summary:

Gosselin et al. use the sybody technology to study effects of in vivo inhibition of the Bacillus subtilis SMC complex. Smc proteins are central DNA binding elements of several complexes that are vital for chromosome dynamics in almost all organisms. Sybodies are selected from three different libraries of the single domain antibodies, using the "transition state" mutant Smc. They identify 14 such mutant sybodies that are lethal when expressed in vivo, because they prevent proper function of Smc. The authors present evidence suggesting that all obtained sybodies bind to a coiled-coil region close to the Smc "neck", and thereby interfere with the Smc activity cycle, as evidenced by defective ATPase activity when Smc is bound to DNA.

The study is well done and presented and shows that the strategy is very potent in finding a means to quickly turn off a protein's function in vivo, much quicker than depleting the protein.

The authors also draw conclusions on the molecular mode of action of the SMC complex. The provide a number of suggestive experiments, but in my view mostly indirect evidence for such mechanism.

My main criticism is that the authors have used a single - and catalytically trapped form of SMC. They speculate why they only obtain sybodies from one library, and then only identify sybodies that bind to a rather small part of the large Smc protein. While the approach is definitely valuable, it is biassed towards sybodies that bind to Smc in a quite special way, it seems. Using wild type Smc would be interesting, to make more robust statements about the action of sybodies potentially binding to different parts of Smc.

Line 105: Alternatively, the other libraries did not produce good binders or these sybodies were 106 not stably expressed in B. subtilis. This could be tested using Western blotting - I am assuming sybody antibodies are commercially available. However, this test is not important for the overall study, it would just clarify a minor point.

Fig. 2B: is odd to count Spo0J foci per cells, as it is clear from the images that several origins must be present within the fluorescent foci. I am fine with the "counting" method, as the images show there is a clear segregation defect when sybodies are expressed, I believe the authors should state, though, that this is not a replication block, but failure to segregate origins.

Testing binding sites of sybodies to the SMC complex is done in an indirect manner, by using chimeric Smc constructs. I am surprised why the authors have not used in vitro crosslinking: the authors can purify Smc, and mass spectrometry analyses would identify sites where sybodies are crosslinked to Smc. Again, I am fine with the indirect method, but the authors make quite concrete statements on binding based on non-inhibition of chimeric Smc; I can see alternative explanations why a chimera may not be targeted.

Smc-disrupting sybodies affect the ATPase activity in one of two ways. Again, rather indirect experiments. This leads to the point Revealing Smc arm dynamics through synthetic binders in the discussion. The authors are quite careful in stating that their experiments are suggestive for a certain mode of action of Smc, which is warranted.

In line 245, they state More broadly, the study demonstrates how synthetic binders can trap, stabilize, or block transient conformations of active chromatin-associated machines, providing a powerful means to probe their mechanisms in living cells. This is off course a possible scenario for the use of sybodies, but the study does not really trap Smc in a transient conformation, at least this is not clearly shown.

Overall, it is an interesting study, with a well-presented novel technology, and a limited gain of knowledge on SMC proteins.

Significance:

The work describes the gaining and use of single-binder antibodies (sybodies) to interfere with the function of proteins in bacteria. Using this technology for the SMC complex, the authors demonstrate that they can obtain a significant of binders that target a defined region is SMC and thereby interfere with the ATPase cycle.

The study does not present a strong gain of knowledge of the mode of action of the SMC complex.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

Gosselin et al., develop a method to target protein activity using synthetic single-domain nanobodies (sybodies). They screen a library of sybodies using ribosome/ phage display generated against bacillus Smc-ScpAB complex. Specifically, they use an ATP hydrolysis deficient mutant of SMC so as to identify sybodies that will potentially disrupt Smc-ScpAB activity. They next screen their library in vivo, using growth defects in rich media as a read-out for Smc activity perturbation. They identify 14 sybodies that mirror smc deletion phenotype including defective growth in fast-growth conditions, as well as chromosome segregation defects. The authors use a clever approach by making chimeras between bacillus and S. pnuemoniae Smc to narrow-down to specific regions within the bacillus Smc coiled-coil that are likely targets of the sybodies. Using ATPase assays, they find that the sybodies either impede DNA-stimulated ATP hydrolysis or hyperactivate ATP hydrolysis (even in the absence of DNA). The authors propose that the sybodies may likely be locking Smc-ScpAB in the "closed" or "open" state via interaction with the specific coiled-coil region on Smc. I have a few comments that the authors should consider:

Major comments:

(1) Lack of direct in vitro binding measurements:

The authors do not provide measurements of sybody affinities, binding/ unbinding kinetics, stoichiometries with respect to Smc-ScpAB. Additionally, do the sybodies preferentially interact with Smc in ATP/ DNA-bound state? And do the sybodies affect the interaction of ScpAB with SMC?

It is understandable that such measurements for 14 sybodies is challenging, and not essential for this study. Nonetheless, it is informative to have biochemical characterization of sybody interaction with the Smc-ScpAB complex for at least 1-2 candidate sybodies described here.

We agree with the reviewer that adding such data would be reassuring and that obtaining solid data using purified components is not trivial, even for a smaller selection of sybodies. We have now incorporated ELISA data as new Table S1, which shows that most sybodies support clear binding to Smc-ScpAB. Curiously, while (only) some sybodies show a clear preference for ATP-bound or unbound Smc, this is not a strong predictor of the strength of phenotype observed in vivo. We have also attempted to characterize the binding of Smc to sybodies by other methods including pull-downs, cross-linking, and by biophysical methods (GCI). However, we prefer to not include these data as the outcomes are not clear due to inconsistencies in the behaviour of purified sybodies.

(2) Many modes of sybody binding to Smc are plausible

The authors provide an elaborate discussion of sybodies locking the Smc-ScpAB complex in open/ closed states. However, in the absence of structural support, the mechanistic inferences may need to be tempered. For example, is it also not possible for the sybodies to bind the inner interface of the coiled-coil, resulting in steric hinderance to coiled-coil interactions. It is also possible that sybody interaction disrupts ScpAB interaction (as data ruling this possibility out has not been provided). Thus, other potential mechanisms would be worth considering/ discussing. In this direction, did AlphaFold reveal any potential insights into putative binding locations?

We have attempted to map the binding by structure prediction, however, so far, even the latest versions of AlphaFold are not able to clearly delineate the binding interface that we have confidently identified by the mapping using chimeric proteins. Indeed, many ways of binding are possible, including disruption of ScpAB interaction. However, since the mapped binding sites are located on the SMC coiled coils, the later scenario seems unlikely and would be an indirect consequence of altered coiled coil configuration, consistent with our current interpretation.

(3) Sybody expression in vivo

Have the authors estimated sybody expression in vivo? Are they all expressed to similar levels?

We have tagged selected sybodies with gfp and performed live cell imaging. This shows that sybodies without strong phenotypes are similarly expressed at least at low inducer concentration. Moreover, many sybodies localize as foci in the cell presumably by binding to Smc complexes loaded onto the chromosome at ParB/parS sites. We have included example data in the revised version of the manuscript as Figure S4 and Figure S5. Notably, a sybody (Sb007) with a weak growth phenotype shows focal localization at low inducer concentration and high expression levels when fully induced, comparable to sybodies with strong phenotypes. Altogether, this suggests that the lack of phenotype is not due to absence of sybody expression or localization.

(4) Sybodies should phenocopy ATP hydrolysis mutant of Smc

The sybodies were screened against an ATP hydrolysis deficient mutant of Smc, with the rationale that these sybodies would interfere this step of the Smc duty cycle. Does the expression of the sybodies in vivo phenocopy the ATP hydrolysis deficient mutant of Smc? Could the authors consider any phenotypic read-outs that can indicate whether the sybody action results in an smc-null effect or specifically an ATP hydrolysis deficient effect?

As alluded to above, we think that our selection gave rise to sybodies that bind various, possibly multiple Smc conformations. Consistent with this idea, the phenotypes of sybody expression are similar to null mutant rather than the ATP-hydrolysis defective EQ mutant, which display even more severe growth phenotypes in B. subtilis. To highlight this point, we have added the following notes to the text:

“These conditions favour ATP-engaged particles alongside the typically predominant ATP-disengaged rod-shaped state.”

“ELISA data revealed that nearly all clones bind purified Smc-ScpAB (Table 1). However, the ELISA signals of only few Sybodies showed clear dependence on the presence or absence of ATP and DNA (Table S1).”

Significance:

Overall, this is an impressive study that uses an elegant strategy to find inhibitors of protein activity in vivo. The manuscript is clearly written and the experiments are logical and well-designed. The findings from the study will be significant to the broad field of genome biology, synthetic biology and also SMC biology. Specifically, the coiled coil domain of SMC proteins have been proposed to be of high functional value. The authors have elegantly identified key coiled-coil regions that may be important for function, and parallelly exhibited potential of the use of synthetic sybody/designed binders for inhibition of protein activity.

Reviewer #2 (Public review):

Summary:

Structural Maintenance of Chromosome proteins (SMCs), a family of proteins found in almost all organisms, are organizers of DNA. They accomplish this by a process known as loop extrusion, wherein double-stranded DNA is actively reeled in and extruded into loops. Although SMCs are known to have several DNA binding regions, the exact mechanism by which they facilitate loop extrusion is not understood but is believed to entail large conformational changes. There are currently several models for loop extrusion, including one wherein the coiled coil (CC) arms open, but there is a lack of insightful experimentation and analysis to confirm any of these models. The work presented aims to provide much-needed new tools to investigate these questions: conformation-selective sybodies (synthetic nanobodies) that are likely to alter the CC opening and closing reactions.

The authors produced, isolated, and expressed sybodies that specifically bound to Bacillus subtilis Smc-ScpAB. Using chimeric Smc constructs, where the coiled coils were partly replaced with the corresponding sequences from Streptococcus pneumoniae, the authors revealed that the isolated sybodies all targeted the same 4N CC element of the Smc arms. This region is likely disrupted by the sybodies either by stopping the arms from opening (correctly) or forcing them to stay open (enough). Disrupting these functional elements is suggested to cause the Smc-dependent chromosome organization lethal phenotype, implying that arm opening and closing is a key regulatory feature of bacterial Smc-ScpAB.

Significance:

The authors present a new method for trapping bacterial Smc's in certain conformations using synthetic antibodies. Using these antibodies, they have pinpointed the (previously suggested) 4N region of the coiled coils as an essential site for the opening and closing of the Smc coiled coil arms and that hindering these reactions blocks Smc-driven chromosomal organization. The work has important implications for how we might elucidate the mechanism of DNA loop extrusion by SMC complexes.

Reviewer #3 (Public review):

Summary:

Gosselin et al. use the sybody technology to study effects of in vivo inhibition of the Bacillus subtilis SMC complex. Smc proteins are central DNA binding elements of several complexes that are vital for chromosome dynamics in almost all organisms. Sybodies are selected from three different libraries of the single domain antibodies, using the "transition state" mutant Smc. They identify 14 such mutant sybodies that are lethal when expressed in vivo, because they prevent proper function of Smc. The authors present evidence suggesting that all obtained sybodies bind to a coiled-coil region close to the Smc "neck", and thereby interfere with the Smc activity cycle, as evidenced by defective ATPase activity when Smc is bound to DNA.

The study is well done and presented and shows that the strategy is very potent in finding a means to quickly turn off a protein's function in vivo, much quicker than depleting the protein.

The authors also draw conclusions on the molecular mode of action of the SMC complex. The provide a number of suggestive experiments, but in my view mostly indirect evidence for such mechanism.

My main criticism is that the authors have used a single - and catalytically trapped form of SMC. They speculate why they only obtain sybodies from one library, and then only identify sybodies that bind to a rather small part of the large Smc protein. While the approach is definitely valuable, it is biassed towards sybodies that bind to Smc in a quite special way, it seems. Using wild type Smc would be interesting, to make more robust statements about the action of sybodies potentially binding to different parts of Smc.

The reviewer reports (Rev. #1 and Rev. #3) made us realize that the manuscript text was misleading on the this point. Although we used the purified ATP hydrolysis–deficient Smc protein for sybody isolation, this is not expected to restrict the selection to a specific conformation. As described in detail in Vazquez-Nunez et al. (Figure 5), this mutant displays the ATP-engaged conformation only in a smaller fraction of complexes (~25% in the presence of ATP and DNA), consistent with prior in vivo observations reported by Diebold-Durand et al. (Figure 5). Rather than limiting the selection to a particular configuration, our aim was to reduce the prevalence of the predominant rod state in order to broaden the range of conformations represented during sybody selection. Consistent with this interpretation, only a small number of isolated sybodies show strong conformation-specific binding in the presence or absence of ATP/DNA, as observed by ELISA (now included in the manuscript). Notably, the effect size of ATP/DNA on ELISA signals was not a strong predictor to the strength of phenotypes observed in vivo. The text has been revised accordingly. See line 84 and line 92.

We are thus quite confident based prior work (and on the now included ELISA data) that the Smc ATPase mutation did not strongly bias the selection in one way or another. The surprising bias towards coiled coil binding sites has likely other explanations, as they likely form a preferred epitope recognized by sybodies from the loop library.

Line 105: Alternatively, the other libraries did not produce good binders or these sybodies were 106 not stably expressed in B. subtilis. This could be tested using Western blotting - I am assuming sybody antibodies are commercially available. However, this test is not important for the overall study, it would just clarify a minor point.

While there are antibody fragments available to augment the size of sybodies (PMID: 40108246), these recognize 3D-epitopes and are thus not suited for Western blotting. We did not follow up on the negative results of two of the three libraries but would like to point out again that there are several biases that likely emerge for the same reason (bias to library, bias to coiled coil binding site). If correct, then sybodies are likely ineffective in inactivating Smc in B. subtilis, with the notable exceptions of the sybodies that we have isolated and characterized in this manuscript. We have added this notion to the manuscript.

Fig. 2B: is odd to count Spo0J foci per cells, as it is clear from the images that several origins must be present within the fluorescent foci. I am fine with the "counting" method, as the images show there is a clear segregation defect when sybodies are expressed, I believe the authors should state, though, that this is not a replication block, but failure to segregate origins.

We agree that this is an important point. We have added the following statement to clarify this point: “These elongated cells are known to harbour expanded nucleoids, consistent with delayed oriC separation rather than delayed DNA replication”

Testing binding sites of sybodies to the SMC complex is done in an indirect manner, by using chimeric Smc constructs. I am surprised why the authors have not used in vitro crosslinking: the authors can purify Smc, and mass spectrometry analyses would identify sites where sybodies are crosslinked to Smc. Again, I am fine with the indirect method, but the authors make quite concrete statements on binding based on non-inhibition of chimeric Smc; I can see alternative explanations why a chimera may not be targeted.

We have made several attempts of testing direct binding with mixed outcomes and decided to not include those results in the light of the stronger and more relevant in vivo mapping. However, we have added ELISA results (new Table S1) that support a direct interaction.

Smc-disrupting sybodies affect the ATPase activity in one of two ways. Again, rather indirect experiments. This leads to the point Revealing Smc arm dynamics through synthetic binders in the discussion. The authors are quite careful in stating that their experiments are suggestive for a certain mode of action of Smc, which is warranted.

In line 245, they state More broadly, the study demonstrates how synthetic binders can trap, stabilize, or block transient conformations of active chromatin-associated machines, providing a powerful means to probe their mechanisms in living cells. This is off course a possible scenario for the use of sybodies, but the study does not really trap Smc in a transient conformation, at least this is not clearly shown.

We agree and have simplified the statement by removing “stabilize” and “transient”.

Overall, it is an interesting study, with a well-presented novel technology, and a limited gain of knowledge on SMC proteins.

We respectfully disagree with the last point, since our unique results highlight the importance of the Smc coiled coils. which are less well represented in the SMC literature (when compared to the heads and hinge domains for example), likely (at least in part) due the mild effect of single point mutations on coiled coil dynamics.

Significance:

The work describes the gaining and use of single-binder antibodies (sybodies) to interfere with the function of proteins in bacteria. Using this technology for the SMC complex, the authors demonstrate that they can obtain a significant of binders that target a defined region is SMC and thereby interfere with the ATPase cycle.

The study does not present a strong gain of knowledge of the mode of action of the SMC complex.

As pointed out above, we respectfully disagree with this assertion.

eLife Assessment

This valuable study focuses on a unique morphogenetic module, the junction-based lamellipodia (JBL). It provides a biomechanical understanding of how JBLs control endothelial cell-cell junctional remodelling to generate lumenised, multicellular blood vessels. The manuscript represents a robust, thoughtfully executed, and convincing study that uses high-resolution time-lapse imaging combined with pharmacological treatments to advance our understanding of lumen formation in vascular development.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers. The authors have addressed the comments raised in the previous round of review.]

Original review:

Summary:

Lumen formation is a fundamental morphogenetic event essential for the function of all tubular organs, notably the vertebrate vascular network, where continuous and patent conduits ensure blood flow and tissue perfusion. The mechanisms by which endothelial cells organize to create and maintain luminal space have historically been categorized into two broad strategies: cell shape changes, which involve alterations in apical-basal polarity and cytoskeletal architecture, and cell rearrangements, wherein intercellular junctions and positional relationships are remodeled to form uninterrupted conduits. The study presented here focuses on the latter process, highlighting a unique morphogenetic module, junction-based lamellipodia (JBL), as the driver for endothelial rearrangements.

Strengths:

The key mechanistic insight from this work is the requirement of the Arp2/3 complex, the classical nucleator of branched actin filament networks, for JBL protrusion. This implicates Arp2/3-mediated actin polymerization in pushing force generation, enabling plasma membrane advancement at junctional sites. The dependence on Arp2/3 positions JBL within the family of lamellipodia-like structures, but the junctional origin and function distinguish them from canonical, leading-edge lamellipodia seen in cell migration.

Weaknesses:

The study primarily presents descriptive observations and includes limited quantitative analyses or genetic modifications. Molecular mechanisms are typically interrogated through the use of pharmacological inhibitors rather than genetic approaches. Furthermore, the precise semantic distinction between JAIL and JBL requires additional clarification, as current evidence suggests their biological relevance may substantially overlap.

Reviewer #2 (Public review):

Original review:

Summary:

In Maggi et al., the authors investigated the mechanisms that regulate the dynamics of a specialized junctional structure called junction-based lamellipodia (JBL), which they have previously identified during multicellular vascular tube formation in the zebrafish. They identified the Arp2/3 complex to dynamically localize at expanding JBLs and showed that the chemical inhibition of Arp2/3 activity slowed junctional elongation. The authors therefore concluded that actin polymerization at JBLs pushes the distal junction forward to expand the JBL. They further revealed the accumulation of Myl9a/Myl9b (marker for MLC) at the junctional pole, at interjunctional regions, suggesting that contractile activity drives the merging of proximal and distal junctions. Indeed, chemical inhibition of ROCK activity decreased junctional mergence. With these new findings, the authors added new molecular and cellular details into the previously proposed clutch mechanism by proposing that Arp2/3-dependent actin polymerization provides pushing forces while actomyosin contractility drives the merging of proximal and distal junctions, explaining the oscillatory protrusive nature of JBLs.

Strengths:

The authors provide detailed analyses of endothelial cell-cell dynamics through time-lapse imaging of junctional and cytoskeletal components at subcellular resolution. The use of zebrafish as an animal model system is invaluable in identifying novel mechanisms that explain the organizing principles of how blood vessels are formed. The data is well presented, and the manuscript is easy to read.

Weaknesses:

While the data generally support the conclusions reached, some aspects can be strengthened. For the untrained eye, it is unclear where the proximal and distal junctions are in some images, and so it is difficult to follow their dynamics (especially in experiments where Cdh5 is used as the junctional marker). Images would benefit from clear annotation of the two junctions. All perturbation experiments were done using chemical inhibitors; this can be further supported by genetic perturbations.

Reviewer #3 (Public review):

Original review:

The paper by Maggi et al. builds on earlier work by the team (Paatero et al., 2018) on oriented junction-based lamellipodia (JBL). They validate the role of JBLs in guiding endothelial cell rearrangements and utilise high-resolution time-lapse imaging of novel transgenic strains to visualise the formation of distal junctions and their subsequent fusion with proximal junctions. Through functional analyses of Arp2/3 and actomyosin contractility, the study identifies JBLs as localized mechanical hubs, where protrusive forces drive distal junction formation, and actomyosin contractility brings together the distal and proximal junctions. This forward movement provides a unique directionality which would contribute to proper lumen formation, EC orientation, and vessel stability during these early stages of vessel development.

Time-lapse live imaging of VEC, ZO-1, and actin reveals that VEC and ZO-1 are initially deposited at the distal junction, while actin primarily localizes to the region between the proximal and distal sites. Using a photoconvertible Cdh5-mClav2 transgenic line, the origin of the VEC aggregates was examined. This convincingly shows that VE-cadherin was derived from pools outside the proximal junctions. However, in addition to de novo VEC derived from within the photoconverted cell, could some VEC also be contributed by the neighbouring endothelial cell to which the JBL is connected?

As seen for JAILs in cultured ECs, the study reveals that Arp2/3 is enhanced when JBLs form by live imaging of Arpc1b-Venus in conjunction with ZO-1 and actin. Therefore Arp2/3 likely contributes to the initial formation of the distal junction in the lamellopodium.

Inhibiting Arp2/3 with CK666 prevents JBL formation, and filopodia form instead of lamellopodia. This loss of JBLs leads to impaired EC rearrangements.

Is the effect of CK666 treatment reversible? Since only a short (30 min) treatment is used, the overall effect on the embryo would be minimal, and thus washing out CK666 might lead to JBL formation and normalized rearrangements, which would further support the role of Arp2/3.

From the images in Figure 4d it appears that ZO-1 levels are increased in the ring after CK666 treatment. Has this been investigated, and could this overall stabilization of adhesion proteins further prevent elongation of the ring?

To explore how the distal and proximal junctions merge, imaging of spatiotemporal imaging of Myl9 and VEC is conducted. It indicates that Myl9 is localized at the interjunctional fusion site prior to fusion. This suggests pulling forces are at play to merge the junctions, and indeed Y 27632 treatment reduces or blocks the merging of these junctions.

For this experiment, a truncated version of VEC was use,d which lacks the cytoplasmic domain. Why have the authors chosen to image this line, since lacking the cytoplasmic domain could also impair the efficiency of tension on VEC at both junction sites? This is as described in the discussion (lines 328-332).

Since the time-lapse movies involve high-speed imaging of rather small structures, it is understandable that these are difficult to interpret. Adding labels to indicate certain structures or proteins at essential timepoints in the movies would help the readers understand these.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

Lumen formation is a fundamental morphogenetic event essential for the function of all tubular organs, notably the vertebrate vascular network, where continuous and patent conduits ensure blood flow and tissue perfusion. The mechanisms by which endothelial cells organize to create and maintain luminal space have historically been categorized into two broad strategies: cell shape changes, which involve alterations in apical-basal polarity and cytoskeletal architecture, and cell rearrangements, wherein intercellular junctions and positional relationships are remodeled to form uninterrupted conduits. The study presented here focuses on the latter process, highlighting a unique morphogenetic module, junction-based lamellipodia (JBL), as the driver for endothelial rearrangements.

Strengths:

The key mechanistic insight from this work is the requirement of the Arp2/3 complex, the classical nucleator of branched actin filament networks, for JBL protrusion. This implicates Arp2/3-mediated actin polymerization in pushing force generation, enabling plasma membrane advancement at junctional sites. The dependence on Arp2/3 positions JBL within the family of lamellipodia-like structures, but the junctional origin and function distinguish them from canonical, leading-edge lamellipodia seen in cell migration.

Weaknesses:

The study primarily presents descriptive observations and includes limited quantitative analyses or genetic modifications. Molecular mechanisms are typically interrogated through the use of pharmacological inhibitors rather than genetic approaches. Furthermore, the precise semantic distinction between JAIL and JBL requires additional clarification, as current evidence suggests their biological relevance may substantially overlap.

We have previously analyzed the effects of different ve-cadherin (cdh5) mutant alleles on EC rearrangements (Paatero et al., 2018; Sauteur et al., 2014).These mutants show complex defects (e.g. hypersprouting, reduced contact inhibition during anastomosis) in EC behavior early in vascular tube formation. We find that analysis of JBL dynamics and function is very difficult in such situations. The use of small molecule inhibitors allows acute interventions within a defined time-window and to avoid pleiotropic effects of genetic ablations. We have expanded our discussion on the distinction between JAIL and JBL and hope that this will clarify why – in our opinion – these terms should be used differentially in different cell biological contexts (see below and lines 348-374 in the manuscript).

Reviewer #2 (Public review):

Summary:

In Maggi et al., the authors investigated the mechanisms that regulate the dynamics of a specialized junctional structure called junction-based lamellipodia (JBL), which they have previously identified during multicellular vascular tube formation in the zebrafish. They identified the Arp2/3 complex to dynamically localize at expanding JBLs and showed that the chemical inhibition of Arp2/3 activity slowed junctional elongation. The authors therefore concluded that actin polymerization at JBLs pushes the distal junction forward to expand the JBL. They further revealed the accumulation of Myl9a/Myl9b (marker for MLC) at the junctional pole, at interjunctional regions, suggesting that contractile activity drives the merging of proximal and distal junctions. Indeed, chemical inhibition of ROCK activity decreased junctional mergence. With these new findings, the authors added new molecular and cellular details into the previously proposed clutch mechanism by proposing that Arp2/3-dependent actin polymerization provides pushing forces while actomyosin contractility drives the merging of proximal and distal junctions, explaining the oscillatory protrusive nature of JBLs.

Strengths:

The authors provide detailed analyses of endothelial cell-cell dynamics through time-lapse imaging of junctional and cytoskeletal components at subcellular resolution. The use of zebrafish as an animal model system is invaluable in identifying novel mechanisms that explain the organizing principles of how blood vessels are formed. The data is well presented, and the manuscript is easy to read.

Weaknesses:

While the data generally support the conclusions reached, some aspects can be strengthened. For the untrained eye, it is unclear where the proximal and distal junctions are in some images, and so it is difficult to follow their dynamics (especially in experiments where Cdh5 is used as the junctional marker). Images would benefit from clear annotation of the two junctions. All perturbation experiments were done using chemical inhibitors; this can be further supported by genetic perturbations.

We have added annotations to several figures and paid particular attention to the proximal and distal junctions.

We have previously analyzed the effects of different ve-cadherin (cdh5) mutant alleles on EC rearrangements (Paatero et al., 2018; Sauteur et al., 2014). These mutants show complex defects (e.g. hypersprouting, reduced contact inhibition during anastomosis) in EC behavior early in vascular tube formation. We find that analysis of JBL dynamics and function is very difficult in such situations. The use of small inhibitors allows acute interventions within a defined time-window and to avoid pleiotropic effects of genetic ablations.

Reviewer #3 (Public review):

The paper by Maggi et al. builds on earlier work by the team (Paatero et al., 2018) on oriented junction-based lamellipodia (JBL). They validate the role of JBLs in guiding endothelial cell rearrangements and utilise high-resolution time-lapse imaging of novel transgenic strains to visualise the formation of distal junctions and their subsequent fusion with proximal junctions. Through functional analyses of Arp2/3 and actomyosin contractility, the study identifies JBLs as localized mechanical hubs, where protrusive forces drive distal junction formation, and actomyosin contractility brings together the distal and proximal junctions. This forward movement provides a unique directionality which would contribute to proper lumen formation, EC orientation, and vessel stability during these early stages of vessel development.

Time-lapse live imaging of VEC, ZO-1, and actin reveals that VEC and ZO-1 are initially deposited at the distal junction, while actin primarily localizes to the region between the proximal and distal sites. Using a photoconvertible Cdh5-mClav2 transgenic line, the origin of the VEC aggregates was examined. This convincingly shows that VE-cadherin was derived from pools outside the proximal junctions. However, in addition to de novo VEC derived from within the photoconverted cell, could some VEC also be contributed by the neighbouring endothelial cell to which the JBL is connected?

Yes, the green (non-converted) VE-cadherin can indeed originate from either of the two cells. The main point we want to make, based on our observations, is that the red (converted) VE-cadherin from the proximal junction (as defined by the ROI) does not contribute to the distal junction.

As seen for JAILs in cultured ECs, the study reveals that Arp2/3 is enhanced when JBLs form by live imaging of Arpc1b-Venus in conjunction with ZO-1 and actin. Therefore Arp2/3 likely contributes to the initial formation of the distal junction in the lamellopodium.

Inhibiting Arp2/3 with CK666 prevents JBL formation, and filopodia form instead of lamellopodia. This loss of JBLs leads to impaired EC rearrangements.

Is the effect of CK666 treatment reversible? Since only a short (30 min) treatment is used, the overall effect on the embryo would be minimal, and thus washing out CK666 might lead to JBL formation and normalized rearrangements, which would further support the role of Arp2/3.

We have performed washout experiments and find that the ectopic filopodia disappear when the inhibitor is removed. This experiment is shown in supplementary Figure 3 and supplementary Movies 12 and 13.

From the images in Figure 4d it appears that ZO-1 levels are increased in the ring after CK666 treatment. Has this been investigated, and could this overall stabilization of adhesion proteins further prevent elongation of the ring?

This is an interesting thought and we haven take a closer look. There is quite a bit of sample-to-sample variation in the ZO1 signal. The quantification (Author response image 1) indicates that there is no increase in the CK666 treated embryos on average.

Author response image 1.

To explore how the distal and proximal junctions merge, imaging of spatiotemporal imaging of Myl9 and VEC is conducted. It indicates that Myl9 is localized at the interjunctional fusion site prior to fusion. This suggests pulling forces are at play to merge the junctions, and indeed Y 27632 treatment reduces or blocks the merging of these junctions.

For this experiment, a truncated version of VEC was use,d which lacks the cytoplasmic domain. Why have the authors chosen to image this line, since lacking the cytoplasmic domain could also impair the efficiency of tension on VEC at both junction sites? This is as described in the discussion (lines 328-332).

This line was used because it labels the entire JBL protrusion more clearly. We have also included an example using the VE-cad-Venus line (supplementary Figure 4b), which shows a Myl-Cherry pattern consistent with the other examples.

Since the time-lapse movies involve high-speed imaging of rather small structures, it is understandable that these are difficult to interpret. Adding labels to indicate certain structures or proteins at essential timepoints in the movies would help the readers understand these.

We have added annotations and labels to all movies. We have also improved annotations in several figures (i.e. Figs. 1, 2, 5, 6 and 7)

Recommendations for the authors:

Reviewing Editor Comments:

Overall, the reviewers are supportive of the manuscript but identify a number of areas where the clarity of the presented data could be improved, and further quantification could be provided to strengthen your conclusions. We would encourage you to address these minor concerns as best you can and to consider the recommendations of all three reviewers when deciding how to revise your manuscript.

Reviewer #1 (Recommendations for the authors):

Lumen formation is a fundamental morphogenetic event essential for the function of all tubular organs, notably the vertebrate vascular network, where continuous and patent conduits ensure blood flow and tissue perfusion. The mechanisms by which endothelial cells organize to create and maintain luminal space have historically been categorized into two broad strategies: cell shape changes, which involve alterations in apical-basal polarity and cytoskeletal architecture, and cell rearrangements, wherein intercellular junctions and positional relationships are remodeled to form uninterrupted conduits. The study presented here focuses on the latter process, highlighting a unique morphogenetic module, junction-based lamellipodia (JBL), as the driver for endothelial rearrangements.

JBL are described as oscillating membrane protrusions emerging at endothelial junctions, operating in a ratchet-like manner to mediate convergent cell movements. This ratchet mechanism allows endothelial cells to approach each other, thereby aligning and joining local luminal segments into a continuous vascular structure. The study employs in vivo high-resolution time-lapse imaging, a technically demanding method that captures spatiotemporal dynamics of cytoskeletal and adhesion complexes during JBL activity with unprecedented detail.

The key mechanistic insight from this work is the requirement of the Arp2/3 complex, the classical nucleator of branched actin filament networks, for JBL protrusion. This implicates Arp2/3-mediated actin polymerization in pushing force generation, enabling plasma membrane advancement at junctional sites. The dependence on Arp2/3 positions JBL within the family of lamellipodia-like structures, but the junctional origin and function distinguish them from canonical, leading-edge lamellipodia seen in cell migration.

An intriguing observation is that a novel junction arises at the distal pole of a JBL. This distal junction is formed from a pool of VE-cadherin that is spatially redistributed from regions outside the initial JBL domain. The distal junction then merges with the proximal junction through a process dependent on actomyosin contractility, as was judged by Myl9 recruitment.

The alternation between pushing forces (Arp2/3-dependent JBL protrusion) and pulling forces (actomyosin-driven junction fusion) defines JBL as a bidirectional mechanical module. Inhibition of actomyosin prevents merging of proximal and distal junctions, thereby stalling lumen continuity. This two-phase system, actin-based extension followed by actomyosin-mediated constriction, ensures both elongation and maturation of endothelial arrangements, ultimately securing vascular patency.

This manuscript represents a robust and thoughtfully executed study that advances our understanding of lumen formation during vascular development. The overarching conclusions are well substantiated, and the results section provides a clear and detailed exposition of the key findings. I appreciate the explanatory movie at the end. Nevertheless, I offer several remarks for further improvement:

(1) The fluorescent images presented are visually compelling, yet lack quantitative analysis in the initial figure. Although quantification is included in Figure 3, it is advisable to incorporate this analysis into Figure 1 as well. Early presentation of quantification will help the reader to appreciate the impact and significance of the findings from the outset.

We appreciate the reviewer’s suggestion and have now added line graphs to measure the spatiotemporal intensities of the Utrophin and ZO-1 reporters in Figure 1b. These measurements demonstrate the sequence of F-actin protrusion and subsequent junctional movement. In Figure 1a, we have added a double-headed arrow which shows the overall movement of the junction towards the dorsal side of the forming DLAV.

(2) For the fluorescence images, further quantitative analysis of membrane overlap, either in terms of width or pixel overlap, would enhance the rigor of the study. Temporal quantification of overlap may provide valuable insights into the stability and reproducibility of the process across experimental replicates.

JBL are quite heterogenous with respect to size, shape and dynamics, which makes quantifications of membrane overlap (JBL size) across experimental replicates difficult. We have published some quantifications on JBL orientation and oscillation in our previous paper (Paatero et al., 2018, Nat. comm. Figures 1 and 2), which are in agreement with our current study.

(3) When referencing the role of Arp2/3, the authors employ an ArpC1b transgenic fish. The results section should thus specifically address the involvement of ArpC1b, rather than generalizing to Arp2/3. In the discussion, it would be appropriate to speculate on the potential involvement of the complete Arp2/3 complex. Notably, the use of CK is acknowledged as a broadly accepted inhibitor of actin polymerization.

As ArpC1b is a subunit of an active Arp2/3 complex (Padrick et al., 2011), we have used an ArpC1b-Venus as a readout for Arp2/3 localization. The construct has been validated before in cell culture (Law et al., 2021) as well as in zebrafish (Malchow et al., 2024) and the spatiotemporal distribution of the reporter shown to be consistent with Arp2/3 complex. We are stating this in the results section (lines 173-178) and subsequently use the term Arp2/3 to facilitate reading of the text. In the corresponding figure legends, we are maintaining the term ArpC1b. CK666 interferes with the dimerization of Arp2 and Arp3 subunits and thus prevents activity of the Arp2/3 complex.

(4) The discussion regarding JAIL versus JBL involvement remains challenging to interpret. If JAIL structures arise from the loss of cell-cell contacts, both JAIL and JBL resemble membrane protrusions and are likely governed by similar molecular mechanisms, predominantly actin polymerization and Arp2/3 activity, with probable contribution from Rac1 signaling. The precise semantic distinction between JAIL and JBL warrants further clarification, as their biological relevance may be overlapping.

We agree with the reviewer. Below we outline the reasons why lamellipodial protrusions that emanate from cell-cell junctions should not be indiscriminately called JAIL, but that JAIL and JBL constitute different cellular activities acting in different tissue contexts. We have modified the text in the Discussion (lines 348-374).

(1) JAIL have originally been described in cell culture experiments (Abu-Taha et al., 2014). According to this and subsequent papers by the same group, local dissolution of endothelial adherens junctions (i.e. downregulation of VE-cadherin) triggers the formation of lamellipodia-like structures. These protrusions eventually retract, followed by the reestablishment of EC junctions.

(2) In our in vivo studies, we observed lamellipodial protrusions during endothelial cell rearrangements, and we call these structures JBL (Paatero et al., 2018). While JBL appear very similar to JAIL in general (i.e. regulation by Arp2/3 and its localization), we also observe two critical differences. For one, JBL form while maintaining the original (proximal) junction. Moreover, a distal junction is formed at the front edge of the JBL, leading to a “double junction” configuration. In our current manuscript, we have examined the role of actomyosin contractility and find that it correlates with and is required for the merging of proximal and distal junctions during JBL cycles. These observations indicates that the proximal and distal junctions are essential components of JBL function during endothelial cell elongation and rearrangements. These salient and distinct features prompted us to adopt the term junction-based-lamellipodia (JBL), in order to differentiate them from JAIL.

(3) We like to argue that JAIL and JBL represent similar but different lamellipodia-like protrusions. JAILs are associated with the maintenance of endothelial integrity, and control permeability and trans-endothelial cell migration, as has been suggested by several publications (Cao et al., 2017; Kipcke et al., 2025; Seebach et al., 2021; Taha et al., 2014). In contrast, JBL drive cell rearrangements, by step-wise elongation of cell junctions leading to convergent cell movements.

(4) Although JAIL have also been implicated in endothelial cell migration (Cao and Schnittler, 2019; Cao et al., 2017; Seebach et al., 2021), neither junctional patterns nor junctional dynamics have been analyzed in this context. We therefore propose that JAIL and JBL are actin-based protrusions forming at endothelial cell-cell junctions, but act in different contexts to provide cell motility (JBL) or endothelial integrity (JAIL).

(5) Some of the quantification plots, specifically in figures 5d and 6c, do not display significant differences or distribution patterns. It would be beneficial to revise these graphs to clearly represent statistical significance and underlying data distributions.

Because of the spatiotemporal heterogeneity, it is difficult to perform statistical quantifications across samples. In Figure 5c/d, we have imaged/analyzed myl9-EGFP in a mosaic situation, in which only one of interacting cells expresses high levels of myl9-EGFP. This is a rare situation and we managed to image only this example. Nevertheless, it is consistent with our other expression data of myl9-reporters and also with our previous photoconversion experiments using photoconvertible UCHD (Paatero et al., 2018, Figure 4), which shows that actin-rich JBL form at the front end of the endothelial cell in the direction of junction elongation. In Figure 5d, we have quantified the average intensity of GFP signal within the region of interest. The newly added error bars indicate the standard deviation between pixel intensities within the ROI.

In Figure 6c, we have analyzed the Myl9b-mCherry intensities and find that it is redistributed during a JBL cycle. The spatial distribution is evident from the heat-map and we have not included a standard deviation. Myl9b-mCherry levels are very heterogenous and is not possible to quantify intensities across samples. We have, however, included four more examples of Myl9b-mCherry distribution in Supplementary Figure 4. The patterns observed in these samples are consistent with those in Figure 6.

(6) The observation of myosin recruitment does not inherently imply a concomitant increase in actomyosin contractile activity. The inclusion of phospho-MLC staining would considerably strengthen the evidence for enhanced actomyosin activity.

This is a good suggestion and we have extensively tried different anti-P-Myl antibodies (and protocols), but did not get them to work reliably on zebrafish embryos. We therefore rely on published work that has established the correlation between the recruitment of myosin light chain and increased actomyosin tension (Fernandez-Gonzalez et al., 2009; Munjal et al., 2015).

Reviewer #2 (Recommendations for the authors):

(1) Figure 1a is not described/mentioned in the Results.

The have corrected this (lines 102-108). We have also added measurements to better present the different dynamics of F-actin (UCHD) and ZO1 within the JBL and the relative endothelial cell movements (see Figure 1b), as suggested by reviewer#1.

(2) In Figure 3a, the authors claim that Arp2/3 is deposited at the distal side of the junction ring. While it is clear where the proximal junction is (ZO1-rich), the distal junction is less so (hardly any ZO1). It is therefore difficult to agree based on this time-lapse imaging that Arpc1b-Venus is at the distal junction. Can the authors please include panels showing merged channels and annotate where the proximal and distal junctions are?

The activation of the Arp2/3 complex and the formation of the distal junction are sequential events. We see that ArpC1b oscillates with an accumulation at the onset and during JBL protrusion. In contrast, the distal junction is formed when the protrusive activity has been stopped. One caveat of the analysis shown in Figure 3a is that our ZO1 reporters label the distal junction only very weakly – this is in particular the case for the ZO1-tdTomato knock-in. The distal junction is better visible in VE-cadherin and UCHD reporters, as shown in Figures 5 to 7.

(3) In Figures 3b and c, it is also difficult to distinguish proximal and distal junctions in these images. Please mark the boundaries in the image panels (Figure 3b) and indicate on the x-axis where the proximal and distal junctions are (Figure 3c).

In Figure 3b, we show ArpC1b-Venus and mRuby-UCHD side-by-side. This Figure demonstrates that the Arp2/3 complex maintains its position at the front of the JBL during the protrusive phase (always distal to the UCHD signal). The imaging is done at very short intervals (1/30sec), which makes it difficult to follow entire oscillations due to photo-bleaching of the ArpC1b reporter.

(4) The treatment of CK666 resulted in perturbed localization of Arpc1b-Venus. Therefore, the inhibition of junctional elongation can also be explained by the mislocalization of Arp2/3, rather than the inhibition of Arp2/3 activity at the junctions. Can the authors discuss this or perform another experiment that is more specific to manipulating Arp2/3 activity?

CK666 is a well-established inhibitor of Arp2/3. Structural and functional analyses have shown that CK666 interferes with the interaction between Arp2 and Arp3, thereby preventing the activation of the complex (Hetrick et al., 2013; Padrick et al., 2011). We therefore conclude that the phenotypes we observe in CK666 treatment are due to Arp2/3 inhibition.

It is possible that CK666 prevents ArpC1b binding to the Arp2/3 complex. However, published work suggests that ArpC1b can bind to Arp2/3 also in its inactive state (Chou et al., 2022). Thus, we can only speculate why we lose localization ArpC1b under CK666. We prefer not to do so.

(5) In Figures 5d and 6c, is the quantification of Myl9 intensity of one cell only? If so, can the authors show the dynamics of average Myl9 intensity i) between forwarding and non-forwarding JBL poles and ii) as the proximal and distal junctions merge several endothelial cells?

Figure 5c/d depicts two interacting cells, expressing different levels of Myl9a-EGFP. This is a rare experimental situation and we managed to image only this example. We quantified the average signal at both poles of the junctional ring within a region of interest. The newly added error bars indicate the standard deviation between pixel intensities within the ROI. The analysis has been done on immunofluorescent images, therefore a dynamic analysis over time is not possible.

In Figure 6c, we have analyzed the Myl9b-mCherry intensities and find that it is redistributed during a JBL cycle. The spatial distribution is evident from the heat-map and we have not included a standard deviation. Myl9b-mCherry levels are very heterogenous and is not possible to quantify intensities across samples. We have, however, included four more examples of Myl9b-mCherry distribution in Supplementary Figure 4. The patterns observed in these samples are consistent with those in Figure 6.

(6) Figure 5. The 'f' in the figure legend should be 'e' since there is no panel 'f'.

We have corrected this.

(7) Figure 7. As the boundaries for proximal and distal junctions are not always clear, especially when Cdh5 appears as clusters, how do you determine where the two junctions are in order to measure the interjunctional space? Please offer a clearer explanation in the Methods.

We have added the following in the M&M. “Junctional merging tracking Speed of junctional merge was evaluated by monitoring isolated junctional rings during DLAV formation. Inhibitor treatment Y-27632 (75 μM) or DMSO (1%) were applied 30 min before mounting. The same concentrations of chemicals were applied to the low-melting-point agarose mounting medium and the E3 medium on top of it before imaging and imaging the junctions for 10-15 min on an Olympus SpinSR spinning disc microscope. Distances were measured using Fiji software. In each frame, the interjunctional distance was defined as the maximum distance between the proximal and distal junctions. A line was manually drawn between the proximal and distal junctions in Fiji, and its length was recorded. The same proximal and distal junction landmarks were used consistently across all time points.”

(8) One would think that upon the inhibition of junctional mergence (by ROCK inhibition), actin polymerization would persist to push the distal junction forward to elongate the JBL. However, there is instead a decrease in junctional elongation (Figure 7b). Can the authors speculate why? Additionally, junction elongation can probably be achieved by continuous "pushing" of the distal junction alone (through actin polymerization). Can the authors speculate why there is a need/what is the benefit of merging proximal and distal junctions for junction elongation?

These are all very interesting questions, but they are quite complex and would require extensive and speculative answers, which is outside the scope of this study. Nevertheless, here are a few quick thoughts on these issues.

(1) When endothelial cells elongate, they have to overcome tensile forces at the junctions (generated by the subjunctional actomyosin belt). JBL are providing a tractive and deforming force, which overcomes the tensile force and thus promotes junctional elongation.

(2) The distal junction is then providing an anchor to which the actin cytoskeleton can attach. The space between proximal and distal junction becomes a compartment of local actomyosin contraction, which provides the force for the ratchet to move the proximal junction forward  junctional mergence.

(3) Thus, it is not the protrusion (pushing) itself that elongates the cell but the elongation of the junction (driven by actomyosin contraction)!

(4) The maintenance of the proximal junction is most likely needed to ensure endothelial integrity during the JBL cycles.

(5) How the frequency and the size of JBLs is regulated is not known. One possible player that might be involved is an internal clock mechanism (e.g. a feedback loop via small GTPases (such as Rac)  Arp2/3 regulation). Another possibility is that JBL size is limited by it sweeping up basally localized VE-cadherin (in cis-configuration). Increasing cell-cell adhesion (by VE-cad trans-interactions between the JBL and the underlying cell) eventually stop the protrusion. It is also possible that an cell-autonomously controlled mechanism of F-actin polymerization (actin pulses) are involved in the regulation of the JBC cycle length.

(9) The animation showing the molecular mechanism of JBL function during endothelial junction elongation (Video 25) is very helpful in understanding the dynamic coupling between junctional proteins, actomyosin cytoskeleton, and junction remodelling. However, I wonder why there are no Myosin II proteins binding to the actin bundles during the merging of proximal and distal junctions (between 0:25 and 0:28), since this is one of the main findings by the authors in this study.

Since we show two JBL cycles, we want to spread the information over both of them.

References:

Cao, J. and Schnittler, H. (2019). Putting VE-cadherin into JAIL for junction remodeling. J. Cell Sci. 132.

Cao, J., Ehling, M., März, S., Seebach, J., Tarbashevich, K., Sixta, T., Pitulescu, M. E., Werner, A. C., Flach, B., Montanez, E., et al. (2017). Polarized actin and VE-cadherin dynamics regulate junctional remodelling and cell migration during sprouting angiogenesis. Nat. Commun. 8, 1–20.

Chou, S. Z., Chatterjee, M. and Pollard, T. D. (2022). Mechanism of actin filament branch formation by Arp2/3 complex revealed by a high-resolution cryo-EM structure of the branch junction. Proc. Natl. Acad. Sci. U. S. A. 119, e2206722119.

Fernandez-Gonzalez, R., Simoes, S. de M., Röper, J. C., Eaton, S. and Zallen, J. A. (2009). Myosin II Dynamics Are Regulated by Tension in Intercalating Cells. Dev. Cell 17, 736–743.

Hetrick, B., Han, M. S., Helgeson, L. A. and Nolen, B. J. (2013). Small molecules CK-666 and CK-869 inhibit actin-related protein 2/3 complex by blocking an activating conformational change. Chem. Biol. 20, 701–712.

Kipcke, J. P., Odenthal-Schnittler, M., Aldirawi, M., Franz, J., Bojovic, V., Seebach, J. and Schnittler, H. (2025). TNF-α induces VE-cadherin-dependent gap/JAIL cycling through an intermediate state essential for neutrophil transmigration. Front. Immunol. 16,.

Law, A. L., Jalal, S., Pallett, T., Mosis, F., Guni, A., Brayford, S., Yolland, L., Marcotti, S., Levitt, J. A., Poland, S. P., et al. (2021). Nance-Horan Syndrome-like 1 protein negatively regulates Scar/WAVE-Arp2/3 activity and inhibits lamellipodia stability and cell migration. Nature Communications 2021 12:1 12, 5687-.

Malchow, J., Eberlein, J., Li, W., Hogan, B. M., Okuda, K. S. and Helker, C. S. M. (2024). Neural progenitor-derived Apelin controls tip cell behavior and vascular patterning. Sci. Adv. 10, 1174.

Munjal, A., Philippe, J. M., Munro, E. and Lecuit, T. (2015). A self-organized biomechanical network drives shape changes during tissue morphogenesis. Nature 524, 351–355.

Paatero, I., Sauteur, L., Lee, M., Lagendijk, A. K., Heutschi, D., Wiesner, C., Guzmán, C., Bieli, D., Hogan, B. M., Affolter, M., et al. (2018). Junction-based lamellipodia drive endothelial cell rearrangements in vivo via a VE-cadherin-F-actin based oscillatory cell-cell interaction. Nat. Commun. 9,.

Padrick, S. B., Doolittle, L. K., Brautigam, C. A., King, D. S. and Rosen, M. K. (2011). Arp2/3 complex is bound and activated by two WASP proteins. Proc. Natl. Acad. Sci. U. S. A. 108, E472–E479.

Sauteur, L., Krudewig, A., Herwig, L., Ehrenfeuchter, N., Lenard, A., Affolter, M. and Belting, H. G. (2014). Cdh5/VE-cadherin promotes endothelial cell interface elongation via cortical actin polymerization during angiogenic sprouting. Cell Rep. 9, 504–513.

Seebach, J., Klusmeier, N. and Schnittler, H. (2021). Autoregulatory “Multitasking” at Endothelial Cell Junctions by Junction-Associated Intermittent Lamellipodia Controls Barrier Properties. Front. Physiol. 11,.

Taha, A. A., Taha, M., Seebach, J. and Schnittler, H. J. (2014). ARP2/3-mediated junction-associated lamellipodia control VE-cadherin-based cell junction dynamics and maintain monolayer integrity. Mol. Biol. Cell 25, 245–256.

eLife Assessment

This important study presents a technically rigorous and carefully controlled analysis of the signalling potential of cancer-associated gain-of-function Notch alleles. The work is clearly presented, and the experiments are robust, comprehensive, and well-controlled. While some data primarily establish the system or report negative findings, the comparative approach in a well-characterized model provides convincing mechanistic evidence for how these Notch variants function. This study will be of interest to researchers in both developmental and cancer biology.

Reviewer #1 (Public review):

Summary:

In their paper, Shimizu and Baron describe the signaling potential of cancer gain-of-function Notch alleles using the Drosophila Notch transfected in S2 cells. These cells do not express Notch or the ligand Dl or Dx, which are all transfected. With this simple cellular system, the authors have previously shown that it is possible to measure Notch signaling levels by using a reporter for the 3 main types of signaling outputs, basal signaling, ligand-induced signaling and ligand-independent signaling regulated by deltex. The authors proceed to test 22 cancer mutations for the above-mentioned 3 outputs. The mutation is considered a cluster in the negative regulatory region (NRR) that is composed of 3 LNR repeats wrapping around the HD domain. This arrangement shields the S2 cleavage site that starts the activation reaction.

The main findings are:

(1) Figure 1: the cell system can recapture ectopic activation of 3 existing Drosophila alleles validated in vivo.

(2) Figure 2: Some of the HD mutants do show ectopic activation that is not induced by Dl or Dx, arguing that these mutations fully expose the S2 site. Some of the HD mutants do not show ectopic activation in this system, a fact that is suggested to be related to retention in the secretory pathway.

(3) Figure 3: Some of the LNR mutants do show ectopic activation that is induced by Dl or Dx, arguing that these might partially expose the S2 site.

(4) Figure 4-6: 3 sites of the LNR3 on the surface that are involved in receptor heterodimerization, if mutated to A, are found to cause ectopic activation that is induced by Dl or Dx. This is not due to changes in their dimerization ability, and these mutants are found to be expressed at a higher level than WT, possibly due to decreased levels of protein degradation.

Strengths and Weaknesses:

The paper is very clearly written, and the experiments are robust, complete, and controlled. It is somewhat limited in scope, considering that Figure 1 and 5 could be supplementary data (setup of the system and negative data). However, the comparative approach and the controlled and well-known system allow the extraction of meaningful information in a field that has struggled to find specific anticancer approaches. In this sense, the authors contribute limited but highly valuable information.

