You'll notice that the links here point to much older, smaller numbers. 1013 3667. These are ideas from thousands of entries ago. ideas I'm still grappling with today. That's what I really love about my link book. The way old and new ideas collide on paper. That's where the real creative sparks happen. And that's why I say the linkbook is an innovation engine. It doesn't just store ideas. It helps them grow, interact, and evolve.<br /> —Michael Herrick [2:45](https://youtu.be/30_v2FHJ9e4?si=HclrmkAMnd6LVca_&t=165

Michael Herrick noticing what others have seen in the past. He doesn't give the idea a new name like he's done with "Linkbook" for commonplacing or various other iterations.

a is agoraphobia but doesn't have to be in panic disorder c is seperation anxiety disorder and d is part of the ocd

wrong. use: .gdpr(gdprSpec -> gdprSpec .uouAndUserSupport(AppDefId, contextBuilder.secrets().appSecret()) ) before the change it is: .gdpr(gdprSpec -> gdprSpec .userSupportOnly(AppDefId, contextBuilder.secrets().appSecret()) )

need to add filterable to email before otherwise error

query wishlist

fix to change in assert function not in the call

add note to put in a new trnaslators file, because when in class mappers the transformers aren't yet build so refrence to the transformer is unknown.

needs to be (WishlistDomain domain) -> domain.size() > 5

change name to transformer

add public static final

dont need to erase the existing, and also Void.class is unnecessary

need to add the step to run the bazel generate model command (in scala exists)

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

Learn more at Review Commons

Reply to the reviewers

Manuscript number: RC-2025-03091

Corresponding author(s): Chia-Tsen, Tsai, Liuh-Yow Chen

1. General Statements [optional]

We thank the reviewers for their valuable time and constructive feedback on our study, which ultimately improved our manuscript. Herein, we provide a detailed response to each of the reviewers' comments, supported by new data that have been integrated into both the main text and the supplementary figures.

2. Point-by-point description of the revisions

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

Summary This manuscript builds upon the authors' prior findings that targeting COUP-TF2 to TRF1 induces ALT-associated phenotypes and G2-mediated synthesis in telomerase-immortalised BJT human fibroblasts. In this study, the authors show that telomere-coupled COUP-TF2 promotes H3K9me3 enrichment in these cells, and that this effect is blocked by TRIM28 depletion. Furthermore, TRIM28 depletion also suppresses the formation of ALT phenotypes in VA13 ALT cells. Given that TRIM28 has been implicated in regulating H3K9me3 deposition via SETDB1, and has been reported to co-purify with TR2 and TR4 (though not previously in the context of ALT telomeres), these findings add mechanistic depth to how heterochromatin regulators contribute to ALT activity. Overall, the manuscript's conclusions are generally supported by the presented data, but several aspects require clarification or additional experimental validation.

The authors report a modest reduction in telomeric H3K9me3 following COUP-TF2 and TR4 depletion in U-2 OS and VA13 cells (Figure 1B). To strengthen the claim that these orphan receptors specifically regulate H3K9me3, the authors should 1) Assess additional heterochromatic histone marks (e.g., H4K20me3) at telomeres, 2) Normalize telomeric signals to both parental histone levels and input, and 3) Evaluate whether global H3K9me3 levels also decrease upon receptor depletion

Response: We appreciate the reviewer's suggestion. To address the concern regarding specificity, we assessed H3K27me3 and H4K20me3 levels upon COUP-TF2/TR4 depletion and found no significant changes (Supplementary Fig. 1C). Furthermore, we reprocessed the telomeric ChIP data, normalizing to both input DNA and parental histone levels (Figure 1B). This refined analysis reinforces our original conclusion. Finally, Western blot analysis showed no significant changes in global H3 or H3K9me3 levels upon COUP-TF2/TR4 depletion (Figure 1A). Altogether, these results further support the specificity of COUP-TF2/TR4 for H3K9me3 at telomeres. We have revised the main text (page 3) and updated Figure 1A, 1B, and Supplementary Figure 1C for these changes.

Most experiments explore chromatin changes in telomerase-positive BJT fibroblasts (Figure 2, Figure 4D). It remains unclear whether similar manipulations in ALT cells yield consistent effects, which would give a broader context for ALT phenotype induction. Are ALT phenotypes similarly induced in ALT cells? Does altered chromatin status affect telomere length or telomerase recruitment/activity? Can these pathways drive ALT phenotypes in non-immortalised cells?

Response: We appreciate the reviewer's suggestion and have explored chromatin changes in telomerase-negative BJ and IMR90 primary fibroblasts (Supplementary Fig. 2C, D). Consistent to the result in BJ-telomerase cells, we found that VP64-TRF1 decreased telomeric H3, H4, and H3K9me3 levels, whereas KRAB-TRF1 increased these marks. Moreover, expression of either VP64-TRF1 or KRAB-TRF1 was sufficient to induce APB formation and ATDs in BJ and IMR90 cells. These results indicate that the chromatin changes at telomeres can drive ALT phenotypes in both primary and telomerase-immortalized fibroblast cells.

    Additionally, regarding whether chromatin alteration affects telomere length or telomere regulation, we have explored telomere length changes in BJT cells expressing vector, TRF1, KRAB-TRF1 or VP64-TRF1. The result of telomere restriction fragment (TRF) assay showed that the cells of all conditions maintained static telomere lengths through 30 days in culture (data shown below), suggesting that the chromatin alterations may not impact telomerase recruitment or activity. As this result is beyond the scope of current study, this data is only shown here in the rebuttal letter for a reference and is not included in the revised manuscript.

    Moreover, according to the reviewer's suggestion, we also carried out VP64-TRF1 or KRAB-TRF1 expression experiments in WI38-VA13/2RA cells that express high TERRA and have altered chromatin structures. Our data revealed that VP64-TRF1 suppresses telomere H3K9me3 and ALT activity, while KRAB-TRF1 increases both (Supplementary Figure 2E), suggesting an association of heterochromatin state with ALT activation in WI38-VA13/2RA cells.

    The observation that VP64-TRF1 reduces ALT activity in WI38-2RA/VA13 cells contrasts with findings in BJT cells. It is worth noting that studies from the Azzalian and Linger groups demonstrated that experimentally induced TERRA expression promotes ALT activity in ALT and non-ALT cells (PMID: 36122232, PMID: 40624280). Therefore, we propose that TERRA upregulation by VP64-TRF1 may contribute to the ALT induction observed in BJT cells (Supplementary Figure 2A, B), whereas the ability of VP64-TRF1 to suppress ALT activity in WI38-2RA/VA13 cells could be attributed to the reduction of telomere H3K9me3 and heterochromatin loss. Importantly, KRAB-TRF1 concurrently enhanced histone H3, H4, and H3K9me3 occupancy and ATL activity in both human fibroblasts and ALT cells. Altogether, these results support the notion that heterochromatin formation triggers ALT.

    We also examined TRIM28 recruitment to telomeres by telomere-ChIP and found that COUP-TF2LBD-TRF1 promotes TRIM28 telomere enrichment in BJ, IMR90 and U2OS, similar to BJT cells (Supplementary Fig. 5A-D).  Moreover, in ALT cell lines WI38-2RA/VA13, U2OS, and Saos-2, depletion of COUP-TF2 or TR4 reduced TRIM28 telomeric association (Figure 4A, B). Together, the data from human fibroblasts and ALT cells supports a role of orphan NRs in recruiting TRIM28 to ALT telomeres.

We acknowledge the reviewer's suggestions, which allow us to clarify and strengthen the conclusions. The corresponding data are presented in Figure 4A-B and Supplementary Figure 2B-D and 5E-F, and the main text has been modified on page 4-6 in the revised manuscript.

When referring to Figure 3G, the authors state that that telomeric H3K9me3 was abolished upon depleting TRIM28 from the U2OS and WI38-VA13/2RA cells. Abolished is a strong word for a 50% decrease, and this sentence should be revised. The reduction appears greater than that seen with COUP-TF2/TR4 depletion. Are the effects additive? If so, might TRIM28 act, at least in part, independently of COUP-TF2/TR4?

Response: We appreciate the reviewer's comments. We have revised the manuscript on page 5, replacing "abolished" with "significantly reduced" to better describe the effect of TRIM28 depletion on telomeric H3K9me3. To further investigate the interplay between TRIM28 and orphan NRs in regulating telomeric H3K9me3, we conducted single and combined knockdown experiments in U2OS and WI38-VA13/2RA cells, followed by telomere-ChIP analysis (Supplementary Figures 4D, E). Our results showed that single depletion of either orphan NRs or TRIM28 lead to a similar decrease in telomeric H3K9me3, and that combined knockdown do not result in any further reduction. These findings support an epistatic interaction between orphan NRs and TRIM28 in the regulation of telomeric H3K9me3. We have expanded on this interpretation in the main text (page 6) and included the relevant data in Supplementary Figures 4D, E.

VA13 cells consistently exhibit stronger effects than U-2 OS (e.g., Figures 1 and 3). This discrepancy could be linked to the high content of variant repeats in VA13 cells. The authors should assess whether variant repeat content underlies the differential response. Repeating key experiments in additional ALT lines with varied repeat compositions would be informative.

Response: We appreciate the reviewer's suggestion and have extended our analyses to two additional ALT osteosarcoma cell lines, SAOS-2 and G292. In both lines, depletion of orphan NRs resulted in a consistent decrease in telomeric H3K9me3 levels (Supplementary Figures 1A, B). We also examined the contribution of TRIM28 to telomeric H3K9me3 in these cells. siRNA-mediated depletion of TRIM28 in SAOS-2 and G292 cells similarly caused a significant reduction in telomeric H3K9me3 and ALT phenotypes (Supplementary Figure 4A-C). Together, these results from 4 ALT cell lines confirm that orphan NRs and TRIM28 promote telomeric H3K9me3 formation in ALT cells. We have modified the main text on page 3 and 5-6 for these results.

In line with the previous point, it would be useful to show whether TRIM28 telomeric enrichment is affected by COUP-TF2/TR4 depletion in U2OS cells (Figure 4C). To improve confidence in these findings, the authors should perform telomeric ChIP assays, especially with the COUP-TF2^LBDΔAF2-TRF1 mutant construct.

Response: Following the reviewer's suggestion, we performed telomere-ChIP assays to assess TRIM28 enrichment at telomeres upon expression of COUP-TF2LBD-TRF1 and its ΔAF2 mutant in U2OS cells. Consistent with our immunofluorescence results, telomere-ChIP revealed that COUP-TF2LBD-TRF1 expression promotes TRIM28 telomere enrichment, while the AF2 deletion mutant failed to recruit TRIM28 (Supplementary Figure 5D). We have modified the main text on page 6 for this result.

The immunoprecipitation experiments showing TRIM28 association with orphan receptors should include benzonase treatment to rule out DNA-mediated co-association (Figure 4F-G).

Response: We appreciate the reviewer's suggestion. To address the possibility of DNA-mediated interactions, we pre-incubated cell lysates with benzonase prior to Co-IP (Page 7). This treatment did not disrupt the association between TRIM28 and COUP-TF2 or TR4 in WI38-VA13/2RA and BJT cells (Supplementary Figures 5E-G), indicating a DNA-independent interaction. We have modified the main text on page 7 for this result.

The study would benefit from a direct assessment of whether COUP-TF2LBDΔAF2-TRF1 fails to induce ALT phenotypes in BJTfibroblasts.

Response: We thank the reviewer for this suggestion. As the role of the COUP-TF2 AF2 domain in ALT induction in BJT fibroblasts has recently been thoroughly investigated and published by our group (PMID: 38752489), we have directed the current study towards a more detailed mechanistic question. Specifically, we have carried out experiments to further demonstrate that COUP-TF2 recruits TRIM28 to telomeres via its AF2 domain in both human fibroblasts and ALT cells (Supplementary Figures 5A-D). On Page 6, we have modified the main text for these results and included a citation to our previous publication to provide the necessary background.

The experiments performed in Figure 5E-H lack a vector-only + siCtrl control.• In Figure 5E, the observation that APB formation is restored in siTRIM28 + Vector-treated cells is unexpected. The authors should address this finding and clarify whether this reflects biological noise or a compensatory effect.

Response: We thank the reviewer for this suggestion. We have repeated the experiments with a revised design, ensuring a consistent vector background across all groups (Vector + siCtrl, Vector + siTRIM28, TRIM28 WT + siTRIM28, and TRIM28 ΔRBCC + siTRIM28) (Figure 5E-H). This improved design confirms that expression of wild-type TRIM28, but not TRIM28 ΔRBCC, restores APB formation, ATDS, ssTeloC, and telomeric H3K9me3 levels in TRIM28-depleted cells. The updated dataset also resolves the previous unexpected increase in APB formation in the siTRIM28 + Vector condition, which is now excluded. We have modified the main text accordingly on page 8.

Reviewer #1 (Significance (Required)):

This work offers valuable mechanistic insight into how COUP-TF2 and TRIM28 coordinate to regulate heterochromatin deposition and ALT phenotype formation. It adds to the growing understanding of chromatin-mediated telomere regulation. What remains unclear is how important this interaction is for ALT maintenance, as H3K9me3 is only moderately altered upon TRIM28 depletion in ALT cells. Depletion of TRIM28 has been shown previously to induce APB formation and telomere elongation in U-2 OS ALT cells (Wang et al., 2021), the opposite to what the authors observed here in VA13 cells (Figure 5E-H). Clarifying whether these differences are variant repeat-dependent, or reflect intrinsic features of specific ALT cell lines, would substantially elevate the study's impact.

Response: We appreciate the reviewer's recognition of the significance of our work in elucidating the molecular basis of ALT regulation through COUP-TF2-TRIM28-mediated heterochromatin formation. In response to the reviewer's insightful comment regarding the importance of this interaction for ALT maintenance, we have expanded our study. We now include data from three additional primary human fibroblasts and a total of four ALT cancer cell lines (Figure 4, Supplementary Figure 4). These new data further strengthen the conclusion that TRIM28 promotes telomeric H3K9me3 and ALT-associated features. Furthermore, our rescue experiments support the model that the ALT-promoting function of TRIM28 in both fibroblasts and ALT cell lines is mediated through its physical interaction with COUP-TF2 (Supplementary Figure 5). We believe these results provide a solid foundation for demonstrating a cooperative role of COUP-TF2 and TRIM28 in ALT maintenance, and address the reviewer's concern regarding the generalizability of our findings.

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

Summary This manuscript investigates the role of orphan nuclear receptors (ORs), specifically COUP-TF2 and TR4, in promoting H3K9me3 enrichment at ALT telomeres via recruitment of TRIM28 (KAP1). The authors propose that the AF2 domain of COUP-TF2, located in its ligand-binding domain (LBD), is sufficient to recruit TRIM28 to telomeres. This, in turn, promotes heterochromatinization and induces hallmarks of the Alternative Lengthening of Telomeres (ALT) pathway, including APB formation and telomeric DNA synthesis outside of S-phase. This study addresses one important and unresolved question in the field: by what mechanism is the heterochromatic state established at ALT telomeres? Another timely question, not addressed here is: how is heterochromatin (specifically H3K9me3) functionally linked to ALT? The findings are potentially novel and mechanistically insightful. However, key elements of the study, particularly the central tethering experiments, require stronger quantification and clarity. Additional mechanistic tests and literature adjustments would also improve the manuscript.

Major Concerns

Central TRF1-COUP-TF2-LBD result lacks quantification and clarity: the tethering of COUP-TF2's LBD to telomeres via TRF1 is a core result of the paper. This experiment demonstrates that this domain is sufficient to induce weak H3K9me3 enrichment and ALT features (APBs and ATDS). However, the supporting ALT data are presented only in Supplementary Figures S1A and S1B, and are not quantified. These data should be quantified with appropriate statistics and moved to a main figure.

Response: The current study builds upon our recent publication (PMID: 38752489), which comprehensively analyzed ALT induction (APBs, ATDS, C-circles, T-SCEs) by orphan NR-TRF1 expression (COUP-TF1, COUP-TF2, TR2, and TR4; full-length and LBD) in various human fibroblast cell lines. To avoid potential duplicate publication concerns, particularly regarding APB and ATDS results for COUP-TF2LBD-TRF1 in BJT cells, we have put the data with revised quantification results in Supplementary Figure 1D-E. We will follow the reviewer's suggestion and move this data to the main figures if the editors agree.

Furthermore, the broader functional implication is not explored. Does this tethering induce a fully functional ALT pathway? For example, can telomerase knockout cells expressing TRF1-COUP-TF2-LBD maintain long-term proliferation? Such evidence would significantly strengthen the impact of the study.

Response: While COUP-TF2LBD-TRF1 expression rapidly induces key ALT phenotypes, we acknowledge that this alone is insufficient to directly promote telomere lengthening and long-term proliferation of primary fibroblasts, as discussed in Gaela et al., 2024 (PMID: 38752489). However, our ongoing, unpublished studies indicate that COUP-TF2LBD-TRF1 can drive immortalization of primary BJ fibroblasts expressing SV40LT by promoting ALT-mediated telomere elongation (Attached Figure A-C; additional data not shown). These findings suggest that COUP-TF2 may cooperate with additional genetic or epigenetic alterations to facilitate ALT development. We appreciate the reviewer's recognition of this critical aspect. As our immortalization study is still in progress and will be the subject of a separate manuscript, we hope the reviewer understands that the data shown in this letter will not be included in the revised manuscript.

Chromatin manipulation experiments lead to ambiguous conclusions: the authors propose that telomeric heterochromatin promotes ALT activity, but their own experiments (e.g., Figure 2) show that both heterochromatin-inducing (KRAB-TRF1) and euchromatin-inducing (VP64-TRF1) tethering can trigger ALT-like features. This makes it difficult to conclude that heterochromatin is specifically required.

To clarify:

-Did the authors express TRF1-VP64 in an ALT cell line? According to their model, this should suppress ALT activity.

-More broadly, do chromatin alterations per se (regardless of direction) trigger ALT features? Clarifying these points is important for interpretation.

Response: In response to the reviewer's suggestion, we expressed VP64-TRF1 and KRAB-TRF1 in WI38-2RA/VA13 cells to investigate telomere chromatin changes and ALT activity. Our data indeed revealed that VP64-TRF1 suppresses telomere H3K9me3 and ALT activity, while KRAB-TRF1 increases both (Supplementary Figure 2E), suggesting that heterochromatin triggers ALT activation.

The observation that VP64-TRF1 reduces ALT activity in WI38-2RA/VA13 cells contrasts with findings in BJT cells. Of note, studies from the Azzalian and Lingner groups demonstrated that experimentally induced TERRA expression promotes ALT activity in ALT and non-ALT cells (PMID: 36122232, PMID: 40624280). Therefore, we propose that TERRA upregulation may contribute to the ALT induction observed in BJT cells (Figure 2A, Supplementary Figure 2A, B). Given the high basal TERRA expression, expression of VP64-TRF1 and KRAB-TRF1 did not result in a consistent change in TERRA levels (Supplementary Figure 2F). Thus, the ability of VP64-TRF1 to suppress ALT activity in WI38-2RA/VA13 cells could be attributed to the reduction of telomere H3K9me3 and heterochromatin loss. Altogether, our results support the hypothesis that heterochromatin formation, rather than euchromatin triggers ALT.

We thank the reviewer's insightful comments, which have allowed us to resolve the ambiguity of our results and strengthen the notion that heterochromatin formation promotes ALT. We think that the heterochromatin features and high TERRA expression represent two independent, coexisting mechanisms within ALT cancer cells to guarantee ALT activation. We have modified the main text on page 4-5 accordingly.

TERRA downregulation contradicts current models: while TERRA upregulation is often observed in ALT cells and is thought to contribute to replication stress and recombination at telomeres, the authors show that TRF1-KAP1 expression induces ALT features while TERRA is downregulated. This observation is not addressed in the manuscript. The authors should at least discuss this discrepancy and propose whether this reflects a cell line-specific phenomenon or a decoupling between TERRA levels and ALT induction in this context.

Response: We thank the reviewer for the comments. As mentioned above (Major Concerns 2), heterochromatin formation and TERRA expression are two mechanisms that can independently promote ALT. Unlike ALT cell lines that have high TERRA levels, human fibroblasts BJ cells have low TERRA that does not induce ALT phenotypes. Thus, the effect of KRAB-TRF1 on ALT induction in BJ cells could be attributed to the heterochromatin formation, but not reduction of TERRA. We have modified the main text on page 5 to clarify the result.

Minor Comments

Introduction (p. 3): The authors cite Episkopou et al. as showing increased H3K9me3 at ALT telomeres. This is incorrect; that paper suggests the opposite. The first study to clearly demonstrate H3K9me3 enrichment at ALT telomeres is Cubiles et al., 2018 and should be cited instead. Results (p. 5, first paragraph): The manuscript should cite Déjardin and Kingston, 2009 as the first to report COUP-TF2 and TR4 localization at ALT telomeres. The studies by Conomos et al., 2012 and Gaela et al., 2024 build on this prior evidence. Please also include this citation in the bibliography.

Response: We appreciate the reviewer's careful reading and for pointing out these errors. The citation errors on pages 2 and 3 have now been corrected.Broader relevance of TRIM28-OR interaction: TRIM28 is a complex protein with roles in SUMOylation, heterochromatin formation, and transcriptional initiation/elongation regulation.

The authors should explore whether similar COUP-TF2/TRIM28 interactions occur at other genomic loci. Public ChIP-seq data for COUP-TF2, TR4, and TRIM28 could be mined to investigate whether these factors co-occupy regulatory regions elsewhere in the genome, and how this relates to gene expression states.

Response: We appreciate the reviewer's insightful suggestion regarding a potential genome-wild functional interaction between TRIM28 and COUP-TF2. To address this, we analyzed public ENCODE ChIP-seq data from K562 cells (TRIM28: ENCSR000BRW; COUP-TF2: ENCSR000BRS). This analysis revealed 3,326 co-binding sites for TRIM28 and COUP-TF2 (Attached Figure A). Interestingly, these co-binding sites were preferentially located within gene bodies (70.7%) and promoter regions (4.3%) (Attached Figures B-D), suggesting a potential cooperative role in gene regulation that aligns with our observation of physical interaction. While the finding is intriguing, a full exploration is beyond the scope of this manuscript, which focuses on ALT telomere regulation. We consider this is an important insight and have briefly noted it in the discussion (p. 9), although the corresponding analyses are not included in the revised manuscript.

Reviewer #2 (Significance (Required)):

This work contributes mechanistic insight into how heterochromatin is established at ALT telomeres-an important and timely question in telomere biology and cancer research. It offers a noncanonical recruitment mechanism for TRIM28, independent of KRAB-ZNFs, and highlights the functional role of orphan nuclear receptors in telomeric chromatin regulation. The study has potential implications for understanding ALT regulation and for identifying new intervention points in ALT-positive cancers. The work is conceptually interesting, but the conclusions are currently limited by insufficient quantification, some interpretative ambiguities, and a few overlooked references. Addressing the concerns listed above would significantly enhance the rigor and impact of the manuscript.

Response: We appreciate the reviewer's recognition of the significance of our work in elucidating the molecular basis of ALT regulation through COUP-TF2-TRIM28-mediated heterochromatin formation. We also thank the reviewer for the valuable feedback, which has significantly strengthened our manuscript.

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

Learn more at Review Commons

Referee #2

Evidence, reproducibility and clarity

Summary

This manuscript investigates the role of orphan nuclear receptors (ORs), specifically COUP-TF2 and TR4, in promoting H3K9me3 enrichment at ALT telomeres via recruitment of TRIM28 (KAP1). The authors propose that the AF2 domain of COUP-TF2, located in its ligand-binding domain (LBD), is sufficient to recruit TRIM28 to telomeres. This, in turn, promotes heterochromatinization and induces hallmarks of the Alternative Lengthening of Telomeres (ALT) pathway, including APB formation and telomeric DNA synthesis outside of S-phase. This study addresses one important and unresolved question in the field: by what mechanism is the heterochromatic state established at ALT telomeres? Another timely question, not addressed here is: how is heterochromatin (specifically H3K9me3) functionally linked to ALT? The findings are potentially novel and mechanistically insightful. However, key elements of the study, particularly the central tethering experiments, require stronger quantification and clarity. Additional mechanistic tests and literature adjustments would also improve the manuscript.

Major Concerns

1. Central TRF1-COUP-TF2-LBD result lacks quantification and clarity: the tethering of COUP-TF2's LBD to telomeres via TRF1 is a core result of the paper. This experiment demonstrates that this domain is sufficient to induce weak H3K9me3 enrichment and ALT features (APBs and ATDS). However, the supporting ALT data are presented only in Supplementary Figures S1A and S1B, and are not quantified. These data should be quantified with appropriate statistics and moved to a main figure. Furthermore, the broader functional implication is not explored. Does this tethering induce a fully functional ALT pathway? For example, can telomerase knockout cells expressing TRF1-COUP-TF2-LBD maintain long-term proliferation? Such evidence would significantly strengthen the impact of the study.
2. Chromatin manipulation experiments lead to ambiguous conclusions: the authors propose that telomeric heterochromatin promotes ALT activity, but their own experiments (e.g., Figure 2) show that both heterochromatin-inducing (KRAB-TRF1) and euchromatin-inducing (VP64-TRF1) tethering can trigger ALT-like features. This makes it difficult to conclude that heterochromatin is specifically required. To clarify:
3. Did the authors express TRF1-VP64 in an ALT cell line? According to their model, this should suppress ALT activity.
4. More broadly, do chromatin alterations per se (regardless of direction) trigger ALT features? Clarifying these points is important for interpretation.
5. TERRA downregulation contradicts current models: while TERRA upregulation is often observed in ALT cells and is thought to contribute to replication stress and recombination at telomeres, the authors show that TRF1-KAP1 expression induces ALT features while TERRA is downregulated. This observation is not addressed in the manuscript. The authors should at least discuss this discrepancy and propose whether this reflects a cell line-specific phenomenon or a decoupling between TERRA levels and ALT induction in this context.

Minor Comments

Introduction (p. 3): The authors cite Episkopou et al. as showing increased H3K9me3 at ALT telomeres. This is incorrect; that paper suggests the opposite. The first study to clearly demonstrate H3K9me3 enrichment at ALT telomeres is Cubiles et al., 2018 and should be cited instead. Results (p. 5, first paragraph): The manuscript should cite Déjardin and Kingston, 2009 as the first to report COUP-TF2 and TR4 localization at ALT telomeres. The studies by Conomos et al., 2012 and Gaela et al., 2024 build on this prior evidence. Please also include this citation in the bibliography. Broader relevance of TRIM28-OR interaction: TRIM28 is a complex protein with roles in SUMOylation, heterochromatin formation, and transcriptional initiation/elongation regulation. The authors should explore whether similar COUP-TF2/TRIM28 interactions occur at other genomic loci. Public ChIP-seq data for COUP-TF2, TR4, and TRIM28 could be mined to investigate whether these factors co-occupy regulatory regions elsewhere in the genome, and how this relates to gene expression states.

Significance

This work contributes mechanistic insight into how heterochromatin is established at ALT telomeres-an important and timely question in telomere biology and cancer research. It offers a noncanonical recruitment mechanism for TRIM28, independent of KRAB-ZNFs, and highlights the functional role of orphan nuclear receptors in telomeric chromatin regulation. The study has potential implications for understanding ALT regulation and for identifying new intervention points in ALT-positive cancers.

The work is conceptually interesting, but the conclusions are currently limited by insufficient quantification, some interpretative ambiguities, and a few overlooked references. Addressing the concerns listed above would significantly enhance the rigor and impact of the manuscript.

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

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

Summary

This manuscript builds upon the authors' prior findings that targeting COUP-TF2 to TRF1 induces ALT-associated phenotypes and G2-mediated synthesis in telomerase-immortalised BJT human fibroblasts. In this study, the authors show that telomere-coupled COUP-TF2 promotes H3K9me3 enrichment in these cells, and that this effect is blocked by TRIM28 depletion. Furthermore, TRIM28 depletion also suppresses the formation of ALT phenotypes in VA13 ALT cells. Given that TRIM28 has been implicated in regulating H3K9me3 deposition via SETDB1, and has been reported to co-purify with TR2 and TR4 (though not previously in the context of ALT telomeres), these findings add mechanistic depth to how heterochromatin regulators contribute to ALT activity. Overall, the manuscript's conclusions are generally supported by the presented data, but several aspects require clarification or additional experimental validation.

• The authors report a modest reduction in telomeric H3K9me3 following COUP-TF2 and TR4 depletion in U-2 OS and VA13 cells (Figure 1B). To strengthen the claim that these orphan receptors specifically regulate H3K9me3, the authors should 1) Assess additional heterochromatic histone marks (e.g., H4K20me3) at telomeres, 2) Normalize telomeric signals to both parental histone levels and input, and 3) Evaluate whether global H3K9me3 levels also decrease upon receptor depletion
• Most experiments explore chromatin changes in telomerase-positive BJT fibroblasts (Figure 2, Figure 4D). It remains unclear whether similar manipulations in ALT cells yield consistent effects, which would give a broader context for ALT phenotype induction. Are ALT phenotypes similarly induced in ALT cells? Does altered chromatin status affect telomere length or telomerase recruitment/activity? Can these pathways drive ALT phenotypes in non-immortalised cells?
• When referring to Figure 3G, the authors state that that telomeric H3K9me3 was abolished upon depleting TRIM28 from the U2OS and WI38-VA13/2RA cells. Abolished is a strong word for a 50% decrease, and this sentence should be revised. The reduction appears greater than that seen with COUP-TF2/TR4 depletion. Are the effects additive? If so, might TRIM28 act, at least in part, independently of COUP-TF2/TR4?
• VA13 cells consistently exhibit stronger effects than U-2 OS (e.g., Figures 1 and 3). This discrepancy could be linked to the high content of variant repeats in VA13 cells. The authors should assess whether variant repeat content underlies the differential response. Repeating key experiments in additional ALT lines with varied repeat compositions would be informative.
• In line with the previous point, it would be useful to show whether TRIM28 telomeric enrichment is affected by COUP-TF2/TR4 depletion in U-2 OS cells (Figure 4C). To improve confidence in these findings, the authors should perform telomeric ChIP assays, especially with the COUP-TF2^LBDΔAF2-TRF1 mutant construct.
• The immunoprecipitation experiments showing TRIM28 association with orphan receptors should include benzonase treatment to rule out DNA-mediated co-association (Figure 4F-G).
• The study would benefit from a direct assessment of whether COUP-TF2LBDΔAF2-TRF1 fails to induce ALT phenotypes in BJT fibroblasts.
• The experiments performed in Figure 5E-H lack a vector-only + siCtrl control.
• In Figure 5E, the observation that APB formation is restored in siTRIM28 + Vector-treated cells is unexpected. The authors should address this finding and clarify whether this reflects biological noise or a compensatory effect.

Significance

This work offers valuable mechanistic insight into how COUP-TF2 and TRIM28 coordinate to regulate heterochromatin deposition and ALT phenotype formation. It adds to the growing understanding of chromatin-mediated telomere regulation. What remains unclear is how important this interaction is for ALT maintenance, as H3K9me3 is only moderately altered upon TRIM28 depletion in ALT cells. Depletion of TRIM28 has been shown previously to induce APB formation and telomere elongation in U-2 OS ALT cells (Wang et al., 2021), the opposite to what the authors observed here in VA13 cells (Figure 5E-H). Clarifying whether these differences are variant repeat-dependent, or reflect intrinsic features of specific ALT cell lines, would substantially elevate the study's impact.

DOI: 10.2139/ssrn.5889036

Resource: (Dominique Alfandari; UMass amherst Cat# Alfandari-3, RRID:AB_2687673)

Curator: @scibot

SciCrunch record: RRID:AB_2687673

What is this?

DOI: 10.1016/j.isci.2025.114416

Resource: (MBL International Cat# M177-3, RRID:AB_10697039)

Curator: @scibot

SciCrunch record: RRID:AB_10697039

What is this?

Keep these key features in mind when it comes to writing. Make a checklist and check your boxes as you go

addは追加するだけどremoveはアンインストールするなので表現が揃っていない。 追加/削除 or インストール/アンインストールに揃えて欲しい

nits: 既存環境を汚さない？というのが気になったので、以下のようなことがわかる文があるとより親切かなと思いました。

「uvxは現在のプロジェクトから分離された一時的な仮想環境にインストールされる」

この辺りは同じファイルを更新している？ので、ファイル名もcaptionに含めると、同いファイルを変更していることが伝わりやすいかなと思いました。 （これ以前のコードも同様に）

作成されるファイルはpythonファイルですが、「sample-script.py」としなくてよいです？ ドキュメントでは、.pyをつけているようでした。 https://docs.astral.sh/uv/guides/scripts/#running-a-script-with-dependencies

が

の方がいいかなー

uvx とバッククォートで囲みたい。以下同様

外部パッケージによって提供されるツール

とか?

.pyはつけない派?

PEP 723で提案されたけどPyPA関連はのちのち変わっていくので、仕様としてはこちらを参照するのが適切です。

(PEPドキュメントの先頭にもそう書いてある)

https://packaging.python.org/en/latest/specifications/inline-script-metadata/#inline-script-metadata

外部

この手順よりは、新規にディレクトリ作成、uv init実行した後に、pyproject.toml, uv.lockだけコピーしてuv sync実行するみたいな手順がいいのではと思った。

実際のユースケースに合わせる

uv pip installした場合はuv addと違ってpyproject.tomlとかuv.lockが更新されないということを書いた方がいいかなぁ

名を

pyproject.tomlの設定を見て、今のuv.lockファイルに存在するパッケージのままか新しいバージョンが存在するかを調べる。

みたいな意味合いだと思います。

あと、これってどっちかというと hoge>4.0.0 とかかいていて、hogeの最新バージョンがでているか(pip list -O)みたいな使い方がメインかなと思ったんですが、そうではない?

pyproject.tomlを書き換えたのはここで例として示しているために必要なだけど、本来はpyproject.toml書き換えたらuv syncの方を使うかなと思ったので(uv 素人なのではずしてたらすいません)

uv.lockファイル

トルでもいいかな(あってもいいけど)

uv lock

uv.lockファイル

uv.lockファイル

ハッシュの例はハッシュ値を省略しつつ1行くらいは載せて欲しいかな

てインストールします。 pyproject.tomlにはdevというセクション(?)が追加され~~

みたいな文章がほしい。

このときもpyproject.tomlの中は書き換えられるよ(ですよね?)とか説明があっていいと思う

依存関係というよりも、「さきほど追加したパッケージが」とかいう説明でよいのではないか

pipコマンドと同様に、とかあってもいいかも

PyPIからダウンロードしてインストールされていることとか触れて欲しい。

依存関係を管理するってちょっとイメージしにくいかなと思いました。

プロジェクトで使用するライブラリを管理する、とか?

コマンド実行後

一文が長いので「利用できます。」で一度切ってもいいかも

なにももう一つの方法なのか

仮想環境下のPythonを実行するもう一つの~

とか?

に

の方がいいかな

さっきとは異なるディレクトリに作ったと言うことですかね。手順で mkdir uv-example がないのでちょっと混乱するかも

では

◯◯な場合に仮想環境を再作成します。

みたいな用途にも触れてくれるとうれしいなと思った。

長いし、トルでも意味は通じるかなと

確認するメッセージが表示されます。

コマンドを

で

でよさそう

仮想環境を有効化した場合は、プロンプトにプロジェクト(ディレクトリ名)が表示されることにも言及して欲しい。

venvは作成したvenv名がプロンプトで表示されていたので

ここで段落を変えてもいいかな

typo: いません

手順にlsを入れるといいかなと。 venvの方にはあった

バージョン0.9.17が出たので更新します。

High-Level Summary

The kidneys are bean-shaped organs protected by three outer layers and organized internally into the cortex, medulla, and renal pelvis. Nephrons in the cortex filter blood supplied by a highly branched vascular network that enters and exits through the renal hilum. Urine formed by nephrons flows through the renal pyramids into calyces, then the renal pelvis, and finally the ureter. Each kidney contains over one million nephrons, which are either cortical or juxtamedullary, depending on their position relative to the medulla.

Study Notes: Kidney Structure 1. External Kidney Structure

The kidney is surrounded by three protective layers (outer → inner):

• Renal fascia Tough connective tissue Anchors kidney to surrounding structures

• Perirenal fat capsule Cushions and stabilizes the kidney

• Renal capsule Thin, tough layer directly covering kidney surface

• Internal Kidney Regions

The kidney has three main internal regions:

Renal Cortex (outer region) Granular appearance Contains nephrons (functional units of the kidney) Site of blood filtration

Renal Medulla (middle region) Made of renal pyramids (cone-shaped tissue masses) Each kidney has ~8 pyramids Renal columns lie between pyramids and carry blood , vessels Pyramid tips = renal papillae, which point toward the , pelvis

Renal Pelvis (inner region) Located at the hilum Funnel-shaped urine collection area Drains urine into the ureter

1. Hilum of the Kidney

2. Concave region of the kidney

3. Entry/exit point for: Renal arteries Renal veins Nerves

4. Exit point for the ureter

5. Urine Flow Pathway

6. Minor calyces → Major calyces → Renal pelvis → Ureter → Urinary bladder

7. Renal Lobes

8. A renal lobe = one renal pyramid + surrounding cortical tissue

9. Functional subdivision of the kidney

Blood Supply of the Kidney (In Order) 1. Aorta 2. Renal arteries 3. Segmental arteries 4. Interlobar arteries (run through renal columns) 5. Arcuate arteries (arch at cortex–medulla boundary) 6. Cortical radiate arteries 7. Afferent arterioles 8. Glomerular capillaries (nephrons)

Venous return: Veins follow the same path in reverse. Same names as arteries except no segmental veins. Drain into the inferior vena cava.

Nephrons (Functional Units) Each kidney contains >1 million nephrons Located mainly in the renal cortex

Types of Nephrons Cortical nephrons (≈85%) Located deep in cortex Short loops of Henle Juxtamedullary nephrons Located near cortex–medulla boundary Long loops of Henle Important for urine concentration

Parts of a Nephron Renal corpuscle Renal tubule Associated capillary network

1-Apperance: how does the person look like wearing and pysical 2-mood: how emotions show itself 3- Cognition: aware of the time and location 4-insight and judgement: aware of the illness itself 5- intellectual functioning: the expression of the toughts are not distrupted and has a flow

詩人

前往

驅逐

羊群 牛群

君主立憲制

酋長

族人

叛亂

條約法

胡格諾派

權利法案

Dus Romanisatie van de kerk: 1. Taal van Grieks naar Latijn 2. Centralisatie van het Christendom 3. Keizers gingen concilies leggen 4. Definiëren van ketterij 5. Bevorderen van bekering -> uiteindelijk dwang 6. Er ontstond een hiërarchische structuur: bisschoppen kregen publieke functies zoals praetoren 7. Er werden meer beslechtingsregels ingevoerd waardoor het meer conflictenrecht werd

Volcanic Ash?

Gas exchange relies on Fick’s Law of Diffusion, which states that the rate of diffusion is directly proportional to the surface area of the membrane. Oxygen cannot enter the bloodstream fast enough to meet the body's demands, and carbon dioxide cannot exit efficiently. The alveolar walls contain the dense network of capillaries where the actual exchange takes place. When the walls are destroyed, the capillaries are destroyed with them. This creates a "dead space" effect where there is air in the lungs, but insufficient blood flow to pick up the oxygen. Healthy alveoli are elastic—they snap back to push air out during exhalation. Smoking destroys the elastin fibers that provide this recoil. When the smoker inhales fresh air, it mixes with this stale, trapped air.10 This lowers the partial pressure of oxygen (PO2) within the alveoli, reducing the driving force that pushes oxygen into the blood.

DOI: 10.1016/j.ccell.2025.11.007

Resource: (MABTECH Cat# 3321-3-1000, RRID:AB_907278)

Curator: @scibot

SciCrunch record: RRID:AB_907278

What is this?

Keep these tips in mind when it comes to writing about your own reflections.

Try to keep it short and to the point so it isn't drawn out and wordy.

Thank you for sharing this dataset and model (as well as the SCVI model). In terms of training cost versus computational budget, how would the smaller training subsets factor in to efficiency for the smaller models? It's interesting to consider training compute normalized by fraction of the data on which the model was trained. Is it possible that training the 3B model on only a subset of the dataset would not hurt performance and therefore improve training efficiency metrics? I appreciate this deep analysis of the training process.

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

Learn more at Review Commons

Reply to the reviewers

1. General Statements

In this study, we mechanistically define a new molecular interaction linking two of the cell's major morphological regulatory pathways-the Rho GTPase and Hippo signaling networks. These two major signaling pathways are both required for life across huge swaths of the tree of life. They are required for the dynamic organization and reorganization of proteins, lipids, and genetic material that occurs in essential cellular processes such as division, motility and differentiation. For decades these pathways have been almost exclusively studied independently, however, they are known to act in concert in cancer to drive cytoskeletal remodeling and morphological changes that promote proliferation and metastasis. However, mechanistic insight into how they are coordinated is lacking.

Our data reveal a mechanistic model where coordination is mediated by the RhoA GTPase-activating protein ARHGAP18, which forms molecular interactions with both the tumor suppressor Merlin (NF2) and the transcriptional co-regulator YAP (YAP1). Using a combination of state-of-the-art super-resolution microscopy (STORM, SORA-confocal) in cultured human cells, biochemical pulldown assays with purified proteins, and analyses of tissue-derived samples, we characterize ARHGAP18's function from the molecular to the tissue level in both native and cancer model systems.

Together, these findings establish a previously unrecognized molecular connection between the RhoA and Hippo pathways and culminate in a working model that integrates our current results with prior work from our group and decades of prior studies. This model provides a new conceptual framework for understanding how RhoA and Hippo signaling are coordinated to regulate cell morphology and tumor progression in human cells.

In this substantially revised manuscript, we have addressed all comments from the expert reviewers described point-by-point below. A shared major comment from the reviewers was the request for direct evidence of the proposed mechanistic model. To address these constructive comments, we've added new experiments, new quantification, new text, new control data, and have added two expert authors, adding super-resolution mouse tissue imaging data for the endogenous study of ARHGAP18 in its native condition. We believe that these additions greatly enhance the manuscript and collectively address the overall message from the reviewer's collective comments.

2. Point-by-point description of the revisions

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

This manuscript describes a dual mechanism by which ARHGAP18 regulates the actin cytoskeleton. The authors propose that in addition to the known role for ARHGAP18 in regulating Rho GTPases, it also affects the cytoskeleton through regulation of the Hippo pathway transcriptional regulator YAP. ARHGAP18 knockout Jeg3 cells are were generated and show a clear loss of basal stress fiber like F-actin bundles. The authors further characterize the effects of ARHGAP18 knockout and overexpression. It is also discovered that ARHGAP18 binds to the Hippo pathway regulator Merlin and to YAP. Ultimately it is concluded that ARHGAP18 regulates the F-actin cytoskeleton through dual regulation of RHO GTPases and of YAP. While the phenotype of the ARHGAP18 knockout and the association of ARHGAP18 with Merlin and YAP is interesting, I found the authors conclusion that these phenotypes are due to ARHGAP18 regulation of both RHO and YAP to be based on largely correlative evidence and sometimes lacking in controls or tests for significance. In addition the authors often make overly strong conclusions based on the experimental evidence. In some instances, the rationale for how the experimental results support the conclusion is insufficiently articulated, making evaluation challenging. In general although the authors have some interesting observations, more definitive experiments with proper controls and statistical tests for significance and reproducibility are needed to justify their overall conclusions.

• *

*We appreciate the reviewers' constructive comments and have added substantial new data and quantifications to address their concerns. We have focused these new data on directly testing the proposed mechanisms, adding controls, and performing quantitative analysis with statistical testing. Additionally, we have edited our language to make our rationale clearer and to present our conclusions as a more moderate assessment of our experimental results. Below we respond to the specific comments made by the reviewer, followed by a list of additional editorial changes we've made based on the reviewer's overarching comments on clarity and rationale. *

Specific Comments

1) The authors make a big point about the effects of ARHGAP18 on myosin light chain phosphorylation. However, this result is not quantified and tested for statistical significance and reproducibility.

*We thank the reviewer for their comments on our western blotting quantification, which in the original submission version had quantification of RhoA downstream signaling of pCofilin/ Cofilin and pLIMK/ LIMK. We had withheld the pMLC and MLC quantification as the result was previously published with quantification, reproducibility, and statistical significance by our group in our prior manuscript on ARHGAP18 published in Elife in 2024 (Fig. 4E of *

https://doi.org/10.7554/eLife.83526 ). However, these prior results lacked the new overexpression data. We recognize the need to add these data to this manuscript as requested by the reviewer.

• *

*To address the reviewer's comment, we have added quantification of pMLC/MLC (Fig. 1F) *

2) Along similar lines in Figure 2C they state that overexpression of ARHGAP18 causes cells to invade over the top of their neighbors. This might be true and interesting, but only a single cell is shown and there is no quantification or controls for simply overexpressing something in that cell. The authors also conclude from this image that the overexpression phenotype is independent of its GAP activity on Rho. It is not clear how this conclusion is made based on the data. It would seem like a more definitive experiment would be to see if a similar phenotype was induced by an ARHGAP18 mutant deficient in GAP activity.

Based on the reviewer's comment, we recognize the qualitative statements made in Figure 2C (now Figure 3) should've been made more quantitative. We have added the control of Jeg 3 WT cells expressed with empty vector flag to show that WT cells do not invade over the top of each other (Fig. 3F). Additionally, we have added the quantification found in Fig. 3E, which shows the % invasive/ non-invasive cells between WT and ARHGAP18 overexpression cells. We have clarified our conclusions to make clear that these data do not directly test if the invasive phenotype derives from a Rho-independent mechanism. The text now states the following conclusion alongside others, which can be seen in our tracked changes:

• *

"These data support the conclusion that ARHGAP18 acts to regulate basal and junctional actin. However, it was not clear whether this activity occurred through a Rho-independent or a Rho-dependent mechanism."

• *

We have added new data of cells expressing an ARHGAP18 mutant deficient in GAP activity, which is explained in detail in the following response below.

3) In Figure 3 the authors compare gene expression profiles of ARHGAP18 knockout cells to wild-type cells. They see lots of differences in focal adhesion and cytoskeletal proteins and conclude that this supports their conclusion that ARHGAP18 is not just acting through RHO. The rationale for this in not clear. In addition, they observe changes in expression profiles consistent with changes in YAP activity. They conclude that the effects are direct. This very well might be true. However RHO is a potent regulator of YAP activity and the results seem quite consistent with ARHGAP18 acting through RHO to affect YAP.

• *

We thank the reviewer for their comment and believe the revised manuscript now presents direct evidence to support the conclusions made through the editing text and the incorporation of new data.

• *

First, the reviewer highlighted that we were not clear in our rationale and explanation of the conclusions made from our RNAseq data in the new Figure 4 (Previously Figure 3). We agree with the reviewer that the RNAseq data alone is not sufficient rationale for the conclusion that ARHGAP18 is acting through YAP directly. In the revised manuscript, the conclusion is now made based on the combination of our multi-faceted investigation of the relationship between ARHGAP18 and YAP (most importantly, new Figure 5). It's important for us to argue that our RNAseq analysis is much more robust and specific than simply reporting a descriptive assay seeing lots of differences in cytoskeletal proteins. We recruited an outside RNAseq expert collaborator; Dr. Yongho Bae, to perform state-of-the-art IPA analysis and a grueling manual curation of the top hit genes to identify the predominant signaling pathways linking the loss of ARHGAP18 to known YAP translational products. We've provided a supplemental table listing each citation supporting the identified YAP pathway associations from this manual curation. We also have added a new discussion paragraph on RNAseq data to clarify our specific RNAseq data results and analysis. In the revised manuscript, we have moderated our language in the results text regarding the RNAseq data to reflect the reviewer's suggestion:

• *

"Our RNAseq data alone could not independently confirm if the alterations to transcriptional signaling and expression of actin cytoskeleton proteins were through a Rho-dependent or Rho-independent mechanism."

• *

• *

Second, in this comment and the above, the reviewer highlights the need for a new experiment to directly test the Rho Independent effects of ARHGAP18, which we now provide in the new Figure 5. In this new data, we've applied an experimental design suggested by reviewer 2 regarding the same concern. In short, we've produced and expressed a point mutant variant ARHGAP18(R365A), which abolishes the Rho GAP activity while maintaining the remainder of the protein intact. This construct allows us to directly test the effects of ARHGAP18 independent from its RhoA GAP activity. We find that the GAP-deficient ARHGAP18 is able to fully rescue basal focal adhesions, indicating that the basal actin phenotype is at least in part regulated through a Rho-independent mechanism.

• *

• *

*We believe the revised manuscript, when taken in totality, provides the definitive proof requested by the reviewer. Specifically, the combination of Figure 5, where we show new data using the ARHGAP18(R365A) variant, and the result that ARHGAP18 forms a stable complex with YAP (Fig. 6G) or Merlin (Fig.6A), is supportive of direct Rho-independent molecular interactions between YAP, Merlin, and ARHGAP18. *

4) In Figure 4A showing Merlin binding to ARHGAP18 there is no control for the amount of Merlin sticking to the column as was done in Figure 4F for binding experiments with YAP. This makes it difficult to determine the significance of the observed binding.

We have performed the requested control experiment and added the results to Figure 6A.

5) The images in Figure 4C showing YAP being maintained in the nucleus more in ARHGAP18 knockout cells compared to wild-type. However the images only show a few cells and YAP localization can be highly variable depending on where you look in a field. Images with more cells and some sort of quantification would bolster this result.

We have provided quantification (Figure 6D) of what was originally Figure 4C (now Figure 6C).

Reviewer #1 (Significance (Required)):

While the phenotype of the ARHGAP18 knockout and the association of ARHGAP18 with Merlin and YAP is interesting, I found the authors conclusion that these phenotypes are due to ARHGAP18 regulation of both RHO and YAP to be based on largely correlative evidence and sometimes lacking in controls or tests for significance. In addition the authors often make overly strong conclusions based on the experimental evidence. In some instances, the rationale for how the experimental results support the conclusion is insufficiently articulated, making evaluation challenging. In general although the authors have some interesting observations, more definitive experiments with proper controls and statistical tests for significance and reproducibility are needed to justify their overall conclusions.

In the above comments, we detail the specific definitive experiments, proper controls, and statistical tests for significance, requested by the reviewer, which we believe greatly strengthen our manuscript.

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

This manuscript investigates the Rho effector, ARHGAP18 in Jegs cells, a trophoblastic cell line. It presents a number of new pieces of data, which increase our understanding of the importance of this GAP on cell function and explains at a molecular level previous results of other workers in the field. ARHGAP18 was originally given the name "conundrum' and continues to stand apart from the majority of other GAP proteins and their functions. Hence the data here is significant and of high standard.

The data is clear, and the images are of high quality and extremely impressive in their resolution. It is significant and adds a further layer to our understanding of the regulation of cell migration, particularly in the formation and resolution of microvilli.

• *

We appreciate the reviewer's comments and supportive insights.

The data is based on the use of the cell line Jeg3. Even the authors previous publication in eLife is based only on this cell line. They need to show the conclusions are general and not specific to this line of cells. As an extension of this, is the ARHGAP18 function shown here only in transformed cells? Does the same mechanisms operate in normal cells, which respond to activation to proliferate or migrate?

• *

• We respectfully point out that the critical experiments of the prior eLife publication were validated in DLD-1 colorectal cells and not Jeg-3 cells alone (Figure 1-figure supplement 2). Our newly independent lab, established just over a year ago, is unable to perform a full expansion of the manuscript using untransformed cells, however, we agree with the reviewer's perspective and wish to address the comment to the best of our current capability. To answer the reviewers' suggestions, we have recruited Dr. Christine Schaner Tooley, an expert in mouse model system studies. In the revised manuscript, we've added new Super-Resolution SORA confocal images of endogenous ARHGAP18's localization in the intact intestinal villi tissue, and apical junctions of WT mice (Fig.1A-C). These data indicate that endogenous ARHGAP18 is enriched (but not exclusively localized) at the apical plasma membranes of normal WT epithelial cells. This localization, where both Merlin and Ezrin are present at apical membrane/ junctions under normal conditions, is a major component of the working model proposed in Fig. 7. These data also indicate that ARHGAP18 is capable of entering the nucleus in WT cells, another critical aspect of our proposed model. Collectively, our DLD-1 studies published previously and or new studies using WT mice tissue samples support the conclusion that at least some of ARHGAP18's functions described in this manuscript are not limited to Jeg3 cells.*

In endothelial cells, Lovelace et al 2017 showed localization to microtubules and that depletion of ARHGAP18 resulted in microtubule instability. The authors may like to comment on the differences. Is this a cell type difference or RhoA versus RhoC difference?

• *

In our previous publication (Lombardo Elife), we validated the finding that ARHGAP18 forms a complex with microtubules, as we detected tubulin in the ARHGAP18 pulldown experiment (Figure 1- Source Data). However, our data indicate that in Jeg3 cells ARHGAP18 does not localize to the same microtubule associated spheres observed in the Lovelace publication. We now comment on the shared conclusions and differences between this manuscript and the Lovelace et al 2017 in the discussion section.

• *

"In endothelial cells, ARHGAP18 has been reported to localize microtubules and plays a role in maintaining proper microtubule stability (Lovelace et al., 2017). In our epithelial cell culture models and WT mouse intestine, we have been unable to detect ARHGAP18 at microtubules suggesting ARHGAP18 may have additional functions is various cell types."

On pages 7,9 they conclude that MLC and basal and junctional actin are regulated through a GAP independent mechanism. The best way to show this is with overexpression of a GAP mutant.

We appreciate the reviewer's insight and have produced and expressed a GAP mutant, ARHGAP18(R365A), in our cells, directly testing our conclusion that ARHGAP18 has a GAP-independent function. These data are now presented in revised Figure 5 and explained further in response to reviewer #1.

There is a huge amount of data presented in Figure 3, but their 2 genes which they focus on, LOP1 and CORO1A, are discussed but no actual data presented in support.

We now validate the CORO1A by qPCR in Figure 4J.

• *

Reviewer #2 (Significance (Required)):

The data is significant and adds a further layer to our understanding of the regulation of cell migration, particularly in the formation and resolution of microvilli. This manuscript will be of significance to an basic science audience in the field of RhoGTPases and cell migration.

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

The study by Murray et al explores the effects of ARHGAP18 on the actin cytoskeleton, Rho effector kinases, non-muscle myosin, and transcription. Using super resolution microscopy, they show that in ARHGAP18 KO cells there is a mixed and unexpected cytoskeleton phenotype where myosin phosphorylation appears to be increased, but actin is disorganised with reduced stress fibres, diminished focal adhesions and augmented invasiveness. They conclude that the underlying mechanisms are likely independent from RhoA. Next, they perform RNAseq using the KO cells and identify an array of dysregulated genes, including those that play crucial roles in microvilli (related to previously published findings). Analysis of the data identify gene expression changes that are relevant for altered focal adhesion (integrins). Further analysis reveals that a large cohort of the dysregulated genes are YAP targets. They then show that in ARHGAP18 KO cells YAP nuclear localization, as detected by immunostaining, is augmented; and demonstrate that immobilized ARHGAP18 protein can bind the Hippo regulator merlin as well as YAP itself.

Major comments:

1, The premise of the study (that ARHGAP18 is a RhoA effector or may acts independently of RhoA) remains not proven.

We have added new evidence of direct RhoA independent activity for ARHGAP18 described in the above comments. Specifically, we've added data using a RhoA-GAP dead variant of ARHGAP18 in Figure 5, which we believe addresses this comment.

• *

At several places (including in the title) the authors refer to ARHGAP18 as a Rho effector, which would suggest that it is downstream form Rho, but the basis for this is not clear. In fact, their own previous study suggested that ARHGAP is a RhoA regulator, rather than an effector. In general, the connection of the described effects to RhoA remains unclear, and not addressed in this study. The authors seem to go back and forth in their conclusions regarding the connection between ARHGAP18 and RhoA. For example, the first section of results is finished by stating (line 194): "These data support the conclusion that ARHGAP18 acts to regulate basal and junctional actin through Rho-independent mechanism". But the next section starts by stating (line 198): "We hypothesized that the invasive and cytoskeletal phenotypes observed at the basal surface of cells devoid of ARHGAP18 may be a result of changes in regulation at the transcriptional level either directly through RhoA signaling or through an additional mechanism specific to ARHGAP18". The paper would be strengthened by adding data that show whether the effects are indeed downstream, from RhoA or RhoA independent. If there is no sufficient demonstration that ARHGAP18 is downstream of RhoA and is an effector, this needs to be stated explicitly, and the wording should be changed.

*We now provide new data in Figure 5, which directly tests the RhoA independent functions of ARHGAP18 as recommended by the reviewer. Our understanding of the term effector is 'a molecule that activates, controls, or inactivates a process or action.' Based on this understanding, we used the term to convey ARHGAP18's functional role within the feedback loop, rather than to imply that it acts exclusively downstream. *

• *

We seek to clarify our perspective with the reviewer's assertion that we go "back and forth" as to if ARHGAP18 functions in a Rho Dependent or Rho Independent manner. It was our intent to propose a model where ARHGAP 18 acts in two separate circuits that regulate cell signaling. The first circuit involves ARHGAP18's canonical RhoA GAP activity, which involves ERMs and LOK/SLK, and is limited to the apical plasma membrane. This first signaling circuit was characterized in our prior Elife manuscript (Lombardo et al., 2024) and in an earlier JCB manuscript (Zaman and Lombardo et al., 2021). In this newly revised manuscript, we provide a partial mechanistic characterization of the second circuit, which we freely admit is much more complex and will likely require additional study to fully characterize.

• *

As both circuits operate as signaling feedback loops, we find the terms 'upstream' and 'downstream' to be of limited value, and we attempt to avoid their use when possible. We retain their use only when referring to the Hippo and ROCK signaling cascades, where these designations are well established. We suggest that the conceptual inconsistencies of Conundrum/ARHGAP18 may have arisen from the tendency to view it in strictly binary terms as upstream or downstream. Here, we propose a third possibility that ARHGAP18 functions as both, participating in a negative feedback loop.

• *

*We have edited and added data testing if the effects are Rho independent and discussion text in response to the reviewer's comments and clarify the molecular function of ARHGAP18.

"Additionally, focal adhesions and basal actin bundles are restored to WT levels when the ARHGAP18(R365A) GAP-ablated mutant is expressed in ARHGAP18 KO cells (Fig. 5A, B). These results represent the strongest argument that ARHGAP18 functions in additional pathways to RhoA/C alone. Our data suggests that at least one of the alternative pathways is through ARHGAP18's interaction with YAP and Merlin. From these data we conclude that ARHGAP18 has important functions in both RhoA signaling through both its GAP activity and in Hippo signaling through its GAP independent binding partners. "*

• *

• *

The study is descriptive and contains a series of observations that are not connected. Because of this, the study's conclusions are not well supported, and key mechanistic insight is limited. The study feels like a set of separate observations, that remain incompletely worked out and have some preliminary feel to them. The model in the last figure also seems to contain hypotheses based on the observations, several of which remains to be proven.

• *

*We present our revised manuscript, in which we've more clearly outlined our rationale and conclusions, as detailed in the above responses, to emphasize the overall connectivity of the study. We have also updated the title of Figure 7 to read "__Theoretical __Model of ARHGAP18's coordination of RhoA and Hippo signaling pathways in Human epithelial cells." To make it clear that we are presenting a working model, which has elements that will require additional investigation. Throughout the manuscript, we highlight the unknown elements that remain to be tested or other outstanding questions. Thus, we do not aim to characterize this complex signaling coordination completely. Instead, this manuscript represents the 3rd iteration in our systematic advances to describe this entirely new signaling pathway. We agree that, despite three separate manuscripts (this one included) to date, this work represents an early stage in understanding the system, many additional studies will be needed to characterize this signaling system fully. Figure 7 is presented as a working model that results from a thoughtful combination of our collective data and that of other researchers, derived from numerous species across decades of study. We firmly believe that proposing such integrative models is valuable for advancing the field. We also recognize the importance of clearly indicating which aspects remain hypothetical. We now explicitly note in several places within the discussion which components of the model will require further validation and experimental confirmation. For example, regarding our theoretical mechanism in Figure 7 we state: *

"Validation of the direct mechanism by which YAP/TAZ transcriptional changes drive basal actin changes in ARHGAP18 KO cells will require further investigation based on predictions from RNAseq results."

• *

Addressing any possible connection between key effects of ARHGAP18 KO (changes in actin, focal adhesion, integrins, Yap and merlin binding) could strengthen the manuscript. One such specific question is the whether the changes in integrin expression (RNAseq) are indeed connected to the actin alterations and reduction ion focal adhesions (Fig 1). Staining for these integrins to show they are indeed altered, and/or manipulating any of them to reproduce changes could provide and exciting addition.

• *

*We attempted to stain cells for Integrins by purchasing three separate antibodies. However, despite extensive optimization and careful selection of the specific integrins using our RNAseq results we were unable to get any of these antibodies to work in any cell type or condition. We believe that there is a technical challenge to staining for integrins due to their transmembrane and extracellular components, which we were unable to overcome. As an attempt to address the reviewers comment, we alternatively stained cells for paxillin which directly binds the cytoplasmic tails of integrins (Fig. 3&5). *

Some of the experimental findings are not convincing or lack controls. Fig 1: some of the western blots are not convincing or poor quality. [...] On the same figure, the quality of LIM kinase blots is poor. [...] The signal is weak, and the blot does not appear to support the quantification. The last condition (expression of flag-ARHGAP18) results in a large drop in pLIMK and pcofilin on the blot, which is not reflected by the graph. Addition of *a better blot and the use of strong positive or negative control would boost confidence in these data. *

• *

In response to this and other reviewers' comments, we have added new western data and quantification to Figure 1. We now focus on MLC/pMLC data as we believe these data highlight the potential Rho-independent mechanism of ARHGAP18, and we were able to greatly improve the quality of the blots through careful optimization. We hope the reviewer finds these blots and quantifications (Fig. 1E and F) more convincing.

*We note that phospho-specific Western blotting presents considerably greater technical challenges than conventional blotting. We believe that the appearance of an attractive looking blot does not always correlate to quality or reproducibility and have focused on taking extraordinarily careful steps in the blotting of our phospho-specific antibodies, which at times comes at the cost of the blot's attractiveness in appearance. For example, all phospho-specific antibodies are run using two color fluorescent markers to blot against both the total protein and the phospho-protein on the same blot. This approach often leads to blots that have reduced signal to noise compared to chemiluminescent Westerns. Additionally, we use phospho-specific blocking buffer reagents which do not contain phosphate-based buffers or agents that attract non-specific phospho-staining signals. These blocking buffers are not as effective as non-fat milk in pbs at blocking the background signal, however, they are ultimately cleaner for phospho-specific primary antibodies. We use carefully optimized protocols, from cell treatment to lysis, transfer, and antibody incubation, including methods developed by laboratories where the corresponding author of the manuscript was trained. Nonetheless, despite these efforts, we have now removed the LIMK and cofilin data because we deemed them unnecessary for the main conclusions of this manuscript and were unable to improve their quality to satisfy the reviewer. *

The changes in pMLC on the western blots are very small, and for any conclusion, these studies require quantification. Further, the expression levels of Flag-ARHGAP18 needs to be shown to support the statement that the protein is expressed, and indeed overexpressed under these conditions (vs just re-expressed).

In continuation of the above comment, we have made significant effort to improve the quality of our pMLC western blots and now provide quantification in Figure 1. We also now provide the Flag-ARHGAP18 signal as requested by the reviewer.

Fig 4: the differences in YAP nuclear localization under the various conditions are not well visible. Quantitation of nuclear/cytosolic signal ratio should be provided. Please provide a rationale and more context for using serum starvation and re-addition. What is the expected effect? Serum removal and addition is referred to as nutrient removal and re-addition, but this is inaccurate, as it does not equal nutrient removal, since serum contains a variety of other important components, e.g. growth factors too.

We have provided new quantification of the nuclear/cytosolic signal ratio in Figure 6D. We have explained our rational for the study through the following new text:

"Merlin is activated and localized to junctions upon signaling, promoting growth and proliferation; among these signals is the availability of growth factors and other components of serum (Bretscher et al., 2002). We hypothesized that since ARHGAP18 formed a complex with Merlin that ARHGAP18's localization may localize to junctions under conditions which promote Merlin activation."

• *

We have altered our use of "nutrient removal" to "serum removal"

The binding between ARHGAP18 and merlin is interesting, but a key limitation is the use of expressed proteins. Can the binding be shown for the endogenous proteins (IP, colocalization). Another important unaddressed question is the relevance of this binding, and the relation of this to altered YAP nuclear localization.

