eLife Assessment

Antibodies that selectively bind distinct amyloid-beta variants are vital tools for Alzheimer's disease research. This valuable manuscript aims to delineate the epitope specificity in a panel of anti-amyloid-beta antibodies, including some with clinical relevance. The experiments were rigorously conducted, employing an interesting combination of established and state-of-the-art methodologies, yielding mostly robust findings. While the data regarding antibody sequence preferences for distinct amyloid-beta regions and aggregation states are convincing, a thorough revision of the manuscript would help to highlight the key results.

Reviewer #1 (Public review):

The manuscript by Ivan et al aimed to identify epitopes on the Abeta peptide for a large set of anti-Abeta antibodies, including clinically relevant antibodies. The experimental work was well done and required a major experimental effort, including peptide mutational scanning, affinity determinations, molecular dynamics simulations, IP-MS, WB, and IHC. Therefore, it is of clear interest to the field. The first part of the work is mainly based on an assay in which peptides (15-18-mers) based on the human Abeta sequence, including some containing known PTMs, are immobilized, thus preventing aggregation. Although some results are in agreement with previous experimental structural data (e.g. for 3D6), and some responses to disease-associated mutations were different when compared to wild-type sequences (e.g. in the case of Aducanumab) - which may have implications for personalized treatment - I have concerns about the lack of consideration of the contribution of conformation (as in small oligomers and large aggregates) in antibody recognition patterns. The second part of the study used full-length Abeta in monomeric or aggregated forms to further investigate the differential epitope interaction between Aducanumab, donanemab, and lecanemab (Figures 5-7). Interestingly, these results confirmed the expected preference of these antibodies for aggregated Abeta, thus reinforcing my concerns about the conclusions drawn from the results obtained using shorter and immobilized forms of Abeta. Overall, I understand that the work is of interest to the field and should be published without the need for additional experimental data. However, I recommend a thorough revision of the structure of the manuscript in order to make it more focused on the results with the highest impact (second part).

Reviewer #2 (Public review):

This paper investigates binding epitopes of different anti-Abeta antibodies. Background information on the clinical outcome of some of the antibodies in the paper, which might be important for readers to know, is lacking. There are no references to clinical outcomes from antibodies that have been in clinical trials. This paper would be much more complete if the status of the antibodies were included. The binding characteristics of aducanumab, donanemab, and lecanemab should be compared with data from clinical phase 3 studies.

Aducanumab was identified at Neurimmune in Switzerland and licensed to Biogen and Eisai. Aducanumab was retracted from the market due to a very high frequency of the side-effect amyloid-related imaging abnormalities-edema (ARIA-E). Gantenerumab was developed by Roche and had two failed phase 3 studies, mainly due to a high frequency of ARIA-E and low efficacy of Abeta clearance. Lecanemab was identified at Uppsala University, humanized by BioArctic, and licensed to Eisai, who performed the clinical studies. Eisai and Biogen are now marketing lecanemab as Leqembi on the world market. Donanemab was developed by Ely Lilly and is sold in the US as Kisunla.

Limitations:

(1) Conclusions are based on Abeta antigens that may not be the primary targets for some conformational antibodies like aducanumab and lecanemab. There is an absence of binding data for soluble aggregated species.

(2) Quality controls and characterization of different Abeta species are missing. The authors need to verify if monomers remain monomeric in the blocking studies for Figures 5 and 6.

(3) The authors should discuss the limitations of studying synthetic Abeta species and how aggregation might hide or reveal different epitopes.

(4) The authors should elaborate on the differences between synthetic Abeta and patient-derived Abeta. There is a potential for different epitopes to be available.

eLife Assessment

This valuable study aims to advance the toolkit of small molecules used for approaches to targeted protein degradation for research and therapeutic applications. The authors provide solid data demonstrating the use of a high-throughput screen of small molecules to target a specific E3 ligase, KLHDC2 (Kelch-like homology domain containing protein 2); the resulting compounds then form the basis for new PROTAC (proteolysis targeting chimera) reagents. The strength of the work lies in expanding the PROTAC reagent inventory. The current work would be strengthened further by confirming that the PROTAC's activity is dependent on KLHDC2 and by a more thorough examination of off-target effects in cellular applications.

Reviewer #1 (Public review):

Summary:

The manuscript "Targeted Protein Degradation by KLHDC2 Ligands Identified by High Throughput Screening" by Zhou, H. et al. describes the development of a high-throughput FP-based screen and the identification of a KLHDC2 ligand from a small molecule library. A counter screen and other filtering criteria led to the identification of lead compounds that contained a tetrahydroquinoline scaffold. Commercially available analogs (52 compounds) that shared this scaffold were characterized by a KLHDC2 competitive binding assay. Optimized compounds were obtained that demonstrated improved potency and increased binding affinity by SPR. Docking of a lead candidate (compound 6) suggested it bound at a distal lipophilic site within the SelK binding pocket of KLHDC2. Based on this model, the authors then synthesized PROTACs that linked the KLHDC2 binder to a BRD4-binding molecule, JQ1. These PROTAC candidates possessed different linker configurations, and PROTAC 8 was able to cause BRD4 degradation in cells, with a half-maximal degradation concentration (DC50) of 80 nM. The authors demonstrate the identification and characterization of small-molecule KLHDC2 ligands that can be used to generate PROTACs that result in BRD4 degradation in cells.

Strengths:

The study by Zhou, H. et al. expands the E3 ligase toolkit by targeting KLHDC2 to identify ligands for PROTAC development, which has predominantly relied on VHL and CRBN. This was accomplished using a described FP-based high-throughput screening strategy (high Z' values in 1536 well format). Both target-specific and counter-specific assays were performed, along with subsequent stringent follow-up assays designed to address non-specific binding/specificity concerns. Label-free direct binding validations by SPR were used to determine binding affinity/kinetics. A strength of the study is the characterization of the interaction between candidate compounds and KLHDC2 versus related KEAP1.

Structural insight into the potential mode of binding was inferred by computational docking studies of the newly discovered KLHDC2 ligands. This was performed to identify where the identified scaffolds could be modified by linker incorporation for the design of PROTACs. The computational predictions were evaluated by linking a solvent-exposed site on the KLHDC2 ligand to JQ1. Three linkers were tested, and two compounds were found to result in BRD4 degradation in cells by HiBiT degradation assay and western blot. These findings demonstrate the feasibility of these compounds for the design of PROTAC-based degraders.

The authors present compelling KLHDC2 binding data for their lead compounds and demonstrate degradation of a target using a PROTAC strategy. Accordingly, the screening approach and compounds identified are likely to be of interest to the field and are likely to be generalizable to other PROTAC targets of interest.

Weaknesses:

The specificity of compounds for KLHDC2 was assessed by using a counter screen against KEAP1 and in vitro binding assays. However, off-target effects might occur in a cellular context, which weren't fully explored in the study. Notably, the authors do not demonstrate that the degradation induced by their PROTACs in cells is KLHDC2-dependent. A requirement for KLHDC2-mediated degradation could be evaluated, for example, by using knockout/knockdown of KLHDC2, or other means, to demonstrate specificity. Addressing specificity is deemed important to evaluate the proposed PROTAC mechanism of action in a cellular context that results in the degradation of BRD4. Specificity is important when considering the utility of these new compounds for PROTAC design.

Additional rationale behind the selection of linkers used to generate candidate PROTACs would be informative and would benefit from additional discussion and/or citation. The reasons for the lack of activity, such as for compound 9, were not fully explored or discussed, such as whether complex assembly is potentially affected by linker choice. Perhaps related to this point, the authors note that a trifluoromethoxy group increased the binding affinity of compound 6. However, the subsequent docking analysis revealed this moiety to be solvent-exposed. The relationship between this site of functionalization, linker selection, and the resulting binding affinity or effect on DC50 was not clear and/or could be developed further.

Minor issues related to the presentation of the manuscript include sections that would benefit from either additional citation and/or description, such as the KI-696 inhibitor used and the BRD4 HiBiT degradation assay that was used to assess PROTAC potency. Figure captions should be reviewed to ensure that the number of independent experiments is indicated, and what data points and error bars represent, as these are not indicated in several figures. BRD4 levels were quantified in 4E; however, error/reproducibility (n) is not indicated.

Reviewer #2 (Public review):

PROTACs are a class of small molecules that induce an interaction between a target protein and a ubiquitin ligase, thereby leading to the target protein's ubiquitination and subsequent proteasomal degradation. Given that the vast majority of PROTACs rely on the cereblon and VHL ubiquitin ligases, a major goal within this field has been to identify and develop ligands for additional ubiquitin ligases, in particular those whose expression affords tissue or subcellular specificity or those whose structure allows them to degrade targets that are otherwise incompatible with cereblon or VHL.

In this work, Zhou and colleagues from the Bollong group at Scripps utilize a high-throughput fluorescence polarization screen of >350,000 compounds to identify and optimize a novel ligand for KLHDC2, a ubiquitin ligase which had previously been discovered to be capable of proximity-induced degradation of target proteins. Zhou et al go on to show that this ligand can be used as the basis for PROTACs capable of degrading BRD4 in a cell line. Of note, prior to this paper, three other groups had also developed ligands to KLHDC2 and used them to generate active PROTACs. Interestingly, docking studies by Zhou suggest that their compound may bind to a different region of the KLHDC2's kelch domain.

The major strengths of this work are its brevity and the clarity of the writing and figures. Their claim that they have discovered a ligand for KLHDC2, which can be used to develop BRD4-degrading PROTACs, is well-supported by their findings from the screen, SPR, and cellular assays. The weakness of the work then, is not so much relevant to the paper at hand but rather stems from the fact that their story leaves me wanting to know more. Indeed, there are a number of interesting experiments that we need as a field in order to assess 1) how generalizable their findings are across cell lines and targets, and 2) how this new KLHDC2 ligand stacks up against the other recently discovered ligands for KLDHC2 as well as the existing standards, cereblon and VHL.

Nonetheless, Zhou and colleagues provide a valuable addition to the emerging repertoire of KLHDC2 ligands, and I'm certain that with time, we will come to understand what ligands work best for KLHDC2-based PROTACs and how they compare to the growing set of ubiquitin ligases in our armamentarium.

eLife Assessment

This study presents a valuable finding on the role of Slit-Robo signaling in cardiac innervation. The evidence supporting the main claims of the authors is solid. The use of several mouse models including constitutive and cell type specific knockout models make the findings more robust. The scope of the presented studies is somewhat limited, as they primarily focus on evaluating the phenotypic changes in cardiac innervation following the loss of various Slit or Robo genes.

Reviewer #1 (Public review):

The study aims to determine the role of Slit-Robo signaling in the development and patterning of cardiac innervation, a key process in heart development. Despite the well-studied roles of Slit axon guidance molecules in the development of the central nervous system, their roles in the peripheral nervous system are less clear. Thus, the present study addresses an important question. The study uses genetic knockout models to investigate how Slit2, Slit3, Robo1, and Robo2 contribute to cardiac innervation.

Using constitutive and cell type-specific knockout mouse models, they show that the loss of endothelial-derived Slit2 reduces cardiac innervation. Additionally, Robo1 knockout, but not Robo2 knockout, recapitulated the Slit2 knockout effect on cardiac innervation, leading to the conclusion that Slit2-Robo1 signaling drives sympathetic innervation in the heart. Finally, the authors also show a reduction in isoproterenol-stimulated heart rate but not basal heart rate in the absence of endothelial Slit2.

The conclusions of this paper are mostly well supported by the data, but some should be modified to account for the study's limitations and discussed in the context of previous literature.

(1) It is well established that Slit ligands undergo proteolytic cleavage, generating N- and C-terminal fragments with distinct biological functions. Full-length Slit proteins and their fragments differ in cell association, with the N-terminal fragment typically remaining membrane-bound, while the C-terminal fragment is more diffusible. This distinction is crucial when evaluating the role of Slit proteins secreted by different cell types in the heart. However, this study does not examine or discuss the specific contributions of different Slit2 fragments, limiting its mechanistic insight into how Slit2 regulates cardiac innervation.

(2) The endothelial-specific deletion of Slit2 leads to its loss in endothelial cells across various organs and tissues in the developing embryo. Therefore, the phenotypes observed in the heart may be influenced by defects in other parts of the embryo, such as the CNS or sympathetic ganglia, and this possibility cannot be ruled out.

Reviewer #2 (Public review):

The aims of investigating Slit-Robo signaling in cardiac innervation were achieved by the experiments designed. While questions remain regarding signal regulation and interplay between established axon guidance signals and further role of other Slit ligands and Robo expression in endothelium, the results strongly support the conclusions drawn.

Writing and presentation are easy to follow and well structured, Appropriate controls are used, statistical analysis applied appropriately, and experiments directly test aims following a logical story.

The authors demonstrate a novel mechanism for Slit-Robo signaling in cardiac sympathetic innervation. The data establishes a framework for future studies.

Recommendations:

Further assessment of interplay between Slit ligands as well as other signaling pathways (Semaphorin, NGF, etc) could be investigated. Is it possible to rescue the phenotype by modulation of other signaling pathways? Can combined Slit2/Slit3 KO rescue? Additionally, as the authors state, conditional Robo1 knockouts will be important to validate the findings of constitutive knockout.

eLife Assessment

This study presents the important finding that lysosomal damage triggers inflammatory signaling through ubiquitination and the TAB-TAK1-IKK-NF-kB axis. The data obtained from the unbiased transcriptomic and proteomic analyses are convincing and provide invaluable information to the field. Although further experiments will be required to clarify how TAB2/3 are activated, this work will be of interest to researchers in the fields of organelle biology and inflammation.

Reviewer #1 (Public review):

Summary:

Lysosomal damage is commonly found in many diseases including normal aging and age-related disease. However, the transcriptional programs activated by lysosomal damage have not been thoroughly characterized. This study aimed to investigate lysosome damage-induced major transcriptional responses and the underlying signaling basis. The authors have convincingly shown that lysosomal damage activates a ubiquitination-dependent signaling axis involving TAB, TAK1, and IKK, which culminates in the activation of NF-kB and subsequent transcriptional upregulation of pro-inflammatory genes and pro-survival genes. Overall, the major aims of this study were successfully achieved.

Strengths:

This study is well-conceived and strictly executed, leading to clear and well-supported conclusions. Through unbiased transcriptomics and proteomics screens, the authors identified NF-kB as a major transcriptional program activated upon lysosome damage. TAK1 activation by lysosome damage-induced ubiquitination was found to be essential for NF-kB activation and MAP kinase signaling. The transcriptional and proteomic changes were shown to be largely driven by TAK1 signaling. Finally, the TAK1-IKK signaling was shown to provide resistance to apoptosis during lysosomal damage response. The main signaling axis of this pathway was convincingly demonstrated.

Weaknesses:

One weakness was the claim of K63-linked ubiquitination in lysosomal damage-induced NF-kB activation. While it was clear that K63 ubiquitin chains were present on damaged lysosomes, no evidence was shown in the current study to demonstrate the specific requirement of K63 ubiquitin chains in the signaling axis being studied. Clarifying the roles of K63-linked versus other types of ubiquitin chains in lysosomal damage-induced NF-kB activation may improve the mechanistic insights and overall impact of this study.

Another weakness was that the main conclusions of this study were all dependent on an artificial lysosomal damage agent. It will be beneficial to confirm key findings in other contexts involving lysosomal damage.

Reviewer #2 (Public review):

Summary:

Endo et al. investigate the novel role of ubiquitin response upon lysosomal damage in activating cellular signaling for cell survival. The authors provide a comprehensive transcriptome and proteome analysis of aging-related cells experiencing lysosomal damage, identifying transcription factors involved in transcriptome and proteome remodeling with a focus on the NF-κB signaling pathway. They further characterized the K63-ubiquitin-TAB-TAK1-NF-κB signaling axis in controlling gene expression, inflammatory responses, and apoptotic processes.

Strengths:

In the aging-related model, the authors provide a comprehensive transcriptome and characterize the K63-ubiquitin-TAB-TAK1-NF-κB signaling axis. Through compelling experiments and advanced tools, they elucidate its critical role in controlling gene expression, inflammatory responses, and apoptotic processes.

Weaknesses:

The study lacks deeper connections with previous research, particularly:<br /> • The established role of TAB-TAK1 in AMPK activation during lysosomal damage<br /> • The potential significance of TBK1 in NF-κB signaling pathways

Reviewer #3 (Public review):

Summary:

The response to lysosomal damage is a fast-moving and timely field. Besides repair and degradation pathways, increasing interest has been focusing on damaged-induced signaling. The authors conducted both transcriptomics and proteomics to characterize the cellular response to lysosomal damage. They identify a signaling pathway leading to activation of NFkappaB. Based on this and supported by Western blot and microscopy data, the authors nicely show that TAB2/3 and TAK1 are activated at damaged lysosomes and kick off the pathway to alter gene expression, which induces cytokines and protect from cell death. TAB2/3 activation is proposed to occur through K63 ubiquitin chain formation. Generally, this is a careful and well conducted study that nicely delineates the pathway under lysosomal stress. The "omics" data serves as a valuable resource for the field. More work should be invested into how TAB2/3 are activated at the damaged lysosomes, also to increase novelty in light of previous reports.

Strengths:

Generally, this is a careful and well-conducted study that nicely delineates the pathway under lysosomal stress. The "omics" data serves as a valuable resource for the field.

Weaknesses:

More work should be invested into how TAB2/3 are activated at the damaged lysosomes, also to increase novelty in light of previous reports. Moreover, different damage types should be tested to probe relevance for different pathophysiological conditions.

Suggestions:

(1) A recent paper claims that NFkappaB is activated by Otulin/M1 chains upon lysosome damage through TBK1 (PMID: 39744815). In contrast, Endo et al. nicely show that ubiquitylation is needed (shown by TAK-243) for NFkB activation but only have correlative data to link it specifically to K63 chains. On page 15, line 11, the authors even argue a "potential" involvement of K63. This point should be better dealt with. Can the authors specifically block K63 formation? K63R overexpression or swapping would be one way. Is the K63 ligase ITCH involved (PMID: 38503285) or any other NEDD4-like ligase? This could be compared to LUBAC inhibition. Also, the point needs to be dealt with more controversially in the discussion as these are alternative claims (M1 vs K63, TAB vs TBK1).

(2) It would be interesting to know what the trigger is that induces the pathway. Lipid perturbation by LLOMe is a good model, but does activation also occur with GPN (osmotic swelling) or lipid peroxidation (oxidative stress) that may be more broadly relevant in a pathophysiological way? Moreover, what damage threshold is needed? Does loss of protons suffice? Can activation be induced with a Ca2+ agonist in the absence of damage?

(3) The authors nicely define JNK and p38 activation. This should be emphasized more, possibly also in the abstract, as it may contribute to the claim of increased survival fitness.

eLife Assessment

This important study describes newly identified light-gated ion channel homologs (channelrhodopsins, ChRs) in several protist species, with a primary focus on the biophysical characterization of ChRs of ancyromonads. The authors employed a powerful combination of bioinformatics, manual and automated patch-clamp electrophysiology, absorption spectroscopy, and flash photolysis. Additionally, they evaluated the applicability of the newly discovered anion-conducting ChRs in cortical neurons of mouse brain slices and in living C. elegans worms. The evidence supporting most of the claims is convincing and this work will be of interest to the microbial rhodopsin community and neuro- and cardioscientists utilizing optogenetics in their research.

Reviewer #1 (Public review):

Summary:

This work by Govorunova et al. identified three naturally blue-shifted channelrhodopsins (ChRs) from ancyromonads, namely AnsACR, FtACR, and NlCCR. The phylogenetic analysis places the ancyromonad ChRs in a distinct branch, highlighting their unique evolutionary origin and potential for novel applications in optogenetics. Further characterization revealed the spectral sensitivity, ionic selectivity, and kinetics of the newly discovered AnsACR, FtACR, and NlCCR. This study also offers valuable insights into the molecular mechanism underlying the function of these ChRs, including the roles of specific residues in the retinal-binding pocket. Finally, this study validated the functionality of these ChRs in both mouse brain slices (for AnsACR and FtACR) and in vivo in Caenorhabditis elegans (for AnsACR), demonstrating the versatility of these tools across different experimental systems.

In summary, this work provides a potentially valuable addition to the optogenetic toolkit by identifying and characterizing novel blue-shifted ChRs with unique properties.

Strengths:

This study provides a thorough characterization of the biophysical properties of the ChRs and demonstrates the versatility of these tools in different ex vivo and in vivo experimental systems. The mutagenesis experiments also revealed the roles of key residues in the photoactive site that can affect the spectral and kinetic properties of the channel.

Weaknesses:

While the novel ChRs identified in this work are spectrally blue-shifted, there still seems to be some spectral overlap with other optogenetic tools. The authors should provide more evidence to support the claim that they can be used for multiplex optogenetics and help potential end-users assess if they can be used together with other commonly applied ChRs. Additionally, further engineering or combination with other tools may be required to achieve truly orthogonal control in multiplexed experiments.

In the C. elegans experiments, partial recovery of pharyngeal pumping was observed after prolonged illumination, indicating potential adaptation. This suggests that the effectiveness of these ChRs may be limited by cellular adaptation mechanisms, which could be a drawback in long-term experiments. A thorough discussion of this challenge in the application of optogenetics tools would prove very valuable to the readership.

Reviewer #2 (Public review):

Summary:

Govorunova et al present three new anion opsins that have potential applications in silencing neurons. They identify new opsins by scanning numerous databases for sequence homology to known opsins, focusing on anion opsins. The three opsins identified are uncommonly fast, potent, and are able to silence neuronal activity. The authors characterize numerous parameters of the opsins.

Strengths:

This paper follows the tradition of the Spudich lab, presenting and rigorously characterizing potentially valuable opsins. Furthermore, they explore several mutations of the identified opsin that may make these opsins even more useful for the broader community. The opsins AnsACR and FtACR are particularly notable, having extraordinarily fast onset kinetics that could have utility in many domains. Furthermore, the authors show that AnsACR is usable in multiphoton experiments having a peak photocurrent in a commonly used wavelength. Overall, the author's detailed measurements and characterization make for an important resource, both presenting new opsins that may be important for future experiments, and providing characterizations to expand our understanding of opsin biophysics in general.

Weaknesses:

First, while the authors frequently reference GtACR1, a well-used anion opsin, there is no side-by-side data comparing these new opsins to the existing state-of-the-art. Such comparisons are very useful to adopt new opsins.

Next, multiphoton optogenetics is a promising emerging field in neuroscience, and I appreciate that the authors began to evaluate this approach with these opsins. However, a few additional comparisons are needed to establish the user viability of this approach, principally the photocurrent evoked using the 2p process, for given power densities. Comparison across the presented opsins and GtACR1 would allow readers to asses if these opsins are meaningfully activated by 2P.

Reviewer #3 (Public review):

Summary:

The authors aimed to develop Channelrhodopsins (ChRs), light-gated ion channels, with high potency and blue action spectra for use in multicolor (multiplex) optogenetics applications. To achieve this, they performed a bioinformatics analysis to identify ChR homologues in several protist species, focusing on ChRs from ancyromonads, which exhibited the highest photocurrents and the most blue-shifted action spectra among the tested candidates. Within the ancyromonad clade, the authors identified two new anion-conducting ChRs and one cation-conducting ChR. These were characterized in detail using a combination of manual and automated patch-clamp electrophysiology, absorption spectroscopy, and flash photolysis. The authors also explored sequence features that may explain the blue-shifted action spectra and differences in ion selectivity among closely related ChRs.

Strengths:

A key strength of this study is the high-quality experimental data, which were obtained using well-established techniques such as manual patch-clamp and absorption spectroscopy, complemented by modern automated patch-clamp approaches. These data convincingly support most of the claims. The newly characterized ChRs expand the optogenetics toolkit and will be of significant interest to researchers working with microbial rhodopsins, those developing new optogenetic tools, as well as neuro- and cardioscientists employing optogenetic methods.

Weaknesses:

This study does not exhibit major methodological weaknesses. The primary limitation of the study is that it includes only a limited number of comparisons to known ChRs, which makes it difficult to assess whether these newly discovered tools offer significant advantages over currently available options. Additionally, although the study aims to present ChRs suitable for multiplex optogenetics, the new ChRs were not tested in combination with other tools. A key requirement for multiplexed applications is not just spectral separation of the blue-shifted ChR from the red-shifted tool of interest but also sufficient sensitivity and potency under low blue-light conditions to avoid cross-activation of the respective red-shifted tool. Future work directly comparing these new ChRs with existing tools in optogenetic applications and further evaluating their multiplexing potential would help clarify their impact.

eLife Assessment

This valuable study examines the role of map3k1, a MAP3K family member that has both kinase and ubiquitin ligase domains, in the differentiation of progenitors in the flatworm Planaria. The convincing analyses demonstrate that map3k1 acts within progenitors to restrict their premature differentiation and to prevent formation of teratomas. This work would be of interest to researchers in the fields of regeneration, developmental biology, and aging.

Reviewer #1 (Public review):

Summary:

The authors assess the role of map3k1 in adult Planaria through whole body RNAi for various periods of time. The authors' prior work has shown that neoblasts (stem cells that can regenerate the entire body) for various tissues are intermingled in the body. Neoblasts divide to produce progenitors that migrate within a "target zone" to the "differentiated target tissues" where they differentiate into a specific cell type. Here the authors show that map3k1-i animals have ectopic eyes that form along the "normal" migration path of eye progenitors (Fig. 1), ectopic neurons and glands along the AP axis (Fig. 2) and pharynx in ectopic anterior positions (Fig. 3). The rest of the study show that positional information is largely unaffected by loss of map3k1 (Fig. 4,5). However, loss of map3k1 leads to premature differentiated of progenitors along their normal migratory route (Fig. 6). They also show that an ill-defined "long-term" whole body depletion of map3k1 results in mis-specified organs and teratomas.

Strengths:

(1) The study has appropriate controls, sample sizes and statistics.<br /> (2) The work appears to be high-quality.<br /> (3) The conclusions are supported by the data.<br /> (4) Planaria is a good system to analyze the function of map3k1, which exists in mammals but not in other invertebrates.

Weaknesses:

(1) The paper is largely descriptive with no mechanistic insights.<br /> (2) Given the severe phenotypes of long-term depletion of map3k1, it is important that this exact timepoint is provided in the methods, figures, figure legends and results.<br /> (3) Fig. 1C, the ectopic eyes are difficult to see, please add arrows.<br /> (4) line 217 - why does the n=2/12 animals not match the values in Fig. 3B, which is 11/12 and 12/12. The numbers don't add up. Please correct/explain.<br /> (5) Figure panels do not match what is written in the results section. There is no Fig. 6E. Please correct.

Reviewer #2 (Public review):

Summary:<br /> The flatworm planarian Schmidtea mediterranea is an excellent model for understanding cell fate specification during tissue regeneration and adult tissue maintenance. Planarian stem cells, known as neoblasts, are continuously deployed to support cellular turnover and repair tissues damaged or lost due to injury. This reparative process requires great precision to recognize the location, timing, and cellular fate of a defined number of neoblast progeny. Understanding the molecular mechanisms driving this process could have important implications for regenerative medicine and enhance our understanding of how form and function are maintained in long-lived organisms such as humans. Unfortunately, the molecular basis guiding cell fate and differentiation remains poorly understood.

In this manuscript, Canales et al. identified the role of the map3k1 gene in mediating the differentiation of progenitor cells at the proper target tissue. The map3k1 function in planarians appears evolutionarily conserved as it has been implicated in regulating cell proliferation, differentiation, and cell death in mammals. The results show that the downregulation of map3k1 with RNAi leads to spatial patterning defects in different tissue types, including the eye, pharynx, and the nervous system. Intriguingly, long-term map3k1-RNAi resulted in ectopic outgrowths consistent with teratomas in planarians. The findings suggest that map3k1 mediates signaling, regulating the timing and location of cellular progenitors to maintain correct patterning during adult tissue maintenance.

Strengths:

The authors provide an entry point to understanding molecular mechanisms regulating progenitor cell differentiation and patterning during adult tissue maintenance.

The diverse set of approaches and methods applied to characterize map3k1 function strengthens the case for conserved evolutionary mechanisms in a selected number of tissue types. The creativity using transplantation experiments is commendable, and the findings with the teratoma phenotype are intriguing and worth characterizing.

Weaknesses:

The article presents a provocative idea related to the importance of positional control for organs and cells, which is at least in part regulated by map3k1. Nonetheless, the role of map3k1 or its potential interaction with regulators of the anterior-posterior, mediolateral axes, and PCGs is somewhat superficial. The authors could elaborate or even speculate more in the discussion section and the different scenarios incorporating these axial modulators into the map3k1 model presented in Figure 8.

The article can be improved by addressing inconsistencies and adding details to the results, including the main figures and supplements. This represents one of the most significant weaknesses of this otherwise intriguing manuscript. Below are some examples of a few figures, but the authors are expected to pay close attention to the remaining figures in the paper.

Details associated with the number of animals per experiment, statistical methods used, and detailed descriptions of figures appear inconsistent or lacking in almost all figures. In some instances, the percentage of animals affected by the phenotype is shown without detailing the number of animals in the experiment or the number of repeats. Figures and their legends throughout the paper lack details on what is represented and sometimes are mislabeled or unrelated. Specifically, the arrows in Figure 1A are different colors. Still, no reasoning is given for this, and in the exact figure, the top side (1A) shows the percentages and the number of animals below. Conversely, in Figures 1B, C, and D, no details on the number of animals or percentages are shown, nor an explanation of why opsin was used in Figure 1A but not 1B. Is Figure 1B missing an image for the respective control? Figure 1C needs details regarding what the two smaller boxes underneath are. Figure 1C could use an AP labeling map in 10 days (e.g., AP6 has one optic cup present). Figure 1C and F counts do not match. In Figure 1C, we do not know the number of animals tested, controls used, the scale bar sizes in the first two images, nor the degree of magnification used despite the pharynx region appearing magnified in the second image. Figure 1C is also shown out of chronological order; 36 days post RNAi is shown before 10 days post RNAi. Moreover, the legends for Figures 1C and 1D are swapped.

Additionally, Figure 1F and many other figures throughout the paper lack overall statistical considerations. Furthermore, Figure 1F has three components, but only one is labeled. Labeling each of them individually and describing them in the corresponding figure legend may be more appropriate.

Figure 2C shows images of gene expression for two genes, but the counts are shown for only one in Figure 2D. It is challenging to follow the author's conclusions without apparent reasoning and by only displaying quantitative considerations for one case but not the other. These inconsistencies are also observed in different figures. In Figure 2D, 24/24 animals were reported to show the phenotype, but only eight were counted (is there a reason for this?). In Figure 2E, the expression for three genes is shown, with some displaying anterior and posterior regions while others only show the anterior picture. Is there a particular reason for this? Also, in Figure 2F, the counts are shown for only the posterior region of two genes out of the three displayed in Figure 2E. It is unclear why the authors do not show counts for the anterior areas considered in Figure 2E. Furthermore, the legend for Figure 2D is missing, and the legend for 2F is mislabeled as a description for Figure 2D.

Supplement Figure 1 B reports data up to 6 weeks, but no text in the manuscript or supplement mentions any experiment going up to 6 weeks. There are no statistics for data in Supplement Figure 1E. Any significance between groups is unclear.

eLife Assessment

This important study is a first report investigating the boundary formation between sensory and non-sensory tissues of the inner ear, which has broad relevance to the developmental field in general. All three reviewers thought the results and data analyses presented are solid. However, the causal relationship between the morphological evidence and the role of Lmx1a is not well supported by the results. The mechanism linking Lmx1a to ROCK is also incomplete, considering ROCK is involved in so many processes.

Reviewer #1 (Public review):

Summary:

This manuscript investigated the mechanism underlying boundary formation necessary for proper separation of vestibular sensory end organs. In both chick and mouse embryos, it was shown that a population of cells abutting the sensory (marked by high Sox2 expression) /nonsensory cell populations (marked by Lmx1a expression) undergo apical expansion, elongation, alignment and basal constriction to separate the lateral crista (LC) from the utricle. Using Lmx1a mouse mutant, organ cultures, pharmacological and viral-mediated Rock inhibition, it was demonstrated that the Lmx1a transcription factor and Rock-mediated actomyosin contractility is required for boundary formation and LC-utricle separation.

Strengths:

Overall, the morphometric analyses were done rigorously and revealed novel boundary cell behaviors. The requirement of Lmx1a and Rock activity in boundary formation was convincingly demonstrated.

Weaknesses:

However, the precise roles of Lmx1a and Rock in regulating cell behaviors during boundary formation were not clearly fleshed out. For example, phenotypic analysis of Lmx1a was rather cursory; it is unclear how Lmx1a, expressed in half of the boundary domain, control boundary cell behaviors and prevent cell mixing between Lmx1a+ and Lmx1a- compartments? Well-established mechanisms and molecules for boundary formation were not investigated (e.g. differential adhesion via cadherins, cell repulsion via ephrin-Eph signaling). Moreover, within the boundary domain, it is unclear whether apical multicellular rosettes and basal constrictions are drivers of boundary formation, as boundary can still form when these cell behaviors were inhibited. Involvement of other cell behaviors, such as radial cell intercalation and oriented cell division, also warrant consideration. With these lingering questions, the mechanistic advance of the present study is somewhat incremental.

Reviewer #2 (Public review):

Summary:

Chen et al. describe the mechanisms that separate the common pan-sensory progenitor region into individual sensory patches, which presage the formation of the sensory epithelium in each of the inner ear organs. By focusing on the separation of the anterior and then lateral cristae, they find that long supra-cellular cables form at the interface of the pan-sensory domain and the forming cristae. They find that at these interfaces, the cells have a larger apical surface area, due to basal constriction, and Sox2 is down-regulated. Through analysis of Lmx1 mutants, the authors suggest that while Lmx1 is necessary for the complete segregation of the sensory organs, it is likely not necessary for the initial boundary formation, and the down-regulation of Sox2.

Strengths:

The manuscript adds to our knowledge and provides valuable mechanistic insight into sensory organ segregation. Of particular interest are the cell biological mechanisms: The authors show that contractility directed by ROCK is important for the maintenance of the boundary and segregation of sensory organs.

Weaknesses:

The manuscript would benefit from a more in-depth look at contractility - the current images of PMLC are not too convincing. Can the authors look at p or ppMLC expression in an apical view? Are they expressed in the boundary along the actin cables? Does Y-27362 inhibit this expression?

The authors suggest that one role for ROCK is the basal constriction. I was a little confused about basal constriction. Are these the initial steps in the thinning of the intervening non-sensory regions between the sensory organs? What happens to the basally constricted cells as this process continues?

The steps the authors explore happen after boundaries are established. This correlates with a down-regulation of Sox2, and the formation of a boundary. What is known about the expression of molecules that may underlie the apparent interfacial tension at the boundaries? Is there any evidence for differential adhesion or for Eph-Ephrin signalling? Is there a role for Notch signalling or a role for Jag1 as detailed in the group's 2017 paper?

A comment on whether cellular intercalation/rearrangements may underlie some of the observed tissue changes.

The change in the long axis appears to correlate with the expression of Lmx1a (Fig 5d). The authors could discuss this more. Are these changes associated with altered PCP/Vangl2 expression?

Reviewer #3 (Public review):

Summary:

Lmx1a is an orthologue of apterous in flies, which is important for dorsal-ventral border formation in the wing disc. Previously, this research group has described the importance of the chicken Lmx1b in establishing the boundary between sensory and non-sensory domains in the chicken inner ear. Here, the authors described a series of cellular changes during border formation in the chicken inner ear, including alignment of cells at the apical border and concomitant constriction basally. The authors extended these observations to the mouse inner ear and showed that these morphological changes occurred at the border of Lmx1a positive and negative regions, and these changes failed to develop in Lmx1a mutants. Furthermore, the authors demonstrated that the ROCK-dependent actomyosin contractility is important for this border formation and blocking ROCK function affected epithelial basal constriction and border formation in both in vitro and in vivo systems.

Strengths:

The morphological changes described during border formation in the developing inner ear are interesting. Linking these changes to the function of Lmx1a and ROCK dependent actomyosin contractile function are provocative.

Weaknesses:

There are several outstanding issues that need to be clarified before one could pin the morphological changes observed being causal to border formation and that Lmx1a and ROCK are involved.

eLife Assessment

The first part of this manuscript describes an interdisciplinary approach to mine the human channelome and discover further ion channel orthologues across diverse organisms. Although the findings and data curation enabled by the new approach are valuable to the ion channel community, as well as to those interested in improved methods for mining sequence space for their protein of interest, this part of the work is incomplete because critical methodological information is missing. Further validation of the improvements this approach shows over others is needed. The second part of the manuscript utilizes the approach described in the first part to delineate co-conserved amino acid patterns in CALHM channels, but the evidence provided to support the role of the identified residues in channel gating is currently inadequate.

Reviewer #1 (Public review):

Summary:

In this manuscript, Taujale et al describe an interdisciplinary approach to mine the human channelome and further discover orthologues across diverse organisms, culminating in delineating co-conserved patterns in an example ion channel: CALHM. Overall, this paper comes in two sections, one where 419 human ion channels and 48,000+ channels from diverse organisms are found through a multidisciplinary data mining approach, and a second where this data is used to find co-conserved sequences, whose functional significance is validated via experiments on CALHM1 and CALHM6. Overall, this is an intriguing data-first approach to better understand even understudied ion channels like CALHM6. However, more needs to be done to pull this story together into a single, coherent narrative.

Strengths:

This manuscript takes advantage of modern-day LLM tools to better mine the literature for ion channel sequences in humans and other species with orthologous ion channel sequences. They explore the 'dark channome' of understudied ion channels to better reveal the information evolution has to tell us about our own proteins, and illustrate the information this provides access to in experimental studies in the final section of the paper. Finally, they provide a wealth of information in the supplementary tables (in the form of Excel spreadsheets) for others to explore. Overall, this is a creative approach to a wide-reaching problem that can be applied to other families of proteins.

Weaknesses:

Overall, while a considerable amount of work has been done for this manuscript, the presentation, both in terms of writing and figures, leaves much to be desired. One can imagine a story that clearly describes the need for a better-curated sequence database of ion channels, and clearly describes how existing resources fall short, but here this is not very clearly illustrated.

One question that arises with the part of the manuscript that discusses the identification and classification of ion channels is whether they plan to make these sequences available to the wider public. For the 419 human sequences, making a small database to share this result so that these sequences can be easily searched and downloaded would be desirable. There are a variety of acceptable formats for this: GitHub/figshare/zenodo/university website that allows a wider community to access their hard work. The authors have included enough information in the supplementary tables that this could be done by a motivated reader, but providing such a resource would greatly expand the impact of this paper. The same question can be asked of the 48,000+ ion channels from diverse organisms. For these, one is even worried that these are not properly sequenced genes? What checks have been done to confirm this? Uniport contains a good deal of unreviewed sequences, especially from single-celled organisms. Potentially, this is covered in the sentence in the Methods: "Finally, the results obtained from both the full-length and pore domains were retained as true orthologous relationships to remove extraneous hits." But this process could be discussed in more detail, clearly illustrating that the risk of gene duplicates and fragments in this final set of ion channel orthologues has been avoided. Related to this, does this analysis include or exclude isoforms?

Another aspect of the identification and classification of ion channel genes that could be improved is the figures for this section. One is relatively used to seeing trees as shown in Figures 3 and 4, which show relationships between genes as distances or evolutionary relationships. The decision to show the families of ion channels in Figure 1 as pie charts within a UMAP embedding is intriguing but somewhat non-intuitive and difficult to understand. Illustrating these results with a standard tree-like visualization of the relationship of these channels to each other would be preferred.

One aspect of the pie-chart/UMAP visualization that works well is the highlighting of the 'dark' ion channels according to the status as designated by IDG, which highlights a strength of this whole paper. However, throughout the paper, this could be emphasized more as the key advantage of this approach and how this or similar approaches could be used for other families of proteins. Specifically, in the initial statement describing 'light' vs 'dark channels', the importance of this distinction and the historical preference in science to study that which has already been studied can be discussed more, even including references to other studies that take this kind of approach. An example of a relevant reference here is to the Structural Genomics Consortium and its goals to achieve structures of proteins for which functions may not be well-characterized. Furthermore, this initial statement mentioning 'light channels' was initially confusing -- does this mean light-sensing channels? As one reads on this is clearly not the case, but for such an important central focus of this paper, these kinds of misunderstandings do not serve the authors well. Clarifying these motivations throughout the entire paper would strengthen it considerably.

Additionally, since the authors have generated this UMAP visualization, it would be interesting to understand how the human vs orthologue gene sets compare in this space. Furthermore, Figure 1, for just the human analysis, should say more clearly that this is an analysis of the human gene set and include more of the information in the text: 419 human ion channel sequences, 75 sequences previously unidentified, 4 major groups and 55 families, 62 outliers, etc. Clearer visualizations of these categories and numbers within the UMAP (and newly included tree) visualization would help guide the reader to better understand these results.

One of the most peculiar aspects of this paper is that it feels like two papers, one about better documenting the ion channel genes across species, and another with well-executed experiments on CALHM channels. One suggestion for how to link these two sections together better is to show that previous methods to analyze conserved residues in CALHM were significantly lacking. What results would that give? Why was this not enough? Were there just not enough identified CALHM orthologues to give strong signals in conservation analysis?

Some of the analysis pipeline is unclear. Specifically, the RAG analysis seems critical, but it is unclear how this works - is it on top of the GPT framework and recursively inquires about the answer to prompts? Some example prompts would be useful to understand this. Furthermore, the existence of 76 auxiliary non-pore containing 'ion channel' genes in this analysis is a little confusing, as it seems a part of the pipeline is looking for pore-lining residues. Furthermore, how many of these are picked up in the larger orthologues search? Are these harder to perform checks on to ensure that they are indeed ion channel genes? A further discussion of the choice to include these auxiliary sequences would be relevant. This could just be further discussion of the literature that has decided to do this in the past.

Overall, this manuscript is a valuable contribution to the field, but it requires a few main things to make it truly useful. Namely, how has this approach really improved the ability to identify conserved residues over a less-involved approach? A better description of their methods and results is required in the first section of the paper, as well as some cosmetic improvements.

Reviewer #2 (Public review):

Summary:

In this paper, the authors defined the "channelome," consisting of 419 predicted human ion channels as well as 48,000 ion channel orthologs from other organisms. Using this information, the ion channels were clustered into groups, which can potentially be used to make predictions about understudied ion channels in the groups. The authors then focused on the CALHM ion channel family, mutating conserved residues and assessing channel function.

Strengths:

The curation of the channelome provides an excellent resource for researchers studying ion channels. Supplemental Table 1 is well organized with an abundance of useful information.

Weaknesses:

There are substantial concerns regarding the analysis of the CALHM channels as detailed below.

(1) There are significant problems with the methodology used for the electrophysiology studies. Pulse protocol is used to assess the current voltage relationship (-100 to +140 mV), which extends far beyond the physiological range; currents for the mutant channels were only assessed at +120 mV. It is also unclear why a holding potential of 0 mV was used for CALHM6 recordings; the channel is already open at this voltage (and in Figure 4, only n = 3 for CALHM6). Further, proper controls were not performed. Inhibitors such as Gd3+ can be used to ensure that only CALHM currents are being measured.

(2) In line 334, the authors state that "expression levels of wild-type proteins and mutants are comparable." However, Western blots showing CALHM protein abundance (Supplementary Figure 3) are not of acceptable quality - in the top blot, WT CALHM1 can't even be seen. Representative blots were not shown for all mutants, and there was no effort to determine if levels were statistically significant compared to the wild-type control. Even if there is more or less protein, what does this mean? The protein could be in an intracellular compartment and not at the plasma membrane. In mammalian cells, CALHM6 is localized to intracellular compartments and only translocates to the plasma membrane upon activating stimulus (Danielli et al, EMBO J, 2023). Thus, if CALHM6 is only intracellular, the protein amount would not change, but the measured current would. Abundant intracellular CALHM1 has also been observed in mammalian cells transfected with this protein (Dreses-Werringloer et al., Cell, 2008). The best way to determine if mutations impact CALHM channel localization is to express GFP-tagged constructs in Xenopus oocytes and look for surface expression.

(3) Since the authors have not definitively shown that there are no defects in localization, they cannot make the claim in lines 346-356 that the mutations "either abolished or markedly reduced channel activity." Further, from their data, there is speculation regarding how these residues impact conformational changes during channel opening and closing. Line 404 - again, there is no concrete evidence that any of these residues play a role in gating function. Lines 406-433 - this entire paragraph is speculation without data to back it up. There is also a lack of specificity with statements such as "all mutants showed either reduced or completely abolished activity." What is meant by activity? Do the authors mean conductance?

(4) Line 303 - 13 aligned amino acids were conserved across all CALHM homologs - are these also aligned in related connexin and pannexin families? It is likely that cysteines and proline in TM2 are since CALHM channels overall share a lot of similarities with connexins and pannexins (Siebert et al, JBC, 2013). As in line 207, it would be expected that pannexins, connexins, and CALHM channel families would group together. Related to this, see Line 406 - in connexins, there is also a proline kink in TM2 that may play a role in mediating conformational changes between channel states (Ri et al, Biophysical Journal, 1999).

eLife Assessment

This useful work employs transition-metal FRET (tmFRET) to study the cyclic nucleotide binding domain (CNBD) of a bacterial ion channel. The authors employ lifetime measurements of fluorescence to extend their own prior study and observe distance changes within the CNBD domains of a full-length channel; they base these measurements on changes in lifetimes due to tmFRET between a metal at an introduced chelator site and a fluorescent non-canonical amino acid at another site within the channel sequence. This allows the authors to show that coupling of the CNBDs to the rest of the channel stabilizes the CNBDs in their active state relative to an isolated CNBD construct. The data are compelling and of high quality, and support the authors' conclusions.

Reviewer #1 (Public review):

Summary:

This useful work extends a prior study from the authors to observe distance changes within the CNBD domains of a full-length CNG channel based on changes in single photon lifetimes due to tmFRET between a metal at an introduced chelator site and a fluorescent non-canonical amino acid at another site. The data are excellent and convincingly support the authors' conclusions. The methodology is of general use for other proteins. The authors also show that coupling of the CNBDs to the rest of the channel stabilizes the CNBDs in their active state, relative to an isolated CNBD construct.

Strengths:

The manuscript is very well written and clear.

Reviewer #2 (Public review):

The manuscript "Domain Coupling in Allosteric Regulation of SthK Measured Using Time-Resolved Transition Metal Ion FRET" by Eggan et al. investigates the energetics of conformational transitions in the cyclic nucleotide-gated (CNG) channel SthK. This lab pioneered transition metal FRET (tmFRET), which has previously provided detailed insights into ion channel conformational changes. Here, the authors analyze tmFRET fluorescence lifetime measurements in the time domain, yielding detailed insights into conformational transitions within the cyclic nucleotide binding domains (CNBDs) of the channel. The integration of tmFRET with time-correlated single-photon counting (TCSPC) represents an advancement of this technique.

The results summarize known conformational transitions of the C-helix and provide distance distributions that agree with predicted values based on available structures. The authors first validated their TCSPC approach using the isolated CNBD construct previously employed for similar experiments. They then study the more complex full-length SthK channel protein. The findings agree with earlier results from this group, demonstrating that the C-helix is more mobile in the closed state than static structures reflect. Upon adding the activating ligand cAMP, the C-helix moves closer to the bound ligand, as indicated by a reduced fluorescence lifetime, suggesting a shorter distance between the donor and acceptor. The observed effects depend on the cAMP concentration, with affinities comparable to functional measurements. Interestingly, a substantial amount of CNBDs appear to be in the activated state even in the absence of cAMP (Figure 6E and F, fA2 ~ 0.4).

This may be attributed to cooperativity among the CNBDs, which the authors could elaborate on further. In this context, the major limitation of this study is that distance distributions are observed only in one domain. While inter-subunit FRET is detected and accounted for, the results focus exclusively on movements within one domain. Thus, the resulting energetic considerations must be assessed with caution. In the absence of the activator, the closed state is favored, while the presence of cAMP favors the open state. This quantifies the standard assumption; otherwise, an activator would not effectively activate the channel. However, the numerical values of approximately 3 kcal/mol are limited by the fact that only one domain is observed in the experiment, and only one distance (C- helix relative to the CNBD) is probed. Additional conformational changes leading to pore opening (including rotation and upward movement of the CNBD, and radial dilation of the tetrameric assembly) are not captured by the current experiments. These limitations should be taken into account when interpreting the results.

Reviewer #3 (Public review):

Summary:

This is a lucidly written manuscript describing the use of transition-metal FRET to assess distance changes during functional conformational changes in a CNG channel. The experiments were performed on an isolated C-terminal nucleotide binding domain (CNBD) and on a purified full-length channel, with FRET partners placed at two positions in the CNBD.

Strengths:

The data and quantitative analysis are exemplary, and they provide a roadmap for use of this powerful approach in other proteins.

Weaknesses/Comments:

A ~3x lower Kd for nucleotide is seen for the detergent-solubilized full-length channel, compared to electrophysiological experiments. This is worth a comment in the Discussion, particularly in the context of the effect of the pore domain on the CNBD energetics.

eLife Assessment

The fundamental findings reported here provide insight into how the viscoelasticity of the fingertip skin influences the activity of mechanoreceptive afferents and thus the neural coding of force in humans. The basic principle studied was whether and to what extent the previous applied force directions impact the firing of FA-1, SA-1 and SA-2 neurons during the current applied force directions. The data and analyses are compelling and will be helpful for modeling the neural representations of force in the context of object grasping and manipulation.

Reviewer #1 (Public review):

The authors investigate how the viscoelasticity of the fingertip skin can affect the firing of mechanoreceptive afferents and they find a clear effect of recent physical skin state (memory), which is different between afferents. The manuscript is extremely well-written and well-presented. It uses a large dataset of low threshold mechanoreceptive afferents in the fingertip, where it is particularly noteworthy that the SA-2s have been thoroughly analyzed and play an important role here. They point out in the introduction the importance of the non-linear dynamics of the event when an external stimulus contacts the skin, to the point at which this information is picked up by receptors. Although clearly correlated, these are different processes, and it has been very well-explained throughout. I have some comments and ideas that the authors could think about that could further improve their already very interesting paper. Overall, the authors have more than achieved their aims, where their results very much support the conclusions and provoke many further questions. This impact of the previous dynamics of skin affecting current state can be explored further in so many ways and may help us in understanding skin aging and the effects of anatomical changes of the skin better.

Comments on revised submission:

The authors have taken all my considerations into account and provided excellent responses to them. They have modified their paper accordingly, which improves its clarity even more. Very interesting work and I have no further comments.

Reviewer #2 (Public review):

Summary:

The authors sought to identify the impact skin viscoelasticity has on neural signalling of contact forces that are representative of those experienced during normal tactile behaviour. The evidence presented in the analyses indicate there is a clear effect of viscoelasticity on the imposed skin movements from a force-controlled stimulus. Both skin mechanics and evoked afferent firing were affected based on prior stimulation, which has not previously been thoroughly explored. This study outlines that viscoelastic effects have an important impact on encoding in the tactile system, which should be considered in the design and interpretation of future studies. Viscoelasticity was shown to affect the mechanical skin deflections and stresses/strains imposed by previous and current interaction force, and also the resultant neuronal signalling. The result of this was an impaired coding of contact forces based upon previous stimulation. The authors may be able to strengthen their findings, by using the existing data to further explore the link between skin mechanics and neural signalling, giving a clearer picture than demonstrating shared variability. This is not a critical addition, but I believe would strengthen the work and make it more generally applicable.

Strengths:

-Elegant design of the study. Direct measurements have been made from the tactile sensory neurons to give detailed information on touch encoding. Experiments have been well designed and the forces/displacements have been thoroughly controlled and measured to give accurate measurements of global skin mechanics during a set of controlled mechanical stimuli.<br /> -Analytical techniques used. Analysis of fundamental information coding and information representation in the sensory afferents reveals dynamic coding properties to develop putative models of the neural representation of force. This advanced analysis method has been applied to a large dataset to study neural encoding of force, the temporal dynamics of this, and the variability in this.

Weaknesses:<br /> -Lack of exploration of the variation in neural responses. Although there is a viscoelastic effect which produces variability in the stimulus effects based on prior stimulation, it is a shame that the variability in neural firing and force induced skin displacements have been presented, and are similarly variable, but there has been no investigation of a link between the two. I believe with these data the authors can go beyond demonstrating shared variability. The force per se is clearly not faithfully represented in the neural signal, being masked by stimulation history, and it is of interest if the underlying resultant contact mechanics are.

Validity of conclusions:

The authors have succeeded in demonstrating skin viscoelasticity has an impact on skin contact mechanics with a given force and that this impacts on the resultant neural coding of force. Their study has been well designed and the results support their conclusions. The importance and scope of the work is adequately outlined for readers to interpret the results and significance.

Impact:

This study will have important implications for future studies performing tactile stimulation and evaluating tactile feedback during motor control tasks. In detailed studies of tactile function, it illustrates the necessity to measure skin contact dynamics to properly understand the effects of a force stimulus on the skin and mechanoreceptors.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public Review):

The authors investigate how the viscoelasticity of the fingertip skin can affect the firing of mechanoreceptive afferents and they find a clear effect of recent physical skin state (memory), which is different between afferents. The manuscript is extremely well-written and well-presented. It uses a large dataset of low threshold mechanoreceptive afferents in the fingertip, where it is particularly noteworthy that the SA-2s have been thoroughly analyzed and play an important role here. They point out in the introduction the importance of the non-linear dynamics of the event when an external stimulus contacts the skin, to the point at which this information is picked up by receptors. Although clearly correlated, these are different processes, and it has been very well-explained throughout. I have some comments and ideas that the authors could think about that could further improve their already very interesting paper. Overall, the authors have more than achieved their aims, where their results very much support the conclusions and provoke many further questions. This impact of the previous dynamics of the skin affecting the current state can be explored further in so many ways and may help us to better understand skin aging and the effects of anatomical changes of the skin.

At the beginning of the Results, it states that FA-2s were not considered as stimuli did not contain mechanical events with frequency components high enough to reliably excite them. Was this really the case, did the authors test any of the FA-2s from the larger dataset? If FA-2s were not at all activated, this is also relevant information for the brain to signal that it is not a relevant Pacinian stimulus (as they respond to everything). Further, afferent receptive fields that were more distant to the stimulus were included, which likely fired very little, like the FA-2s, so why not consider them even if their contribution was low?

Thank you for bringing this up, we have now clarified in the text that while FA-2s did respond at a low rate during the experiment, their responses were not reliably driven by the force stimuli. In the Methods section we have included the following text:

“Initially, 10 FA-2 neurons were also included in the analysis. But their responsiveness during the experiment was remarkably low, and unlike the other neuron types, their responses were rarely affected by force stimuli. Specifically, only one of the observed FA-2 neurons responded during the force protraction phases. Due to the lack of clear stimulus-driven responses, FA-2 neurons were subsequently excluded from further analysis.”

One question that I wondered throughout was whether you have looked at further past history in stimulation, i.e. not just the preceding stimulus, but 2 or 3 stimuli back? It would be interesting to know if there is any ongoing change that can be related back further. I do not think you would see anything as such here, but it would be interesting to test and/or explore in future work (e.g. especially with sticky, forceful, or sharp indentation touch). However, even here, it could be that certain directions gave more effects.

This is a very interesting question! A discernible effect from the previous stimulus could persist at the end of the current stimulation (see Figure 4C), potentially influencing the next one—a 2-stimuli-back effect. Unfortunately, our experimental design did not allow for rigorous testing of this effect. While all possible pairs of stimulus directions were included in immediately consecutive trials, this was not the case for pairs separated by additional trials. Hence, the combination of a likely weak effect and limited variation in history precluded a thorough analysis of a 2-stimuli-back effect. Future work should delve into the time course of the viscoelastic effect in greater detail.