Comments on revised version:

I reviewed the changes and response to criticism, and it seems to me that all has been reasonably addressed.

Reviewer #3 (Public review):

Summary:

This manuscript by Shimizu et al., systematically analyzes cancer-associated mutations in the Negative Regulatory Region (NRR) of Drosophila Notch to reveal diverse regulatory mechanisms with implications for cancer modelling and therapy development. The study introduces cancer-associated mutations equivalent to human NOTCH1 mutations, covering a broad spectrum across the LNR and HD domains. By linking mutant-specific mechanistic diversity to differential signaling properties, the work directly informs targeted approaches for modulating Notch activity in cancer cells. These are an exciting set of observations from S2 cells, which should be taken up further for further assessment in any physiological implications.

Strengths:

This manuscript by Shimizu et al., systematically analyzes cancer-associated mutations in the Negative Regulatory Region (NRR) of Drosophila Notch to reveal diverse regulatory mechanisms with implications for cancer modelling and therapy development. The study introduces cancer-associated mutations equivalent to human NOTCH1 mutations, covering a broad spectrum across the LNR and HD domains. The authors use rigorous phenotypic assays to classify their functional outcomes. By leveraging the S2 cell-based assay platform, the work identifies mechanistic differences between mutations that disrupt the LNR-HD interface, core HD, and LNR surface domains, enhancing understanding of Notch regulation. The discovery that certain HD and LNR-HD interface mutations (e.g., R1626Q and E1705P) in Drosophila mirror the constitutive activation and synergy with PEST deletion seen in mammalian T-ALL is nice and provides a platform for future cancer modelling. Surface-exposed LNR-C mutations were shown to increase Notch protein stability and decrease turnover, suggesting a previously unappreciated regulatory layer distinct from canonical cleavage-exposure mechanisms. By linking mutant-specific mechanistic diversity to differential signaling properties, the work directly informs targeted approaches for modulating Notch activity in cancer cells.

Weaknesses:

This is an exciting set of observations, however the work is entirely cell line based, and is the primary weakness. I list my main specific concerns herewith:

(1) The analysis is confined to Drosophila S2 cells, which may not fully recapitulate tissue or organism-level regulatory complexity observed in vivo.

(2) And perhaps for this reason too, some Drosophila HD domain mutants accumulate in the secretory pathway and do not phenocopy human T-ALL mutations. Possibly due to limitations on physiological inputs that S2 cells cannot account for or species-specific differences such as the absence of S1 cleavage. Thus, the findings may not translate directly to understanding Notch 1 function in mammalian cancer models.

(3) Also, while the manuscript highlights mechanistic variety, the functional significance of these mutations for hematopoietic malignancies or developmental contexts in live animals remains untested. Thus even though the changes are evident in Notch signaling, any impact on blood cells or hematopoiesis leading to aberrant malignancies remains to be seen.

(4) Which hematopoietic cell type, progenitor or differentiating cells, would be most sensitive to this kind of altered Notch signaling also remains unclear.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

In their paper, Shimizu and Baron describe the signaling potential of cancer gain-of-function Notch alleles using the Drosophila Notch transfected in S2 cells. These cells do not express Notch or the ligand Dl or Dx, which are all transfected. With this simple cellular system, the authors have previously shown that it is possible to measure Notch signaling levels by using a reporter for the 3 main types of signaling outputs, basal signaling, ligand-induced signaling and ligand-independent signaling regulated by deltex. The authors proceed to test 22 cancer mutations for the above-mentioned 3 outputs. The mutation is considered a cluster in the negative regulatory region (NRR) that is composed of 3 LNR repeats wrapping around the HD domain. This arrangement shields the S2 cleavage site that starts the activation reaction.

The main findings are:

(1) Figure 1: the cell system can recapture ectopic activation of 3 existing Drosophila alleles validated in vivo.

(2) Figure 2: Some of the HD mutants do show ectopic activation that is not induced by Dl or Dx, arguing that these mutations fully expose the S2 site. Some of the HD mutants do not show ectopic activation in this system, a fact that is suggested to be related to retention in the secretory pathway.

(3) Figure 3: Some of the LNR mutants do show ectopic activation that is induced by Dl or Dx, arguing that these might partially expose the S2 site.

(4) Figure 4-6: 3 sites of the LNR3 on the surface that are involved in receptor heterodimerization, if mutated to A, are found to cause ectopic activation that is induced by Dl or Dx. This is not due to changes in their dimerization ability, and these mutants are found to be expressed at a higher level than WT, possibly due to decreased levels of protein degradation.

Strengths and Weaknesses:

The paper is very clearly written, and the experiments are robust, complete, and controlled. It is somewhat limited in scope, considering that Figure 1 and 5 could be supplementary data (setup of the system and negative data). However, the comparative approach and the controlled and well-known system allow the extraction of meaningful information in a field that has struggled to find specific anticancer approaches. In this sense, the authors contribute limited but highly valuable information.

Reviewer #2 (Public review):

Summary:

This ambitious study introduced 22 mutations corresponding to amino acid substitution mutations known to induce cancer in human Notch1, located within the Negative Regulatory Region, into the Drosophila Notch gene. It comprehensively examined their effects on activity, intracellular transport, protein levels, and stability. The results revealed that the impact of amino acid substitutions within the Negative Regulatory Region can be grouped based on their location, differing between the Heterodimerization Domain and the Lin12/Notch Repeat. These findings provide important insights into elucidating the mechanisms by which amino acid substitution mutations in human Notch1 cause leukemia and cancer.

Strengths:

In this study, the authors successfully measured the activity of amino acid-substituted Notch with high precision by effectively leveraging the advantages of their previously established experimental system. Furthermore, they clearly demonstrated ligand-dependent and Deltex-dependent properties.

Weaknesses:

Amino acid substitution mutations exhibit interesting effects depending on their position, so interest naturally turns to the mechanisms generating these differences. Unfortunately, however, elucidating these mechanisms will require considerable time in the future. Therefore, it is reasonable to conclude that questions regarding the mechanism fall outside the scope of this paper.

We thank the editors and reviewers for their initial reviews and constructive suggestions. We have revised the manuscript with some additional data contained in two additional supplementary figures and by the inclusion of additional text.

Reviewer #3 (Public review):

While this is indeed an exciting set of observations, the work is entirely cell-line-based, and is the primary reason why this approach dampens the enthusiasm for the study. The analysis is confined to Drosophila S2 cells, which may not fully recapitulate tissue or organism-level regulatory complexity observed in vivo. Some Drosophila HD domain mutants accumulate in the secretory pathway and do not phenocopy human T-ALL mutations. Possibly due to limitations on physiological inputs that S2 cells cannot account for, or species-specific differences such as the absence of S1 cleavage.

Thus, the findings may not translate directly to understanding Notch 1 function in mammalian cancer models. While the manuscript highlights mechanistic variety, the functional significance of these mutations for hematopoietic malignancies or developmental contexts in live animals remains untested. Overall, the work does not yet provide evidence for altered Notch signaling that is physiologically relevant.

S2 cells are a standard cell culture model which have been extensively used for analysing Notch signalling mechanisms and by and large are found to recapitulate the mechanisms of Notch activation and its regulation in vivo. However, we agree that it will be desirable in future work to build on our current findings by generating Notch mutants in vivo in Drosophila as the in vivo context may introduce additional nuances in the behaviour of the mutants.This can be done by overexpressing cDNA constructs in particular tissues, or more physiologically by generating endogenous gene mutations using CRISPR/Cas9 based gene editing. However, the likely outcome of the latter approach is embryo lethality due to constitutive over-activation during development. Therefore, methods of genetic manipulation need to be applied which allow the final activating mutant form to be generated in somatic clones. We feel that this would be considerable amount of additional work and is out of scope for the current study, but we look forward to developing this approach in future work.

Recommendations for the authors:

Reviewing Editor Comments:

(a) Table 1: Explain the rationale for mapping non-conserved residues between human and fly Notch; consider adding an alignment or supplementary figure.

We have added a new Supplementary figure S2 showing an alignment of Notch sequences from different species to indicate the degree of conservation at the sites chosen for our mutagenesis study. Some locations were highly conserved and some locations less so. Both conserved and non-conserved residues were included to examine how structural perturbations at equivalent positions affect signalling activity, independent of sequence conservation. In addition to the new supplementary figure, we have changed the text in the Table 1 legend to clarify.

(b) Add or discuss data connecting LNR and HD mutant expression levels with stability and degradation mechanisms.

We have added additional text in the results section referring to Fig6A/B regarding the varying Notch protein levels between the different mutants. With regard to the slower degradation kinetics of certain LNR-C mutants in Fig6 E/F, we have also added a new supplementary figure S3 which shows that mutants from the LNR/HD interface do not behave similarly to the LNR-C mutants with respect to their degradation kinetics.

(c) Some mutants, especially those retained in the secretory pathway, are insufficiently characterized. The mechanism underlying their differential trafficking and stability remains underexplored.

We have added some extra text to the discussion section which explores the issue of secretory pathway retention of HD mutants in Drosophila cells further.

(2) Figure Legends:

(a) Figure 1A - Explain the ribbon vs. space-filling representation and color coding; include a definition of the Heterodimerization Domain.

We have added extra text to the Figure 1A legend

(b) Figure 2E - Clarify mutant selection; if possible, include additional examples for consistency.

We added extra text regarding selection of mutants for study into the legend of Figure 2

(c) Figure 3-4 - Explain logic for alanine substitutions; discuss difference at residue 1570 (P vs. A).

We added the following text to the result section. “Y1532 and Y1535 are not mutated in human cancers and therefore could not be assessed through patient-derived variants. Alanine substitution provides a controlled way to probe their contribution to NRR integrity and activation sensitivity by selectively removing their side-chain interactions while preserving overall fold.” We added extra text in the discussion section regarding the differences in the outcomes of the 1570 to A and P mutations.

(d) Figure 4 - Improve resolution and legibility.

We have replaced figure 4.

(e) Figure 6C - Correct residue numbering (1563, 1566).

Thank you for spotting this. This has been corrected.

(f) Figure 6F - Include control where protein levels do not increase.

A new supplementary figure S3 has been added which included this control data.

(3) Contextual and Conceptual Framing:

(a) Incorporate the limitations of the S2 system, and delineate which mechanistic insights are likely conserved versus those that may be species- or context-specific.

We have incorporated text to discuss S2 cell limitations.

(b) The study does not test functional consequences in hematopoietic or developmental contexts. Expand the discussion to emphasize how these cell-based findings could inform future in vivo studies or mammalian cancer modeling.

eLife Assessment

This manuscript offers valuable structural and mechanistic insights into the assembly of the Type II internal ribosome entry site (IRES) from encephalomyocarditis virus (EMCV) and the translation initiation complex, revealing a direct interaction between the IRES and the 40S ribosomal subunit. A solid experimental strategy, combining cryo-EM analysis, complementary biochemistry, and detailed structural comparisons, provides mechanistic insights into IRES-based translation initiation systems. This paper will attract researchers in cap-independent translation, host-pathogen interactions, and virology.

Reviewer #1 (Public review):

Summary:

The authors have studied how a virus (EMCV) uses its RNA (Type 2 IRES) to hijack the host's protein-making machinery. They use cryo-EM to extract structural information about the recruitment of viral Type 2 IRES to ribosomal pre-IC. The authors propose a novel interaction mechanism in which the EMCV Type 2 IRES mimics 28S rRNA and interacts with ribosomal proteins and initiator tRNA (tRNAi).

Strengths:

(1) Getting structural insights about the Type 2 IRES-based initiation is novel.

(2) The study allows a good comparison of other IRES-based initiation systems.

(3) The manuscript is well-written and clearly explains the background, methods, and results.

Comments on revised version:

I have gone through the revised manuscript by Das and Hussain along with the rebuttal comments. While the poor resolution of the ribosomal complex limits detailed analysis of the molecular interactions, addition of the luciferase reporter assay in the supplementary has enriched the paper.

Reviewer #2 (Public review):

Summary:

The field of protein translation has long sought the structure of a Type 2 Internal Ribosome Entry Site (IRES). In this work, Das and Hussain pair cryo-EM with algorithmic RNA structure prediction to present a structure of the Type 2 IRES found in Encephalomyocarditis virus (EMCV). Using medium to low resolution cryo-EM maps, they resolve the overall shape of a critical domain of this Type 2 IRES. They use algorithmic RNA prediction to model this domain onto their maps and attempt to explain previous results using this model.

Strengths:

(1) This study reveals a previously unknown/unseen binding modality used by IRESes: a direct interaction of the IRES with the initiator tRNA.

(2) Use of an IRES-associated factor to assemble and pull down an IRES bound to the small subunit of the ribosome from cellular extracts is innovative.

(3) Algorithmic modeling of RNA structure to complement medium to low resolution cryo-EM maps, as employed here, can be implemented for other RNA structures.

Comments on revised version:

Thanks to the authors for providing thorough responses to the reviewer questions and comments. I appreciate their attempts of improving overall resolution of the complex via various processing strategies that the reviewers suggested.

The authors interpretations of their cryo-EM data match those reported by Bhattacharjee et al. 2025 (EMCV-IRES 48S) and can be contextualized in the light of Velazquez et al. 2025 (poliovirus IRES-48S).

The authors' contextualization of their results with previously published studies (Discussion section lines 355-402) is satisfactory to me but can be improved.

Reviewer #3 (Public review):

Summary:

Type II IRES, such as those from encephalomyocarditis virus (EMCV) and foot-and-mouth disease virus (FMDV), mediate cap-independent translation initiation by using the full complement of eukaryotic initiation factors (eIFs), except the cap-binding protein eIF4E. The molecular details of how IRES type II interacts with the ribosome and initiation factors to promote recruitment have remained unclear. Das and Hussain used cryo-electron microscopy to determine the structure of a translation initiation complex assembled on the EMCV IRES. The structure reveals a direct interaction between the IRES and the 40S ribosomal subunit, offering mechanistic insight into how type II IRES elements recruit the ribosome.

Strengths:

The structure reveals a direct interaction between the IRES and the 40S ribosomal subunit, offering mechanistic insight into how type II IRES elements recruit the ribosome.

Comments on revised version:

The revised manuscript does not improve the resolution; however, the authors provide a detailed and well-reasoned rationale that directly addresses the concerns I raised about their structural interpretation. In addition, two independent preprints have been released since the initial submission. In one case, the authors report a higher-resolution, and importantly, all three studies present consistent assignments and interpretations. Together, these observations strengthen confidence in the authors' conclusions. I therefore do not have major concerns regarding the publication of this revised manuscript.

Author response:

The following is the authors’ response to the original reviews.

eLife Assessment

This manuscript offers valuable structural and mechanistic insights into the structure and assembly of the Type II internal ribosome entry site (IRES) from encephalomyocarditis virus (EMCV) and the translation initiation complex, revealing a direct interaction between the IRES and the 40S ribosomal subunit. While a solid cryo-EM method was used, enhancing the overall resolution or adding complementary biochemical data would further improve the clarity and impact of this study. This manuscript will attract researchers in cap-independent translation, host-pathogen interactions, and virology.

We thank the editorial team for a favourable assessment and for mentioning our work as ‘valuable’. In the following sections, we have addressed the weaknesses and recommendations pointed out by the Reviewers and hope for an improvement in the description of this work.

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors have studied how a virus (EMCV) uses its RNA (Type 2 IRES) to hijack the host's protein-making machinery. They use cryo-EM to extract structural information about the recruitment of viral Type 2 IRES to ribosomal pre-IC. The authors propose a novel interaction mechanism in which the EMCV Type 2 IRES mimics 28S rRNA and interacts with ribosomal proteins and initiator tRNA (tRNAi).

Strengths:

(1) Getting structural insights about the Type 2 IRES-based initiation is novel.

(2) The study allows a good comparison of other IRES-based initiation systems.

(3) The manuscript is well-written and clearly explains the background, methods, and results.

We thank Reviewer 1 for appreciating our efforts and finding structural insights about the Type 2 IRES-based initiation presented in this study as novel.

Weaknesses:

(1) The main weakness of the work is the low resolution of the structure. This limits the possibility of data interpretation at the molecular level.

However, despite the moderate resolution of the cryo-EM reconstructions, the model fits well into the density. The analysis of the EMCV IRES-48S PIC structure is thorough and includes meaningful comparisons to previously published structures (e.g., PDB IDs - 7QP6 and 7QP7). These comparisons showed that Map B1 represents a closed conformation, in contrast to Map A in the open state (Figure 2). Additionally, the proposed 28S rRNA mimicry strategy supported by structural superposition with the 80S ribosome and sequence similarity between the I domain of the IRES and the h38 region of 28S rRNA (Fig. 4) is well-justified.

We agree that the low resolution of the map has compromised the data interpretation at the molecular level, and we thank the reviewer for appreciating our findings at this resolution. Due to the low resolution, we have reported findings for stretches or regions such as the domain I loops and stems, rather than individual nucleotides.

(2) The lack of experimental validation of the functional importance of regions like the GNRA and RAAA loops is another limitation of this study.

We agree about the lack of additional experiments other than Cryo-EM for probing the importance of regions such as GNRA and RAAA loops in this study. Previously, multiple studies have reported on the importance of GNRA and RAAA loops and we have cited them in the manuscript. The essentiality of RAAA loop for type 2 IRES was demonstrated in earlier report López de Quinto and Martínez-Salas, 1997 (Cited in manuscript). Further, the conservation of this loop across the type 2 IRES family adds to the importance of this loop (Manuscript Figure 6B). This loop and its flanking G-C stem are similar to h38 of 28S rRNA, and it appears that RAAA loop adopts a mimicry mechanism to interact with the 40S ribosomal protein- uS19, thus highlighting its importance for interaction with 40S. Experiments destabilising the G-C stem also compromise IRES activity, as shown for the case of FMDV IRES (Fernández et al 2011). Previous studies related to the mutation of the GNRA or GCGA loop in EMCV IRES have shown a deficiency in IRES activity (Roberts and Belsham, 1997; Robertson et al 1999), suggesting the importance of these regions in the viral IRES biology, and these reports are cited in the manuscript. Not only EMCV IRES, but mutation in the GUAA (representative of GNRA) loop of FMDV IRES also showed a significant reduction in IRES activity (López de Quinto and Martínez-Salas, 1997). In this work, we observe that the GCGA loop interacts with tRNA_i in the EMCV IRES-48S PIC, thus implicating the importance of this loop. Moreover, incubation of FMDV IRES with 40S ribosomes has shown a decrease in SHAPE reactivity in domain 3 apex (position 170- 200 nucleotides) (Lozano et al 2018), which corresponds to EMCV IRES domain I apex.

However, to address this concern in the revised manuscript we mutated these loops and performed luciferase assay (Supplementary figure 4 A). The results showed decreased IRES activity (Pg 10) and correlated with previous reports demonstrating the importance of these regions for overall IRES activity.

(3) Minor modifications related to data processing and biochemical studies will further validate and strengthen the findings.

(a) In the cryo-EM data section, the authors should include an image showing rejected particles during 2D classification. This would help readers understand why, despite having over 22k micrographs with sufficient particle distribution and good contrast, only a smaller number of particles were used in the final reconstruction. Additionally, employing map-sharpening tools such as Ewald sphere correction, Bayesian polishing, or reference-based motion correction might further improve the quality of the maps. Targeting high-resolution structures would be particularly informative.

We have included the image for rejected 2D classes (Author response image 1). We agree with the Reviewer’s query related to the huge number of micrographs and relatively smaller number of particles for the final reconstruction. Since the total number of micrographs (22000) is the summation of multiple datasets, prepared and collected at different times, the distribution of the particles per micrograph was not uniform in all sessions, ranging from good to poor. Among these, around 8000 micrographs have poor particle number and distribution. As a result, the number of particles per micrograph is heterogeneous across the compiled dataset, and only 237054 ribosomal particles were obtained after multiple rounds of 2D and 3D classification. Further, the final reconstruction was performed using particles obtained after masked classification for IRES and ternary complex density. Only the particles that show the best density for both IRES and ternary complex are used for this map. Another set of particles that have only a portion of IRES and NO density for ternary complex forms another map. And we have a third map with an empty 40S.

We thank the reviewer for the suggestions to improve the quality of the maps further. As suggested, we started with the processing of the data. However, during this process the common computational cluster that were using for this data processing had to be physically relocated, and unfortunately after the relocation we faced technical issues in accessing and continuing with the processing. Several attempts to resolve the issue with the help of IT team failed. Thus, we lost 3-4 months without any progress. Therefore, we used Relion on our in-house workstation to process the data files from the start, as our in-house computational resources are unequipped to run cryoSPARC processes (for large dataset due to memory limitations).

We reprocessed the datasets in Relion5 and did ‘Bayesian Processing’, for reference-based beam-induced motion correction per-particle. Post-processing, we used cryoSPARC to merge the particles and tried classifying the good ribosome particles using focus-based masked classification, as shown in Supplementary Figure 1.1. However, this processing did not improve the resolution, as Map B (containing 40S, tRNA, IRES) had an overall resolution of 4.8 Å (Author response image 2). Therefore, we would like to report the same maps as given in the initial submission.

We estimated the time to redo the entire processing using cryoSPARC on the common computational cluster, and it would take us another 3-4 months or more and we do not anticipate a massive improvement in the extra density.

Author response image 1.

The selected 2D classes and the rejected 2D classes from initial round of classification, and the final selected 2D classes, which were subjected to Ab-initio reconstruction to get the good ribosome particles.

Author response image 2.

Reprocessing of the entire dataset using Relion5 for polishing of selected particles, followed by 3D classification and refinements in cryoSPARC.

(b) The strategic modelling of different IRES domains into the density, particularly the domain into the region above the 40S head, is appreciable. However, providing the full RNA tertiary structure (RNAfold) of the EMCV IRES (nucleotides 280-905) would better explain the logic behind the model building and its molecular interpretation.

We thank the reviewer for appreciating the modelling of the domain I apex in the cryo-EM density. We tried to predict the full tertiary structure of the IRES using Alphafold3; however, inclusion of the full-length sequence from 280-905 gave models of extremely low confidence (Author response image 3), and a few domains do not abide by the secondary structure of EMCV IRES as reported in Duke et al 1992.

Author response image 3.

Prediction of tertiary structure of EMCV IRES (280-905 nucleotides) and zoomed features for each domain present in the IRES. The predicted aligned error plot for the RNA structure is shown.

We used individual domains of EMCV IRES and predicted the tertiary structure, independent of other IRES domain using Alphafold3. As a result, the confidence scores improved, and the tertiary structures also correlated with the experimentally determined EMCV IRES secondary structure (Duke et al 1992; Maloney and Joseph, 2024). Although the overall tertiary structure of EMCV IRES is lacking, recent studies were able to solve the structures of EMCV IRES domains in complex with their respective binding proteins. We superimposed the independently predicted domains D, E, and F tertiary structure on the NMR ensemble of IRES domain D to F with PTB1 (Dorn et al 2023), where the predicted domains fit in the experimental model. Similarly, we used the cryo-EM structure of domain J-K-eIF4G-eIF4A (Imai et al 2023) and found a close fit with the predicted structures. The analysis highlighted that the domain I apex serves as the best fit with the extra density with respect to architecture and fitting. This analysis is now added in the revised manuscript in Supplementary figure- 3.2.

Furthermore, 3D structural models of FMDV IRES domains 2, 3, and 4 (corresponding to EMCV IRES domains- H, I, and J-K) were predicted from SHAPE reactivity values and RNAComposer server (Figure 3, Lozano et al 2018). The predicted architecture of domain 3 apex (FMDV IRES) coincides with our domain I apex model (EMCV IRES).

(c) Although the authors compare their findings with other types of IRESs (Types 1, 3, and 4), there is no experimental validation of the functional importance of regions like the GNRA and RAAA loops. Including luciferase-based assays or mutational studies of these regions for validation of structural interpretations is strongly recommended.

We have discussed the possibility of how the other IRESs, such as type 1 and type 5, might use similar strategies as EMCV IRES to assemble the 48S PIC, given the similarity in the motif sequence and position across the viral IRESs. Like EMCV IRES, the type 1 IRES (Poliovirus, Coxsackie virus, etc.) also harbours the GNRA loop, preceded by a C-rich loop at its longest domain, known for long-range RNA-RNA interactions. The segment harbouring GNRA loop is highly conserved across the type 1 family of IRESs (Kim et al 2015). The Aichi viral IRES harbours a GNRA loop in its longest domain, that is, domain J. Deletion of the GNRA loop has compromised the IRES activity; however, substitution mutations in this region have elevated the IRES activity or remained unaltered (Yu et al 2011). We have hypothesized that these IRESs might use the GNRA motifs in their longest domain (domain IV in type 1, and domain J in Aichi virus- type 5) based on the location and architecture to that of EMCV IRES, where GNRA is present in the longest domain (I) and preceded by a C-rich loop where it can potentially mediate long-range interactions with tRNA_i, as all these IRESs require eIF2-ternary complex for the formation of 48S PIC. Parallelly, like EMCV IRES, type 1 and type 5 IRESs have the placement of this GNRA motif-containing domain before the eIF4G-binding domain. Thus, we suggest the possibility of adoption of a similar strategy by these IRESs to interact with tRNA_i during the formation of 48S PIC. During the revision of this work a preprint reported the structure of polioviral IRES-48S PIC where domain IV apex (similar to domain I apex in EMCV IRES) interacts with uS13 and uS19, and the GNRA loop directly interacts with tRNA_i during start codon recognition (Velazquez et al 2025). We hypothesize that Aichiviral IRES might use this motif to mediate long-range interactions with tRNA_i, similar to type 1 and type 2 IRESs, as all these IRESs require eIF2-ternary complex for the formation of 48S PIC.

Reviewer #2 (Public review):

Summary:

The field of protein translation has long sought the structure of a Type 2 Internal Ribosome Entry Site (IRES). In this work, Das and Hussain pair cryo-EM with algorithmic RNA structure prediction to present a structure of the Type 2 IRES found in Encephalomyocarditis virus (EMCV). Using medium to low resolution cryo-EM maps, they resolve the overall shape of a critical domain of this Type 2 IRES. They use algorithmic RNA prediction to model this domain onto their maps and attempt to explain previous results using this model.

Strengths:

(1) This study reveals a previously unknown/unseen binding modality used by IRESes: a direct interaction of the IRES with the initiator tRNA.

(2) Use of an IRES-associated factor to assemble and pull down an IRES bound to the small subunit of the ribosome from cellular extracts is innovative.

(3) Algorithmic modeling of RNA structure to complement medium to low resolution cryo-EM maps, as employed here, can be implemented for other RNA structures.

We thank Reviewer 2 for positive and encouraging comments on our work, appreciating our ‘innovative’ approach of using IRES-associated factor to assemble and pull down the IRES-bound ribosomal complex.

Weaknesses:

(1) Maps at the resolution presented prevent unambiguous modelling of the EMCV-IRES. This, combined with the lack of any biochemical data, calls into question any inferences made at the level of individual nucleotides, such as the GNRA loop and CAAA loop (Figure 4).

We understand the concerns raised by the reviewer related to the resolution of the EMCV IRES-48S PIC map. We refrained from commenting on individual nucleotides or molecular interactions in the manuscript. Instead, we discuss loops, RNA stretches or motifs that could be inferred with more confidence in the IRES density as shown in Figure 4. The EMCV IRES can directly interact with the 40S ribosome using its domain H and I (Chamond et al 2014), however, the details of this interaction were unknown. We observe that the CAAA loop of domain I apex interacts with 40S ribosome based on the placement of a portion of domain I in the cryo-EM map. This is also reflected in the SHAPE data (Chamond et al 2014-Supplementary figures 2, and 8), where a decrease in reactivity is evident in the presence of 40S ribosome. In addition, incubation of EMCV IRES with rabbit reticulocyte lysate (RRL) offered protection to domain I apex regions, which included the CAAA loop (Maloney and Joseph, 2024- Figure 4b).

Furthermore, this decrease in SHAPE reactivity pattern is evident for FMDV IRES domain 3 apex (similar to domain I in EMCV IRES) in the presence of 40S ribosome (Lozano et al 2018). Thus, these studies are consistent with the placement of IRES model in the cryo-EM map. Moreover, we performed structural analysis (mentioned above) which showed that the domain I apex serves as the best fit with the extra density with respect to architecture and fitting (Supplementary figure- 3.2).

(2) The EMCV IRES contains an upstream AUG at position 826, where the PIC can assemble (Pestova et al 1996; PMID 8943341). It is unclear if this start codon was mutated in this study. If it were not mutated, placement of AUG-834 over AUG-826 in the P-site is unexplained.

We thank the reviewer for bringing up this point, as we missed mentioning this in the initial submission. The EMCV IRES does not require scanning and directly positions the AUG-834 at the P site (Pestova et al 1996). In Pestova et al 1996, the intensity of the toeprint at AUG-834 is more intense than that of AUG-826. Further, AUG-834 lies in the Kozak context, whereas AUG-826 has a poor Kozak context, and AUG-826 codon is not in-frame with AUG-834. Therefore, the synthesis of the polypeptide requires AUG-834 at the P site. In our cryo-EM map, we observed that the tRNA_i is in a P_IN state, which indicates the recognition of the start codon, and we reasoned that it is more likely that AUG-834 is placed at the P site than AUG-826. We have mentioned this in the revised manuscript as we had NOT mutated AUG-826 (Pg 8).

(3) The claims the authors make about (i) the general overall shape and binding site of the IRES, (ii) its gross interaction with the two ribosomal proteins, (iii) the P-in state of the 48S, (iv) the rearrangement of the ternary complex are all warranted. Their claims about individual nucleotides or smaller stretches of the IRES-without any supporting biochemical data-is not warranted by the data.

We thank the reviewer for warranting major claims, and due to the low-resolution we have reported findings for stretches or regions such as the domain I loops and stems, rather than individual nucleotides. The interaction of domain I apical region with uS13, uS19, and tRNA_i is also observed the high-resolution structure of reconstituted EMCV IRES-48S PIC that was reported in a preprint while our work was under peer review process (Bhattacharjee et al 2025). Thus, the reconstituted EMCV IRES-48S PIC (Bhattacharjee et al 2025) also supports our assignment of domain I and its conserved loops, interacting with ribosome and tRNA_i.

Reviewer #3 (Public review):

Summary:

Type II IRES, such as those from encephalomyocarditis virus (EMCV) and foot-and-mouth disease virus (FMDV), mediate cap-independent translation initiation by using the full complement of eukaryotic initiation factors (eIFs), except the cap-binding protein eIF4E. The molecular details of how IRES type II interacts with the ribosome and initiation factors to promote recruitment have remained unclear. Das and Hussain used cryo-electron microscopy to determine the structure of a translation initiation complex assembled on the EMCV IRES. The structure reveals a direct interaction between the IRES and the 40S ribosomal subunit, offering mechanistic insight into how type II IRES elements recruit the ribosome.

Strengths:

The structure reveals a direct interaction between the IRES and the 40S ribosomal subunit, offering mechanistic insight into how type II IRES elements recruit the ribosome.

Weaknesses:

While this reviewer acknowledges the technical challenges inherent in determining the structure of such a highly flexible complex, the overall resolution remains insufficient to fully support the authors' conclusions, particularly given that cryo-EM is the sole experimental approach presented in the manuscript.

The study is biologically significant; however, the authors should improve the resolution or include complementary biochemical validation.

We thank Reviewer 3 for acknowledging the technical challenges in this study and finding our study biologically significant. We understand the concerns related to low resolution and the requirement of complementary biochemical validation for our reported observations and interpretations in the manuscript. We tried to improve the resolution, but the improvement was not sufficient to resolve the IRES at the nucleotide level. Independently, another group has reported the same findings at a higher resolution while our work was under peer review process (Bhattacharjee et al 2025), which corroborates our structural data on EMCV IRES and its interaction with ribosome and tRNA_i in its 48S PIC stage. Further, in the revised manuscript we also present biochemical validation for GNRA and RAAA loops in EMCV IRES. We mutated these loops and performed luciferase assay (Supplementary figure 4 A). The results showed decreased IRES activity (Pg 10) and correlated with previous reports (Roberts and Belsham, 1997; López de Quinto and Martínez-Salas, 1997; Robertson et al 1999) demonstrating the importance of these regions for overall IRES activity.

Reviewing Editor Comments:

The reviewers' comments are appended. While the reviewers acknowledge the complexity associated with this system, they also raised concerns about the modeling of RNA and registering its sequence in low-resolution maps. We believe that the strength of evidence and overall impact of your study can be elevated by providing higher-resolution cryo-EM data or complementary biochemical studies and addressing reviewers' concerns.

Reviewer #2 (Recommendations for the authors):

(1) Science:

Have the authors tried a focused refinement (local refinement in cryoSPARC) using a generous mask that encloses the head and the IRES but excludes the ternary complex and the body of the 40S? This can be done with all the particles in map B (~55K) and has the possibility of improving the resolution of domain I which can be subsequently used to build a better model of the IRES. See the middle right panel, light yellow colored mask in Figure 1A in PMID 37659578 for the type of mask being suggested.

We did another round of 2D classification to eliminate any residual junk in the ~55k particle set, corresponding to Map B. Post classification, 49439 particles were selected and refined using non-uniform refinement to get Map B11. The overall resolution of Map B11 was 4.6 Å. Thereafter, we made a mask around the 40S head-IRES-tRNA on Map B11 and subjected the class for local refinement. The overall local resolution in the masked region improved to 4.5 Å (Author response image 4).

Author response image 4.

Data processing- Map B particles were 2D classified, and further junk was cleared as rejected particles. The selected particles were refined using non-uniform refinement to get Map B11, and later, a focused mask circling the head-tRNA-IRES region was used for local refinement in the region to yield map B111.

We estimated the local resolution across the focused region in Map B111 and compared this with that of Map B (Author response image 5). The local refinement shows minor improvement in the local resolution in this region, and is not sufficient to resolve the IRES density at the level of nucleotides.

Author response image 5.

Comparison of local resolution across head-IRES-tRNA in map B1 (as reported in the manuscript) and Map B111.

(2) Presentation:

(a) Please use the previously established convention of naming the domains: "domain I", "domain H", etc, instead of "I domain" or "J-K domain" while describing parts of the IRES.

We have made the changes as per the established convention.

(b) Figure 2B reports a 6.9 A distance vs. 7 A in the text. Please use ~ or approximately to keep numbers consistent.

We have used ~ symbol to suggest the approximate distance.

(c) References missing on page 15 when referring to "previously determined HCV and CrPV structures".

We have added the references (Pg 12).

(d) Please edit the text for typos and sentence structure.

The typos and sentence structure were corrected wherever necessary.

(e) Some phrases and sentences (e.g. last few sentences of the first paragraph in the discussion) could be rewritten for clarity.

Previous sentence- “The domain I of EMCV IRES is similar to domain IV of polioviral IRES (or other type 1 IRESs such as Coxsackie viral IRES) in terms of length, secondary structure, and conserved motifs (GNRA, C-rich) positioning (Fig. 6C), therefore, anticipating a similar interaction with tRNA_i, highlighting a sequestering tendency by competing with cellular mRNAs.”

Rephrased sentence- “Like EMCV IRES, the type 1 IRES (Poliovirus, Coxsackie virus, etc.) also harbours the GNRA loop, preceded by a C-rich loop at its longest domain, known for long-range RNA-RNA interactions. The segment harbouring GNRA loop is highly conserved across the type 1 family of IRESs (Kim et al 2015). The domain I of EMCV IRES is similar to domain IV of polioviral IRES or other type 1 IRESs in terms of length, secondary structure, and conserved motifs (GNRA, C-rich) positioning (Fig. 6C). Therefore, we anticipate a similar interaction of domain IV (in type 1 IRES class) with tRNA_i. Also, this interaction of IRES with tRNA_i could be a strategy by which these IRESs can sequester the tRNA_i pool in the cell, rendering them unavailable for capped cellular mRNAs.”

Reviewer #3 (Recommendations for the authors):

(1) For the revision process, the authors provided three atomic models alongside their corresponding cryo-EM density maps, including a 48S complex in closed conformation. Given this conformation, it is reasonable to interpret the structure as representing a post-start codon recognition state (late-stage initiation). However, this reviewer finds that the local resolution within the mRNA channel is insufficient to support the atomic model building as presented. The density does not allow for an unambiguous assignment of nucleotides in this region; the authors should either improve the local resolution or remove the modeled mRNA from the structure.

We understand the concern of the Reviewer. Although the mRNA density in the channel is poor, we modelled the mRNA with AUG-834 at the P site because the known biology of EMCV IRES. The EMCV IRES does not require scanning and directly positions the AUG-834 at the P site (Pestova et al 1996). In Pestova et al 1996, the intensity of the toeprint at AUG-834 is more intense than that of AUG-826. Further, AUG-834 lies in the Kozak context, whereas AUG-826 has a poor Kozak context, and AUG-826 codon is not in-frame with AUG-834. Therefore, the synthesis of the polypeptide requires AUG-834 at the P site. In our cryo-EM map, we observed that the tRNA_i is in a P_IN state, which indicates the recognition of the start codon, and we reasoned that it is very likely that AUG-834 is placed at the P site.

(2) As noted by the authors, the start codon in the EMCV IRES is positioned within a strong Kozak sequence. The nucleotide at position -3 is known to interact with eIF2α, yet, in the current model, A831 is positioned such that physical contact with eIF2α would be structurally impossible. This discrepancy raises concerns about the accuracy of the modeled eIF2α, which, like other regions of the structure, is not clearly supported by the cryo-EM density. The authors should revise the atomic model of eIF2α to ensure it is consistent with the experimental map and established molecular interactions.

In our analysis of EMCV IRES-48S PIC, we could observe eIF2α and eIF2γ in Map B and B1. However, the local resolution was low to model the entire protein with side-chains (Supplementary figure 1.2 A). So, we used rigid body fitting of eIF2α and eIF2γ (Author response image 6). From the model, we could trace the backbone of Arg55, however could not resolve the side chain. Similarly, the mRNA in the channel was modelled based on placement of AUG-834 at the P site for EMCV IRES, which enabled us to model the flanking residues, rather than at the nucleotide-level resolution. We anticipate that a higher resolution structure will be able to capture this interaction of eIF2α with mRNA nucleotide (-3), therefore refrained from commenting on this interaction in the manuscript. In the revised manuscript, we have removed the side chains of eIF2α and eIF2γ, and kept the Cα-backbone only. The map-model statistics of map B1 is updated in table 1.

Author response image 6.

(left) Fitting of eIF2α model in the map. (right) Fitting of Cα backbone of eIF2α and mRNA in the map.

(3) The authors observed additional density interacting with ribosomal proteins uS19 and uS13, and tRNA, which they tentatively assign to domain I of the IRES. Although the local resolution in this region does not allow an unambiguous assignment, the interpretation is reasonable. However, further structural and functional validation is necessary to support this assignment. The authors should improve the local resolution, either by performing focused refinement or by increasing the number of particles used in the reconstruction.

The assignment of the extra density to domain I of the IRES was based on the architecture of the density. This density allows no other IRES domain to fit in this region (Supplementary figure 3.2). We tried to improve the local resolution using focused refinement, but the resolution was insufficient to resolve the IRES at the nucleotide level. Please see the above-mentioned comments in this regard on Pg 12.

(4) Figure 5 shows a slight shift in the position of the ternary complex. Is the observed tRNA conformation compatible with the structural rearrangements required for 60S subunit joining?

During the transition of 48S PIC to 80S elongation-competent complex, there are major changes in the conformation of tRNA_i, due to the joining of eIF5B, and release of eIF2 (Petrychenko et al 2024). This joining event of eIF5B positions the tRNA_i elbow and acceptor stem towards the 40S body to aid 60S ribosomal subunit joining (Petrychenko et al 2024). However, in the context of EMCV IRES-48S PIC, we observed that the position of tRNA_i elbow and acceptor stem is towards the 40S head, and away from the body. On superimposing the human 48S PIC structure (before 60S joining), 48S-5 (PDB Id- 8PJ5- Petrychenko et al 2024), we note that tRNA_i in EMCV IRES-48S PIC is away from the canonical tRNA_i position (in contact with eIF5B). Therefore, we anticipate a change in tRNA_i conformation during eIF5B joining and eIF2 release. This hypothesis coincides with the fact that the IRES interacting with the tRNA_i elbow needs to be displaced from the position to facilitate the interaction of tRNA_i with eIF5B. Moreover, this rearrangement would also aid in 60S joining and prevent any clash with the IRES domain I. We have added this in Results selection 5 and Figure 5D.

(5) In the discussion section, the authors state: "eIF3-eIF4G interaction is dispensable for EMCV IRES-48S PIC formation, so we do not rule out the possibility that EMCV IRES may dislodge eIF3 from its position on the solvent surface as observed in the case of HCV IRES (Hashem et al, 2013)." This statement is highly speculative. Is there any experimental or structural evidence to support this proposed mechanism in the context of EMCV IRES?

Previous biochemical reports on the eIF3-eIF4G interaction suggested that eIF4G residues from 1011-1104 interact with eIF3 (Villa et al 2013). In the context of EMCV IRES, this region of eIF4G is not required to form 48S PIC on the IRES, suggesting the eIF3-eIF4G interaction is dispensable for EMCV IRES-48S PIC formation. However, the recent structure of the human canonical 48S PIC has shown that the eIF4G-HEAT1 domain can interact with eIF3 subunits c, h, and l, and that eIF4G-bound eIF4A can interact with 40S ribosomal protein eS7, thus mediating the interaction between eIF4-bound mRNA and the 43S PIC (Brito Querido et al 2024) but the known eIF3-binding region in eIF4G was not captured in the map. Although the canonical eIF3-eIF4G interaction is essential in the case of cap-dependent initiation, this interaction could be dispensable for 48S PIC formation on EMCV IRES. In case of HCV IRES-mediated initiation, eIF3 is displaced from its canonical position that facilitates the binding of HCV IRES to 40S ribosomal subunit (Hashem et al 2013). We did not see any density corresponding to eIF3 in the obtained maps. Further, we have used focused classification using a mask on the canonical eIF3 position; however, we do not see any density corresponding to eIF3 in the EMCV IRES-48S PIC complex. Therefore, we hypothesized the possibility that eIF3 might be dislodged from its canonical binding site on the 40S ribosomal subunit. However, as per the recent independent report on EMCV IRES-48S PIC, eIF3 is present in the complex (Bhattarcharjee et al 2025).

Hence, we have rephrased the existing sentence- “However, eIF3-eIF4G interaction is dispensable for EMCV IRES-48S PIC formation, so we do not rule out the possibility that EMCV IRES may dislodge eIF3 from its position on the solvent surface as observed in case of HCV IRES (Hashem et al 2013).”

Rephrased sentence- “However, the canonical eIF3-eIF4G interaction (Villa et al 2013) is dispensable for EMCV IRES-48S PIC formation (Lomakin et al 2000; Sweeney et al 2014), and we do not see any density for eIF3 even after focused classification. However, as per the recent independent report on reconstituted EMCV IRES-48S PIC, eIF3 is present in the complex at the canonical position (Bhattarcharjee et al 2025). This position of eIF3 further highlights the possibility that eIF4G-eIF4A proteins are also placed similarly to the canonical eIF3-eIF4G-eIF4A position (Brito Querido et al 2024) in context to EMCV IRES-48S PIC. Thus, placing eIF4G-domain J-K close to ES6 of 40S ribosome, which coincides with the previous hydroxyl radical cleavage assay (Yu et al 2011).”

(6) eIF4A has been shown to directly interact with eIF3 and facilitate recruitment of the 43S PIC. Does the interaction of the J-K domain with eIF4G/eIF4A, compatible with the known eIF4A-eIF3 interaction within the 43S PIC? In other words, during EMCV IRES-mediated initiation, could the eIF4A-eIF3 interaction functionally substitute for the eIF4G-eIF3 interaction?

Reports on EMCV IRES-mediated translation initiation have shown eIF4G as an essential component of 48S PIC formation (Pestova et al 1996; Lomakin et al 2000; Kolupaeva et al 2003; Sweeney et al 2014), where eIF4G directly interacts with domain J-K of IRES and eIF4A, thus enabling loading of eIF4A on the IRES. In our study, the cryo-EM map of EMCV IRES-48S PIC lacks density for eIF3 and eIF4 proteins, and locating eIF4F is challenging due to the inherent flexibility associated with the complex. Previous studies on EMCV IRES-48S PIC have mapped the location of eIF4G close to ES6 towards the platform side of the body and eIF3 using the hydroxyl radical cleavage assay (Yu et al 2011). The human 48S initiation complex structures have shown a similar location for eIF4G, which is at the mRNA exit site, contacting eIF3 (Brito Querido et al 2020; Brito Querido et al 2024). On overlapping the 18S rRNA of EMCV IRES-48S PIC to that of the human 48S PIC in closed conformation (PDB Id- 8OZ0), and further superimposing the J-K-St- eIF4G- eIF4A (PDB Id- 8HUJ) on human 48S PIC (PDB Id- 8OZ0) with respect to HEAT1 of eIF4G, the domain J-K becomes positioned at the subunit face of 40S body, close to ES6 (Author response image 7). This correlates with the previously reported position for eIF4G with respect to EMCV IRES-48S PIC (Yu et al 2011). The predicted model shows no clashes with the canonical eIF4A-eIF3/ eIF4G-eIF4A-eIF3 interaction, or with the domain J-K-eIF4G-eIF4A model. Thus, highlighting a possibly compatible interaction axis among eIF3-eIF4G-eIF4A-domain J-K of IRES.

Author response image 7.

(upper left) Location of eIF4G-eIF4A in canonical human 48S PIC (PDB Id- 8OZ0). (upper right) Superimposition of 18S rRNA from human 48S and EMCV IRES 48S. (lower left) Superimposition of Human Closed 48S PIC structure (PDB Id- 8OZ0) on EMCV IRES-48S PIC model and placement of EMCV IRES- J-K domain-HEAT1-eIF4A structure (PDB Id- 8HUJ) with respect to eIF4G-HEAT1 domain. (lower right) Predicting location of eIF3 and eIF4 proteins in EMCV IRES-48S PIC.

(7) Assuming that the additional density near the ternary complex corresponds to Domain I of the IRES and that the codon in the P site represents the EMCV AUG start codon, what is the authors' mechanistic model for EMCV IRES-mediated initiation? Specifically, how is the mRNA positioned or inserted into the 40S mRNA channel in the absence of canonical scanning? As it stands, the discussion does not sufficiently address this key aspect of the EMCV initiation mechanism.

The EMCV IRES start codon (A-834) is directly placed in the P site (Pestova et al 1996), and the captured complex harboured the initiator tRNA in P_IN state with AUG at the P site. This start codon is preceded by domains J-K-L, where the J-K domain interacts with eIF4 proteins via eIF4G1-HEAT1 domain, and L domain is 20 residues upstream of the AUG and known to interact with eIF4B (Pestova et al 1996; de Quinto et al 2001). Based on the position and binding partners for these domains, the domain L could be placed at the mRNA exit site, preceded by domain J-K, which could be placed close to eIF4G-eIF4A position on EMCV IRES 48S PIC, near expansion segment 6 (ES6). The domain J-K can interact with eIF4G, localized close to the left foot or ES6 as per previous biochemical experiments (Yu et al 2011). This suggests that position of eIF4G and eIF4A could be the same as that of cap-dependent initiation where it can interact with eIF3 core subunits as well as the IRES domain J-K and the predicted path of mRNA from the exit site can follow the path of mRNA in human closed 48S PIC (PDB Id- 8OZ0), where it interacts with eIF3 core.

Examining the path of RNA in channel from the G-825 (exit site) to C-785 (domain J-K), we found the shortest distance is ~ 173 Å. This bridge could be filled by a single-stranded stretch of 40 nucleotides. However, the presence of domain L (stem loop- residues- 782 to 810) might hinder the placement of A-834 in the P-site (Author response image 8). We anticipate that to accommodate the start codon at the P site, either the domain L stem loop is resolved, which is an energetically expensive process (free energy of the thermodynamic ensemble is -11.12 kcal/mol, predicted using RNAfold). Another way could be a change in the orientation or conformation of domain J-K such that the start codon is directly placed at the P site without resolving domain L.

Author response image 8.

(left) The shortest distance between the last fitted residue- 825th of EMCV IRES to 785th of J-K domain of IRES (keeping eIF4G position same as that of PDB Id- 8OZ0) is 173 Å. (right) Tracing the path of mRNA (red) upstream of AUG coming out of the exit site of 40S ribosome and the possible position of eIF4G on EMCV IRES-48S PIC. Addition of nucleotides between C-785 and G-825 would fill the gap. The route of predicted mRNA from the exit channel is based on the mRNA (green) exiting the channel (PDB Id- 8OZ0).

The domain I is followed by domain J-K, close to the left foot of the 40S ribosomal subunit as per previous biochemical experiments (Yu et al 2011). However, the minimum distance connecting the I domain at 601st nucleotide to 682nd nucleotide of domain J-K (at the predicted location) is ~300 Å, which might be difficult to be covered by 80 nucleotides (from 601 to 682), present as a double helical strand. We suppose there could be instances of J-K domain repositioning in the EMCV IRES-48S PIC such that the I domain apical region can contact the 40S head and simultaneously place the start codon at the P site (Author response image 9).

Author response image 9.

Rotated views of EMCV IRES domains- I apical part in contact with 40S head and tRNAi and predicted location of J-K domain in contact with eIF4G, close to the left foot of 40S (predicted from PDB Id- 8OZ0). The minimum distance connecting 601st nucleotide in I domain to 682nd nucleotide in J-K domain is 295.5 Å.

We lack any details on the other IRES domains, such as domain I lower stem, domain J-K, or L; therefore, we refrained from commenting on these in our manuscript.

(8) Supplementary Figure 1 is missing labels for the RNA ladders.

The size of the DNA ladder used is mentioned.

References:

Bhattacharjee S, Abaeva IS, Brown ZP, Arhab Y, Fallah H, Hellen CUT, Frank J, Pestova TV. The mechanism of ribosomal recruitment during translation initiation on Type 2 IRESs. bioRxiv [Preprint]. 2025 Jun 11:2025.06.11.659010. doi: 10.1101/2025.06.11.659010. PMID: 40568087; PMCID: PMC12191231.

Brito Querido J, Sokabe M, Díaz-López I, Gordiyenko Y, Fraser CS, Ramakrishnan V. The structure of a human translation initiation complex reveals two independent roles for the helicase eIF4A. Nat Struct Mol Biol. 2024 Mar;31(3):455-464. doi: 10.1038/s41594-023-01196-0. Epub 2024 Jan 29. PMID: 38287194; PMCID: PMC10948362.

Brito Querido J, Sokabe M, Kraatz S, Gordiyenko Y, Skehel JM, Fraser CS, Ramakrishnan V. Structure of a human 48S translational initiation complex. Science. 2020 Sep 4;369(6508):1220-1227. doi: 10.1126/science.aba4904. PMID: 32883864; PMCID: PMC7116333.

Chamond N, Deforges J, Ulryck N, Sargueil B. 40S recruitment in the absence of eIF4G/4A by EMCV IRES refines the model for translation initiation on the archetype of Type II IRESs. Nucleic Acids Res. 2014;42(16):10373-84. doi: 10.1093/nar/gku720. Epub 2014 Aug 26. PMID: 25159618; PMCID: PMC4176346.

Dorn G, Gmeiner C, de Vries T, Dedic E, Novakovic M, Damberger FF, Maris C, Finol E, Sarnowski CP, Kohlbrecher J, Welsh TJ, Bolisetty S, Mezzenga R, Aebersold R, Leitner A, Yulikov M, Jeschke G, Allain FH. Integrative solution structure of PTBP1-IRES complex reveals strong compaction and ordering with residual conformational flexibility. Nat Commun. 2023 Oct 13;14(1):6429. doi: 10.1038/s41467-023-42012-z. PMID: 37833274; PMCID: PMC10576089.

Duke GM, Hoffman MA, Palmenberg AC. Sequence and structural elements that contribute to efficient encephalomyocarditis virus RNA translation. J Virol. 1992 Mar;66(3):1602-9. doi: 10.1128/JVI.66.3.1602-1609.1992. PMID: 1310768; PMCID: PMC240893.