• *

*Our data in Fig. 6G shows binding of a resin bound human ARHGAP18 to endogenous YAP from human cells as suggested by the reviewer. In Fig. 6A, we have selected to use GFP-Merlin as Merlin shares approximately 60% sequence identity with Ezrin, Radixin, and Moesin (ERMs). Their similarity is such that Merlin was named for Moesin-Ezrin-Radixin-Like Protein. In our experience, nearly all Merlin or ERM antibodies have some cross-contaminating signal. Thus, a major concern is that if we were to blot for endogenous Merlin in the pull-down experiment, we may see a band that could in fact be ERMs. To avoid this, we tagged Merlin with GFP to ensure that the product pulled down by ARHGAP18 was Merlin, not an ERM. Regarding the ARHGAP18-resin bound column, our homemade ARHGAP18 antibody is polyclonal. We have extensive experience in pulldown assays and have found that the binding of a polyclonal antibody to the bait protein can produce less accurate results, as the binding site for the antibody is unknown and can sterically hinder attachment of target proteins like Merlin. In our experience, attachment to a flag-tag, which is expressed after a flexible linker at the N- or C-terminus, allows us to overcome this limitation, which we've used in this manuscript. *

Minor comments:

Introduction line 99: "When localized to the nucleus, YAP/TAZ promotes the activation of cytoskeletal transcription factors associated with cell proliferation and actin polymerization" Please clarify what you mean by this statement, that is inaccurate in its present for. Did you mean effects on transcription factors that control cytoskeletal proteins, or do you mean that Yap/Taz affect these proteins? Please also provide reference for this.

We've altered the sentence as suggested by the reviewer, which now reads the following:

"When localized to the nucleus, YAP/TAZ promotes transcriptional changes associated with cell proliferation and actin polymerization."

• *

*The full mechanism for how YAP/TAZ promotes proliferation and actin polymerization is a currently debated issue. We do not think introducing the various current proposed models is required for this manuscript, and we simply intend to convey that when in the nucleus, YAP/TAZ promotes transcriptional changes that drive actin polymerization and cell proliferation. *

-What is the cell confluence in these experiments? For epithelial cells confluence affects actin structure. Please comment on similarity of confluency across experimental conditions?

• *

All cellular experiments are paired where WT and ARHGAP18 KO cells are plated at the same time under identical conditions. For imaging, we plate all cells onto glass coverslips in a 6 well dish so that each condition is literally in the same cell culture plate and gets identical treatment. In our prior Elife paper studying ARHGAP18, we characterized that ARHGAP18 KO cells and WT cells divide at a similar rate and have similar proliferation characteristics. The epithelial cell cultures are maintained for experiments around 70-80% confluency. For the focal adhesion staining experiments, the confluency is slightly lower, between 50-60% to capture the focal adhesions towards the leading edge. We have added the following new text to further describe these methods: "Cell cultures for experiments were maintained at 70%-80% confluency. For focal adhesion experiments, the cell cultures were maintained at 50%-60% confluency."

-Fig 2 legend: please indicate that the protein detected was non-muscle myosin heavy chain (distinct from the light chain detected in Fig 1).

• *

We have altered original Figure 2 (new Figure 3) legend.

-Line 339-340: please check the syntax of this sentence -Western blot quantification: the comparison of experiments with samples run on different gels/blots requires careful normalization and experimental consistency. Please describe how this was achieved.

• *

We have added the following new text to further describe these methods:

"For blots which required quantification of antibodies that were only rabbit primaries (e.g., pMLC/MLC antibodies listed above), samples were loaded onto a single gel and transferred onto a single membrane at the same time. After transfer, the membrane was cut in half and subsequent steps were done in parallel. All quantified blots were checked for equal loading using either anti-tubulin as a housekeeping protein or total protein as detected by Coomassie staining"

Reviewer #3 (Significance (Required)):

Rho signalling is a central regulator of an array of normal and pathological cell functions, and our understanding of the context dependent regulation of this key pathway remains very incomplete. Therefore, new knowledge on the role of specific regulators, such as ARHGAP18, is of interest to a very broad range of researchers. A further exciting aspect of this protein, that despite indications by many studies that it acts as a GAP (inhibitor) for Rho proteins, there are findings in the literature that suggest that its manipulation can affect actin in unexpected (opposite) manner. These point to possible Rho-independent roles, and warranted further in-depth exploration.

One of the strength of the study is that it explores possible roles of ARHGAP18 beyond RhoA and describes some new and interesting observations, which advance our knowledge. The authors use some excellent tools (e.g. ARHGAP KO cells and re-expression) and approaches (e.g. super resolution microscopy to analyze actin changes, RNAseq and bioinformatics to find genes that may be downstream from ARHGAP18). A key limitation of the study however, is that it is not clear whether the observed findings are indeed independent from RhoA. Further limitation is that potential causal relationships between the described findings are not studied, and therefore the findings are in some cases overinterpreted, and limited mechanistic insights are provided. In some cases the exclusive use of expressed proteins is also a limitation. Finally, some of the experiments also need improvement.

Reviewer expertise: RhoA signalling, guanine nucleotide exchange factors, epithelial biology, cell migration, intercellular junctions.

In the above comments, we detail the new experimental data addressing reviewer 3's listed key limitations. We've added new data using the Rho GAP deficient ARHGAP18(R365A) variant which allows for the direct characterization of ARHGAP18's Rho independent activity. We have introduced new data in WT cells studying endogenous proteins to address the limitations from expressed proteins. Finally, we have moderated our language to address overinterpretation. Collectively, we believe that our revised manuscript addresses the constructive reviewer's comments.

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

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

The study by Murray et al explores the effects of ARHGAP18 on the actin cytoskeleton, Rho effector kinases, non-muscle myosin, and transcription. Using super resolution microscopy, they show that in ARHGAP18 KO cells there is a mixed and unexpected cytoskeleton phenotype where myosin phosphorylation appears to be increased, but actin is disorganised with reduced stress fibres, diminished focal adhesions and augmented invasiveness. They conclude that the underlying mechanisms are likely independent from RhoA. Next, they perform RNAseq using the KO cells and identify an array of dysregulated genes, including those that play crucial roles in microvilli (related to previously published findings). Analysis of the data identify gene expression changes that are relevant for altered focal adhesion (integrins). Further analysis reveals that a large cohort of the dysregulated genes are YAP targets. They then show that in ARHGAP18 KO cells YAP nuclear localization, as detected by immunostaining, is augmented; and demonstrate that immobilized ARHGAP18 protein can bind the Hippo regulator merlin as well as YAP itself.

Major comments:

1. The premise of the study (that ARHGAP18 is a RhoA effector or may acts independently of RhoA) remains not proven. At several places (including in the title) the authors refer to ARHGAP18 as a Rho effector, which would suggest that it is downstream form Rho, but the basis for this is not clear. In fact, their own previous study suggested that ARHGAP is a RhoA regulator, rather than an effector. In general, the connection of the described effects to RhoA remains unclear, and not addressed in this study. The authors seem to go back and forth in their conclusions regarding the connection between ARHGAP18 and RhoA. For example, the first section of results is finished by stating (line 194): "These data support the conclusion that ARHGAP18 acts to regulate basal and junctional actin through Rho-independent mechanism". But the next section starts by stating (line 198): "We hypothesized that the invasive and cytoskeletal phenotypes observed at the basal surface of cells devoid of ARHGAP18 may be a result of changes in regulation at the transcriptional level either directly through RhoA signaling or through an additional mechanism specific to ARHGAP18". The paper would be strengthened by adding data that show whether the effects are indeed downstream, from RhoA or RhoA independent. If there is no sufficient demonstration that ARHGAP18 is downstream of RhoA and is an effector, this needs to be stated explicitly and the wording should be changed.
2. The study is descriptive and contains a series of observations that are not connected. Because of this, the study's conclusions are not well supported, and key mechanistic insight is limited. The study feels like a set of separate observations, that remain incompletely worked out and have some preliminary feel to them. The model in the last figure also seems to contain hypotheses based on the observations, several of which remains to be proven. Addressing any possible connection between key effects of ARHGAP18 KO (changes in actin, focal adhesion, integrins, Yap and merlin binding) could strengthen the manuscript. One such specific question is the whether the changes in integrin expression (RNAseq) are indeed connected to the actin alterations and reduction ion focal adhesions (Fig 1). Staining for these integrins to show they are indeed altered, and/or manipulating any of them to reproduce changes could provide and exciting addition.
3. Some of the experimental findings are not convincing or lack controls.

Fig 1: some of the western blots are not convincing or poor quality. The changes in pMLC on the western blots are very small, and for any conclusion, these studies require quantification. Further, the expression levels of Flag-ARHGAP18 needs to be shown to support the statement that the protein is expressed, and indeed overexpressed under these conditions (vs just re-expressed). On the same figure, the quality of LIM kinase blots is poor. The signal is weak, and the blot does not appear to support the quantification. The last condition (expression of flag-ARHGAP18) results in a large drop in pLIMK and pcofilin on the blot, which is not reflected by the graph. Addition of a better blot and the use of a strong positive or negative control would boost confidence in these data.

Fig 4: the differences in YAP nuclear localization under the various conditions are not well visible. Quantitation of nuclear/cytosolic signal ratio should be provided. 4. Please provide a rationale and more context for using serum starvation and re-addition. What is the expected effect? Serum removal and addition is referred to as nutrient removal and re-addition, but this is inaccurate, as it does not equal nutrient removal, since serum contains a variety of other important components, e.g. growth factors too. 5. The binding between ARHGAP18 and merlin is interesting, but a key limitation is the use of expressed proteins. Can the binding be shown for the endogenous proteins (IP, colocalization). Another important unaddressed question is the relevance of this binding, and the relation of this to altered YAP nuclear localization.

Minor comments:

• Introduction line 99: "When localized to the nucleus, YAP/TAZ promotes the activation of cytoskeletal transcription factors associated with cell proliferation and actin polymerization" Please clarify what you mean by this statement, that is inaccurate in its present for. Did you mean effects on transcription factors that control cytoskeletal proteins, or do you mean that Yap/Taz affect these proteins? Please also provide reference for this.
• What is the cell confluence in these experiments? For epithelial cells confluence affects actin structure. Please comment on similarity of confluency across experimental conditions?
• Fig 2 legend: please indicate that the protein detected was non-muscle myosin heavy chain (distinct from the light chain detected in Fig 1).
• Line 339-340: please check the syntax of this sentence
• Western blot quantification: the comparison of experiments with samples run on different gels/blots requires careful normalization and experimental consistency. Please describe how this was achieved.

Significance

Rho signalling is a central regulator of an array of normal and pathological cell functions, and our understanding of the context dependent regulation of this key pathway remains very incomplete. Therefore, new knowledge on the role of specific regulators, such as ARHGAP18, is of interest to a very broad range of researchers. A further exciting aspect of this protein, that despite indications by many studies that it acts as a GAP (inhibitor) for Rho proteins, there are findings in the literature that suggest that its manipulation can affect actin in unexpected (opposite) manner. These point to possible Rho-independent roles, and warranted further in-depth exploration. One of the strength of the study is that it explores possible roles of ARHGAP18 beyond RhoA and describes some new and interesting observations, which advance our knowledge. The authors use some excellent tools (e.g. ARHGAP KO cells and re-expression) and approaches (e.g. super resolution microscopy to analyze actin changes, RNAseq and bioinformatics to find genes that may be downstream from ARHGAP18). A key limitation of the study however, is that it is not clear whether the observed findings are indeed independent from RhoA. Further limitation is that potential causal relationships between the described findings are not studied, and therefore the findings are in some cases overinterpreted, and limited mechanistic insights are provided. In some cases the exclusive use of expressed proteins is also a limitation. Finally, some of the experiments also need improvement.<br /> Reviewer expertise: RhoA signalling, guanine nucleotide exchange factors, epithelial biology, cell migration, intercellular junctions.

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

Learn more at Review Commons

Referee #2

Evidence, reproducibility and clarity

This manuscript investigates the Rho effector, ARHGAP18 in Jegs cells, a trophoblastic cell line. It presents a number of new pieces of data, which increase our understanding of the importance of this GAP on cell function and explains at a molecular level previous results of other workers in the field. ARHGAP18 was originally given the name "conundrum' and continues to stand apart from the majority of other GAP proteins and their functions. Hence the data here is significant and of high standard.

The data is clear, and the images are of high quality and extremely impressive in their resolution. It is significant and adds a further layer to our understanding of the regulation of cell migration, particularly in the formation and resolution of microvilli.

The data is based on the use of the cell line Jeg3. Even the authors previous publication in eLife is based only on this cell line. They need to show the conclusions are general and not specific to this line of cells. As an extension of this, is the ARHGAP18 function shown here only in transformed cells? Does the same mechanisms operate in normal cells, which respond to activation to proliferate or migrate? In endothelial cells, Lovelace et al 2017 showed localisation to microtubules and that depletion of ARHGAP18 resulted in microtubule instability. The authors may like to comment on the differences. Is this a cell type difference or RhoA versus RhoC difference?

On pages 7,9 they conclude that MLC and basal and junctional actin are regulated through a GAP independent mechanism. The best way to show this is with overexpression of a GAP mutant.

There is a huge amount of data presented in Figure 3, but their 2 genes which they focus on, LOP1 and CORO1A, are discussed but no actual data presented in support.

Significance

The data is significant and adds a further layer to our understanding of the regulation of cell migration, particularly in the formation and resolution of microvilli.

This manuscript will be of significance to an basic science audience in the field of RhoGTPases and cell migration.

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

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

This manuscript describes a dual mechanism by which ARHGAP18 regulates the actin cytoskeleton. The authors propose that in addition to the known role for ARHGAP18 in regulating Rho GTPases, it also affects the cytoskeleton through regulation of the Hippo pathway transcriptional regulator YAP. ARHGAP18 knockout Jeg3 cells are were generated and show a clear loss of basal stress fiber like F-actin bundles. The authors further characterize the effects of ARHGAP18 knockout and overexpression. It is also discovered that ARHGAP18 binds to the Hippo pathway regulator Merlin and to YAP. Ultimately it is concluded that ARHGAP18 regulates the F-actin cytoskeleton through dual regulation of RHO GTPases and of YAP. While the phenotype of the ARHGAP18 knockout and the association of ARHGAP18 with Merlin and YAP is interesting, I found the authors conclusion that these phenotypes are due to ARHGAP18 regulation of both RHO and YAP to be based on largely correlative evidence and sometimes lacking in controls or tests for significance. In addition the authors often make overly strong conclusions based on the experimental evidence. In some instances, the rationale for how the experimental results support the conclusion is insufficiently articulated, making evaluation challenging. In general although the authors have some interesting observations, more definitive experiments with proper controls and statistical tests for significance and reproducibility are needed to justify their overall conclusions.

Specific Comments

1) The authors make a big point about the effects of ARHGAP18 on myosin light chain phosphorylation. However this result is not quantified and tested for statistical significance and reproducibility.

2) Along similar lines in Figure 2C they state that overexpression of ARHGAP18 causes cells to invade over the top of their neighbors. This might be true and interesting, but only a single cell is shown and there is no quantification or controls for simply overexpressing something in that cell. The authors also conclude from this image that the overexpression phenotype is independent of its GAP activity on Rho. It is not clear how this conclusion is made based on the data. It would seem like a more definitive experiment would be to see if a similar phenotype was induced by an ARHGAP18 mutant deficient in GAP activity.

3) In Figure 3 the authors compare gene expression profiles of ARHGAP18 knockout cells to wild-type cells. They see lots of differences in focal adhesion and cytoskeletal proteins and conclude that this supports their conclusion that ARHGAP18 is not just acting through RHO. The rationale for this in not clear. In addition, they observe changes in expression profiles consistent with changes in YAP activity. They conclude that the effects are direct. This very well might be true. However RHO is a potent regulator of YAP activity and the results seem quite consistent with ARHGAP18 acting through RHO to affect YAP.

4) In Figure 4A showing Merlin binding to ARHGAP18 there is no control for the amount of Merlin sticking to the column as was done in Figure 4F for binding experiments with YAP. This makes it difficult to determine the significance of the observed binding.

5) The images in Figure 4C showing YAP being maintained in the nucleus more in ARHGAP18 knockout cells compared to wild-type. However the images only show a few cells and YAP localization can be highly variable depending on where you look in a field. Images with more cells and some sort of quantification would bolster this result.

Significance

While the phenotype of the ARHGAP18 knockout and the association of ARHGAP18 with Merlin and YAP is interesting, I found the authors conclusion that these phenotypes are due to ARHGAP18 regulation of both RHO and YAP to be based on largely correlative evidence and sometimes lacking in controls or tests for significance. In addition the authors often make overly strong conclusions based on the experimental evidence. In some instances, the rationale for how the experimental results support the conclusion is insufficiently articulated, making evaluation challenging. In general although the authors have some interesting observations, more definitive experiments with proper controls and statistical tests for significance and reproducibility are needed to justify their overall conclusions.

for - DNA replication experiment - in vitro - lots of errors - 1/10,000 pairs error rate - our genomes 3 billion base pairs long each - so 300,000 errors - would be fatal to any cell

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

Learn more at Review Commons

Reply to the reviewers

Dear editor and reviewers,

We sincerely thank you for your thoughtful comments and constructive suggestions, which have greatly improved the quality and clarity of our manuscript. In response, we have implemented all requested changes, which are highlighted in yellow throughout the revised text, and updated several figures accordingly. Furthermore, we have performed all additional experiments recommended by the reviewers and incorporated the new data into the manuscript. To enhance clarity, we have also included a schematic representation of our proposed model in an additional figure, providing a concise visual summary of our findings.

We hope that these revisions fully address all concerns raised by the reviewers and meet all the expectations for publication.

Below, we answer the reviewers point by point (in blue).

---

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

In this paper, the authors address the important question of the role of centrosomes during neuronal development. They use Drosophila as an in vivo model. The field is somewhat unclear on the role and importance of centrosomes during neuronal development, although the current data would suggest they are dispensable for axon specification and growth. Early studies in cultured mammalian neurons showed that centrosomes are active and that their microtubules can be cut and transported into the neurites. But a study then showed that centrosomes in these cultured neurons are deactivated relatively early during neuronal development in vitro and that ablating centrosomes even when they are active had no obvious effect on axon specification and growth. Consistent with this, a study in Drosophila provided evidence that centrosomes were not active or necessary in different types of neurons. More recently, a study showed that centrosomal microtubules are dispensable for axon specification and growth in mice in vivo but are required for neuronal migration in the cerebral cortex. However, another study has linked the generation of acetylated microtubules at centrosomes with axon development. In this current study, the authors examine the effect of centrosome loss on various motor and sensory neurons and muscles mainly by examining mutants in essential centriole duplication genes. They associate axonal routing and morphology defects with centrosome loss and provide some evidence that centrosomes could still be active in the developing neurons. Overall, they conclude that centrosomes are active during at least early neuronal development and that this activity is important for proper axonal morphology and routing.

While I think this study addressing a very interesting and important question, I think as it stands the data is not sufficient to be conclusive on a role for centrosomes during neuronal development. My biggest concern is that most phenotypes have not yet been shown to be cell autonomous, as whole animal mutants have been analysed rather than analysing the effect of cell-specific depletion, and the evidence for active centrosomes needs to be strengthened. If the authors can provide stronger evidence for a role of centrosomes in axonal development then the paper will certainly be of interest to a broad readership.

We thank the reviewer for the clear and concise summary and fully agree that our study addresses a critical gap in understanding. Centrosomes have long been implicated in morphogenesis, yet their precise contribution to nervous system development has remained unclear. Our findings provide compelling evidence that centrosomes are indispensable for proper nervous system formation and that their absence also triggers muscular defects, highlighting their broader role in tissue organization.

We acknowledge that the original manuscript lacked some key details; therefore, we have now strengthened our conclusions with additional experiments. Specifically, we demonstrate that these effects are cell-autonomous by using two independent RNAi lines targeted to a subset of motor neurons. Furthermore, we present new data showing that neuronal centrosomes remain active during the early stages of axonal development, emphasising their functional relevance in morphogenesis. All new experiments, figures, and corresponding text revisions are detailed below.

Major comments 1) The sas-6 transallelic combination shows only 17% embryonic lethality compared to 50% embryonic lethality with sas-4 mutants. Given that both mutants should result in the same degree of centrosome loss (this should be quantified in sas-6 mutants) it would suggest that either sas-4 has other roles away from centrosomes or that the sas-4 mutant chromosome used in the experiment has other mutations that affect viability. The effect of picking up "second-site lethal" mutations on mutant chromosomes is common and so I would not be surprised if this is the reason for the difference in phenotypes. This can be addressed either by "cleaning up" the sas-4 mutant chromosome by backcrossing to wild-type lines, allowing recombination to occur and replace the potential second site mutations, or by using transallelic combinations of sas-4, as they did for sas-6. The "easier" option may just be to analyse all the phenotypes with the sas-6 transallelic combination.

We appreciate this comment, as it brought to light an issue with the CRISPR line Sas-6-Δa. Upon reanalysing all the data, we determined that this line is embryonic lethal both in homozygosis and when combined with the deficiency uncovering the genomic region, Df(3R)BSC794. In contrast, Sas-6-Δb homozygotes are viable. The inconsistency between these results raised concerns about whether the Δa and Δb Sas-6 mutants carry deletions confined to the Sas-6 coding region. Although this would not hinder our cell biology analysis, it could represent a problem in viability tests. To address this, we repeated all analyses using Sas-6-Δb homozygotes and Sas-6-Δb combined with Df(3R)BSC794. These new results are more consistent and indicate that approximately 50% of Sas-6/Def individuals hatch as adults. Fig. 3 was redone and the manuscript text changed in view of these results.

2) Using "whole animal" mutants for assessing neuronal morphology is risky due to non-cell-autonomous effects. The authors have carried out some phenotypic analysis of neurons depleted of Sas-4 by cell-specific RNAi, but I feel they need to do this for all of their analysis. This includes embryonic lethality measures, quantification of centrosome numbers, and all axonal phenotypes in Sas-4 RNAi neurons. It would also be prudent to use 2 distinct RNAi lines to help ensure any phenotypes are not off-target effects (and this may help clarify why the authors see some additional phenotypes with RNAi). Indeed, there are relatively weak phenotypes in muscles when using RNAi compared to the mutants and these potential non-cell-autonomous effects could then have a knock-on effect on neuronal morphology. If the authors were concerned that RNAi is not very efficient (explaining any potential weaker phenotypes than in mutants) the authors could examine the effectiveness of RNAi lines by analysing protein depletion by western blotting or mRNA depletion by rt-qPCR (although this has to be done in a different cell type due to the difficulty in obtaining a neuronal extract).

We have now added a new panel to supplementary Figure 1, showing how the expression of a different Sas-4 RNAi line (2) induces similar nervous system phenotypes when expressed only in aCC, pCC and RP2 pioneer neurons (Sup. Fig. 1 M-O).

3) When analysing centriole presence or absence it is a good idea to stain with two different centriole markers e.g. Asl and Plp. This helps rule out unspecific staining. It is clear from the images that similar sized foci can be observed outside of the cells (see Figure 5A for example), so clearly some of the foci that appear to be within the cells may also be unspecific staining.

In a new supplementary figure, we now show that Asl and Plp colocalize and quantify the number of times we find this colocalization in neurons (Supl. Fig 3). In addition, and we apologise for the confusion, but the reason why there are foci outside the marked cells is because these are wholemount embryonic stainings and the anti-Plp antibody marks all centrosomes in all cells in the embryo.

4) The evidence for active centrosomes is not that convincing. Acetylated tubulin is associated with stable MTs, which are not normally organised by "active" centrosomes that nucleate dynamic microtubules. Moreover, it is plausible that centriole foci happen to overlap with the acetylated tubulin staining by chance. This would explain why not all centrosomes colocalise with acetylated tubulin signal. The authors could better test centrosome activity by performing live imaging with EB1-GFP. If centrosomes are active, it is very easy to observe the many comets produced by the centrosomes.

We appreciate the reviewer’s comment and agree that acetylated tubulin alone is not an ideal marker for centrosome activity. To address this, we performed live imaging of aCC neurons expressing EB1-GFP together with Asl-Tomato. This was technically challenging because we were imaging only two neurons per segment in live embryos, under significant limitations in fluorescence detection and timing. Despite these constraints, we were able to clearly observe EB1 comets emerging from the centrosome and moving toward the cell periphery, providing direct evidence of microtubule nucleation from centrosomes in neurons.

Importantly, we complemented this with a microtubule depolymerization/polymerization assay, which provides unequivocal evidence that polymerization initiates at the centrosome. After depolymerization, we observed microtubule regrowth from the centrosome, confirming its role as an active microtubule-organizing centre in these neurons. Together, we hope that these results are enough to demonstrate that neuronal centrosomes are functionally active during early axonal development. These experiments are presented in Figure 6 and corresponding text in the manuscript.

5) If the authors believe that centrosomes have a role in axon pathfinding in sensory neurons, they should show that these centrosomes are active, at least during early stages (again using EB1-GFP imaging).

We appreciate the reviewer’s suggestion and agree that EB1-GFP imaging would be the most direct way to assess centrosome activity in sensory neurons. However, performing time-lapse imaging in these neurons is technically very demanding due to their location and accessibility in live embryos, and we did not attempt this approach. Instead, we now provide new evidence showing that sensory neuron centrosomes colocalize with both α-tubulin and γ-tubulin. This strongly supports that these centrosomes are associated with microtubule nucleation machinery and are as likely as motor neuron centrosomes to be active during early stages of axon development. These new data have been included in the revised manuscript (see Figure 5 and corresponding text).

6) The authors mention in the discussion that "increased JNK activity, can result in axonal wiggliness (Karkali et al, 2023)". I therefore wonder whether centrosome loss may induce JNK activation (the stress response), as this would then indicate an indirect effect of centrosome loss on axonal structure rather than a direct influence of centrosome-generated microtubules. The authors could assess whether the DNK-JNK pathway is activated in neurons lacking centrosomes by expression UAS-Puc-GFP and quantifying the nuclear signal.

In a new supplementary figure, we now show by using a reporter for JNK signalling, as requested, that Sas-4 neurons do not activate the JNK pathway (Supl. Fig 4).

7) In Figure 5, the authors claim that they find "a correlation between axonal guidance phenotypes and the numbers of centrioles per embryo". I don't think this is a strong correlation. The difference in centriole number between embryos with no defects and those with defects is very small. In contrast, the difference between centriole numbers in control (no defects) and mutant (no defects) is very large. So, there does not appear to be a strong correlation between centrosome number and phenotype.

We agree and we have corrected this sentence to better explain the results.

Minor comments

1) I don't understand Figure 3C - why do the % of surviving homozygotes and heterozygotes add up to 100%? Should the grey boxes not relate to dead and the white to surviving?

Thank you for pointing this out. Figures 1B and 3C represent only the surviving individuals. The grey boxes correspond to surviving homozygotes, and the white boxes correspond to surviving heterozygotes. The percentages add up to 100% only at embryonic stages because all embryos reach late embryonic stages. The grey and white boxes reflect the proportion of these two genotypes among the survivors, not the total number of embryos including those that died. We have changed the text to convey this.

2) "In mouse fibroblasts, myoblasts and endothelial cells, centrosome orientation is important for nuclear positioning and cell migration(Chang et al, 2015; Gomes et al, 2005; Kushner et al, 2014)." Do you mean "centrosome position"?

Yes, text changed, thank you for spotting it.

3) In the introduction, the authors mention Meka et al. when saying the centrosomal microtubules are important for axonal development, but they should also discuss the counter argument from Vinopal et al., 2023 (Neuron) that showed how centrosomes were required for neuronal migration but not axon growth, which was instead mediated by Golgi-derived microtubules.

Done, thank you very much.

4) Lines 228-230 - repeated sentence

Corrected, thank you very much.

5) Additionally, we did not detect centrioles in the quadrant opposite the axon exit point (Fig. 2B n=75) - this data is not in Fig 2B

Correct, it is in figure 4B, thank you very much.

6) "This significant decrease in the humber of centrioles further supports the critical role of Sas-4 in pioneer neurons of the ventral nerve cord (VNC) during Drosophila embryogenesis". It rather highlights that Sas-4 is required for centriole formation in these neurons. Also, humber = number.

We agree, and have changed the text, thank you very much.

7) Result title: Non-ciliated sensory neurons have centrioles. This is kind of obvious. A better title may be "axon phenotypes correlate with centriole numbers in sensory neurons" but unfortunately i don't think there is good evidence for this (See major point above).

We agree and we have changed. We now believe we have strong evidence to support it. We hope the additional data presented in the revision convincingly demonstrate this point.

Reviewer #1 (Significance (Required)):

As mentioned above, the advance will be important if more evidence is provided. In this case, the paper will be interesting to a broad readership. But currently the paper is limited by the lack of evidence for centrosome function and activity in the neurons.

We hope that reviewer 1, now considers that the manuscript is not limited anymore and that it shows convincing evidence for centrosome function and activity in embryonic neurons.

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

Summary: In this manuscript, Gonzalez et al. examine the potential function of centrosomes in the neurons and muscle cells of Drosophila embryos. By studying various mutant and RNAi lines in which centriole duplication has been disrupted, they conclude that the loss of centrioles disrupts axonal pathfinding and muscle integrity.

Major points: 1. Throughout the manuscript, the phenotypes presented are often quite subtle. For this reason, I would really recommend that these experiments are scored blind. Perhaps the authors did this, but I didn't see any mention of this.

All our phenotypic analyses are performed blind. We apologize for not having originally included this information in the Methods section; it has now been added. Embryos are stained using colorimetric methods (DAB) to label the nervous system, while balancer chromosomes are marked with a fluorescent antibody. This approach allows us to assess and quantify phenotypes using white light without knowing whether the embryos are homozygous mutants or heterozygous, which can only be detected by changing the channels to fluorescence.

1. The authors conclude that neurons have active centrioles that function as centrosomes (Figure 6), but the data here is confusing. The authors state that in these cells they observe acetylated MTs extending from the centrosomes and these colocalised with g-tubulin. But the authors don't show the overlap between centrosomes, g-tubulin and MTs, as they stain for these separately. This is problematic, as it was not clear from these images that the majority of the MTs really are extending from the centrosome: the centrosome may just associate or be close by to these MT cables (Figure 6A,B). Moreover, the authors show that only a fraction of the centrosomes in these cells associate with g-tubulin, so presumably in cells where the centrosomes lack g-tubulin they would not expect the centrosomes to be associated with the MTs-but they do not show that this is the case. Perhaps the authors can't test this, but an alternative would be to show that these MT arrays are absent in Sas-4 mutants. This would give more confidence that these MTs arise from the centrosomes.

We agree that the initial data based on acetylated microtubules and γ-tubulin colocalization were not sufficient to conclude that microtubules originate from the centrosome, as these markers can only suggest association. To address this, we have now included additional experiments that provide direct evidence of centrosome activity.

First, we performed live imaging of aCC neurons expressing EB1-GFP together with Asl-Tomato. Despite the technical challenges of imaging only two neurons per segment in live embryos under strict fluorescence and timing constraints, we were able to clearly observe EB1 comets emerging from the centrosome and moving toward the cell periphery. This demonstrates active microtubule nucleation from centrosomes rather than mere proximity to microtubule bundles.

Second, we carried out a microtubule depolymerization/polymerization assay, which provides unequivocal evidence that polymerization initiates at the centrosome. After depolymerization, microtubules regrew from the centrosome, confirming its role as an active microtubule-organizing center. These experiments go beyond colocalization and directly address the concern that centrosomes might simply be adjacent to microtubule cables.

Regarding the suggestion to use Sas-4 mutants, while we did not perform this experiment, the regrowth assay combined with EB1 imaging strongly supports that these microtubules originate from the centrosome. All new data are presented in Figure 6 and the corresponding text in the revised manuscript.

1. The authors show that muscle cell integrity is compromised by centriole-loss (Figure 2). This is very surprising as it is widely believed that centrosomes are non-functional in muscle cells, and the MTs are instead organised around the nuclear envelope. I'm not aware of the situation in Drosophila muscle cells, but the authors should ideally try to examine if the centrioles are functioning as centrosomes in these cells. At the very least they should discuss how they think centriole-loss is influencing the muscle integrity when it is widely believed they are inactive in these cells.

We do not claim that centrosomes are active in muscle cells at these developmental stages. The observed muscle defects could result from earlier processes such as cell division, migration, or muscle fusion. We agree that this is an intriguing observation; however, pursuing this question further would go beyond the scope of the current manuscript. As requested by the reviewer, we have now expanded the discussion to consider how centriole loss might impact muscle integrity.

Regardless of the strength of the supporting data, I think the authors should tone down their conclusions. The title and abstract led me to believe that centriole loss would cause significant problems in axonal pathfinding and muscle integrity. In all the mutant specimens examined (and certainly the low magnification views shown in Figure 1D'-F', Figure 1I'-K' and Figure 2D'-F') the mutants look very similar to the WT. Many readers may not get past the title and abstract, so the authors should make it clearer that these defects are very subtle.

We have changed the text to convey this idea.

Minor points: 1. In Figures 4 and 5, CP309 staining is relied on to identify centrioles, but there is quite a background of non-specific dots, making it hard to be certain what is a centriole and what isn't. For example, in Figure 5D' there are lots of dots within some of the cells - are any of these centrioles? How can the authors be certain which dot is a centriole in some of the cells shown in Figure 5C'? Is it possible to use a second marker and only count as centrioles dots that are recognised by both antibodies?

We thank the reviewer for this suggestion and agree that using a second marker improves confidence in centriole identification. In a new supplementary figure (Supplementary Fig. 3), we now show that Asl and Plp colocalize in neurons and provide a quantification of the frequency of this colocalization. This dual labelling confirms the identity of centrioles and addresses the concern about non-specific background.

We also apologize for any confusion regarding the presence of foci outside the marked cells. These images are whole-mount embryonic stainings, and the anti-Plp antibody labels all centrosomes in all cells of the embryo, which explains the additional foci observed.

In the abstract that authors state that traditionally centrosomes have been considered to be non-essential in terminally differentiated cells. I don't think this is correct. In the standard "textbook" view of a cell, the centrosome is normally positioned in the centre of the cell organising an extensive array of MTs that are thought play an important role in organising intracellular transport, the positioning and movement of organelles and the maintenance and establishment of cell polarity. I don't think it is only recent evidence that suggests they play vital roles in terminally differentiated cells.

We thank the reviewer for this correction and we have changed the text accordingly.

1. Line 162 the authors state that in the RNAi knockdown lines they observe several additional phenotypes, but then in the same sentence (Line 164) they say that these defects were also observed in the original mutant and mutant/Df lines.

We apologise for this confusion, we have rearranged the sentence for clearance.

The sentences in Line281-287 don't reference any of the Figures, so it seems the authors are just stating these results without presenting any data (e.g. "Significantly, we also found a correlation between axonal guidance phenotypes and the numbers of centrioles per embryo". If they've tested this correlation, they should show it.

We have rearranged the sentences for better understanding.

In Figure 7 I did not understand how the authors measured tortuosity (wiggliness) and could see no description in the methods. This is important as, again the defect seems quite subtle, but perhaps I am not understanding which bits of the axon are being measures. Is it just the small bit of the axons close to the asterixis that is being measured, or the whole FasII track?

We have now added another quantification and additional descriptions in the methods section.

Reviewer #2 (Significance (Required)):

The potential function of centrosomes in axonal outgrowth is quite controversial, so this study is potentially of considerable interest.

However, several aspects of the data presented here were confusing or not terribly convincing. In its present state, I don't think the main conclusions are strongly enough supported by the data.

We hope that reviewer 2, now considers that the manuscript is not confusing anymore and that it shows convincing evidence for centrosome function and activity in embryonic neurons.

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

The manuscript of González et al. entitled "Centriole Loss in Embryonic Development Disrupts Axonal Pathfinding and Muscle Integrity" deals with the role of centrosomes in shaping axonal morphology. To this aim the AA analysed Drosophila Sas-4 mutants that are reported to develop until adult stage without centrioles. Remarkably, the AA observe that 50% of the homozygous mutant embryos fail to hatch as larvae. The present observations suggest that centrosome loss results in axonemal shaping defects and muscle developmental abnormalities. Finally, the AA show the presence of functional centrosomes in neurons. In my opinion, the manuscript is interesting because shows unexpected findings. However, to justify these new findings the AA are required to improve some experimental observations.

We thank the reviewer for his summary of our work and for considering it interesting. We have taken into account all the comments and believe that these have helped improve our manuscript.

Major: Abstract- It is unclear in which phenotypic condition the observations of centrosome loss or centrosome presence have been found. Please better explain. l.36. embryos, larvae, adult, from Sas4 or controls? If mutants, the observations are very interesting since Sas4 would be without centrioles. Indeed, Basto et al., show that chemosensory neurons do not develop an axoneme in the absence of centrioles, but extend dendrites toward the sensory bristle.

We have made clear which refer to wild-type and which are Centriole Loss (CL) conditions. CL conditions refer to mutant and downregulation conditions, whereas targeted downregulation refers to RNAi downregulation only in neurons.

I do not think appropriate the use of "centriole" in the main title since the centrioles would be localized by true centriolar antigens rather than by centrosomal antigens. This problem occurs throughout the text and some figures where the AA image centrioles by centrosomal material. In Gig. 5A only the AA properly look at Asl localization. The other pictures of presumptive centrioles or centriole quantification report CP309 dots. This localization does not unequivocally reveal centrioles, since CP309 is essentially required for centrosome-mediated Mt nucleation. There are differentiated Drosophila tissues in which centrioles are present, but inactivated, and unable to recruit pericentriolar material. Mt are nucleated by ncMTOCs that contain centrosomal material and gamma-tubulin. Thus, the centrosomal antigens do not colocalize with centrioles.

We have changed centrioles to centrosomes in the title and most sections in the manuscript. We have also included an extra control, showing that Asl and Plp colocalize and quantify the number of times we find this colocalization in neurons (Supl. Fig 3). Asl is a reliable and widely used marker for centrioles, as it localizes specifically to the centriole structure (Varmark H, Llamazares S, Rebollo E, Lange B, Reina J, Schwarz H, Gonzalez C. Asterless is a centriolar protein required for centrosome function and embryo development in Drosophila. Curr Biol. 2007 Oct 23;17(20):1735-45. doi: 10.1016/j.cub.2007.09.031. PMID: 17935995.)

Minor: l. 58. The early arrest is mainly due to a checkpoint control. In double mutant for Sas4 and P53 the embryos survive longer, even if their further development is asrrested.

We thank the reviewer for this comment, and we have changed the text accordingly.

1. Previous works, also quoted by the AA, reported that in mature neurons the centrosome are inactivated, whereas the present manuscript describes functional centrosomes in Drosophila motor and peripheral nervous system. This is an intriguing observations that needs a better explanation in Discussion section.

We thank the reviewer for this comment, and we have changed the discussion accordingly.

l.143-145. I understand that 50% of the Sas4 embryos that reach the adult stage have centrioles. Is it correct? But if it is so, how the AA explain the absence of centrioles in sensory neurons of adult flies as reported by Basto et al. ?

According to our results they have less centrioles than controls already at embryonic stages. In addition, as reported in Basto et al. they continue losing centrioles during larval stages and metamorphosis, which explains why centrioles are not detected at adult stages.

l.215. It is unclear for me why the AA analyse Sas6 flies, unless explain the mutant phenotype.

To strengthen our conclusions with Sas-4 and exclude the possibility that the observed phenotypes arise from a centrosome-independent function of Sas-4. For this reason, we have taken additional steps to confirm that the effects are specifically due to centrosome loss and we used Sas-6 mutants as one of these.

l.221. How the centrioles have been quantified? What antibody, the AA used.

We have quantified centrosomes using antibodies agains Plp (CP309) and Asl-YFP expression.

l.244. and Fig 4C,D. I see high background with CP309. As reported previously I think better to use antibodies against centriolar proteins, such as Sas6, Ana1, Asl, or Sas4 ( if centrioles are present in 50% of mutants as the AA claim, the antibody could be also useful). In addition, I can see some CP309 spots in Fig 4E,F. Are they centrioles?

Indeed, as we report, Sas-4 mutant embryos are not totally devoid of centrosomes. In addition, and we apologise for the confusion, but the reason why there are foci outside the marked cells in control embryos is because these are wholemount embryonic stainings and the anti-Plp antibody marks all centrosomes in all cells in the embryo, not just in the neurons.

l.270 and Fig. 5A and Fig.5 C-E. Why the AA localize Cp309 and not Asl (Fig. 5A) to detect centrioles?

In a new supplementary figure, we now show that Asl and Plp colocalize and quantify the number of times we find this colocalization in neurons (Supl. Fig 3). So, we can use CP309 in neurons to the same effect as Asl-

L295-296. I cannot see Mts, but only a diffuse staining. I am expecting to see distinct Mt bundles.

In figure 5 it is now easier to see the MT bundles in the new experiment in Fig. 5F-I , where we performed MT depolymerisation/repolymerisation: Nevertheless, we need to stress out that we are doing these analyses in wholemount embryonic stainings.

326-327. How the AA explain this different lethality, even if both the proteins are involved in centriole assembly?

We have now redone all the viability and mutant phenotype analysis using Sas-6 CRISPR mutant over the Deficiency, which is a better way to access the phenotype.

335-337. In my opinion the quoted publications are not relevant.

We believe that these references back up our hypothesis because:

• Metzger et al 2012 stress the importance of nuclear position in muscle development in Drosophila
• Loh et al 2023, relate centrosomes with nuclear migration in Drosophila
• Tillery et al 2018, is a review describing MTs in muscle development in Drosophila.

358-359. Does maternal contribution persist after gastrulation?

While bulk degradation occurs by midblastula transition, some stable maternal products persist beyond gastrulation. In our case, if centrioles are formed due to the maternal contribution, they will only be diluted by cell division, which explains why we can detect centrioles at late embryonic stages.

l.366. This is an intriguing point, but as previously observed I have some problem with centriole localization. References. Please uniform Journal abbreviations and control page numbers.

I hope we have clarified this problem with the new experiments showing MT repolarization from the centrosomes in neurons.

Reviewer #3 (Significance (Required)):

The manuscript is potentially interesting for peoples working of cell and molecular biology, and development. However, the paper needs an additional working to be suitable for publication.

We hope that reviewer 3, considers that the additional work and revision make this manuscript suitable for publication.

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

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

The manuscript of González et al. entitled "Centriole Loss in Embryonic Development Disrupts Axonal Pathfinding and Muscle Integrity" deals with the role of centrosomes in shaping axonal morphology. To this aim the AA analysed Drosophila Sas-4 mutants that are reported to develop until adult stage without centrioles. Remarkably, the AA observe that 50% of the homozygous mutant embryos fail to hatch as larvae. The present observations suggest that centrosome loss results in axonemal shaping defects and muscle developmental abnormalities. Finally, the AA show the presence of functional centrosomes in neurons. In my opinion, the manuscript is interesting because shows unexpected findings. However, to justify these new findings the AA are required to improve some experimental observations.

Major:

Abstract- It is unclear in which phenotypic condition the observations of centrosome loss or centrosome presence have been found. Please better explain. l.36. embryos, larvae, adult, from Sas4 or controls? If mutants, the observations are very interesting since Sas4 would be without centrioles. Indeed, Basto et al., show that chemosensory neurons do not develop an axoneme in the absence of centrioles, but extend dendrites toward the sensory bristle.

I do not think appropriate the use of "centriole" in the main title since the centrioles would be localized by true centriolar antigens rather than by centrosomal antigens. This problem occurs throughout the text and some figures where the AA image centrioles by centrosomal material. In Gig. 5A only the AA properly look at Asl localization. The other pictures of presumptive centrioles or centriole quantification report CP309 dots. This localization does not unequivocally reveal centrioles, since CP309 is essentially required for centrosome-mediated Mt nucleation. There are differentiated Drosophila tissues in which centrioles are present, but inactivated, and unable to recruit pericentriolar material. Mt are nucleated by ncMTOCs that contain centrosomal material and gamma-tubulin. Thus, the centrosomal antigens do not colocalize with centrioles.

Minor:

l. 58. The early arrest is mainly due to a checkpoint control. In double mutant for Sas4 and P53 the embryos survive longer, even if their further development is asrrested.

l. 102. Previous works, also quoted by the AA, reported that in mature neurons the centrosome are inactivated, whereas the present manuscript describes functional centrosomes in Drosophila motor and peripheral nervous system. This is an intriguing observations that needs a better explanation in Discussion section.

l.143-145. I understand that 50% of the Sas4 embryos that reach the adult stage have centrioles. Is it correct? But if it is so, how the AA explain the absence of centrioles in sensory neurons of adult flies as reported by Basto et al. ?

l.215. It is unclear for me why the AA analyse Sas6 flies, unless explain the mutant phenotype.

l.221. How the centrioles have been quantified? What antibody, the AA used.

l.244. and Fig 4C,D. I see high background with CP309. As reported previously I think better to use antibodies against centriolar proteins, such as Sas6, Ana1, Asl, or Sas4 ( if centrioles are present in 50% of mutants as the AA claim, the antibody could be also useful). In addition, I can see some CP309 spots in Fig 4E,F. Are they centrioles?

l.270 and Fig. 5A and Fig.5 C-E. Why the AA localize Cp309 and not Asl (Fig. 5A) to detect centrioles?

L295-296. I cannot see Mts, but only a diffuse staining. I am expecting to see distinct Mt bundles.

L. 326-327. How the AA explain this different lethality, even if both the proteins are involved in centriole assembly?

l. 335-337. In my opinion the quoted publications are not relevant.

l. 358-359. Does maternal contribution persist after gastrulation?

l.366. This is an intriguing point, but as previously observed I have some problem with centriole localization.

References. Please uniform Journal abbreviations and control page numbers.

Significance

The manuscript is potentially interesting for peoples working of cell and molecular biology, and development. However, the paper needs an additional working to be suitable for publication.

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

Learn more at Review Commons

Referee #2

Evidence, reproducibility and clarity

Summary: In this manuscript, Gonzalez et al. examine the potential function of centrosomes in the neurons and muscle cells of Drosophila embryos. By studying various mutant and RNAi lines in which centriole duplication has been disrupted, they conclude that the loss of centrioles disrupts axonal pathfinding and muscle integrity.

Major points:

1. Throughout the manuscript, the phenotypes presented are often quite subtle. For this reason, I would really recommend that these experiments are scored blind. Perhaps the authors did this, but I didn't see any mention of this.
2. The authors conclude that neurons have active centrioles that function as centrosomes (Figure 6), but the data here is confusing. The authors state that in these cells they observe acetylated MTs extending from the centrosomes and these colocalised with g-tubulin. But the authors don't show the overlap between centrosomes, g-tubulin and MTs, as they stain for these separately. This is problematic, as it was not clear from these images that the majority of the MTs really are extending from the centrosome: the centrosome may just associate or be close by to these MT cables (Figure 6A,B). Moreover, the authors show that only a fraction of the centrosomes in these cells associate with g-tubulin, so presumably in cells where the centrosomes lack g-tubulin they would not expect the centrosomes to be associated with the MTs-but they do not show that this is the case. Perhaps the authors can't test this, but an alternative would be to show that these MT arrays are absent in Sas-4 mutants. This would give more confidence that these MTs arise from the centrosomes.
3. The authors show that muscle cell integrity is compromised by centriole-loss (Figure 2). This is very surprising as it is widely believed that centrosomes are non-functional in muscle cells, and the MTs are instead organised around the nuclear envelope. I'm not aware of the situation in Drosophila muscle cells, but the authors should ideally try to examine if the centrioles are functioning as centrosomes in these cells. At the very least they should discuss how they think centriole-loss is influencing the muscle integrity when it is widely believed they are inactive in these cells.
4. Regardless of the strength of the supporting data, I think the authors should tone down their conclusions. The title and abstract led me to believe that centriole loss would cause significant problems in axonal pathfinding and muscle integrity. In all the mutant specimens examined (and certainly the low magnification views shown in Figure 1D'-F', Figure 1I'-K' and Figure 2D'-F') the mutants look very similar to the WT. Many readers may not get past the title and abstract, so the authors should make it clearer that these defects are very subtle.

Minor points:

1. In Figures 4 and 5, CP309 staining is relied on to identify centrioles, but there is quite a background of non-specific dots, making it hard to be certain what is a centriole and what isn't. For example, in Figure 5D' there are lots of dots within some of the cells - are any of these centrioles? How can the authors be certain which dot is a centriole in some of the cells shown in Figure 5C'? Is it possible to use a second marker and only count as centrioles dots that are recognised by both antibodies?
2. In the abstract that authors state that traditionally centrosomes have been considered to be non-essential in terminally differentiated cells. I don't think this is correct. In the standard "textbook" view of a cell, the centrosome is normally positioned in the centre of the cell organising an extensive array of MTs that are thought play an important role in organising intracellular transport, the positioning and movement of organelles and the maintenance and establishment of cell polarity. I don't think it is only recent evidence that suggests they play vital roles in terminally differentiated cells.
3. Line 162 the authors state that in the RNAi knockdown lines they observe several additional phenotypes, but then in the same sentence (Line 164) they say that these defects were also observed in the original mutant and mutant/Df lines.
4. The sentences in Line281-287 don't reference any of the Figures, so it seems the authors are just stating these results without presenting any data (e.g. "Significantly, we also found a correlation between axonal guidance phenotypes and the numbers of centrioles per embryo". If they've tested this correlation, they should show it.
5. In Figure 7 I did not understand how the authors measured tortuosity (wiggliness) and could see no description in the methods. This is important as, again the defect seems quite subtle, but perhaps I am not understanding which bits of the axon are being measures. Is it just the small bit of the axons close to the asterixis that is being measured, or the whole FasII track?

Significance

The potential function of centrosomes in axonal outgrowth is quite controversial, so this study is potentially of considerable interest.

However, several aspects of the data presented here were confusing or not terribly convincing. In its present state, I don't think the main conclusions are strongly enough supported by the data.

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

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

In this paper, the authors address the important question of the role of centrosomes during neuronal development. They use Drosophila as an in vivo model. The field is somewhat unclear on the role and importance of centrosomes during neuronal development, although the current data would suggest they are dispensable for axon specification and growth. Early studies in cultured mammalian neurons showed that centrosomes are active and that their microtubules can be cut and transported into the neurites. But a study then showed that centrosomes in these cultured neurons are deactivated relatively early during neuronal development in vitro and that ablating centrosomes even when they are active had no obvious effect on axon specification and growth. Consistent with this, a study in Drosophila provided evidence that centrosomes were not active or necessary in different types of neurons. More recently, a study showed that centrosomal microtubules are dispensable for axon specification and growth in mice in vivo but are required for neuronal migration in the cerebral cortex. However, another study has linked the generation of acetylated microtubules at centrosomes with axon development. In this current study, the authors examine the effect of centrosome loss on various motor and sensory neurons and muscles mainly by examining mutants in essential centriole duplication genes. They associate axonal routing and morphology defects with centrosome loss and provide some evidence that centrosomes could still be active in the developing neurons. Overall, they conclude that centrosomes are active during at least early neuronal development and that this activity is important for proper axonal morphology and routing.

While I think this study addressing a very interesting and important question, I think as it stands the data is not sufficient to be conclusive on a role for centrosomes during neuronal development. My biggest concern is that most phenotypes have not yet been shown to be cell autonomous, as whole animal mutants have been analysed rather than analysing the effect of cell-specific depletion, and the evidence for active centrosomes needs to be strengthened. If the authors can provide stronger evidence for a role of centrosomes in axonal development then the paper will certainly be of interest to a broad readership.

Major comments

1. The sas-6 transallelic combination shows only 17% embryonic lethality compared to 50% embryonic lethality with sas-4 mutants. Given that both mutants should result in the same degree of centrosome loss (this should be quantified in sas-6 mutants) it would suggest that either sas-4 has other roles away from centrosomes or that the sas-4 mutant chromosome used in the experiment has other mutations that affect viability. The effect of picking up "second-site lethal" mutations on mutant chromosomes is common and so I would not be surprised if this is the reason for the difference in phenotypes. This can be addressed either by "cleaning up" the sas-4 mutant chromosome by backcrossing to wild-type lines, allowing recombination to occur and replace the potential second site mutations, or by using transallelic combinations of sas-4, as they did for sas-6. The "easier" option may just be to analyse all the phenotypes with the sas-6 transallelic combination.
2. Using "whole animal" mutants for assessing neuronal morphology is risky due to non-cell-autonomous effects. The authors have carried out some phenotypic analysis of neurons depleted of Sas-4 by cell-specific RNAi, but I feel they need to do this for all of their analysis. This includes embryonic lethality measures, quantification of centrosome numbers, and all axonal phenotypes in Sas-4 RNAi neurons. It would also be prudent to use 2 distinct RNAi lines to help ensure any phenotypes are not off-target effects (and this may help clarify why the authors see some additional phenotypes with RNAi). Indeed, there are relatively weak phenotypes in muscles when using RNAi compared to the mutants and these potential non-cell-autonomous effects could then have a knock-on effect on neuronal morphology. If the authors were concerned that RNAi is not very efficient (explaining any potential weaker phenotypes than in mutants) the authors could examine the effectiveness of RNAi lines by analysing protein depletion by western blotting or mRNA depletion by rt-qPCR (although this has to be done in a different cell type due to the difficulty in obtaining a neuronal extract).
3. When analysing centriole presence or absence it is a good idea to stain with two different centriole markers e.g. Asl and Plp. This helps rule out unspecific staining. It is clear from the images that similar sized foci can be observed outside of the cells (see Figure 5A for example), so clearly some of the foci that appear to be within the cells may also be unspecific staining.
4. The evidence for active centrosomes is not that convincing. Acetylated tubulin is associated with stable MTs, which are not normally organised by "active" centrosomes that nucleate dynamic microtubules. Moreover, it is plausible that centriole foci happen to overlap with the acetylated tubulin staining by chance. This would explain why not all centrosomes colocalise with acetylated tubulin signal. The authors could better test centrosome activity by performing live imaging with EB1-GFP. If centrosomes are active, it is very easy to observe the many comets produced by the centrosomes.
5. If the authors believe that centrosomes have a role in axon pathfinding in sensory neurons, they should show that these centrosomes are active, at least during early stages (again using EB1-GFP imaging).
6. The authors mention in the discussion that "increased JNK activity, can result in axonal wiggliness (Karkali et al, 2023)". I therefore wonder whether centrosome loss may induce JNK activation (the stress response), as this would then indicate an indirect effect of centrosome loss on axonal structure rather than a direct influence of centrosome-generated microtubules. The authors could assess whether the DNK-JNK pathway is activated in neurons lacking centrosomes by expression UAS-Puc-GFP and quantifying the nuclear signal.
7. In Figure 5, the authors claim that they find "a correlation between axonal guidance phenotypes and the numbers of centrioles per embryo". I don't think this is a strong correlation. The difference in centriole number between embryos with no defects and those with defects is very small. In contrast, the difference between centriole numbers in control (no defects) and mutant (no defects) is very large. So, there does not appear to be a strong correlation between centrosome number and phenotype.

Minor comments

1. I don't understand Figure 3C - why do the % of surviving homozygotes and heterozygotes add up to 100%? Should the grey boxes not relate to dead and the white to surviving?
2. "In mouse fibroblasts, myoblasts and endothelial cells, centrosome orientation is important for nuclear positioning and cell migration(Chang et al, 2015; Gomes et al, 2005; Kushner et al, 2014)." Do you mean "centrosome position"?
3. In the introduction, the authors mention Meka et al. when saying the centrosomal microtubules are important for axonal development, but they should also discuss the counter argument from Vinopal et al., 2023 (Neuron) that showed how centrosomes were required for neuronal migration but not axon growth, which was instead mediated by Golgi-derived microtubules.
4. Lines 228-230 - repeated sentence
5. Additionally, we did not detect centrioles in the quadrant opposite the axon exit point (Fig. 2B n=75) - this data is not in Fig 2B
6. "This significant decrease in the humber of centrioles further supports the critical role of Sas-4 in pioneer neurons of the ventral nerve cord (VNC) during Drosophila embryogenesis". It rather highlights that Sas-4 is required for centriole formation in these neurons. Also, humber = number.
7. Result title: Non-ciliated sensory neurons have centrioles. This is kind of obvious. A better title may be "axon phenotypes correlate with centriole numbers in sensory neurons" but unfortunately i don't think there is good evidence for this (See major point above).

Significance

As mentioned above, the advance will be important if more evidence is provided. In this case, the paper will be interesting to a broad readership. But currently the paper is limited by the lack of evidence for centrosome function and activity in the neurons.

Reviewer #1 (Public Review):

Summary:

Argunşah et al. describe and investigate the mechanisms underlying the differential response dynamics of barrel vs septa domains in the whisker-related primary somatosensory cortex (S1). Upon repeated stimulation, the authors report that the response ratio between multi- and single-whisker stimulation increases in layer (L) 4 neurons of the septal domain, while remaining constant in barrel L4 neurons. The authors attribute this divergence to differences in short-term synaptic plasticity, particularly within somatostatin-expressing (SST⁺) interneurons. This interpretation is supported by 1) the increased density of SST+ neurons in L4 of the septa compared to barrel domain, 2) the stronger response of (L2/3) SST+ neurons to repeated multi- vs single-whisker stimulation and 3) the reduced functional difference in single- versus multi-whisker response ratios across barrel and septal domains in Elfn1 KO mice, which lack a synaptic protein that confers characteristic short-term plasticity, notably in SST+ neurons. Consistently, a decoder trained on WT data fails to generalize to Elfn1 KO responses. Finally, the authors report a relative enrichment of S2- and M1-projecting cell densities in L4 of the septal domain compared to the barrel domain, suggesting that septal and barrel circuits may differentially route information about single vs multi-whisker stimulation downstream of S1.

Strengths:

This paper describes and aims to study a circuit underlying differential response between barrel columns and septal domains of the primary somatosensory cortex. This work supports the view these two domains contribute distinctly to the processing single versus multi-whisker inputs and highlight the role of SST+ neuron and their short-term plasticity. Together, this study suggests that the barrel cortex multiplexes whisker-derived sensory information across its domains, enabling parallel processing within S1.

Weaknesses:

Although the divergence in responses to repeated single- versus multi-whisker stimulation between barrel and septal domains is consistent with a role for SST⁺ neuron short-term plasticity, the evidence presented does not conclusively demonstrate that this mechanism is the critical driver of the difference. The lack of targeted recordings and manipulations limits the strength of this conclusion: SST⁺ neuron activity is not measured in L4, nor is it assessed in a domain-specific manner. The Elfn1 knockout manipulation does not appear to selectively affect either stimulus condition, domain or interneuron subtype. Finally, all experiments were performed under anesthesia, which raises concerns about how well the reported dynamics generalize to awake cortical processing.

Reviewer #2 (Public review):

Summary:

Argunsah and colleagues demonstrate that SST expressing interneurons are concentrated in the mouse septa and differentially respond to repetitive multi-whisker inputs. Identifying how a specific neuronal phenotype impacts responses is an advance.

Strengths:

(1) Careful physiological and imaging studies.

(2) Novel result showing the role of SST+ neurons in shaping responses.

(3) Good use of a knockout animal to further the main hypothesis.

(4) Clear analytical techniques.

Comments on revisions:

The authors have effectively responded to my initial critiques - I have no further concerns.

Reviewer #3 (Public review):

Summary:

This study investigates the functional differences between barrel and septal columns in the mouse somatosensory cortex, focusing on how local inhibitory dynamics (particularly involving SST⁺ interneurons) may mediate temporal integration of multi-whisker (MW) stimuli in septa. Using a combination of in vivo multi-unit recordings, calcium imaging, and anatomical tracing, the authors propose a model in which Elfn1-dependent synaptic facilitation onto SST⁺ interneurons contributes to the distinct sensory responses to MW input in barrels and septa, enabling functional segregation between these domains.

Strengths:

The study presents a thought-provoking and useful conceptual model for understanding sensory processing in the somatosensory cortex. While barrel columns have been widely studied, septal regions remain relatively understudied in mice. If septa indeed act as selective integrators of distributed sensory input, this would suggest a novel computational role for cortical microcircuits beyond the classical view focused on barrels. Although still hypothetical, the proposed model in which SST⁺ interneurons contribute to domain-specific sensory responses between barrel and septal domains is intriguing and opens new avenues for investigating inhibitory circuit mechanisms.

Weaknesses:

The primary limitation of this study lies in the spatial and cellular specificity of the recording techniques. The physiological data rely predominantly on unsorted multi-unit activity (MUA) recorded with low-channel-count silicon probes. Because MUA aggregates signals from multiple neurons over a radius of approximately 50-100 µm (often wider than the typical septal width in mice), this approach makes it difficult to confidently isolate activity originating strictly from within septal domains. The manuscript would benefit from additional analyses to validate the spatial specificity of these recordings, such as systematically varying spike detection thresholds to test the robustness of domain attribution, as suggested by the reviewer. Furthermore, although the authors now appropriately frame their findings in the Elfn1 knockout mice as indirect evidence, it is worth emphasizing that the study lacks direct in vivo, cell-type-specific recordings and manipulations to more definitively test the proposed mechanism.

Author response:

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

Public Reviews:

Reviewer #1 (Public Reviews):

Summary:

Argunşah et al. describe and investigate the mechanisms underlying the differential response dynamics of barrel vs septa domains of the whisker-related primary somatosensory cortex (S1). Upon repeated stimulation, the authors report that the response ratio between multi- and single-whisker stimulation increases in layer (L) 4 neurons of the septal domain, while remaining constant in barrel L4 neurons. This difference is attributed to the short-term plasticity properties of interneurons, particularly somatostatin-expressing (SST+) neurons. This claim is supported by the increased density of SST+ neurons found in L4 of the septa compared to barrels, along with a stronger response of (L2/3) SST+ neurons to repeated multi- vs single-whisker stimulation. The role of the synaptic protein Elfn1 is then examined. Elfn1 KO mice exhibited little to no functional domain separation between barrel and septa, with no significant difference in single- versus multi-whisker response ratios across barrel and septal domains. Consistently, a decoder trained on WT data fails to generalize to Elfn1 KO responses. Finally, the authors report a relative enrichment of S2- and M1-projecting cell densities in L4 of the septal domain compared to the barrel domain.

Strengths:

This paper describes and aims to study a circuit underlying differential response between barrel columns and septal domains of the primary somatosensory cortex. This work supports the view that barrel and septal domains contribute differently to processing single versus multi-whisker inputs, suggesting that the barrel cortex multiplexes sensory information coming from the whiskers in different domains.

We thank the reviewer for the very neat summary of our findings that barrel cortex multiplexes converging information in separate domains.

Weaknesses:

While the observed divergence in responses to repeated SWS vs MWS between the barrel and septal domains is intriguing, the presented evidence falls short of demonstrating that short-term plasticity in SST+ neurons critically underpins this difference. The absence of a mechanistic explanation for this observation limits the work’s significance. The measurement of SST neurons’ response is not specific to a particular domain, and the Elfn1 manipulation does not seem to be specific to either stimulus type or a particular domain.

We appreciate the reviewer’s perspective. Although further research is needed to understand the circuit mechanisms underlying the observed phenomenon, we believe our data suggest that altering the short-term dynamics of excitatory inputs onto SST neurons reduces the divergent spiking dynamics in barrels versus septa during repetitive single- and multi-whisker stimulation. Future work could examine how SST neurons, whose somata reside in barrels and septa, respond to different whisker stimuli and the circuits in which they are embedded. At this time, however, the authors believe there is no alternative way to test how the short-term dynamics of excitatory inputs onto SST neurons, as a whole, contribute to the temporal aspects of barrel versus septa spiking.

The study's reach is further constrained by the fact that results were obtained in anesthetized animals, which may not generalize to awake states.

We appreciate the reviewer’s concern regarding the generalizability of our findings from anesthetized animals to awake states. Anesthesia was employed to ensure precise individual whisker stimulation (and multi-whisker in the same animal), which is challenging in awake rodents due to active whisking. While anesthesia may alter higher-order processing, core mechanisms, such as short and long term plasticity in the barrel cortex, are preserved under anesthesia (Martin-Cortecero et al., 2014; Mégevand et al., 2009).

The statistical analysis appears inappropriate, with the use of repeated independent tests, dramatically boosting the false positive error rate.