Did the authors analyze or take into account the difference between receptive field locations? For example, did afferents more on the sides have lower responses and a lesser effect of history?

An investigation into the potential impact of the relationship between the receptive field location on the fingertip skin and the primary contact site of the stimulus surface revealed no discernible influence for SA-1 and SA-2 neurons. In contrast, FA-1 neurons, particularly those predominantly sensitive to the previous stimulation or displaying mixed sensitivity, exhibited a tendency to terminate near the primary stimulation site. We have added these observations to the text:

“We found no straightforward relationship between a neuron's sensitivity to current and previous stimulation and its termination site in fingertip skin. Specifically, there was no statistically significant effect of the distance between a neuron's receptive field center and the primary contact site of the stimulus surface on whether neurons signaled current, prior, or mixed information for SA-1 (Kruskal-Wallis test H(2)=3.86, p= 0.15) or SA-2 neurons (H(2)=0.75, p=0.69). However, a significant difference emerged for FA-1 neurons (H(2)=8.66, p=0.01), indicating that neurons terminating closer to the stimulation site on the flat part of the fingertip were more likely to signal past or mixed information.”

Was there anything different in the firing patterns between the spontaneous and non-spontaneously active SA-2s? For example, did the non-spontaneous show more dynamic responses?

The firing patterns of both spontaneously and non-spontaneously active SA-2 neurons shared similarities in terms of adaptation and range of firing rate modulation in response to force stimuli, i.e., ‘dynamic response’. The distinction lay in the pattern of modulation of the firing rate associated with stimulus presentations. For spontaneously active SA-2 neurons, this modulation occurred around a significant background discharge, implying that a force stimulus could either decrease or increase the firing rate, depending on how it deformed the fingertip. This characteristic is well illustrated by the firing pattern of the neuron depicted in the lower panels of Figure 3D. Conversely, in non-spontaneously active SA-2 neurons, a force stimulus could only induce an increase in the firing rate or no change. Although the neuron depicted in the upper panels of Figure 3D exhibited some background activity, it serves to exemplify this characteristic. In the text, we have elucidated the dynamics of the SA-2 neuron response by highlighting that force stimulation can either decrease or increase the firing rate in neurons with spontaneous activity through the following addition/change:

“This increased variability was most evident during the force protraction phase where most neurons exhibited the most intense responses. Increased variability was also observed in instances where the dynamic response to force stimulation involved a decrease in the firing rate (lower panels of Figure 3D). This phenomenon was observed in SA-2 neurons that maintained an ongoing discharge during intertrial periods (cf. Fig. 2A). In these cases, the response to a force stimulus constituted a modulation of the firing rate around the background discharge, signifying that a force stimulus could either decrease or increase the firing rate depending on the prevailing stimulus direction.”

Were the spontaneously active SA-2 afferents firing all the time or did they have periods of rest - and did this relate to recent stimulation? Were the spontaneously active SA-2s located in a certain part of the finger (e.g. nail) or were they randomly spread throughout the fingertip? Any distribution differences could indicate a more complicated role in skin sensing.

SA-2 neurons, in general, are well-known for undergoing significant post-stimulation depression (e.g., Knibestöl and Vallbo, 1970; Chambers et al., 1972; Burgess and Perl, 1973). In our force stimulations, this post-excitatory depression manifested as a reduced or absent response during the latter part of the stimulus retraction period for stimuli in directions that markedly excited the neuron. The excitability recovered when the fingertip relaxed during the subsequent intertrial period, and for "spontaneously active" neurons, the firing resumed (see examples in Figure 7A). Furthermore, some “spontaneously active” neurons could be silenced or exhibit a near-silent period during force stimulation for certain force directions, while the spontaneous firing returned during the upcoming intertrial period when the fingertip shape recovered (for example, see responses to stimulation in the proximal and especially ulnar directions in the top panel in Figure 7A).

Regarding the location of the receptive field centres of spontaneously active and non-spontaneously active SA-2 neurons on the fingertip we did not observe any obvious spatial segregation. To illustrate this, we have revised Figure 1A by color-marking SA-2 neurons that exhibited ongoing activity in intertrial periods, and the figure caption has been modified accordingly:

“Figure 1. Experimental setup. A. Receptive field center locations shown on a standardized fingertip for all first-order tactile neurons included in the study, categorized by neuron type. Purple symbols denote spontaneously active SA-2 neurons exhibiting ongoing activity without external stimulation.”

Did the authors look to see if the spontaneous firing in SA-2s between trials could predict the extent to which the type 1 afferents encode the proceeding stimulus? Basically, does the SA-2 state relate to how the type 1 units fire?

We found no clear indications that the responses of FA-1 and SA-1 could be readily anticipated based on the firing patterns of SA-2 neurons.

In the discussion, it is stated that "the viscoelastic memory of the preceding loading would have modulated the pattern of strain changes in the fingertip differently depending on where their receptor organs are situated in the fingertip". Can the authors expand on this or make any predictions about the size of the memory effect and the distance from the point of stimulation?

We have explored this topic further in the text, referring to recent studies modeling essential aspects of fingertip mechanics. However, in our view, current models lack the capability to predict the specific nature sought by the reviewer. These models should include a detailed understanding of the intricate networks of collagen fibers anchoring the pulp tissue at the distal phalangeal bone and the nail. They should also consider potential inherent directional preferences of the receptor organs, attributed to their microanatomy. The text modifications are as follows:

“In addition to the receptor organ locations, the variation in sensitivity among neurons to fingertip deformations in response to both previous and current loadings would stem from the fingertip’s geometry and its complex composite material properties. Possible inherent directional preferences of the receptor organs, attributed to their microanatomy, could also be significant. However, mechanical anisotropy, particularly within the viscoelastic subcutaneous tissue of the fingertip induced by intricately oriented collagen fiber strands forming fat columns in the pulp (Hauck et al., 2004), are likely to play a crucial role. This anisotropy would shape the dynamic pattern of strain changes at neurons' receptor sites, intricately influencing a neuron's sensitivity not only to current but also to preceding loadings. Indeed, recent modeling efforts suggest that such mechanical anisotropy strongly influences the spatiotemporal distribution of stresses and strains across the fingertip (Duprez et al., 2024).”

Relatedly, we have included additional text to provide a more comprehensive explanation of the “bulk deformation” of the fingertip that occurs during the loadings:

“As pressure increases in the pulp, the pulp tissue bulges at the end and sides of the fingertip. Simultaneously, the tangential force component amplifies the bulging in the direction of the force while stretching the skin on the opposite side.”

In the discussion, it would be good if the authors could briefly comment more on the diversity of the mechanoreceptive afferent firing and why this may be useful to the system.

The diversity in responses among neurons is instrumental in enhancing the information transmitted to the brain by averting redundancy in information acquisition. This diversity thereby contributes to an overall increase in information. We've included a brief statement, along with several references, underscoring this concept:

"The resulting diversity in the sensitivities of neurons might enhance the overall information collected and relayed to the brain by the neuronal population, facilitating the discrimination between tactile stimuli or mechanical states of the fingertip (see Rongala et al., 2024; Corniani et al., 2022; Tummala et al., 2023, for more extensive explorations of this idea)."

Also, the authors could briefly discuss why this memory (or recency) effect occurs - is it useful, does it serve a purpose, or it is just a by-product of our skin structure? There are examples of memory in the other senses where comparisons could be drawn. Is it like stimulus adaptation effects in the other senses (e.g. aftereffects of visual motion)?

We have expanded the concluding paragraph of the discussion, specifically delving into the question of whether the mechanical memory effect serves a deliberate purpose or is simply an incidental byproduct of our skin structure:

“In any case, the viscoelastic deformability of the fingertips plays a pivotal role in supporting the diverse functions of the fingers. For example, it allows for cushioned contact with objects featuring hard surfaces and allows the skin to conform to object shapes, enabling the extraction of tactile information about objects' 3D shapes and fine surface properties. Moreover, deformability is essential for the effective grasping and manipulation of objects. This is achieved, among other benefits, by expanding the contact surface, thereby reducing local pressure on the skin under stronger forces and enabling tactile signaling of friction conditions within the contact surface for control of grasp stability. Throughout, continuous acquisition of information about various aspects of the current state of the fingertip and its skin by tactile neurons is essential for the functional interaction between the brain and the fingers. In light of this, the viscoelastic memory effect on tactile signaling of fingertip forces can be perceived as a by-product of an overall optimization process within prevailing biological constraints.”

One point that would be nice to add to the discussion is the implications of the work for skin sensing. What would you predict for the time constant of relaxation of fingertip skin, how long could these skin memory effects last? Two main points to address here may be how the hydration of the skin and anatomical skin changes related to aging affect the results. If the skin is less viscoelastic, what would be the implications for the firing of mechanoreceptors?

It is likely that the time constant depends to some extent on mechanical factors of the skin, which will likely change due to age or environmental factors. However, while these questions are intriguing, they fall outside the scope of the current study and we are not aware of studies that have addressed these issues directly in experiments either.

How long does it take for the effect to end? Again, this will likely depend on the skin's viscoelasticity. However, could the authors use it in a psychophysical paradigm to predict whether participants would be more or less sensitive to future stimuli? In this way, it would be possible to test whether the direction modifies touch perception.

Time constants for tissue viscoelasticity have been estimated to extend up to several seconds (see citations in the introduction). While direct perceptual effects could indeed be explored through psychophysical experimental paradigms, we are currently unaware of any studies specifically addressing the type of effect described in this study. In addition to the statement that, concerning manipulation and haptic tasks, "to our knowledge, a possible influence of fingertip viscoelasticity on task performance has not been systematically investigated," we have now also addressed tactile psychophysical tasks conducted during passive touch with the following sentence in the text:

“Similarly, there is a lack of systematic investigation of potential effects of fingertip viscoelasticity on performance in tactile psychophysical tasks conducted during passive touch.”

Reviewer #2 (Public Review):

Summary:

The authors sought to identify the impact skin viscoelasticity has on neural signalling of contact forces that are representative of those experienced during normal tactile behaviour. The evidence presented in the analyses indicates there is a clear effect of viscoelasticity on the imposed skin movements from a force-controlled stimulus. Both skin mechanics and evoked afferent firing were affected based on prior stimulation, which has not previously been thoroughly explored. This study outlines that viscoelastic effects have an important impact on encoding in the tactile system, which should be considered in the design and interpretation of future studies. Viscoelasticity was shown to affect the mechanical skin deflections and stresses/strains imposed by previous and current interaction force, and also the resultant neuronal signalling. The result of this was an impaired coding of contact forces based on previous stimulation. The authors may be able to strengthen their findings, by using the existing data to further explore the link between skin mechanics and neural signalling, giving a clearer picture than demonstrating shared variability. This is not a critical addition, but I believe would strengthen the work and make it more generally applicable.

Strengths:

- Elegant design of the study. Direct measurements have been made from the tactile sensory neurons to give detailed information on touch encoding. Experiments have been well designed and the forces/displacements have been thoroughly controlled and measured to give accurate measurements of global skin mechanics during a set of controlled mechanical stimuli.

- Analytical techniques used. Analysis of fundamental information coding and information representation in the sensory afferents reveals dynamic coding properties to develop putative models of the neural representation of force. This advanced analysis method has been applied to a large dataset to study neural encoding of force, the temporal dynamics of this, and the variability in this.

Weaknesses:

- Lack of exploration of the variation in neural responses. Although there is a viscoelastic effect that produces variability in the stimulus effects based on prior stimulation, it is a shame that the variability in neural firing and force-induced skin displacements have been presented, and are similarly variable, but there has been no investigation of a link between the two. I believe with these data the authors can go beyond demonstrating shared variability. The force per se is clearly not faithfully represented in the neural signal, being masked by stimulation history, and it is of interest if the underlying resultant contact mechanics are.

Thank you for this suggestion. We have added a new section investigating the link between skin deformation and neural firing in more depth via a simple neural model. Please see our answer below in the ‘Recommendations’ section for further details.

Validity of conclusions:

The authors have succeeded in demonstrating skin viscoelasticity has an impact on skin contact mechanics with a given force and that this impacts the resultant neural coding of force. Their study has been well-designed and the results support their conclusions. The importance and scope of the work is adequately outlined for readers to interpret the results and significance.

Impact:

This study will have important implications for future studies performing tactile stimulation and evaluating tactile feedback during motor control tasks. In detailed studies of tactile function, it illustrates the necessity to measure skin contact dynamics to properly understand the effects of a force stimulus on the skin and mechanoreceptors.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

(Very) minor comments

- The authors say at the beginning of the Results that, "The fourth type of tactile neurons in the human glabrous skin, fast adapting type II neurons...". Although generally written that there are four types of afferent in the glabrous skin, it would be better to state that these are low-threshold A-beta myelinated mechanoreceptive afferents, at least one time, as there are other types of afferent in the glabrous skin that respond to mechanical stimulation (e.g. low and high threshold C-fibers).

This is now clarified at the start of the Results section:

“We recorded action potentials in the median nerve of individual low-threshold A-beta myelinated first-order human tactile neurons innervating the glabrous skin of the fingertip…”

- Fig. 3: Could you add '(N)' as the measurement of force for Fig. 3A for Fz, Fy, and Fz? Also, please change 'Data was recorded' to 'Data were recorded' in the legend.

Fixed.

- At the beginning of the Methods, you say that your study conforms to the Declaration of Helsinki, which actually requires pre-registration in a database. If you did not pre-register your study, please can you add '... in accordance with the Declaration of Helsinki, apart from pre-registration in a database'.

Thanks for making us aware of this. We have added the suggested qualifier to the ethics statement.

Reviewer #2 (Recommendations For The Authors):

The neural representation/encoding of the actual displacement vectors would be a useful addition to the analyses. These vectors have been demonstrated to systematically change with the condition in the irregular series (Figure 2E) and will thus significantly act on the dynamics of induced mechanical changes in the skin with a given interaction force. Thus, it could be examined how the neurons code the magnitude of displacements as well as their direction. An evaluation of the extent to which the imposed displacement magnitudes are encoded in the neural responses would be a useful addition in explaining the signalling of the force events and how the central nervous system decodes these. Evaluating an alternative displacement encoding for comparison to pure force encoding may reveal more about how contact events are represented in the tactile system, which must decode these variable afferent signals to reconstruct a percept of the interaction. It could then be explored how the central nervous system may then scale the dynamic afferent responses based on the background viscoelastic state likely to be present in the SA-II afferent signals (Figure 7) for a context in which to evaluate the dynamic contact forces. This may of course be a complex relationship for the type-I afferents, where the underlying mechanical events evoking the firing (microslips not represented in global forces) have not been measured here. Such a model could be more widely applicable, as the skin viscoelasticity and displacement magnitudes are a straightforward measurement metric and could perhaps be used as a better proxy for neural signalling. This would allow the investigation of a wider variety of forces, and the study of the timing of the viscoelastic effect, both of which have been fixed here. This would give the work a broader impact, rather than just highlighting that this effect produces variability, it could reveal if this mechanical feature is structured in the neural representation. The categorical encoding/decoding tested here is specific to the stimuli used (magnitudes, intervals), but there is the possibility that this may be more generally applicable (within the bounds of forces/speeds) if the underlying basis of the variability in the signalling produced by the viscoelasticity is identified. Since the time course of the viscoelasticity has not been measured here (fixed forces and intervals), further study is required to fully understand the implications this has for a wider variety of situations.

We agree that a better understanding of how the mechanical deformations are reflected in the resulting spike trains would be valuable. While ultimately a full understanding will need precise measurements of skin deformation across the whole fingertip to account for mechanical propagation to mechanoreceptor locations, relating the deformations at the contact location with neural firing patterns directly can provide useful hints into which aspects of deformation are encoded and how. To this end, we ran a new analysis that aimed to predict the time-varying neural responses directly from the recorded mechanical movements of the contactor.

Below we have reproduced the new results and methods text along with the additional figures for this analysis. Note that we have also added text in the Discussion to interpret these findings in the context of our other results.

New section in Results titled Predicting neural responses from contactor movements: “The similarity in the history-dependent variation in neural firing and fingertip deformation at a given force stimulus suggests that neuronal firing is determined by how the fingertip deforms rather than the applied force itself. However, this similarity does not clarify the relationship between fingertip deformation dynamics and neural signaling. To investigate further, we fit cross-validated multiple linear regression models to evaluate how well distinct aspects of contactor movement could predict the time-varying firing rates of individual neurons during the protraction phases of the irregular sequence. The models used predictors based on (1) the three-dimensional position of the contactor, (2) its three-dimensional velocity, (3) a combination of position and velocity signals, and, finally, (4) position and velocity signals along with all possible two-way interactions between them, capturing potentially complex relationship between fingertip deformations and neural signaling.

Comparing the variance explained (R²) by each regression model for each neuron type revealed clear differences between the models (Figure 5A). A two-way mixed design ANOVA, with regression model as within-group effects and neuron type as a between-group effect revealed a main effect of model on variance explained (F(3,462) = 815.5, p < 0.001, η_p² = 0.84). Model prediction accuracy overall increased with the number of predictors, with the two-way interaction model outperforming all others (p < 0.001 for all comparisons, Tukey’s HSD). Additionally, a significant main effect of neuron type (F(2,154) = 29.8, p < 0.001, η_p² = 0.28) and a significant interaction between regression model and neuron type were observed (F(6,462) = 50.8, p < 0.001, η_p² = 0.40).

For neuron type, model predictions were most accurate for SA-2 neurons, followed by SA-1 neurons, with FA-1 neurons showing the lowest accuracy (p < 0.003 for all comparisons, Tukey’s HSD). The interaction between model and neuron type revealed distinct patterns. For SA-1 and SA-2 neurons, position-only and velocity-only models had similar prediction accuracy (p ≥ 0.996, Tukey’s HSD) with no significant differences between these neuron types (p ≥ 0.552, Tukey’s HSD). FA-1 neurons performed poorly with the position-only model but showed higher accuracy with the velocity-only model (p < 0.001, Tukey’s HSD) and better than SA-1 neurons (p = 0.006, Tukey’s HSD). Models combining position and velocity predictors (without interactions) surpassed both position-only and velocity-only models for SA-1 and SA-2 neurons (p < 0.001, Tukey’s HSD). Overall, the differences between neuron types broadly match their tuning to static and dynamic stimulus properties.

The two-way interaction model, accounting for most variance in neural responses, produced mean R² values of 0.75 for FA-1, 0.88 for SA-1, and 0.91 for SA-2 neurons (Figure 5A). To evaluate the contribution of the different predictors, we ranked them using the permutation feature importance method, focusing on the six most important ones. Regression analyses using only these variables explained almost all of the variance explained by the full model, with a median R² reduction of just 0.055 across all neurons. Across all neuron types, at least half included all three velocity components (dPx, dPy, dPz) among the top six, with FA-1 neurons showing the highest prevalence (Figure 5B). Interactions between normal position (Pz) and each velocity component were also frequently observed, while interactions involving tangential position and velocity components were less common. Interactions among velocity components were relatively well represented, followed by interactions limited to position components. Position signals were generally less represented, except for normal position (Pz) in slowly adapting neurons, where it appeared in 50% of SA-1 and 68% of SA-2 neurons. Despite these broad trends, important predictors varied widely across ranks even within a given neuron class (see Figure 5-figure supplement 1), and even the most frequent variables appeared in only a subset of cases, suggesting broad variability in sensitivity across neurons.”

New methods paragraph titled Predicting time-varying firing rates from skin deformations:

“This analysis was conducted in Python (v3.13) with pandas for data handling, numpy for numerical operations, and scikit-learn for model fitting and evaluation.

To assess how well individual neurons' time-varying firing rates could be predicted from simultaneous contactor movements, we fitted multiple linear regression models (see Khamis et al., 2015, for a similar approach}. This analysis focused on the force protraction phase of the irregular sequence, where neurons were most responsive and sensitive to stimulation history. Data from 100 ms before to 100 ms after the protraction phase (between -0.100 s and 0.225 s relative to protraction onset) were included for each trial. Neurons were included if they fired at least two action potentials during the force protraction phase and the following 100 ms in at least five of the 25 trials. This ensured sufficient variability in firing rates for meaningful regression analysis, resulting in 68 SA-1, 38 SA-2, and 51 FA-1 neurons being included.

Contractor position signals digitized at 400 Hz were linearly interpolated to 1000 Hz. Instantaneous firing rates, derived from action potentials sampled at 12.8 kHz, were resampled at 1000 Hz to align with position signals. A Gaussian filter (σ = 10 ms, cutoff ~16 Hz) was applied to the firing rate as well as to the position signals before differentiation. To account for axonal conduction (8–15 ms) and sensory transduction delays (1–5 ms), firing rates were advanced by 15 ms to align approximately with independent variables.

Regressions were performed using scikit-learn's Ridge and RidgeCV regressors, which apply L2 regularization to mitigate overfitting. Hyperparameter tuning for the regularization parameter (alpha) was performed using GridSearchCV with a predefined range (0.001–1000.0), incorporating five-fold cross-validation to select the best value. To minimize overfitting risks, model performance was further validated with independent five-fold cross-validation (KFold), and R² scores were computed using cross_val_score.

We constructed four linear regression models with increasing complexity: (1) Position-only, using three-dimensional contactor positions (Px, Py, Pz); (2) Velocity-only, using three-dimensional velocities (dPx, dPy, dPz); (3) Combined, including all position and velocity signals (6 predictors); and (4) Interaction, including all signals and their two-way interactions (21 predictors). All features were standardized using StandardScaler to improve regularization and model convergence. PolynomialFeatures generated second-order interaction terms for the interaction model. Feature importance was evaluated with permutation_importance, and simpler models were built using the most important features. These models were validated through cross-validation to assess retained explanatory power.”

Minor:

- It would be useful to add a brief description of the material aspects of the contactor tip to the methods (as per Birznieks 2001).

We have added the following statement:

“To ensure that friction between the contactor and the skin was sufficiently high to prevent slips, the surface was coated with silicon carbide grains (50–100 μm), approximating the finish of smooth sandpaper.”

- The axes labelling on Figure 3A and legend description is ambiguous, probably placing the Px, Py, and Pz labels on the far left axes and the Fx, Fy, and Fz on the right side of the far right axes would make this clearer.

Label placement has been improved along with some other minor fixes.

- For the quasi-static phase analysis, the phrase "absence of loading" used in reference to the interstimulus period and SA-II afferents does not seem to be a correct description. The finger is still loaded (at least in the normal direction), with a magnitude of imposed displacement that counteracts the viscoelastic force exerted by the skin mechanics of the fingertip. Although there is a zero net-force load, a mechanical stimulus is still being actively applied to the skin.

We have changed the wording throughout the text and now consistently refer either to the “interstimulus period” directly or to an “absence of externally applied stimulation” to avoid confusion.

eLife Assessment

This fundamental work presents two clinically relevant BMP4 mutations that contribute to vertebrate development. The compelling evidence, both from wet lab and AI generated predictions, supports that the site-specific cleavage at the BMP4 pro-domain precisely regulates its function and provides mechanistic insight how homodimers and heterodimers behave differently. The work will be of broad interest to researchers working on growth factor signaling mechanisms and vertebrate development.

Reviewer #1 (Public review):

Summary:

The authors demonstrate that two human preproprotein human mutations in the BMP4 gene cause a defect in proprotein cleavage and BMP4 mature ligand formation, leading to hypomorphic phenotypes in mouse knock-in alleles and in Xenopus embryo assays.

Strengths:

They provide compelling biochemical and in vivo analyses supporting their conclusions, showing the reduced processing of the proprotein and concomitant reduced mature BMP4 ligand protein from impressively mouse embryonic lysates. They perform excellent analysis of the embryo and post-natal phenotypes demonstrating the hypomorphic nature of these alleles. Interesting phenotypic differences between the S91C and E93G mutants are shown with excellent hypotheses for the differences. Their results support that BMP4 heterodimers act predominantly throughout embryogenesis whereas BMP4 homodimers play essential roles at later developmental stages.

Weaknesses:

In the revision the authors have appropriately addressed the previous minor weaknesses.

Reviewer #2 (Public review):

Summary:

The revised paper by Kim et al. reports two disease mutations in proBMP4, S91C and E93G, disrupt the FAM20C phosphorylation site at Ser91, blocking the activation of proBMP4 homodimers, while still allowing BMP4/7 heterodimers to function. Analysis of DMZ explants from Xenopus embryos expressing the proBMP4 S91C or E93G mutants showed reduced expression of pSmad1 and tbxt1. The expert amphibian tissue transplant studies were expanded to in vivo studies in Bmp4S91C/+ and Bmp4E93G/+ mice, highlighting the impact of these mutations on embryonic development, particularly in female mice, consistent with patient studies. Additionally, studies in mouse embryonic fibroblasts (MEFs) demonstrated that the mutations did not affect proBMP4 glycosylation or ER-to-Golgi transport but appeared to inhibit the furin-dependent cleavage of proBMP4 to BMP4. Based on these findings and AI modeling using AlphaFold of proBMP4, the authors speculate that pSer91 influences access of furin to its cleavage site at Arg289AlaLysArg292 in a new "Ideas and Speculation" section. Overall, the authors addressed the reviewers' comments, improving the presentation.

Strengths:

The strengths of this work continue to lie in the elegant Xenopus and mouse studies that elucidate the impact of the S91C and E93G disease mutations on BMP signaling and embryonic development. Including an "Ideas and Speculation" subsection for mechanistic ideas reduces some shortcomings regarding the analysis of the underlying mechanisms.

Reviewer #3 (Public review):

Summary:

The authors describe important new biochemical elements in the synthesis of a class of critical developmental signaling molecules, BMP4. They also present a highly detailed description of developmental anomalies in mice bearing known human mutations at these specific elements.

Strengths:

This paper presents exceptionally detailed descriptions of pathologies occurring in BMP4 mutant mice. Novel findings are shown regarding the interaction of propeptide phosphorylation and convertase cleavage, both of which will move the field forward. Lastly, a provocative hypothesis regarding furin access to cleavage sites is presented, supported by Alphafold predictions.

Author response:

The following is the authors’ response to the previous reviews

Reviewer #2 (Public review):

Summary:

The revised paper by Kim et al. reports two disease mutations in proBMP4, S91C and E93G, disrupt the FAM20C phosphorylation site at Ser91, blocking the activation of proBMP4 homodimers, while still allowing BMP4/7 heterodimers to function. Analysis of DMZ explants from Xenopus embryos expressing the proBMP4 S91C or E93G mutants showed reduced expression of pSmad1 and tbxt1. The expert amphibian tissue transplant studies were expanded to in vivo studies in Bmp4S91C/+ and Bmp4E93G/+ mice, highlighting the impact of these mutations on embryonic development, particularly in female mice, consistent with patient studies. Additionally, studies in mouse embryonic fibroblasts (MEFs) demonstrated that the mutations did not affect proBMP4 glycosylation or ER-to-Golgi transport but appeared to inhibit the furin-dependent cleavage of proBMP4 to BMP4. Based on these findings and AI modeling using AlphaFold of proBMP4, the authors speculate that pSer91 influences access of furin to its cleavage site at Arg289AlaLysArg292 in a new "Ideas and Speculation" section. Overall, the authors addressed the reviewers' comments, improving the presentation.

Strengths:

The strengths of this work continue to lie in the elegant Xenopus and mouse studies that elucidate the impact of the S91C and E93G disease mutations on BMP signaling and embryonic development. Including an "Ideas and Speculation" subsection for mechanistic ideas reduces some shortcomings regarding the analysis of the underlying mechanisms.

Weaknesses:

(1)  (Minor) In Figure S1 and lines 165-174 and 179-180, the authors should consider that, unlike the wild-type protein (Ser), which can be reversibly phosphorylated or dephosphorylated, phosphomimic mutations are locked into mimicking either the phosphorylated state (Asp) or the non-phosphorylated state (Ala). Consequently, if the S91D mutant exhibits lower activity than WT, it could imply that S91D interferes with other regulatory constraints, as the authors suggest. However, it may also be inhibiting activation. Therefore, caution is warranted when comparing S91D with S91C to conclude that Ser91 phosphorylation increases BMP4 activity. While additional experiments are not necessary, further consideration is essential.

(Minor) In lines 394-399, the authors cleverly speculate that pS91 interacts with Arg289-the essential P4 arginine for furin processing. If so, this interaction could hinder the cleavage of proBMP4, as indicated by the results in Figure S1. The discussion would benefit from considering that, contrary to their favored model, dephosphorylation at Ser91 might actually facilitate cleavage.

We have added a paragraph raising this possibility but explaining why it is unlikely and inconsistent with our in vivo data. The S91D construct was a simple control that was tested in ectopic expression assays and not in vivo.  We can make no conclusions about whether this construct resembles the phosphorylated state or whether it hinders or facilitates cleavage in vivo. The conclusion that dephosphorylation promotes BMP4 cleavage or activity is not compatible with the finding that two mutations associated with birth defects in humans (p.S91C or p.E93G) that are predicted to prevent FAM20C-mediated phosphorylation of the BMP4 prodomain lead to impaired proteolytic maturation of endogenous BMP4 and reduced BMP activity in vivo. 

(2)  In Figure 4, panels A, E, and I, the proBMP bands in the mouse embryonic lysates and MEFs expressing the mutations show a clear size shift. Are these shifts a cause or a consequence of the lack of cleavage? Regardless, the size shifts should be explicitly noted.

These intriguing shifts were observed in some but not all biological replicates.  When present, the shifts were not reversed by treatment with phosphatases or deglycosylases, and the shifts were never observed in epitope tagged wild type controls.  We have added a paragraph noting the shifts and our tests of whether they might be due to glycosylation, phosphorylation or epitope tags. 

(3)  (Minor) In line 314, the authors should consider modifying the wording to: "is required for modulating proprotein convertase..."

The original wording (“Collectively, our findings are consistent with a model in which FAM20C-mediated phosphorylation of the BMP4 prodomain is not required for folding or exit of the precursor protein from the ER, but is required for proprotein convertase recognition and/or for trafficking to post-TGN compartment(s) where BMP4 is cleaved”) more accurately reflects the model that is supported by our findings. Stating that “phosphorylation ……is required to modulate proprotein convertase recognition and/or trafficking” is vague and leaves open the possibility that it modulates in either direction, which our data do not support as described in point 1 above.

eLife Assessment

In their valuable study, Bracey et al. investigate how microtubule organization within pancreatic islet beta cells supports optimal insulin secretion. Using a combination of live imaging and photo-kinetic assays in an in vitro culture system, they provide convincing evidence that kinesin-1-mediated microtubule sliding, which plays critical roles in neurons and embryos, also plays a critical role in forming the sub-membranous microtubule band in response to glucose in beta cells. This work will be of interest to cell biologists studying cytoskeletal dynamics and organelle trafficking, as well as to translational biologists focused on diabetes.

Reviewer #1 (Public review):

This study investigates the role of microtubules (MT) in regulating insulin secretion from pancreatic islet beta cells. This is of great importance considering that controlled secretion of insulin is essential to prevent diabetes. Previously, it has been shown that KIF5B plays an essential role in insulin secretion by transporting insulin granules to the plasma membrane. High glucose activates KIF5B to increase insulin secretion resulting in cellular uptake of glucose. In order to prevent hypoglycemia, insulin secretion needs to be tightly controlled. Notably, it is known that KIF5B plays a role in MT sliding. This is important, as the authors described previously that beta cells establish a peripheral sub-membrane MT array, which is critical for withdrawal of excessive insulin granules from the secretion sites. At high glucose, the sub-membrane MT array is destabilized to allow for robust insulin secretion. Here the authors aim to answer the question how the peripheral array is formed. Based on the previously published data the authors hypothesize that KIF5B organizes the sub-membrane MT array via microtubule sliding.

General comment:<br /> This manuscript provides data that indicate that KIF5B, like in many other cells, mediates MT sliding in beta cells to establish a non-radial sub-membrane MT array. This study is based mainly on in vitro assays and one cell line. To demonstrate the importance of KIF5B in vivo/under physiological conditions, the MT pattern and directionality in beta cells within whole isolated pancreatic islets from KIF5B KO mice was analyzed in comparison to their WT littermates. While the presented effects appear often rather small, it is important to note that small changes in MT configuration can have strong effects. However, the authors provide no link to insulin secretion and glucose uptake. Finally, it remains unclear whether a KIF5B-dependent mechanism regulating microtubule sliding plays a major role in controlling insulin secretion.

Specific comments:<br /> (1) It is difficult to appreciate that there is a "peripheral sub-membrane microtubule array" as it is not well defined in the manuscript. This reviewer assumes that this is in the respective field clear. Yet, while it is appreciated that there is an increased amount of MTs close to the cytoplasmic membrane, the densities appear very variable along the membrane. Please provide a clear description in the Introduction what is meant with "peripheral sub-membrane microtubule array".<br /> (2) The authors described a "consistent presence of a significant peripheral array in the C57BL/6J control mice, while the KO counterparts exhibited a partial loss of this peripheral bundle. Specifically, the measured tubulin intensity at the cell periphery was significantly reduced in the KO mice compared to their wild-type counterparts". In vitro "control cells had convoluted non-radial MTs with a prominent sub-membrane array, typical for β cells (Fig. 2A), KIF5B-depleted cells featured extra-dense MTs in the cell center and sparse receding MTs at the periphery (Fig. 2B,C)". Please comment/discuss why in vivo there are no "extra-dense MTs in the cell center".<br /> (3) Authors should include in the Discussion a paragraph discussing the fact that small changes in MT configuration can have strong effects.

Reviewer #2 (Public review):

This elegant study provides significant and impactful insights into the factors contributing to the distinct arrangement of sub-membrane microtubules within mouse β-cells of the pancreas. The authors propose that in these cells, the motor protein KIF5B plays a crucial role in sliding existing microtubules toward the cell periphery and aligning them with one another along the plasma membrane. Furthermore, similar to other physiological features of β-cells, high glucose levels enhance this microtubule sliding process. A precise arrangement of microtubules beneath the cell membrane in β-cells is vital for the regulated secretion of pancreatic enzymes and hormones; thus, KIF5B has a significant role in pancreatic activity in both healthy conditions and diseases. The authors support their model by demonstrating that the levels of KIF5B mRNA in MIN6 cells are higher than those of other known kinesins. They show that microtubule sliding becomes less efficient when KIF5B is genetically silenced using two different short hairpin RNAs (shRNAs). Additionally, silencing of KIF5A in the same cells results in a general reorganization of microtubules throughout the cell. Specifically, while control cells exhibit a convoluted and non-radial arrangement of microtubules near the cell membrane, KIF5B-depleted cells display a sparse and less dense sub-membrane array of microtubules. Based on these findings, the authors conclude that the loss of KIF5B strongly affects the localization of microtubules to the cell periphery. Using a dominant-negative approach, the authors also demonstrate that KIF5B facilitates the sliding of microtubules by binding to cargo microtubules through the kinesin-1 tail binding domain. They present evidence suggesting that KIF5B-mediated microtubule sliding is glucose-dependent, similar to the activity levels of kinesin-1, which increase in the presence of glucose. Lastly, they show that this is glucose-dependent.

Strengths:

This study unveils a previously unexplained mechanism that regulates the specific rearrangement of microtubules beneath the cell membrane in pancreatic β-cells. The findings have significant implications because the precise regulation of the microtubule array at the secretion zone plays a critical role in controlling pancreatic function in both healthy and diseased states. The provided data supports the authors' conclusions well, and the study demonstrates the use of state-of-the-art methodologies, including quantification techniques and elegant dominant-negative experiments.

Weaknesses: None

Author response:

The following is the authors’ response to the original reviews

Public Reviews: 

Reviewer #1 (Public Review): 

This study investigates the role of microtubules in regulating insulin secretion from pancreatic islet beta cells. This is of great importance considering that controlled secretion of insulin is essential to prevent diabetes. Previously, it has been shown that KIF5B plays an essential role in insulin secretion by transporting insulin granules to the plasma membrane. High glucose activates KIF5B to increase insulin secretion resulting in the cellular uptake of glucose. In order to prevent hypoglycemia, insulin secretion needs to be tightly controlled. Notably, it is known that KIF5B plays a role in microtubule sliding. This is important, as the authors described previously that beta cells establish a peripheral sub-membrane microtubule array, which is critical for the withdrawal of excessive insulin granules from the secretion sites. At high glucose, the sub-membrane microtubule array is destabilized to allow for robust insulin secretion. Here the authors aim to answer the question of how the peripheral array is formed. Based on the previously published data the authors hypothesize that KIF5B organizes the sub-membrane microtubule array via microtubule sliding. 

General comment: 

This manuscript provides data that indicate that KIF5B, like in many other cells, mediates microtubule sliding in beta cells. This study is limited to in vitro assays and one cell line. Furthermore, the authors provide no link to insulin secretion and glucose uptake and the overall effects described are moderate. Finally, the overall effect of microtubule sliding upon glucose stimulation is surprisingly low considering the tight regulation of insulin secretion. Moreover, the authors state "the amount of MT polymer on every glucose stimulation changes only slightly, often undetectable…. In fact, we observe a prominent effect of peripheral MT loss only after a long-term kinesin depletion (three-four days)". This challenges the view that a KIF5Bdependent mechanism regulating microtubule sliding plays a major role in controlling insulin secretion. 

(1) Our initial study was indeed done in a cell line, which is a normal approach to addressing molecular mechanisms of a phenomenon in a challenging cell model: primary pancreatic beta cells are prone to rapidly dedifferentiate outside of the organism and are hard to genetically modify. To address this reviewer’s comment, in the revised manuscript we now confirm the phenotype in beta cells within intact pancreatic islets from a KIF5B KO mouse model (New Figure 2 – Supplemental Figure 1).

(2) We agree that testing the effect of microtubule sliding on insulin secretion is an important question. Unfortunately, the experimental design needed to accomplish this task is not straighDorward. Importantly, besides microtubule sliding, KIF5B is heavily engaged in insulin granule transport, and GSIS deficiency upon KIF5B inactivation is well documented (e.g. Varadi et al 2002). In this study, we choose not to repeat this GSIS assay because of ample existing data. However, this reported GSIS deficiency could result from a combination of lack of insulin granule delivery to the periphery (previous data) and from the depletion of insulin granules from the periphery due to the loss of the submembrane MT bundle (this study and Bracey et al 2020).  In order to exclusively test the role of MT sliding in secretion, a significant investment in mutant tool development would be needed. Ideally, a new mutant mouse model where insulin granule transport is allowed by MT sliding in blocked must be developed to specifically address this question. To conclude, answering this question will be the subject for another, follow-up study. 

(3) We respecDully disagree with the reviewer’s opinion that the effect of MT sliding in beta cells is moderate. As MT networks go, even a slight change in MT configuration often has dramatic consequences. For example, in mitotic spindles, a tiny overgrowth of microtubule ends during metaphase, which causes them to attach to both kinetochores rather than just one, is very significant for the efficiency of chromosome segregation, causing aneuploidy and cancer. The changes in beta-cell MT networks that we are reporting are much stronger: the effect on the peripheral MT network accumulated over three days of KIF5B depletion is dramatic (Fig 2 B, C). Short-term gross MT network configurations after a single glucose stimulation are harder to detect, but MTs at the cell periphery are, in fact, destabilized and fragmented, as we and others have previously reported (Ho et al 2020, Mueller et al 2021). Preventing this MT rearrangement completely blocks GSIS (Zhu et al 2015, Ho et al 2020). 

One of the most fascinating features of insulin secretion regulation is that the amount of generated insulin granules significantly exceeds the normal physiological needs for insulin secretion (~100 times more than needed). At the same time, even slightly facilitated glucose depletion can be devastating. Accordingly, the excessive insulin content of a beta cell resulted in the development of multiple levels of control, preventing excessive secretion. Our previous data suggest that the peripheral MT array provides one of those mechanisms. This study indicates that microtubule sliding is necessary to form the proper peripheral network in the long term. Short-term glucose-induced changes in the peripheral MT array likely need to be subtle to prevent over-secretion. Thus, we are not surprised that a dramatic effect of sliding inhibition is only detectable by our approaches after the changes in the MT network accumulate over time. In the revised paper, we now discuss the potential impact of peripheral MT sliding on positive and negative regulation of secretion and add a schematic model illustrating these processes.

Specific comments: 

(1) Notably, the authors have previously reported that high glucose-induced remodeling of microtubule networks facilitates robust glucose-stimulated insulin secretion. This remodeling involves the disassembly of old microtubules and the nucleation of new microtubules. Using real-time imaging of photoconverted microtubules, they report that high levels of glucose induce rapid microtubule disassembly preferentially in the periphery of individual β-cells, and this process is mediated by the phosphorylation of microtubule-associated protein tau. Here, they state that the sub-membrane microtubule array is destabilized via microtubule sliding. What is the relevance of the different processes? 

In this comment, the summary of our previous conclusions is correct, but the conclusion of this current study is re-stated incorrectly. Indeed, we have previously shown that in high glucose, MTs are destabilized at the cell periphery and nucleated in the cell interior. However, this current paper does not state that “the sub-membrane microtubule array is destabilized via microtubule sliding”. To answer this reviewer’s question, our data support a model where, during glucose stimulation, MT sliding within the peripheral bundle might move fragments of MTs severed by other mechanisms. Importantly, we propose that MT sliding restores the partially destabilized peripheral bundle by delivery of MTs that are nucleated at the cell interior and incorporating them into that bundle. In our overall model, three processes (destabilization, nucleation, and sliding to restore the bundle) are coordinated to maintain beta cell fitness on each GSIS cycle.

(2) On one hand the authors describe how KIF5B depletion prevents sliding and the transport of microtubules to the plasma membrane to form the sub-membrane microtubule array. This indicates KIF5B is required to form this structure. On the other hand, they describe that at high glucose concentration, KIF5B promotes microtubule sliding to destabilize the sub-membrane microtubule array to allow robust insulin secretion. This appears contradictory. 

We never intended to make an impression that MT sliding destabilized the sub-membrane bundle. Apologies if there was a reason in our wording that caused this misunderstanding of our model. We propose that while the bundle is destabilized downstream of glucose signaling (e.g. due to tau phosphorylation, please see Ho et al Diabetes 2020), MT sliding remodels the bundle and thereafter rebuilds it to prevent over-secretion. In the revised manuscript, we have doublechecked the whole text to make sure that such misunderstanding is avoided. 

(3) Previously, it has been shown that KIF5B induces tubulin incorporation along the microtubule shaft in a concentration-dependent manner. Moreover, running KIF5B increases microtubule rescue frequency and unlimited growth of microtubules. Notably, KIF5B regulates microtubule network mass and organization in cells (PMID: 34883065). Consequently, it appears possible that the here observed phenomena of changes in the microtubule network might be due to alterations in these processes. 

We thank the reviewer for proposing this alternative explanation to the observed change in microtubule networks after KIF5B depletion. We have now directly tested this possibility. Namely, we have re-expressed the kinesin-1 motor domain in MIN6 cells depleted of KIF5B. This motor domain construct by itself is not capable of driving microtubule sliding because it lacks the tail domain. At the same time, it is known to move very efficiently at microtubules and should provide the effects as reported in the article cited by the reviewer. We found that the reexpression of the kinesin motor domain does not rescue microtubule network defects in beta cells (see new Figure 2 – Supplemental Figure 2). Thus, we conclude that the effects of kinesin depletion on the microtubule network in beta cells are due to the lack of microtubule sliding, as reported here.

(4) The authors provide data that indicate that microtubule sliding is enhanced upon glucose stimulation. They conclude that these data indicate that microtubule sliding is an integral part of glucose-triggered microtubule remodeling. Yet, the authors fail to provide any evidence that this process plays a role in insulin secretion or glucose uptake. 

We would like to point out that we do not “fail” but rather choose not to overload our study by repeating insulin secretion assays in KIF5B-inactivated cells because this would not have been very informative. It has been found previously that kinesin-1 inactivation or knockout significantly attenuates insulin secretion because kinesin-1 is actively transporting insulin granules and kinesin-1 activity is enhanced under high glucose conditions (e.g. Varadi et al 2002, Cui et al., 2011, Donelan et al, 2002). That said, our current finding is very much in line with these previous data. When kinesin is depleted, two things would be happening at the same time: in the absence of sub-membrane microtubule bundle pre-existing insulin granules would be over-secreted, and new insulin would not be delivered to the periphery, both decreasing GSIS. Unfortunately, we do not have tools yet that would allow us to dissect which part of the insulin secretion defect is due to prior over-secretion (the consequence of deficient MT sliding) and which part is due to the lack of new granule delivery. We plan to develop such tools in the future and elaborate on them in a follow-up study. Here, our goal is to understand microtubule organization principles in beta cells, and we choose not to extend the scope of the current study to metabolic assays.  

(5) The authors speculate that the sub-membrane microtubule array prevents the over-secretion of insulin. Would one not expect in this case a change in the distribution of insulin granules at the plasma membrane when this array is affected? Or after glucose stimulation? Notably, it has been reported that "the defects of β-cell function in KIF5B mutant mice were not coupled with observable changes in islet morphology, islet cell composition, or β-cell size" and "the subcellular localization of insulin vesicles was found to not be affected significantly by the decreased Kif5b level. The cytoplasm of both wild-type and mutant β-cells was filled with insulin vesicles. Insulin vesicle numbers per square μm were determined by counting all insulin vesicles in randomly photographed β-cells. More insulin granules were found in Kif5b knockout β-cells compared with control cells. This phenomenon is consistent with the observation that insulin secretion by β-cells is affected" whereby "Insulin vesicles (arrowheads) were distributed evenly in both mutant and control cells" (PMID: 20870970).  

Quantitative analyses in the study cited by the reviewer do not include assays that would be relevant to our study. Particularly, in that study neither the amount of insulin granules at the cell periphery nor the ratio between the number of granules at the periphery and the beta cell interior has been analyzed. In addition, in our preliminary observations not shown here, insulin content in beta cells in KIF5B KO mice is highly heterogeneous, with a subpopulation of cells severely depleted of insulin. This opens a new avenue of investigation into beta cell heterogeneity, which is out of the scope of this current study. Thus, we chose to restrict this current study to microtubule organization data.   

(6) Does the sub-membrane microtubule array exist in primary beta cells (in vitro and/or in vivo) and how it is affected in KIF5B knockout mice?  

Yes, it does exist. In fact, we have first reported it in mouse islets (Bracey et al 2020, Ho et al 2020). Now, we report that the sub-membrane bundle is defective, and microtubules are misaligned in KIF5B KO mice (new Figure 2 – Supplemental Figure 1).

Reviewer #2 (Public Review): 

In this article, Bracey et al. provide insights into the factors contributing to the distinct arrangement observed in sub-membrane microtubules (MTs) within mouse β-cells of the pancreas. Specifically, they propose that in clonal mouse pancreatic β-cells (MIN6), the motor protein KIF5B plays a role in sliding existing MTs towards the cell periphery and aligning them with each other along the plasma membrane. Furthermore, similar to other physiological features of β-cells, this process of MTs sliding is enhanced by a high glucose stimulus. Because a precise alignment of MTs beneath the cell membrane in β-cells is crucial for the regulated secretion of pancreatic enzymes and hormones, KIF5B assumes a significant role in pancreatic activity, both in healthy conditions and during diseases. 

The authors provide evidence in support of their model by demonstrating that the levels of KIF5B mRNA in MIN6 cells are higher compared to other known KIFs. They further show that when KIF5B is genetically silenced using two different shRNAs, the MT sliding becomes less efficient. Additionally, silencing of KIF5A in the same cells leads to a general reorganization of MTs throughout the cell. Specifically, while control cells exhibit a convoluted and non-radial arrangement of MTs near the cell membrane, KIF5B-depleted cells display a sparse and less dense sub-membrane array of MTs. Based on these findings, the Authors conclude that the loss of KIF5B strongly affects the localization of MTs to the periphery of the cell. Using a dominant-negative approach, the authors also demonstrate that KIF5B facilitates the sliding of MTs by binding to cargo MTs through the kinesin-1 tail binding domain. Additionally, they present evidence suggesting that KIF5B-mediated MT sliding is dependent on glucose, similar to the activity levels of kinesin-1, which increase in the presence of glucose. Notably, when the glucose concentrations in the culturing media of MIN6 cells are reduced from 20 mM to 5 mM, a significant decrease in MT sliding is observed. 

Strengths:

This study unveils a previously unexplained mechanism that regulates the specific rearrangement of MTs beneath the cell membrane in pancreatic β-cells. The findings of this research have implications and are of significant interest because the precise regulation of the MT array at the secretion zone plays a critical role in controlling pancreatic function in both healthy and diseased states. In general, the author's conclusions are substantiated by the provided data, and the study demonstrates the utilization of state-of-the-art methodologies including quantification techniques, and elegant dominant-negative experiments. 

Weaknesses:

A few relatively minor issues are present and related to data interpretation and the conclusions drawn in the study. Namely, some inconsistencies between what appears to be the overall and sub-membrane MT array in scramble vs. KIF5B-depleted cells, the lack of details about the sub-cellular localization of KIF5B in these cells and the physiological significance of the effect of glucose levels in beta-cells of the pancreas. 

We thank the reviewer for this insighDul review. In the revised version, we provided re-worded and extended interpretations and conclusions to prevent any issues or misunderstandings.  We trust that while some noted apparent inconsistencies may reflect the intrinsic heterogeneity of the beta cell population, all data presented here indicate the same trend in phenotypes.  In the revised version, we have provided additional cell views and, in places, alternative representative images and videos, to clear out any apparent inconsistencies. We also would like to point out that we in fact reported KIF5B localization: not surprisingly, KIF5B predominantly localized to insulin granules and the punctate staining fills the whole cytoplasm (Figure 2A, bottom panel). However, as pointed out in detail in our response to reviewer 1, we choose to leave out an extensive study of the physiological and metabolic consequences of the reported microtubule network dynamics to a follow-up study. 

Reviewer #3 (Public Review): 

Prior work from the Kaverina lab and others had determined that beta-cells build a microtubule network that differs from the canonical radial organization typical in most mammalian cell types and that this organization facilitates the regulated secretion of insulin-containing secretory granules (IGs). In this manuscript, the authors tested the hypothesis that kinesin-driven microtubule sliding is an underlying mechanism that establishes a sub-membranous microtubule array that regulates IG secretion. They employed knock-down and dominant-negative strategies to convincingly show microtubule sliding does, in fact, drive the assembly of the sub-membranous microtubule band. They also used live cell imaging assays to demonstrate that kinesin-mediated microtubule sliding in beta-cells is triggered by extracellular high glucose. Overall, this is an interesting and important study that relates microtubule dynamics to an important physiological process. The experiments were rigorous and well-controlled. 

We truly appreciate this reviewer’ opinion. 

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors): 

Figures: 

(1) Figure 1: 

a) Why can one not see here, and in most following images, the peripheral sub-membrane microtubule array? One can also not see an accumulation of microtubules in the cell interior. 

Microtubule pattern in beta cells is variable, and the sub-membrane array is seen in the whole population to a variable extent (see directionality histogram in Figure 2E for statistics). In fact, an array of peripheral MTs parallel to the cell border is present in the example shown in Figure 1 and in all following control images. To make it clearer, we now show the pre-bleach images in Figure 1 D-F at a lower magnification, so that the differences in MT density at the cell periphery and cell center are more clearly seen: MTs lack at the periphery in KF5B-depleted but not the control cells.  

b) 5 min appears to be a long time and enough time to polymerize a significant number of new microtubules. 

We interpret this comment as the reviewer’s concern that in FRAP assays, fluorescently-labeled MTs moving into the bleached area might be newly polymerizing MTs rather than preexisting MT relocated into that area. However, this is not the case because newly polymerized MTs contain predominantly quenched “dark” tubulin molecules and only a small percent of fluorescent tubulin. These dim MTs are not included in MT sliding assay analysis, where a threshold for bright MTs is introduced. Now, we added more details for the quantification of these data to Materials and Methods section.

c) The overall effects appear minor. It is unclear how Fig. 1-Suppl-Fig.1, where no significant difference is shown, is translated into Figure 1 J and K showing a significant difference. 

With all due respect, we do not agree that the effect is minor. Please see our response to the Public Review where we discuss the major consequences of MT defects in detail. 

To answer this specific comment, we show that there are significant differences in the number of rapidly moving MTs (5-sec displacement over 0.3 µm) and in the amount of stationary MTs (5sec displacement is below 0.15 µm). There is no significant difference in the amount of slightly displaced MTs (displacements between 0.15 and 0.3 µm; the central part of the histogram). This might indicate that these slight displacements do not depend on kinesin-1 motor but rather are caused by experimental noise, pushing by moving organelles, and/or myosin-dependent forces in the cell. In the revised manuscript, we have this quantification more clearly detailed in Methods and included in Figure legends.

d) The authors utilize single molecule tracking to further strengthen their conclusion that KIF5B promotes microtubule sliding. The observed effects are weaker than the data obtained from photobleaching experiments. The videos clearly show that there is still significant movement also in KIF5B-depleted cells. If K560RigorE236A binds irreversibly to a microtubule and this microtubule is growing (not only by the addition of tubulin dimers to the plus end; see PMID: 34883065) wouldn't that also result in movement of the tagged K560RigorE236A? As KIF5B is also required in the transport of insulin granules, it should also label "interior microtubules". And in Video 2 it appears that pretty much all "labeled" microtubules are moving. 

K560RigorE236A forms fiducial marks along the whole MTs lattice, as previously shown in (Tanenbaum et al., 2014). When it is bound to MT lattice, K560RigorE236A moves with the whole MT if it is being relocated. The mechanism described in (PMID: 34883065) appears to be absent or minor in beta cells (see Figure 2- Supplemental Figure 2), thus, even if this mechanism would displace already polymerized MTs, this is not happening in this cell type.

The reviewer is correct, K560RigorE236A does mark all MTs throughout a beta cell. All MTs are moving slightly in a living cell because they are pushed around by moving organelles, actin contractility, etc. MTs may also be slid by other MT-dependent motors (dynein against the membrane and such). So, it is not surprising that the MT network is “breezing,” and kinesindependent sliding is only a part of MT movement. What we show here is that the KIF5Bdependent MT sliding is responsible for a relatively “long-distance” relocation of MTs manifested in long, directional displacement of fiducial marks.  This does not exclude other movements. This makes extraction of kinesin-dependent MT movements somewhat challenging, of course, that is why we needed to do those extensive analyses. 

e) Figure 1 G to K is misleading, at least in the context of the provided videos. There are several microtubules that move extensively in shRNA#2-treated cells and overall there appears more movement in this cell as in the control cell. Figure 1I is clearly not representative of the movement shown in Video 2. 

We apologize if our selection of representative movies/figures for this experiment was imperfect. Indeed, in all depleted cells, SunTag puncta still move to a certain extent, either due to incomplete depletion or to alternative intracellular forces dislocating microtubules. However, there is a clear difference in the fraction of persistently moving puncta (please see Figure 1K and  histogram in Figure 1 - Supplemental Figure 1B). Unfortunately, when the number of SunTag puncta per a cell is variable, it sometimes prevents a good visual perception of the actual distribution of moving versus stationary microtubules. We now show an alternative representative movie for the Figure 1I and the corresponding Video 2, with a goal to compare cells with more consistent numbers of Sun-Tag puncta.

(2) Figure 2A. 

a) This is the only image that clearly shows the existence of a sub-membrane microtubule array and the concentration of microtubules in the cell interior. The differences are unclear between the experimental setups including the length of cultivation and knockdown of KIF5B or expression of mutants. 