Fernández N, Fernandez-Miragall O, Ramajo J, García-Sacristán A, Bellora N, Eyras E, Briones C, Martínez-Salas E. Structural basis for the biological relevance of the invariant apical stem in IRES-mediated translation. Nucleic Acids Res. 2011 Oct;39(19):8572-85. doi: 10.1093/nar/gkr560. Epub 2011 Jul 8. PMID: 21742761; PMCID: PMC3201876.

Hashem Y, des Georges A, Dhote V, Langlois R, Liao HY, Grassucci RA, Pestova TV, Hellen CU, Frank J. Hepatitis-C-virus-like internal ribosome entry sites displace eIF3 to gain access to the 40S subunit. Nature. 2013 Nov 28;503(7477):539-43. doi: 10.1038/nature12658. Epub 2013 Nov 3. PMID: 24185006; PMCID: PMC4106463.

Imai S, Suzuki H, Fujiyoshi Y, Shimada I. Dynamically regulated two-site interaction of viral RNA to capture host translation initiation factor. Nat Commun. 2023 Aug 28;14(1):4977. doi: 10.1038/s41467-023-40582-6. PMID: 37640715; PMCID: PMC10462655.

Kim H, Kim K, Kwon T, Kim DW, Kim SS, Kim YJ. Secondary structure conservation of the stem-loop IV sub-domain of internal ribosomal entry sites in human rhinovirus clinical isolates. Int J Infect Dis. 2015 Dec;41:21-8. doi: 10.1016/j.ijid.2015.10.015. Epub 2015 Oct 27. PMID: 26518063.

Lomakin IB, Hellen CU, Pestova TV. Physical association of eukaryotic initiation factor 4G (eIF4G) with eIF4A strongly enhances binding of eIF4G to the internal ribosomal entry site of encephalomyocarditis virus and is required for internal initiation of translation. Mol Cell Biol. 2000 Aug;20(16):6019-29. doi: 10.1128/mcb.20.16.6019-6029.2000. PMID: 10913184; PMCID: PMC86078.

López de Quinto S, Martínez-Salas E. Conserved structural motifs located in distal loops of aphthovirus internal ribosome entry site domain 3 are required for internal initiation of translation. J Virol. 1997 May;71(5):4171-5. doi: 10.1128/JVI.71.5.4171-4175.1997. PMID: 9094703; PMCID: PMC191578.

Lozano G, Francisco-Velilla R, Martinez-Salas E. Ribosome-dependent conformational flexibility changes and RNA dynamics of IRES domains revealed by differential SHAPE. Sci Rep. 2018 Apr 3;8(1):5545. doi: 10.1038/s41598-018-23845-x. PMID: 29615727; PMCID: PMC5882922.

Maloney A, Joseph S. Validating the EMCV IRES Secondary Structure with Structure-Function Analysis. Biochemistry. 2024 Jan 2;63(1):107-115. doi: 10.1021/acs.biochem.3c00579. Epub 2023 Dec 11. PMID: 38081770; PMCID: PMC10896073.

Pestova TV, Hellen CU, Shatsky IN. Canonical eukaryotic initiation factors determine initiation of translation by internal ribosomal entry. Mol Cell Biol. 1996 Dec;16(12):6859-69. doi: 10.1128/MCB.16.12.6859. PMID: 8943341; PMCID: PMC231689.

Petrychenko V, Yi SH, Liedtke D, Peng BZ, Rodnina MV, Fischer N. Structural basis for translational control by the human 48S initiation complex. Nat Struct Mol Biol. 2024 Sep 17. doi: 10.1038/s41594-024-01378-4. Epub ahead of print. PMID: 39289545.

Roberts LO, Belsham GJ. Complementation of defective picornavirus internal ribosome entry site (IRES) elements by the coexpression of fragments of the IRES. Virology. 1997 Jan 6;227(1):53-62. doi: 10.1006/viro.1996.8312. PMID: 9007058.

Robertson ME, Seamons RA, Belsham GJ. A selection system for functional internal ribosome entry site (IRES) elements: analysis of the requirement for a conserved GNRA tetraloop in the encephalomyocarditis virus IRES. RNA. 1999 Sep;5(9):1167-79. doi: 10.1017/s1355838299990301. PMID: 10496218; PMCID: PMC1369840.

Sweeney TR, Abaeva IS, Pestova TV, Hellen CU. The mechanism of translation initiation on Type 1 picornavirus IRESs. EMBO J. 2014 Jan 7;33(1):76-92. doi: 10.1002/embj.201386124. Epub 2013 Dec 15. PMID: 24357634; PMCID: PMC3990684.

Velazquez MA, Nuthalapati SS, Hankinson J, Fominykh K, Lulla V, Sweeney TR, Hill CH. Structural and mechanistic insights into translation initiation on the enterovirus Type 1 IRES. bioRxiv [Preprint]. 2025 Oct 3: 2025.10.04.680434. doi: 10.1101/2025.10.04.680434.

Yu Y, Sweeney TR, Kafasla P, Jackson RJ, Pestova TV, Hellen CU. The mechanism of translation initiation on Aichivirus RNA mediated by a novel type of picornavirus IRES. EMBO J. 2011 Aug 26;30(21):4423-36. doi: 10.1038/emboj.2011.306. PMID: 21873976; PMCID: PMC3230369.

eLife Assessment

This is a valuable study on single-cell transcriptomic analyses, focused on morphogenesis of the zebrafish inner ear in wildtype and lmx1bb mutants. The supporting evidence is mostly convincing, but incomplete in parts.

Reviewer #1 (Public review):

Summary:

The authors dissected the ears with some surrounding tissue from 600 embryos at 4 developmental time points of wild-type larvae, as well as from an lmx1bb mutant, performed scRNA-seq analyses, and subclustered the ear/neuromast clusters. They identified cluster markers and performed PAGA pseudotime analyses to build developmental timelines of lineages. They validated some of the cluster markers with HCRs. Many of the clusters are not annotated in detail, but the data sets are still valuable for the community.

Strengths:

Using scRNA-Seq, the authors identified cluster markers for tissues of the developing zebrafish ear and validated some of them with HCRs. The data they compiled and submitted to public databases is a valuable resource for the community.

Weaknesses:

Many of the clusters have not been annotated or rely on published data. For the ones for which no HCRs or UMAPs are shown, it is therefore difficult to estimate which of the markers are indeed the most cell type/state-specific ones.

Major comments:

(1) It would be very useful if the cluster numbers in the Excel files also had the associated cell type annotations as a second column (at least for the ones that are known). E.g., in Supplemental Table 2, the text states which clusters represent which neuromast and ear cell type, but these are not mentioned in the Excel table.

(2) Many of the clusters have not been annotated or rely on published data. For the ones for which no HCRs or UMAPs are shown, it is therefore difficult to estimate which of the markers are indeed the most cell-type/state-specific ones.

(3) Uploading the data to gEAR (https://umgear.org/dataset_explorer.html), a web-based, publicly available ear database, would further increase the usefulness of this study to the broader community.

Method:

The authors should provide the details about how many cells were sequenced for each ear developmental stage, how many cells were present per cluster (page 8), and how many cells were present in each subcluster of ear and lateral line clusters (page 10).

Reviewer #2 (Public review):

Summary:

Munjal and colleagues present a single-cell RNAseq atlas of otic tissue at 4 developmental stages, generate coarse-grained PAGA graphs to describe the development of various otic cell types, rigorously validate their scRNAseq annotations using fluorescent in situ hybridization, and identify changes in epcam expression in lmx1bb mutants that potentially cause the dramatic defects in otic vesicle formation in these mutants.

Strengths:

The data set is very nice, and the annotations are extremely rigorous and more in-depth than other datasets that include these tissues, since these investigators have enriched significantly for this tissue of interest. Their use of PAGA to identify potential developmental relationships within the data is rigorous. I also would like to specifically point out how incredibly gorgeous the microscopy of the lmx1bb phenotype is in Figure 7. Wow.

Weaknesses:

A missed opportunity is that the authors describe creating an additional scRNAseq dataset from lmx1bb mutants, but do not show any comparative scRNAseq analyses that would identify broader sets of differentially expressed genes. It seems almost as if a key element of the study was removed at the last minute, and as a result, the discussion of changes in epcam expression in lmx1bb mutants in Figure 7 seems somewhat tacked onto the end of the study and not motivated by the analyses presented in the manuscript.

Overall, I do not think this study requires any major revisions to be appropriate and useful to the community. This study would be potentially stronger with a more formal analysis of what gene expression changes occurred in otic tissue in lmx1bb mutants, but it is also useful without this. I did have a couple of minor suggestions for the presentation of some aspects that would have made it easier for me as a reader.

Reviewer #3 (Public review):

Summary:

The authors use single-cell transcriptomic analysis to identify distinct cell types in the zebrafish inner ear. They identify markers of hair cells and supporting cells associated with sensory patches, cells that generate the semicircular canals, endolymphatic duct and sac, and periotic mesenchymal cells.

Strengths:

The computational analysis is thorough, and the findings are clearly described. In situ hybridization provides corroboration of cell identities in many cases. This resource atlas will be of particular interest for studies of inner ear morphogenesis. Indeed, the identification of a smooth muscle marker in the endolymphatic sac suggests future analysis of the degree to which this structure undergoes contraction. Identification of cell signaling components in BMP, Wnt, FGF, and other signaling pathways will also provide a resource for understanding signals coordinating ear development.

Weaknesses:

The manuscript is incomplete. Important details that would allow replicable analysis are not provided, with notebooks not available on the referenced GitHub site, and additional files are missing.

The authors make a detailed description of hair cells and supporting cells that are consistent with previous findings (Figures 2 and 3). By contrast, the analysis of distinct cell types that have not been previously well characterized in zebrafish is somewhat incomplete. Markers are described for cells forming the semicircular canals, including ccn1l1 (Figure 4). The authors report an intriguing pattern of its expression before overt bud formation; however, they provide no detailed expression analysis to support this assertion.

The authors also identify new markers for subsets of periotic mesenchyme (Figure 6). These include epyc and otos, which mark distinct populations within the mammalian inner ear - cochlea supporting cells, spiral limbus, and ligament, respectively. Identification of the equivalent of the spiral ligament would be of particular interest. However, the expression analysis is not of sufficient resolution to identify which cell types these represent in the zebrafish inner ear.

Differences in gene expression are reported for lmx1bb mutants. However, none of the single-cell data for mutants is provided, and the table (S8) of differential gene expression is missing. Significantly more detail would be needed to interpret these findings.

Author response:

We thank the editors and reviewers for their careful consideration of our manuscript and for their constructive feedback, which we will address in detail in our revised version. We value that Reviewer 1 considered that “data they compiled and submitted to public databases is a valuable resource for the community.” We are also encouraged by Reviewer #2 when they stated that “The data set is very nice, and the annotations are extremely rigorous and more in-depth than other datasets that include these tissues, since these investigators have enriched significantly for this tissue of interest. Their use of PAGA to identify potential developmental relationships within the data is rigorous. I also would like to specifically point out how incredibly gorgeous the microscopy of the lmx1bb phenotype is in Figure 7. Wow.” We were encouraged by Reviewer #3’s comments that “The computational analysis is thorough, and the findings are clearly described. In situ hybridization provides corroboration of cell identities in many cases. This resource atlas will be of particular interest for studies of inner ear morphogenesis.”

We spent a significant effort and time considering and addressing the reviewers’ public criticisms.

Below we address the criticisms of the reviewers’ Public Reviews individually.

Public Reviews:

Reviewer #1 (Public review):

Weaknesses:

Many of the clusters have not been annotated or rely on published data. For the ones for which no HCRs or UMAPs are shown, it is therefore difficult to estimate which of the markers are indeed the most cell type/state-specific ones.

Major comments:

(1) It would be very useful if the cluster numbers in the Excel files also had the associated cell type annotations as a second column (at least for the ones that are known). E.g., in Supplemental Table 2, the text states which clusters represent which neuromast and ear cell type, but these are not mentioned in the Excel table.

Thank you for the suggestion, we will include additional annotations in the revised version.

(2) Many of the clusters have not been annotated or rely on published data. For the ones for which no HCRs or UMAPs are shown, it is therefore difficult to estimate which of the markers are indeed the most cell-type/state-specific ones.

We recognize the need to evaluate potential new markers, we will include a heat map of markers and clusters to assess cell-type/state specificity in the revised version.

(3) Uploading the data to gEAR (https://umgear.org/dataset_explorer.html), a web-based, publicly available ear database, would further increase the usefulness of this study to the broader community.

We appreciate the suggestion to upload to gEAR and will upload to the database in the near future.

Method:

The authors should provide the details about how many cells were sequenced for each ear developmental stage, how many cells were present per cluster (page 8), and how many cells were present in each subcluster of ear and lateral line clusters (page 10).

We will add cell numbers for each cluster in the revised version as an additional column in the supplemental tables.

Reviewer #2 (Public review):

Weaknesses:

A missed opportunity is that the authors describe creating an additional scRNAseq dataset from lmx1bb mutants, but do not show any comparative scRNAseq analyses that would identify broader sets of differentially expressed genes. It seems almost as if a key element of the study was removed at the last minute, and as a result, the discussion of changes in epcam expression in lmx1bb mutants in Figure 7 seems somewhat tacked onto the end of the study and not motivated by the analyses presented in the manuscript.

Overall, I do not think this study requires any major revisions to be appropriate and useful to the community. This study would be potentially stronger with a more formal analysis of what gene expression changes occurred in otic tissue in lmx1bb mutants, but it is also useful without this. I did have a couple of minor suggestions for the presentation of some aspects that would have made it easier for me as a reader.

We will include analysis of the lmx1bb mutant data in the revised version and value the suggestions for improved presentation. We will work on irmpoving presentation of the mutant data, including a UMAP with the WT cells in one color and the mutant cells in another color.

Reviewer #3 (Public review):

Weaknesses:

The manuscript is incomplete. Important details that would allow replicable analysis are not provided, with notebooks not available on the referenced GitHub site, and additional files are missing.

Python notebooks will be added shortly, and files for mapping in Drops data will be provided at the GitHub site.

The authors make a detailed description of hair cells and supporting cells that are consistent with previous findings (Figures 2 and 3). By contrast, the analysis of distinct cell types that have not been previously well characterized in zebrafish is somewhat incomplete. Markers are described for cells forming the semicircular canals, including ccn1l1 (Figure 4). The authors report an intriguing pattern of its expression before overt bud formation; however, they provide no detailed expression analysis to support this assertion.

The authors also identify new markers for subsets of periotic mesenchyme (Figure 6). These include epyc and otos, which mark distinct populations within the mammalian inner ear - cochlea supporting cells, spiral limbus, and ligament, respectively. Identification of the equivalent of the spiral ligament would be of particular interest. However, the expression analysis is not of sufficient resolution to identify which cell types these represent in the zebrafish inner ear.

Thank you for your input regarding the analysis of the periotic mesenchyme. In the revised version, we will attempt to improve resolution of different populations, first by comparing epyc and otos expression by HCR. It is unclear how to correlate any patterns with structures that have yet to evolve, but we will look for similarities and differences to studies performed in mice (PMID: 37720106).

Differences in gene expression are reported for lmx1bb mutants. However, none of the single-cell data for mutants is provided, and the table (S8) of differential gene expression is missing. Significantly more detail would be needed to interpret these findings.

We will include analysis of the lmx1bb mutant data in the revised version and value the suggestions for improved presentation.

eLife Assessment

This study analyzes the temporal dynamics of gene expression following TNF stimulation in macrophages. The work brings valuable data and new methodological approaches to implicate the splicing rate of certain introns as a mechanism regulating mature mRNA expression. This will be of interest to audiences in RNA biology and innate immune response regulation. The experimental design is solid for the core findings, although in places the data limit the conclusions.

Reviewer #1 (Public review):

Summary:

In this work, the authors revisit a well-defined experimental system for studying temporal gene expression mechanisms in TNF-alpha-stimulated macrophages, bringing new tools to the process. Using a hybrid-capture approach, they are able to obtain deeper RNA sequencing of target genes, which allows them to identify potential differences in splicing kinetics of individual introns. Further implementing transcriptional blocks to measure intron half-lives, and predictive machine learning models to identify potential contributing cis-acting RNA elements, they define a group of 'bottleneck' introns whose delayed splicing is a rate-limiting step in mRNA maturation.

Strengths:

(1) The hybrid-capture approach enables deeper RNA sequencing of target transcripts.

(2) The neural network application to identify motifs outside of splice sites could be related to intron removal kinetics.

(3) The paper uses splicing reporters with modulation of 5' splice sites to test the effect on reporter gene expression in the context of 'bottleneck' introns.

Weaknesses:

(1) While evidence is provided that these introns are distinct from previously published splicing kinetics studies, 'bottleneck' introns are not adequately placed in context for assessment of how they are similar or different.

(2) Splicing reporters are a good approach, but the complexities of post-transcriptional gene expression regulation are not adequately addressed

(3) Deep learning models are a potentially powerful tool for identifying novel regulatory sequences; however, their use here is underdeveloped.

Reviewer #2 (Public review):

Summary:

The authors analyzed the temporal dynamics of gene expression patterns within the inflammatory response transcriptome following TNF stimulation, and proposed that the splicing rate of certain introns is a key mechanism of regulating mature mRNA expression rate.

Strengths:

The measurement strategy is generally well-designed to understand the core question of splicing rate and gene expression. The following computation analysis, as well as the mutation or repair studies, further supported the claims. The writing and presentation of the results are also generally clear and easy to follow. I think this manuscript will be of interest to a wide audience.

Weaknesses: 

I do have some questions regarding some of the results and conclusions, and I think either more analysis or more explanation and discussion can make the claims more solid. Please see below for details:<br /> <br /> (1) On the hybrid capture method and the RNA coverage results: The strategy of enriching for the last exon before sequencing does have significance in linking pre-mRNA and mature mRNA. If I understand correctly, this enriches for pre-mRNA molecules that are about to finish the full-length elongation of RNA polymerase. However, is this strategy biased towards measuring the splicing rate variation on introns closer to the 3-prime end? For example, if a gene takes 5 minutes for the RNA polymerase to elongate through the full length of the gene, for intron #1 that's very close to the 5' end, you can't tell if it takes 20s to be spliced out or 4 minutes, as both will show as fully spliced out in the sequencing library. In other words, for introns near the 5' end, a consistent "CoSI=1" pattern in the data doesn't necessarily suggest a true consistent fast splicing of that intron. Do you observe any general pattern of the measured "slowliness" in relation to the 5'-3' location of the introns? If so, should the 5' introns be specially considered or even excluded from certain analyses that use all introns?<br /> <br /> (2) Following on my last point, it may benefit the readers if the author can provide a more detailed comparison of possible sequencing library construction choices. For example, is it feasible to also enrich for other exons for the sequencing library, etc?<br /> <br /> (3) Figure 1C: Are there biological replicates, and should there be error bars and statistics on the plot? Similarly, in places like Figure 2, Supplemental Figure 4C, Supplemental Figure 6, etc., is there any statistical analysis that can be done to show if the claimed differences are statistically significant?<br /> <br /> (4) The logic behind measuring the half-lives of introns seems a little unclear to me.  From the time-dependent RNA coverage plots in Figure 2, it seems that, if we assume a constant transcription elongation rate, then the splicing rate of a specific intron can vary across time after TNF stimulation, as represented by the temporal change of CoSI values, or the heights of the coverage plot relative to neighboring exons. This means the splicing rate or half-life of an intron is not necessarily constant but may be time-dependent, at least in the case of TNF stimulation. Shouldn't the half-life measurements be designed in a way to measure the half-life at multiple time points after TNF stimulation? And maybe the measured half-lives of some introns will show as time-dependent?<br /> <br /> (5) In Supplemental Figure 6, the interpretation is a little confusing to me: If delayed splicing is causing delayed expression of the corresponding gene, shouldn't the non-immediate gene groups (early/intermediate/Late) have low CoSI beginning from the early time points (e.g. 4 minutes)? Why does the slowdown of splicing seem to peak at a later time point? Does it mean immediately after TNF stimulation, there's a different mechanism in delaying the expression of the non-immediate gene groups? Maybe it's better to have more explanation or use a different visualization to show what non-immediate gene groups are experiencing at very early time points.<br /> <br /> (6) On the fine-tuning of the deep sequence model: it's a little unclear whether the input and output are time-dependent. It's stated that expression at multiple time points is used for training, but it's unclear whether the model outputs time-dependent expression patterns and whether the time information is used as input.

Reviewer #3 (Public review):

Summary:

The manuscript by Dearborn et al investigates the kinetics of intron splicing in inflammation-associated transcripts after TNF-stimulation of macrophages, using targeted sequencing of chromatin-associated RNA to obtain high coverage across a focused set of induced genes. The authors' main conclusion is that splicing kinetics are heterogeneous across these transcripts, and that delayed introns (which they term "bottleneck introns") are associated with weak donor sequences. Using a deep learning approach, they have also identified additional sequence features that might contribute to intron splicing kinetics.

Overall, I think the findings in the manuscript are very intriguing and will be of interest to readers working on RNA biology. The changes the authors have made to the manuscript in response to some very valid comments from reviewers have strengthened the manuscript. While the existing data might not be sufficient to directly address some of the broader mechanistic claims made by the authors, I think the findings are nonetheless very interesting and should contribute towards a better understanding of the post-transcriptional regulation of gene expression.

Strengths:

A strength of the manuscript is the experimental design. The targeted capture approach is innovative and well-suited to the goal of measuring intron-specific splicing behaviour across time. The inclusion of experimental validation in minigene assays of some of the computational predictions also strengthens the claims made by the authors.

The authors have made a constructive effort to address some of the concerns raised in a previous round of review. The revised manuscript reads as a balanced text.

Weaknesses:

The study still does not fully resolve the downstream consequences of delayed splicing. In particular, it remains unclear whether the bottleneck introns lead primarily to delayed production of mature transcripts, reduced productive transcript output, or some combination of the two.

On a related point, the minigene reporter assays measure a steady-state level of the transcript and don't provide insights into the kinetics directly.

Lastly, given that the detailed analyses were performed on a selected subset of (inflammation-induced) transcripts, a broader evolutionary interpretation needs to be restrained given the current data.

Author response:

We thank the Reviewing Editor and reviewers for their thoughtful and constructive evaluation of our manuscript, Programmed Delayed Splicing: A Mechanism for Timed Inflammatory Gene Expression. We are encouraged that the reviewers found the study valuable, the experimental design strong for the core findings. We appreciate the reviewers’ careful attention to the limits of inference in several parts of the manuscript, and will address these points in a revised version. We especially want to acknowledge that this paper has benefited from the abiding interest in splicing regulation by the editors and reviewers who have meticulously improved nearly every aspect of this multifaceted work in its present state.

Our planned revisions will focus on five areas. First, we will more carefully evaluate and discuss the extent to which the hybrid-capture strategy may impose position-dependent constraints on apparent splicing behavior, particularly across 5′ and 3′ introns. Second, we will clarify the use of the term “bottleneck introns,” distinguishing descriptive use in the main text from the ranked subsets used in downstream analyses. Third, we will revise the framing of the reporter assays to make explicit that these measure steady-state reporter output and do not, on their own, resolve all downstream kinetic consequences of delayed splicing. Fourth, we will clarify the interpretation of the actinomycin D experiments as providing estimates of intron excision behavior under transcriptional arrest rather than a complete time-resolved model of splicing during TNF induction. Fifth, we will substantially revise the scope and stated limitations of the deep learning-aided interpretations of data in this work.

Reviewer #1

We thank Reviewer #1 for the positive assessment of the hybrid-capture strategy, the splice-site reporter experiments, and the potential value of the neural-network-based analysis. We appreciate the reviewer’s view that these approaches help extend a well-established system for studying temporal gene expression in TNF-stimulated macrophages. We address the main concerns raised in the public review below.

(1) While evidence is provided that these introns are distinct from previously published splicing kinetics studies, “bottleneck” introns are not adequately placed in context for assessment of how they are similar or different.

We appreciate this point and agree that the current manuscript does not yet place these introns in sufficiently clear context relative to prior literature. Our study builds on foundational work describing regulated changes in splicing kinetics, widespread intron retention, and detained introns as biologically meaningful modes of gene regulation, including transcript-specific regulation of splicing in response to stress (Pleiss, Mol Cell., 2007), widespread functional intron retention in mammals (Braunschweig, Genome Res., 2014), and the definition of detained introns as a distinct class of post-transcriptionally spliced introns (Boutz, Genes Dev., 2015). In revision, we will expand the comparison to previously described classes of delayed or retained introns and clarify more explicitly how the introns studied here are defined in the setting of inducible inflammatory transcripts and their temporal resolution over the course of stimulation. We will also revise the relevant Results and Discussion text so that the distinction is made directly in the manuscript rather than relying on inference from the broader presentation.

(2) Splicing reporters are a good approach, but the complexities of post-transcriptional gene expression regulation are not adequately addressed.

We agree that the interpretive limits of the reporter assays should be stated more clearly and consistently. In revision, we will revise the presentation of the minigene experiments to make explicit that these are steady-state reporter assays and therefore do not, on their own, resolve all downstream kinetic consequences of delayed splicing in the endogenous context. At the same time, we believe the assay remains informative because it provides a controlled system in which the contribution of splice donor sequence can be tested directly in matched reporter constructs. In that sense, the reporter experiments are valuable as a reductionist test of whether weak donor sequences are sufficient to alter reporter output, even if they do not fully recapitulate the broader endogenous post-transcriptional environment. We will emphasize that these data support an association between weak donor sites and altered reporter output, while moderating any broader mechanistic claims that extend beyond what the assay directly measures.

(3) Deep learning models are a potentially powerful tool for identifying novel regulatory sequences; however, their use here is underdeveloped.

We appreciate this concern and agree that the deep-learning section should be revised substantially. In a revised manuscript, we will clarify the training setup, the definition of the slow-intron subsets used in downstream analyses, and the interpretation of the attribution and motif analyses. Alongside, we believe the assay remains informative because it provides a controlled system in which the contribution of splice donor sequence can be tested directly in matched reporter constructs. In that respect, the reporter experiments are valuable as a reductionist test of whether weak donor sequences are sufficient to alter reporter output, even if they do not fully recapitulate the broader endogenous post-transcriptional environment. We will revise the framing of these results so that they are presented more explicitly as identifying candidate sequence features associated with delayed splicing, rather than as direct evidence of specific causal regulatory mechanisms.

Reviewer #2

We thank Reviewer #2 for the thoughtful and detailed comments, and for recognizing the strengths of the measurement strategy and the clarity of the manuscript. We appreciate the reviewer’s view that the study will be of interest to a broad audience, and we agree that several conclusions will be strengthened by additional analysis and clearer explanation. We address the main concerns raised in the public review below.

(1) Concern regarding possible bias of the hybrid-capture strategy toward introns closer to the 3′ end, and whether 5′ introns should be treated separately in some analyses.

We thank the reviewer for this careful and important point. We agree that this is a potential limitation of the approach and that it should be addressed more explicitly in the manuscript. Our assay begins with poly(A)-selected RNA and then enriches transcripts of interest through terminal-exon capture, so the molecules analyzed are completed, polyadenylated transcripts rather than nascent partial transcripts. This feature is important for reducing ambiguity arising from incomplete transcription, particularly in the chromatin-associated fraction. At the same time, we agree that for introns near the 5′ end, the assay may have limited power to distinguish very rapid splicing from moderately rapid splicing if excision is largely complete by the time the transcript is fully synthesized and polyadenylated.

In revision, we will address this concern directly in two ways. First, we will revise the Results and Discussion to clarify that the assay provides a population-level measure of splice completion in completed transcripts and that interpretation is strongest for introns whose excision is not already fully resolved before transcript completion. Second, we will more systematically evaluate whether apparent slow splicing covaries with transcript position, distance from the 3′ end, and intron length, and we will perform sensitivity analyses with and without the most 5′ introns to determine which conclusions are robust to these positional constraints. We will also examine transcript coverage patterns in greater detail to better assess the extent to which library construction and  cDNA generation may contribute to apparent positional bias. Our preliminary inspection suggests that transcript position is not the sole determinant of the observed heterogeneity, but we agree that a more explicit treatment of this issue is warranted in the revised manuscript.

(2) Request for more detailed discussion of alternative library-construction choices.

We appreciate this suggestion and agree that the revised manuscript would benefit from a fuller discussion of the strengths and limitations of the current enrichment strategy. We chose poly(A) selection followed by terminal-exon capture because this design enriches completed transcripts of interest and reduces ambiguity from nascent partial transcripts, which is particularly important in the chromatin-associated fraction. This approach also provides greater read depth over the selected inflammatory transcripts, enabling more informative intron-level comparisons within the targeted dataset. In revision, we will clarify this rationale more explicitly in the manuscript. We will also discuss the tradeoffs of this design relative to alternative exon-targeting strategies and how those alternatives might provide different, but complementary, views of splicing kinetics.

(3) Questions regarding biological replicates, error bars, and statistical analysis in Figure 1C and other plots.

We agree that the replicate structure and intended interpretation of these plots should be clarified more explicitly. In revision, we will revise the figure legends and Methods to distinguish panels that display a single bulk RNA-seq time course (for example, Figure 1C) from panels that summarize distributions across many introns (for example, Figure 2 and Supplementary Figure 6). We will also add statistical comparisons where they are most appropriate and informative, such as in sequence-feature comparisons like Supplementary Figure 4C, while making clear that some CoSI panels are intended as descriptive summaries of intron-level heterogeneity rather than replicate-based inferential plots.

(4) Concern that intron half-lives may be time-dependent during TNF induction, and that the logic of the actinomycin D measurements is therefore unclear.

We appreciate this point and agree that the manuscript should distinguish more clearly between two related but non-identical quantities: the CoSI trajectories observed during ongoing TNF induction, and the interruption-based half-life estimates derived from actinomycin D treatment. The actinomycin D experiments were performed using multiple post-treatment timepoints, but they were designed to estimate intron excision behavior after transcriptional arrest under a defined set of conditions, rather than to measure whether an individual intron’s effective splicing rate changes across all phases of the TNF response. We agree that these estimates should therefore be interpreted as constrained measurements under the assay conditions used, rather than as a complete time-resolved model of splicing kinetics during induction. In revision, we will clarify this point in the Results, Methods, and Discussion, and we will more explicitly acknowledge that effective splicing behavior could vary across the induction time course.

(5) Concern that the interpretation of Supplementary Figure 6 is unclear, particularly why delayed splicing in non-immediate groups appears to peak later rather than at the earliest time points.

We appreciate this point and agree that the current presentation of Supplementary Figure 6 does not explain this behavior clearly enough. Our interpretation is not that delayed splicing is the sole determinant of early versus later induction classes. Rather, the earliest time points reflect a combination of transcriptional induction timing and RNA processing state. In this framework, the dip in CoSI shortly after stimulation reflects the appearance of newly induced, incompletely spliced transcripts, and the later kinetic groups appear to recover from this dip more slowly than the immediate-early group. Thus, the strongest signal of delayed splicing may become most apparent only after sufficient transcript accumulation, rather than necessarily at the very earliest time point. In revision, we will revise the text to make this logic clearer and will consider a more intuitive visualization of these group-specific CoSI trajectories.

(6) Concern that the deep-learning setup does not make clear whether the model input and output are time-dependent.

We appreciate this concern and agree that the current manuscript does not explain the model setup clearly enough. Briefly, we will clarify the role of the three TNF timepoints in model training, including the fact that these outputs were modeled jointly and that time itself was not provided as an explicit input to the model. We will also revise the Results and Methods so that the scope and interpretation of the resulting analyses are more explicit.

Reviewer #3

We thank Reviewer #3 for the positive assessment of the targeted capture design, the evaluation of overall interest of the findings, and the improvements in the current version. We appreciate the reviewer’s view that the study is intriguing and that the manuscript has been strengthened in revision. We agree, however, that the manuscript should more clearly distinguish what is directly demonstrated from what remains mechanistically unresolved. We address the main concerns raised in the public review below.

(1) The study still does not fully resolve the downstream consequences of delayed splicing, including whether bottleneck introns lead primarily to delayed production of mature transcripts, reduced productive transcript output, or some combination of the two.

We agree with this assessment. The current data do not fully resolve whether delayed splicing primarily delays mature transcript production, reduces productive transcript output, or reflects some combination of the two. In revision, we will further moderate the framing of the downstream consequences of delayed splicing and will revise the Abstract, Results, and Discussion to make clear that the present data do not fully distinguish among delayed mature transcript production, reduced productive transcript output, or a combination of both. We will ensure that the manuscript consistently presents these possibilities as alternatives not fully resolved by the current data.

(2) The minigene reporter assays measure a steady-state level of the transcript and do not provide direct insight into kinetics.

We agree and will revise the manuscript to make this limitation explicit throughout. In particular, we will ensure that the reporter assays are described consistently as steady-state reporter assays that support an association between splice donor strength and altered reporter output, while avoiding stronger claims that they directly resolve endogenous splicing kinetics or downstream transcript fate.

(3) Given that the detailed analyses were performed on a selected subset of inflammation-induced transcripts, a broader evolutionary interpretation should be restrained.

We agree that the broader evolutionary and mechanistic framing should be more carefully defined. In revision, we will restrain these interpretations so that they remain closely aligned with the inflammation-focused and targeted-transcript scope of the current study, and we will moderate language that extends beyond what is directly supported by the present dataset.

Closing Remarks

We again thank the reviewers for their constructive comments. We believe that the planned revisions will strengthen the manuscript by clarifying the scope of the mechanistic conclusions, sharpening the interpretation of the experimental approaches, and more carefully defining the role of the computational analyses. We appreciate the opportunity to revise the work and to provide this provisional response to accompany the Reviewed Preprint.

eLife Assessment

In this important manuscript, Matsuda and colleagues present a model describing the regulation of tracheal tubulogenesis in Drosophila melanogaster embryos. The authors support this model using convincing approaches that combine novel experimental results with previously published work from their group. While some conclusions are consistent with earlier studies, the present manuscript introduces distinct molecular markers not previously reported, which reinforce the authors' prior findings. In addition, the manuscript analyses, using experimental strategies, the requirement of the Dpp and EGFR signalling pathways for the maintenance of trachealess (trh), one of the key transcription factors governing tracheal development.

Reviewer #1 (Public review):

Summary:

In this manuscript, Matsuda and collaborators present a model of how tracheal tubulogenesis is controlled in Drosophila embryos. Some of the results backing the model are new, but others are based on information already published by the authors. However, the results in this manuscript present different molecular markers not published before, which agree with previous conclusions. The manuscript also analyses the requirement of the dpp and EGFR signalling pathways for trachealess (trh) maintenance, one of the main tracheal transcription factors.

Strengths:

The two most interesting novel points of the manuscript are:

(1) Its contribution to the analysis of how the dpp and EGFR pathways contribute to the maintenance of trh expression.

(2) The experimental evidence showing that mechanical invagination is not a requirement for trh maintenance in the tracheal cells, an intriguing hypothesis previously suggested by (Kondo Hayashi 2019 eLife 8:e45145) that can now be discarded by the data presented in this work.

Weaknesses:

Because of the mixture of new and already published data, this manuscript can be considered as a review/experimental paper.

Already known data:<br /> - The results showing that hh and vvl drive tracheal invaginaton independently of trh are reported in Figure 5 of (Matsuda et al. 2015 eLife 4:e09646).<br /> - The results showing dpp requirement for trh maintenance are partially reported in Figure 6 of (Matsuda 2015 eLife 4:e09646).

Reviewer #2 (Public review):

Summary:

Matsuda et al. investigate the regulatory mechanisms controlling gene expression and morphogenesis in the Drosophila embryonic trachea. Building on previous findings that tracheal invagination can occur independently of trh, they identify extrinsic hh and intrinsic vvl as key regulators that cooperatively promote this process. The study also integrates major signaling pathways (Dpp/BMP and EGFR) in defining tracheal cell identity and demonstrates that Ras activation can upregulate trh. Overall, the work supports a model in which multiple transcription factors and signaling inputs coordinate airway progenitor specification.

Strengths:

This study uses genetic analysis of various mutants to dissect regulatory relationships underlying tracheal development. While the uncoupling of tracheal invagination from trh function has been previously recognized, this work advances the field by identifying hh and vvl as key regulators of invagination independent of trh. The study also integrates multiple signaling pathways, such as Dpp/BMP and EGFR, into a coherent framework for tracheal cell specification. In addition, the demonstration that Ras activation can upregulate trh provides a clear mechanistic link between RTK signaling and transcriptional regulation. Overall, the work offers important and broadly relevant insights into how gene expression and morphogenesis are coordinated during development.

Weaknesses:

Data presentation and clarity of interpretation could be improved. Many images primarily show lateral views of whole embryos, which can make it difficult to fully assess some phenotypes; higher-magnification or sectional views would enhance clarity. There are also some minor inconsistencies in the description of invagination phenotypes, particularly regarding whether all trh+ cells remain in a 2D plane versus indications of partial invagination in hh vvl double mutants blocking apoptosis, which would benefit from further clarification. Finally, some statements in the abstract, especially regarding the role of grn, are not directly supported by data in this study and could be better aligned with the scope of the presented results.

Author response:

Reviewer #1<br /> - The results showing that hh and vvl drive tracheal invaginaton independently of trh are reported in Figure 5 of (Matsuda et al. 2015 eLife 4:e09646).

Reviewer #2

Many images primarily show lateral views of whole embryos, which can make it difficult to fully assess some phenotypes; higher-magnification or sectional views would enhance clarity. There are also some minor inconsistencies in the description of invagination phenotypes, particularly regarding whether all trh+ cells remain in a 2D plane versus indications of partial invagination in hh vvl double mutants blocking apoptosis, which would benefit from further clarification.

The data in our previous eLife publication (DOI: 10.7554/eLife.09646)1 were mostly projection views. Therefore, it is hard to conclude if the airway progenitors of hh vvl double mutants failed to invaginate or they invaginated to form sacs. We will provide magnified views of the progenitor invagination in hh vvl double mutants and describe the degrees of their invagination phenotypes.

Reviewer #1

The results showing dpp requirement for trh maintenance are partially reported in Figure 6 of (Matsuda 2015 eLife 4:e09646).

Reviewer #2

Finally, some statements in the abstract, especially regarding the role of grn, are not directly supported by data in this study and could be better aligned with the scope of the presented results.

trh-lacZ (1-eve-1) has been used as the earliest and the strongest enhancer trap line to mark the airway primordia and the airway progenitors. Perdurance of beta-galactocidase proteins makes it difficult to conclude if the marker signals result from the active transcriptional state of the trh locus. In our previous eLife publication we showed that Trh proteins and trh_transcripts are not detectable in _H99 grn hh vvl quadruple mutants and in grn hh vvl triple mutants (Figure 5H and Figure 5-figure supplement 2A of DOI: 10.7554/eLife.09646, respectively)1, although trh-LacZ signals are detected in grn hh vvl triple mutants.

Similarly, although we previously showed trh-LacZ expression in dpp mutant combinations, Figure 2 in the current manuscript, shows that even strong trh-LacZ signals do not always correlate with trh transcripts in dpp mutants. Therefore, in the current manuscript we included the data of dpp-driven positive regulation of trh transcripts at later stages since they have not been shown before.

Assessments and advices of the Editors and the Reviewers are indispensable for improving the manuscript. We will address all the Reviewers comments (Weakness of Public review, major and minor issues of Recommendations for the authors) both experimentally and in the text.

Sincerely yours,

Christos Samakovlis on behalf of all authors

• (1) Matsuda, R., Hosono, C., Samakovlis, C. & Saigo, K. Multipotent versus differentiated cell fate selection in the developing Drosophila airways. eLife 4 (2015).

eLife Assessment

This important work advances our understanding of intraflagellar transport, ciliogenesis, and ciliary-based signaling, by identifying the interactions of IFT172 with IFT-A components, ubiquitin-binding, and ubiquitination, mediated by IFT172 C-terminus and its role in ciliogenesis and ciliary signaling. The evidence supporting the findings is convincing. This paper will be of interest to cell biologists and biochemists, especially those working on cilia and signaling.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers. The authors have addressed the comments raised in the previous round of review.]

Summary:

Zacharia and colleagues investigate the role of the C-terminus of IFT172 (IFT172c), a component of the IFT-B subcomplex. IFT172 is required for proper ciliary trafficking and mutations in its C-terminus are associated with skeletal ciliopathies. The authors begin by performing a pull-down to identify binding partners of His-tagged CrIFT172968-C in Chlamydomonas reinhardtii flagella. Interactions with three candidates (IFT140, IFT144, and a UBX-domain containing protein) are validated by AlphaFold Multimer with the IFT140 and IFT144 predictions in agreement with published cryo-ET structures of anterograde and retrograde IFT trains. They present a crystal structure of IFT172c and find that a part of the C-terminal domain of IFT172 resembles the fold of a non-canonical U-box domain. As U-box domains typically function to bind ubiquitin-loaded E2 enzymes, this discovery stimulates the authors to investigate the ubiquitin-binding and ubiquitination properties of IFT172c. Using in vitro ubiquitination assays with truncated IFT172c constructs, the authors demonstrate partial ubiquitination of IFT172c in the presence of the E2 enzyme UBCH5A. The authors also show a direct interaction of IFT172c with ubiquitin chains in vitro. Finally, the authors demonstrate that deletion of the U-box-like subdomain of IFT172 impairs ciliogenesis and TGFbeta signaling in RPE1 cells.

However, some of the conclusions of this paper are only partially supported by the data, and presented analyses are potentially governed by in vitro artifacts. In particular, the data supporting autoubiquitination and ubiquitin-binding are inconclusive. Without further evidence supporting a ubiquitin-binding role for the C-terminus, the title is potentially misleading.

Strengths:

(1) The pull-down with IFT172 C-terminus from C. reinhardtii cilia lysates is well performed and provides valuable insights into its potential roles.

(2) The crystal structure of the IFT172 C-terminus is of high quality.

(3) The presented AlphaFold-multimer predictions of IFT172c:IFT140 and IFT172c:IFT144 are convincing and agree with experimental cryo-ET data.

Reviewer #2 (Public review):

Summary:

Cilia are antenna-like extensions projecting from the surface of most vertebrate cells. Protein transport along the ciliary axoneme is enabled by motor protein complexes with multimeric so-called IFT-A and IFT-B complexes attached. While the components of these IFT complexes have been known for a while, precise interactions between different complex members, especially how IFT-A and IFT-B subcomplexes interact, are still not entirely clear. Likewise, the precise underlying molecular mechanism in human ciliopathies resulting from IFT dysfunction has remained elusive.

Here, the authors investigated the structure and putative function of the to-date poorly characterised C-terminus of IFT-B complex member IFT172 using alpha-fold predictions, crystallography and biochemical analyses including proteomics analyses followed by mass spectrometry, pull-down assays, and TGFbeta signalling analyses using chlamydomonas flagellae and RPE cells. The authors hereby provide novel insights into the crystal structure of IFT172 and identify novel interaction sites between IFT172 and the IFT-A complex members IFT140/IFT144. They suggest a U-box-like domain within the IFT172 C-terminus could play a role in IFT172 auto-ubiquitination as well as for TGFbeta signalling regulation.

As a number of disease-causing IFT72 sequence variants resulting in mammalian ciliopathy phenotypes in IFT172 have been previously identified in the IFT172 C-terminus, the authors also investigate the effects of such variants on auto-ubiquitination. This revealed no mutational effect on mono-ubiquitination which the authors suggest could be independent of the U-box-like domain but reduced overall IFT172 ubiquitination.

Strengths:

The manuscript is clear and well written and experimental data is of high quality. The findings provide novel insights into IFT172 function, IFT complex-A and B interactions, and they offer novel potential mechanisms that could contribute to the phenotypes associated with IFT172 C-terminal ciliopathy variants.

Reviewer #3 (Public review):

Summary:

Zacharia et al report on the molecular function of the C-terminal domain of the intraflagellar transport IFT-B complex component IFT172 by structure determination and biochemical in vitro and cell culture-based assays. The authors identify an IFT-A binding site that mediates a mutually exclusive interaction to two different IFT-A subunits, IFT144 and IFT140, consistent with interactions suggested in anterograde and retrograde IFT trains by previous cryo-electron tomography studies. Additionally, the authors identify a U-box-like domain that binds ubiquitin and conveys ubiquitin conjugation activity in the presence of the UbcH5a E2 enzyme in vitro. RPE1 cell lines that lack the U-box domain show a reduction in ciliation rate with shorter cilia, and heterozygous cells manifest TGF-beta signaling defects, suggesting an involvement of the U-box domain in cilium-dependent signaling.

Strengths:

(1) The structural analyses of the C-terminal domain of IFT172 combine crystallography with structure prediction using state-of-the-art algorithms, which gives high confidence in the presented protein structures. The structure-based predictions of protein interactions are validated by further biochemical experiments to assess the specific binding of the IFT172 C-terminal domains with other proteins.

(2) The finding that the IFT172 C-terminus interactions with the IFT-A components IFT140 and IFT144 appear mutually exclusive confirm a suggested role in mediating the binding of IFT-B to IFT-A in anterograde and retrograde IFT trains, which is of very high scientific value.

(3) The suggested molecular mechanism of IFT train coordination explains previous findings in Chlamydomonas IFT172 mutants, in particular an IFT172 mutant that appeared defective in retrograde IFT, as well as mutations identified in ciliopathy patients.

(4) The identification of other IFT172 interactors by unbiased mass spectrometry-based proteomics is very exciting. Analysis of stoichiometries between IFT components suggests that these interactors could be part of IFT trains, either as cargos or additional components that may fulfill interesting functions in cilia and flagella.

(5) The authors unexpectedly identify a U-box-like fold in the IFT172 C-terminus and thoroughly dissect it by sequence and mutational analyses to reveal unexpected ubiquitin binding and potential intrinsic ubiquitination activity.

(6) The overall data quality is very high. The use of IFT172 proteins from different organisms suggests a conserved function.

Overall, the authors achieved to characterize an understudied protein domain of the ciliary intraflagellar transport machinery and gained important molecular insights into its role in primary cilia biology, beyond IFT. By identifying an unexpected functional protein domain and novel interaction partners the work makes an important contribution to further our understanding of how ciliary processes might be regulated by ubiquitination on a molecular level. Based on this work it will be important for future studies in the cilia community to consider direct ubiquitin binding by IFT complexes.

Conceptually, the study highlights that protein transport complexes can exhibit additional intrinsic structural features for potential auto-regulatory processes. Moreover, the study adds to the functional diversity of small U-box and ubiquitin-binding domains, which will be of interest to a broader cell biology and structural biology audience.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

Summary:

Zacharia and colleagues investigate the role of the C-terminus of IFT172 (IFT172c), a component of the IFT-B subcomplex. IFT172 is required for proper ciliary trafficking and mutations in its C-terminus are associated with skeletal ciliopathies. The authors begin by performing a pull-down to identify binding partners of His-tagged CrIFT172968-C in Chlamydomonas reinhardtii flagella. Interactions with three candidates (IFT140, IFT144, and a UBX-domain containing protein) are validated by AlphaFold Multimer with the IFT140 and IFT144 predictions in agreement with published cryo-ET structures of anterograde and retrograde IFT trains. They present a crystal structure of IFT172c and find that a part of the C-terminal domain of IFT172 resembles the fold of a non-canonical U-box domain. As U-box domains typically function to bind ubiquitin-loaded E2 enzymes, this discovery stimulates the authors to investigate the ubiquitin-binding and ubiquitination properties of IFT172c. Using in vitro ubiquitination assays with truncated IFT172c constructs, the authors demonstrate partial ubiquitination of IFT172c in the presence of the E2 enzyme UBCH5A. The authors also show a direct interaction of IFT172c with ubiquitin chains in vitro. Finally, the authors demonstrate that deletion of the U-box-like subdomain of IFT172 impairs ciliogenesis and TGFbeta signaling in RPE1 cells.

However, some of the conclusions of this paper are only partially supported by the data, and presented analyses are potentially governed by in vitro artifacts. In particular, the data supporting autoubiquitination and ubiquitin-binding are inconclusive. Without further evidence supporting a ubiquitin-binding role for the C-terminus, the title is potentially misleading.

Strengths:

(1) The pull-down with IFT172 C-terminus from C. reinhardtii cilia lysates is well performed and provides valuable insights into its potential roles.

(2) The crystal structure of the IFT172 C-terminus is of high quality.

(3) The presented AlphaFold-multimer predictions of IFT172c:IFT140 and IFT172c:IFT144 are convincing and agree with experimental cryo-ET data.

Weaknesses:

(1) The crystal structure of HsIFT172c reveals a single globular domain formed by the last three TPR repeats and C-terminal residues of IFT172. However, the authors subdivide this globular domain into TPR, linker, and U-box-like regions that they treat as separate entities throughout the manuscript. This is potentially misleading as the U-box surface that is proposed to bind ubiquitin or E2 is not surface accessible but instead interacts with the TPR motifs. They justify this approach by speculating that the presented IFT172c structure represents an autoinhibited state and that the U-box-like domain can become accessible following phosphorylation. However, additional evidence supporting the proposed autoinhibited state and the potential accessibility of the U-box surface following phosphorylation is needed, as it is not tested or supported by the current data.

We thank the reviewer for this comment. IFT172C contains TPR region and Ubox-like region, which are admittedly tightly bound to each other. While there is a possibility that this region functions and exists as one domain, below are the reasons why we chose to classify these regions as two different domains.

(1) TPR and Ubox-like regions are two different structural classes

(2) TPR region is linked to Ubox-like region via a long linker which seems poised to regulate the relative movement between these regions.

(3) Many ciliopathy mutations are mapped to the interface of TPR region and the Ubox region hinting at a regulatory mechanism governed by this interface.

That said, we agree that the proposed autoinhibited state and its potential relief by phosphorylation remains a hypothesis that requires experimental validation. We have revised the manuscript to present this more clearly as a speculative model rather than an established mechanism. We clearly acknowledge this limitation on pg. 16-17 of the revised discussion: ‘The IFT172 U-box domain appears to be in an auto-inhibited state in our crystal structure of HsIFT172C2 (Fig. 2E), potentially explaining the absence of a robust auto-ubiquitination activity in-vitro. This structural inhibition is reminiscent of the RING ubiquitin ligase CBL [59], where phosphorylation and substrate binding trigger a conformational change that activates ligase activity [59,75]. Intriguingly, the phosphosite database [76] lists four residues (T1533, S1549, T1689, Y1691) at the U-box/TPR interface as phosphorylation sites (Fig. S2D). Phosphorylation of these residues could potentially alleviate the auto-inhibited state, suggesting a possible regulatory mechanism. Furthermore, a 30-residue linker connects the U-box domain to the last TPR of IFT172, likely providing significant conformational flexibility (Fig. 2A-B). This flexibility may be functionally crucial for the U-box domain, allowing it to adopt different conformations as needed for its various roles. However, we note that the proposed autoinhibition model and its potential regulation by phosphorylation remain hypothetical and require future experimental validation.

(2) While in vitro ubiquitination of IFT172 has been demonstrated, in vivo evidence of this process is necessary to support its physiological relevance.

We thank the reviewer for this important point. We agree that in vivo evidence of IFT172 ubiquitination would strengthen the physiological relevance of our findings. While our current study focuses on the in vitro characterization of this activity, we have revised the manuscript to more clearly state that demonstration of IFT172 ubiquitination activity in cells, including identification of bona fide substrates, is required to establish its physiological significance (p. 16). We consider this an important direction for future studies.

(3) The authors describe IFT172 as being autoubiquitinated. However, the identified E2 enzymes UBCH5A and UBCH5B can both function in E3-independent ubiquitination (as pointed out by the authors) and mediate ubiquitin chain formation in an E3-independent manner in vitro (see ubiquitin chain ladder formation in Figure 3A). In addition, point mutation of known E3-binding sites in UBCH5A or TPR/U-box interface residues in IFT172 has no effect on the mono-ubiquitination of IFT172c1. Together, these data suggest that IFT172 is an E3-independent substrate of UBCH5A in vitro. The authors should state this possibility more clearly and avoid terminology such as "autoubiquitination" as it implies that IFT172 is an E3 ligase, which is misleading. Similarly, statements on page 10 and elsewhere are not supported by the data (e.g. "the low in vitro ubiquitination activity exhibited by IFT172" and "ubiquitin conjugation occurring on HsIFT172C1 in the presence of UBCH5A, possibly in coordination with the IFT172 U-box domain").

We now consider this possibility and tone down our statements about the autoubiquitination activity of IFT172 in both the abstract and results/discussion parts of the revised version of the manuscript. We no longer refer to IFT172 as having auto-ubiquitination activity in the manuscript.