Thank you for your feedback on our analysis using independent rank-based tests for each time point in wild-type (WT) animals. To address concerns regarding multiple comparisons and temporal dependencies (for Figure 1F and 4D for now but we will add more in our revision), we performed a repeated measures ANOVA for WT animals (13 Barrel, 8 Septa, 20 time points), which revealed a significant main effect of Condition (F(1,19) = 16.33, p < 0.001) and a significant Condition-Time interaction (F(19,361) = 2.37, p = 0.001). Post-hoc tests confirmed significant differences between Barrel and Septa at multiple time points (e.g., p < 0.0025 at times 3, 4, 6, 7, 8, 10, 11, 12, 16, 19 after Bonferroni posthoc correction), supporting a differential multi-whisker vs. single-whisker ratio response in WT animals. In contrast, a repeated measures ANOVA for knock-out (KO) animals (11 Barrel, 7 Septa, 20 time points) showed no significant main effect of Condition (F(1,14) = 0.17, p = 0.684) or Condition-Time interaction (F(19,266) = 0.73, p = 0.791), indicating that the BarrelSepta difference observed in WT animals is absent in KO animals.

Furthermore, the manuscript suffers from imprecision; its conclusions are occasionally vague or overstated. The authors suggest a role for SST+ neurons in the observed divergence in SWS/MWS responses between barrel and septal domains. However, this remains speculative, and some findings appear inconsistent. For instance, the increased response of SST+ neurons to MWS versus SWS is not confined to a specific domain. Why, then, would preferential recruitment of SST+ neurons lead to divergent dynamics between barrel and septal regions? The higher density of SST+ neurons in septal versus barrel L4 is not a sufficient explanation, particularly since the SWS/MWS response divergence is also observed in layers 2/3, where no difference in SST+ neuron density is found.

Moreover, SST+ neuron-mediated inhibition is not necessarily restricted to the layer in which the cell body resides. It remains unclear through which differential microcircuits (barrel vs septum) the enhanced recruitment of SST+ neurons could account for the divergent responses to repeated SWS versus MWS stimulation.

We fully appreciate the reviewer’s comment. We currently do not provide any evidence on the contribution of SST neurons in the barrels versus septa in layer 4 on the response divergence of spiking observed in SWS versus MWS. We only show that these neurons differentially distribute in the two domains in this layer. It is certainly known that there is molecular and circuit-based diversity of SST-positive neurons in different layers of the cortex, so it is plausible that this includes cells located in the two domains of vS1, something which has not been examined so far. Our data on their distribution are one piece of information that SST neurons may have a differential role in inhibiting barrel stellate cells versus septa ones. Morphological reconstructions of SST neurons in L4 of the somatosensory barrel cortex has shown that their dendrites and axons project locally and may confine to individual domains, even though not specifically examined (Fig. 3 of Scala F et al., 2019). The same study also showed that L4 SST cells receive excitatory input from local stellate cells) and is known that they are also directly excited by thalamocortical fibers (Beierlein et al., 2003; Tan et al., 2008), both of which facilitate.

As shown in our supplementary figure, the divergence is also observed in L2/3 where, as the reviewer also points out, where we do not have a differential distribution of SST cells, at least based on a columnar analysis extending from L4. There are multiple scenarios that could explain this “discrepancy” that one would need to examine further in future studies. One straightforward one is that the divergence in spiking in L2/3 domains may be inherited from L4 domains, where L4 SST act on. Another is that even though L2/3 SST neurons are not biased in their distribution their input-output function is, something which one would need to examine by detailed in vitro electrophysiological and perhaps optogenetic approaches in S1. Despite the distinctive differences that have been found between the L4 circuitry in S1 and V1 (Scala F et al., 2019), recent observations indicate that small but regular patches of V1 marked by the absence of muscarinic receptor 2 (M2) have high temporal acuity (Ji et al., 2015), and selectively receive input from SST interneurons (Meier et al., 2025). Regions lacking M2 have distinct input and output connectivity patterns from those that express M2 (Meier et al., 2021; Burkhalter et al., 2023). These findings, together with ours, suggest that SST cells preferentially innervate and regulate specific domains columns- in sensory cortices.

Regardless of the mechanism, the Elfn1 knock-out mouse line almost exclusively affects the incoming excitability onto SST neurons (see also reply to comment below), hence what can be supported by our data is that changing the incoming short-term synaptic plasticity onto these neurons brings the spiking dynamics between barrels and septa closer together.

The Elfn1 KO mouse model seems too unspecific to suggest the role of the short-term plasticity in SST+ neurons in the differential response to repeated SWS vs MWS stimulation across domains. Why would Elfn1-dependent short-term plasticity in SST+ neurons be specific to a pathway, or a stimulation type (SWS vs MWS)? Moreover, the authors report that Elfn1 knockout alters synapses onto VIP+ as well as SST+ neurons (Stachniak et al., 2021; previous version of this paper)-so why attribute the phenotype solely to SST+ circuitry? In fact, the functional distinctions between barrel and septal domains appear largely abolished in the Elfn1 KO.

Previous work by others and us has shown that globally removing Elfn1 selectively removes a synaptic process from the brain without altering brain anatomy or structure. This allows us to study how the temporal dynamics of inhibition shape activity, as opposed to inhibition from particular cell types. We will nevertheless update the text to discuss more global implications for SST interneuron dynamics and include a reference to VIP interneurons that contain Elfn1.

When comparing SWS to MWS, we find that MWS replaces the neighboring excitation which would normally be preferentially removed by short-term plasticity in SST interneurons, thus providing a stable control comparison across animals and genotypes. On average, VIP interneurons failed to show modulation by MWS. We were unable to measure a substantial contribution of VIP cells to this process and also note that the Elfn1 expressing multipolar neurons comprise only ~5% of VIP neurons (Connor and Peters, 1984; Stachniak et al., 2021), a fraction that may be lost when averaging from 138 VIP cells. Moreover, the effect of Elfn1 loss on VIP neurons is quite different and marginal compared to that of SST cells, suggesting that the primary impact of Elfn1 knockout is mediated through SST+ interneuron circuitry. Therefore, even if we cannot rule out that these 5% of VIP neurons contribute to barrel domain segregation, we are of the opinion that their influence would be very limited if any.

Reviewer #2 (Public Reviews):

Summary:

Argunsah and colleagues demonstrate that SST-expressing interneurons are concentrated in the mouse septa and differentially respond to repetitive multi-whisker inputs. Identifying how a specific neuronal phenotype impacts responses is an advance.

Strengths:

(1)  Careful physiological and imaging studies.

(2)  Novel result showing the role of SST+ neurons in shaping responses.

(3)  Good use of a knockout animal to further the main hypothesis.

(4)  Clear analytical techniques.

We thank the reviewer for their appreciation of the study.

Weaknesses:

No major weaknesses were identified by this reviewer. Overall, I appreciated the paper but feel it overlooked a few issues and had some recommendations on how additional clarifications could strengthen the paper. These include:

(1) Significant work from Jerry Chen on how S1 neurons that project to M1 versus S2 respond in a variety of behavioral tasks should be included (e.g. PMID: 26098757). Similarly, work from Barry Connor’s lab on intracortical versus thalamocortical inputs to SST neurons, as well as excitatory inputs onto these neurons (e.g. PMID: 12815025) should be included.

We thank the reviewer for these valuable resources that we overlooked. We will include Chen et al. (2015), Cruikshank et al. (2007) and Gibson et al. (1999) to contextualize S1 projections and SST+ inputs, strengthening the study’s foundation as well as Beierlein et al. (2003) which nicely show both local and thalamocortical facilitation of excitatory inputs onto L4 SST neurons, in contrast to PV cells. The paper also shows the gradual recruitment of SST neurons by thalamocortical inputs to provide feed-forward inhibition onto stellate cells (regular spiking) of the barrel cortex L4 in rat.

(2) Using Layer 2/3 as a proxy to what is happening in layer 4 (~line 234). Given that layer 2/3 cells integrate information from multiple barrels, as well as receiving direct VPm thalamocortical input, and given the time window that is being looked at can receive input from other cortical locations, it is not clear that layer 2/3 is a proxy for what is happening in layer 4.

We agree with the reviewer that what we observe in L2/3 is not necessarily what is taking place in L4 SST-positive cells. The data on L2/3 was included to show that these cells, as a population, can show divergent responses when it comes to SWS vs MWS, which is not seen in L2/3 VIP neurons. Regardless of the mechanisms underlying it, our overall data support that SST-positive neurons can change their activation based on the type of whisker stimulus and when the excitatory input dynamics onto these neurons change due to the removal of Elfn1 the recruitment of barrels vs septa spiking changes at the temporal domain. Having said that, the data shown in Supplementary Figure 3 on the response properties of L2/3 neurons above the septa vs above the barrels (one would say in the respective columns) do show the same divergence as in L4. This suggests that a circuit motif may exist that is common to both layers, involving SST neurons that sit in L4, L5 or even L2/3. This implies that despite the differences in the distribution of SST neurons in septa vs barrels of L4 there is an unidentified input-output spatial connectivity motif that engages in both L2/3 and L4. Please also see our response to a similar point raised by reviewer 1.

(3) Line 267, when discussing distinct temporal response, it is not well defined what this is referring to. Are the neurons no longer showing peaks to whisker stimulation, or are the responses lasting a longer time? It is unclear why PV+ interneurons which may not be impacted by the Elfn1 KO and receive strong thalamocortical inputs, are not constraining activity.

We thank the reviewer for their comment and will clarify the statement.

This convergence of response profiles was further clear in stimulus-aligned stacked images, where the emergent differences between barrels and septa under SWS were largely abolished in the KO (Figure 4B). A distinction between directly stimulated barrels and neighboring barrels persisted in the KO. In addition, the initial response continued to differ between barrel and septa and also septa and neighbor (Figure 4B). This initial stimulus selectivity potentially represents distinct feedforward thalamocortical activity, which includes PV+ interneuron recruitment that is not directly impacted by the Elfn1 KO (Sun et al., 2006; Tan et al., 2008). PV+ cells are strongly excited by thalamocortical inputs, but these exhibit short-term depression, as does their output, contrasting with the sustained facilitation observed in SST+ neurons. These findings suggest that in WT animals, activity spillover from principal barrels is normally constrained by the progressive engagement of SST+ interneurons in septal regions, driven by Elfn1-dependent facilitation at their excitatory synapses. In the absence of Elfn1, this local inhibitory mechanism is disrupted, leading to longer responses in barrels, delayed but stronger responses in septa, and persistently stronger responses in unstimulated neighbors, resulting in a loss of distinction between the responses of barrel and septa domains that normally diverge over time (see Author response image 1 below).

Author response image 1.

(A) Barrel responses are longer following whisker stimulation in KO. (B) Septal responses are slightly delayed but stronger in KO. (C) Unstimulated neighbors show longer persistent responses in KO.

(4) Line 585 “the earliest CSD sink was identified as layer 4…” were post-hoc measurements made to determine where the different shank leads were based on the post-hoc histology?

Post hoc histology was performed on plane-aligned brain sections which would allow us to detect barrels and septa, so as to confirm the insertion domains of each recorded shank. Layer specificity of each electrode therefore could therefore not be confirmed by histology as we did not have coronal sections in which to measure electrode depth.

(5) For the retrograde tracing studies, how were the M1 and S2 injections targeted (stereotaxically or physiologically)? How was it determined that the injections were in the whisker region (or not)?

During the retrograde virus injection, the location of M1 and S2 injections was determined by stereotaxic coordinates (Yamashita et al., 2018). After acquiring the light-sheet images, we were able to post hoc examine the injection site in 3D and confirm that the injections were successful in targeting the regions intended. Although it would have been informative to do so, we did not functionally determine the whisker-related M1 and whisker-related S2 region in this experiment.

(6) Were there any baseline differences in spontaneous activity in the septa versus barrel regions, and did this change in the KO animals?

Thank you for this interesting question. Our previous study found that there was a reduction in baseline activity in L4 barrel cortex of KO animals at postnatal day (P)12, but no differences were found at P21 (Stachniak et al., 2023).

Reviewer #3 (Public Reviews):

Summary:

This study investigates the functional differences between barrel and septal columns in the mouse somatosensory cortex, focusing on how local inhibitory dynamics, particularly involving Elfn1-expressing SST⁺ interneurons, may mediate temporal integration of multiwhisker (MW) stimuli in septa. Using a combination of in vivo multi-unit recordings, calcium imaging, and anatomical tracing, the authors propose that septa integrate MW input in an Elfn1-dependent manner, enabling functional segregation from barrel columns.

Strengths:

The core hypothesis is interesting and potentially impactful. While barrels have been extensively characterized, septa remain less understood, especially in mice, and this study's focus on septal integration of MW stimuli offers valuable insights into this underexplored area. If septa indeed act as selective integrators of distributed sensory input, this would add a novel computational role to cortical microcircuits beyond what is currently attributed to barrels alone. The narrative of this paper is intellectually stimulating.

We thank the reviewer for finding the study intellectually stimulating.

Weaknesses:

The methods used in the current study lack the spatial and cellular resolution needed to conclusively support the central claims. The main physiological findings are based on unsorted multi-unit activity (MUA) recorded via low-channel-count silicon probes. MUA inherently pools signals from multiple neurons across different distances and cell types, making it difficult to assign activity to specific columns (barrel vs. septa) or neuron classes (e.g., SST⁺ vs. excitatory).

The recording radius (~50-100 µm or more) and the narrow width of septa (~50-100 µm or less) make it likely that MUA from "septal" electrodes includes spikes from adjacent barrel neurons.

The authors do not provide spike sorting, unit isolation, or anatomical validation that would strengthen spatial attribution. Calcium imaging is restricted to SST⁺ and VIP⁺ interneurons in superficial layers (L2/3), while the main MUA recordings are from layer 4, creating a mismatch in laminar relevance.

We thank the reviewer for pointing out the possibility of contamination in septal electrodes. Importantly, it may not have been highlighted, although reported in the methods, but we used an extremely high threshold (7.5 std, in methods, line 583) for spike detection in order to overcome the issue raised here, which restricts such spatial contaminations. Since the spike amplitude decays rapidly with distance, at high thresholds, only nearby neurons contribute to our analysis, potentially one or two. We believe that this approach provides a very close approximation of single unit activity (SUA) in our reported data. We will include a sentence earlier in the manuscript to make this explicit and prevent further confusion.

Regarding the point on calcium imaging being performed on L2/3 SST and VIP cells instead of L4. Both reviewer 1 and 2 brought up the same issue and we responded as follows. As shown in our supplementary figure, the divergence is also observed in L2/3 where we do not have a differential distribution of SST cells, at least based on a columnar analysis extending from L4. There are multiple scenarios that could explain this “discrepancy” that one would need to examine further in future studies. One straightforward one is that the divergence in spiking in L2/3 domains may be inherited from L4 domains, where L4 SST act on. Another is that even though L2/3 SST neurons are not biased in their distribution their input-output function is, something which one would need to examine by detailed in vitro electrophysiological and perhaps optogenetic approaches in S1. Despite the distinctive differences that have been found between the L4 circuitry in S1 and V1 (Scala F et al., 2019), recent observations indicate that small but regular patches of V1 marked by the absence of muscarinic receptor 2 (M2) have high temporal acuity (Ji et al., 2015), and selectively receive input from SST interneurons (Meier et al., 2025). Regions lacking M2 have distinct input and output connectivity patterns from those that express M2 (Meier et al., 2021; Burkhalter et al., 2023). These findings, together with ours, suggest that SST cells preferentially innervate and regulate specific domains -columns- in sensory cortices.

Furthermore, while the role of Elfn1 in mediating short-term facilitation is supported by prior studies, no new evidence is presented in this paper to confirm that this synaptic mechanism is indeed disrupted in the knockout mice used here.

We thank Reviewer #3 for noting the absence of new evidence confirming Elfn1’s disruption of short-term facilitation in our knockout mice. We acknowledge that our study relies on previously strong published data demonstrating that Elfn1 mediates short-term synaptic facilitation of excitatory inputs onto SST+ interneurons (Sylwestrak and Ghosh, 2012; Tomioka et al., 2014; Stachniak et al., 2019, 2023). These studies consistently show that Elfn1 knockout abolishes facilitation in SST+ synapses, leading to altered temporal dynamics, which we hypothesize underlies the observed loss of barrel-septa response divergence in our Elfn1 KO mice (Figure 4). Nevertheless, to address the point raised, we will clarify in the revised manuscript (around lines 245-247 and 271-272) that our conclusions are based on these established findings, stating: “Building on prior evidence that Elfn1 knockout disrupts short-term facilitation in SST+ interneurons (Sylwestrak and Ghosh, 2012; Tomioka et al., 2014; Stachniak et al., 2019, 2023), we attribute the abolished barrel-septa divergence in Elfn1 KO mice to altered SST+ synaptic dynamics, though direct synaptic measurements were not performed here.”

Additionally, since Elfn1 is constitutively knocked out from development, the possibility of altered circuit formation-including changes in barrel structure and interneuron distribution, cannot be excluded and is not addressed.

We thank Reviewer #3 for raising the valid concern that constitutive Elfn1 knockout could potentially alter circuit formation, including barrel structure and interneuron distribution. To address this, we will clarify in the revised manuscript (around line ~271 and in the Discussion) that in our previous studies that included both whole-cell patch-clamp in acute brain slices ranging from postnatal day 11 to 22 (P11 - P21) and in vivo recordings from barrel cortex at P12 and P21, we saw no gross abnormalities in barrel structure, with Layer 4 barrels maintaining their characteristic size and organization, consistent with wildtype (WT) mice (Stachniak et al., 2019, 2023). While we cannot fully exclude subtle developmental changes, prior studies indicate that Elfn1 primarily modulates synaptic function rather than cortical cytoarchitecture (Tomioka et al., 2014). Elfn1 KO mice show no gross morphological or connectivity differences and the pattern and abundance of Elfn1 expressing cells (assessed by LacZ knock in) appears normal (Dolan and Mitchell, 2013).

We will add the following to the Discussion: “Although Elfn1 is constitutively knocked out, we find here and in previous studies that barrel structure is preserved (Stachniak et al., 2019, 2023). Further, the distribution of Elfn1 expressing interneurons is not different in KO mice, suggesting minimal developmental disruption (Dolan and Mitchell, 2013).

Nonetheless, we acknowledge that subtle circuit changes cannot be ruled out without the usage of time-depended conditional knockout of the gene.”

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors): 

(1) My biggest concern is regarding statistics. Did the authors repeatedly apply independent tests (Mann-Whitney) without any correction for multiple comparisons (Figures 1 and 4)? In that case, the chances of a spurious "significant" result rise dramatically. 

In response to the reviewer’s comment, we now present new statistical results by utilizing ANOVA and blended these results in the manuscript between lines 172 and 192 for WT data and 282 and 298 for Elfn1 KO data. This new statistical approach shows the same differences as we had previously reported, hence consolidating the statements made. 

(2) The findings only hint at a mechanism involving SST+ neurons for how SWS and MWS are processed differently in the barrel vs septal domains. As a direct test of SST+ neuron involvement in the divergence of barrel and septal responses, the authors might consider SST-specific manipulations - for example, inhibitory chemo- or optogenetics during SWS and MWS stimulation.

We thank the reviewer for this comment and agree that a direct manipulation of SST+ neurons via inhibitory chemo- or opto-genetics could provide further supporting evidence for the main claims in our study. We have opted out from performing these experiments for this manuscript as we feel they can be part of a future study.  At the same time, it is conceivable that such manipulations and depending on how they are performed may lead to larger and non-specific effects on cortical activity, since SST neurons will likely be completely shut down. So even though we certainly appreciate and value the strengths of such approaches, our experiments have addressed a more nuanced hypothesis, namely that the synaptic dynamics onto SST+ neurons matter for response divergence of septa versus barrels, which could not have been easily and concretely addressed by manipulating SST+ cell firing activity.  

(3) In general, it is hard to comprehend what microcircuit could lead to the observed divergence in the MWS/SWS ratio in the barrel vs septal domain. There preferential recruitment of SST+ neurons during MWS is not specific to a particular domain, and the higher density of SST+ neurons specifically in L4 septa cannot per se explain the diverging MWS/SWS ratio in L4 septal neurons since similar ratio divergence is observed across domains in L2/3 neurons without increase SST+ neuron density in L2/3. This view would also assume that SST+ inhibition remains contained to its own layer and domain. Is this the case? Is it that different microcircuits between barrels and septa differently shape the response to repeated MWS? This is partially discussed in the paper; can the authors develop on that? What would the proposed mechanism be? Can the short-term plasticity of the thalamic inputs (VPM vs POm) be part of the picture?

We thank the reviewer for raising this important point. We propose that the divergence in MWS/SWS ratios across barrel and septal domains arises from dynamic microcircuit interactions rather than static anatomical features such as SST+ density, which we describe and can provide a hint. In L2/3, where SST+ density is uniform, divergence persists, suggesting that trans-laminar and trans-domain interactions are key. Barrel domains, primarily receiving VPM inputs, exhibit short-term depression onto excitatory cells and engage PV+ and SST+ neurons to stabilize the MWS/SWS ratio, with Elfn1-dependent facilitation of SST+ neurons gradually increasing inhibition during repetitive SWS. Septal domains, in contrast, are targeted by facilitating POm inputs, combined with higher L4 SST+ density and Elfn1-mediated facilitation, producing progressive inhibitory buildup that amplifies the MWS/SWS ratio. SST+ projections in septa may extend trans-laminarly and laterally, influencing L2/3 and neighboring barrels, thereby explaining L2/3 divergence despite uniform SST+ density in L2/3. In this regards, direct laminar-dependent manipulations will be required to confirm whether L2/3 divergence is inherited from L4 dynamics. In Elfn1 KO mice, the loss of facilitation in SST+ neurons likely flattens these dynamics, disrupting functional segregation. Future experiments using VPM/POm-specific optogenetic activation and SST+ silencing will be critical to directly test this model.

We expanded the discussion accordingly.

(4) Can the decoder generalize between SWS and MWS? In this condition, if the decoder accuracy is higher for barrels than septa, it would support the idea that septa are processing the two stimuli differently. 

Our results show that septal decoding accuracy is generally higher than barrel accuracy when generalizing from multi-whisker stimulation (MWS) to single-whisker stimulation (SWS), indicating distinct information processing in septa compared to barrels.

In wild-type (WT) mice, septal accuracy exceeds barrel accuracy across all time windows (150ms, 51-95ms, 1-95ms), with the largest difference in the 51-95ms window (0.9944 vs. 0.9214 at pulse 20, 10Hz stimulation). This septal advantage grows with successive pulses, reflecting robust, separable neural responses, likely driven by the posterior medial nucleus (POm)’s strong MWS integration contrasting with minimal SWS activation. Barrel responses, driven by consistent ventral posteromedial nucleus (VPM) input for both stimuli, are less distinguishable, leading to lower accuracy.

In Elfn1 knockout (KO) mice, which disrupt excitatory drive to somatostatin-positive (SST+) interneurons, barrel accuracy is higher initially in the 1-50ms window (0.8045 vs. 0.7500 at pulse 1), suggesting reduced early septal distinctiveness. However, septal accuracy surpasses barrels in later pulses and time windows (e.g., 0.9714 vs. 0.9227 in 51-95ms at pulse 20), indicating restored septal processing. This supports the role of SST+ interneurons in shaping distinct MWS responses in septa, particularly in late-phase responses (51-95ms), where inhibitory modulation is prominent, as confirmed by calcium imaging showing stronger SST+ activation during MWS.

These findings demonstrate that septa process SWS and MWS differently, with higher decoding accuracy reflecting structured, POm- and SST+-driven response patterns. In Elfn1 KO mice, early deficits in septal processing highlight the importance of SST+ interneurons, with later recovery suggesting compensatory mechanisms. 

We have added Supplementary Figure 4 and included this interpretation between lines 338353. 

We thank the reviewer for suggesting this analysis.

(5) It is not clear to me how the authors achieve SWS. How is it that the pipette tip "placed in contact with the principal whisker" does not detach from the principal whisker or stimulate other whiskers? Please clarify the methods. 

Targeting the specific principal whisker is performed under the stereoscope.  

Specifically, we have added this statement in line 628:

“We trimmed the whiskers where necessary, to avoid them touching each other and to avoid stimulating other whiskers. By putting the pipette tip very close (almost touching) to the principal whisker, the movement of the tip (limited to 1mm) would reliably move the targeted whisker. The specificity of the stimulation of the selected principal whisker was observed under the stereoscope.”

(6) The method for calculating decoder accuracy is not clearly described-how can accuracy exceed 1? The authors should clarify this metric and provide measures of variability (e.g., confidence intervals or standard deviations across runs) to assess the significance of their comparisons. Additionally, using a consistent scale across all plots would improve interoperability. 

We thank the reviewer for raising this point. We have now changed the way accuracies are calculated and adopted a common scale among different plots (see updated Figure 5). We have also changed the methods section accordingly.

(7) Figure 1: The sample size is not specified. It looks like the numbers match the description in the methods, but the sample size should be clearly stated here. 

These are the numbers the reviewer is inquiring about. 

WT: (WT) animals: a 280 × 95 × 20 matrix for the stimulated barrel (14 Barrels, 95ms, 20 pulses), a 180 × 95 × 20 matrix for the septa (9 Septa, 95ms, 20 pulses), and a 360 × 95 × 20 matrix for the neighboring barrel (18 Neighboring barrels, 95ms, 20 pulses). N=4 mice.

KO: 11-barrel columns, 7 septal columns, 11 unstimulated neighbors from N=4 mice.

Panels D-F are missing axes and axis labels (firing rate, p-value). Panel D is mislabeled (left, middle, and right). I can't seem to find the yellow line. 

Thank you for this observation. We made changes in the figures to make them easier to navigate based on the collective feedback from the reviewers.

Why is changing the way to compare the differences in the responses to repeated stimulation between SWS and MWS? 

To assess temporal accumulation of information, we compared responses to repeated single-whisker stimulation (SWS) and multi-whisker stimulation (MWS) using an accumulative decoding approach rather than simple per-pulse firing rates. This method captures domain-specific integration dynamics over successive pulses.

The use of the term "principal whisker" is confusing, as it could refer to the whisker that corresponds to the recorded barrel. 

When we use the term principal whisker, the intention is indeed to refer to the whisker corresponding to the recorded barrel during single whisker stimulation. The term principal whisker is removed from Figure legend 1 and legend S1C where it may have led to  ambiguity.    

Why the statement "after the start of active whisking"? Mice are under anesthesia here; it does not appear to be relevant for the figure. 

“After the start of active whisking” refers to the state of the barrel cortex circuitry at the time of recordings. The particular reference we use comes from the habit of assessing sensory processing also from a developmental point of view. The reviewer is correct that it has nothing to do the with the status of the experiment. Nevertheless, since the reviewer found that it may create confusion, we have now taken it out. 

(8) Figure 3: The y-axis label is missing for panel C. 

This is now fixed. (dF/F).

(9) Figure 4: Axis labels are missing.

Added.

Minor: 

(10) Line 36: "progressive increase in septal spiking activity upon multi-whisker stimulation". There is no increase in septal spiking activity upon MWS; the ratio MWS/SWS increases.

We have changed the sentence as follows: Genetic removal of Elfn1, which regulates the incoming excitatory synaptic dynamics onto SST+ interneurons, leads to the loss of the progressive increase in septal spiking ratio (MWS/SWS) upon stimulation.

(11) Line 105: domain-specific, rather than column-specific, for consistency.

We have changed it.

(12) Lines 173-174: "a divergence between barrel and septa domain activity also occurred in Layer 4 from the 2nd pulse onward (Figure 1E)". The authors only show a restricted number of comparisons. Why not show the p-values as for SWS?

The statistics is now presented in current Figure 1E.

(13) Lines 151-153: "Correspondingly, when a single whisker is stimulated repeatedly, the response to the first pulse is principally bottom-up thalamic-driven responses, while the later pulses in the train are expected to also gradually engage cortico-thalamo-cortical and cortico-cortical loops." Can the authors please provide a reference?

We have now added the following references : (Kyriazi and Simons, 1993; Middleton et al., 2010; Russo et al., 2025).

(14) Lines 184-186: "Our electrophysiological experiments show a significant divergence of responses over time upon both SWS and MWS in L4 between barrels (principal and neighboring) and adjacent septa, with minimal initial difference". The only difference between the neighboring barrel and septa is the responses to the initial pulse. Can the author clarify? 

We have now changed the sentence as follows: Our electrophysiological experiments show a significant divergence of responses between domains upon both SWS and MWS in L4. (Line 198 now)

(15) Line 214: "suggest these interneurons may play a role in diverging responses between barrels and septa upon SWS". Why SWS specifically?

We have changed the sentence as follows: These results confirmed that SST+ and VIP+ interneurons have higher densities in septa compared to barrels in L4 and suggest these interneurons may play a role in diverging responses between barrels and septa. (Line 231 now).

(16) Line 235: "This result suggests that differential activation of SST+ interneurons is more likely to be involved in the domain-specific temporal ratio differences between barrels and septa". Why? The results here are not domain-specific.

We have now revised this statement to: This result suggested that temporal ratio differences specific to barrels and septa might involve differential activation of SST+ interneurons rather than VIP+ interneurons.

(17) Lines 241-243: "SST+ interneurons in the cortex are known to show distinct short-term synaptic plasticity, particularly strong facilitation of excitatory inputs, which enables them to regulate the temporal dynamics of cortical circuits." Please provide a reference.

We have now added the following references: (Grier et al., 2023; Liguz-Lecznar et al., 2016).

(18) Lines 245-247: "A key regulator of this plasticity is the synaptic protein Elfn1, which mediates short-term synaptic facilitation of excitation on SST+ interneurons (Stachniak et al., 2021, 2019; Tomioka et al., 2014)". Is Stachniak et al., 2021 not about the role of Elf1n in excitatory-to-VIP+ neuron synapses?

The reviewer correctly spotted this discrepancy . This reference has now been removed from this statement.

(19) Lines 271-272: "Building on our findings that Elfn1-dependent facilitation in SST+ interneurons is critical for maintaining barrel-septa response divergence". The authors did not show that.

We have now changed the statement to: Building on our findings that Elfn1 is critical for maintaining barrel-septa response divergence  

(20) Line 280: second firing peak, not "peal".

Thank you, it is now fixed.

(21) Lines 304-305: "These results highlight the critical role of Elfn1 in facilitating the temporal integration of 305 sensory inputs through its effects on SST+ interneurons". This claim is also overstated. 

We have now changed the statement to: These results highlight the contribution of Elfn1 to the temporal integration of sensory inputs. (Line 362)

(22) Line 329: Any reason why not cite Chen et al., Nature 2013?

We have now added this reference, as also pointed out by reviewer 1.

(23) Line 341-342: "wS1" and "wS2" instead of S1 and S2 for consistency.

Thanks, we have now updated the terms.

Reviewer #2 (Recommendations for the authors): 

(1) Figure 3D - the SW conditions are labeled but not the MW conditions (two right graphs) - they should be labeled similarly (SSTMW, VIPMW). 

The two right graphs in Figure 3D represent paired SW vs MW comparisons of the evoked responses for SST and VIP populations, respectively.

(2) Figure 6 D and E I think it would be better if the Depth measurements were to be on the yaxis, which is more typical of these types of plots. 

We thank the reviewer for this comment. Although we appreciate this may be the case, we feel that the current presentation may be easier for the reader to navigate, and we have hence kept it. 

(3) Having an operational definition of septa versus barrel would be useful. As the authors point out, this is a tough distinction in a mouse, and often you read papers that use Barrel Wall versus Barrel Hollow/Center - operationally defining how these areas were distinguished would be helpful. 

We thank the reviewer for this comment and understand the point made.

We have now updated the methods section in line 611: 

DiI marks contained within the vGlut2 staining were defined as barrel recordings, while DiI marks outside vGlut2 staining were septal recordings.

Reviewer #3 (Recommendations for the authors): 

To support the manuscript's major claims, the authors should consider the following:

(1) Validate the septal identity of the neurons studied, either anatomically or functionally at the single-cell level (e.g., via Ca²⁺ imaging with confirmed barrel/septa mapping). 

We thank the reviewer for this suggestion, but we feel that these extensive experiments are beyond the scope of this study. 

(2) Provide both anatomical and physiological evidence to assess the possibility of altered cortical development in Elfn1 KO mice, including potential changes in barrel structure or SST⁺ cell distribution. 

To address the reviewer’s point, we have now added the following to the Discussion: “Although Elfn1 is constitutively knocked out, we find here and in previous studies that barrel structure is preserved (Stachniak et al., 2019, 2023). Further, the distribution of Elfn1 expressing interneurons is not different in KO mice, suggesting minimal developmental disruption (Dolan and Mitchell, 2013). Nonetheless, we acknowledge that subtle circuit changes cannot be ruled out without conditional knockouts.”,

(3) Examine the sensory responses of SST⁺ and VIP⁺ interneurons in deeper cortical layers, particularly layer 4, which is central to the study's main conclusions.

We thank the reviewer for this suggestion and appreciate the value it would bring to the study. We nevertheless feel that these extensive experiments are beyond the scope of this study and hence opted out from performing them. 

Minor Comments:

(1)  The authors used a CLARITY-based passive clearing protocol, which is known to sometimes induce tissue swelling or distortion. This may affect anatomical precision, especially when assigning neurons to narrow domains such as septa versus barrels. Please clarify whether tissue expansion was measured, corrected, or otherwise accounted for during analysis.

Yes, the tissue expansion was accounted during analysis for the laminar specification. We excluded the brains with severe distortion. 

(2) While the anatomical data are plotted as a function of "depth from the top of layer 4," the manuscript does not specify the precise depth ranges used to define individual cortical layers in the cleared tissue. Given the importance of laminar specificity in projection and cell type analyses, the criteria and boundaries used to delineate each layer should be explicitly stated.

Thank you for pointing this out. We now include the criteria for delineating each layer in the manuscript. “Given that the depth of Layer 4 (L4) can be reliably measured due to its welldefined barrel boundaries, and that the relative widths of other layers have been previously characterized (El-Boustani et al., 2018), we estimated laminar boundaries proportionally. Specifically, Layer 2/3 was set to approximately 1.3–1.5 times the width of L4, Layer 5a to ~0.5 times, and Layer 5b to a similar width as L4. Assuming uniform tissue expansion across the cortical column, we extrapolated the remaining laminar thicknesses proportionally.”

(3)  In several key comparisons (e.g., SST⁺ vs. VIP⁺ interneurons, or S2-projecting vs. M1projecting neurons), it is unclear whether the same barrel columns were analyzed across conditions. Given the anatomical and functional heterogeneity across wS1 columns, failing to control for this may introduce significant confounds. We recommend analyzing matched columns across groups or, if not feasible, clearly acknowledging this limitation in the manuscript.

We thank the reviewer for raising this important point. For the comparison of SST⁺ versus VIP⁺ interneurons, it would in principle have been possible to analyze the same barrel columns across groups. However, because some of the cleared brains did not reach the optimal level of clarity, our choice of columns was limited, and we were not always able to obtain sufficiently clear data from the same columns in both groups. Similarly, for the analysis of S2- versus M1-projecting neurons, variability in the position and spread of retrograde virus injections made it difficult to ensure measurements from identical barrel columns. We have now added a statement in the Discussion to acknowledge this limitation.

(4) Figure 1C: Clarify what each point in the t-SNE plot represents-e.g., a single trial, a recording channel, or an averaged response. Also, describe the input features used for dimensionality reduction, including time windows and preprocessing steps.

In response to the reviewer’s comment, we have now added the following in the methods: In summary, each point in the t-SNE plots represents an averaged response across 20 trials for a specific domain (barrel, septa, or neighbor) and genotype (WT or KO), with approximately 14 points per domain derived from the 280 trials in each dataset. The input features are preprocessed by averaging blocks of 20 trials into 1900-dimensional vectors (95ms × 20), which are then reduced to 2D using t-SNE with the specified parameters. This approach effectively highlights the segregation and clustering patterns of neural responses across cortical domains in both WT and KO conditions.

(5) Figures 1D, E (left panels): The y-axes lack unit labeling and scale bars. Please indicate whether values are in spikes/sec, spikes/bin, or normalized units.

We have now clarified this. 

(6) Figures 1D, E (right panels): The color bars lack units. Specify whether the values represent raw firing rates, z-scores, or other normalized measures. Replace the vague term "Matrix representation" with a clearer label such as "Pulse-aligned firing heatmap."

Thank you, we have now done it.

(7) Figure 1E (bottom panel): There appears to be no legend referring to these panels. Please define labels such as "B" and "S." 

Thank you, we have now done it.

(8) Figure 1E legend: If it duplicates the legend from Figure 1D, this should be made explicit or integrated accordingly. 

We have changed the structure of this figure.

(9) Figure 1F: Define "AUC" and explain how it was computed (e.g., area under the firing rate curve over 0-50 ms). Indicate whether the plotted values represent percentages and, if so, label the y-axis accordingly. If normalization was applied, describe the procedure. Include sample sizes (n) and specify what each data point represents (e.g., animal, recording site). 

The following paragraph has been added in the methods section:

The Area Under the Curve (AUC) was computed as the integral of the smoothed firing rate (spikes per millisecond) over a 50ms window following each whisker stimulation pulse, using trapezoidal integration. Firing rate data for layer 4 barrel and septal regions in wild-type (WT) and knockout (KO) mice were smoothed with a 3-point moving average and averaged across blocks of 20 trials. Plotted values represent the percentage ratio of multi-whisker (MW) to single whisker (SW) AUC with error bars showing the standard error of the mean. Each data point reflects the mean AUC ratio for a stimulation pulse across approximately 11 blocks (220 trials total). The y-axis indicates percentages.

(10) Figure 3C: Add units to the vertical axis.

We have added them.

(11) Figure 3D: Specify what each line represents (e.g., average of n cells, individual responses?). 

Each line represents an average response of a neuron.  

(12) Figure 4C legend: Same with what?". No legend refers to the bottom panels - please revise to clarify. 

Thank you. We have now changed the figure structure and legends and fixed the missing information issue.

(13) Supplementary Figure 1B: Indicate the physical length of the scale bar in micrometers. 

This has been fixed. The scale bar is 250um.

(14) Indicate the catalog number or product name of the 8×8 silicon probe used for recordings.

We have added this information. It is the A8x8-Edge-5mm-100-200-177-A64

References

(1) Beierlein, M., Gibson, J. R. & Connors, B. W. (2003). Two dynamically distinct inhibitory networks in layer 4 of the neocortex. J. Neurophysiol. 90, 2987–3000.

(2) Burkhalter, A., D’Souza, R. D. & Ji, W. (2023). Integration of feedforward and feedback information streams in the modular architecture of mouse visual cortex. Annu. Rev. Neurosci. 46, 259–280.

(3) Chen, J. L., Margolis, D. J., Stankov, A., Sumanovski, L. T., Schneider, B. L. & Helmchen, F. (2015). Pathway-specific reorganization of projection neurons in somatosensory cortex during learning. Nat. Neurosci. 18, 1101–1108.

(4) Connor, J. R. & Peters, A. (1984). Vasoactive intestinal polypeptide-immunoreactive neurons in rat visual cortex. Neuroscience 12, 1027–1044.

(5) Cruikshank, S. J., Lewis, T. J. & Connors, B. W. (2007). Synaptic basis for intense thalamocortical activation of feedforward inhibitory cells in neocortex. Nat. Neurosci. 10, 462–468.

(6) Dolan, J. & Mitchell, K. J. (2013). Mutation of Elfn1 in mice causes seizures and hyperactivity. PLoS One 8, e80491.

(7) Gibson, J. R., Beierlein, M. & Connors, B. W. (1999). Two networks of electrically coupled inhibitory neurons in neocortex. Nature 402, 75–79.

(8) Ji, W., Gămănuţ, R., Bista, P., D’Souza, R. D., Wang, Q. & Burkhalter, A. (2015). Modularity in the organization of mouse primary visual cortex. Neuron 87, 632–643.

(9) Martin-Cortecero, J. & Nuñez, A. (2014). Tactile response adaptation to whisker stimulation in the lemniscal somatosensory pathway of rats. Brain Res. 1591, 27–37.

(10) Mégevand, P., Troncoso, E., Quairiaux, C., Muller, D., Michel, C. M. & Kiss, J. Z. (2009). Long-term plasticity in mouse sensorimotor circuits after rhythmic whisker stimulation. J. Neurosci. 29, 5326–5335.

(11) Meier, A. M., Wang, Q., Ji, W., Ganachaud, J. & Burkhalter, A. (2021). Modular network between postrhinal visual cortex, amygdala, and entorhinal cortex. J. Neurosci. 41, 4809– 4825.

(12) Meier, A. M., D’Souza, R. D., Ji, W., Han, E. B. & Burkhalter, A. (2025). Interdigitating modules for visual processing during locomotion and rest in mouse V1. bioRxiv 2025.02.21.639505.

(13) Scala, F., Kobak, D., Shan, S., Bernaerts, Y., Laturnus, S., Cadwell, C. R., Hartmanis, L., Froudarakis, E., Castro, J. R., Tan, Z. H., et al. (2019). Layer 4 of mouse neocortex differs in cell types and circuit organization between sensory areas. Nat. Commun. 10, 4174.

(14) Stachniak, T. J., Sylwestrak, E. L., Scheiffele, P., Hall, B. J. & Ghosh, A. (2019). Elfn1induced constitutive activation of mGluR7 determines frequency-dependent recruitment of somatostatin interneurons. J. Neurosci. 39, 4461–4475.

(15) Stachniak, T. J., Kastli, R., Hanley, O., Argunsah, A. Ö., van der Valk, E. G. T., Kanatouris, G. & Karayannis, T. (2021). Postmitotic Prox1 expression controls the final specification of cortical VIP interneuron subtypes. J. Neurosci. 41, 8150–8166.

(16) Stachniak, T. J., Argunsah, A. Ö., Yang, J. W., Cai, L. & Karayannis, T. (2023). Presynaptic kainate receptors onto somatostatin interneurons are recruited by activity throughout development and contribute to cortical sensory adaptation. J. Neurosci. 43, 7101–7118.

(17) Sun, Q.-Q., Huguenard, J. R. & Prince, D. A. (2006). Barrel cortex microcircuits: Thalamocortical feedforward inhibition in spiny stellate cells is mediated by a small number of fast-spiking interneurons. J. Neurosci. 26, 1219–1230.

(18) Sylwestrak, E. L. & Ghosh, A. (2012). Elfn1 regulates target-specific release probability at CA1-interneuron synapses. Science 338, 536–540.

(19) Tan, Z., Hu, H., Huang, Z. J. & Agmon, A. (2008). Robust but delayed thalamocortical activation of dendritic-targeting inhibitory interneurons. Proc. Natl. Acad. Sci. USA 105, 2187–2192.

(20) Tomioka, N. H., Yasuda, H., Miyamoto, H., Hatayama, M., Morimura, N., Matsumoto, Y., Suzuki, T., Odagawa, M., Odaka, Y. S., Iwayama, Y., et al. (2014). Elfn1 recruits presynaptic mGluR7 in trans and its loss results in seizures. Nat. Commun. 5, 4501.

(21) Yamashita, T., Vavladeli, A., Pala, A., Galan, K., Crochet, S., Petersen, S. S. & Petersen, C. C. (2018). Diverse long-range axonal projections of excitatory layer 2/3 neurons in mouse barrel cortex. Front. Neuroanat. 12, 33.

Reviewer #2 (Public review):

Summary:

In this manuscript, Pereira de Castro and coworkers are studying potential competition between a more standard splicing factor SF1 and an alternative splicing factor called QK1. This is interesting because they bind to overlapping sequence motifs and could potentially have opposing effects on promoting the splicing reaction. To test this idea, the authors KD either SF1 or QK1 in mammalian cells and uncover several exons whose splicing regulation follows the predicted pattern of being promoted for splicing by SF1 and repressed by QK1. Importantly, these have introns enriched in SF1 and QK1 motifs. The authors then focus on one exon in particular with two tandem motifs to study the mechanism of this in greater detail and their results confirm the competition model. Mass spec analysis largely agrees with their proposal; however, it is complicated by apparently quick transition of SF1 bound complexes to later splicing intermediates. An inspired experiment in yeast shows how QK1 competition could potentially have a determinental impact on splicing in an orthogonal system. Overall these results show how splicing regulation can be achieved by competition between a "core" and alternative splicing factor and provide additional insight into the complex process of branch site recognition. The manuscript is exceptionally clear and the figures and data very logically presented. The work will be valuable to those in the splicing field who are interested in both mechanism and bioinformatics approaches to deconvolve any apparent "splicing code" being used by cells to regulate gene expression.

Strengths:

(1) The main discovery of the manuscript involving evidence for SF1/QK1 competition is quite interesting and important for this field. This evidence has been missing and may change how people think about branch site recognition.

(2) The experiments and the rationale behind them are clearly and logically presented.

(3) The experiments are carried out to a high standard and well-designed controls are included.

(4) The extrapolation of the result to yeast in order to show the potentially devastating consequences of QK1 competition was creative and informative.

Weaknesses:

Overall the weaknesses are relatively minor and involve cases where conclusions could potentially have been strengthened with additional experimentation. For example, pull-down of the U2 snRNP could be strengthened by detection of the snRNA whereas the proteins may themselves interact with these factors in the absence of the snRNA. In addition the discussion is a bit speculative given the data, but compelling nonetheless.

Reviewer #3 (Public review):

Summary:

In this manuscript the authors were trying to establish whether competition between the RNA binding proteins SF1 and QKI controlled splicing outcomes. These two proteins have similar binding sites and protein sequences, but SF1 lacks a dimerization motif and seems to bind a single version of the binding sequence. Importantly, these binding sequences correspond to branchpoint consensus sequences, with SF1 binding leading to productive splicing, but QKI binding leading instead to association with paraspeckle proteins. They show that in human cells SF1 generally activates exons and QKI represses, and a large group of the jointly regulated exons (43% of joint targets) are reciprocally controlled by SF1 and QKI. They focus on one of these exons RAI14 that shows this reciprocal pattern of regulation, and has 2 repeats of the binding site that make it a candidate for joint regulation, and confirm regulation within a minigene context. The authors used assembly of proteins within nuclear extracts to explain the effect of QKI versus SF1 binding. Finally the authors show that expression of QKI is lethal in yeast, and causes splicing defects.

How this fits in the field. This study is interesting and provides a conceptual advance by providing a general rule how SF1 and QKI interact with relation to binding sites, and the relative molecular fates followed, so is very useful. Most of the analysis seems to focus on one example, but the choice of this example was carefully explained in the text. The molecular analysis and global work significantly adds to the picture from the previously published paper about NUMB joint regulation by QKI and SF (Zong et al, cited in text as reference 50, that looked at SF1 and QKI binding in relation to a duplicated binding site/branchpoint sequence in NUMB).

Strengths:

The data presented are strong and clear. The ideas discussed in this paper are of wide interest, and present a simple model where two binding sites generates a potentially repressive QKI response, whereas exons that have a single upstream sequence are just regulated by SF1. The assembly of splicing complexes on RNAs derived from RAI14 in nuclear extracts, followed by mass spec gave interesting mechanistic insight into what was occurring as a result of QKI versus SF1 binding.

Weaknesses:

The authors have addressed the previous weaknesses of the study, resulting in a much stronger manuscript

Author response:

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

eLife Assessment

This important manuscript provides insights into the competition between Splicing Factor 1 (SF1) and Quaking (QKI) for binding at the ACUAA branch point sequence in a model intron, regulating exon inclusion. The study employs rigorous transcriptomic, proteomic, and reporter assays, with both mammalian cell culture and yeast models. Nevertheless, while the data are convincing, broadening the analysis to additional exons and narrowing the manuscript's title to better align with the experimental scope would strengthen the work.

Public Reviews:

Reviewer #1 (Public review):

In this manuscript, the authors aimed to show that SF1 and QKI compete for the intron branch point sequence ACUAA and provide evidence that QKI represses inclusion when bound to it.

Major strengths of this manuscript include:

(1) Identification of the ACUAA-like motif in exons regulated by QKI and SF1.

(2) The use of the splicing reporter and mutant analysis to show that upstream and downstream ACUAAC elements in intron 10 of RAI are required for repressing splicing.

(3) The use of proteomic to identify proteins in C2C12 nuclear extract that binds to the wild type and mutant sequence.

(4) The yeast studies showing that ectopic lethality when Qki5 expression was induced, due to increased mis-splicing of transcripts that contain the ACUAA element.

The authors conclusively show that the ACUAA sequence is bound by QKI and provide strong evidence that this leads to differences in exons inclusion and exclusion. In animal cells, and especially in human, branchpoint sequences are degenerate but seem to be recognized by specific splicing factors. Although a subset of splicing factors shows tissue-specific expression patterns most don't, suggesting that yet-to-be-identified mechanisms regulate splicing. This work suggests that an alternate mechanism could be related to the binding affinity of specific RNA binding factors for branchpoint sequences coupled with the level of these different splicing factors in a given cell.

We thank the reviewer for the positive comments.

Reviewer #2 (Public review):

Summary:

In this manuscript, Pereira de Castro and coworkers are studying potential competition between a more standard splicing factor SF1, and an alternative splicing factor called QK1. This is interesting because they bind to overlapping sequence motifs and could potentially have opposing effects on promoting the splicing reaction. To test this idea, the authors KD either SF1 or QK1 in mammalian cells and uncover several exons whose splicing regulation follows the predicted pattern of being promoted for splicing by SF1 and repressed by QK1. Importantly, these have introns enriched in SF1 and QK1 motifs. The authors then focus on one exon in particular with two tandem motifs to study the mechanism of this in greater detail and their results confirm the competition model. Mass spec analysis largely agrees with their proposal; however, it is complicated by the apparently quick transition of SF1-bound complexes to later splicing intermediates. An inspired experiment in yeast shows how QK1 competition could potentially have a detrimental impact on splicing in an orthogonal system. Overall, these results show how splicing regulation can be achieved by competition between a "core" and alternative splicing factor and provide additional insight into the complex process of branch site recognition. The manuscript is exceptionally clear and the figures and data are very logically presented. The work will be valuable to those in the splicing field who are interested in both mechanism and bioinformatics approaches to deconvolve any apparent "splicing code" being used by cells to regulate gene expression. Criticisms are minor and the most important of them stem from overemphasis on parts of the manuscript on the evolutionary angle when evolution itself wasn't analyzed per se.

We thank the reviewer for the positive comments and very clear and fair critical points.

Strengths:

(1) The main discovery of the manuscript involving evidence for SF1/QK1 competition is quite interesting and important for this field. This evidence has been missing and may change how people think about branch site recognition.

(2) The experiments and the rationale behind them are exceptionally clearly and logically presented. This was wonderful!

Thank you so much. We felt the overall flow of the paper and data make for a nice “story” that conveys a relatively easy-to-understand explanation for a complex subject.

(3) The experiments are carried out to a high standard and well-designed controls are included.

(4) The extrapolation of the result to yeast in order to show the potentially devastating consequences of the QK1 competition was very exciting and creative.

We agree this is a very exciting result and finding! Thanks.

Weaknesses:

Overall the weaknesses are relatively minor and involve cases where clarification is necessary, some additional analysis could bolster the arguments, and suggestions for focusing the manuscript on its strengths.

(1) The title (Ancient...evolutionary outcomes), abstract, and some parts of the discussion focus heavily on the evolutionary implications of this work. However, evolutionary analysis was not performed in these studies (e.g., when did QK1 and SF1 proteins arise and/or diverge? How does this line up with branch site motifs and evolution of U2? Any insight from recent work from Scott Roy et al?). I think this aspect either needs to be bolstered with experimental work/data or this should be tamped down in the manuscript. I suggest highlighting the idea expressed in the sentence "A nuanced implication of this model is that loss-of-function...". To me, this is better supported by the data and potentially by some analysis of mutations associated with human disease.

We have revised the title and dampened the evolutionary aspects of the previous version of the manuscript.

(2) One paper that I didn't see cited was that by Tanackovic and Kramer (Mol Biol Cell 2005). This paper is relevant because they KD SF1 and found it nonessential for splicing in vivo. Do their results have implications for those here? How do the results of the KD compare? Could QK1 competition have influenced their findings (or does their work influence the "nuanced implication" model referenced above?)?

This is an interesting point, and thank you for the suggestion. We have now included a brief description of this study in the Introduction of the revised manuscript and do note that the authors measured intron retention of a beta globin reporter and SF3A1, SF3A2, and SF3A3 during SF1 knockdown, but did not detect elevated unspliced RNA in these targets.

(3) Can the authors please provide a citation for the statement "degeneracy is observed to a higher degree in organisms with more alternative splicing"? Does recent evolutionary analysis support this?

We have removed the statement, as it did not add much to the content and I am not sure I can state the concept I was attempting to convey in a simple manner with few citations.

(4) For the data in Figure 3, I was left wondering if NMD was confounding this analysis. Can the authors respond to this and address this concern directly?

We have not measured if the reporters used in Figure 3 produce protein(s). Presumably, though, all spliced reporter RNA would be degraded equally (the included/skipped isoforms’ “reading frames” are not altered from one another). This would not be case for unspliced nuclear reporter RNA, however. Given this difference, and that our analysis can not resolve the subcellular localization of the different reporter species, we have removed the measurement of and subsequent results describing unspliced reporter RNA from Figure 3.

(5) To me, the idea that an engaged U2 snRNP was pulled down in Figure 4F would be stronger if the snRNA was detected. Was that able to be observed by northern or primer extension? Would SF1 be enriched if the U2 snRNA was degraded by RNaseH in the NE?

We did not measure any co-associating RNAs in this experimental approach, but agree that this approach would strengthen the evidence for it.

(6) I'm wondering how additive the effects of QK1 and SF1 are... In Figure 2, if QK1 and SF1 are both knocked down, is the splicing of exon 11 restored to "wt" levels?

This is an interesting question that we were unfortunately unable to address experimentally here.

(7) The first discussion section has two paragraphs that begin "How does competition between SF1..." and "Relatively little is known about how...". I found the discussion and speculation about localization, paraspekles, and lncRNAs interesting but a bit detracting from the strengths of the manuscript. I would suggest shortening these two paragraphs into a single one.

We have revised the Discussion.

Reviewer #3 (Public review):

Summary:

In this manuscript, the authors were trying to establish whether competition between the RNA-binding proteins SF1 and QKI controlled splicing outcomes. These two proteins have similar binding sites and protein sequences, but SF1 lacks a dimerization motif and seems to bind a single version of the binding sequence. Importantly, these binding sequences correspond to branchpoint consensus sequences, with SF1 binding leading to productive splicing, but QKI binding leading instead to association with paraspeckle proteins. They show that in human cells SF1 generally activates exons and QKI represses, and a large group of the jointly regulated exons (43% of joint targets) are reciprocally controlled by SF1 and QKI. They focus on one of these exons RAI14 that shows this reciprocal pattern of regulation, and has 2 repeats of the binding site that make it a candidate for joint regulation, and confirm regulation within a minigene context. The authors used the assembly of proteins within nuclear extracts to explain the effect of QKI versus SF1 binding. Finally, the authors show that the expression of QKI is lethal in yeast, and causes splicing defects.

How this fits in the field. This study is interesting and provides a conceptual advance by providing a general rule on how SF1 and QKI interact in relation to binding sites, and the relative molecular fates followed, so is very useful. Most of the analysis seems to focus on one example, although the molecular analysis and global work significantly add to the picture from the previously published paper about NUMB joint regulation by QKI and SF (Zong et al, cited in text as reference 50, that looked at SF1 and QKI binding in relation to a duplicated binding site/branchpoint sequence in NUMB).

Thank you for the encouraging remarks.

Strengths:

The data presented are strong and clear. The ideas discussed in this paper are of wide interest, and present a simple model where two binding sites generate a potentially repressive QKI response, whereas exons that have a single upstream sequence are just regulated by SF1. The assembly of splicing complexes on RNAs derived from RAI14 in nuclear extracts, followed by mass spec gave interesting mechanistic insight into what was occurring as a result of QKI versus SF1 binding.

Weaknesses:

I did not think the title best summarises the take-home message and could be perhaps a bit more modest. Although the authors investigated splicing patterns in yeast and human cells, yeast do not have QKI so there is no ancient competition in that case, and the study did not really investigate physiological or evolutionary outcomes in splicing, although it provides interesting speculation on them. Also as I understood it, the important issue was less conserved branchpoints in higher eukaryotes enabling alternative splicing, rather than competition for the conserved branchpoint sequence. So despite the the data being strong and properly analysed and discussed in the paper, could the authors think whether they fit best with the take-home message provided in the title? Just as a suggestion (I am sure the authors can do a better job), maybe "molecular competition between variant branchpoint sequences predict physiological and evolutionary outcomes in splicing"?

Thank you for this point (Reviewer 2 had a similar comment) and the suggestion. We have revised the title.

Although the authors do provide some global data, most of the detailed analysis is of RAI14. It would have been useful to examine members of the other quadrants in Figure 1C as well for potential binding sites to give a reason why these are not co-regulated in the same way as RAI14. How many of the RAI14 quadrants had single/double sites (the motif analysis seemed to pull out just one), and could one of the non-reciprocally regulated exons be moved into a different quadrant by addition or subtraction of a binding site or changing the branchpoint (using a minigene approach for example).

This is an interesting point that we have considered. Our intent with the focus on RAI14 was to use a naturally occurring intron bps with evidence of strong QKI binding that did not require a high degree of sequence manipulation or engineering.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) Most of my recommendations are really centered on the figures. In their current state, they detract from the data shown and could be improved: I recommend the authors use a uniform font. For example, Figure 1E and F have at least three different fonts of varying sizes making it very messy. In Figure 1C, the authors could bold the Ral14 ex11 or simply indicate that the blue is this exon in the legend, thus removing the text from this very busy graph. In Figure 4F, I would recommend, having all the labels the same size and putting those genes of interest like Sf3a1 in bold. This could also be done in Figure 4E.

Thank you for the suggestion and we have edited these (FYI the font in Fig’s 1E and 1F were from the rMAPS default output, but I agree, it gives a sloppy appearance).

(2) In Figures 4D and 4G, is there QKI binding to the downstream deletion mutant after 30 minutes? Also, in Figure 4G, are these all from the same blot? The band sizes seem to be very different between lanes. If these were not on the same blot, the original gels should be submitted.

A small amount of Qki appears to be binding after 30 min. All lanes/blots are from the same gels/membranes; see new Supplemental Figure 4 for the original (uncropped) images of the blots.

(3) The authors should indicate, the source and concentration of the antibodies used for their WB. They should also indicate the primers used for RT-PCRs.

We have revised the methods to include the antibody information and have uploaded a supplemental table 8 with all oligonucleotide sequences used (which I (Sam Fagg) neglected to do initially, so that’s my bad).

Reviewer #2 (Recommendations for the authors):

(1) This may come down to the author's preference but branch point and branch site are frequently two words, not a single compound word (branch point vs. branchpoint). In addition, the authors may want to use branchsite with the abbreviation BS more frequently since they often don't describe the specific point of branching, and bp and bps could be confused for the more frequent abbreviations for base pair(s).

Good suggestion; we have edited the text accordingly.

(2) In general the addition of page numbers and line numbers to the manuscript would greatly aid reviewers!

Point taken…

(3) Introduction; "...under normal growth conditions they are efficiently spliced". I would say MOST introns in yeast are efficiently spliced. This is definitely not universal.

Text edited to indicate that most are efficiently spliced.

(4) Introduction; " recognition of the bps by SF1 (mammals) (20)". The choice of reference 20 is an odd one here. I think the Robin Reed and Michael Rosbash paper was the first to show SF1 was the human homolog of BBP.

Got it, thanks (added #14 here and kept #20 also since it shows the structure of SF1 in complex with a UACUAAC bps.)

(5) Results; "QK1 and SF1 co-regulate.."; it may be useful for the reader if you could explain in more detail why exon inclusion and intron retention are expected outcomes for QK1 knockdown and vice versa for SF1. The exon inclusion here is more obvious than the intron retention phenotype. (In other words, if more exons are included shouldn't it follow that more introns are removed?)

We explain the expected results for exon inclusion in the Introduction and this paragraph of the Results. Although we have observed more intron retention under QKI loss-of-function approaches before, I am uncertain where the reviewer sees that we indicate any expected result for intron retention from either QKI or SF1 knockdown. I believe the statement you refer to might be on line 162 and starts with: “Consistent with potentially opposing functions in splicing…” ?

Also, I agree that if SF1 is a “splicing activator,” one might expect more IR in its absence (but this is not the case; there is, in fact, less), but nonetheless, the opposite outcome is observed with QKI knockdown (more IR). It is unclear why this is the case, and we did not investigate it.

(6) Results; "QK1 and SF1 co-regulate.."; "Thus the most highly represented set.." To me, the most highly represented set is those which are not both QK1-repressed and SF1-activated. Does this indicate that other factors are involved at most sites than simple competition between these two?

We have revised the sentence in question to include the text “by quadrant” in order to convey our meaning more precisely.

(7) Throughout the manuscript, 5 apostrophes and 3 apostrophes are used instead of 5 prime symbols and 3 prime symbols.

Thank you for pointing that out. We have fixed each instance of this.

(8) Sometimes SF1 is written as Sf1. (also Tatsf1)

This was a mouse/human gene/protein nomenclature error that we have fixed; thank you for pointing this out.

(9) You may want to make sure that figures are labeled consistently with the manuscript text. In Figure 1B, it is RI rather than IR. In Figure 4 it is myoblast NE rather than C2C12 nuclear extract.

We have fixed these, checked for other examples, and where relevant, edited those too.

(10) I think Figure 1A could be improved by also including a depiction of the domain arrangements of SF1 and QK1.

Done.

(11) I was a bit confused with all the lines in Figure 1E and 1F. What is the difference between the log (pVal) and upregulated plots? Can these figures be simplified or explained more thoroughly?

Based on this comment and one from Reviewer 1, we have slightly revised the wording (and font) on the output, which hopefully clarifies. These are motif enrichment plots generated by rMAPS (Refs 61 and 62) analysis of rMATS (Ref 60) data for exons more included (depicted by the red lines) or more skipped (depicted by the blue lines) compared to control versus a “background” set of exons that are detectable but unchanged. The -log₁₀ is P-value (dotted line) indicates the significance of exons more included in shRNA treatment vs control shRNA (previously read “upregulated”) compared to background exons that are detectable but unchanged; the solid lines indicate the motif score; these are described in the references indicated.

(12) Figure 1B, it is a bit hard to conclude that there is more AltEx or "RI/IR" in one sample vs. the other from these plots since the points overlay one another. Can you include numbers here?

Added (and deleted Suppl Fig S1, which was simply a chart showing the numbers).

(13) How was PSI calculated in Figure 2A?

VAST-tools (we state this in the legend in the revised version).

You may want to include rel protein (or the lower limit of detection) for Figure 2B to be consistent with 2C. Why is KD of SF1 so poor and variable between 2C and 2D?

We have not investigated this, but these blots show an optimized result that we were able to obtain for the knockdown in each cell type. It may be that HEK293 cells (Fig 2B) have a stronger requirement for SF1 than C2C12 cells…? I would argue that it is not necessarily “poor” in Fig 2C, as we observe ~70% depletion of the protein.

Why are two bands present in the gel?

Two to three isoforms of SF1 are present in most cell types.

A good (or bad, really) example of an SF1 western blot (and knockdown of ~35% in K562 or ~45% in HepG2 can also be seen on the ENCODE project website, for reference:

https://www.encodeproject.org/documents/6001a414-b096-4073-94ff-3af165617eb5/@@download/attachment/SF1_BGKLV28-49.pdf

By comparison, I think ours are much more cosmetically pleasing, and our knockdown (especially in C2C12) is much more efficient.

(14) Figure 3, The asterisk refers to a cryptic product. Can the uaAcuuuCAG be used as a branch point? Presumably the natural 3' SS is now too close so this would result in activation of a downstream 3'SS?

We did not pursue determining the identity of this minor and likely artefactual product, but we (and others) have observed a similar phenomenon when using splicing reporter-based mutational approaches.

(15) For the methods. The "RNA extraction, RT -PCR,..." subheading needs to be on its own line. Please add (w/v) or (v/v) to percentages where appropriate. Please convert ug to the symbol for "micro".

Thank you, we have made these changes.

(16) In Figure 4B, the text here and legend are microscopic. Even with reading glasses, I couldn't make anything out!

We have increased the font sizes for the text and scale bar…when referring to “legend” does the reviewer mean the scale bar?

(17) As a potential discussion item, it is worth noting that SF1 could also repress splicing if it could either not engage with U2AF or be properly displaced by U2 snRNP so the snRNA could pair. I was wondering if QK1 could similarly be activating if it could engage with U2AF. I'm unsure if this could be tested by domain swaps (and is beyond the scope of this paper). It just may be worth speculating about.

Good point and suggestion…we are looking into this.

Reviewer #3 (Recommendations for the authors):

(1) Is the reference in the text to Figure 5F correct for actin splicing (this is just before the discussion)?

I see references several lines up from this, but I do not see a reference just before the discussion…?