We now provide a more detailed description of each image acquisition and processing in Materials and Methods. In brief, while the morphology of MT patterns is intrinsically variable in beta cells, all control cells have populated peripheral MTs that exhibit a more parallel configuration as compared to depletions and mutants.

b) The authors state "While control cells had convoluted non-radial MTs with a prominent sub-membrane array, typical for beta cells (Fig. 2A), KIF5B-depleted cells featured extra-dense MTs in the cell center and sparse reseeding MTs at the periphery (Fig. 2B, C)". Could that not be explained with the observation that "Kinesin-1 controls microtubule length" (PMID: 34883065)? 

Thank you for this interesting alternative idea. It does not appear to be the case for beta cells.

Please see Figure 2-Supplemental Figure 2  and our response to Public Review Comment #3.

Also, our apologies for the typo in the original manuscript: this is “receding” nor “reseeding”.

(3) Figure 3: 

a) This is an elegant way to determine whether KIF5B is involved in microtubule sliding independent of the fact that the effect appears very small. 

Thank you!            

b) The assay depends on ectopic expression of a dominant negative mutant. It appears important to show that KIFDNwt is high enough expressed to indeed block the binding of endogenous KIF5B. The authors need to provide a control for this. Furthermore, authors need to provide evidence that other functions of KIF5B are not impaired such as transport of insulin granules and tubulin incorporation or microtubule stability and length.

Expression of cargo-binding motor domains routinely causes a dominant-negative effect of their cargo transport. This exact construct has been used for the purpose of dominant-negative action previously (Ravindran et al., 2017). It does prevent the membrane cargo binding of KIF5B (Ravindran et al., 2017), thus the transport of insulin granules is also impaired in overexpression cells. Confirming this fact would not influence our study conclusions, so we chose not to repeat these assays for the sake of time.

c) N-numbers should be similar. The data for KIFDNmut are difficult to interpret with possibly 2 experiments showing little to no displacement and 3 showing displacement. 

In the revised manuscript, additional data have been added to increase N-numbers.

(4) Figure 4 and supplements: The morphology of the KIFDNwt cells is greatly affected and this makes it difficult to say whether the effect on microtubules at the cell periphery is a direct or indirect effect. 

Yes, these cells often have less spread appearance, obscuring visual perception of MT distribution. We have now replaced the image of KIFDNwt cell (Figure 4, Supplemental Figure 1 A) to a more visually representative example.

Things to do: 

(1) Notably, the authors have previously reported that high glucose-induced remodeling of microtubule networks facilitates robust glucose-stimulated insulin secretion. This remodeling involves the disassembly of old microtubules and the nucleation of new microtubules. Here, they state that the sub-membrane microtubule array is destabilized via microtubule sliding. What is the relevance of the different processes? Please discuss these in the manuscript. 

Thank you, we have now extended our discussion of these points and our prior findings. We have also added a schematic model figure for clarity (Figure 7).  

(2) 5 min appears to be a long time and enough time to polymerize a significant number of new microtubules. Do the authors have any information about the speed of MT formation in MIN6 cells? Can the authors repeat this experiment by preventing MT polymerization? Or repeat the experiment with EB1/EB3 reporter to visualize microtubule growth in the same experimental setting? 

While some MT polymerization will happen in this timeframe, newly polymerized MTs contain predominantly quenched “dark” tubulin molecules and only a small percent of fluorescent tubulin. These dim MTs are not included in MT sliding assay analysis, where a threshold for bright MTs is introduced. We apologize for initially omitting certain details from the FRAP assay analysis. Now these details have been added.   

Are the microtubules shown on the cell surface (TIRF microscopy) or do we see here all microtubules? 

Please see Materials and Methods for microscopy methods and image processing for each figure. Specifically, FRAP assays show a maximum intensity projection of spinning disk confocal stacks over 2.4µm in height (approximately the ventral half of a cell).

(3) Previously, it has been shown that KIF5B induces tubulin incorporation along the microtubule shaft in a concentration-dependent manner. Moreover, running KIF5B increases microtubule rescue frequency and unlimited growth of microtubules. Notably, KIF5B regulates microtubule network mass and organization in cells (PMID: 34883065). Consequently, it appears possible that the here observed phenomena of changes in the microtubule network might be due to alterations in these processes. Authors need to exclude these possibilities and discuss them. 

Thank you for this interesting alternative idea. It does not appear to be the case for beta cells. Please see Figure 2-Supplemental Figure 2  and our response to Public Review Comment #3.

(4) It is important that the authors describe in the text and possibly in the figure legends the differences between the experimental set-ups including the length of cultivation and knock down of KIF5B or expression of mutants. 

Thank you, please see these details in the text (Materials and Methods section).

(5) Figure 5: Does KIF5B depletion rescue the kinesore-induced defects 

Thank you for suggesting this control. We have now conducted corresponding experiments. The answer is yes, it does. Kinesore does not induce detectable changes in MT patterns in KIF5Bdepleted cells (new Figure 5-Supplemental Figure 2). 

(6) Can the authors block kinesin-1 resulting in microtubule accumulation in the cell center and then release the block, and best inhibiting microtubule formation, to see whether the microtubules accumulated in the cell center will be transported to the periphery? 

This proposed experiment would have been a nice illustration to the study, however it has proven to be too challenging. Unfortunately we have to leave it for the future studies. However,  the experiments already included in the paper are sufficient to prove our conclusions. 

Minor comments: 

(1) The English needs to be improved. Oaen it is unclear what the authors try to convey. The manuscript is difficult to read and contains several overstatements. 

The revised manuscript has been through several rounds of proof-reading for clarity.

(2) It is important to describe in more detail in the introduction what is known about KIF5B in beta cells. Previously, it has been demonstrated that silencing, or inactivation by a dominant negative form of KIF5B, blocks the sustained phase of glucose-stimulated insulin secretion (PMID: 9112396, PMID: 12356920, PMID: 20870970). 

Yes, this is of course very important and have been cited in the original manuscript. Now, we have expanded the discussion on the matter.

(3) Figure 1B and Fig. 1 Suppl Fig.1: Please provide band sizes and provide information on the size of KIF5B. 

We have replaced Fig. 1B and Suppl Fig 1A with quantitative analysis of KIF5B depletion, not found in new Fig. 1B and Suppl Fig. 1A-C. 

(4) It is important to state the used glucose concentrations in Figure 1D (based on the methods section it is probably 25 mM glucose) and all subsequent experiments. Is this correct and comparable to Figure 6A or B? For the non-specialized reader, more information should be provided on why initial glucose starvation is performed.  

Cell culture models of pancreatic beta cells are routinely maintained at glucose levels that at considered “high”, or stimulatory for secretion. This is needed to prevent the loss of cells’ capacity to respond to glucose stimulation over generations. In order to test GSIS, cells need to be equilibrated at low (fasting, standardly 2.8mM) glucose levels for several hours, so that they are capable of secreting insulin upon glucose addition. 25mM glucose is normally used to stimulate GSIS in cell culture models of beta cells, like MIN6. This is a higher concentration as compared to what is needed to stimulate primary beta cells in islets.

Reviewer #2 (Recommendations For The Authors): 

I have the following specific questions that pertain to data interpretation and the conclusions drawn.

(1) The morphology of the overall MT array before the bleach treatment in both control cells and KIF5B-KD cells depicted in Figure 1D-F and Figure 2A-C appears to be distinct. In Figure 1, it seems that the absence of KIF5B results in a general augmentation of MT mass, whereas the arrangement presented in Figure 2 indicates the contrary. Even in the sub-membrane areas, this phenomenon appears to hold true. However, the images used in this study, which depict entire cells or a significant portion of cells, may not be ideal for visualizing the sub-membrane regions.

It would be beneficial if the author could offer some explanations for this apparent inconsistency. 

While beta cell population is intrinsically heterogeneous, all data presented here indicate the same trend in phenotypes. Possibly, some apparent inconsistency between figure 1 and 2 appeared because in the original manuscript we did not show the pre-bleach whole-cell overview in Figure 1. In the revised version, we now show the whole cells for pre-bleach so that MT organization at the cell periphery can be assessed. Please note that in the control cell, MTs are more or less equally distributed over the cell, while in KIF5B depletions the cell periphery is significantly less populated than the cell center. Furthermore, we did not detect MT mass augmentation or increase in KIF5B depletions. One possible explanation for such reviewer’s impression from Figure 2 is that Figure 2 F-H shows thresholded images where threshold was adjusted to highlight peripheral MTs in each cell. Please note that this is not the same threshold for each cell (see Figure 2 - Supplemental Figure 2 and 3). Thus, KIF5B-depleted cells that have fewer MTs at the periphery appear brighter in these thresholded images. For the true comparison of MT intensity, please see Figure 2 A-C (grayscale image, not the threshold).

(2) It would be helpful if the author could provide a visual representation or comment on the sub-cellular localization of KIF5B in MIN6 cells. Is it predominantly localized in the submembrane region, or is it more evenly distributed throughout the cytoplasm? 

Please see Fig 2A, lower panel. KIF5B is seen across the cell as a punctate staining, in agreement with previous findings that it mostly localize at IGs.

(3) The alteration in microtubule (MT) organization and sliding in the absence of KIF5B seems to initiate in proximity to the apparent microtubule organizing center (MTOC) depicted in Figure 2A, and then "simply" extends towards the sub-membrane region. Although the authors acknowledge it, it would be advantageous for the readers to have a clearer indication that the sub-membrane microtubule (MT) reorganization in the absence of KIF5B is a result of a broader MT reorganization rather than a specific occurrence restricted to the sub-membrane regions. 

Thank you for this comment. We now extend our discussion to clearer state our conclusions and interpretations of this point. We also have added a schematic Figure 7 as an illustration. 

(4) Regarding the "glucose experiments," it is common to add 20-25 mM glucose to culture media, but physiological concentrations of glucose typically hover around 5 mM. Therefore, it is somewhat unclear what the implications are when investigating the impact of KIF5B depletion on MT sliding at 2.8 mM of glucose. It would be helpful if the authors could provide some commentary on this matter, particularly in relation to physiological and pathological conditions. 

2.8 mM glucose is a standard low glucose condition used to model glucose deprivation/fasting. For functional primary beta cells within pancreatic islets, GSIS can be triggered by glucose stimulation as low as 8-12 mM glucose. However, for glucose stimulation of cultured beta cells such as MIN6 used in this paper, 20-25 mM glucose is standardly used because these cell lines have a higher threshold of stimulation compared to primary beta cells and whole islets.

(5) In supplementary Figure 1A, it would be helpful if the lanes in the WB were marked indicating what is what. In my observation, it appears that Supplementary Figure 1A, particularly lanes #2, 3, and 4, display the GAPDH protein (MW 36 kDa) (or is it alpha-tubulin, as mentioned in the Material and Methods section and indicated in lane #409?) relative to Figure 1A. I am curious about KIF5B (MW 108 kDa). Is it represented by the upper band? Did the author probe the same membrane simultaneously with two different primary antibodies? This should be clarified, and the author should indicate the molecular weight of the ladder. 

Indeed, in the original WB two antibodies have been used together, due to a challenge in collecting a sufficient number of shRNA-expressing beta cells. It caused a confusion and improper interpretation of the loading control. We thank the reviewer for catching this.  We have now replaced old Fig. 1B and Suppl. Fig. 1A with quantitative analysis of KIF5B depletion based on single-cell immunofluorescent staining. It is now found in new Fig. 1B and Suppl Fig. 1A-C.  

Reviewer #3 (Recommendations For The Authors): 

In all of the figures that present microtubule orientations (e.g. Figure 2E) the error bars obscure the vertical bins making them difficult to read or interpret. If they were rendered at a larger scale, it would be easier to read and interpret these results. 

Thank you pointing this out. We now show these histograms with a different format of error bars and without outliers that obscure the view. A variant with outliers is now shown in the supplement. 

Some of the callouts to the videos in the paper are inaccurate. Perhaps the authors reordered sections of the paper but failed to correctly renumber the video citations? 

Thank you for this comment, we have corrected all callouts now.

eLife Assessment

This important study provides evidence of a deeply conserved role for the gene Mirror in providing positional identity in the posterior part of butterfly and fly wings, despite increased morphological complexity of butterfly wings. The findings are solid for the field of evo-devo. However, the tools in butterflies are more limited than in Drosophila and it is more difficult to determine which specific cells are mutant and whether the effect of mutation is cell-intrinsic. The work will be of interest to evolutionary and developmental biologists working on insect wing evolution and the evolution of patterning more generally.

Reviewer #1 (Public review):

Summary:

This short report shows that the transcription factor gene mirror is specifically expressed in the posterior region of the butterfly wing imaginal disk, and uses CRISPR mosaic knock-outs to show it is necessary to specify the morphological features (scales, veins, and surface) of this area.

Strengths:

The data and figures support the conclusions. The article is swiftly written and makes an interesting evolutionary comparison to the function of this gene in Drosophila. Based on the data presented, it can now be established that mirror likely has a similar selector function for posterior-wing identity in a plethora of insects.

Comments on revisions:

The revision is satisfactory. I agree with the authors that this article provides interesting insights on the evolution of insect wings. Of note, butterfly and fly wing imaginal disks differ in their mode of development: while fly wing disks grow as epithelial sacs that evaginate during metamorphosis, butterfly wing disks develop as relatively flat epithelial sheets that expand and differentiate progressively. This makes the similar role of mirror all the more interesting.

The revised text appropriately discuss how selector genes like mirror regionalize the wing during larval and pupal development. This article makes a reasonable use of CRISPR mosaic knock outs and uses contralateral controls to show the nature of the phenotypic transformations.

Reviewer #2 (Public review):

This is a short and unpretentious paper. It is an interesting area and therefore, although much of this area of research was pioneered in flies, extending basic findings to butterflies would be worthwhile. Indeed, there is an intriguing observation but it is technically flawed and these flaws are far too serious to allow us to recommend publication

The authors show that mirror is expressed at the back of the wing in butterflies (as in flies). They present some evidence that is required for the proper development of the back of the wing in butterflies (a region dubbed the vannus by the ancient guru Snodgrass). But there are problems with that evidence. First, concerning the method, using CRISP they treat embryos and the expectation is that the mirror gene will be damaged in groups of cell lineages, giving a mosaic animal in which some lines of cells are normal for mirror and others not. We do not know where the clones or patches of cells that are defective for mirror are because they are not marked. Also, we do not know what part of the wing is wildtype and what part is mutant for mirror. When the mirror mutant cells colonise the back of the wing and that butterfly survives (many butterflies fail to develop), the back of the wing is altered in some selected butterflies. This raises a second problem: we do not know whether the rear of the wing is missing or transformed. From the images the appearance of the back of the wing is clearly different from wild type, but is that due to transformation or not? And then I believe we need to know specifically what us difference between the rear of the wing and the main part. What we see is a silvery look at the back that is not present in the main part, is it the structure of the scales? We are not told. There are other problems. Mirror is only part of a group of genes in flies and in flies both iroquois and mirror are needed to make the back of the wing, the alula (Kehl et al). What is known about iro expression in butterflies?

In flies, mirror regulates a late and local expression of dpp that seems to be responsible of making the alula. What happens in butterflies? Would a study of expression of Dpp in wildtype and mirror compromised wings be useful?

Thus, I find the paper to be disappointing for a general journal as it does little more than claim what was discovered in Drosophila is at least partly true in butterflies. Also it fails to explain what the authors mean by "wing domains" and "domain specification". They are not alone, butterfly workers in general appear vague about these concepts, their vagueness allowing too much loose thinking. Since these matters are at the heart of the purpose and meaning of the work reported here, we readers need a paper containing more critical thought and information. I would like to have a better and more logical introduction and discussion.

They do define what they mean by the vannus of the wing. In flies the definition of compartments is clear and abundantly demonstrated, with gene expression and requirement being limited precisely to sets of cells that display lineage boundaries. It is true that domains of gene expression in flies, for example, of the iroquois complex, which includes mirror, can only be related to pattern with difficulty. Some recap of what is known plus the opinion of the authors on how they interpret papers on possible lineage domains in butterflies might also be useful as the reader, is no wiser about what the authors might mean at the end of it!

The references are sometimes inappropriate. The discovery of the AP compartments should not be referred to Guillen et al 1995, but to Morata and Lawrence 1975.

Comments on revisions:

Nearly all the previous criticisms remain valid and are not discussed or overcome in the revision. The authors wish to draw their conclusions and we think they can do that, but they should make clear that key evidence is lacking. Thus their conclusions are speculative. But they present them more or less as facts. This is not justified. Let us suppose that clones lacking mirror do not survive or do not develop properly in the rear part of the wing and what they are seeing is occasional damage due to incomplete regeneration or to regenerative duplication?

Many clones in flies only include parts of one surface of the wing, could this happen here and how would it affect interpretations?

The null phenotype in the wing is not known but deduced from abnormal wings which "even in mKO..... appeared to have a mutant phenotype across the entire posterior region", a nice example of circular logic.

We believe the authors should be more objective and explain that their interpretations are not solid and that they should ideally be tested by finding ways of independently marking the clones. Other clonal mosaic experiments in butterflies have been done (eg https://journals.biologists.com/dev/article/150/18/dev201868/329659/Frizzled2-receives-WntA-signaling-during-butterfly) without cell autonomous independent markers, but they are more solid as transformed spots are made visible cell by cell by scale colour changes etc.

Their deduction that "mirror acts as a selector gene necessary to define the far posterior wing domain" is a speculative hypothesis, not a deduction and the readers should be so informed.

Reviewer #3 (Public review):

Summary:

The manuscript by Chatterjee et al., examines the role of the mirror locus in patterning butterfly wings. The authors examine the pattern of mirror expression in the common buckeye butterfly, Junonia coenia and then employ CRISPR mutagenesis to generate mosaic butterflies carrying clones of mirror mutant cells. They find that mirror is expressed in a well-defined posterior sector of final-instar wing discs from both hindwings and forewings and that CRISPR-injected larvae display a loss of adult wing structures presumably derived from the mirror expressing region of hindwing primordium (the case for forewings is a bit less clear since the mirror domain is narrower than in the hindwing, but there also do seem to be some anomalies in posterior regions of forewings in adults derived from CRISPR injected larvae). The authors conclude that wings of these butterflies have at least three different fundamental wing compartments, the mirror domain, a posterior domain defined by engrailed expression, and an anterior domain expressing neither mirror or engrailed. They speculate that this most posterior compartment has been reduced to a rudiment in Drosophila and thus has not been adequately recognized as a such a primary regional specialization.

Critique: This is a very straight-forward study and the experimental results presented support the key claims that mirror is expressed in a restricted posterior section of the wing primordium and that mosaic wings from CRISPR injected larvae display loss of adult wing structures presumably derived from cells expressing mirror (or at least nearby). The major issue I have with this paper is the strong interpretation of these findings that lead the authors to conclude that mirror is acting as a high level gene akin to engrailed in defining a separate extreme posterior wing compartment. To place this claim in context, it is important in my view to consider what is known about engrailed, for which there is ample evidence to support the claim that this gene does play a very ancestral and conserved function in a defining posterior compartments of all body segments (including the wing) across arthropods.

(A) engrailed is expressed in a broad posterior domain with a sharp anterior border in all segments of virtually all arthropods examined (broad use of a very good pan-species anti-En antibody makes this case very strong).

(B) In Drosophila, marked clones of wing cells (generated during larval stages) strictly obey a straight anterior-posterior border indicating that cells in these two domains do not normally intermix, thus, supporting the claim that a clear A/P lineage compartment exists.

In my opinion, mirror does not seem to be in the same category of regulator as engrailed for the following reasons:

(1) There is no evidence that I am aware of, either from the current experiments, or others that the mirror expression domain corresponds to a clonal lineage compartment. It is also unclear from the data shown in this study whether engrailed is co-expressed with mirror in posterior most cells of J. coenia wing discs? If so, it does not seem justified to infer that mirror acts as an independent determinant of the region of the wing where it is expressed.

(2) The mirror is not only expressed in a posterior region of the wing in flies but also in the ventral region of the eye. In Drosophila, mirror mutants not only lack the alula (derived approximately from cells where mirror is expressed), but also lacks tissue derived from the ventral region of the eye disc (although this ventral tissue loss phenotype may extend beyond the cells expressing mirror).

In summary, it seems most reasonable to me to think of mirror as a transcription factor that provides important development information for a diverse set of cells in which it can be expressed (posterior wing cells and ventral eye cells) but not that it acts as a high level regulator as engrailed.

Recommendation:

While the data provided in this succinct study are solid and interesting, it is not clear to me that these findings support the major claim that mirror defines an extreme posterior compartment akin to that specified by engrailed. Minimally, the authors should address the points outlined above in their discussion section and greatly tone down their conclusion regarding mirror being a conserved selector-like gene dedicated to establishing posterior-most fates of the wing. They also should cite and discuss the original study in Drosophila describing the mirror expression pattern in the embryo and eye and the corresponding eye phenotype of mirror mutants: McNeill et al., Genes & Dev. 1997. 11: 1073-1082; doi:10.1101/gad.11.8.1073.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public Review):

Summary:

This short report shows that the transcription factor gene mirror is specifically expressed in the posterior region of the butterfly wing imaginal disk, and uses CRISPR mosaic knock-outs to show it is necessary to specify the morphological features (scales, veins, and surface) of this area.

Strengths:

The data and figures support the conclusions. The article is swiftly written and makes an interesting evolutionary comparison to the function of this gene in Drosophila. Based on the data presented, it can now be established that mirror likely has a similar selector function for posterior-wing identity in a plethora of insects.

We thank the reviewer for their feedback.

Weaknesses:

This first version has minor terminological issues regarding the use of the terms "domains" and "compartment".

We acknowledge that the terminologies “domains” and “compartments” might lead to confusion. To avoid confusion we have removed the term “compartment” from the manuscript.

Reviewer #2 (Public Review):

This is a short and unpretentious paper. It is an interesting area and therefore, although much of this area of research was pioneered in flies, extending basic findings to butterflies would be worthwhile. Indeed, there is an intriguing observation but it is technically flawed and these flaws are serious.

The authors show that mirror is expressed at the back of the wing in butterflies (as in flies). They present some evidence that is required for the proper development of the back of the wing in butterflies (a region dubbed the vannus by the ancient guru Snodgrass). But there are problems with that evidence. First, concerning the method, using CRISP they treat embryos and the expectation is that the mirror gene will be damaged in groups of cell lineages, giving a mosaic animal in which some lines of cells are normal for mirror and others are not. We do not know where the clones or patches of cells that are defective for mirror are because they are not marked. Also, we do not know what part of the wing is wild type and what part is mutant for mirror. When the mirror mutant cells colonise the back of the wing and that butterfly survives (many butterflies fail to develop), the back of the wing is altered in some selected butterflies. This raises a second problem: we do not know whether the rear of the wing is missing or transformed. From the images, the appearance of the back of the wing is clearly different from the wild type, but is that due to transformation or not? And then I believe we need to know specifically what the difference is between the rear of the wing and the main part. What we see is a silvery look at the back that is not present in the main part, is it the structure of the scales? We are not told.

Thank you for this feedback. We appreciate that many readers may not accustomed to looking at mosaic knockouts. As discussed in a previous review article (Zhang & Reed 2017), we rely on a combination of contralateral asymmetry and replicates to infer mutant phenotypes. For many genes (e.g. pigmentation enzymes) mutant clones are obvious, but for other types of genes (e.g. ligands) clone boundaries are sometimes not directly diagnosable. It is simply a limitation of our study system. Nonetheless, you see for yourself that “the back of the wing is altered in some butterflies” – the effects of deleting mirror are clear and repeatable.

In terms of interpreting mutant phenotypes, we agree that that paper would benefit from a better description of the specific effects. Therefore, we have included an improved, more systematic description of phenotypes, along with better-annotated figures showing changes in wing shape and venation, scale coloration, and color pattern transformation (e.g. posterior elongation of the orange marginal stripes).

There are other problems. Mirror is only part of a group of genes in flies and in flies both iroquois and mirror are needed to make the back of the wing, the alula (Kehl et al). What is known about iro expression in butterflies?

In Drosophila mirror, araucan, and caupolican comprise the so-called Iroqouis Complex of genes. As denoted in Figure S4 and in Kerner et al (doi: https://doi.org/10.1186/1471-2148-9-74) the divergence of araucan and caupolican into two separate paralogs is restricted to Drosophila. As in most insects, butterflies have only two Iroquois Complex genes: araucan and mirror. We tested the role of araucan in Junonia coenia as shown in our pre-print: https://doi.org/10.1101/2023.11.21.568172. Its expression appears to be restricted to early pupal wings where it is transcribed in all scale-forming cells. Mosaic araucan KOs resulted in a change in scale iridescent coloration associated with changes in the laminar thickness of scale cells.  

In flies, mirror regulates a late and local expression of dpp that seems to be responsible for making the alula. What happens in butterflies? Would a study of the expression of Dpp in wildtype and mirror compromised wings be useful?

We thank the reviewer for the proposal and agree that a future study comparing Dpp in wild-type versus mirror KO butterflies would be useful to clarify the mechanism of Dpp signalling in wing development. It is not clear, however, that the results of a Dpp experiment would change the conclusions of our current study therefore we decided not to undertake these additional experiments for our revision.

Thus, I find the paper to be disappointing for a general journal as it does little more than claim what was discovered in Drosophila is at least partly true in butterflies. 

We respect that the reviewer does not have a strong interest in the comparative aspects of this study. Fair enough. This report is primarily aimed at biologists interested in the evolutionary history of insect wings.

Also, it fails to explain what the authors mean by "wing domains" and "domain specification". They are not alone, butterfly workers, in general, appear vague about these concepts, their vagueness allowing too much loose thinking.

A domain is “a region distinctively marked by some physical feature”. This term is used extensively in the developmental biology literature (e.g. “expression domain”, “embryonic domain”, “tissue domain”, “domain specification”) and is found throughout popular textbooks (e.g. Alberts et al. “The Cell”, Gilbert “Developmental Biology”). We prefer the term “domain” because of its association in the Drosophila literature with transcription factors that define fields of cells. We specifically avoided using the term “compartment” because of its association with cell lineage, which we have not tested. 

Since these matters are at the heart of the purpose and meaning of the work reported here, we readers need a paper containing more critical thought and information. I would like to have a better and more logical introduction and discussion.

We would like the very same thing, of course, and we hope the reviewer finds our revised manuscript to be more satisfying to read.

The authors do define what they mean by the vannus of the wing. In flies the definition of compartments is clear and abundantly demonstrated, with gene expression and requirement being limited precisely to sets of cells that display lineage boundaries. It is true that domains of gene expression in flies, for example of the iroquois complex, which includes mirror, can only be related to patterns with difficulty. Some recap of what is known plus the opinion of the authors on how they interpret papers on possible lineage domains in butterflies might also be useful as the reader, is no wiser about what the authors might mean at the end of it!

We thank the reviewer for this suggestion. However, our experiments have little to contribute to the topic of cell lineage compartmentalization. We have therefore opted to avoid speculating on this topic to prevent confusion and to keep the manuscript focused on our experimental results.

The references are sometimes inappropriate. The discovery of the AP compartments should not be referred to Guillen et al 1995, but to Morata and Lawrence 1975. Proofreading is required.

We thank the reviewer for suggesting this important reference. We have included it in our revision.

Reviewer #3 (Public Review):

Summary:

The manuscript by Chatterjee et al. examines the role of the mirror locus in patterning butterfly wings. The authors examine the pattern of mirror expression in the common buckeye butterfly, Junonia coenia, and then employ CRISPR mutagenesis to generate mosaic butterflies carrying clones of mirror mutant cells. They find that mirror is expressed in a well-defined posterior sector of final-instar wing discs from both hindwings and forewings and that CRISPR-injected larvae display a loss of adult wing structures presumably derived from the mirror expressing region of hindwing primordium (the case for forewings is a bit less clear since the mirror domain is narrower than in the hindwing, but there also do seem to be some anomalies in posterior regions of forewings in adults derived from CRISPR injected larvae). The authors conclude that the wings of these butterflies have at least three different fundamental wing compartments, the mirror domain, a posterior domain defined by engrailed expression, and an anterior domain expressing neither mirror nor engrailed. They speculate that this most posterior compartment has been reduced to a rudiment in Drosophila and thus has not been adequately recognized as such a primary regional specialization.

Critique:

This is a very straightforward study and the experimental results presented support the key claims that mirror is expressed in a restricted posterior section of the wing primordium and that mosaic wings from CRISPR-injected larvae display loss of adult wing structures presumably derived from cells expressing mirror (or at least nearby). The major issue I have with this paper is the strong interpretation of these findings that lead the authors to conclude that mirror is acting as a high-level gene akin to engrailed in defining a separate extreme posterior wing compartment. To place this claim in context, it is important in my view to consider what is known about engrailed, for which there is ample evidence to support the claim that this gene does play a very ancestral and conserved function in defining posterior compartments of all body segments (including the wing) across arthropods.

(1) Engrailed is expressed in a broad posterior domain with a sharp anterior border in all segments of virtually all arthropods examined (broad use of a very good panspecies anti-En antibody makes this case very strong).

(2) In Drosophila, marked clones of wing cells (generated during larval stages) strictly obey a straight anterior-posterior border indicating that cells in these two domains do not normally intermix, thus, supporting the claim that a clear A/P lineage compartment exists.

In my opinion, mirror does not seem to be in the same category of regulator as engrailed for the following reasons:

(1) There is no evidence that I am aware of, either from the current experiments, or others that the mirror expression domain corresponds to a clonal lineage compartment. It is also unclear from the data shown in this study whether engrailed is co-expressed with mirror in the posterior-most cells of J. coenia wing discs. If so, it does not seem justified to infer that mirror acts as an independent determinant of the region of the wing where it is expressed.

(2) Mirror is not only expressed in a posterior region of the wing in flies but also in the ventral region of the eye. In Drosophila, mirror mutants not only lack the alula (derived approximately from cells where mirror is expressed), but also lack tissue derived from the ventral region of the eye disc (although this ventral tissue loss phenotype may extend beyond the cells expressing mirror).

In summary, it seems most reasonable to me to think of mirror as a transcription factor that provides important development information for a diverse set of cells in which it can be expressed (posterior wing cells and ventral eye cells) but not that it acts as a high-level regulator as engrailed.

Recommendation:

While the data provided in this succinct study are solid and interesting, it is not clear to me that these findings support the major claim that mirror defines an extreme posterior compartment akin to that specified by engrailed. Minimally, the authors should address the points outlined above in their discussion section and greatly tone down their conclusion regarding mirror being a conserved selector-like gene dedicated to establishing posterior-most fates of the wing. They also should cite and discuss the original study in Drosophila describing the mirror expression pattern in the embryo and eye and the corresponding eye phenotype of mirror mutants: McNeill et al., Genes & Dev. 1997. 11: 1073-1082; doi:10.1101/gad.11.8.1073.

We thank the reviewer for their summary, critique, and recommendations. We agree with everything the reviewer says. Honestly, however, we were surprised by these comments because we took great care in the paper to never refer to mirror as a compartmentalization gene or claim it has a function in cell lineage compartmentalization like engrailed. As pointed out, we lack clonal analyses to test for compartmentalization. This is why we used the term “domain” instead of “compartment” in the title and throughout the manuscript. Nevertheless, we have recrafted the discussion in the manuscript, including completely removing the term “compartment”, to better avoid implications that mirror plays a role in cell lineage compartmentalization. 

We also thank the reviewer for recommending the paper about the role of mirror in eye development. For the sake of keeping the paper focused, however, we decided not to broach the topic of mirror functions outside the context of wing development.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

I have minor comments for improvement.

The abstract and introductions are terminologically problematic when they refer to the concept of compartment and compartment boundaries. Allegedly this confusion has previously propagated in several articles related to butterfly wing development, which keeps alienating this literature from being taken seriously by fly specialists, for example. So it is important to use the right terms. I will try to explain point by point here, but I would appreciate it if the authors could undertake a significant rewrite taking these comments into account. The authors use the terms compartment and compartment boundary. This has a very specific use in developmental genetics: mitotic clones never cross a boundary (or compartment). I think the authors can keep referring to the equivalent of the A-P boundary, which is situated somewhere between M1-M2 based on unpublished data from the Patel Lab, and is not very well defined (Engrailed expression moves a little bit during development in this area). Domain is a looser term and can be used more liberally to describe genetically defined regions.

- "Classical morphological work subdivides insect wings into several distinct domains along the antero-posterior (AP) axis, each of which can evolve relatively independently." Yes. This concept of domain and individuation seems important. You could make a proposed link to selector genes here.

- "There has been little molecular evidence, however, for AP subdivision beyond a single compartment boundary described from Drosophila melanogaster." Incorrect, and this conflates "domain" and "compartment".

Flies have wing AP domains too, that pattern their veins (see the cited Banerjee et al). 

- "Our results confirm that insect wings can have more than one posterior developmental domain, and support models of how selector genes may facilitate evolutionarily individuation of distinct AP domains in insect wings". Yes, and I like the second part of the sentence. Still, I would recommend simply deleting "confirm that insect wings can have more than one posterior developmental domain, and" because this is neglecting previous work on AP genetic regionalization in both flies (vein literature) and butterflies (e.g. McKenna and Nijhout, Banerjee et al).

- "Analyses of wing pattern diversity across butterflies, considering both natural variation and genetic mutants, suggest that wings can be subdivided into at least five AP domains, bounded by the M1, M3, Cu2, and 2A veins respectively, within each of which there are strong correlations in color pattern variation and wing morphology (Figure 1A)". Yes, and I would recommend emphasizing they correspond to welldefined gene expression domains as mentioned in Banerjee et al, or McKenna and Nijhout.

- "The anterior-most of these domains, bordered by the M1 vein, appears to correspond to an AP compartment boundary originally described by cell lineage tracing in Drosophila melanogaster, and later supported in butterfly wings by expression of the Engrailed transcription factor. Interestingly, however, D. melanogaster work has yet to reveal clear evidence for additional AP domain boundaries in the wing." Confusingly, because the first sentence is about compartments while the second is about AP domains. I also think the claim that Dmel has no other known AP domains is dubious because Spalt is highly regionalized in flies.

- "Previous authors have proposed the existence of such individuated domains, and speculated that they may be specified by selector genes.5,10 Our data provide experimental support for this model, and now motivate us to identify factors that specify other domain boundaries between the M1 and A2 veins." Yes, I completely agree with this way to emphasize the selector effect, and to link it to the concept of "individuated domain"

We cannot thank the reviewer enough for the time and thought they devoted to giving helpful suggestions to improve our manuscript. We have applied all of the above recommendations to the revision.

Fig. S1: the field needs to move away from Red/Green microscopy images, for accessibility reasons.

The easiest fix here would be to change the red channels to magenta.

Green/Magenta provides excellent contrast and accessibility in general in 2-channel images.

We thank the reviewer for this suggestion. We have improved the color accessibility of Fig. S1.

eLife Assessment

This important study examines the neuronal mechanisms underlying visual perception of integrated face and body cues. The innovative paradigm, which employs monkey avatars in combination with electrophysiological recordings from fMRI-defined brain areas, provides compelling evidence on face and body integration. These results should be of wide interest to system and cognitive neuroscientists, psychologists, and behavioural biologists working on visual and social cognition.

Reviewer #1 (Public review):

The study addresses how faces and bodies are integrated in two STS face areas revealed by fMRI in the primate brain. It is building upon recordings and analysis of the responses of large populations of neurons to three sets of images, that vary face and body positions. These sets allowed the author to thoroughly investigate invariance to position on the screen (MC HC), to pose (P1 P2), to rotation (0 45 90 135 180 225 270 315), to inversion, to possible and impossible postures (all vs straight), to presentation of head and body together or in isolation. By analyzing neuronal responses, they find that different neurons showed preferences for body orientation, or head orientation or for the interaction between the two. By using a linear support vector machine classifier, they show that the neuronal population can decode head-body angle presented across orientations, in the anterior aSTS patch (but not middle mSTS patch), except for mirror orientation. On the contrary, mSTS neurons show less invariance for head-body angle and more specialization for head or body orientation.

Strengths:

These results expand prior work on the role of Anterior STS fundus face area in face-body integration and its invariance to mirror symmetry, with a rigorous set of stimuli revealing the workings of these neuronal populations in processing individuals as a whole, in an important series of carefully designed conditions.

It also raises questions for future investigations comparing humans and monkeys expertise with upright and inverted configurations, in light of monkey-specific hanging upside-down behavior. Further, using two types of body postures (sitting, standing), they show a correlation in head-body angle between postures, suggesting that monkey orientation might be more meaningful to these neurons than precise posture.

Reviewer #2 (Public review):

Summary:

This paper investigates the neuronal encoding of the relationship between head and body orientations in the brain. Specifically, the authors focus on the angular relationship between the head and body by employing virtual avatars. Neuronal responses were recorded electrophysiologically from two fMRI-defined areas in the superior temporal sulcus and analyzed using decoding methods. They found that: (1) anterior STS neurons encode head-body angle configurations; (2) these neurons distinguish aligned and opposite head-body configurations effectively, whereas mirror-symmetric configurations are more difficult to differentiate; and (3) an upside-down inversion diminishes the encoding of head-body angles. These findings advance our understanding of how visual perception of individuals is mediated, providing a fundamental clue as to how the primate brain processes the relationship between head and body-a process that is crucial for social communication.

Strengths:

The paper is clearly written, and the experimental design is thoughtfully constructed and detailed. The use of electrophysiological recordings from fMRI-defined areas elucidated the mechanism of head-body angle encoding at the level of local neuronal populations. Multiple experiments, control conditions, and detailed analyses thoroughly examined various factors that could affect the decoding results. The decoding methods effectively and consistently revealed the encoding of head-body angles in the anterior STS neurons. Consequently, this study offers valuable insights into the neuronal mechanisms underlying our capacity to integrate head and body cues for social cognition-a topic that is likely to captivate readers in this field.

Weaknesses:

I did not identify any major weaknesses in this paper.

Reviewer #3 (Public review):

Summary:

Zafirova et al. investigated the interaction of head and body orientation in the macaque superior temporal sulcus (STS). Combining fMRI and electrophysiology, they recorded responses of visual neurons to a monkey avatar with varying head and body orientations. They found that STS neurons integrate head and body information in a nonlinear way, showing selectivity for specific combinations of head-body orientations. Head-body configuration angles can be reliably decoded, particularly for neurons in the anterior STS, suggesting a transformation of face/body orientation signals from the middle to the anterior STS. Furthermore, body inversion resulted in reduced decoding of head-body configuration angles. Compared to previous work that examined face or body alone, this study demonstrates how head and body information are integrated to compute a socially meaningful signal.

Strengths:

This work presents an elegant design of visual stimuli, with a monkey avatar of varying head and body orientations, making the analysis and interpretation straightforward. Together with several control experiments, the authors systematically investigated different aspects of head-body integration in the macaque STS. The results and analyses of the paper are convincing.

Weakness:

While this work has characterized the neural integration of head and body information in detail, it's unclear how the neural representation relates to the animal's perception. Behavioural experiments using the same set of stimuli could help address this question, but I agree that these additional experiments may be beyond the scope of the current paper.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

The study addresses how faces and bodies are integrated in two STS face areas revealed by fMRI in the primate brain. It builds upon recordings and analysis of the responses of large populations of neurons to three sets of images, that vary face and body positions. These sets allowed the authors to thoroughly investigate invariance to position on the screen (MC HC), to pose (P1 P2), to rotation (0 45 90 135 180 225 270 315), to inversion, to possible and impossible postures (all vs straight), to the presentation of head and body together or in isolation. By analyzing neuronal responses, they found that different neurons showed preferences for body orientation, head orientation, or the interaction between the two. By using a linear support vector machine classifier, they show that the neuronal population can decode head-body angle presented across orientations, in the anterior aSTS patch (but not middle mSTS patch), except for mirror orientation.

Strengths:

These results extend prior work on the role of Anterior STS fundus face area in face-body integration and its invariance to mirror symmetry, with a rigorous set of stimuli revealing the workings of these neuronal populations in processing individuals as a whole, in an important series of carefully designed conditions.

Minor issues and questions that could be addressed by the authors:

(1) Methods. While monkeys certainly infer/recognize that individual pictures refer to the same pose with varying orientations based on prior studies (Wang et al.), I am wondering whether in this study monkeys saw a full rotation of each of the monkey poses as a video before seeing the individual pictures of the different orientations, during recordings.

The monkeys had not been exposed to videos of a rotating monkey pose before the recordings. However, they were reared and housed with other monkeys, providing them with ample experience of monkey poses from different viewpoints.

(2) Experiment 1. The authors mention that neurons are preselected as face-selective, body-selective, or both-selective. Do the Monkey Sum Index and ANOVA main effects change per Neuron type?

We have performed a new analysis to assess whether the Monkey Sum Index is related to the response strength for the face versus the body as measured in the Selectivity Test of Experiment 1. To do this we selected face- and body-category selective neurons, as well as neurons responding selectively to both faces and bodies. First, we selected those neurons that responded significantly to either faces, bodies, or the two control object categories, using a split-plot ANOVA for these 40 stimuli. From those neurons, we selected face-selective ones having at least a twofold larger mean net response to faces compared to bodies (faces > 2 * bodies) and the control objects for faces (faces  > 2* objects). Similarly, a body-selective neuron was defined by a twofold larger mean net response to bodies compared to faces and the control objects for bodies. A body-and-face selective neuron was defined as having a twofold larger net response to the faces compared to their control objects, and to bodies compared to their control objects, with the ratio between mean response to bodies and faces being less than twofold. Then, we compared the distribution of the Monkey Sum Index (MSI) for each region (aSTS; mSTS), pose (P1, P2), and centering (head- (HC) or monkey-centered (MC)) condition. Too few body-and-face selective neurons were present in each combination of region, pose, and centering (a maximum of 7) to allow a comparison of their MSI distribution with the other neuron types. The Figure below shows the distribution of the MSI for the different orientation-neuron combinations for the body- and face-selective neurons (same format as in Figure 3a, main text). The number of body-selective neurons, according to the employed criteria, varied from 21 to 29, whereas the number of face-selective neurons ranged from 14 to 24 (pooled across monkeys). The data of the two subjects are shown in a different color and the number of cases for each subject is indicated (n1: number of cases for M1; n2: number of cases for M2). The arrows indicate the medians for the data pooled across the monkey subjects. For the MC condition, the MSI tended to be more negative (i.e. relatively less response to the monkey compared to the sum of the body and face responses) for the face compared to the body cells, but this was significant only for mSTS and P1 (p = 0.043; Wilcoxon rank sum test; tested after averaging the indices per neuron to avoid dependence of indices within a neuron). No consistent, nor significant tendencies were observed for the HC stimuli. This absence of a consistent relationship between MSI and face- versus body-selectivity is in line with the absence of a correlation between the MSI and face- versus body-selectivity using natural images of monkeys in a previous study (Zafirova Y, Bognár A, Vogels R. Configuration-sensitive face-body interactions in primate visual cortex. Prog Neurobiol. 2024 Jan;232:102545).

We did not perform a similar analysis for the main effects of the two-way ANOVA because the very large majority of neurons showed a significant effect of body orientation and thus no meaningful difference between the two neuron types can be expected.

Author response image 1.

(3) I might have missed this information, but the correlation between P1 and P2 seems to not be tested although they carry similar behavioral relevance in terms of where attention is allocated and where the body is facing for each given head-body orientation.

Indeed, we did not compute this correlation between the responses to the sitting (P1) and standing (P2) pose avatar images. However, as pointed out by the reviewer, one might expect such correlations because of the same head orientations and body-facing directions. Thus, we computed the correlation between the 64 head-body orientation conditions of P1 and P2 for those neurons that were tested with both poses and showed a response for both poses (Split-plot ANOVA). This was performed for the Head-Centered and Monkey-Centered tests of Experiment 1 for each monkey and region. Note that not all neurons were tested with both poses (because of failure to maintain isolation of the single unit in both tests or the monkey stopped working) and not all neurons that were recorded in both tests showed a significant response for both poses, which is not unexpected since these neurons can be pose selective. The distribution of the Pearson correlation coefficients of the neurons with a significant response in both tests is shown in Figure S1. The median correlation coefficient was significantly larger than zero for each region, monkey, and centering condition (outcome of Wilcoxon tests, testing whether the median was different from zero (p1 = p-value for M1; p2: p-value for M2) in Figure), indicating that the effect of head and/or body orientation generalizes across pose. We have noted this now in the Results (page 12) and added the Figure (New Figure S1) in the Suppl. Material.

(4) Is the invariance for position HC-MC larger in aSTS neurons compared to mSTS neurons, as could be expected from their larger receptive fields?

Yes, the position tolerance of the interaction of body and head orientation was significantly larger for aSTS compared to mSTS neurons, as we described on pages 11 and 12 of the Results. This is in line with larger receptive fields in aSTS than in mSTS. However, we did not plot receptive fields in the present study.

(5) L492 "The body-inversion effect likely results from greater exposure to upright than inverted bodies during development". Monkeys display more hanging upside-down behavior than humans, however, does the head appear more tilted in these natural configurations?

Indeed, infant monkeys do spend some time hanging upside down from their mother's belly. While we lack quantitative data on this behavior, casual observations suggest that even young monkeys spend more time upright. The tilt of the head while hanging upside down can vary, just as it does in standing or sitting monkeys (as when they search for food or orient to other individuals). To our knowledge, no quantitative data exist on the frequency of head tilts in upright versus upside-down monkeys. Therefore, we refrain from further speculation on this interesting point, which warrants more attention.

(6) Methods in Experiment 1. SVM. How many neurons are sufficient to decode the orientation?

The number of neurons that are needed to decode the head-body orientation angle depends on which neurons are included, as we show in a novel analysis of the data of Experiment 1. We employed a neuron-dropping analysis, similar to Chiang et al. (Chiang FK, Wallis JD, Rich EL. Cognitive strategies shift information from single neurons to populations in prefrontal cortex. Neuron. 2022 Feb 16;110(4):709-721) to assess the positive (or negative) contribution of each neuron to the decoding performance. We performed cross-validated linear SVM decoding N times, each time leaving out a different neuron (using N-1 neurons; 2000 resamplings of pseudo-population vectors). We then ranked decoding accuracies from highest to lowest, identifying the ‘worst’ (rank 1) to ‘best’ (rank N) neurons. Next, we conducted N decodings, incrementally increasing the number of included neurons from 1 to N, starting with the worst-ranked neuron (rank 1) and sequentially adding the next (rank 2, rank 3, etc.). This analysis focused on zero versus straight angle decoding in the aSTS, as it yielded the highest accuracy. We applied it when training on MC and testing on HC for each pose. Plotting accuracy as a function of the number of included neurons suggested that less than half contributed positively to decoding. We show also the ten “best” neurons for each centering condition and pose. These have a variety of tuning patterns for head and body orientation suggesting that the decoding of head-body orientation angle depends on a population code. Notably, the best-ranked (rank N) neuron alone achieved above-chance accuracy. We have added this interesting and novel result to the Results (page 16) and Suppl. Material (new Figure S3).

(7) Figure 3D 3E. Could the authors please indicate for each of these neurons whether they show a main effect of face, body, or interaction, as well as their median corrected correlation to get a flavor of these numbers for these examples?

We have indicated these now in Figure 3.

(8) Methods and Figure 1A. It could be informative to precise whether the recordings are carried in the lateral part of the STS or in the fundus of the STS both for aSTS and mSTS for comparison to other studies that are using these distinctions (AF, AL, MF, ML).

In experiment 1, the recording locations were not as medial as the fundus. For experiments 2 and 3, the ventral part of the fundus was included, as described in the Methods. We have added this to the Methods now (page 31).

Wang, G., Obama, S., Yamashita, W. et al. Prior experience of rotation is not required for recognizing objects seen from different angles. Nat Neurosci 8, 1768-1775 (2005). https://doi-org.insb.bib.cnrs.fr/10.1038/nn1600

Reviewer #2 (Public review):

Summary:

This paper investigates the neuronal encoding of the relationship between head and body orientations in the brain. Specifically, the authors focus on the angular relationship between the head and body by employing virtual avatars. Neuronal responses were recorded electrophysiologically from two fMRI-defined areas in the superior temporal sulcus and analyzed using decoding methods. They found that: (1) anterior STS neurons encode head-body angle configurations; (2) these neurons distinguish aligned and opposite head-body configurations effectively, whereas mirror-symmetric configurations are more difficult to differentiate; and (3) an upside-down inversion diminishes the encoding of head-body angles. These findings advance our understanding of how visual perception of individuals is mediated, providing a fundamental clue as to how the primate brain processes the relationship between head and body - a process that is crucial for social communication.

Strengths:

The paper is clearly written, and the experimental design is thoughtfully constructed and detailed. The use of electrophysiological recordings from fMRI-defined areas elucidated the mechanism of head-body angle encoding at the level of local neuronal populations. Multiple experiments, control conditions, and detailed analyses thoroughly examined various factors that could affect the decoding results. The decoding methods effectively and consistently revealed the encoding of head-body angles in the anterior STS neurons. Consequently, this study offers valuable insights into the neuronal mechanisms underlying our capacity to integrate head and body cues for social cognition-a topic that is likely to captivate readers in this field.

Weaknesses:

I did not identify any major weaknesses in this paper; I only have a few minor comments and suggestions to enhance clarity and further strengthen the manuscript, as detailed in the Private Recommendations section.

Reviewer #3 (Public review):

Summary:

Zafirova et al. investigated the interaction of head and body orientation in the macaque superior temporal sulcus (STS). Combining fMRI and electrophysiology, they recorded responses of visual neurons to a monkey avatar with varying head and body orientations. They found that STS neurons integrate head and body information in a nonlinear way, showing selectivity for specific combinations of head-body orientations. Head-body configuration angles can be reliably decoded, particularly for neurons in the anterior STS. Furthermore, body inversion resulted in reduced decoding of head-body configuration angles. Compared to previous work that examined face or body alone, this study demonstrates how head and body information are integrated to compute a socially meaningful signal.

Strengths:

This work presents an elegant design of visual stimuli, with a monkey avatar of varying head and body orientations, making the analysis and interpretation straightforward. Together with several control experiments, the authors systematically investigated different aspects of head-body integration in the macaque STS. The results and analyses of the paper are mostly convincing.

Weaknesses:

(1) Using ANOVA, the authors demonstrate the existence of nonlinear interactions between head and body orientations. While this is a conventional way of identifying nonlinear interactions, it does not specify the exact type of the interaction. Although the computation of the head-body configuration angle requires some nonlinearity, it's unclear whether these interactions actually contribute. Figure 3 shows some example neurons, but a more detailed analysis is needed to reveal the diversity of the interactions. One suggestion would be to examine the relationship between the presence of an interaction and the neural encoding of the configuration angle.

This is an excellent suggestion. To do this, one needs to identify the neurons that contribute to the decoding of head-body orientation angles. For that, we employed a neuron-dropping analysis, similar to Chiang et al. (Chiang FK, Wallis JD, Rich EL. Cognitive strategies shift information from single neurons to populations in prefrontal cortex. Neuron. 2022 Feb 16;110(4):709-721.) to assess the positive (or negative) contribution of each neuron to the decoding performance. We performed cross-validated linear SVM decoding N times, each time leaving out a different neuron (using N-1 neurons; 2000 resamplings of pseudo-population vectors). We then ranked decoding accuracies from highest to lowest, identifying the ‘worst’ (rank 1) to ‘best’ (rank N) neurons. Next, we conducted N decodings, incrementally increasing the number of included neurons from 1 to N, starting with the worst-ranked neuron (rank 1) and sequentially adding the next (rank 2, rank 3, etc.). This analysis focused on zero versus straight angle decoding in the aSTS, as it yielded the highest accuracy. We applied it when training on MC and testing on HC for each pose. Plotting accuracy as a function of the number of included neurons suggested that less than half contributed positively to decoding (see Figure S3). We examined the tuning for head and body orientation of the 10 “best” neurons (Figure S3). For half or more of those the two-way ANOVA showed a significant interaction. These are indicated by the red color in the Figure. They showed a variety of tuning patterns for head and body orientation, suggesting that the decoding of the head-body orientation angle results from a combination of neurons with different tuning profiles. Based on a suggestion from reviewer 2, we performed for each neuron of experiment 1 a one-way ANOVA with as factor head-body orientation angle. To do that, we combined all 64 trials that had the same head-body orientation angle. The percentage of neurons (required to be responsive in the tested condition) for which this one-way ANOVA was significant was low but larger than the expected 5% (Type 1 error), with a median of 16.5% (range: 3 to 23%) in aSTS and 8% for mSTS (range: 0-19%). However, a higher percentage of the 10 best neurons for each pose (indicated by the star) showed a significant one-way ANOVA for angle (for P1, MC: 50% (95% confidence interval (CI): 19% – 81%); P1, HC: 70% (CI: 35% - 93%); P2, MC: 70% (CI: 35% – 93%); P2: HC: 50% (CI: 19%-81%)). These percentages were significantly higher than expected for a random sample from the population of neurons for each pose-centering combination (expected percentages listed in the same order as above: 16%, 13%, 16%, and 10%; all outside CI). Thus, for at least half of the “best” neurons, the response differed significantly among the head-orientation angles at the single neuron level. Nonetheless, the tuning profiles were diverse, suggesting a populationl code for head-body orientation angle. We have added this interesting and novel result to the Results (page 16) and Suppl. Material (Figure S3).

(2) Figure 4 of the paper shows a better decoding of the configuration angle in the anterior STS than in the middle STS. This is an interesting result, suggesting a transformation in the neural representation between these two areas. However, some control analyses are needed to further elucidate the nature of this transformation. For example, what about the decoding of head and body orientations - dose absolute orientation information decrease along the hierarchy, accompanying the increase in configuration information?

We have performed now two additional analyses, one in which we decoded the orientation of the head and another one in which we decoded the orientation of the body. We employed the responses to the avatar of experiment 1, using the same sample of neurons of which we decoded the head-body orientation angle. To decode the head orientation, the trials with identical head orientation, irrespective of their body orientation, were given the same label. For this, we employed only responses in the head-centered condition. To decode the body orientation, the trials with identical body orientation, irrespective of their head orientation, had the same label, and we employed only responses in the body-centered condition. The decoding was performed separately for each pose (P1 and P2) and region. We decoded either the responses of 20 neurons (10 randomly sampled from each monkey for each of the 1000 resamplings), 40 neurons (20 randomly sampled per monkey), or 60 neurons (30 neurons per monkey) since the sample of 60 neurons yielded close to ceiling performance for the body orientation decoding. For each pose, the body orientation decoding was worse for aSTS than for mSTS, although this difference reached significance only for P1 and for the 40 neurons sample of P2 (p < 0.025; two-tailed test; same procedure as employed for testing the significance of the decoding of whole-body orientation for upright versus inverted avatars (Experiment 3))). Face orientation decoding was significantly worse for aSTS compared to mSTS. These results are in line with the previously reported decreased decoding of face orientation in the anterior compared to mid-STS face patches (Meyers EM, Borzello M, Freiwald WA, Tsao D. Intelligent information loss: the coding of facial identity, head pose, and non-face information in the macaque face patch system. J Neurosci. 2015 May 6;35(18):7069-81), and decreased decoding of body orientation in anterior compared to mid-STS body patches (Kumar S, Popivanov ID, Vogels R. Transformation of Visual Representations Across Ventral Stream Body-selective Patches. Cereb Cortex. 2019 Jan 1;29(1):215-229). As mentioned by the reviewer, this contrasts with the decoding of the head-body orientation angle, which increases when moving more anteriorly. We mention this finding now in the Discussion (page 27) and present the new Figure S10 in the Suppl. Material.    