(4) Related to the above point, the conclusion on page 11, that mono-ubiquitination of IFT172 is U-box-independent while polyubiquitination of IFT172 is U-box-dependent appears implausible. The authors should consider that UBCH5A is known to form free ubiquitin chains in vitro and structural rearrangements in F1715A/C1725R variants could render additional ubiquitination sites or the monoubiquitinated form of IFT172 inaccessible/unfavorable for further processing by UBCH5A.

We agree and the conclusion on pg. 11 has now been changed to: Therefore, while mutations in the IFT172 U-box domain affect the formation of higher molecular weight ubiquitin conjugates, the prominent mono-ubiquitination of IFT172 is likely attributable to the E3-independent activity of UbcH5a, as this event is not impacted by these U-box mutations, rather than indicating an intrinsic auto-ubiquitination capacity of IFT172 itself.

(5) Identification of the specific ubiquitination site(s) within IFT172 would be valuable as it would allow targeted mutation to determine whether the ubiquitination of IFT172 is physiologically relevant. Ubiquitination of the C1 but not the C2 or C3 constructs suggests that the ubiquitination site is located in TPRs ranging from residues 969-1470. Could this region of TPR repeats (lacking the IFT172C3 part) suffice as a substrate for UBCH5A in ubiquitination assays?

We thank the reviewer for raising this important point about ubiquitination site identification. While not included in our manuscript, we did perform mass spectrometry analysis of ubiquitination sites using wild-type IFT172 and several mutants (P1725A, C1727R, and F1715A). As shown in Author response image 1, we detected multiple ubiquitination sites across these constructs. The wild-type protein showed ubiquitination at positions K1022, K1237, K1271, and K1551, while the mutants displayed slightly different patterns of modification. However, we should note that the MS intensity signals for these ubiquitinated peptides were relatively low compared to unmodified peptides, making it difficult to draw strong conclusions about site specificity or physiological relevance.

Author response image 1.

Consistent with the reviewer's suggestion, all detected ubiquitination sites fall within the TPR-containing region (residues 1022-1551), which is present in the C1 construct but absent from C2 and C3, explaining the construct-dependent ubiquitination pattern. We did not test the TPR region alone as a UBCH5A substrate, but this would be an informative experiment for future studies.

(6) The discrepancy between the molecular weight shifts observed in anti-ubiquitin Western blots and Coomassie-stained gels is noteworthy. The authors show the appearance of a mono-ubiquitinated protein of ~108 kDa in anti-ubiquitin Western blots. However, this molecular weight shift is not observed for total IFT172 in the corresponding Coomassie-stained gels (Figures 3B, D, F). Surprisingly, this MW shift is visible in an anti-His Western blot of a ubiquitination assay (Fig 3C). Together, this raises the concern that only a small fraction of IFT172 is being modified with ubiquitin. Quantification of the percentage of ubiquitinated IFT172 in the in vitro experiments could provide helpful context.

We acknowledge that the ubiquitin conjugation of IFT172 in vitro is weak, as stated in the manuscript (p. 16). The discrepancy between anti-ubiquitin Western blots and Coomassie-stained gels is consistent with only a small fraction of IFT172 being modified, which is expected given that the reaction likely reflects E3-independent ubiquitination by UBCH5A rather than a robust enzymatic activity of IFT172 itself. The anti-His Western blot (Fig. 3C) is more sensitive than Coomassie staining, explaining why the shift is visible there but not on Coomassie. We have not performed formal quantification of the ubiquitinated fraction, but based on the Coomassie data, we estimate it to be a minor proportion of total IFT172, consistent with the toned-down conclusions in our revised manuscript. The identification of physiological substrates and in vivo validation will be important future directions to establish the biological relevance of these observations.

(7) The authors propose that IFT172 binds ubiquitin and demonstrate that GST-tagged HsIFT172C2 or HsIFT172C3 can pull down tetra-ubiquitin chains. However, ubiquitin is known to be "sticky" and to have a tendency for weak, nonspecific interactions with exposed hydrophobic surfaces. Given that only a small proportion of the ubiquitin chains bind in the pull-down, specific point mutations that identify the ubiquitin-binding site are required to convincingly show the ubiquitin binding of IFT172.

We appreciate the reviewer's point regarding the potential for non-specific ubiquitin interactions and the value of mutational analysis for confirming specificity. While further mutagenesis of the predicted ubiquitin-binding interface was not performed for this revision, we note that our data show comparable tetra-ubiquitin pull-down by both the larger HsIFT172C2 construct and, importantly, the isolated HsIFT172C3 U-box domain itself (Fig. 4D). This localization of binding to the smaller U-box domain, coupled with our AlphaFold model predicting a specific interface with ubiquitin (Fig. 4E-F) and the observation that a mutation elsewhere (D1605R, Fig. 4C) does not abrogate this binding, collectively suggest a degree of specificity. We have revised the manuscript to more cautiously present these findings and acknowledge the need for future studies to definitively map the binding site. Specifically, we have now toned down the conclusion in the section on pg. 12-13 of the revised manuscript including a toned down heading: “IFT172 U-box domain pulls down ubiquitin in vitro”.

(8) The authors generated structure-guided mutations based on the predicted Ub-interface and on the TPR/U-box interface and used these for the ubiquitination assays in Fig 3. These same mutations could provide valuable insights into ubiquitin binding assays as they may disrupt or enhance ubiquitin binding (by relieving "autoinhibition"), respectively. Surprisingly, two of these sites are highlighted in the predicted ubiquitin-binding interface (F1715, I1688; Figure 4E) but not analyzed in the accompanying ubiquitin-binding assays in Figure 4.

We thank the reviewer for emphasizing the importance of mutational analysis to confirm the specificity of ubiquitin binding and for specifically inquiring about residues like F1715 and I1688 at the predicted ubiquitin interface. We tested purified HsIFT172C1 constructs containing the F1715A mutation (along with P1725A and C1727R variants) in pull-down assays with GST-Ubiquitin, see Author response image 2.

Author response image 2.

However, these experiments did not reveal a conclusive difference in ubiquitin binding for any of the tested variants compared to wild-type IFT172. The I1688A mutant, unfortunately, yielded insoluble protein and could not be evaluated. It is conceivable that the F1715A mutation was not disruptive enough to significantly alter binding, and future studies with different substitutions might be more informative. Nevertheless, our observations that the isolated HsIFT172C3 U-box domain itself pulls down tetra-ubiquitin (Fig. 4D), that our AlphaFold model predicts a specific interface (Fig. 4E-F), and that a mutation elsewhere (D1605R, Fig. 4C) does not abrogate this binding, collectively suggest a degree of specificity. We have revised the manuscript to present these ubiquitin binding findings cautiously, acknowledging the need for further investigation to definitively map the binding site and its functional relevance.

(9) If IFT172 is a ubiquitin-binding protein, it might be expected that the pull-down experiments in Figure S1 would identify ubiquitin, ubiquitinated proteins, or E2 enzymes. These were not observed, raising doubt that IFT172 is a ubiquitin-binding protein.

We acknowledge that the absence of ubiquitin or ubiquitinated proteins in our pull-down/MS experiment (Fig. S1) could raise questions about the ubiquitin-binding capacity of IFT172. However, several technical factors likely explain this. First, IFT172 appears to bind ubiquitin with low affinity, as indicated by our in vitro pull-downs and the AF-predicted interface. Second, we used extensive washes to remove non-specific interactors, which would also remove weak but potentially genuine ubiquitin interactions. Third, we did not include ubiquitination-preserving reagents such as NEM in our pull-down buffers, exposing ubiquitinated proteins to DUB-mediated deubiquitination during the experiment. These factors combined would strongly select against the detection of ubiquitin-related interactors under our experimental conditions.

(10) The cell-based experiments demonstrate that the U-box-like region is important for the stability of IFT172 but does not demonstrate that the effect on the TGFb pathway is due to the loss of ubiquitin-binding or ubiquitination activity of IFT172.

We acknowledge that our current data cannot definitively distinguish whether the TGFβ pathway defects arise from reduced IFT172 protein stability or from specific loss of ubiquitin-related functions of the U-box domain. Our experiments demonstrate that the U-box region is required for both IFT172 stability and proper TGFβ signaling, but we agree that establishing a direct mechanistic link between ubiquitin-binding/conjugation and signaling would require additional experiments such as point mutations that selectively disrupt ubiquitin-related activity without affecting protein stability. We have revised the discussion (p. 18-19) to more clearly acknowledge this limitation. Addition to text: “However, we note that our current experiments cannot distinguish whether these signaling effects result specifically from loss of ubiquitin-related functions of the U-box domain or from the reduced levels of functional IFT172 protein in the heterozygous U-box deleted cells. Targeted point mutations that selectively disrupt ubiquitin binding without affecting protein stability would be required to resolve this question.”

(11) The challenges in experimentally validating the interaction between IFT172 and the UBX-domain-containing protein are understandable. Alternative approaches, such as using single domains from the UBX protein, implementing solubilizing tags, or disrupting the predicted binding interface in Chlamydomonas flagella pull-downs, could be considered. In this context, the conclusion on page 7 that "The uncharacterized UBX-domain-containing protein was validated by AF-M as a direct IFT172 interactor" is incorrect as a prediction of an interaction interface with AF-M does not validate a direct interaction per se.

We agree with the reviewer that our AlphaFold-Multimer (AF-M) predictions alone do not constitute experimental validation of a direct interaction. We appreciate the reviewer's understanding of the technical challenges in validating this interaction experimentally. We have revised our text (p. 7) to state that "The uncharacterized UBX-domain-containing protein was predicted by AF-M as a potential direct IFT172 interactor" and discuss the AF-M predictions as computational evidence that suggests, but does not prove, a direct interaction.

Reviewer #2 (Public review):

Summary:

Cilia are antenna-like extensions projecting from the surface of most vertebrate cells. Protein transport along the ciliary axoneme is enabled by motor protein complexes with multimeric so-called IFT-A and IFT-B complexes attached. While the components of these IFT complexes have been known for a while, precise interactions between different complex members, especially how IFT-A and IFT-B subcomplexes interact, are still not entirely clear. Likewise, the precise underlying molecular mechanism in human ciliopathies resulting from IFT dysfunction has remained elusive.

Here, the authors investigated the structure and putative function of the to-date poorly characterised C-terminus of IFT-B complex member IFT172 using alpha-fold predictions, crystallography and biochemical analyses including proteomics analyses followed by mass spectrometry, pull-down assays, and TGFbeta signalling analyses using chlamydomonas flagellae and RPE cells. The authors hereby provide novel insights into the crystal structure of IFT172 and identify novel interaction sites between IFT172 and the IFT-A complex members IFT140/IFT144. They suggest a U-box-like domain within the IFT172 C-terminus could play a role in IFT172 auto-ubiquitination as well as for TGFbeta signalling regulation.

As a number of disease-causing IFT72 sequence variants resulting in mammalian ciliopathy phenotypes in IFT172 have been previously identified in the IFT172 C-terminus, the authors also investigate the effects of such variants on auto-ubiquitination. This revealed no mutational effect on mono-ubiquitination which the authors suggest could be independent of the U-box-like domain but reduced overall IFT172 ubiquitination.

Strengths:

The manuscript is clear and well written and experimental data is of high quality. The findings provide novel insights into IFT172 function, IFT complex-A and B interactions, and they offer novel potential mechanisms that could contribute to the phenotypes associated with IFT172 C-terminal ciliopathy variants.

Weaknesses:

Some suggestions/questions are included in the comments to the authors below.

Reviewer #3 (Public review):

Summary:

Zacharia et al report on the molecular function of the C-terminal domain of the intraflagellar transport IFT-B complex component IFT172 by structure determination and biochemical in vitro and cell culture-based assays. The authors identify an IFT-A binding site that mediates a mutually exclusive interaction to two different IFT-A subunits, IFT144 and IFT140, consistent with interactions suggested in anterograde and retrograde IFT trains by previous cryo-electron tomography studies. Additionally, the authors identify a U-box-like domain that binds ubiquitin and conveys ubiquitin conjugation activity in the presence of the UbcH5a E2 enzyme in vitro. RPE1 cell lines that lack the U-box domain show a reduction in ciliation rate with shorter cilia, and heterozygous cells manifest TGF-beta signaling defects, suggesting an involvement of the U-box domain in cilium-dependent signaling.

Strengths:

(1) The structural analyses of the C-terminal domain of IFT172 combine crystallography with structure prediction using state-of-the-art algorithms, which gives high confidence in the presented protein structures. The structure-based predictions of protein interactions are validated by further biochemical experiments to assess the specific binding of the IFT172 C-terminal domains with other proteins.

(2) The finding that the IFT172 C-terminus interactions with the IFT-A components IFT140 and IFT144 appear mutually exclusive confirm a suggested role in mediating the binding of IFT-B to IFT-A in anterograde and retrograde IFT trains, which is of very high scientific value.

(3) The suggested molecular mechanism of IFT train coordination explains previous findings in Chlamydomonas IFT172 mutants, in particular an IFT172 mutant that appeared defective in retrograde IFT, as well as mutations identified in ciliopathy patients.

(4) The identification of other IFT172 interactors by unbiased mass spectrometry-based proteomics is very exciting. Analysis of stoichiometries between IFT components suggests that these interactors could be part of IFT trains, either as cargos or additional components that may fulfill interesting functions in cilia and flagella.

(5) The authors unexpectedly identify a U-box-like fold in the IFT172 C-terminus and thoroughly dissect it by sequence and mutational analyses to reveal unexpected ubiquitin binding and potential intrinsic ubiquitination activity.

(6) The overall data quality is very high. The use of IFT172 proteins from different organisms suggests a conserved function.

Weaknesses:

(1) Interaction studies were carried out by pulldown experiments, which identified more IFT172 interaction partners. Whether these interactions can be seen in living cells remains to be elucidated in subsequent studies.

We agree with the reviewer that validation of protein-protein interactions in living cells provides important physiological context. While our pulldown experiments have identified several promising interaction partners and the AF-M predictions provide computational support for these interactions, we acknowledge that demonstrating these interactions in vivo would strengthen our findings. However, we believe our current biochemical and structural analyses provide valuable insights into the molecular basis of IFT172's interactions, laying important groundwork for future cell-based studies.

(2) The cell culture-based experiments in the IFT172 mutants are exciting and show that the U-box domain is important for protein stability and point towards involvement of the U-box domain in cellular signaling processes. However, the characterization of the generated cell lines falls behind the very rigorous analysis of other aspects of this work.

We thank the reviewer for noting that the characterization of our cell lines could be more rigorous. In the revised version of the manuscript, we have addressed this by providing additional validation data for all four engineered RPE1 cell lines. First, we performed Sanger sequencing to confirm precise in-frame integration of the GFP tag at the targeted loci and to exclude unintended insertions or deletions (indels), both for the full-length IFT172-eGFP lines (Fig. S6) and for the IFT172∆U-box-eGFP lines (Fig. S7). Second, we performed anti-IFT172 immunoblotting on all four cell lines alongside parental RPE1 cells, confirming expression of both the full-length and U-box-truncated IFT172 proteins (Fig. S8). Notably, the immunoblot revealed reduced steady-state levels of the IFT172∆U-box protein compared to full-length IFT172, providing direct biochemical evidence that loss of the U-box domain compromises IFT172 protein stability consistent with the ciliogenesis phenotype described in the main text. Together, these data verify the integrity of the edited loci at both the genomic and protein levels, and strengthen the validation of the cellular models used in this study.

Overall, the authors achieved to characterize an understudied protein domain of the ciliary intraflagellar transport machinery and gained important molecular insights into its role in primary cilia biology, beyond IFT. By identifying an unexpected functional protein domain and novel interaction partners the work makes an important contribution to further our understanding of how ciliary processes might be regulated by ubiquitination on a molecular level. Based on this work it will be important for future studies in the cilia community to consider direct ubiquitin binding by IFT complexes.

Conceptually, the study highlights that protein transport complexes can exhibit additional intrinsic structural features for potential auto-regulatory processes. Moreover, the study adds to the functional diversity of small U-box and ubiquitin-binding domains, which will be of interest to a broader cell biology and structural biology audience.

Additional comments:

The authors investigate the consequences of the U-box deletion on ciliary TGF-beta signaling. While a cilium-dependent effect of TGF-beta signaling on the phosphorylation of SMAD2 has been demonstrated, the precise function of cilia in AKT signaling has not been fully established in the field. Therefore, the relevance of this finding is somewhat unclear. It may help to discuss relevant literature on the topic, such as Shim et al., PNAS, 2020.

We appreciate the reviewer's comment highlighting that the role of primary cilia in AKT signaling is not as well established as for SMAD2/3. However, we note that a direct functional link between AKT signaling and ciliogenesis has been demonstrated, showing that AKT regulates ciliogenesis initiation through a Rab11-effector switch mechanism (Walia et al., 2019; PMID: 31204173, co-authored by the corresponding author of this study). Furthermore, Shim et al. (PMID: 33753495) demonstrated a cilia-dependent reciprocal activation of AKT1 and SMAD2/3. In the revised manuscript (p. 19, ref. 97), we have expanded the discussion to cite these studies and provide a clearer literature context for the cilia-AKT connection, while acknowledging that the precise mechanism by which the IFT172 U-box domain influences AKT activation requires further investigation.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Points for the discussion:

(1) The discussion should mention that IFT-A subunits IFT121, IFT122 and IFT144 share a similar domain organization to IFT172 (TPRs terminating in Zn-finger-like domains). Do the authors consider these as potential ubiquitin-binding proteins with E3 ligase activity? The possibility that these Zn-finger-like regions share a common origin, and function to stabilize the proteins or mediate IFT subunit interactions without a role in ubiquitin biology should be considered.

We appreciate this important point. We agree that the shared domain architecture across IFT121, IFT122, IFT144, and IFT172 raises the question of whether these C-terminal domains primarily serve structural rather than ubiquitin-related roles. We have added a discussion paragraph (p. 16) acknowledging that a structural/stabilizing function is the more parsimonious explanation, while noting that whether IFT172's U-box-like domain has additionally acquired ubiquitin-related activity remains an open question.

(2) From their modeling data, do the authors have an explanation for why a substitution as conservative as D1605E would cause disease?

The D1605E substitution maps to the IFT172-IFT-A interaction interface (Fig. 1F). While this is a conservative change, D1605 is located at a tightly packed protein-protein interface where even the addition of a single methylene group (the difference between aspartate and glutamate) could introduce steric clashes with residues of IFT140 or IFT144, or alter the precise geometry of hydrogen bonds or salt bridges critical for the interaction. Unfortunately, this level of detail is beyond the resolution of AlphaFold models. However, the fact that this residue is positioned directly at the binding interface provides a plausible structural rationale for its pathogenicity.

(3) The authors speculate that the L1615P mutation in the Chlamydomonas fla11 strain causes a faulty switch to retrograde IFT and this provides a molecular basis for the retrograde IFT phenotype. However, because the mutation is also within the IFT144 binding site, why is anterograde IFT also not affected?

The fla11 L1615P mutation resides in helix αA, which participates in both IFT144 (anterograde) and IFT140 (retrograde) interactions. The predominantly retrograde phenotype can be rationalized by the fundamentally different structural roles of the IFT172 C-terminus in anterograde versus retrograde trains. In anterograde trains, the IFT172 C-terminus acts as a flexible tether in stoichiometric excess (2:1 IFT-B:IFT-A ratio), providing an avidity effect that likely compensates for reduced binding affinity caused by L1615P (Lacey et al., 2023). Additional lateral interactions between IFT-B subunits further stabilize the anterograde polymer independently of the IFT172-IFT144 link. In contrast, the retrograde train requires the IFT172 C-terminus to adopt a rigid, resolved conformation that is integral to the IFT-A dimeric interface, with no redundant lateral interactions to compensate (Lacey et al., 2024). The helix-breaking L1615P mutation would specifically disrupt this precise structural requirement, explaining the selective retrograde IFT defect in fla11. We have added this discussion to the revised manuscript (p. 16).

Minor:

(1) On page 5, the authors describe the fla11 phenotypes including accumulation of IFT particles at the tip and accumulation of ubiquitinated proteins in the cilium. Could the authors please expand on how this suggests that IFT172 could be involved in ciliary ubiquitination events and discuss an alternative scenario of impaired assembly of functional retrograde IFT in this strain leading to accumulation of ubiquitinated proteins?

In the revised manuscript (p. 16), we have expanded the discussion of the fla11 phenotype to address this point. We now discuss how the distinct structural roles of the IFT172 C-terminus in anterograde versus retrograde trains explain the selective retrograde IFT defect in fla11, and explicitly note that the accumulation of ubiquitinated proteins in fla11 cilia may reflect impaired retrograde IFT-mediated clearance rather than a direct role of IFT172 in ciliary ubiquitination.

(2) The authors should also expand on the literature of known UBX-IFT interactions in their manuscript (e.g. Raman et al. PMID 26389662).

We have expanded the discussion of UBX-IFT interactions in the revised manuscript (p. 7) by citing the work of Raman et al. (PMID 26389662), who identified a direct interaction between the UBX-domain protein UBXN10 and IFT-B via CLUAP1/IFT38 for VCP-mediated regulation of IFT complex integrity. This provides important context for our identification of a UBX-domain protein as an IFT172 interactor.

(3) On page 11, I1688 is incorrectly referred to as I688.

Fixed.

Reviewer #2 (Recommendations for the authors):

(1) The finding that the interaction with IFT140/144 is mutually exclusive is very interesting. Could you speculate on or do you have any data regarding the effects to the overall IFT-complex conformation and downstream biological effects depending on which partner is bound?

I am not a structural biologist so this may be an irrelevant/impossible-to-answer question: I was also wondering as Ref 46 has shown that the dynein-2 motor complex binds to the edge of IFT-B2 (for assembled trains): Could the IFT172 C-terminus be involved here or somehow influence this interaction? In your mass spec data from Cr cilia using CrIFT172_968-C you don`t mention pulling down dynein-2 components so there doesn`t seem to be a direct interaction, but could the IFT-B2 conformation depend on if IFT172 has bound IFT-140 or IFT144 and hence this interaction influence the dynein-2 binding?

We thank the reviewer for this insightful question. Based on recent cryo-ET structures of anterograde and retrograde IFT trains (Lacey et al., 2023; 2024), the switch from IFT144 to IFT140 binding fundamentally changes IFT172's structural role. In anterograde trains, the IFT172 C-terminus acts as a flexible tether tolerating the 2:1 IFT-B:IFT-A stoichiometry and permitting long polymer formation. In retrograde trains, it adopts a rigid conformation integral to the IFT-A dimeric interface, driving the formation of discrete retrograde units with distinct architecture.

Regarding Dynein-2: while IFT172 does not directly bind Dynein-2 (consistent with our MS data), the reviewer's intuition is correct that IFT172's binding partner influences Dynein-2 association. In anterograde trains, autoinhibited Dynein-2 binds a composite surface formed between adjacent IFT-B2 repeats. When IFT172 switches to IFT140 at the ciliary tip, the resulting train depolymerization destroys this composite binding site, releasing Dynein-2 from its cargo mode to function as an active retrograde motor. The IFT172 binding switch may thus indirectly acts as a structural checkpoint for Dynein-2 activation.

(2) The data provided regarding TGFbeta signalling effects in cells with heterozygous U-box-like domain deletions is interesting. While secondary effects of impaired ciliogenesis due to homozygous deletion of the U-box-like domain can cause difficulties to analysing cell signalling effects, it would still be interesting to check the effects of bi-allelic human IFT172 disease variants in this region as well (the human disease phenotype is recessive and human mutations are likely hypomorphic variants still allowing for ciliogenesis).

Also, while there may be secondary effects, it would still be interesting to check homozygous U-box deleted cells as an aggravated effect would further support the data from the het cells.

We agree that testing bi-allelic human disease variants would strengthen the physiological relevance of our findings. While generating knock-in RPE1 lines was beyond the scope of this revision, we have obtained preliminary data from patient-derived fibroblasts carrying bi-allelic IFT172 missense variants in the U-box region (NPH2161). TGF-β1 stimulation time courses in these fibroblasts show altered p-SMAD2 kinetics compared to control fibroblasts, consistent with the phenotype observed in our heterozygous U-box deleted RPE1 cells (see Author response image 3).

Author response image 3.

While these results are preliminary and require further replication, they support the involvement of the IFT172 U-box domain in TGF-β signaling regulation in a disease-relevant context. Regarding homozygous U-box deleted cells, the severe reduction in IFT172 protein levels and ciliogenesis defects (Fig. 5B,D) make it difficult to separate U-box-specific effects from secondary consequences of impaired cilia formation, as the reviewer notes. We consider this an important direction for future studies using targeted point mutations rather than domain deletions.

(3) Figure 5 E-G: Overall, the effects upon TGFB1 addition are rather small compared to previously published data eg Clement et al Cell reports 2013 where one of the authors is the senior. Are RPE cells less responsive or do you have another theory? Did you check TGFB receptor levels to ensure the differences are not due to different levels of receptor expression? I feel it could be interesting to also check ciliary phopsho-SMAD localisation by IF. In Clement et al, loss of IFT88 results in reduced phospho-SMAD2 levels, do you have any theory why these opposite effects compared to the IFT172 loss of function could occur?

We thank the reviewer for this insightful comment. The Tg737orpk fibroblasts used in Clement et al. (2013), which harbor a hypomorphic mutation in IFT88, exhibit severely stunted cilia. This defect broadly disrupts cilium-dependent signaling pathways, including R-SMAD activation, and is therefore expected to produce more pronounced signaling phenotypes. In contrast, our study utilizes RPE-1 cells with structurally intact cilia, enabling us to investigate more specific alterations in ciliary signaling associated with IFT172 function rather than the global effects of cilia loss. Consequently, the more modest effects observed in our system are consistent with the less severe structural and functional perturbation. Both fibroblasts and RPE-1 cells are known to express TGF-β receptors and to respond robustly to TGF-β stimulation, making it unlikely that differences in receptor abundance alone account for the observed discrepancies. We also note that increasing evidence supports a role for the primary cilium in fine-tuning TGF-β signaling output by coordinating both canonical (R-SMAD-mediated) and non-canonical (e.g., AKT/ERK-mediated) pathways. Our data raise the possibility that loss of the IFT172 U-box domain, or reduced IFT172 levels, may differentially affect this balance, rather than simply attenuating signaling uniformly, as seen with more severe ciliary defects such as IFT88 disruption in Tg737orpk cells. We agree that the current dataset does not fully resolve the underlying mechanism. We therefore consider it an important direction for future work to examine, in greater detail, the localization and phosphorylation status of key canonical and non-canonical signaling components in context of the primary cilium by IF analyses.

(4) In the summary conclusion at the end of the discussions, the authors propose that IFT72 could directly influence the fate of ubiquitinated TGFB receptors. Do you have any data supporting the theory that TGFB ubiquitination is influenced by IFT172 ?

We acknowledge that our current data are insufficient to establish a direct link between IFT172-dependent ubiquitination events and TGF-β receptor regulation. Accordingly, we have revised the Discussion (page 19) to remove our previous hypothesis proposing a role for IFT172 in modulating TGF-β receptor ubiquitination.

While our experiments demonstrate that the U-box region is required for both IFT172 stability and proper TGF-β signaling, we agree that establishing a direct mechanistic connection between ubiquitin-related activity of IFT172 and signaling outcomes would require additional approaches such as targeted point mutations that selectively disrupt ubiquitin-binding or conjugation functions.

Furthermore, we note that our current data do not allow us to distinguish whether the observed signaling phenotypes arise specifically from the loss of ubiquitin-related functions of the U-box domain or from reduced levels of functional IFT172 protein in the heterozygous U-box–deleted cells.

(5) Wording:

Abstract

"IFT72..is associated with several disease variants causing ciliopathies". I would change this to "..and several disease-causing IFT172 variants have been identified in ciliopathy patients".

Corrected.

Introduction

"Another cohort of patients with milder ciliopathy resembling BBS also presented with ...". I would reword this to "Another cohort of patients with phenotypically slightly different ciliopathy features resembling BBS also presented with ...". It`s not necessarily less severe (they may die of cardiovascular complications in their early thirties for example due to metabolic syndrome, they are intellectually impaired, become blind...), but rather different.

Changed according to the reviewer’s recommendations.

Reviewer #3 (Recommendations for the authors):

(1) Recommended modifications:

(a) The RPE lines generated should be described better, i.e. sequencing information should be provided, or some kind of evidence that the lines are what they are supposed to be.

As also noted above, we acknowledge that the characterization presented for the RPE cell lines was insufficient in the initial version of the manuscript. In the revised version, we have addressed this limitation by including detailed sequencing analyses to validate the modifications introduced. Specifically, we provide sequencing data confirming both the integration of the GFP tag and the successful deletion of the U-box domain in all four engineered RPE cell lines. These data verify the integrity of the edited loci and exclude the presence of unintended insertions or deletions at the targeted regions. The corresponding results are presented in Figures S6 and S7 of the revised manuscript, thereby strengthening the validation of the cellular models used in this study.

(b) It would be more convincing if more than one clone of the RPE lines were presented, as this could rule out possible clonal effects.

We acknowledge that only a single clone was characterized for each of the four genotypes (IFT172-FL homozygous, IFT172-FL heterozygous, IFT172∆U-box homozygous, IFT172∆U-box heterozygous), and we agree that independent clones would provide stronger protection against clonal artifacts. Generating and validating additional clones was not feasible within the scope of this revision. However, several features of our data mitigate this concern. First, the phenotypes scale with allele dosage: the homozygous ∆U-box line shows the strongest reduction in IFT172 protein level, ciliation, and cilium length, while the heterozygous line shows intermediate defects (Fig. 5B, D and Fig. S8). A clonal off-target effect would not be expected to produce this dose-dependent pattern across two independently isolated lines. Second, the reduced steady-state IFT172 level in the ∆U-box lines (Fig. S8) is consistent with our in vitro observation that the U-box/TPR interface is required for protein stability, providing an independent biochemical rationale for the cellular phenotype. Third, Sanger sequencing of all four lines confirmed precise in-frame integration with no indels at the targeted locus (Figs. S6, S7). We have added a sentence to the Discussion (p. 20) acknowledging that confirmation in additional independent clones remains an important goal for follow-up work.

(c) Figure 5C: distribution of the GFP-tagged IFT172∆U-box protein could be quantified to support the statement.

In the revised version of the manuscript, we have included additional quantification of GFP fluorescence across all four cell lines to support our conclusions regarding IFT172 ciliary localization. The corresponding data for each cell line are presented in Figure S5C–F.

(d) The final sentences include quite bold statements about a general function of IFT172 in signal regulation. Yet, the evidence is the weakest part of the work. It is only shown in i) one cell line, ii) in one cell clone that is not extensively characterized, and iii) for one signaling pathway that is not the best-studied cilia signaling pathway. Therefore, I recommend a more moderate statement.

Abstract last sentence has now been toned down and reads: Our findings suggest that IFT172, beyond its structural role in bridging IFT-A and IFT-B complexes within IFT trains, harbors a conserved U-box-like domain with potential involvement in ciliary ubiquitination processes and signaling, providing new insights into the molecular mechanisms underlying IFT172-related ciliopathies.

(e) The order of the figures is not followed in the main text, which is distracting.

The order of figures is now consecutive in the revised manuscript.

(2) Questions and comments to consider:

(a) It is unclear why tetra-ubiquitin chains have been used.

We thank the reviewer for this question. Recent evidence suggests that ubiquitin chains, rather than monomeric ubiquitin, act as sorting and signaling cues at the primary cilium (Shinde et al., 2020). To probe the ubiquitin-binding activity of IFT172, we therefore used a tetrameric ubiquitin chain as a model substrate, which better reflects the multivalent nature and binding avidity expected for physiological polyubiquitin signals than a ubiquitin monomer. Specifically, we used a recombinantly expressed linear (Met1-linked) tetra-ubiquitin chain, generated as a genetically encoded fusion. Linear ubiquitin chains are well-established non-degradative signaling chains recognized by a dedicated class of ubiquitin-binding domains, making them a suitable probe for detecting ubiquitin-binding activity outside the canonical proteasomal pathway. In addition, monomeric ubiquitin (~8 kDa) is poorly retained during membrane transfer in Western blotting, which further precluded its reliable use as a probe in our pull-down assays. Together, these considerations motivated the use of tetrameric ubiquitin as a biologically and technically appropriate substrate for assessing IFT172's ubiquitin-binding activity.

(b) Figure 4D: described in the text as "pulldown tetraubiquitin at comparable levels", which is not obvious from the figure presented, it appears reduced by at least 30%.

We thank the reviewer for this observation. As described on page 10 of the manuscript and evident from Figure 4D, the purified GST–HsIFT172C3 construct underwent substantial proteolytic cleavage during purification. This degradation limited our ability to include amounts of intact GST–HsIFT172C3 comparable to those of the full-length GST–HsIFT172C2 construct in the pull-down assays. Importantly, when accounting for the reduced proportion of full-length GST–HsIFT172C3 present in the assay, the observed differences in tetra-ubiquitin pull-down efficiency between the two constructs are expected to be comparable. This is supported by the Coomassie staining shown in Figure 4D, which reflects the relative abundance of the intact protein species used in the experiment.

(c) With the proposed model, why would the fla11 mutant only affect retrograde IFT?

We have revised our manuscript in page 16 of the discussion section providing a plausible explanation of why only retrograde IFT is affected in the fla11 mutant.

(3) Minor copy-editing:

(a) Page 3, first paragraph: led := leads.

(b) Kinesin-2 and Dynein-2 should be hyphenated.

(c) Page 4: wwp1 should be WWP1.

(d) Bonafide should be italicized: bona fide.

(e) Some abbreviations appear uncommon and therefore somewhat distracting: TGFB instead of TGF-beta, Cr in instances where specifically referred to the organism.

(f) Unprecise lab jargon: "very C-terminal".

(g) Lab jargon: "purified a C-terminal construct".

(h) Lab jargon: "pull-downs".

(i) Page 8: "DALI" only abbreviated.

(j) Page 9: "Appearance ... were observed" should be "was".

(k) Page 11: "I688" should be "I1688".

(l) Page 12: "PDs" unclear.

These minor points have been corrected.

We have revised the text and figures to ensure using the widely accepted nomenclature, using TGF-β to refer to the signaling pathway and TGF-β1 specifically when referring to the ligand.

We further revised the text to reflect the use “Chlamydomonas reinhardtii” in instances when referring to the organism and “Cr” when referring to the protein.

We have removed the informal phrases "very C-terminal" and "purified a C-terminal construct" from the revised manuscript. We have retained the term "pull-down," as this is well-established and widely used terminology in the biochemistry literature to describe the affinity-based co-isolation assays used here. PD has been replaced with pull-down.

The grammatical error on page 9 ("Appearance... “were observed") has been corrected to "was observed”.

eLife Assessment

This valuable study reports the architectural reorganization of the uterine luminal epithelium during the implantation period. The data presented are solid, although improvements are needed. This work is of interest to reproductive biologists and physicians practicing reproductive medicine.

Reviewer #1 (Public review):

Summary:

This manuscript asks how the uterine lumen is remodeled across the peri-implantation window and whether this remodeling is functionally linked to embryo attachment and subsequent pregnancy establishment. The authors combine whole-organ three-dimensional imaging of optically cleared mouse uteri with single-cell and spatial transcriptomic profiling, conditional deletion of p38α at the uterine-wide versus epithelial-restricted level, and rescue experiments using progesterone and leukemia inhibitory factor. Based on these datasets, the authors propose that the luminal epithelium undergoes a previously underappreciated phase of organ-scale architectural reorganization before attachment, and that a p38α-dependent stress-responsive program coordinates epithelial remodeling together with epithelial-stromal communication required for implantation competence.

Strengths:

By moving beyond local attachment-site morphology to a horn-level representation of luminal topology, the work provides anatomical context that is difficult to reconstruct from conventional section-based approaches and should be broadly useful to the implantation community. The integration of organ-scale morphology with single-cell and spatial transcriptomic datasets, together with genetic perturbation and rescue experiments, adds breadth and increases the potential long-term utility of the dataset for investigators interested in uterine receptivity and embryo-uterine interactions.

Weaknesses:

(1) The whole-uterus analysis of luminal folds and creases requires stronger methodological support. Given the mechanical compliance of the uterine lumen, it is difficult to evaluate from the current description whether dissection, fixation, clearing, and/or mounting could influence the observed luminal topography. This issue is particularly important because several key conclusions depend on the spatial distribution of folds across the uterine horn. A fuller account of tissue handling and reconstruction, together with validation that the preparation preserves native morphology, would substantially strengthen confidence in the organ-scale conclusions.

(2) Several of the central morphological claims are supported primarily only by representative reconstructions. This includes the proposed flattening/creasing dynamics, alternating stretched and shrunken regions, persistence of abnormal folding in the mutant uterus, and the extent of structural rescue following progesterone supplementation. The authors could extract objective measures from the reconstructed luminal surface and provide more statistical analysis to demonstrate the reproducibility of the results.

(3) The manuscript appears to over-reach in concluding that luminal remodeling zones embryos before attachment from day 4 to 5. As presented, the data support a correlation between luminal architecture and embryo position, but do not discriminate between (i) luminal remodeling directing embryo positioning, (ii) embryos locally shaping the lumen, or (iii) parallel regulation of both. The evidence is based on observations of the uterus and the inside blastocysts at certain time points around implantation. Without the time-lapse analysis within the uterus, the dynamic interactions between embryos and the uterus couldn't be determined.

(4) The key conclusion of the manuscript is that uterine p38α regulates luminal epithelial remodeling required for embryo attachment, as shown in the title. Against this background, the finding that epithelial-restricted loss of p38α does not overtly impair fertility is notable, as it suggests that the major function of p38α may not be epithelial cell-autonomous but instead may arise through other uterine compartments that secondarily influence the epithelium. At present, however, this conclusion remains insufficiently supported: the epithelial-specific model is not characterized in sufficient depth during the peri-implantation period, and the transcriptomic evidence for altered epithelial-stromal communication does not by itself explain the phenotypic difference between uterine-wide and epithelial-specific deletion. If stromal p38α is proposed as the critical upstream regulator, more direct testing, such as stromal-specific deletion, would be needed.

Reviewer #2 (Public review):

Summary:

In this study, the authors aimed to characterize the architectural reorganization of the uterine luminal epithelium during the implantation period. Using 3D histological reconstruction, single-cell RNA sequencing, and spatial transcriptomics, the authors characterize luminal remodeling during the peri-implantation period and employ a mouse model to explore the role of p38α in regulating luminal flattening.

Strengths:

This study clearly described the changes in luminal architecture during implantation. Moreover, they also used integration of multiple advanced techniques, including 3D tissue reconstruction, single-cell transcriptomics, and spatial transcriptomics, which together provide a detailed description of the molecular characteristics of the uterine architecture during implantation.

Weaknesses:

The authors used PR-Cre to generate uterine p38α knockout mice. This Cre driver deletes p38α not only in epithelial cells but also in stromal compartments. Therefore, it remains unclear whether the observed phenotype arises from epithelial cells, stromal cells, or a combination of both. Previous studies have shown that p38α regulates epithelial polarity, cytoskeletal organization, and E-cadherin localization. However, the current study does not examine changes in cell adhesion or epithelial junction integrity. Previous studies have reported that uterine fluid absorption during implantation is closely associated with luminal closure and remodeling. It would be important to determine whether epithelial transport-related genes are altered in the mutant uterus. Could dysregulated fluid homeostasis contribute to the implantation defects observed in the p38α-deficient mice?

eLife Assessment

This manuscript provides valuable high-resolution structural insights into the interaction between vaccine-elicited antibodies and SARS‑CoV‑2 evolution. The evidence is solid; however, the conclusions could be strengthened with further experimentation and analysis.

Reviewer #1 (Public review):

Summary:

The authors provide high-resolution cryoEM structures to map and functionally characterize human antibodies against SARS-CoV-2 elicited by a standard mRNA vaccine. Here, they report high-resolution structural information on seven previously documented neutralizing antibodies from this response, which were produced from early plasmablasts and which engage diverse targets on the viral spike glycoprotein. This structural information is then integrated with functional assays to define how antibody epitope specificity, geometry, and conformational dynamics may shape neutralization outcomes.

Strengths:

A core strength of the study is a technically-well executed analysis of multiple 'ectopically balanced' mAbs elicited by early B cell plasmablast responses. These antibodies engage different neutralizing targets on the S-trimer of SARS-CoV-2, including the RBD and NTD domains. This has resolved a core distinction in terms of how nAbs engaging these features (and subfeatures, e.g., more conserved hydrophobic pocket within NTD) neutralize the virus.

Weaknesses:

A general weakness is that these antibody classes have been structurally characterized already (albeit individually), and much of this work has been done in the context of understanding susceptibility to escape mutations (delta, omicron, and subvariants therein; class I-IV antibody crossreactivity on Wuhan SARS-CoV-2 to present). It is exceptionally fine technical work presenting the antibodies in a collection like this, but perhaps the new predictive power of this analysis is somewhat overstated.

The early plasmablast angle seems like it could be better fleshed out. Many of the known SARS-CoV-2 nAbs are from the plasmablast pool, but how does this predict the antibody profile at latter stages, as per the stated goal and claim of the current study? Does the paratope pattern of plasmablast antibodies then change within the immune sera at later time points? New or existing cryoEMPEM data could shed light on this.

Reviewer #2 (Public review):

Summary:

This manuscript provides important insights into the interaction between early vaccine-elicited antibodies and SARS‑CoV‑2 evolution. The work will be of broad interest to researchers in structural virology, immunology, and vaccine development. However, several conclusions-particularly those involving neutralization breadth and spike destabilization-require additional functional and biophysical validation.

Strengths:

The manuscript provides an unusually comprehensive structural dataset, resolving all neutralizing antibodies in complex with the SARS‑CoV‑2 spike and enabling direct mechanistic comparison across epitope classes. Its integration of cryo‑EM structures with variant binding, sequence analysis, and fusion‑inhibition assays offers a coherent, multidimensional explanation for antibody breadth and escape. Notably, the identification of a conserved NTD hydrophobic pocket targeted by broad-reactive antibodies represents a conceptually important advance with clear implications for future vaccine design.

Weaknesses:

The study lacks variant-specific neutralization assays, limiting the ability to directly correlate binding breadth with functional viral inhibition. It also omits kinetic affinity measurements, leaving important mechanistic questions, such as why certain antibodies retain breadth, only partially resolved. Additionally, reliance on HEK293T-based spike display raises concerns about glycosylation-related artifacts, especially for NTD loop-dependent antibodies.

Reviewer #3 (Public review):

Summary:

In this manuscript by Jaiswal et al., the authors used structural biology combined with cellular assays to determine the molecular basis underlying the neutralizing ability of the SARS-CoV-2 antibodies. The authors compared the binding mode of the neutralizing antibodies that have two distinct binding interfaces and identified key sites that determine their vulnerability to virus evolution. Interestingly, the author also demonstrated that the trimer-disrupting antibody has the broadest activity as the variations at the trimer interface are limited in evolution.

Strengths:

This manuscript reported a large collection of structures and covered a broad range of binding modes and mechanisms of action. Many of the cryo-EM structures are of good quality. The authors' hypothesis regarding the molecular determinants of evolution vulnerability is solid.

Weaknesses:

However, in my opinion, several points listed below need to be addressed.

(1) At the beginning of the results section, the authors started by determining the structures of Fab PVI.V3-9 and Fab PVI.V6-4 in complex with the ancestral SARS-CoV-2 spike. However, the readers could benefit from a brief introduction of the Fabs PVI.V3-9 and PVI.V6-4. The same applies to the anti-NTD Fabs.

(2) In Figure 1A and E, the spike protein is shown with two different views. It is best to show the same view for comparison.

(3) Throughout the manuscript, the map quality of some Fabs (e.g., V6-11, V6-7, V6-2) is suboptimal. Does the map support the claims on the residues that form the interface? The authors should provide a figure showing the cryo-EM density for all side-chain residues involved at the interface.

(4) Line 152, the terminology "NTD top binders" could be ambiguous, as it could mean those Fabs have the strongest binding affinity. Maybe the authors can change the "top" to "tip".

(5) The authors described the interface between the spike protein and the Fabs in great detail. However, it would be nice if the authors could summarize the common binding strategy for each group of antibodies that utilize the same binding surface.

(6) Line 275, the authors should define what strain of Omicron is in Figure 4. The authors should also explain that the strains in Figure 4A are ordered by evolutionary age.

(7) Lines 286-287, isn't this conclusion already made from the cell-based flow cytometry binding assay? This sentence could be deleted.

(8) In both Figures S10 and S11, the readers could benefit from an additional row highlighting the residues interacting with ACE2.

(9) Lines 298-301, based on Figure S11, no contact is made between the N2 loop and the Fab. The authors may elaborate on why the mutations observed in the N2 loop indirectly influenced Fab recognition.

(10) Lines 321-323, even though this is a well-established assay, it is probably better to clearly explain that one pool of cells expresses the spike and the other pool of cells expresses ACE2.

eLife Assessment

This valuable study compares orthogonal approaches for detecting RNA chemical modifications and provides a helpful framework for improving the reliability of direct RNA sequencing-based identification of RNA modifications. The evidence supporting the technical benchmarking claims is solid. However, support for the broader biological conclusions is not as strong, and the quantitative interpretation of the results, as well as the limitations of the underlying models, would benefit from further clarification.

Reviewer #1 (Public review):

Summary:

The authors set out to evaluate how accurately direct sequencing of RNA can identify and quantify several chemical modifications on RNA molecules, focusing primarily on m6A. A central goal of the work is to compare this approach with an independent chemical-based method (glyoxal and nitrite-mediated deamination of unmethylated adenosines (GLORI), using the same RNA samples, in order to assess reproducibility, false-positive signals, and sensitivity across a range of detection strategies. The authors further aim to demonstrate the biological utility of this approach by applying it to two human cell types, primary human fibroblasts and HD10.6 neurons. While the manuscript also reports detection of additional RNA modifications (pseudouridine and m5C, the depth of analysis and strength of controls are greatest for m6A, which forms the primary focus of the study

Strengths:

A strength of this work is the direct comparison of two distinct measurement approaches performed on the same RNA input material; this has not been done in other recently published benchmarking studies evaluating the utility of the recent direct RNA sequencing for calling m6A. The authors systematically test multiple analysis models and show that, when appropriate filtering is applied, detection of modified sites is reproducible across software versions. The use of synthetic RNA standards and METTL3 inhibitors as negative controls helps to reinforce the overall results.

The data show good agreement between the two methods at higher m6A modification levels, supporting the conclusion that direct RNA sequencing can reliably detect high-confidence modification sites. The authors also demonstrate that this approach can, in principle, provide information at the level of individual RNA variants (although only one example was provided), which is difficult to achieve with short-read methods. The methodology described here is likely to be useful to others seeking to apply similar approaches to identify and quantify m6A. The study also explores the detection of other RNA modifications, which highlights the broader potential of the approach, although these analyses are necessarily more exploratory given the more limited controls and data available.

Weaknesses:

Despite these strengths, several issues limit the interpretation of the results and should be clarified for readers.

First, the authors appropriately address false-positive signals by estimating expected false-positive rates and by quantitatively comparing sequence motif enrichment before and after filtering. These analyses provide important support for the use of stoichiometry-based thresholds and demonstrate that filtering substantially improves specificity. However, even after filtering, a subset of detected sites remains outside the expected sequence context. It therefore remains unclear to what extent these non-canonical sites reflect genuine biology versus residual technical artifacts.

Second, claims regarding the ability of direct RNA sequencing to resolve modification patterns across different RNA variants are supported by very limited evidence. The conclusion that this approach provides superior isoform-level quantification relative to short-read methods is based largely on a single gene example. While this case is interesting, it does not establish how widespread or general this advantage is. A broader analysis indicating how many genes show isoform-specific modification patterns detectable by this method, and how often these are missed by the comparison approach, would be necessary to support a general claim.

Third, the biological interpretation of cell type-specific differences in modification levels remains underdeveloped. Although differences in modification stoichiometry are reported between fibroblasts and neuron-derived cells, the functional consequences of these differences are not addressed. It is unclear whether changes in modification levels are associated with differences in RNA abundance, stability, or translation. As a result, statements suggesting that these modifications fine-tune core cellular pathways are speculative and should either be supported with additional analyses or framed more cautiously.

Related to this point, differences in gene expression between the two cell types are a potential confounding factor. The pathway enrichment patterns presented appear biased toward particular functional categories, but without clear control for differential gene expression, it is difficult to determine whether the observed enrichment reflects cell type-specific regulation of RNA modification or simply differences in which genes are expressed. Clarifying how background gene sets were defined for these analyses would help readers interpret the results.

The manuscript also suggests broader differences in overall modification levels between cell types, but this is not validated using an independent global assay. An orthogonal measurement of total modification levels on polyadenylated RNA (for example, dot blot) would help place site-specific stoichiometry differences in a clearer biological context.

Finally, the effects of the METTL3 inhibitor on these cell types are not fully characterized. While changes in m6A modification patterns are reported following treatment, the manuscript does not address whether the treatment affects cell growth or viability.

Appraisal of conclusions and impact:

Overall, the study provides an informative technical assessment of direct RNA sequencing for modification detection and establishes clear conditions under which the method performs well. The evidence strongly supports conclusions related to technical benchmarking, reproducibility, and the importance of filtering and controls, particularly for m6A. In contrast, conclusions regarding isoform-specific regulation and cell type-specific biological roles of RNA modification are less well supported by the data currently presented, and would benefit from either additional analysis or more restrained interpretation.

The work is likely to have a meaningful impact as a practical reference for researchers using direct RNA sequencing, particularly by clarifying sources of false positives and the value of appropriate controls. With clearer limits placed on biological interpretation or more data presented in support of the biological interpretation, the study would serve as a valuable reference for users seeking to apply these technologies reliably.

Reviewer #2 (Public review):

Summary:

In this study, the authors aim to establish a calibrated framework for detecting RNA modifications using long-read sequencing and apply it to compare modification patterns between fibroblasts and neuron-like cells. The work combines long-read sequencing, in vitro transcribed controls, methyltransferase inhibition, and comparison to an orthogonal sequencing-based method in an attempt to derive filtering strategies that reduce false positive modification calls. The authors further apply this framework to explore differences in modification levels between the two cell types.

The resulting dataset may be of interest to researchers working on RNA modification detection using long-read sequencing technologies. Independent datasets across additional cellular systems can be useful for benchmarking computational methods and evaluating the behavior of modification detection models. However, the conceptual advance of the analytical framework presented here remains somewhat unclear, as many aspects of the analysis closely resemble strategies that have already been described in recent benchmarking studies.

Strengths:

A clear strength of the study is the generation of a relatively large long-read sequencing dataset together with several useful experimental controls, including in vitro transcribed RNA and pharmacological inhibition of the methyltransferase enzyme responsible for installing this modification. These controls are helpful for illustrating the challenges associated with distinguishing high-confidence modification sites from background signals. The inclusion of two different human cellular systems also provides an additional dataset that may be useful for benchmarking and cross-validation in the field. The study addresses a practically relevant question for the community, namely, how to reduce false positive calls in long-read sequencing-based RNA modification analyses.

Weaknesses:

The main weakness of the manuscript is its limited methodological novelty. Much of the analytical framework presented here closely follows benchmarking strategies that have already been described in recent studies of RNA modification detection using long-read sequencing. Several previous studies have evaluated modification-aware basecalling approaches, discussed the need for stringent filtering strategies, and compared long-read sequencing-based predictions with orthogonal mapping approaches. The manuscript would therefore benefit from a deeper engagement with the recent benchmarking literature and a clearer explanation of what conceptual or methodological advance the present study provides beyond these earlier analyses.

A second concern relates to the filtering strategy that forms the core of the proposed workflow. The manuscript applies several thresholds, including modification probability, stoichiometry, and read coverage cutoffs, but it is not clearly explained how these thresholds were determined. It remains unclear whether these cutoffs were derived from statistical calibration, empirical optimization using the presented dataset, or adopted from previous studies. Because the downstream conclusions depend strongly on these filtering choices, a clearer methodological justification would strengthen the work and help readers assess the robustness of the proposed framework.

The interpretation of the comparison between the two modification detection approaches also appears somewhat overstated. Differences between the methods are frequently interpreted as evidence that one approach produces large numbers of false positive calls, but the analyses presented do not fully exclude alternative explanations such as differences in sensitivity, sequencing depth, or methodological biases. A more cautious interpretation of these discrepancies would therefore be appropriate.