(2) I was not sure why the minigene experiments showed such high levels of intron retention that seemed to be impacted also by deletion of the branchpoint sequences, and suggest that the two branchpoints are not equal in strength.

Neither were we, but Reviewer 2 has suggested that degradation of the spliced products could be rapid (NMD substrates) which could complicate the interpretation of what appears to be higher levels of intron retention. Given the possibility that this could be a non-physiological artefact, we have removed the measurement of unspliced reporter and now only show the spliced products (equally subject to degradation) and report their percent inclusion.

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors investigate the nanoscopic distribution of glycine receptor subunits in the hippocampus, dorsal striatum, and ventral striatum of the mouse brain using single-molecule localization microscopy (SMLM). They demonstrate that only a small number of glycine receptors are localized at hippocampal inhibitory synapses. Using dual-color SMLM, they further show that clusters of glycine receptors are predominantly localized within gephyrin-positive synapses. A comparison between the dorsal and ventral striatum reveals that the ventral striatum contains approximately eight times more glycine receptors and this finding is consistent with electrophysiological data on postsynaptic inhibitory currents. Finally, using cultured hippocampal neurons, they examine the differential synaptic localization of glycine receptor subunits (α1, α2, and β). This study is significant as it provides insights into the nanoscopic localization patterns of glycine receptors in brain regions where this protein is expressed at low levels. Additionally, the study demonstrates the different localization patterns of GlyR in distinct striatal regions and its physiological relevance using SMLM and electrophysiological experiments. However, several concerns should be addressed.

Specific comments on the original version:

(1) Colocalization analysis in Figure 1A. The colocalization between Sylite and mEos-GlyRβ appears to be quite low. It is essential to assess whether the observed colocalization is not due to random overlap. The authors should consider quantifying colocalization using statistical methods, such as a pixel shift analysis, to determine whether colocalization frequencies remain similar after artificially displacing one of the channels.

(2) Inconsistency between Figure 3A and 3B. While Figure 3B indicates an ~8-fold difference in the number of mEos4b-GlyRβ detections per synapse between the dorsal and ventral striatum, Figure 3A does not appear to show a pronounced difference in the localization of mEos4b-GlyRβ on Sylite puncta between these two regions. If the images presented in Figure 3A are not representative, the authors should consider replacing them with more representative examples or providing an expanded images with multiple representative examples. Alternatively, if this inconsistency can be explained by differences in spot density within clusters, the authors should explain that.

(3) Quantification in Figure 5. It is recommended that the authors provide quantitative data on cluster formation and colocalization with Sylite puncta in Figure 5 to support their qualitative observations.

(4) Potential for pseudo replication. It's not clear whether they're performing stats tests across biological replica, images, or even synapses. They often quote mean +/- SEM with n = 1000s, and so does that mean they're doing tests on those 1000s? Need to clarify.

(5) Does mEoS effect expression levels or function of the protein? Can't see any experiments done to confirm this. Could suggest WB on homogenate, or mass spec?

(6) Quantification of protein numbers is challenging with SMLM. Issues include i) some of FP not correctly folded/mature, and ii) dependence of localisation rate on instrument, excitation/illumination intensities, and also the thresholds used in analysis. Can the authors compare with another protein that has known expression levels- e.g. PSD95? This is quite an ask, but if they could show copy number of something known to compare with, it would be useful.

(7) Rationale for doing nanobody dSTORM not clear at all. They don't explain the reason for doing the dSTORM experiments. Why not just rely on PALM for coincidence measurements, rather than tagging mEoS with a nanobody, and then doing dSTORM with that? Can they explain? Is it to get extra localisations- i.e. multiple per nanobody? If so, localising same FP multiple times wouldn't improve resolution. Also, no controls for nanobody dSTORM experiments- what about non-spec nb, or use on WT sections?

(8) What resolutions/precisions were obtained in SMLM experiments? Should perform Fourier Ring Correlation (FRC) on SR images to state resolutions obtained (particularly useful for when they're presenting distance histograms, as this will be dependent on resolution). Likewise for precision, what was mean precision? Can they show histograms of localisation precision.

(9) Why were DBSCAN parameters selected? How can they rule out multiple localisations per fluor? If low copy numbers (<10), then why bother with DBSCAN? Could just measure distance to each one.

(10) For microscopy experiment methods, state power densities, not % or "nominal power".

(11) In general, not much data presented. Any SI file with extra images etc.?

(12) Clarification of the discussion on GlyR expression and synaptic localization: The discussion on GlyR expression, complex formation, and synaptic localization is sometimes unclear, and needs terminological distinctions between "expression level", "complex formation" and "synaptic localization". For example, the authors state: "What then is the reason for the low protein expression of GlyRβ? One possibility is that the assembly of mature heteropentameric GlyR complexes depends critically on the expression of endogenous GlyR α subunits." Does this mean that GlyRβ proteins that fail to form complexes with GlyRα subunits are unstable and subject to rapid degradation? If so, the authors should clarify this point. The statement "This raises the interesting possibility that synaptic GlyRs may depend specifically on the concomitant expression of both α1 and β transcripts." suggests a dependency on α1 and β transcripts. However, is the authors' focus on synaptic localization or overall protein expression levels? If this means synaptic localization, it would be beneficial to state this explicitly to avoid confusion. To improve clarity, the authors should carefully distinguish between these different aspects of GlyR biology throughout the discussion. Additionally, a schematic diagram illustrating these processes would be highly beneficial for readers.

(13) Interpretation of GlyR localization in the context of nanodomains. The distribution of GlyR molecules on inhibitory synapses appears to be non-homogeneous, instead forming nanoclusters or nanodomains, similar to many other synaptic proteins. It is important to interpret GlyR localization in the context of nanodomain organization.

Significance:

The paper presents biological and technical advances. The biological insights revolve mostly on the documentation of Glycine receptors in particular synapses in forebrain, where they are typically expressed at very low levels. The authors provide compelling data indicating that the expression is of physiological significance. The authors have done a nice job of combining genetically tagged mice with advanced microscopy methods to tackle the question of distributions of synaptic proteins. Overall, these advances are more incremental than groundbreaking.

Comments on revised version:

The authors have addressed the majority of the significant issues raised in the review and revised the manuscript appropriately. One issue that can be further addressed relates to the issue of pseudo-replication. The authors state in their response that "All experiments were repeated at least twice to ensure reproducibility (N independent experiments). Statistical tests were performed on pooled data across the biological replicates; n denotes the number of data points used for testing (e.g., number of synaptic clusters, detections, cells, as specified in each case).". This suggests that they're not doing their stats on biological replicates, and instead are pseudo replicating. It's not clear how they have ensured reproducibility, when the stats seem to have been done on pooled data across repeats.

Reviewer #3 (Public review):

In this study, Camuso et al., make use of a knock-in mouse model expressing endogenously mEos4b-tagged GlyRβ subunits to detect endogenous glycine receptors in mouse brain using single-molecule localization microscopy (SMLM). At synapses in the hippocampus GlyRβ molecules are detected at very low copy numbers. Assuming that each detected GlyRβ molecule is incorporated in a pentameric glycine receptor, it was estimated that while the majority of hippocampal inhibitory synapses do not contain glycine receptors, a small population of inhibitory synapses contain a few (up to 10) glycine receptors. Using dual-color SMLM approaches it is furthermore confirmed that the detected GlyRβ molecules are embedded in the postsynaptic domain marked by gephyrin. In contrast to the hippocampus, at inhibitory synapses in the striatum GlyRβ molecules were detected at considerably higher copy numbers. Interestingly, the observed number of GlyRβ detections was significantly higher in the ventral striatum compared to the dorsal striatum. These findings are corroborated by electrophysiological recordings showing that postsynaptic glycinergic currents can be readily detected in the ventral striatum but are almost absent in the dorsal striatum. Using lentiviral overexpression of recombinant GlyRalpha1, alpha2, and beta subunits in cultured hippocampal neurons, it is shown that GlyR alpha1 subunits are readily detectable at synapses, but overexpressed GlyRalpha2 and beta subunits did not strongly enrich at synapses. This could indicate that GlyRa1 expression is limiting the synaptic expression of GlyRβ-containing glycine receptors in hippocampal neurons.

Comments on revised version:

This revised manuscript is significantly improved. New experimental and quantitative analysis is presented that strengthen the conclusions. Overall, the results presented in this manuscript are based on carefully performed SMLM experiments and are well-presented and described. The knock-in mouse with endogenously tagged GlyRβ molecules is a very strong aspect of this study and provides confidence in the labeling, the combination with SMLM is very strong as it provides high sensitivity and spatial resolution. These results confirm previous studies and will be of interest to a specialised audience interested in glycine receptors, inhibitory synapse biology and super-resolution microscopy.

Author response:

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

We thank the editors of eLife and the reviewers for their thorough evaluation of our study. As regards the final comments of reviewer 1 please note that all experimental replicates were first analyzed separately, and were then pooled, since the observed changes were comparable between experiments. This mean that statistical analyses were done on pooled biological replicates.

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

General Statements

We thank the reviewers for their thorough and constructive evaluation of our work. We have revised the manuscript carefully and addressed all the criticisms raised, in particular the issues mentioned by several of the reviewers (see point-by-point response below). We have also added a number of explanations in the text for the sake of clarity, while trying to keep the manuscript as concise as possible.

In our view, the novelty of our research is two-fold. From a neurobiological point of view, we provide conclusive evidence for the existence of glycine receptors (GlyRs) at inhibitory synapses in various brain regions including the hippocampus, dentate gyrus and sub-regions of the striatum. This solves several open questions and has fundamental implications for our understanding of the organisation and function of inhibitory synapses in the telencephalon. Secondly, our study makes use of the unique sensitivity of single molecule localisation microscopy (SMLM) to identify low protein copy numbers. This is a new way to think about SMLM as it goes beyond a mere structural characterisation and towards a quantitative assessment of synaptic protein assemblies.

Point-by-point description of the revisions

Reviewer #1 (Evidence, reproducibility and clarity): 

In this manuscript, the authors investigate the nanoscopic distribution of glycine receptor subunits in the hippocampus, dorsal striatum, and ventral striatum of the mouse brain using single-molecule localization microscopy (SMLM). They demonstrate that only a small number of glycine receptors are localized at hippocampal inhibitory synapses. Using dual-color SMLM, they further show that clusters of glycine receptors are predominantly localized within gephyrinpositive synapses. A comparison between the dorsal and ventral striatum reveals that the ventral striatum contains approximately eight times more glycine receptors and this finding is consistent with electrophysiological data on postsynaptic inhibitory currents. Finally, using cultured hippocampal neurons, they examine the differential synaptic localization of glycine receptor subunits (α1, α2, and β). This study is significant as it provides insights into the nanoscopic localization patterns of glycine receptors in brain regions where this protein is expressed at low levels. Additionally, the study demonstrates the different localization patterns of GlyR in distinct striatal regions and its physiological relevance using SMLM and electrophysiological experiments. However, several concerns should be addressed. 

The following are specific comments: 

(1) Colocalization analysis in Figure 1A. The colocalization between Sylite and mEos-GlyRβ appears to be quite low. It is essential to assess whether the observed colocalization is not due to random overlap. The authors should consider quantifying colocalization using statistical methods, such as a pixel shift analysis, to determine whether colocalization frequencies remain similar after artificially displacing one of the channels. 

Following the suggestion of reviewer 1, we re-analysed CA3 images of Glrb^eos/eos hippocampal slices by applying a pixel-shift type of control, in which the Sylite channel (in far red) was horizontally flipped relative to the mEos4b-GlyRβ channel (in green, see Methods). As expected, the number of mEos4b-GlyRβ detections per gephyrin cluster was markedly reduced compared to the original analysis (revised Fig. 1B), confirming that the synaptic mEos4b detections exceed chance levels (see page 5). 

(2) Inconsistency between Figure 3A and 3B. While Figure 3B indicates an ~8-fold difference in the number of mEos4b-GlyRβ detections per synapse between the dorsal and ventral striatum, Figure 3A does not appear to show a pronounced difference in the localization of mEos4bGlyRβ on Sylite puncta between these two regions. If the images presented in Figure 3A are not representative, the authors should consider replacing them with more representative examples or providing an expanded images with multiple representative examples. Alternatively, if this inconsistency can be explained by differences in spot density within clusters, the authors should explain that. 

The pointillist images in Fig. 3A are essentially binary (red-black). Therefore, the density of detections at synapses cannot be easily judged by eye. For clarity, the original images in Fig. 3A have been replaced with two other examples that better reflect the different detection numbers in the dorsal and ventral striatum. 

(3) Quantification in Figure 5. It is recommended that the authors provide quantitative data on cluster formation and colocalization with Sylite puncta in Figure 5 to support their qualitative observations. 

This is an important point that was also raised by the other reviewers. We have performed additional experiments to increase the data volume for analysis. For quantification, we used two approaches. First, we counted the percentage of infected cells in which synaptic localisation of the recombinant receptor subunit was observed (Fig. 5C). We found that mEos4b-GlyRa1 consistently localises at synapses, indicating that all cells express endogenous GlyRb. When neurons were infected with mEos4b-GlyRb, fewer cells had synaptic clusters, meaning that indeed, GlyR alpha subunits are the limiting factor for synaptic targeting. In cultures infected with mEos4b-GlyRa2, only very few neurons displayed synaptic localisation (as judged by epifluorescence imaging). We think this shows that GlyRa2 is less capable of forming heteromeric complexes than GlyRa1, in line with our previous interpretation (see pp. 9-10, 13). 

Secondly, we quantified the total intensity of each subunit at gephyrin-positive domains, both in infected neurons as well as non-infected control cultures (Fig. 5D). We observed that mEos4bGlyRa1 intensity at gephyrin puncta was higher than that of the other subunits, again pointing to efficient synaptic targeting of GlyRa1. Gephyrin cluster intensities (Sylite labelling) were not significantly different in GlyRb and GlyRa2 expressing neurons compared to the uninfected control, indicating that the lentiviral expression of recombinant subunits does not fundamentally alter the size of mixed inhibitory synapses in hippocampal neurons. Interestingly, gephyrin levels were slightly higher in hippocampal neurons expressing mEos4b-GlyRa1. In our view, this comes from an enhanced expression and synaptic targeting of mEos4b-GlyRa1 heteromers with endogenous GlyRb, pointing to a structural role of GlyRa1/b in hippocampal synapses (pp. 10, 13).

The new data and analyses have been described and illustrated in the relevant sections of the manuscript.

(4) Potential for pseudo replication. It's not clear whether they're performing stats tests across biological replica, images, or even synapses. They often quote mean +/- SEM with n = 1000s, and so does that mean they're doing tests on those 1000s? Need to clarify. 

All experiments were repeated at least twice to ensure reproducibility (N independent experiments). Statistical tests were performed on pooled data across the biological replicates; n denotes the number of data points used for testing (e.g., number of synaptic clusters, detections, cells, as specified in each case). We have systematically given these numbers in the revised manuscript (n, N, and other experimental parameters such as the number of animals used, coverslips, images or cells). Data are generally given as mean +/- SEM or as mean +/- SD as indicated.

(5) Does mEoS effect expression levels or function of the protein? Can't see any experiments done to confirm this. Could suggest WB on homogenate, or mass spec? 

The Glrb^eos/eos knock-in mouse line has been characterised previously and does not to display any ultrastructural or functional deficits at inhibitory synapses (Maynard et al. 2021 eLife). GlyRβ expression and glycine-evoked responses were not significantly different to those of the wildtype. The synaptic localisation of mEos4b-GlyRb in KI animals demonstrates correct assembly of heteromeric GlyRs and synaptic targeting. Accordingly, the animals do not display any obvious phenotype. We have clarified this in the manuscript (p. 4). In the case of cultured neurons, long-term expression of fluorescent receptor subunits with lentivirus   has proven ideal to achieve efficient synaptic targeting. The low and continuous supply of recombinant receptors ensures assembly with endogenous subunits to form heteropentameric receptor complexes (e.g. [Patrizio et al. 2017 Sci Rep]). In the present study, lentivirus infection did not induce any obvious differences in the number or size of inhibitory synapses compared to control neurons, as judged by Sylite labelling of synaptic gephyrin puncta (new Fig. 5D).

(6) Quantification of protein numbers is challenging with SMLM. Issues include i) some of FP not correctly folded/mature, and ii) dependence of localisation rate on instrument, excitation/illumination intensities, and also the thresholds used in analysis. Can the authors compare with another protein that has known expression levels- e.g. PSD95? This is quite an ask, but if they could show copy number of something known to compare with, it would be useful. 

We agree that absolute quantification with SMLM is challenging, since the number of detections depends on fluorophore maturation, photophysics, imaging conditions, and analysis thresholds (discussed in Patrizio & Specht 2016, Neurophotonics). For this reason, only very few datasets provide reliable copy numbers, even for well-studied proteins such as PSD-95. One notable exception is the study by Maynard et al. (eLife 2021) that quantified endogenous GlyRβcontaining receptors in spinal cord synapses using SMLM combined with correlative electron microscopy. The strength of this work was the use of a KI mouse strain, which ensures that mEos4b-GlyRβ expression follows intrinsic regional and temporal profiles. The authors reported a stereotypic density of ~2,000 GlyRs/µm² at synapses, corresponding to ~120 receptors per synapse in the dorsal horn and ~240 in the ventral horn, taking into account various parameters including receptor stoichiometry and the functionality of the fluorophore. These values are very close to our own calculations of GlyR numbers at spinal cord synapses that were obtained slightly differently in terms of sample preparation, microscope setup, imaging conditions, and data analysis, lending support to our experimental approach. Nevertheless, the obtained GlyR copy numbers at hippocampal synapses clearly have to be taken as estimates rather than precise figures, because the number of detections from a single mEos4b fluorophore can vary substantially, meaning that the fluorophores are not represented equally in pointillist images. This can affect the copy number calculation for a specific synapse, in particular when the numbers are low (e.g. in hippocampus), however, it should not alter the average number of detections (Fig. 1B) or the (median) molecule numbers of the entire population of synapses (Fig. 1C). We have discussed the limitations of our approach (p. 11).

(7) Rationale for doing nanobody dSTORM not clear at all. They don't explain the reason for doing the dSTORM experiments. Why not just rely on PALM for coincidence measurements, rather than tagging mEoS with a nanobody, and then doing dSTORM with that? Can they explain? Is it to get extra localisations- i.e. multiple per nanobody? If so, localising same FP multiple times wouldn't improve resolution. Also, no controls for nanobody dSTORM experiments- what about non-spec nb, or use on WT sections? 

As discussed above (point 6), the detection of fluorophores with SMLM is influenced by many parameters, not least the noise produced by emitting molecules other than the fluorophore used for labelling. Our study is exceptional in that it attempts to identify extremely low molecule numbers (down to 1). To verify that the detections obtained with PALM correspond to mEos4b, we conducted robust control experiments (including pixel-shift as suggested by the reviewer, see point 1, revised Fig. 1B). The rationale for the nanobody-based dSTORM experiments was twofold: (1) to have an independent readout of the presence of low-copy GlyRs at inhibitory synapses and (2) to analyse the nanoscale organisation of GlyRs relative to the synaptic gephyrin scaffold using dual-colour dSTORM with spectral demixing (see p. 6). The organic fluorophores used in dSTORM (AF647, CF680) ensure high photon counts, essential for reliable co-localisation and distance analysis. PALM and dSTORM cannot be combined in dual-colour mode, as they require different buffers and imaging conditions. 

The specificity of the anti-Eos nanobody was demonstrated by immunohistochemistry in spinal cord cultures expressing mEos4b-GlyRb and wildtype control tissue (Fig. S3). In response to the reviewer's remarks, we also performed a negative control experiment in Glrb^eos/eos slices (dSTORM), in which the nanobody was omitted (new Fig. S4F,G). Under these conditions, spectral demixing produced a single peak corresponding to CF680 (gephyrin) without any AF647 contribution (Fig. S4F). The background detection of "false" AF647 detections at synapses was significantly lower than in the slices labelled with the nanobody. We conclude that the fluorescence signal observed in our dual-colour dSTORM experiments arises from the specific detection of mEos4b-GlyRb by the nanobody, rather than from background, crossreactivity or wrong attribution of colour during spectral demixing. We have added these data and explanations in the results (p. 7) and in the figure legend of Fig. S4F,G.

(8) What resolutions/precisions were obtained in SMLM experiments? Should perform Fourier Ring Correlation (FRC) on SR images to state resolutions obtained (particularly useful for when they're presenting distance histograms, as this will be dependent on resolution). Likewise for precision, what was mean precision? Can they show histograms of localisation precision. 

This is an interesting question in the context of our experiments with low-copy GlyRs, since the spatial resolution of SMLM is limited also by the density of molecules, i.e. the sampling of the structure in question (Nyquist-Shannon criterion). Accordingly, the priority of the PALM experiments was to improve the sensibility of SMLM for the identification of mEos4b-GlyRb subunits, rather than to maximize the spatial resolution. The mean localisation precision in PALM was 33 +/- 12 nm, as calculated from the fitting parameters of each detection (Zeiss, ZEN software), which ultimately result from their signal-to-noise ratio. This is a relatively low precision for SMLM, which can be explained by the low brightness of mEos4b compared to organic fluorophores together with the elevated fluorescence background in tissue slices.

In the case of dSTORM, the aim was to study the relative distribution of GlyRs within the synaptic scaffold, for which a higher localisation precision was required (p. 6). Therefore, detections with a precision ≥ 25 nm were filtered during analysis with NEO software (Abbelight). The retained detections had a mean localisation precision of 12 +/- 5 for CF680 (Sylite) and 11 +/- 4 for AF647 (nanobody). These values are given in the revised manuscript (pp. 18, 22).

(9) Why were DBSCAN parameters selected? How can they rule out multiple localisations per fluor? If low copy numbers (<10), then why bother with DBSCAN? Could just measure distance to each one. 

Multiple detections of the same fluorophore are intrinsic to dSTORM imaging and have not been eliminated from the analysis. Small clusters of detections likely represent individual molecules (e.g. single receptors in the extrasynaptic regions, Fig. 2A). DBSCAN is a robust clustering method that is quite insensitive to minor changes in the choice of parameters. For dSTORM of synaptic gephyrin clusters (CF680), a relatively low length (80 nm radius) together with a high number of detections (≥ 50 neighbours) were chosen to reconstruct the postsynaptic domain with high spatial resolution (see point 8). In the case of the GlyR (nanobody-AF647), the clustering was done mostly for practical reasons, as it provided the coordinates of the centre of mass of the detections. The low stringency of this clustering (200 nm radius, ≥ 5 neighbours) effectively filters single detections that can result from background noise or incorrect demixing. An additional reference explaining the use of DBSCAN including the choice of parameters is given on p. 22 (see also R2 point 4).

(10) For microscopy experiment methods, state power densities, not % or "nominal power". 

Done. We now report the irradiance (laser power density) instead of nominal power (pp. 18, 21). 

(11) In general, not much data presented. Any SI file with extra images etc.? 

The original submission included four supplementary figures with additional data and representative images that should have been available to the reviewer (Figs. S1-S4). The SI file has been updated during revision (new Fig. S4E-G). 

(12) Clarification of the discussion on GlyR expression and synaptic localization: The discussion on GlyR expression, complex formation, and synaptic localization is sometimes unclear, and needs terminological distinctions between "expression level", "complex formation" and "synaptic localization". For example, the authors state:"What then is the reason for the low protein expression of GlyRβ? One possibility is that the assembly of mature heteropentameric GlyR complexes depends critically on the expression of endogenous GlyR α subunits." Does this mean that GlyRβ proteins that fail to form complexes with GlyRα subunits are unstable and subject to rapid degradation? If so, the authors should clarify this point. The statement "This raises the interesting possibility that synaptic GlyRs may depend specifically on the concomitant expression of both α1 and β transcripts." suggests a dependency on α1 and β transcripts. However, is the authors' focus on synaptic localization or overall protein expression levels? If this means synaptic localization, it would be beneficial to state this explicitly to avoid confusion. To improve clarity, the authors should carefully distinguish between these different aspects of GlyR biology throughout the discussion. Additionally, a schematic diagram illustrating these processes would be highly beneficial for readers. 

We thank the reviewer to point this out. We are dealing with several processes; protein expression that determines subunit availability and the assembly of pentameric GlyRs complexes, surface expression, membrane diffusion and accumulation of GlyRb-containing receptor complexes at inhibitory synapses. We have edited the manuscript, particularly the discussion and tried to be as clear as possible in our wording.

We chose not to add a schematic illustration for the time being, because any graphical representation is necessarily a simplification. Instead, we preferred to summarise the main numbers in tabular form (Table 1). We are of course open to any other suggestions.

(13) Interpretation of GlyR localization in the context of nanodomains. The distribution of GlyR molecules on inhibitory synapses appears to be non-homogeneous, instead forming nanoclusters or nanodomains, similar to many other synaptic proteins. It is important to interpret GlyR localization in the context of nanodomain organization. 

The dSTORM images in Fig. 2 are pointillist representations that show individual detections rather than molecules. Small clusters of detections are likely to originate from a single AF647 fluorophore (in the case of nanobody labelling) and therefore represent single GlyRb subunits. Since GlyR copy numbers are so low at hippocampal synapses (≤ 5), the notion of nanodomain is not directly applicable. Our analysis therefore focused on the integration of GlyRs within the postsynaptic scaffold, rather than attempting to define nanodomain structures (see also response to point 8 of R1). A clarification has been added in the revised manuscript (p. 6).

Reviewer #1 (Significance): 

The paper presents biological and technical advances. The biological insights revolve mostly on the documentation of Glycine receptors in particular synapses in forebrain, where they are typically expressed at very low levels. The authors provide compelling data indicating that the expression is of physiological significance. The authors have done a nice job of combining genetically-tagged mice with advanced microscopy methods to tackle the question of distributions of synaptic proteins. Overall these advances are more incremental than groundbreaking. 

We thank the reviewer for acknowledging both the technical and biological advances of our study. While we recognize that our work builds upon established models, we consider that it also addresses important unresolved questions, namely that GlyRs are present and specifically anchored at inhibitory synapses in telencephalic regions, such as the hippocampus and striatum. From a methodological point of view, our study demonstrates that SMLM can be applied not only for structural analysis of highly abundant proteins, but also to reliably detect proteins present at very low copy numbers. This ability to identify and quantify sparse molecule populations adds a new dimension to SMLM applications, which we believe increases the overall impact of our study beyond the field of synaptic neuroscience.

Reviewer #2 (Evidence, reproducibility and clarity): 

In their manuscript "Single molecule counting detects low-copy glycine receptors in hippocampal and striatal synapses" Camuso and colleagues apply single molecule localization microscopy (SMLM) methods to visualize low copy numbers of GlyRs at inhibitory synapses in the hippocampal formation and the striatum. SMLM analysis revealed higher copy numbers in striatum compared to hippocampal inhibitory synapses. They further provide evidence that these low copy numbers are tightly linked to post-synaptic scaffolding protein gephyrin at inhibitory synapses. Their approach profits from the high sensitivity and resolution of SMLM and challenges the controversial view on the presence of GlyRs in these formations although there are reports (electrophysiology) on the presence of GlyRs in these particular brain regions. These new datasets in the current manuscript may certainly assist in understanding the complexity of fundamental building blocks of inhibitory synapses. 

However I have some minor points that the authors may address for clarification: 

(1) In Figure 1 the authors apply PALM imaging of mEos4b-GlyRß (knockin) and here the corresponding Sylite label seems to be recorded in widefield, it is not clearly stated in the figure legend if it is widefield or super-resolved. In Fig 1 A - is the scale bar 5 µm? Some Sylite spots appear to be sized around 1 µm, especially the brighter spots, but maybe this is due to the lower resolution of widefield imaging? Regarding the statistical comparison: what method was chosen to test for normality distribution, I think this point is missing in the methods section. 

This is correct; the apparent size of the Sylite spots does not reflect the real size of the synaptic gephyrin domain due to the limited resolution of widefield imaging including the detection of outof-focus light. We have clarified in the legend of Fig. 1A that Sylite labelling was with classic epifluorescence microscopy. The scale bar in Fig. 1A corresponds to 5 µm. Since the data were not normally distributed, nonparametric tests (Kruskal- Wallis one-way ANOVA with Dunn’s multiple comparison test or Mann-Whitney U-test for pairwise comparisons) were used (p. 23). 

Moreover I would appreciate a clarification and/or citation that the knockin model results in no structural and physiological changes at inhibitory synapses, I believe this model has been applied in previous studies and corresponding clarification can be provided. 

The Glrbeos/eos mouse model has been described previously and does not exhibit any structural or physiological phenotypes (Maynard et al. 2021 eLife). The issue was also raised by reviewer R1 (point 5) and has been clarified in the revised manuscript (p. 4).

(2) In the next set of experiments the authors switch to demixing dSTORM experiments - an explanation why this is performed is missing in the text - I guess better resolution to perform more detailed distance measurements? For these experiments: which region of the hippocampus did the authors select, I cannot find this information in legend or main text. 

Yes, the dSTORM experiments enable dual-colour structural analysis at high spatial resolution (see response to R1 point 7). An explanation has been added (p. 6).

(3) Regarding parameters of demixing experiments: the number of frames (10.000) seems quite low and the exposure time higher than expected for Alexa 647. Can the authors explain the reason for chosing these particular parameters (low expression profile of the target - so better separation?, less fluorophores on label and shorter collection time?) or is there a reference that can be cited? The laser power is given in the methods in percentage of maximal output power, but for better comparison and reproducibility I recommend to provide the values of a power meter (kW/cm2) as lasers may change their maximum output power during their lifetime. 

Acquisition parameters (laser power, exposure time) for dSTORM were chosen to obtain a good localisation precision (~12 nm; see R1 point 8). The number of frames is adequate to obtain well sampled gephyrin scaffolds in the CF680 channel. In the case of the GlyR (nanobody-AF647), the concept of spatial resolution does not really apply due to the low number of targets (see R1, point 13). Power density (irradiance) values have now been given (pp. 18, 21).

(4) For analysis of subsynaptic distribution: how did the authors decide to choose the parameters in the NEO software for DBSCAN clustering - was a series of parameters tested to find optimal conditions and did the analysis start with an initial test if data is indeed clustered (K-ripley) or is there a reference in literature that can be provided? 

DBSCAN parameters were optimised manually, by testing different values. Identification of dense and well-delimited gephyrin clusters (CF680) was achieved with a small radius and a high number of detections (80 nm, ≥ 50 neighbours), whereas filtering of low-density background in the AF647 channel (GlyRs) required less stringent parameters (200 nm, ≥ 5) due to the low number of target molecules. Similar parameters were used in a previous publication (Khayenko et al. 2022, Angewandte Chemie). The reference has been provided on p. 22 (see also R1 point 9).

(5) A conclusion/discussion of the results presented in Figure 5 is missing in the text/discussion. 

This part of the manuscript has been completely overhauled. It includes new experimental data, quantification of the data (new Fig.5), as well as the discussion and interpretation of our findings (see also R1, point 3). In agreement with our earlier interpretation, the data confirm that low availability of GlyRa1 subunits limits the expression and synaptic targeting of GlyRa1/b heteropentamers. The observation that GlyRa1 overexpression with lentivirus increases the size of the postsynaptic gephyrin domain further points to a structural role, whereby GlyRs can enhance the stability (and size) of inhibitory synapses in hippocampal neurons, even at low copy numbers (pp. 13-14). 

(6) In line 552 "suspension" is misleading, better use "solution" 

Done.

Reviewer #2 (Significance): 

Significance: The manuscript provides new insights to presence of low-copy numbers by visualizing them via SMLM. This is the first report that visualizes GlyR optically in the brain applying the knock-in model of mEOS4b tagged GlyRß and quantifies their copy number comparing distribution and amount of GlyRs from hippocampus and striatum. Imaging data correspond well to electrophysiological measurements in the manuscript. 

Field of expertise: Super-Resolution Imaging and corresponding analysis 

Reviewer #4 (Evidence, reproducibility and clarity): 

In this study, Camuso et al., make use of a knock-in mouse model expressing endogenously mEos4b-tagged GlyRβ to detect endogenous glycine receptors using single-molecule localization microscopy. The main conclusion from this study is that in the hippocampus GlyRβ molecules are barely detected, while inhibitory synapses in the ventral striatum seem to express functionally relevant GlyR numbers. 

I have a few points that I hope help to improve the strength of this study. 

- In the hippocampus, this study finds that the numbers of detections are very low. The authors perform adequate controls to indicate that these localizations are above noise level. Nevertheless, it remains questionable that these reflect proper GlyRs. The suggestion that in hippocampal synapses the low numbers of GlyRβ molecules "are important in assembly or maintenance of inhibitory synaptic structures in the brain" is on itself interesting, but is not at all supported. It is also difficult to envision how such low numbers could support the structure of a synapse. A functional experiment showing that knockdown of GlyRs affects inhibitory synapse structure in hippocampal neurons would be a minimal test of this. 

It is not clear what the reviewer means by “it remains questionable that these reflect proper GlyRs”. The PALM experiments include a series of stringent controls (see R1, point 1) demonstrating the existence of low-copy GlyRs at inhibitory synapses in the hippocampus (Fig. 1) and in the striatum (Fig. 3), and are backed up by dSTORM experiments (Fig. 2). We have no reason to doubt that these receptors are fully functional (as demonstrated for the ventral striatum (Fig. 4). However, due to their low number, a role in inhibitory synaptic transmission is clearly limited, at least in the hippocampus and dorsal striatum. 

We therefore propose a structural role, where the GlyRs could be required to stabilise the postsynaptic gephyrin domain in hippocampal neurons. This is based on the idea that the GlyRgephyrin affinity is much higher than that of the GABAAR-gephyrin interaction (reviewed in Kasaragod & Schindelin 2018 Front Mol Neurosci). Accordingly, there is a close relationship between GlyRs and gephyrin numbers, sub-synaptic distribution, and dynamics in spinal cord synapses that are mostly glycinergic (Specht et al. 2013 Neuron; Maynard et al. 2021 eLife; Chapdelaine et al. 2021 Biophys J). It is reasonable to assume that low-copy GlyRs could play a similar structural role at hippocampal synapses. A knockdown experiment targeting these few receptors is technically very challenging and beyond the scope of this study. However, in response to the reviewer's question we have conducted new experiments in cultured hippocampal neurons (new Fig. 5). They demonstrate that overexpression of GlyRa1/b heteropentamers increases the size of the postsynaptic domain in these neurons, supporting our interpretation of a structural role of low-copy GlyRs (p. 14).

- The endogenous tagging strategy is a very strong aspect of this study and provides confidence in the labeling of GlyRβ molecules. One caveat however, is that this labeling strategy does not discriminate whether GlyRβ molecules are on the cell membrane or in internal compartments. Can the authors provide an estimate of the ratio of surface to internal GlyRβ molecules? 

Gephyrin is known to form a two-dimensional scaffold below the synaptic membrane to which inhibitory GlyRs and GABAARs attach (reviewed in Alvarez 2017 Brain Res). The majority of the synaptic receptors are therefore thought to be located in the synaptic membrane, which is supported by the close relationship between the sub-synaptic distribution of GlyRs and gephyrin in spinal cord neurons (e.g. Maynard et al. 2021 eLife). To demonstrate the surface expression of GlyRs at hippocampal synapses we labelled cultured hippocampal neurons expressing mEos4b-GlyRa1 with anti-Eos nanobody in non-permeabilised neurons (see Author response image 1). The close correspondence between the nanobody (AF647) and the mEos4b signal confirms that the majority of the GlyRs are indeed located in the synaptic membrane.

Author response image 1.

Left: Lentivirus expression of mEos4b-GlyRa1 in fixed and non-permeabilised hippocampal neurons (mEos4b signal). Right: Surface labelling of the recombinant subunit with anti-Eos nanoboby (AF647). 

- “We also estimated the absolute number of GlyRs per synapse in the hippocampus. The number of mEos4b detections was converted into copy numbers by dividing the detections at synapses by the average number of detections of individual mEos4b-GlyRβ containing receptor complexes”. In essence this is a correct method to estimate copy numbers, and the authors discuss some of the pitfalls associated with this approach (i.e., maturation of fluorophore and detection limit). Nevertheless, the authors did not subtract the number of background localizations determined in the two negative control groups. This is critical, particularly at these low-number estimations. 

We fully agree that background subtraction can be useful with low detection numbers. In the revised manuscript, copy numbers are now reported as background-corrected values. Specifically, the mean number of detections measured in wildtype slices was used to calculate an equivalent receptor number, which was then subtracted from the copy number estimates across hippocampus, spinal cord and striatum. This procedure is described in the methods (p. 20) and results (p. 5, 8), and mentioned in the figure legends of Fig. 1C, 3C. The background corrected values are given in the text and Table 1.

- Furthermore, the authors state that "The advantage of this estimation is that it is independent of the stoichiometry of heteropentameric GlyRs". However, if the stoichometry is unknown, the number of counted GlyRβ subunits cannot simply be reported as the number of GlyRs. This should be discussed in more detail, and more carefully reported throughout the manuscript. 

The reviewer is right to point this out. There is still some debate about the stoichiometry of heteropentameric GlyRs. Configurations with 2a:3b, 3a:2b and 4a:1b subunits have been advanced (e.g. Grudzinska et al. 2005 Neuron; Durisic et al. 2012 J Neurosci; Patrizio et al. 2017 Sci Rep; Zhu & Gouaux 2021 Nature). We have therefore chosen a quantification that is independent of the underlying stoichiometry. Since our quantification is based on very sparse clusters of mEos4b detections that likely originate from a single receptor complex (irrespective of its stoichiometry), the reported values actually reflect the number of GlyRs (and not GlyRb subunits). We have clarified this in the results (p. 5) and throughout the manuscript (Table 1). 

- The dual-color imaging provides insights in the subsynaptic distribution of GlyRβ molecules in hippocampal synapses. Why are similar studies not performed on synapses in the ventral striatum where functionally relevant numbers of GlyRβ molecules are found? Here insights in the subsynaptic receptor distribution would be of much more interest as it can be tight to the function. 

This is an interesting suggestion. However, the primary aim of our study was to identify the existence of GlyRs in hippocampal regions. At low copy numbers, the concept of sub-synaptic domains (SSDs, e.g. Yang et al. 2021 EMBO Rep) becomes irrelevant (see R1 point 13). It should be pointed out that the dSTORM pointillist images (Fig. 2A) represent individual GlyR detections rather than clusters of molecules. In the striatum, our specific purpose was to solve an open question about the presence of GlyRs in different subregions (putamen, nucleus accumbens).

- It is unclear how the experiments in Figure 5 add to this study. These results are valid, but do not seem to directly test the hypothesis that "the expression of α subunits may be limiting factor controlling the number of synaptic GlyRs". These experiments simply test if overexpressed α subunits can be detected. If the α subunits are limiting, measuring the effect of α subunit overexpression on GlyRβ surface expression would be a more direct test. 

Both R1 and R2 have also commented on the data in Fig. 5 and their interpretation. We have substantially revised this section as described before (see R1 point 3) including additional experiments and quantification of the data (new Fig. 5). The findings lend support to our earlier hypothesis that GlyR alpha subunits (in particular GlyRa1) are the limiting factor for the expression of heteropentameric GlyRa/b in hippocampal neurons (pp. 13-14). Since the GlyRa1 subunit itself does not bind to gephyrin (Patrizio et al. 2017 Sci Rep), the synaptic localisation of the recombinant mEos4b-GlyRa1 subunits is proof that they have formed heteropentamers with endogenous GlyRb subunits and driven their membrane trafficking, which the GlyRb subunits are incapable of doing on their own.

Reviewer #4 (Significance): 

These results are based on carefully performed single-molecule localization experiments, and are well-presented and described. The knockin mouse with endogenously tagged GlyRβ molecules is a very strong aspect of this study and provides confidence in the labeling, the combination with single-molecule localization microscopy is very strong as it provides high sensitivity and spatial resolution. 

The conceptual innovation however seems relatively modest, these results confirm previous studies but do not seem to add novel insights. This study is entirely descriptive and does not bring new mechanistic insights. 

This study could be of interest to a specialized audience interested in glycine receptor biology, inhibitory synapse biology and super-resolution microscopy. 

My expertise is in super-resolution microscopy, synaptic transmission and plasticity 

As we have stated before, the novelty of our study lies in the use of SMLM for the identification of very small numbers of molecules, which requires careful control experiments. This is something that has not been done before and that can be of interest to a wider readership, as it opens up SMLM for ultrasensitive detection of rare molecular events. Using this approach, we solve two open scientific questions: (1) the demonstration that low-copy GlyRs are present at inhibitory synapses in the hippocampus, (2) the sub-region specific expression and functional role of GlyRs in the ventral versus dorsal striatum.

The following review was provided later under the name “Reviewer #4”. To avoid confusion with the last reviewer from above we will refer to this review as R4-2.

Reviewer #4-2 (Evidence, reproducibility and clarity):  

Summary:

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

The authors investigate the presence of synaptic glycine receptors in the telencephalon, whose presence and function is poorly understood. 

Using a transgenically labeled glycine receptor beta subunit (Glrb-mEos4b) mouse model together with super-resolution microscopy (SLMM, dSTORM), they demonstrate the presence of a low but detectable amount of synaptically localized GLRB in the hippocampus. While they do not perform a functional analysis of these receptors, they do demonstrate that these subunits are integrated into the inhibitory postsynaptic density (iPSD) as labeled by the scaffold protein gephyrin. These findings demonstrate that a low level of synaptically localized glycerine receptor subunits exist in the hippocampal formation, although whether or not they have a functional relevance remains unknown.

They then proceed to quantify synaptic glycine receptors in the striatum, demonstrating that the ventral striatum has a significantly higher amount of GLRB co-localized with gephyrin than the dorsal striatum or the hippocampus. They then recorded pharmacologically isolated glycinergic miniature inhibitory postsynaptic currents (mIPSCs) from striatal neurons. In line with their structural observations, these recordings confirmed the presence of synaptic glycinergic signaling in the ventral striatum, and an almost complete absence in the dorsal striatum. Together, these findings demonstrate that synaptic glycine receptors in the ventral striatum are present and functional, while an important contribution to dorsal striatal activity is less likely.

Lastly, the authors use existing mRNA and protein datasets to show that the expression level of GLRA1 across the brain positively correlates with the presence of synaptic GLRB.

The authors use lentiviral expression of mEos4b-tagged glycine receptor alpha1, alpha2, and beta subunits (GLRA1, GLRA1, GLRB) in cultured hippocampal neurons to investigate the ability of these subunits to cause the synaptic localization of glycine receptors. They suggest that the alpha1 subunit has a higher propensity to localize at the inhibitory postsynapse (labeled via gephyrin) than the alpha2 or beta subunits, and may therefore contribute to the distribution of functional synaptic glycine receptors across the brain.

Major comments:

- Are the key conclusions convincing?

The authors are generally precise in the formulation of their conclusions.

(1) They demonstrate a very low, but detectable, amount of a synaptically localized glycine receptor subunit in a transgenic (GlrB-mEos4b) mouse model. They demonstrate that the GLRB-mEos4b fusion protein is integrated into the iPSD as determined by gephyrin labelling. The authors do not perform functional tests of these receptors and do not state any such conclusions.

(2) The authors show that GLRB-mEos4b is clearly detectable in the striatum and integrated into gephyrin clusters at a significantly higher rate in the ventral striatum compared to the dorsal striatum, which is in line with previous studies.

(3) Adding to their quantification of GLRB-mEos4b in the striatum, the authors demonstrate the presence of glycinergic miniature IPSCs in the ventral striatum, and an almost complete absence of mIPSCs in the dorsal striatum. These currents support the observation that GLRB-mEos4b is more synaptically integrated in the ventral striatum compared to the dorsal striatum.

(4) The authors show that lentiviral expression of GLRA1-mEos4b leads to a visually higher number of GLR clusters in cultured hippocampal neurons, and a co-localization of some clusters with gephyrin. The authors claim that this supports the idea that GLRA1 may be an important driver of synaptic glycine receptor localization. However, no quantification or statistical analysis of the number of puncta or their colocalization with gephyrin is provided for any of the expressed subunits. Such a claim should be supported by quantification and statistics 

A thorough analysis and quantification of the data in Fig.5 has been carried out as requested by all the other reviewers (e.g. R1, point 3). The new data and results have been described in the revised manuscript (pp. 9-10, 13-14).

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

One unaddressed caveat is the fact that a GLRB-mEos4b fusion protein may behave differently in terms of localization and synaptic integration than wild-type GLRB. While unlikely, it is possible that mEos4b interacts either with itself or synaptic proteins in a way that changes the fused GLRB subunit’s localization. Such an effect would be unlikely to affect synaptic function in a measurable way, but might be detected at a structural level by highly sensitive methods such as SMLM and STORM in regions with very low molecule numbers (such as the hippocampus). Since reliable antibodies against GLRB in brain tissue sections are not available, this would be difficult to test. Considering that no functional measures of the hippocampal detections exist, we would suggest that this possible caveat be mentioned for this particular experiment.

This question has also been raised before (R1, point 5). According to an earlier study the mEos4b-GlyRb knock-in does not cause any obvious phenotypes, with the possible exception of minor loss of glycine potency (Maynard et al. 2021 eLife). The fact that the synaptic levels in the spinal cord in heterozygous animals are precisely half of those of homozygous animals argues against differences in receptor expression, heteropentameric assembly, forward trafficking to the plasma membrane and integration into the synaptic membrane as confirmed using quantitative super-resolution CLEM (Maynard et al. 2021 eLife). Accordingly, we did not observe any behavioural deficits in these animals, making it a powerful experimental model. We have added this information in the revised manuscript (p. 4). 

In addition, without any quantification or statistical analysis, the author’s claims regarding the necessity of GLRA1 expression for the synaptic localization of glycine receptors in cultured hippocampal neurons should probably be described as preliminary (Fig. 5).

As mentioned before, we have substantially revised this part (R1, point 3). The quantification and analysis in the new Fig. 5 support our earlier interpretation.

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

The authors show that there is colocalization of gephyrin with the mEos4b-GlyRβ subunit using the Dual-colour SMLM. This is a powerful approach that allows for a claim to be made on the synaptic location of the glycine receptors. The images presented in Figure 1, together with the distance analysis in Figure 2, display the co-localization of the fluorophores. The co-localization images in all the selected regions, hippocampus and striatum, also show detections outside of the gephyrin clusters, which the authors refer to as extrasynaptic. These punctated small clusters seem to have the same size as the ones detected and assigned as part of the synapse. It would be informative if the authors analysed the distribution, density and size of these nonsynaptic clusters and presented the data in the manuscript and also compared it against the synaptic ones. Validating this extrasynaptic signal by staining for a dendritic marker, such as MAP-2 or maybe a somatic marker and assessing the co-localization with the non-synaptic clusters would also add even more credibility to them being extrasynaptic. 

The existence of extrasynaptic GlyRs is well attested in spinal cord neurons (e.g. Specht et al. 2013 Neuron; this study see Fig. S2). The fact that these appear as small clusters of detections in SMLM recordings results from the fact that a single fluorophore can be detected several times in consecutive image frames and because of blinking. Therefore, small clusters of detections likely represent single GlyRs (that can be counted), and not assemblies of several receptor complexes. Due to their diffusion in the neuronal membrane, they are seen as diffuse signals throughout the somatodendritic compartment in epifluorescence images (e.g. Fig. 5A). SMLM recordings of the same cells resolves this diffuse signal into discrete nanoclusters representing individual receptors (Fig. 5B). It is not clear what information co-localisation experiments with specific markers could provide, especially in hippocampal neurons, in which the copy numbers (and density) of GlyRs is next to zero.

In addition we would encourage the authors to quantify the clustering and co-localization of virally expressed GLRA1, GLRA2, and GLRB with gephyrin in order to support the associated claims (Fig. 5). Preferably, the density of GLR and gephyrin clusters (at least on the somatic surface, the proximal dendrites, or both) as well as their co-localization probability should be quantified if a causal claim about subunit-specific requirements for synaptic localization is to be made.

Quantification of the data have been carried out (new Fig.5C,D). The results have been described before (R1, point 3) and support our earlier interpretation of the data (pp. 13-14).

Lastly, even though it may be outside of the scope of such a study analysing other parts of the hippocampal area could provide additional important information. If one looks at the Allen Institute’s ISH of the beta subunit the strongest signal comes from the stratum oriens in the CA1 for example, suggesting that interneurons residing there would more likely have a higher expression of the glycine receptors. This could also be assessed by looking more carefully at the single cell transcriptomics, to see which cell types in the hippocampus show the highest mRNA levels. If the authors think that this is too much additional work, then perhaps a mention of this in the discussion would be good. 

We have added the requested information from the ISH database of the Allen Institute in the discussion as suggested by the reviewer (p. 12). However, in combination with the transcriptomic data (Fig. S1) our finding strongly suggest that the expression of synaptic GlyRs depends on the availability of alpha subunits rather than on the presence of the GlyRb transcript. This is obvious when one compares the mRNA levels in the hippocampus with those in the basal ganglia (striatum) and medulla. While the transcript concentrations of GlyRb are elevated in all three regions and essentially the same, our data show that the GlyRb copy numbers at synapses differ over more than 2 orders of magnitude (Fig. 1B, Table 1). 

- Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

Since the labeling and some imaging has been performed already, the requested experiment would be a matter of deploying a method of quantification. In principle, it should not require any additional wet-lab experiments, although it may require additional imaging of existing samples.

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

Yes, for the most part.

- Are the experiments adequately replicated and statistical analysis adequate?

Yes

Minor comments:

- Specific experimental issues that are easily addressable.

N/A

- Are prior studies referenced appropriately?

Yes

- Are the text and figures clear and accurate?

Yes, although quantification in figure 5 is currently not present.

A quantification has been added (see R1, point 3).

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

This paper presents a method that could be used to localize receptors and perhaps other proteins that are in low abundance or for which a detailed quantification is necessary. I would therefore suggest that Figure S4 is included into Figure 2 as the first panel, showcasing the demixing, followed by the results. 

We agree in principle with this suggestion. However, the revised Fig. S4 is more complex and we think that it would distract from the data shown in Fig. 2. Given that Fig. S4 is mostly methodological and not essential to understand the text, we have kept it in the supplement for the time being. We leave the final decision on this point to the editor.

Reviewer #4-2 (Significance): 

[This review was supplied later]

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

Using a novel and high resolution method, the authors have provided strong evidence for the presence of glycine receptors in the murine hippocampus and in the dorsal striatum. The number of receptors calculated is small compared to the numbers found in the ventral striatum. This is the first study to quantify receptor numbers in these region. In addition it also lays a roadmap for future studies addressing similar questions. 

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

This is done well by the authors in the curation of the literature. As stated above, the authors have filled a gap in the presence of glycine receptors in different brain regions, a subject of importance in understanding the role they play in brain activity and function. 

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

Neuroscientists working at the synaptic level, on inhibitory neurotransmission and on fundamental mechanisms of expression of genes at low levels and their relationship to the presence of the protein would be interested. Furthermore, researchers in neuroscience and cell biology may benefit from and be inspired by the approach used in this manuscript, to potentially apply it to address their own aims. 

We thank the reviewer for the positive assessment of the technical and biological implications of our work, as well as the interest of our findings to a wide readership of neuroscientists and cell biologists. 

- Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

Synaptic transmission, inhibitory cells and GABAergic synapses functionally and structurally, cortex and cortical circuits. No strong expertise in super-resolution imaging methods.

Focus on Political Repression in Southern Africa was a serial publication produced by the International Defense and Aid Fund (IDAF), an organization dedicated to documenting and challenging apartheid-era injustice. Published from the 1960s through the 1980s, Focus provided clear, accessible reporting on political trials, security laws, censorship, prisoner conditions, and everyday experiences of state repression across South Africa and neighboring countries.

Written for an international audience, the publication combined investigative reporting, legal analysis, and first-hand testimony to expose abuses that were otherwise hidden from the global public. This digital collection brings together issues of Focus to support research, teaching, and general understanding of how apartheid systems operated—and how people and organizations resisted them.

Reviewer #3 (Public review):

Summary:

This work provides an overview of the motor neuron landscape in the male reproductive system. Some work had been done to elucidate the circuits of ejaculation in the spine, as well as, the cord but this work fills a gap of knowledge at the level of the reproductive organs. Using complementary approaches the authors show that there are two types of motor neurons that are mutually exclusive: neurons that co-express octopamine and glutamate and neurons that co-express serotonin and glutamate. They also show evidence that both types of neurons express large dense core vesicles indicating that neuropeptides play a role in male fertility. This paper provides a thorough characterization of expression of the different glutamate, octopamine and serotonin receptors in the different organs and tissues of the male reproductive system. The differential expression in different tissues and organs allows building initial theories on the control of emission and expulsion. Additionally, the authors characterize the expression of synaptic proteins and the neuromuscular junction sites. On a mechanistic level, the authors show that neither octopamine/glutamate neuron transmission nor glutamate transmission in serotonin/glutamate neurons are required for male fertility. This final result is quite surprising and opens up many questions on how ejaculation is coordinated.

Strengths:

This work fills an important gap on characterization of innervation of the male reproductive system by providing an extensive characterization of the motor neurons and the potential receptors of motor neuron release.The authors show convincing evidence of glutamate/monoamine co-release and of mutual exclusivity of serotonin/glutamate and octopamine/glutamate neurons.

Weaknesses:

The experiment looking at peristaltic waves in the male organs is missing labeling of the different regions and quantification of the observed waves.

Author response:

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

Reviewer #1 (Public review): 

Summary: 

This very thorough anatomical study addresses the innervation of the Drosophila male reproductive tract. Two distinct glutamatergic neuron types were classified: serotonergic (SGNs) and octopaminergic (OGNs). By expansion microscopy, it was established that glutamate and serotonin /octopamine are co-released. The expression of different receptors for 5-HT and OA in muscles and epithelial cells of the innervation target organs was characterized. The pattern of neurotransmitter receptor expression in the target organs suggests that seminal fluid and sperm transport and emission are subjected to complex regulation. While silencing of abdominal SGNs leads to male infertility and prevents sperm from entering the ejaculatory duct, silencing of OGNs does not render males infertile. 

Strengths: 

The studied neurons were analysed with different transgenes and methods, as well as antibodies against neurotransmitter synthesis enzymes, building a consistent picture of their neurotransmitter identity. The careful anatomical description of innervation patterns together with receptor expression patterns of the target organs provides a solid basis for advancing the understanding of how seminal fluid and sperm transport and emission are subjected to complex regulation. The functional data showing that SGNs are required for male fertility and for the release of sperm from the seminal vesicle into the ejaculatory duct is convincing. 

Weaknesses: 

The functional analysis of the characterized neurons is not as comprehensive as the anatomical description, and phenotypic characterization was limited to simple fertility assays. It is understandable that a full functional dissection is beyond the scope of the present work. The paper contains experiments showing neuron-independent peristaltic waves in the reproductive tract muscles, which are thematically not very well integrated into the paper. Although very interesting, one wonders if these experiments would not fit better into a future work that also explores these peristaltic waves and their interrelation with neuromodulation mechanistically. 

Reviewer #2 (Public review): 

Summary: 

Cheverra et al. present a comprehensive anatomical and functional analysis of the motor neurons innervating the male reproductive tract in Drosophila melanogaster, addressing a gap in our understanding of the peripheral circuits underlying ejaculation and male fertility. They identify two classes of multi-transmitter motor neurons-OGNs (octopamine/glutamate) and SGNs (serotonin/glutamate)-with distinct innervation patterns across reproductive organs. The authors further characterize the differential expression of glutamate, octopamine, and serotonin receptors in both epithelial and muscular tissues of these organs. Behavioral assays reveal that SGNs are essential for male fertility, whereas OGNs and glutamatergic transmission are dispensable. This work provides a high-resolution map linking neuromodulatory identity to organ-specific motor control, offering a valuable framework to explore the neural basis of male reproductive function. 

Strengths: 

Through the use of an extensive set of GAL4 drivers and antibodies, this work successfully and precisely defines the neurons that innervate the male reproductive tract, identifying the specific organs they target and the nature of the neurotransmitters they release. It also characterizes the expression patterns and localization of the corresponding neurotransmitter receptors across different tissues. The authors describe two distinct groups of dual-identity neurons innervating the male reproductive tract: OGNs, which co-express octopamine and glutamate, and SGNs, which co-express serotonin and glutamate. They further demonstrate that the various organs within the male reproductive system differentially express receptors for these neurotransmitters. Based on these findings, the authors propose that a single neuron capable of co-releasing a fast-acting neurotransmitter alongside a slower-acting one may more effectively synchronize and stagger events that require precise timing. This, together with the differential expression of ionotropic glutamate receptors and metabotropic aminergic receptors in postsynaptic muscle tissue, adds an additional layer of complexity to the coordinated regulation of fluid secretion, organ contractility, and directional sperm movement-all contributing to the optimization of male fertility. 

Weaknesses: 

The main weakness of the manuscript is the lack of detail in the presentation of the results. Specifically, all microscopy image figures are missing information about the number of samples (N), and in the case of colocalization experiments, quantitative analyses are not provided. Additionally, in the first behavioral section, it would be beneficial to complement the data table with figures similar to those presented later in the manuscript for consistency and clarity. 

Wider context: 

This study delivers the first detailed anatomical map connecting multi-transmitter motor neurons with specific male reproductive structures. It highlights a previously unrecognized functional specialization between serotonergic and octopaminergic pathways and lays the groundwork for exploring fundamental neural mechanisms that regulate ejaculation and fertility in males. The principles uncovered here may help explain how males of Drosophila and other organisms adjust reproductive behaviors in response to environmental changes. Furthermore, by shedding light on how multi-transmitter systems operate in reproductive control, this model could provide insights into therapeutic targets for conditions such as male infertility and prostate cancer, where similar neuronal populations are involved in humans. Ultimately, this genetically accessible system serves as a powerful tool for uncovering how multi-transmitter neurons orchestrate coordinated physiological actions necessary for the functioning of complex organs. 

Reviewer #3 (Public review): 

Summary: 

This work provides an overview of the motor neuron landscape in the male reproductive system. Some work had been done to elucidate the circuits of ejaculation in the spine, as well as the cord, but this work fills a gap in knowledge at the level of the reproductive organs. Using complementary approaches, the authors show that there are two types of motor neurons that are mutually exclusive: neurons that co-express octopamine and glutamate and neurons that co-express serotonin and glutamate. They also show evidence that both types of neurons express large dense core vesicles, indicating that neuropeptides play a role in male fertility. This paper provides a thorough characterization of the expression of the different glutamate, octopamine, and serotonin receptors in the different organs and tissues of the male reproductive system. The differential expression in different tissues and organs allows building initial theories on the control of emission and expulsion. Additionally, the authors characterize the expression of synaptic proteins and the neuromuscular junction sites. On a mechanistic level, the authors show that neither octopamine/glutamate neuron transmission nor glutamate transmission in serotonin/glutamate neurons is required for male fertility. This final result is quite surprising and opens up many questions on how ejaculation is coordinated. 

Strengths: 

This work fills an important gap in the characterization of innervation of the male reproductive system by providing an extensive characterization of the motor neurons and the potential receptors of motor neuron release. The authors show convincing evidence of glutamate/monoamine co-release and of mutual exclusivity of serotonin/glutamate and octopamine/glutamate neurons. 

Weaknesses: 

(1) Often, it is mentioned that the expression is higher or lower or regional without quantification or an indication of the number of samples analysed. 

(2) The experiment aimed at tracking sperm in the male reproductive system is difficult to interpret when it is not assessed whether ejaculation has occurred. 

(3) The experiment looking at peristaltic waves in the male organs is missing labeling of the different regions and quantification of the observed waves. 

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors): 

(1) While the peripheral innervations are very carefully described, it is not clear to which SGNs and OGNs (i.e., cell bodies in the central nervous system) these innervations belong. Are SV, AG, and ED innervated by branches of one neuron or by separate neurons? Multi-color flip-out experiments could provide an answer to this. 

We agree this is important and are planning these experiments for follow-up study.

(2) In contrast, for the analysis of the VT19028 split line (Figure 9), only vnc and cell body images are shown. How do the arborisations of these split combinations look in the periphery? Are the same reproductive organs innervated as shown in Figure 2?

Figure 9S3 was inadvertently omitted from the initial submission.  That figure is now included and shows that the VT019028 split broadly innervates the SV, AG, and ED.

(3) In the discussion, I think it would be helpful to offer some potential explanations for the role of octopaminergic and glutamatergic signaling. If not required for basic fertility, they probably have some other role.

Thank you, we have included speculation in the Discussion section "Potential for adaptation to environment".

(4) Line 543: Figure 8S4 E, (not 8E). 

Correction made.

Reviewer #2 (Recommendations for the authors): 

(1) Line 213-217 

Comment:

The use of "significantly less expression" may be misleading, as no quantification or statistical analysis is provided to support this comparison. 

Suggestion:

Consider using a more neutral term, such as "markedly less" or "noticeably less," unless quantitative data and statistical analysis are included to substantiate the claim.

Good recommendation.This suggestion has been incorporated.

(2) Line 264-267 

Comment:

The observation regarding the distinct morphology of SGNs and OGNs is interesting and could strengthen the argument regarding functional differences. 

Suggestion: 

Consider including a quantification of morphological complexity (e.g., branching) to support the claim. A method such as Sholl analysis (Sholl, 1953), as adapted in Fernández et al., 2008, could be applied. 

This is a good suggestion, and we will consider it as part of a follow-up study.

(3) Line 269-271 

Comment:

The anatomical context of the observation is not explicitly stated. 

Suggestion:

Add "in the ED" for clarity: "With the TRH-GAL4 experiment in the ED, vGlut-40XMYC (Figure 5S1, A and E) and 6XV5-vMAT (Figure 5S1, B and F) were both present with a highly overlapping distribution (Figure 5S1, I)." 

Suggestion has been incorporated.

(4) Line 275-276 

Comment:

The claim about the reduced ability to distinguish SGNs and OGNs in the ED would benefit from quantitative support. 

Suggestion:

Include a morphological comparison or quantification between SGNs and OGNs in the ED and SV to reinforce this point.

Certain information on morphological comparison can be inferred within the images themselves, and we will include quantitation in a follow-up study.

(5) Line 277-279 

Comment:

As with line 269, the anatomical site could be specified more clearly. 

Suggestion: 

Rephrase as: "With the Tdc2-GAL4 experiment in the ED, vGlut-40XMYC (Figure 5S1, M and Q) and 6XV5-vMAT (Figure 5S1, N and R) were both observed in a highly overlapping distribution (Figure 5S1, U)." 

Suggestion has been incorporated.

(6) Line 348-350 

Comment:

The phrase "significantly higher density" implies a statistical comparison that is not shown. 

Suggestion:

If no quantification is provided, replace with a qualitative term such as "visibly higher" or "notably more dense." Alternatively, add a quantitative analysis with statistical testing to justify the use of "significantly." 

Suggestion has been incorporated.

(7) Lines 415-458 (Section comment) 

Comment:

There appears to be differential localization of neurotransmitter receptor expression (glutamate in muscle vs. 5-HT in epithelium or neurons), which could have functional implications. 

Suggestion:

Expand this section to briefly discuss the differential localization patterns of these receptors and potential implications for signal transduction in male reproductive tissues. 

(8) Lines 638-682 (Section comment) 

Comment:

The table summarizing fertility phenotypes would be more informative with additional detail on experimental outcomes. 

Suggestion:

Add a column showing the number of fertile males over the total tested (e.g., "n fertile / n total"). Also, clarify whether the fertility assays are identical to those reported in Figure 10S2, and whether similar analyses were conducted for females. Consider including a figure summarizing fertility results for all genotypes listed in the table, similar to Figure 10S2. 

The fertility tests reported in Table 1 were separate from those reported in Figure 10S2.  For these tests, the results were clear-cut with 100% of males and females reported as infertile exhibiting the infertile phenotype.  For the males and females reported as fertile, it was also clear-cut with nearly 100% showing fertility at a high level.  In subsequent figures we attempted to assess degrees of fertility.

(9) Line 724-727 

Comment:

There seems to be a mistake in the identification of the driver lines used to silence OA neurons. Also, figure references might be incorrect. 

Suggestion:

The OA neuron driver line should be corrected to "Tdc2-GAL4-DBD ∩ AbdB-AD" instead of TRH-GAL4. Additionally, the figure references should be verified; specifically, the letter "B" (in "Figure 10B, D" and "10B, E") appears to be unnecessary or misplaced.

Thanks for catching this, the corrections have been made.

(10) Line 872-877 

Comment:

The discussion on the co-release of fast-acting glutamate and slower aminergic neurotransmitters is interesting and well-articulated. However, it remains somewhat disconnected from the behavioral findings. 