(3) While this work has characterized the neural integration of head and body information in detail, it's unclear how the neural representation relates to the animal's perception. Behavioural experiments using the same set of stimuli could help address this question, but I agree that these additional experiments may be beyond the scope of the current paper. I think the authors should at least discuss the potential outcomes of such experiments, which can be tested in future studies.

Unfortunately, we do not have behavioral data. One prediction would be that the discrimination of head-body orientation angle, irrespective of the viewpoint of the avatar, would be more accurate for zero versus straight angles compared to the right versus left angles. We have added this to the Discussion (page 28).

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) P22 L373. It should read Figure S5C instead of S4C.

Thanks; corrected.

(2) Figure 7B. All inverted decoding accuracies, although significantly lower than upright decoding accuracies, appear significantly above baseline. Should the title be amended accordingly?

Thanks for pointing this out. To avoid future misunderstanding we have changed the title to:

“Integration of head and body orientations in the macaque superior temporal sulcus is stronger for upright bodies”

(3) Discussion L432-33. "with some neurons being tuned to a particular orientation of both the head and the body". Wouldn't that be visible as a diagonal profile on the normalized net responses in Fig 3D? Or can the Anova evidence such a tuning?

We meant to say that some neurons were tuned to a particular combination of head and body orientation, like the third aSTS example neuron shown in Figure 3D. We have corrected the sentence.

Reviewer #2 (Recommendations for the authors):

Major comment:

This paper effectively demonstrates that the angular relationship between the head and body can be decoded from population responses in the anterior STS. In other words, these neurons encode information about the head-body angle. However, how exactly do these neurons encode this information? Given that the study employed electrophysiological recordings from a local population of neurons, it might be possible to provide additional data on the response patterns of individual neurons to shed light on the underlying encoding mechanisms.

Although the paper already presents example response patterns (Figures 3D, E) and shows that STS neurons encode interactions between head and body orientations (Figure 3B), it remains unclear whether the angle difference between the head and body has a systematic effect on neuronal responses. For instance, a description of whether some neurons preferentially encode specific head-body angle differences (e.g., a "45-degree angle neuron"), or additional population analyses such as a one-way ANOVA with angle difference as the main effect (or two-way ANOVA with angle difference as one of the main effect), would be very informative. Such data could offer valuable insights into how individual neurons contribute to the encoding of head-body angle differences-a detail that may also be reflected in the decoding results. Alternatively, it is possible that the encoding of head-body angle is inherently complex and only discernible via decoding methods applied to population activity. Either scenario would provide interesting and useful information to the field.

We have performed two additional analyses which are relevant to this comment. First, we attempted to relate the tuning for body and head orientation with the decoding of the head-body orientation angle. To do this, one needs to identify the neurons that contribute to the decoding of head-body orientation angles. For that, we employed a neuron-dropping analysis, similar to Chiang et al. (Chiang FK, Wallis JD, Rich EL. Cognitive strategies shift information from single neurons to populations in prefrontal cortex. Neuron. 2022 Feb 16;110(4):709-721.) to assess the positive (or negative) contribution of each neuron to the decoding performance. We performed cross-validated linear SVM decoding N times, each time leaving out a different neuron (using N-1 neurons; 2000 resamplings of pseudo-population vectors). We then ranked decoding accuracies from highest to lowest, identifying the ‘worst’ (rank 1) to ‘best’ (rank N) neurons. Next, we conducted N decodings, incrementally increasing the number of included neurons from 1 to N, starting with the worst-ranked neuron (rank 1) and sequentially adding the next (rank 2, rank 3, etc.). This analysis focused on zero versus straight angle decoding in the aSTS, as it yielded the highest accuracy. We applied it when training on MC and testing on HC for each pose. Plotting accuracy as a function of the number of included neurons suggested that less than half contributed positively to decoding (see Figure S3). We examined the tuning for head and body orientation of the 10 “best” neurons (Figure S3). For half or more of those the two-way ANOVA showed a significant interaction. These are indicated by the red color in the Figure. They showed a variety of tuning patterns for head and body orientation, suggesting that the decoding of the head-body orientation angle results from a combination of neurons with different tuning profiles.

Second, we have followed the suggestion of the reviewer to perform for each neuron of experiment 1 a one-way ANOVA with as factor head-body orientation angle. To do that, we combined all 64 trials that had the same head-body orientation angle. The percentage of neurons (required to be responsive in the tested condition) for which this one-way ANOVA was significant is shown in the Tables below for each region, separately for each pose (P1, P2), centering condition (MC = monkey-centered; HC = head-centered) and monkey subject (M1, M2). The percentages were low but larger than the expected 5% (Type 1 error), with a median of 16.5% (range: 3 to 23%) in aSTS and 8% for mSTS (range: 0-19%).

Author response table 1.

Interestingly, a higher percentage of the 10 best neurons for each pose (indicated by the star in the Figure above) showed a significant one-way ANOVA for angle (for P1, MC: 50% (95% confidence interval (CI): 19% – 81%); P1, HC: 70% (CI: 35% - 93%); P2, MC: 70% (CI: 35% – 93%); P2: HC: 50% (CI: 19%-81%)). These percentages were significantly higher than expected for a random sample from the population of neurons for each pose-centering combination (expected percentages listed in the same order as above: 16%, 13%, 16%, and 10%; all outside CI). Thus, for at least half of the “best” neurons, the response differed significantly among the head-orientation angles at the single neuron level. Nonetheless, the tuning profiles were quite diverse, suggesting population coding of head-body orientation angle. We have added this interesting and novel result to the Results (page 16) and Suppl. Material (Figure S3).    

Minor comments:

(1) Figure 4A, Fourth Row Example (Zero Angle vs. Straight Angle, Bottom of the P2 Examples): The order of the example stimuli might be incorrect- the 0{degree sign} head with 180{degree sign} body stimulus (leftmost) might be swapped with the 180{degree sign} head with 0{degree sign} body stimulus (5th from the left). While this ordering may be acceptable, please double-check whether it reflects the authors' intended arrangement.

We have changed the order of the two stimuli in Figure 4A, following the suggestion of the reviewer.

(2) Page 12, Lines 192-194: The text states, "Interestingly, some neurons (e.g. Figure 3D) were tuned to a particular combination of a head and body irrespective of centering." However, Figure 3D displays data for a total of 10 neurons. Could you please specify which of these neurons are being referred to in this context?

The wording was not optimal. We meant to say that some neurons were tuned to a particular combination of head and body orientation, like the third aSTS example neuron of Figure 3D. We have rephrased the sentence and clarified which example neuron we referred to.

(3) Page 28, Lines 470-471: The text states, "We observed no difference in response strength between anatomically possible and impossible configurations." Please clarify which data were compared for response strength, as I could not locate the corresponding analyses.

The anatomically possible and impossible configurations differ in the head-body orientation angle. However, as we reported before in the Results, there was no effect of head-body orientation angle on mean response strength across poses (Friedman ANOVA; all p-values for both poses and centerings > 0.1). We have clarified this now in the Discussion (page 28).

(4) Pages 40-43, Decoding Analyses: In experiments 2 and 3, were the decoding analyses performed on simultaneously recorded neurons? If so, such analyses might leverage trial-by-trial correlations and thus avoid confounds from trial-to-trial variability. In contrast, experiment 1, which used single-shank electrodes, would lack this temporal information. Please clarify how trial numbers were assigned to neurons in each experiment and how this assignment may have influenced the decoding performance.

For the decoding analyses of experiments 2 and 3, we combined data from different daily penetrations, with only units from the same penetration being recorded simultaneously. In the decoding analyses of each experiment, the trials were assigned randomly to the pseudo-population vectors, shuffling on each resampling the trial order per neuron. This shuffling abolishes noise correlations in the analysis of each experiment.

(5) Page 41, Lines 792-802: The authors state that "To assess the significance of the differences in classification scores between pairs of angles ... we computed the difference in classification score between the two pairs for each resampling and the percentile of 0 difference corresponded to the p-value." In a two-sided test under the null hypothesis of no difference between the distributions, the conventional approach would be to compute the p-value as the proportion of resampled differences that are as extreme or more extreme than the observed difference. Since a zero difference might be relatively rare, relying solely on its percentile could potentially misrepresent the tail probabilities relevant to a two-sided test. Could you clarify how their method addresses this issue?

This test is based on the computation of the distribution of the difference between classification accuracies across resamplings. This is similar to the computation of the confidence interval of a  difference. Thus, we assess whether the theoretical zero value (= no difference; = null hypothesis) is outside the 2.5 and 97.5 percentile interval of the computed distribution of the empirically observed differences. We clarified now in the Methods (page 41) that for a two-tailed test the computed p-value (the percentile of the zero value) should be smaller than 0.025.

(6) Page 43, Lines 829-834: The manuscript explains: "The mean of 10 classification accuracies (i.e., of 10 resamplings) was employed to obtain a distribution (n=100) of the differences in classification accuracy ... The reported standard deviations of the classification accuracies are computed using also the means of 10 resamplings." I am unfamiliar with this type of analysis and am unclear about the rationale for calculating distributions and standard deviations based on the means of 10 resamplings rather than using the original distribution of classification accuracies. This resampling procedure appears to yield a narrower distribution and smaller standard deviations than the original data. Could you please justify this approach?

The logic of the analysis is to reduce the noise in the data, by averaging across 10 randomly selected resamplings, but still keeping a sufficient number of data (100 values) for a test.

Reviewer #3 (Recommendations for the authors):

(1) Some sentences are too long and difficult to parse. For example, in line 177: "the correlations between the responses to the 64 head-body orientation conditions of the two centerings for the neuron and pose combinations showing significant head-body interactions for the two centerings were similar to those observed for the whole population."

We have modified this sentence: For neuron and pose combinations with significant head-body interactions in both centerings, the correlations between responses to the 64 head-body orientation conditions were similar to those observed in the whole population.

(2) The authors argue in line 485: "in our study, a search bias cannot explain the body-inversion effect since we selected responsive units using both upright and inverted images." However, the body-selective patches were localized using upright images, correct?

The monkey-selective patches were localized using upright images indeed. However, we recorded in experiment 3 (and 2) also outside the localized patches (as we noted before in the Methods:  “In experiments 2 and 3 we recorded from a wider region, which overlapped with the two monkey patches and the recording locations of experiment 1”). Furthermore, the preference for upright monkey images is not an all-or-nothing phenomenon: most units still responded to inverted monkeys. Also, we believe it is likely that the mean responses to the inverted bodies in the monkey patches, defined by upright bodies versus objects, would be larger than those to objects and we would be surprised to learn that there is a patch selective for inverted bodies that we would have missed with our localizer.

(3) Typo: line 447, "this independent"->"is independent"?

Corrected.

eLife Assessment

Some delayed rectifier currents in neurons are formed by the combination of Kv2 and silent subunits, KvS. However, we lack the tools to identify these heteromeric channels in vivo. In this important study by the Sack group, the authors identify a pharmacological tool that can reveal the presence of KvS subunits as components of the delayed rectifier potassium currents in selected neurons. The experimental evidence presented in the manuscript is compelling and represents a significant advance that should be of interest to a wide community of neuroscientists and channel physiologists.

Reviewer #1 (Public review):

Summary:

Kv2 subfamily potassium channels contribute to delayed rectifier currents in virtually all mammalian neurons and are encoded by two distinct types of subunits: Kv2 alpha subunits that have the capacity to form homomeric channels (Kv2.1 and Kv2.2), and KvS or silent subunits (Kv5,6,8.9) that can assemble with Kv2.1 or Kv2.2 to form heteromeric channels with novel biophysical properties. Many neurons express both types of subunits and therefore have the capacity to make both homomeric Kv2 channels and heteromeric Kv2/KvS channels. Determining the contributions of each of these channel types to native potassium currents has been very difficult because the differences in biophysical properties are modest and there are no Kv2/KvS-specific pharmacological tools. The authors set out to design a strategy to separate Kv2 and Kv2/KvS currents in native neurons based on their observation that Kv2/KvS channels have little sensitivity to the Kv2 pore blocker RY785 but are blocked by the Kv2 VSD blocker GxTx. They clearly demonstrate that Kv2/KvS currents can be differentiated from Kv2 currents in native neurons using a two-step strategy to first selectively block Kv2 with RY785, and then block both with GxTx. The manuscript is beautifully written; takes a very complex problem and strategy and breaks it down so both channel experts and the broad neuroscience community can understand it.

Strengths:

The compounds the authors use are highly selective and unlikely to have significant confounding cross-reactivity to other channel types. The authors provide strong evidence that all Kv2/KvS channels are resistant to RY785. This is a strength of the strategy - it can likely identify Kv2/KvS channels containing any of the 10 mammalian KvS subunits and thus be used as a general reagent on all types of neurons. The limitation then of course is that it can't differentiate the subtypes, but at this stage, the field really just needs to know how much Kv2/KvS channels contribute to native currents and this strategy provides a sound way to do so.

Weaknesses:

The authors are very clear about the limitations of their strategy, the most important of which is that they can't differentiate different subunit combinations of Kv2/KvS heteromers. This study is meant to be a start to understanding the roles of Kv2/KvS channels in vivo. As such, this is a minor weakness, far outweighed by the potential of the strategy to move the field through a roadblock that has existed since its inception.

The study accomplishes exactly what it set out to do: provide a means to determine the relative contributions of homomeric Kv2 and heteromeric Kv2/KvS channels to native delayed rectifier K+ currents in neurons. It also does a fabulous job laying out the case for why this is important to do.

Comments on revisions:

I liked this manuscript the first time and thought it was a great attempt to address a difficult problem, made more difficult by confusing background literature and conventions. The authors have kept all the strong points I liked from the first round and made it even stronger with their thoughtful and substantive responses to reviews. My first review was strongly supportive, and my initial short assessment/public review was written with the assumption that they would be public and the paper would be published essentially in its original form. All those points still apply so I am going to leave the initial reviews as is. The paper is a pleasure to read and a nice contribution to the field.

Reviewer #2 (Public review):

The authors used combined blockers/modulators to dissect the potassium currents mediated by inter-subunit heteromeric Kv channels. The method is robust given that the researchers know their limitations. Nevertheless, the authors elegantly tested their hypotheses, making this manuscript friendly to read despite the depth of all aspects they dealt with.

The quality of the data presented will positively impact the science involved in the study heteromeric channels, with clear developments in the field. Finally, the approach presented may unlock new studies related to these channels.

Comments on revisions:

The authors clarified all my points and beyond, specifically by adding some computational work that will also contribute to the subfield of heteromeric Kv channels.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public Review):

Summary:

Kv2 subfamily potassium channels contribute to delayed rectifier currents in virtually all mammalian neurons and are encoded by two distinct types of subunits: Kv2 alpha subunits that have the capacity to form homomeric channels (Kv2.1 and Kv2.2), and KvS or silent subunits (Kv5,6,8.9) that can assemble with Kv2.1 or Kv2.2 to form heteromeric channels with novel biophysical properties. Many neurons express both types of subunits and therefore have the capacity to make both homomeric Kv2 channels and heteromeric Kv2/KvS channels. Determining the contributions of each of these channel types to native potassium currents has been very difficult because the differences in biophysical properties are modest and there are no Kv2/KvS-specific pharmacological tools. The authors set out to design a strategy to separate Kv2 and Kv2/KvS currents in native neurons based on their observation that Kv2/KvS channels have little sensitivity to the Kv2 pore blocker RY785 but are blocked by the Kv2 VSD blocker GxTx. They clearly demonstrate that Kv2/KvS currents can be differentiated from Kv2 currents in native neurons using a two-step strategy to first selectively block Kv2 with RY785, and then block both with GxTx. The manuscript is beautifully written; takes a very complex problem and strategy and breaks it down so both channel experts and the broad neuroscience community can understand it.

Strengths:

The compounds the authors use are highly selective and unlikely to have significant confounding cross-reactivity to other channel types. The authors provide strong evidence that all Kv2/KvS channels are resistant to RY785. This is a strength of the strategy - it can likely identify Kv2/KvS channels containing any of the 10 mammalian KvS subunits and thus be used as a general reagent on all types of neurons. The limitation then of course is that it can't differentiate the subtypes, but at this stage, the field really just needs to know how much Kv2/KvS channels contribute to native currents and this strategy provides a sound way to do so.

Weaknesses:

The authors are very clear about the limitations of their strategy, the most important of which is that they can't differentiate different subunit combinations of Kv2/KvS heteromers. This study is meant to be a start to understanding the roles of Kv2/KvS channels in vivo. As such, this is a minor weakness, far outweighed by the potential of the strategy to move the field through a roadblock that has existed since its inception.

The study accomplishes exactly what it set out to do: provide a means to determine the relative contributions of homomeric Kv2 and heteromeric Kv2/KvS channels to native delayed rectifier K+ currents in neurons. It also does a fabulous job laying out the case for why this is important to do.

Reviewer #2 (Public Review):

Summary:

Silent Kv subunits and the channels containing these Kv subunits (Kv2/KvS heteromers) are in the process of discovery. It is believed that these channels fine-tune the voltage-activated K+ currents that repolarize the membrane potential during action potentials, with a direct effect on cell excitability, mostly by determining action potentials firing frequency.

Strengths:

What makes silent Kv subunits even more important is that, by being expressed in specific tissues and cell types, different silent Kv subunits may have the ability to fine-tune the delayed rectifying voltage-activated K+ currents that are one of the currents that crucially determine cell excitability in these cells. The present manuscript introduces a pharmacological method to dissect the voltage-activated K+ currents mediated by Kv2/KvS heteromers as a means of starting to unveil their importance, together with Kv2-only channels, to the cells where they are expressed.

Weaknesses:

While the method is effective in quantifying these currents in any isolated cell under an electric voltage clamp, it is ineffective as a modulating maneuver to perhaps address these currents in an in vivo experimental setting. This is an important point but is not a claim made by the authors.

We agree. We have now stated in the introduction that this study does not address the roles of Kv2/KvS currents in an in vivo setting.

Manuscript revisions:

While this study does not address the impact of GxTX or RY785 on action potentials or in vivo, the distinct pharmacology of Kv2/KvS heteromers presented here suggests that KvS conductances could be targeted to selectively modulate discrete subsets of cell types.  

There are other caveats with the methods and data:

(i) The need for a 'cocktail' of blockers to supposedly isolate Kv2 homomers and Kv2/KvS heteromers' currents from others may introduce errors in the quantification Kv2/KvS heteromers-mediated K+ currents and that is due to possible blockers off targets.

We now point out that is possible that off target effects of blockers may introduce errors, include references that identify the selectivity of the blockers used in the cocktail, and specifically note that 4-aminopyridine in the cocktail is expected to block 2% of Kv2 homomers yet have a lesser impact Kv2/KvS heteromers. Additionally, to test whether the KvS isolation strategy requires the cocktail in neurons, we performed new experiments on a different subclass of nociceptors without the blocker cocktail and identified a substantial KvS-like component (new Fig 7 Supplement 3).

Manuscript revisions:

“After whole-cell voltage clamp was established, non-Kv2/KvS conductances were suppressed by changing to an external solution containing a cocktail of inhibitors: 100 nM alpha-dendrotoxin (Alomone) to block Kv1 (Harvey and Robertson, 2004), 3 μM AmmTX3 (Alomone) to block Kv4 (Maffie et al., 2013; Pathak et al., 2016), 100 μM 4-aminopyridine to block Kv3 (Coetzee et al., 1999; Gutman et al., 2005), 1 μM TTX to block TTX sensitive Nav channels, and 10 μM A803467 (Tocris) to block Nav1.8 (Jarvis et al., 2007). It is possible that off target effects of blockers may introduce errors in the quantification Kv2/KvS heteromer-mediated K⁺ currents. For example, 4-aminopyridine is expected to block a small fraction, 2%, of Kv2 homomers and have a lesser impact on Kv2/KvS heteromers (Post et al., 1996; Thorneloe and Nelson, 2003; Stas et al., 2015) which could result in a slight overestimation of the ratio of Kv2/KvS heteromers to Kv2 homomers.”

“We also tested the other major mouse C-fiber nociceptor population, peptidergic nociceptors, to determine if this subpopulation also has conductances resistant to RY785 yet sensitive to GxTX. We voltage clamped DRG neurons from a CGRP^GFP mouse line that expresses GFP in peptidergic nociceptors (Gong et al., 2003). Deep sequencing has identified mRNA transcripts for Kv6.2, Kv6.3, Kv8.1 and Kv9.3 present in GFP+ neurons, an overlapping but distinct set of KvS subunits from the Mrgprd^GFP non-peptidergic population (Zheng et al., 2019). In GFP+ neurons from CGRP^GFP mice, we found that a fraction of outward current was inhibited by 1 µM RY785 and additional current inhibited by 100 nM GxTX (Fig 7 Supplement 3 A-C). In these experiments, 58 ± 2% (mean ± SEM) was KvS-like (Fig 7 Supplement 3 D) identifying that KvSlike conductances are present in these peptidergic nociceptors. For CGRP^GFP neurons we did not include the Kv1, Kv3, Kv4, Nav and Cav channel inhibitor cocktail used for other neuron experiments, indicating that the cocktail of inhibitors is not required to identify KvS-like conductances.”

(ii) During the electrophysiology experiments, the authors use a holding potential that is not as negative as it is needed for the recording of the full population of the Kv2/KvS channels. Depolarized holding potentials lead to a certain level of inactivation of the channels, that vary according to the KvS involved/present in that specific population of channels. As a reminder, some KvS promote inactivation and others prevent inactivation. Therefore, the data must be interpreted as such.

We agree. We now point out that the physiological holding potentials used are insufficiently negative to relieve inactivation from all Kv2/KvS heteromeric channels. We also note that the ratio of Kv2-like to KvS-like conductance is expected to vary with voltage protocols.

Manuscript revisions:

“Neurons were held at a membrane potential of –74 mV to mimic a physiological resting potential. KvS subunits can profoundly shift the voltage-inactivation relation (Salinas et al., 1997a; Kramer et al., 1998; Kerschensteiner and Stocker, 1999) and this potential is likely insufficiently negative to relieve inactivation from all Kv2/KvS heteromeric channels. Also, the activation membrane potential is close to the half-maximal point of Kv2/KvS conductances. Thus the ratio of Kv2-like to KvS-like conductance is expected to vary with voltage protocols.”

(iii) The analysis of conductance activation by using tail currents is only accurate when dealing with non-inactivating conductances. Also, in dealing with a heterogenous population of Kv2/KvS heteromers, heterogenous K+ conductance deactivation kinetics is a must. Indeed, different KvS may significantly relate to different deactivation kinetics as well.

We now discuss that the bi-exponential fit of tail currents is likely inadequate to capture the deactivation kinetics of all underlying components of a heterogenous population of Kv2/KvS heteromers.

Manuscript revisions:

“We note that the analysis of conductance activation by using tail currents is only accurate when dealing with non-inactivating conductances. We expect that inactivation of Kv2/KvS conductances during the 200 ms pre-pulse is minimal (Salinas et al., 1997a; Kramer et al., 1998; Kerschensteiner and Stocker, 1999) and did not notice inactivation during the activation pulse. Also, deactivation kinetics can vary in a heterogenous population of Kv2/KvS heteromers. While analysis of tail currents could skew the quantification of total Kv2 like and KvS-like conductances, our data supports that mouse nociceptors and human neurons have tail currents that are resistant to RY785 and sensitive to GxTX consistent with the presence of Kv2/KvS heteromers.”

(iv) Silent Kv subunits may be retained in the ER, in heterologous systems like CHO cells. This aspect may subestimate their expression in these systems. Nevertheless, the authors show similar data in CHO cells and in primary neurons.

We agree. We now note that in heterologous systems, including CHO cells, transfection of KvS subunits can result in KvS subunits that are retained intracellularly.

Manuscript revisions:

“While a fraction of KvS subunits appear to be retained intracellularly, immunofluorescence for Kv5.1, Kv9.3 and Kv2.1 also appeared localized to the perimeter of transfected Kv2.1-CHO cells (Figure 1 Supplement).”

(v) The hallmark of silent Kv subunits is their effect on the time inactivation of K+ currents. As such, data should be shown throughout, preferably, from this perspective, but it was only done so in Figure 4G.

Indeed, effects on inactivation are a hallmark of KvS subunits. However, quantifying inactivation of Kv2/KvS channels requires steps to positive voltages for approximately 10 seconds. In neurons steps this long usually resulted in irreversible changes in leak currents/input resistance that degraded the accuracy of RY785/GxTX subtraction currents. Consequently, we did not acquire inactivation data in neurons, and we now explain in the manuscript why such data was not obtained.

Manuscript revisions:

“While changes in inactivation are prominent with KvS subunits, we did not investigate inactivation in neurons because the lengthy depolarizations required often resulted in irreversible leak current increases that degraded the accuracy of RY785/GxTX subtraction current quantification.”

(vi) Functional characterization of currents only, as suggested by the authors as a bona fide of Kv2 and Kv2/KvS currents, should not be solely trusted to classify the currents and their channel mediators.

We agree, and now state explicitly that functional characterization cannot be trusted to classify their channel mediators of conductances, and we try to be clear about this throughout the manuscript by using soft terms such as "KvS-like" when identity is uncertain.

Manuscript revisions:

“As functional characterization alone cannot be trusted to classify their channel mediators of conductances, we define conductances consistent with Kv2/KvS heteromers as 'KvS-like' and conductances consistent with Kv2 homomers as 'Kv2-like'.”

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

There is not a lot to do here - this was a real pleasure to read and very easy to understand, as written. Here are a few minor things to consider:

(1) The naming of the KvS subunits has always been confusing - it is not clear that Kv5,6,8,9 are members of the Kv2 subfamily from the names. KvS does a good job of differentiating them by assembly phenotype and has been used a lot in the literature, but it doesn't solve the misconception of what subfamily they belong to. This might not matter so much for mammals, where all KvS channels are in the Kv2 subfamily, but it makes it impossible to extend the naming system to other animals where subunits requiring heteromeric assembly are common in most subfamilies. How about trying the name Kv2S? It would have continuity with KvS in the reader's mind, make it clear that they are Kv2 subfamily, and make a naming system that could be extended beyond vertebrates. This is not a problem the authors created - just a completely optional suggestion on how to solve it if so inclined.

We agree that naming conventions for these subunits are problematic, and agonized quite a bit about nomenclature. In the end we chose to stick with the precedent of KvS.

(2) Another naming issue they should definitely change is the use of "subfamily" for the different KvS subtypes (Kv5, Kv6, Kv8, and Kv9). This really creates confusion with the higher-order subfamilies that have a very clear functional definition: a subfamily of Kv genes is a group of related genes that have assembly compatibility. Those are Kv1, Kv2, Kv3 and Kv4. KvS genes are assembly compatible with Kv2, evolutionarily derived from the Kv2 lineage, and thus clearly a part of the Kv2 subfamily. Using a subfamily for the next lower level of the naming hierarchy confuses this. The authors should use different terms like sub-type or class or subgroups for the divisions within KvS.

Thank you. We have standardized to Kv2/KvS as a subfamily; Kv5, Kv6, Kv8, and Kv9 as subtypes; and individual proteins, e.g. Kv8.1, as subunits.

(3) When you discuss whether the KvS subunit directly disrupts Ry785 binding in the pore or works allosterically and you said you know which KvS residues point into the pore from models, I thought that maybe you could tell from a sequence alignment whether the KvS channels you didn't test look the same in the conduction pathway as the ones you did test. If so, you could mention that if the binding site is the pore, they should all be resistant. Alternatively, if one you didn't test looks fundamentally more similar to the Kv2s in this region, then maybe it could be fingered as a possible exception that needs to be tested later.

Great ideas. We now assess sequence KvS variability near the proposed RY785 binding site in all KvS subunits. We generated structural models of RY785 docking to Kv2.1 and Kv2.1/Kv8.1 and found that residues near RY785 are different in all KvS subunits.

Manuscript revisions:

“We analyzed computational structural models of RY785 docked to a Kv2.1 homomer and a 3:1 Kv2.1:Kv8.1 heteromer (Fig 9) to gain structural insight into how KvS subunits might interfere with RY785 binding. We used Rosetta to dock RY785 to a cryo-EM structure of a Kv2.1 homomer in an apparently open state (Fernández-Mariño et al., 2023). The top-scoring docking pose has RY785 positioned below the selectivity filter and off-axis of the pore (Fig 9 A), similar to a stable pose observed in molecular dynamic simulations (Zhang et al., 2024). In this pose, RY785 contacts a collection of Kv2.1 residues that vary in every KvS subtype (Fig 9 B,D,E). Notably, RY785 bound similarly to a 3:1 model of Kv2.1/Kv8.1, in contact with the three Kv2.1 subunits, yet avoided the Kv8.1 subunit (Fig 9C). This is consistent with RY785 binding less well to Kv2.1/Kv8.1 heteromers, and also suggests that a 3:1 Kv2:KvS channel could retain a RY785 binding site when open.”

(4) Future suggestion or tip - not for this paper. Your data shows your isolation strategy works really well on Kv6 channels, and these are also the Kv2/KvS channels that have the most pronounced biophysical changes. Working on neurons that have a prominent Kv2/Kv6 component would really show how well the strategy outlined here works to describe the physiology of native neurons. The highest KvS expression I have seen in public data in a wellstudied cell type is Kv6.4 in spinal motor neurons.

Wonderful tip, thank you. We are indeed very interested in Kv6.4 in spinal motor neurons.

Reviewer #2 (Recommendations For The Authors):

The manuscript makes a good contribution to the identification of Kv2/KvS channels in primary cells. The pharmacological method proposed by the authors to dissect the currents in an experimental setting seems proper. Although meritorious in themselves, the findings are heavily phenomenological in the opinion of this reviewer. The manuscript should be improved with some level of mechanistic data and/or the demonstration of different levels of expression in different cell types.

Thank you for the suggestions. This manuscript now demonstrates strikingly higher levels of the KvS-like component of Kv2 currents in somatosensory (DRG nonpeptidergic and peptidergic nociceptor) versus autonomic (SCG) neuron types. The mechanistic question of what electrophysiological properties the KvS subunits are providing to the neuronal circuit is an exciting one that we are pursuing separately.

Manuscript revisions:

“While we found only RY785-sensitive Kv2-like conductances in SCG neurons, Kv2/KvS heteromer-like conductances were dominant in DRG neurons.”

At present, the manuscript says that the combination of RY785 and guangxitoxin-1E can be used to define Kv2/KvS-mediated K+ currents. Importantly, this method cannot be used in a way that one can functionally determine the function of Kv2/KvS channels, since it depends on the pre-blocking of Kv2-mediated K+ currents prior. In the opinion of this reviewer, this fact decreases the attention of a potential reader.

Indeed, our study is focused on revealing KvS heteromers by voltage clamp, and we now clarify in the introduction that we do not determine the function of Kv2/KvS channels in this study, so as not to lead the reader to expect studies of neuronal signaling.

However, the selective pharmacology we identify suggests RY785 application could reveal the function of Kv2 homomers, and for RY785-insensitive signaling, GxTX application of could reveal the function of Kv2/KvS heteromers. We now mention these possible applications in the Discussion.

Manuscript revisions:

“While this study does not address the impact of GxTX or RY785 on action potentials or in vivo, the distinct pharmacology of Kv2/KvS heteromers presented here suggests that KvS conductances could be targeted to selectively modulate discrete subsets of cell types.”

Please find below suggestions for improving the manuscript:

(1) The term "Kv2/KvS heteromers" should be used throughout instead of variations such as "Kv2/KvS channels", "Kv2/KvS" and others. Standardization of the term to refer to heteromers would make the manuscript easier to read.

Thank you. We have standardized terms to consistently refer to Kv2/KvS heteromers.

(2) Confusing terms like KvS conductances, KvS-like conductances, KvS-like (RY785-resistant, GxTX-sensitive) currents, and KvS channels should be avoided because they disregard the current belief that KvS cannot form functional homomeric channels. The term KvS-containing channels, and Kv2/KvS channels, seem more accurate. Uniformization in this regard will also make the manuscript more easily readable.

Thank you. We have standardized terms to Kv2/KvS heteromers and KvS-containing channels when channel subunits are known and the use terms KvS-like and Kv2-like for functionally identified endogenous conductances with unknown channel subunits.

(3) Referring to KvS as a regulatory subunit is inaccurate. It is clear that KvS is part of, and it makes up the alpha pore. KvS therefore is a part of the conductive pathway and not a regulatory (suggesting accessory) subunit. KvS take part in selectivity filter (fully conserved), but they also make up an important part of the conducting pathway with non-conserved amino acid residues.

We felt it important to include the descriptor “regulatory” to connect our nomenclature with prior use of the descriptor in the literature, and now only use the term at the start of the introduction.

Manuscript revisions:

“A potential source of molecular diversity for Kv2 channels are a group of Kv2-related proteins which have been referred to as regulatory, silent, or KvS subunits.”

(4) The use of a cocktail of channel inhibitors may affect the quantification of Kv2/KvS heteromers-mediated K+ currents because they may interact with RY785 and/or GxTx or they may even interact with the sites for these two drugs on Kv2-containing channels.

This is an interesting point worth considering, thank you. We now alert readers to this possibility in the discussion when considering the limitations of our approach.

Manuscript revisions:

“Also, the cocktail of inhibitors used in most neuron experiments here could potentially alter RY785 or GxTX action against KvS/Kv2 channels.”

(5) The graphical representation of fractional blocking and other parameters (e.g., Fig 1D), is hard to read in these slim plots. In my opinion, tall bars would be more meaningfully visualized.

Thank you for pointing out that the graphs were hard to read, we have made the graph easier to read and added tall bars.

(6) Vehicle control for IHC and electrophysiology. Please state what is the vehicle used in the electrophysiology experiments.

Thank you. The composition of vehicle has now been stated in the methods.

Manuscript revisions:

“All RY785 solutions contained 0.1% DMSO. Vehicle control solutions also contained 0.1% DMSO but lacked RY785.”

“Sections were incubated in vehicle solution (4% milk, 0.2% triton diluted in PB) for 1 hr at RT.”

(7) The reference Trapani & Korn, 2003 (?) is not included in the list. This reference is important since it sets what are the Kv2.1-CHO cells. In this regard it is also important to mention, even better to address, the expressing qualities of this system in the face of a co-expression with a plasmid-based expression of silent Kv subunits. Are these two ways of expressing Kv subunits, meant to come together (or not) in heteromers, balanced? This question is critical here. Still, in regard to Kv2.1-CHO cells, it was not clear in the manuscript if the term "transfection" refers only to the plasmids used to temporarily induce the expression of silent Kv subunits and potentially Kv channels accessory subunits.

We now include the Trapani & Korn, 2003 reference (thank you for pointing out this accidental omission), and better explain expression methods. The benefit of the inducible Kv2.1 expression is control of Kv conductance densities which can otherwise become so large as to be refractory to voltage clamp. The beauty of the expression system is that cells recently transfected with KvS subunits can be induced to express just enough Kv2.1 to get a substantial but not clampoverwhelming RY785-resistant Kv2/KvS conductance. We also discuss that our expression methods are distinct from past studies. We stop short of comparing the expression systems, as this is beyond the scope of what we set out to study.

Manuscript revisions: See next response

(8) Kv2.1-CHO cells transfection procedures, induction, and validation are unclear. This validation is important here.

We have clarified transfection procedures, induction, and validation in the methods section.

Manuscript revisions:

“The CHO-K1 cell line transfected with a tetracycline-inducible rat Kv2.1 construct (Kv2.1-CHO) (Trapani and Korn, 2003) was cultured as described previously (Tilley et al., 2014).”

Transfections were achieved with Lipofectamine 3000 (Life Technologies, L3000001). 1 μl Lipofectamine was diluted, mixed, and incubated in 25 μl of Opti-MEM (Gibco, 31985062).”

“Concurrently, 0.5 μg of KvS or AMIGO1 or Navβ2, 0.5 μg of pEGFP, 2 μl of P3000 reagent and 25 μl of Opti-MEM were mixed. DNA and Lipofectamine 3000 mixtures were mixed and incubated at room temperature for 15 min. This transfection cocktail was added to 1 ml of culture media in a 24 well cell culture dish containing Kv2.1-CHO cells and incubated at 37 °C in 5% CO2 for 6 h before the media was replaced. Immediately after media was replaced, Kv2.1 expression was induced in Kv2.1-CHO cells with 1 μg/ml minocycline (Enzo Life Sciences, ALX380-109-M050), prepared in 70% ethanol at 2 mg/ml. Voltage clamp recordings were performed 12-24 hours later. We note that the expression method of Kv2/KvS heteromers used here is distinct from previous studies which show that the KvS:Kv2 mRNA ratio can affect the expression of functional Kv2/KvS heteromers (Salinas et al., 1997b; Pisupati et al., 2018). We validated the functional Kv2/KvS heteromer expression using voltage clamp to establish distinct channel kinetics and the presence of RY785-resistant conductance in KvS-transfected cells and using immunohistochemistry to label apparent surface localization of KvS subunits (Figure 4, Figure 1 Supplement, Figure 1 and Figure 5).”

(9) It is important for readers to add some context to Kv2.1/Kv8.1 channels (and other Kv2/KvS heteromers) used to test the combination of RY785 and GxTx. In my opinion, this enriches the discussion.

Good idea. We have added context about each of the KvS subunits we test.

Manuscript revisions:

“To test the pharmacological response of KvS we began with Kv8.1, a subunit that creates heteromers with biophysical properties distinct from Kv2 homomers (Salinas et al., 1997a), and modulates motor neuron vulnerability to cell death (Huang et al., 2024).

Each of these KvS subunits create Kv2/KvS heteromers that have distinct biophysical properties (Kramer et al., 1998; Kerschensteiner and Stocker, 1999; Bocksteins et al., 2012). Kv5.1/Kv2.1 heteromers play an important role in controlling the excitability of mouse urinary bladder smooth muscle (Malysz and Petkov, 2020), mutations in Kv6.4 have been shown to influence human labor pain (Lee et al., 2020b), and deficiency of Kv9.3 disrupts parvalbumin interneuron physiology in mouse prefrontal cortex (Miyamae et al., 2021).”

(10) In general, the membrane potential used to activate Kv2 only channels and Kv2/KvS channels is too close to the activation V1/2. In case the comparing curves are displaced in their relative voltage dependence and voltage sensitivity, using that range of membrane potential may introduce a crucial error in the estimation of the conductance's relative amplitudes.

We now note that the relative conductances of Kv2-only vs Kv2/KvS channels are expected to vary with voltage protocol, as KvS inclusion results in channels with altered voltage responses.

Manuscript revisions:

“…the activation membrane potential is close to the half-maximal point of Kv2/KvS conductances. Thus the ratio of Kv2-like to KvS-like conductance is expected to vary with voltage protocols.”

(11) The use of tail currents to estimate conductance is problematic if i) lack of current inactivation is not assured, and ii) if the different currents, with possible different deactivation kinetics at the used membrane potential (e.g., mV), are not assured. Why was the activation peak used at times, and at different elapsed times the tail currents were used instead? These aspects of conductance's amplitude estimation methods should be well defined.

In CHO cells peak currents were analyzed because outward currents seem to offer the best signal/noise. In neurons, we restricted analysis to tail currents at elapsed times to minimize complications from non-Kv2 endogenous voltage-gated channels which deactivate more quickly. We have clarified this analysis in the methods section.

Manuscript revisions:

“In CHO cells peak currents were analyzed because outward currents seem to offer the best signal/noise. In neurons, we restricted analysis to tail currents at elapsed times to minimize complications from non-Kv2 endogenous voltage-gated channels which deactivate more quickly. In neurons, voltage gated currents remained in the toxin cocktail + RY785 and GxTX, that were sometimes unstable. To minimize complications from these currents, we restricted analysis of RY785 and GxTX subtraction experiments to tail currents at elapsed times to minimize complications from non-Kv2 endogenous voltage-gated channels which deactivate more quickly. We note that the analysis of conductance activation by using tail currents is only accurate when dealing with non-inactivating conductances. We expect that inactivation of Kv2/KvS conductances during the 200 ms pre-pulse is minimal (Salinas et al., 1997a; Kramer et al., 1998; Kerschensteiner and Stocker, 1999) and did not notice inactivation during the activation pulse. Also, deactivation kinetics can vary in a heterogenous population of Kv2/KvS heteromers. While analysis of tail currents could skew the quantification of total Kv2 like and KvS-like conductances, our data supports that mouse nociceptors and human neurons have tail currents that are resistant to RY785 and sensitive to GxTX consistent with the presence of Kv2/KvS heteromers.”

(12) Were the experiments including different conditions such as control, RY, and RY+GxTx done pair-wised? This could potentially better the statistics and strengthen the data and the conclusions drawn from them.

The control, RY, and RY+GxTX in neurons were done pairwise and the statistical tests performed for these experiments were pairwise tests. We have clarified this in the figure legends.

Manuscript revisions:

“Wilcoxon rank tests were paired, except the comparison of RY785 to vehicle which was unpaired.”

(13) The holding potential of the experiments, mostly -89 mV, may be biasing the estimation of Kv2 only channels vs. Kv2/KvS channels conductances. Figure 4I exemplifies this concern.

We agree. Figure 4I reveals that a holding potential of -89 mV vs -129 mV reduces conductance of Kv2.1/Kv8.1 heteromers vs Kv2.1 homomers in CHO cells by ~20%. We have now alerted readers that the ratio of Kv2 only channels vs. Kv2/KvS conductances can vary with holding voltage.

Manuscript revisions:

“Under these conditions, 58 ± 3 % (mean ± SEM) of the delayed rectifier conductance was resistant to RY785 yet sensitive to GxTX (KvS-like) (Fig 7 F). We note that the ratio of KvS- to Kv2-like conductances is expected to vary with holding potential, as KvS subunits can change the degree and voltage-dependence of steady state inactivation (e.g. Fig 4I).”

(14) It is possible that Figure 6A (control trace) and Figure 6C ("Kv2-like" trace) are the same, by mistake, since their noise pattern looks too similar.

Indeed the noise pattern of the Figure 6A (control trace) and Figure 6C ("Kv2-like" trace) are related, as they have inputs from the same trace, with Figure 6C ("Kv2-like" trace) being a subtraction of Figure 6A (+RY trace) from Figure 6A (control trace).

(15) For example, in Figure 7A, what is the identity of the current remaining after the RY+GxTx application? In Figure 7B, a supposed outlier in the group of data referring to "veh" in the right panel is what possibly is making this group different from +RY in the left panel (p=0.02, Wilcoxon rank test). I would recommend parametric tests only since the data is essentially quantitative.

In Figure 7A, we do not know the identity of the current remaining after the RY+GxTX application, the kinetics of the residual current appeared distinct from the Kv2/KvS-like currents blocked by RY or GxTX, but we did not analyze these.

The date in Figure 7B, was indeed the positive outlier in the group of data referring to "veh" in the right panel and contributes to the p-value, but we saw no reason to exclude it. We have now replaced the representative trace in 7B with a non-outlier trace. We respectfully disagree with the suggestion to use parametric statistical tests as we do not know the distribution underlying the variance our data.

Manuscript revisions:

“Subsequent application of 100 nM GxTX decreased tail currents by 68 ± 5% (mean ± SEM) of their original amplitude before RY785. We do not know the identity of the outward current that remains in the cocktail of inhibitors + RY785 + GxTX.”

(16) Please state the importance of using nonpeptidergic neurons to study silent Kv5.1 and Kv9.1 subunits. RNA data may not necessarily work to probe function or protein abundance, which is crucial in heteromeric complexes.

We have now more thoroughly explained our rationale for choosing the nonpeptidergic neurons.

RNA is not predictive of protein abundance, and we have not yet been successful in measuring KvS protein abundance in these neurons, so we've probed KvS abundance by assessing RY785 resistance.

Manuscript revisions:

“Mouse dorsal root ganglion (DRG) somatosensory neurons express Kv2 proteins (Stewart et al., 2024), have GxTX-sensitive conductances (Zheng et al., 2019), and express a variety of KvS transcripts (Bocksteins et al., 2009; Zheng et al., 2019), yet transcript abundance does not necessarily correlate with functional protein abundance. To record from a consistent subpopulation of mouse somatosensory neurons which has been shown to contain GxTXsensitive currents and have abundant expression of KvS mRNA transcripts (Zheng et al., 2019), we used a Mrgprd^GFP transgenic mouse line which expresses GFP in nonpeptidergic nociceptors (Zylka et al., 2005; Zheng et al., 2019). Deep sequencing identified that mRNA transcripts for Kv5.1, Kv6.2, Kv6.3, and Kv9.1 are present in GFP+ neurons of this mouse line (Zheng et al., 2019) and we confirmed the presence of Kv5.1 and Kv9.1 transcripts in GFP+ neurons from Mrgprd^GFP mice using RNAscope (Fig 7 Supplement 1).”

(17) In Figure 8B, were +RY data different from veh data? The figure shows no Wilcoxon (nonparametric) comparison and this is important to be stated. What conductance(s) is the vehicle solution blocking or promoting? What is RY dissolved in, DMSO? What is the DMSO final concentration?

We now state that in Figure 8B, +RY amplitudes were not statistically different from veh data in this limited data set. However, the RY-subtraction currents always had Kv2-like biophysical properties, whereas vehicle-subtraction currents had variable properties precluding biophysical analysis for Fig 8D.

In Figure 8B, we do not know what conductance(s) the vehicle solution is affecting, we think the changes observed are likely merely time dependent or due to the solution exchange itself. RY stock is in DMSO. All recording solutions have 0.1% DMSO final concentration, this is now noted in methods.

Manuscript revisions:

“Unlike mouse neurons, we did not detect a significant difference in tail currents of RY785 versus vehicle controls. However, RY785-subtracted currents always had Kv2-like biophysical properties whereas vehicle-subtraction currents had variable properties that precluded the same biophysical analysis. Overall, these results show that human DRG neurons can produce endogenous voltage-gated currents with pharmacology and gating consistent with Kv2/KvS heteromeric channels.”

“All RY785 solutions contained 0.1% DMSO. Vehicle control solutions also contained 0.1% DMSO but lacked RY785.”

(18) METHODS. The electrophysiology approach should be unified in all aspects as applicable and possible.

We have unified the mouse dorsal root ganglion and mouse superior cervical ganglion methods sections. We have kept CHO cells and mouse/human neurons section separate because the methods were substantially different.

(19) DISCUSSION. The discussion section spends half of its space trying to elaborate on possible blocking/inhibiting/modulating mechanisms for RY785. The present manuscript shows no data, at least not that I have noticed, that would evoke such discussion.

We have shortened this section, and enhance the discussion with structural models (new Fig 9), and our functional data indicating perturbed RY785 interaction with Kv2.1/8.1.

Manuscript revisions:

“In this pose, RY785 contacts a collection of Kv2.1 residues that vary in every KvS subtype (Fig 9 B,D,E). Notably, RY785 bound similarly to a 3:1 model of Kv2.1/Kv8.1, in contact with the three Kv2.1 subunits, yet avoided the Kv8.1 subunit (Fig 9C). This is consistent with RY785 binding less well to Kv2.1/Kv8.1 heteromers, and also suggests that a 3:1 Kv2:KvS channel could retain a RY785 binding site when open. However, the RY785 resistance of Kv2/KvS heteromers may primarily arise from perturbed interactions with the constricted central cavity of closed channels. In homomeric Kv2.1, RY785 becomes trapped in closed channels and prevents their voltage sensors from fully activating, indicating that RY785 must interact differently with closed channels (Marquis and Sack, 2022). Here we found that Kv2.1/Kv8.1 current rapidly recovers following washout of RY785, suggesting that Kv2.1/Kv8.1 heteromers do not readily trap RY785 (Figure 2 Supplement). Overall, the structural modeling suggests that KvS subunits sterically interfere with RY785 binding to the central cavity, while functional data suggest KvS subunits disrupt RY785 trapping in closed states.”

(20) DISCUSSION. Topics like ER retention and release upon certain conditions would be a better enrichment for the manuscript in my opinion.

ER retention of KvS subunits is indeed an important topic! However, we have opted not to delve into it here.

(21) DISCUSSION. Speculation about the binding site for RY on Kv2/KvS channels is also not touched by the data shown in the manuscript.

We have shortened this section of discussion, and now present this with structural models of RY785 docked to a Kv2.1 homomer and 3:1 Kv2.1: Kv8.1 heteromer (new Fig 9) to ground speculations. See manuscript changes noted in response to comment (19) above.

(22) DISCUSSION. An important reference is missing in regard to stoichiometry: Bocksteins et al., 2017. This work is the only one using a non-optical technique to add knowledge to that question.

Good point, and an excellent study we didn’t realize we’d not included before. We now include Bocksteins et al. 2017 as a reference in the Introduction.

(23) In my opinion, allosterism and orthosterism are concepts not yet useful for the discussion of RY binding sites without even a general piece of data.

We now include structural models of RY785 docked to a Kv2.1 homomer and 3:1 Kv2.1: Kv8.1 heteromer (new Fig 9) to ground blocking speculations. See manuscript changes noted in response to comment (19).

(24) The term "homogeneously susceptible" associated with a Hill slope close to 1 needs to be more elaborated.

Thank you, we have elaborated.

Manuscript revisions:

“Also, the degree of resistance to RY785 may vary if Kv2:KvS subunit stoichiometry varies. With high doses of RY785, we found that the concentration-response characteristics of Kv2.1/Kv8.1 in CHO cells revealed hallmarks of a homogenous channel population with a Hill slope close to 1 (Fig 2B). However, other KvS subunits might assemble in multiple stoichiometries and result in pharmacologically-distinct heteromer populations.”

(25) Stating the KvS are resistant to RY785 is not proper in my opinion. This opinion relates to the fact that the RY binding site in the channels is certainly not restricted to a binding site residing only on the Kv subunit.

Good point. We have now changed phrasing to convey that KvS subunits are a component of a heteromer that imbues RY785 resistance.

Manuscript revisions:

“These results show that voltage-gated outward currents in cells transfected with members from each KvS subtype have decreased sensitivity to RY785 but remain sensitive to GxTX. While we did not test every KvS subunit, the ubiquitous resistance suggests that all KvS subunits may provide resistance to 1 μM RY785 yet remain sensitive to GxTX, and that RY785 resistance is a hallmark of KvS-containing channels.”

eLife Assessment

The study presents compelling evidence that the melanocortin system originating in the arcuate nucleus of the hypothalamus plays a crucial role in puberty onset, representing a significant advance in our understanding of reproductive biology. The research employs innovative approaches and benefits from the combined expertise of two respected laboratories, enhancing the robustness of the findings. Given the potential impact on human health and the strength of the evidence presented, this fundamental work will likely influence the field substantially and may inform future clinical applications.

Reviewer #2 (Public review):

Summary:

I found this an interesting manuscript describing a study investigating the role of MC4R signalling on kisspeptin neurons. The initial question is a good one. Infertility associated with MC4 mutations in humans has typically been ascribed to the consequent obesity and impaired metabolic regulation. Whether there is a direct role for MC4 in regulating the HPG axis has not been thoroughly examined. Here, the researchers have put together an elegant combination of targeted loss of function and gain of function in vivo experiments, specifically targeting MC4 expression in kisspeptin neurons. This excellent experimental design should provide compelling evidence for whether melanocortin signalling has a direct role in arcuate kisspeptin neurons to support normal reproductive function. There were definite effects on reproductive function (irregular estrous cycle, reduced magnitude of LH surge induced by exogenous estradiol). However, the magnitude of these responses and the overall effect on fertility were relatively minor. The mice lacking MC4R in kisspeptin neurons remained fertile despite these irregularities. The second part of the manuscript describes a series of electrophysiological studies evaluating the pharmacological effects of melanocortin signalling in kisspeptin cells in ex-vivo brain slides. These studies characterised interesting differential actions of melanocortins in two different populations of kisspeptin neurons. Collectively, I think the study provides novel insights into how direct actions of melanocortin signalling, via the MC4 receptor in kisspeptin neurons, contribute to the metabolic regulation of the reproductive system. Importantly, however, it is clear that other mechanisms are also at play.

Strengths:

The loss of function/gain of function experiments provide a conceptually simple but hugely informative experimental design. This is the key strength of the current paper - especially the knock-in study that showed improved reproductive function even in the presence of ongoing obesity. This is a very convincing result that documents that reproductive deficits in MC4R knockout animals (and humans with deleterious variants of the MC4R gene) can be ascribed to impaired signalling in the hypothalamic kisspeptin neurons and not necessarily simply caused as a consequence of obesity. As concluded by the authors: "reproductive impairments observed in MC4R deficient mice, which replicate many of the conditions described in humans, are largely mediated by the direct action of melanocortins via MC4R on Kiss1 neurons and not to their obese phenotype." This is important, as it might change the way such fertility problems are treated.

Limitation:

The mechanistic studies evaluating melanocortin signalling in kisppetin neurons were all completed in ovariectomized animals (with and without exogenous hormones). This reductionist approach allowed a focus on the direct actions of estradiol to regulate responses but missed an opportunity to evaluate how cyclical changes in hormones might impact the system. Such cyclical changes are fundamental to how these neurons function in vivo and may dynamically alter the way they respond to hormones and neuropeptides. However, the inclusion of gonad-intact animals would have significantly increased the complexity of experiments and can reasonably be considered outside of the scope of the present study.

Author response:

The following is the authors’ response to the previous reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors investigate the role of the melanocortin system in puberty onset. They conclude that POMC neurons within the arcuate nucleus of the hypothalamus provide important but differing input to kisspeptin neurons in the arcuate or rostral hypothalamus.

Strengths:

Innovative and novel

Technically sound

Well-designed

Thorough

Weaknesses:

There were no major weaknesses identified.

Reviewer #2 (Public review):

Summary:

This interesting manuscript describes a study investigating the role of MC4R signalling on kisspeptin neurons. The initial question is a good one. Infertility associated with MC4 mutations in humans has typically been ascribed to the consequent obesity and impaired metabolic regulation. Whether there is a direct role for MC4 in regulating the HPG axis has not been thoroughly examined. Here, the researchers have assembled an elegant combination of targetted loss of function and gain of function in vivo experiments, specifically targetting MC4 expression in kisspeptin neurons. This excellent experimental design should provide compelling evidence for whether melanocortin signalling dirently affects arcuate kisspeptin neurons to support normal reproductive function. There were definite effects on reproductive function (irregular estrous cycle, reduced magnitude of LH surge induced by exogenous estradiol). However, the magnitude of these responses and the overall effect on fertility were relatively minor. The mice lacking MC4R in kisspeptin neurons remained fertile despite these irregularities. The second part of the manuscript describes a series of electrophysiological studies evaluating the pharmacological effects of melanocortin signalling in kisspeptin cells in ex-vivo brain slides. These studies characterised interesting differential actions of melanocortins in two different populations of kisspeptin neurons. Collectively, the study provides some novel insights into how direct actions of melanocortin signalling via the MC4 receptor in kisspeptin neurons contribute to the metabolic regulation of the reproductive system. Importantly, however, it is clear that other mechanisms are also at play.

Strengths:

The loss of function/gain of function experiments provides a conceptually simple but hugely informative experimental design. This is the key strength of the current paper - especially the knock-in study that showed improved reproductive function even in the presence of ongoing obesity. This is a very convincing result that documents that reproductive deficits in MC4R knockout animals (and humans with deleterious MC4R gene variants) can be ascribed to impaired signalling in the hypothalamic kisspeptin neurons and not necessarily caused as a consequence of obesity. As concluded by the authors: "reproductive impairments observed in MC4R deficient mice, which replicate many of the conditions described in humans, are largely mediated by the direct action of melanocortins via MC4R on Kiss1 neurons and not to their obese phenotype." This is important, as it might change how such fertility problems are treated.

I would like to see the validation experiments for the genetic manipulation studies given greater prominence in the manuscript because they are critical to interpretation. Presently, only single unquantified images are shown, and a much more comprehensive analysis should be provided.