Some discussion points also appear speculative. In particular, certain interpretations propose mechanistic explanations without presenting analyses that would allow these possibilities to be distinguished. Such interpretations would benefit from either additional supporting analyses or more cautious phrasing.

From a methodological perspective, the statistical robustness of the thresholds used throughout the analysis could also be discussed in more detail. Given the relatively modest read coverage cutoff applied in the study, low stoichiometry estimates may be strongly influenced by sampling noise, and fixed stoichiometry thresholds may therefore not correspond to a consistent level of confidence across sites. In addition, the manuscript relies heavily on fixed modification probability cutoffs to define high-confidence calls, but it does not discuss whether these scores are statistically calibrated or how they relate to expected error rates. Neural network outputs are often not well-calibrated probabilities, and interpreting these values as direct confidence estimates can therefore be problematic. Finally, modification detection models trained on known modification sites may capture sequence-context patterns present in the training data, meaning that motif enrichment or positional distributions along transcripts may partly reflect model biases rather than purely biological signals. A brief discussion of these limitations would help readers better interpret the robustness of the proposed filtering strategy and the downstream biological conclusions.

Overall, while the dataset may be of interest to the community, the extent to which the study advances current methodological understanding beyond recent benchmarking efforts remains limited.

Minor comments:

The discussion of the "DRACH" versus "all-context" outputs would benefit from greater technical precision. The statement that the number of sites within DRACH motifs identified by the all-context approach was nearly identical to the number reported by the DRACH model may suggest that these outputs derive from fundamentally different predictive models. As I understand it, the underlying neural network is the same, whereas the distinction lies primarily in the classification context. Clarifying this explicitly in the manuscript would improve interpretability and avoid potential confusion for readers.

The manuscript compares results obtained with different basecalling and modification settings but refers primarily to Dorado software versions. This may be misleading, as software version and model version are not necessarily equivalent. Different basecalling or modification models can be used with the same software release, and newer software versions may still use older models. For clarity and reproducibility, the authors should report the exact basecalling and modification model names used in the analyses rather than referring only to the Dorado software version.

Reviewer #3 (Public review):

In this study, the authors aim to establish a calibrated framework for identifying RNA chemical marks from direct RNA sequencing data using a modification-aware basecalling workflow, with a particular focus on N6-methyladenosine. By combining native RNA sequencing with an unmodified control transcriptome, enzyme inhibition, comparison across multiple software versions, and orthogonal validation using an independent mapping approach, the authors seek to define a best-practice pipeline for reducing false-positive calls and improving confidence in quantitative interpretation across cell types.

A major strength of the work is the rigor of the benchmarking strategy. In particular, the inclusion of an unmodified control transcriptome is both important and useful, and the study provides compelling evidence that this control remains necessary for robust interpretation, despite being omitted in many current workflows. The comparison across software versions and the matched analysis with an independent sequencing-based approach also substantially strengthen the evidence presented. The work therefore makes a valuable contribution to the community by offering a more stringent analytical framework that will likely be broadly useful to groups applying native RNA sequencing to study RNA chemical marks.

The evidence supporting the main conclusions is solid overall. The authors convincingly show that stringent filtering substantially reduces false-positive calls and improves agreement with orthogonal approaches, particularly at highly modified sites. The observation that many sites are conserved across cell types, while showing differences in relative modification levels, is also supported by the presented analyses.

At the same time, several conceptual issues limit the strength of some downstream interpretations. Most importantly, the manuscript repeatedly refers to the reported values as "stoichiometry," whereas the underlying software output is more appropriately interpreted as a statistical estimate of the proportion of aligned reads classified as modified. This distinction is important because the conclusions regarding cell-type differences rely on quantitative comparisons of these values. In addition, the current calling framework depends on successful canonical base assignment before modification calling, which raises an important limitation: sites with the strongest signal deviations may be underrepresented if they are more likely to be miscalled during basecalling. This issue may be especially relevant for RNA marks that induce stronger mismatch signatures than N6-methyladenosine and should be more explicitly discussed.

Overall, the authors largely achieve their primary aim of establishing a more rigorous and broadly applicable analytical framework for direct RNA sequencing-based modification detection. The work is likely to have a meaningful impact on the field, particularly by reinforcing the importance of appropriate negative controls and benchmarking standards. With clearer framing of the quantitative outputs and explicit discussion of current software limitations, this study will serve as a highly useful resource for the community.

eLife Assessment

This well-designed study offers important insights into the development of infants' responses to music based on the exploration of EEG neural auditory responses and video-based movement analysis. The compelling results revealed that evoked responses emerge between 3 and 12 months of age, but no age group demonstrated evidence of coordinated movements to music. This study will be of significant interest to developmental psychologists and neuroscientists, as well as researchers interested in music processing and in the translation of perception into action.

Reviewer #1 (Public review):

Summary:

This study aims to investigate the development of infants' responses to music by examining neural activity via EEG and spontaneous body kinematics using video-based analysis. The authors also explore the role of musical pitch in eliciting neural and motor responses, comparing infants at 3, 6, and 12 months of age.

Strengths:

A key strength of the study lies in its analysis of body kinematics and modeling of stimulus-motor coupling, demonstrating how the amplitude envelope of music predicts infant movement, and how higher musical pitch may enhance auditory-motor synchronization.

EEG data provide evidence for enhanced neural responses to music compared to shuffled auditory sequences. These findings ecourage further investigation of the proposed developmental trajectory of neural responses to music and their link to musical behavior in infants.

Comments on revisions:

The authors have addressed my questions in their revision. I have no other questions. Thank you for the opportunity to read and evaluate this interesting study and also for all the work carried out in response to the comments.

Author response:

The following is the authors’ response to the previous reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study aims to investigate the development of infants' responses to music by examining neural activity via EEG and spontaneous body kinematics using video-based analysis. The authors also explore the role of musical pitch in eliciting neural and motor responses, comparing infants at 3, 6, and 12 months of age.

Strengths:

A key strength of the study lies in its analysis of body kinematics and modeling of stimulus-motor coupling, demonstrating how the amplitude envelope of music predicts infant movement, and how higher musical pitch may enhance auditory-motor synchronization.

EEG data provide evidence for enhanced neural responses to music compared to shuffled auditory sequences. These findings ecourage further investigation of the proposed developmental trajectory of neural responses to music and their link to musical behavior in infants.

Comments on revisions:

I thank the authors for the considerable effort devoted to revising the manuscript and addressing the raised questions and comments. I particularly appreciate the additional analyses and the extended arguments included in the discussion. I believe that this paper represents a valuable contribution to the literature on music development.

One remaining comment concerns the evoked response observed in the shuffled condition, which I still find intriguing. Considering that the auditory events in the shuffled condition display a clear rise time, particularly for those events that were selected based on being preceded and followed by longer periods of silence, one would expect to observe an evoked response emerging from baseline. However, this pattern is not evident in the presented curves. The authors may further examine and discuss the shape and characteristics of these response patterns.

We thank the Reviewer for highlighting this intriguing aspect of our data. We entirely agree that from a purely bottom-up, acoustic perspective, one would expect a clear onset-locked evoked response, such as an P1/P2 complex in adults or its developmental equivalent, given the prominent acoustic rise times and the surrounding periods of silence (such as those accounted for in the control analyses)

The fact that these responses are not present in the curves for the shuffled condition was striking to us as well. We interpret this severe attenuation not as a failure of sensory perception, but potentially as a consequence of higher-level cognitive modulation. Specifically, because the shuffled condition completely lacks structural regularities, the brain might be unable to build reliable temporal and/or melodic expectations. In the absence of a learnable structure, the auditory system likely down-weights the processing of these random sequences to conserve cognitive resources, leading participants to attentionally disengage.

This phenomenon aligns with both developmental and adult models of auditory processing. For instance, the "Goldilocks effect" demonstrates that infants systematically withdraw attention from auditory sequences that are entirely unpredictable (Kidd et al., 2014). Similarly, adult auditory literature suggests that while predictable patterns automatically capture attention, random and unpredictable acoustic streams could be actively tuned out (Dayan et al., 2000; Esber & Haselgrove, 2011).

Following the Reviewer’s helpful suggestion to further discuss the characteristics of these response patterns, we have expanded our description and interpretation of the shuffled condition curves in the revised manuscript. We added the following text to the Methods and Discussion to explicitly address the dampened shape of these responses:

p. 9: “Importantly, and in line with the adults’ data, all infant groups exhibited enhanced P1 amplitudes in response to music compared to shuffled music. Actually, across all groups, shuffled music did not elicit clear ERPs as the ones elicited by music”.

p.20: “This process was markedly dampened or interrupted by shuffled music (Bianco et al., 2024, 2025; Lense et al., 2022), a finding that could be interpreted as evidence of disengagement from such highly unpredictable sequences (Dayan et al., 2000; Esber & Haselgrove, 2011; Kidd et al., 2014).”

Reviewer #2 (Public review):

Summary:

Infants' auditory brain responses reveal processing of music (clearly different from shuffled music patterns) from the age of 3 months; however, they do not show related increase in spontaneous movement activity to music until the age of 12 months.

Strengths:

This is a nice paper, well designed, with sophisticated analyses and presenting clear results filling an important gap about early infant sensitivity, detection, and differentiation of musical sounds. The addition of EEG recordings (specifically ERPs) in response to music presentations at 3 different infant ages in the first postnatal year is important, and the manipulation of the music stimuli into shuffled, high and low pitch to capture differences in brain response processing and spontaneous movements is interesting. Further, the movement analysis based on Quantity of Movements (QoM) and movement subdivision into 10 distinct Principal Movements (PMs) is novel and creative.

Overall, results show that ERPs responses to music occurs earlier than QoM in early development, and that even at 12 months, motor responses to music remain coarse and not rhythmically aligned with the music tempo. This work increases our fundamental understanding of infants' early music perception in relation to auditory processing and motor response.

Comments on revisions:

The authors have addressed my questions in their revision. I have no other questions. Thanks again for the opportunity to read and evaluate this interesting work.

We thank the Reviewer for their time, their positive evaluation of our revised manuscript, and their constructive feedback throughout the review process, which has greatly helped us to strengthen this paper.

Reviewer #3 (Public review):

Summary

This study provides a detailed investigation of neural auditory responses and spontaneous movements in infants listening to music. Analyses of EEG data (event-related potentials and steady-state responses) first highlighted that infants at 3, 6 and 12 months of age and adults showed enhanced auditory responses to music than shuffled music. 6-month-olds also exhibited enhanced P1 response to high-pitch vs low-pitch stimuli, but not the other groups. Besides, whole body spontaneous movements of infants were decomposed into 10 principal components. Kinematic analyses revealed that the quantity of movement was higher in response to music than shuffled music only at 12 months of age. Although Granger causality analysis suggested that infants' movement was related to the music intensity changes, particularly in the high-pitch condition, infants did not exhibit phase-locked movement responses to musical events, and the low movement periodicity was not coordinated with music.

Strengths

This study investigates an important topic on the development of music perception and translation to action and danse. It targets a crucial developmental period that is difficult to explore. It evaluates two modalities by measuring neural auditory responses and kinematics, while cross-modal development is rarely evaluated. Overall, the study fills a clear gap in the literature.

Besides, the study uses state-of-the-art analyses. Detailed investigations were performed, as well as exploratory analyses in supplementary information. The discussion is rich in neurodevelopmental interpretations and comparisons with the literature. All steps are clearly detailed. The manuscript is very clear, well-written and pleasant to read. Figures are well-designed and informative. The authors' responses to previous reviews are also detailed and informative.

Comments on revisions:

The authors answered all my questions.

Thank you very much for your positive evaluation and for taking the time to review our revisions. We deeply appreciate your insightful comments across the review rounds, which have helped us improve the clarity and rigor of our paper.

eLife Assessment

This study provides important insights into how tumorous germline stem cells (GSCs) in the Drosophila melanogaster ovary can mimic niche function and suppress the differentiation of neighboring cells. The findings that GSC tumors can incorporate non-mutant cells and inhibit their differentiation are compelling and extend current understanding of stem cell-niche interactions. However, the evidence supporting the conclusion that GSC tumors produce BMP ligands to mediate this effect remains incomplete, due to concerns regarding the quality and interpretation of the HCR-FISH data.

Reviewer #1 (Public review):

Summary:

This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche. This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche.

Strengths:

(1) Use of an excellent and established model for tumorous cells in a stem cell microenvironment

(2) Powerful genetics allow them to test various factors in the tumorous vs non-tumorous cells

(3) Appropriate use of quantification and statistics

Weaknesses:

(1) What is the frequency of SGCs in nos>flp; bam-mutant tumors? For example, are they seen in every germarium, or in some germaria, etc or in a few germaria.

This concern was addressed in the rebuttal. The line number is 106, not line 103.

(2) Does the breakdown in clonality vary when they induce hs-flp clones in adults as opposed to in larvae/pupae?

This concern was addressed in the rebuttal. However, these statements are no on lines 331-335 but instead starting on line 339. Please be accurate about the line numbers cited in the rebuttal. They need to match the line numbers in the revised manuscript.

(3) Approximately 20-25% of SGCs are bam+, dad-LacZ+. Firstly, how do the authors explain this? Secondly, of the 70-75% of SGCs that have no/low BMP signaling, the authors should perform additional characterization using markers that are expressed in GSCs (i.e., Sex lethal and nanos).

The authors did not perform additional staining for GSC-enriched protein like Sex lethal and nanos.

(4) All experiments except Fig. 1I (where a single germarium with no quantification) were performed with nos-Gal4, UASp-flp. Have the authors performed any of the phenotypic characterizations (i.e., figures other than figure 1) with hs-flp?

In the rebuttal, the authors stated that they used nos>flp for all figures except for Fig. 1I. It would be more convincing for them to prove in Fig. 1 than there is not phenoytpic difference between the two methods and then switch to the nos>FLP method for the rest of the paper.

(5) Does the number of SGCs change with the age of the female? The experiments were all performed in 14-day old adult females. What happens when they look at young female (like 2-day old). I assume that the nos>flp is working in larval and pupal stages and so the phenotype should be present in young females. Why did the authors choose this later age? For example, is the phenotype more robust in older females? or do you see more SGCs at later time points?

The authors did not supply any data to prove that the clones were larger in 14-day-old flies than in younger flies. Additionally, the age of "younger" flies was not specified. Therefore, the authors did not satisfactorily answer my concern.

(6) Can the authors distinguish one copy of GFP versus 2 copies of GFP in germ cells of the ovary? This is not possible in the Drosophila testis. I ask because this could impact on the clonal analyses diagrammed in Fig. 4A and 4G and in 6A and B. Additionally, in most of the figures, the GFP is saturated so it is not possible to discern one vs two copies of GFP.

In the rebuttal, the authors stated that they cannot differential one vs two copies of GFP. They used other clone labeling methods in Fig. 4 and 6. I think that the authors should make a statement in the manuscript that they cannot distinguish one vs two copies of GFP for the record.

(7) More evidence is needed to support the claim of elevated Dpp levels in bam or bgcn mutant tumors. The current results with dpp-lacZ enhancer trap in Fig 5A,B are not convincing. First, why is the dpp-lacZ so much brighter in the mosaic analysis (A) than in the no-clone analysis (B); it is expected that the level of dpp-lacZ in cap cells should be invariant between ovaries and yet LacZ is very faint in Fig. 5B. I think that if the settings in A matched those in B, the apparent expression of dpp-lacZ in the tumor would be much lower and likely not statistically significantly. Second, they should use RNA in situ hybridization with a sensitive technique like hybridization chain reactions (HCR) - an approach that has worked well in numerous Drosophila tissues including the ovary.

The HCR FISH in Fig.5 of the revised manuscript needs an explanation for how the mRNA puncta were quantified. Currently, there is no information in the methods. What is meant but relative dpp levels. I think that the authors should report in and unbiased manner "number" of dpp or gbb puncta in TFs. For the germaria, I think that they should report the number of puncta of dpp or gbb divide by the total area in square pixels counted. Additionally, the background fluorescence is noticeably much higher in bamBG/delta86 germaria, which would (falsely) increase the relative intensity of dpp and gbb in bam mutants. Although, I commend the authors for performing HCR FISH, these data are still not convincing to me.

(8) In Fig 6, the authors report results obtained with the bamBG allele. Do they obtain similar data with another bam allele (i.e., bamdelta86)?

The authors did not try any experiments with the bamdelta86 allele, despite this allele being molecularly defined, where the bamBG allele is not defined.

Comments on second revision:

The authors have adequately addressed several points. However, there is still no information in the material and methods for how they measured and quantified the HCR-FISH probe signal. They have the same size region that they use for each genotype, but they do not control for the number of nuclei in each square. I would also be helpful if they provided a different image for the gbb probe stained in the mutant background. It is the only panel that does not have other germaria in very close proximity. I am still not fully convinced of the HCR data, esp for gbb.

Reviewer #2 (Public review):

In the current version, Zhang et al. have made substantial improvements to the manuscript. It is now easier to read, and the data are more solid compared with the previous version, supporting their conclusion that tumor GSCs secrete stemness factors (BMPs and Dpp) to suppress the differentiation of neighboring wild-type GSCs. This study should benefit a broad readership across developmental biology, germ cell biology, stem cell biology, and cancer biology.

Comments on revision:

If the exact number of germaria was not recorded (as described), an approximate number can be provided in the Materials and Methods; for example, stating that more than 10 germaria were analyzed per biological replicate.

Reviewer #3 (Public review):

Zhang et al. investigated how germline tumors influence the development of neighboring wild-type (WT) germline stem cells (GSC) in the Drosophila ovary. They report that germline tumors generated by differentiation-arrested mutations (bam and bgcn) inhibit the differentiation of neighboring WT GSCs by arresting them in an undifferentiated state, resulting from reduced expression of the differentiation-promoting factor Bam. They find that these tumor cells produce low levels of the niche-associated signaling molecules Dpp and Gbb, which suppress bam expression and consequently inhibit the differentiation of neighboring WT GSCs non-cell-autonomously. Based on these findings, the authors propose that germline tumors mimic the niche to suppress the differentiation of the neighboring wild-type germline stem cells.

Strengths:

The study uses a well-established in vivo model to addresses an important biological question concerning the interaction between germline tumor cells and wild-type (WT) germline stem cells in the Drosophila ovary. If the findings are substantiated, this study could provide valuable insights that are applicable to other stem cell systems.

Weaknesses:

The authors have addressed some of my concerns in the revised submission. However, the data presented do not allow the authors to distinguish whether the failed differentiation of WT stem cells/germline cells results from "arrested differentiation due to the loss of the differentiation niche" or from "direct inhibition by tumor-derived expression of niche-associated molecules Dpp and Gbb". The critical supporting data, HCR in situ results, are not sufficiently convincing.

Author response:

The following is the authors’ response to the previous reviews

eLife Assessment

This study presents results supporting a model that tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the stem cell niche and inhibit the differentiation of neighboring cells. The valuable findings show that GSC tumors often contain non-mutant cells whose differentiation is suppressed by the GSC tumorous cells. However, the evidence showing that the GSC tumors produce BMP ligands to suppress differentiation of non-mutant cells is incomplete due to concerns about the new HCR data.

Thanks for this assessment. All concerns raised by the reviewers regarding the HCR data and others are followed by our responses below.

Public Reviews:

Reviewer #1 (Public review):

Summary:

This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche. This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche.

Strengths:

(1) Use of an excellent and established model for tumorous cells in a stem cell microenvironment

(2) Powerful genetics allow them to test various factors in the tumorous vs non-tumorous cells

(3) Appropriate use of quantification and statistics

Thank you for your valuable comments, and we greatly appreciate them.

Weaknesses:

(1) What is the frequency of SGCs in nos>flp; bam-mutant tumors? For example, are they seen in every germarium, or in some germaria, etc or in a few germaria.

This concern was addressed in the rebuttal. The line number is 106, not line 103.

(2) Does the breakdown in clonality vary when they induce hs-flp clones in adults as opposed to in larvae/pupae?

This concern was addressed in the rebuttal. However, these statements are no on lines 331-335 but instead starting on line 339. Please be accurate about the line numbers cited in the rebuttal. They need to match the line numbers in the revised manuscript.

We have rechecked the line numbers and confirmed that the mismatch arose from the Word-to-PDF conversion process on the eLife website. As this issue has recurred and reviewers’ file-format preferences are unknown to us, we have added a clarifying note at the beginning of each response letter: “Please note that the line numbers cited refer to the revised manuscript in the Microsoft Word format”.

(3) Approximately 20-25% of SGCs are bam+, dad-LacZ+. Firstly, how do the authors explain this? Secondly, of the 70-75% of SGCs that have no/low BMP signaling, the authors should perform additional characterization using markers that are expressed in GSCs (i.e., Sex lethal and nanos).

The authors did not perform additional staining for GSC-enriched protein like Sex lethal and nanos.

The 70-75% of SGCs that have low BMP signaling display the following characteristics: 1) dot-like spectrosomes, 2) positivity for Dad-lacZ, and 3) absence of bamP-GFP expression. This combination of traits is sufficient to classify them as GSC-like cells. Neither Sex lethal nor Nanos is expressed exclusively in GSCs (Chau et al., 2009; Li et al., 2009), rendering them unsuitable for distinguishing GSC-like from cystoblast-like cells.

(4) All experiments except Fig. 1I (where a single germarium with no quantification) were performed with nos-Gal4, UASp-flp. Have the authors performed any of the phenotypic characterizations (i.e., figures other than figure 1) with hs-flp?

In the rebuttal, the authors stated that they used nos>flp for all figures except for Fig. 1I. It would be more convincing for them to prove in Fig. 1 than there is not phenoytpic difference between the two methods and then switch to the nos>FLP method for the rest of the paper.

We appreciate this suggestion. These data are included in Figure 1-figure supplement 3 in the revised manuscript.

(5) Does the number of SGCs change with the age of the female? The experiments were all performed in 14-day old adult females. What happens when they look at young female (like 2-day old). I assume that the nos>flp is working in larval and pupal stages and so the phenotype should be present in young females. Why did the authors choose this later age? For example, is the phenotype more robust in older females? or do you see more SGCs at later time points?

The authors did not supply any data to prove that the clones were larger in 14-day-old flies than in younger flies. Additionally, the age of "younger" flies was not specified. Therefore, the authors did not satisfactorily answer my concern.

We appreciate this critical comment. Figure 1J includes the SGC phenotype data from 1-, 7-, and 14-day-old flies. Both 1- and 7-day-old flies are younger flies in our analyses. The evidence that germline clones were larger in 14-day-old flies than in younger flies was provided in Figure 1-figure supplement 2 in the revised manuscript.

(6) Can the authors distinguish one copy of GFP versus 2 copies of GFP in germ cells of the ovary? This is not possible in the Drosophila testis. I ask because this could impact on the clonal analyses diagrammed in Fig. 4A and 4G and in 6A and B. Additionally, in most of the figures, the GFP is saturated so it is not possible to discern one vs two copies of GFP.

In the rebuttal, the authors stated that they cannot differential one vs two copies of GFP. They used other clone labeling methods in Fig. 4 and 6. I think that the authors should make a statement in the manuscript that they cannot distinguish one vs two copies of GFP for the record.

Thank you for this suggestion. Such statement has been added in the revised manuscript (Lines 177-178).

(7) More evidence is needed to support the claim of elevated Dpp levels in bam or bgcn mutant tumors. The current results with dpp-lacZ enhancer trap in Fig 5A,B are not convincing. First, why is the dpp-lacZ so much brighter in the mosaic analysis (A) than in the no-clone analysis (B); it is expected that the level of dpp-lacZ in cap cells should be invariant between ovaries and yet LacZ is very faint in Fig. 5B. I think that if the settings in A matched those in B, the apparent expression of dpp-lacZ in the tumor would be much lower and likely not statistically significantly. Second, they should use RNA in situ hybridization with a sensitive technique like hybridization chain reactions (HCR) - an approach that has worked well in numerous Drosophila tissues including the ovary.

The HCR FISH in Fig.5 of the revised manuscript needs an explanation for how the mRNA puncta were quantified. Currently, there is no information in the methods. What is meant but relative dpp levels. I think that the authors should report in and unbiased manner "number" of dpp or gbb puncta in TFs. For the germaria, I think that they should report the number of puncta of dpp or gbb divide by the total area in square pixels counted. Additionally, the background fluorescence is noticeably much higher in bamBG/delta86 germaria, which would (falsely) increase the relative intensity of dpp and gbb in bam mutants. Although, I commend the authors for performing HCR FISH, these data are still not convincing to me.

We appreciate these critical comments. Due to variable puncta sizes and frequent clustering in TF and cap cells (see Figure 5A, C), direct quantification of puncta number was unreliable. Therefore, we quantified mean fluorescence intensity instead, as described in the revised figure legend of Figure 5 (Lines 603-604). In Author response image 1 1A, B (modified from Figure 5A, C) , magenta ovals indicate empty background fluorescence areas, which appear similar between w¹¹¹⁸ (wild-type control) and bam^-/- germaria. In Author response image 1, the yellow oval outlines a neighboring germarium, not an empty area (see the DAPI channel).

Author response image 1.

In situ-HCR results of dpp and gbb in wild-type and bam mutant germaria. Magenta ovals indicate empty areas displaying only background fluorescence. In panel (B), the yellow oval outlines a neighboring germarium, not an empty area (see the DAPI channel below).

(8) In Fig 6, the authors report results obtained with the bamBG allele. Do they obtain similar data with another bam allele (i.e., bamdelta86)?

The authors did not try any experiments with the bamdelta86 allele, despite this allele being molecularly defined, where the bamBG allele is not defined.

While we agree that repeating the experiments in Figure 6 with bam^Δ86 would be helpful, our mosaic analysis strategy for two genes on different chromosome arms is technically complex (see genotypes in Source data 1). Switching from bam^BG to bam^Δ86 would necessitate extensive and time-consuming genetic recombination. Given that both alleles induce the SGC phenotype indistinguishably (Figure 1J), we believe that repeating these experiments with bam^Δ86 would not alter our key conclusion. We appreciate your understanding regarding this technical complexity.

Reviewer #2 (Public review):

In the current version, Zhang et al. have made substantial improvements to the manuscript. It is now easier to read, and the data are more solid compared with the previous version, supporting their conclusion that tumor GSCs secrete stemness factors (BMPs and Dpp) to suppress the differentiation of neighboring wild-type GSCs. This study should benefit a broad readership across developmental biology, germ cell biology, stem cell biology, and cancer biology.

Thank you for your valuable comments, and we greatly appreciate them.

However, the following suggestions may further improve the clarity and rigor of the research content:

(1) Clarification of sample size (n).

Each germarium can contain highly variable numbers of SGCs, sometimes reaching 50-100. When reporting "n" values, the authors are encouraged to also indicate the number of germaria analyzed. For example, in lines 126-128:

"Notably, 74% of SGCs (n = 132) were GFP-negative, while the remaining 26% were GFP-positive (Figure 2B, C). This suggests that SGCs can be categorized into two distinct groups: those resembling GSCs (GSC-like) and those resembling cystoblasts (cystoblast-like)." Please clarify how many germaria were examined to obtain n = 132.

We appreciate this comment. In 14-day-old fly ovaries, each germarium that met our criterion for quantifying the SGC phenotype contains approximately 1.5 SGCs (see Figure 1K). For the specific analysis of the “132” SGCs presented in Figure 2C, we did not record the number of germaria from which they originated.

In addition, it is unclear whether the authors intend to suggest that the GFP-negative SGCs are GSC-like or cystoblast-like; this point should be clarified.

Thank you for this suggestion. We intend to suggest that the bamP-GFP-negative SGCs are GSC-like, which information has been added in the revised manuscript (Line 129).

(2) Improvement of Fig. 6 in situ hybridization images.

The in situ hybridization images in Fig. 6 are not fully convincing. The control images, in particular, would benefit from higher resolution and enlarged views of the germarium region.

Thank you for this valuable suggestion. The enlarged views of both the control and bam^-/- germarium regions were included in Figure 5A, C in the revised manuscript.

In panel C, abundant signals are also present outside the germarium, which may complicate interpretation and should be clarified or controlled for.

In the right panel of Figure 5C, the abundant signals noted by the reviewer originate from neighboring germaria (see the DAPI channel), not from empty areas, which would be expected to show only background fluorescence. For more details, please refer to our response to Question (7) raised by Reviewer #1.

Alternatively, the authors could strengthen the in situ analysis by using bam mutants or bam dpp / bam gbb double mutants as controls to better define signal specificity.

We appreciate this comment. Homozygous dpp or gbb mutants are lethal, precluding the generation of dpp bam or gbb bam double-mutant flies. Additionally, the GFP signal was drastically reduced during our HCR processing, preventing mosaic clone analysis.

Reviewer #3 (Public review):

Zhang et al. investigated how germline tumors influence the development of neighboring wild-type (WT) germline stem cells (GSC) in the Drosophila ovary. They report that germline tumors generated by differentiation-arrested mutations (bam and bgcn) inhibit the differentiation of neighboring WT GSCs by arresting them in an undifferentiated state, resulting from reduced expression of the differentiation-promoting factor Bam. They find that these tumor cells produce low levels of the niche-associated signaling molecules Dpp and Gbb, which suppress bam expression and consequently inhibit the differentiation of neighboring WT GSCs non-cell-autonomously. Based on these findings, the authors propose that germline tumors mimic the niche to suppress the differentiation of the neighboring wild-type germline stem cells.

Strengths:

The study uses a well-established in vivo model to address an important biological question concerning the interaction between germline tumor cells and wild-type (WT) germline stem cells in the Drosophila ovary. If the findings are substantiated, this study could provide valuable insights that are applicable to other stem cell systems.

Thank you for your valuable comments, and we greatly appreciate them.

Weaknesses:

The authors have addressed some of my concerns in the revised submission. However, the data presented do not allow the authors to distinguish whether the failed differentiation of WT stem cells/germline cells results from "arrested differentiation due to the loss of the differentiation niche" or from "direct inhibition by tumor-derived expression of niche-associated molecules Dpp and Gbb".

Blocking Dpp or Gbb secretion specifically from germline tumor cells promoted differentiation of neighboring wild-type germ cells (Figure 6). This indicates that BMP ligands secreted by germline tumors are required to inhibit this differentiation. However, we cannot rule out the possibility that disruption of the differentiation niche also contributes to the SGC phenotype, a point highlighted in the manuscript (Line 204).

The critical supporting data, HCR in situ results, are not sufficiently convincing.

Below, we provide a point-by-point reply addressing each of your specific recommendations.

Recommendations for the authors:

Reviewer #3 (Recommendations for the authors):

It's a surprising that the authors failed to induce germline tumors at the adult stage, as this has been reported by many labs and would allow for time course analysis of SGC phenotype. As a result, the data in this manuscript address only events occurring after the germline tumor formation (with clonal induction at larval stage) and and focus on the already presene "arrested wild-type germ cells", without providing insight into the process of by which these arrested germ cells are formed.

In our hands, inducing germline clones by the hs-FLP method at the adult stage was efficient in males but not in females, despite subjecting adult flies to intensive heat-shock at 37°C.

The HCR in situ data exhibit a high background.

Regarding the background issue, please see our response to Reviewer #1’s Question (7).

First, the signal appears stronger in TF cells than in cap cells.

As demonstrated by Li et al. (Li et al., 2016), dpp-lacZ (P4-lacZ) signals are also stronger in TF cells than in cap cells (see their Figure 4D').

Second, both dpp and gbb are detected broadly in somatic cells including escort cells. These observations are inconsistent with published data.

As shown in Figure 5A and C, dpp and gbb were detected broadly in somatic cells of bam^-/- germaria, but not in those of w¹¹¹⁸ (wild-type) controls. To our knowledge, no previous study has reported the expression pattern of these ligands in a bam mutant background.

To demonstrate the tumor-derived dpp and gbb, the HCR in situ analysis could be performed in the germarium with mosaic clones. If these niche-associated molecules are indeed expressed in tumor cells, the authors should observe a mosaic expression pattern of these molecules, with signal "ON" in tumor cells and "OFF" in neighbouring arrested germ cells.

This is a great idea and was indeed our original approach. However, GFP signal was drastically reduced during our HCR processing, ultimately precluding mosaic clone analysis.

References

Chau, J., Kulnane, L.S., and Salz, H.K. (2009). Sex-lethal facilitates the transition from germline stem cell to committed daughter cell in the Drosophila ovary. Genetics 182, 121-132.

Li, X., Yang, F., Chen, H., Deng, B., Li, X., and Xi, R. (2016). Control of germline stem cell differentiation by Polycomb and Trithorax group genes in the niche microenvironment. Development 143, 3449-3458.

Li, Y., Minor, N.T., Park, J.K., McKearin, D.M., and Maines, J.Z. (2009). Bam and Bgcn antagonize Nanos-dependent germ-line stem cell maintenance. Proc Natl Acad Sci U S A 106, 9304-9309.

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potentially unique contribution to local field potentials (LFPs). The biophysical foundations of transmembrane electric dipoles, and the associated argument points, are generally compelling. Experimental constraints on gap junctions and strictly quantitative matching between chemical vs. junctional inputs have been hard to achieve. This computational investigation thus offers a specific way to enhance conceptual understanding and provides interesting testable predictions, which would be of great interest to experimental neurophysiologists who interpret relevant recordings.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

Strengths:

Novelty and Scope:

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience.<br /> It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

Methodological Rigor:

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance.<br /> Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

Biological Relevance:

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses.<br /> The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

Clarity and Depth:

The manuscript is well-structured, with logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

Comments on revised version:

The authors have addressed all my concerns in the revised version of the manuscript.

Reviewer #2 (Public review):

Summary:

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to differences in the amplitudes and time courses of the input currents). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

Strengths:

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

Weaknesses:

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known, and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S4, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies, although it is closely related to models used in earlier modeling work from the same laboratory. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

Comments on revised version:

The authors made mainly cosmetic changes in the manuscript (primarily by adding more discussion), and most of these do not affect my earlier assessment. I have updated my Public Review in a few places to reflect those few changes that substantially address my previous concerns.

Author response:

The following is the authors’ response to the original reviews.

eLife Assessment:

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence can be more convincing if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

We gratefully thank the editors and the reviewers for the time and effort in rigorously assessing our manuscript, for the constructive review process, for their enthusiastic responses to our study, and for the encouraging and thoughtful comments. We especially thank you for deeming our study to be a valuable exploration on the differential contributions of active dendritic gap junctions vs. chemical synapses to local field potentials. We thank you for your appreciation of the quantitative biophysical demonstration on the differences in electric dipoles that appear in extracellular potentials with gap junctions vs. chemical synapses.

However, we are surprised by aspects of the assessment that resulted in deeming the approach incomplete, especially given the following with specific reference to the points raised:

(1) Testing against experiments: With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

In addition, the complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

Together, we emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials.

(2) Model validation: The model used in this study was adopted from a physiologically validated model from our laboratory (Roy & Narayanan, 2021). Please note that the original model was validated against several physiological measurements along the somatodendritic axis. We sincerely regret our oversight in not mentioning clearly that we have used an existing, thoroughly physiologically-validated model from our laboratory in this study.

(3) Comparisons between chemical vs. junctional inputs: We had taken elaborate precautions in our experimental design to match the intracellular electrophysiological signatures with reference to synchronous as well as oscillatory inputs, irrespective of whether inputs arrived through gap junctions or chemical synapses. A new Supplementary Figure S3 has been added to address this concern raised by the reviewers.

In the revised manuscript, we have addressed all the concerns raised by the reviewers in detail. We have provided point-by-point responses to reviewers’ helpful and constructive comments below. We thank the editors and the reviewers for this constructive review process, which helped us in improving our manuscript with specific reference to emphasizing the novelty of our approach and conclusions. The specific changes incorporated into the revised manuscript are detailed below.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

Novelty and Scope

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience. It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

We thank you for the positive comments on the novelty of our approach and how our study addresses an underexplored area in neuroscience. The assumptions about the passive nature of dendritic structures had indeed resulted in an underestimation of the contributions of gap junctions to extracellular potentials. Once the realities of active structures are accounted for, the contributions of gap junctions increases by several orders of magnitude compared to passive structures (Fig. 1D).

Methodological Rigor

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance. Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

We thank you for your encouraging comments on the experimental design and methodological rigor of our approach.

Biological Relevance

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses. The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

We thank you for your positive comments on the biological relevance of our approach. We also gratefully thank you for emphasizing the two striking novelties unveiling the dichotomy between gap junctions and chemical synapses in their contributions to field potentials: polarity differences and spectral characteristics.

Clarity and Depth

The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

We sincerely thank you for the positive comments on the structure and comprehensive coverage of our manuscript encompassing different types of inputs that neurons typically receive.

Weaknesses and Areas for Improvement

Generality and Validation

The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings. Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

We thank you for raising this important point. The prime novelty and the principal conclusion of this study is that gap junctional contributions to extracellular field potentials are orders of magnitude higher when the active nature of cellular compartments are accounted for. The lacuna in the literature has been consequent to the assumption that cellular compartments are passive, resulting in the dogma that gap junctional contributions to field potentials are negligible. Despite knowledge about active dendritic structures for decades now, this assumption has kept studies from understanding or even exploring the contributions of gap junctions to field potentials. The rationale behind the choice of a computational approach to address the lacuna were as follows:

(1) The complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

(2) With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non-specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). 'The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

We highlight the novelty of our approach and of the conclusions about differences in extracellular signatures associated with active-dendritic chemical synapses and gap junctions, against these experimental difficulties. We emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials. Our analyses clearly demonstrates that gap junctions do contribute to extracellular potentials if the active nature of the cellular compartments is explicitly accounted for (Fig. 1D). We also show theoretically well-grounded and mechanistically elucidated differences in polarity (Figs. 1–3) as well as in spectral signatures (Figs. 5–8) of extracellular potentials associated with gap junctional vs. chemical synaptic inputs. Together, our fundamental demonstration in this study is the critical need to account for the active nature of cellular compartments in studying gap junctional contributions of extracellular potentials, with CA1 pyramidal neuronal dendrites used as an exemplar.

In the revised version of the manuscript, we have emphasized the motivations for the approach we took, highlighting the specific novelties both in methodological and conceptual aspects, finally emphasizing the need to account for other cell types and gap junctional contributions therein. Importantly, we have emphasized the non-specificities associated with gap-junctional blockers as the reason why experimental delineation of gap junctional vs. chemical synaptic contributions to LFP becomes tedious. We believe that these points underscore the need for the computational approach that we took to address this important question, apart from the novelties of the study.

In response to your constructive comments, we have added the following to the revised version of the manuscript, in the Introduction section as motivation for the specific route we took:

“Given the complexity arising from the concurrent activity of chemical synapses, gap junctions, and active dendritic conductances across multiple neuronal populations, experimentally isolating the contributions of individual components to extracellular potentials remains highly challenging. To address this limitation, we employed a computational modeling approach, which provides a quantitative framework for systematically dissecting the distinct roles of specific cellular and synaptic elements. This strategy is consistent with previous studies that have successfully used computational methods to elucidate the contributions of active dendritic mechanisms to LFPs (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to LFPs (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). In addition, experimentally isolating the contribution of gap junctions is complicated by non-specific effects of available pharmacological modulators targeting these connections (Behrens et al., 2011; Rouach et al., 2003). Most genetic knockouts of gap junctional proteins are either lethal or trigger functional compensatory mechanisms (Bedner et al., 2012; Lo, 1999), thereby rendering causal attribution of specific gap junctional contributions infeasible with currently available experimental approaches. Consequently, biophysically and morphologically detailed computational modeling provides a crucial means to evaluate the impact of individual neuronal components on extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).”

We thank you for raising this point as this allowed us to expand on the specific motivations for the approach we took, and to present the specific novelties of our study to the analyses of extracellular field potentials. Thank you.

Role of Active Dendritic Currents

The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

We thank you for this constructive suggestion. We agree that it is important to consider the implications for broader network dynamics of the outward HCN currents that are observed with synchronous inputs. In the revised manuscript, we have elaborated on the implications of the outward HCN current to network dynamics in detail. The following paragraph has been added to Discussion subsection on “Outward HCN currents regulate extracellular potentials”:

“HCN channels play a critical role in shaping hippocampal network dynamics by modulating neuronal excitability, oscillatory behavior, and susceptibility to pathological states (Kessi et al., 2022; Magee, 1998; Mishra & Narayanan, 2025; Nolan et al., 2004). The outward-like properties of the HCN current we observed may have specific functional implications at different scales. At the cellular scale, the manifestation of outward current during action potentials or plateau potentials could contribute to after hyperpolarization thereby regulating firing properties. In cortical and hippocampal pyramidal neurons, most single-neuron processing occurs in their elaborate dendritic branches, where there is spatiotemporal summation of different synaptic potentials, plateau potentials, back propagating action potentials, and dendritic spikes (Johnston & Narayanan, 2008; Major et al., 2013; Stuart & Spruston, 2015). Considering the heavy expression of HCN channels in the dendrites of hippocampal and cortical pyramidal neurons (Kole et al., 2006; Lorincz et al., 2002; Magee, 1998; Williams & Stuart, 2000), the back propagating action potentials, plateau potentials, or dendritic spikes at dendritic location could yield outward currents. These outward currents could act as a hyperpolarizing mechanism that suppresses spatiotemporal summation of the different dendritic potentials.

At the network scale, such regulation of dendritic potentials and somatic firing could contribute to overall reduction in firing rates of different neurons in the network. For instance, as inhibitory neurons typically elicit action potentials at higher frequencies, somatic outward HCN currents would occur more frequently in inhibitory neurons that express HCN channels compared to excitatory neurons. However, the heavy expression of HCN channels in the dendrites and the higher prevalence of dendritic spikes and plateau potentials in dendrites (Basak & Narayanan, 2018; Larkum et al., 2022; Moore et al., 2017) imply that the impact on outward HCN currents might be higher. Thus, the presence of outward HCN currents would regulate network balance of excitation inhibition in an activity-dependent manner. Additionally, the outward component of the current through HCN channels could contribute to stabilization of network synchrony by promoting spike phase coherence and to modulation of spike-LFP phase relationships (Das et al., 2017; Ness et al., 2016, 2018; Seenivasan & Narayanan, 2020; Sinha & Narayanan, 2015, 2022).

Together, the outward HCN current could play critical roles in regulating several cellular and network functions including spatiotemporal summation within single neurons, amplitude and phase of different oscillations, excitatory-inhibitory interactions, and rate and temporal coding involved in spatial navigation (Hussaini et al., 2011; Nolan et al., 2004; O'Keefe & Recce, 1993). In the context of brain rhythms, future investigations are needed to explore ripple-frequency oscillations, specifically to assess whether high-frequency network interactions are modulated by HCN outward currents. Importantly, future studies could specifically focus on delineating the prevalence and specific contributions of outward currents through HCN channels to single-neuron and network physiology.”

We thank you for highlighting this point, as it allowed us to elaborate the broader roles of HCN channels to single-cell computation, network dynamics, and field potentials. Thank you.

Analysis of Plasticity

While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

We thank you for this constructive suggestion. Please note that we have presented consistent results for both fewer and more gap junctions in our analyses (Figure 1 with 217 gap junctions and Supplementary Figure 1 with 99 gap junctions). Thus, our fundamentally novel result that gap junctions onto active dendrites differentially shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron. Thus, these results demonstrate that the conclusions about their contributions to LFP are invariant to plasticity in their gap junctional numerosity.

We had only briefly mentioned plasticity in the Introduction to highlight the different modes of synaptic transmission and to emphasize that plasticity has been studied in both chemical synapses and gap junctions, playing a role in learning and adaptation. However, it seems that this wording inadvertently suggested that our study includes plasticity simulations. Therefore, we have removed that sentence from Introduction in the revised manuscript to ensure clarity.

In the ‘Limitations of analyses and future studies’ section in Discussion, we suggested investigating the impact of plasticity mechanisms—specifically, activity-dependent plasticity of ion channels—on synaptic receptors vs. gap junctions and their effects on extracellular field potentials under various input conditions and plasticity combinations across different structures. We fully agree with the reviewer that such studies would offer valuable insights and further enhance the broader relevance of our findings. However, while our study implies this direction, it was not the primary focus of our investigation.

In the revised manuscript, we have also expanded on intrinsic/synaptic plasticity and how they could contribute to LFPs (Sinha & Narayanan, 2015, 2022), while also pointing to simulations with different numbers of gap junction in this context. The following specific changes have been incorporated to the revised manuscript:

Discussion subsection “Limitations of analyses and future directions”

“We demonstrated that the contribution of gap junctions to extracellular field potentials remains consistent regardless of the number of gap junctions. Specifically, we showed that the distinct positive LFP deflections persisted irrespective of their relative density on neurons (Fig. 1 with 217 gap junctions and Supplementary Fig. 1 with 99 gap junctions). Previous studies have quantitatively demonstrated that intrinsic and synaptic plasticity modulate hippocampal LFPs and phase coding (Sinha & Narayanan, 2015, 2022). Future analyses should also assess the impact of activity-dependent plasticity in ion channels (on dendrites, axonal initial segments, and other compartments), in synaptic receptors, and in gap junctions (Andersen et al., 2006; Coulon & Landisman, 2017; Johnston & Narayanan, 2008; Magee & Grienberger, 2020; Mishra & Narayanan, 2021; Neves et al., 2008; O'Brien, 2014; Pereda, 2014; Vaughn & Haas, 2022) on extracellular potentials with various kinds of gap junctional inputs and different combinations of plasticity in various structures. Interactions among different forms of plasticity and how co-dependent plasticity in different components alters extracellular field potentials could provide deeper insights about physiological changes during learning and pathological changes observed in different neurological disorders (Sinha & Narayanan, 2022).”

We thank you for highlighting this as this allowed us to improve on the specific focus of the manuscript and the study. Thank you.

Frequency-Dependent Effects

The study demonstrates that gap junctional inputs suppress highfrequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

We sincerely thank you for these insightful comments that we totally agree with. As it so happens, this manuscript forms the first part of a broader study where we explore the implications of gap junctions to ripple frequency oscillations. The ripple oscillations part of the work was presented as a poster in the Society for Neuroscience (SfN) annual meeting 2024 (Sirmaur & Narayanan, 2024). There, we simulate a neuropil made of hundreds of morphologically realistic neurons to assess the role of different synaptic inputs excitatory, inhibitory, and gap junctional and active dendrites to ripple frequency oscillations. We demonstrate there that the conclusions from single-neuron simulations in this current manuscript extend to a neuropil with several neurons, each receiving excitatory, inhibitory and gap-junctional inputs, especially with reference to high-frequency oscillations. Our network based analyses unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions in conjunction with return-current contributions from active dendrites playing regulatory roles in determining ripple characteristics (Sirmaur & Narayanan, 2024).

Our principal goal in this study, therefore, was to lay the single-neuron foundation for network analyses of the impact of gap junctions on LFPs. We are preparing the network part of the study, with a strong focus on ripple-frequency oscillations, for submission for peer review separately. Please see abstract of our poster presented at the Society for Neuroscience annual meeting 2024 on the topic here: https://tinyurl.com/57ehvsep).

In the revised manuscript, we have mentioned the results from our SfN abstract with reference to network simulations and high-frequency oscillations, while also presenting discussions from other studies on the role of gap junctions in synchrony and LFP oscillations. The following has been added to the revised manuscript under the Discussion subsection “High-frequency LFP power was suppressed with gap junctional inputs”:

“In this context, our analyses lay the foundation for network analyses of the impact of gap junctions on LFPs. The conclusions from the single-neuron simulations in this study extend to a neuropil with several neurons, each receiving synaptic and gap junctional inputs, especially with reference to high-frequency ripple oscillations (Sirmaur & Narayanan, 2024). A neuropil made of hundreds of morphologically realistic pyramidal neurons was used to assess the role of different synaptic inputs excitatory, inhibitory, and gap junctional with different patterns of stimulation and active dendritic contributions to ripple-frequency oscillations. Network-based analyses have unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions, in conjunction with return-current contributions from active dendrites, playing modulatory roles in governing ripple characteristics (Sirmaur & Narayanan, 2024). Future studies could expand on these conclusions to explore the implications of frequency-dependent filtering (with reference to gap junctional coupling) on high-frequency extracellular oscillations.”

We thank you for highlighting this point as it allowed us to expand on the implications for our analyses to brain rhythms, especially with reference to high-frequency oscillations. Thank you.

Visualization

Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

We thank you for this constructive suggestion. We used the specific visualization throughout, where we place the outcomes associated with chemical synapses and gap junctions in the same figure, adjacent to each other. We believe that this offers visually intuitive distinction between the outcomes for chemical synapses and gap junctions, rather than placing them in different figures. Splitting them would place the outcomes in different figures and requires turning pages or placing two different figures adjacent to each other for quantitative comparison. We respectfully request that we be allowed to retain this form of visualization in the figures. Thank you.

Contextual Relevance

The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

We sincerely appreciate your valuable suggestion and acknowledge the importance of integrating our results into established neural network dynamics, particularly their implications for gamma rhythms. We have addressed this aspect in the revised version of our manuscript. We have added this to the Discussion subsection on “High-frequency LFP power was suppressed with gap junctional inputs” of the revised manuscript:

“In the context of oscillations and gap-junctional coupling, electrical synapses have been shown to regulate the emergence and stability of the network interactions underlying rhythms of different frequencies, especially gamma-frequency oscillations (Bocian et al., 2009; Buhl et al., 2003; Draguhn et al., 1998; Hormuzdi et al., 2001; Konopacki et al., 2004; LeBeau et al., 2003; Posluszny, 2014; Traub et al., 2003). Specifically, both genetic and pharmacological manipulations of gap junctions have been shown to disrupt gamma rhythms. Genetic deletion of connexin-36 impairs the gamma oscillations associated with awake, active behavioral states (Buhl et al., 2003; Hormuzdi et al., 2001). High-frequency oscillations in the hippocampus have been shown to be sensitive to pharmacological agents like carbenoxolone and octanol that are known to inhibit gap junctions. Carbenoxolone has been known to reduce the transient gamma-frequency oscillations while octanol abolishes the persistent gamma rhythm (Draguhn et al., 1998; Hormuzdi et al., 2001; Posluszny, 2014; Traub et al., 2003). In the context of our results, where we demonstrate that the relative contributions of gap-junctional coupling to high-frequency extracellular potentials is low (Figs. 6–7), how do gap junctions contribute to enhanced extracellular gamma oscillations in these circuits?

It should be noted that in hippocampal circuits, gamma oscillations emerge predominantly due to interactions between inhibitory interneurons through GABAA103046 receptors (Buzsaki & Wang, 2012; Colgin, 2016; Colgin & Moser, 2010; Wang, 2010; Wang & Buzsaki, 1996; Whittington et al., 1995). Thus, the presence of additional gap junctional coupling between these inhibitory neurons allows for tighter synchrony between these reciprocally inhibition-coupled neurons. In other words, the presence of gap junctions increases the probability of action potential generation in other neurons that are electrically coupled to them, together increasing the population of inhibitory neurons that elicit synchronous action potentials. When these synchronous action potentials act on the adjacent cells, both excitatory and inhibitory, the transmembrane GABAA receptor currents yield stronger gamma-frequency oscillations in the extracellular potentials (Draguhn et al., 1998; Hormuzdi et al., 2001; Posluszny, 2014; Traub et al., 2003). Thus, the stronger high-frequency oscillations observed in these scenarios is owing to the enhanced synchrony that is brought about the gap-junctional coupling, which translates to stronger transmembrane inhibitory receptor currents.

These observations also strongly emphasize the utility of the computational approach we took in this study towards discerning the specific roles of gap junctions. Gap junctional coupling have strong physiological roles in terms of enhancing synchronous activity across the neurons that they couple and often express along with other receptors that connect the sets of neurons. Thus, the specific contributions of different neuronal components need to be studied with reference to how they contribute to physiological characteristics vs. their contributions to extracellular potentials. Thus, computational modeling offers an ideal route to understand the specific contributions of different neural-circuit components to extracellular field potentials and rhythms therein (Buzsaki et al., 2012; Einevoll et al., 2019; Einevoll et al., 2013; Sinha & Narayanan, 2022).”

We thank you for highlighting this point as this allowed us to delineate the impact of gap junctions to regulating synchrony across connected neurons vs. modulating field potentials. Thank you.

Reviewer #2 (Public review):

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

We gratefully thank you for the positive comments and the encouraging words about the novel contributions of our study. We are particularly thankful to you for your comment on the generality of our conclusions that hold for different cell types and neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

We sincerely thank you for the positive comments about the readability of the paper.