Suggestion:

Consider linking this proposed mechanism to the results observed in the mating duration assays. For instance, the sequential action of neurotransmitters described here could potentially underlie the prolonged mating observed when specific neuromodulators are active, helping to functionally integrate molecular and behavioral data. 

(11) Line 926-928 

Comment:

The interpretation of 5-HT7 receptor expression in the sphincter is compelling, suggesting a role in regulating its function. However, this anatomical observation could be further contextualized with the functional data. 

Suggestion:

It may strengthen the interpretation to explicitly connect this finding with the fertility assays, where SGNs - presumably acting via serotonergic signaling - are shown to be necessary for male fertility. This would support a functional role for 5-HT7 in reproductive success via sphincter regulation.

This has been added. 

(12) Figure 1 

Comment:

The figure legend is generally clear, but could benefit from more consistency and precision in the color-coded labeling. Additionally, the naming of some structures could be more explicit. 

Suggestion: 

Revise the figure and the legend as follows:

Figure 1. The Drosophila male reproductive system. A) Schematic diagram showing paired testes (colour), SVs (green), AGs (purple), Sph (red), ED (gray), and EB (colour). B) Actual male reproductive system. Te - testes, SV - seminal vesicle, AG - accessory gland, Sph - singular sphincter, ED - ejaculatory duct, EB - ejaculatory bulb. Scale bar: 200 µm.

This suggestion has been incorporated.

(13) Figure 3S2 

Comment:

There appears to be a typographical error in the description of the genotypes, which may lead to confusion. 

Suggestion:

Correct the legend to reflect the appropriate genotypes:

Figure 3S2. Expression of vGlut-LexA and Tdc2-GAL4 in the Drosophila male reproductive system. A, D, G, J, M, P) vGlut-LexA, LexAop-6XmCherry; B, E, H, K, N, Q) Tdc2-GAL4, UAS-6XGFP; C, F, I, L, O, R) Overlay. Scale bars: O - 50 µm; R - 10 µm.

The corrections have been made.

(14) Figure 3S3

Comment:

The genotypes for panels D and E appear to be incomplete; the DBD component of the split-GAL4 drivers is missing. 

Suggestion:

Update the figure legend to: 

Figure 3S3. Fruitless and Doublesex expression in the Drosophila male reproductive system. A) fru-GAL4, UAS-6XGFP; B) vGlut-LexA, LexAop-6XmCherry; C) Overlay; D) Tdc2-AD ∩ dsx-GAL4-DBD; E) TRH-AD ∩ dsx-GAL4-DBD. Scale bar: 200 µm.

The corrections have been made.

(15) Figure 4S4 

Comment: 

There is a repeated segment in the figure legend, which makes it unclear and redundant. 

Suggestion:

Edit the legend to remove the duplicated lines: 

Figure 4S4. Expression of vGlut, TβH-GFP, and 5-HT at the junction of the SV and AGs with the ED of the Drosophila male reproductive system. A) vGlut-40XV5; B) TβH-GFP; C) 5-HT; D) vGlut-40XV5, TβH-GFP overlay; E) vGlut-40XV5, 5-HT overlay; F) TβH-GFP, 5-HT overlay. Scale bar: 50 µm.

The correction has been made.

(16) Figure 6S5 

Comment:

Within this figure, the orientation and/or scale of the tissue varies noticeably between individual panels, making it difficult to directly compare the different experimental conditions. 

Suggestion:

For improved clarity and interpretability, consider standardizing the orientation and size of the tissue shown across all panels within the figure. Consistent presentation will facilitate direct comparisons between treatments or genotypes. 

There is often variation in the size of the male reproductive organs. They were all acquired at the same magnification. The only point of this figure is there is no vGAT or vAChT at these NMJs and the result is unambiguously negative. 

(17) Figure 10 

Comment:

Panel A appears redundant, as it shows the same information as the other panels but without indicating statistical significance. 

Suggestion:

Consider removing panel A and keeping only the remaining four graphs, which include relevant statistical comparisons and clearly show significant differences.

We realize there is some redundancy of panel A with the other panels, but we feel there is value in having all the genotypes in a single panel for comparison.

Reviewer #3 (Recommendations for the authors): 

Here are some suggestions to improve the manuscript: 

(1) Prot B GFP experiment: the authors should explain better the time chosen to look at the sperm content of the male reproductive system. At 10 minutes, it is expected that the male has already ejaculated, and therefore, a failure to ejaculate would result in more sperm in the reproductive system, not less. Since we are not certain when the male ejaculates, it would be important to do the analysis at different time points.

In the Prot-GFP experiments, the 10-minute time point was chosen because we nearly always observe sperm in the ejaculatory duct of control males.  In the experimental males, we never observed sperm in the ejaculatory duct at this time point.  Also, no Prot-GFP sperm were observed in the reproductive tract of females mated to experimental males even when mating was allowed to go to completion, while abundant sperm were found in females mated to Prot-GFP controls.  Figure 10S1 has been updated to include Images of these female reproductive systems.  The results showing the absence of Prot-GFP sperm in the female reproductive tract mated to experimental males indicates sperm transfer in these males isn't occurring earlier during the copulation process than in control males and that we didn't miss it by only examining at the ejaculatory duct.

(2) Discuss what may be the role of the octopamine/glutamate neurons and glutamate transmission in serotonin/glutamate neurons in the male reproductive system, given that they are not required for fertility (at least under the context in which it was tested). It is quite a striking result that deserves some attention. 

We agree it is a surprising result and have included speculation on the role of glutamate and octopamine in male reproduction in the Discussion section "Potential for adaptation to environment".

(3) Very important: 

(a) Figure 3 is present in the Word document but not the PDF. 

(b) Figure 9S3 is not present 

(c) In Figure 5 X), the legend does not correspond to the panel.

All of these corrections have been made. 

(4) Other suggestions:

(a) A summary schematic (or several) of the findings would make it an easier read.

(b) Explain why the ejaculatory bulb was left out of the analysis.

(c) Explain in the main text some of the tools, such as, BONT-C and the conditional vGlut mutation.

Reviewer #2 (Public review):

Summary:

In this manuscript, Arnould et. al. develop an unbiased, affinity-guided reagent to label P2X7 receptor and use super-resolution imaging to monitor P2X7 redistribution in response to inflammatory signaling.

Strengths:

I think the X7-uP probe that they developed is very useful for visualizing localization of P2X7 receptor. They convincingly show that under inflammatory conditions, there is a reorganization of P2X7 localization into receptor clusters. Moreover, I think they have shown a very clever way to specifically label any receptor of interest. This has broad appeal.

I think the authors have done a very nice job addressing my original concerns. Here are those original concerns and my new comments related to how the authors address them.

(1) While the authors state that chemical modification of AZ10606120 to produce the X7-UP reagent has "minimal impact" on the inhibition of P2X7, we can see from Figure 2A and 2B that it does not antagonize P2X7 as effectively as the original antagonist. For the sake of completeness and quantitation, I think it would be great if the authors could determine the IC50 for X7-uP and compare it to the IC50 of AZ10606120.

The authors now show the relative inhibition of X7-uP compared to AZ10606120 at different concentrations. This provides a nice comparison to give the reader an idea of how effectively X7-uP inhibits P2X7 receptor. This is great.

(2) Do the authors know whether modification of the lysines with biotin affects the receptor's affinity for ATP (or ability to be activated by ATP)? What about P2X7 that has been modified with biotin and then labeled with Alexa 647? For the sake of completeness and quantitation, I think it would be great if the authors could determine the EC50 of biotinylated P2X7 for ATP as well as biotinylated and then Alexa 647 labeled P2X7 for ATP and compare these values to the affinity of unmodified WT P2X7 for ATP.

I agree with the authors that assessing the functional integrity of P2X7 following biotinylation and fluorophore labeling is outside the scope of this paper but would be important for studies involving dynamic or post-labeling functional analyses such as live trafficking.

(3) It is a little misleading to color the fluorescence signal from mScarlet green (for example, in Figure 3 and Figure 4). The fluorescence is not at the same wavelength as GFP. In fact, the wavelength (570 nm - 610 nm) for emission is closer to orange/red than to green. I think this color should be changed to differentiate the signal of mScarlet from the GFP signal used for each of the other P2X receptor subtypes.

The authors have now changed the mScarlet color to orange, which solves my concern.

(4) It is my understanding that P2X6 does not form homotrimers. Thus, I was a little surprised to see that the density and distribution of P2X6-GFP in Figure 3 looks very similar to the density and distribution of the other P2X subtypes. Do the authors have an explanation for this? Are they looking at P2X6 protomers inserted into the plasma membrane? Does the cell line have endogenous P2X receptor subtypes? Is Figure 3 showing heterotrimers with P2X6 receptor? A little explanation might be helpful.

The authors address this point very well and include nice data to show that P2X6 does not insert into the plasma membrane as a homotrimer.

(5) It is easy to overlook the fact that the antagonist leaves the binding pocket once the biotin has been attached to the lysines. It might be helpful if the authors made this a little more apparent in Figure 1 or in the text describing the NASA chemistry reaction.

The authors have modified Figure 1 to make it easier to understand the NASA chemistry reaction.

I congratulate the authors on an outstanding paper!

Author response:

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

Public Reviews:

Reviewer #1 (Public review): 

Summary: 

In this paper, the authors developed a chemical labeling reagent for P2X7 receptors, called X7-uP. This labeling reagent selectively labels endogenous P2X7 receptors with biotin based on ligand-directed NASA chemistry (Ref. 41). After labeling the endogenous P2X7 receptor with biotin, the receptor can be fluorescently labeled with streptavidin-AlexaFluor647. The authors carefully examined the binding properties and labeling selectivity of X7-uP to P2X7, characterized the labeling site of P2X7 receptors, and demonstrated fluorescence imaging of P2X7 receptors. The data obtained by SDS-PAGE, Western blot, and fluorescence microscopy clearly show that X7-uP labels the P2X7 receptor. Finally, the authors fluorescently labeled the endogenous P2X7 in BV2 cells, which are a murine microglia model, and used dSTORM to reveal a nanoscale P2X7 redistribution mechanism under inflammatory conditions at high resolution. 

Strengths: 

X7-uP selectively labels endogenous P2X7 receptors with biotin. Streptavidin-AlexaFluor647 binds to the biotin labeled to the P2X7 receptor, allowing visualization of endogenous P2X7 receptors. 

We thank the reviewer for their positive comment.

Weaknesses: 

Weaknesses & Comments 

(1) The P2X7 receptor exists in a trimeric form. If it is not a monomer under the conditions of the pull-down assay in Figure 2C, the quantitative values may not be accurate. 

We thank the reviewer for this comment. As shown in Figure 2C, the band observed on the denaturing SDS-PAGE corresponds to the monomeric form of the P2X7 receptor. While we cannot exclude the presence of non-monomeric species under native conditions, no such higher-order forms are visible in the gel. This observation supports the conclusion that the quantitative values presented are based on the monomeric form and are therefore reliable.

(2) In Figure 3, GFP fluorescence was observed in the cell. Are all types of P2X receptors really expressed on the cell surface ? 

We thank the reviewer for this excellent comment, which was also raised by reviewer 2. To address this concern, we performed a commercial cell-surface protein biotinylation assay to assess whether GFP-tagged P2X receptors reach the plasma membrane. As expected, all P2X subtypes except P2X6 were detected at the cell surface in HEK293T cells, thereby validating our confocal fluorescence microscopy assay. These new data are now included in Figure 3 — figure supplement 1.

(3) The reviewer was not convinced of the advantages of the approach taken in this paper, because the endogenous receptor labeling in this study could also be done using conventional antibody-based labeling methods. 

We thank the reviewer for raising this important point and would like to highlight several advantages of our approach compared to conventional antibody-based labeling.

First, commercially available P2X7 antibodies often suffer from poor specificity and are generally not suitable for reliably detecting endogenous P2X7 receptors, as documented in previous studies (e.g., PMID: 16564580 and PMID: 15254086). While recent advances have been made using nanobodies with improved specificity for P2X7 (e.g., PMID: 30074479 and PMID: 38953020), our strategy is distinct and complementary to nanobody-based approaches.

Second, antibodies rely on non-covalent interactions with the receptor, which can result in dissociation over time. In contrast, our X7-uP probe covalently biotinylates lysine residues on the P2X7 receptor through stable amide bond formation. This covalent labeling ensures that the biotin moiety remains permanently attached, an advantage not afforded by reversible binding strategies.

Third, by selectively biotinylating P2X7 receptors, our method provides a versatile platform for the chemical attachment of a wide range of probes or functional moieties. Although we did not demonstrate this application in the current study, we believe this modularity represents an additional advantage of our approach.

We have now revised the discussion to highlight these key advantages, allowing the reader to form their own opinion. We hope this addresses the reviewer’s concerns and clarifies the benefits of our approach.

(4) Although P2X7 was successfully labeled in this paper, it is not new as a chemistry. There is a need for more attractive functional evaluation such as live trafficking analysis of endogenous P2X7. 

We agree with the reviewer that the underlying chemistry is not novel per se. However, to our knowledge, it has not previously been applied to the P2X7 receptor, and thus constitutes a novel application with specific relevance for studying native P2X7 biology.

We also appreciate the reviewer’s suggestion regarding live trafficking analysis of endogenous P2X7. While this is indeed a valuable and interesting direction, we believe it lies beyond the scope of the present study, as it would first require demonstrating that the labeling itself does not affect P2X7 function (see below). This important step would necessitate additional experiments, which we consider more appropriate for a follow-up investigation.

(5) The reviewer has concerns that the use of the large-size streptavidin to label the P2X7 receptor may perturbate the dynamics of the receptor. 

We thank the reviewer for raising this important point. Although we did not directly measure receptor dynamics, it is indeed possible that tetrameric streptavidin (tStrept-A 647) could promote P2X7 clustering by cross-linking nearby receptors due to its tetravalency (see also point 7 raised by the reviewer). To address this concern, we performed additional dSTORM experiments using a monomeric form of streptavidin-Alexa 647 (mSA) (see PMID: 26979420). Owing to its reduced size and lack of tetravalency, mSA has been shown to minimize artificial crosslinking of synaptic receptors (PMID: 26979420). A drawback of using mSA, however, is that the monomeric form carries only two fluorophores (estimated degree of labeling, DOL ≈ 2, PMID: 26979420), whereas the tetrameric form, according to the manufacturer’s certificate of analysis (Invitrogen S21374), has an average DOL of three fluorophores per monomer, resulting in a total of ~12 fluorophores per streptavidin.

We tested three conditions with mSA incubation: (i) control BV2 cells (without X7-uP), (ii) untreated X7-uP-labeled BV2 cells, and (iii) X7-uP-labeled BV2 cells treated with LPS and ATP (using the same concentrations and incubation times described in the manuscript). As shown in Author response image 1, only LPS+ATP treatment induced a clear increase in the mean cluster density compared to quiescent (untreated) BV2 cells. This effect closely matches the results obtained with tStrept-A 647, supporting the conclusion the tetrameric streptavidin does not artificially promote P2X7 clustering. It is also possible that the cellular environment of BV2 microglia differs from the confined architecture of synapses, which may further explain why cross-linking effects are less pronounced in our system.

As expected, the overall fluorescence signal with mSA was about tenfold lower than with tStrept-A 647, consistent with the expected fluorophore stoichiometry. This lower signal may explain why the values for the untreated condition appeared slightly higher than for the control, although the difference was not statistically significant (P = 0.1455).

We hope these additional experiments adequately address the reviewer’s concerns.

Author response image 1.

BV2 labeling with monomeric streptavidin–Alexa 647 (mSA).(A) Bright-field and dSTORM images of BV2 cells labeled with mSA in the presence (untreated and LPS+ATP) or absence (control) of 1 µM X7-uP. Treatment: LPS (1 µg/mL for 24 hours) and ATP (1 mM for 30 minutes). Scale bars, 10 µm. Insets: Magnified dSTORM images. Scale bars, 1 µm.(B) Quantification of the number of localizations (n = 2 independent experiments). Bars represent mean ± s.e.m. One-way ANOVA with Tukey’s multiple comparisons (P values are indicated above the graph).

(6) It is better to directly label Alexa647 to the P2X7 receptor to avoid functional perturbation of P2X7. 

Directly labeling of Alexa647 to the P2X7 receptor would require the design and synthesis of a novel probe, which is currently not available. Implementing such a strategy would involve substantial new experimental work that lies beyond the scope of the present study.

(7) In all imaging experiments, the addition of streptavidin, which acts as a cross-linking agent, may induce P2X7 receptor clustering. This concern would be dispelled if the receptors were labeled with a fluorescent dye instead of biotin and observed. 

We refer the reviewer to our response in point 5, where we addressed this concern by comparing tetrameric and monomeric streptavidin conjugates. As noted above (see also point 6), directly labeling the receptor with a fluorescent dye would require the development of a new probe, which is outside the scope of the present study.

(8) There are several mentions of microglia in this paper, even though they are not used. This can lead to misunderstanding for the reader. The author conducted functional analysis of the P2X7 receptor in BV-2 cells, which are a model cell line but not microglia themselves. The text should be reviewed again and corrected to remove the misleading parts that could lead to misunderstanding. e.g. P8. lines 361-364

First, it combines N-cyanomethyl NASA chemistry with the high-affinity AZ10606120 ligand, enabling rapid labeling in microglia (within 10 min)

P8. lines 372-373 

Our results not only confirm P2X7 expression in microglia, as previously reported (6, 26-33), but also reveal its nanoscale localization at the cell surface using dSTORM. 

We agree with the reviewer’s comment. We have now modified the text, including the title.

Reviewer #2 (Public review): 

Summary: 

In this manuscript, Arnould et. al. develop an unbiased, affinity-guided reagent to label P2X7 receptor and use super-resolution imaging to monitor P2X7 redistribution in response to inflammatory signaling. 

Strengths: 

I think the X7-uP probe that they developed is very useful for visualizing localization of P2X7 receptor. They convincingly show that under inflammatory conditions, there is a reorganization of P2X7 localization into receptor clusters. Moreover, I think they have shown a very clever way to specifically label any receptor of interest. This has broad appeal 

We thank the reviewer for their positive comment.

Weaknesses: 

Overall, the manuscript is novel and interesting. However, I do have some suggestions for improvement. 

(1) While the authors state that chemical modification of AZ10606120 to produce the X7-UP reagent has "minimal impact" on the inhibition of P2X7, we can see from Figure 2A and 2B that it does not antagonize P2X7 as effectively as the original antagonist. For the sake of completeness and quantitation, I think it would be great if the authors could determine the IC50 for X7-uP and compare it to the IC50 of AZ10606120. 

We thank the reviewer for this insightful comment. Unfortunately, due to the limited availability of X7-uP, we were not able to establish a complete concentration–response curve to determine its IC₅₀, which would require testing at concentrations >1 µM. Nevertheless, to estimate the effect of the modification, we assessed current inhibition at 300 µM X7-uP and compared it with the reported IC₅₀ of AZ10606120 (10 nM). Under these conditions, both compounds produced a similar level of inhibition, indicating that while the chemical modification reduces potency relative to AZ10606120, X7-uP still functions as an effective probe for P2X7. We have now included these data in Figure 2 and revised the text accordingly.

(2) Do the authors know whether modification of the lysines with biotin affects the receptor's affinity for ATP (or ability to be activated by ATP)? What about P2X7 that has been modified with biotin and then labeled with Alexa 647? For the sake of completeness and quantitation, I think it would be great if the authors could determine the EC50 of biotinylated P2X7 for ATP as well as biotinylated and then Alexa 647 labeled P2X7 for ATP and compare these values to the affinity of unmodified WT P2X7 for ATP.

We thank the reviewer for raising this important point. At present, we have not determined whether modification of lysine residues with biotin, or subsequent labeling with Alexa647, affects the ATP sensitivity or functional properties of P2X7. However, we believe this does not impact the conclusions of the current study, as all functional assays were conducted prior to X7-uP labeling. The labeling is used here as a terminal "snapshot" to visualize the endogenous receptor without interfering with the functional characterization.

We fully agree that assessing the functional integrity of P2X7 following biotinylation and fluorophore labeling—such as by determining the EC₅₀ for ATP—would be essential for studies involving dynamic or post-labeling functional analyses, such as live trafficking. However, as noted earlier in our response to Reviewer 1 (point 4), these experiments lie beyond the scope of the current study.

(3) It is a little misleading to color the fluorescence signal from mScarlet green (for example, in Figure 3 and Figure 4). The fluorescence is not at the same wavelength as GFP. In fact, the wavelength (570 nm - 610 nm) for emission is closer to orange/red than to green. I think this color should be changed to differentiate the signal of mScarlet from the GFP signal used for each of the other P2X receptor subtypes. 

As suggested, we changed the mScarlet color to orange for all relevant figures.

(4) It is my understanding that P2X6 does not form homotrimers. Thus, I was a little surprised to see that the density and distribution of P2X6-GFP in Figure 3 looks very similar to the density and distribution of the other P2X subtypes. Do the authors have an explanation for this? Are they looking at P2X6 protomers inserted into the plasma membrane? Does the cell line have endogenous P2X receptor subtypes? Is Figure 3 showing heterotrimers with P2X6 receptor? A little explanation might be helpful.

We thank the reviewer for raising this important point. Indeed, it is well established that P2X6 does not form functional channels, which supports the conclusion that it does not form homotrimeric complexes. Although previous studies have shown that P2X6–GFP expression is generally lower, more diffuse, and not efficiently targeted to the cell surface compared with other P2X subtypes (see PMID: 12077178), the similar fluorescence distribution and density observed in our Figure 3 do not imply that P2X6 forms homotrimers.

We did not directly assess the presence of endogenous P2X6 in our HEK293T cells; however, according to the Human Protein Atlas, there is no detectable P2X6 RNA expression in HEK293 cells (nTPM = 0), indicating that endogenous P2X6 is not expressed in this cell line. To further investigate surface expression (see also point 2 of reviewer 1), we performed a commercial cell-surface protein biotinylation assay to assess whether GFP-tagged P2X6 reaches the plasma membrane. As expected, P2X6 was not detected at the cell surface in HEK293T cells, whereas GFP-tagged P2X1 to P2X5 were readily detected. These results further support the conclusion that P2X6 does not insert into the plasma membrane as a homotrimer, thereby validating our confocal fluorescence microscopy assay. These new data are now included in Figure 3 — figure supplement 1.

(5) It is easy to overlook the fact that the antagonist leaves the binding pocket once the biotin has been attached to the lysines. It might be helpful if the authors made this a little more apparent in Figure 1 or in the text describing the NASA chemistry reaction.

We thank the reviewer for this insightful suggestion. To address this, we have modified Figure 1A and updated the legend.

Reviewer #3 (Public review): 

Summary: 

This manuscript describes the development of a covalent labeling probe (X7-uP) that selectively targets and tags native P2X7 receptors at the plasma membrane of BV2 microglial cells. Using super-resolution imaging (dSTORM), the authors demonstrate that P2X7 receptors form nanoscale clusters upon microglial activation by lipopolysaccharide (LPS) and ATP, correlating with synergistic IL-1β release. These findings advance understanding of P2X7 reorganization during inflammation and provide a generalizable labeling strategy for monitoring endogenous P2X7 in immune cells. 

Strengths: 

(1) The authors designed X7-uP by coupling a high-affinity, P2X7-specific antagonist (AZ10606120) with N-cyanomethyl NASA chemistry to achieve site-directed biotinylation. This approach offers high specificity, minimal off-target reactivity, and a straightforward pull-down/imaging readout. 

(2) The results connect P2X7's nanoscale clustering directly with IL-1β secretion in microglia, reinforcing the role of P2X7 in inflammation. By localizing endogenous P2X7 at single-molecule resolution, the authors reveal how LPS priming and ATP stimulation synergistically reorganize the receptor. 

(3) The authors systematically validate their method in recombinant systems (HEK293 cells) and in BV2 cells, showing selective inhibition, mutational confirmation of the binding site, and Western blot pulldown experiments.

We thank the reviewer for their positive comment.

Weaknesses: 

(1) While the data strongly indicate that P2X7 clustering contributes to IL-1β release, the manuscript would benefit from additional experiments (if feasible) or discussion on how receptor clustering interfaces with downstream inflammasome assembly. Clarification of whether the P2X7 clusters physically colocalize with known inflammasome proteins would solidify the mechanism. 

We thank the reviewer for this valuable suggestion. Determining the physical colocalization of P2X7 clusters with known inflammasome components would provide important insight into the molecular partners involved in inflammasome activation. However, we believe that such an investigation would constitute a substantial study on its own and therefore lies beyond the scope of the present work.

Nevertheless, in response to the reviewer’s suggestion, we have added a short paragraph at the end of the Discussion section addressing potential mechanisms by which P2X7 clustering may contribute to downstream inflammasome activation. We also revised the text to tone down the hypothesis of physical colocalization.

(2) The authors might expand on the scope of X7-uP in other native cells that endogenously express P2X7 (e.g., macrophages, dendritic cells). Although they mention the possibility, demonstrating the probe's applicability in at least one other primary immune cell type would strengthen its general utility. 

We thank the reviewer for this valuable suggestion. Again, we believe that such an investigation would constitute a substantial study on its own and therefore lies beyond the scope of the present work.

(3) The authors do include appropriate negative controls, yet providing additional details (e.g., average single-molecule on-time or blinking characteristics) in supplementary materials could help readers assess cluster calculations. 

As suggested, we have included additional data showing single-molecule blinking events in untreated and LPS+ATP-treated BV2 cells, along with the corresponding movies. The data are now presented in Figure 5—supplement figure 3A and B and Figure 5—Videos 1 and 2.

Recommendations for the authors:

Reviewer #2 (Recommendations for the authors): 

(1) On line 96, the authors refer to the "ballast" domain of P2X7 receptor but do not cite the original article from which this nomenclature originated (McCarthy et al., 2019, Cell). This article should be cited to give appropriate credit. 

Done.

(2) On line 602, the authors state that they use models from PDB 1MK5 and 6U9W to generate the cartoons in Figure 6. The manuscripts from which these PDB files were generated need to be appropriately cited. 

Done.

(3) On line 319, the authors say "300 mM BzATP" but I think they mean 300 uM.

Done. Thank you for catching the typo.

Reviewer #3 (Recommendations for the authors): 

Overall, excellent data quality. The paper would benefit from a discussion of the physiological implications of clustering. It would also be helpful to elaborate about the potential mechanisms for clustering: diffusion and/or insertion. Finally, the authors should comment on work by Mackinnon's (PMID: 39739811) and Santana lab (PMID: 31371391) on two distinct models for clustering of proteins. 

As suggested by the reviewer, we have revised the discussion to incorporate their comments. First, we have added the following text:

“Upon BV2 activation, we observed significant nanoscale reorganization of P2X7. Both LPS and ATP (or BzATP) trigger P2X7 upregulation and clustering, increasing the overall number of surface receptors and the number of receptors per cluster, from one to three (Figure 6). By labeling BV2 cells with X7-uP shortly after IL-1b release, we were able to correlate the nanoscale distribution of P2X7 with the functional state of BV2 cells, consistent with the two-signal, synergistic model for IL-1b secretion observed in microglia and other cell types (Ferrari et al, 1996; Perregaux et al, 2000; Ferrari et al, 2006; Di Virgilio et al, 2017; He et al, 2017; Swanson et al, 2019). In this model, LPS priming leads to intracellular accumulation of pro-IL-1b, while ATP stimulation activates P2X7, triggering NLRP3 inflammasome activation and the subsequent release of mature IL-1b.

What is the mechanism underlying P2X7 upregulation that leads to an overall increase in surface receptors—does it result from the lateral diffusion of previously masked receptors already present at the plasma membrane, or from the insertion of newly synthesized receptors from intracellular pools in response to LPS and ATP? Although our current data do not distinguish between these possibilities, a recent study suggests that the a1 subunit of the Na⁺/K</sup>+</sup>-ATPase (NKAa1) forms a complex with P2X7 in microglia, including BV2 cells, and that LPS+ATP induces NKAa1 internalization (Huang et al, 2024). This internalization appears to release P2X7 from NKAa1, allowing P2X7 to exist in its free form. We speculate that the internalization of NKAa1 induced by both LPS and ATP exposes previously masked P2X7 sites, including the allosteric AZ10606120 sites, thus making them accessible for X7-uP labeling.”

Second, we have added a short paragraph at the end of the Discussion section addressing potential mechanisms by which P2X7 clustering may contribute to downstream inflammasome activation:

“What mechanisms underlie P2X7 clustering in response to inflammatory signals? Several models have been proposed to explain membrane protein clustering, including recruitment to structural scaffolds (Feng & Zhang, 2009), partitioning into membrane domains enriched in specific chemical components such as lipid rafts (Simons & Ikonen, 1997), and self-assembly mechanisms (Sieber et al, 2007). These self-assembly mechanisms include an irreversible stochastic model (Sato et al, 2019) and a more recent reversible self-oligomerization model which gives rise to higher-order transient structures (HOTS) (Zhang et al, 2025). Supported by cryogenic optical localization microscopy with very high resolution (~5 nm), the HOTS model has been observed in various membrane proteins, including ion channels and receptors (Zhang et al, 2025). Furthermore, HOTS are suggested to be dynamically modulated and to play a functional role in cell signaling, potentially influencing both physiological and pathological processes (Zhang & MacKinnon, 2025). While this hypothesis is compelling, our current dSTORM data lack sufficient spatial resolution to confirm whether P2X7 trimers form HOTS via self-oligomerization. Further biophysical and ultra-high-resolution imaging studies are required to test this model in the context of P2X7 clustering.”

Reviewer #3 (Public review):

Summary:

The authors present a clearly written and beautifully presented piece of work demonstrating clear evidence to support the idea that BK channels and Cav1.3 channels can co-assemble prior to their assertion in the plasma membrane.

Strengths:

The experimental records shown back up their hypotheses and the authors are to be congratulated for the large number of control experiments shown in the ms.

Author response:

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

Reviewer #1 (Public review):

Summary:

This manuscript by Pournejati et al investigates how BK (big potassium) channels and CaV1.3 (a subtype of voltage-gated calcium channels) become functionally coupled by exploring whether their ensembles form early-during synthesis and intracellular trafficking-rather than only after insertion into the plasma membrane. To this end, the authors use the PLA technique to assess the formation of ion channel associations in the different compartments (ER, Golgi or PM), single-molecule RNA in situ hybridization (RNAscope), and super-resolution microscopy.

Strengths:

The manuscript is well written and addresses an interesting question, combining a range of imaging techniques. The findings are generally well-presented and offer important insights into the spatial organization of ion channel complexes, both in heterologous and endogenous systems.

Weaknesses:

The authors have improved their manuscript after revisions, and some previous concerns have been addressed.

Still, the main concern about this work is that the current experiments do not quantitatively or mechanistically link the ensembles observed intracellularly (in the endoplasmic reticulum (ER) or Golgi) to those found at the plasma membrane (PM). As a result, it is difficult to fully integrate the findings into a coherent model of trafficking. Specifically, the manuscript does not address what proportion of ensembles detected at the PM originated in the ER. Without data on the turnover or halflife of these ensembles at the PM, it remains unclear how many persist through trafficking versus forming de novo at the membrane. The authors report the percentage of PLApositive ensembles localized to various compartments, but this only reflects the distribution of pre-formed ensembles. What remains unknown is the proportion of total BK and Ca_V1.3 channels (not just those in ensembles) that are engaged in these complexes within each compartment. Without this, it is difficult to determine whether ensembles form in the ER and are then trafficked to the PM, or if independent ensemble formation also occurs at the membrane. To support the model of intracellular assembly followed by coordinated trafficking, it would be important to quantify the fraction of the total channel population that exists as ensembles in each compartment. A comparable ensemble-to-total ratio across ER and PM would strengthen the argument for directed trafficking of pre-assembled channel complexes.

We appreciate the reviewer’s thoughtful comment and agree that quantitatively linking intracellular hetero-clusters to those at the plasma membrane is an important and unresolved question. Our current study does not determine what proportion of ensembles at the plasma membrane originated during trafficking. It also does not quantify the fraction of total BK and Ca_V1.3 channels engaged in these complexes within each compartment. Addressing this requires simultaneous measurement of multiple parameters—total BK channels, total Ca_V1.3 channels, hetero-cluster formation (via PLA), and compartment identity—in the same cell. This is technically challenging. The antibodies used for channel detection are also required for the proximity ligation assay, which makes these measurements incompatible within a single experiment.

To overcome these limitations, we are developing new genetically encoded tools to enable real-time tracking of BK and Ca_V1.3 dynamics in live cells. These approaches will enable us to monitor channel trafficking and the formation of hetero-clusters, as detected by colocalization. This kind of experiments will provide insight into their origin and turnover. While these experiments are beyond the scope of the current study, the findings in our current manuscript provide the first direct evidence that BK and CaV channels can form hetero-clusters intracellularly prior to reaching the plasma membrane. This mechanistic insight reveals a previously unrecognized step in channel organization and lays the foundation for future work aimed at quantifying ensemble-to-total ratios and determining whether coordinated trafficking of pre-assembled complexes occurs.

This limitation is acknowledged in the discussion section, page 23. It reads: “Our findings highlight the intracellular assembly of BK-Ca_V1.3 hetero-clusters, though limitations in resolution and organelle-specific analysis prevent precise quantification of the proportion of intracellular complexes that ultimately persist on the cell surface.”

Reviewer #2 (Public review):

Summary:

The co-localization of large conductance calcium- and voltage activated potassium (BK) channels with voltage-gated calcium channels (CaV) at the plasma membrane is important for the functional role of these channels in controlling cell excitability and physiology in a variety of systems.

An important question in the field is where and how do BK and CaV channels assemble as 'ensembles' to allow this coordinated regulation - is this through preassembly early in the biosynthetic pathway, during trafficking to the cell surface or once channels are integrated into the plasma membrane. These questions also have broader implications for assembly of other ion channel complexes

Using an imaging based approach, this paper addresses the spatial distribution of BKCaV ensembles using both overexpression strategies in tsa201 and INS-1 cells and analysis of endogenous channels in INS-1 cells using proximity ligation and superesolution approaches. In addition, the authors analyse the spatial distribution of mRNAs encoding BK and Cav1.3.

The key conclusion of the paper that BK and Ca_V1.3 are co-localised as ensembles intracellularly in the ER and Golgi is well supported by the evidence.However, whether they are preferentially co-translated at the ER, requires further work. Moreover, whether intracellular pre-assembly of BK-Ca_V1.3 complexes is the major mechanism for functional complexes at the plasma membrane in these models requires more definitive evidence including both refinement of analysis of current data as well as potentially additional experiments.

The reviewer raises the question of whether BK and Ca_V1.3 channels are preferentially co-translated. In fact, I would like to propose that co-translation has not yet been clearly defined for this type of interaction between ion channels. In our current work, we 1) observed the colocalization between BK and Ca_V1.3 mRNAs and 2) determined that 70% of BK mRNA in active translation also colocalizes with Ca_V1.3 mRNA. We think these results favor the idea of translational complexes that can underlie the process of co-translation. However, and in total agreement with the Reviewer, the conclusion that the mRNA for the two ion channels is cotranslated would require further experimentation. For instance, mRNA coregulation is one aspect that could help to define co-translation. 

To avoid overinterpretation, we have revised the manuscript to remove references to “co-translation” in the Results section and included the word “potential” when referring to co-translation in the Discussion section. We also clarified the limitations of our evidence in the Discussion that can be found on page 25: “It is important to note that while our data suggest mRNA coordination, additional experiments are required to directly assess co-translation.”

Strengths & Weaknesses

(1) Using proximity ligation assays of overexpressed BK and CaV1.3 in tsa201 and INS1 cells the authors provide strong evidence that BK and CaV can exist as ensembles (ie channels within 40 nm) at both the plasma membrane and intracellular membranes, including ER and Golgi. They also provide evidence for endogenous ensemble assembly at the Golgi in INS-1 cells and it would have been useful to determine if endogenous complexes are also observe in the ER of INS-1 cells. There are some useful controls but the specificity of ensemble formation would be better determined using other transmembrane proteins rather than peripheral proteins (eg Golgi 58K).

We thank the reviewer for their thoughtful feedback and for recognizing the strength of our proximity ligation assay data supporting BK–Ca_V1.3 hetero-clusters formation at both the plasma membrane and intracellular compartments. As for specificity controls, we appreciate the suggestion to use transmembrane markers. To strengthen our conclusion, we have performed an additional experiment comparing the number of PLA puncta formed by the interaction of Ca_V1.3 and BK channels with the number of PLA puncta formed by the interaction of Ca_V1.3 channels and ryanodine receptors in INS-1 cells. As shown in the figure below, the number of interactions between Ca_V1.3 and BK channels is significantly higher than that between Ca_V1.3 and RyR₂. Of note, RyR₂ is a protein resident of the ER. These results provide additional evidence of the existence of endogenous complex formation in INS-1 cells. We have added this figure as a supplement.

(2) Ensemble assembly was also analysed using super-resolution (dSTORM) imaging in INS-1 cells. In these cells only 7.5% of BK and CaV particles (endogenous?) co-localise that was only marginally above chance based on scrambled images. More detailed quantification and validation of potential 'ensembles' needs to be made for example by exploring nearest neighbour characteristics (but see point 4 below) to define proportion of ensembles versus clusters of BK or Cav1.3 channels alone etc. For example, it is mentioned that a distribution of distances between BK and Cav is seen but data are not shown.

We thank the reviewer for this comment. To address the request for more detailed quantification and validation of ensembles, we performed additional analyses:

Proportion of ensembles vs isolated clusters: We quantified clusters within 200 nm and found that 37 ± 3% of BK clusters are near one or more CaV1.3 clusters, whereas 15 ± 2% of CaV1.3 clusters are near BK clusters. Figure 8– Supplementary 1A

Distance distribution: As shown in Figure 8–Supplementary 1B, the nearestneighbor distance distribution for BK-to-CaV1.3 in INS-1 cells (magenta) is shifted toward shorter distances compared to randomized controls (gray), supporting preferential localization of BK–CaV1.3 hetero-clusters.

Together, these analyses confirm that BK–CaV1.3 ensembles occur more frequently than expected by chance and exhibit an asymmetric organization favoring BK proximity to CaV1.3 in INS-1 cells. We have included these data and figures in the revised manuscript, as well as description in the Results section. 

(3) The evidence that the intracellular ensemble formation is in large part driven by cotranslation, based on co-localisation of mRNAs using RNAscope, requires additional critical controls and analysis. The authors now include data of co-localised BK protein that is suggestive but does not show co-translation. Secondly, while they have improved the description of some controls mRNA co-localisation needs to be measured in both directions (eg BK - SCN9A as well as SCN9A to BK) especially if the mRNAs are expressed at very different levels. The relative expression levels need to be clearly defined in the paper. Authors also use a randomized image of BK mRNA to show specificity of co-localisation with Cav1.3 mRNA, however the mRNA distribution would not be expected to be random across the cell but constrained by ER morphology if cotranslated so using ER labelling as a mask would be useful?

We thank the reviewer for these constructive suggestions. We measured mRNA colocalization in both directions as recommended. As shown in the figure below, colocalization between KCNMA1 and SCN9A transcripts was comparable in both directions, with no statistically significant difference, supporting the specificity of the observed associations. We decided not to add this to the original figure to keep the figure simple. 

We agree that co-localization of BK protein with BK mRNA is not conclusive evidence of co-translation, and we do not intend to mislead readers in our conclusion. Consequently, we were careful in avoiding the use of co-translation in the result section and added the word “potential” when referring to co-translation in the Discussion section. We added a sentence in the discussion to caution our interpretation: “It is important to note that while our data suggest mRNA coordination, additional experiments are required to directly assess cotranslation.”

Author response image 1.

(4) The authors attempt to define if plasma membrane assemblies of BK and CaV occur soon after synthesis. However, because the expression of BK and CaV occur at different times after transient transfection of plasmids more definitive experiments are required. For example, using inducible constructs to allow precise and synchronised timing of transcription. This would also provide critical evidence that co-assembly occurs very early in synthesis pathways - ie detecting complexes at ER before any complexes 

We appreciate the reviewer’s insightful suggestion regarding the use of inducible constructs to synchronize transcription timing. This is an excellent approach and would allow direct testing of whether co-assembly occurs early in the synthesis pathway, including detection of complexes at the ER prior to plasma membrane localization. These experiments are beyond the scope of the present work but represent an important direction for future studies.

We have added the following sentence to the Discussion section (page 24) to highlight this idea. “Future experiments using inducible constructs to precisely control transcription timing will enable more precise quantification of heterocluster formation in the ER compartment prior to plasma membrane insertion and reduce the variability introduced by differences in expression timing after plasmid transfection.” 

(5) While the authors have improved the definition of hetero-clusters etc it is still not clear in superesolution analysis, how they separate a BK tetramer from a cluster of BK tetramers with the monoclonal antibody employed ie each BK channel will have 4 binding sites (4 subunits in tetramer) whereas Cav1.3 has one binding site per channel. Thus, how do authors discriminate between a single BK tetramer (molecular cluster) with potential 4 antibodies bound compared to a cluster of 4 independent BK channels.

We appreciate the reviewer’s thoughtful comment regarding the interpretation of super-resolution data. We agree that distinguishing a single BK tetramer from a cluster of multiple BK channels is challenging when using an antibody that can bind up to four sites per channel. To clarify, our analysis does not attempt to resolve individual subunits within a tetramer; rather, it focuses on the nanoscale spatial proximity of BK and Ca_V1.3 signals.

We want to note that this limitation applies only to the super-resolution maps in Figures 8C and 9D and does not affect Airyscan-based analyses or measurements of BK–Ca_V1.3 proximity.

To address how we might distinguish between a single BK tetramer and a cluster of multiple BK channels, we considered two contrasting scenarios. In the first case, we assume that all four α-subunits within a tetramer are labeled. Based on cryoEM structures, a BK tetramer measures approximately 13 nm × 13 nm (≈169 nm²). Adding two antibody layers (primary and secondary) would increase the footprint by ~14 nm in each direction, resulting in an estimated area of ~41 nm × 41 nm (≈1681 nm²). Under this assumption, particles smaller than ~1681 nm² would likely represent individual tetramers, whereas larger particles would correspond to clusters of multiple tetramers. 

In the second scenario, we propose that steric constraints at the S9–S10 segment, where the antibody binds, limit labeling to a single antibody per tetramer. If true, the localization precision would approximate 14 nm × 14 nm—the combined size of the antibody complex and the channel—close to the resolution limit of the microscope. To test this, we performed a control experiment using two antibodies targeting the BK C-terminal domain, raised in different species and labeled with distinct fluorophores. Super-resolution imaging revealed that only ~12% of particles were colocalized, suggesting that most channels bind a single antibody.

If multiple antibodies could bind each tetramer, we would expect much greater colocalization.

Although these data are not included in the manuscript, we have added the following clarification to the Results section (page 19): “It is important to note that this technique does not allow us to distinguish between labeling of four BK αsubunits within a tetramer and labeling of multiple BK channel clusters. Hence, particles smaller than ~1680 nm² may represent either a single tetramer or a cluster. This limitation applies to Figures 8C and 9D and does not affect measurements of BK–Ca_V1.3 proximity.”

Author response image 2.

(6) The post-hoc tests used for one way ANOVA and ANOVA statistics need to be defined throughout

We thank the reviewer for highlighting the need for clarity regarding our statistical analyses. We have now specified the post-hoc tests used for all one-way ANOVA and ANOVA comparisons throughout the manuscript, and updated figure legends.

Reviewer #3 (Public review):

Summary:

The authors present a clearly written and beautifully presented piece of work demonstrating clear evidence to support the idea that BK channels and Cav1.3 channels can co-assemble prior to their assertion in the plasma membrane.

Strengths:

The experimental records shown back up their hypotheses and the authors are to be congratulated for the large number of control experiments shown in the ms.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

The authors have sufficiently addressed the specific points previously raised and the manuscript has improved clarity in those aspects. My main concern, which still remains, is stated in the public review.

Reviewer #3 (Recommendations for the authors):

I am content that the authors have attempted to fully address my previous criticisms.

I have only three suggestions

(1) I think the word Homo-clusters at the bottom right of Figure 1 is erroneously included.

We thank the reviewer for bringing this to our attention. The figure has been corrected accordingly.

(2) The authors should, for completeness, to refer to the beta, gamma and LINGO subunit families in the Introduction and include appropriate references:

Knaus, H. G., Folander, K., Garcia-Calvo, M., Garcia, M. L., Kaczorowski, G. J., Smith, M., & Swanson, R. (1994). Primary sequence and immunological characterization of betasubunit of high conductance Ca2+-activated K+ channel from smooth muscle. The Journal of Biological Chemistry, 269(25), 17274-17278.

Brenner, R., Jegla, T. J., Wickenden, A., Liu, Y., & Aldrich, R. W. (2000a). Cloning and functional characterization of novel large conductance calcium-activated potassium channel beta subunits, hKCNMB3 and hKCNMB4. The Journal of Biological Chemistry, 275(9), 6453-6461.

Yan, J & R.W. Aldrich. (2010) LRRC26 auxiliary protein allows BK channel activation at resting voltage without calcium. Nature. 466(7305):513-516

Yan, J & R.W. Aldrich. (2012) BK potassium channel modulation by leucine-rich repeatcontaining proteins. Proceedings of the National Academy of Sciences 109(20):7917-22

Dudem, S, Large RJ, Kulkarni S, McClafferty H, Tikhonova IG, Sergeant, GP, Thornbury, KD, Shipston, MJ, Perrino BA & Hollywood MA (2020). LINGO1 is a novel regulatory subunit of large conductance, Ca2+-activated potassium channels. Proceedings of the National Academy of Sciences 117 (4) 2194-2200

Dudem, S., Boon, P. X., Mullins, N., McClafferty, H., Shipston, M. J., Wilkinson, R. D. A., Lobb, I., Sergeant, G. P., Thornbury, K. D., Tikhonova, I. G., & Hollywood, M. A. (2023). Oxidation modulates LINGO2-induced inactivation of large conductance, Ca2+-activated potassium channels. The Journal of Biological Chemistry, 299 (3) 102975.

We agree with the reviewer’s suggestion and have revised the Introduction to include references to the beta, gamma, and LINGO subunit families. Appropriate citations have been added to ensure completeness and contextual relevance.

Additionally, BK channels are modulated by auxiliary subunits, which fine-tune BK channel gating properties to adapt to different physiological conditions. The β, γ, and LINGO1 subunits each contribute distinct structural and regulatory features: β-subunits modulate Ca²⁺ sensitivity and can induce inactivation; γ-subunits shift voltage-dependent activation to more negative potentials; and LINGO1 reduces surface expression and promotes rapid inactivation (18-24). These interactions ensure precise control over channel activity, allowing BK channels to integrate voltage and calcium signals dynamically in various cell types.

(3) I think it may be more appropriate to include the sentence "The probes against the mRNAs of interest and tested in this work were designed by Advanced Cell Diagnostics." (P16, right hand column, L12-14) in the appropriate section of the Methods, rather than in Results.

We thank the reviewer for this helpful suggestion. In response, we have relocated the sentence to the appropriate section of the Methods, where it now appears with relevant context.

estructura sugerida: 1. Welfare states, commodification & policy feedbacks 2. Distributive justice & market justice 3. Key studies on market justice & main findings 4. Pensions 5. Operationalization

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

Learn more at Review Commons

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity):

A previous study by Komada et al. demonstrated that MAP7 is expressed in both Sertoli and germ cells, and that Map7 gene-trap mutant mice display disrupted microtubule bundle formation in Sertoli cells, accompanied by defects in spermatid manchettes and germ cell loss. In the current study, Kikuchi et al. investigated the role of MAP7 in the formation of the Sertoli cell apical domain during the first wave of spermatogenesis. They generated a GFP-tagged MAP7 mouse line and demonstrated that the endogenous MAP7 protein localizes to the apical microtubules in Sertoli cells and to the manchette microtubules in step 9-11 spermatids. They also generated a new Map7 knockout (KO) mouse line in a genetic background distinct from the one used in the previous study. Focusing on stages before the emergence of step 9-11 spermatids, the authors aimed to isolate defects caused by the function of MAP7 in Sertoli cells. They report that loss of MAP7 impairs Sertoli cell polarity and apical domain formation, accompanied by the microtubule remodeling defect. Using the GFP-tagged MAP7 line, they performed immunoprecipitation-mass spectrometry and identified several MAP7-interacting proteins in the testis, including MYH9. They further observed that MAP7 deletion alters the distribution of MYH9. Single-cell RNA sequencing revealed that the loss of MAP7 in Sertoli cells resulted in slight transcriptomic shifts but had no significant impact on their functional differentiation. Single-cell RNA sequencing analysis also showed delayed meiotic progression in the MAP7-deficient testis. Overall, while the study provides some interesting discoveries of early Sertoli cell defects in MAP7-deficient testes, some conclusions are premature and not fully supported by the presented data. The mechanistic investigations remain limited in depth.

Response: We thank the reviewer for this insightful summary. We agree that some of our initial interpretations were speculative and have revised the relevant sections to more accurately reflect the limitations of the current data. We also acknowledge that further mechanistic studies will be important to strengthen our conclusions, and we have outlined these plans in the individual responses below.

Major comments:

• Although the infertility phenotype of the Map7 gene-trap mutant mice has been reported previously, it remains essential to assess fertility in this newly generated MAP7 knockout line. While the authors present testis size and histological differences between WT and KO mice (Extended Fig. 2e and 2f), there is no corresponding description or interpretation in the main text regarding fertility outcomes.

Response: We thank the reviewer for raising this point. Although we had presented the differences in testis size and histology between wild-type and Map7-/- mice, we agree that a description of the corresponding fertility outcomes was missing from the main text. We have now revised the relevant part of the Results section as follows: “Consistent with observations in Map7 gene-trap mice, Map7-/- males exhibited reduced testis size and spermatogenic defects (Supplemental Fig. 2E, F). Notably, the cauda epididymis of Map7-/- males contained no mature spermatozoa (Supplemental Fig. 2F), indicating male infertility.” (page 5, line 33–page 6, line 2)

• In Figure 2C, the authors identified Sertoli cells, spermatogonia cells, and spermatocytes using SEM, based on their cell morphology and adhesion to the basement membrane. Given that the loss of MAP7 disrupts the polarity and architecture of Sertoli cells, the position of germ cells will be affected, making this identification criterion less reliable.

Response: We appreciate the reviewer’s comment. While the reviewer notes that cell identification was based on cell morphology and adhesion to the basement membrane, we clarify that nuclear morphology was also considered, as described in the original manuscript. Specifically, germ cells have spherical nuclei, whereas Sertoli cell nuclei are irregularly shaped (representative segmentation results can be provided as an additional Supplemental Figure upon request). Round spermatids at P21 can be distinguished from spermatocytes by their smaller nuclear size. In addition, spermatogonia remain attached to the basement membrane even in Map7-/- testes, as confirmed by GFRα1-positive spermatogonial stem cells (Figure 6A). Together, these features ensure reliable identification of each cell type, independent of the altered polarity observed in Map7-deficient Sertoli cells.

• In Figure 2e, the number of Sox9-positive Sertoli cells in MAP7 knockout mice appears higher than that in the control at P17. Quantification of total Sox9-positive cells should be done to determine whether MAP7 deletion increases Sertoli cell numbers.

Response: As suggested by the reviewer, we will quantify the density of SOX9-positive Sertoli cells per unit area of seminiferous tubule at P10 and P17 in Map7+/- and Map7-/- testes, and include the results in the revised manuscript.

• To determine whether MAP7's role in regulating Sertoli cell polarity relies on germ cells, the authors treated mice with busulfan at P28 to delete germ cells, a stage after Sertoli cell polarity defect has developed in MAP7 knockout mice. This data is insufficient to support the conclusion that MAP7 regulates Sertoli cell polarity independently of the presence of germ cells. Germ cell deletion should be done before the Sertoli cell defect develops to address this question.

Response: We appreciate the reviewer’s thoughtful comment regarding the interpretation of the busulfan experiments. While depletion of germ cells at P28 enabled us to assess Sertoli cell polarity in the absence of postnatal spermatogonia, these experiments do not definitively determine whether MAP7 regulates Sertoli cell polarity independently of germ cells. Neonatal germ-cell depletion would more directly test germ cell–independent effects; however, systemic busulfan administration at early developmental stages is highly toxic, often causing bone marrow failure and multi-organ damage, which precludes survival and confounds analysis of testis-specific effects. Although germ cell ablation could, in principle, be achieved using transgenic approaches that exploit the natural resistance of mice to diphtheria toxin (DTX) (reviewed in Smith et al., Andrology, 2015), these strategies require multiple transgenes and show minor variability in efficiency, making them impractical for our current experiments. Generating the necessary genetic combinations would require considerable time. We therefore plan to pursue alternative genetic approaches in future work.

In the revised manuscript, we have modified the relevant section to more accurately reflect the limitations of the current experiments, as follows: “Busulfan was administered at P28, and testes were analyzed 6 weeks later, after complete elimination of germ cell lineages. Following treatment, Map7+/- mice showed testis-to-body weight ratios comparable to untreated Map7-/- mice (Supplemental Fig. 3D), and hematoxylin-eosin (HE) staining confirmed germ cell depletion (Fig. 2F; Supplemental Fig. 3E). In Map7+/- testes, most Sertoli nuclei remained basally positioned, indicating that once apical–basal polarity is established, it is stably maintained even in the absence of germ cells. In contrast, Map7-/- Sertoli nuclei were frequently misoriented toward the lumen under the same conditions (Fig. 2F; Supplemental Fig. 3E), suggesting that polarity defects in Map7-deficient Sertoli cells occur independently of germ cell presence.” (page 7, lines 20–28)

In addition, we have added the following sentences to the Discussion section to highlight the implication of these findings: “In addition, even after germ cell depletion by busulfan treatment, Map7-deficient Sertoli cells failed to reestablish basal nuclear positioning, indicating that loss of MAP7 causes an intrinsic polarity defect. These findings suggest that MAP7 acts as a cell-autonomous regulator of Sertoli cell polarity, rather than mediating effects indirectly through germ cell–Sertoli cell interactions.” (page 15, lines17–21)

• The resolution of the SEM images in Figure 3c is insufficient to evaluate tight and adherens junctions clearly. As such, these images do not convincingly support the claim that adherens junctions are absent in the KO testes.

Response: We thank the reviewer for this insightful comment. Tight junctions can be reliably identified in SEM images as dense intercellular structures accompanied by endoplasmic reticulum aligned along the cell boundaries. The region immediately apical to the tight junctions likely corresponds to adherens junctions, which are also associated with the endoplasmic reticulum. Unlike tight junctions, these regions exhibit wider intercellular spaces, consistent with the looser membrane apposition characteristic of adherens junctions, although they cannot be unambiguously distinguished from gap junctions or desmosomes based on morphology alone. In the original figure, 2× binning reduced image resolution, which may have contributed to the reviewer’s concern.

In the revised manuscript, we have re-acquired the SEM images in high-resolution mode, focusing on the relevant regions. The new high-resolution images have replaced the original panels in revised Figure 3C, providing clearer visualization of junctional structures at P10 and P21 in Map7+/- and Map7-/- testes. The original Figure 3C images have been moved to Supplemental Figure 4B for reference.

The corresponding section in the Results has been revised as follows in the updated manuscript: “We then performed SEM to examine the effects of Map7 KO. In P21 Map7+/- testes, electron-dense regions along the basal side of Sertoli–Sertoli junctions corresponded to tight junctions closely associated with the endoplasmic reticulum, consistent with previous reports (Luaces et al. 2023) (Fig. 3C; Supplemental Fig. 4B). The region immediately apical to the tight junctions likely represents adherens junctions, which were also associated with the endoplasmic reticulum. Unlike tight junctions, these regions displayed wider intercellular spaces, reflecting the looser membrane apposition typical of adherens junctions, though they could not be definitively distinguished from gap junctions or desmosomes based on morphology alone (Fig. 3C; Supplemental Fig. 4B). At P10, both Map7+/- and Map7-/- testes lacked clearly defined tight junctions and adherens junction–like structures (Fig. 3C; Supplemental Fig. 4B). In P21 Map7-/- mice, Sertoli cells formed expanded basal tight junctions but failed to establish adherens junction–like structures (Fig. 3C; Supplemental Fig. 4B).” (page 8, line 34–page 9, line 12)

• GFP-tagged reporter mice and HeLa cells were used for immunoprecipitation-mass spectrometry to identify proteins that interact with MAP7. Given that the authors aimed to elucidate the mechanism by which MAP7 regulates Sertoli cell cytoskeleton organization, the rationale for including HeLa cells is unclear and should be better justified or reconsidered.

Response: We thank the reviewer for this comment. MAP7-egfpKI HeLa cells were used as a complementary system to identify MAP7-associated proteins, providing sufficient material and a controlled environment for robust detection. By comparing IP-MS results from MAP7-egfpKI HeLa cells and P17–P20 Map7-egfpKI testes, we can distinguish proteins that are specific to polarized Sertoli cells: proteins detected exclusively in P17–P20 testes may be involved in Sertoli cell polarization, whereas proteins detected in both systems likely represent general MAP7-associated factors that are not specific to Sertoli cell polarity.

This rationale has been clarified in the revised manuscript by adding the following sentence to the Results section: “MAP7-egfpKI HeLa cells were used as a complementary system, providing sufficient material and a controlled environment for robust detection of MAP7-associated proteins. Comparison of IP-MS results between MAP7-egfpKI HeLa cells and P17–P20 Map7-egfpKI testes allows identification of MAP7-associated proteins that are specific to polarized Sertoli cells, whereas proteins detected in both systems likely represent general MAP7-associated proteins.” (page 9 lines 27-32)

• The authors observed that MYH9, one of the MAP7-interacting proteins, does not colocalize with ectopic microtubule and F-actin structures in MAP7 KO testes and concluded that MAP7 facilitates the integration of microtubules and F-actin via interaction with NMII heavy chains. This conclusion is speculative and not adequately supported by the presented data.

Response: We thank the reviewer for this insightful comment. We agree that our initial conclusion was speculative and have revised the relevant section to more accurately reflect the limitations of the current data. The revised text now reads as follows: “These findings indicate that MYH9 localization at the luminal interface depends on MAP7, and suggest that MAP7 helps coordinate microtubules and F-actin, potentially via its association with NMII heavy chains.” (page 10, lines 13–15)

To further elucidate this mechanism, we will perform biochemical domain-mapping to define the MAP7 region responsible for MYH9 complex formation. We have already established a series of human MAP7 deletion mutants (as reported previously, EMBO Rep., 2018) and will conduct co-immunoprecipitation assays in HEK293 cells to identify the specific MAP7 domain required for complex formation with MYH9. Based on these results, we plan to use AlphaFold3 to predict the three-dimensional structure of the MAP7–MYH9 complex. These analyses will help clarify how MAP7 associates with the actomyosin network and provide additional mechanistic insights that complement our in vivo observations of MYH9 mislocalization in Map7-/- testes.

• The authors used Spearman correlation coefficients to analyze six Sertoli cell clusters and generated a minimum spanning tree to infer differentiation trajectories. However, details on the method used for constructing the tree are lacking. Moreover, relying solely on Spearman correlation to define differentiation topology is oversimplified.

Response: We appreciate the reviewer’s valuable feedback. We agree that Spearman correlation alone is insufficient to infer differentiation topology. In response, we reanalyzed the data using Monocle3, which implements branch-aware pseudotime inference to capture both cluster continuity and differentiation directionality. This reanalysis provides a more accurate reconstruction of differentiation trajectories among the six Sertoli cell clusters. Although the overall trajectories appeared different and a higher proportion of Map7-/- Sertoli cells exhibited very low pseudotime values, comparison of the control and Map7-/- trajectories revealed that the average node degree was nearly identical, indicating that the local graph structure—reflecting the connectivity among neighboring cells—was largely preserved. The numbers of branch points and the graph diameter differed slightly, likely due to differences in sample size (311 control vs. 434 Map7-/- Sertoli cells) and distribution bias rather than major topological changes. Accordingly, Figures 5C and 5D have been replaced with the updated Monocle3-based trajectory analysis, and the corresponding text in the Results section and figure legend have been revised as follows:

“To reconstruct differentiation trajectories among the six Sertoli cell clusters, we reanalyzed the datasets using Monocle3, which incorporates branch-aware pseudotime inference. Cluster C1 was selected as the root based on shared specificity and entropy scores, consistent with its metabolically active and transcriptionally diverse profile (Fig. 5B, C; Supplemental Fig. 7). While the overall trajectories appeared altered, the proportion of Map7-/- Sertoli cells with very low pseudotime values was only modestly increased (Fig. 5D). Comparison with controls showed that the average node degree was nearly identical (Fig. 5C), indicating that the local graph structure, reflecting connectivity among neighboring cells, remained largely intact. Minor differences in branch points and graph diameter likely reflect inherent variability in the data rather than major topological changes (Supplemental Fig. 6B). Consistent with this, the relative proportions of the six clusters showed only modest shifts, suggesting that the overall architecture of Sertoli cell differentiation is largely preserved in the absence of MAP7.” (page 11, lines 7-18)

“(C) Control and Map7-/- Sertoli cells were visualized separately using UMAPs constructed in Seurat. Using the same datasets, pseudotime trajectories were inferred with Monocle3. For root selection, shared_score (cluster overlap), specificity_score (cluster uniqueness), and entropy_score (transcriptional diversity) were computed, resulting in cluster 1 being selected as the root. The numbers of nodes, edges, branch points, average degree, and diameter of each trajectory are shown below the corresponding UMAPs. (D) Parallel comparison of pseudotime distributions between control and Map7-/- populations.” (page 30, lines 5-12)

Minor comments:

• Several extended data figures are redundant with main figures and do not provide additional value (e.g., Fig. 2d vs. Extended Data Fig. 3a; Fig. 2f vs. Extended Data Fig. 3d; Fig. 2C vs. Extended Data Fig. 4b; Fig. 3d vs. Extended Data Fig. 4c). The authors should consolidate or remove duplicates.

Response: Regarding the concerns about redundancy between main and Supplemental figures, we would like to clarify the rationale for retaining certain Supplemental figures.

Fig. 2D vs. Supplemental Fig. 3A: Due to space limitations in the main figure, only the merged three-color image was shown. We believe that the single-color grayscale images in Supplemental Fig. 3A provide additional clarity, allowing easier visualization of SOX9-positive Sertoli cell distribution and differences in F-actin structure.

Fig. 2F vs. Supplemental Fig. 3E: In the main figure, only the high-magnification image was shown due to space constraints. The lower-magnification image in Supplemental Fig. 3E demonstrates that the selected field was not chosen arbitrarily, providing context for the observed structures. In addition, Supplemental Fig. 3E includes both low- and high-magnification images of age-matched busulfan (-) testes as a control for the busulfan (+) condition, further supporting the validity of the comparison.

For the above-mentioned cases (Fig. 2D vs. Supplemental. 3A; Fig. 2F vs. Supplemental Fig. 3E), as well as other potentially overlapping figures (e.g., Fig. 3D vs. Supplemental Fig. 4C), we believe that the additional single-channel and lower-magnification images provide important context that cannot be fully conveyed in the main figures due to space limitations. Nevertheless, to address the reviewer’s concern, we will (i) clearly state the purpose of each Supplemental figure in the corresponding legends, and (ii) re-evaluate all figures to consolidate or remove any truly redundant panels. Our goal is to ensure that all figures collectively convey the data in the most concise and informative manner.

• Figure citations in the main text do not consistently match figure content. For example, on page 7 (lines 5-6), the text refers to Extended Data Fig. 4a for SOX9 staining. Yet, it is the extended Data Fig. 3a that contains the relevant data. Similarly, the reference to Extended Data Fig. 4b and 4c on page 7 (lines 7-8) for adult defects is inaccurate.

Response: We thank the reviewer for drawing attention to these inconsistencies. We have carefully checked all figure citations throughout the main text and corrected them so that they consistently match the figure content. The revised manuscript reflects these corrections.

• In Figure 2e, percentages of Sertoli cells across three layers are shown. The figure legend should specify which layer(s) show statistically significant differences between WT and KO.

Response: We are grateful to the reviewer for highlighting this point. Statistical comparisons were performed between Map7+/- and Map7-/- mice within each corresponding layer at P17. Statistical significance was assessed using Student’s t-test, and all three layers showed significant differences between Map7+/- and Map7-/- (P < 2.20 × 10⁻⁴). The figure legend has been revised accordingly as follows: “Statistical comparisons between Map7+/- and Map7-/- mice were performed for each corresponding layer at P17 using Student’s t-test. All three layers showed significant differences between Map7+/- and Map7-/- mice (*, P<2.20 × 10⁻⁴).” (page 28, lines 5-8)

• The current color scheme for F-actin and TUBB3 in Figure 3 lacks sufficient contrast. Adjusting to more distinguishable colors would improve readability.