Weaknesses:

(1) Given that mice lacking MC4R in kisspeptin neurons remained fertile despite some reproductive irregularities, this can be described as a contributing pathway, but other mechanisms must also be involved in conveying metabolic information to the reproductive system. This is now appropriately covered in the discussion.

(2) The mechanistic studies evaluating melanocortin signalling in kisspeptin neurons were all completed in ovariectomised animals (with and without exogenous hormones) that do not experience cyclical hormone changes. Such cyclical changes are fundamental to how these neurons function in vivo and may dynamically alter how they respond to hormones and neuropeptides. Eliminating this variable makes interpretation difficult, but the authors have justified this as a reductionist approach to evaluate estradiol actions specifically. However, this does not reflect the actual complexity of reproductive function.

For example, the authors focus on a reduced LH response to exogenous estradiol in ovariectomised mice as evidence that there might be a sub-optimal preovulatory LH surge. However, the preovulatory LH sure (in intact animals) was not measured.

They have not assessed why some follicles ovulated, but most did not. They have focused on the possibility that the ovulation signal (LH surge) was insufficient rather than asking why some follicles responded and others did not. This suggests some issue with follicular development, likely due to changes in gonadotropin secretion during the cycle and not simply due to an insufficient LH surge.

Reviewer #3 (Public review):

The manuscript by Talbi R et al. generated transgenic mice to assess the reproduction function of MC4R in Kiss1 neurons in vivo and used electrophysiology to test how MC4R activation regulated Kiss1 neuronal firing in ARH and AVPV/PeN. This timely study is highly significant in neuroendocrinology research for the following reasons.

(1) The authors' findings are significant in the field of reproductive research. Despite the known presence of MC4R signaling in Kiss1 neurons, the exact mechanisms of how MC4R signaling regulates different Kiss1 neuronal populations in the context of sex hormone fluctuations are not entirely understood. The authors reported that knocking out Mc4r from Kiss1 neurons replicates the reproductive impairment of MC4RKO mice, and Mc4r expression in Kiss1 neurons in the MC4R null background partially restored the reproductive impairment. MC4R activation excites Kiss1 ARH neurons and inhibits Kiss1 AVPV/PeN neurons (except for elevated estradiol).

(2) Reproduction dysfunction is one of obesity comorbidities. MC4R loss-of-function mutations cause obesity phenotype and impaired reproduction. However, it is hard to determine the causality. The authors carefully measured the body weight of the different mouse models (Figure 1C, Figure 2A, Figure 3B). For example, the Kiss1-MC4RKO females showed no body weight difference at puberty onset. This clearly demonstrated the direct function of MC4R signaling in reproduction but was not a consequence of excessive adiposity.

(3) Gene expression findings in the "KNDy" system align with the reproduction phenotype.

(4) The electrophysiology results reported in this manuscript are innovative and provide more details of MC4R activation and Kiss1 neuronal activation.

Overall, the authors have presented sufficient background in a clear, logical, and organized structure, clearly stated the key question to be addressed, used the appropriate methodology, produced significant and innovative main findings, and made a justified conclusion.

Comments on revisions:

The authors have addressed my comments.

Recommendations for the authors:

The reviewers noted that they received comments in response to their concerns, and some improvements have been made to the manuscript. However, as described below, in some cases, a rebuttal was provided, but changes were not made to the manuscript. It is suggested that these issues be addressed to improve the quality of the manuscript.

We thank the reviewers and editor for the assessment of the manuscript and recommendations for its improvement. We have addressed the remaining comments from reviewer #2 below, and hope that they find our revisions satisfactory.

Reviewer #2 (Recommendations for the authors):

The manuscript convincingly shows that MC4R in kisspeptin-producing cells can influence reproductive function. This suggests that fertility problems associated with melanocortin mutations are likely due to direct effects on the reproductive systems rather than simply being side effects of the resultant obesity.

We are pleased that this reviewer finds the data convincing and thank them for the careful review of the manuscript, which has helped to improve its published version.

The authors have responded to the reviewer's comments and made several improvements to the manuscript.

The authors are correct in pointing out that the POMC-Cre animals should be fine for studies involving the administration of AAVs to adult animals. I have misinterpreted how these mice were being used, and this concern is fully addressed.

Unfortunately, in some cases, the authors rebutted the reviewer's comments but did not change the manuscript. I suggest addressing several issues in the manuscript (after all, it is not the reviewer's opinion that counts; this process is about improving the manuscript).

(1) Validation of the KO is insufficiently reported. From the methods, it appears that this was done thoroughly, but currently, only a single image of the arcuate nucleus is shown, and no image of the AVPV is shown. There is no quantitative information provided. The authors can keep these data as supplementary material, but they should be comprehensive and convincing, as so much depends on the degree of knockout in this model. One cannot assume complete KO based simply on the relevant genetics, as there are examples in this system where different Cre lines produce different outcomes with various floxed genes in the two major populations of kisspeptin neurons. This figure should show the quantitation of the RNAscope analysis from each of the two regions regarding the percentage of kisspeptin cells showing expression of MC4R mRNA. In addition, the lack of MC4 labelling in the arcuate nucleus, outside of kisspeptin neurons, is a concern. One would expect to see AgRP or POMC cells at this level, but are they still showing expression of MC4? A single image is insufficient to be convinced of the model's efficacy.

We appreciate the reviewer’s concerns regarding the validation of the MC4RKO model. Below, we provide clarification and additional justification for our approach.

(1) Quantification of MC4R in the Arcuate Nucleus (ARC): As noted by the reviewer, we were unable to detect sufficient MC4R signal in the ARC of KO mice to perform meaningful quantification. This is consistent with the expected outcome of a successful MC4R deletion. Given the low endogenous expression levels of MC4R in this region, even in control animals, and the technical limitations of RNAscope in detecting very low-abundance transcripts, especially for receptors, the absence of MC4R signal in the ARC of KO mice strongly supports effective deletion. Moreover, the MC4R loxP mouse has been published and validated by many labs including Brad Lowell’s lab who’s done extensive work using these mice for selective deletion of Mc4r from various neuronal populations such as Sim1 and Vglut2 neurons (Shah et al., 2014, de Souza Cordeiro et al., 2020). To further strengthen our validation, we provide additional images from another animal (Fig_S1) to illustrate the consistency of the MC4R KO in the ARC. These will be included as supplementary material, as suggested.Regarding AgRP and POMC neurons, MC4R is not highly expressed in these neurons (as per previous literature, e.g., Garfield et al., Nat Neurosci. 2015; Padilla SL et al, Endocrinology 2012; Henry et al, Nature, 2015). Instead, MC4R is predominantly found in downstream neurons in the paraventricular nucleus (PVN) and other hypothalamic regions (which is intact in our KO mice as shown in our validation figure). Thus, the absence of MC4R labeling in AgRP or POMC cells in our images aligns with known expression patterns and does not contradict the validity of our model.

(2) MC4R Expression in the AVPV and OVX Effect on Kiss1 Expression: We acknowledge the reviewer’s request for MC4R expression analysis in the anteroventral periventricular nucleus (AVPV). However, due to the timing of tissue collection after ovariectomy (OVX), Kiss1 expression in the AVPV is significantly suppressed, making it technically unfeasible to perform co-staining of MC4R with Kiss1 in this region. This is a well-documented effect of estrogen depletion following OVX (Smith et al., 2005; Lehman et al., 2010). While we acknowledge that an ideal validation would include AVPV co-labeling, the experimental constraints related to OVX preclude this analysis in our dataset.

Given these considerations and validations, we are confident that the KO is effective and specific.

(2) Line 88: "... however, conflicting reports exist". Expand on this sentence to describe what these conflicting reports show. The authors responded to my comment but made no changes to the introduction. As a reader, I dislike being told there are conflicting reports, but then I have to go and look up the reference to see what that actual point of conflict is.

By conflicting reports we meant that other studies have shown no association between MC4R and reproductive disorders, this has now been included in the revised manuscript (Line 89).

(3) Could the authors explain how a decrease in AgRP would be interpreted as a "decrease in hypothalamic melanocortin tone" in line 142 and line 364? These overly simplistic interpretations of qPCR data detract from the overall quality of the paper.

The reference to a decrease in melanocortin tone referred to the decrease in the expression of melanocortin receptor signaling, this has been clarified in the revised manuscript (lines 142 and 360).

(4) Please show the individual cycle patterns for all animals, as in Figure 2B. This can be a supplemental figure, but the current bar charts are not informative.

We respectfully disagree that the bar charts are not informative as they include the critical statistical analysis. We have now included all individual estrous cycle data in new separate supplemental figure (Sup. Figure 3). Therefore, we have excluded the representative cycles from the main figures as they are now in the new Supplemental. We have changed the orders of the figures in the text accordingly.

(5) In their rebuttal, the authors state: "Mice lack true follicular and luteal phases, and therefore, it is impossible to separate estrogen-mediated changes from progesterone-mediated changes (e.g., in a proestrous female). Therefore, we use an ovariectomized female model in which we can generate an LH surge with an E2-replacement regimen [1]. This model enables us to focus on estrogen effects, exclude progesterone effects, and minimize variability. Inclusion of cycling females would make interpretation much more difficult." I disagree, but the authors can take this position if they wish. However, they should not report the responses to exogenous estradiol in an ovariectomised mouse as a "preovulatory LH surge" (line 380). An ovariectomised mouse cannot ovulate, and the estrogen-induced LH surge is significantly different in magnitude and timing from the endogenous preovulatory LH surge (likely due to the actions of progesterone). One goal of these studies is to understand why the ovulation rate appears to be low in the MC4-KO animals. Hence, evaluating whether the preovulatory LH surge is typical is important. This has not been done. The authors have shown that the response to exogenous estradiol is sub-normal. Such an effect might lead to a reduced preovulatory LH surge, but this has not been measured.

We appreciate this reviewer’s concern about the nature of the preovulatory LH surge. We have clarified this in the revised manuscript and described it as “an induced LH surge” throughout the text (Lines 163, 533, 6560).

(6) I believe that the ovulation process should be considered "all or none," and I do not quite understand the rebuttal discussion. The authors describe that "numerous follicles mature at the same time....". That is not disputed. My point was that each mature follicle will receive the identical endocrine ovulatory signal (correct? Or do the authors believe something different?). If it were sufficient for one follicle to ovulate, then all of those mature follicles (the number of which will be variable between animals and between cycles) would be expected to undergo ovulation. The fact that they do not raise several possibilities. One that the authors favor is that an insufficient ovulatory signal might approach a threshold where some follicles ovulate and others do not. This possibility is supported by the apparent increase in cystic follicles, which might be preovulatory follicles that did not complete the ovulation process. Such variation might be stochastic, within normal variation for sensitivity to LH. However, it is also possible that the follicles have not matured at the same rate, perhaps influenced by abnormal secretion of LH or FSH during earlier phases of the cycle, and hence are not in the appropriate condition to respond to the ovulation signal when it arrives. Some may even have matured prematurely due to the elevated gonadotropins reported in this study. Given the data and the partial fertility, the most likely explanation is that the genetic manipulation has resulted in fewer follicles being available for ovulation due to changes in follicular development rather than a deficit of the ovulation signal, although the latter mechanism might also contribute. A third possibility is that genetic manipulation has directly affected the ovary. The authors did not answer whether Kiss1 and MC4 are co-expressed in the ovary. I think the authors might want to rule this out by showing no change in MC4R expression in the ovary.

We thank the reviewer for this thoughtful comment and agree that these are possible outcomes. We have now acknowledged them in the Discussion.

To answer the reviewer’s question, we have not investigated the co-expression of Kiss1 and Mc4r in the ovary. While MC4R has indeed been documented in the ovary (Chen et al. Reproduction, 2017), the changes in gonadotropin release and supporting in vitro data included in this manuscript clearly document a central effect, however, an additional effect at the level of the ovary cannot be completely ruled out. This has now been added to the discussion (Line 378-387).

(7) Lines 390, 454 " impaired LH pulse" What was the evidence for impaired LH pulse (see figure 2D)?

Thank you for pointing this out. This comment referred to augmented LH release. This has been corrected in the revised manuscript (Line 394).

The paper's strengths remain, as outlined in my original review. The authors have addressed what I perceived to be weaknesses, predominantly by changing the tone of discussion and interpretation of the data. This is appropriate. I consider the focus on the LH surge as the primary mechanism too narrow, and the authors should be considering how other changes during the cycle might influence ovarian function.

We sincerely appreciate the reviewer’s thoughtful evaluation of our manuscript and their constructive feedback. We are pleased that our revisions have addressed the perceived weaknesses and that the adjustments to the discussion and interpretation were deemed appropriate.

We acknowledge the reviewer’s perspective on broadening the discussion beyond the LH surge to consider additional cycle-dependent influences on ovarian function. While our current study focuses on this specific mechanism, we recognize that ovarian function is influenced by multiple physiological changes throughout the cycle. We have refined our discussion to reflect this broader context and appreciate the suggestion to consider these additional factors in future studies.

We have addressed all of the reviewer’s comments to the best of our ability and hope they find the revised manuscript satisfactory.

eLife Assessment

Combining experiments in microfluidic devices and computer simulation, this study provides a valuable analysis of the relevant parameters that determine the motility of (multicellular) magnetotactic bacteria in sediment-like environments. The study presents convincing evidence that there is an optimum in the biological parameters for motile life under such conditions.

Reviewer #1 (Public review):

Summary:

The authors track the motion of multiple consortia of Multicellular Magnetotactic Bacteria moving through an artificial network of pores and report a discovery of a simple strategy for such consortia to move fast through the network: an optimum drift speed is attained for consortia that swim a distance comparable to the pore size in the time it takes to align the with an external magnetic field. The authors rationalize their observations using dimensional analysis and numerical simulations. Finally, they argue that the proposed strategy could generalize to other species by demonstrating the positive correlation between the swimming speed and alignment time based on theoretical analysis and parameters derived from literature.

Strengths:

The underlying dimensional analysis and model convincingly rationalize the experimental observation of an optimal drift velocity: the optimum balances the competition between the trapping in pores at large magnetic fields and random pore exploration for weak magnetic fields.

Weaknesses:

The convex pore geometry studied here creates convex traps for cells, which I expect enhances their trapping. Natural environments may create a much smaller concentration of such traps. In this case, whether a non-monotonic dependence of the drift velocity on the Scattering number would persist is unclear.

Comments on revisions:

Thank you very much for addressing my comments. I think the revisions have improved the paper.

Reviewer #2 (Public review):

Summary:

The authors have made microfluidic arrays of pores and obstacles with a complex shape and studied the swimming of multicellular magnetotactic bacteria through this system. They provide a comprehensive discussion of the relevant parameters of this system and identify one dimensionless parameter, which they call the scattering number and which depends on the swimming speed and magnetic moment of the bacteria as well as the magnetic field and the size of the pores, as the most relevant. They measure the effective speed through the array of pores and obstacles as a function of that parameter, both in their microfluidic experiments and in simulations, with good agreement between the two. They find an optimal scattering number, which they estimate to reflect the parameters of the studied multicellular bacteria in their natural environment. They finally use this knowledge to compare different species. Despite the variability of bacteria parameters, they estimate the scattering number to be rather narrowly distributed, suggesting that their results apply to a broad range of species.

Strengths:

This is a beautiful experimental approach and the observation of an optimal scattering number (likely reflecting an optimal magnetic moment) is very convincing. The results here improve on similar previous work in two respects: On the one hand, the tracking of bacteria does not have the limitations of previous work, and on the other hand, the effective motility is quantified. Both features are enabled by choices of the experimental system: the use the multicellular bacteria which are larger than the usual single-celled magnetotactic bacteria and the design of the obstacle array which allows the quantification of transition rates due to the regular organization as well as the controlled release of bacteria into this array through a clever mechanism.

Weaknesses:

Some of the key experimental choices on which the strength of the approach is based also come at a price and impose some limitations, namely the use of a non-culturable organism and the regular, somewhat unrealistic artificial obstacle array, but the advantages of these choices outweigh the drawbacks.

Comments on revisions:

The paper has been improved with respect to presentation and content. In particular, I appreciate the new plots comparing the simulation and experiments directly and the estimate of the scattering number for different species. In my opinion, all issues raised by the reviewers have been addressed in a productive way.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

The authors track the motion of multiple consortia of Multicellular Magnetotactic Bacteria moving through an artificial network of pores and report a discovery of a simple strategy for such consortia to move fast through the network: an optimum drift speed is attained for consortia that swim a distance comparable to the pore size in the time it takes to align the with an external magnetic field. The authors rationalize their observations using dimensional analysis and numerical simulations. Finally, they argue that the proposed strategy could generalize to other species by demonstrating the positive correlation between the swimming speed and alignment time based on parameters derived from literature.

Strengths:

The underlying dimensional analysis and model convincingly rationalize the experimental observation of an optimal drift velocity: the optimum balances the competition between the trapping in pores at large magnetic fields and random pore exploration for weak magnetic fields.

Weaknesses:

The convex pore geometry studied here creates convex traps for cells, which I expect enhances their trapping. The more natural concave geometries, resulting from random packing of spheres, would create no such traps. In this case, whether a non-monotonic dependence of the drift velocity on the Scattering number would persist is unclear.

We agree that convex walls increase the time that consortia remain trapped in pores at high magnetic fields. Since the non-monotonic behavior of the drift velocity with the Scattering number arises largely due to these long trapping times, we agree that experiments using concave pores are likely to show a peak drift velocity that is diminished or erased.

However, we disagree that a random packing of spheres or similar particles provides an appropriate model for natural sediment, which is not composed exclusively of hard particles in a pure fluid. Pore geometry is also influenced by clogging. Biofilms growing within a network of convex pillars in two-dimensional microfluidic devices have been observed to connect neighboring pillars, thereby forming convex pores. Similar pore structures appear in simulations of biofilm growth between spherical particles in three dimensions. Moreover, the salt marsh sediment in which MMB live is more complex than simple sand grains, as cohesive organic particles are abundant. Experiments in microfluidic channels show that cohesive particles clog narrow passageways and form pores similar to those analyzed here. Thus, we expect convex pores to be present and even common in natural sediment where clogging plays a role.

The concentration of convex pores in the experiments presented here is almost certainly much higher than in nature. Nonetheless, since magnetotactic bacteria continuously swim through the pore space, they are likely to regularly encounter such convexities. Efficient navigation of the pore space thus requires that magnetotactic bacteria be able to escape these traps. In the original version of this manuscript, this reasoning was reduced to only one or two sentences. That was a mistake, and we thank the reviewer for prompting us to expand on this point. As the reviewer notes, this reasoning is central to the analysis and should have been featured more prominently. In the final version, we will devote considerable space to this hypothesis and provide references to support the claims made above.

The reviewer suggests that the generality of this work depends on our finding a ”positive correlation between the swimming speed and alignment [rate] based on parameters derived from literature.” We wish to emphasize that, in addition to predicting this correlation, our theory also predicts the function that describes it. The black line in Figure 3 is not fitted to the parameters found in the literature review; it is a pure prediction.

Reviewer #2 (Public review):

The authors have made microfluidic arrays of pores and obstacles with a complex shape and studied the swimming of multicellular magnetotactic bacteria through this system. They provide a comprehensive discussion of the relevant parameters of this system and identify one dimensionless parameter, which they call the scattering number and which depends on the swimming speed and magnetic moment of the bacteria as well as the magnetic field and the size of the pores, as the most relevant. They measure the effective speed through the array of pores and obstacles as a function of that parameter, both in their microfluidic experiments and in simulations, and find an optimal scattering number, which they estimate to reflect the parameters of the studied multicellular bacteria in their natural environment. They finally use this knowledge to compare different species to test the generality of this idea.

Strengths:

This is a beautiful experimental approach and the observation of an optimal scattering number (likely reflecting an optimal magnetic moment) is very convincing. The results here improve on similar previous work in two respects: On the one hand, the tracking of bacteria does not have the limitations of previous work, and on the other hand, the effective motility is quantified. Both features are enabled by choices of the experimental system: the use the multicellular bacteria which are larger than the usual single-celled magnetotactic bacteria and the design of the obstacle array which allows the quantification of transition rates due to the regular organization as well as the controlled release of bacteria into this array through a clever mechanism.

Weaknesses:

Some of the reported results are not as new as the authors suggest, specifically trapping by obstacles and the detrimental effect of a strong magnetic field have been reported before as has the hypothesis that the magnetic moment may be optimized for swimming in a sediment environment where there is a competition of directed swimming and trapping. Other than that, some of the key experimental choices on which the strength of the approach is based also come at a price and impose some limitations, namely the use of a non-culturable organism and the regular, somewhat unrealistic artificial obstacle array.

In the “Recommendations for the Authors,” this reviewer drew our attention to a manuscript that absolutely should have been prominently cited. As the reviewer notes, our manuscript meaningfully expands upon this work. We are pleased to learn that the phenomena discussed here are more general than we initially understood. It was an oversight not to have found this paper earlier. The final version will better contextualize our work and give due credit to the authors. We sincerely appreciate the reviewer for bringing this work to our attention.

We disagree that the use of non-culturable organisms and our unrealistic array should be considered serious weaknesses. While any methodological choice comes with trade-offs, we believe these choices best advance our aims. First, the goal of our research, both within and beyond this manuscript, is to understand the phenotypes of magnetotactic bacteria in nature. While using pure cultures enables many useful techniques, phenotypic traits may drift as strains undergo domestication. We therefore prioritize studying environmental enrichments.

Clearly, an array of obstacles does not fully represent natural heterogeneity. However, using regular pore shapes allows us to average over enough consortium-wall collisions to enable a parameter-free comparison between theory and experiment. Conducting an analysis like this with randomly arranged obstacles would require averaging over an ensemble of random environments, which is practically challenging given the experimental constraints. Since we find good agreement between theory and experiment in simple geometries, we are now in a position to justify extending our theory to more realistic geometries. Additionally, we note that a microfluidic device composed of a random arrangement of obstacles would also be a poor representation of environmental heterogeneity, as pore shape and network topology differ between two and three dimensions.

Recommendations for the Authors: 

Reviewer #1 (Recommendations for the authors):

My main suggestion is for the authors to describe the limitations of their approach in the case of concave pores.

As we noted in our public comments, this was a very useful comment to hear from you and one that has been repeated as we have spoken about these results to colleagues. Convexities here represent an experimentally simple way to force bacteria to back track through the maze, as they must through natural sediment. We have greatly expanded this discussion to clarify this reasoning (lines 84–105). We provide references to three types of physical processes that may lead to such traps. First, as in figure 1 of Kurz et al, biofilm (white) can fill the spaces between convex pillars to create covexities. Additionally, clogging by cohesive particles can make narrow passageways between convex particles impassible. An example of clogging is shown in figure 6 of Dressaire & Sauret 2017. Finally, air bubbles trapped in the sediment can create pore-scale dead ends that require bacteria to backtrack. The full references are provided in the main text.

Small points:

(1) How many trajectories were used to produce Figures 2 b and c?

We have modified the caption to note that these data represent the measured transition rates of a total 938 consortia at various Scattering numbers. Each consortium may pass between pores many times.

(2) Can the authors describe in more detail how Equation (3) is derived? Why doesn’t it depend on the gap size between the pores?

We have provided a derivation of this equation in Appendix 2 of the new version. This derivation shows that the drift velocity U_drift is proportional to the pore diameter and difference between the transition rates.

The proportionality constant α depends on how the pores are connected together in space. In the original version, we wanted to highlight the role of the asymmetry of the transition rates, so we imagined a one dimensional network of pores without gaps. In this case, α \= 1. This reasoning was poorly explained in the previous version and we thank the reviewer for pointing this deficiency out. In the new version, we include the gap size and use the layout of pores in a square lattice with gaps, which is shown in figure 1. The proportionality constant for a square lattice in the absence of gaps√ would be 1/2. The limitations of photolithography require some gap that increase the proportionality constant to α \= 0.8344.

We have updated the text, equation (3), and the figures to account for the finite gap sizes.

(3) I found the second part of the abstract, related to the comparison between diverse bacteria, to be slightly misleading. Upon first reading, my expectation was that the authors carried out experiments with different species.

We have modified the abstract to make clear that we rely on values taken from a literature review.

(4) More information is needed on how many trajectories were used to produce the probability densities in Figures 1b-d. How were the densities computed?

The probability distributions give the probability that a pixel in a pore is covered by a consortium. They reflect between 1.2 and 7 million measurements (depending on the panel) of the instantaneous positions of consortia. We have added a section (Lines 453–469) to Materials and Methods that describes exactly how these distributions were calculated.

Reviewer #2 (Recommendations for the authors):

(1) As mentioned under Weaknesses in the Public review, some results are less new than claimed here. The existence of an optimal magnetic moment has been shown by Codutti et al eLife eLife13:RP98001 in very similar experiments, where it was also proposed that this may be an evolutionary adaptation to the sediment habitat. The paper here provides additional evidence for this, and with better tracking and quantification, but previous work should be discussed. Likewise, the work by Dekharghani et al. that is mentioned rather suddenly in the Results section appears to be a crucial previous state of the art and could already be mentioned in the introduction.

We thank the reviewer for bringing this paper, which came out as we were writing this manuscript, to our attention. The hypothesis that there is an optimal phenotype that balances magnetotaxis with obstacle avoidance—and that natural selection could guide organisms to this optimum—goes back to at least 2022. It seems that Codutti et al independently came up with this same hypothesis and provided the first test.

We have substantively rewritten the introduction (Lines 46–58) to better contextualize our work and give due attention to Dekharghani et al.

(2) The first paragraph of Results also contains background information and could be moved into the introduction.

As part of the rewrite to better contextualize our work, we moved the first two paragraphs of results to the introduction.

(3) I found Figure 1 a bit confusing and it took me some time to understand the geometry. I think the black obstacles are very dominant to the viewer’s eye and draw attention away from the essentially circular shape of the pores. Likewise, I am not sure that cutting the neighboring pores off in a circular fashion in Figures 1b-d was the best choice. The authors should think about whether the presentation can be improved. Likewise, when describing the direction of the field in the text, I would suggest adding that it is along the horizontal direction in Figure 1.

We have modified the figure and the text as the reviewer suggests.

(4) That collisions with a pore wall are an important mechanism of changing direction is clear and it is nice to see the paper demonstrate that this mechanism is dominant over rotational diffusion. However, this may not be universal, as (i) rotational diffusion is more important for smaller cells and (ii) interaction with walls can result in all kinds of different behaviors than complete randomization (e.g. swimming along the walls as shown in microfluidic chambers, Ostapenko et al. Phys Rev Lett 2018, Codutti et al. eLife 2022, or reversals, Kuhn et al PNAS 2017). Here, it appears that complete randomization of the direction is an assumption, but this could be tested/quantified by analyzing the trajectories.

This is an excellent point. We have modified the text to describe qualitatively how these tendencies would shift the Critical Scattering number. We also note in the text that there is evidence of these differences in Fig 3. The Desulfobacterota are shifted upwards in Fig 3 relative to the α-proteobacteria. This shift indicates that Desulfobacterota tend to live at slightly greater scattering numbers of 0.9±0.3 than the α-proteobacteria, which live at scattering number 0.37 ± 0.03. It is likely that this difference reflects taxonomic differences in rotational diffusion and cell-wall interactions.

It is true that total randomization of the direction is indeed an assumption, and it is stated as such in line 189. We performed all of the numerics to find the solid curves in Fig 2 before we got any experimental data and so, at the time, total randomization seemed like a fair choice. Looking at Fig 2b, it is clear that these numerics systematically overestimate k₋. We believe that this error is do to the assumption of total randomization.

As this effect is small and does not change any of the conclusions of the paper and Codutti et al were able to publish their paper in the time that we were writing ours, we feel some urgency to move forward.

(5) From the manuscript it is not fully clear to what extent experiments and simulations are or can be quantitatively compared. For example: is the curve (“fit”) in Figure 2c based on the simulations? Is there an explicit expression or is this just a spline or something like that? Why does Figure 5 (simulation) show the velocity as a function of Sc⁻¹and Figure 2 (experiment) as a function of Sc? It looks to me as if a quantitative comparison could be achieved.

The original version of Figure 2 shows a quantitative comparison between theory and experiment with no fit parameters. The data points are the result of experiments in which consortia are tracked as they as they move between connected pores. The solid line is a found by interpolating a smooth curve through the data from simulations. As we make clear in the new version (Lines 537–551), this blue curve is the most probable smooth curve that explains the simulations.

We have added the simulations to figure 2 so that a single panel includes the data, the simulations, and the smooth curve. To further make clear that this comparison is quantitative and parameter free, we have added a panel to Figure 2. This panel directly compares the prediction to observation and is independent of the blue curve.

As was noted (deep within the methods section) in the original version, our numerics can exactly simulate Sc = ∞. Consequently, it was reasonable to simulate parameters that are uniformly spaced in Sc⁻¹.

(6) While I like the idea behind Figure 3, the data shown here is not as convincing as suggested. If one looks at the data without the black line, I think one gets a weaker dependence. The correlation between U₀ and γ_geo is likely not as strong as it seems. Calculating a correlation coefficient might be helpful here. In any case, the assumptions going into this figure should be discussed more explicitly and the results should in my opinion be phrased more cautiously (I tend to believe what the authors claim, but I don’t think the evidence for this point is very strong).

We appreciate the reviewer’s skepticism. However, we believe that the data are stronger than one might understand from the previous text. We have rewritten the text (Lines 219–291) and included new analysis, figures, and explanation to make three points clear.

(a) It is surprising that speed, magnetic moment, and mobility all vary tremendously(between one and three orders of magnitude) across taxa and environment, however, their dimensionless combination Sc is narrowly distributed. We have added a panel to Fig. 3 to show the measured Scattering numbers.

It is notable that there are no adjusted parameters in the calculation of the Scattering numbers: it is a simple dimensionless combination of phenotypic and environmental parameters. All but one of these parameters (the pore size) is measured either by us or by other authors. The pore radius is likely narrowly distributed. We measure it at our field site and, when it is not reported, we use a value typical of the geological and fluvial environment. Just as the size of sand grains does not vary greatly between the beaches of Australia, Africa, and California, it is a good assumption that the pore spaces that host these magnetotactic bacteria do not vary tremendously in size.

(b) In the new version we compare the Scattering number statistics to a parameterfree null model of phenotypic diversity. We argue in the text that it is appropriate to bootstrap over the phenotypic diversity of species. This null model provides the correct method to calculate p-values as the variability in the Scattering numbers is neither identically distributed nor normally distributed.

We use this null model to show that—given the measured phenotypic diversity across species—the probability that fifteen random species would fall within the measured range of Scattering numbers that is consistent with optimal navigation is ∼ 10⁻⁶. This result is strong evidence that the phenotypic variables exhibit the correlations that are predicted by our analysis.

(c) The correlation between U₀/r and γ_geo is reasonably strong. I think that our choice of axes in Fig 3, which were chosen to fit the legend, make the data look flatter than then they actually are. Here are the same data plotted without the line with tighter axes:

Author response image 1.

With the exception of the very first point and the very last point, the data appear to our eyes to be pretty correlated. This impression is born out by a calculation of the correlation coefficient which gives 0.77. The p-value is 4 × 10⁻⁴. We have included these values in the main text to clarify that this correlation is both statistically significant and of primary importance.

(7) There is a comment at the end of the discussion that the evolutionary hypothesis could be tested by transferring the magnetotaxis genes to nonmagnetotactic organisms. This would indeed be highly desirable, but this is very difficult as indicated by the successful efforts in that direction (which often are only moderately magnetic/magnetotactic), see Kolinko et al Nature Nanotech 2014, Dziuba et al Nature Nanotech 2024.

Thank you for highlighting these references, which we have included. We agree that these experiments will be challenging. Our results make a prediction about the evolution of these strains, so it seems worth mentioning this fact. We feel that this manuscript is not the correct space for a detailed description of challenges that we will encounter should we pursue this direction of study.

(8) A section on how the bacterial samples were obtained could be added in Methods.

We have done so.

Additional Changes

(1) In the original version, we feared that the consortia in the microfluidic device arepoorly representative of the natural population. Consequently, we used the values from previous experiments, which we performed using consortia taken from the same pond. Since submitting this manuscript we have undertaken new experiments that allowed us to measure the Scattering number of individual consortia. It turns out the effect is smaller than we worried. We have included these measurements in the new version. We find that even as the most common phenotypes vary over the course of time, the Scattering number remains constant. This result is additional evidence that there is strong selective pressure to optimally navigate.

As a result of these additions, we have added an author, Julia Hernandez, who contributed to these experiments and analysis.

(2) We have expanded the table of phenotypic variable in Appendix 1 to make it easier forother researchers to reproduce our analysis.

eLife Assessment

This manuscript reports findings of fundamental significance on how bacteria might load helicase for DNA replication when normal DnaA-based loading pathway is defective. It provides convincing genetic and biochemical evidence that helicase loading at the E. coli oriC is not (as previously assumed) exclusively performed by the DnaA initiator protein but can also be executed by PriC (whether this occurs specifically at oriC has not been addressed in vivo). This is a significant step forward in our understanding of bacterial replication initiation.

Reviewer #1 (Public review):

Summary:

This manuscript reports the investigation of PriC activity during DNA replication initiation in Escherichia coli. It is reported that PriC is necessary for growth and control of DNA replication initiation under diverse conditions where helicase loading is perturbed at the chromosome origin oriC. A model is proposed where PriC loads helicase onto ssDNA at the open complex formed by DnaA at oriC. Reconstituted helicase loading assays in vitro are consistent with the model.

Strengths:

The complementary combination of genetics in vivo and reconstituted assays in vitro provide solid evidence to support the role of PriC at a replication origin.

The manuscript is well written and has a logical narrative.

The data provide new insight to how bacteria might load helicase at the replication origin when the wild-type DnaA-dependent loading pathway is perturbed.

Weakness:

It has not yet been established whether PriC localises at oriC in vivo under the conditions tested.

Reviewer #2 (Public review):

This is a great paper. Yoshida et al. convincingly show that DnaA does not exclusively do loading of the replicative helicase at the E. coli oriC, but that PriC can also perform this function. Importantly, PriC seems to contribute to helicase loading even in wt cells albeit to a much lesser degree than DnaA. On the other hand, PriC takes a larger role in helicase loading during aberrant initiation, i.e. when the origin sequence is truncated or when the properties of initiation proteins is suboptimal. Here highlighted by mutations in dnaA or dnaC.

This a major finding because it clearly demonstrates that the two roles of DnaA in the initiation process can be separated into initially forming an open complex at the DUE region by binding/nucleation onto DnaA-boxes and second in loading of the helicase. Whereas these two functions are normally assumed to be coupled, the present data clearly show that they can be separated and that PriC can perform at least part of the helicase loading provided that an area of duplex opening is formed by DnaA.<br /> This puts into questions the interpretation of a large body of previous work on mutagenesis of oriC and dnaA to find a minimal oriC/DnaA complex in many bacteria. In other words, mutants in which oriC is truncated/mutated may support initiation of replication and cell viability only in the presence of PriC. Such mutants are capable to generate single strand opening but may fail to load the helicase in absence of PriC. Similarly, dnaA mutants may generate aberrant complex on oriC that trigger strand opening but are incapable of loading DnaB unless PriC is present.

In the present work, the sequence of experiments presented is logical and the manuscript is clearly written and easy to follow. The very last part regarding PriC in cSDR replication does not add much to the story and may be omitted.

I have a few specific questions/comments

The partial complementation of the dnaC2 strain by PriC seems quite straightforward since this particular mutation leads to initiation arrest at the open complex stage and this sets the stage for PriC to load the helicase. The situation is somewhat different for dnaA46. Why is this mutation partly complemented by PriC at 37C? DnaA46 binds neither ATP nor ADP, yet it functions in initiation at permissive temperature. At nonpermssive temperature, it binds oriC as well but does not lead to initiation. Does the present data imply that the true initiation defect of DnaA46 lies in helicase loading? The authors need to comment on this in the text.

Relating to the above. In Fig. 3 it is shown that the pFH plasmid partly complement dnaA46 in a PriC dependent manner. Again, it would be nice to know the nature of the DnaA46 protein defect. It would be interesting to see how a pING1-dnaA46 plasmid performs in the experiment presented in Fig. 3.

Reviewer #3 (Public review):

Summary:

At the abandoned replication fork, loading of DnaB helicase requires assistance from PriABC, repA, and other protein partners, but it does not require replication initiator protein, DnaA. In contrast, nucleotide-dependent DnaA binding at the specific functional elements is fundamental for helicase loading, leading to the DUE region's opening. However, the authors questioned in this study that in case of impeding replication at the bacterial chromosomal origins, oriC, a strategy similar to an abandoned replication fork for loading DnaB via bypassing the DnaA interaction step could be functional. The study by Yoshida et al. suggests that PriC could promote DnaB helicase loading on the chromosomal oriC ssDNA without interacting with the DnaA protein. The conclusions drawn supported by the evidence provided are compelling.

Strengths:

Understanding the mechanism of how DNA replication restarts via reloading the replisomes onto abandoned DNA replication forks is crucial. Notably, this knowledge becomes crucial to understanding how bacterial cells maintain DNA replication from a stalled replication fork when challenging or non-permissive conditions prevail. This critical study combines experiments to address a fundamental question of how DnaB helicase loading could occur when replication initiation impedes at the chromosomal origin, leading to replication restart.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This manuscript reports the investigation of PriC activity during DNA replication initiation in Escherichia coli. It is reported that PriC is necessary for the growth and control of DNA replication initiation under diverse conditions where helicase loading is perturbed at the chromosome origin oriC. A model is proposed where PriC loads helicase onto ssDNA at the open complex formed by DnaA at oriC. Reconstituted helicase loading assays in vitro support the model. The manuscript is well-written and has a logical narrative.

Thank you for understanding this study.

Major Questions/Comments:

An important observation here is that a ΔpriC mutant alone displays under-replication, suggesting that this helicase loading pathway is physiologically relevant. Has this PriC phenotype been reported previously? If not, would it be possible to confirm this result using an independent experimental approach (e.g. marker frequency analysis or fluorescent reporter-operator systems)?

We thank Reviewer 1 for this comment. This study provides the first direct evidence for PriC’s role in initiation of chromosome replication. Given the change of the oriC copy number of ∆priC cells in non-stressed conditions is only slight, resolution of the suggested methods is clearly not high enough to distinguish the differences in the oriC copy number between priC⁺ (WT) and ∆priC cells. Thus, to corroborate the ∆priC phenotype, we additionally analyzed using flow cytometry priC⁺ and ∆priC cells growing under various nutrition and thermal conditions.

As shown in Figure 2-figure supplement 1 of the revised version, the fraction of cells with non-2ⁿ oriC copies was slightly higher in ∆priC cells compared to priC⁺ cells. Furthermore, when grown in M9 minimal medium at 37˚C, ∆priC mutant cells exhibited slightly reduced ori/mass values. These are supportive to the idea that inhibition of replication initiation occurs at low frequency even in the WT dnaA and dnaC background, and that PriC function is necessary to ensure normal replication initiation. Related descriptions have been revised accordingly.

Is PriA necessary for the observed PriC activity at oriC? Is there evidence that PriC functions independently of PriA in vivo?

As described in Introduction of the original manuscript, PriA is a 3’-to-5’ helicase which specifically binds to the forked DNA with the 3’-end of the nascent DNA strand. Thus, structural specificity of target DNA is essentially different between PriA and PriC. Consistent with this, our in vitro data indicate that PriC alone is sufficient to rescue the abortive helicase loading at oriC (Figure 7), indicating that PriA is principally unnecessary for PriC activity at oriC. Consistently, as described in Introduction, PriC can interact with ssDNA to reload DnaB (Figure 1E). Nevertheless, a possibility that PriA might participate in the PriC-dependent DnaB loading rescue at oriC in vivo can not be completely excluded. However, elucidation of this possibility is clearly beyond the scope of the present study and should be analyzed in the future. An additional explanation has been included in Discussion of the revised version.

Is PriC helicase loading activity in vivo at the origin direct (the genetic analysis leaves other possibilities tenable)? Could PriC enrichment at oriC be detected using chromatin immunoprecipitation?

These are advanced questions about genomic dynamics of PriC. Given that PriC facilitates DnaB reloading at stalled replication forks (Figure 1E) (Heller and Marians, Mol Cell., 2005; Wessel et al., J Biol Chem, 2013; Wessel et al., J Biol Chem, 2016), PriC might interact with the whole genome and its localization might not necessarily exhibit a preference for oriC in growing cells. Analysis about these advanced questions is interesting but is beyond the scope of the present study and should be analyzed in the future study.

Reviewer #2 (Public review):

This is a great paper. Yoshida et al. convincingly show that DnaA does not exclusively do loading of the replicative helicase at the E. coli oriC, but that PriC can also perform this function. Importantly, PriC seems to contribute to helicase loading even in wt cells albeit to a much lesser degree than DnaA. On the other hand, PriC takes a larger role in helicase loading during aberrant initiation, i.e. when the origin sequence is truncated or when the properties of initiation proteins are suboptimal. Here highlighted by mutations in dnaA or dnaC.

This is a major finding because it clearly demonstrates that the two roles of DnaA in the initiation process can be separated into initially forming an open complex at the DUE region by binding/nucleation onto DnaA-boxes and second by loading of the helicase. Whereas these two functions are normally assumed to be coupled, the present data clearly show that they can be separated and that PriC can perform at least part of the helicase loading provided that an area of duplex opening is formed by DnaA. This puts into question the interpretation of a large body of previous work on mutagenesis of oriC and dnaA to find a minimal oriC/DnaA complex in many bacteria. In other words, mutants in which oriC is truncated/mutated may support the initiation of replication and cell viability only in the presence of PriC. Such mutants are capable of generating single-strand openings but may fail to load the helicase in the absence of PriC. Similarly, dnaA mutants may generate an aberrant complex on oriC that trigger strand opening but are incapable of loading DnaB unless PriC is present.

We would like to thank Revierwer#2 for the very positive comments about our work.

In the present work, the sequence of experiments presented is logical and the manuscript is clearly written and easy to follow. The very last part regarding PriC in cSDR replication does not add much to the story and may be omitted.

Given that the role PriC in stimulating cSDR was unclear, we believe that our finding that PriC has little or no role in cSDR, despite being a negative result, is valuable for the general readership of eLife. To further assess impact of PriC on cSDR and as recommended by Referee #1, we carried out the chromosome loci copy-number analysis by the whole-genome sequencing. As shown in Figure 8-supplement 1 of the revised version, the results support our conclusion from the original version.

Reviewer #3 (Public review):

Summary:

At the abandoned replication fork, loading of DnaB helicase requires assistance from PriABC, repA, and other protein partners, but it does not require replication initiator protein, DnaA. In contrast, nucleotide-dependent DnaA binding at the specific functional elements is fundamental for helicase loading, leading to the DUE region's opening. However, the authors questioned in this study that in case of impeding replication at the bacterial chromosomal origins, oriC, a strategy similar to an abandoned replication fork for loading DnaB via bypassing the DnaA interaction step could be functional. The study by Yoshida et al. suggests that PriC could promote DnaB helicase loading on the chromosomal oriC ssDNA without interacting with the DnaA protein. However, the conclusions drawn from the primarily qualitative data presented in the study could be slightly overwhelming and need supportive evidence.

Thank you for your understanding and careful comments.

Strengths:

Understanding the mechanism of how DNA replication restarts via reloading the replisomes onto abandoned DNA replication forks is crucial. Notably, this knowledge becomes crucial to understanding how bacterial cells maintain DNA replication from a stalled replication fork when challenging or non-permissive conditions prevail. This critical study combines experiments to address a fundamental question of how DnaB helicase loading could occur when replication initiation impedes at the chromosomal origin, leading to replication restart.

Thank you for your understanding.

Weaknesses:

The term colony formation used for a spotting assay could be misleading for apparent reasons. Both assess cell viability and growth; while colony formation is quantitative, spotting is qualitative. Particularly in this study, where differences appear minor but draw significant conclusions, the colony formation assays representing growth versus moderate or severe inhibition are a more precise measure of viability.

We used serial dilutions of the cell culture for the spotting assay and thus this assay should be referred as semi-quantitative rather than simply qualitative. For more quantitative assessment of viability, we analyzed the growth rates of cells and the chromosome replication activity using flow cytometry.

Figure 2

The reduced number of two oriC copies per cell in the dnaA46priC-deficient strain was considered moderate inhibition. When combined with the data suggested by the dnaAC2priC-deficient strain containing two origins in cells with or without PriC (indicating no inhibition)-the conclusion was drawn that PriC rescue blocked replication via assisting DnaC-dependent DnaB loading step at oriC ssDNA.

The results provided by Saifi B, Ferat JL. PLoS One. 2012;7(3):e33613 suggests the idea that in an asynchronous DnaA46 ts culture, the rate by which dividing cells start accumulating arrested replication forks might differ (indicated by the two subpopulations, one with single oriC and the other with two oriC). DnaA46 protein has significantly reduced ATP binding at 42C, and growing the strain at 42C for 40-80 minutes before releasing them at 30 C for 5 minutes has the probability that the two subpopulations may have differences in the active ATP-DnaA. The above could be why only 50% of cells contain two oriC. Releasing cells for more time before adding rifampicin and cephalexin could increase the number of cells with two oriCs. In contrast, DnaC2 cells have inactive helicase loader at 42 C but intact DnaA-ATP population (WT-DnaA at 42 or 30 C should not differ in ATP-binding). Once released at 30 C, the reduced but active DnaC population could assist in loading DnaB to DnaA, engaged in normal replication initiation, and thus should appear with two oriC in a PriC-independent manner.

This is a question about dnaA46 Δ_priC_ mutant cells. Inhibition of the replication forks causes inhibition of RIDA (the DNA-clamp complex-dependent DnaA-ATP hydrolysis) system, resulting in the increase of ATP-DnaA molecules (Kurokawa et al. (1999) EMBO J.). Thus, if Δ_priC_ inhibits the replication forks significantly, the ATP-DnaA level should increase and initiation should be stimulated. However, the results of Figure 2BC are opposite, indicating inhibition of initiation by Δ_priC_. Thus, we infer that the inhibition of initiation in the Δ_priC_ cells is not related to possible changes in the ATP-DnaA level. Even if the ATP-DnaA levels are different in subpopulations in dnaA46 cells, Δ_priC_ mutation should not affect the ATP-DnaA levels significantly. Thus, we infer that even in dnaA46 Δ_priC_ mutant cells, Δ_priC_ mutation directly affect initiation mechanisms, rather than indirectly through the ATP-DnaA levels.

Broadly, the evidence provided by the authors may support the primary hypothesis. Still, it could call for an alternative hypothesis: PriC involvement in stabilizing the DnaA-DnaB complex (this possibility could exist here). To prove that the conclusions made from the set of experiments in Figures 2 and 3, which laid the foundations for supporting the primary hypothesis, require insights using on/off rates of DnaB loading onto DnaA and the stability of the complexes in the presence or absence of PriC, I have a few other reasons to consider the latter arguments.

This is a very careful consideration. However, we infer that stabilization of the DnaA-DnaB interaction by PriC, even if present, does not always result in stimulation of DnaB loading to oriC. Given that interactions between DnaA and DnaB during DnaB loading to oriC are highly dynamic and complicated with multiple steps, stabilization of the DnaA-DnaB interaction by PriC, even if it occurs, has a considerable risk of inhibiting the DnaB loading by constructing abortive complexes. In addition, DnaA-DiaA binding is very tight and stable (Keyamura et al., 2007, 2009). Even if WT DnaA and WT DnaB are present, PriC can rescue the initiation defects of oriC mutants. Based on these facts and the known characteristics of PriC as explained in Introduction, it is more reasonable to infer that PriC provides a bypass of DnaB loading even at oriC, as proposed for the mechanism at the stalled replication fork. However, we cannot completely rule out the indicated possibility and these explanations are included in the revised version.

Figure 3

One should consider the fact that dnA46 is present in these cells. Overexpressing pdnaAFH could produce mixed multimers containing subunits of DnaA46 (reduced ATP binding) and DnaAFH (reduced DnaB binding). Both have intact DnaA-DnaA oligomerization ability. The cooperativity between the two functions by a subpopulation of two DnaA variants may compensate for the individual deficiencies, making a population of an active protein, which in the presence of PriC could lead to the promotion of the stable DnaA: DnaBC complexes, able to initiate replication. In the light of results presented in Hayashi et al. and J Biol Chem. 2020 Aug 7;295(32):11131-11143, where mutant DnaBL160A identified was shown to be impaired in DnaA binding but contained an active helicase function and still inhibited for growth; how one could explain the hypothesis presented in this manuscript. If PriC-assisted helicase loading could bypass DnaA interaction, then how growth inhibition in a strain carrying DnaBL160A should be described. However, seeing the results in light of the alternative possibility that PriC assists in stabilizing the DnaA: DnaBC complex is more compatible with the previously published data.

Unfortunately, in this comment, there is a crucial misunderstanding in the growth of cells bearing DnaA L160A. Hayashi et al. reported that the dnaB(Ts) cells bearing the dnaB L160A allele grew slowly and formed colonies even at 42°C. This feature is similar to the growth of dnaA46 cells bearing dnaA F46A H136A allele (Figure 2). Thus, the results of dnaB L160A cells are consistent with our model and support the idea that PriC partially rescues the growth inhibition of cells bearing the DnaB L160A allele by bypassing the strict requirement for the DnaA-DnaB interaction. Nevertheless, we have to be careful about a possibility that DnaB L160A could affect interaction with PriC, which we are going to investigate for a future paper.

As suggested, if mixed complexes of DnaA46 and DnaA F46A H136A proteins are formed, those might retain partial activities in oriC unwinding and DnaB interaction although those cells are inviable at 42°C without PriC. It is noteworthy that in the specific oriC mutants which are impaired in DnaB loading (e.g., Left-oriC), PriC effectively rescues the initiation and cell growth. In these cells, both DnaA and DnaB are intact. Thus, the idea that only mutant DnaA (or DnaB) protein is simulated specifically via PriC interaction is invalid. Even in cells bearing wild-type oriC, DnaA and DnaB, contribution of PriC for initiation is detected.

In addition, as described in the above response, given that interactions between DnaA and DnaB during DnaB loading to oriC are very dynamic and complicated with multiple steps, stabilization of the DnaA-DnaB interaction by PriC, even if present, would not simply result in stimulation of DnaB loading to oriC; rather we think a probability that it would inhibit the DnaB loading by constructing abortive complexes. Based on the known characteristics of PriC as explained in Introduction, it is more reasonable to infer that PriC provides a bypass of DnaB loading even at oriC, as proposed for the mechanism at the stalled replication fork.

However, we cannot completely rule out the indicated possibility and this explanation has been described in the revised version as noted in response to the above question.

Figure 4

Overexpression of DiaA could contribute to removing a higher number of DnaA populations. This could be more aggravated in the absence of PriC (DiaA could titrate out more DnaA)-the complex formed between DnaA: DnaBC is not stable, therefore reduced DUE opening and replication initiation leading to growth inhibition (Fig. 4A ∆priC-pNA135). Figure 7C: Again, in the absence of PriC, the reduced stability of DnaA: DnaBC complex leaves more DnaA to titrate out by DiaA, and thus less Form I*. However, adding PriC stabilizes the DnaA: DnaBC hetero-complexes, with reduced DnaA titration by DiaA, producing additional Form I*. Adding a panel with DnaBL160A that does not interact with DnaA but contains helicase activity could be helpful. Would the inclusion of PriC increase the ability of mutant helicase to produce additional Form I*?

Unfortunately, the proposed idea is biased disregarding the fact that DiaA effectively stimulates assembling processes of DnaA molecules at oriC. As oriC contains multiple DnaA boxes and multiple DnaA molecules are recruited there, DiaA will efficiently facilitate assembling of DnaA molecules on oriC. Even DnaA molecules of DnaA-DiaA complexes can efficiently bind to oriC. This is consistent with in vitro experiments showing that higher levels of DiaA stimulate assembly of DnaA molecules and oriC unwinding (i.e., DUE opening) but even excessive levels of DiaA do not inhibit those reactions (Keyamura et al., J. Biol. Chem. (2009) 284, 25038-25050). However, as shown in Figure 9, DiaA tightly binds to the specific site of DnaA which is the same as the DnaB L160-binding site, which causes inhibition of DnaA-DnaB binding (ibid). These are consistent with in vivo experiments, and concordantly consistent with the idea that the excessive DiaA level inhibits interaction and loading of DnaB by the DnaA-oriC complexes, but not oriC unwinding (i.e., DUE opening) in vivo. Also, as mentioned above, we do not consider that stabilization of DnaA-DnaBC complex simply results in stimulation of DnaB loading to oriC. Based on the known characteristics of PriC, it is more reasonable to infer that PriC provides a bypass of DnaB loading even at oriC, as proposed for the mechanism at the stalled replication fork (Figure 1E), as described in the above response.

As for DnaB L160A, as mentioned above, we are currently investigating interaction modes between DnaB and PriC. While investigating DnaB L160A could further support our model, we believe its contribution to the present manuscript would be incremental. In addition, there is a possibility that DnaA L160A could affect interaction with PriC. Thus, analysis of DnaB mutants in this PriC rescue mechanisms should be addressed in future study.

Figure 5

The interpretation is that colony formation of the Left-oriC ∆priC double mutant was markedly compromised at 37˚C (Figure 5B), and 256 the growth defects of the Left-oriC mutant at 25{degree sign}C and 30{degree sign}C were aggravated. However, prima facia, the relative differences in the growth of cells containing and lacking PriC are similar. Quantitative colony-forming data is required to claim these results. Otherwise, it is slightly confusing.

The indicated concern was raised due to our typing error lacking ∆priC. In the revised manuscript, we have amended as follows: the cell growth of the Left-oriC ∆priC double mutant was markedly compromised at 37˚C and moderately reduced at 25°C and 30°C (Figure 5B).

A minor suggestion is to include cells expressing PriC using plasmid DNA to show that adding PriC should reverse the growth defect of dnaA46 and dnaC2 strains at non-permissive temperatures. The same should be added at other appropriate places.

Even in the presence of PriC, unwinding of oriC and DnaB helicase loading to the wound oriC require DnaA and DnaC activities as indicated by previous studies (see for a review, Windgassen et al., (2018) Nucleic Acids Res. 46, 504-519). Thus, dnaA46 cells and dnaC2 cells bearing pBR322-priC can not grow at 42°C and 37°C (as follows). These are reasonable results. However, at semi-permissive temperatures (37°C for dnaA46 and 35°C for dnaC2), slight stimulation of the cell growth by pBR322-priC might be barely observed (Figure 2-supplement 1 of the revised version). These suggest that the intrinsic level of PriC is functionally nearly sufficient. This explanation has been included in the revised version.

Author response image 1.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Line 38. "in assembly of the replisome".

Corrected.

Line 137. "specifically" rather than specificity.

Corrected.

Line 139. "at" rather than by.

Corrected.

The DnaA46 protein variant contains two amino acid substitutions (A184V and H252Y) within the AAA+ motif. H136 appears to reside adjacent to A184 in structure. Is A184V mutation causative?

The DnaA H136A and A184V alleles are responsible for different defects. Indeed, the DnaA A184V variant is thermolabile and defective in ATP binding whereas the H136A variant retains ATP binding but impairs DnaB loading (Carr and Kaguni, Mol. Microbiol., 1996; Sakiyama et al., Front. Microbiol., 2018). These observations strongly support the view that the phenotype of the DnaA H136A allele is independent of that of the DnaA A184V allele.

Figure 2A. Regarding the dnaA46 allele grown at 37°C.

Individual colonies cannot be resolved. Is an image from a later time-point available?