Weaknesses

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

We thank you for raising this important point. We would like to emphasize that our experimental design and analyses quantitatively account for the spatial distribution and temporal pattern of specific kinds of inputs that arrive through gap junctions and chemical synapses. We submit that our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. We elucidate these points below:

(1) Spatial distribution: The inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were of the same nature: excitatory.

(2) Different numbers of inputs: We have presented consistent results for both fewer and more gap junctions or chemical synapses in our analyses (see Figure 1 with 217 gap junctions or 245 chemical synapses and Supplementary Figure 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

(3) Synchronous inputs (Figs. 1–3): For chemical synapses, the waveforms are in the shape of postsynaptic potentials. For gap junctional inputs, the waveforms are in the shape of postsynaptic potentials or dendritic spikes (to respect the active nature of inputs from the other cell). Here, the electrical response of the postsynaptic cell is identical irrespective of whether inputs arrive through gap junctions or chemical synapses: an action potential. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic and gap junctional inputs. We mechanistically analyzed the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Figs. 2–3). In the revised manuscript, we have shown the intracellular responses to demonstrate that they are electrically matched (new Supplementary Figure 3).

(4) Random inputs (Fig. 4): For random inputs, we did not account for the number of action potentials that arrived, as the only observation we made here was with reference to the biphasic nature of the extracellular potentials with gap junctional inputs in the “No Sodium” scenario. We note that in the “No Sodium” scenario, the time-domain amplitudes were comparable for the field potentials (Fig. 4B, Fig. 4D).

(5) Rhythmic inputs (Fig. 5–8): For rhythmic inputs, please note that the intracellular and extracellular waveforms for every frequency are provided in supplementary figures S5– S11. It may be noted that the intracellular responses are comparable. In simulations for assessing spike-LFP comparison, we tuned the strengths to produce a single spike per cycle, ensuring fair comparison of LFPs with gap junctions vs. chemical synapses.

Taken together, we demonstrate through explicit sets of simulations and analyses that the differences in LFPs were not driven by the strength or patterns of the inputs but rather by the differences in direct transmembrane currents, which are subsequently reflected in the LFPs. In the revised manuscript, we have emphasized these points in the Discussion section, apart from providing intracellular traces for cases where they were not provided before (new Supplementary Figure 3):

Discussion subsection “Dominance of active dendritic currents with LFP associated with gap junctions”

“Our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. A multitude of factors suggests that the observed LFP differences result not from variations in input strength or patterns but rather from differences in direct transmembrane currents, which are subsequently reflected in the LFP signals.

First, the inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were exclusively excitatory in nature.

Second, the results remained consistent regardless of the number of gap junctions or chemical synapses. (Fig. 1 with 217 gap junctions or 245 chemical synapses and Supplementary Fig. 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

Third, for synchronous chemical synaptic inputs, the waveforms resembled typical postsynaptic potentials. Whereas, for gap junctional inputs, the waveforms showed characteristics of postsynaptic potentials or dendritic spikes (accounting the active nature of inputs from the potential presynaptic cells). Electrical response of postsynaptic cell remains identical, producing an action potential regardless of whether inputs arrive via gap junctions or chemical synapses. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic or gap junctional inputs. We mechanistically analyzed the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Fig. 23).

Fourth, for random inputs, the models were not specifically tuned to generate a single action potential. Here, the inputs served as a proxy for asynchronous inputs arriving from other subregions at random times.

Finally, the intracellular responses were comparable for chemical synaptic and gap junctional rhythmic inputs (Supplementary Fig. S5-S11). Here, the model was tuned to elicit a single spike per cycle in simulations evaluating spike-LFP interactions, ensuring a fair comparison between LFPs from gap junctional and chemical synaptic inputs.”

We have added a new Supplementary Figure 3 to the revised manuscript and have referred to this figure in the Results subsection. We thank you for raising these points as it allowed to elaborate on the several novelties and implications of our methodology and conclusions. Thank you.

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

We thank you for raising this important question. Leak channels were among the several contributors to the positive deflection observed in LFPs associated with gap junctions. This effect was present not only in gap junctional models with intact sodium conductance but also in the no-sodium model, where the amplitude of the positive deflection was reduced across other models as well (Fig. 2F, I). Furthermore, even in the absence of leak conductance, a small positive deflection was still observed (Fig. 2F), leading us to further investigate other transmembrane currents over time and across spatial locations, from the proximal to the distal dendritic ends relative to the soma (Fig. 3D). We had observed that the dominant contributor in the case of chemical synapses was the inward synaptic current (Fig. 3A), whereas for gap junctions, the primary contributors were leak conductance along with other outward currents, such as potassium and HCN currents (Fig. 3D). Together, the direct transmembrane component of chemical synapses provides a dominant contribution to extracellular potentials. This dominance translates to differences in the relative contributions of indirect currents (including leak currents) to extracellular potentials associated chemical synaptic vs. gap junctional inputs. Our analyses of the exact ionic mechanisms (Fig. 3) demonstrates the involvement of several ion channels contributing to the indirect component in either scenario.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

We thank you for raising this important point. We agree with the analyses presented by the reviewer on the importance of network simulations and bidirectional gap junctions that respect the voltages in both neurons. However, the complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses “post-synaptic” currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a pointneuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; Martinez-Canada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications becomes essential.

Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations. Thus, the inputs were not pre-determined by “pre” neurons. Instead, the recorded voltages from potential synaptic partner neurons were randomized across locations and scaled using factors at the dendrites before being injected into the target neuron (Supplementary Fig. S1). While incorporating a network of interconnected neurons is indeed important, we utilized biophysical, morphologically realistic CA1 neuron model with different sets of input patterns to model LFPs, which were derived from the total transmembrane currents across all compartments of the multi-compartmental neuron model. Given the complexity of this approach, adding further network-level interactions or pre-post connections would have been computationally demanding.

In the revised manuscript, we have elaborated on the general methodology used in LFP modeling studies to introduce synaptic currents. We have emphasized that our study extends this approach to modeling gap junctional inputs, while also highlighting randomization of locations and the scaling process in assigning gap junctional synaptic strengths. The following paragraphs were specifically added to the revised version of the manuscript:

Methods subsection “Chemical synaptic and gap junctional inputs: Characteristics and temporal dynamics”:

“The complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses post-synaptic currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a point-neuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; MartinezCanada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications become essential.”

“Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations from potential presynaptic sources.”

We thank for you highlighting these points as it allowed us to place our methodology in the specific context of the literature. Thank you.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

We thank you for raising this point and agree with you on every point. Please note that we do not assert that the outward HCN currents are exclusively associated with gap junctional inputs. Rather, our results show that synchronous inputs generate outward HCN currents in both chemical synapses (Fig. 3B; positive/outward HCN currents, except in the no sodium or leak model) and gap junctions (Fig. 3D; positive/outward HCN currents). We emphasized this in the case of gap junctions because, in the absence of inward synaptic currents, HCN (acting as outward currents with synchronous inputs) contributed to the positive deflection observed in the LFPs. While HCN would also contribute in the case of chemical synapses, its effect was negligible due to the presence of large inward synaptic currents. Since LFPs reflect the collective total transmembrane currents, the dominant contributors differ between these two scenarios, which we aimed to highlight. Since HCN exhibited outward currents in our synchronous input simulations, we have elaborated on this mechanism in the supplementary figure (Fig. S3). Our intention was not to emphasize this effect for only one synaptic mode but rather to highlight HCN's contribution to the positive deflection as one of the contributing factors.

We agree that HCN currents are relatively small in magnitude; therefore, our conclusions were based on HCN being one of the several contributing factors. Leak conductance and other outward conductances, including HCN currents (Fig. 3D), collectively contribute to the positive deflections observed in the case of gap junctional synchronous inputs.

In the revised manuscript, we have provided the following clarifications in the Results subsection on” Synchronous inputs: Outward transmembrane currents from active dendrites contribute to positive deflection in extracellular potentials associated with gap junctional inputs”:

“It is important to note that despite their relatively small magnitude, the outward HCN currents (Fig. 3D) substantially contribute to positive extracellular potential deflections associated with gap junctional inputs (Fig. 2), together with leak and other outward conductances.”

“While outward HCN currents (Fig. 3B) are also expected to influence LFPs under chemical synaptic input, their impact was minimal due to the predominance of large inward synaptic currents (Fig. 3A). As LFPs reflect the summation of all transmembrane currents, the dominant contributors vary across different modes of synaptic transmission.”

We thank you for emphasizing this point. This allowed us to expand on the specific roles of HCN channels and potential contributions of the outward nature of the HCN current. We have also expanded the Discussion subsection on “Outward HCN currents regulate extracellular potentials” to elaborate on this aspect as well. Thank you.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

We thank you for raising this important point. The CA1 pyramidal neuronal model used in this study is built with ion-channel models derived from biophysical and electrophysiological recordings from these cells. As mentioned in the Methods section “Dynamics and distribution of active channels” and Supplementary Table S1, models for individual channels, their gating kinetics, and channel distributions across the somatodendritic arbor (wherever known) are all derived from their physiological equivalents. Importantly, these values were derived from previously validated models from the laboratory, which contain these very ion channel models and the exact same morphology (Roy & Narayanan, 2021). Please compare Supplementary Table S1 with Table 1 from (Roy & Narayanan, 2021). Please note that this model was validated against several physiological measurements along the somatodendritic axis (Fig. 1 of (Roy & Narayanan, 2021)).

In the revised manuscript, we have explicitly mentioned this while also mentioning the different physiological properties that were used for the validation process from (Roy & Narayanan, 2021):

Methods subsection “Pyramidal neuron model”

“All parameters and their corresponding values for the neuronal model were derived from previously validated models (Roy & Narayanan, 2021). These CA1 models were validated against several physiological measurements along the somato dendritic axis (Roy & Narayanan, 2021).”

“These channel distributions and the associated parametric values (Supplementary Table S1) were demonstrated to satisfy 22 different somato-dendritic measurements (Roy & Narayanan, 2021). Specifically, six physiological measurements input resistance, maximal impedance amplitude, resonance frequency, resonance strength, total inductive phase, and back-propagating action potential were validated with respective electrophysiological ranges at three somato-dendritic locations (Soma, ~150 µm dendrite, and ~300 µm dendrite) each (6×3=18 measurements). In addition, action potential firing frequency at each of 100 pA, 150 pA, 200 pA, and 250 pA (4 measurements) were also matched in the model to fall within the respective ranges of corresponding electrophysiological measurements. The electrophysiological ranges of intrinsic measurements were derived from respective somato-dendritic recordings (Malik et al., 2016; Narayanan et al., 2010; Narayanan & Johnston, 2007, 2008; Spruston et al., 1995). Together, the CA1 pyramidal model neuron used in this study was tuned to match several electrophysiological characteristics and ion-channel distributions (Roy & Narayanan, 2021).”

We thank you for pointing us to this slip in elaborating on how the model was validated. We have now rectified this. Thank you.

References

Andersen, P., Morris, R., Amaral, D., Bliss, T., & O'Keefe, J. (2006). The hippocampus book. Oxford University Press.

Basak, R., & Narayanan, R. (2018). Spatially dispersed synapses yield sharply-tuned place cell responses through dendritic spike initiation. Journal of Physiology, 596(17), 4173-4205. https://doi.org/10.1113/JP275310

Bedner, P., Steinhauser, C., & Theis, M. (2012). Functional redundancy and compensation among members of gap junction protein families? Biochim Biophys Acta, 1818(8), 1971-1984. https://doi.org/10.1016/j.bbamem.2011.10.016

Behrens, C. J., Ul Haq, R., Liotta, A., Anderson, M. L., & Heinemann, U. (2011). Nonspecific effects of the gap junction blocker mefloquine on fast hippocampal network oscillations in the adult rat in vitro. Neuroscience, 192, 11-19. https://doi.org/10.1016/j.neuroscience.2011.07.015

Bocian, R., Posluszny, A., Kowalczyk, T., Golebiewski, H., & Konopacki, J. (2009). The effect of carbenoxolone on hippocampal formation theta rhythm in rats: in vitro and in vivo approaches. Brain Res Bull, 78(6), 290-298. https://doi.org/10.1016/j.brainresbull.2008.10.005

Buhl, D. L., Harris, K. D., Hormuzdi, S. G., Monyer, H., & Buzsaki, G. (2003). Selective impairment of hippocampal gamma oscillations in connexin-36 knock-out mouse in vivo. J Neurosci, 23(3), 1013-1018. https://doi.org/10.1523/jneurosci.23-03-01013.2003

Buzsaki, G., Anastassiou, C. A., & Koch, C. (2012). The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nat Rev Neurosci, 13(6), 407-420. https://doi.org/10.1038/nrn3241

Buzsaki, G., & Wang, X. J. (2012). Mechanisms of gamma oscillations. Annual Review of Neuroscience, Vol 36, 35, 203-225. https://doi.org/10.1146/annurev-neuro-062111150444

Colgin, L. L. (2016). Rhythms of the hippocampal network. Nat Rev Neurosci, 17(4), 239249. https://doi.org/10.1038/nrn.2016.21

Colgin, L. L., & Moser, E. I. (2010). Gamma oscillations in the hippocampus. Physiology (Bethesda), 25(5), 319-329. https://doi.org/10.1152/physiol.00021.2010

Coulon, P., & Landisman, C. E. (2017). The Potential Role of Gap Junctional Plasticity in the Regulation of State. Neuron, 93(6), 1275-1295. https://doi.org/10.1016/j.neuron.2017.02.041

Das, A., Rathour, R. K., & Narayanan, R. (2017). Strings on a Violin: Location Dependence of Frequency Tuning in Active Dendrites. Front Cell Neurosci, 11, 72. https://doi.org/10.3389/fncel.2017.00072

Draguhn, A., Traub, R. D., Schmitz, D., & Jefferys, J. G. (1998). Electrical coupling underlies high-frequency oscillations in the hippocampus in vitro. Nature, 394(6689), 189-192. https://doi.org/10.1038/28184

Einevoll, G. T., Destexhe, A., Diesmann, M., Grun, S., Jirsa, V., de Kamps, M., Migliore, M., Ness, T. V., Plesser, H. E., & Schurmann, F. (2019). The Scientific Case for Brain Simulations. Neuron, 102(4), 735-744. https://doi.org/10.1016/j.neuron.2019.03.027

Einevoll, G. T., Kayser, C., Logothetis, N. K., & Panzeri, S. (2013). Modelling and analysis of local field potentials for studying the function of cortical circuits. Nat Rev Neurosci, 14(11), 770-785. https://doi.org/10.1038/nrn3599

Gold, C., Henze, D. A., Koch, C., & Buzsaki, G. (2006). On the origin of the extracellular action potential waveform: A modeling study. J Neurophysiol, 95(5), 3113-3128. https://doi.org/10.1152/jn.00979.2005

Hagen, E., Dahmen, D., Stavrinou, M. L., Linden, H., Tetzlaff, T., van Albada, S. J., Grun, S., Diesmann, M., & Einevoll, G. T. (2016). Hybrid Scheme for Modeling Local Field Potentials from Point-Neuron Networks. Cereb Cortex, 26(12), 4461-4496. https://doi.org/10.1093/cercor/bhw237

Halnes, G., Ness, T. V., Næss, S., Hagen, E., Pettersen, K. H., & Einevoll, G. T. (2024). Electric Brain Signals: Foundations and Applications of Biophysical Modeling. Cambridge University Press. https://doi.org/10.1017/9781009039826

Hormuzdi, S. G., Pais, I., LeBeau, F. E., Towers, S. K., Rozov, A., Buhl, E. H., Whittington, M. A., & Monyer, H. (2001). Impaired electrical signaling disrupts gamma frequency oscillations in connexin 36-deficient mice. Neuron, 31(3), 487-495. https://doi.org/10.1016/s0896-6273(01)00387-7

Hussaini, S. A., Kempadoo, K. A., Thuault, S. J., Siegelbaum, S. A., & Kandel, E. R. (2011). Increased size and stability of CA1 and CA3 place fields in HCN1 knockout mice. Neuron, 72(4), 643-653. https://doi.org/10.1016/j.neuron.2011.09.007

Johnston, D., & Narayanan, R. (2008). Active dendrites: colorful wings of the mysterious butterflies. Trends Neurosci, 31(6), 309-316. https://doi.org/10.1016/j.tins.2008.03.004

Kessi, M., Peng, J., Duan, H., He, H., Chen, B., Xiong, J., Wang, Y., Yang, L., Wang, G., Kiprotich, K., Bamgbade, O. A., He, F., & Yin, F. (2022). The Contribution of HCN Channelopathies in Different Epileptic Syndromes, Mechanisms, Modulators, and Potential Treatment Targets: A Systematic Review. Front Mol Neurosci, 15, 807202. https://doi.org/10.3389/fnmol.2022.807202

Kole, M. H., Hallermann, S., & Stuart, G. J. (2006). Single Ih channels in pyramidal neuron dendrites: properties, distribution, and impact on action potential output [Research Support, Non-U.S. Gov't]. J Neurosci, 26(6), 1677-1687. https://doi.org/10.1523/JNEUROSCI.3664-05.2006

Konopacki, J., Kowalczyk, T., & Golebiewski, H. (2004). Electrical coupling underlies theta oscillations recorded in hippocampal formation slices. Brain Res, 1019(1-2), 270-274. https://doi.org/10.1016/j.brainres.2004.05.083

Larkum, M. E., Wu, J., Duverdin, S. A., & Gidon, A. (2022). The Guide to Dendritic Spikes of the Mammalian Cortex In Vitro and In Vivo. Neuroscience, 489, 15-33. https://doi.org/10.1016/j.neuroscience.2022.02.009

LeBeau, F. E., Traub, R. D., Monyer, H., Whittington, M. A., & Buhl, E. H. (2003). The role of electrical signaling via gap junctions in the generation of fast network oscillations. Brain Res Bull, 62(1), 3-13. https://doi.org/10.1016/j.brainresbull.2003.07.004

Lo, C. W. (1999). Genes, gene knockouts, and mutations in the analysis of gap junctions. Dev Genet, 24(1-2), 1-4. https://doi.org/10.1002/(SICI)1520-6408(1999)24:1/2%3C1::AID-DVG1%3E3.0.CO;2-U

Lorincz, A., Notomi, T., Tamas, G., Shigemoto, R., & Nusser, Z. (2002). Polarized and compartment-dependent distribution of HCN1 in pyramidal cell dendrites. Nat Neurosci, 5(11), 1185-1193. https://doi.org/10.1038/nn962

Magee, J. C. (1998). Dendritic hyperpolarization-activated currents modify the integrative properties of hippocampal CA1 pyramidal neurons. J Neurosci, 18(19), 7613-7624. https://doi.org/10.1523/jneurosci.18-19-07613.1998

Magee, J. C., & Grienberger, C. (2020). Synaptic Plasticity Forms and Functions. Annual Review of Neuroscience, Vol 36, 43, 95-117. https://doi.org/10.1146/annurev-neuro090919-022842

Major, G., Larkum, M. E., & Schiller, J. (2013). Active properties of neocortical pyramidal neuron dendrites [Review]. Annual Review of Neuroscience, Vol 36, 36, 1-24. https://doi.org/10.1146/annurev-neuro-062111-150343

Malik, R., Dougherty, K. A., Parikh, K., Byrne, C., & Johnston, D. (2016). Mapping the electrophysiological and morphological properties of CA1 pyramidal neurons along the longitudinal hippocampal axis. Hippocampus, 26(3), 341-361. https://doi.org/10.1002/hipo.22526

Martinez-Canada, P., Ness, T. V., Einevoll, G. T., Fellin, T., & Panzeri, S. (2021). Computation of the electroencephalogram (EEG) from network models of point neurons. PLoS Comput Biol, 17(4), e1008893. https://doi.org/10.1371/journal.pcbi.1008893

Mazzoni, A., Linden, H., Cuntz, H., Lansner, A., Panzeri, S., & Einevoll, G. T. (2015). Computing the Local Field Potential (LFP) from Integrate-and-Fire Network Models. PLoS Comput Biol, 11(12), e1004584. https://doi.org/10.1371/journal.pcbi.1004584

Mishra, P., & Narayanan, R. (2021). Stable continual learning through structured multiscale plasticity manifolds. Curr Opin Neurobiol, 70, 51-63. https://doi.org/10.1016/j.conb.2021.07.009

Mishra, P., & Narayanan, R. (2025). The enigmatic HCN channels: A cellular neurophysiology perspective. Proteins, 93(1), 72-92. https://doi.org/10.1002/prot.26643

Moore, J. J., Ravassard, P. M., Ho, D., Acharya, L., Kees, A. L., Vuong, C., & Mehta, M. R. (2017). Dynamics of cortical dendritic membrane potential and spikes in freely behaving rats. Science, 355(6331). https://doi.org/10.1126/science.aaj1497

Narayanan, R., Dougherty, K. J., & Johnston, D. (2010). Calcium Store Depletion Induces Persistent Perisomatic Increases in the Functional Density of h Channels in Hippocampal Pyramidal Neurons. Neuron, 68(5), 921-935. https://doi.org/10.1016/j.neuron.2010.11.033

Narayanan, R., & Johnston, D. (2007). Long-term potentiation in rat hippocampal neurons is accompanied by spatially widespread changes in intrinsic oscillatory dynamics and excitability. Neuron, 56(6), 1061-1075. https://doi.org/10.1016/j.neuron.2007.10.033

Narayanan, R., & Johnston, D. (2008). The h channel mediates location dependence and plasticity of intrinsic phase response in rat hippocampal neurons. J Neurosci, 28(22), 5846-5860. https://doi.org/10.1523/JNEUROSCI.0835-08.2008

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2016). Active subthreshold dendritic conductances shape the local field potential. Journal of Physiology, 594(13), 38093825. https://doi.org/10.1113/JP272022

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2018). h-Type Membrane Current Shapes the Local Field Potential from Populations of Pyramidal Neurons. J Neurosci, 38(26), 6011-6024. https://doi.org/10.1523/jneurosci.3278-17.2018

Neves, G., Cooke, S. F., & Bliss, T. V. (2008). Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. Nat Rev Neurosci, 9(1), 65-75. https://doi.org/10.1038/nrn2303

Nolan, M. F., Malleret, G., Dudman, J. T., Buhl, D. L., Santoro, B., Gibbs, E., Vronskaya, S., Buzsaki, G., Siegelbaum, S. A., Kandel, E. R., & Morozov, A. (2004). A behavioral role for dendritic integration: HCN1 channels constrain spatial memory and plasticity at inputs to distal dendrites of CA1 pyramidal neurons. Cell, 119(5), 719-732. https://doi.org/10.1016/j.cell.2004.11.020

O'Brien, J. (2014). The ever-changing electrical synapse. Curr Opin Neurobiol, 29, 64-72. https://doi.org/10.1016/j.conb.2014.05.011

O'Keefe, J., & Recce, M. L. (1993). Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus, 3(3), 317-330. https://doi.org/10.1002/hipo.450030307

Pereda, A. E. (2014). Electrical synapses and their functional interactions with chemical synapses. Nat Rev Neurosci, 15(4), 250-263. https://doi.org/10.1038/nrn3708

Posluszny, A. (2014). The contribution of electrical synapses to field potential oscillations in the hippocampal formation. Front Neural Circuits, 8, 32. https://doi.org/10.3389/fncir.2014.00032

Reimann, M. W., Anastassiou, C. A., Perin, R., Hill, S. L., Markram, H., & Koch, C. (2013). A biophysically detailed model of neocortical local field potentials predicts the critical role of active membrane currents. Neuron, 79(2), 375-390. https://doi.org/10.1016/j.neuron.2013.05.023

Rouach, N., Segal, M., Koulakoff, A., Giaume, C., & Avignone, E. (2003). Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions. Journal of Physiology, 553(Pt 3), 729-745. https://doi.org/10.1113/jphysiol.2003.053439

Roy, A., & Narayanan, R. (2021). Spatial information transfer in hippocampal place cells depends on trial-to-trial variability, symmetry of place-field firing, and biophysical heterogeneities. Neural Netw, 142, 636-660. https://doi.org/10.1016/j.neunet.2021.07.026

Schomburg, E. W., Anastassiou, C. A., Buzsaki, G., & Koch, C. (2012). The spiking component of oscillatory extracellular potentials in the rat hippocampus. J Neurosci, 32(34), 11798-11811. https://doi.org/10.1523/JNEUROSCI.0656-12.2012

Seenivasan, P., & Narayanan, R. (2020). Efficient phase coding in hippocampal place cells. Physical Review Research, 2(3), 033393. https://doi.org/10.1103/PhysRevResearch.2.033393

Sinha, M., & Narayanan, R. (2015). HCN channels enhance spike phase coherence and regulate the phase of spikes and LFPs in the theta-frequency range. Proc Natl Acad Sci U S A, 112(17), E2207-2216. https://doi.org/10.1073/pnas.1419017112

Sinha, M., & Narayanan, R. (2022). Active Dendrites and Local Field Potentials: Biophysical Mechanisms and Computational Explorations. Neuroscience, 489, 111-142. https://doi.org/10.1016/j.neuroscience.2021.08.035

Sirmaur, R., & Narayanan, R. (2024). Distinct extracellular signatures of chemical and electrical synapses impinging on active dendrites differentially contribute to ripplefrequency oscillations. Society for Neuroscience annual meeting, Chicago, USA.

Spruston, N., Schiller, Y., Stuart, G., & Sakmann, B. (1995). Activity-dependent action potential invasion and calcium influx into hippocampal CA1 dendrites [Research Support, Non-U.S. Gov't]. Science, 268(5208), 297-300. https://doi.org/10.1126/science.7716524

Stuart, G. J., & Spruston, N. (2015). Dendritic integration: 60 years of progress. Nat Neurosci, 18(12), 1713-1721. https://doi.org/10.1038/nn.4157

Szarka, G., Balogh, M., Tengolics, A. J., Ganczer, A., Volgyi, B., & Kovacs-Oller, T. (2021). The role of gap junctions in cell death and neuromodulation in the retina. Neural Regen Res, 16(10), 1911-1920. https://doi.org/10.4103/1673-5374.308069

Traub, R. D., Cunningham, M. O., Gloveli, T., LeBeau, F. E., Bibbig, A., Buhl, E. H., & Whittington, M. A. (2003). GABA-enhanced collective behavior in neuronal axons underlies persistent gamma-frequency oscillations. Proc Natl Acad Sci U S A, 100(19), 11047-11052. https://doi.org/10.1073/pnas.1934854100

Vaughn, M. J., & Haas, J. S. (2022). On the Diverse Functions of Electrical Synapses. Front Cell Neurosci, 16, 910015. https://doi.org/10.3389/fncel.2022.910015

Wang, X. J. (2010). Neurophysiological and computational principles of cortical rhythms in cognition. Physiol Rev, 90(3), 1195-1268. https://doi.org/10.1152/physrev.00035.2008

Wang, X. J., & Buzsaki, G. (1996). Gamma oscillation by synaptic inhibition in a hippocampal interneuronal network model. J Neurosci, 16(20), 6402-6413. https://doi.org/10.1523/jneurosci.16-20-06402.1996

Whittington, M. A., Traub, R. D., & Jefferys, J. G. (1995). Synchronized oscillations in interneuron networks driven by metabotropic glutamate receptor activation. Nature, 373(6515), 612-615. https://doi.org/10.1038/373612a0

Williams, S. R., & Stuart, G. J. (2000). Site independence of EPSP time course is mediated by dendritic I(h) in neocortical pyramidal neurons [In Vitro]. J Neurophysiol, 83(5), 3177-3182. https://doi.org/10.1152/jn.2000.83.5.3177

eLife Assessment

In this useful paper, the authors present a comprehensive method for the purification of recombinant Snake Venom Metalloproteinases (SVMPs) using the MultiBac expression system, explain the self-activation of the enzymes by Zn2+ incubation, and establish high-throughput screening (HTS) techniques. The authors addressed a key problem: producing a substantial amount of pure and enzymatically active SVMPs required for structural and functional studies. Altogether, this work builds a solid foundation for the large-scale production of active SVMPs for future biochemical and structural characterization as well as for drug discovery, albeit leaving certain caveats about the universal applicability of the described methodology for the production of any recombinant SVMPs.

Reviewer #1 (Public review):

Summary:

The authors Hall et al. establish a purification method for snake venom metalloproteinases (SVMPs). By generating a generic approach to purify this divergent class of recombinant proteins, they enhance the field's accessibility to larger quantity SVMPs with confirmed activity and, for some, characterized kinetics. In some cases, the recombinant protein displayed comparable substrate specificity and substrate recognition compared to the native enzyme, providing convincing evidence of the authors' successful recombinant expression strategy. Beyond describing their route towards protein purification, they further provide evidence for self-activation upon Zn2+ incubation. They further provide initial insights on how to design high throughput screening (HTS) methods for drug discovery and outline future perspectives for the in-depth characterization of these enzyme classes to enable the development of novel biomedical applications.

Strengths:

The study is well presented and structured in a compelling way and the universal applicability of the approach is nicely presented.<br /> The purification strategy results in highly pure protein products, well characterized by size exclusion chromatography, SDS page as well as confirmed by mass spectrometry analysis. Further, a significant portion of the manuscript focuses on enzyme activity, thereby validating function. Particularly convincing is the comparability between recombinant vs. native enzymes; this is successfully exemplified by insulin B digestion. By testing the fluorogenic substrate, the authors provide evidence that their production method of recombinant protein can open up possibilities in HTS. Since their purification method can be applied to three structurally variable SVMP classes, this demonstrates the robust nature of the approach.

Weakness

The product obtained from the purification protocol appears to be a heterogenous mixture of self-activated and intact protein species. The protocol would benefit from improved control over the self-activation process. The authors explain well why they cannot deplete Zn2+ in cell culture or increase the pH to prevent autoactivation during the current purification steps. However, this leads me to the suggestion, if the His tag could be exchanged to a different tag that is less pH sensitive and not dependent on divalent ions (Strep-Tactin XT?) to allow for removal of divalent ions and low pH during purification steps. Another suggestion would be if they could replace the endogenous protease cleavage site in their expression construct design to a TEV protease recognition site, for example, to have more control over activation of the recombinant proteins.

The graphic to explain the universal applicability of the approach, Figure S1, has some mistakes, like duplication of text, an arrow without a meaning and should be revised.

Overall, the authors successfully purified active SVMP proteins of all three structurally diverse classes in high quality and provided convincing evidence throughout the manuscript to support their claims. The described method will be of use for a broader community working with self-activating and cytotoxic proteases.

Comment on the revised version:

I find that the clarity and overall structure of the manuscript have improved. However, the weakness I previously highlighted has neither been addressed experimentally nor convincingly explained. Therefore, the assessment stayed unchanged from my side.

Reviewer #2 (Public review):

Summary:

The aim of the study by Hall et al. was to establish a generic method for production of Snake Venom Metalloproteases (SVMPs). These have been difficult to purify in the mg quantities required for mechanistic biochemical and structural studies.

Strengths:

The authors have successfully applied the MultiBac system and describe with a high level of details, the downstream purification methods applied to purify the SVMP PI, PII and PIII. The paper carefully presents the non-successful approaches taken (such as expression of mature proteins, the use of protease inhibitors, prodomain segments and co-expression of disulfide-isomerases) before establishing the construct and expression conditions required. The authors finally convincingly describe various activity assays to demonstrate the activity of the purified enzymes in a variety of established SVMP assays.

Weaknesses:

Some experiments are difficult to perform with relevant controls (i.e. native SVMP from the venome), but authors have explained this and provided the best possible assessment.

Overall, the data presented demonstrates a very credible path for production of active SVMP for further downstream characterization. The generality of the approach to all SVMP from different snakes remains to be demonstrated by the community, but if generally applicable, the method will enable numerous studies with the aim of either utilizing SVMPS as therapeutic agents or to enable generation of specific anti-venom reagents such as antibodies or small molecule inhibitors.

Comment on the revised version:

I think the manuscript has benefited from the review and the revised version provides more clarity, is more concise and reads significantly better with the preliminary data/experiments moved to the supplements. My overall assessment of the manuscript remains unchanged.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public review):

Summary:

The authors Hall et al. establish a purification method for snake venom metalloproteinases (SVMPs). By generating a generic approach to purify this divergent class of recombinant proteins, they enhance the field's accessibility to larger quantities of SVMPs with confirmed activity and, for some, characterized kinetics. In some cases, the recombinant protein displayed comparable substrate specificity and substrate recognition compared to the native enzyme, providing convincing evidence of the authors' successful recombinant expression strategy. Beyond describing their route towards protein purification, they further provide evidence for self-activation upon Zn2+ incubation. They further provide insights on how to design high-throughput screening (HTS) methods for drug discovery and outline future perspectives for the in-depth characterization of these enzyme classes to enable the development of novel biomedical applications.

Strengths:

The study is well-presented and structured in a compelling way. The purification strategy results in highly pure protein products, well characterized by size exclusion chromatography, SDS page as well as confirmed by mass spectrometry analysis. Further, a significant portion of the manuscript focuses on enzyme activity, thereby validating function. Particularly convincing is the comparability between recombinant vs. native enzymes; this is successfully exemplified by insulin B digestion. By testing the fluorogenic substrate, the authors provide evidence that their production method of recombinant protein can open up possibilities in HTS. Since their purification method can be applied to three structurally variable SVMP classes, this demonstrates the robust nature of the approach.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The universal applicability of the approach could be emphasized more clearly. The potential for this generic protocol for recombinant SVMP zymogen production to be adapted to other SVMPs is somewhat obscured by the detailed optimization steps. A general schematic overview would strengthen the manuscript, presented as a final model, to illustrate how this strategy can be extended to other targets with similar features. Such a schematic might, for example, outline the propeptide fusion design, including its tags, relevant optimizations during expression, lysis, purification (e.g., strategies for metal ion removal and maintenance of protease inactivity), as well as the controllable auto-activation.

In the revised version of the manuscript, we moved the detailed description of the optimisation of SVMP expression, including mature SVMP expression, Marimastat addition, active site mutations and fusion of propeptides, into the supplement as supplementary text. We hope this improves the clarity and flow. As suggested, we now include a new figure outlining the SVMP production strategy and optimisation steps in the revised manuscript (new Figure S1).

The product obtained from the purification protocol appears to be a heterogeneous mixture of selfactivated and intact protein species. The protocol would benefit from improved control over the selfactivation process. The Methods section does not indicate whether residual metal ions were attempted to be removed during the purification, which could influence premature activation.

We agree that improved control of self-activation would be desirable. However, there is an issue: Previous studies reported that (1) SVMP zymogens are processed within secretory cells of the venom gland (Portes-Junior et al., 2014), and (2) mature SVMPs accumulate in secretory vesicles during venom production (Carneiro et al., 2002). Accordingly, preventing the auto-processing of SVMP zymogens is difficult to achieve because this would require Zn²⁺ depletion within the insect cells during production which would result in cytotoxicity. We have included this information in the updated Discussion section of the revised manuscript.

Additionally, it has not been discussed whether the shift to pH 8 in the purification process is necessary from the initial steps onwards, given that a lower pH would be expected to maintain enzyme latency.

The shift to pH 8 is required for the affinity purification of the SVMP zymogens from the medium, involving the poly-histidine-tag and immobilized metal affinity chromatography (IMAC). At lower pH, the histidines would become protonated, preventing binding of the His-tag to the column. Thus, with the His-tag the shift to pH 7.5 or pH 8 is necessary.

The characterization of PIII activity using the fluorogenic peptide effectively links the project to its broader implications for drug design. However, the absence of comparable solutions for PI and PII classes limits the overall scope and impact of the finding.

We agree that such assays would be extremely useful. However, the development of fluorescence based high-throughput assays to test for PI and PII SVMP activity is beyond the scope of this study. Here, our overarching objective is to report a broadly applicable production method for PI, PII and PIII SVMPs.

Overall, the authors successfully purified active SVMP proteins of all three structurally diverse classes in high quality and provided convincing evidence throughout the manuscript to support their claims. The described method will be of use for a broader community working with self-activating and cytotoxic proteases.

Thank you.

Reviewer #2 (Public review):

Summary:

The aim of the study by Hall et al. was to establish a generic method for the production of Snake Venom Metalloproteases (SVMPs). These have been difficult to purify in the mg quantities required for mechanistic, biochemical, and structural studies.

Strengths:

The authors have successfully applied the MultiBac system and describe with a high level of detail the downstream purification methods applied to purify the SVMP PI, PII, and PIII. The paper carefully presents the non-successful approaches taken (such as expression of mature proteins, the use of protease inhibitors, prodomain segments, and co-expression of disulfide-isomerases) before establishing the construct and expression conditions required. The authors finally convincingly describe various activity assays to demonstrate the activity of the purified enzymes in a variety of established SVMP assays.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The manuscript suffers from a lack of bottoming out and stringent scientific procedures in the methodology and the characterization of the generated enzymes.

As an example, a further characterization of the generated protein fragments in Figure 3 by intact mass spectroscopy would have aided in accurate mass determination rather than relying on SEC elution volumes against a standard. Protein shape and charge can affect migration in SEC.

We agree that intact MS would be useful to determine the mass of the produced SVMPs. In this manuscript, we performed SEC as a purification step, removing aggregates. Furthermore, SEC allowed determining if the SVMPs form monomers or dimers. MS characterisation of intact SVMPs (and their PTMs) is not trivial and beyond the scope of this manuscript (see below).

Also, the analysis of N-linked glycosylation demonstrates some reactivity of PIII to PNGase F, but fails to conclude whether one or more sites are occupied, or whether other types of glycosylation is present. Again, intact mass experiments would have resolved such issues.

We concur that glycosylation of SVMPs is an important question. However, analysing the glycosylation of the SVMPs is beyond the scope of this manuscript; it is actually a project on its own: Intact MS can indeed provide information on glycosylation but is not very precise. Unambiguous assignment of the number and occupancy of glycosylation sites is more challenging, especially for large, glycosylated proteins such as our PIII SVMP zymogen. In practice, confident mapping of glycosylation sites would require peptide-level mass spectrometry following enzymatic digestion (Trypsin and Multi-Enzymatic Limited Digestion, ideally). Sample preparation, method optimization, MS acquisition, and data analysis together would require a significant investment. Moreover, we do not have access to the native PIII SVMP from Echis carinatus sochureki venom - this is the main point of our manuscript: we describe a protocol to produce SVMPs which could not be purified from venom. Therefore, a comparison of the glycosylation of the recombinant SVMP and the native SVMP cannot be performed unfortunately (see below).

The activity assays in Figure 4 are not performed consistently with kinetic assays and degradation assays performed for some, but not all, enzymes, and there is no Echis ocellatus comparison in Figure 4h.

This is correct. The suggested control experiment is not possible for the PII SVMP and PIII SVMP because we cannot purify the native PII and PIII SVMPs from Echis venom. We have highlighted this information in the revised manuscript in the insulin B degradation section.

Overall, whilst not affecting the main conclusion, this leaves the reader with an impression of preliminary data being presented. For consistency, application of the same assays to all enzymes (high-grade purified) would have provided the reader with a fuller picture.

In the revised manuscript, we included new data showing the requested characterisations of all three SVMPs.

We have included the respective assays in Figure 5 and Supplementary Figure S11. In the original manuscript, we had omitted these assays as the data show no enzymatic activity in the respective assays. Specifically, we show that (1) PII does not cause insulin B degradation (Fig. S11b), (2) that the PI and PII SVMPs do not degrade the fluorogenic peptide which is prototypic for PIII SVMPs and MMPs (Fig. S11a), (3) PI and PIII do not cause platelet aggregation because they lack the entire disintegrin domain (PI) or the RGD motif (PIII) (Fig. 5a), and (4) that the PI and PII SVMPs, like the PIII SVMP, are not pro-coagulant and do not cause blood clotting (Fig. 5d,5e and Fig. S11c). We also included this new information in the main text of our revised manuscript.

Overall, the data presented demonstrates a very credible path for the production of active SVMP for further downstream characterization. The generality of the approach to all SVMP from different snakes remains to be demonstrated by the community, but if generally applicable, the method will enable numerous studies with the aim of either utilizing SVMPS as therapeutic agents or to enable the generation of specific anti-venom reagents, such as antibodies or small molecule inhibitors.

Thank you.

Reviewer #3 (Public review):

Summary:

The presented study describes the long journey towards the expression of members' SVMP toxins from snake venom, which are toxins of major importance in a snakebite scenario. As in the past, their functional analysis relied on challenging isolation; the toxins' heterologous expression offers a potential solution to some major obstacles hindering a better understanding of toxin pathophysiology. Through a series of laborious and elegantly crafted experiments, including the reporting of various failed attempts, the authors establish the expression of all three SVMP subtypes and prove their activity in bioassays. The expression is carried out as naturally occurring zymogens that autocleave upon exposure to zinc, which is a novel modus operandi for yielding fusion proteins and sheds also some new light on the potential mechanism that snakes use to activate enzymatic toxins from zymogenic preforms.

Strengths:

The manuscript draws from an extensive portfolio of well-reasoned and hypothesis-driven experiments that lead to a stepwise solution. The wetlands data generated is outstanding, although not all experiments along this rocky road to victory were successful. A major strength of the paper is that, translationally speaking, it opens up novel routes for biodiscovery since a first reliable platform for expression of an understudied, yet potent toxin class is established. The discovered strategy to pursue expression as zymogens could see broad application in venom biotechnology, where several toxin types are pending successful expression. The work further provides better insights into how snake toxins are processed.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The manuscript contains several chapters reporting failed experiments, which makes it difficult to follow in places.

Based on a similar comment of Reviewer 1, we now moved the ‘failed’ experiments reporting on SVMP expression optimisation to the supplement as new supplementary text. We hope that the revisions have improved the clarity and overall readability of our manuscript.

The reporting of experimental details, especially sample sizes and replicates, could be optimised.

The number of replicates has now been added to the figure legends in the revised manuscript. Detailed experimental information is found in the revised Methods part.

At the time of writing, it remains unclear whether the glycosilations detected at a pIII SVMP could have an impact on the bioactivities measured, which is a major aspect, and future follow-ups should clarify this.

A detailed analysis of glycosylation of the PIII SVMP is beyond the scope of our manuscript (see above, response to Reviewer 2). Our manuscript describes a generic protocol to produce active SVMPs. Importantly, we cannot purify the native PIII SVMP from Echis carinatus sochureki venom. Therefore, it is not possible to compare our PIII SVMP with the native PIII SVMP.

We agree that this is an important question, and we will aim in the future to perform such a comparison of a different insect cell-produced PIII with a native PIII SVMP that can be readily purified from venom.

Finally, the work, albeit of critical importance, would benefit from a more down-to-earth evaluation of its findings, as still various persistent obstacles that need to be overcome.

We consider cytotoxicity to be the principal bottleneck in SVMP production. In this study, we present a strategy to overcome this bottleneck.

Major comments to the manuscript:

(1) Lines 148-149: "indicating that expressing inactivated SVMPs could be a viable, although inefficient, approach". I think this text serves a good purpose to express some thoughts on the nature of how the current draft is set up. It is quite established that various proteases cause extreme viability losses to their expression host (whether due to toxicity, but surely also because of metabolic burden), which is why their expression as inactive fusion proteins is the default strategy in all cases I have thus far seen. I believe that, especially in venom studies, this is of importance given the increased toxicity often targeting cellular integrity, and especially here, because Echis are known to feed on arthropods at younger life history stages, making it very likely that some venom components are especially active against insects and other invertebrates. With that in mind, I would argue that exploring their production in inactive form is the obvious strategy one would come up with and not really the conclusion of a series of (well-conducted and scientifically sound!) experiments. For me, the insight of inactive expression is largely confirmatory of what is established, unless I miss something in the authors' rationale. If yes, it would be important to clarify that in the online version.

We agree that producing zymogens represents a straightforward strategy and now, in hindsight, would have wished we had tested this first thing, it would have saved us and apparently many others significant effort. However, realising this, and implementing this approach took us considerable time and insight as we described in this manuscript. The alternative strategies we describe in the manuscript, in particular the use of inhibitors and active-site mutation, have been successfully applied for recombinant production of diverse enzymes before, including enzymes that are toxic to host cells.

We have revised the manuscript as requested and moved the optimisation of SVMP expression to the Supplement. We hope this improved the clarity, overall readability of the text and thus addressed the reviewer’s comment.

(2) Line 173: Here, Alphafold 3 was used, whereas in previous sections (e.g., line 153, line 210), it was Alphafold 2. I suggest using one release across the manuscript.

Thank you for bringing this to our attention. In the revised version of the manuscript, we clarified that all models were generated using AlphaFold 3.

(3) Line 252-254: I fully agree, the PIII SVMP is glycosylated. Glycosylation is an important mediator of snake venom activity, and several works have described their importance in the field. This raises the question, which glycosylations have been introduced here in the SVMP, and to verify that these are glycosylations that belong to those found in snakes. This is important as insects facilitate thousands of N- and O- O-glycosylations to modulate the activity of their proteome, of which many are specific to insects. If some of these were integrated into the SVMP, this could have an impact on downstream produced bioassays and also antigenicity (the surface would be somewhat different from natural toxins, causing different selection).

We agree that glycosylation is important and warrants a follow-up in the future.

However, most publications we found reported that de-glycosylation has a negative effect on stability and solubility of SVMPs, which is expected to have a knock-on effect on toxin activity (e.g. AndradeSilva et al., 2025; DOI: 10.1021/acs.jproteome.5c00249). It will be difficult to separate the two effects from each other. We found only a few examples where SVMP glycosylation (sialylation and Nglycosylation) modulated proteolytic and haemorrhagic functions, including interaction with substrates such as e.g. fibrinogen (Schluga et al., 2024; https://doi.org/10.3390/toxins16110486; Chen et al., 2008; 10.1111/j.1742-4658.2008.06540.x; Nikai et al., 2000; DOI: 10.1006/abbi.2000.1795. PMID: 10871038). In our manuscript, we show that our PIII SVMP is very cytotoxic and highly active in casein, fibrinogen and ESO10 degradation assays, with a K_M and k_cat/K_M comparing favourably with other SVMPs and MMPs. We are not aware of a specific substrate for this particular PIII SVMP that depends on a distinct glycosylation pattern. Recombinant production of such SVMPs with specific glycosylation pattern requirement would be a challenge in all commonly used expression systems (yeast, plant, insect cells and mammalian cells). In fact, insect cell expression systems could be advantageous in this respect because the Sf21 and High Five (Hi5) lepidopteran cell lines we utilised are well-characterized for their ability to perform posttranslational modifications on complex secreted proteins:

(1) N-Glycan conservation: Both Sf21 and Hi5 cells typically produce N-glycans that are trimmed to a core 'paucimannose' structure (Man3GlcNAc2), often with an alpha1,6-fucosylation. While snakes can produce more complex, sialylated N-glycans, glycomic studies of native venoms (e.g., Bothrops venom) have demonstrated that high-mannose and paucimannose structures are also prevalent in native SVMPs. Therefore, the recombinant glycoforms produced in our system are not 'unnatural' in the snake venom context but rather represent a subset of the native glycan microheterogeneity.

(2) Occupancy vs structure: The critical function of glycosylation in PIII SVMPs is thought to be often structural, facilitating correct folding and protecting the large metalloprotease and disintegrin-like domains from proteolytic degradation. Because Sf21 and Hi5 cells recognize the same Nglycosylation sequon (Asn-X-Ser/Thr) as reptilian cells, the site-occupancy remains consistent with the native protein, preserving the overall topography of the toxin.

(3) Activity and authentic self-processing: We acknowledge that insect-specific alpha1,3-fucosylation can occur in Hi5 cells and is potentially antigenic. As the recombinant SVMPs will be used for binder selections and for testing in silico designed binders, useful binders will be selected based on neutralising activity against venom toxins. Here, our assays focused on auto-activation and proteolytic activity, which is primarily driven by the catalytic Zn²⁺-site and the protein backbone.

As stated above, analysis of glycosylation pattern of the PIII SVMP is a project on its own and beyond the scope of this manuscript.

We have incorporated some of the above information into the discussion section of the revised manuscript to clarify that insect cell glycosylation does not recapitulate the full diversity of SVMP glycosylation observed in native venoms.

(4) General comment for the bioassays: It would be good to specify the replicates again and report the data, including standard deviations.

We included this information in the figure legends.

Discussion:

I think the data generated in the study is very valuable and will be instrumental for pushing the frontiers in SVMP research, but still I would like to see a bit of modesty in their discussion. As I have pointed out above, it is unclear which effect the glycosilations may have (i.e., are the glycosilations found reminiscent of natural ones?), despite their being functionally important. Also, yes, isolation of SVMPs is challenging, but the reality is that their expression is equally challenging, as evidenced by the heaps of presented negative data (with which I have no problems, I think reporting such is actually important). So far, the "generic" protocol has been used to express one member per structural class of Echis SVMP, but no evidence is provided that it would work equally well on other members from taxonomically more distant snakes (e.g., the pIII known from Naja oxiana). It is very likely, but at the time of writing, purely speculative.

We have expressed additional PIII SVMPs from Echis and Daboia species and will report their production and characterisation in due course.

Lastly, the reality is also that the expression in insect cells can only be carried out by highly specialized labs (even in the expression world, as most laboratories work with bacterial or fungal hosts), whereas the isolation can be attempted in most venom labs. That said, production in insect cells also has economic repercussions as it will be very challenging to generate yields that are economically viable versus other systems, which is pivotal because the authors talk about bioprospecting and the toxins used in snakebite agent research.

We thank the reviewer for this perspective on the practicalities of protein expression. However, we respectfully disagree with the characterization of insect cell expression as an inaccessible or economically non-viable platform for toxin research. We offer the following points:

(1) Prevalence and accessibility: Contrary to the suggestion that insect cell expression is restricted to highly specialized labs, the Baculovirus Expression Vector System (BEVS) has become a cornerstone of modern biologics production, structural biology and biochemistry. For instance, our MultiBac system (which is but one of several systems currently widely in use) is utilised by over 1,000 laboratories and institutions, academic and pharma/biotech, worldwide. The maturation of commercially available kits, automated platforms, and standardized protocols has moved this technology into the mainstream, making it a standard tool for any lab requiring high-quality eukaryotic proteins.

(2) Biological necessity: Bacterial (E. coli) and fungal (P. pastoris) systems are widely accessible, however, they appear to be fundamentally incapable of producing functional SVMPs. SVMPs require complex disulfide-bond formation, intricate folding, and N-glycosylation for stability and solubility. Bacterial systems have been widely tried by us and others but typically result in very low expression or misfolded inclusion bodies. Of note, originally, we had invested significant effort to adapt P. pastoris to the production of eukaryotic proteins we are interested in, without success, before moving on to the MultiBac system. The SVMPs that we analysed here are highly cytotoxic, rendering the baculovirus/insect cell system in a way a logical choice given that the cells are no longer 'living' after infection with the baculovirus (but more akin membrane-enveloped bioreactors). Thus, one can make the argument that insect cells represent the most accessible middle ground that provides folding apparatus and necessary post-translational modifications (PTMs) required for biological relevance, and it is possible to produce mg amounts of SVMP proteins per litre cell culture as reported here in our manuscript.

(3) Economic viability and bioprospecting: Regarding the economic argument, we contend that viability in bioprospecting is defined by functional yield rather than simple volume. Producing large quantities of non-functional or misfolded protein in a cheaper system is economically inefficient. Furthermore, for snakebite research, the ability to produce specific, pure isoforms recombinantly without the contamination of other toxic venom components found in native isolations is essential for high-throughput screening and drug design.

(4) Scalability: Historically, insect cell production was seen as expensive, but current bioreactor technology and reduction in consumables and media costs allow for significant scaling. Many therapeutic reagents (vaccines, viral vectors, protein biologics) are produced routinely in baculovirus/insect cells. For the purposes of bioprospecting and lead identification, the yields provided by our Hi5/Sf21 system are sufficient for rigorous downstream bioassays and structural characterization.

Again, I believe the paper is highly important and excellently crafted, but I think especially the discussion should see some refinement to address the drawbacks and to evaluate the paper's findings with more modesty.

Thank you. We included the discussion about glycosylation patterns.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) It is not entirely clear to me if the final constructs are indeed "fusion-proteins" (line 172, 974), in the sense of chimeric proteins. From the current description, it appears that the prodomain is encoded in the same gene rather than fused as a separate domain. Thus, referring to these constructs as fusion proteins may overstate the degree of protein engineering involved in the study.