Response: Response: We thank the reviewer for this helpful suggestion. In the original merged images, four channels (DNA, TUBB3, F-actin, and β-catenin) were displayed together, which reduced contrast between cytoskeletal signals. To improve clarity, we generated new merged images showing only TUBB3 and F-actin, allowing better visual distinction between these components. In addition, β-catenin and DNA are now displayed together as a separate merged image (β-catenin in yellow and DNA in blue) in the final column, highlighting the altered localization of β-catenin in Map7-/- testes.

• Since multiple scale bars with different units are present within the same figures, adding units directly above or beside each scale bar would improve readability.

Response: We thank the reviewer for the suggestion. Following this recommendation, we have added units directly above each scale bar in all figures to improve readability.

• It is recommended to directly mark Sertoli cells, spermatogonia, and spermatocytes on the SEM images in Figure 2C for clearer visualization.

Response: We thank the reviewer for the suggestion. We will follow this recommendation by performing segmentation and directly marking Sertoli cells, spermatogonia, and spermatocytes on the SEM images in Figure 2C to improve visualization.

• The quantification of Sertoli cell positioning shown in Fig. 2C is already described in the main text and is unnecessary in the figure.

Response: We appreciate the reviewer’s comment regarding the quantification of Sertoli cell positioning. Although the results are described in the main text, we believe that the visual presentation in Figure 2C is essential for conveying the spatial distribution pattern in an intuitive and comparative manner. To address the concern about redundancy, we have slightly revised the figure legend (page 27, lines 28–29) to clarify that this panel provides a visual summary of the quantitative data described in the text, thereby improving clarity without unnecessary duplication.

_Referee cross-commenting_

I concur with Reviewer 2 that the Map7-eGFP mouse model is a valuable tool for the research community. I also agree that performing MAP7-MYH9 double immunofluorescence staining to demonstrate their colocalization would further strengthen the authors' conclusions regarding their interaction. My overall assessment of the manuscript remains unchanged: the study represents an incremental advance that extends previous findings on MAP7 function but provides limited new mechanistic insight.

Reviewer #1 (Significance):

This study investigates the role of the microtubule-associated protein MAP7 in Sertoli cell polarity and apical domain formation during early stages of spermatogenesis. Using GFP-tagged and MAP7 knockout mouse models, the authors show that MAP7 localizes to apical microtubules and is required for Sertoli cell cytoskeletal organization and germ cell development. While the study identifies early Sertoli cell defects and candidate MAP7-interacting proteins, the mechanistic insights remain limited, and several conclusions require stronger experimental support. Overall, the discovery represents an incremental advance that extends prior findings on MAP7 function, providing additional but modest insights into the role of MAP7 in cytoskeletal regulation in male reproduction.

Response: We thank the reviewer for their constructive comments and thoughtful evaluation of our manuscript. We appreciate the positive feedback regarding the value of the Map7-egfpKI mouse model for the research community. We also thank the reviewer for the suggestion to perform MAP7–MYH9 double immunofluorescence staining to demonstrate colocalization, which we agree will further strengthen the mechanistic support.

We would like to clarify that several aspects of our findings represent novel contributions within a field where the mechanisms of microtubule remodeling during apical domain formation have remained largely unresolved. In particular, our study provides evidence that MAP7 is asymmetrically enriched at the apical microtubule network in Sertoli cells and contributes to the directional organization of these microtubules—an aspect of Sertoli cell polarity that has not been previously characterized. Our results further indicate that dynamic microtubule turnover, rather than stabilization alone, is required for proper apical domain formation, addressing a gap in current understanding of how microtubules are reorganized during early polarity establishment. In addition, the data support a role for MAP7 in coordinating microtubule and actomyosin organization, suggesting a scaffolding function that links these cytoskeletal systems. We also observe that Sertoli cell polarity can be functionally separated from cell identity and that disruptions in apical domain architecture precede delays in germ cell developmental progression. Taken together, these observations provide mechanistic insight that expands upon previous studies of MAP7 function at the cellular level.

The conclusions are supported by multiple, complementary lines of evidence, including knockout and Map7-egfpKI mouse models, high-resolution electron microscopy, immunoprecipitation–mass spectrometry, and single-cell RNA sequencing. While we agree that further experiments, such as MAP7–MYH9 double staining, will strengthen the mechanistic framework, we will also perform complementary biochemical analyses to provide additional insight. Specifically, we plan to conduct domain-mapping experiments to identify the MAP7 region required for MYH9 complex formation, coupled with co-immunoprecipitation assays in cultured cells to validate this association.

Although generating new mutant mouse lines is not feasible within the scope of this revision, and no in vitro system fully recapitulates Sertoli cell polarization, these complementary approaches will provide further mechanistic support. We believe that these planned experiments, together with the current dataset, will clarify the underlying mechanisms and reinforce the significance of our findings, while appropriately acknowledging the current limits of experimental evidence.

Reviewer #2 (Evidence, reproducibility and clarity):

In this manuscript the authors evaluate the role of Microtubule Associated Protein 7 (MAP7) in postnatal Sertoli cell development. The authors build two novel transgenic mouse lines (Map7-eGFP, Map7 knockout) which will be useful tools to the community. The transgenic mouse lines are used in paired advanced sequencing experiments and advanced imaging experiments to determine how Sertoli cell MAP7 is involved in the first wave of spermatogenesis. The authors identify MAP7 as an important regulator of Sertoli cell polarity and junction formation with loss of MAP7 disrupting intracellular microtubule and F-actin arrangement and Sertoli cell morphology. These structural issues impact the first wave of spermatogenesis causing a meiotic delay that limits round spermatid numbers. The authors also identify possible binding partners for MAP7, key among those MYH9.

The authors did a great job building a complex multi-modal project that addressed the question of MAP7 function from many angles. The is an excellent balance of using many advanced methods while still keeping the project narrowed, to use only tools to address the real questions. The lack of quality testing on the germ cells outside of TUNEL is disappointing, but the Conclusion section implies that this sort of work is being done currently so the omission in this manuscript is acceptable. However, there is an issue with the imaging portion of the work on MYH9. The conclusions from the MYH9 data is currently overstated, super-resolution imaging of Map7 knockouts with microtubule and F-actin stains, and imaging that uses MYH9 with either Map7-eGFP or anti-MAP7 are also needed to both support the MAP7-MYH9 interaction normally and lack of interaction with failure of MYH9 to localize to microtubules and F-actin in knockouts. Since a Leica SP8 was used for the imaging, using either Leica LIGHTNING or just higher magnification will likely be the easiest solution.

Response: We sincerely appreciate the reviewer’s thorough and positive evaluation of our study. We are encouraged that the reviewer recognized the overall strength of our multi-modal approach and the scientific value of the Map7-egfp knock-in and Map7 knockout genome-edited mouse models that we generated. We also thank the reviewer for highlighting the balance between methodological breadth and focused, hypothesis-driven investigation in our work.

Regarding the reviewer’s valuable comments on the imaging data, we have addressed them as follows. We improved the cytoskeletal imaging data as described in response to the reviewer’s minor comments. Specifically, in the revised Figure 3B, we replaced the original images with higher-resolution confocal images to provide a clearer view of cytoskeletal organization. In addition, following Reviewer #1’s suggestion, we modified the panel layout to enlarge each field and enhance the contrast between TUBB3 and F-actin channels, allowing better visualization of their altered localization in Map7-/- testes.

We agree that super-resolution imaging comparing control and Map7-/- testes stained for TUBB3 and F-actin would further strengthen the analysis. If the current resolution is still considered insufficient, we plan to perform additional imaging using a Carl Zeiss Airyscan or Leica Stellaris 5 system to further improve spatial resolution and confirm the observed cytoskeletal phenotypes. Finally, we will perform co-imaging of MYH9 with MAP7 to validate their spatial relationship under normal conditions, complementing the existing data obtained from Map7-/- testes.

This manuscript is nicely organized with almost all of the results spelled out very clearly and almost always paired with figures that make compelling and convincing support for the conclusions. There are minor revision suggestions for improving the manuscript listed below. These include synching up Figure and Supplemental Figure reference mismatches. There are also many minor, but important, details that need to be added to the Methods section including many catalog numbers and some references.

- Some of the imaging, especially Fig4F could benefit and be more convincing with super-resolution imaging in the 150nm range (SIM, Airyscan, LIGHTNING, SoRa) possibly even just imaging with a higher magnification objective (60x or 100x)

Response: We appreciate the reviewer’s suggestion to improve the resolution of the imaging data. In addition to revising Figure 3B as described above, we have also replaced the images in Figure 4F with higher-resolution confocal images to provide a clearer view of MYH9 localization relative to microtubules and F-actin. These revised images highlight that MYH9 specifically accumulates at apical regions where microtubules and F-actin intersect, forming the apical ES, but is not localized to the basal ES-associated F-actin structures. To retain spatial context and allow readers to appreciate the overall distribution pattern, the original lower-magnification images from Figure 4F have been moved to Supplemental Figure 5.

- SuppFig1D: Please add context in the legend to the meaning of the Yellow Stars and "O->U" labels. The latter would seem to be to indicate the Ovarian and Uterine sides of the image

Response: In response to this comment, we revised the figure legend to clarify the annotations. The legend now states: “O, ovary side; U, uterus side. Asterisks indicate secretory cells that lack planar cell polarity.”

- Pg6Line7: up to P23 or up to P35?

Response: We appreciate the reviewer’s attention to this detail. The text has been revised for clarity as follows: “To examine the temporal dynamics of Sertoli cell polarity establishment, we analyzed seminiferous tubule morphology across the first wave of spermatogenesis, from postnatal day (P)10 to P35. To specifically assess the role of MAP7 in Sertoli cells while minimizing contributions from germ cells, our analysis focused on stages up to P23, before MAP7 expression becomes detectable in step 9–11 spermatids (Fig. 1), to exclude potential secondary effects resulting from MAP7 loss in germ cells.” (page 6, lines 5-10)

- SuppFig4B: Does SuppFig4B reference back to Fig3B or Fig3C? If the latter please update this in the legend.

- Pg7Line21-23: Is SuppFig3D,E meant to be referenced and not SuppFig5A,B?

- Pg8Line22-25: Is SuppFig4A meant to be reference and not SuppFig5?

- Pg8Line34-Pg9Line: Is SuppFig4B meant to be reference and not SuppFig5B?

Response: We appreciate the reviewer’s careful reading. All mismatches in Supplemental figure references have been corrected, ensuring that each reference in the text now accurately corresponds to the appropriate data.

- Pg9Line28-33: Would the authors be willing to rework this figure to include images that more closely match the reported findings? The current version does not strongly support the idea that MYH9 fails to localize to microtubule and F-actin domains in Map7 knockout P17 seminiferous tubules. This could also just be a matter of acquiring these images at a higher magnification or with a lower-end (150nm range) super-resolution system (SIM, Airyscan, LIGHTNING, SoRa etc)

Response: Following the reviewer’s recommendation, we replaced the images in Figure 4F with higher-resolution confocal images to better visualize MYH9 localization relative to microtubules and F-actin in Map7+/- and Map7-/- testes. These revised images demonstrate that MYH9 specifically accumulates at apical regions where microtubules and F-actin intersect, but not at the basal ES-associated F-actin structures. To preserve spatial context, the original low-magnification images have been moved to Supplemental Figure 5. If additional resolution is required, we are prepared to acquire further images using an Airyscan or Stellaris 5 system.

- SuppFig7A: The legend notes these are P23 samples but the image label says 8W. Please update this to whichever is the correct age.

Response: We thank the reviewer for pointing out this discrepancy. The figure legend for Supplemental Figure 7A (now revised as Supplemental Figure 8A) has been corrected to indicate that the samples are from 8-week-old mice, consistent with the image label.

- Pg16Line4-5: Please include in the text the vendor and catalog number for the C57BL/6 mice

Response: The text now specifies: “C57BL/6NJcl mice were purchased from CLEA Japan (Tokyo, Japan)” (page 17, line 4). CLEA Japan does not assign catalog numbers to mouse strains.

- Pg16Line18-19: Please include in the text the catalog number for the DMEM

- Pg16Line19-20: Please include in the text the vendor and catalog number for the FBS

- Pg16Line20: Please include in the text the vendor and catalog number for the Pen-Strep

Response: We have added vendor and catalog information as follows: “Wild-type and MAP7-EGFPKI HeLa cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, 043-30085; Fujifilm Wako Pure Chemical, Osaka, Japan) supplemented with 10% fetal bovine serum (FBS, 35-015-CV; Corning, Corning, NY, USA) and penicillin–streptomycin (26253-84; Nacalai, Kyoto, Japan) at 37 °C in a humidified atmosphere containing 5% CO₂ 18.” (page 17, lines 18-22)

- Pg17Line6-12: Thank you for including organized and detailed information about the primers, please also define the PCR protocol used including temperatures, timing, and cycles for Map7 knockout genotyping

- Pg17Line20-27: Thank you for including organized and detailed information about the primers, please also define the PCR protocol used including temperatures, timing, and cycles for Map7-eGFP genotyping

Response: The text has been updated to include the PCR conditions used for genotyping as follows: “Genotyping PCR was routinely performed as follows. Genomic DNA was prepared by incubating a small piece of the cut toe in 180 µL of 50 mM NaOH at 95 °C for 15 min, followed by neutralization with 20 µL of 1 M Tris-HCl (pH 8.0). After centrifugation for 20 min, 1 µL of the resulting DNA solution was used as the PCR template. Each reaction (8 µL total volume) contained 4 µL of Quick Taq HS DyeMix (DTM-101; Toyobo, Osaka, Japan) and a primer mix. PCR cycling conditions were as follows: 94 °C for 2 min; 35 cycles of 94 °C for 30 s, 65 °C for 30 s, and 72 °C for 1 min; followed by a final extension at 72 °C for 2 min and a hold at 4 °C. PCR products were analyzed using agarose gel electrophoresis. This protocol was also applied to other mouse lines and alleles generated in this study.” (page 18, lines 17–25)

- Pg17Line30: Please include in the text the vendor and catalog number for the Laemmli sample buffer

Response: We clarified that the buffer was prepared in-house.

- Pg17Line32&SuppTable1: Thank you for including an organized and detailed table for the primary antibodies used, please also make either a similar table or expand the current table to include secondary antibody information

- Pg17Line32: Please note in the text which primary antibodies and secondary antibodies from Supp Table 1

Response: Supplementary Table 1 has been updated to include both primary and HRP-conjugated secondary antibodies. In the Immunoblotting section of the Materials and Methods, we specified the antibodies used: “The following primary antibodies were used: mouse anti-Actin (C4, 0869100-CF; MP Biomedicals, Irvine, CA, USA), mouse anti-Clathrin heavy chain (610500; BD Biosciences, Franklin Lakes, NJ, USA), rat anti-GFP (GF090R; Nacalai, 04404-84), rabbit anti-MAP7 (SAB1408648; Sigma-Aldrich, St. Louis, MO, USA), rabbit anti-MAP7 (C2C3, GTX120907; GeneTex, Irvine, CA, USA), and mouse anti-α-tubulin (DM1A, T6199; Sigma-Aldrich). Corresponding HRP-conjugated secondary antibodies were used for detection: goat anti-mouse IgG (12-349; Sigma-Aldrich), goat anti-rabbit IgG (12-348; Sigma-Aldrich), and goat anti-rat IgG (AP136P; Sigma-Aldrich). Detailed information for all primary and secondary antibodies is provided in Supplementary Table 1.” (page 19, lines 14-22)

- Pg18Line2: Please include in the text the vendor and catalog number for the Bouin's

Response: The text has been updated to indicate that Bouin’s solution was prepared in-house

- Pg18Line3: Please include in the text the catalog number for the CREST-coated glass slides

- Pg18Line7: Please include in the text the catalog number for the OCT compound

- Pg18Line11: Please include in the text the vendor and catalog number for the Donkey Serum

- Pg18Line11: Please include in the text the vendor and catalog number for the Goat Serum

Response: The text now includes vendor and catalog information for all these reagents, including CREST-coated slides (SCRE-01; Matsunami Glass, Osaka, Japan), OCT compound (4583; Sakura Finetechnical, Tokyo, Japan), donkey serum (017-000-121; Jackson ImmunoResearch Laboratories, PA, USA), and goat serum (005-000-121; Jackson ImmunoResearch Laboratories).

- Pg18Line13: Thank you for including an organized and detailed table for the primary antibodies used, please also make either a similar table or expand the current table to include secondary antibody information

Response: We thank the reviewer for the suggestion. Supplementary Table 1 already includes information for the antibodies used for immunoblotting, and we have now added information for the Alexa Fluor-conjugated secondary antibodies used for immunofluorescence in this study.

- Pg18Line18: Please include in the text the vendor and catalog number for the DAPI

Response: The text has been updated to include the vendor and catalog number for DAPI (D9542; Sigma-Aldrich).

- Pg18Line19: Please also include information about the objectives used including catalog numbers, detectors used (PMT vs HyD)

Response: We thank the reviewer for the suggestion. The following information has been added to the Histological analysis section in Materials and Methods: “Objectives used were HC PL APO 40×/1.30 OIL CS2 (11506428; Leica) and HC PL APO 63×/1.40 OIL CS2 (11506350; Leica), with digital zoom applied as needed for high-magnification imaging. DAPI was detected using PMT detectors, while Alexa Fluor 488, 594, and 647 signals were captured using HyD detectors. Images were acquired in sequential mode with detector settings adjusted to prevent signal bleed-through.” (page 20, lines 13-17)

- Pg18Line23: Please cite in the text the reference paper for Fiji (Schindelin et al. 2012 Nature Methods PMID: 22743772) and note the version of Fiji used

- Pg18Line24: Please note the version of Aivia used

Response: We have revised the text accordingly by citing the reference paper for Fiji (Schindelin et al., 2012, Nature Methods, PMID: 22743772) and noting the version used (v.2.16/1.54p). In addition, we have added the version of Aivia used in this study (version 14.1).

- Pg18Line25: If possible, please use a more robust and reliable system than Microsoft Excel to do statistics (Graphpad Prism, Stata, R, etc), if this is not possible please note the version of Microsoft Excel used

Response: We appreciate the reviewer’s suggestion. For basic statistical analyses such as the Student’s t-test, we used Microsoft Excel (Microsoft Office LTSC Professional Plus 2021), which has been sufficient for these standard calculations. For more advanced analyses, including ANOVA and single-cell RNA-seq analyses, we used R. These details have now been added to the text.

- Pg18Line25: Please cite in the text the reference paper for R (R Core Team 2021 R Foundation for Statistical Computing "R: A Language and Environment for Statistical Computing") and note the version of R used

- Pg18Line25: Please note the specific R package with version used to do ANOVA, and cite in the text the reference for this package

Response: We have cited the reference for R (R Core Team, 2021. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria) and noted the version used (version 4.4.0) in the text. In addition, regarding ANOVA, we have added the following description: “For ANOVA analysis, linear models were fitted using the base stats package (lm function), and analysis of variance was conducted with the anova function.” (page 20, lines 23-25)

- Pg18Line25: Please clarify, was a R package called "AVNOVA" used to do ANOVA or is this a typo?

Response: We thank the reviewer for pointing this out. It was a typographical error — the correct term is “ANOVA”. The text has been corrected accordingly.

- Pg18Line32: Please include in the text the catalog number for the EPON 812 Resin

- Pg19Line3: Please include the version number for Stacker Neo

- Pg19Line5: Please include the vendor and version number for Amira 2022

- Pg19Line5: Please include the version number for Microscopy Image Browser

- Pg19Line5: Please include the version number for MATLAB that was used to run Microscopy Image Browser

Response: We added the catalog number for the EPON 812 resin and the vendor and version information for the software used. The following details have been included in the revised text:

EPON 812 resin: TAAB Embedding Resin Kit with DMP-30 (T004; TAAB Laboratory and Microscopy, Berks, UK)

Stacker Neo: version 3.5.3.0; JEOL

Amira 2022: version 2022.1; Thermo Fisher Scientific

Microscopy Image Browser: version 2.91

Note that although Microscopy Image Browser is written in MATLAB, we used the standalone version that does not require a separate MATLAB installation.

- Pg19Line: 9-10: Please include in the text the catalog number for the complete protease inhibitor

- Pg19Line14: Please include in the text the catalog number for the Magnetic Agarose Beads

- Pg19Line16: Please include in the text the catalog number for the GFP-Trap Magnetic Agarose Beads

Response: We have added the catalog numbers for the complete protease inhibitor (4693116001), control magnetic agarose beads (bmab), and GFP-Trap magnetic agarose beads (gtma).

- Pg19Line21: Please note in the text which primary antibodies and secondary antibodies from Supp Table 1

- Pg19Line21-22: Please include in the text the catalog number for the ECL Prime

Response: We thank the reviewer for the helpful suggestions. The description regarding immunoblotting (“Eluted samples were separated by SDS–PAGE, transferred to PVDF membranes…”) was reorganized: overlapping content has been removed, and the necessary information has been integrated into the “Immunoblotting” section, where details of the primary and secondary antibodies (listed in Supplementary Table 1) are already provided. In addition, the information for ECL Prime has been updated to “Amersham ECL Prime (RPN2236; Cytiva, Tokyo, Japan)”.

- Pg20Line2: Please include the version number for Xcalibur

Response: The version of Xcalibur used in this study (version 4.0.27.19) has been added to the text.

- Pg20Line5: Please cite in the text the reference paper for SWISS-PROT (Bairoch and Apweiler 1999 Nucleic Acid Research PMID: 9847139)

Response: The reference paper for SWISS-PROT (Bairoch and Apweiler, 1999, Nucleic Acids Research, PMID: 9847139) has been cited in the text.

- Pg19Line26: Please include in the text the catalog number for the NuPAGE gels

- Pg19Line28: Please include in the text the catalog number for the SimpleBlue SafeStain

Response: Both catalog numbers have been added in the Mass spectrometry section as follows: 4–12% NuPAGE gels (NP0321PK2; Thermo Fisher Scientific) and SimplyBlue SafeStain (LC6060; Thermo Fisher Scientific).

- Pg20Line26: Please include in the text the catalog number for the Chromium Singel Cell 3' Reagent Kits v3

Response: The catalog number for the Chromium Single Cell 3′ Reagent Kits v3 (PN-1000075; 10x Genomics) has been added to the text.

- Pg21Line3: Please cite in the text the reference paper for R (R Core Team 2021 R Foundation for Statistical Computing "R: A Language and Environment for Statistical Computing")

Response: The reference for R (R Core Team, 2021. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria) has already been cited in the “Histological analysis” section, where ANOVA analysis is described.

- Pg21Line3 Please cite in the text the reference for RStudio (Posit team (2025). RStudio: Integrated Development Environment for R. Posit Software, PBC, Boston, MA. URL http://www.posit.co/.)

Response: The reference for RStudio (Posit team, 2025. RStudio: Integrated Development Environment for R. Posit Software, PBC, Boston, MA, USA. URL: http://www.posit.co/) has been added to the text.

- Pg21Line23: Please include the version number for Metascape

Response: The version of Metascape used in this study (v3.5.20250701) has been added to the text.

- SuppFig12: please update the legend to include a description after the title and update the figure labeling to correspond to the legend. Also, this figure is currently not referenced anywhere in the text.

Response: We have updated the legend for Supplemental Figure 12 (Supplemental Figure 13) to include a descriptive sentence after the title and have adjusted the figure labeling to match the legend. The revised legend now reads: “Full-scan images of the agarose gels shown in Supplemental Figs. 1B and 2C are displayed in the upper and lower left panels, respectively, while the corresponding full-scan images of the immunoblots shown in Supplemental Figs. 1C and 2D are presented in the upper and lower right panels, respectively.”

As these images serve as source data, they are not referenced directly in the main text.

_Referee cross-commenting_

I generally agree with Reviewer 1 and specifically concur related to adding details about fertility assessment of the Map7 Knockout line, and enhancing the SEM imaging.

Response: As noted in our response to Reviewer #1, we have re-acquired the SEM images in high-resolution mode, focusing on the relevant regions. The new high-resolution images have replaced the original panels in revised Figure 3C, providing clearer visualization of junctional structures at P10 and P21 in Map7+/- and Map7-/- testes. The original Figure 3C images have been moved to Supplemental Figure 4B for reference.

Reviewer #2 (Significance):

There are mouse lines, and datasets that will be useful resources to the field. This work also advances our understanding of a period in Sertoli cell development that is critical to fertility but very understudied.

Response: We thank the reviewer for the positive comments and for recognizing the potential value of our mouse lines and datasets to the field, as well as the significance of our work in advancing the understanding of this critical but understudied period in Sertoli cell development.

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

Learn more at Review Commons

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity):

Summary:

The manuscript titled "Unravelling the Progression of the Zebrafish Primary Body Axis with Reconstructed Spatiotemporal Transcriptomics" presents a comprehensive analysis of the development of the primary body axis in zebrafish by integrating bulk RNA-seq, 3D images, and Stereo-Seq. The authors first clearly demonstrate the application of Palette for integrating RNA-seq and Stereo-Seq using published spatial transcriptomics data of Drosophila embryos. Subsequently, they produced serial bulk RNA-seq data for certain developmental stages of Danio rerio embryos and utilized published Stereo-Seq data. Through robust validation, the authors observe the molecular network involved in AP axis formation. While the authors show that integrating bulk RNA-seq data with Stereo-Seq improves spatial resolution, additional proof is required to demonstrate the extent of this improvement.

Response: We thank the reviewer for the positive feedback on our Palette pipeline, zSTEP construction and analysis of primary body axis development. We appreciate the constructive suggestions provided, which we can implement to improve our manuscript. As pointed out by the reviewer, some analysis procedures were not described in sufficient detail. To address this, we have added more explanatory texts and additional schematic diagrams to make the methods clearer and more understandable. We also thank the reviewer for the meticulous reading and for reminding us to include parameters, references and essential texts, which significantly improve the manuscript quality and make the manuscript more rigorous. Furthermore, as suggested by the reviewer, the extent of the improvement on the spatial resolution was not clearly demonstrated in the manuscript. Therefore, we have provided an additional figure to show the original expression on the stacked Stereo-seq slices and 3D live image compared to the expression from zSTEP, and the results indicate that zSTEP provides better, more continuous expression patterns. We still have two remaining tasks that are expected to be completed within the next month. We hope our responses have address the concerns raised by the reviewer, and we are pleased to provide any additional proof as needed.

Major Comments:

1. Lines 66-68: Discuss the limitations of existing tools and explicitly state the advantages of using Palette.

Response: We thank the reviewer for the valuable suggestion. We have added the following new texts after line 68 to emphasize the features and advantages of Palette.

"Newly developed tools are committed to integrating bulk and/or scRNA-seq data with ST data to enhance spatial resolution, focusing on expression at the spot level. However, gene expression patterns are closely correlated to the biological functions and are more critical for understanding biological processes. Therefore, a tool focusing on inferring spatial gene expression patterns would be desirable."

1. Body Pattern Genes Analysis: For both Drosophila and Danio rerio, it would be valuable to examine body pattern genes in Stereo-Seq and apply Palette to determine if the resolution of the segments improves or merges. The resolution of the A-P axis is convincing, but further evidence for other segments would be beneficial.

Response: We thank the reviewer for the suggestions. For the Drosophila data, we only used two adjacent slices for Palette performance assessment, and thus were only able to evaluate the expression patterns within the slice.

For the zebrafish data, although we have construct zSTEP as a 3D transcriptomic atlas, we have to admit that the left-right (LR) and dorsal-ventral (DV) patterning is not satisfactory enough. Here we show a section from the dorsal part of 16 hpf zSTEP that displays a relatively well-defined left-right pattern (Fig. 2). Along the left-right axis, the notochord cells are centrally located, flanked by somite cells on either side, with the outermost cells being pronephros.

One reason for the limited LR and DV patterning is that the original annotation of the ST data does not clearly distinguish all the cell types. Another reason is likely due to the disordered cell positions when stacking ST slices. Thus, our zSTEP is most suitable for investigating the AP patterns, while the performances on LR and DV patterns may not achieve the same level of accuracy.

See response letter for the figure.

1. Figure 2d: Include the A-P line for which the intensity profile was plotted in the main figure, rather than just in the supplementary material. Additionally, consider simplifying the plot by not combining three lines into one, as it complicates the interpretation of observations.

Response: We thank the reviewer for the helpful suggestions. We have updated Figure 2d and Figure S1b by adding a A-P line on each subfigure (Fig. 3). Additionally, as the reviewer suggested, we have separated the intensity plots so that each subfigure now includes a dedicated intensity plot along A-P axis.

See response letter for the figure.

1. Drosophila Data Analysis: While the alignment and validation of Danio rerio sections are clearly explained, the analysis and validation of Drosophila data are insufficiently detailed. Provide a more thorough explanation of how the intensity profiles between BDGP in situ data and Stereo-Seq data are adjusted.

Response: We thank the reviewer for raising this issue. To make the analysis procedure clearer, we have updated Figure 2a (Fig. 4) and added explanatory texts in the figure legends to describe the processing procedure for the Drosophila ST data.

See response letter for the figure.

Additionally, the following sentences have been added into the Methods section to describe the generation of the intensity profiles.

"The intensity plot profiles along AP axis were generated through the following steps: The expression pattern plot images or in situ hybridization images were imported into ImageJ and converted to grayscale. The colour was then inverted, and a line of a certain width (here set as 10) was drawn across from the anterior part to the posterior part (Fig. S1a). The signal intensities along the width of the line were measured and imported into R for generating intensity plots."

1. Figure 3d: Present a plot with the expected expression profiles of the three genes if the embryo is aligned as anticipated.

Response: We thank the reviewer for this helpful suggestion, which improves the clarity of our manuscript. We have added the following subfigure in as Figure 3d (Fig. 5) to show the expected expression profiles of the three midline genes along left-right axis.

See response letter for the figure.

1. Analysis Without Palette: Between lines 277-438, the outcome of using Palette with bulk RNA-seq and Stereo-Seq is convincing. However, consider the following:

o What would be the observations if the analysis were conducted solely with Stereo-Seq data, without incorporating bulk RNA-seq data and employing Palette?

Response: We thank the reviewer for raising this important question. Here we show the comparison of ST expression on stacked Stereo-seq slices, ST expression projected on 3D live images, and the Palette-inferred expression (Fig. 6). The stacked ST slices do not fully reflect the zebrafish morphology, and the gene expression appears sparse, making it look massive (the first row). While after projecting ST expression onto the live image, the expression patterns can be observed on zebrafish morphology, but the expression is still sparsely distributed in spots (the second row). However, the expression patterns captured by Palette in zSTEP show more continuous expression patterns (the third row), which are more similar to the observations in in situ hybridization images (the fourth row). We are considering put these analyses into the supplementary figure.

See response letter for the figure.

o This study uses only Stereo-Seq as the spatial transcriptomics reference. It would strengthen the argument to use at least one other spatial transcriptomics method, such as Visium or MERFISH, in conjunction with bulk RNA-seq and Palette, to demonstrate whether Palette consistently improves gene expression resolution.

Response: We thank the reviewer for raising this professional question. To demonstrate a broad application of Palette, it would be necessary to test Palette performance using different types of ST references. We plan to perform extra analyses to evaluate Palette performance using Visium and MERFISH data as ST references, respectively. Additionally, our Palette pipeline only takes the overlapped genes for inference. As only hundreds of genes can be detected by MERFISH, Palette can only infer the expression patterns of these genes. As mentioned in the work of Liu et al. (2023), MERFISH can independently resolve distinct cell types and spatial structures, and thus we believe Palette will also show great performance when using MERFISH as ST reference. We've already started the analyses and expect to accomplish it within the next month. And we will update the analyses as separated tutorials to the GitHub repository.

Reference:

Liu, J. et al. Concordance of MERFISH spatial transcriptomics with bulk and single-cell RNA sequencing. Life Sci Alliance 6 (2023).

1. PDAC Data Analysis: Provide a more detailed explanation of the PDAC data analysis and use appropriate colors in the tissue images to clearly distinguish cell types.

Response: We thank the reviewer for the suggestions. We have updated the colours used in the tissue images to be consistent to the colours in tissue clustering analysis. Additionally, we have added an additional subfigure in supplementary figure (Fig. 7) with more explanatory texts in the figure legends to provide a more thorough explanation for the analysis.

See response letter for the figure.

1. Comparison with Other Methods: State the limitations of not using STitch3D and Spateo for alignment and explain why these methods were not employed.

Response: We thank the reviewer for raising this constructive comment. We fully agree with you that the introduction of published alignment algorithms would be helpful in our analysis. Currently, the slice alignment is adjusted manually, and thus the main limitation of not using these tools is that manual operation may induce bias compared to the alignment generated by computational algorithm. Unfortunately, STitch3D and Spateo are not included in this study because of two reasons. First, these two newly developed tools have been recently posted, and our analyses were largely completed before that. Therefore, we only mentioned these tools in the Discussion section. Second, we do not want to embed too many external tools into our analysis, which may increase the difficulties for researchers' operation. Specifically, STitch3D and Spateo are configured to run in Python environment, while Palette is based on R packages. Moreover, without these tools, our current manual alignment also achieves desired performance. However, we value this enlightening suggestion by the reviewer and therefore plan to further compare the performance of manual alignment versus the mentioned two alignment tools. At present, we have a preliminary comparison scheme and collected relevant datasets. Hopefully, we will complete this analysis within the next 1 to 2 weeks.

Minor Comments:

1. References: Add references to the statements in lines 51-53.

Response: We thank the reviewer for reminding us of the missing references. We have added the works of Junker et al. (2014), Liu et al. (2022), Chen et al. (2022), Wang et al. (2022), Shi et al. (2023) and Satija et al. (2015) as references in line 53 as follows.

"Thus, great efforts are ongoing to construct gene expression maps of these models with higher resolution, depth, and comprehensiveness1-6."

References:

1. Junker, J.P. et al. Genome-wide RNA Tomography in the zebrafish embryo. Cell 159, 662-675 (2014).
2. Liu, C. et al. Spatiotemporal mapping of gene expression landscapes and developmental trajectories during zebrafish embryogenesis. Dev Cell 57, 1284-1298 e1285 (2022).
3. Chen, A. et al. Spatiotemporal transcriptomic atlas of mouse organogenesis using DNA nanoball-patterned arrays. Cell 185, 1777-1792 e1721 (2022).
4. Wang, M. et al. High-resolution 3D spatiotemporal transcriptomic maps of developing Drosophila embryos and larvae. Dev Cell 57, 1271-1283 e1274 (2022).
5. Shi, H. et al. Spatial atlas of the mouse central nervous system at molecular resolution. Nature 622, 552-561 (2023).
6. Satija, R. et al. Spatial reconstruction of single-cell gene expression data. Nature biotechnology 33, 495-502 (2015)

1. Scientific Name Consistency: Ensure consistency in using either "Danio rerio" or "zebrafish" throughout the manuscript.

Response: We thank the reviewer for this suggestion. We have changed "Danio rerio" to "zebrafish" to make "zebrafish" consistent throughout the manuscript.

1. Related References: Include the following relevant references:

o https://academic.oup.com/bib/article/25/4/bbae316/7705532

o https://www.life-science-alliance.org/content/6/1/e202201701

Response: We thank the reviewer for bringing these two relevant works to us. Baul et al. (2024) presented STGAT leveraging Graph Attention Networks for integrating spatial transcriptomics and bulk RNA-seq, and Liu et al. (2023) demonstrated the concordance of MERFISH ST with bulk and single-cell RNA-seq. Both are excellent works and relevant to our work. We have added these two references in line 61 and line 68, respectively.

References:

Baul, S. et al. Integrating spatial transcriptomics and bulk RNA-seq: predicting gene expression with enhanced resolution through graph attention networks. Brief Bioinform 25 (2024).

Liu, J. et al. Concordance of MERFISH spatial transcriptomics with bulk and single-cell RNA sequencing. Life Sci Alliance 6 (2023).

1. Figure 1a: In the Venn diagram, include the number of genes in the bulk and Stereo-Seq datasets, as well as the number of overlapping genes.

Response: We thank the reviewer reminding us to include these important numbers. And in our current manuscript, we have added the following sentences in the Methods section to provide the gene numbers (Fig. 8). While the Venn diagram in Figure 1a serves as a schematic representation, so we did not include the gene numbers, as these may vary depending on the actual data.

"Palette was performed on the aligned slices using the overlapped genes. For the 10 hpf embryo, there were 24,658 genes in the bulk data, 18,698 genes in the Stereo-seq data, and 16,601 overlapped genes. For the 12 hpf embryo, there were 23,018 genes in the bulk data, 18,948 genes in the Stereo-seq data, and 16,401 overlapped genes. For the 16 hpf embryo, there were 24,357 genes in the bulk data, 23,110 genes in the Stereo-seq data, and 19,539 overlapped genes."

See response letter for the figure.

1. Figure 1 Improvement: Enlarge Figure 1 and reduce repetitive elements, such as parts of the deconvolution and Figure 1b.

Response: We thank the reviewer for the helpful suggestion. We agree with the reviewer that the deconvolution sections appear repetitive. We have updated Figure 1 (Fig. 9) by replacing these repetitive elements with a clearer and simpler diagram.

See response letter for the figure.

1. Figure 3f: Explain the black discontinuous line in the plot.

Response: We thank the reviewer for the reminder. We are sorry about the lack of the explanation. We have added the below explanation for the black discontinuous line in the legend of Figure 3 (Fig. 10) as follows.

See response letter for the figure.

1. Line 610: State the percentage of unpaired imaging spots.

Response: We thank the review for the reminder. We are sorry about not including the paired and unpaired spot number. We have added the number of paired spots with the percentage in the total spots in the Method section as follows.

"The numbers of mapped spots for the 10 hpf, 12 hpf and 16 hpf embryos are 15,379 (69.4% of the total spots), 14,697 (70.5% of the total spots) and 21,605 (77.2% of the total spots), respectively."

1. Lines 616-618: Specify the unit for the spot diameter.

Response: We thank the reviewer for the reminder. Again, we are sorry about not including the spot diameter information in our previous version of manuscript. We have added the spot diameter in Method section as follows.

"In the Stereo-seq data, each spot contained 15 × 15 DNA nanoball (DNB) spots (The diameter of each spot is near 10 μm)."

Reviewer #1 (Significance):

This algorithm will be useful not only for the field of developmental biology but also for wider applications in spatial omics. Although I have expertise in spatial omics technology development, my understanding of computational biology is limited, which restricts my ability to fully evaluate the Palette algorithm presented in this paper.

Response: We thank the reviewer for recognizing our work, and we greatly appreciate the constructive suggestions from the reviewer. Although the reviewer acknowledged limited expertise in computational biology, the comments from the reviewer are highly professional and valuable. Following the suggestions from the reviewer, we have not only included more explanatory texts and figures to make the analysis procedures clearer and more understandable, but also supplemented the important parameters that were missing in our previous manuscript. We also provided extra figure to demonstrate the improvements of zSTEP on gene expression patterns. We believe that our work is now more scientific and more understandable, and we will continue working to solve the remaining issues as planned. We express our thanks for the reviewer again.

Reviewer #2 (Evidence, reproducibility and clarity):

The authors of the study introduce the Palette method, a novel approach designed to infer spatial gene expression patterns from bulk RNA-sequencing (RNA-seq) data. This method is complemented by the development of the DreSTEP 3D spatial gene expression atlas of zebrafish embryos, establishing a comprehensive resource for visualizing gene expression and investigating spatial cell-cell interactions in developmental biology.

Response: We sincerely appreciate the reviewer's positive feedback on our Palette pipeline and the zSTEP 3D spatial expression atlas of zebrafish embryos. We also thank the reviewer for the professional comments and constructive suggestions. The reviewer raised the concerns from the aspect of algorithm design and computational biology, which we did not address well in our previous manuscript. We agree with the reviewer that we did not clarify the selection criteria of the parameters in detail, and we are now working on the additional analyses to address this issue.

We also agree with the reviewer that we did not provide enough discussion of the strategies used in the pipeline, the features of Palette and the application scenarios of Palette and zSTEP. For wide use of our tools, it is significantly important to state these aspects. In this revised version, we have added more paragraphs in the Discussion section to address this issue. Additionally, we acknowledge that we did not adequately demonstrate the computational efficacy and computational requirements, which are important for researchers. We are also working on the additional analyses to address this issue.

Finally, we thank the reviewer again for the professional and constructive suggestions. These suggestions are addressable, and by following them, we believe our manuscript will see a significant improvement, especially in the Palette pipeline part, making the pipeline more rigorous and easier to access. We are confident that we can complete the planned additional tasks within the next 1-2 months.

1. The efficacy of the Palette method may be compromised by its dependency on the quality of the reference spatial transcriptomics data. As highlighted in the study, variations in data quality can lead to significant challenges in reconstructing accurate spatial expression patterns from bulk data. This underscores the necessity of evaluating quality parameters, such as the number of gene detections and spatial resolution, to ensure reliable outcomes. Additional studies should rigorously assess how these quality factors influence the accuracy and efficiency of the algorithm in various data contexts, particularly under diverse conditions of gene detection.

Response: We thank the reviewer for this valuable suggestion. We agree with the reviewer that the quality of the reference ST data may greatly influence the performance and efficacy of the Palette, and we have added paragraphs in the Discussion section to further discuss the impact of ST data quality on Palette performance. As mentioned by the reviewer, gene detections and spatial resolution are two important parameters that can influence the Palette performance. Low gene detection may impact the clustering process, making the cell types of spots not distinguished well. To evaluate the performance of Palette when ST data shows low gene detection, we plan to applied Palette using MERFISH data as the ST reference, which only captures hundreds of genes. Moreover, we will also investigate the impact of spatial resolution on Palette performance by merging ST spots to simulate lower resolution scenarios, as well as the impact of gene detection by randomly reducing detected genes. Through the comparison among the inferred expression patterns with ST data of different spatial resolutions or different numbers of detected genes, we can better access the performance of Palette and provide guidance to researchers on the appropriate ST data requirements for optimal performance. These analyses will take another one month to accomplish after this round of revision due to the limited response time.

1. The methodology raises pertinent questions regarding how the clustering results from different algorithms may affect the reconstructions by the Palette method. The authors would better provide a detailed discussion/comparison of clustering processes that optimize the reconstruction of spatial patterns, ensuring precision in the downstream analyses.

Response: We thank the reviewer for the constructive comments. We agree with the reviewer that the differences in clustering results would impact the inference of the Palette. In our Palette pipeline, rather than develop a new methodology for clustering, we employ the BayesSpace for spot clustering, which considers both spot transcriptional similarity and neighbouring structure for clustering. In this case, researchers may adjust the parameters in the BayesSpace package to achieve optimal clustering results. Actually, in most cases, the spot identities were achieved through UMAP analysis, which only considers the transcriptional differences but does not consider the spatial information. This kind of clustering strategy will potentially lead to an intricate arrangement of spots belonging to different clusters, and may result in sparse gene expression in Palette outcome, which is different from the patterns in bona fide tissues. Therefore, a suitable clustering strategy will definitely help capture the local patterns.

Moreover, our Palette pipeline also can use the clustering results from the tissue histomorphology. Using tissue histomorphology for clustering would be a good choice, as it is closer to the real case. The following Figure (Fig. 11) displays the Palette performance on PDAC datasets using both spatial clustering and histomorphology clustering strategies. The result using histomorphology clustering captures the weak pattern (indicated by the red circle) that were missed when using the spatial clustering (Fig. 11d).

See response letter for the figure.

1. The choice to utilize only highly expressed genes in the initial stages of the Palette algorithm also warrants further exploration. Addressing the criteria for determining which genes qualify as "highly expressed" and outlining robust cutoff will enhance the algorithm's rigor and applicability. Similarly, in the iterative estimation of gene expression across spatial spots, establishing optimal iteration conditions is crucial. Implementing a loss function may offer a systematic method for concluding iterations, thus refining computational efficiency.

Response: We thank the reviewer for the professional suggestions. As pointed out by the reviewer, the selection of highly expressed genes and the iteration times are two important parameters in our pipeline. The definition of highly expressed genes and the number of highly expressed genes are important for achieving a satisfactory clustering performance. We tested the impact of different numbers of highly expressed genes on cluster performance in our preliminary analyses, while we did not summarize these tests and specify the parameters. Therefore, we plan to include a supplementary figure showing the clustering performances under different definitions of highly expressed genes and different numbers of highly expressed genes. Additionally, for the iteration conditions, we have tested different iteration numbers to find out a suitable iteration number to achieve a stable expression in each spot. The following figure (Fig. 1) shows the results after performing Palette with different iteration times. We randomly selected 20 cells and compared their expression across tests with varying iteration times. The results indicate that for a ST dataset with 819 spots, the expression in each spot becomes nearly stable after 5000 iteration times. We previously did not consider the computational efficiency, while here the reviewer raises a valuable and professional suggestion to implement a loss function to determine the optimal number of iterations. We greatly appreciate this suggestion, and plan to apply a loss function to summarize the optimal iteration times for ST datasets of different sizes. This will provide guidance for potential researchers in selecting iteration times and enhance computational efficiency.

See response letter for the figure.

1. Performance metrics relating to processing speed and computational demands remain inadequately addressed in the current framework. Understanding how the Palette method scales across varying gene counts and bulk RNA-seq datasets will be essential for potential applications in larger biological contexts. Notably, the quantitative demands of analyzing 20,000 genes when processing 10, 100, or 1,000 bulk RNA profiles must be articulated to guide researchers in planning accordingly.

Response: We thank the reviewer for this valuable and professional suggestion. In our previous analyses, we did not consider the computation efficiency, processing speed and computational demands, which are important information for potential researchers. To address this issue, we will list our computer configuration first. And under this configuration, we plan to run Palette on datasets with different numbers of overlapped genes or ST references with varying spot numbers, and then summarize the running times into a metrics table. This will help researchers estimate the running time for their datasets and guide them in planning the analyses. We will begin the analyses soon and expect to complete the analysis within the next 1 to 2 months.

Minor opinions:

1. Despite the promising advances offered by the zebrafish 3D reconstruction, there is a lack of details regarding numbers of the spatial transcriptomics (ST) data utilized, and the number of bulk RNA-seq data employed in the analyses. These parameters need to be clarified.

Response: We thank the reviewer for reminding us of these parameters. We are sorry for not including these parameters in our previous manuscript. We have now included the numbers of bulk, ST and overlap genes in the Methods section as follows (Fig. 12).

"Palette was performed on the aligned slices using the overlapped genes. For the 10 hpf embryo, there were 24,658 genes in the bulk data, 18,698 genes in the Stereo-seq data, and 16,601 overlapped genes. For the 12 hpf embryo, there were 23,018 genes in the bulk data, 18,948 genes in the Stereo-seq data, and 16,401 overlapped genes. For the 16 hpf embryo, there were 24,357 genes in the bulk data, 23,110 genes in the Stereo-seq data, and 19,539 overlapped genes."

See response letter for the figure.

1. Issues regarding spatial cell-cell communication, especially concerning interactions over longer distances, necessitate careful consideration. Introducing spatial distance constraints could help formulate more realistic models of cellular interactions, a vital aspect of embryonic development.

Response: We thank the reviewer for this essential comment. We agree with the reviewer that the spatial distance is an essential factor to investigate in vivo cell-cell communication during embryonic development. Therefore, in our analyses, we employed CellChat for spatial cell-cell communication analysis, which can be used to infer and visualize spatial cell-cell communication network for ST datasets, considering the spatial distance as constrains of the computed communication probability. However, during our analyses, we observed that there were interactions between cell types over longer distances, as mentioned by the reviewer. We then investigated how these interactions of longer distances occurred. Here, we show the FGF interaction between tail bud and neural crest cells from our spatial cell-cell analysis as an example, and the distance between these two cell types appears quite significant (Fig. 13). We labelled tail bud cells and neural crest cells on the selected midline section and observed that, although most neural crest cells are distributed anteriorly, a small number of neural crest cells are located at tail, close to the tail bud cells. Therefore, the observed interaction between tail bud and neural crest cells is likely due to their adjacent distribution in the tail region, while the anteriorly distributed of neural crest spot in spatial cell-cell communication analysis reflects the anterior positioning of most neural crest cells. As a result, the distances shown on the spatial cell-cell communication analysis are not the real distance between two cell types.

In most cases in our spatial cell-cell communication analyses, the observed interactions over longer distances are likely influenced by this visualization strategy. Additionally, pre-processing the dataset may enhance the performance of the analyses. Here we performed systematic analyses of the entire embryo, which can make the interactions between cell types appear massive. To investigate specific biological questions, researchers can subset cell types of interest or categorize them into different subtypes based on their positions.

See response letter for the figure.

1. Evaluation metrics such as the Adjusted Rand Index (ARI) and Root Mean Square Error (RMSE) represent critical tools for systematically measuring the similarity of inferred spatial patterns, yet their specific application within this context should be elaborated.

Response: We thank the reviewer for recommending these two tools. We have applied them to evaluate the similarity between the expression patterns (Fig. 14). The inclusion of these statistical values makes our comparisons of expression patterns more scientific and convincing. And we have added the following texts in the Methods section to describe the calculation of these two values.

"The Adjusted Rand Index (ARI) and Root Mean Square Error (RMSE) were used to evaluate the similarity of the expression patterns. The expression patterns of in situ hybridization images were considered as the expected values, and the expression patterns of ST data and inferred expression patterns were compared to the expected values. Common positions along the AP axis within all three expression profiles were used, and the RMSE were calculated based on the scaled intensity of these positions. Values greater than the threshold were set to 1; otherwise, they were set to 0, and the ARI was then calculated based on the intensity category. Higher ARI and lower RMSE indicate greater similarity."

See response letter for the figure.

1. The study's limitations surrounding ST data quality cannot be overstated. Discussing scenarios where only limited or poor-quality ST data are available will be crucial for guiding future studies. Furthermore, a clear explanation of how enhanced specificity and accuracy translate into tangible biological insights is essential for demystifying the underlying mechanisms driving developmental processes.

Response: We thank the reviewer for raising this essential suggestion. We have realized that in our previous manuscript, our discussion on the advantages and limitations of Palette and zSTEP was neither broad nor detailed enough.

Therefore, in our revised manuscript, we have added the following paragraphs to further discuss the advantages and limitations of Palette and zSTEP, as well as the potential application of zSTEP in developmental biology.

In this section, we have emphasized again the impact of ST data quality on the performance of Palette and zSTEP, and then compared Palette with the strategy that uses well-established marker genes to infer spatial information. We demonstrated that although Palette cannot achieve single cell resolution, it captures the major expression patterns, which are closely correlated to biological functions and critical for embryonic development. Furthermore, we further discussed that zSTEP is not only a valuable tool for investigating gene expression patterns, but also has the potential in evaluating the reaction-diffusion model to investigate the complicated and well-choreographed pattern formation during embryonic development.

As here we have provided a more comprehensive discussion about Palette and zSTEP, we think that the potential researchers will better understand the application scenarios of our inference pipeline and our datasets. We hope our study can assist and inspire further research in the field of spatial transcriptomics and developmental biology.

"Thirdly, the performance of Palette and zSTEP heavily relied on the quality of ST data. If the quality of ST data is not of sufficient quality, the low-expression genes may not be detected or only appear in very few scattered spots, and the performance of spot clustering could also be affected. Moreover, in this study, for example, the Stereo-seq data of 12 hpf zebrafish embryo had fewer slices on the right side (Fig. S3b), resulting in more blank spots in the right part of zSTEP for the 12 hpf embryo. However, with the ongoing advancements in spatial resolution and data quality, the performance of Palette is expected to be enhanced and demonstrate even greater potential for analysing spatiotemporal gene expression.

On the other hand, compared to the brilliant strategy that infers spatial information of scRNA-seq data from well-established genes, our Palette pipeline cannot achieve single cell resolution. However, our Palette pipeline is based on the ST reference, and thus preserves the real positional relationships between spots. Furthermore, the focus of our pipeline is to infer the gene expression patterns, which are closely correlated to biological functions and critical for embryonic development, rather than the sparse expression within individual spots. In this regard, our Palette pipeline can be advantageous, as it allows for reconstruction of the major expression profiles, which are often more relevant for understanding developmental processes. Additionally, our Palette can be applied to serial sections, enabling the construction of 3D ST atlas.

Finally, while the current analyses demonstrated that zSTEP can serve as a valuable tool for identifying genes having specific patterns at certain developmental stages, the exploration of zSTEP is still limited. During animal development, pattern formation is always one of the most important developmental issues. As demonstrated by the reaction-diffusion (RD) model, morphogen molecules are produced at specific regions of the embryo, forming morphogen gradients to guide cell specification, while interactions between different morphogens instruct more complicated and well-choreographed pattern formation. Our Palette constructed zSTEP, as a comprehensive transcriptomic expression pattern during development, could be leveraged to evaluate and prove the RD model during development, including AP patterning. Moreover, the investigation of gene expression patterns should not be limited to morphogens and TFs, and further investigation of their roles in AP patterning is desirable. Additionally, here a random forest model may be sufficient for investigating the most essential morphogens and TFs for AP axis refinement, while more sophisticated machine learning models may be required for addressing more specific biological questions."

Reviewer #2 (Significance):

The Palette pipeline demonstrates a marked improvement in specificity and accuracy when predicting spatial gene expression patterns. Evaluative studies on Drosophila and zebrafish datasets affirm its enhanced performance compared to existing methodologies. By effectively reconstructing spatial information from bulk transcriptomic data, the Palette method innovatively merges the philosophy of leveraging single-cell transcriptomic data for deconvolution analyses. This integration is pivotal, advancing traditional bulk RNA-seq approaches while laying the groundwork for future research.

One of the notable achievements in this work is the construction of the DreSTEP atlas, which integrates serial bulk RNA-seq data with advanced 3D imaging techniques. This resource grants researchers unprecedented access to the visualization of gene expression patterns across the zebrafish embryo, facilitating the investigation of spatial relationships and cell-cell interactions critical for developmental processes. Such capabilities are invaluable for understanding the intricate dynamics of embryogenesis and the distinct roles of individual cell types.

Response: We thank the reviewer for the positive evaluation of our work, either the Palette pipeline or zSTEP. The reviewer has strong expertise in algorithm development and computational biology, and the concerns and suggestions from the reviewer are significantly precious and valuable for us. Regarding the bioinformatics tool development, we did not have extensive experiences, and thus we did not thoroughly address the selection criteria or clarify the parameters used in the pipeline, which may influence the application by other researchers. Therefore, we sincerely appreciate the professional suggestions from the reviewer, which we can follow to address these issues, improve our manuscript and make our work more impactful for researchers. Additionally, we did not consider computation efficiency, processing speed and computational demands, which would be important factors for other researchers to use Palette. We would like to add extra analyses to address these aspects.

Currently, based on the suggestions from the reviewer, we have added extra texts discussing the clustering strategy in Palette pipeline, the advantages and limitations of Palette, and the potential application of zSTEP in developmental biology. We believe that readers will now have a clearer understanding of the performance of Palette and the application scenarios of both Palette and zSTEP. We have not fully addressed the comments raised by the reviewer yet, while we are working on the planned additional analyses and expect to complete all these tasks within the next 1-2 months. We sincerely thank the reviewer for the professional and valuable suggestions, which definitely improve our work and will make it accessible for a wide range of researchers.

Finally, through this review process, we have learned a lot about the important considerations and requirements when designing bioinformatics tools, and we benefit a lot from the thoughtful guidance. We express our thanks to the reviewer again for the guidance, and we will try our best to address the remaining issues to further improve our manuscript.

Reviewer #3 (Evidence, reproducibility and clarity):

Evidence, reproducibility and clarity

In this study, Dong and colleagues developed a computational pipeline to use spatial transcriptomics (ST) datasets as a reference to infer the spatial patterns of gene expression from bulk RNA sequencing data. This approach aims to overcome the low read depth and limited gene detection capabilities in current ST datasets, while exploiting its ability to provide highly resolved spatial information. By combining bulk RNA-seq datasets from 3 developmental stages during early zebrafish development with previously available ST and imaging datasets, the authors build DreSTEP (Danio rerio spatiotemporal expression profiles). Using this approach, they go on to identify the morphogens and transcription factors involved in anteroposterior patterning.

The paper is well written, and the pipeline presented in this study is likely to be useful beyond the case studies included in this study. There are a few questions that, in my view, would be important to clarify to increase the impact of this work:

Response: We sincerely appreciate the positive feedback from the reviewer on the Palette pipeline and zebrafish spatiotemporal expression profiles zSTEP. We thank the reviewer for the constructive suggestions, which have inspired us to think deeply about application and advantages of Palette and zSTEP for future studies.

We fully agree with the reviewer that we do not sufficiently clarify the advantages and limitations of our inference pipeline in the original manuscript. The questions raised by the reviewer are very insightful. For example, while the inference expression patterns may closely resemble the in situ hybridization observation, which we consider as good performance, the reviewer pointed out that we should consider whether weak, yet real expression may have been removed. These questions have motivated us to think more deeply about the underlying principles and assumptions of our inference pipeline. Following the reviewer's questions, we have expanded our discussion on the application of zSTEP in developmental biology and the features of Palette compared to the existing strategies.

We believe that after incorporating the revisions, our current manuscript now demonstrates the application scenario of Palette clearer and suggested the application of zSTEP for investigating biological questions in developmental biology. We are grateful for the reviewer's guidance, which helps us increase the impact of our work.

1. The authors mention that they used a variable factor to adjust expression differences between the ST and bulk RNA-seq datasets. It would be important for the authors to comment on how much overlap in gene expression is necessary between the datasets for an accurate calculation of this variable factor? Can this be directly tested, for instance, by testing how their conclusions vary if expression is adjusted by a variable factor calculated from only a smaller set of genes?

Response: We thank the reviewer for the professional questions. We are sorry about not including the gene numbers in our previous manuscript. And now we have provided the numbers of genes in bulk and ST data and the numbers of the overlapped genes (Fig. 15).

"Palette was performed on the aligned slices using the overlapped genes. For the 10 hpf embryo, there were 24,658 genes in the bulk data, 18,698 genes in the Stereo-seq data, and 16,601 overlapped genes. For the 12 hpf embryo, there were 23,018 genes in the bulk data, 18,948 genes in the Stereo-seq data, and 16,401 overlapped genes. For the 16 hpf embryo, there were 24,357 genes in the bulk data, 23,110 genes in the Stereo-seq data, and 19,539 overlapped genes."

See response letter for the figure.

For Palette implementation, we took all the overlapped genes. To calculate the variable factor, we aggregated the expression of each gene in the ST data, and then used the expression of the bulk data to divide the aggregated expression for variable factor calculation. As a result, each overlapped gene was assigned a variable factor to adjust its expression, based on its difference between bulk and ST data. The rationale behind this approach is that by considering the ST data as a whole, we can effectively reduce the variations among individual spots. This allows the variable factors to provide reasonable adjustment to gene expression.

Above all, the variable factors can be directly calculated. Currently Palette only can infer the expression patterns of overlapped genes. It means when the number of overlapped genes is small, such as MERFISH only detecting hundreds of genes, Palette can only infer the expression patterns of these genes. However, if the MERFISH data have good quality, which enable resolving distinct cell types, we believe Palette will also show good performance when using MERFISH as ST reference. Additionally, we plan to perform Palette using MERFISH as ST reference to further demonstrate its broad application when using different ST references.

1. Palette gives rise to highly spatially precise patterns, which closely match those found in ISH. However, the smoothening of the expression can also remove weak, yet real, local expression patterns, as shown for idgf6 in Fig. 2a. Can the authors test this more extensively for other genes?

Response: We thank the reviewer for this essential question. We agree with the reviewer that weak, yet real expression might be removed in our Palette inference pipeline. The weak, sparse expression may be due to the ST technique itself or the variations in samples. However, that sparse gene expression may not have biological meaning, and the focus of our pipeline in to capture the expression patterns, which are closely correlated with functions and crucial for embryonic development. Therefore, our algorithm considers spot characteristics and emphasize cluster-specific expression, resulting in spatial-specific expression patterns. In most cases, the main gene expression patterns can be captured, which can help understand gene functions and roles in embryonic development. We have updated Supplementary Figure S1a (Fig. 16) to include more gene patterns to demonstrate this point.

See response letter for the figure.

1. Using adjacent slices for ST and "bulk RNA-seq" may provide better results than those obtained when comparing two independent datasets. Could the authors also extend the analysis of Palette's functionalities by using separate, previously available but independent datasets, for ST and bulk RNA-seq in Drosophila as well?

Response: We thank the reviewer for the valuable question. We agree with the reviewer that using adjacent slices may provide better results. The idea here is that the inferred spatial expression patterns from pseudo bulk RNA-seq can be used to compare with the real expression of ST to evaluate Palette performance. We have updated our Figure 2a (Fig. 17) to illustrate the analysis clearer.

See response letter for the figure.

To demonstrate the Palette's functionalities, we have used Palette to infer zebrafish bulk RNA-seq slice (Junker et al., 2014) using Stereo-seq slice (Liu et al., 2022) as ST reference, and these two datasets are separate and independent. We agree with the reviewer that it would be good to use separate datasets to test in Drosophila to further demonstrate the Palette's functionalities. However, unfortunately, we did not find the Drosophila serial bulk RNA-seq data along left-right axis of the corresponding stages, and thus we might be unable to perform the extra analyses using independent Drosophila datasets.

References:

Junker, J.P. et al. Genome-wide RNA Tomography in the zebrafish embryo. Cell 159, 662-675 (2014).

Liu, C. et al. Spatiotemporal mapping of gene expression landscapes and developmental trajectories during zebrafish embryogenesis. Dev Cell 57, 1284-1298 e1285 (2022).

1. The DreSTEP analysis in zebrafish embryos is interesting and validates well-established observations in the field. Can the authors also discuss whether and how their dataset allows them to refine our understanding of the spatial or temporal pattern of the morphogens and TFs involved in AP patterning? This would further validate their approach.

Response: We appreciate the reviewer for recognition of our zSTEP and raising this valuable question, which has inspired us to think more deeply about the potential application of zSTEP in developmental biology. As the reviewer noted, our zSTEP analyses have validated well-established observations in the field. Rather than focusing on the sparse expression detected in ST data, zSTEP emphasizes the gene expression patterns that are closely correlated with biological functions and critical for embryonic development. Therefore, zSTEP can serve as a valuable tool for identifying the genes having specific patterns at certain developmental stages.

Pattern formation is one of the most important developmental issues for all animals. The reaction-diffusion (RD) model is a widely recognized theoretical framework used to explain self-regulated pattern formation in developing animal embryos (Kondo & Miura, 2010). Morphogen molecules are produced at specific regions of the embryo, forming morphogen gradients to guide cell specification. Most importantly, interactions between different morphogens instruct more complicated and well-choreographed pattern formation. Our Palette-constructed zSTEP provides a comprehensive transcriptomic expression pattern, including all morphogens and TFs, across the whole embryo during development. These valuable resources, in our opinion, could be leveraged to evaluate and prove the RD model during development, including AP patterning. In our current zSTEP analyses, we have already identified genes that exhibit specific expression patterns along AP axis, some of which have not been fully characterized. These genes could be potential targets for further investigation into their roles in AP patterning, although they are not the primary focus of this study. Additionally, our analyses only focused on morphogens and TFs, but zSTEP can be used to investigate the expression patterns of other genes as well. Moreover, we employed a random forest model to investigate the most essential morphogens and TFs for AP axis refinement, which is one of the basic applications of zSTEP. To investigate specific biological questions of interest, it would be worth exploring the use of more sophisticated machine learning models.

We have added the following paragraph in the Discussion section to discuss the potential application of zSTEP in future studies.

"Finally, while the current analyses demonstrated that zSTEP can serve as a valuable tool for identifying genes having specific patterns at certain developmental stages, the exploration of zSTEP is still limited. During animal development, pattern formation is always one of the most important developmental issues. As demonstrated by the reaction-diffusion (RD) model, morphogen molecules are produced at specific regions of the embryo, forming morphogen gradients to guide cell specification, while interactions between different morphogens instruct more complicated and well-choreographed pattern formation. Our Palette constructed zSTEP, as a comprehensive transcriptomic expression pattern during development, could be leveraged to evaluate and prove the RD model during development, including AP patterning. Moreover, the investigation of gene expression patterns should not be limited to morphogens and TFs, and further investigation of their roles in AP patterning is desirable. Additionally, here a random forest model may be sufficient for investigating the most essential morphogens and TFs for AP axis refinement, while more sophisticated machine learning models may be required for addressing more specific biological questions."

Reference

Kondo, S. & Miura, T. Reaction-Diffusion model as a framework for understanding biological pattern formation. Science 329, 1616-1620 (2010).

1. Can the authors comment on the limits of this inference pipeline? And how it performs as compared to single-cell RNA sequencing datasets where spatial information is inferred from well-established marker genes?