We have replaced the original image with one from another replicate that provides better resolution. Please see Figure 2A in the revised version.

Figure 2C. Quantification of the number of cells with more than one chromosome equivalent in the dnaC2 ΔpriC strain. The plot from flow cytometry appears to show >20% of cells with only 1 genome. Are these numbers correct?

Thank you for this careful comment. We quantified the peaks more strictly, but the percentages were noy largely changed. To improve resolution of the DNA profiles, we have changed the range of the x-axis in panels B and C of Figure 2 in the revised version.

Figure 3. Are both F46A and H136A mutations in the plasmid-encoded dnaA necessary?

Yes. The related explanation is included in the Discussion section (the third paragraph) of the original manuscript. As described there, dnaA46 cells expressing the DnaA H136A single mutant exhibited severe defects in cell growth even in the presence of PriC (Sakiyama et al., 2018). The His136 residue is located within the weak, secondary DnaB interaction region in DnaA, and is crucial for DnaB loading onto oriC ssDNA. Given domain I in DnaA H136A can stably tether DnaB-DnaC complexes to DnaA complexes on oriC (Sakiyama et al., 2018), we infer that oriC-DnaA complexes including DnaA H136A stably bind DnaB via DnaA domain I as an abortive complex, which inhibits functional interaction between PriC and DnaB as well as DnaB loading to oriC DNA.

As for DnaA F46A mutant, our previous studies show that DnaA F46A has a limited residual activity in vivo (unlike in vitro), and allows slow growth of cells. As the stable DnaA-DnaB binding is partially impaired in vivo in DnaA F46A, this feature is consistent with the above ideas. Thus, both F46A and H136A mutations are required for severer inhibition of DnaB loading. This is additionally described in the revised Discussion.

Figure 3. Is the DnaA variant carrying F46A and H136A substitutions stably expressed in vivo?

We have performed western blotting, demonstrating that the DnaA variant carrying F46A and H136A substitutions is stable in vivo. In the revised version, we have added new data to Figure 3-figure supplement 1 and relevant description to the main text as follows:

Western blotting demonstrated that the expression levels were comparable between WT DnaA and DnaA F46A H136A double mutant (Figure 3-figure supplement 1).

Figure 5A. Should the dashed line extending down from I2 reach the R4Tma construct?

We have amended the indicated line appropriately.

Figure 6C. It was surprising that the strain combining the subATL mutant with ΔpriC displayed a pronounced under-initiation profile by flow cytometry, and yet there was no growth defect observed (see Figure 6B). This seems to contrast with results using the R4Tma origin, where the ΔpriC mutant produced a relatively modest change to the flow cytometry profile, and yet growth was perturbed (Figure 5C-D). How might these observations be interpreted? Is the absolute frequency of DNA replication initiation critical?

Please note that, in E. coli, initiation activity corelates closely with the numbers of oriC copies per cell mass (ori/mass), rather than the apparent DNA profiles measured by flow cytometer. When cells were grown in LB at 30˚C, the mean ori/mass values were as follows: 0.34 for R4Tma priC, 0.51 for R4Tma, 0.82 for DATL priC, 0.99 for DATL (Figures 5 & 6 in the original manuscript). These values closely correspond to the cell growth ability shown in Figure 5C in the original manuscript.

In the revised manuscript, we have cited appropriate references for introduction of the ori/mass values as follows.

To estimate the number of oriC copies per unit cell mass (ori/mass) as a proxy for initiation activity (Sakiyama et al., 2017, 2022),

Line 295. Reference for Form I* assay should cite the original publication.

Done. The following paper is additionally cited.

Baker, T. A., Sekimizu, K., Funnell, B. E., and Kornberg, A. (1986). Extensive unwinding of the plasmid template during staged enzymatic initiation of DNA replication from the origin of the Escherichia coli chromosome. Cell 45, 53–64.doi: 10.1016/0092-8674(86)90537-4

Reviewer #2 (Recommendations for the authors):

The partial complementation of the dnaC2 strain by PriC seems quite straightforward since this particular mutation leads to initiation arrest at the open complex stage and this sets the stage for PriC to load the helicase. The situation is somewhat different for dnaA46. Why is this mutation partly complemented by PriC at 37C? DnaA46 binds neither ATP nor ADP, yet it functions in initiation at permissive temperature. At nonpermissive temperature, it binds oriC as well but does not lead to initiation. Does the present data imply that the true initiation defect of DnaA46 lies in helicase loading? The authors need to comment on this in the text.

Given the thermolabile propensity of the DnaA46 protein, it is presumable that DnaA46 protein becomes partially denatured at the sub-permissive temperature of 37˚C. This partial denaturation should impair both origin unwinding and helicase loading, though not to the extent that cell viability is lost. The priC deletion should further exacerbate helicase loading defects by inhibiting the bypass mechanism, resulting in the lethality of dnaA46 cells at this temperature. This explanation is included in the revised Discussion section.

Relating to the above. In Figure 3 it is shown that the pFH plasmid partly complements dnaA46 in a PriC-dependent manner. Again, it would be nice to know the nature of the DnaA46 protein defect. It would be interesting to see how a pING1-dnaA46 plasmid performs in the experiment presented in Figure 3.

A previous paper showed that multicopy supply of DnaA46 can suppress temperature sensitivity of the dnaA46 cells (Rao and Kuzminov, G3, 2022). This is reasonable in that DnaA46 has a rapid degradation rate unlike wild-type DnaA. As DnaA46 preserves the intact sequences in DnaB binding sites such as G21, F46 and H136, the suppression would not depend on PriC but would be due to the dosage effect.

Figure 8 B: The authors should either remove the data or show a genome coverage: it is not clear that yapB is a good reference. A genome coverage would be nice, and show whether initiation can occur at oriC even if it is not the major place of initiation in a rnhA mutant.

As suggested, we carried out the chromosome loci copy-number analysis by whole-genome sequencing to assess impact of PriC on cSDR. The new data are shown in Figure 8-supplement 1 with relevant descriptions of the main text of the revised version as shown below. Briefly, results of the chromosome loci copy-number analysis are consistent with those of real-time qPCR (Figure 8B). Given that the role PriC in stimulating cSDR was unclear, we believe that our finding that PriC has little or no role in cSDR, despite being a negative result, is valuable for the general readership of eLife.

Line 38-39: .....resulting in replisome assembly.

Corrected.

Line 48: Something is wrong with the Michel reference. Also in the reference list.

Corrected

Line 156: replace retarded with reduced.

Corrected.

Line 171 and elsewhere: WT priC cells is somewhat misleading. Isn't this simply PriC+ cells?

Yes. We have revised the wording to “priC⁺” for clarity.

Line 349-350: "the oriC copy number ratio of the dnaA46 DpriC double mutant was lower than that of the dnaA46 single mutant....". This is only provided growth rate of the strains is the same.

These strains exhibited similar growth rates. This is included in the Result section of the revised manuscript as follows: At the permissive temperature, despite having similar growth rates, the oriC copy number ratio of the dnaA46 ∆priC double mutant strain was lower than that of the dnaA46 single mutant.

Reviewer #3 (Recommendations for the authors):

I would suggest improved or additional experiments, data, or analyses.

The revised version includes improved or additional experiments, data, or analyses.

eLife Assessment

This manuscript presents a fundamental advance in our understanding of nuclear receptor pharmacology by expanding on previous work demonstrating dual ligand occupancy in the peroxisome proliferator-activated receptor-gamma (PPARγ). Using a compelling combination of biophysical, structural, and cellular approaches, the authors show that covalent inhibitors with inverse agonist activities modulate receptor conformation to permit co-binding with additional ligands, leading to a finely tuned transcriptional response. The data support a model of proximal, bi-directional allostery that challenges traditional views of nuclear receptor regulation. These findings will be of broad interest to researchers in structural biology, transcriptional control, and drug discovery.

Reviewer #1 (Public review):

Summary:

This paper focuses on understanding how covalent inhibitors of peroxisome proliferator-activated receptor-gamma (PPARg) show improved inverse agonist activities. This work is important because PPARg plays essential roles in metabolic regulation, insulin sensitization, and adipogenesis. Like other nuclear receptors, PPARg, is a ligand-responsive transcriptional regulator. Its important role, coupled with its ligand-sensitive transcriptional activities, makes it an attractive therapeutic target for diabetes, inflammation, fibrosis, and cancer. Traditional non-covalent ligands like thiazolininediones (TZDs) show clinical benefit in metabolic diseases, but utility is limited by off-target effects and transient receptor engagement. In previous studies, the authors characterized and developed covalent PPARg inhibitors with improved inverse agonist activities. They also showed that these molecules engage unique PPARg ligand binding domain (LBD) conformations whereby the c-terminal helix 12 penetrates into the orthosteric binding pocket to stabilize a repressive state. In the nuclear receptor superclass of proteins, helix 12 is an allosteric switch that governs pharmacologic responses, and this new conformation was highly novel. In this study, the authors did a more thorough analysis of how two covalent inhibitors, SR33065 and SR36708 influence the structural dynamics of PPARg LBD.

Strengths:

(1) The authors employed a compelling integrated biochemical and biophysical approach.

(2) The cobinding studies are unique for the field of nuclear receptor structural biology, and I'm not aware of any similar structural mechanism described for this class of proteins.

(3) Overall, the results support their conclusions.

(4) The results open up exciting possibilities for the development of new ligands that exploit the potential bidirectional relationship between the covalent versus non-covalent ligands studied here.

Weaknesses:

(1) The major weakness in this work is that it is hard to appreciate what these shifting allosteric ensembles actually look like on the protein structure. Additional graphical representations would really help convey the exciting results of this study.

Reviewer #2 (Public review):

Summary:

The authors use ligands (inverse agonists, partial agonists) for PPAR, and coactivators and corepressors, to investigate how ligands and cofactors interact in a complex manner to achieve functional outcomes (repressive vs. activating).

Strengths:<br /> The data (mostly biophysical data) are compelling from well-designed experiments. Figures are clearly illustrated. The conclusions are supported by these compelling data. These results contribute to our fundamental understanding of the complex ligand-cofactor-receptor interactions.

Weaknesses:

This is not the weakness of this particular paper, but the general limitation in using simplified models to study a complex system.

eLife Assessment

This important paper shows that the moment at which rats acquire an appetitive Pavlovian conditioned response is determined by the ratio of the reward rate during the Pavlovian cue to the overall reward rate in the context. The exact quantitative relationship between reward rate during the Pavlovian cue, reward rate in the context and acquisition of conditioned responding is very similar to that observed over 40 years ago in a pigeon auto-shaping procedure. This similarity suggests that the mathematical laws that determine acquisition of conditioned responding in different species may be perfectly general. Claims about the processes underlying learning and conditional behavior in rats are supported by convincing evidence. There is solid evidence for the claim that the same relationship describes data from pigeons by Gibbon and Balsam (1981) and the rats in this study. This study will be of interest to those interested in learning, motivated behavior, and related disease states and brain mechanisms.

Reviewer #1 (Public review):

Summary:

This manuscript by Harris and Gallistel investigates how the rate of learning and strength of conditioned behavior post learning depend on the various temporal parameters of Pavlovian conditioning. They replicate results from Gibbon and Balsam (1981) in rats to show that the rate of learning is proportional to the ratio between the cycle duration and the cue duration. They further show that the strength of conditioned behavior post learning is proportional to the cue duration, and not the above ratio. The overall findings here are interesting, provide context to many conflicting recent results on this topic, and are supported by reasonably strong evidence. Nevertheless, there are some major weaknesses in the evidence presented for some of the stronger claims in the manuscript.

Strengths:

This manuscript has many strengths including a rigorous experimental design, several different approaches to data analysis, careful consideration of prior literature, and a thorough introduction and discussion. The central claim-that animals track the rates of events in their environment, and that the ratio of two rates determine the rate of learning-is supported with solid evidence.

Weaknesses:

Despite the above major strengths, some key aspects of the paper need major improvement. These are listed below.

(1) A key claim made here is that the same relationship (including the same parameter) describes data from pigeons by Gibbon and Balsam (1981) and the rats in this study. I think the evidence for this claim is weak as presented here. First, the exact measure used for identifying trials to criterion makes a big difference in Fig 3. As best as I understand, the authors do not make any claims about which of these approaches is the "best" way. Second, the measure used for identifying trials to criterion in Fig 1 appears different from any of the criteria used in Fig 3. If so, to make the claim that the quantitative relationship is one and the same in both datasets, the authors need to use the same measure of learning rate on both datasets and show that the resultant plots are statistically indistinguishable. Currently, the authors simply plot the dots from the current dataset on the plot in Fig 1 and ask the readers to notice the visual similarity. This is not at all enough to claim that both relationships are the same. In addition to the dependence of the numbers on the exact measure of learning rate used, the plots are in log-log axis. Slight visual changes can mean a big difference in actual numbers. For instance, between Fig 3 B and C, the highest information group moves up only "slightly" on the y-axis but the difference is a factor of 5. The authors need to perform much more rigorous quantification to make the strong claim that the quantitative relationships obtained here and in Gibbon and Balsam 1981 are identical.

(2) Another interesting claim here is that the rates of responding during ITI and the cue are proportional to the corresponding reward rates with the same proportionality constant. This too requires more quantification and conceptual explanation. For quantification, it would be more convincing to calculate the regression slope for the ITI data and the cue data separately and then show that the corresponding slopes are not statistically distinguishable from each other. Conceptually, I am confused why the data used to the test the ITI proportionality come from the last 5 sessions. Specifically, if the model is that animals produce response rates during the ITI (a period with no possible rewards) based on the overall rate of rewards in the context, wouldn't it be better to test this before the cue learning has occurred? Before cue learning, the animals would presumably only have attributed rewards in the context to the context and thus, produce overall response rates in proportion to the contextual reward rate. After cue learning, the animals could technically know that the rate of rewards during ITI is zero. Why wouldn't it be better to test the plotted relationship for ITI before cue learning has occurred? Further, based on Fig 1, it seems that the overall ITI response rate reduces considerably with cue learning. What is the expected ITI response rate prior to learning based on the authors' conceptual model? Why does this rate differ pre and post cue learning? Finally, if the authors' conceptual framework predicts that ITI response rate after cue learning should be proportional to contextual reward rate, why should the cue response rate be proportional to cue reward rate instead of cue reward rate plus contextual reward rate?

(3) I think there was a major conceptual disconnect between the gradual nature of learning shown in Figs 7 and 8 and the information theoretic model proposed by the authors. To the extent that I understand the model, the animals should simply learn the association once the evidence crosses a threshold (nDKL > threshold) and then produce behavior in proportion to the expected reward rate. If so, why should there be a gradual component of learning as shown in these figures? In terms of the proportional response rule to rate of rewards, why is it changing as animals go from 10% to 90% of peak response? I think the manuscript would be much strengthened if these results are explained within the authors' conceptual framework. If these results are not anticipated by the authors' conceptual framework, please do explicitly state this in the manuscript.

(4) I find the idea stated in the Conclusion section that any model considering probability of reinforcement cannot be correct because it doesn't have temporal units to be weak. I think the authors might mean that existing models based on probability do not work and not that no possible model can work. For any point process, the standard mathematical treatment of continuous time is to compute the expected count of events as p*dt where p is the probability of occurrence of the event in that time bin and dt is an infinitesimal time bin. There is obviously a one-to-one mapping between probability of an event in a point process and its rate. Existing models use an arbitrary time bin/trial and thus, I get the authors' argument in the discussion. However, I think their conclusion is overstated.

(5) The discussion states that the mutual information defined in equation 1 does not change during partial reinforcement. I am confused by this. The mean delay between reinforcements increases in inverse proportion to the probability of reinforcement, but doesn't the mean delay between cue and next reinforcement increase by more than this amount (next reinforcement is greater than or equal to the cue-to-cue interval away from the cue for many trials)? Why is this ratio invariant to partial reinforcement?

Comments on revisions:

Update following revision

(1) This point is discussed in more detail in the attached file, but there are some important details regarding the identification of the learned trial that require more clarification. For instance, isn't the original criterion by Gibbon et al. (1977) the first "sequence of three out of four trials in a row with at least one response"? The authors' provided code for the Wilcoxon signed rank test and nDkl thresholds looks for a permanent exceeding of the threshold. So, I am not yet convinced that the approaches used here and in prior papers are directly comparable. Also, there's still no regression line fitted to their data (Fig 3's black line is from Fig 1, according to the legends). Accordingly, I think the claim in the second paragraph of the Discussion that the old data and their data are explained by a model with "essentially the same parameter value" is not yet convincing without actually reporting the parameters of the regression. Related to this, the regression for their data based on my analysis appears to have a slope closer to -0.6, which does not support strict timescale invariance. I think that this point should be discussed as a caveat in the manuscript.

(2) The authors report in the response that the basis for the apparent gradual/multiple step-like increases after initial learning remains unclear within their framework. This would be important to point out in the actual manuscript. Further, the responses indicating the fact that there are some phenomena that are not captured by the current model would be important to state in the manuscript itself.

(3) There are several mismatches between results shown in figures and those produced by the authors' code, or other supplementary files. As one example, rat 3 results in Fig 11 and Supplementary Materials don't match and neither version is reproduced by the authors' code. There are more concerns like this, which are detailed in the attached review file.

Reviewer #2 (Public review):

A long-standing debate in the field of Pavlovian learning relates to the phenomenon of timescale invariance in learning i.e. that the rate at which an animal learns about a Pavlovian CS is driven by the relative rate of reinforcement of the cue (CS) to the background rate of reinforcement. In practice, if a CS is reinforced on every trial, then the rate of acquisition is determined by the relative duration of the CS (T) and the ITI (C = inter-US-interval = duration of CS + ITI), specifically the ratio of C/T. Therefore, the point of acquisition should be the same with a 10s CS and a 90s ITI (T = 10; C = 90 + 10 = 100, C/T = 100/10 = 10) and with a 100s CS and a 900s ITI (T = 100; C = 900 + 100 = 1000, C/T = 1000/100 = 10). That is to say, the rate of acquisition is invariant to the absolute timescale as long as this ratio is the same. This idea has many other consequences, but is also notably different from more popular prediction-error based associative learning models such as the Rescrola-Wagner model. The initial demonstrations that the ratio C/T predicts the point of acquisition across a wide range of parameters (both within and across multiple studies) was conducted in Pigeons using a Pavlovian autoshaping procedure. What has remained under contention is whether or not this relationship holds across species, particularly in the standard appetitive Pavlovian conditioning paradigms used in rodents. The results from rodent studies aimed at testing this have been mixed, and often the debate around the source of these inconsistent results focuses on the different statistical methods used to identify the point of acquisition for the highly variable trial-by-trial responses at the level of individual animals.<br /> The authors successfully replicate same effect found in pigeon autoshaping paradigms decades ago (with almost identical model parameters) in a standard Pavlovian appetitive paradigm in rats. They achieve this through a clever change the experimental design, using a convincingly wide range of parameters across 14 groups of rats, and by a thorough and meticulous analysis of these data. It is also interesting to note that the two author's have published on opposing sides of this debate for many years, and as a result have developed and refined many of the ideas in this manuscript through this process.

Main findings

(1) The present findings demonstrate that the point of initial acquisition of responding is predicted by the C/T ratio.

(2) The terminal rates of responding to the CS appears to be related to the reinforcement rate of the CS (T; specifically, 1/T) but not its relation to the reinforcement rate of the context (i.e. C or C/T). In the present experiment, all CS trials were reinforced so it is also the case that the terminal rate of responding was related to the duration of the CS.

(3) An unexpected finding was that responding during the ITI was similarly related to the rate of contextual reinforcement (1/C). This novel finding suggests that the terminal rate of responding during the ITI and the CS are related to their corresponding rates of reinforcement. This finding is surprising as it suggests that responding during the ITI is not being driven by the probability of reinforcement during the ITI.

(4) Finally, the authors characterised the nature of increased responding from the point of initial acquisition until responding peaks at a maximum. Their analyses suggest that nature of this increase was best described as linear in the majority of rats, as opposed to the non-linear increase that might be predicted by prediction error learning models (e.g. Rescorla-Wagner). However, more detailed analyses revealed that these changes can be quite variable across rats, and more variable when the CS had lower informativeness (defined as C/T).

Strengths and Weaknesses:

There is an inherent paradox regarding the consistency of the acquisition data from Gibbon & Balsam's (1981) meta-analysis of autoshaping in pigeons, and the present results in magazine response frequency in rats. This consistency is remarkable and impressive, and is suggestive of a relatively conserved or similar underlying learning principle. However, the consistency is also surprising given some significant differences in how these experiments were run. Some of these differences might reasonably be expected to lead to differences in how these different species respond. For example:

- The autoshaping procedure commonly used in the pigeons from these data were pretrained to retrieve rewards from a grain hopper with an instrumental contingency between head entry into the hopper and grain availability. During Pavlovian training, pecking the key light also elicited an auditory click feedback stimulus, and when the grain hopper was made available the hopper was also illuminated.

- In the present experimental procedure, the rats were not given contextual exposure to the pellet reinforcers in the magazine (e.g. a magazine training session is typically found in similar rodent procedures). The Pavlovian CS was a cue light within the magazine itself.

These design features in the present rodent experiment are clearly intentional. Pretraining with the reinforcer in the testing chambers would reasonably alter the background rate of reinforcement (parameter), so it make sense not to include this but differs from the paradigm used in pigeons. Having the CS inside the magazine where pellets are delivered provides an effective way to reduce any potential response competition between CS and US directed responding and combines these all into the same physical response. This makes the magazine approach response more like the pecking of the light stimulus in the pigeon autoshaping paradigm. However, the location of the CS and US is separated in pigeon autoshaping, raising questions about why the findings across species are consistent despite these differences.

Intriguingly, when the insertion of a lever is used as a Pavlovian cue in rodent studies, CS directed responding (sign-tracking) often develops over training such that eventually all animals bias their responding towards the lever than towards the US (goal-tracking at the magazine). However, the nature of this shift highlights the important point that these CS and US directed responses can be quite distinct physically as well as psychologically. Therefore, by conflating the development of these different forms of responding, it is not clear whether the relationship between C/T and the acquisition of responding describes the sum of all Pavlovian responding or predominantly CS or US directed responding.

Another interesting aspect of these findings is that there is a large amount of variability that scales inversely with C/T. A potential account of the source of this variability is related to the absence of preexposure to the reward pellets. This is normally done within the animals' homecage as a form of preexposure to reduce neophobia. If some rats take longer to notice and then approach and finally consume the reward pellets in the magazine, the impact of this would systematically differ depending on the length of the ITI. For animals presented with relatively short CSs and ITIs, they may essentially miss the first couple of trials and/or attribute uneaten pellets accumulating in the magazine to the background/contextual rate of reinforcement. What is not currently clear is whether this was accounted for in some way by confirming when the rats first started retrieving and consuming the rewards from the magazine.

While the generality of these findings across species is impressive, the very specific set of parameters employed to generate these data raise questions about the generality of these findings across other standard Pavlovian conditioning parameters. While this is obviously beyond the scope of the present experiment, it is important to consider that the present study explored a situation with 100% reinforcement on every trial, with a variable duration CS (drawn form a uniform distribution), with a single relatively brief CS (maximum of 122s) CS and a single US. Again, the choice of these parameters in the present experiment is appropriate and very deliberately based on refinements from many previous studies from the authors. This includes a number of criteria used to define magazine response frequency that includes discarding specific responses (discussed and reasonably justified clearly in the methods section). Similarly, the finding that terminal rates of responding are reliably related to 1/T is surprising, and it is not clear whether this might be a property specific to this form of variable duration CS, the use of a uniform sampling distribution, or the use of only a single CS. However, it is important to keeps these limitations in mind when considering some of the claims made in the discussion section of this manuscript that go beyond what these data can support.

The main finding demonstrating the consistent findings across species is presented in Figure 3. In the analysis of these data, it is not clear why the correlations between C, T, and C/T and the measure of acquisition in Figure 3A were presented as r values, whereas the r2 values were presented in the discussion of Figure 3B, and no values were provided in discussing Figure 3C. The measure of acquisition in Figure 3A is based on a previously established metric, whereas the measure in Figure 3B employs the relatively novel nDKL measure that is argued to be a better and theoretically based metric. Surprisingly, when r and r2 values are converted to the same metric across analyses, it appears that this new metric (Figure 3B) does well but not as well as the approach in Figure 3A. This raises questions about why a theoretically derived measure might not be performing as well on this analysis, and whether the more effective measure is either more reliable or tapping into some aspect of the processes that underlie acquisition that is not accounted for by the nDKL metric. Unfortunately, the new metric is discussed and defined at great length but its utility is not considered.<br /> An important analysis issue that is unclear in the present manuscript is exactly how the statistics were run (how the model was defined, were individual subjects or group medians used, what software was used etc...). For example, it is not clear whether the analyses conducted in relation to Figure 3 used the data from individual rats or the group medians. Similarly, it appears that each rat contributes four separate data points, and a single regression line was fit to all these data despite the highly likely violation of the assumption independent observations (or more precisely, the assumption of uncorrelated errors) in this analysis. Furthermore, it is claimed that the same regression line fit the IT and CS period data in this figure, however this<br /> If the data in figure 3 were analyzed with log(ITI) or log(C/ITI) i.e. log(C/(T-C)), would this be a better fit for these data? Is it the case that the ratio of C/T the best predictor of the trial/point of acquisition, or is it the case that another metric related to reinforcement rates provides a better fit?

Based on the variables provided in Supplementary file 3, containing the acquisition data, I was unable to reproduce the values reported in the analysis of Figure 3.<br /> In relation to Figure 3: I am curious about whether the authors would be able to comment on whether the individual variability in trials to acquisition would be expected to scale differently based on C/T, or C, or (if a less restricted range was used) T?<br /> It is not clear why Figure 3C is presented but not analyzed, and why the data presented in Figure 4 to clarify the spread of the distribution of the data observed across the plots in Figure 3 uses the data from Figure 3C. This would seem like the least representative data to illustrate the point of Figure 4. It also appears to my eye that the data actually plotted in Figure 4 correspond to Figure 3A and 3B rather than the odds 10:1 data indicated in text.

What was the decision criteria used to decide on averaging the final 5 conditioning sessions as terminal responding for the analyses in Figure 5? This is an oddly specific number. Was this based on consistency with previous work, or based on the greatest number of sessions where stable data for all animals could be extracted?<br /> In the analysis corresponding to Figures 7-8: If I understand the description of this analysis correctly, for each rat the data are the cumulative response data during the CS, starting from the trial on which responding to the CS > ITI (t = 1), and ending at the trial on which CS responding peaked (maximum over 3 session moving average window; t = end). This analysis does not seem to account for changes (decline) in the ITI response rates over this period of acquisition, and it is likely that responding during the ITI is still declining after t=1. Are the 4 functions that were fit to these data to discriminate between different underlying generative processes still appropriate on total CS responding instead of conditional CS responding after accounting for changes in baseline response rates during ITI?

Page 27, Procedure, final sentence: The magazine responding during the ITI is defined as the 20s period immediately before CS onset. The range of ITI values (Table 1) always starts as low as 15s in all 14 groups. Even in the case of an ITI on a trial that was exactly 20s, this would also mean that the start of this period overlaps with the termination of the CS from the previous trial and delivery (and presumably consumption) of a pellet. Please indicate if the definition of the ITI period was modified on trials where the preceding ITI was <20s, and if any other criteria were used to define the ITI.

Were the rats exposed to the reinforcers/pellets in their home cage prior to acquisition? Please indicate whether rats where pre-exposed to the reward pellets in their home cages e.g. as is often done to reduce neophobia. Given the deliberate absence of a magazine-training phase, this information is important when assessing the experienced contingency between the CS and the US.

For all the analyses, please provide the exact models that were fit and the software used. For example, it is not necessarily clear to the reader (particularly in the absence of degrees of freedom) that the model fits discussed in Figure 3 are fit on the individual subject data points or the group medians. Similarly, in Figure 6 there is no indication of whether a single regression model was fit to all the plotted data or whether tests of different slopes for each of the conditions were compared. With regards to the statistics in Figure 6, depending on how this was run, it is also a potential problem that the analyses does not correct for the potentially highly correlated multiple measurements from the same subjects i.e. each rat provides 4 data points which are very likely not to be independent observations.

A number of sections of the discussion are speculative or not directly supported by the present experimental data (but may well be supported by previous findings that are not the direct focus of the present experiment). For example, Page 19, Paragraph 2: this entire paragraph is not really clearly explained and is presenting an opinion rather than a strong conclusion that follows directly from the present findings. Evidence for an aspect of RET in the present paper (i.e. the prediction of time scale invariance on the initial point of acquisition, but not necessarily the findings regarding the rate of terminal acquisition) - while supportive - does not necessarily provide unconditional evidence for this theory over all the alternatives.

Similarly, the Conclusion section (Page 23) makes the claim that "the equations have at most one free parameter", which may be an oversimplification that is conditionally true in the narrow context of the present experiment where many things were kept constant between groups and run in a particular way to ensure this is the case. While the equations do well in this narrow case, it is unlikely that additional parameters would not need to be added to account for more general learning situations. To clarify, I am not contending that this kind of statement is necessarily untrue, merely that it is being presented in a narrow context and may require a deeper discussion of much more of the literature to qualify/support properly - and the discussion section of the present experiment/manuscript may not be the appropriate place for this.

- Consider taking advantage of an "Ideas and Speculation" subsection within the Discussion that is supported by eLife [ https://elifesciences.org/inside-elife/e3e52a93/elife-latest-including-ideas-and-speculation-in-elife-papers ]. This might be more appropriate to qualify the tone of much of the discussion from page 19 onwards.

It seems like there are entire analyses and new figures being presented in the discussion e.g. Page 20: Information-Theoretic Contingency. These sections might be better placed in the methods section or a supplementary section/discussion.

Author response:

The following is the authors’ response to the original reviews

ANALYTICAL

(1) A key claim made here is that the same relationship (including the same parameter) describes data from pigeons by Gibbon and Balsam (1981; Figure 1) and the rats in this study (Figure 3). The evidence for this claim, as presented here, is not as strong as it could be. This is because the measure used for identifying trials to criterion in Figure 1 appears to differ from any of the criteria used in Figure 3, and the exact measure used for identifying trials to criterion influences the interpretation of Figure 3***. To make the claim that the quantitative relationship is one and the same in the Gibbon-Balsam and present datasets, one would need to use the same measure of learning on both datasets and show that the resultant plots are statistically indistinguishable, rather than simply plotting the dots from both data sets and spotlighting their visual similarity. In terms of their visual characteristics, it is worth noting that the plots are in log-log axis and, as such, slight visual changes can mean a big difference in actual numbers. For instance, between Figure 3B and 3C, the highest information group moves up only "slightly" on the y-axis but the difference is a factor of 5 in the real numbers. Thus, in order to support the strong claim that the quantitative relationships obtained in the Gibbon-Balsam and present datasets are identical, a more rigorous approach is needed for the comparisons.

***The measure of acquisition in Figure 3A is based on a previously established metric, whereas the measure in Figure 3B employs the relatively novel nDKL measure that is argued to be a better and theoretically based metric. Surprisingly, when r and r2 values are converted to the same metric across analyses, it appears that this new metric (Figure 3B) does well but not as well as the approach in Figure 3A. This raises questions about why a theoretically derived measure might not be performing as well on this analysis, and whether the more effective measure is either more reliable or tapping into some aspect of the processes that underlie acquisition that is not accounted for by the nDKL metric.

Figure 3 shows that the relationship between learning rate and informativeness for our rats was very similar to that shown with pigeons by Gibbon and Balsam (1981). We have used multiple criteria to establish the number of trials to learn in our data, with the goal of demonstrating that the correspondence between the data sets was robust. In the revised Figure 3, specifically 3C and 3D, we have plotted trials to acquisition using decision criterion equivalent to those used by Gibbon and Balsam. The criterion they used—at least one peck at the response key on at least 3 out of 4 consecutive trials—cannot be directly applied to our magazine entry data because rats make magazine entries during the inter-trial interval (whereas pigeons do not peck at the response key in the inter-trial interval). Therefore, evidence for conditioning in our paradigm must involve comparison between the response rate during CS and the baseline response rate, rather than just counting responses during the CS. We have used two approaches to adapt the Gibbon and Balsam criterion to our data. One approach, plotted in Figure 3C, uses a non-parametric signed rank test for evidence that the CS response rate exceeds the pre-CS response rate, and adopting a statistical criterion equivalent to Gibbon and Balsam’s 3-out-of-4 consecutive trials (p<.3125). The second method (Figure 3D) estimates the nDkl for the criterion used by Gibbon and Balsam and then applies this criterion to the nDkl for our data. To estimate the nDkl of Gibbon and Balsam’s data, we have assumed there are no responses in the inter-trial interval and the response probability during the CS must be at least 0.75 (their criterion of at least 3 responses out of 4 trials). The nDkl for this difference is 2.2 (odds ratio 27:1). We have then applied this criterion to the nDkl obtained from our data to identify when the distribution of CS response rates has diverged by an equivalent amount from the distribution of pre-CS response rates. These two analyses have been added to the manuscript to replace those previously shown in Figures 3B and 3C.

(2) Another interesting claim here is that the rates of responding during ITI and the cue are proportional to the corresponding reward rates with the same proportionality constant. This too requires more quantification and conceptual explanation. For quantification, it would be more convincing to calculate the regression slope for the ITI data and the cue data separately and then show that the corresponding slopes are not statistically distinguishable from each other. Conceptually, it is not clear why the data used to test the ITI proportionality came from the last 5 conditioning sessions. What were the decision criteria used to decide on averaging the final 5 sessions as terminal responses for the analyses in Figure 5? Was this based on consistency with previous work, or based on the greatest number of sessions where stable data for all animals could be extracted?

If the model is that animals produce response rates during the ITI (a period with no possible rewards) based on the overall rate of rewards in the context, wouldn't it be better to test this before the cue learning has occurred? Before cue learning, the animals would presumably only have attributed rewards in the context to the context and thus, produce overall response rates in proportion to the contextual reward rate. After cue learning, the animals could technically know that the rate of rewards during ITI is zero. Why wouldn't it be better to test the plotted relationship for ITI before cue learning has occurred? Further, based on Figure 1, it seems that the overall ITI response rate reduces considerably with cue learning. What is the expected ITI response rate prior to learning based on the authors' conceptual model? Why does this rate differ from pre and post-cue learning? Finally, if the authors' conceptual framework predicts that ITI response rate after cue learning should be proportional to contextual reward rate, why should the cue response rate be proportional to the cue reward rate instead of the cue reward rate plus the contextual reward rate?

A single regression line, as shown in Figure 5, is the simplest possible model of the relationship between response rate and reinforcement rate and it explains approximately 80% of the variance in response rate. Fixing the log-log slope at 1 yields the maximally simple model. (This regression is done in the logarithmic domain to satisfy the homoscedasticity assumption.) When transformed into the linear domain, this model assumes a truly scalar relation (linear, intercept at the origin) and assumes the same scale factor and the same scalar variability in response rates for both sets of data (ITI and CS). Our plot supports such a model. Its simplicity is its own motivation (Occam’s razor).

If separate regression lines are fitted to the CS and ITI data, there is a small increase in explained variance (R₂ = 0.82). These regression lines have been added to the plot in the revised manuscript (Figure 5). We leave it to further research to determine whether such a complex model, with 4 parameters, is required. However, we do not think the present data warrant comparing the simplest possible model, with one parameter, to any more complex model for the following reasons:

· When a brain—or any other machine—maps an observed (input) rate to a rate it produces (output rate), there is always an implicit scalar. In the special case where the produced rate equals the observed rate, the implicit scalar has value 1. Thus, there cannot be a simpler model than the one we propose, which is, in and of itself, interesting.

· The present case is an intuitively accessible example of why the MDL (Minimum Description Length) approach to model complexity (Barron, Rissanen, & Yu, 1998; Grünwald, Myung, & Pitt, 2005; Rissanen, 1999) can yield a very different conclusion from the conclusion reached using the Bayesian Information Criterion (BIC) approach. The MDL approach measures the complexity of a model when given N data specified with precision of B bits per datum by computing (or approximating) the sum of the maximum-likelihoods of the model’s fits to all possible sets of N data with B precision per datum. The greater the sum over the maximum likelihoods, the more complex the model, that is, the greater its measured wiggle room, it’s capacity to fit data. Recall that von Neuman remarked to Fermi that with 4 parameters he could fit an elephant. His deeper point was that multi-parameter models bring neither insight nor predictive power; they explain only post-hoc, after one has adjusted their parameters in the light of the data. For realistic data sets like ours, the sums of maximum likelihoods are finite but astronomical. However, just as the Sterling approximation allows one to work with astronomical factorials, it has proved possible to develop readily computable approximations to these sums, which can be used to take model complexity into account when comparing models. Proponents of the MDL approach point out that the BIC is inadequate because models with the same number of parameters can have very different amounts of wiggle room. A standard illustration of this point is the contrast between logarithmic model and power-function model. Log regressions must be concave; whereas power function regressions can be concave, linear, or convex—yet they have the same number of parameters (one or two, depending on whether one counts the scale parameter that is always implicit). The MDL approach captures this difference in complexity because it measures wiggle room; the BIC approach does not, because it only counts parameters.

· In the present case, one is comparing a model with no pivot and no vertical displacement at the boundary between the black dots and the red dots (the 1-parameter unilinear model) to a bilinear model that allows both a change in slope and a vertical displacement for both lines. The 4-parameter model is superior if we use the BIC to take model complexity into account. However, 4-parameter has ludicrously more wiggle room. It will provide excellent fits—high maximum likelihood—to data sets in which the red points have slope > 1, slope 0, or slope < 0 and in which it is also true that the intercept for the red points lies well below or well above the black points (non-overlap in the marginal distribution of the red and black data). The 1-parameter model, on the other hand, will provide terrible fits to all such data (very low maximum likelihoods). Thus, we believe the BIC does not properly capture the immense actual difference in the complexity between the 1-parameter model (unilinear with slope 1) to the 4-parameter model (bilinear with neither the slope nor the intercept fixed in the linear domain).

· In any event, because the pivot (change in slope between black and red data sets), if any, is small and likewise for the displacement (vertical change), it suffices for now to know that the variance captured by the 1-parameter model is only marginally improved by adding three more parameters. Researchers using the properly corrected measured rate of head poking to measure the rate of reinforcement a subject expects can therefore assume that they have an approximately scalar measure of the subject’s expectation. Given our data, they won’t be far wrong even near the extremes of the values commonly used for rates of reinforcement. That is a major advance in current thinking, with strong implications for formal models of associative learning. It implies that the performance function that maps from the neurobiological realization of the subject’s expectation is not an unknown function. On the contrary, it’s the simplest possible function, the scalar function. That is a powerful constraint on brain-behavior linkage hypotheses, such as the many hypothesized relations between mesolimbic dopamine activity and the expectation that drives responding in Pavlovian conditioning (Berridge, 2012; Jeong et al., 2022; Y.  Niv, Daw, Joel, & Dayan, 2007; Y. Niv & Schoenbaum, 2008).

The data in Figures 4 and 5 are taken from the last 5 sessions of training. The exact number of sessions was somewhat arbitrary but was chosen to meet two goals: (1) to capture asymptotic responding, which is why we restricted this to the end of the training, and (2) to obtain a sufficiently large sample of data to estimate reliably each rat’s response rate. We have checked what the data look like using the last 10 sessions, and can confirm it makes very little difference to the results. We now note this in the revised manuscript. The data for terminal responding by all rats, averaged over both the last 5 sessions and last 10 sessions, can be downloaded from https://osf.io/vmwzr/

Finally, as noted by the reviews, the relationship between the contextual rate of reinforcement and ITI responding should also be evident if we had measured context responding prior to introducing the CS. However, there was no period in our experiment when rats were given unsignalled reinforcement (such as is done during “magazine training” in some experiments). Therefore, we could not measure responding based on contextual conditioning prior to the introduction of the CS. This is a question for future experiments that use an extended period of magazine training or “poor positive” protocols in which there are reinforcements during the ITIs as well as during the CSs. The learning rate equation has been shown to predict reinforcements to acquisition in the poor-positive case (Balsam, Fairhurst, & Gallistel, 2006).

(3) There is a disconnect between the gradual nature of learning shown in Figures 7 and 8 and the information-theoretic model proposed by the authors. To the extent that we understand the model, the animals should simply learn the association once the evidence crosses a threshold (nDKL > threshold) and then produce behavior in proportion to the expected reward rate. If so, why should there be a gradual component of learning as shown in these figures? In terms of the proportional response rule to the rate of rewards, why is it changing as animals go from 10% to 90% of peak response? The manuscript would be greatly strengthened if these results were explained within the authors' conceptual framework. If these results are not anticipated by the authors' conceptual framework, this should be explicitly stated in the manuscript.

One of us (CRG) has earlier suggested that responding appears abruptly when the accumulated evidence that the CS reinforcement rate is greater than the contextual rate exceeds a decision threshold (C.R.  Gallistel, Balsam, & Fairhurst, 2004). The new more extensive data require a more nuanced view. Evidence about the manner in which responding changes over the course of training is to some extent dependent on the analytic method used to track those changes. We presented two different approaches. The approach shown in Figures 7 and 8 (now 6 and 7), extending on that developed by Harris (2022), assumes a monotonic increase in response rate and uses the slope of the cumulative response rate to identify when responding exceeds particular milestones (percentiles of the asymptotic response rate). This analysis suggests a steady rise in responding over trials. Within our theoretical model, this might reflect an increase in the animal’s certainty about the CS reinforcement rate with accumulated evidence from each trial. While this method should be able to distinguish between a gradual change and a single abrupt change in responding (Harris, 2022) it may not distinguish between a gradual change and multiple step-like changes in responding and cannot account for decreases in response rate.

The other analytic method we used relies on the information theoretic measure of divergence, the nDkl (Gallistel & Latham, 2023), to identify each point of change (up or down) in the response record. With that method, we discern three trends. First, the onset tends to be abrupt in that the initial step up is often large (an increase in response rate by 50% or more of the difference between its initial value and its terminal value is common and there are instances where the initial step is to the terminal rate or higher). Second, there is marked within-subject variability in the response rate, characterized by large steps up and down in the parsed response rates following the initial step up, but this variability tends to decrease with further training (there tend to be fewer and smaller steps in both the ITI response rates and the CS response rate as training progresses). Third, the overall trend, seen most clearly when one averages across subjects within groups is to a moderately higher rate of responding later in training than after the initial rise. We think that the first tendency reflects an underlying decision process whose latency is controlled by diminishing uncertainty about the two reinforcement rates and hence about their ratio. We think that decreasing uncertainty about the true values of the estimated rates of reinforcement is also likely to be an important part of the explanation for the second tendency (decreasing within-subject variation in response rates). It is less clear whether diminishing uncertainty can explain the trend toward a somewhat greater difference in the two response rates as conditioning progresses. It is perhaps worth noting that the distribution of the estimates of the informativeness ratio is likely to be heavy tailed and have peculiar properties (as witness, for example, the distribution of the ratio of two gamma distributions with arbitrary shape and scale parameters) but we are unable at this time to propound an explanation of the third trend.

(4) Page 27, Procedure, final sentence: The magazine responding during the ITI is defined as the 20 s period immediately before CS onset. The range of ITI values (Table 1) always starts as low as 15 s in all 14 groups. Even in the case of an ITI on a trial that was exactly 20 s, this would also mean that the start of this period overlaps with the termination of the CS from the previous trial and delivery (and presumably consumption) of a pellet. It should be indicated whether the definition of the ITI period was modified on trials where the preceding ITI was < 20 s, and if any other criteria were used to define the ITI. Were the rats exposed to the reinforcers/pellets in their home cage prior to acquisition?

There was an error in the description provided in the original text. The pre-CS period used to measure the ITI responding was 10 s rather than 20 s. There was always at least a 5-s gap between the end of the previous trial and the start of the pre-CS period. The statement about the pre-CS measure has been corrected in the revised manuscript.

(5) For all the analyses, the exact models that were fit and the software used should be provided. For example, it is not necessarily clear to the reader (particularly in the absence of degrees of freedom) that the model discussed in Figure 3 fits on the individual subject data points or the group medians. Similarly, in Figure 6 there is no indication of whether a single regression model was fit to all the plotted data or whether tests of different slopes for each of the conditions were compared. With regards to the statistics in Figure 6, depending on how this was run, it is also a potential problem that the analyses do not correct for the potentially highly correlated multiple measurements from the same subjects, i.e. each rat provides 4 data points which are very unlikely to be independent observations.

Details about model fitting have been added to the revision. The question about fitting a single model or multiple models to the data in Figure 6 (now 5) is addressed in response 2 above. In Figure 5, each rat provides 2 behavioural data points (ITI response rate and CS response rate) and 2 values for reinforcement rate (1/C and 1/T). There is a weak but significant correlation between the ITI and CS response rates (r = 0.28, p < 0.01; log transformed to correct for heteroscedasticity). By design, there is no correlation between the log reinforcement rates (r = 0.06, p = .404).

CONCEPTUAL

(1) We take the point that where traditional theories (e.g., Rescorla-Wagner) and rate estimation theory (RET) both explain some phenomenon, the explanation in terms of RET may be preferred as it will be grounded in aspects of an animal's experience rather than a hypothetical construct. However, like traditional theories, RET does not explain a range of phenomena - notably, those that require some sort of expectancy/representation as part of their explanation. This being said, traditional theories have been incorporated within models that have the representational power to explain a broader array of phenomena, which makes me wonder: Can rate estimation be incorporated in models that have representational power; and, if so, what might this look like? Alternatively, do the authors intend to claim that expectancy and/or representation - which follow from probabilistic theories in the RW mould - are unnecessary for explanations of animal behaviour?***

It is important for the field to realize that the RW model cannot be used to explain the results of Rescorla’s (Rescorla, 1966; Rescorla, 1968, 1969) contingency-not-pairing experiments, despite what was claimed by Rescorla and Wagner (Rescorla & Wagner, 1972; Wagner & Rescorla, 1972) and has subsequently been claimed in many modelling papers and in most textbooks and reviews (Dayan & Niv, 2008; Y. Niv & Montague, 2008). Rescorla programmed reinforcements with a Poisson process. The defining property of a Poisson process is its flat hazard function; the reinforcements were equally likely at every moment in time when the process was running. This makes it impossible to say when non-reinforcements occurred and, a fortiori, to count them. The non-reinforcements are causal events in RW algorithm and subsequent versions of it. Their effects on associative strength are essential to the explanations proffered by these models. Non-reinforcements—failures to occur, updates when reinforcement is set to 0, hence also the lambda parameter—can have causal efficacy only when the successes may be predicted to occur at specified times (during “trials”). When reinforcements are programmed by a Poisson process, there are no such times. Attempts to apply the RW formula to reinforcement learning soon foundered on this problem (Gibbon, 1981; Gibbon, Berryman, & Thompson, 1974; Hallam, Grahame, & Miller, 1992; L.J. Hammond, 1980; L. J. Hammond & Paynter, 1983; Scott & Platt, 1985). The enduring popularity of the delta-rule updating equation in reinforcement learning depends on “big-concept” papers that don’t fit models to real data and discretize time into states while claiming to be real-time models (Y. Niv, 2009; Y. Niv, Daw, & Dayan, 2005).

The information-theoretic approach to associative learning, which sometimes historically travels as RET (rate estimation theory), is unabashedly and inescapably representational. It assumes a temporal map and arithmetic machinery capable in principle of implementing any implementable computation. In short, it assumes a Turing-complete brain. It assumes that whatever the material basis of memory may be, it must make sense to ask of it how many bits can be stored in a given volume of material. This question is seldom posed in associative models of learning, nor by neurobiologists committed to the hypothesis that the Hebbian synapse is the material basis of memory. Many—including the new Nobelist, Geoffrey Hinton— would agree that the question makes no sense. When you assume that brains learn by rewiring themselves rather than by acquiring and storing information, it makes no sense.

When a subject learns a rate of reinforcement, it bases its behavior on that expectation, and it alters its behavior when that expectation is disappointed. Subjects also learn probabilities when they are defined. They base some aspects of their behavior on those expectations, making computationally sophisticated use of their representation of the uncertainties (Balci, Freestone, & Gallistel, 2009; Chan & Harris, 2019; J. A. Harris, 2019; J.A. Harris & Andrew, 2017; J. A. Harris & Bouton, 2020; J. A. Harris, Kwok, & Gottlieb, 2019; Kheifets, Freestone, & Gallistel, 2017; Kheifets & Gallistel, 2012; Mallea, Schulhof, Gallistel, & Balsam, 2024 in press).

(2) The discussion of Rescorla's (1967) and Kamin's (1968) findings needs some elaboration. These findings are already taken to mean that the target CS in each design is not informative about the occurrence of the US - hence, learning about this CS fails. In the case of blocking, we also know that changes in the rate of reinforcement across the shift from stage 1 to stage 2 of the protocol can produce unblocking. Perhaps more interesting from a rate estimation perspective, unblocking can also be achieved in a protocol that maintains the rate of reinforcement while varying the sensory properties of the US (Wagner). How does rate estimation theory account for these findings and/or the demonstrations of trans-reinforcer blocking (Pearce-Ganesan)? Are there other ways that the rate estimation account can be distinguished from traditional explanations of blocking and contingency effects? If so, these would be worth citing in the discussion. More generally, if one is going to highlight seminal findings (such as those by Rescorla and Kamin) that can be explained by rate estimation, it would be appropriate to acknowledge findings that challenge the theory - even if only to note that the theory, in its present form, is not all-encompassing. For example, it appears to me that the theory should not predict one-trial overshadowing or the overtraining reversal effect - both of which are amenable to discussion in terms of rates.

I assume that the signature characteristics of latent inhibition and extinction would also pose a challenge to rate estimation theory, just as they pose a challenge to Rescorla-Wagner and other probability-based theories. Is this correct?

The seemingly contradictory evidence of unblocking and trans-reinforcer blocking by Wagner and by Pearce and Ganesan cited above will be hard for any theory to accommodate. It will likely depend on what features of the US are represented in the conditioned response.

RET predicts one-trial overshadowing, as anyone may verify in a scientific programming language because it has no free parameters; hence, no wiggle room. Overtraining reversal effects appear to depend on aspects of the subjects’ experience other than the rate of reinforcement. It seems unlikely that it can proffer an explanation.

Various information-theoretic calculations give pretty good quantitative fits to the relatively few parametric studies of extinction and the partial-reinforcement extinction effect (see Gallistel (2012, Figs 3 & 4); Wilkes & Gallistel (2016, Fig 6) and Gallistel (2025, under review, Fig 6). It has not been applied to latent inhibition, in part for want of parametric data. However, clearly one should not attribute a negative rate to a context in which the subject had never been reinforced. An explanation, if it exists, would have to turn on the effect of that long period on initial rate estimates AND on evidence of a change in rate, as of the first reinforcement.

Recommendations for authors:

MINOR POINTS

(1) It is not clear why Figure 3C is presented but not analyzed, and why the data presented in Figure 4 to clarify the spread of the distribution of the data observed across the plots in Figure 3 uses the data from Figure 3C. This would seem like the least representative data to illustrate the point of Figure 4. It also appears that the data plotted in Figure 4 corresponds to Figure 3A and 3B rather than the odds 10:1 data indicated in the text.

Figures 3 has changed as already described. The data previously plotted in Figure 4 are now shown in 3B and corresponds to that plotted in Figure 3A.

(2) Log(T) was not correlated with trials to criterion. If trials to criterion is inversely proportional to log(C/T) and C is uncorrelated with T, shouldn't trials to criterion be correlated with log(T)? Is this merely a matter of low statistical power?

Yes. There is a small, but statistically non-significant, correlation between log(T) and trials to criterion, r = 0.35, p = .22. That correlation drops to .08 (p = .8) after factoring out log(C/T), which demonstrates that the weak correlation between log(T) and trials to criterion is based on the correlation between log(t) and log(C/T).

(3) The rationale for the removal of the high information condition samples in the Fig 8 "Slope" plot to be weak. Can the authors justify this choice better? If all data are included, the relationship is clearly different from that shown in the plot.

We have now reported correlations that include those 3 groups but noted that the correlations are largely driven by the much lower slope values of those 3 groups which is likely an artefact of their smaller number of trials. We use this to justify a second set of correlations that excludes those 3 groups.

(4) The discussion states that there is at most one free parameter constrained by the data - the constant of proportionality for response rate. However, there is also another free parameter constrained by data-the informativeness at which expected trials to acquisition is 1.

I think this comment is referring to two different sets of data. The constant of proportionality of the response rate refers to the scalar relationship between reinforcement rate and terminal response rate shown in Figure 5. The other parameter, the informativeness when trials to acquisition equals 1, describes the intercept of the regression line in Figure 1 (and 3).

(5) The authors state that the measurement of available information is not often clear. Given this, how is contingency measurable based on the authors' framework?

(6) Based on the variables provided in Supplementary File 3, containing the acquisition data, we were unable to reproduce the values reported in the analysis of Figure 3.

Figure 3 has changed, using new criteria for trials to acquisition that attempt to match the criterion used by Gibbon and Balsam. The data on which these figures are based has been uploaded into OSF.

GRAPHICAL AND TYPOGRAPHICAL

(1) Y-axis labels in Figure 1 are not appropriately placed. 0 is sitting next to 0.1. 0 should sit at the bottom of the y-axis.

If this comment refers to the 0 sitting above an arrow in the top right corner of the plot, this is not misaligned. The arrow pointing to zero is used to indicate that this axis approaches zero in the upward direction. 0 should not be aligned to a value on the axis since a learning rate of zero would indicate an infinite number of learning trials. The caption has been edited to explain this more clearly.

(2) Typo, Page 6, Final Paragraph, line 4. "Fourteen groups of rats were trained with for 42 session"

Corrected. Thank you.

(3) Figure 3 caption: Typo, should probably be "Number of trials to acquisition"?

This change has now been made. The axis shows reinforcements to acquisition to be consistent with Gibbon and Balsam, but trials and number of reinforcements are identical in our 100% reinforcement schedule.

(4) Typo Page 17 Line 1: "Important pieces evidence about".

Correct. Thank you.

(5) Consider consistent usage of symbols/terms throughout the manuscript (e.g. Page 22, final paragraph: "iota = 2" is used instead of the corresponding symbol that has been used throughout).

Changed.

(6) Typo Page 28, Paragraph 1, Line 9: "We used a one-sample t-test using to identify when this".

This section of text has been changed to reflect the new analysis used for the data in Figure 3.

(7) Typo Page 29, Paragraph 1, Line 2: "problematic in cases where one of both rates are undefined" either typo or unclear phrasing.

“of” has been corrected to “or”

(8) Typo Page 30: Equation 3 appears to have an error and is not consistent with the initial printing of Equation 3 in the manuscript.

The typo in initial expression of Eq 3 (page 23) has been corrected.

(9) Typo Page 33, Line 5: "Figures 12".

Corrected.

(10) Typo Page 34, Line 10: "and the 5 the increasingly"? Should this be "the 5 points that"?

Corrected.

(11) Typo Page 35, Paragraph 2: "estimate of the onset of conditioned is the trial after which".

Corrected.

(12) Clarify: Page 35, final paragraph: it is stated that four-panel figures are included for each subject in the Supplementary files, but each subject has a six-panel figure in the Supplementary file.

The text now clarifies that the 4-panel figures are included within the 6-panel figures in the Supplementary materials.

(13) It is hard to identify the different groups in Figure 2 (Plot 15).

The figure is simply intended to show that responding across seconds within the trial is relatively flat for each group. Individuation of specific groups is not particularly important.

(14) It appears that the numbering on the y-axis is misaligned in Figure 2 relative to the corresponding points on the scale (unless I have misunderstood these values and the response rate measure to the ITI can drop below 0?).