This is correct. In the revised manuscript, ‘fusion protein’ is only used in the context of the propeptide SVMP fusion construct to avoid confusion.

(2) Figure 2J: It is difficult to assess how much protein is secreted relative to the intracellular amounts. The blot is surely misleading, as the effective protein dilution differs substantially between intracellularly vs. extracellularly. Providing an estimate of the relative dilution of extracellular protein would help clarify the extent of secretion.

We estimate that the SNP and SN fractions are at least 10-times more concentrated than the media fraction. The blot is analytical and not quantitative.

(3) The manuscript appears to use both alphafold 2 and alphafold 3 for structural predictions. Clarification on the choice of the version and its impact on results would improve consistency.

In the revised version of the manuscript, we clarify that all structural models were generated using AlphaFold 3.

(4) Figure S3b and others: a clear description of the antibodies used in the Western blots would be appreciated (including in the methods).

We included this information in the figure legends and a paragraph in the methods section for Western blots in the revised manuscript.

(5) MTT cytotoxicity testing would be more convincing if done in a concentration-dependent manner.

We repeated this assay using different concentrations of SVMPs and show the results as a new Figure 5f in the revised manuscript.

(6) Figure S3c: It could be interesting to show the sequence coverage to get an impression of what part of the protein is there.

We have included this information as Supplementary Figure S4d in the revised manuscript.

Reviewer #2 (Recommendations for the authors):

Overall, the study is presented in a step-by-step manner, and its conclusions are valid.

(1) As suggested in the public review, further characterization of the purified material would be good, for example, by intact mass-spectroscopy to characterize the enzymes in further detail.

Preliminary MALDI-MS analysis (performed in Loic Quinton’s laboratory) of our PIII SVMP revealed a broad and heterogeneous mass distribution, consistent with heterogeneity caused by the presence of multiple glycoforms (which is not unlike the microheterogeneity in native snake venom). However, owing to the inherent limitations of MALDI-MS for the analysis of glycoproteins, our data do not allow determination of the number of occupied N-glycosylation sites or the identification of additional types of glycosylation.

Moreover, the relatively large molecular mass of these proteins (zymogen 70.2 kDa protein only, mature PIII 50.6 kDa protein only) makes analysis by electrospray ionisation mass spectrometry technically challenging.

An MS-based deep analysis of the glycosylation patterns would therefore be a project on its own, and beyond the scope of the present manuscript.

(2) The studies involving PII appear challenging due to low yields and stability of the enzyme and the mentioned self-degradation. Some studies, such as the casein-degradation, would benefit from working with a well-characterized batch of enzymes to ensure, it is not auto-degrading during the experiment.

We believe that the finding that the PII SVMP degrades itself after incubation with Zn²⁺ is an important observation. It is novel to the best of our knowledge. Moreover, the key message of our manuscript is that we can produce and characterise novel SVMPs that cannot be readily purified from venom (and thus are not well characterised).

Besides, there are very few intact PII SVMPs in venom (e.g. Suntravat et al. BMC Molecular Biol 2016); the vast majority cleaves itself into a PI and a disintegrin.

(3) Figure 4h. Degradation of insulin is only shown for recombinant PIII, not the native enzyme, and therefore doesn't convey any information with respect to how well they compare.

We do not have available any native PII and PIII SVMPs for a comparison with the recombinant SVMPs (in our manuscript we show expression of new, uncharacterised SVMPs). We have included the PIII SVMP in the original manuscript to show that the enzyme is active and has a different specificity compared to PI SVMP. In the revised manuscript, we also included the PII SVMP insulin B degradation assay in Supplementary Figure S11b.

(4) Figure 5a. Inconsistent use of enzymes - data for PII is presented (both as mature protein and Zymogen) and compared to PIII, but not PI, as both zymogen and mature protein. The current data presentation is confusing and gives the idea of the manuscript assembled with figures produced during the exploratory phase of the study, and not from subsequent experiments systematically conducted for the purposes of clarity and completeness.

In the revised manuscript, we included the missing enzymatic characterisations in Figure 5 (panel a and e) and Supplementary Figure S11a-c. These data were initially not included because the respective enzymes are inactive in these assays.

(5) The manuscript would benefit from editing to make it more concise. For an early-career reader, it is of interest and utility to follow the thought and experimental processes that led to the successful solution, but there is a risk of losing the reader's interest along the way by going through expression experiments that did not "work" in the typical sense of the word. To this reviewer, there is no added value in a full paragraph around co-expression with disulfide isomerase, as it did not improve the protein yield. A single sentence, "co-expression with PDI did not improve yields," with a reference to a supplemental figure would convey that message.

We have moved the optimisation of SVMP expression to the Supplementary Information, which we hope has improved the clarity and flow of the main text.

We note that the hypothesis that co-expression of protein disulfide isomerases (PDIs) enhances yields of functional SVMPs, given the high expression of PDIs in snake venom gland cells, is well established in the field. While we consider PDIs (and other chaperones) likely to play an important role in SVMP expression, we were unable to demonstrate this effect using the baculovirus-insect cell expression system and hypothesize that efficient insect and/or baculoviral PDIs are already present.

(6) Similarly with N-linked glycosylation, the section needs a headline (line 241) and firming up of a sentence like "and possibly not all of the glycosylation..." which is vague and appears to state that it was not really of interest to pursue this further. My view is that either an experiment is done properly with a stated aim and purpose, interpreted, and then, based on whether the results are of interest to the main story or not, they are included. If N-linked glycosylation is to be included in the manuscript, it should be with a purpose (e.g., N-linked glycosylation affects enzyme activity). As it stands, the message is "there is some N-linked glycosylation" without further explanation, and this generates information without justifying the inclusion hereof.

Please see our reply above regarding an in-depth characterisation of insect cell glycosylation of the recombinant PIII SVMP without access to the native enzyme for comparison. In our revised manuscript, we confirm that the PIII SVMP is glycosylated and that this at least partly accounts for the apparent discrepancy in molecular weight observed in SEC and SDS PAGE. We have modified the text to clarify the purpose of the PNGase deglycosylation experiment.

(7) The manuscript, in its current form, appears to have been copied from a Thesis with very detailed step-by-step logic and description. While this is useful in a scholarly context, a scientific manuscript should be presented more compactly, assuming the readers know basic biochemistry.

We trust that this Reviewer finds the revised version of our manuscript more compact and concise. 

Reviewer #3 (Recommendations for the authors):

(1) Material and Methods plus Figures:

Please report the number of replicates per experiment and how data is presented (means/ medians/ standard deviation/ others), and add error bars to the plots where needed.

In the revised manuscript we have included the number of repeats in the figure legends.

(2) Abstract

Line 4: I would not say that SVMPs are the most potent viper toxins. This place is probably taken by some of the highly neurotoxic PLA2, such as Crotoxin. Nevertheless, SVMPs are surely some of the most important toxins responsible for pathophysiological effects stemming from viper envenoming, but I would suggest rephrasing for accuracy.

In the revised manuscript, we have modified this sentence.

(3) Introduction

Lines 27-31: I would like to see a reference supporting the existence of all SVMP types across vipers.

We have included references supporting the existence of PI, PII and PIII SVMPs in viper venom. We also rewrote the sentence to state that “representatives of all three sub-classes are present in different viper venoms.” This clarifies that we do not say that all classes are present in all venoms.

Lines 59-60: I am not sure if this should be considered such an important impediment. Essentially, many vipers yield double- to triple-digit mg amounts of crude venom per specimen from only a single milking.

We have rewritten this text in the revised manuscript.

Currently, it is not possible to purify any given SVMP of interest from venom; in particular for E. ocellatus SVMP isoform mixtures are typically purified rather than individual enzymes (see also introduction section of our manuscript line 57ff). Also, many SVMPs are not present in sufficient amounts in the venom. Here, we provide an approach to recombinantly produce any SVMP of interest, independent of its abundance in the venom.

(4) Results

Line 102: The army-fallworms name is Spodoptera, not Spotoptera. Please correct the typo.

Done. Apologies for our oversight.

Line 311: Please provide the data at least as a supplement.

In the revised manuscript, we have included this experiment in Supplementary Figure S6c.

Line 432- 433: It would be useful to clarify whether the protein should have a pro-coagulant activity (or not).

We have changed this sentence as follows in the revised manuscript: This shows that our recombinantly produced SVMPs have no pro-coagulant activity, which was unknown before.

eLife Assessment

This study provides valuable data on the role of Hsd17b7, a gene involved in cholesterol biosynthesis, as a potential regulator of mechanosensory hair cell function. The authors used both zebrafish and the HEI cell line to examine the effects of deletion of Hsd17b7 on hair cell function and survival. While the study presents convincing evidence, the effect sizes observed across several experiments, including functional readouts such as the acoustic startle response, are modest, which raises questions about the biological significance of the proposed mechanism.

Reviewer #1 (Public review):

Summary:

This study identifies HSD17B7 as a cholesterol biosynthesis gene enriched in sensory hair cells, with demonstrated importance for auditory behavior and potential involvement in mechanotransduction. Using zebrafish knockdown and rescue experiments, the authors show that loss of hsd17b7 reduces cholesterol levels and impairs hearing behavior. They also report a heterozygous nonsense variant in a patient with hearing loss. The gene mutation has a complex and somewhat inconsistent phenotype, appearing to mislocalize, reduce mRNA and protein levels, and alter cholesterol distribution, supporting HSD17B7 as a potential deafness gene.

The study presents an interesting deafness candidate with a complex mechanism, and highlights an underexplored role for cholesterol (and lipids) in hair cell function.

The authors were very responsive to the initial reviews, and the manuscript is now significantly stronger.

Strengths:

- HSD17B7 is a new candidate deafness gene with plausible biological relevance.

- Cross-species RNAseq convincingly shows hair-cell enrichment.

- Lipid metabolism, particularly cholesterol homeostasis, is an emerging area of interest in auditory function.

- The connection between cholesterol levels and MET is potentially impactful and, if substantiated, would represent a significant advance.

- The localization of HSD17B7 is reasonably convincing, despite the lack of a KO control: In HEI-OC1 cells, HSD17B7 localizes to the ER, as expected. In mouse hair cells, the staining pattern is cytosolic. The developmental increase between P1 and P4, and the higher expression in OHCs aligns nicely with RNAseq data.

Weaknesses:

- The pathogenic mechanism of the E182STOP variant is unclear: The mutant protein presumably does not affect WT protein localization, arguing against a dominant-negative effect. Yet, overexpression of HSD17B7-E182* alone causes toxicity in zebrafish and it binds and mislocalizes cholesterol in HEI-OC1 cells, suggesting some gain-of-function or toxic effect. In addition, the mRNA of the variant has low expression level, suggesting nonsense-mediated decay. The mechanistic conclusions of the study are therefore not as clear cut as one would had hoped, but it might just be a reflection of real biological complexity.

- The link to human deafness is based on a single heterozygous patient with no syndromic features. Given that nearly all known cholesterol metabolism disorders are syndromic, this raises concerns about causality or specificity. HSD17B7 is therefore, at this point, a candidate deafness gene, and not a fully established "novel deafness gene". This is acknowledged by the authors.

- This study does not directly investigate how reduced cholesterol levels affect MET. However, this is not a significant limitation given the study's scope, and it is reasonable that such detailed functional analyses are left to specialists in hair cell physiology.

Reviewer #2 (Public review):

A summary of what the authors were trying to achieve.

The authors aim to determine whether the gene Hsb17b7 is essential for hair cell function and, if so, to elucidate the underlying mechanism, specifically the HSB17B7 metabolic role in cholesterol biogenesis. They use animal, tissue, or data from zebrafish, mouse, and human patients.

Strengths:

(1) This is the first study of Hsb17b7 in the zebrafish (a previous report identified this gene as a hair cell marker in the mouse utricle).

(2) The authors demonstrate that Hsb17b7 is expressed in hair cells of zebrafish and the mouse cochlea.

(3) In zebrafish larvae, a likely KO of the Hsb17b7 gene causes a mild phenotype in an acoustic/vibrational assay, which also involves a motor response.

(4) In zebrafish larvae, a likely KO of the Hsb17b7 gene causes a mild reduction in lateral line neuromast hair cell number and a mild decrease in the overall mechanotransduction activity of hair cells, assayed with a fluorescent dye entering the mechanotransduction channels.

(5) When HSB17B7 is overexpressed in a cell line, it goes to the ER, and an increase in Cholesterol cytoplasmic puncta is detected. Instead, when a truncated version of HSB17B7 is overexpressed, HSB17B7 forms aggregates that co-localize with cholesterol.

(6) It seems that the level of cholesterol in crista and neuromast hair cells decreases when Hsb17b7 is defective

Comments on the revised version:

Overall, the paper has been improved, but it still needs to be moderated regarding the observed effects and their qualification. I suggest expressing each effect as % {plus minus} SD and indicating it in the main text to inform the reader.

- The title " HSD17B7 is required for the function of sensory hair cells by regulating cholesterol Synthesis" should be moderated: "affects" instead of "required" would be better.

- In the abstract "conserved and essential role for HSD17B7-mediated cholesterol biosynthesis", the term essential seems overstated and premature

- In the discussion: "Collectively, these results suggest that the heterozygous c.544G>T (p.E182*) variant contributes to auditory dysfunction through potential pathogenic mechanisms: haploinsufficiency caused by reduced"...; "could contribute" would be safer.

- In the discussion: "In summary, our study identifies HSD17B7 as a critical regulator of cholesterol synthesis in sensory hair cells and as an essential factor in normal MET and sound-evoked sensory responses. "This part is still an overstatement. The effect in zebrafish is not directly shown to affect hearing, and startle reflex impairment is mild. It is not essential.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

(1) The pathogenic mechanism of the E182STOP variant is unclear. The mutant protein does not appear to affect WT protein localization, arguing against a dominant-negative effect. Yet, overexpression of HSD17B7-E182* alone causes toxicity in zebrafish and mislocalizes cholesterol in HEI-OC1 cells, suggesting a gain-of-function or toxic effect. In addition, the variant mRNA is expressed at a low level, consistent with nonsense-mediated decay. This apparent complexity and inconsistency need clearer explanation.

We appreciate the reviewer’s careful evaluation of this mechanistic complexity. Based on our combined molecular, cellular, and in vivo data, we propose that the pathogenic effect of the HSD17B7-E182* variant reflects a composite mechanism, rather than a classical dominant-negative effect.

At the transcript level, the E182* variant introduces a premature termination codon and shows markedly reduced mRNA abundance, consistent with partial degradation by nonsense-mediated mRNA decay. This reduction is expected to decrease overall HSD17B7 dosage, contributing a loss-of-function component. Unlike HSD17B7, the truncated HSD17B7^E182* mislocalizes cholesterol in HEI-OC1 cells, and overexpression alone reduces hair cell MET function and startle response in zebrafish embryos. We therefore propose that the truncated protein disturbing local cholesterol homeostasis, thereby exerts a toxic or ectopic gain-of-function.

We have revised the manuscript to clarify the dual-mechanism model.

(2) The link to human deafness is based on a single heterozygous patient with no syndromic features. Given that nearly all known cholesterol metabolism disorders are syndromic, this raises concerns about causality or specificity. The term "novel deafness gene" is premature without additional cases or segregation data.

We thank the reviewer for this important point. We fully agree that, based on a single heterozygous case without segregation data, it is premature to designate HSD17B7 as a novel deafness gene. Therefore, we have revised the manuscript to use the description of "candidate deafness genes".

(3) The localization of HSD17B7 should be clarified better: In HEI-OC1 cells, HSD17B7 localizes to the ER, as expected. In mouse hair cells, the staining pattern is cytosolic and almost perfectly overlaps with the hair cell marker used, Myo7a. This needs to be discussed. Without KO tissue, HSD17B7 antibody specificity remains uncertain.

We thank the reviewer for the constructive comments regarding HSD17B7 localization and antibody specificity.

Regarding subcellular localization, the original Figure 1K was intended to demonstrate the expression of HSD17B7 in mouse hair cells. To address this concern, we performed additional immunostaining on dissected organ of Corti sections at P1, P4, and P7 using higher magnification. Using parvalbumin as a hair cell marker, HSD17B7 displayed a partially punctate intracellular pattern in hair cells (revised Figure 1K). This pattern is consistent with localization to membrane-associated compartments, including the endoplasmic reticulum, and agrees with the ER-associated localization observed in HEI-OC1 cells and zebrafish hair cells. In mature hair cells, ER-associated signals may appear cytosolic and overlap with general hair cell markers such as Myo7a.

Regarding antibody specificity, although HSD17B7 knockout tissue was not available, we performed complementary validation experiments in HEI-OC1 cells. Cells were transfected with pCMV-Flag, pCMV-Flag-hHSD17B7WT, or pCMV-hHSD17B7WT-EGFP constructs and stained with anti-Flag, anti-EGFP, and anti-HSD17B7 antibodies. The HSD17B7 antibody signal showed strong co-localization with both FLAG- and EGFP-tagged HSD17B7 (revised Figure S1A and B), supporting its specificity.

Reviewer #2 (Public review):

(1) The statement that HSD17B7 is "highly" expressed in sensory hair cells in mice and zebrafish seems incorrect for zebrafish:

(a) The data do not support the notion that HSB17B7 is "highly expressed" in zebrafish. Compared to other genes (TMC1, TMIE, and others), the HSB17B7 level of expression in neuromast hair cells is low (Figure 1F), and by extension (Figure 1C), also in all hair cells. This interpretation is in line with the weak detection of an mRNA signal by ISH (Figure 1G I"). On this note, the staining reported in I" does not seem to label the cytoplasm of neuromast hair cells. An antisense probe control, along with a positive control (such as TMC1 or another), is necessary to interpret the ISH signal in the neuromast.

We thank the reviewer for this detailed evaluation and agree that the description of HSD17B7 expression in zebrafish hair cells requires clarification.

To address this, we performed a quantitative comparison of average expression levels within neuromast hair cells using log-normalized single-cell RNA-seq data. This analysis shows that hsd17b7 is expressed at a level comparable to several known MET-associated genes (e.g., tmc1 and lhfpl5a) (revised Figure 1D). Regarding the pseudotime heatmap (Figure 1F), we now state that this analysis illustrates temporal expression dynamics within neuromast hair cell development.

In addition, we have clarified the interpretation of the whole-mount in situ hybridization data by emphasizing that the signal indicates spatial enrichment rather than high transcript abundance.

We have updated the figure panels, legends, and corresponding text in the Results section to reflect these changes.

(b) However, this is correct for mouse cochlear hair cells, based on single-cell RNA-seq published databases and immunostaining performed in the study. However, the specificity of the anti-HSD17B7 antibody used in the study (in immunostaining and western blot) is not demonstrated. Additionally, it stains some supporting cells or nerve terminals. Was that expression expected?

To assess antibody specificity, we performed validation experiments using distinct epitopes. In HEI-OC1 cells transfected with pCMV-Flag-HSD17B7, or pCMV-HSD17B7-EGFP constructs, immunostaining with anti-HSD17B7 showed strong co-localization with both FLAG- and EGFP-tag (revised Figure S1B). In addition, western blot analyses using the same constructs confirmed the specific detection of HSD17B7 protein (revised Figure S1B). These validation data have now been included as supplementary figures in the revised manuscript and provide independent supporting evidence for the specificity of the anti-HSD17B7 antibody.

(2) A previous report showed that HSD17B7 is expressed in mouse vestibular hair cells by single-cell RNAseq and immunostaining in mice, but it is not cited: Spatiotemporal dynamics of inner ear sensory and non-sensory cells revealed by single-cell transcriptomics. Jan TA, Eltawil Y, Ling AH, Chen L, Ellwanger DC, Heller S, Cheng AG. Cell Rep. 2021 Jul 13;36(2):109358. doi: 10.1016/j.celrep.2021.109358.

We have now cited this reference in the revised manuscript.

(3) Overexpressed HSD17B7-EGFP C-terminal fusion in zebrafish hair cells shows a punctiform signal in the soma but apparently does not stain the hair bundles. One limitation is the consequence of the C-terminal EGFP fusion to HSD17B7 on its function, which is not discussed.

We thank the reviewer for raising this important technical point. The apparent absence of an HSD17B7-EGFP signal in hair bundles is primarily due to the imaging strategy and the selection of representative images. In zebrafish hair cells, the EGFP signal within hair bundles is extremely strong. To better visualize the intracellular distribution of HSD17B7 within the hair cell soma, we selected representative confocal optical sections that were focused on the cell body rather than on the apical hair bundle plane. As a result, the hair bundle signal is not visible in the images shown.

Importantly, we agree that C-terminal EGFP fusion may potentially influence protein localization or function. We have therefore revised the Discussion to discuss this limitation and to clarify that our central conclusions regarding HSD17B7 function are primarily supported by loss-of-function analyses, rescue experiments using untagged mRNA, and cholesterol perturbation phenotypes, rather than relying solely on EGFP-tagged overexpression constructs.

(4) A mutant Zebrafish CRISPR was generated, leading to a truncation after the first 96 aa out of the 340 aa total. It is unclear why the gene editing was not done closer to the ATG. This allele may conserve some function, which is not discussed.

Targeting regions close to the ATG is indeed a commonly used strategy for CRISPR-mediated gene disruption. In this study, sgRNA selection was guided by online CRISPR design tools (CRISPRscan), prioritizing predicted cutting efficiency and specificity. This strategy resulted in a frameshift mutation introducing a premature stop codon after amino acid 96 of the 340-aa Hsd17b7 protein.

Importantly, this truncation removes most of the conserved catalytic core required for 17β-hydroxysteroid dehydrogenase activity, including key motifs involved in NAD(P)-binding and substrate recognition. Therefore, although the mutation does not occur immediately adjacent to the ATG, the resulting allele is predicted to lack enzymatic function. We have clarified this rationale and discussed the functional consequences of the truncation in the revised manuscript.

(5) The hsd17b7 mutant allele has a slightly reduced number of genetically labeled hair cells (quantified as a 16% reduction, estimated at 1-2 HC of the 9 HC present per neuromast). On a note, it is unclear what criteria were used to select HC in the picture. Some Brn3C:mGFP positive cells are apparently not included in the quantifications (Figure 2F, Figure 5A).

Upon re-evaluation, we recognized that the original figure annotations were not sufficiently clear and may have led to confusion regarding hair cell selection. In the original images, the absence of dashed outlines around some Brn3c:mGFP⁺ cells may have been misinterpreted as their exclusion from analysis. To address this issue, we have revised Figures 2F and 5A by updating the annotations to ensure that all Brn3c:mGFP⁺ hair cells within each neuromast are clearly visible and unambiguously included (revised Figures 2F and 6A). Corresponding figure legends have also been revised to clarify the criteria used for hair cell identification and quantification.

(6) The authors used FM4-64 staining to evaluate the hair cell mechanotransduction activity indirectly. They found a 40% reduction in labeling intensity in the HCs of the lateral line neuromast. Because the reduction of hair cell number (16%) is inferior to the reduction of FM4-64 staining, the authors argue that it indicates that the defect is primarily affecting the mechanotransduction function rather than the number of HCs. This argument is insufficient. Indeed, a scenario could be that some HC cells died and have been eliminated, while others are also engaged in this path and no longer perform the MET function. The numbers would then match. If single-cell staining can be resolved, one could determine the FM4-64 intensity per cell. It would also be informative to evaluate the potential occurrence of cell death in this mutant. On another note, the current quantification of the FM4-64 fluorescence intensity and its normalization are not described in the methods. More importantly, an independent and more direct experimental assay is needed to confirm this point. For example, using a GCaMP6-T2A-RFP allele for Ca2+ imaging and signal normalization. 

We have revised the FM4-64 quantification strategy. Instead of measuring fluorescence intensity at the neuromast level, FM4-64 uptake was re-quantified at the single hair cell level. Hair cells within each neuromast were identified based on mGFP labeling, and the mean FM4-64 fluorescence intensity was measured for each individual hair cell. The average FM4-64 intensity per hair cell was then calculated for each neuromast and used for group comparisons (revised Figures 2F, 6B, and 8B, Figure S5B). The updated quantification method, normalization procedure, and analysis pipeline have now been described in the revised Methods section.

As supportive evidence, we further analyzed single-cell RNA-seq data from control and hsd17b7 mutant hair cells (revised Figure 3). This analysis revealed dysregulation of multiple genes involved in the MET machinery, including reduced expression of tip-link–associated components and altered expression of other MET-related genes. While these transcriptional changes do not constitute a direct functional assay, they are consistent with perturbation of MET-associated pathways and complement the FM4-64 findings.

(7) The authors used an acoustic startle response to elicit a behavioral response from the larvae and evaluate the "auditory response". They found a significative decrease in the response (movement trajectory, swimming velocity, distance) in the hsd17b7 mutant. The authors conclude that this gene is crucial for the "auditory function in zebrafish".

This is an overstatement:

(a) First, this test is adequate as a screening tool to identify animals that have lost completely the behavioral response to this acoustic and vibrational stimulation, which also involves a motor response. However, additional tests are required to confirm an auditory origin of the defect, such as Auditory Evoked Potential recordings, or for the vestibular function, the Vestibulo-Ocular Reflex. 

We thank the reviewer for highlighting the limitations in interpreting the acoustic startle assay. We have revised the manuscript to avoid overstatement and now describe the observed phenotype as a reduction in the behavioral response to acoustic and vibrational stimulation, rather than concluding a specific impairment of auditory function.

(b) Secondly, the behavioral defects observed in the mutant compared to the control are significantly different, but the differences are slight, contained within the Standard Deviation (20% for velocity, 25% for distance). To this point, the Figure 2 B and C plots are misleading because their y-axis do not start at 0.

We have corrected Figures 2B and 2C so that the y-axes start at zero, thereby providing a more transparent visualization of the behavioral differences. The figure legends have also been revised to clarify the presentation of the data.

(8) Overexpression of HSD17B7 in cell line HEI-OC1 apparently "significantly increases" the intensity of cholesterol-related signal using a genetically encoded fluorescent sensor (D4H-mCherry). However, the description of this quantification (per cell or per surface area) and the normalization of the fluorescent signal are not provided. 

The quantification of the D4H-mCherry signal in HEI-OC1 cells was performed at the single-cell level. Specifically, individual cells were segmented based on morphology, and the mean fluorescence intensity of D4H-mCherry per cell was measured. To account for variability in cell size and imaging conditions, fluorescence intensity was normalized to the background signal measured from cell-free regions in the same field of view. We have now clarified the quantification strategy and normalization procedure in the revised Methods and Results sections.

(9) When this experiment is conducted in vivo in zebrafish, a reduction in the "DH4 relative intensity" is detected (same issue with the absence of a detailed method description). However, as the difference is smaller than the standard deviation, this raises questions about the biological relevance of this result.

We have now clarified the quantification strategy and normalization procedure in the revised Methods and Results sections.

(10) The authors identified a deaf child as a carrier of a nonsense mutation in HSB17B7, which is predicted to terminate the HSB17B7 protein before the transmembrane domain. However, as no genetic linkage is possible, the causality is not demonstrated.

We thank the reviewer for raising this important point. Unfortunately, we were unable to obtain the parents' genetic testing data to perform formal genetic and linkage analysis. To address this limitation, we have revised the manuscript to avoid causal overstatement and now describe the HSD17B7 E182* variant as a candidate pathogenic variant associated with hearing loss. Importantly, our functional analyses in zebrafish and cell-based systems demonstrate that the E182* truncation abolishes key biological activities of HSD17B7, including subcellular localization, cholesterol regulation, mechanotransduction-related activity, and behavioral responses. These convergent functional data provide biological support for the potential pathogenic relevance of this variant.

(11) Previous results obtained from mouse HSD17B7-KO (citation below) are not described in sufficient detail. This is critical because, in this paper, the mouse loss-of-function of HSD17B7 is embryonically lethal, whereas no apparent phenotype was reported in heterozygotes, which are viable and fertile. Therefore, it seems unlikely that heterozygous mice exhibit hearing loss or vestibular defects; however, it would be essential to verify this to support the notion that the truncated allele found in one patient is causal.

Hydroxysteroid (17beta) dehydrogenase 7 activity is essential for fetal de novo cholesterol synthesis and for neuroectodermal survival and cardiovascular differentiation in early mouse embryos.

Jokela H, Rantakari P, Lamminen T, Strauss L, Ola R, Mutka AL, Gylling H, Miettinen T,

Pakarinen P, Sainio K, Poutanen M. Endocrinology. 2010 Apr;151(4):1884-92. doi: 10.1210/en.2009-0928. Epub 2010 Feb 25.

We thank the reviewer for raising this important point. We acknowledge that previous work has shown that complete loss of Hsd17b7 in mice is embryonically lethal, whereas heterozygous animals are viable and fertile (Jokela et al., 2010). Notably, this study primarily focused on embryonic development, cholesterol metabolism, and cardiovascular and neuroectodermal survival, and auditory or vestibular functions were not specifically examined. Therefore, subtle or sensory organ–specific phenotypes in heterozygous mice cannot be excluded.

The human variant identified in this study (E182*) is a nonsense mutation predicted to truncate the HSD17B7 protein prior to the transmembrane and cytoplasmic domains. We therefore present it as a candidate loss-of-function variant, providing supportive human genetic evidence that is consistent with our functional analyses in zebrafish hair cells, rather than as definitive proof of causality. We have revised the manuscript to clarify these points and to acknowledge this limitation.

(12) The authors used this truncated protein in their startle response and FM4-64 assays. First, they show that contrary to the WT version, this truncated form cannot rescue their phenotypes when overexpressed. Secondly, they tested whether this truncated protein could recapitulate the startle reflex and FM4-64 phenotypes of the mutant allele. At the homozygous level (not mentioned by the way), it can apparently do so to a lesser degree than the previous mutant. Again, the differences are within the Standard Deviation of the averages. The authors conclude that this mutation found in humans has a "negative effect" on hearing, which is again not supported by the data. 

We thank the reviewer for this important comment. We agree that the overexpression strategy employed in this study does not fully replicate the endogenous heterozygous state observed in patients, and that the magnitude of the observed effects varies across samples. Accordingly, our experiments were not intended to demonstrate a definitive causal role of the HSD17B7 ^E182* variant in hearing loss.

Instead, the overexpression assays were designed to assess whether the truncated HSD17B7 protein displays abnormal cellular properties and whether its presence can interfere with processes relevant to hair cell function. Under these conditions, HSD17B7^E182* exhibited aberrant subcellular localization, altered intracellular cholesterol distribution, and was associated with reduced FM4-64 uptake and changes in startle-associated behaviors, whereas the wild-type protein did not.

We revised the manuscript to moderate our conclusions. Rather than claim that the E182* mutation has a definitive “negative effect on auditory function,” we now describe it as a functionally compromised allele that disrupts cholesterol distribution and MET-related activity under overexpression conditions, providing mechanistic support consistent with our zebrafish loss-of-function data and the identification of this variant in a patient with hearing loss. In addition, the "negative effect" statement was based on the result that overexpression of the E182* mutation in wild-type embryos caused the compromised MET function and startle response defect.

(13) The authors looked at the distribution of the HSB17B7 in a cell line. The WT version goes to the ER, while the truncated one forms aggregates. An interesting experiment consisted of co-expressing both constructs (Figure S6) to see whether the truncated version would mislocalize the WT version, which could be a mechanism for a dominant phenotype. However, this is not the case.

We thank the reviewer for raising this important point regarding a potential dominant-negative mechanism. Consistent with the reviewer’s interpretation, we found that HSD17B7^WT predominantly localizes to the endoplasmic reticulum, whereas the truncated HSD17B7^E182* protein forms intracellular aggregates. Importantly, we further observed that the E182* mutation markedly reduces the stability of both HSD17B7 mRNA and protein, resulting in substantially decreased abundance of the truncated protein (Figure S6B–E). As a consequence, the cellular levels of HSD17B7^E182* are abnormally low.

Based on these findings, we consider it unlikely that the E182* variant exerts its effect through interference with the wild-type protein. Our results suggest that the heterozygous c.544G>T (p.E182*) variant contributes to auditory dysfunction through potential pathogenic mechanisms: 1, haploinsufficiency caused by reduced HSD17B7 expression, 2, functional impairment due to altered protein subcellular localization and cholesterol distribution.

We have revised the Results and Discussion sections. Our conclusions now emphasize that the functional impact of this variant is attributable to decreased effective HSD17B7 dosage, consistent with the observed defects in cholesterol synthesis, MET-related activity, and auditory-associated phenotypes in our model.

(14) Through mass spectrometry of HSB17B7 proteins in the cell line, they identified a protein involved in ER retention, RER1. By biochemistry and in a cell line, they show that truncated HSB17B7 prevents the interaction with RER1, which would explain the subcellular localization.

Consistent with the reviewer’s interpretation, wild-type HSD17B7 interacts with RER1, a protein known to participate in ER retention, whereas this interaction is lost in the truncated HSD17B7 variant. We propose that RER1 is an interacting partner of HSD17B7, providing a mechanistic explanation for the protein's subcellular localization.

(15) Information and specificity validation of the HSB17B7 antibody are not presented. It seems that it is the same used on mice by IF and on zebrafish by Western. If so, the antibody could be used on zebrafish by IF to localize the endogenous protein (not overexpression as done here). Secondly, the specificity of the antibody should be verified on the mutant allele. That would bring confidence that the staining on the mouse is likely specific.

We thank the reviewer for raising this important point regarding antibody specificity and validation. Information on the HSD17B7 antibody and its validation has been provided in our response to comment 1, where we described the use of antibodies recognizing different epitopes and the experimental strategies employed to assess specificity (revised Figure S1A and B).

Although the same antibody was used for Western blot analysis in zebrafish samples, its performance in immunofluorescence staining of zebrafish tissues was suboptimal, with relatively high background. For this reason, we did not rely on this antibody for endogenous Hsd17b7 localization in zebrafish by immunofluorescence and instead employed tagged constructs for subcellular localization analyses. This approach provides more reliable and interpretable localization information under the current experimental conditions.

Recommendations for the authors:

Reviewing Editor Comments:

Suggested revisions to help improve the study and the eLife Assessment:

(1) FM4-64 uptake: Isolate the effect of hair cell loss and MET reduction.

(2) Clarify the mechanistic model: Is the mutant protein pathogenic due to toxicity, lack of expression or function, or both? Come up with a clearer causal chain of events.

(3) Mouse immunostaining: Validate the HSD17B7 antibody, and since mouse RNAseq data (gEAR database) suggest that HSD17B7 expression increases dramatically between P0-P5, show this developmental progression by immunostaining of the mouse organ of Corti at P0, P3, and P5.

(4) The HSD17B7-E182* expression disrupts cholesterol (D4H staining) in OC1 cells. This should also be demonstrated in the mutant zebrafish.

(5) Structural modeling of E182* is uninformative; half the protein is absent. This kind of analysis is better suited for missense variants. Suggest removing this analysis.

We thank the Reviewing Editor for these constructive suggestions. The major points raised here substantially overlap with the concerns raised in the public reviews. In response, we have:

(1) revised FM4-64 quantification and interpretation to better distinguish hair cell loss from MET impairment;

(2) Clarify the mechanistic mode. Mechanistically, the mutation decreases mRNA abundance and significantly reduces protein levels. Moreover, expression of the p.E182* mutation disrupted the interaction between HSD17B7 and the ER retention receptor RER1, leading to aberrant subcellular localization and altered cholesterol distribution, thereby exacerbating HC dysfunction.

(3) provided additional validation of the HSD17B7 antibody using antibodies targeting distinct epitopes, and extended mouse organ of Corti immunostaining to postnatal stages P1, P4, and P7 to demonstrate the developmental upregulation of HSD17B7 expression;

(4) added in vivo zebrafish experiments demonstrating that expression of HSD17B7^E182* disrupts cholesterol distribution in hair cells, consistent with the effects observed in HEI-OC1 cells using D4H staining;

(5) removed the structural modeling of the E182* variant.

Recommendations for the authors:

The recommendations from Reviewer #1 and Reviewer #2 were carefully considered and addressed. Most of these points overlap with the public reviews and the Reviewing Editor's comments and have been addressed through a revised mechanistic interpretation, additional clarifications in the Methods, more moderate claims regarding auditory function and human genetics, and the removal or revision of potentially misleading analyses. In addition, a number of minor issues were corrected, including missing or incorrect references, repetitive or unclear statements in the Introduction, insufficient methodological details, imprecise terminology, and typographical or formatting errors. Collectively, these revisions improve the clarity, rigor, and transparency of the study without altering its central conclusions.

eLife Assessment

This important study describes computationally designed proteins that bind to the chemokine CCL25. The authors present evidence that some binders simply prevent chemokine binding to the CCR9 receptor, while one binder changes the downstream signaling triggered by chemokine binding. The evidence is solid overall, but some uncertainty remains with respect to functional selectivity due to sensitivity differences between functional assays and the degree of binder selectivity between the large family of chemokine ligands.

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors describe the use of BindCraft computational protein design to create a series of binders to the chemokine CCL25. This chemokine normally mediates CCR9-dependent trafficking of immune cells to the gut, making it a potential target for the treatment of inflammatory bowel disease and related conditions. Importantly, CCL25 also binds a scavenging receptor, ACKR4. The computational protein design approach used does not involve defining particular epitopes to be targeted, allowing a free search for any potential interaction surface.

Among four designs tested, three were predicted to interact at a similar site on the chemokine, while a fourth clone, VUP25111, was predicted to bind to a different site. All four designs showed binding to CCL25, with similar high-nM KD values in all cases. The first three clones showed evidence of direct competition with the receptor for CCL25 binding, while VUP25111 showed incomplete inhibition of binding. In functional assays, all clones acted as antagonists except for VUP25111, which inhibited arrestin recruitment by CCR9, but did not affect G protein activation by CCR9 or arrestin recruitment by ACKR4 (which signals exclusively through arrestin and not G protein).

Strengths:

The work is completed to a high technical standard, and the functional diversity of the clones is intriguing. It is exciting to see pathway-selective modulation of signaling, and this basic paradigm is likely to generalize to other chemokine/receptor systems. The exceptional complexity of chemokine signaling makes this an excellent area to explore the development of modulators that can restrict signaling to a specific subset of receptors.

Weaknesses:

No major weaknesses were noted by this reviewer.

Reviewer #2 (Public review):

This study from de Boer, Lamme, Verdwaald and Schafer describes the de novo AI-guided design of miniproteins that target the chemokine CCL25, with the aim to modulate the activation and signalling of the chemokine receptors CCR9 and ACKR4. The study focuses on characterising four miniproteins that all bind CCL25 with good affinity. Three designs appear to prevent CCL25 binding to both CCR9 and ACKR4, with increasing concentrations of miniproteins resulting in decreased arrestin (both receptors) and mini G protein recruitment (CCR9), less inhibition of forskolin-stimulated cAMP (CCR9), and decreased GRK3 recruitment and receptor internalisation (CCR9). One miniprotein, VUP25111, changes the properties of CCL25 rather than preventing ligand/receptor interactions, resulting in greater selectivity for CCR9 over ACKR4 and a G protein-biased signalling profile (maintenance of mini G protein recruitment, GRK3 recruitment, inhibition of cAMP and receptor internalisation, but loss of arrestin recruitment). VUP25111 also maintained chemotactic migration in MOLT-4 T lymphoblast cells, whereas this response was lost in the presence of the other three miniproteins.

Overall, this is a very interesting application of AI-designed de novo miniproteins to modulate GPCR responses by directly binding the ligand rather than the receptor. This is a conceptually very intriguing approach that could, in principle, be extended to other GPCR systems beyond the chemokine family. The authors deploy an impressive array of assays spanning multiple signalling endpoints, providing a thorough picture of how each miniprotein influences receptor activation and downstream signalling. The presentation of concentration-response relationships for CCL25 alone and in the presence of each miniprotein is particularly informative, and the figures are very well constructed throughout. The inclusion of clear cartoons illustrating the basis of each assay is a nice touch that will help readers from outside the immediate field follow the logic of each experiment.

There are two main conclusions that are not currently as well-supported by the evidence as they might be, and that would benefit from some qualification. The first concerns the selectivity of the miniproteins for CCL25. Testing the impact of the miniproteins on CXCL12 activation of CXCR4 is an important and welcome experiment, but it addresses selectivity against only one other chemokine system, and the current claim of specificity is therefore stronger than the data allow. Additionally, at the highest concentration tested (10 µM), the more potent miniproteins (VUP25101, VUP25107) appear to show some inhibition of arrestin recruitment to CXCR4 - perhaps unsurprising given the degree of structural conservation among chemokines. The statements regarding selectivity and the lack of effect on the CXCL12/CXCR4 system would benefit from revision to more accurately reflect these observations.

The second concern relates to the interpretation of the preserved GRK3 recruitment, but the complete loss of arrestin recruitment observed with VUP25111. In the GRK3 recruitment experiments, 20 nM CCL25 was used, representing an EC40 concentration in this assay. VUP25111 causes a concentration-dependent reduction in CCL25-induced GRK3 recruitment, down to approximately 15% of the maximal response. It is worth considering whether this degree of reduction in GRK3 recruitment could itself be sufficient to disrupt patterns of receptor phosphorylation and thereby prevent observable arrestin recruitment. Both interpretations are complicated by the fact that the GRK3 recruitment and arrestin recruitment assays likely differ in their sensitivity and dynamic windows, making direct quantitative comparisons between them difficult. In the absence of direct measurements of CCR9 phosphorylation in the presence of VUP25111, the alternative interpretation remains open and would benefit from acknowledgement. Given recent work from the same group demonstrating that receptor internalisation is only partially dependent on arrestins (Lamme et al., 2025, J Biol Chem), further discussion of the relationship between GRK and arrestin recruitment and CCR9 internalisation would be of value to the broader GPCR audience this work is likely to attract.

Finally, some additional justification for the use of 20 nM CCL25 across all assays would strengthen the study, as this concentration represents different points on the concentration-response curve depending on the assay and receptor in question. It ranges from an EC40 for CCR9 GRK3 recruitment and internalisation, to an EC50 for CCR9 arrestin and mini-Gi recruitment, an EC80 for CCR9 cAMP inhibition, and an EMax for ACKR4 arrestin recruitment. This has potential consequences for the interpretation and cross-assay comparison of miniprotein potency, and the authors are encouraged to acknowledge and discuss this in the context of their conclusions.

Reviewer #3 (Public review):

Summary:

The authors employed the BindCraft platform to develop binders targeting the chemokine CCL25, a selective activator of the chemokine receptor CCR9. They successfully generated two classes of binders: one that inhibits all CCL25-mediated CCR9 activation, and another that permits CCR9 G protein signaling while simultaneously preventing arrestin recruitment. These tools will enable the dissection of arrestin involvement in regulating cell migration.

My comments, in the order of reading:

(1) Title: I strongly recommend removing the term "biasing" from the title. In this context, it does not convey a specific mechanistic concept. The term "biased signaling" has been used for a very broad range of mechanistically distinct pharmacological phenomena, and without a precise definition, it adds more confusion than clarity. I therefore suggest refraining from using it in the title.

(2) Abstract, line 34: The term "bias" should be replaced. As currently used, it appears to suggest a dichotomy between G protein signaling and arrestin recruitment. However, arrestin recruitment is a consequence of G protein signaling, and it is not conceptually sound to compare a signaling event mediated by one protein family with a protein-protein interaction involving another protein family. A meaningful comparison requires experimental paradigms that differ by a single variable; in this case, there are two - distinct protein families and fundamentally different types of readouts (signaling versus protein-protein interaction).

(3) Abstract, line 34: The term "balanced agonist" should be removed. Any chosen reference ligand is, by definition, the "balanced" agonist for that analysis, regardless of its actual signaling profile. Consequently, the expression "balanced agonist" adds no mechanistic information beyond "the agonist used as reference in a particular bias calculation" and is potentially misleading, as it implies that this ligand possesses a uniquely unbiased, system‑independent signaling profile, which is not the case.

(4) Abstract, line 36: I also recommend removing the term "bias" at this point. The concept of bias typically arises from experiments that quantitatively compare more than one variable. As currently written, the phrasing suggests a dichotomy between G protein- and arrestin-mediated signaling, yet the study does not assess arrestin signaling, only arrestin recruitment. Under these conditions, the use of "bias" is not appropriate. The data are clear and compelling on their own without the need for this potentially misleading terminology.

(5) Introduction: This is interesting to read and generally well written, though certain statements would benefit from improved semantic precision. For example, in lines 110-111, the phrase "G protein-biased complex" should be reconsidered, as it relies on the notion of G protein- versus arrestin-mediated signaling. Arrestins themselves do not signal; what is measured here is their recruitment. Comparing G protein signaling with arrestin recruitment is therefore conceptually unsound, since arrestin engagement is a downstream consequence of G protein activation. Comparisons become meaningful only when designed to differentiate between G protein-dependent and G protein-independent arrestin recruitment, which is not the case in this study.

(6) Results, 122,123: The authors should consider being more precise; possibly, the truncated CCL25 is somewhat less potent on CCR9. The authors should make a statistical test and then decide whether to rephrase or not for enhanced precision.

(7) Figure S5: This figure is currently confusing and needs clarification. The authors state in the main text that CXCR4 is stimulated with CXCL12, yet the figure legend refers to CCL25; this discrepancy should be corrected to ensure consistency. In addition, inhibition of CXCR4 by the miniprotein binders should be analyzed and presented with normalization to CXCR4 responses, not to CCL25-stimulated CCR9. To avoid misinterpretation, inhibition by the miniproteins should be quantified separately for CCR9 and CXCR4, each normalized to its own receptor-specific and functionally equivalent stimulation condition, rather than to the "other" receptor.

(8) Results, lines 211-213: The authors should be more semantically precise. They state that no binder has any effect on arrestin recruitment to CXCR4. If I see the data, this is not really true, as 25101 and 25107 inhibit arrestin recruitment by about 50 % or more at the highest applied concentrations; only 111 and 112 are completely inactive. As already commented, normalization should be done to arrestin recruitment of CXCR4 and not CCR9.

Author response:

We thank the editors and reviewers for thoroughly reviewing our manuscript and offering thoughtful and constructive feedback. We appreciate the positive reception of our work and welcome the opportunity to address the lingering concerns. In the coming revisions, we will be directly addressing the question of the miniprotein’s specificity and increase the precision in the language used to discuss our findings.

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence would be stronger if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

Strengths:

Novelty and Scope:<br /> The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience.<br /> It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

Methodological Rigor:<br /> The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance.<br /> Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

Biological Relevance:<br /> The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses.<br /> The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

Clarity and Depth:<br /> The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

Weaknesses and Areas for Improvement:

Generality and Validation:<br /> The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings.<br /> Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

Role of Active Dendritic Currents:<br /> The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

Analysis of Plasticity:<br /> While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

Frequency-Dependent Effects:<br /> The study demonstrates that gap junctional inputs suppress high-frequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

Visualization:<br /> Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

Contextual Relevance:<br /> The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

Suggestions for Improvement:

Broader Application:<br /> Simulate EFPs in multi-neuron networks to assess how the findings extend to network-level interactions, particularly in regions with mixed synaptic connectivity.

Experimental Correlation:<br /> Collaborate with experimental groups to validate the computational predictions using in vivo or in vitro recordings.

Mechanistic Insights:<br /> Provide a more detailed mechanistic explanation of how specific ionic currents (e.g., HCN, sodium, leak) interact during gap junctional vs. chemical synaptic inputs.

Implications for Neural Coding:<br /> Discuss how the observed differences in EFP signatures might influence neural coding, especially in circuits with heavy gap junctional connectivity.

Reviewer #2 (Public review):

Summary:

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

Strengths:

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

Weaknesses:

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

Author response:

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence can be more convincing if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

We gratefully thank the editors and the reviewers for the time and effort in rigorously assessing our manuscript, for the constructive review process, for their enthusiastic responses to our study, and for the encouraging and thoughtful comments. We especially thank you for deeming our study to be a valuable exploration on the differential contributions of active dendritic gap junctions vs. chemical synapses to local field potentials. We thank you for your appreciation of the quantitative biophysical demonstration on the differences in electric dipoles that appear in extracellular potentials with gap junctions vs. chemical synapses.

However, we are surprised by aspects of the assessment that resulted in deeming the approach incomplete, especially given the following with specific reference to the points raised:

(1) Testing against experiments: With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established nonspecificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021), reproduced below. In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

In addition, the complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

Together, we emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials.

(2) Model validation: The model used in this study was adopted from a physiologically validated model from our laboratory (Roy & Narayanan, 2021). Please note that the original model was validated against several physiological measurements along the somatodendritic axis. We sincerely regret our oversight in not mentioning clearly that we have used an existing, thoroughly physiologically-validated model from our laboratory in this study.

(3) Comparisons between chemical vs. junctional inputs: We had taken elaborate precautions in our experimental design to match the intracellular electrophysiological signatures with reference to synchronous as well as oscillatory inputs, irrespective of whether inputs arrived through gap junctions or chemical synapses.

In a revised manuscript, we will address all the concerns raised by the reviewers in detail. We have provided point-by-point responses to reviewers’ helpful and constructive comments below. We thank the editors and the reviewers for this constructive review process, which we believe will help us in improving our manuscript with specific reference to emphasizing the novelty of our approach and conclusions.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

Novelty and Scope

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience. It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

We thank you for the positive comments on the novelty of our approach and how our study addresses an underexplored area in neuroscience. The assumptions about the passive nature of dendritic structures had indeed resulted in an underestimation of the contributions of gap junctions to extracellular potentials. Once the realities of active structures are accounted for, the contributions of gap junctions increases by several orders of magnitude compared to passive structures (Fig. 1D).

Methodological Rigor

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance. Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

We thank you for your encouraging comments on the experimental design and methodological rigor of our approach.

Biological Relevance

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses. The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

We thank you for your positive comments on the biological relevance of our approach. We also gratefully thank you for emphasizing the two striking novelties unveiling the dichotomy between gap junctions and chemical synapses in their contributions to field potentials: polarity differences and spectral characteristics.

Clarity and Depth

The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

We sincerely thank you for the positive comments on the structure and comprehensive coverage of our manuscript encompassing different types of inputs that neurons typically receive.

Weaknesses and Areas for Improvement

Generality and Validation

The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings. Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

We thank you for raising this important point. The prime novelty and the principal conclusion of this study is that gap junctional contributions to extracellular field potentials are orders of magnitude higher when the active nature of cellular compartments are accounted for. The lacuna in the literature has been consequent to the assumption that cellular compartments are passive, resulting in the dogma that gap junctional contributions to field potentials are negligible. Despite knowledge about active dendritic structures for decades now, this assumption has kept studies from understanding or even exploring the contributions of gap junctions to field potentials. The rationale behind the choice of a computational approach to address the lacuna were as follows:

(1) The complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

(2) With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non-specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

We highlight the novelty of our approach and of the conclusions about differences in extracellular signatures associated with active-dendritic chemical synapses and gap junctions, against these experimental difficulties. We emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials. Our analyses clearly demonstrates that gap junctions do contribute to extracellular potentials if the active nature of the cellular compartments is explicitly accounted for (Fig. 1D). We also show theoretically well-grounded and mechanistically elucidated differences in polarity (Figs. 1–3) as well as in spectral signatures (Figs. 5–8) of extracellular potentials associated with gap junctional vs. chemical synaptic inputs. Together, our fundamental demonstration in this study is the critical need to account for the active nature of cellular compartments in studying gap junctional contributions of extracellular potentials, with CA1 pyramidal neuronal dendrites used as an exemplar.

In a revised version of the manuscript, we will emphasize the motivations for the approach we took, highlighting the specific novelties both in methodological and conceptual aspects, finally emphasizing the need to account for other cell types and gap junctional contributions therein. Importantly, we will emphasize the non-specificities associated with gap-junctional blockers as the reason why experimental delineation of gap junctional vs. chemical synaptic contributions to LFP becomes tedious. We hope that these points will underscore the need for the computational approach that we took to address this important question, apart from the novelties of the manuscript.

Role of Active Dendritic Currents

The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

We thank you for this constructive suggestion. We agree that it is important to consider the implications for broader network dynamics of the outward HCN currents that are observed with synchronous inputs. In a revised manuscript, we will elaborate on the implications of the outward HCN current to network dynamics in detail.

Analysis of Plasticity

While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

We thank you for this constructive suggestion. Please note that we have presented consistent results for both fewer and more gap junctions in our analyses (Figure 1 with 217 gap junctions and Supplementary Figure 1 with 99 gap junctions). Thus, our fundamentally novel result that gap junctions onto active dendrites differentially shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron. Thus, these results demonstrate that the conclusions about their contributions to LFP are invariant to plasticity in their gap junctional numerosity.