Response: We appreciate the reviewer for this insightful question, which has inspired us to further explore the advantages and limitations of the Palette pipeline in comparison with other inference strategies. As mentioned in the Discussion section, a key limitation of the inference pipeline is its heavy reliance on the quality of ST data. It is obvious that if the quality of ST data is not of sufficient quality, the low-expression genes may not be detected or only appear in very few scattered spots. We think it is a common issue for any inference tools using ST data as the reference. However, with the ongoing advancements in spatial resolution and data quality, the performance of Palette is expected to be improved.

As a comparison, the single-cell RNA sequencing datasets where spatial information is inferred from well-established marker genes do not face this limitation. The ground-breaking work by Satija et al. (2015) used such a strategy that combined scRNA-seq and in situ hybridizations of well-established marker genes to infer spatial location, enabling single cell resolution, as it maintains the high read depth and gene detection. One advantages of this scRNA-seq-based strategy is that it provides the transcriptomics of individual cells, rather than a combination of cell within a ST spot, although the positional relationships between cells are not real.

However, compared to the inference from ST data, the positional relationships between cells are not directly captured. On the other hand, as the embryonic development progresses, more cell types will be specified, and the body patterning becomes more complex. In this scenario, using well-established marker gene to infer spatial information would be much more challenging. Additionally, there are not many scRNA-seq datasets of serial sections, and thus this strategy may not be used to construct 3D ST atlas.

In contrast, our Palette inference pipeline is based on the ST data, which preserves the real positional relationships between spots. Although our inference pipeline cannot achieve single cell resolution, it focuses on the gene expression patterns rather than the sparse expression within individual spots. By applying Palette to paired serial sections, we were able to generated a 3D spatial expression atlas of zebrafish embryos, which has showed promising performance for investigating gene expression patterns and their involvement in AP patterning.

Reference

Satija, R. et al. Spatial reconstruction of single-cell gene expression data. Nature biotechnology 33, 495-502 (2015)

We have updated the following paragraphs to further demonstrating the limitation of the inference pipeline in details in the Discussion section.

"Thirdly, the performance of Palette and zSTEP heavily relied on the quality of ST data. If the quality of ST data is not of sufficient quality, the low-expression genes may not be detected or only appear in very few scattered spots, and the performance of spot clustering could also be affected. Moreover, in this study, for example, the Stereo-seq data of 12 hpf zebrafish embryo had fewer slices on the right side (Fig. S3b), resulting in more blank spots in the right part of zSTEP for the 12 hpf embryo. However, with the ongoing advancements in spatial resolution and data quality, the performance of Palette is expected to be enhanced and demonstrate even greater potential for analysing spatiotemporal gene expression.

On the other hand, compared to the brilliant strategy that infers spatial information of scRNA-seq data from well-established genes, our Palette pipeline cannot achieve single cell resolution. However, our Palette pipeline is based on the ST reference, and thus preserves the real positional relationships between spots. Furthermore, the focus of our pipeline is to infer the gene expression patterns, which are closely correlated to biological functions and critical for embryonic development, rather than the sparse expression within individual spots. In this regard, our Palette pipeline can be advantageous, as it allows for reconstruction of the major expression profiles, which are often more relevant for understanding developmental processes. Additionally, our Palette can be applied to serial sections, enabling the construction of 3D ST atlas."

Reviewer #3 (Significance):

This study tackles an important challenge in biology - the difficult to resolve gene expression patterns with high spatial precision and in a high-throughput manner. By integrating sequencing datasets from previously published studies, as well as newly-generated datasets, the authors provide evidence that their novel inference pipeline enables them to obtain high-quality spatial information simply from bulk RNA-seq datasets, using ST as a reference. The development of this pipeline - Palette - is a major part of this manuscript and its applicability is validated using datasets from Drosophila and zebrafish embryos. This in an important advance for the field, but it would be nice for the authors to further comment on i) the validity of some of their approaches and how they may influence the quality of their inference, as well as, ii) potential pitfalls/limitations of this approach as compared to others available in the field. This would synthetize both previous and current findings into a conceptual and technological framework that would have a strong impact well beyond cell and developmental biology.

Audience: This study would be relevant for a broad audience of biologists, interested in morphogen signaling, gene regulatory networks and cell fate specification.

Expertise in zebrafish development, gastrulation, morphogen signaling and morphogenesis.

Response: We thank the reviewer for providing the positive feedback, arising these valuable questions, which have motivated us to deeply consider the design concept and further application of Palette and zSTEP. Based on the insightful questions from the reviewer, we have added two extra paragraphs in the Discussion section to further discuss the potential application of zSTEP in developmental biology and application scenarios of the Palette pipeline. Specially, we have demonstrated that the performance of the inference pipeline relies on the spatial resolution and data quality of the ST data. We have then compared the advantages and limitations of Palette with the existing brilliant spatial inference strategy, which infers spatial information of scRNA-seq from well-established marker genes. Although our inference pipeline cannot achieve single cell resolution, it can capture the major expression patterns, which are closely correlated to functions and critical for embryonic development. We believe this will help readers gain a clearer understanding of the advantage and limitations of our pipeline compared to other tools, as well as the tasks for which Palette and our constructed zSTEP can be utilized. We express our thanks to the reviewer again for the valuable comments.

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

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

In this study, Dong and colleagues developed a computational pipeline to use spatial transcriptomics (ST) datasets as a reference to infer the spatial patterns of gene expression from bulk RNA sequencing data. This approach aims to overcome the low read depth and limited gene detection capabilities in current ST datasets, while exploiting its ability to provide highly resolved spatial information. By combining bulk RNAseq datasets from 3 developmental stages during early zebrafish development with previously available ST and imaging datasets, the authors build DreSTEP (Danio rerio spatiotemporal expression profiles). Using this approach, they go on to identify the morphogens and transcription factors involved in anteroposterior patterning.

The paper is well written, and the pipeline presented in this study is likely to be useful beyond the case studies included in this study. There are a few questions that, in my view, would be important to clarify to increase the impact of this work:

1. The authors mention that they used a variable factor to adjust expression differences between the ST and bulk RNAseq datasets. It would be important for the authors to comment on how much overlap in gene expression is necessary between the datasets for an accurate calculation of this variable factor? Can this be directly tested, for instance, by testing how their conclusions vary if expression is adjusted by a variable factor calculated from only a smaller set of genes?
2. Palette gives rise to highly spatially precise patterns, which closely match those found in ISH. However, the smoothening of the expression can also remove weak, yet real, local expression patterns, as shown for idgf6 in Fig. 2a. Can the authors test this more extensively for other genes?
3. Using adjacent slices for ST and "bulk RNAseq" may provide better results than those obtained when comparing two independent datasets. Could the authors also extend the analysis of Palette's functionalities by using separate, previously available but independent datasets, for ST and bulk RNAseq in Drosophila as well?
4. The DreSTEP analysis in zebrafish embryos is interesting and validates well-established observations in the field. Can the authors also discuss whether and how their dataset allows them to refine our understanding of the spatial or temporal pattern of the morphogens and TFs involved in AP patterning? This would further validate their approach.
5. Can the authors comment on the limits of this inference pipeline? And how it performs as compared to single-cell RNA sequencing datasets where spatial information is inferred from well-established marker genes?

Significance

This study tackles an important challenge in biology - the difficult to resolve gene expression patterns with high spatial precision and in a high-throughput manner. By integrating sequencing datasets from previously published studies, as well as newly-generated datasets, the authors provide evidence that their novel inference pipeline enables them to obtain high-quality spatial information simply from bulk RNAseq datasets, using ST as a reference. The development of this pipeline - Palette - is a major part of this manuscript and its applicability is validated using datasets from Drosophila and zebrafish embryos. This in an important advance for the field, but it would be nice for the authors to further comment on i) the validity of some of their approaches and how they may influence the quality of their inference, as well as, ii) potential pitfalls/limitations of this approach as compared to others available in the field. This would synthetize both previous and current findings into a conceptual and technological framework that would have a strong impact well beyond cell and developmental biology.

Audience: This study would be relevant for a broad audience of biologists, interested in morphogen signaling, gene regulatory networks and cell fate specification.

Expertise in zebrafish development, gastrulation, morphogen signaling and morphogenesis.

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

Learn more at Review Commons

Referee #2

Evidence, reproducibility and clarity

The authors of the study introduce the Palette method, a novel approach designed to infer spatial gene expression patterns from bulk RNA-sequencing (RNA-seq) data. This method is complemented by the development of the DreSTEP 3D spatial gene expression atlas of zebrafish embryos, establishing a comprehensive resource for visualizing gene expression and investigating spatial cell-cell interactions in developmental biology.

Major concerns:

1. The efficacy of the Palette method may be compromised by its dependency on the quality of the reference spatial transcriptomics data. As highlighted in the study, variations in data quality can lead to significant challenges in reconstructing accurate spatial expression patterns from bulk data. This underscores the necessity of evaluating quality parameters, such as the number of gene detections and spatial resolution, to ensure reliable outcomes. Additional studies should rigorously assess how these quality factors influence the accuracy and efficiency of the algorithm in various data contexts, particularly under diverse conditions of gene detection.
2. The methodology raises pertinent questions regarding how the clustering results from different algorithms may affect the reconstructions by the Palette method. The authors would better provide a detailed discussion/comparison of clustering processes that optimize the reconstruction of spatial patterns, ensuring precision in the downstream analyses.
3. The choice to utilize only highly expressed genes in the initial stages of the Palette algorithm also warrants further exploration. Addressing the criteria for determining which genes qualify as "highly expressed" and outlining robust cutoff will enhance the algorithm's rigor and applicability. Similarly, in the iterative estimation of gene expression across spatial spots, establishing optimal iteration conditions is crucial. Implementing a loss function may offer a systematic method for concluding iterations, thus refining computational efficiency.
4. Performance metrics relating to processing speed and computational demands remain inadequately addressed in the current framework. Understanding how the Palette method scales across varying gene counts and bulk RNA-seq datasets will be essential for potential applications in larger biological contexts. Notably, the quantitative demands of analyzing 20,000 genes when processing 10, 100, or 1,000 bulk RNA profiles must be articulated to guide researchers in planning accordingly.

Minor opinions:

1. Despite the promising advances offered by the zebrafish 3D reconstruction, there is a lack of details regarding numbers of the spatial transcriptomics (ST) data utilized, and the number of bulk RNA-seq data employed in the analyses. These parameters need to be clarified.
2. Issues regarding spatial cell-cell communication, especially concerning interactions over longer distances, necessitate careful consideration. Introducing spatial distance constraints could help formulate more realistic models of cellular interactions, a vital aspect of embryonic development.
3. Evaluation metrics such as the Adjusted Rand Index (ARI) and Root Mean Square Error (RMSE) represent critical tools for systematically measuring the similarity of inferred spatial patterns, yet their specific application within this context should be elaborated.
4. The study's limitations surrounding ST data quality cannot be overstated. Discussing scenarios where only limited or poor-quality ST data are available will be crucial for guiding future studies. Furthermore, a clear explanation of how enhanced specificity and accuracy translate into tangible biological insights is essential for demystifying the underlying mechanisms driving developmental processes.

Significance

The Palette pipeline demonstrates a marked improvement in specificity and accuracy when predicting spatial gene expression patterns. Evaluative studies on Drosophila and zebrafish datasets affirm its enhanced performance compared to existing methodologies. By effectively reconstructing spatial information from bulk transcriptomic data, the Palette method innovatively merges the philosophy of leveraging single-cell transcriptomic data for deconvolution analyses. This integration is pivotal, advancing traditional bulk RNA-seq approaches while laying the groundwork for future research.

One of the notable achievements in this work is the construction of the DreSTEP atlas, which integrates serial bulk RNA-seq data with advanced 3D imaging techniques. This resource grants researchers unprecedented access to the visualization of gene expression patterns across the zebrafish embryo, facilitating the investigation of spatial relationships and cell-cell interactions critical for developmental processes. Such capabilities are invaluable for understanding the intricate dynamics of embryogenesis and the distinct roles of individual cell types.

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

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

Summary:

The manuscript titled "Unravelling the Progression of the Zebrafish Primary Body Axis with Reconstructed Spatiotemporal Transcriptomics" presents a comprehensive analysis of the development of the primary body axis in zebrafish by integrating bulk RNA-seq, 3D images, and Stereo-Seq. The authors first clearly demonstrate the application of Palette for integrating RNA-seq and Stereo-Seq using published spatial transcriptomics data of Drosophila embryos. Subsequently, they produced serial bulk RNA-seq data for certain developmental stages of Danio rerio embryos and utilized published Stereo-Seq data. Through robust validation, the authors observe the molecular network involved in AP axis formation. While the authors show that integrating bulk RNA-seq data with Stereo-Seq improves spatial resolution, additional proof is required to demonstrate the extent of this improvement.

Major Comments:

1. Lines 66-68: Discuss the limitations of existing tools and explicitly state the advantages of using Palette.
2. Body Pattern Genes Analysis: For both Drosophila and Danio rerio, it would be valuable to examine body pattern genes in Stereo-Seq and apply Palette to determine if the resolution of the segments improves or merges. The resolution of the A-P axis is convincing, but further evidence for other segments would be beneficial.
3. Figure 2d: Include the A-P line for which the intensity profile was plotted in the main figure, rather than just in the supplementary material. Additionally, consider simplifying the plot by not combining three lines into one, as it complicates the interpretation of observations.
4. Drosophila Data Analysis: While the alignment and validation of Danio rerio sections are clearly explained, the analysis and validation of Drosophila data are insufficiently detailed. Provide a more thorough explanation of how the intensity profiles between BDGP in situ data and Stereo-Seq data are adjusted.
5. Figure 3d: Present a plot with the expected expression profiles of the three genes if the embryo is aligned as anticipated.
6. Analysis Without Palette: Between lines 277-438, the outcome of using Palette with bulk RNA-seq and Stereo-Seq is convincing. However, consider the following:<br /> o What would be the observations if the analysis were conducted solely with Stereo-Seq data, without incorporating bulk RNA-seq data and employing Palette?<br /> o This study uses only Stereo-Seq as the spatial transcriptomics reference. It would strengthen the argument to use at least one other spatial transcriptomics method, such as Visium or MERFISH, in conjunction with bulk RNA-seq and Palette, to demonstrate whether Palette consistently improves gene expression resolution.
7. PDAC Data Analysis: Provide a more detailed explanation of the PDAC data analysis and use appropriate colors in the tissue images to clearly distinguish cell types.
8. Comparison with Other Methods: State the limitations of not using STitch3D and Spateo for alignment and explain why these methods were not employed.

Minor Comments:

1. References: Add references to the statements in lines 51-53.
2. Scientific Name Consistency: Ensure consistency in using either "Danio rerio" or "zebrafish" throughout the manuscript.
3. Related References: Include the following relevant references:
4. https://academic.oup.com/bib/article/25/4/bbae316/7705532
5. https://www.life-science-alliance.org/content/6/1/e202201701
6. Figure 1a: In the Venn diagram, include the number of genes in the bulk and Stereo-Seq datasets, as well as the number of overlapping genes.
7. Figure 1 Improvement: Enlarge Figure 1 and reduce repetitive elements, such as parts of the deconvolution and Figure 1b.
8. Figure 3f: Explain the black discontinuous line in the plot.
9. Line 610: State the percentage of unpaired imaging spots.
10. Lines 616-618: Specify the unit for the spot diameter.

Significance

This algorithm will be useful not only for the field of developmental biology but also for wider applications in spatial omics. Although I have expertise in spatial omics technology development, my understanding of computational biology is limited, which restricts my ability to fully evaluate the Palette algorithm presented in this paper.

Reviewer #1 (Public review):

Summary:

Adult (4mo) rats were tasked to either press one lever for an immediate reward or another for a delayed reward. The task had an adjusting amount structure in which (1) the number of pellets provided on the immediate reward lever changed as a function of the decisions made, (2) rats were prevented from pressing the same lever three times in a row.

While the authors have been very responsive to the reviews, and I appreciate that, unfortunately, the new analyses reported in this revision actually lead me to deeper concerns about the adequacy of the data to support the conclusions. In this revision, it has become clear that the conclusions are forced and not supported by the data. Alternative theories are not considered or presented. This revision has revealed deep problems with the task, the analyses, and the modeling.

Data Weaknesses

Most importantly, the inclusion of the task behavior data has revealed a deep problem with the entire structure of the data. As is obvious in Figure 1D, there is a slow learning effect that is changing over the sessions as the animals learn to stop taking the delayed outcome. Unfortunately, the 8s delays came *after* the 4s. The first 20 sessions contain 19 4s delays and 1 8s delay, while the last 20 sessions contain 14 8s delays and 6 4s delays. Given the changes across sessions, it is likely that a large part of the difference is due to across-session learning (which is never addressed or considered).

These data are not shown by subject and I suspect that individual subjects did all 4s then all 8s and some subjects switched tasks at different times. If my suspicion is true, then any comparisons between the 4s and 8s conditions (which are a major part of the author's claims) may have nothing to do with the delays, but rather with increased experience on the task.

Furthermore, the four "groups", which are still poorly defined, seem to have been assessed at a session-by-session level. So when did each animal fall into a given group? Why is Figure 1D not showing which session fell into which group and why are we not seeing each animal's progression? They also admit that animals used a mixture of strategies, which implies that the "group" assignment is an invalid analysis, as the groups do not accommodate strategy mixing.

Figure 2 shows that none of the differences of the group behavior against random choice with a basic p(delay) are significant. The use a KS test to measure these differences. KS tests are notoriously sensitive as KS tests simply measure whether there are any statistical differences between two distributions. They do not report the full statistics for Figure 2, but only say that the 4HI group was not significant (KS p-value = 0.72) and the 8LO showed a p-value of 0.1 (which they interpret as significant). p=0.1 is not significant. They don't report the value of the 4LO or 8HI groups (why not?), but say they are in-between these two extremes. That means *none* of the differences are significant.

They then test a model with additional parameters, and say that the model includes more than the minimal p_D parameter, but never report BIC or AIC model comparisons. In order to claim that the model is better than the bare p_D assumption, they should be reporting model-comparison statistics. But given that the p_D parameters are enough (q.v. Figure 2), this entire model seems unnecessary

It took me a while to determine what was being shown in Figure 3, but I was eventually able to determine that 0 was the time after the animal made the choice to wait out the delay side, so the 4s in Figure 3A1 with high power in the low-frequency (<5 Hz) range is the waiting time. They don't show the full 8s time. Nor do they show the spectrograms separated by group (assuming that group is the analytical tool they are using). In B they show only show theta power, but it is unclear how to interpret these changes over time.

In Figure 4, panel A is mostly useless because it is just five sample sessions showing firing rate plotted on the same panels as the immediate reward amount. If they want to claim correlation, they should show and test it. But moreover, this is not how neural data should be presented - we need to know what the cells are doing, population-wise. We need to have an understanding of the neural ensemble. These data are clearly being picked and chosen, which is not OK.

Figure 4, panels B and C show that the activity trivially reflects the reward that has been delivered to the animal, if I am understanding the graphs correctly. (The authors do not interpret it this way, but the data is, to my eyes, clear.) The "immediate" signal shows up immediately at choice and reflects the size of the immediate reward (which is varying). The "delay" signal shows up after the delay and does not, which makes sense as the animals get 6 pellets on the delayed side no matter what. In fact, the max delayed side activity = the max immediate side activity, which is 6 pellets. This is just reward-related firing.

Figure 5 is poorly laid out, switching the order in 5C to be 2 1 3 in E and F. (Why?!) The statistics for Figure 5 on page 17 should be asking whether there are differences between neuron types, not whether there is a choice x time interaction in a given neuron type. When I look at Figure 5F1-3, all three types look effectively similar with different levels of noise. It is unclear why they are doing this complicated PC analysis or what we should be drawing from it.

Figure 6 mis-states pie charts as "total number" rather than proportions.

Interpretation Weaknesses

The separation of cognitive effort into "resource-based" and "resistance-based" seems artificial to me. I still do not understand why the ability to resist a choice does not also depend on resource or why using resources are not a form of resistance. Doesn't every action in the end depend on the resources one has available? And doesn't every use of a resource resist one option by taking another? Even if one buys these two separate cognitive control processes (which at this point in reading the revision, I do not), the paper starts from the assumption that a baseline probability of waiting out the delays is a "resistance-based cognitive control" (why?) and a probability of choice that takes into account the size of the immediate value (confusingly abbreviated as ival) is a "resource-based cognitive control" (again, why?)

Reviewer #2 (Public review):

Summary:

I appreciate the considerable work the authors have done on the revision. The manuscript is markedly improved.

Strengths still include the strong theoretical basis, well-done experiments, and clear links to LFP / spectral analyses that have links to human data. The task is now more clearly explained, and the neural correlates better articulated.

Weaknesses:

I had remaining questions, many related to my previous questions.<br /> (1) The results have some complexity, but I still had questions about which is resource and which is resistance based. The authors say in the last sentence of the discussion: "Prominent pre-choice theta power was associated with a behavioral strategy characterized by a strong bias towards a resistance-based strategy, whereas the neural signature of ival-tracking was associated with a strong bias towards a resource-based strategy.".<br /> I might suggest making this simpler and clear in the abstract and the first paragraph of the discussion. A simple statement like 'pre-choice theta was biased towards resistance whereas single neurons were biased towards resources" might make this idea come across?

(2) I think most readers would like to see raw single trial LFP traces in Figure 3, single unit rasters in Figure 4, and spike-field records in Figure 5.

(3) What limitations are there to this work? I wonder if readers might benefit from some contextualization - the sample size, heterogenous behavior - lack of cell-type specificity - using PC3 to define spectral relationships - I might suggest pointing these out.

(4) I still wasn't sure what 4 Hz vs. theta 6-12 Hz meant - is it all based on PC3's pos/neg correlation? I wonder if showing a scatter plot with the y-axis being PC3 and the x-axis being theta 4 Hz power would help distinguish these? Is this the first time this sort of analysis has been done? If so, it requires clearer definitions.

Reviewer #3 (Public review):

Summary:

The study investigated decision making in rats choosing between small immediate rewards and larger delayed rewards, in a task design where the size of the immediate rewards decreased when this option was chosen and increased when it was not chosen. The authors conceptualise this task as involving two different types of cognitive effort; 'resistance-based' effort putatively needed to resist the smaller immediate reward, and 'resource-based' effort needed to track the changing value of the immediate reward option. They argue based on analyses of the behaviour, and computational modelling, that rats use different strategies in different sessions, with one strategy in which they preferentially choose the delayed reward option irrespective of the current immediate reward size, and another strategy in which they preferentially choose the immediate reward option when the immediate reward size is large, and the delayed reward option when the immediate reward size is small. The authors recorded neural activity in anterior cingulate cortex. They propose that oscillatory activity in the 6-12Hz theta band occurs when subjects use a 'resistance-based' strategy of choosing the delayed option irrespective of the current value of the immediate reward option. They also examine neural representation of the current value of the immediate reward option, and suggest that this value is more strongly represented when subjects are using this value information to guide choice. They further argue that neurons whose activity is modulated by theta oscillations are less involved in tracking the value of the immediate reward option than neurons whose activity is not theta modulated. If solid, these findings will be of interest to researchers working on cognitive control and ACCs involvement in decision making. However, there are some issues with the modelling and analysis which preclude high confidence in the validity of the conclusions.

Strengths:

The behavioural task used is interesting and the recording methods used (64 channel silicon probes) should enable the collection of good quality single unit and LFP electrophysiology data. The authors recorded from a sizable sample of subjects for this type of study. The approach of splitting the data into sessions where subjects used different strategies and then examining the neural correlates of each is in principle interesting, though I have some reservations about the strength of evidence for the existence of multiple strategies.

Limitations:

The dataset is unbalanced in terms of both the number of sessions contributed by each subject, and their distribution across the different putative behavioural strategies (see Table 1), with some subjects contributing 7 sessions to a given strategy and others 0. Further, only 2 of 10 subjects contribute any sessions to one of the behavioural strategies (8LO), and a single subject contributes >50% of the sessions (7 of 13) sessions to another strategy (8HI). Apparent differences in brain activity between the strategies could therefore in fact reflect differences between subjects, which could arise due to e.g. differences in electrode placement. To make firm conclusions that neural activity is different in sessions where different strategies are thought to be employed, it would be necessary to account for potential cross-subject variation in the data. The current statistical methods don't appear to do this as they use within subject measures (e.g. trials or neurons) as the experimental unit and ignore which subject the neuron/trial came from.

The starting point for the analysis was the splitting of sessions into 4 groups based on the duration of the delay (4 vs 8 seconds) and then clustering within each delay category into two sub-groups. It was not clear why 2 clusters per delay category were used, nor whether the data did in fact have a clear split into two distinct clusters or continuous variation across the population of sessions. The simplified RL model used in the revised manuscript (which is an improvement from that used in the previous version) could in principle help to quantify variation across the populations of sessions, by using model fitting and comparison methods to evaluate variation in strategy across subjects. However, as far as I could tell no model-fitting or comparison was performed, and the only attempt to link the model to data was by simulating data using a fixed probability of choosing the delayed lever (i.e. with no learning across trials) and comparing the distribution of total rewards obtained per session with that of the subjects in each group (Figure 2). Total reward per session is a very coarse behavioural metric and using likelihood-based methods to fit model parameters to subjects trial-by-trial choice data would provide a more sensitive way of using the modelling to assess behavioural strategy across sessions.

Conceptually, it is not obvious that choices towards the delayed vs immediate lever reflect use of different strategies employing different types of cognitive effort. Rather these could reflect a single strategy which compares the estimated value of the two levers, with differences in behaviour between sessions accounted for either by differences in the task itself (between the 8s and 4s delay condition) or differences in the parameters of the strategy, such as the strength of temporal discounting.

Even if one accepts the claim that the task recruits two distinct types of cognitive control, the argument that theta oscillations, which occur on delay choice trials in the 4s delay condition, are a correlate of a 'resistance-based' strategy (resisting the immediate reward), is hard to reconcile with the fact that theta oscillations do not occur on delay choice trials in the 8s delay condition (Figure 3). The authors note this discrepancy, but state that 'The reason was because these groups largely avoided the delayed lever (Figure 1) and thereby abandoned the need to implement resistance-based control altogether.' However, the data in Figure 1D show that even in the 8s condition the subjects choose the delayed lever on around 50% of trials. It is not obvious why choosing the delayed lever on 50% of trials in the 8s condition does not require 'resistance-based' cognitive effort, while choosing it in the 4s delay condition does.

The other main claims regarding the neural data are that the neuronal representation of the value of the immediate reward lever (ival) is stronger in sessions where subjects are choosing that lever more often, particularly the 8LO group, and that neurons whose activity tracks ival are a different population from neurons whose activity is theta modulated. However, the analysis methods used to make these claims are rather convoluted and make it hard to assess the strength of the evidence for them.

To evaluate the strength of ival representation in neural activity, the authors first fit a regression model predicting each neuron's activity at different timepoints as a function of behavioural variables including ival, which is a sensible first step. However, they then perform clustering on the regression coefficients and then plot neural activity only for the cluster which they state 'provided the clearest example of value tracking'. It is not clear how the clustering was done, whether there were in fact well defined clusters in the neural activity, how the clusters whose activity is plotted were chosen, nor the proportion of neurons in this cluster for each group of sessions. The analysis therefore provides only limited information about the strength of ival representation in different session groups. It would be useful to quantify the variance explained by ival in neural activity for each group of sessions using a simpler quantification of the regression analysis, such as cross-validated coefficient of partial determination.

The analysis of how theta modulation related to representation of ival across neurons was also complicated and non-standard. To determine whether individual neurons were theta modulated, the authors did PCA on a matrix comprised of spike train autocorrelations for individual neurons, and then grouped neurons according to the projection of their autocorrelation function onto the 3rd Principal Component, on the basis that neurons with negative projection onto this component showed a peak roughly at theta frequency in the power spectrum of their autocorrelation. Even ignoring the fact that the peak in the power spectrum is broad and centred above the standard theta frequency (see figure 5B3), this is an arbitrary and unnecessarily complex way to determine if neurons are theta modulated. It would be much simpler and greatly preferable to either directly assess the modulation depth of individual neurons spike train autocorrelation in the theta band, or to use a metric of spike-LFP coupling in the theta band instead. The authors do include some analysis of spike field coherence in Figure 6 and this is a much more sensible approach. However, it is worth noting that the only session group which shows a difference in coherence at theta frequency relative to the other groups is 8LO, to which only 2 of 8 animals contribute any data and 70% of sessions come from one animal. It is therefore unclear whether differences in this group are due to differences in behavioural strategy, or reflect other sources of cross-animal variation.

Author response:

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

We would like to thank the reviewers for their efforts and feedback on our preprint. We have elected to rework the manuscript for publication in a different journal. In this process we will alter many of the approaches and re-evaluate the conclusions. With this, many of the points raised by the reviewers will be no longer relevant and therefore do not require a response. Again, we thank the reviewers for their time and helpful feedback.

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

eLife Assessment:

The authors present a potentially useful approach of broad interest arguing that anterior cingulate cortex (ACC) tracks option values in decisions involving delayed rewards. The authors introduce the idea of a resource-based cognitive effort signal in ACC ensembles and link ACC theta oscillations to a resistance-based strategy. The evidence supporting these new ideas is incomplete and would benefit from additional detail and more rigorous analyses and computational methods.

We are extremely grateful for the several excellent and comments of the reviewers. To address these concerns, we have completely reworked the manuscript adding more rigorous approaches in each phase of the analysis and computational model. We realize that this has taken some time to prepare the revision. However, given the comments of the reviewers, we felt it necessary to thoroughly rework the paper based on their input. Here is a (nonexhaustive) overview of the major changes we made:

We have developed a way to more adequately capture the heterogeneity in the behavior

We have completely reworked the RL model

We have added additional approaches and rigor to the analysis of the value-tracking signal. 

Reviewer #1 (Public Review):

Summary:

Young (2.5 mo [adolescent]) rats were tasked to either press one lever for immediate reward or another for delayed reward. 

Please note that at the time of testing and training that the rats were > 4 months old. 

The task had a complex structure in which (1) the number of pellets provided on the immediate reward lever changed as a function of the decisions made, (2) rats were prevented from pressing the same lever three times in a row. Importantly, this task is very different from most intertemporal choice tasks which adjust delay (to the delayed lever), whereas this task held the delay constant and adjusted the number of 20 mg sucrose pellets provided on the immediate value lever.

Several studies parametrically vary the immediate lever (PMID: 39119916, 31654652, 28000083, 26779747, 12270518, 19389183). While most versions of the task will yield qualitatively similar estimates of discounting, the adjusting amount is preferred as it provides the most consistent estimates (PMID: 22445576). More specifically this version of the task avoids contrast effects of that result from changing the delay during the session (PMID: 23963529, 24780379, 19730365, 35661751) which complicates value estimates. 

Analyses are based on separating sessions into groups, but group membership includes arbitrary requirements and many sessions have been dropped from the analyses. 

We have updated this approach and now provide a more comprehensive assessment of the behavior. The updated approach applies a hierarchical clustering model to the behavior in each session. This was applied at each delay to separate animals that prefer the immediate option more/less. This results in 4 statistically dissociable groups (4LO, 4HI, 8LO, 8HI) and includes all sessions. Please see Figure 1. 

Computational modeling is based on an overly simple reinforcement learning model, as evidenced by fit parameters pegging to the extremes. 

We have completely reworked the simulations in the revision. In the updated RL model we carefully add parameters to determine which are necessary to explain the experimental data. We feel that it is simplified yet more descriptive. Please see Figure 2 and associated text. 

The neural analysis is overly complex and does not contain the necessary statistics to assess the validity of their claims.

We have dramatically streamlined the spike train analysis approach and added several statistical tests to ensure the rigor of our results. Please see Figures 4,5,6 and associated text. 

Strengths:

The task is interesting.

Thank you for the positive comment

Weaknesses:

Behavior:

The basic behavioral results from this task are not presented. For example, "each recording session consisted of 40 choice trials or 45 minutes". What was the distribution of choices over sessions? Did that change between rats? Did that change between delays? Were there any sequence effects? (I recommend looking at reaction times.) Were there any effects of pressing a lever twice vs after a forced trial? 

Please see the updated statistics and panels in Figures 1 and 2. We believe these address this valid concern.  

This task has a very complicated sequential structure that I think I would be hard pressed to follow if I were performing this task. 

Human tasks implement a similar task structure (PMID: 26779747). Please note the response above that outlines the benefits of using of this task.   

Before diving into the complex analyses assuming reinforcement learning paradigms or cognitive control, I would have liked to have understood the basic behaviors the rats were taking. For example, what was the typical rate of lever pressing? If the rats are pressing 40 times in 45 minutes, does waiting 8s make a large difference?

Thank you for this suggestion. Our additions to Figure 1 are intended to better explain and quantify the behavior of the animals. Note that this task is designed to hold the rate of reinforcement constant no matter the choices of the animals. Our analysis supports the long-held view in the literature that rats do not like waiting for rewards, even at small delays. Going from the 4 à 8 sec delay results in significantly more immediate choices, indicating that the rats will forgo waiting 8 sec for a larger reinforcer and take a smaller reinforcer at 4 sec.  

For that matter, the reaction time from lever appearance to lever pressing would be very interesting (and important). Are they making a choice as soon as the levers appear? Are they leaning towards the delay side, but then give in and choose the immediate lever? What are the reaction time hazard distributions?

This is an excellent suggestion, we have added a brief analysis of reaction times (Please see the section entitled “4 behavioral groups are observed across all sessions” in the Results). Please note that an analysis of the reaction times has been presented in a prior analysis of this data set (White et al., 2024). In addition, an analysis of reaction times in this task was performed in Linsenbardt et al. (2017). In short, animals tend to choose within 1 second of the lever appearing. In addition, our prior work shows that responses on the immediate lever tend to be slower, which we viewed as evidence of increased deliberation requirements (possibly required to integrate value signals).   

It is not clear that the animals on this task were actually using cognitive control strategies on this task. One cannot assume from the task that cognitive control is key. The authors only consider a very limited number of potential behaviors (an overly simple RL model). On this task, there are a lot of potential behavioral strategies: "win-stay/lose-shift", "perseveration", "alternation", even "random choices" should be considered.

The strategies the Reviewer mentioned are descriptors of the actual choices the rats made. For example, perseveration means the rat is choosing one of the levers at an excessively high rate whereas alternation means it is choosing the two levers more or less equally, independent of payouts. But the question we are interested in is why? We are arguing that the type of cognitive control determines the choice behavior, but cognitive control is an internal variable that guides behavior, rather than simply a descriptor of the behavior. For example, the animal opts to perseverate on the delayed lever because the cognitive control required to track ival is too high. We then searched the neural data for signatures of the two types of cognitive control.

The delay lever was assigned to the "non-preferred side". How did side bias affect the decisions made?

The side bias clearly does not impact performance as the animals prefer the delay lever at shorter delays, which works against this bias.  

The analyses based on "group" are unjustified. The authors compare the proportion of delayed to immediate lever press choices on the non-forced trials and then did k-means clustering on this distribution. But the distribution itself was not shown, so it is unclear whether the "groups" were actually different. They used k=3, but do not describe how this arbitrary number was chosen. (Is 3 the optimal number of clusters to describe this distribution?) Moreover, they removed three group 1 sessions with an 8s delay and two group 2 sessions with a 4s delay, making all the group 1 sessions 4s delay sessions and all group 2 sessions 8s delay sessions. They then ignore group 3 completely. These analyses seem arbitrary and unnecessarily complex. I think they need to analyze the data by delay. (How do rats handle 4s delay sessions? How do rats handle 6s delay sessions? How do rats handle 8s delay sessions?). If they decide to analyze the data by strategy, then they should identify specific strategies, model those strategies, and do model comparison to identify the best explanatory strategy. Importantly, the groups were session-based, not rat based, suggesting that rats used different strategies based on the delay to the delayed lever.

We have completely reworked our approach for capturing the heterogeneity in behavior. We have taken care to show more of the behavioral statistics that have gone into identifying each of the groups. All sessions are included in this analysis. As the reviewer suggests, we used the statistics from each of the behavioral groups to inform the RL model that explores neural signals that underly decisions in this task. We strongly disagree that groups should be rat and not session based as the behavior of the animal can, and does, change from day to day. This is important to consider when analyzing the neural data as rat-based groupings would ignore this potential source of variance. 

The reinforcement learning model used was overly simple. In particular, the RL model assumes that the subjects understand the task structure, but we know that even humans have trouble following complex task structures. Moreover, we know that rodent decision-making depends on much more complex strategies (model-based decisions, multi-state decisions, rate-based decisions, etc). There are lots of other ways to encode these decision variables, such as softmax with an inverse temperature rather than epsilon-greedy. The RL model was stated as a given and not justified. As one critical example, the RL model fit to the data assumed a constant exponential discounting function, but it is well-established that all animals, including rodents, use hyperbolic discounting in intertemporal choice tasks. Presumably this changes dramatically the effect of 4s and 8s. As evidence that the RL model is incomplete, the parameters found for the two groups were extreme. (Alpha=1 implies no history and only reacting to the most recent event. Epsilon=0.4 in an epsilongreedy algorithm is a 40% chance of responding randomly.)

While we agree that the approach was not fully justified, we do not agree that it was invalid. Simply stated, a softmax approach gives the best fit to the choice behavior, whereas our epsilon-greedy approach attempted to reproduce the choice behavior using a naïve agent that progressively learns the values of the two levers on a choice-by-choice basis. Nevertheless, we certainly appreciate that important insights can be gained by fitting a model to the data as suggested. We feel that the new modeling approach we have now implemented is optimal for the present purposes and it replaces the one used in the original manuscript.

The authors do add a "dbias" (which is a preference for the delayed lever) term to the RL model, but note that it has to be maximal in the 4s condition to reproduce group 2 behavior, which means they are not doing reinforcement learning anymore, just choosing the delayed lever.

The dbias term was dropped in the new model implementation

Neurophysiology:

The neurophysiology figures are unclear and mostly uninterpretable; they do not show variability, statistics or conclusive results.

While the reviewer is justified in criticizing the clarity of the figures, the statement that “they do not show variability, statistics or conclusive results” is not correct. Each of the figures presented in the first draft of the manuscript, except Figure 3, are accompanied by statistics and measures of variability. Nonetheless we have updated each of the neurophysiology analyses. We hope that the reviewer will find our updates more rigorous and thorough.   

As with the behavior, I would have liked to have seen more traditional neurophysiological analyses first. What do the cells respond to? How do the manifolds change aligned to the lever presses? Are those different between lever presses?

We have added several figures that plot the mean +/- SEM of the neural activity (see Figures 4 and 5). Hopefully this provides a more intuitive picture of the changes in neural activity throughout the task.  

Are there changes in cellular information (both at the individual and ensemble level) over time in the session? 

We provide several analyses of how firing rate changes over trials in relation to ival over time and trials in the session. In addition, we describe how these signals change in each of the behavioral groups. 

How do cellular responses differ during that delay while both levers are out, but the rats are not choosing the immediate lever?

We were somewhat unclear about this suggestion as the delay follows the lever press. In addition, there is no delay after immediate presses 

Figure 3, for example, claims that some of the principal components tracked the number of pellets on the immediate lever ("ival"), but they are just two curves. No statistics, controls, or justification for this is shown. BTW, on Figure 3, what is the event at 200s?

This comment is no longer relevant based on the changes we’ve made to the manuscript. 

I'm confused. On Figure 4, the number of trials seems to go up to 50, but in the methods, they say that rats received 40 trials or 45 minutes of experience.

This comment is no longer relevant based on the changes we’ve made to the manuscript. 

At the end of page 14, the authors state that the strength of the correlation did not differ by group and that this was "predicted" by the RL modeling, but this statement is nonsensical, given that the RL modeling did not fit the data well, depended on extreme values. Moreover, this claim is dependent on "not statistically detectable", which is, of course, not interpretable as "not different".

This comment is no longer relevant based on the changes we’ve made to the manuscript. 

There is an interesting result on page 16 that the increases in theta power were observed before a delayed lever press but not an immediate lever press, and then that the theta power declined after an immediate lever press. 

Thank you for the positive comment. 

These data are separated by session group (again group 1 is a subset of the 4s sessions, group 2 is a subset of the 8s sessions, and group 3 is ignored). I would much rather see these data analyzed by delay itself or by some sort of strategy fit across delays.

Thank you for the excellent suggestion. Our new group assignments take delay into account. 

That being said, I don't see how this description shows up in Figure 6. What does Figure 6 look like if you just separate the sessions by delay?

We are unclear what the reviewer means by “this description”.  

Discussion:

Finally, it is unclear to what extent this task actually gets at the questions originally laid out in the goals and returned to in the discussion. The idea of cognitive effort is interesting, but there is no data presented that this task is cognitive at all. The idea of a resourced cognitive effort and a resistance cognitive effort is interesting, but presumably the way one overcomes resistance is through resourcelimited components, so it is unclear that these two cognitive effort strategies are different.

The basis for the reviewers assertation that “the way one overcomes resistance is through resourcelimited components” is not clear. In the revised version, we have taken greater care to outline how each type of effort signal facilitates performance of the task and articulate these possibilities in our stochastic and RL models. We view the strong evidence for ival tracking presented herein as a critical component of resource based cognitive effort. 

The authors state that "ival-tracking" (neurons and ensembles that presumably track the number of pellets being delivered on the immediate lever - a fancy name for "expectations") "taps into a resourced-based form of cognitive effort", but no evidence is actually provided that keeping track of the expectation of reward on the immediate lever depends on attention or mnemonic resources. They also state that a "dLP-biased strategy" (waiting out the delay) is a "resistance-based form of cognitive effort" but no evidence is made that going to the delayed side takes effort.

We challenge the reviewers that assertation ival tracking is a “fancy name for expectations”. We make no claim about the prospective or retrospective nature of the signal. Clearly, expectations should be prospective and therefore different from ival tracking. Regarding the resistance signal: First, animals avoid the delay lever more often at the 8 sec delay (Figure 1). We have shown that increasing the delay systematically biases responses AWAY from the delay (Linsenbardt et al., 2017). This is consistent with a well-developed literature that rats and mice do not like waiting for delayed reinforcers. We contend that enduring something you don’t like takes effort. 

The authors talk about theta synchrony, but never actually measure theta synchrony, particularly across structures such as amygdala or ventral hippocampus. The authors try to connect this to "the unpleasantness of the delay", but provide no measures of pleasantness or unpleasantness. They have no evidence that waiting out an 8s delay is unpleasant.

We have added spike-field coherence to better contact the literature on synchrony. Note that we never refer to our results as “synchrony”. However, we would be remiss to not address the growing literature on theta synchrony in effort allocation. There is a well-developed literature that rats and mice do not like waiting for delayed reinforcers. If waiting out the delay was not pleasant then why do the animals forgo larger rewards to avoid it? 

The authors hypothesize that the "ival-tracking signal" (the expectation of number of pellets on the immediate lever) "could simply reflect the emotional or autonomic response". Aside from the fact that no evidence for this is provided, if this were to be true, then, in what sense would any of these signals be related to cognitive control?

This is proposed as an alternative explanation to the ival signal in the discussion. It was added as our due diligence. Emotional state could provide feedback to the currently implemented control mechanism. If waiting for reinforcement is too unpleasant this could drive them to ival tracking and choosing the immediate option more frequently. We provide this option only as a possibility, not a conclusion. We have clarified this in the revised text. Nevertheless, based on our review of the literature, autonomic tracking in some form, seems to be the most likely function of ACC (Seamans & Floresco 2022). While the reviewer may disagree with this, we feel it is at least as valid as all the complex, cognitively-based interpretations that commonly appear in the literature.

Reviewer #2 (Public Review):

Summary:

This manuscript explores the neuronal signals that underlie resistance vs resource-based models of cognitive effort. The authors use a delayed discounting task and computational models to explore these ideas. The authors find that the ACC strongly tracks value and time, which is consistent with prior work. Novel contributions include quantification of a resource-based control signal among ACC ensembles, and linking ACC theta oscillations to a resistance-based strategy.

Strengths:

The experiments and analyses are well done and have the potential to generate an elegant explanatory framework for ACC neuronal activity. The inclusion of local-field potential / spike-field analyses is particularly important because these can be measured in humans.

Thank you for the endorsement of our work.

Weaknesses:

I had questions that might help me understand the task and details of neuronal analyses.

(1) The abstract, discussion, and introduction set up an opposition between resource and resistancebased forms of cognitive effort. It's clear that the authors find evidence for each (ACC ensembles = resource, theta=resistance?) but I'm not sure where the data fall on this dichotomy.

(a) An overall very simple schematic early in the paper (prior to the MCML model? or even the behavior) may help illustrate the main point.

(b) In the intro, results, and discussion, it may help to relate each point to this dichotomy.

(c) What would resource-based signals look like? What would resistance based signals look like? Is the main point that resistance-based strategies dominate when delays are short, but resource-based strategies dominate when delays are long?

(d) I wonder if these strategies can be illustrated? Could these two measures (dLP vs ival tracking) be plotted on separate axes or extremes, and behavior, neuronal data, LFP, and spectral relationships be shown on these axes? I think Figure 2 is working towards this. Could these be shown for each delay length? This way, as the evidence from behavior, model, single neurons, ensembles, and theta is presented, it can be related to this framework, and the reader can organize the findings.

These are excellent suggestions, and we have implemented them, where possible. 

(2) The task is not clear to me.

(a) I wonder if a task schematic and a flow chart of training would help readers.

Yes, excellent idea, we have now included this in Figure 1. 

(b) This task appears to be relatively new. Has it been used before in rats (Oberlin and Grahame is a mouse study)? Some history / context might help orient readers.

Indeed, this task has been used in rats in several prior studies in rats. Please see the following references (PMID: 39119916, 31654652, 28000083, 26779747, 12270518, 19389183).

(c) How many total sessions were completed with ascending delays? Was there criteria for surgeries? How many total recording sessions per animal (of the 54?)

Please note that the delay does not change within a session. There were no criteria for surgery. 

(d) How many trials completed per session (40 trials OR 45 minutes)? Where are there errors? These details are important for interpreting Figure 1.

Every animal in this data set completed 40 trials and we have updated the task description to clarify this issue. There are no errors in this task, but rather the task is designed to the tendency to make an impulsive choice (smaller reward now). 

(3) Figure 1 is unclear to me.

(a) Delayed vs immediate lever presses are being plotted - but I am not sure what is red, and what is blue. I might suggest plotting each animal.

We have updated Figure 1 considerably for clarity. 

(b) How many animals and sessions go into each data point?

We hope this is clarified now with our new group assignments as all sessions were included in the analysis. 

(c) Table 1 (which might be better referenced in the paper) refers to rats by session. Is it true that some rats (2 and 8) were not analyzed for the bulk of the paper? Some rats appear to switch strategies, and some stay in one strategy. How many neurons come from each rat?

We have updated Table 1 based on our new groupings. The rats that contribute the most sessions also tend to be represented across the behavioral groups therefore it is unlikely that effort allocation strategies across groupings are an esoteric feature of an animal. 

(d) Task basics - RT, choice, accuracy, video stills - might help readers understand what is going into these plots

(e) Does the animal move differently (i.e., RTs) in G1 vs. G2?

Excellent suggestion. We have added more analysis of the task variables in the revision (e.g. RT, choice comparisons across delays, etc…)

(4) I wasn't sure how clustered G1 vs. G2 vs G3 are. To make this argument, the raw data (or some axis of it) might help.

(a) This is particularly important because G3 appears to be a mix of G1 and G2, although upon inspection, I'm not sure how different they really are

(b) Was there some objective clustering criteria that defined the clusters?

(c) Why discuss G3 at all? Can these sessions be removed from analysis?

Based on our updates to the behavioral analysis these comments are no longer relevant. 

(5) The same applies to neuronal analyses in Fig 3 and 4

(a) What does a single neuron peri-event raster look like? I would include several of these.

(b) What does PC1, 2 and 3 look like for G1, G2, and G3?

(c) Certain PCs are selected, but I'm not sure how they were selected - was there a criteria used? How was the correlation between PCA and ival selected? What about PCs that don't correlate with ival?

(d) If the authors are using PCA, then scree plots and PETHs might be useful, as well as comparisons to PCs from time-shuffled / randomized data.

We hope that our reworking of the neural data analysis has clarified these issues. We now include several firing rate examples and aggregate data.   

(6) I had questions about the spectral analysis

(a) Theta has many definitions - why did the authors use 6-12 Hz? Does it come from the hippocampal literature, and is this the best definition of theta? What about other bands (delta - 1-4 Hz), theta (4-7 Hz); and beta - 13- 30 Hz? These bands are of particular importance because they have been associated with errors, dopamine, and are abnormal in schizophrenia and Parkinson's disease.

This designation comes mainly from the hippocampal and ACC literature in rodents. In addition, this range best captured the peak in the power spectrum in our data. Note that we focus our analysis on theta give the literature regarding theta in the ACC as a correlate of cognitive controls (references in manuscript). We did interrogate other bands as a sanity check and the results were mostly limited to theta. Given the scope of our manuscript and the concerns raised regarding complexity we are concerned that adding frequency analyses beyond theta obfuscates the take home message.

However, the spectrograms in Figure 3 show a range of frequencies and highlight the ones in the theta band as the most dynamic prior to the choice. 

(b) Power spectra and time-frequency analyses may justify the authors focus. I would show these (yaxis - frequency, x-axis - time, z-axis, power).

Thank you for the suggestion. We have added this to Figure 3.    

(7) PC3 as an autocorrelation doesn't seem the to be right way to infer theta entrainment or spikefield relationships, as PCA can be vulnerable to phantom oscillations, and coherence can be transient. It is also difficult to compare to traditional measures of phase-locking. Why not simply use spike-field coherence? This is particularly important with reference to the human literature, which the authors invoke.

Excellent suggestion. Note that PCA provided a way to classify neurons that exhibited peaks in the autocorrelation at theta frequencies. We have added spike-field coherence, and this analysis confirms the differences in theta entrainment of the spike trains across the behavioral groups. Please see Figure 6D.   

Reviewer #3 (Public Review):

Summary:

The study investigated decision making in rats choosing between small immediate rewards and larger delayed rewards, in a task design where the size of the immediate rewards decreased when this option was chosen and increased when it was not chosen. The authors conceptualise this task as involving two different types of cognitive effort; 'resistance-based' effort putatively needed to resist the smaller immediate reward, and 'resource-based' effort needed to track the changing value of the immediate reward option. They argue based on analyses of the behaviour, and computational modelling, that rats use different strategies in different sessions, with one strategy in which they consistently choose the delayed reward option irrespective of the current immediate reward size, and another strategy in which they preferentially choose the immediate reward option when the immediate reward size is large, and the delayed reward option when the immediate reward size is small. The authors recorded neural activity in anterior cingulate cortex (ACC) and argue that ACC neurons track the value of the immediate reward option irrespective of the strategy the rats are using. They further argue that the strategy the rats are using modulates their estimated value of the immediate reward option, and that oscillatory activity in the 6-12Hz theta band occurs when subjects use the 'resistancebased' strategy of choosing the delayed option irrespective of the current value of the immediate reward option. If solid, these findings will be of interest to researchers working on cognitive control and ACCs involvement in decision making. However, there are some issues with the experiment design, reporting, modelling and analysis which currently preclude high confidence in the validity of the conclusions.

Strengths:

The behavioural task used is interesting and the recording methods should enable the collection of good quality single unit and LFP electrophysiology data. The authors recorded from a sizable sample of subjects for this type of study. The approach of splitting the data into sessions where subjects used different strategies and then examining the neural correlates of each is in principle interesting, though I have some reservations about the strength of evidence for the existence of multiple strategies.

Thank you for the positive comments. 

Weaknesses:

The dataset is very unbalanced in terms of both the number of sessions contributed by each subject, and their distribution across the different putative behavioural strategies (see table 1), with some subjects contributing 9 or 10 sessions and others only one session, and it is not clear from the text why this is the case. Further, only 3 subjects contribute any sessions to one of the behavioural strategies, while 7 contribute data to the other such that apparent differences in brain activity between the two strategies could in fact reflect differences between subjects, which could arise due to e.g. differences in electrode placement. To firm up the conclusion that neural activity is different in sessions where different strategies are thought to be employed, it would be important to account for potential cross-subject variation in the data. The current statistical methods don't do this as they all assume fixed effects (e.g. using trials or neurons as the experimental unit and ignoring which subject the neuron/trial came from).

In the revised manuscript we have updated the group assignments. We have improved our description of the logic and methods for employing these groupings as well. With this new approach, all sessions are now included in the analysis. The group assignments are made purely on the behavioral statistics of an animal in each session. We feel this approach is preferable to eliminating neurons or session with the goal of balancing them, which may introduce bias. Further, the rats that contribute the most sessions also tend to be represented across the behavioral groups therefore it is unlikely that effort allocation strategies across groupings are an esoteric feature of an animal. As neurons are randomly sampled from each animal on a given session, we feel that we’re justified in treating these as fixed effects.   

It is not obvious that the differences in behaviour between the sessions characterised as using the 'G1' and 'G2' strategies actually imply the use of different strategies, because the behavioural task was different in these sessions, with a shorter wait (4 seconds vs 8 seconds) for the delayed reward in the G1 strategy sessions where the subjects consistently preferred the delayed reward irrespective of the current immediate reward size. Therefore the differences in behaviour could be driven by difference in the task (i.e. external world) rather than a difference in strategy (internal to the subject). It seems plausible that the higher value of the delayed reward option when the delay is shorter could account for the high probability of choosing this option irrespective of the current value of the immediate reward option, without appealing to the subjects using a different strategy.

Further, even if the differences in behaviour do reflect different behavioural strategies, it is not obvious that these correspond to allocation of different types of cognitive effort. For example, subjects' failure to modify their choice probabilities to track the changing value of the immediate reward option might be due simply to valuing the delayed reward option higher, rather than not allocating cognitive effort to tracking immediate option value (indeed this is suggested by the neural data). Conversely, if the rats assign higher value to the delayed reward option in the G1 sessions, it is not obvious that choosing it requires overcoming 'resistance' through cognitive effort.

The RL modelling used to characterise the subject's behavioural strategies made some unusual and arguably implausible assumptions:

Thank you for the feedback, based on these comments (and those above) we have completely reworked the RL model. In addition, we’ve taken care to separate out the variables that correspond to a resistance- versus a resource-based signal. 

There were also some issues with the analyses of neural data which preclude strong confidence in their conclusions:

Figure 4I makes the striking claim that ACC neurons track the value of the immediately rewarding option equally accurately in sessions where two putative behavioural strategies were used, despite the behaviour being insensitive to this variable in the G1 strategy sessions. The analysis quantifies the strength of correlation between a component of the activity extracted using a decoding analysis and the value of the immediate reward option. However, as far as I could see this analysis was not done in a cross-validated manner (i.e. evaluating the correlation strength on test data that was not used for either training the MCML model or selecting which component to use for the correlation). As such, the chance level correlation will certainly be greater than 0, and it is not clear whether the observed correlations are greater than expected by chance.

We have added more rigorous methods to assess the ival tracking signal (Figure 4 and 5). In addition, we’ve dropped the claim that ival tracking is the same across the behavioral groups. We suspect that this was an artifact of a suboptimal group assignment approach in the previous version. 

An additional caveat with the claim that ACC is tracking the value of the immediate reward option is that this value likely correlates with other behavioural variables, notably the current choice and recent choice history, that may be encoded in ACC. Encoding analyses (e.g. using linear regression to predict neural activity from behavioural variables) could allow quantification of the variance in ACC activity uniquely explained by option values after controlling for possible influence of other variables such as choice history (e.g. using a coefficient of partial determination).

We agree that the ival tracking signal may be influenced by other variables – especially ones that are not cognitive but rather more generated by the autonomic system. We have included a discussion of this possibility in the Discussion section. Our previous work has explored the role of choice history on neural activity, please see White et al., (2024). 

Figure 5 argues that there are systematic differences in how ACC neurons represent the value of the immediate option (ival) in the G1 and G2 strategy sessions. This is interesting if true, but it appears possible that the effect is an artefact of the different distribution of option values between the two session types. Specifically, due to the way that ival is updated based on the subjects' choices, in G1 sessions where the subjects are mostly choosing the delayed option, ival will on average be higher than in G2 sessions where they are choosing the immediate option more often. The relative number of high, medium and low ival trials in the G1 and G2 sessions will therefore be different, which could drive systematic differences in the regression fit in the absence of real differences in the activity-value relationship. I have created an ipython notebook illustrating this, available at: https://notebooksharing.space/view/a3c4504aebe7ad3f075aafaabaf93102f2a28f8c189ab9176d48 07cf1565f4e3. To verify that this is not driving the effect it would be important to balance the number of trials at each ival level across sessions (e.g. by subsampling trials) before running the regression.

This is an excellent point and lead us to abandon the linear regression-based approach to quantify differences in ival coding across behavioral groups.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

This paper was extremely hard to read. In addition to the issues raised in the public review (overly complex and incomplete analyses), one of the hardest things to deal with was the writing.

Thank you for the feedback. Hopefully we have addressed this with our thorough rewrite. 

The presentation was extremely hard to follow. I had to read through it several times to figure out what the task was. It wasn't until I got to the RL model Figure 2A that I realized what was really going on with the task. I strongly recommend having an initial figure that lays out the actual task (without any RL or modeling assumptions) and identifies the multiple different kinds of sessions. What is the actual data you have to start with? That was very unclear.

Excellent idea. We have implemented this in Figure 1.  

Labeling session by "group" is very confusing. I think most readers take "group" as the group of subjects, but that's not what you mean at all. You mean some sessions were one way and some were another. (And, as I noted in the public review, you ignore many of the sessions, which I think is not OK.) I think a major rewrite would help a lot. Also, I don't think the group analysis is necessary at all. In the public review, I recommend doing the analyses very differently and more classically.

We have updated the group assignments in a manner that is more intuitive, reflects the delays, and includes all sessions.  

The paper is full of arbitrary abbreviations that are completely unnecessary. Every time I came to "ival", I had to translate that into "number of pellets delivered on the immediate lever" and every time I came to dLP, I had to translate that into "delayed lever press". Making the text shorter does not make the text easier to read. In general, I was taught that unless the abbreviation is the common term (such as "DNA" not "deoxyribonucleic acid"), you should never use an abbreviation. While there are some edge cases (ACC probably over "anterior cingulate cortex"), dLP, iLP, dLPs, iLPs, ival, are definitely way over the "don't do that" line.

We completely agree here and apologize for the excessive use of abbreviations. We have removed nearly all of them

The figures were incomplete, poorly labeled, and hard to read. A lot of figures were missing, for example

Basic task structure

Basic behavior on the task

Scatter plot of the measures that you are clustering (lever press choice X number of pellets on the immediate lever, you can use color or multiple panels to indicate the delay to the delayed lever) Figure 3 is just a couple of examples. That isn't convincing at all.

Figure 4 is missing labels. In Figure 4, I don't understand what you are trying to say.

I don't see how the results on page 16 arise from Figure 6. I strongly recommend starting from the actual data and working your way to what it means rather than forcing this into this unreasonable "session group" analysis.

We have completely reworked the Figures for clarity and content. 

The statement that "no prior study has explored the cellular correlates of cognitive effort" is ludicrous and insulting. There are dozens of experiments looking at ACC in cognitive effort tasks, in humans, other primates, and rodents. There are many dozens of experiments looking at cellular correlates in intertemporal choice tasks, some with neural manipulations, some with ensemble recordings. There are many dozens of experiments looking at cellular relationships to waiting out a delay.

We agree that our statement was extremely imprecise. We have updated this to say:  “Further, a role for theta oscillations in allocating physical effort has been identified. However, the cellular

mechanisms within the ACC that control and deploy types of cognitive effort have not been identified.”

Reviewer #2 (Recommendations For The Authors):

In Figure 2, the panels below E and F are referred to as 'right' - but they are below? I would give them letters.

I would make sure that animal #s, neuron #s, and LFP#s are clearly presented in the results and in each figure legend. This is important to follow the results throughout the manuscript.

Some additional proofreading ('Fronotmedial') might help with clarity.

Based on our updates, this is no longer relevant.  

Reviewer #3 (Recommendations For The Authors):

In addition to the suggestions above to address specific issues, it would be useful to report some additional information about aspects of the experiments and analyses:

Specify how spike sorting was performed and what metrics were used to select well isolated single units.

Done.

Provide histology showing the recording locations for each subject.

Histological assessments of electrodes placements are provided in White et al. 2024, but we provide an example placement. This has been added to the text. 

Indicate the sequence of recording sessions that occurred for each subject, including for each session what delay duration was used and which dataset the session contributed to, and indicate when the neural probes were advanced between sessions.

We feel that this adds complexity unnecessarily as we make no claims about holding units across sessions for differences in coding in the dorsoventral gradient of ACC. 

Indicate the experimental unit when reporting uncertainty measures in figure legends (e.g. mean +/- SEM across sessions).

Done.

Reviewer #1 (Public review):

The paper from Hudait and Voth details a number of coarse-grained simulations as well as some experiments focused on the stability of HIV capsids in the presence of the drug lenacapavir. The authors find that LEN hyperstabilizes the capsid, making it fragile and prone to breaking inside the nuclear pore complex.

I found the paper interesting. I have a few suggestions for clarification and/or improvement.

(1) How directly comparable are the NPC-capsid and capsid-only simulations? A major result rests on the conclusion that the kinetics of rupture are faster inside the NPC, but are the numbers of LENs bound identical? Is the time really comparable, given that the simulations have different starting points? I'm not really doubting the result, but I think it could be made more rigorous/quantitative.

(2) Related to the above, it is stated on page 12 that, based on the estimated free-energy barrier, pentamer dissociation should occur in ~10 us of CG time. But certainly, the simulations cover at least this length of time?

(3) At first, I was surprised that even in a CG simulation, LEN would spontaneously bind to the correct site. But if I read the SI correctly, LEN was parameterized specifically to bind to hexamers and not pentamers. This is fine, but I think it's worth describing in the main text.

Reviewer #2 (Public review):

Here, Hudait et al. use CG modeling to investigate the mechanism by which lenacapavir (LEN) treats HIV capsids that dock to the nuclear pore complex (NPC). However, the manuscript fails to present meaningful findings that were previously unreported in the literature, and is thus of low impact. Many claims made in the manuscript are not substantiated by the presented data. Key mechanistic details that the work purports to reveal are artifacts of the parameterization choices or simulation/analysis design, with the simulations said to reveal details that they were specifically biased to reproduce. This makes the manuscript highly problematic, as its contributions to the literature would represent misconceptions based on oversights in modeling, and thus mislead future readers.

(1) Considering the literature, it is unclear that the manuscript presents new scientific discoveries. The following are results from this paper that have been previously reported:

(a) LEN-bound capsid can dock to the nuclear pore (Figure 2; see e.g. 10.1016/j.cell.2024.12.008 or 10.1128/mbio.03613-24).

(b) NUP98 interacts with the docked capsid (Figure 2; see e.g. 10.1016/j.virol.2013.02.008 or 10.1038/s41586-023-06969-7 or 10.1016/j.cell.2024.12.008).

(c) LEN and NUP98 compete for a binding interface (Figure 2; see e.g. 10.1126/science.abb4808 or 10.1371/journal.ppat.1004459).

(d) LEN creates capsid defects (Figure 3 and 5, see e.g. 10.1073/pnas.2420497122).

(e) RNP can emerge from a damaged capsid (Figure 3 and 5; see e.g. 10.1073/pnas.2117781119 or 10.7554/eLife.64776).

(f) LEN hyperstabilizes/reduces the elasticity of the capsid lattice (Figure 6; see e.g. 10.1371/journal.ppat.1012537).

(2) The mechanistic findings related to how these processes occur are problematic, either based on circular reasoning or unsubstantiated, based on the presented data. In some cases, features of parameterization and simulation/analysis design are erroneously interpreted as predictions by the CG models.

(a) Claim: LEN-bound capsids remain associated with the NPC after rupture. CG simulations did not reach the timescale needed to demonstrate continued association or failure to translocate, leaving the claim unsubstantiated.

(b) Claim: LEN contributes to loss of capsid elasticity. The authors do not measure elasticity here, only force constants of fluctuations between capsomers in freely diffusing capsids. Elasticity is defined as the ability of a material to undergo reversible deformation when subjected to stress. Other computational works that actually measure elasticity (e.g., 0.1371/journal.ppat.1012537) could represent a point of comparison, but are not cited. The changes in force constants in the presence of LEN are shown in Figure 6C, but the text of the scale bar legend and units of k are not legible, so one cannot discern the magnitude or significance of the change.