The numbers on the Y axes had become misaligned. That has now been corrected.

(15) Please include the data from Figure 3A in the spreadsheet supplementary file 3. If it has already been included as one of the columns of data, please consider a clearer/consistent description of the relevant column variable in Supplementary File 1.

The data from Figure 3 are now available from the linked OSF site, referenced in the manuscript.

(16) Errors in supplementary data spreadsheets such that the C/T values are not consistent with those provided in Table 1 (C/T values of 4.5, 54, 180, and 300 are slightly different values in these spreadsheets). A similar error/mismatch appears to have occurred in the C/T labels for Figures (e.g. Figure 10) and the individual supplementary figures.

The C/T values on the figures in the supplementary materials have been corrected and are now consistent with those in Table 1.

(17) Currently the analysis and code provided at https://osf.io/vmwzr/ are not accessible without requesting access from the author. Please consider making these openly available without requiring a request for authorization. As such, a number of recommendations made here may already have been addressed by the data and code deposited on OSF. Apologies for any redundant recommendations.

Data and code are now available in at the OSF site which has been made public without requiring request.

(18) Please consider a clearer and more specific reference to supplementary materials. Currently, the reader is required to search through 4 separate supplementary files to identify what is being discussed/referenced in the text (e.g. Page 18, final line: "see Supplementary Materials" could simply be "see Figure S1").

We have added specific page numbers in references to the Supplementary Materials.

eLife Assessment

This valuable study presents an analysis of evolutionary conservation in intrinsically disordered regions, identified as key drivers of phase separation, leveraging a protein language model. The strength of evidence is potentially compelling, but a clearer justification of the methods and analyses is needed to fully support the main claims.

Reviewer #1 (Public review):

The manuscript by Zhang et al describes the use of a protein language model (pLM) to analyse disordered regions in proteins, with a focus on those that may be important in biological phase separation. While the paper is relatively easy to read overall, my main comment is that the authors could perhaps make it clearer which observations are new, and which support previous work using related approaches. Further, while the link to phase separation is interesting, it is not completely clear which data supports the statements made, and this could also be made clearer.

Major comments:

(1) With respect to putting the work in a better context of what has previously been done before, this is not to say that there is not new information in it, but what the authors do is somewhat closely related to work by others. I think it would be useful to make those links more directly. Some examples:

(1a) Alderson et al (reference 71) analysed in detail the conservation of IDRs (via pLDDT, which is itself related to conservation) to show, for example, that conserved residues fold upon binding. This analysis is very similar to the analysis used in the current study (using ESM2 as a different measure of conservation). Thus, the approach (pages 7-8) described as "This distinction allows us to classify disordered regions into two types: "flexible disordered" regions, which show high ESM2 scores and greater mutational tolerance, and "conserved disordered" regions, which display low ESM2 scores, indicating varying levels of mutational constraint despite a lack of stable folding." is fundamentally very similar to that used by Alderson et al. Thus, the result that "Given that low ESM2 scores generally reflect mutational constraint in folded proteins, the presence of region a among disordered residues suggests that certain disordered amino acids are evolutionarily conserved and likely functionally significant" is in some ways very similar to the results of that paper.

(1b) Dasmeh et al (https://doi.org/10.1093/genetics/iyab184), Lu et al (https://doi.org/10.1371/journal.pcbi.1010238) and Ho & Huang (https://doi.org/10.1002/pro.4317) analysed conservation in IDRs, including aromatic residues and their role in phase separation

(1c) A number of groups have performed proteomewide saturation scans using pLMs, including variants of the ESM family, including Meier (reference 89, but cited about something else) and Cagiada et al (https://doi.org/10.1101/2024.05.21.595203) that analysed variant effects in IDRs using a pLM. Thus, I think statements such as "their applicability to studying the fitness and evolutionary pressures on IDRs has yet to be established" should possibly be qualified.

(2) On page 4, the authors write, "The conserved residues are primarily located in regions associated with phase separation." These results are presented as a central part of the work, but it is not completely clear what the evidence is.

(3) It would be useful with an assessment of what controls the authors used to assess whether there are folded domains within their set of IDRs.

Reviewer #2 (Public review):

This manuscript uses the ESM2 language model to map the evolutionary fitness landscape of intrinsically disordered regions (IDRs). The central idea is that mutational preferences predicted by these models could be useful in understanding eventual IDR-related behavior, such as disruption of otherwise stable phases. While ESM2-type models have been applied to analyze such mutational effects in folded proteins, they have not been used or verified for studying IDRs. Here, the authors use ESM2 to study membraneless organelle formation and the related fitness landscape of IDRs.

Through this, their key finding in this work is the identification of a subset of amino acids that exhibit mutation resistance. Their findings reveal a strong correlation between ESM2 scores and conservation scores, which if true, could be useful for understanding IDRs in general. Through their ESM2-based calculations, the authors conclude that IDRs crucial for phase separation frequently contain conserved sequence motifs composed of both so-called sticker and spacer residues. The authors note that many such motifs have been experimentally validated as essential for phase separation.

Unfortunately, I do not believe that the results can be trusted. ESM2 has not been validated for IDRs through experiments. The authors themselves point out its little use in that context. In this study, they do not provide any further rationale for why this situation might have changed. Furthermore, they mention that experimental perturbations of the predicted motifs in in vivo studies may further elucidate their functional importance, but none of that is done here. That some of the motifs have been previously validated does not give any credibility to the use of ESM2 here, given that such systems were probably seen during the training of the model.

I believe that the authors should revamp their whole study and come up with a rigorous, scientific protocol where they make predictions and test them using ESM2 (or any other scientific framework).

Reviewer #3 (Public review):

Summary:

This is a very nice and interesting paper to read about motif conservation in protein sequences and mainly in IDRs regions using the ESM2 language model. The topic of the paper is timely, with strong biological significance. The paper can be of great interest to the scientific community in the field of protein phase transitions and future applications using the ESM models. The ability of ESM2 to identify conserved motifs is crucial for disease prediction, as these regions may serve as potential drug targets. Therefore, I find these findings highly significant, and the authors strongly support them throughout the paper. The work motivates the scientific community towards further motif exploration related to diseases.

Strengths:

(1) Revealing conserved regions in IDRs by the ESM-2 language model.

(2) Identification of functionally significant residues within protein sequences, especially in IDRs.

(3) Findings supported by useful analyses.

Weaknesses:

(1) Lack of examples demonstrating the potential biological functions of these conserved regions

(2) Very limited discussion of potential future work and of limitations.

eLife Assessment

In this valuable study, the authors demonstrate that TCF7L2 plays a role in the pathogenesis of cachexia in a mouse model of GI cancer. The results are solid, although future studies will need further mechanistic analyses. These data will be interesting to cancer biologists, especially those trying to understand late-stage complications such as cachexia and wasting, a major cause of cancer morbidity and mortality.

Reviewer #1 (Public review):

Summary:

Systemic and partial Tcf7l2 repression is effective in protecting cancer mice from cachexia-induced death. Hence, this is a promising treatment strategy for cancer patients suffering from cachexia.

Strengths:

The method is well-designed and clearly explained.

Weaknesses:

(1) Abbreviations should be mentioned in full terms for the first time.

(2) Relatively old or even very old references in the Introduction and Discussion.

(3) The result section contains discussion with references, as well.

(4) The number of mice in individual groups is relatively small (3 mice in some groups).

Reviewer #2 (Public review):

Summary:

This study by Leong and colleagues examines the role of the TCF7L2 transcription factor in the Wnt signaling pathway as a regulator of colon/small intestinal cancers and cachexia. Investigators utilize a Tet off repressor genetic system in mice under Dox regulation to silence TCF7L2. Results show DSS-treated APCMin/+ mice lose body weight that can be rapidly rescued by Dox treatment and suppression of TCF7L2 expression. Reduction of TCF7L2 rescues features of cachexia, including body weight, gastrocnemius muscle and adipose mass, as well as molecular markers of cachexia such as the E3 Ub ligases, MuRF1, and Atrogin-1. The most significant finding in the study is that loss of TCF7L2 reduces but does not eliminate tumor progression, as tumors go from adenomas to adenocarcinomas over time while mice are treated with Dox, yet cachexia persists. This implies that TCF7L2 has a direct effect on cachexia. Overall, the study provides insight into the role of TCFL2 in the development of intestinal cancers and muscle atrophy in cachexia.

Strengths:

The study uses an elegant genetic mouse model to provide significant new insight into the role of TCFL2 in colon and small intestinal cancers. In addition, the authors describe the role of TCF7L2 as a regulator of muscle wasting in cachexia. This, too, can be viewed as a new finding for the cachexia field.

Weaknesses:

However, in its current form, the study lacks sufficient data to support the authors' claim regarding the relevance of TCF7L2 as a regulator of cachexia.

eLife Assessment

This is a fundamental discovery revealing two independent IFNγ-induced pathways that restrict bacterial motility: one GBP1-dependent and the other GVIN1-dependent. The findings are supported by compelling evidence. While the paper is already very strong, there are a few points that could be addressed editorially or through the addition of a few key experiments.

Reviewer #1 (Public review):

This is a very elegant and convincing study. Using systematic screening of actin tail formation in two bacterial strains and employing a panel of CRISPR-CAS ko cell lines, the authors identify a novel dynamin-related GTPase GVIN, which forms an oligomeric coat around an intracellular Burkholderia strain. The bacterial O-antigen LPS layer is required for the formation of the GVIN coat, which disturbs the polar localization of the bacterial actin-polymerizing BimA protein.

I am not an expert in infection studies, but the experiments appear to be of high quality, the figures are well prepared, and clean and statistically significant results are provided. I have no criticism of the presented approaches.

The identification of a novel GBP1-independent pathway targeting intracellular bacteria is not only of fundamental importance for the immunity field but also of high interest to researchers in other areas, for example, evolutionary or structural biologists.

Reviewer #2 (Public review):

Summary:

The authors wanted to investigate how cells defend against intracellular pathogens, such as Shigella and Burkholderia species, that co-opt the host actin machinery to spread from cell-to-cell. Previous work has identified IFNg-inducible GTPase of the Guanylate Binding Protein (GBP) family in cytosolic defence against Gram-negative bacteria. By forming a coat around Shigella, human GBP1 suppresses actin-based motility by displacing IcsA, which is the actin-polymerising virulence factor present at bacterial poles. In addition, GBP1 recruits the cytosolic LPS-sensor, caspase-4, to the bacterial surface, which results in the removal of bacterial replicative niches via pyroptotic cell death. Here, they followed up their finding that GBP1 can reduce actin-based motility of Shigella in HeLa cells and, surprisingly, fails to do so during Burkholderia infection. In contrast, in T24 bladder epithelial cells, GBP1 is competent in blocking Burkholderia actin-tails. They therefore wanted to identify the GBP1-independent factor that blocks actin-based motility in IFNg-treated cells that is absent in HeLa cells.

Major strengths and weaknesses of the methods and results:

The authors report a second IFNg-dependent pathway involving the protein product of the gene GVIN1, which was previously thought to be a pseudogene. GVIN1 (GTPase, very large interferon inducible 1) is thus the first human member of this family of ~250 kDa putative GTPases to be demonstrated to be functional and have potential antimicrobial roles. The discovery that GVIN1 is indeed functional, forms coats on Burkholderia in an LPS O-antigen-dependent manner, and limits actin-dependent motility are the main strengths of this paper. The authors use CRISPR/Cas9-based knockouts in HeLa and T24 cells, and complement them to demonstrate that GBP1 and GVIN1 are both required to inhibit actin-based motility.

An appraisal of whether the authors achieved their aims and whether the results support their conclusions:

The authors achieved their main goals through well-planned experiments and unbiased screens. They succeeded in finding the factor that blocks actin-based motility independently of GBP1. This is driven by GVIN1, which coats bacteria and limits actin-tail formation by reducing the expression of BimA through currently unknown mechanisms. Further, they found that an O-antigen mutant can escape coating by GVIN1, indicating the requirement for these polysaccharides in GVIN1-dependent bacterial sensing. However, the authors have not investigated whether GVIN1, which has two GTPase-domains, does indeed have GTPase activity and whether GVIN1 and GBP1 together completely block cell-to-cell spread by Burkholderia and thereby restrict bacterial numbers over the infection time course. They also do not show whether GBP1 and GVIN1 target the same bacterial cell or different populations of bacteria.

A discussion of the likely impact of the work on the field, and the utility of the methods and data to the community:

This work uncovers the antimicrobial actions of a member of yet another family of IFNg-induced GTPases, which potentially acts against other intracellular pathogens. GVIN1 appears to operate independently and in parallel to GBP1, pointing to the breadth and complexity of the IFNg-inducible GTPase families.

Reviewer #3 (Public review):

Summary:

Here, Guo et al. (2025) propose that the IFN-induced GTPase GVIN1 forms a coat on cytosolic Burkholderia thailandensis, blocking actin tail formation through a mechanism analogous to GBP1-mediated restriction of Shigella motility.

Their study was prompted by the intriguing observation that IFNγ priming and GBP1 coat formation fail to inhibit B. thailandensis actin-based motility in HeLa cells, yet IFNγ restricts the motility of Burkholderia in T24 cells. Further investigation revealed that IFNγ restricts B. thailandensis motility in T24 cells via both GBP1-dependent and -independent mechanisms, suggesting that HeLa cells lack a critical GBP1 co-factor required to inhibit actin tail formation.

To identify the GBP1-independent mechanism, the authors performed an siRNA screen of interferon-stimulated genes (ISGs) and identified GVIN1, a large IFN-induced GTPase, as essential for restricting B. thailandensis motility. To identify the GBP1-independent mechanism, perform a knock-down screen for ISGs and find that the loss of GVIN, a very large IFN-induced GTPase, results in higher actin tail-positive B. thailandensis in T24 cells. They further demonstrate that GVIN forms coats on the surface of B. thailandensis, which prevent the polar localization of BimA and thus actin tail formation. In summary, the data reveal two independent IFNγ-induced pathways that restrict bacterial motility: one GBP1-dependent and the other GVIN1-dependent, each relying on distinct host co-factors.

Global assessment:

This is a well-executed study that convincingly demonstrates how GVIN1 restricts the actin-based motility of B. thailandensis through the assembly of coatomers. The results are clearly described, the manuscript is easy to follow, and the data are overall compelling and well presented. I have only a few suggestions on how the manuscript could be further improved.

eLife Assessment

This study has fundamental findings that support the potential application of exogenous stem cell therapy as a viable therapeutic option for the management of intraocular pressure (IOP) and to increase outflow facility. The evidence supporting the clinical application of stem cells is compelling, using a combination of established in vivo and ex vivo experimental techniques. The work will be of interest to both basic stem cell biologists and clinical glaucoma specialists.

Reviewer #1 (Public review):

Summary:

This manuscript describes a novel magnetic steering technique to target human adipose derived mesenchymal stem cells (hAMSC) or induce pluripotent stem cells to the TM (iPSC-TM). The authors show delivery of the stem cells lowered IOP, increased ouflow facility, and increased TM cellularity.

Strengths:

The technique is novel and shows promise as a novel therapeutic to lower in IOP in glaucoma. hAMSC are able to lower IOP below baseline as well as increase outflow facility above baseline with no tumorigenicity. These data will have a positive impact on the field and will guide further research using hAMSC in glaucoma models.

Weaknesses:

The transgenic mouse model of glaucoma the authors used did not show ocular hypertensive phenotypes as previously reported; therefore, the Tg-MYOCY437H model should be used with caution in the future. However, the results presented here clearly show magnetically steered cell therapy as a viable treatment strategy to lower intraocular pressure even from baseline. Future studies are needed to demonstrate the effects in ocular hypertensive eyes.

Reviewer #2 (Public review):

This observational study investigates the efficacy of intracameral injected human stems cells as a means to re-functionalize the trabecular meshwork for the restoration of intraocular pressure homeostasis. Using a murine model of glaucoma, human adipose-derived mesenchymal stem cells are shown to be biologically safer and functionally superior at eliciting a sustained reduction in intraocular pressure (IOP). The authors conclude that the use of magnetically-steered human adipose-derived mesenchymal stem cells has potential for long-term treatment of ocular hypertension in glaucoma.

Comments on revisions: Previously noted concerns have been thoughtfully and sincerely considered by the authors and are now clearly addressed in the revised manuscript. No further concerns/comments.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This manuscript describes a novel magnetic steering technique to target human adipose derived mesenchymal stem cells (hAMSC) or induce pluripotent stem cells to the TM (iPSC-TM). The authors show that delivery of the stem cells lowered IOP, increased outflow facility, and increased TM cellularity.

Strengths:

The technique is novel and shows promise as a novel therapeutic to lower IOP in glaucoma. hAMSC are able to lower IOP below the baseline as well as increase outflow facility above baseline with no tumorigenicity. These data will have a positive impact on the field and will guide further research using hAMSC in glaucoma models.

Weaknesses:

The transgenic mouse model of glaucoma the authors used did not show ocular hypertensive phenotypes at 6-7 months of age as previously reported. Therefore, if there is no pathology in these animals the authors did not show a restoration of function, but rather a decrease in pressure below normal IOP.

We appreciate the reviewer’s feedback and agree with the statement of weakness. Accordingly, we have revised the language to improve clarity. Specifically, all references to "restoration of IOP" or "restoration of conventional outflow function" have been replaced with more precise phrases, in the following locations: 

• lines 2-3 (title): Magnetically steered cell therapy for reduction of intraocular pressure  as a treatment strategy for open-angle glaucoma

• lines 36-8 (abstract): We observed a 4.5 [3.1, 6.0] mmHg or 27% reduction in intraocular pressure (IOP) for nine months after a single dose of only 1500 magnetically-steered hAMSCs, explained by increased conventional outflow facility and associated with higher TM cellularity.

• lines 45-6 (one-sentence summary): A novel magnetic cell therapy provided effective intraocular pressure reduction in mice, motivating future translational studies.

• lines 123-4 (introduction): Despite the absence of ocular hypertension in our MYOC^Y437H mice, our data demonstrate sustained IOP lowering and a significant benefit of magnetic cell steering in the eye, particularly for hAMSCs, strongly indicating further translational potential.

• line 207 (results): The observed reductions in IOP and increases in outflow facility after delivery of both cell types suggested functional changes in the conventional outflow pathway.

• line 509-10 (discussion): In summary, this work shows the effectiveness of our novel magnetic TM cell therapy approach for long-term IOP reduction through functional changes in the conventional outflow pathway.

It is very important to note that at the 23rd annual Trabecular Meshwork Study Club meeting (San Diego, December 2024), Dr. Zode, the lead author of reference 26 originally describing the transgenic myocilin mouse model, announced during his talk that this model no longer demonstrates the glaucomatous phenotype in his hands, which incidentally has motivated him to create a new, CRISPR MYOC mouse model. Dr. Zode also stated that he was uncertain of the reason for this loss of phenotype. His observation is consistent with our report. However, other investigators continue to observe the desired phenotype in their colonies of this mouse (Dr. Wei Zhu, personal communication). Continued use of this mouse model should therefore be approached with caution. 

Reviewer #2 (Public review):

Summary:

This observational study investigates the efficacy of intracameral injected human stem cells as a means to re-functionalize the trabecular meshwork for the restoration of intraocular pressure homeostasis. Using a murine model of glaucoma, human adiposederived mesenchymal stem cells are shown to be biologically safer and functionally superior at eliciting a sustained reduction in intraocular pressure (IOP). The authors conclude that the use of human adipose-derived mesenchymal stem cells has the potential for long-term treatment of ocular hypertension in glaucoma.

Strengths:

A noted strength is the use of a magnetic steering technique to direct injected stem cells to the iridocorneal angle. An additional strength is the comparison of efficacy between two distinct sources of stem cells: human adipose-derived mesenchymal vs. induced pluripotent cell derivatives. Utilizing both in vivo and ex vivo methodology coupled with histological evidence of introduced stem cell localization provides a consistent and compelling argument for a sustainable impact exogenous stem cells may have on the refunctionalization of a pathologically compromised TM.

Weaknesses:

A noted weakness of the study, as pointed out by the authors, includes the unanticipated failure of the genetic model to develop glaucoma-related pathology (elevated IOP, TM cell changes). While this is most unfortunate, it does temper the conclusion that exogenous human adipose derived mesenchymal stem cells may restore TM cell function. Given that TM cell function was not altered in their genetic model, it is difficult to say with any certainty that the introduced stem cells would be capable of restoring pathologically altered TM function. A restoration effect remains to be seen. 

We acknowledge that the phrase “restoration of TM function” is not fully supported by our results, given the absence of ocular hypertension in our animal model. Accordingly, we have revised the language to more precisely describe our findings. For specific details regarding these changes, please refer to our response to Reviewer 1’s public comments above.

Another noted complication to these findings is the observation that sham intracameralinjected saline control animals all showed elevated IOP and reduced outflow facility, compared to WT or Tg untreated animals, which allowed for more robust statistically significant outcomes. Additional comments/concerns that the authors may wish to address are elaborated in the Private Review section.

We agree that sham-injected animals tended to have higher average IOPs than transgenic animals in our study. However, these differences did not reach statistical significance and therefore remain inconclusive. Further, an increase in IOP following placebo injection has been previously reported (Zhu et al., 2016). 

Prompted by the Referee’s comments and also a private comment from Referee 1, we further investigated this effect by analyzing IOP in uninjected contralateral eyes at the mid-term time point and comparing the IOPs in these eyes to other cohorts, as now presented as additional data in Supplementary Tables 1 and 2 and Supplementary Figure 4 (see below). In brief, the uninjected contralateral transgenic eyes (10 months old) showed an IOP of 16.5 [15.9, 17.1] mmHg, which was intermediate between the IOP levels of the 6–7-month-old Tg group (15.4 [14.7, 16.1] mmHg) and the sham group (16.9 [15.5, 18.2] mmHg). However, none of these differences reached statistical significance. Additionally, we cannot rule out potential contralateral effects induced by the injections.

Regarding the best way to assess the effect of cell treatment, we feel very strongly that the most relevant IOP comparison is between cell-injected eyes and control (vehicle)-injected eyes, since this provides the most direct accounting for the effects of injection itself on IOP. Other comparisons, such as WT or untreated Tg eyes vs. cell-treated eyes, are interesting but harder to interpret. However, in response to the referee’s comment, we have added comparisons between cell-treated groups and untreated Tg eyes to Table 2, adjusting the post-hoc corrections accordingly. All hAMSC treated groups show statistically significant decrease in IOP even compared to Tg untreated eyes, while iPSC-TMs fail to reach such significance.

The following changes were made to the manuscript:

Lines 326 et seq.: Eyes subjected to saline injection exhibited marginally higher IOPs and lower outflow facilities on average, in comparison to the transgenic animals at baseline. However, due to the lack of statistical significance in these differences and the inherent age difference between the saline-injected animals and the non-injected controls at baseline, no conclusive inference can be drawn regarding the effect of saline injection. To investigate this phenomenon further, we also analyzed IOPs in uninjected contralateral eyes at the midterm time point (Supplementary Tables 1 and 2, Supplementary Figure 4). The uninjected contralateral transgenic eyes (10 months old) showed an IOP of 16.5 [15.9, 17.1] mmHg, which was intermediate between the IOP levels of the 6–7-month-old Tg group (15.4 [14.7, 16.1] mmHg) and the sham-injected group (16.9 [15.5, 18.2] mmHg). However, none of these differences reached statistical significance. Of note, contralateral hypertension has been previously reported after subconjunctival and periocular injection of dexamethasoneloaded nanoparticles (34), and we similarly cannot definitively rule out potential contralateral effects induced by our stem cell injections. Thus, we cannot draw any definite conclusions from these additional IOP comparisons at this time.

Reviewer #3 (Public review):

Summary:

The purpose of the current manuscript was to investigate a magnetic cell steering technique for efficiency and tissue-specific targeting, using two types of stem cells, in a mouse model of glaucoma. As the authors point out, trabecular meshwork (TM) cell therapy is an active area of research for treating elevated intraocular pressure as observed in glaucoma. Thus, further studies determining the ideal cell choice for TM cell therapy is warranted. The experimental protocol of the manuscript involved the injection of either human adipose derived mesenchymal stem cells (hAMSCs) or induced pluripotent cell derivatives (iPSC-TM cells) into a previously reported mouse glaucoma model, the transgenic MYOCY437H mice and wild-type littermates followed by the magnetic cell steering. Numerous outcome measures were assessed and quantified including IOP, outflow facility, TM cellularity, retention of stem cells, and the inner wall BM of Schlemm's canal.

Strengths:

All of these analyses were carefully carried out and appropriate statistical methods were employed. The study has clearly shown that the hAMSCs are the cells of choice over the iPSC-TM cells, the latter of which caused tumors in the anterior chamber. The hAMSCs were shown to be retained in the anterior segment over time and this resulted in increased cellular density in the TM region and a reduction in IOP and outflow facility. These are all interesting findings and there is substantial data to support it.

Weaknesses:

However, where the study falls short is in the MYOCY437H mouse model of glaucoma that was employed. The authors clearly state that a major limitation of the study is that this model, in their hands, did not exhibit glaucomatous features as previously reported, such as a significant increase in IOP, which was part of the overall purpose of the study. The authors state that it is possible that "the transgene was silenced in the original breeders". The authors did not show PCR, western blot, or immuno of angle tissue of the tg to determine transgenic expression (increased expression of MYOC was shown in the angle tissue of the transgenics in the original paper by Zode et al, 2011). This should be investigated given that these mice were rederived. Thus, it is clearly possible that these are not transgenic mice.

All MYOC mice that were used in this study were genotyped and confirmed to carry the transgene as noted in the original version of the paper (see lines 590-2). However, the transgene seems not to have been active, based on the lack of ocular hypertension as well as the lack of differences in supporting endpoints such as outflow facility and TM cellularity. While it would have been possible to carry out their recommended assays to investigate the root cause of this loss of phenotype this was not an objective of our study. Thus we instead here focus simply on communicating the observed loss of phenotype to readers. We also refer the referee to the final paragraph of our response to Referee 1. 

If indeed they are transgenics, the authors may want to consider the fact that in the Zode paper, the most significant IOP elevation in the mutant mice was observed at night and thus this could be examined by the authors. 

This is a good point. However, while the dark-phase IOP does exhibit a distinctly larger elevation (as previously observed in hypertonic saline sclerosis), Zode et al. also reported a notable 3 mmHg IOP increase during the light phase. The complete absence of such daytime (light phase) IOP elevation in our animals diminished our enthusiasm for pursuing darkphase IOP measurements. 

Other glaucomatous features of these mice could also have been investigated such as loss of RGCs, to further determine their transgenic phenotype. 

We agree that these other phenotypes could be studied, but in the absence of any detectable IOP elevation (and thus lack of mechanical insult on RGC axons), loss of RGC is extremely unlikely. We also note that the loss of retinal ganglion cells (RGCs) in the Myocilin model remains a subject of controversy. For example, despite a significant increase in IOP (>10 mmHg) in this model across four mouse strains, three, including C57BL6/J, did not exhibit any signs of optic nerve damage (McDowell et al., 2012). In contrast, Zhu et al. observed considerable nerve damage in this model, which was reversed following iPSC-TM cell transplantation (Zhu et al., 2016). Given these conflicting findings, we directed our efforts toward outcome measures directly related to aqueous humor dynamics.

Finally, while increased cellular density in the TM region was observed, proliferative markers could be employed to determine if the transplanted cells are proliferating.

We agree that identifying the source of the increased trabecular meshwork (TM) cellularity we observed is interesting and we plan to pursue that in future studies. 

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

The sham-injected transgenic animals showed elevated IOP 3-4 weeks after the baseline measurements in the transgenic mice. The authors justify this may be due to the increase in age in these animals. However, this seems unlikely due to the short duration of time between measurement of the baseline IOP and the Short time point (3-4 weeks). The authors do not provide IOP data for any WT sham injected eyes or naïve Tg eyes at these time points. These data are essential to determine if the elevation is due to the sham injection, age, or the transgene. Could it be that the IOP in this cohort of Tg mice didn't increase until 7-8 months of age instead of 6-7 months of age? The methods state only unilateral injections of the stem cells were done so it is assumed the contralateral eye was uninjected. What was the IOP in these eyes? These data would clarify the confusion in the data from sham-injected animals compared to baseline (naive) measurements.

We agree that the average IOP in saline-injected groups is higher than in WT or non-treated Tg mice, although the difference is inconclusive due to a lack of statistical significance. It is important to note, however, that this difference is subtle and not comparable to the 3 mmHg light-phase IOP elevation previously observed in this model (Zode et al., 2011). 

We appreciate the reviewer’s suggestion to include IOP data from the contralateral uninjected eyes, and we have now provided this information along with the comparative statistics in the supplementary materials. Additional details can be found in our response to a similar comment from Reviewer 2’s public review. In summary, the IOP difference in contralateral non-injected ten-month-old transgenic eyes was even smaller than in the original Tg group. IOP elevation following saline injection in mice has been reported previously (Zhu et al., 2016). As a potential confounding factor, we highlight possible contralateral effects of the injection itself (which is why we initially did not analyze IOP in the contralateral eyes).

The hAMSC-treated eyes appear to lower IOP even from baseline (although stats were only provided compared to the sham-injected eyes, which as stated above appear to have increased).

However, the iPSC-TM-treated eyes had IOPs equal to that of the baseline measurements taken 3 weeks prior. The significance is coming from the "sham-treated" eyes which had elevated IOPs. The controls listed above should be included to make these conclusions.

The reviewer makes an astute observation. Please refer to our response to a similar observation by Reviewer 2 under public reviews, where we provide and discuss the comparative statistics noted by the reviewer. However, we feel very strongly that the most relevant IOP comparison is between cell-injected eyes and control-injected eyes. 

If the transgenic mouse model truly did not have a phenotype, then the authors are testing the ability of the stem cells to lower IOP from baseline normal pressures. Therefore, the authors are not "restoring function of the conventional outflow pathway" as there is no damage to begin with. The language in the manuscript should be corrected to reflect this if the transgenics have no phenotype.

We agree and have adjusted the language accordingly. For further details, please refer to our response to your public review.

The authors noted in the iPSC-TM-treated eyes there was a high rate of tumorigenicity. If the magnetic steering of these cells is specific and targeted to the TM, why do the tumors form near the central iris?

While magnetic steering is more specific to the trabecular meshwork (TM) than previouslyused approaches (Bahrani Fard et al., 2023), it is not perfect, and a modest amount of offtarget delivery to the iris, including its central portion, still occurs. Apparently, it took only a few mis-directed iPSC-TM cells to lead to tumors in this work, which is a serious concern for future translational approaches. 

Reviewer #2 (Recommendations for the authors):

(1) It appears that mice were injected unilaterally (Line 590). I may have missed this, but was the companion un-injected eye analyzed in this study? If not analyzed, was there a confounding concern or limitation that necessitated omitting this possible control option?

Contralateral effects, such as hypertension in the untreated eye after subconjunctival and periocular injection of dexamethasone-loaded nanoparticles, have previously been reported in the literature (Li et al., 2019) and also reported anecdotally by other leaders in the field to the senior authors, which is why we did not initially analyze contralateral eyes in this study. However, prompted by this comment and others, we have now included the IOP measurements for contralateral uninjected ten-month-old transgenic eyes in the supplementary materials. For further details, please refer to our response to your public review.

(2) Were all these mice the same gender? Would gender be expected to alter the findings of this study?

Animals of both sexes were randomly chosen and included in the study. We added the following statement to the Materials and Methods section (line 530): After breeding and genotyping, mice, regardless of sex, were maintained to age 6-7 months, when transgenic animals were expected to have developed a POAG phenotype.

(3) As noted in the public review, the use of PBS for a control seems to have resulted in a slight elevation in IOP (Figure 2) as well as a reduction in outflow facility (Figure 3B) when compared to WT or Tg mice. Was this difference statistically significant? 

The differences between the sham (saline)-injected groups at any time point and untreated Tg mice did not reach statistical significance for IOP, facility, or TM cellularity and for facility, did not even show clear trends. For example, WT mice had, on average, 0.2 mmHg higher IOP and 0.6 nl/min/mmHg greater facility than the Tg group. Meanwhile on a similar scale, the long-term sham group exhibited 0.4 nl/min/mmHg higher facility compared to the Tg group. As the statistical tests indicate, these differences should be interpreted more as noise than meaningful signal. 

If so, then it should be noted as to whether the observed decrease in IOP following stem cell injection remained statistically significant when compared to these un-injected control animals. If significance was lost, then this should be appropriately noted and discussed. It is not apparently obvious why sham controls should have elevated IOP. This is a design and statistical concern.

Please refer to our response to a similar observation by Reviewer 1. We believe that comparing the treatment (cell suspension in saline) with its age-matched vehicle (saline) is the appropriate approach which maintains rigor by most directly accounting for the effects of injection. 

(4) The tonicity of the PBS used as a vehicle control was not stated and I did not see within the methods whether the stem cells were suspended using this same PBS vehicle. I assume isotonic phosphate buffered saline was used and that the stem cells were resuspended using the same sterile PBS. 

Thanks for catching this. We added “sterile PBS (1X, Thermo Fisher Scientific, Waltham, MA)” to the Methods section of the manuscript (line 567). 

With regards to using PBS as an injection control, I wonder if a better comparable control might have been to use mesenchymal stem cells that were rendered incapable of proliferating prior to intracameral injection. This, of course, addresses the unexplained mechanism(s) by which mesenchymal stem cells elicit a decrease in IOP.

This is an interesting idea, and represents another level of control. However, we explicitly chose not to use non-proliferating hAMSCs as a control, for several reasons. Firstly, a saline injection is the simplest control and in this initial study with multiple groups, we did not feel another experimental group should be added. Second, this control would not rule out paracrine effects from injected cells, which our data suggested are an important effect. Third, rendering injected cells truly non-proliferative could introduce unwanted/unknown phenotypes in these cells that would need to be carefully characterized. That being said, if an efficient method could be developed to render an entire population of these cells irreversibly non-proliferating, the reviewer’s suggestion would be worth pursuing to better understand the mechanism of TM cell therapies. 

(5) As noted in Figure 4C, TM cellular density as quantified was not altered in the sham control, so a loss of cellular density can not explain the elevated IOP with this group. Injecting viable (not determined?) mesenchymal stem cells did show, over the short term, a noted increase in TM cellular density. 

Thank you for noting this. We agree that changes in cell density do not explain the mild IOP elevation in the sham group. As the referee certainly is aware, there are multiple reasons that IOP can be elevated (changes in trabecular meshwork extracellular matrix, changes in trabecular meshwork stiffness) that are not necessarily related to cell density.  Since we do not know definitively the cause of this mild elevation, we would prefer to not speculate about it in the manuscript. 

Thanks for pointing out our omission of a statement about injected cell viability. We have now included the following statement in the Materials and Methods section (564-566): “For all the experiments where animals received hAMSC, cell count and >90% viability was verified using a Countess II Automated Cell Counter (Thermo Fisher Scientific, Waltham, MA).”

I'm confused, as clearly stated (Lines 431-432), mesenchymal stem cells accumulated close to, but not within, the TM. How is it that TM cellular density increased if these stem cells did not enter the TM? The authors may wish to clarify this distinction. Given that mesenchymal stem cells did not increase the risk of tumorigenicity, do the authors have any evidence that these cells actually proliferated post-injection or did they undergo senesce thereby displaying senescence-associated secretory phenotype as a source of paracrine support?

As the reviewer correctly noted, our observations show that hAMSCs primarily accumulated close to, but outside, the TM (likely caught up in the pectinate ligaments). Based on observations of increased TM cellularity, we think that the most likely explanation of these findings is paracrine signaling, as the reviewer suggests and which was discussed at length in the original version of the manuscript (lines 453-477). 

We agree that, despite observing little signal from hAMSCs within the TM, labeling with proliferation markers (e.g., Ki-67) and searching for co-localization with exogenous cells, and/or labeling for senescence markers would have provided more mechanistic information. This is an excellent topic for future study, which we plan to pursue, but was outside the scope of this study. 

(6) As noted in the public review, I think it is a bit of a stretch to even suggest that the findings of this study support stem cell restoration of TM function given that the model apparently did not produce TM cell dysfunction as anticipated. A restoration effect remains to be seen.

We agree and have adjusted the language accordingly. For further details, please refer to our response to Reviewer 1’s public comment.

Reviewer #3 (Recommendations for the authors):

(1) Show PCR, western blot, or immuno of angle tissue of the MYOC tg to confirm transgenic expression.

(2) Examine the IOP of mice at night.

(3) Investigate other glaucomatous features in the mice to determine if they have any of the transgenic phenotypes previously reported.

(4) Examine proliferative markers in the TM region of angles injected with stem cells.

Please see our responses to all four of these comments in the public section.

Bibliography (for this response letter only)

Bahrani Fard, M.R., Chan, J., Sanchez Rodriguez, G., Yonk, M., Kuturu, S.R., Read, A.T., Emelianov, S.Y., Kuehn, M.H., Ethier, C.R., 2023. Improved magnetic delivery of cells to the trabecular meshwork in mice. Exp. Eye Res. 234, 109602. https://doi.org/10.1016/j.exer.2023.109602

Li, G., Lee, C., Agrahari, V., Wang, K., Navarro, I., Sherwood, J.M., Crews, K., Farsiu, S., Gonzalez, P., Lin, C.-W., Mitra, A.K., Ethier, C.R., Stamer, W.D., 2019. In vivo measurement of trabecular meshwork stiffness in a corticosteroid-induced ocular hypertensive mouse model. Proc. Natl. Acad. Sci. U. S. A. 116, 1714–1722.

https://doi.org/10.1073/pnas.1814889116

Zhu, W., Gramlich, O.W., Laboissonniere, L., Jain, A., Sheffield, V.C., Trimarchi, J.M., Tucker, B.A., Kuehn, M.H., 2016. Transplantation of iPSC-derived TM cells rescues glaucoma phenotypes in vivo. Proc. Natl. Acad. Sci. 113, E3492–E3500.

Zode, G.S., Kuehn, M.H., Nishimura, D.Y., Searby, C.C., Mohan, K., Grozdanic, S.D., Bugge, K., Anderson, M.G., Clark, A.F., Stone, E.M., Sheffield, V.C., 2011. Reduction of ER stress via a chemical chaperone prevents disease phenotypes in a mouse model of primary open angle glaucoma. J. Clin. Invest. 121, 3542–3553. https://doi.org/10.1172/JCI58183

eLife Assessment

This important work combines self-report, neural and physiology data to examine the efficacy and mechanisms of counter conditioning versus extinction in reducing re-emergence of conditioned threat responses and show that this appears to rely on the nucleus accumbens rather than the ventromedial prefrontal cortex. These findings are supported by convincing evidence, though some areas could benefit from a few targeted refinements. The findings will be of interest to researchers across multiple subfields, including neuroscientists, cognitive theory researchers, and clinicians, particularly those with an interest in clinical applications in trauma therapies.

Reviewer #1 (Public review):

The authors attempted to replicate previous work showing that counterconditioning leads to more persistent reduction of threat responses, relative to extinction. They also aimed to examine the neural mechanisms underlying counterconditioning and extinction. They achieved both of these aims, and were able to provide some additional information, such as how counterconditioning impacts memory consolidation. Having a better understanding of which neural networks are engaged during counterconditioning may provide novel pharmacological targets to aid in therapies for traumatic memories. It will be interesting to follow up by examining the impact of varying amounts of time between acquisition and counterconditioning phases, to enhance replicability to real world therapeutic settings.

Major strengths

• This paper is very well written and attempts to comprehensively assess multiple aspects counterconditioning and extinction processes. For instance, the addition of memory retrieval tests is not core to the primary hypotheses, but provides additional mechanistic information on how episodic memory is impacted by counterconditioning. This methodical approach is commonly seen in animal literature, but less so in human studies.

• The Group x Cs-type x Phase repeated measure statistical tests with 'differentials' as outcome variables are quite complex, however the authors have generally done a good job of teasing out significant F test findings with post hoc tests and presenting the data well visually. It is reassuring that there is convergence between self-report data on arousal and valence and the pupil dilation response. Skin conductance is a notoriously challenging modality, so it is not too concerning that this was placed in the supplementary materials. Neural responses also occurred in logical regions with regards to reward learning.

• Strong methodology with regards to neuroimaging analysis, and physiological measures.

• The authors are very clear on documenting where there were discrepancies from their pre-registration and providing valid rationales for why.

Major Weaknesses

• The statistics showing that counterconditioning prevents differential spontaneous recovery are the weakest p values of the paper (and using one tailed tests, although this is valid due to directions being pre-hypothesised). This may be due to relatively small number of participants and some variability in responses.

Reviewer #2 (Public review):

Summary:

The present study sets out to examine the impact of counterconditioning (CC) and extinction on conditioned threat responses in humans, particularly looking at neural mechanisms involved in threat memory suppression. By combining behavioral, physiological, and neuroimaging (fMRI) data, the authors aim to provide a clear picture of how CC might engage unique neural circuits and coding dynamics, potentially offering a more robust reduction in threat responses compared to traditional extinction.

Strengths:

One major strength of this work lies in its thoughtful and unique design - integrating subjective, physiological, and neuroimaging measures to capture the variouse aspects of counterconditiong (CC) in humans. Additionally, the study is centered on a well-motivated hypothesis and the findings have potentials for improving the current understanding of pathways associated with emotional and cognitive control.

The data presentation is systematic, and the results on behavioral and physiological measures fit well with the hypothesized outcomes. The neuroimaging results also provide strong support for distinct neural mechanisms underlying CC versus extinction.

Weaknesses:

Overall, this study is a well-conducted and thought-provoking investigation into counterconditioning, with strong potential to advance our understanding of threat modulation mechanisms. Two minor weaknesses concern the scope and decisions regarding analysis choices. First, while the findings are solid, the topic of counterconditioning is relatively niche and may have limited appeal to a broader audience. Expanding the discussion to connect counterconditioning more explicitly to widely studied frameworks in emotional regulation or cognitive control would enhance the paper's accessibility and relevance to a wider range of readers. This broader framing could also underscore the generalizability and broader significance of the results. In addition, detailed steps in the statistical procedures and analysis parameters seem to be missing. This makes it challenging for readers to interpret the results in light of potential limitations given the data modality and/or analysis choices.

Comments on revisions: My previous concerns and questions have been sufficiently addressed.

Reviewer #3 (Public review):

In this manuscript, Wirz et al use neuroimaging (fMRI) to show that counterconditioning produces a longer lasting reduction in fear conditioning relative to extinction and appears to rely on the nucleus accumbens rather than the ventromedial prefrontal cortex. These important findings are supported by convincing evidence and will be of interest to researchers across multiple subfields, including neuroscientists, cognitive theory researchers, and clinicians.

In large part, the authors achieved their aims of giving a qualitative assessment of the behavioural mechanisms of counterconditioning versus extinction, as well as investigating the brain mechanisms. The results support their conclusions and give interesting insights into the psychological and neurobiological mechanisms of the processes that underlie the unlearning, or counteracting, of threat conditioning.

Strengths:

* Clearly written with interesting psychological insights<br /> * Excellent behavioural design, well-controlled and tests for a number of different psychological phenomena (e.g. extinction, recovery, reinstatement, etc).<br /> * Very interesting results regarding the neural mechanisms of each process.<br /> * Good acknowledgement of the limitations of the study.

Weaknesses:

* I am not sure that the memories tested were truly episodic<br /> * Twice as many female participants than males

Comments on revisions: I have no remaining concerns

Author response:

The following is the authors’ response to the original reviews

Public reviews:

Reviewer #1:

The authors attempted to replicate previous work showing that counterconditioning leads to more persistent reduction of threat responses, relative to extinction. They also aimed to examine the neural mechanisms underlying counterconditioning and extinction. They achieved both of these aims and were able to provide some additional information, such as how counterconditioning impacts memory consolidation. Having a better understanding of which neural networks are engaged during counterconditioning may provide novel pharmacological targets to aid in therapies for traumatic memories. It will be interesting to follow up by examining the impact of varying amounts of time between acquisition and counterconditioning phases, to enhance replicability to real-world therapeutic settings.

Major strengths

· This paper is very well written and attempts to comprehensively assess multiple aspects of counterconditioning and extinction processes. For instance, the addition of memory retrieval tests is not core to the primary hypotheses but provides additional mechanistic information on how episodic memory is impacted by counterconditioning. This methodical approach is commonly seen in animal literature, but less so in human studies.

· The Group x Cs-type x Phase repeated measure statistical tests with 'differentials' as outcome variables are quite complex, however, the authors have generally done a good job of teasing out significant F test findings with post hoc tests and presenting the data well visually. It is reassuring that there is a convergence between self-report data on arousal and valence and the pupil dilation response. Skin conductance is a notoriously challenging modality, so it is not too concerning that this was placed in the supplementary materials. Neural responses also occurred in logical regions with regard to reward learning.

· Strong methodology with regards to neuroimaging analysis, and physiological measures.

·The authors are very clear on documenting where there were discrepancies from their pre-registration and providing valid rationales for why.

We thank reviewer 1 for the positive feedback and for pointing out the strengths of our work. We agree that future research should investigate varying times between acquisition and counterconditioning to assess its success in real-life applications.

Major Weaknesses

(1) The statistics showing that counterconditioning prevents differential spontaneous recovery are the weakest p values of the paper (and using one-tailed tests, although this is valid due to directions being pre-hypothesized). This may be due to a relatively small number of participants and some variability in responses. It is difficult to see how many people were included in the final PDR and neuroimaging analyses, with exclusions not clearly documented. Based on Figure 3, there are relatively small numbers in the PDR analyses (n=14 and n=12 in counterconditioning and extinction, respectively). Of these, each group had 4 people with differential PDR results in the opposing direction to the group mean. This perhaps warrants mention as the reported effects may not hold in a subgroup of individuals, which could have clinical implications.

General exclusion criteria are described on page 17. We have added more detailed information on the reasons for exclusion (see page 17). All exclusions were in line with pre-registered criteria. For the analysis, the reviewer is referring to (PDR analysis that investigated whether CC can prevent the spontaneous recovery of differential conditioned threat responses), 18 participants were excluded from this analysis: 2 participants did not show evidence for successful threat acquisition as was already indicated on page 17, and 16 participants were excluded due to (partially) missing data. We now explicitly mention the exclusion of the additional 16 participants on page 7 and have updated Figure 3 to improve visibility of the individual data points. Therefore, for this analysis both experimental groups consisted of 15 participants (total N=30).

It is true that in both groups a few participants show the opposite pattern. Although this may also be due to measurement error, we agree that it is relevant to further investigate this in future studies with larger sample sizes. It will be crucial to identify who will respond to treatments based on the principles of standard extinction or counterconditioning. We have added this point in the discussion on page 14.

Reviewer #2:

Summary:

The present study sets out to examine the impact of counterconditioning (CC) and extinction on conditioned threat responses in humans, particularly looking at neural mechanisms involved in threat memory suppression. By combining behavioral, physiological, and neuroimaging (fMRI) data, the authors aim to provide a clear picture of how CC might engage unique neural circuits and coding dynamics, potentially offering a more robust reduction in threat responses compared to traditional extinction.

Strengths:

One major strength of this work lies in its thoughtful and unique design - integrating subjective, physiological, and neuroimaging measures to capture the various aspects of counterconditioning (CC) in humans. Additionally, the study is centered on a well-motivated hypothesis and the findings have the potential to improve the current understanding of pathways associated with emotional and cognitive control. The data presentation is systematic, and the results on behavioral and physiological measures fit well with the hypothesized outcomes. The neuroimaging results also provide strong support for distinct neural mechanisms underlying CC versus extinction.

We thank reviewer 2 for the feedback and for valuing the thoughtfulness that went into designing the study.

Weaknesses:

(1) Overall, this study is a well-conducted and thought-provoking investigation into counterconditioning, with strong potential to advance our understanding of threat modulation mechanisms. Two main weaknesses concern the scope and decisions regarding analysis choices. First, while the findings are solid, the topic of counterconditioning is relatively niche and may have limited appeal to a broader audience. Expanding the discussion to connect counterconditioning more explicitly to widely studied frameworks in emotional regulation or cognitive control would enhance the paper's accessibility and relevance to a wider range of readers. This broader framing could also underscore the generalizability and broader significance of the results. In addition, detailed steps in the statistical procedures and analysis parameters seem to be missing. This makes it challenging for readers to interpret the results in light of potential limitations given the data modality and/or analysis choices.

In this updated version of the manuscript, we included the notion that extinction has been interpreted as a form of implicit emotion regulation. In addition to our discussion on active coping (avoidance), we believe that our discussion has an important link to the more general framework of emotion regulation, while remaining within the scope of relevance. Please see pages 14 and 15 for the changes. In addition to being informative to theories of emotion regulation, our findings are also highly relevant for forms of psychotherapy that build on principles of counterconditioning (e.g. the use of positive reinforcement in cognitive behavioral therapy), as we point out in the introduction. We believe this relevance shows that counterconditioning is more than a niche topic. In line with the recommendation from reviewer 2, we added more details and explanations to the statistical procedures and analyses where needed (see responses to recommendations).

Reviewer #3:

Summary:

In this manuscript, Wirz et al use neuroimaging (fMRI) to show that counterconditioning produces a longer lasting reduction in fear conditioning relative to extinction and appears to rely on the nucleus accumbens rather than the ventromedial prefrontal cortex. These important findings are supported by convincing evidence and will be of interest to researchers across multiple subfields, including neuroscientists, cognitive theory researchers, and clinicians.

In large part, the authors achieved their aims of giving a qualitative assessment of the behavioural mechanisms of counterconditioning versus extinction, as well as investigating the brain mechanisms. The results support their conclusions and give interesting insights into the psychological and neurobiological mechanisms of the processes that underlie the unlearning, or counteracting, of threat conditioning.

Strengths:

· Mostly clearly written with interesting psychological insights

· Excellent behavioural design, well-controlled and tests for a number of different psychological phenomena (e.g. extinction, recovery, reinstatement, etc).

· Very interesting results regarding the neural mechanisms of each process.

· Good acknowledgement of the limitations of the study.

We thank reviewer 3 for the detailed feedback and suggestions.

Weaknesses:

(1) I think the acquisition data belongs in the main figure, so the reader can discern whether or not there are directional differences prior to CC and extinction training that could account for the differences observed. This is particularly important for the valence data which appears to differ at baseline (supplemental figure 2C).

Since our design is quite complex with a lot of results, we left the fear acquisition results as a successful manipulation check in the Supplementary Information to not overload the reader with information that is not the main focus of this manuscript. If the editor would like us to add the figure to the main text, we are happy to do so. During fear acquisition, both experimental groups showed comparable differential conditioned threat responses as measured by PDRs and SCRs. Subjective valence ratings indeed differed depending on CS category. Importantly, however, the groups only differed with respect to their rating to the CS- category, but not the CS+ category, which suggests that the strength of the acquired fear is similar between the groups. To make sure that these baseline differences cannot account for the differences in valence after CC/Ext, we ran an additional group comparison with differential valence ratings after fear acquisition added as a covariate. Results show that despite the baseline difference, the group difference in valence after CC/Ext is still significant (main effect Group: F_(1,43)=7.364, p=0.010, η²=0.146). We have added this analysis to the manuscript (see page 7).

(2) I was confused in several sections about the chronology of what was done and when. For instance, it appears that individuals went through re-extinction, but this is just called extinction in places.

We understand that the complexity of the design may require a clearer description. We therefore made some changes throughout the manuscript to improve understanding. Figure 1 is very helpful in understanding the design and we therefore refer to that figure more regularly (see pages 6-7). We also added the time between tasks where appropriate (e.g. see page 7). Re-extinction after reinstatement was indeed mentioned once in the manuscript. Given that the reinstatement procedure was not successful (see page 9), we could not investigate re-extinction and it is therefore indeed not relevant to explicitly mention and may cause confusion. We therefore removed it (see page 12).

(3) I was also confused about the data in Figure 3. It appears that the CC group maintained differential pupil dilation during CC, whereas extinction participants didn't, and the authors suggest that this is indicative of the anticipation of reward. Do reward-associated cues typically cause pupil dilation? Is this a general arousal response? If so, does this mean that the CSs become equally arousing over time for the CC group whereas the opposite occurs for the extinction group (i.e. Figure 3, bottom graphs)? It is then further confusing as to why the CC group lose differential responding on the spontaneous recovery test. I'm not sure this was adequately addressed.

Indeed, reward and reward anticipation also evoke an increase in pupil dilation. This was an important reason for including a separate valence-specific response characterization task. Independently from the conditioning task, this task revealed that both threat and reward-anticipation induced strong arousal-related PDRs and SCRs. This was also reflected in the explicit arousal ratings, which were stronger for both the shock-reinforced (negative valence) and reward-reinforced (positive valence) stimuli. Therefore, it is not surprising that reward anticipation leads to stronger PDRs for CS+ (which predict reward) compared to CS- stimuli (which do not predict reward) during CC, but is reduced during extinction due to a decrease in shock anticipation. During the spontaneous recovery test, a return of stronger PDRs for CS+ compared to CS- stimuli in the standard extinction group can only reflect a return of shock anticipation. Importantly, the CC group received no rewards during the spontaneous recovery task and was aware of this, so it is to be expected that the effect is weakened in the CC group. However, CS+ and CS- items were still rated of similar valence and PDRs did not differ between CS+ and CS- items in the CC group, whereas the Ext group rated the CS+ significantly more negative and threat responses to the CS+ did return. It therefore is reasonable to conclude that associating the CS+ with reward helps to prevent a return of threat responses. We have added some clarifications and conclusions to this section on page 8.

(4) I am not sure that the memories tested were truly episodic

In line with previous publications from Dunsmoor et al.[1-4], our task allows for the investigation of memory for elements of a specific episode. In the example of our task, retrieval of a picture probes retrieval of the specific episode, in which the picture was presented. In contrast, fear retrieval relies on the retrieval of the category-threat association, which does not rely on retrieval of these specific episodic elements, but could be semantic in nature, as retrieval takes place at a conceptual level. We have added a small note on what we mean with episodic in this context on page 4. We do agree that we cannot investigate other aspects of episodic memories here, such as context, as this was not manipulated in this experiment.

(5) Twice as many female participants than males

It is indeed unfortunate that there is no equal distribution between female and male participants. Investigating sex differences was not the goal of this study, but we do hope that future studies with the appropriate sample sizes are able to investigate this specifically. We have added this to the limitations of this study on page 17.

(6) No explanation as to why shocks were varied in intensity and how (pseudo-randomly?)

The shock determination procedure is explained on pages 18-19 (Peripheral stimulation). As is common in fear conditioning studies in humans (see references), an ascending staircase procedure was used. The goal of this procedure is to try and equalize the subjective experience of the electrical shocks to be “maximally uncomfortable but not painful”.

Recommendations for the authors:

Reviewer #1:

Very well written. No additional comments

We thank reviewer 1 for valuing our original manuscript version. To further improve the manuscript, we adapted the current version based on the reviewer’s public review (see response to reviewer #1 public review comment 1).

Reviewer #2:

(1) I feel that more justification/explanation is needed on why other regions highly relevant to different aspects of counterconditioning (e.g., threat, memory, reward processing) were not included in the analyses.

We first performed whole-brain analyses to get a general idea of the different neural mechanisms of CC compared to Ext. Clusters revealing significant group differences were then further investigated by means of preregistered ROI analyses. We included regions that have previously been shown to be most relevant for affective processing/threat responding (amygdala), memory (hippocampus), reward processing (NAcc) and regular extinction (vmPFC). We restricted our analyses to these most relevant ROIs as preregistered to prevent inflated or false-positive findings[5]. Beyond these preregistered ROIs, we applied appropriate whole-brain FEW corrections. The activated regions are listed in Supplementary Table 1 and include additional regions that were expected, such as the ACC and insula.