We had only briefly mentioned plasticity in the Introduction to highlight the different modes of synaptic transmission and to emphasize that plasticity has been studied in both chemical synapses and gap junctions, playing a role in learning and adaptation. However, if this wording inadvertently suggests that our study includes plasticity simulations, we would remove it from Introduction in the updated manuscript to ensure clarity.

In the ‘Limitations of analyses and future studies’ section in Discussion, we suggested investigating the impact of plasticity mechanisms—specifically, activity-dependent plasticity of ion channels—on synaptic receptors vs. gap junctions and their effects on extracellular field potentials under various input conditions and plasticity combinations across different structures. We fully agree with the reviewer that such studies would offer valuable insights and further enhance the broader relevance of our findings. However, while our study implies this direction, it was not the primary focus of our investigation.

In the revised manuscript, we will expand on intrinsic/synaptic plasticity and how they could contribute to LFPs (Sinha & Narayanan, 2015, 2022), while also pointing to simulations with different numbers of gap junction in this context.

Frequency-Dependent Effects

The study demonstrates that gap junctional inputs suppress highfrequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

We sincerely thank you for these insightful comments that we totally agree with. As it so happens, this manuscript forms the first part of a broader study where we explore the implications of gap junctions to ripple frequency oscillations. The ripple oscillations part of the work was presented as a poster in the Society for Neuroscience (SfN) annual meeting 2024 (Sirmaur & Narayanan, 2024). There, we simulate a neuropil made of hundreds of morphologically realistic neurons to assess the role of different synaptic inputs — excitatory, inhibitory, and gap junctional — and active dendrites to ripple frequency oscillations. We demonstrate there that the conclusions from single-neuron simulations in this current manuscript extend to a neuropil with several neurons, each receiving excitatory, inhibitory and gap-junctional inputs, especially with reference to high-frequency oscillations. Our networkbased analyses unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions in conjunction with return-current contributions from active dendrites playing regulatory roles in determining ripple characteristics (Sirmaur & Narayanan, 2024).

Our principal goal in this study, therefore, was to lay the single-neuron foundation for network analyses of the impact of gap junctions on LFPs. We are preparing the network part of the study, with a strong focus on ripple-frequency oscillations, for submission for peer review separately.

In a revised manuscript, we will mention the results from our SfN abstract with reference to network simulations and high-frequency oscillations, while also presenting discussions from other studies on the role of gap junctions in synchrony and LFP oscillations.

Visualization

Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

We thank you for this constructive suggestion. In the revised manuscript, we will enhance the visualization of the figures to ensure a clearer and more intuitive distinction between chemical synapses and gap junctions.

Contextual Relevance

The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

We sincerely appreciate your valuable suggestion and acknowledge the importance of integrating our results into established neural network dynamics, particularly their implications for gamma rhythms. We will address this aspect more comprehensively in the revised version of our manuscript.

Reviewer #2 (Public review):

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

We gratefully thank you for the positive comments and the encouraging words about the novel contributions of our study. We are particularly thankful to you for your comment on the generality of our conclusions that hold for different cell types and neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

We sincerely thank you for the positive comments about the readability of the paper.

Weaknesses

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

We thank you for raising this important point. We would like to emphasize that our experimental design and analyses quantitatively account for the spatial distribution and temporal pattern of specific kinds of inputs that arrive through gap junctions and chemical synapses. We submit that our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. We elucidate these points below:

(1) Spatial distribution: The inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were of the same nature: excitatory.

(2) Different numbers of inputs: We have presented consistent results for both fewer and more gap junctions or chemical synapses in our analyses (see Figure 1 with 217 gap junctions or 245 chemical synapses and Supplementary Figure 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

(3) Synchronous inputs (Figs. 1–3): For chemical synapses, the waveforms are in the shape of postsynaptic potentials. For gap junctional inputs, the waveforms are in the shape of postsynaptic potentials or dendritic spikes (to respect the active nature of inputs from the other cell). Here, the electrical response of the postsynaptic cell is identical irrespective of whether inputs arrive through gap junctions or chemical synapses: an action potential. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic and gap junctional inputs. We mechanistically analyze the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Figs. 2–3). In a revised manuscript, we will show the intracellular responses to demonstrate that they are electrically matched.

(4) Random inputs (Fig. 4): For random inputs, we did not account for the number of action potentials that arrived, as the only observation we made here was with reference to the biphasic nature of the extracellular potentials with gap junctional inputs in the “No Sodium” scenario. We note that in the “No Sodium” scenario, the time-domain amplitudes were comparable for the field potentials (Fig. 4B, Fig. 4D).

(5) Rhythmic inputs (Fig. 5–8): For rhythmic inputs, please note that the intracellular and extracellular waveforms for every frequency are provided in supplementary figures S5– S11. It may be noted that the intracellular responses are comparable. In simulations for assessing spike-LFP comparison, we tuned the strengths to produce a single spike per cycle, ensuring fair comparison of LFPs with gap junctions vs. chemical synapses.

Taken together, we demonstrate through explicit sets of simulations and analyses that the differences in LFPs were not driven by the strength or patterns of the inputs but rather by the differences in direct transmembrane currents, which are subsequently reflected in the LFPs. In a revised manuscript, we will add a section to emphasize these points apart from providing intracellular traces for cases where they are not provided.

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

We thank you for raising this important question. Leak channels were among the several contributors to the positive deflection observed in LFPs associated with gap junctions. This effect was present not only in gap junctional models with intact sodium conductance but also in the no-sodium model, where the amplitude of the positive deflection was reduced across other models as well (Fig. 2F, I). Furthermore, even in the absence of leak conductance, a small positive deflection was still observed (Fig. 2F), leading us to further investigate other transmembrane currents over time and across spatial locations, from the proximal to the distal dendritic ends relative to the soma (Fig. 3D). We had observed that the dominant contributor in the case of chemical synapses was the inward synaptic current (Fig. 3A), whereas for gap junctions, the primary contributors were leak conductance along with other outward currents, such as potassium and HCN currents (Fig. 3D). Together, the direct transmembrane component of chemical synapses provides a dominant contribution to extracellular potentials. This dominance translates to differences in the relative contributions of indirect currents (including leak currents) to extracellular potentials associated chemical synaptic vs. gap junctional inputs. Our analyses of the exact ionic mechanisms (Fig. 3) demonstrates the involvement of several ion channels contributing to the indirect component in either scenario.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

We thank you for raising this important point. We agree with the analyses presented by the reviewer on the importance of network simulations and bidirectional gap junctions that respect the voltages in both neurons. However, the complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses “post-synaptic” currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a pointneuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; Martinez-Canada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications becomes essential.

Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations. Thus, the inputs were not pre-determined by “pre” neurons. Instead, the recorded voltages from potential synaptic partner neurons were randomized across locations and scaled using factors at the dendrites before being injected into the target neuron (Supplementary Fig. S1). While incorporating a network of interconnected neurons is indeed important, we utilized biophysical, morphologically realistic CA1 neuron model with different sets of input patterns to model LFPs, which were derived from the total transmembrane currents across all compartments of the multi-compartmental neuron model. Given the complexity of this approach, adding further network-level interactions or pre-post connections would have been computationally demanding.

In a revised manuscript, we will introduce the general methodology used in LFP modeling studies to introduce synaptic currents. We will emphasize that our study extends this approach to modeling gap junctional inputs, while also highlighting randomization of locations and the scaling process in assigning gap junctional synaptic strengths.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

We thank you for raising this point and agree with you on every point. Please note that we do not assert that the outward HCN currents are exclusively associated with gap junctional inputs. Rather, our results show that synchronous inputs generate outward HCN currents in both chemical synapses (Fig. 3B; positive/outward HCN currents, except in the no sodium or leak model) and gap junctions (Fig. 3D; positive/outward HCN currents). We emphasized this in the case of gap junctions because, in the absence of inward synaptic currents, HCN (acting as outward currents with synchronous inputs) contributed to the positive deflection observed in the LFPs. While HCN would also contribute in the case of chemical synapses, its effect was negligible due to the presence of large inward synaptic currents. Since LFPs reflect the collective total transmembrane currents, the dominant contributors differ between these two scenarios, which we aimed to highlight. Since HCN exhibited outward currents in our synchronous input simulations, we have elaborated on this mechanism in the supplementary figure (Fig. S3). Our intention was not to emphasize this effect for only one synaptic mode but rather to highlight HCN's contribution to the positive deflection as one of the contributing factors.

We agree that HCN currents are relatively small in magnitude; therefore, our conclusions were based on HCN being one of the several contributing factors. Leak conductance and other outward conductances, including HCN currents (Fig. 3D), collectively contribute to the positive deflections observed in the case of gap junctional synchronous inputs.

We will ensure that we will account for all the points appropriately in a revised manuscript.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

We thank you for raising this important point. The CA1 pyramidal neuronal model used in this study is built with ion-channel models derived from biophysical and electrophysiological recordings from these cells. As mentioned in the Methods section “Dynamics and distribution of active channels” and Supplementary Table S1, models for individual channels, their gating kinetics, and channel distributions across the somatodendritic arbor (wherever known) are all derived from their physiological equivalents. Importantly, these values were derived from previously validated models from the laboratory, which contain these very ion channel models and the exact same morphology (Roy & Narayanan, 2021). Please compare Supplementary Table S1 with the Table 1 from (Roy & Narayanan, 2021). Please note that this model was validated against several physiological measurements along the somatodendritic axis (Fig. 1 of (Roy & Narayanan, 2021)).

In a revised manuscript, we will explicitly mention this while also mentioning the different physiological properties that were used for the validation process from (Roy & Narayanan, 2021). We sincerely regret not mentioning these details in the current version of our manuscript.

We will fix these in a revised version of the manuscript.

References

Bedner, P., Steinhauser, C., & Theis, M. (2012). Functional redundancy and compensation among members of gap junction protein families? Biochim Biophys Acta, 1818(8), 1971-1984. https://doi.org/10.1016/j.bbamem.2011.10.016

Behrens, C. J., Ul Haq, R., Liotta, A., Anderson, M. L., & Heinemann, U. (2011). Nonspecific effects of the gap junction blocker mefloquine on fast hippocampal network oscillations in the adult rat in vitro. Neuroscience, 192, 11-19. https://doi.org/10.1016/j.neuroscience.2011.07.015

Buzsaki, G., Anastassiou, C. A., & Koch, C. (2012). The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nat Rev Neurosci, 13(6), 407-420. https://doi.org/10.1038/nrn3241

Einevoll, G. T., Destexhe, A., Diesmann, M., Grun, S., Jirsa, V., de Kamps, M., Migliore, M., Ness, T. V., Plesser, H. E., & Schurmann, F. (2019). The Scientific Case for Brain Simulations. Neuron, 102(4), 735-744. https://doi.org/10.1016/j.neuron.2019.03.027

Gold, C., Henze, D. A., Koch, C., & Buzsaki, G. (2006). On the origin of the extracellular action potential waveform: A modeling study. J Neurophysiol, 95(5), 3113-3128. https://doi.org/10.1152/jn.00979.2005

Hagen, E., Dahmen, D., Stavrinou, M. L., Linden, H., Tetzlaff, T., van Albada, S. J., Grun, S., Diesmann, M., & Einevoll, G. T. (2016). Hybrid Scheme for Modeling Local Field Potentials from Point-Neuron Networks. Cereb Cortex, 26(12), 4461-4496. https://doi.org/10.1093/cercor/bhw237

Halnes, G., Ness, T. V., Næss, S., Hagen, E., Pettersen, K. H., & Einevoll, G. T. (2024). Electric Brain Signals: Foundations and Applications of Biophysical Modeling. Cambridge University Press. https://doi.org/DOI: 10.1017/9781009039826

Lo, C. W. (1999). Genes, gene knockouts, and mutations in the analysis of gap junctions. Dev Genet, 24(1-2), 1-4. https://doi.org/10.1002/(SICI)1520-6408(1999)24:1/2%3C1::AID-DVG1%3E3.0.CO;2-U

Martinez-Canada, P., Ness, T. V., Einevoll, G. T., Fellin, T., & Panzeri, S. (2021). Computation of the electroencephalogram (EEG) from network models of point neurons. PLoS Comput Biol, 17(4), e1008893. https://doi.org/10.1371/journal.pcbi.1008893

Mazzoni, A., Linden, H., Cuntz, H., Lansner, A., Panzeri, S., & Einevoll, G. T. (2015). Computing the Local Field Potential (LFP) from Integrate-and-Fire Network Models. PLoS Comput Biol, 11(12), e1004584. https://doi.org/10.1371/journal.pcbi.1004584

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2018). h-Type Membrane Current Shapes the Local Field Potential from Populations of Pyramidal Neurons. J Neurosci, 38(26), 6011-6024. https://doi.org/10.1523/jneurosci.3278-17.2018

Reimann, M. W., Anastassiou, C. A., Perin, R., Hill, S. L., Markram, H., & Koch, C. (2013). A biophysically detailed model of neocortical local field potentials predicts the critical role of active membrane currents. Neuron, 79(2), 375-390. https://doi.org/10.1016/j.neuron.2013.05.023

Rouach, N., Segal, M., Koulakoff, A., Giaume, C., & Avignone, E. (2003). Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions. Journal of Physiology, 553(Pt 3), 729-745. https://doi.org/10.1113/jphysiol.2003.053439

Roy, A., & Narayanan, R. (2021). Spatial information transfer in hippocampal place cells depends on trial-to-trial variability, symmetry of place-field firing, and biophysical heterogeneities. Neural Netw, 142, 636-660. https://doi.org/10.1016/j.neunet.2021.07.026

Schomburg, E. W., Anastassiou, C. A., Buzsaki, G., & Koch, C. (2012). The spiking component of oscillatory extracellular potentials in the rat hippocampus. J Neurosci, 32(34), 11798-11811. https://doi.org/10.1523/JNEUROSCI.0656-12.2012

Sinha, M., & Narayanan, R. (2015). HCN channels enhance spike phase coherence and regulate the phase of spikes and LFPs in the theta-frequency range. Proc Natl Acad Sci U S A, 112(17), E2207-2216. https://doi.org/10.1073/pnas.1419017112

Sinha, M., & Narayanan, R. (2022). Active Dendrites and Local Field Potentials: Biophysical Mechanisms and Computational Explorations. Neuroscience, 489, 111-142. https://doi.org/10.1016/j.neuroscience.2021.08.035

Sirmaur, R., & Narayanan, R. (2024). Distinct extracellular signatures of chemical and electrical synapses impinging on active dendrites differentially contribute to ripple-frequency oscillations. Society for Neuroscience annual meeting (https://www.abstractsonline.com/pp8/?_gl=1*1bxo7m*_gcl_au*MTc5MTQ0NjE0NC4xNzI3MDcwOTMw*_ga*MTMxMTE5OTcyMy4xNzI3MDcwOTMx*_ga_T09K 3Q2WDN*MTcyNzA3MDkzMS4xLjEuMTcyNzA3MDkzNy41NC4wLjA.#!/20433/ presentation/13949), Chicago, USA.

Szarka, G., Balogh, M., Tengolics, A. J., Ganczer, A., Volgyi, B., & Kovacs-Oller, T. (2021). The role of gap junctions in cell death and neuromodulation in the retina. Neural Regen Res, 16(10), 1911-1920. https://doi.org/10.4103/1673-5374.308069

eLife Assessment

This study resolves a cryo-EM structure of the GPCR, human GPR30, which responds to bicarbonate and regulates cellular responses to pH and ion homeostasis. Understanding the ligand and the mechanism of activation is important to the field of receptor signaling and potentially facilitates drug development targeting this receptor. Structures and functional assays provide solid evidence for a potential bicarbonate binding site.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers.]

Summary:

This study resolves a cryo-EM structure of the GPCR, GPR30, in the presence of bicarbonate, which the author's lab recently identified as the physiological ligand. Understanding the ligand and the mechanism of activation is of fundamental importance to the field of receptor signaling. This solid study provides important insight into the overall structure and suggests a possible bicarbonate binding site.

Strengths:

The overall structure, and proposed mechanism of G-protein coupling are solid. Based on the structure, the authors identify a binding pocket that might accommodate bicarbonate. Although assignment of the binding pocket is speculative, extensive mutagenesis of residues in this pocket identifies several that are important to G-protein signaling. The structure shows some conformational differences with a previous structure of this protein determined in the absence of bicarbonate (PMC11217264). To my knowledge, bicarbonate is the only physiological ligand that has been identified for GPR30, making this study an important contribution to the field. However, the current study provides novel and important circumstantial evidence for the bicarbonate binding site based on mutagenesis and functional assays.

Weaknesses:

Bicarbonate is a challenging ligand for structural and biochemical studies, and because of experimental limitations, this study does not elucidate the exact binding site. Higher resolution structures would be required for structural identification of bicarbonate. The functional assay monitors activation of GPR30, and thus reports on not only bicarbonate binding, but also the integrity of the allosteric network that transduces the binding signal across the membrane. However, biochemical binding assays are challenging because the binding constant is weak, in the mM range.

The authors appropriately acknowledge the limitations of these experimental approaches, and they build a solid circumstantial case for the bicarbonate binding pocket based on extensive mutagenesis and functional analysis. However, the study does fall short of establishing the bicarbonate binding site.

Reviewer #2 (Public review):

Summary:

In this manuscript, "Cryo-EM structure of the bicarbonate receptor GPR30," the authors aimed to enrich our understanding of the role of GPR30 in pH homeostasis by combining structural analysis with a receptor function assay. This work is a natural development and extension of their previous work on Nature Communications (PMID: 38413581). In the current body of work, they solved the cryo-EM structure of the human GPR30-G-protein (mini-Gsqi) complex in the presence of bicarbonate ions at 3.15 Å resolution. From the atomic model built based on this map, they observed the overall canonical architecture of class A GPCR and also identified 3 extracellular pockets created by ECLs (Pockets A-C). Based on the polarity, location, size, and charge of each pocket, the authors hypothesized that pocket A is a good candidate for the bicarbonate binding site. To identify the bicarbonate binding site, the authors performed an exhaustive mutant analysis of the hydrophilic residues in Pocket A and analyzed receptor reactivity via calcium assay. In addition, the human GPR30-G-protein complex model also enabled the authors to elucidate the G-protein coupling mechanism of this special class A GPCR, which plays a crucial role in pH homeostasis.

Strengths:

As a continuation of their recent Nature Communications publication, the authors used cryo-EM coupled with mutagenesis and functional studies to elucidate bicarbonate-GPR30 interaction. This work provided atomic-resolution structural observations for the receptor in complex with G-protein, allowing us to explore its mechanism of action, and will further facilitate drug development targeting GPR30. There were 3 extracellular pockets created by ECLs (Pockets A-C). The authors were able to filter out 2 of them and hypothesized that pocket A was a good candidate for the bicarbonate binding site based on the polarity, location, and charge of each pocket. From there, the authors identified the key residues on GPR30 for its interaction with the substrate, bicarbonate. Together with their previous work, they mapped out amino acids that are critical for receptor reactivity.

Weaknesses:

When we see a reduction of a GPCR-mediated downstream signaling, several factors could potentially contribute to this observation: 1) a reduced total expression of this receptor due to the mutation (transcription and translation issue); 2) a reduced surface expression of this receptor due to the mutation (trafficking issue); and 3) a dysfunctional receptor that doesn't signal due to the mutation.

Altogether, the wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

Reviewer #3 (Public review):

Summary

GPR30 responds to bicarbonate and plays a role in regulating cellular pH and ion homeostasis. However, the molecular basis of bicarbonate recognition by GPR30 remains unresolved. This study reports the cryo-EM structure of GPR30 bound to a chimeric mini-Gq in the presence of bicarbonate, revealing mechanistic insights into its G-protein coupling. Nonetheless, the study does not identify the bicarbonate-binding site within GPR30.

Strengths

The work provides strong structural evidence clarifying how GPR30 engages and couples with Gq.

Weaknesses

Several GPR30 mutants exhibited diminished responses to bicarbonate, but their expression levels were also reduced. As a result, the mechanism by which GPR30 recognizes bicarbonate remains uncertain.

Author Response:

The following is the authors’ response to the previous reviews

Reviewer #1 (Public review):

Summary:

This study resolves a cryo-EM structure of the GPCR, GPR30, in the presence of bicarbonate, which the author's lab recently identified as the physiological ligand. Understanding the ligand and the mechanism of activation is of fundamental importance to the field of receptor signaling. This solid study provides important insight into the overall structure and suggests a possible bicarbonate binding site.

Strengths:

The overall structure, and proposed mechanism of G-protein coupling are solid. Based on the structure, the authors identify a binding pocket that might accommodate bicarbonate. Although assignment of the binding pocket is speculative, extensive mutagenesis of residues in this pocket identifies several that are important to G-protein signaling. The structure shows some conformational differences with a previous structure of this protein determined in the absence of bicarbonate (PMC11217264). To my knowledge, bicarbonate is the only physiological ligand that has been identified for GPR30, making this study an important contribution to the field. However, the current study provides novel and important circumstantial evidence for the bicarbonate binding site based on mutagenesis and functional assays.

Weaknesses:

Bicarbonate is a challenging ligand for structural and biochemical studies, and because of experimental limitations, this study does not elucidate the exact binding site. Higher resolution structures would be required for structural identification of bicarbonate. The functional assay monitors activation of GPR30, and thus reports on not only bicarbonate binding, but also the integrity of the allosteric network that transduces the binding signal across the membrane. However, biochemical binding assays are challenging because the binding constant is weak, in the mM range.

The authors appropriately acknowledge the limitations of these experimental approaches, and they build a solid circumstantial case for the bicarbonate binding pocket based on extensive mutagenesis and functional analysis. However, the study does fall short of establishing the bicarbonate binding site.

We thank the reviewer for this thoughtful and constructive assessment of our revised manuscript. We are grateful for the recognition of the overall quality of the cryo-EM structure and the proposed mechanism of G-protein coupling, as well as for highlighting the importance of identifying bicarbonate as a physiological ligand for GPR30 and the contribution this work makes to the receptor signaling field. We also appreciate the reviewer’s careful and balanced discussion of the inherent challenges posed by bicarbonate as a low-affinity, small, negatively charged ligand, and we fully agree that, given current experimental limitations, our data provide circumstantial—rather than definitive—evidence for the binding site and that higher-resolution structures would be required for direct visualization. Importantly, we value the reviewer’s acknowledgement that we transparently describe these limitations and that our extensive mutagenesis and functional analyses nonetheless build a solid case for the proposed bicarbonate-binding pocket, which we believe will serve as a useful framework for future biochemical and structural investigation

Reviewer #1 (Recommendations for the authors):

Overall, the authors do a good job responding to the previous review, with updated structures and experimental data. I have two comments on the current version:

(1) When the authors compare their structure to a previously published structure of the same receptor, they say that the previous structure came out while the current manuscript was in revision (line 255). This is not correct. The previous manuscript was published May 14, 2024, and the current manuscript was received by eLife on May 20, 2024. This sentence should be corrected to "During the preparation of this manuscript..."

We corrected the sentence accordingly (line 259).

(2) Line 173: what other structures are the authors referring to? Citations should be included here.

Is Line 193 correct? We added citations (line 190).

Reviewer #2 (Public review):

Summary:

In this manuscript, "Cryo-EM structure of the bicarbonate receptor GPR30," the authors aimed to enrich our understanding of the role of GPR30 in pH homeostasis by combining structural analysis with a receptor function assay. This work is a natural development and extension of their previous work on Nature Communications (PMID: 38413581). In the current body of work, they solved the cryo-EM structure of the human GPR30-G-protein (mini-Gsqi) complex in the presence of bicarbonate ions at 3.15 Å resolution. From the atomic model built based on this map, they observed the overall canonical architecture of class A GPCR and also identified 3 extracellular pockets created by ECLs (Pockets A-C). Based on the polarity, location, size, and charge of each pocket, the authors hypothesized that pocket A is a good candidate for the bicarbonate binding site. To identify the bicarbonate binding site, the authors performed an exhaustive mutant analysis of the hydrophilic residues in Pocket A and analyzed receptor reactivity via calcium assay. In addition, the human GPR30-G-protein complex model also enabled the authors to elucidate the G-protein coupling mechanism of this special class A GPCR, which plays a crucial role in pH homeostasis.

Strengths:

As a continuation of their recent Nature Communications publication, the authors used cryo-EM coupled with mutagenesis and functional studies to elucidate bicarbonate-GPR30 interaction. This work provided atomic-resolution structural observations for the receptor in complex with G-protein, allowing us to explore its mechanism of action, and will further facilitate drug development targeting GPR30. There were 3 extracellular pockets created by ECLs (Pockets A-C). The authors were able to filter out 2 of them and hypothesized that pocket A was a good candidate for the bicarbonate binding site based on the polarity, location, and charge of each pocket. From there, the authors identified the key residues on GPR30 for its interaction with the substrate, bicarbonate. Together with their previous work, they mapped out amino acids that are critical for receptor reactivity.

Weaknesses:

When we see a reduction of a GPCR-mediated downstream signaling, several factors could potentially contribute to this observation: 1) a reduced total expression of this receptor due to the mutation (transcription and translation issue); 2) a reduced surface expression of this receptor due to the mutation (trafficking issue); and 3) a dysfunctional receptor that doesn't signal due to the mutation. In the current revision, based on the gating strategy, the surface expression of the HA-positive WT GPR30-expressing cells is only 10.6% of the total population, while the surface expression levels of the mutants range from 1.89% (P71A) to 64.4% (D111A). Combining this information with the functional readout in Figure 3F and G, as well as their previous work, the authors concluded that mutations at P71, E115, D125, Q138, C207, D210, and H307 would decrease bicarbonate responses. Among those sites,

E115, Q138, and H307 were from their previous Nature Comm paper.

Authors claim P71 and C207 make a structural-stability contribution, as their mutations result in a significant reduction in surface expression: P71A (1.89%) and C207A (2.71%). However, compared to 10.6% of the total population in the WT, (P71A is 17.8% of the WT, and C207A is 25.6% of the WT), this doesn't rule out the possibility that the mutated receptor is also dysfunctional: at 10 mM NaHCO3, RFU of WT is ~500, RFU of P71 and C207 are ~0.

The authors also interpret "The D125ECL1A mutant has lost its activity but is located on the surface" and only mention "D125 is unlikely to be a bicarbonate binding site, and the mutational effect could be explained due to the decreased surface expression". Again, compared to 10.6% of the total population in the WT, D125A (3.94%) is 37.2% of the WT. At 10 mM NaHCO3, the RFU of the WT is ~500, the RFU of D125 is ~0. This doesn't rule out the possibility that the mutated receptor is also dysfunctional. It is not clear why D125A didn't make it to the surface.

Other mutants that the authors didn't mention much in their text: D111A (64.4%, 607.5% of WT surface expression), E121A (50.4%, 475.5% of WT surface expression), R122 (41.0%, 386.8% of WT surface expression), N276A (38.9%, 367.0% of WT surface expression) and E218A (24.6%, 232.1% of WT surface expression) all have similar RFU as WT, although the surface expression is about 2-6 times more. On the other hand, Q215A (3.18%, 30% of WT surface expression) has similar RFU as WT, with only a third of the receptor on the surface.

Altogether, the wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

We sincerely thank the reviewer for their careful reading and thoughtful evaluation of our manuscript on the cryo-EM structure of the bicarbonate receptor GPR30. We greatly appreciate the reviewer’s positive assessment of the overall significance of combining structural determination with extensive mutagenesis and functional assays to advance understanding of bicarbonate–GPR30 interactions and G-protein coupling, as well as their recognition that these atomic-level insights will be valuable for future mechanistic studies and drug-development efforts. We are also grateful for the reviewer’s constructive critique regarding the interpretation of reduced signaling in the context of variable surface expression across mutants, which highlights an important point about disentangling effects of expression/trafficking from intrinsic receptor dysfunction; these comments are highly insightful and will help us strengthen the clarity and rigor of our presentation and conclusions in the revised manuscript.

Reviewer #2 (Recommendations for the authors):

In this revision, the authors have made a significant effort to improve and validate the structural observations, as well as address the comments in the previous submission. They updated the functional assays and evaluated the receptor function by measuring intracellular calcium mobilization, which is a more direct measurement for the downstream signaling of hGPR30-Gq signaling. They also used flow cytometry with an HA-antibody for a more direct measurement of the surface expression of the receptor, replacing their previous assay that normalized to the housekeeping gene Na-K-ATPase.

I appreciate the effort the authors made to address the previous comments made by the reviewers. However, there are still some concerns about the current data.

(1) The authors have addressed my previous comment on untangling the mixture of their previous and new data in the "insights into bicarbonate binding" section. They have made it clear that the importance of E115, Q138, and H307 in the receptor-bicarbonate interaction was shown in their Nature Communications paper.

(2) The authors have addressed my previous comment on adding some content about the physiological concentration of HCO3, or referring more to their previous work about the rationale to select the bicarbonate dose in their functional assay.

(3) The authors have updated Figure 3

(4) The authors have updated Supplemental Figure 1 to show the full gel with molecular weight markers in the supplemental data to demonstrate the sample purity.

(5) The authors have updated the predicted model using AF3

(6) The authors added E218A as suggested before.

Some new suggestions for this R1:

(1) The wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

We acknowledge this limitation. The wide range of surface expression among cell lines, together with differences in assay modalities, may introduce variability that complicates direct quantitative comparisons and therefore only partially supports the structural observations. Future work using more standardized expression systems and matched functional readouts will be important to strengthen the structure–function linkage.

(2) Line 101, "ICL1 and ECL1 contain short α helices", no α helix of ICL1 is shown in Figure 2C

We removed the word “ICL1” (line 98).

(3) For the unsolved region of ECL2, could the author put a dashed line connecting ECL2 with TM4? In the current Figure 2B, it looks like ECL2 connects TM3 and TM5.

According to the suggestion, we corrected Figure 2B.

(4) I appreciate that the authors updated the predicted model with AF3, but they didn't make it clear why they had the comparison between their cryo-EM structure (bicarbonate-activated G-protein-incorporated GPR30) and the predicted AF3 model (inactive GPR30)

We wish to assert the usefulness of experimental structures, not merely predictions. These include structures independent of receptor activation, such as SS bonds.

(5) I appreciate that the authors have addressed my previous comment on adding some content about the physiological concentration of HCO3, but it was still not clear to me why they picked 11 mM in Figure 3G for the bar graph. Also, since a dose-response curve was made in Figure 3F, why not just calculate and report the EC50 of NaHCO3 for each mutant?

Thank you for your comment. Thank you for the comment. We’ve calculated the EC50 of the calcium response and assessed its correlation with receptors’ cell surface expression. We chose 11 mM in Fig .3G since our previous paper in Nature Communications showed the EC50 value of IPs assay was around 11 mM. However, the calcium response was more sensitive and gave a lower value than expected. Therefore, according to your advice, we deleted the bar graph with 11 mM responses, calculated EC50, and drew pictures of the correlation among cell surface expression, EC50, and maximum responses (Figure 3F-I, Supplementary File 1). Moreover, we revised the explanation about this mutagenesis study (lines139-154 and 217-230).

(6) In the previous submission and comments, E218 was in close contact with bicarbonate in the previous Figure 4D (the bicarbonate is deleted in the new structure). I thank the authors for making an E218A mutant and performing the functional assay. As mentioned above, E218A (24.6%, 232.1% of WT surface expression) has a similar functional readout as WT. Doesn't this also indicate that E218A is partially broken, so you will need twice as much as WT to have the same downstream signal?

Thank you for your comment. In our revised manuscript, we described the correlation between cell surface expression and EC50 and found that cell surface expression and the response to bicarbonate are not correlated, which you mentioned in your review comment (Figure 3F-I, Supplementary File 1). There are many possibilities that could explain this: GPR30 localization in specific spots on the plasma membrane might limit the response stoichiometry, GPR30 might also work intracellularly to blunt the increased response because of more GPR30 expression on PM, redundant GPR30 on PM might be broken, or E118A might be less functional and need twice as much as WT. We will examine cell surface expression of GPR30 and its response to bicarbonate in a future study.

I would suggest that the authors in future studies consider using the Tet-on inducible cell lines, such as HEK293 Flp-In Trex. These cell lines will allow the authors to fine-tune the surface expression of their mutants to the same level with different doses of Tetracycline in their stable cell lines.

We appreciate your advice. We’ll introduce Tet-on inducible cell lines for future research.

Reviewer #3 (Public review):

Summary

GPR30 responds to bicarbonate and plays a role in regulating cellular pH and ion homeostasis. However, the molecular basis of bicarbonate recognition by GPR30 remains unresolved. This study reports the cryo-EM structure of GPR30 bound to a chimeric mini-Gq in the presence of bicarbonate, revealing mechanistic insights into its G-protein coupling. Nonetheless, the study does not identify the bicarbonate-binding site within GPR30.

Strengths

The work provides strong structural evidence clarifying how GPR30 engages and couples with Gq.

Weaknesses

Several GPR30 mutants exhibited diminished responses to bicarbonate, but their expression levels were also reduced. As a result, the mechanism by which GPR30 recognizes bicarbonate remains uncertain, leaving this aspect of the study incomplete.

We sincerely thank the reviewer for this thoughtful and balanced assessment of our manuscript, including the clear summary of the central advance and the constructive identification of remaining limitations. We particularly appreciate the recognition that our cryo-EM analysis provides strong structural evidence for how GPR30 engages and couples with Gq, and we agree that pinpointing the bicarbonate-binding site remains a critical open question. In the revised manuscript, we will make this point more explicit, clarify the interpretation of the mutagenesis results in light of reduced receptor expression for some variants, and further strengthen the presentation and discussion of what our current data do—and do not—allow us to conclude regarding bicarbonate recognition by GPR30

Reviewer #3 (Recommendations for the authors):

The authors have removed the bicarbonate assignment from their model and have addressed all of my concerns. In this study, or in future work, it would be advisable for the authors to explore the use of bicarbonate mimetics with higher binding affinity to facilitate more definitive structural characterization.

Thank you for this constructive suggestion. We agree that exploring bicarbonate mimetics with higher binding affinity would be an important next step to enable more definitive structural characterization of GPR30 and to strengthen mechanistic conclusions. In future work, we plan to pursue the identification and/or design of such mimetics, guided by the architecture and mutational landscape of the extracellular pocket described here, and to combine these ligands with optimized cryo-EM sample preparation and complementary functional assays to better stabilize and visualize the bound state.

eLife Assessment

This is an important study on the role of the neurokinin-2 receptor (NK2R) as a regulatory node connecting intestinal lipid metabolism, mucosal immunity, and the gut microbiome, bidirectionally regulating enterocyte lipid uptake, lipid droplet storage, chylomicron output, and systemic metabolic parameters in DIO mice. The authors present solid evidence linking Tacr2 deletion to reprogrammed epithelial lineage allocation, dampened immune gene expression, and male-biased protection from DSS colitis, despite dysbiotic microbiota. However, the causal evidence for some mechanistic and pro-inflammatory NK2R claims remains incomplete and potentially confounding, requiring additional cell-type-specific and functional experiments.

Reviewer #1 (Public review):

Summary:

This study identifies NK2R as an intestinal GPCR that tunes enterocyte lipid uptake, lipid droplet storage, and chylomicron output, with loss or antagonism enhancing post‑prandial triglyceridemia and epithelial lipid stores, and agonism reducing adiposity and improving glycemia in DIO mice. Through bulk RNA‑seq, deconvolution, DSS colitis, and 16S profiling, the authors link Tacr2 deletion to coordinated induction of epithelial lipid‑metabolic programs, dampened immune gene expression, sex‑specific remodeling of secretory lineages, and male‑biased protection from experimental colitis despite dysbiotic microbiota. This is an overall important and thorough paper on an emerging obesity drug target, but it should temper some interpretations, and the following points would be needed to strengthen the claims in the manuscript.

Strengths:


The study uses an impressive combination of genetic loss‑of‑function, pharmacological agonism/antagonism, transcriptomics, and in vivo physiology to establish NK2R as a bidirectional regulator of epithelial lipid handling. The integration of RNA‑seq, epithelial cell‑type deconvolution, DSS colitis, and microbiome profiling provides a rich, systems‑level view of how Tacr2 deletion reshapes epithelial metabolism, lineage allocation, and inflammatory responsiveness in a sex‑specific manner. The gain- and loss‑of‑function data particularly support a model in which NK2R acts as an epithelial metabolic rheostat that restrains lipid absorption and chylomicron export, with downstream consequences for barrier fitness and immune tone.

Weaknesses:

Major points

While the data convincingly establish NK2R's role in epithelial lipid handling, the manuscript arguably overstates a primary "pro‑inflammatory" function for NK2R, given that Tacr2‑/‑ mice show enhanced enterocyte lipid uptake and storage, higher post‑prandial triglycerides, and a dysbiotic microbiota yet reduced mucosal immune gene expression and, in males, protection from DSS colitis. It remains equally plausible that the apparent "protection" reflects a mucosa that is less reactive to unfavorable microbiota rather than genuinely protected, and that NK2R's main function is metabolic, with immune changes emerging secondarily. Such a model would actually help reconcile the long-standing question as to why NK2R antagonism has not translated into clear benefit in clinical trials for GI inflammation over the past several decades.

Without temporal resolution, it is equally plausible that antagonists primarily perturb epithelial lipid homeostasis rather than directly and beneficially modulating immune tone. To discriminate between these possibilities and strengthen the potential direct inflammatory claims, the authors should:

(1) generate epithelial‑specific, immune‑cell-specific, and nociceptor‑specific Tacr2 deletions in the DSS model

(2) test gut‑restricted NK2R agonism versus antagonism under controlled dietary fat conditions for effects on LD load, barrier integrity, and colitis severity

(3) perform ex vivo tachykinin/NK2R stimulation of isolated epithelial versus immune compartments with functional readouts

(4) assess whether microbiota transfer from Tacr2‑/‑ versus WT donors into germ‑free or antibiotic‑treated recipients can recapitulate protection or susceptibility independently of epithelial NK2R status.

Minor points

Additional clarifications on Tac1 and tachykinin receptor expression in male/female colitis models, and validation of the NK2R antibody in KO tissue (or in situ hybridization), would also be needed to strengthen key mechanistic and localization claims.

Reviewer #2 (Public review):

Summary:

This manuscript- "NK2R signaling governs intestinal lipid mobilization and mucosal inflammation" by Perez et al investigates the role of the neurokinin-2 receptor (NK2R) as a regulatory node connecting intestinal lipid metabolism, mucosal immunity, and the gut microbiome. The authors utilized a ubiquitous deleted Tacr2 mouse model alongside targeted pharmacological treatments to demonstrate that NK2R limits luminal lipid uptake and chylomicron secretion. Additionally, the study uncovers that Tacr2 deficiency promotes male-biased protection against DSS-induced colitis and drives distinct diet- and genotype-dependent shifts in the fecal microbiota.

Strengths:

(1) The authors successfully utilized both a genetic whole-body knockout model (Tacr2-/-) and targeted pharmacological agents, such as the antagonist GR159897 and the agonist EB1002. This dual approach effectively corroborates the core phenotypic findings.

(2) The study provides a compelling case for targeting the tachykinin-NK2R axis therapeutically. The remarks that NK2R agonists could be leveraged to treat obesity, while antagonists might be used for inflammatory bowel disease, will be an exciting clinical outcome if further validated.

(3) The integration of RNAseq for epithelial lineage analysis, combined with in vivo gut permeability assays, lipid tolerance assays, and 16S microbiome sequencing, provides a robust and highly detailed physiological picture.

Weaknesses:

This manuscript has some notable limitations. While the transcriptomic data show an upregulation of the enterocyte lipid droplet program in Tacr2-/- mice, the manuscript lacks biochemical experiments to conclude the downstream signaling mechanism driving such changes. The reliance on a global whole-body knockout model confounds the ability to definitively conclude that the observed metabolic and inflammatory phenotypes are linked to the intestinal epithelium. The authors discuss a male-biased protection against DSS-induced colitis, but they rely on speculation regarding sex hormones rather than providing experimental data to explain this dimorphism.

eLife Assessment

This useful study identified XAP5 as an ancient transcriptional regulator critical for primary ciliogenesis. The evidence supporting the conceptual framework linking evolutionary conservation to functional specialization in primary ciliogenesis remains incomplete. This work will be of interest to developmental biologists and to those studying diseases caused by ciliopathies.

Reviewer #1 (Public review):

Summary:

The authors have attempted to establish a role for XAP5, a transcriptional regulator they have previously identified for flagellar biogenesis in Chlamydomonas and mice, in primary cilia differentiation.

Strengths:

Genetic and biochemical analysis using a cultured mouse cell line, NIH3T3.

Weaknesses:

(1) The authors have ignored established data that, like in C. elegans and Drosophila, there is in vivo genetic evidence that primary cilia formation is regulated by the RFX transcriptional module (for example, PMID 19887680, PMID 29510665).

(2) The analysis with one mammalian cell line, NIH3T3, while done quite rigorously, is not sufficient. Also, the effect on cilia differentiation is very modest - a shortening of cilia length on XAP5, NONO and SOX5 knockout - which can happen for a variety of reasons, especially in culture conditions. In my view, this relatively mild phenotype does not establish that the XAP5/NONO and SOX5 axis is an important regulator of primary cilia differentiation.

(3) The lack of any data that validates the findings in the model vertebrate is a major weakness of this paper. Validation using clean genetics (whole body knockouts or tissue-specific conditional knockouts) is absolutely essential for these data to be acceptable.

Reviewer #2 (Public review):

Summary:

This study investigates how evolutionarily conserved transcription factors are repurposed to regulate the functional diversification of cilia. Building on previous work identifying Xap5 as a regulator of motile ciliogenesis during spermatogenesis, the authors now propose a broader role for Xap5 as a master regulator of primary ciliogenesis. Through extensive mechanistic analyses, they identify an Xap5-NONO-SOX transcriptional axis and suggest that this module contributes to ciliary diversity and may be implicated in ciliopathies.

Overall, the work addresses an important and timely question regarding the transcriptional control of primary ciliogenesis. However, additional evidence is required to fully support the proposed conceptual framework linking evolutionary conservation to functional specialization.

Strengths:

(1) Addresses a timely and fundamental question in cilia biology.

(2) Extends Xap5 function beyond motile ciliogenesis.

(3) Identifies a novel regulatory axis (Xap5-NONO-SOX).

(4) Combines multiple well-designed mechanistic approaches.

(5) Proposes an interesting conceptual framework linking evolution and ciliogenesis.

Weaknesses:

(1) Specificity for primary ciliogenesis not demonstrated.

(2) No data on motile ciliogenesis in somatic MCCs.

(3) Conclusions drawn from NIH/3T3 cells (murine stromal cells).

(4) GC-rich motif identified but underexplored.

(5) Link to ciliopathies is speculative.

eLife Assessment

These findings are important because they suggest that more selective JAK inhibition, particularly targeting JAK1 or JAK2, can effectively reduce organ pathology and pathogenic IFN-γ-producing immune cells in AIRE deficiency, refining therapeutic strategies beyond broad JAK inhibition. The work highlights JAK2 inhibition as a promising and potentially more targeted clinical approach for treating autoimmunity in this setting. The evidence is solid and moderately strong, building on the prior efficacy of ruxolitinib and supported by comparative studies in Aire-deficient models, though further validation in human systems would strengthen translational confidence.

Reviewer #1 (Public review):

Summary:

Heller et al use a murine model of AIRE deficiency, a disease that leads to systemic autoimmune disease, to demonstrate differential effects of selective JAK inhibitors. This group and others have previously demonstrated the efficacy of the JAK1/2 inhibitor ruxolitinib in patients with AIRE deficiency. Here, they focus on the ability of ruxolitinib versus drugs inhibiting either JAK1, JAK2, or JAK3 to alter organ pathology and accumulation of interferon-gamma producing immune cells in the lungs, which are important mediators of inflammation in patients with this disease. The current study provides evidence that selective JAK2 or JAK1 both reduce disease in this mouse model. There is potentially a more beneficial effect of selective JAK2 inhibition, although these differences are minor, and it is uncertain whether this is clinically relevant for patients. They demonstrate that inhibition of JAK3 alone in the mouse was clearly not beneficial for disease. Overall, this study provides evidence for consideration of more selective JAK inhibition in patients with AIRE deficiency.

Strengths:

(1) Robust model for investigating AIRE deficiency.

(2) They combine cellular studies (immune cell production of IFN-g) and robust organ pathology scoring to evaluate the effects of the drugs tested here.

(3) Data clearly demonstrates that JAK3 inhibition, at least as used here, may increase IFN-g production and does not reduce organ pathology.

Weaknesses:

(1) There is no direct comparison of the effects of JAK2 vs. JAK1 inhibition to support that JAK2 inhibition is clearly superior.

(2) They were not able to perform pharmacokinetic studies or measure the efficacy of JAK inhibition in their model, and it is uncertain how the doses of drug used here will translate to the treatment of patients.

(3) It is uncertain whether this study, performed in a murine model, will correspond to tissue/cell specificity of JAK inhibition in patients.

Reviewer #2 (Public review):

Summary:

This work from Heller et al. examines the differential responses of treatment with selective JAK inhibitors in Aire knockout mice, which develop several autoimmune diseases. The authors had previously shown efficacious responses in both mice and humans with a broader JAK-I, Ruxolitinib, that had Aire-deficiency. Because of the side effect profile, it may be better to determine if selective JAK-I therapy could continue to work with less of the side effects of Ruxolitinib. Here, they develop a protocol of treating mice for four weeks with JAK1,2, and 3 inhibitors and then examining tissues for infiltration of T cells and gamma-interferon-producing T cells. They also perform analyses of infiltration of the tissues versus intravascular localization of T cells. They find that JAK2 inhibition provided the most robust results for decreasing infiltrates and gamma interferon-producing T cells. All JAK-I's resulted in decreased T cell infiltration of tissues, and somewhat paradoxically, the JAK3 inhibitor caused an increased accumulation of gamma-interferon-producing T cells in tissues.

Strengths:

This is a nice set of studies that makes some inroads on a more refined approach to treating autoimmunity in the Aire knockout model. The work here will be important for developing the next clinical trial for patients with APS1 and represents an advance for efforts in that space.

Weaknesses:

The increase in gamma-interferon-producing cells in tissues with JAK3 inhibition is interesting, but essentially remains unanswered in any way. There is a minimal assessment of the broad STAT pathways that the selective JAK-i's could be hitting, and perhaps that could be assessed more systematically. Finally, there is no pharmacokinetic data, which makes comparisons between the treatments a bit limited.

eLife Assessment

This is a valuable study that investigates the role of the long non-coding RNA Dreg1 for the development, differentiation, or maintenance of group 2 ILC (ILC2). The authors generate Dreg1-/- mice and show solid evidence for a reduction of group 2 innate lymphoid cells (ILC2). However, the strength of evidence supporting and analysing the impact of Dreg1 on Gata3 expression, a transcription factor required for ILC2 cell fate decisions, remains incomplete. This study will be of interest to immunologists.

Reviewer #1 (Public review):

Summary:

This study examines the role of the long non-coding RNA Dreg1 in regulating Gata3 expression and ILC2 development. Using Dreg1 deficient mice, the authors show a selective loss of ILC2s but not T or NK cells, suggesting a lineage-specific requirement for Dreg1. By integrating public chromatin and TF-binding datasets, they propose a Tcf1-Dreg1-Gata3 regulatory axis. The topic is relevant for understanding epigenetic regulation of ILC differentiation.

Strengths:

(1) Clear in vivo evidence for a lineage-specific role of Dreg1.

(2) Comprehensive integration of genomic datasets.

(3) Cross-species comparison linking mouse and human regulatory regions.

Weaknesses:

(1) Mechanistic conclusions remain correlative, relying on public data.

(2) Lack of direct chromatin or transcriptional validation of Tcf1-mediated regulation.

(3) Human enhancer function is not experimentally confirmed.

(4) Insufficient methodological detail and limited mechanistic discussion.

Comments on revisions:

The authors have provided clear evidence that Dreg1 is necessary for ILC2 development, but their refusal to perform any mechanistic experiment remains a significant weakness. While their appeal to the 3Rs and the use of public datasets is noted, re-analyzing external data from heterogeneous sources cannot substitute for direct, internal validation of the Tcf1-Dreg1-Gata3 axis in their specific knockout model. This is particularly problematic because ILC2 progenitors, though rare, can be isolated from bone marrow, especially since assays like CUT&Tag and others are specifically designed for low cell numbers. By relying on public T-cell CRISPR screens to justify human ILC2 functions, the authors are substituting cross-cell-type correlation for definitive functional proof. Consequently, the manuscript currently describes a discovery of necessity without providing a verified molecular mechanism, which should be more explicitly reflected in the title and conclusions.

Reviewer #2 (Public review):

The authors investigate the role of the long non-coding RNA Dreg1 for the development, differentiation or maintenance of group 2 ILC (ILC2). Dreg1 is encoded close to the Gata3 locus, a transcription factor implicated in the differentiation of T cells and ILC, and in particular of type 2 immune cells (i.e., Th2 cells and ILC2). The center of the paper is the generation of a Dreg1-deficient mouse. The role of Dreg1 in ILC2 was documented by mixed bone marrow experiments. While Dreg1-/- mice did not show any profound ab T or gd T cell, ILC1, ILC3 and NK cell phenotypes, ILC2 frequencies were reduced in various organs tested (small intestine, lung, visceral adipose tissue). In the bone marrow, immature ILC2 or ILC2 progenitors were reduced whereas a common ILC progenitor was overrepresented suggesting a differentiation block. Using ATAC-seq, the authors find the promoter of Dreg1 is open in early lymphoid progenitors and the acquisition of chromatin accessibility downstream correlates with increased Dreg1 expression in ILC2 progenitors. Examining publicly available Tcf1 CUT&Run data, they find that Tcf1 was specifically bound to the accessible sites of the Dreg1 locus in early innate lymphoid progenitors. Finally, the syntenic region in the human genome contains two non-coding RNA genes with an expression pattern resembling mouse Dreg1.

The topic of the manuscript is interesting. The article is focused on the first description of the Dreg1 knockout mouse and the specific effect of Dreg1 deficiency on ILC2 development.

(1) The data of how Dreg1 contributes to the differentiation and or maintenance of ILC2 is not addressed at a very definitive level. Does Dreg1 affect Gata3 expression, mRNA stability or turnover in ILC2? Previous work of the authors indicated that knock-down of Dreg1 does not affect Gata3 expression (PMID: 32970351). The current data (Figure 2H) showed small differences in Gata3 expression in CHILP which were, however, not statistically significant. No differences were found in ILCP and ILC2P.

(2) How Dreg1 exactly affects ILC2 differentiation remains unclear.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study examines the role of the long non-coding RNA Dreg1 in regulating Gata3 expression and ILC2 development. Using Dreg1-deficient mice, the authors show a selective loss of ILC2s but not T or NK cells, suggesting a lineage-specific requirement for Dreg1. By integrating public chromatin and TF-binding datasets, they propose a Tcf1-Dreg1-Gata3 regulatory axis. The topic is relevant for understanding epigenetic regulation of ILC differentiation.

Strengths:

(1) Clear in vivo evidence for a lineage-specific role of Dreg1.

(2) Comprehensive integration of genomic datasets.

(3) Cross-species comparison linking mouse and human regulatory regions.

Weaknesses:

(1) Mechanistic conclusions remain correlative, relying on public data.

We agree that the mechanistic conclusions are of our study are indeed correlative and we mention this in the discussion. The primary work of the study is the discovery of Dreg1's necessity for ILC2 development via the new knockout mouse model. Re-analysing good quality publicly available data on rare cell populations is an appropriate approach and in line with DORA guidelines for ethical research.

(2) Lack of direct chromatin or transcriptional validation of Tcf1-mediated regulation.

The most appropriate way to examine direct Tcf1 target genes in primary cells is to examine the association of Tcf1 binding with the changes that occur in Tcf1-bound genes after Tcf7 knockout. By analysing publicly available data on ILC progenitors we indeed did this. We revealed that Tcf1 bound to Dreg1 and that Dreg1 was not expressed when Tcf1 was knocked out in ILC progenitors. In addition we examined H3K27ac at the Dreg1 locus in the same ILC progenitors to demonstrate that Tcf1 appears to be important for decorating the Dreg1 gene with this histone modification. We believe that this analysis is sufficient to conclude that Tcf1 is required for the expression of Dreg1 in ILC progenitors.

(3) Human enhancer function is not experimentally confirmed.