(c) Claim: Capsid defects are formed along striated patterns of capsid disorder. Data is not presented that correlates defects/cracks with striations.

(d) Claim: Typically 1-2 LEN, but rarely 3 bind per capsid hexamer. The authors state: "The magnitude of the attractive interactions was adjusted to capture the substoichiometric binding of LEN to CA hexamers (Faysal et al., 2024). ... We simulated LEN binding to the capsid cone (in the absence of NPC), which resulted in a substoichiometric binding (~1.5 LEN per CA hexamer), consistent with experimental data (Singh et al., 2024)." This means LEN was specifically parameterized to reproduce the 1-2 binding ratio per hexamer apparent from experiments, so this was a parameterization choice, not a prediction by CG simulations as the authors erroneously claim: "This indicates that the probability of binding a third LEN molecule to a CA hexamer is impeded, likely due to steric effects that prevent the approach of an incoming molecule to a CA hexamer where 2 LEN molecules are already associated. ... Approximately 20% of CA hexamers remain unoccupied despite the availability of a large excess of unbound LEN molecules. This suggests a heterogeneity in the molecular environment of the capsid lattice for LEN binding." These statements represent gross over-interpretation of a bias deliberately introduced during parameterization, and the "finding" represents circular reasoning. Also, if "steric effects" play any role, the authors could analyze the model to characterize and report them rather than simply speculate.

(e) Claim: Competition between NUP98 and LEN regulates capsid docking. The authors state: "A fraction of LEN molecules bound at the narrow end dissociate to allow NUP98 binding to the capsid ... Therefore, LEN can inhibit the efficient binding of the viral cores to the NPC, resulting in an increased number of cores in the cytoplasm." Capsid docking occurs regardless of the presence of LEN, and appears to occur at the same rate as the LEN-free capsid presented in the authors' previous work (Hudait &Voth, 2024). The presented data simply show that there is a fluctuation of bound LEN, with about 10 fewer (<5%) bound at the end of the simulation than at the beginning, and the curve (Figure 2A) does not clearly correlate with increased NUP98 contact. In that case, no data is shown that connects LEN binding with the regulation of the docking process. Further, the two quoted statements contradict each other. The presented data appear to show that NUP outcompetes LEN binding, rather than LEN inhibiting NUP binding. The "Therefore" statement is an attempt to reconcile with experimental studies, but is not substantiated by the presented data.

(f) Claim: LEN binding leads to spontaneous dissociation of pentamers. The CG simulation trajectories show pentamer dissociation. However, it is quite difficult to believe that a pentamer in the wide end of the capsid would dissociate and diffuse 100 nm away before a hexamer in the narrow end (previously between two pentamers and now only partially coordinated, also in a highly curved environment, and further under the force of the extruding RNA) would dissociate, as in Figure 2B. A more plausible explanation could be force balance between pent-hex versus hex-hex contacts, an aspect of CG parameterization. No further modeling is presented to explain the release of pentamers, and changes in pent-hex stiffness are not apparent in the force constant fluctuation analysis in Figure 6C.

(g) Claim: WTMetaD simulations predict capsid rupture. The authors state: "In WTMetaD simulations, we used the mean coordination number (Figure S6) between CA proteins in pentamers and in hexamers as the reaction coordinate." This means that the coordination number, the number of pent-hex contacts, is the bias used to accelerate simulation sampling. Yet the authors then interpret a change in coordination number leading to capsid rupture as a discovery, representing a fundamental misuse of the WTMetaD method. Changes in coordination number cannot be claimed as an emergent property when they are in fact the applied bias, when the simulation forced them to sample such states. The bias must be orthogonal to the feature of interest for that feature to be discoverable. While the reported free energies are orthogonal to the reaction coordinate, the structural and stepwise-mechanism "findings" here represent circular reasoning.

(3) Another major concern with this work is the excessive self-citation, and the conspicuous lack of engagement with similar computational modeling studies that investigate the HIV capsid and its interactions with LEN, capsid mechanical properties relevant to nuclear entry, and other capsid-NPC simulations (e.g., 10.1016/j.cell.2024.12.008 and 10.1371/journal.ppat.1012537). Other such studies available in the literature include examination of varying aspects of the system at both CG and all-atom levels of resolution, which could be highly complementary to the present work and, in many cases, lend support to the authors' claims rather than detract from them. The choice to omit relevant literature implies either a lack of perspective or a lack of collegiality, which the presentation of the work suffers from. Overall, it is essential to discuss findings in the context of competing studies to give readers an accurate view of the state of the field and how the present work fits into it. It is appropriate in a CG modeling study to discuss the potential weaknesses of the methodology, points of disagreement with alternative modeling studies, and any lack of correlation with a broader range of experimental work. Qualitative agreement with select experiments does not constitute model validation.

(4) Other critiques, questions, concerns:

(a) The first Results sub-heading presents "results", complete with several supplementary figures and a movie that are from a previous publication about the development of the HIV capsid-NPC model in the absence of LEN (Hudait &Voth, 2024). This information should be included as part of the introduction or an abbreviated main-text methods section rather than being included within Results as if it represents a newly reported advancement, as this could be misleading.

(b) The authors say the unbiased simulations of capsid-NPC docking were run as two independent replicates, but results from only one trajectory are ever shown plotted over time. It is not mentioned if the time series data are averaged or smoothed, so what is the shadow in these plots (e.g., Figures 1,2, and Supplementary Figure 5)?

(c) Why do the insets showing LEN binding in Figure 2A look so different from the models they are apparently zoomed in on? Both instances really look like they are taken from different simulation frames, rather than being a zoomed-in view.

(d) What are the sudden jerks apparent in the SI movies? Perhaps this is related to the rate at which trajectory frames are saved, but occasionally, during the relatively smooth motion of the capsid-NPC complex, something dramatic happens all of a sudden in a frame. For example, significant and apparently instantaneous reorientation of the cone far beyond what preceding motions suggest is possible (SI movie 2, at timestamp 0.22), RNP extrusion suddenly in a single frame (SI movie 2, at timestamp 0.27), and simultaneous opening of all pentamers all at once starting in a single frame (SI movie 2, at timestamp 0.33). This almost makes the movie look generated from separate trajectories or discontinuous portions of the same trajectory. If movies have been edited for visual clarity (e.g., to skip over time when "nothing" is happening and focus on the exciting aspects), then the authors should state so in the captions.

(e) Figure 3c presents a time series of the degree of defects at pent-hex and hex-hex interfaces, but I do not understand the normalization. The authors state, "we represented the defects as the number of under-coordinated CA monomers of the hexamers at the pentamer-hexamer-pentamer and hexamer-hexamer interface as N_Pen-Hex and N_Hex-Hex ... Note that in N_Pen-Hex and N_Hex-Hex are calculated by normalizing by the total number of CA pentamer (12) and hexamer rings (209) respectively." Shouldn't the number of uncoordinated monomers be normalized by the number of that type of monomer, rather than the number of capsomers/rings? E.g., 12*5 and 209*6, rather than 12 and 209?

(f) The authors state that "Although high computational cost precluded us from continuing these CG MD simulations, we expect these defects at the hexamer-hexamer interface to propagate towards the high curvature ends of the capsid." The defects being reported are apparently propagating from (not towards) the high curvature ends of the capsid.

(g) The first half of the paper uses the color orange in figures to indicate LEN, but the second half uses orange to indicate defects, and this could be confusing for some readers. Both LEN and "defects" are simply a cluster of spheres, so highlighted defects appear to represent LEN without careful reading of captions.

(h) SI Figure S3 captions says "The CA monomers to which at least one LEN molecule is bound are shown in orange spheres. The CA monomers to which no LEN molecule is bound are shown in white spheres. " While in contradiction, the main-text Fig 2 says "The CA monomers to which at least one LEN molecule is bound are shown in white spheres. The CA monomers to which no LEN molecule is bound are shown in orange spheres. " One of these must be a typo.

(i) The authors state that: "CG MD simulations and live-cell imaging demonstrate that LEN-treated capsids dock at the NPC and rupture at the narrow end when bound to the central channel and then remain associated to the NPC after rupture." However, the live cell imaging data do not show where rupture occurs, such that this statement is at least partially false. It is also unclear that CG simulations show that cores remain bound following rupture, given that simulations were not extended to the timescale needed to observe this, again rendering the statement partially false.

(j) The authors state: "We previously demonstrated that the RNP complex inside the capsid contributes to internal mechanical strain on the lattice driven by CACTD-RNP interactions and condensation state of RNP complex (Hudait &Voth, 2024). " In that case, why do the present CG models detect no difference in results for condensed versus uncondensed RNP?

(k) The authors state: "The distribution demonstrates that the binding of LEN to the distorted lattice sites is energetically favorable. Since LEN localizes at the hydrophobic pocket between two adjoining CA monomers, it is sterically favorable to accommodate the incoming molecule at a distorted lattice site. This can be attributed to the higher available void volume at the distorted lattice relative to an ordered lattice, the latter being tightly packed. This also allows the drug molecule to avoid the multitude of unfavorable CA-LEN interactions and establish the energetically favorable interactions leading to a successful binding event. " What multitude of unfavorable interactions are the authors referring to? Data is not presented to substantiate the claim of increased void volume between hexamers in the distorted lattice. Capsomer distortion is shown as a schematic in Figure 6A rather than in the context of the actual model.

(l) The authors state that "These striated patterns also demonstrate deviations from ideal lattice packing. " What does ideal lattice packing mean in this context, where hexamers are in numerous unique environments in terms of curvature? What is the structural reference point?

(m) If pentamer-hexamer interactions are weakened in the presence of LEN, why are differences at these interfaces not apparent in the Figure 6C data that shows stiffening of the interactions between capsomer subunits?

(n) The authors state: "Lattice defects arising from the loss of pentamers and cracks along the weak points of the hexameric lattice drive the uncoating of the capsid." The word rupture or failure should be used here rather than uncoating; it is unclear that the authors are studying the true process of uncoating and whether the defects induced by LEN binding relate in any way to uncoating.

(o) The authors state: "LEN-treated broken cores are stabilized by the interaction with the disordered FG-NUP98 mesh at the NPC." But no data is presented to demonstrate that capsid stability is increased by NUP98 interaction. In fact, the presented data could suggest the opposite since capsids in contact with NUP98 in the NPC appeared to rupture faster than freely diffusing capsids.

(p) The authors state: "LEN binding stimulates similar changes in free capsids, but they occur with lower frequency on similar time scales, suggesting that the cores docked at the NPC are under increased stress, resulting in more frequent weakening of the hexamer-pentamer and hexamer-hexamer interactions, as well as more nucleation of defects at the hexamer-hexamer<br /> Interface. ... Our results suggest that in the presence of the LEN, capsid docking into the NPC central channel will increase stress, resulting in more frequent breaks in the capsid lattice compared to free capsids." The first is a run-on sentence. The results shown support that LEN stimulates changes in free capsids to happen faster, but not more frequently. The frequency with which an event occurs is separate from the speed with which the event occurs.

(q) The authors state: "A possible mechanistic pathway of capsid disassembly can be that multiple pentamers are dissociated from the capsid sequentially, and the remaining hexameric lattice remains stabilized by bound LEN molecules for a time, before the structural integrity of the remaining lattice is compromised." This statement is inconsistent with experimental studies that say LEN does not lead to capsid disassembly, and may even prevent disassembly as part of its disruption of proper uncoating (e.g., 10.1073/pnas.2420497122 previously published by the authors).

(r) Finally, it remains a concern with the authors' work that the bottom-up solvent-free CG modeling software used in this and supporting works is not open source or even available to other researchers like other commonly used molecular dynamics software packages, raising significant questions about transparency and reproducibility.

Author response:

Before providing a brief provisional response to the two reviews, it is important to reiterate a few key points about our work. First, our paper is largely a computational biophysics paper, augmented by experimental results. Generally speaking, computational biophysics work intends to achieve one of two things (or both). One is to provide more molecular level insight into various behaviors of biomolecular systems that have not been (or cannot be) provided by qualitative experimental results alone. The second general goal of computational biophysics it to formulate new hypotheses to be tested subsequently by experiment. In our paper, we have achieved both of these goals and then confirmed the key computational results by experiment..

The first reviewer has some valuable points, which can be addressed as follows (and will be emphasized in the revised version of the paper): (1) Yes the simulations of capsid rupture in the NPC and capsid-only are directly comparable as both have approximately the same number of bound LEN, as determined by following the LEN-capsid interaction protocol described in the main text (around Fig 6) and in the SI section S3; (2) While we have stressed this point in several places in the manuscript, here again we stress that coarse-grained (CG) MD time is not the same as real time. The point of CG simulations is to accelerate the timescale of the MD and the associated sampling, so the CG “time” from the MD integrator needs to be rescaled to associate a real time to it. As such, our CG simulation is not representing a microsecond of real time but rather something much longer. We will emphasize this again in the revised text. (3) Actually, we think that the parameterization of the LEN model and the LEN-capsid interactions is well described in the text associated with Fig 6 and in SI section S3. It is true that this one part of the CG model was parameterized “top-down” given the good experimental structures of bound LEN to capsid and other data, but the rest of the CG model is “bottom-up” (meaning developed from well-defined coarse-graining statistical mechanics as applied to molecular level structures and interactions, see also below). 

As for the second reviewer, this review is quite problematic in our view as the reviewer seems to think that quoting a number of qualitative experimental results is sufficient to undermine the impact of our paper (they are not) and, furthermore, the reviewer appears to have a very minimal understanding of “bottom-up” CG modeling, which we have utilized. This modeling does not in fact rely on the “assumptions” this reviewer alleges we have relied on. (As an aside, it could be helpful for this reviewer to study the review by Jin et al, https://doi.org/10.1021/acs.jctc.2c00643) in order to become more familiar with the field and our approach before criticizing it.) We also note that our main HIV capsid-NPC docking model is already published in PNAS (https://doi.org/10.1073/pnas.2313737121), where it underwent rigorous peer review. In our forthcoming full response to the reviews and in the revised paper we will attempt to address a number of this reviewers comments, but the number, extent, and tone of this collection of criticisms, for us, calls into question the objectivity of this reviewer, not to mention the reviewer’s rather weak understanding of what we have done and how we have done it.

Finally, while we certainly appreciate the overall positive eLife assessment, we are disappointed by the statement “some mechanistic interpretations rely on assumptions embedded in the simulations, leaving parts of the evidence incomplete”. Of course, all simulations (and experiments) rely on certain assumptions, but we have gone to great length to provide a “bottomup” approach to our modeling, based on underlying molecular level structures and interactions, and we have provided experimental validation of the main simulation predictions. It seems that the comments of the second reviewer may have influenced this point of view, but we do not feel it is justified.

Reviewer #1 (Public review):

Summary:

Spinal projection neurons in the anterolateral tract transmit diverse somatosensory signals to the brain, including touch, temperature, itch, and pain. This group of spinal projection neurons is heterogeneous in their molecular identities, projection targets in the brain, and response properties. While most anterolateral tract projection neurons are multimodal (responding to more than one somatosensory modality), it has been shown that cold-selective projection neurons exist in lamina I of the spinal cord dorsal horn. Using a combination of anatomical and physiological approaches, the authors discovered that the cold-selective lamina I projection neurons are heavily innervated by Trpm8+ sensory neuron axons, with calb1+ spinal projection neurons primarily capturing these cold-selective lamina I projection neurons. These neurons project to specific brain targets, including the PBNrel and cPAG. This study adds to the ongoing effort in the field to identify and characterize spinal projection neuron subtypes, their physiology, and functions.

Strengths:

(1) The combination of anatomical and physiological analyses is powerful and offers a comprehensive understanding of the cold-selective lamina I projection neurons in the spinal cord dorsal horn. For example, the authors used detailed anatomical methods, including EM imaging of Trpm8+ axon terminals contacting the Phox2a+ lamina I projection neurons. Additionally, they recorded stimulus-evoked activity in Trpm8-recipient neurons, carefully selected by visual confirmation of tdTomato and GFP juxtaposition, which is technically challenging.

(2) This study identifies, for the first time, a molecular marker (calb1) that labels cold-selective lamina I projection neurons. Although calb1+ projection neurons are not entirely specific to cold-selective neurons, using an intersectional strategy combined with other genes enriched in this ALS group or cold-induced FosTRAP may further enhance specificity in the future.

(3) This study shows that cold-selective lamina I projection neurons specifically innervate certain brain targets of the anterolateral tract, including the NTS, PBNrel, and cPAG. This connectivity provides insights into the role of these neurons in cold sensation, which will be an exciting area for future research.

Weaknesses:

(1) The sample size for the ex vivo electrophysiology is small. Given the difficulty and complexity of the preparation, this is understandable. However, a larger sample size would have strengthened the authors' conclusions.

(2) The authors used tdTomato expression to identify brain targets innervated by these cold-selective lamina I projection neurons. Since tdTomato is a soluble fluorescent protein that fills the entire cell, using synaptophysin reporters (e.g., synaptophysin-GFP) would have been more convincing in revealing the synaptic targets of these projection neurons.

(3) The summary cartoon shown in Figure 7 can be misleading because this study did not determine whether these cold-selective lamina I projection neurons have collateral branches to multiple brain targets or if there are anatomical subtypes that may project exclusively to specific targets. For example, a recent study (Ding et al., Neuron, 2025) demonstrated that there are PBN-projecting spinal neurons that do not project to other rostral brain areas. Furthermore, based on the authors' bulk labeling experiments, the three main brain targets are NTS, PBNrel, and cPAG. The VPL projection is very sparse and almost negligible.

Reviewer #2 (Public review):

Lang et al. investigate the contribution of individual neuronal encoding of specific task features to population dynamics and behavior. Using a taste-based decision-making behavioral task with electrophysiology from the mouse gustatory cortex and computational modeling, the authors reveal that neurons encoding sensory, perceptual, and decision-related information with linear and categorical patterns are essential for driving neural population dynamics and behavioral performance. Their findings suggest that individual linear and categorical coding units have a significant role in cortical dynamics and perceptual decision-making behavior.

Overall, the experimental and analytical work is of very high quality, and the findings are of great interest to the taste coding field, as well as to the broader systems neuroscience field.

I have a couple of suggestions to further enhance the authors' important conclusions:

My main comment is the distinction between constrained and unconstrained units. The authors train a small percentage of units to match the real neural data (constrained units), and then find some unconstrained units that are similar to the real neural data and some that are not. As far as I could tell, the relative fraction of constrained and unconstrained units in the trained RNN is not reported; I assume the constrained ones are a much smaller population, but this is unclear. The selection of different groups of neurons for the RNN ablation experiments appears to be based on their response profiles only. Therefore, if I understood correctly, both constrained and unconstrained units and ablated together for a given response category (e.g., linear or step-perception). It would be useful, therefore, to separately compare the effects of constrained vs. unconstrained RNN units.

Specifically:

(1) For the analyses in the initial version of the manuscript, the authors should specify how many units in each ablation category are constrained and unconstrained.

(2) The authors should repeat Figure 6, but only for unconstrained units to test how much of the effects in the initial version of Figure 6 are driven by constrained vs. unconstrained RNN units.

(3) The authors should repeat Figure 7, but performing ablations separately on the constrained and unconstrained units to examine how the network behaves in each case and the resulting "behavioral" effect.

Reviewer #3 (Public review):

Primary taste cortex neurons show a variety of dynamic response profiles during taste decision-making tasks, reflecting both sensory and decision variables. In the present study, Lang et al. set out to determine how neurons with distinct response profiles contribute to perceptual decisions about taste stimuli.

The methods, with reference to the behavioral task and electrophysiological recordings/data analysis, are straightforward, solid, and appropriate. The computational model is presented in a clear and conceptually intuitive manner, although the details are outside of my area of expertise.

The experimental design features a simple 2-alternative forced-choice design that yielded clear psychometric curves across a range of stimuli. In vivo recordings were performed using Neuropixels and yielded an appropriate sample of single neuron responses. The strength of the model lies in the fact that it consists of single neurons whose response profiles mimic those recorded in vivo, and allows neuron-selective manipulation.

By virtually lesioning specific subsets of neurons in the network, the authors demonstrate that a relatively small population of neurons with specific tuning profiles was sufficient to produce the observed neural dynamics and behavioral responses. This effect was selective as lesioning other responsive neurons did not affect overall response dynamics or performance.

These findings provide new insight into the relation between the response profiles of single neurons in sensory cortex, their population-level activity dynamics, and the perceptual decisions they inform.

The approach is particularly innovative as it uses computational modeling to target functionally-defined "cell types", which cannot necessarily be targeted by more conventional genetic approaches.

Author response:

Reviewer #1 (Public review):

This manuscript provides several important findings that advance our current knowledge about the function of the gustatory cortex (GC). The authors used high-density electrophysiology to record neural activity during a sucrose/NaCl mixture discrimination task. They observed population-based activity capable of representing different mixtures in a linear fashion during the initial stimulus sampling period, as well as representing the behavioral decision (i.e., lick left or right) at a later time point. Analyzing this data at the single neuron level, they observed functional subpopulations capable of encoding the specific mixture (e.g., 45/55), tastant (e.g., sucrose), and behavioral choice (e.g., lick left). To test the functional consequences of these subpopulations, they built a recurrent neural network model in order to "silence" specific functional subpopulations of GC neurons. The virtual ablation of these functional subpopulations altered virtual behavioral performance in a manner predicted by the subpopulation's presumed contribution.

Strengths:

Building a recurrent neural network model of the gustatory cortex allows the impact of the temporal sequence of functionally identifiable populations of neurons to be tested in a manner not otherwise possible. Specifically, the author's model links neural activity at the single neuron and population level with perceptual ability. The electrophysiology methods and analyses used to shape the network model are appropriate. Overall, the conclusions of the manuscript are well supported.

Weaknesses:

One potential concern is the apparent mismatch between the neural and behavioral data. Neural analyses indicate a clear separation of the activity associated with each mixture that is independent of the animal's ultimate choice. This would seemingly indicate that the animals are making errors despite correctly encoding the stimulus. Based solely on the neural data, one would expect the psychometric curve to be more "step-like" with a significantly steeper slope. One potential explanation for this observation is the concentration of the stimuli utilized in the mixture discrimination task. The authors utilize equivalent concentrations, rather than intensity-matched concentrations. In this case, a single stimulus can (theoretically) dominate the perception of a mixture, resulting in a biased behavioral response despite accurate concentration coding at the single neuron level. Given the difficulty of isointensity matching concentrations, this concern is not paramount. However, the apparent mismatch between the neural and behavioral data should be acknowledged/addressed in the text.

We thank the Reviewer for the insightful comments and thoughtful suggestions. Our electrophysiological recordings show that GC dynamically encodes stimulus concentration of mixture elements, dominant perceptual quality, and decisions of directional lick. With regard to the encoding of mixtures, the clear separation of activity associated with each mixture (Figure 3) is present at a trial-averaged pseudo-population level, and average activities associated with more similar, intermediate mixtures are closer to each other in this space. In fact, at a single trial level activity evoked by similar, intermediate mixtures can be hard to separate. This increased similarity can lead to behavioral errors resulting from either incorrect encoding of the stimulus or from the inability to interpret the stimuli to guide the correct decision.

The psychometric function, which shows that more distinct stimuli (100/0 vs 0/100) lead to fewer mistakes than more ambiguous, intermediate mixtures (55/45 vs 55/45), is consistent with the increased ambiguity of responses to intermediate mixtures and with the possibility that, compared to pure stimuli, intermediate mixtures lead to more trials in which the binary choice component of neural activity is inverted, resulting in more directional errors.

The Reviewer is correct that there could be a slight mismatch in the perceived intensity of the mixture components. This mismatch could be the reason for the slight asymmetry in our psychometric function (Figure 1B). However, it is not uncommon for mice in these 2AC tasks to also have a motor laterality bias in their responses that manifests itself for the more ambiguous stimuli. We chose not to model this bias given its subtlety and its unknown origin. Rather, we chose to model an ideal scenario in which stimuli have matched intensity and no motor bias exists. In the revised version we will discuss this issue.

Reviewer #2 (Public review):

Lang et al. investigate the contribution of individual neuronal encoding of specific task features to population dynamics and behavior. Using a taste-based decision-making behavioral task with electrophysiology from the mouse gustatory cortex and computational modeling, the authors reveal that neurons encoding sensory, perceptual, and decision-related information with linear and categorical patterns are essential for driving neural population dynamics and behavioral performance. Their findings suggest that individual linear and categorical coding units have a significant role in cortical dynamics and perceptual decision-making behavior.

Overall, the experimental and analytical work is of very high quality, and the findings are of great interest to the taste coding field, as well as to the broader systems neuroscience field.

I have a couple of suggestions to further enhance the authors' important conclusions:

My main comment is the distinction between constrained and unconstrained units. The authors train a small percentage of units to match the real neural data (constrained units), and then find some unconstrained units that are similar to the real neural data and some that are not. As far as I could tell, the relative fraction of constrained and unconstrained units in the trained RNN is not reported; I assume the constrained ones are a much smaller population, but this is unclear. The selection of different groups of neurons for the RNN ablation experiments appears to be based on their response profiles only. Therefore, if I understood correctly, both constrained and unconstrained units and ablated together for a given response category (e.g., linear or step-perception). It would be useful, therefore, to separately compare the effects of constrained vs. unconstrained RNN units.

We thank the Reviewer for the constructive feedback and are pleased that the work is considered of broad interest. The Reviewer is correct that ablations were carried out with respect to response categories only and included both constrained and unconstrained units.

The ratio of total units to constrained units is fixed at 5.88, thus constrained units are ~17% of the network and unconstrained units are ~83%. This value is specified in the Methods (RNN: Components and dynamics), but we will report it in the Results of the revised manuscript as well for clarity.

Specifically:

(1) For the analyses in the initial version of the manuscript, the authors should specify how many units in each ablation category are constrained and unconstrained.

In the revised manuscript, we will specify the fractions of constrained and unconstrained units within each response category. For convenience, they are reported here: Linear = 194 constrained and 691 unconstrained units; Step-perception = 147 constrained and 840 unconstrained units; Step-choice = 129 constrained and 814 unconstrained units; Other = 353 constrained and 1739 unconstrained units.

(2) The authors should repeat Figure 6, but only for unconstrained units to test how much of the effects in the initial version of Figure 6 are driven by constrained vs. unconstrained RNN units.

In the revised version we will add a Supplemental Figure in which the contribution of constrained vs unconstrained units is addressed.

(3) The authors should repeat Figure 7, but performing ablations separately on the constrained and unconstrained units to examine how the network behaves in each case and the resulting "behavioral" effect.

The revised version will include a Supplemental Figure with these simulations.

Reviewer #3 (Public review):

Primary taste cortex neurons show a variety of dynamic response profiles during taste decision-making tasks, reflecting both sensory and decision variables. In the present study, Lang et al. set out to determine how neurons with distinct response profiles contribute to perceptual decisions about taste stimuli.

The methods,with reference to the behavioral task and electrophysiological recordings/data analysis, are straightforward, solid, and appropriate. The computational model is presented in a clear and conceptually intuitive manner, although the details are outside of my area of expertise.

The experimental design features a simple 2-alternative forced-choice design that yielded clear psychometric curves across a range of stimuli. In vivo recordings were performed using Neuropixels and yielded an appropriate sample of single neuron responses. The strength of the model lies in the fact that it consists of single neurons whose response profiles mimic those recorded in vivo, and allows neuron-selective manipulation.By virtually lesioning specific subsets of neurons in the network, the authors demonstrate that a relatively small population of neurons with specific tuning profiles was sufficient to produce the observed neural dynamics and behavioral responses. This effect was selective as lesioning other responsive neurons did not affect overall response dynamics or performance.These findings provide new insight into the relation between the response profiles of single neurons in sensory cortex, their population-level activity dynamics, and the perceptual decisions they inform.

The approach is particularly innovative as it uses computational modeling to target functionally-defined "cell types", which cannot necessarily be targeted by more conventional genetic approaches.

We thank the Reviewer for the positive assessment of our study.

Reviewer #1 (Public review):

Summary:

In this manuscript, Green et al. attempt to use large-scale protein structure analysis to find signals of selection and clustering related to antibiotic resistance. This was applied to the whole proteome of Mycobacterium tuberculosis, with a specific focus on the smaller set of known antibiotic-resistance-related proteins.

Strengths:

The use of geospatial analysis to detect signals of selection and clustering on the structural level is really intriguing. This could have a wider use beyond the AMR-focussed work here and could be applied to a more general evolutionary analysis context. Much of the strength of this work lies in breaking ground into this structural evolution space, something rarely seen in such pathogen data. Additional further research can be done to build on this foundation, and the work presented here will be important for the field.

The size of the dataset and use of protein structure prediction via AlphaFold, giving such a consistent signal within the dataset, is also of great interest and shows the power of these approaches to allow us to integrate protein structure more confidently into evolution and selection analyses.

Weaknesses:

There are several issues with the evolutionary analysis and assumptions made in the paper, which perhaps overstate the findings, or require refining to take into account other factors that may be at play.

(1) The focus on antimicrobial resistance (AMR) throughout the paper contains the findings within that lens. This results in a few different weaknesses:

(a) While the large size of the analysis is highlighted in the abstract and elsewhere, in reality, only a few proteins are studied in depth. These are proteins already associated with AMR by many other studies, somewhat retreading old ground and reducing the novelty.

(b) Beyond the AMR-associated proteins, the proteome work is of great interest, but only casually interrogated and only in the context of AMR. There appears to be an assumption that all signals of positive selection detected are related to AMR, whereas something like cas10 is part of the CRISPR machinery, a set of proteins often under positive selection, and thus unlikely to be AMR-related.

(2) The strength of the signal from the structural information and the novelty of the structural incorporation into prediction are perhaps overstated.

(a) A drop of 13% in F1 for a gain of 2% in PPV is quite the trade-off. This is not as indicative of a strong predictor that could be used as the abstract claims. While the approach is novel and this is a good finding for a first attempt at such complex analysis, this is perhaps not as significant as the authors claim

(b) In relation to this, there is a lack of situating these findings within the wider research landscape. For instance, the use of structure for predicting resistance has been done, for example, in PncA (https://academic.oup.com/jacamr/article/6/2/dlae037/7630603, https://www.sciencedirect.com/science/article/pii/S1476927125003664, https://www.nature.com/articles/s41598-020-58635-x) and in RpoB (https://www.nature.com/articles/s41598-020-74648-y). These, and other such works, should be acknowledged as the novelty of this work is perhaps not as stark as the authors present it to be.

(3) The authors postulate that neutral AA substitutions would be randomly distributed in the protein structure and thus use random mutations as a negative control to simulate this neutral evolution. However, I am unsure if this is a true negative control for neutral evolution. The vast majority of residues would be under purifying selection, not neutral selection, especially in core proteins like rpoB and gyrA. Therefore, most of these residues would never be mutated in a real-world dataset. Therefore, you are not testing positive selection against neutral selection; you are testing positive against purifying, which will have a much stronger signal. This is likely to, in turn, overestimate the signal of positive selection. This would be better accounted for using a model of neutral evolution, although this is complex and perhaps outside the scope. Still, it needs to be made clear that these negative controls are not representative of neutral evolution.

(4) In a similar vein, the use of 15 Å as a cut-off for stating co-localisation feels quite arbitrary. The average radius of a globular protein is about 20 Å, so this could be quite a large patch of a protein. I think it may be good to situate the cut-off for a 'single location' within a size estimator of the entire protein, as 15 Å could be a neighbourhood in a large protein, but be the whole protein for smaller ones.

Reviewer #1 (Public review):

Summary:

This manuscript by Xie and colleagues presents an intriguing behavioral finding for the field of perceptual learning (PL): combining the reactivation-based training paradigm with anodal tDCS induces complete generalization of the learning effect. Notably, this generalization is achieved without compromising the magnitude of learning effects and with an 80% reduction in total training time. The experimental design is well-structured, and the observed complete generalization is robustly replicated across two stimulus dimensions (orientation and motion direction).

However, while the empirical results are methodologically valid and scientifically surprising, the theoretical framework proposed to explain them appears underdeveloped and, in some cases, difficult to reconcile with the existing literature. Several arguments are insufficiently justified. In addition, the introduction of a non-standard metric (NGI: normalized learning gain index) raises concerns about the interpretability and comparability with existing PL literature.

Strengths:

(1) Rigorous experimental design

In this study, Xie and colleagues employed a 2×2 factorial design (Training paradigm: Reactivation vs. Full-Practice × tDCS protocols: Anodal vs. Sham), which allowed clear dissociation of the main and interaction effects.

(2) High statistical credibility

Sample sizes were predetermined using G*Power, non-significant effects were evaluated using the Bayes factor, and the core behavioral findings were replicated in a second stimulus dimension. These strengthen the credibility of the findings.

(3) Strong translational potential

The observed complete generalization could have useful implications for sensory rehabilitation. The large reduction (80%) in total training time is particularly compelling.

Weaknesses:

(1) NGI (Normalized learning gain index) is a non-standard behavioral metric and may distort interpretability.

NGI (pre - post / ((pre + post) / 2)) is rarely used in PL studies to measure learning effects. Almost all PL studies rely on raw thresholds and percent improvements (pre - post / pre), making it difficult to contextualize the current NGI-based results within the broader field. The current manuscript provides no justification for adopting NGI.

A more critical issue is the NGI's nonlinearity: by normalizing to the mean of pre- and post-test thresholds, it disproportionately inflates learning effects for participants with lower post-test thresholds. Notably, the "complete generalization" claims are illustrated mainly with NGI plots. Although the authors also analyze thresholds directly and the results also support the core claim, the interpretation in the text relies heavily on NGI.

The authors may consider rerunning key analyses using the standard percent improvement metric. If retaining NGI, the authors should provide explicit justification for why NGI is superior to standard measures.

(2) The proposed theoretical framework is sometimes unclear and insufficiently supported.

The authors propose the following mechanistic chain:

(a) reactivation-based learning depends on offline consolidation mediated by GABA (page 4 line 73);

(b) online a-tDCS reduces GABA (page 4, line 76), thereby disrupting offline consolidation (page 11, line 225);

(c) disrupted offline consolidation reduces perceptual overfitting (page 4, line 77; page 11, line 225), thereby enabling generalization;

(d) under full-practice training, a-tDCS increases specificity via a different mechanism (page 11 line 235).

While this framework is plausible in broad terms, several components are speculative at best in the absence of neurochemical or neural measurements.

(3) Several reasoning steps require further clarification.

(a) Mechanisms of Reactivation-based Learning.

The manuscript focuses on the neurochemical basis of reactivation-based learning. However, reactivation-induced neurochemical changes differ across brain regions. In the motor cortex, Eisenstein et al. (2023) reported that after reactivation, increased GABA and decreased E/I ratio were associated with offline gains. In contrast, Bang et al. (2018) demonstrated that, in the visual cortex, reactivation decreased GABA and increased E/I ratio. While both studies are consistent with GABA involvement, the direction of GABA modulation differs. The authors should clarify this discrepancy.<br /> More importantly, Bang et al. (2018) demonstrated that reactivation-based (3 blocks) and full-practice (16 blocks) training produced similar time courses of E/I ratio changes in V1: an initial increase followed by a decrease. Given this similarity, the manuscript would benefit from a more thorough discussion of how the two paradigms diverge mechanistically. For example, behaviorally, Song et al. (2021) reported greater generalization with reactivation-based training than with full-practice training, aligning with Kondat et al. (2025). Neurally, Kondat et al. (2024) showed that reactivation-based training increased activity in higher-order brain regions (e.g., IPS), whereas full practice training reduced connectivity between temporal and parietal regions.

(b) tDCS Mechanisms and Protocols.

The effect of a-tDCS on GABA is not consistent across brain regions. While a-tDCS reliably reduces GABA in the motor cortex, recently, a more related work (Abuleli et al., 2025) reports no significant modulation of GABA or Glx in V1, challenging the authors' assumption of tDCS-induced GABA reduction in the visual cortex.

The manuscript proposes that online a-tDCS disrupts offline consolidation is somewhat difficult to interpret conceptually. Online tDCS typically modulates processes occurring during stimulation (e.g., encoding process, attentional state), whereas consolidation occurs afterward. Thus, stating that online tDCS protocols only disrupt offline consolidation without considering the possibility that they first modulate the encoding process is difficult to interpret. Even if tDCS has prolonged effects, the link between online stimulation and disruption of offline consolidation remains unelucidated.

(c) Missing links between GABA modulation and perceptual overfitting.

The proposed chain ("tDCS disrupts consolidation → reduced overfitting → improved generalization") skips a critical step: how GABA modulation translates to changes in neural representational properties (e.g., tuning width, representational overlap between trained/untrained stimuli) that define "perceptual overfitting." The PL literature has not established a link between GABA levels and these representational changes, leaving a key component of the mechanistic explanation underspecified.

(d) Insufficient explanation of the opposite effects.

The manuscript does not fully explain why the same a-tDCS promotes generalization in reactivation-based training but increases specificity in full-practice training. Both paradigms engage offline consolidations, and, as mentioned above, the time courses of E/I ratio changes are similar for 3-block reactivation-based or 16-block training. Thus, if offline consolidation mechanisms (and their associated E/I changes) are comparable across paradigms, it is unclear why identical a-tDCS would produce opposite outcomes in the two paradigms.

Reviewer #2 (Public review):

Xie et al., combined transcranial direct current brain stimulation (tDCS) and a reactivation-based training protocol to investigate the generalization of learning. Using visual perceptual learning as a model, they found that a reactivation-based training protocol, when combined with anodal tDCS over the visual cortex, can induce learning transfer to untrained visual orientations and motion directions. Interestingly, extending reactivation-based training to a full-training protocol with more training trials did not induce generalization of learning. Furthermore, even when paired with tDCS, extending the training protocol did not provide benefits for generalization of learning. This study provides interesting insights into the mechanisms of brain plasticity and how future training protocols could be designed to achieve robust and generalizable learning outcomes.

The authors supported their arguments with a series of well-constructed experiments. The conclusions are largely supported by the data, although some clarifications about their hypotheses and control analyses could strengthen the work:

(1) The authors hypothesize that tDCS can reduce perceptual overfitting through reduced GABA concentrations in the visual cortex, which leads to learning transfer. However, without a clear description of the role of GABA in perceptual learning and perceptual overfitting, it is difficult for the reader to understand why reduced GABA concentrations would contribute to generalization. Do the authors imply that increased GABA can lead to specificity? Are there studies that can support this argument? The authors also did not describe clearly how reactivation-based visual perceptual learning can modify GABA levels in the visual cortex differently (compared to full-practice) during training and during the offline consolidation phase. In order for the reader to better understand their hypotheses and the motivation of the current study, it is beneficial for the authors to provide a concise but clearer description of the roles of GABA in perceptual learning with a focus on the roles of GABA in generalization and during off-line consolidation for different types of training protocols (see for instance Bang et al., 2018; Frangou et al., 2019; Frank et al., 2022; Jia et al., 2024; Shibata et al., 2011; Tamaki et al., 2020; Yamada et al., 2024).

(2) Based on the results, an alternative explanation is that the amount of transfer to the untrained visual feature might be related to the amount of learning for the trained visual feature, which might be different depending on the training protocol and brain stimulation combination. Is it beneficial to compare the amount of learning gains across different training and stimulation protocols to rule out this possibility? Would more learning gains for the trained visual feature predict less transfer for the untrained visual feature? Are there correlations between learning gains and learning transfer?

(3) The authors argued that a reactivation-based training protocol, rather than the amount of training, was critical for the generalization of learning. The control experiment in the study showed that full-practice training combined with tDCS did not lead to transfer, as in reactivation-based training. However, in order to rule out the confounding effects from the amount of training, it is crucial to examine whether a training protocol in which a similar number of trials as in the reactivation-based training but not separated across training sessions would lead to similar generalization of learning.

Reviewer #3 (Public review):

Summary:

This research focuses on a long-lasting and interesting phenomenon in human plasticity. When humans learn basic perceptual skills such as judging the orientation of a simple line, the learned abilities are often limited to the trained condition but not generalizable to untrained conditions. The authors hypothesized that this learning specificity was related to GABA, an inhibitory neurotransmitter in the brain. Using a novel training method that combines reactivation and a brain stimulation method (tDCS) that hypothetically inactivates GABA, the authors hypothesized that learned visual perceptual skills would show greater transfer.

Strengths:

The authors conducted a list of well-conceived behavior studies to demonstrate the effectiveness of their proposed method in enabling learning transfer in two different visual tasks, and carefully conducted comparison studies to elucidate other possible explanations. The sample size was adequate to convey convincing results, and the analyses were thorough.

Weaknesses:

While the authors built their training paradigm on

(1) the hypothetical role GABA plays in inhibiting learning transfer, and

(2) the hypothetical impact tDCS may have on GABA, there was no direct evidence supporting these hypotheses in the current study.

Further, learning specificity takes many formats from features to locations to tasks; it is not yet clear the scope of the observed transfer with the proposed method.

Reviewer #2 (Public review):

The authors describe the first deep neurological characterization of WAC mutation in two vertebrate species (zebrafish and mouse). They examine these at various levels, guided by the work in humans that has associated a heterozygous WAC mutation with DeSantos Shinawi Syndrome (DESSH). Therefore, they investigate the animals for a variety of phenotypes, following a template for what is seen when characterizing a new mouse/fish model of a developmental disability gene. Investigations include analysis of skull and jaw for abnormalities(both species), MRI of brain structure(in mice), electrophysiology(mice), assessment of signaling pathways (by Western blot, in mice), cell counts (both, more in mice), transcriptomics (mice), and behavior (both).

Generally, this describes an important first characterization of the consequences of the mutation. Most of the studies appear well-conducted and reasonably powered, thus solid or convincing. However, there are a few places where the data presentation could be improved for clarity, and a few concerns about some choices in analytical approach for a couple of the experiments, where improved statistical approaches could improve their sensitivity and/or better rule out false positives, and thus the support of some of these claims is currently incomplete. There is also some lack of clarity about the rationale for some decisions regarding the fish genetics. Nonetheless, this is an important and useful first characterization of many phenotypes of these lines. Such experiments form a baseline for future mechanistic studies in the same lines and a platform to test approaches to reverse phenotypes.

Individual claims and their strength & weaknesses:

(1) The authors developed mouse and zebrafish models of WAC deletion

They used the existing KOMP floxed WAC line to generate a null allele. For the mouse, there is a Western showing that it is indeed null for the protein. The fish data is less robustly validated - they don't confirm the allele in null at the protein or RNA level, and fish have two paralogs (waca and wacb), and this paper only characterizes one of these. So this evidence is less clear. The evaluated mice are heterozygous (Het), similar to patients, while the fish appear to be evaluated as homozygous mutants.

(2) The authors show that both species show altered craniofacial features

These data appear well powered, and the findings are robust.

(3) Each model altered GABAergic neurons

In mice, the authors stained with PV antibodies and saw a decrease in cells positive for this staining. A second marker, Lhx6, does not show a difference, suggesting this might be a change in PV expression rather than cell number. They could maybe look into the literature to see if this loss of just the protein also occurs in other models. Overall, the sample size here is a bit smaller than other parts of the paper (n=3), and the methods on the cell counts were less clear, so it is not as clear that this finding is as robust. The authors counted several other broad classes of cells, and those appear normal. Interestingly, there might also be some TBR1 mislocalization in layer 6 that might be significant with added power.

The fish data is based on an in situ hybridization for GAD. The measure shown is the width of the positive area in the forebrain. This measure is not one I have seen much before, and has potential to be driven by something unrelated to GABA (e.g., if the whole forebrain were simply a bit smaller). So this analysis could use a couple of other approaches (density of signal?) and/or a control probe for some other brain gene showing the measure is normal, and thus it is not just a size issue.

(4) Mice were more susceptible to the seizure-inducing agent PTZ

These data appear well powered, and the findings are robust. The authors also did a fair amount of useful electrophysiology that was all normal, but appeared to be well executed.

(5) Mice had changes in brain volume that interact with sex

The authors conducted an MRI on a good number of mice and reported a slight increase in global volume just in males. Sample size is fair, but the statistical approach here may be better if it puts males and females in the same model (to boost power and explicitly test for sex by genotype interaction that they report), and there is some chance that the brain region level differences that they report could include some false positives. They tested many regions, and it is not clear whether or not they corrected for the number of tests. Often, an FDR correction would be used in such imaging studies. It may be that only the most robust regional findings will survive those corrections. It is interesting data either way, but the analysis could be improved.

(6) Several behaviors are altered in the mice as well

These studies were fairly well-powered (n=15,16), and they found several positive and negative results, including alterations in memory and sociability in both species. There is a minor statistical flaw in the three-chamber analysis (they don't actually compare the Hets directly to the wildtypes in their statistical testing - a common mistake in neuroscience that should be addressed. But the data look like they will probably still be significant when correctly analyzed. In the supplement, the authors could do a bit more with the data they have to look at hyperactivity (i.e., show total motion in open field, not just time in center vs. periphery), and adding sex to their model might improve sensitivity for genotype effects.

(7) Some biochemical signaling pathways are altered in the brain

These are n=4 immunoblots, and show altered phospho ERK, but no changes in other signaling events predicted from prior WAC literature like H2B ubiquitination. They appear well done, and the authors share the full blots in the supplement.

(8) WAC deletion also alters gene expression in the brain

These studies were well-powered for RNAseq, with 10 and 14 samples, using neonates (P2), just the forebrain. The sequencing quality metrics all looked good, and the approach to analysis was okay. It would be stronger to again include sex in the model, rather than separate by sex. There were some typos in this part of the paper that made part of the conclusions unclear, but the RNAseq nicely confirmed the mutation of the mice, and discovered many differentially expressed genes, consistent with the role of this gene as a regulator of transcription. The presentation could be expanded to make more use of the data. Overall, though, this is a useful first characterization of the transcriptome in the line.

Reviewer #2 (Public review):

Summary:

Previous studies by some of the same authors of the actual manuscript showed that healthy human newborns memorize recently learned nonsense words. They exposed neonates to a familiarization period (several minutes) when multiple repetitions of a bisyllabic word were presented, uttered by the same speaker. Then they exposed neonates to an "interference period" when newborns listened to music or the same speaker uttering a different pseudoword. Finally, neonates were exposed to a test period when infants hear the familiarized word again. Interestingly, when the interference was music, the recognition of the word remained. The word recognition of the word was measured by using the NIRS technique, which estimates the regional brain oxygenation at the scalp level. Specifically, the brain response to the word in the test was reduced, unveiling a familiarity effect, while an increase in regional brain oxygenation corresponds to the detection of a "new word" due to a novelty effect. In previous studies, music does not erase the memory traces for a word (familiarity effect), while a different word uttered by the same speaker does.

The current study aims at exploring whether and how word memory is interfered with by other speech properties, specifically the changes in the speaker, while young children can distinguish speakers by processing the speech. The author's main hypothesis anticipates that new speaker recognition would produce less interference in the familiarized word because somehow neonates "separate" the processing of both words (familiarized uttered by one speaker, and interfering word, uttered by a different speaker), memorizing both words as different auditory events.

From my point of view, this hypothesis is interesting, since the results would contribute to estimating the role of the speaker in word learning and speech processing early in life.

Strengths:

(1) New data from neonates. Exploring neonates' cognitive abilities is a big challenge, and we need more data to enrich the knowledge of the early steps of language acquisition.

(2) The study contributes new data showing the role of speaker (recognition) on word learning (word memory), a quite unexplored factor. The idea that neonates include speakers in speech processing is not new, but its role in word memory has not been evaluated before. The possible interpretation is that neonates integrate the process of the linguistic and communicative aspects of speech at this early age.

(3) The study proposes a quite novel analytic approach. The new mixed models allow exploring the brain response considering an unbalanced design. More than the loss of data, which is frequent in infants' studies, the familiarization, interference and learning processes may take place at different moments of the experiment (e.g. related to changes in behavioural states along the experiment) or expressed in different regions (e.g. related to individual variations in optodes' locations and brain anatomy).

Weaknesses:

I did not find major weaknesses. However, I would like to have more discussion or explanation on the following points.

(1) It would be fine to report the contribution of each infant to the analysis, i.e. how many good blocks, 1 to 5 in sequence 1 and 2, were provided by each infant.

(2) Why did the factor "blocknumber" range from 0 to 4? The authors should explain what block zero means and why not 1 to 5.

(3) I may suggest intending to integrate the changes in brain activity across the 3 phases. That is, whether changes in familiarization relate to changes in the test and interference phases. For instance, in Figure 2, the brain response distinguishes between same and novel words that occurred over IFG and STG in both hemispheres. However, in the right STG there was no initial increase in the brain response, and the response for the same was higher than the one for novels in the 5th block.

(4) Similarly, it is quite amazing that the brain did not increase the activity with respect to the familiarization during the interference phase, mainly over the left hemisphere, even if both the word and speaker changed. Although the discussion considers these findings, an integrated discussion of the detection of novel words and the detection of a novel speaker over time may benefit from a greater integration of the results.

Appraisal:

The authors achieved their aims because the design and analytic approaches showed significant differences. The conclusions are based on these results. Specifically, the hypothesis that neonates would memorize words after interference, when interfered speech is pronounced by a different speaker, was supported by the data in blocks 2 and 5, and the potential mechanisms underlying these findings were discussed, such as separate processing for different speakers, likely related to the recognition of speaker identity.

I think the discussion is well-structured, although I may suggest integrating the changes into the three phases of the study. Maybe comparing with other regions, not related to speech processing.

Evaluating neonates is a challenge. Because physiology is constantly changing. For instance, in 9 minutes, newborns may transit from different behavioral states and experience different physiological needs.

This study offers the opportunity to inspire looking for commonalities and individual differences when investigating early memory capacities of newborns.

<< search

100k CIDs within a 3-hour window single thread

After winning against the Xiongnu, Emperor Wu sent Zhang Qian on another mission with 300 people. They brought valuable gifts like silk, gold, and lacquerware to give to the leaders of the Wusun, another nomadic group.

créances

Version 3 of this preprint has been peer-reviewed and recommended by Peer Community in Evolutionary Biology.<br /> See the peer reviews and the recommendation.

Do you see evidence of these actin-dense patches staying at the base of newly formed branchlets? I was curious whether you have observations that help distinguish patches functioning as nucleation sites for new protrusions versus representing increased endocytic activity in the nutrient-enriched conditions. Also, do you think these puncta might correspond to Arp2/3-mediated branched actin networks, or is that still unclear?

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

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

This is a good manuscript, well performed and well presented. I have several suggestions/questions to enhance the clarity of the concept, as technically the work is rather well performed.

1. I suggest that the authors explain better the mesenchymal-to-epithelial (MET) transition in reprogramming. Perhaps, explaining that epithelial gene acquisition (e.g., CDH1) and epidermal cell fate are not exactly the same. This approach could also be used to divide the genes they study further in their analyses.
2. KLF4 is both a repressor and an activator in different cell contexts including reprogramming. Does HIC2 act only as repressor? Is it possible that HIC2 is repressing KLF4-activated genes bad for reprogramming (including epidermal genes) and activating KLF4-suppressed genes ncessary for reprogramming? This should not be too difficult to explore with their current dataset and they also could look at available datasets for histone modifications in reprogramming.
3. Does HIC2 bind to genes related to somatic cell identify that need to be suppressed in reprogramming before the MET phase takes place?
4. Does HIC2 influence proliferation during reprogramming?

Referee cross-commenting

Comments by the other reviewers are sound and will help improve the manuscript.

Significance

In this manuscript, Kaji and colleagues perform a CRISPR/Cas9 screen to identify genes involved in mouse somatic cell reprogramming, identifying HIC2 as a target that they further validate. They conclude that HIC2 acts by repressing the epidermal/epithelial program induced by KLF4 during reprogramming. Studying the complex role of transcription factor interactions in the context of cell fate conversions (of any kind and not just somatic cell reprogramming) is highly relevant. This work helps clarify such complexity in a specific context but the work has wider conceptual implications.

We should move this to point 3, and point 3 should shift to point 4 and onwards. All of these items should be clickable.

Reviewer #1 (Public review):

Summary:

In this paper, the authors develop a biologically plausible recurrent neural network model to explain how the hippocampus generates and uses barcode-like activity to support episodic memory. They address key questions raised by recent experimental findings: how barcodes are generated, how they interact with memory content (such as place and seed-related activity), and how the hippocampus balances memory specificity with flexible recall. The authors demonstrate that chaotic dynamics in a recurrent neural network can produce barcodes that reduce memory interference, complement place tuning, and enable context-dependent memory retrieval, while aligning their model with observed hippocampal activity during caching and retrieval in chickadees.

Strengths:

(1) The manuscript is well-written and structured.

(2) The paper provides a detailed and biologically plausible mechanism for generating and utilizing barcode activity through chaotic dynamics in a recurrent neural network. This mechanism effectively explains how barcodes reduce memory interference, complement place tuning, and enable flexible, context-dependent recall.

(3) The authors successfully reproduce key experimental findings on hippocampal barcode activity from chickadee studies, including the distinct correlations observed during caching, retrieval, and visits.

(4) Overall, the study addresses a somewhat puzzling question about how memory indices and content signals coexist and interact in the same hippocampal population. By proposing a unified model, it provides significant conceptual clarity.

Weaknesses:

The recurrent neural network model incorporates assumptions and mechanisms, such as the modulation of recurrent input strength, whose biological underpinnings remain unclear. The authors acknowledge some of these limitations thoughtfully, offering plausible mechanisms and discussing their implications in depth. It may be worth exploring the robustness of the results to certain modeling assumptions. For instance, the choice to run the network for a fixed amount of time and then use the activity at the end for plasticity could be relaxed.

Author response:

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

Reviewer #1 (Public review):

Summary:

In this paper, the authors develop a biologically plausible recurrent neural network model to explain how the hippocampus generates and uses barcode-like activity to support episodic memory. They address key questions raised by recent experimental findings: how barcodes are generated, how they interact with memory content (such as place and seed-related activity), and how the hippocampus balances memory specificity with flexible recall. The authors demonstrate that chaotic dynamics in a recurrent neural network can produce barcodes that reduce memory interference, complement place tuning, and enable context-dependent memory retrieval, while aligning their model with observed hippocampal activity during caching and retrieval in chickadees.

Strengths:

(1) The manuscript is well-written and structured.

(2) The paper provides a detailed and biologically plausible mechanism for generating and utilizing barcode activity through chaotic dynamics in a recurrent neural network. This mechanism effectively explains how barcodes reduce memory interference, complement place tuning, and enable flexible, context-dependent recall.

(3) The authors successfully reproduce key experimental findings on hippocampal barcode activity from chickadee studies, including the distinct correlations observed during caching, retrieval, and visits.

(4) Overall, the study addresses a somewhat puzzling question about how memory indices and content signals coexist and interact in the same hippocampal population. By proposing a unified model, it provides significant conceptual clarity.

Weaknesses:

The recurrent neural network model incorporates assumptions and mechanisms, such as the modulation of recurrent input strength, whose biological underpinnings remain unclear. The authors acknowledge some of these limitations thoughtfully, offering plausible mechanisms and discussing their implications in depth.

One thread of questions that authors may want to further explore is related to the chaotic nature of activity that generates barcodes when recurrence is strong. Chaos inherently implies sensitivity to initial conditions and noise, which raises questions about its reliability as a mechanism for producing robust and repeatable barcode signals. How sensitive are the results to noise in both the dynamics and the input signals? Does this sensitivity affect the stability of the generated barcodes and place fields, potentially disrupting their functional roles? Moreover, does the implemented plasticity mitigate some of this chaos, or might it amplify it under certain conditions? Clarifying these aspects could strengthen the argument for the robustness of the proposed mechanism.

In our model, chaos is used to produce a random barcode when forming memories, but memory retrieval depends on attractor dynamics. Specifically, the plasticity update at the end of the cache creates an attractor state, and then afterwards for successful memory retrieval the network activity must settle into this attractor rather than remaining chaotic. This attractor state is a conjunction of memory content (place and seed activity) and memory index (barcode activity). Thus a barcode is ‘reactivated’ when network dynamics during retrieval settle into this cache attractor, or in other words chaotic dynamics do not need to generate the same barcode twice.

The reviewer raises an important point, which is how sensitivity to initial conditions and noise would affect the reliability of our proposed mechanism. The key question here is how noise will affect the network’s dynamics during retrieval. Would adding noise to the dynamics make memory retrieval more difficult? We thank the reviewer for suggesting we investigate this further, and below describe our experiments and changes to the manuscript to better address this topic.

We first experimented with adding independent gaussian distributed noise into each unit, drawn independently at each timestep. We analyzed recall accuracy using the same task and methods as Fig. 4F while varying the magnitude of noise. Memory recall was quite robust to this form of noise, even as the magnitude of noise approached half of the signal amplitude. This first experiment added noise into the temporal dynamics of the network. We subsequently examined adding static noise into the network inputs, which can also be thought of as introducing noise into initial conditions. Specifically, we added independent gaussian distributed noise into each unit, with the random value held constant for the extent of temporal dynamics. This perturbation decreased the likelihood of memory recall in a graded manner with noise magnitude, without dramatically changing the spatial profile. Examination of dynamics on individual trials revealed that the network failed to converge onto a cache attractor on some random fraction of trials, with other trials appearing nearly identical to noiseless results. We now include these results in the text and as a new supplementary figure, Figure S4AB.

To clarify the network dynamics and the purpose of chaos in our model, we make the following modifications in text:

Section 2.3, paragraph 2 (starting at “To store memories…”):

“…place inputs arrive into the RNN, recurrent dynamics generate an essentially random barcode, seed inputs are activated, and then Hebbian learning binds a particular pattern of barcode activity to place- and seed-related activity.”

Section 2.3, paragraph 3 (starting at “Memory recall in our network…”): As an example, consider a scenario in which an animal has already formed a memory at some location l, resulting in the storage of an attractor \vec{a} into the RNN. The attractor \vec{a} can be thought of as a linear combination of place input-driven activity $p(l)$, seed input-driven activity $s$, and a recurrent-driven barcode component $b$. Later, the animal returns to the same location and attempts recall (i.e. sets r \= 1, Figure 3B). Place inputs for location l drive RNN activity towards $p(l)$, which is partially correlated with attractor \vec{a}, and the recurrent dynamics cause network activity to converge onto attractor \vec{a}. In this way, barcode activity $b$ is reactivated, along with the place and seed components stored in the attractor state, $p(l)$ and $s$. The seed input can also affect recall, as discussed in the following section.

Section 2.4, final paragraph (starting “We further examined how model hyperparameters affected performance on these tasks”), added the following describing new results on adding noise: We found that adding noise to the network's temporal dynamics had little effect on memory recall performance (Figure S4A). However, large static noise vectors added to the network's input and initial state decreased the overall probability of memory recall, but not its spatial profile (Figure S4B).

It may also be worth exploring the robustness of the results to certain modeling assumptions.  For instance, the choice to run the network for a fixed amount of time and then use the activity  at the end for plasticity could be relaxed.

As described above, chaotic dynamics are necessary to generate a barcode during a cache, but not to reactivate that barcode during retrieval. During a successful memory retrieval, network activity settles into an attractor state and thus does not depend on the duration of simulated dynamics. The choice of duration to run dynamics during caching is important, but only insofar as activity significantly decorrelates from the initial state. We show in Figure S1B that decorrelation saturates ~t=25, and thus any random time point t > 25 would be similarly effective. We used a fixed duration runtime for caches only to avoid introducing unnecessary complication into our model.

Reviewer #2 (Public review):

Summary:

Striking experimental results by Chettih et al 2024 have identified high-dimensional, sparse patterns of activity in the chickadee hippocampus when birds store or retrieve food at a given site. These barcode-like patterns were interpreted as "indexes" allowing the birds to retrieve from memory the locations of stored food.

The present manuscript proposes a recurrent network model that generates such barcode activity and uses it to form attractor-like memories that bind information about location and food. The manuscript then examines the computational role of barcode activity in the model by simulating two behavioral tasks, and by comparing the model with an alternate model in which barcode activity is ablated.

Strengths of the study:

Proposes a potential neural implementation for the indexing theory of episodic memory - Provides a mechanistic model of striking experimental findings: barcode-like, sparse patterns of activity when birds store a grain at a specific location

A particularly interesting aspect of the model is that it proposes a mechanism for binding discrete events to a continuous spatial map, and demonstrates the computational advantages of this mechanism.

Weaknesses:

The relation between the model and experimentally recorded activity needs some clarification

The relation with indexing theory could be made more clear

The importance of different modeling ingredients and dynamical mechanisms could be made more clear

The paper would be strengthened by focusing on the most essential aspects

Comments:

The model distinguishes between "barcode activity" and "attractors". Which of the two corresponds to experimentally-recorded barcodes? I would presume the attractors. A potential issue is that the attractors are, as explained in the text (l.137), conjunctions of place activity, barcode activity and "seed" inputs. The fact that the seed activity is shared across attractors seems to imply that they have a non-zero correlation independent of distance. Is that the case in the model? If I understand correctly, Fig 3D shows correlations between an attractor and barcodes at different locations, but correlations between attractors at different locations are not shown. Fig 1 F instead shows that correlations between recorded retrieval activities decay to zero with distance.

More generally, the fact that the expression "barcode" is apparently used with different meanings in the model and in the experiments is potentially confusing (in the model they correspond to activity generating during caching, and this activity is distinct from the memories; my understanding is that in the experiments barcodes correspond to both caching and retrieval, but perhaps I am mistaken?).

Our intent is to use the expression “barcode” as similarly as possible between model and experimental work. The reviewer points out that the connection between barcodes in experimental and modeling work is unclear, as well as the relation of “attractors” in our model to previous experimental results. The meaning of ‘barcode’ is absolutely critical—we clarify below our intended meaning, and then describe changes to the manuscript to highlight this.

In experiments, we observed that activity during caching looked different than ordinary hippocampal activity (i.e. typical “place activity” observed during visits). Empirically there were two major differences. First, there was a pattern of neural activity which was present during every cache . This pattern was also present when birds visually inspected sites containing a cached seed, but not when visually inspecting an empty site. This is what we refer to as “seed activity”. Second, there was a pattern of neural activity which was unique to each cache. This pattern re-occurred during retrieval, and was orthogonal to place activity (see Fig. 1E-F). This is what we refer to as “barcode activity”. In summary, activity during a cache (or retrieval) contains a combination of three components: place activity, seed activity, and barcode activity.

These experimental findings are recapitulated in our model, as activity during a cache contains a combination of three components: place activity driven by place inputs, seed activity driven by seed inputs, and barcode activity generated by recurrent dynamics. Cache activity in the model corresponds to cache activity in experiments, and barcodes in the model correspond to barcodes in experiments. Our model additionally has “attractors”, meaning that network connectivity changes so that the activity generated during a simulated cache becomes an attractor state of network dynamics. “Attractors” refers to a feature of network dynamics, not a distinct activity state, and we do not yet know if these attractors exist in experimental data.

Figure 3D, as described in the figure legend, is a correlation of activity during cache and retrieval (in purple), for cache-retrieval pairs at the same or at different sites. We believe this is what the reviewer asks to see: the correlation between attractor states for different cache locations. The reviewer makes an important point: seed activity is shared across all attractors, so then why are correlations not high for all locations? This is because attractors also have a place component, which is anti-correlated for distant locations. This is evident in Fig. 3D by noticing that visit-visit correlations (black line, corresponding to place activity only) are negative for distant locations, and the correlation between attractors (purple line, cache-retrieval pairs) is subtly shifted up relative to the black line (place code only) for these distant locations. The size of this shift is due to the relative magnitude of place and seed inputs. For example, if we increase the strength of the seed input during caching (blue line), we can further increase the correlation between attractors even for quite distant sites:

Author response image 1.

To clarify the manuscript, we made the following modifications:

Section 2.2, first paragraph: We model the hippocampus as a recurrent neural network (RNN) (Alvarez and Squire, 1994; Tsodyks, 1999; Hopfield, 1982) and propose that recurrent dynamics can generate barcodes from place inputs. As in experiments, the model’s population activity during a cache should exhibit both place and barcode activity components.

Section 2.3, paragraph 3 (starting at “Memory recall in our network…”): As an example, consider a scenario in which an animal has already formed a memory at some location l , resulting in the storage of an attractor \vec{a} into the RNN . The attractor \vec{a} can be thought of as a linear combination of place input-driven activity $p(l)$, seed input-driven activity $s$, and a recurrent-driven barcode component $b$. Later, the animal returns to the same location and attempts recall (i.e. sets r \= 1, Figure 3B). Place inputs for l drive RNN activity towards $p(l)$, which is partially correlated with attractor \vec{a}, and the recurrent dynamics cause network activity to converge onto attractor \vec{a}. In this way, barcode activity $b$ is reactivated as part of attractor \vec{a}, along with the place and seed components stored in the attractor state, $p(l)$ and $s$. The seed input can also affect recall, as discussed in the following section.