(2) Were there observed differences across participants in the experiment? Any information on variance in the data such as how individual differences might influence these findings would provide a richer understanding of counterconditioning and increase the depth of interpretation for a broad readership.

We agree that investigating individual differences is crucial to gain a better understanding of treatment efficacy in the framework of personalized medicine. Specifically, future research should aim to identify factors that help predict which treatment will be most effective for a particular patient. The results of this study provide a good basis for this, as we could show that the vmPFC in contrast to regular extinction, is not required in CC to improve the retention of safety memory. Therefore, this provides a viable option for patients who are not responding to treatments that rely on the vmPFC. In addition, as noted by Reviewer 1, in both groups a few participants show the opposite pattern (see Figure 3). It will be crucial to identify who will respond to treatments based on the principles of standard extinction or counterconditioning. We have added this point in the discussion on page 14.

(3) While most figures are informative and clear, Figure 3 would benefit from detailed axis labels and a more descriptive caption. Currently, it is challenging to navigate the results presented to support the findings related to differential PDRs. A supplementary figure consolidating key patterns across conditions might also further facilitate understanding of this rather complicated result.

We have made some changes to the figure to improve readability and understanding. Specifically, we changed the figure caption to “Change from last 2 trials CC/Ext to first 2 trials Spontaneous recovery test”, to give more details on what exactly is shown here. We also simplified the x-axis labels to “counterconditioning”, “recovery test” and “extinction”. With the addition of a clearer figure description, we hope to have improved understanding and do not think that another supplemental figure is needed.

(4) Additional details on the statistical tests are needed. For example, please clarify whether p-values reported were corrected across all experimental conditions. Also, it would be helpful for the authors to discuss why for example repeated measures ANOVA or mixed-effects conditions were not used in this study. Might those tests not capture variance across participants' PDRs and SCRs over time better?

We added that significant interactions were followed by Bonferroni-adjusted post-hoc tests where applicable (see page 21). We have used repeated measures ANOVAs to capture early versus late phases of acquisition and CC/extinction, as well as to compare late CC/extinction (last 2 trials) compared to early spontaneous recovery (first 2 trials) as is often done in the literature. A trial-level factor in a small sample would cost too many degrees of freedom and is not expected to provide more information. We have added this information and our reasoning to the methods section on page 21.

Reviewer #3:

(1) Suggest putting acquisition data into the main figures. In fact many of the supplemental figures could be integrated into the main figures in my opinion.

See response to reviewer #3 public review comment 1.

(2) Include explanations for why shock intensity was varied

See response to reviewer #3 public review comment 6.

(3) Include a better explanation for the change in differential responding from training to spontaneous recovery in the CC group (I think the loss of such responding in extinction makes more sense and is supported by the notion of spontaneous recovery, but I'm not sure about the loss in the CC group. There is some evidence from the rodent literature - which I am most familiar with - regarding a loss in contextual gradient across time which could account for some loss in specificity, could it be something like this?).

See response to reviewer #3 public review comment 3.

If we understand the reviewer correctly in that the we see a loss of differential responding due to a generalization to the CS-, this would imply an increase in responding to the CS-, which is not what we see. Our data should therefore be correctly interpreted as a loss of the specific response to the CS+ from the CC phase to the recovery test. Therefore, there is no spontaneous recovery in the CC group, and also not a non-specific recovery. To clarify this we relabeled Figure 3 by indicating “recovery test” instead of “spontaneous recovery”.

(4) Is there a possibility that baseline differences, particularly that in Supplemental Figure 2C, could account for later differences? If differences persist after some transformation (e.g. percentage of baseline responding) this would be convincing to suggest that it doesn't.

See response to reviewer #3 public review comment 1.

(5) As I mentioned, I got confused by the chronology as I read through. Maybe mention early on when reporting the spontaneous recovery results that testing occurred the next day and that participants were undergoing re-extinction when talking about it for the second time.

See response to reviewer #3 public review comment 2.

(6) Page 8 - I was confused as to why it is surprising that the CC group were more aroused than the extinction group, the latter have not had CSs paired with anything with any valence, so doesn't this make sense? Or perhaps I am misunderstanding the results - here in text the authors refer back to Figure 2B, but I'm not sure if this is showing data from the spontaneous recovery test or from CC/extinction. If it is the latter, as the caption suggests, why are the authors referring to it here?

Participants in the CC group showed increased differential self-reported arousal after CC, whereas arousal ratings did not differ between CS+ and CS- items after extinction. We interpret this in line with the valence and PDR results as an indication of reward-induced arousal. At the start of the next day, however, participants from the CC and extinction groups gave comparable ratings. It may therefore be surprising why participants in the CC group do not still show stronger ratings since nothing happened between these two ratings besides a night’s sleep (see design overview in Figure 1A). We removed the “suprisingly” to prevent any confusion.

(7) I suggest that the authors comment on whether there were any gender differences in their results.

See response to reviewer #3 public review comment 5.

(8) The study makes several claims about episodic memory, but how can the authors be sure that the memories they are tapping into are episodic? Episodic has a very specific meaning - a biographical, contextually-based memory, whereas the information being encoded here could be semantic. Perhaps a bit of clarification around this issue could be helpful.

See response to reviewer #3 public review comment 4.

References

(1) Dunsmoor, J. E. & Kroes, M. C. W. Episodic memory and Pavlovian conditioning: ships passing in the night. Curr Opin Behav Sci 26, 32-39 (2019). https://doi.org/10.1016/j.cobeha.2018.09.019

(2) Dunsmoor, J. E. et al. Event segmentation protects emotional memories from competing experiences encoded close in time. Nature Human Behaviour 2, 291-299 (2018). https://doi.org/10.1038/s41562-018-0317-4

(3) Dunsmoor, J. E., Murty, V. P., Clewett, D., Phelps, E. A. & Davachi, L. Tag and capture: how salient experiences target and rescue nearby events in memory. Trends Cogn Sci 26, 782-795 (2022). https://doi.org/10.1016/j.tics.2022.06.009

(4) Dunsmoor, J. E., Murty, V. P., Davachi, L. & Phelps, E. A. Emotional learning selectively and retroactively strengthens memories for related events. Nature 520, 345-348 (2015). https://doi.org/10.1038/nature14106

(5) Gentili, C., Cecchetti, L., Handjaras, G., Lettieri, G. & Cristea, I. A. The case for preregistering all region of interest (ROI) analyses in neuroimaging research. Eur J Neurosci 53, 357-361 (2021). https://doi.org/10.1111/ejn.14954

eLife Assessment

This study presents valuable findings on the relative cerebral blood volume of non-human primates that move us closer to uncovering the functional and architectonic principles that govern the interplay between neuronal and vascular networks. The evidence of areal variations and of vessel counting and laminar analysis is solid. The lack of a direct comparison of their approach against better-established MRI-based methods for measuring hemodynamics and vascular structure somewhat weakens the evidence provided in the current paper version, but the current work is an significant step forward. The work will be of interest to NHP imaging scientists.

Reviewer #1 (Public review):

Summary:

Audio et al. present an interesting study examining cerebral blood volume (CBV) across cortical areas and layers in non-human primates (NHPs) using high-resolution MRI. While with contrast agents are frequently employed to improve fMRI sensitivity in NHP research, its application for characterizing baseline CBV distribution is less common. This study quantifies large-vessel distribution as well as regional and laminar CBV variations, comparing them with other metrics.

Strengths:

(1) Noninvasive mapping of relative cerebral blood volume is novel for non-human primates.<br /> (2) A key finding was the observation of variations in CBV across regions; primary sensory cortices had high CBV, whereas other higher areas had low CBV.<br /> (3) The measured relative CBV values correlated with previously reported neuronal and receptor densities, potentially providing valuable physiological insights.

Weaknesses:

(1) A weakness of this manuscript is that the quantification of CBV with postprocessing approaches to remove susceptibility effects from pial and penetrating vessels is not fully validated, especially on a laminar scale.<br /> (2) High-resolution MRI with a critical sampling frequency estimated from previous studies (Weber 2008, Zheng 1991) was performed to separate penetrating vessels. However, this approach depends on multiple factors, including spatial resolution, contrast agent dosage, and data processing methods. This raises concerns about the generalizability of these findings to other experimental setups or populations.<br /> (3) Baseline R2* is sensitive to baseline R2, vascular volume, iron content, and susceptibility gradients. Additionally, it is sensitive to imaging parameters; higher spatial resolution tends to result in lower R2* values (closer to the R2 value). Although baseline R2* correlates with several physiological parameters, drawing direct physiological inferences from it remains challenging.<br /> (4) CBV-weighted deltaR2*, which depends on both CBV and contrast agent dose, correlates with various metrics (cytoarchitectural parcellation, myelin/receptor density, cortical thickness, CO, cell-type specificity, etc.). While such correlations may be useful for exploratory analyses, all comparisons depend on measurement accuracy. A fundamental question remains whether CBV-weighted ΔR2* can provide reliable and biologically meaningful insights into these metrics, particularly in diseased or abnormal brain states.

Reviewer #2 (Public review):

Summary:

This manuscript presents a new approach for non-invasive, MRI-based, measurements of cerebral blood volume (CBV). Here, the authors use ferumoxytol, a high-contrast agent and apply specific sequences to infer CBV. The authors then move to statistically compare measured regional CBV with known distribution of different types of neurons, markers of metabolic load and others. While the presented methodology captures and estimated 30% of the vasculature, the authors corroborated previous findings regarding lack of vascular compartmentalization around functional neuronal units in the primary visual cortex.

Strengths:

Non-invasive methodology geared to map vascular properties in vivo.

Implementation of a highly sensitive approach for measuring blood volume.

Ability to map vascular structural and functional vascular metrics to other types of published data.

Weaknesses:

The key issue here is the underlying assumption about the appropriate spatial sampling frequency needed to captures the architecture of the brain vasculature. Namely, ~7 penetrating vessels / mm2 as derived from Weber et al 2008 (Cer Cor). The cited work, begins by characterizing the spacing of penetrating arteries and ascending veins using vascular cast of 7 monkeys (Macaca mulatta, same as in the current paper). The ~7 penetrating vessels / mm2 is computed by dividing the total number of identified vessels by the area imaged. The problem here is that all measurements were made in a "non-volumetric" manner and only in V1. Extrapolating from here to other brain areas is therefore not possible without further exploration with independent methodologies.

Please note that these are comments on the revised version.

Author response:

The following is the authors’ response to the previous reviews

Reviewer #1 (Public review):

Summary:

Audio et al. measured cerebral blood volume (CBV) across cortical areas and layers using high-resolution MRI with contrast agents in non-human primates. While the non-invasive CBV MRI methodology is often used to enhance fMRI sensitivity in NHPs, its application for baseline CBV measurement is rare due to the complexities of susceptibility contrast mechanisms. The authors determined the number of large vessels and the areal and laminar variations of CBV in NHP, and compared those with various other metrics.

Strengths:

Noninvasive mapping of relative cerebral blood volume is novel for non-human primates. A key finding was the observation of variations in CBV across regions; primary sensory cortices had high CBV, whereas other higher areas had low CBV. The measured CBV values correlated with previously reported neuronal and receptor densities.

We appreciate your recognition of the novelty of our non-invasive relative cerebral blood volume (CBV) mapping in non-human primates, as well as the observed areal variations and their correlations with neuronal and receptor densities. However, we are concerned that key contributions of our work—such as cortical layer-specific vasculature mapping and benchmarking surface vessel density estimations against anatomical ground truth—are being framed as limitations rather than significant advances in the field pushing the boundaries of current neuroimaging capabilities and providing a valuable foundation for future research. Additionally, we would like to clarify that dynamic susceptibility contrast (DSC) MRI using gadolinium is the gold standard for CBV measurement in clinical settings and the argument that “baseline CBV measurements are rare due to the complexities of susceptibility contrast” is simply not true. The limited use of ferumoxytol for CBV imaging is primarily due to previous FDA regulatory restrictions, rather than inherent methodological shortcomings.

Changes in text:

Compared to clinically used gadolinium-based agents, ferumoxytol's substantially longer half-life and stronger R₂* effect allows for higher-resolution and more sensitive vascular volume measurements (Buch et al., 2022), albeit these methodologies are hampered by confounding factors such as vessel orientation relative to the magnetic field (B₀) direction (Ogawa et al., 1993).

Weaknesses:

A weakness of this manuscript is that the quantification of CBV with postprocessing approaches to remove susceptibility effects from pial and penetrating vessels is not fully validated, especially on a laminar scale. Further specific comments follow.

(1) Baseline CBV indices were determined using contrast agent-enhanced MRI (deltaR₂*). Although this approach is suitable for areal comparisons, its application at a laminar scale poses challenges due to significant contributions from large vessels including pial vessels. The primary concern is whether large-vessel contributions can be removed from the measured deltaR₂* through processing techniques.

Eliminating the contribution of large vessels completely is unlikely, and we agree with the reviewer that ΔR₂* results likely reflect a weighted combination of signals from both large vessels and capillaries. However, the distribution of ΔR₂* more closely aligns with capillary density in areas V1–V5 than with large vessel distributions (Weber et al., 2008), suggesting that our ΔR₂* results are more weighted toward capillaries. Moreover, we demonstrated that the pial vessel induced signal-intensity drop-outs are clearly limited to the superficial layers and exhibit smaller spatial extent than generally thought (Supp. Figs. 2 and 4).

(2) High-resolution MRI with a critical sampling frequency estimated from previous studies (Weber 2008, Zheng 1991) was performed to separate penetrating vessels. However, this approach is still insufficient to accurately identify the number of vessels due to the blooming effects of susceptibility and insufficient spatial resolution. The reported number of penetrating vessels is only applicable to the experimental and processing conditions used in this study, which cannot be generalized.

Our intention was not to suggest that our measurements provide a general estimate of vessel density across the macaque cerebral cortex. At 0.23 mm isotropic resolution, we successfully delineated approximately 30% of the penetrating vessels in V1. Our primary objective was to demonstrate a proof-of-concept quantifiable measurement rather than to establish a generalized vessel density metric for all brain regions. We have consistently emphasized this throughout the manuscript, but if there is a specific point of misunderstanding, we would be happy to consider revisions for clarity.

(3) Baseline R₂* is sensitive to baseline R₂, vascular volume, iron content, and susceptibility gradients. Additionally, it is sensitive to imaging parameters; higher spatial resolution tends to result in lower R₂* values (closer to the R₂ value). Thus, it is difficult to correlate baseline R₂* with physiological parameters.

The observed correlation between R₂* and neuron density is likely indirect, as R₂* is strongly influenced by iron, myelin, and deoxyhemoglobin densities. However, the robust correlation between R₂* and neuron density, peaking in the superficial layers (R = 0.86, p < 10^-10), is striking and difficult to ignore (revised Supp. Fig. 6D-E). Upon revision, we identified an error in Supp. Fig. 6D-E, where the previous version used single-subject R₂* and ΔR₂* maps instead of the group-averaged maps. The revised correlations are slightly stronger than in the earlier version.

Given that the correlation between neuron density and R₂* is strongest in the superficial layers, we suggest this relationship reflects an underlying association with tissue cytochrome oxidase (CO) activity and cumulative effect of deoxygenated venous blood drainage toward the pial network. The superficial cortical layers are also less influenced by myelin and iron densities, which are more concentrated in the deeper cortical layers. Additional factors may contribute to this relationship, including the iron dependence of mitochondrial CO activity, as iron is an essential component of CO’s heme groups. Moreover, myelin maintenance depends on iron, which is predominantly stored in oligodendrocytes. The presence of myelinated thin axons and a higher axonal surface density may, in turn, be a prerequisite for high neuron density.

In this context, it is also valuable to note the absolute range of superficial R₂* values (≈ 6 s^-1; Supp. Fig. 6D). This variation in cortical surface R₂* is about 12-30 times larger compared to the signal changes observed during task-based fMRI (6 vs. 0.2-0.5 s^-1). This relation seems reasonable because regional increases in absolute blood flow associated with imaging signals, as measured by PET, typically do not exceed 5%–10% of the brain's resting blood flow (Raichle and Mintum 2016; Brain work and brain imaging). The venous oxygenation level is typically 60%, with task-induced activation increasing it by only a few percent. We suggest that this is ~40% oxygen extraction is reflected in the superficial R₂*. Finally, the large intercept (≈ 14.5 1/s; Supp. Fig. 6D), which is not equivalent to the water R₂* (≈ 1 1/s), suggests that R₂* is influenced by substantial non-neuron density factors, such as receptor, myelin, iron, susceptibility gradients and spatial resolution.

The R₂* values are well known to be influenced by intra-voxel phase coherence and thus spatial resolution. However, our view is that the proposed methodology of acquiring cortical-layer thickness adjusted high-resolution (spin-echo) R₂ maps poses more methodological limitations and is less practical. Notwithstanding, to further corroborate the relationship between R₂* and neuron density, we investigated whether a similar correlation exists in non-quantitative T2w SPACE-FLAIR images (0.32 mm isotropic) signal-intensity and neuron density. Using B₁ bias-field and B₀ orientation bias corrected T2w SPACE-FLAIR images (N=7), we parcellated the equivolumetric surface maps using Vanderbilt sections. Our findings showed that signal intensity—where regions with high signal intensity correspond to low R₂ values, and areas with low signal intensity correspond to high R₂ values—was positively correlated with neuron density, particularly in the superficial layers (R = 0.77, p = 10^-11; Author response image 1).This analysis confirmed the correlation with neuron density and R₂ peaks at superficial layers. However, this correlation was slightly weaker compared to quantitative R₂* (Supp. Fig. 6D), suggesting the variable flip-angle spin-echo train refocused signal-phase coherence loss from large draining vessels or that non-quantitative T2w-FLAIR images may be confounded by other factors such as B₁ transmission field biases (Glasser et al., 2022). Notwithstanding, this non-quantitative fast spin-echo with variable flip-angles approach, which is in principle less dependent on image resolution and closer to R_2,intrinsic than R₂*, yields similar findings in comparison to quantitative gradient-echo.

Author response image 1.

(A) T2w-FLAIR SPACE normalized signal-intensity plotted vs neuron density. Note that low signal-intensity corresponds to high R₂ and high neuron density, consistent with findings using ME-GRE. (B) Correlation between T2w-FLAIR SPACE and neuron density across equivolumetric layers. Notably, a similar relationship with neuron density was observed using a variable spin-echo pulse sequence as with quantitative gradient-echo-based imaging.

Changes in text:

Results:

“Because the Julich cortical area atlas covers only a section of the cerebral cortex, and the neuron density estimates are interpolated maps, we extended our analysis using the original Collins sample borders encompassing the entire cerebral cortex (Supp. Fig. 6A-C). This analysis reaffirmed the positive correlation with ΔR₂* (peak at EL2, R = 0.80, p < 10^-11) and baseline R₂* (peak at EL2a, R = 0.86, p < 10^-13), yielding linear coefficients of ΔR₂* = 102 × 10³ neurons/s and R₂* = 41 × 10³ neurons/s (Supp. Fig. 6D-G). This suggests that the sensitivity of quantitative layer R₂* MRI in detecting neuronal loss is relatively weak, and the introduction of the Ferumoxytol contrast agent has the potential to enhance this sensitivity by a factor of 2.5.”

A new paragraph was added into discussion section 4.3 corroborating the relation between R₂* and neuron density:

“Another key finding of this study was the strong correlation between baseline R₂* and neuron density (Supp. Fig. 6D, E). While R₂* is well known to be influenced by iron, myelin, and deoxyhemoglobin densities, this correlation peaks in the superficial layers (Supp. Fig. 6E), suggesting a link to CO activity and the accumulation of deoxygenated venous blood draining from all cortical layers toward the pial network. Notably, the absolute range of superficial R₂* values (max - min ≈ 6 s^-1; Supp. Fig. 6D) is approximately 12-30 times larger than the ΔR₂* observed during task-based BOLD fMRI at 3T (0.2-0.5 1/s) (Yablonskiy and Haacke 1994). Since venous oxygenation is around 60% and task-induced changes in blood flow account for only 5%–10% of the brain's resting blood flow (Raichle & Mintun, 2006), these results suggest that superficial R₂* (Fig. 1D) may serve as a more accurate proxy for total deoxyhemoglobin content (and thus total oxygen consumption), which scales with the neuron density of the underlying cortical gray matter. Importantly, superficial layers may also provide a more specific measure of deoxyhemoglobin, as they are less influenced by myelin and iron, which are more concentrated in deeper cortical layers. Additionally, smaller but direct contributors, such as mitochondrial CO density—an iron-dependent factor—may also play a role in this relationship.”

References:

Raichle, M.E., Mintun, M.A., 2006. BRAIN WORK AND BRAIN IMAGING. Annu. Rev. Neurosci. 29, 449–476. https://doi.org/10.1146/annurev.neuro.29.051605.112819

(4) CBV-weighted deltaR₂* is correlated with various other metrics (cytoarchitectural parcellation, myelin/receptor density, cortical thickness, CO, cell-type specificity, etc.). While testing the correlation between deltaR₂* and these other metrics may be acceptable as an exploratory analysis, it is challenging for readers to discern a causal relationship between them. A critical question is whether CBV-weighted deltaR₂* can provide insights into other metrics in diseased or abnormal brain states.

We acknowledge that having multivariate analysis using dense histological maps would be valuable to establish causality among these several metrics:

“To comprehensively understand the factors contributing to the vascular organization of the brain, experimental disentanglement through multivariate analysis of laminar cell types and receptor densities is needed (Hayashi et al., 2021, Froudist-Walsh et al., 2023). Moreover, employing more advanced statistical modeling, including considerations for synapse-neuron interactions, may be important for refined evaluations.”

We think the primary contributors to the brain's energy budget are neurons and receptors, as shown in several references and stated in the manuscript. To investigate relationship between neuron density and CBV, we estimated the energy budget allocated to neurons and extrapolated the remaining CBV to other contributing factors:

Changes in text:

“However, this is a simplified estimation, and a more comprehensive assessment would need to account for an aggregate of biophysical factors such as neuron types, neuron membrane surface area, firing rates, dendritic and synaptic densities (Fig. 6F-G), neurotransmitter recycling, and other cell types (Kageyama 1982; Elston and Rose 1997; Perge et al., 2009; Harris et al., 2012). Indeed, the majority of the mitochondria reside in the dendrites and synaptic transmission is widely acknowledged to drive the majority of the energy consumption and blood flow (Wong-Riley, 1989; Attwell et al., 2001).

Extrapolating cortical ΔR₂* to zero neuron density results in a large intercept (~35 1/s), corresponding to 60% of the maximum cortical CBV (57 1/s; Supp. Fig. 6F). This supports the view that the majority of energy consumption occurs in the neuropil—comprising dendrites, synapses, and axons—which accounts for ~80–90% of cortical gray matter volume, whereas neuronal somata constitute only ~10–20% (Wong-Riley, 1989). Although neuronal cell bodies exhibit higher CO activity per unit volume due to their dense mitochondrial content, these results suggest their overall contribution to the total CBV per mm³ tissue remains lower than that of the neuropil, given the latter's substantially larger volume fraction in cortical tissue.

Contrary to our initial expectations, we observed a relatively smaller CBV in regions and layers with high receptor density (Fig. 6B, D, F). This relationship extends to other factors, such as number of spines (putative excitatory inputs) and dendrite tree size across the entire cerebral cortex (Supp. Fig. 7) (Froudist-Walsh et al., 2023, Elston 2007). These results align with the work of Weber and colleagues, who reported a similar negative correlation between vascular length density and synaptic density, as well as a positive correlation with neuron density in macaque V1 across cortical layers (Weber et al., 2008).”

Variations in neurons and receptors are reflected in cytoarchitecture, myelin (axon density likely scales with neuron density and myelin inhibits synaptic connections), and cell-type composition. For example, fast-spiking parvalbumin interneurons, which target the soma or axon hillock, are well-suited for regulating activity in regions with high neuron density, whereas bursting calretinin interneurons, which target distal dendrites, are more adapted to areas with high synaptic density. These factors in turn, gradually change along the cortical hierarchy level (higher levels have thinner cortical layer IV, more complex dendrite trees and more numerous inter-areal connectivity patterns). In our view, these factors are tightly interlinked and explain the strong correlations and metabolic demands observed across different metrics.

We also agree that cortical layer imaging of vasculature in diseased or abnormal brain states is an intriguing direction for future research; however, it falls beyond the scope of the present study.

Reviewer #2 (Public review):

Summary:

This manuscript presents a new approach for non-invasive, MRI-based, measurements of cerebral blood volume (CBV). Here, the authors use ferumoxytol, a high-contrast agent and apply specific sequences to infer CBV. The authors then move to statistically compare measured regional CBV with known distribution of different types of neurons, markers of metabolic load and others. While the presented methodology captures and estimated 30% of the vasculature, the authors corroborated previous findings regarding lack of vascular compartmentalization around functional neuronal units in the primary visual cortex.

Strengths:

Non invasive methodology geared to map vascular properties in vivo.

Implementation of a highly sensitive approach for measuring blood volume.

Ability to map vascular structural and functional vascular metrics to other types of published data.

Weaknesses:

The key issue here is the underlying assumption about the appropriate spatial sampling frequency needed to captures the architecture of the brain vasculature. Namely, ~7 penetrating vessels / mm2 as derived from Weber et al 2008 (Cer Cor). The cited work, begins by characterizing the spacing of penetrating arteries and ascending veins using vascular cast of 7 monkeys (Macaca mulatta, same as in the current paper). The ~7 penetrating vessels / mm2 is computed by dividing the total number of identified vessels by the area imaged. The problem here is that all measurements were made in a "non-volumetric" manner and only in V1. Extrapolating from here to the entire brain seems like an over-assumption, particularly given the region-dependent heterogeneity that the current paper reports.

We appreciate the reviewer’s concerns regarding spatial sampling frequency and its implications for characterizing brain vasculature, which we investigated in this study. To clarify, our analysis of surface vessel density was explicitly restricted to V1 precisely due to the limitations of our experimental precision. While we reported the total number of vessels identified in the cortex, we intentionally chose not to present density values across regions in this manuscript. Although these calculations are feasible, we focused on the data directly analyzed and avoided extrapolating density values beyond the scope of our findings. Thus, we are uncertain about the suggestion that we extrapolated vessel density values across the entire brain, as we have taken care to limit our conclusions of our vessel density precision to V1.

Regarding methodology, we conducted two independent analyses of vessel density specifically in V1. The first involved volumetric analysis using the Frangi filter, while the second used surface-based analysis of local signal-intensity gradients (as illustrated in Fig. 2E and Supp. Figs. 3 and 4), albeit the final surface density analysis is performed using the ultra-high resolution equivolumetric layers. Notably, these two approaches produced consistent and comparable vessel density estimates, supporting the reliability of our findings within the scope of V1 (we found 30% of the vessels relative to the ground-truth).

Comments on revisions:

I appreciate the effort made to improve the manuscript. That said, the direct validation of the underlying assumption about spatial resolution sampling remains unaddressed in the final version of this manuscript. With the only intention to further strengthen the methodology presented here, I would encourage again the authors to seek a direct validation of this assumption for other brain areas.

In their reply, the authors stated "... line scanning or single-plane sequences, at least on first impression, seem inadequate for whole-brain coverage and cortical surface mapping. ". This seems to emanate for a misunderstanding as the method could be used to validate the mapping, not to map per-se.

We apologize for any misunderstanding in our previous response and appreciate your clarification. We now understand that you were suggesting the use of line-scanning or single-plane sequences as a method to validate, rather than map, our spatial sampling assumptions.

We agree that single-plane sequences at very high in-plane resolution (e.g., 50 × 50 × 1000 µm) have great potential to detect penetrating vessels and even vessel branching patterns. These techniques could indeed provide valuable insights into region-specific vessel density variations which could then be used to validate whole brain 3D acquisitions. However, as noted above, we have refrained from reporting vessel densities outside V1 precisely due to sampling limitations (we only found 30% of the penetrating vessels in V1, or only 2 mm²/30mm² ≈ 7% of branching vessel ground-truth, see discussion).

We acknowledge the merit of incorporating such methods to validate regional vessel densities and agree that this would be an important avenue for future research. Thank you for suggesting this point, we have briefly mentioned the advantage of single-plane EPI at discussion.

Changes in text:

“4.1 Methodological considerations - vessel density informed MRI

…anatomical studies accounting for branching patterns have reported much higher vessel densities up to 30 vessels/mm² (Keller et al., 2011; Adams et al., 2015). Further investigations are warranted, taking into account critical sampling frequencies associated with vessel branching patterns (Duverney 1981), and achieving higher SNR through ultra-high B₀ MRI (Bolan et al., 2006; Harel et al., 2010; Kim et al., 2013) and utilize high-resolution single-plane sequences and prospective motion correction schemes to accurately characterize regional vessel densities. Such advancements hold promise for improving vessel quantification, classifications for veins and arteries and constructing detailed cortical surface maps of the vascular networks which may have diagnostic and neurosurgical utilities (Fig. 2A, B) (Iadecola, 2013; Qi and Roper, 2021; Sweeney et al., 2018).”

During the revision we found a typo and corrected it in Supp. Fig. 8: Dosal -> Dorsal.

eLife Assessment

This study explores how heterozygosity for specific neurodevelopmental disorder-associated TRIO variants affects brain function in mice. The authors conducted thorough analyses on mouse lines harboring TRIO-variants associated with autism spectrum disorder, schizophrenia, and bipolar disorder, and the results provide compelling evidence demonstrating unique alterations of each variant in synaptic functions and behavior. These findings highlight a fundamental aspect of TRIO variants contributing to brain functions and neuropsychiatric disorders.

Reviewer #1 (Public review):

Summary:

This study explores how heterozygosity for specific neurodevelopmental disorder-associated Trio variants affects mouse behavior, brain structure, and synaptic function, revealing distinct impacts on motor, social, and cognitive behaviors linked to clinical phenotypes. Findings demonstrate that Trio variants yield unique changes in synaptic plasticity and glutamate release, highlighting Trio's critical role in presynaptic function and the importance of examining variant heterozygosity in vivo.

Strengths:

This study generated multiple mouse lines to model each Trio variant, reflecting point mutations observed in human patients with developmental disorders. The authors employed various approaches to evaluate the resulting behavioral, neuronal morphology, synaptic function, and proteomic phenotypes.

Reviewer #2 (Public review):

Summary:

The authors generated three mouse lines harboring ASD, Schizophrenia, and Bi-polar-associated variants in the TRIO gene. Anatomical, behavioral, physiological, and biochemical assays were deployed to compare and contrast the impact of these mutations in these animals. In this undertaking the authors sought to identify and characterize the cellular and molecular mechanisms responsible for ASD, Schizophrenia, and Bi-polar disorder development.

Strengths:

The establishment of TRIO dysfunction in the development of ASD, Schizophrenia, and Bi-polar disorder is very recent and of great interest. Disorder-specific variants have been identified in the TRIO gene, and this study is the first to compare and contrast the impact of these variants in vivo in preclinical models. The impact of these mutations was carefully examined using an impressive host of methods. The authors achieved their goal of identifying behavioral, physiological, and molecular alterations that are disorder/variant specific. The impact of this work is extremely high given the growing appreciation of TRIO dysfunction in a large number of brain-related disorders. This work is very interesting in that it begins to identify the unique and subtle ways brain function is altered in ASD, Schizophrenia, and Bi-polar disorder.

Weaknesses:

(1) Most assays were performed in older animals and perhaps only capture alterations that result from homeostatic changes resulting from prodromal pathology that may look very different.

(2) Identification of upregulated (potentially compensating) genes in response to these disorder specific Trio variants is extremely interesting. However, a functional demonstration of compensation is not provided.

(3) There are instances where data is not shown in the manuscript. See "data not shown". All data collected should be provided even if significant differences are not observed.

I consider weaknesses 1 and 2 minor. While they would very interesting to explore, these experiments might be more appropriate for a follow up study. The missing data in 3 should be provided in the supplemental material.

Revised Manuscript:

All of my above concerns were well addressed by the authors in the revised submission.

Author response:

The following is the authors’ response to the original reviews

eLife Assessment

This study provides useful findings about the effects of heterozygosity for Trio variants linked to neurodevelopmental and psychiatric disorders in mice. However, the strength of the evidence is limited and incomplete mainly because the experimental flow is difficult to follow, raising concerns about the conclusions' robustness. Clearer connections between variables, such as sex, age, behavior, brain regions, and synaptic measures, and more methodological detail on breeding strategies, test timelines, electrophysiology, and analysis, are needed to support their claims.

We appreciate the opportunity to address the constructive feedback provided by eLife and the reviewers. Below, we respond to the overall assessment and individual reviewers' comments, clarifying our experimental approach, addressing concerns, and providing additional details where necessary.

We thank the editors for highlighting the significance of our findings regarding the effects of Trio variant heterozygosity in mice. We acknowledge the feedback concerning the experimental flow and agree that clarity is paramount. To address these concerns:

(1) Connections between variables: The word limit of the initial submission constrained our ability to provide adequate details and connections between variables. We have revised the manuscript to explicitly outline and extend explanations and the relationships between sex, age, behavior, brain regions, and synaptic measures, ensuring that the rationale for each experiment and its relevance to the overall conclusions are improved.

(2) Methodological details: The Methods section of our initial submission was condensed, with key details provided in the Supplemental Methods section. We have merged all into an extended section to improve clarity. We have expanded our description of breeding strategies, test timelines, electrophysiological protocols, and data analysis methods in the revised Methods section. We believe the additions have enhanced the transparency and reproducibility of our study and ensured full support of our conclusions.

(3) Experimental flow: We have revised and extended our results, methods, and discussion sections to clarify the rationale and experimental design to guide readers through the experimental sequence and rationale.

We are confident these revisions address the concerns raised and enhance the robustness and coherence of our findings.

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study explores how heterozygosity for specific neurodevelopmental disorder-associated Trio variants affects mouse behavior, brain structure, and synaptic function, revealing distinct impacts on motor, social, and cognitive behaviors linked to clinical phenotypes. Findings demonstrate that Trio variants yield unique changes in synaptic plasticity and glutamate release, highlighting Trio's critical role in presynaptic function and the importance of examining variant heterozygosity in vivo.

Strengths:

This study generated multiple mouse lines to model each Trio variant, reflecting point mutations observed in human patients with developmental disorders. The authors employed various approaches to evaluate the resulting behavioral, neuronal morphology, synaptic function, and proteomic phenotypes.

Weaknesses:

While the authors present extensive results, the flow of experiments is challenging to follow, raising concerns about the strength of the experimental conclusions. Additionally, the connection between sex, age, behavioral data, brain regions, synaptic transmission, and plasticity lacks clarity, making it difficult to understand the rationale behind each experiment. Clearer explanations of the purpose and connections between experiments are recommended. Furthermore, the methodology requires more detail, particularly regarding mouse breeding strategies, timelines for behavioral tests, electrophysiology conditions, and data analysis procedures.

We appreciate the reviewer’s recognition of the novelty and comprehensiveness of our approach, particularly the generation of multiple mouse lines and our efforts to model Trio variant effects in vivo.

Weaknesses

(1) Experimental flow and rationale and connection between variables: We have expanded on the connections between behavioral data, neuronal morphology, synaptic function, and proteomics in the Results and Discussion sections to clarify how each experiment informs the reasoning and the conclusions and to highlight the relationships between sex, age, behavior, and synaptic measures.

(2) Methodological details: Our initial Methods section was formatted to be short to fulfill word limits on the submitted version, with additional details provided in the Supplemental Methods section. We have merged our Methods and Supplemental Methods sections and expanded on our breeding strategies, test timelines, electrophysiological protocols, and data analysis. We believe these additions enhance the transparency and reproducibility of our study.

(3) Recommendations for the authors: We thank Reviewer #1 for providing several recommendations to improve our manuscript. We have addressed their comments in the revision, as detailed below, adding key experiments that bolster our findings.

Reviewer #2 (Public review):

Summary:

The authors generated three mouse lines harboring ASD, Schizophrenia, and Bipolar-associated variants in the TRIO gene. Anatomical, behavioral, physiological, and biochemical assays were deployed to compare and contrast the impact of these mutations in these animals. In this undertaking, the authors sought to identify and characterize the cellular and molecular mechanisms responsible for ASD, Schizophrenia, and Bipolar disorder development.

Strengths:

The establishment of TRIO dysfunction in the development of ASD, Schizophrenia, and Bipolar disorder is very recent and of great interest. Disorder-specific variants have been identified in the TRIO gene, and this study is the first to compare and contrast the impact of these variants in vivo in preclinical models. The impact of these mutations was carefully examined using an impressive host of methods. The authors achieved their goal of identifying behavioral, physiological, and molecular alterations that are disorder/variant specific. The impact of this work is extremely high given the growing appreciation of TRIO dysfunction in a large number of brain-related disorders. This work is very interesting in that it begins to identify the unique and subtle ways brain function is altered in ASD, Schizophrenia, and Bipolar disorder.

Weaknesses:

(1) Most assays were performed in older animals and perhaps only capture alterations that result from homeostatic changes resulting from prodromal pathology that may look very different.

(2) Identification of upregulated (potentially compensating) genes in response to these disorder-specific Trio variants is extremely interesting. However, a functional demonstration of compensation is not provided.

(3) There are instances where data is not shown in the manuscript. See "data not shown". All data collected should be provided even if significant differences are not observed.

I consider weaknesses 1 and 2 minor. While they would be very interesting to explore, these experiments might be more appropriate for a follow-up study. I would recommend that the missing data in 3 should be provided in the supplemental material.

We are grateful for the reviewer’s recognition of our study’s significance and methodological rigor. The acknowledgment of Trio dysfunction as a novel and impactful area of research is deeply appreciated.

Weaknesses:

We agree that focusing on older animals limits insights into early-stage pathophysiology. However, our goal in this study was to examine the functional impacts of Trio heterozygosity at an adolescent stage and to reveal the ultimate impact of these alleles on synaptic function. Our choice of age aligns with our objectives. Future studies of earlier developmental stages will be beneficial and complement these findings.

Functional compensation:

We tested functional compensation through rescue experiments in +/K1431M brain slices using a Rac1-specific inhibitor, NSC23766, which prevents Rac1 activation by Trio or Tiam1. Our finding that direct Rac1 inhibition normalizes deficient neurotransmitter release in +/K1431M mice strongly suggests that increased Rac1 activity drives this phenotype.

Data not shown:

We will incorporate all previously shown data into the Supplemental Materials, even when results are nonsignificant. We agree that this ensures full transparency and facilitates a more comprehensive evaluation of our findings.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) In Figure 1K-N, the lack of observed differences in +/M2145T mice across all tests raises questions about its validity as a BPD model. Furthermore, the differences in female behavior data compared to males, as shown in the Supplemental section, lack clarification-specifically, whether these variations are due to sex differences or sample size disparities, which is not discussed. Additionally, it's unclear if the same mice were used in tests K through L-N, as the reported numbers differ without explanation; if relevant, any mortality should be reported. Given the observed body weight differences, it is important to display locomotor data, despite the mention of no change in open field results. Lastly, a detailed breeding strategy and timeline for behavioral testing would enhance clarity.

We thank Reviewer 1 for recognizing these confusing points in our behavioral data and seek to add clarification in our Revision as below:

(a) We have revised the text to emphasize our goal to evaluate the impact of NDD-related Trio alleles that have discrete and measurable effects on brain development and function, and not to model specific NDDs (e.g. ASD, SCZ, or BPD). The three specific Trio mutations were chosen based on strong evidence of these mutations impairing the biochemical functions of Trio. We reasoned our approach would reveal how impairing Trio in different ways – i.e. altering protein level or GEF1/GEF2 function – and under genetic conditions (heterozygosity) that mimic those found in individuals with Trio-related disorders impacts brain development and function. The lack of behavioral phenotypes in +/M2145T mice is indeed intriguing, especially given the alterations in electrophysiology and biochemistry experiments. It remains possible that further behavioral analyses of these mice will reveal behavioral phenotypes.

(b) Given that the prevalence and clinical presentation of individuals with various NDDs are influenced by sex, it is possible that the behavioral differences we see in male versus female Trio variant mice reflect human sex difference phenotypes. We have reorganized the Figure panels to clarify these sex differences in behaviors (new Fig. 2, Supp. Fig. 2). We focused on the most significant behavioral phenotypes shared by both sexes in the main text, or in males alone, as our anatomical and electrophysiological experiments were restricted to males to reduce variation due to estrus. The observed behavioral sex differences are not likely due to sample size disparities as power analyses were performed for all experimental results to ensure adequate sample size. A comprehensive study of the mechanisms underlying these behavioral findings merits examination but is outside the scope of this study.

(c) All mice were subjected to all behavioral tests described. No sudden mortality was observed during the behavioral experiments. Outliers in post-hoc statistical analyses were removed, which explains the apparent sample size differences between behavioral tests. We have revised the Data analysis section in our Methods to include these details (Lines 216-289, 450-457).

(d) Results of the open field test have been added to the Supplemental Data (new Supp. Fig. 2) and Results (Lines 532-537)

(e) The Methods section was expanded to include more detail on the breeding strategy (Lines 98-106). A timeline for behavioral testing has also been included in the Figures to enhance clarity (new Fig. 2A).

(2) In Figure 2A-E, head width and brain weight showed significant differences, but not body weight, how come the ratio does not change? Comparing with female results in Supplementary Figure 2A-E, it does show a difference between males and females. It is essential to clarify which sex authors use in all follow-up experiments, including synapse, transmission, and plasticity. Since the males and females have different phenotypes, why do the authors focus on males only? The E plot has no data points on the bar graph. In Figure 2I, it lacks example images for all four conditions.

We greatly appreciate this Reviewer’s attention to details in our brain and body weight data and revised the manuscript to address these concerns.

(a) The ratios of head width/body weight were calculated for each individual mouse. Hence the distribution of the ratio data (old Fig. 2D; new Fig. 3D) differs from the distribution of head width or body weight data alone (old Fig. 2A, 2C, resp.; now Fig. 3A, 3C), and therefore can affect the p-value for statistical significance. The body weight of +/M2145T males is 21.217 ±0.327 g, while for WT males is 21.745 ±0.224 g, a non-significant decrease of 0.528 g (adjusted p=0.3806). These values have been added to the Fig 3. figure legend (Lines 1020-1034) for clarity.

(b) Similar to the behavioral experiments in comment (1), we observed sex differences in head width, brain weight, and body weight in Trio heterozygous variant mice compared to WT counterparts. The differences in the ratios of head width/body weight or brain weight/body weight were the same for both males and females (i.e. head width/body weight ratio is decreased in +/K1431M mice compared to WT regardless of sex, and brain weight/body weight ratio is decreased in both +/K1431M and +/K1918X mice compared to WT regardless of sex). These findings affirm the impact of Trio mutations on these phenotypes across both sexes. We have modified the text to draw more attention to this key point (Lines 554-566 and 777-801).

(c) All experiments (excluding behavior and weight data) were performed in males only to minimize the variation in spine and synapse morphology and physiological activity that can occur due to estrus. We have clarified this in the ‘Animal Work’ section of the Methods (Lines 103-106) as well as in the Figure Legends.

(d) We thank the Reviewer for pointing out Fig. 3E lacks individual data points on the bar graph. Fig. 3E has been modified to now include the brain weight/body weight ratio for each individual mouse rather than across the population, to be consistent with the calculation of head width/body weight ratio (see point 2a).

On original submission, only a representative WT image was selected due to space constraints. The figure (new Fig. 3H and 3K) and figure legend have been revised to include representative traces for all genotypes examined.

(3) In lines 315-320, "None of the Trio variant heterozygotes exhibited altered dendritic spine density on M1 L5 pyramidal neurons compared to WT mice on either apical or basal arbors (Supplementary Figure 3L, M). Electron microscopy of cortical area M1 L5 revealed that synapse density was significantly increased in +/K1918X mice compared to WT (Figure 3A, B), possibly due to a net reduction in neuropil resulting from smaller dendritic arbors." The proposed explanation does not adequately address the observed discrepancy between spine density and synapse density reported in these two experiments. A more thorough analysis is needed to reconcile these conflicting findings and clarify how these distinct measurements may relate to each other in the context of the study's conclusions.

We acknowledge the apparent discrepancy between our dendritic spine density data, which is unchanged from WT for all three Trio variant heterozygotes, and our synapse density data, which showed an increase in +/K1918X M1 L5 compared to WT. We have expanded the explanation for this discrepancy below and added this to the Discussion (Lines 802-811):

a) Because spine density can vary by dendritic branch order and distance from the soma, only protrusions from secondary dendritic arbors of M1 L5 pyramidal neurons were quantified for consistency in analyses. However, all synapses meeting criteria were quantified in EM images, regardless of where they were located along an individual neuron’s arbors. It is possible that the density and distribution of spines along other arbors are different between genotypes but was not captured in our current data.

b) +/K1918X L5 pyramidal neurons are smaller and less complex than WT neurons, especially in the basal compartment corresponding to L5 where EM images were obtained, consistent with the smaller brain size and reduced cortical thickness of +/K1918X mice. We posit that due to their smaller dendritic field size, L5 neurons pack more densely contributing to the increased synapse density observed in +/K1918X M1 L5 cortex. Consistent with this hypothesis, we observed a trend toward increased DAPI+ cell density in M1 L5 of +/K1918X neurons (Supp. Fig. 3N).

(4) In Figure 4, one potential rationale for measuring AMPAR mEPSC frequency is to infer synapse density changes. However, the findings show no frequency change in +/K1431M and +/K1918X, with an increase only in +/M2145T, which contradicts Figure 3 results indicating a trend toward increased density across variants.

This inconsistency is confusing, especially since the authors claim to follow the methodology from the study "Trio Haploinsufficiency Causes Neurodevelopmental Disease-Associated Deficits"; yet, the observed mEPSC amplitude differs significantly from that study, while the frequency remains unaffected. Additionally, the NMDAR mEPSCs reflect combined AMPAR and NMDAR responses at positive holding potentials, with peak amplitude dominated by AMPAR. This inconsistency between holding potential results is unclear, as frequency should theoretically align across negative and positive potentials. For accurate NMDAR mEPSC measurement, it would be optimal to assess amplitude 50 ms post-initial peak and, if possible, increase the holding potential to enhance the driving force given the typically low signal of NMDAR response.

We thank the Reviewer for highlighting these important points.

a) Previous work from our lab and others demonstrate that Trio regulates synaptic AMPA receptor levels, which is why we chose to focus on AMPAR-mediated evoked and miniature EPSC frequencies and amplitudes in the current study. We acknowledge Reviewer 1’s comment on seemingly contradictory results regarding AMPAR mEPSC frequency and synapse density; however, the unchanged AMPAR mEPSC frequency in +/K1431M and +/K1918X mice is consistent with our finding of unaltered dendritic spine density in these mice compared to WT (Supp. Fig. 4L,M). The differences between dendritic spine counts and synapse density is addressed in Response (3) above.

b) While synapse density changes can be inferred from AMPAR mEPSC frequency, mEPSCs are also measures of spontaneous neurotransmitter release changes especially in the absence of changes in synaptic numbers. Notably, the increased mEPSC frequency in the +/M2145T variant is linked to enhanced spontaneous release, not to spine or synapse density changes. These findings are reinforced by increase in counts of synaptic vesicles, calculated PPR changes, and estimates of the Pr and RRP from HFS train analysis. We have included these points in the Discussion (Lines 861-863).

c) While it is tempting to compare the current study to our previously published conditional Trio haploinsufficiency model, we highlight key distinctions that may underlie phenotypic differences between these two mouse models. First, our prior model used a NEX-Cre transgene to ablate one Trio allele from excitatory neurons only beginning at embryonic day 11. In contrast, our Trio variants are expressed in all cell types throughout development, akin to the genetic variants found in individuals with TRIO-related disorders. Second, the Trio variant mice in this study are on a C57BL/6 background, while the Trio haploinsufficient mice were on a mixed 129Sv/J X C57BL/6 background. These differences in the current study may explain why some measures, such as mEPSC amplitude, may not align with those from the Trio conditional haploinsufficiency model.

d) Recordings were performed using specific inhibitors to isolate AMPA and NMDA mEPSCs; these missing methodological details have now been clarified in the updated Methods section (Lines 353-360).

(5) In Supplementary Figure 4, the sample traces indicate a higher NMDA/AMPA ratio, raising the question of whether the AMPA EPSC amplitude changes, as this could reflect PSD length. In Figure 4B, the increased AMPAR mEPSC amplitude in the +/K1918X condition compared to WT suggests an enhanced postsynaptic response, yet the PSD length is reduced in Figure 3C. Can the authors provide a potential hypothesis to explain this?

We appreciate the Reviewer’s feedback. Yes, both evoked and miniature recordings indicate increased AMPAR amplitudes in the +/K1918X variants compared to WT. While PSD length is often linked to synaptic strength, the observed reduction in PSD length in EM PSD length reduction in +/K1918X synapses is small (~6% of WT) and clearly does not correlate with significant changes in synaptic strength. We also note that the whole cell recordings of mEPSCs represent input from all active synapses on the neuron, while PSD length is measured only in synapses of the L5.

(6) In Figure 4, synaptic plasticity appears to decrease to around 50% of baseline; could this reduction be attributed to LTD, or might it result from changes in pipette resistance? Additionally, is the observed potentiation due to changes in presynaptic release probability? Measuring paired-pulse ratio (PPR) before and after induction would clarify this aspect.

We thank the Reviewer for highlighting these important points.

a) We used a well-established theta burst stimulation method for LTP induction in M1 L5 pyramidal neurons. This protocol reliably evokes LTP in WT neurons, as shown in Fig. 5J and K. Both +/K1431M and +/K1918X variants exhibit a slight but discernible increase in evoked excitatory postsynaptic currents (eEPSCs), indicative of the initiation of LTP. Although this increase is smaller compared to WT, the presence of potentiation indicates that long-term depression (LTD) is an unlikely explanation for the observed reduction.

b) To rule out the influence of technical artifacts, pipette resistance was carefully monitored before and after LTP induction. Any cells exhibiting resistance changes exceeding 20% during electrophysiological recordings were excluded from the analysis, ensuring that fluctuations in pipette resistance did not confound LTP measurements. These technical details are denoted in the Methods (Lines 344-346 and 364-366).

c) The potentiation in the +/M2145T variant may stem from increased release probability (Pr) and greater synaptic vesicle availability, but is beyond the scope of this work. We agree this is an intriguing question, not only for +/M2145T but also for +/K1431M mice. Future studies should address this, ideally using models where the Trio variant is selectively introduced into the presynaptic neuron.

(7) In lines 377-380, "The +/M2145T PPR curve was unusual, with significantly reduced PPF at short ISIs, yet clearly increased PPF at longer ISI (Figure 5A, B) compared to WT." The unusual PPR observed at the 100 ms ISI appears unexpected. Can the authors provide an explanation for this anomaly? This finding could suggest atypical presynaptic dynamics or modulation at this specific interval, which may differ from typical synaptic behavior. Further insights into possible mechanisms or experimental conditions affecting this result would be valuable.

"The decreased PPF at initial ISI in +/M2145T mice correlated with increased mEPSC frequency (Fig. 4A-C), suggestive of a possible increase in spontaneous glutamate Pr." If this is the case, it raises the question of why the increased PPR at the initial ISI in +/K1431M does not correspond to the result shown in Figure 4C. This discrepancy suggests that factors beyond initial presynaptic release probability might be influencing the observed synaptic response, or that compensatory mechanisms could be affecting PPR and mEPSC frequency differently in this variant. Further clarification on the interplay between these measurements would help resolve this inconsistency.

We appreciate the Reviewer’s critical reading and genuine interest on this phenotype in +/M2145T mice.

a) The unusual shift of the PPR in +/M2145T at ISI 100ms is fascinating and will require significant additional experimentation that lies beyond the scope of this report to address. We propose it results from altered presynaptic regulators, including increased Syt3 and reduced RhoA activity. Notably, Syt3 influences calcium-dependent SV replenishment, which can cause similar PPR defects (Weingarten DJ et al., 2022); this is now included in the Discussion. (Lines 915-918).

Weingarten DJ, Shrestha A, Juda-Nelson K, Kissiwaa SA, Spruston E, Jackman SL. Fast resupply of synaptic vesicles requires synaptotagmin-3. Nature. 2022 Nov;611(7935):320-325. doi: 10.1038/s41586-022-05337-1. Epub 2022 Oct 19. PMID: 36261524.

b) Thank you for raising the concern in clarity of this statement "The decreased PPF at initial ISI in +/M2145T mice correlated with increased mEPSC frequency (Fig. 4A-C), suggestive of a possible increase in spontaneous glutamate Pr." We have edited the sentence to be more clear (Lines 701-703). First, the K1431M and M2145T variants impact different TRIO catalytic activities disrupting distinct GTPase pathways and differentially affecting presynaptic regulators, which can lead to non-overlapping phenotypes. Also, we expand our discussion that +/K1431M variant data suggest increased AMPAR numbers and fewer silent synapses (Lines 850-855), potentially increasing AMPAR mEPSC frequency and masking the expected decrease in spontaneous release (Lines 905-910). Further experiments are needed, ideally using mixed cultures with TRIO variants in presynaptic neurons with synapses on WT neurons, as minimal stimulation variance analysis in slices would be inconclusive due to its reflection of both Pr and silent synapse changes, similar to mEPSC frequency.

(8) In Figure 5, there is no evidence demonstrating that the NSC inhibitor functions specifically in the +/K1431M condition without affecting other conditions. To verify its specificity, the authors should test the NSC inhibitor's effects across other conditions in parallel, including a control group. Additionally, cumulative RRP measurements should be provided for a more comprehensive assessment of the inhibitor's impact on synaptic function.

We appreciate the Reviewer’s feedback.

a) Previous studies have shown that Rac1 activity can bidirectionally regulate synchronous release probability (Pr). We used the Rac1-specific inhibitor NSC23766 (NSC) to test how Rac1 inhibition impacted the neurotransmitter release deficits observed in +/K1431M mice. We also added control experiments testing the impact of NSC on WT slices. These new experiments are now presented in new Fig. 8 of the revised manuscript, with expanded details in the Results (Lines 737-750) and Discussion (Lines 892-900).

b) To estimate Pr and the RRP, we employed the Decay method as described by (Ruiz et al., 2011), which does not rely on cumulative EPSC plots for RRP estimation. This approach was chosen to account for the initial facilitation in these synapses and fits are done using EPSCs plotted against stimulus number. Additional details have been provided in the Methods section  (Lines 367-373).

Ruiz R, Cano R, Casañas JJ, Gaffield MA, Betz WJ, Tabares L. Active zones and the readily releasable pool of synaptic vesicles at the neuromuscular junction of the mouse. J Neurosci. 2011 Feb 9;31(6):2000-8. doi: 10.1523/JNEUROSCI.4663-10.2011. PMID: 21307238; PMCID: PMC6633039.

(9) Given the relevance to NDD, specifying the age window of the mice used is crucial. It is confusing that the synaptic function studies were conducted at P42, while the proteomic analysis was performed at P21. Could the authors clarify the rationale behind using different age points for these analyses? Consistency in age selection, or an explanation for this variation, would help in interpreting the developmental relevance of the findings.

P42 was chosen as the age as it represents young adulthood, by which time clinical features will have already presented in individuals with neurodevelopmental disorders. Our prior studies of NEX-Cre Trio^-/- mice found significant measurable differences from WT at this age, after neuronal migration, differentiation, synaptogenesis and pruning have occurred. An earlier developmental timepoint, P21, which coincides with juvenile age in mice, was chosen for proteomics studies to identify earlier changes and potentially targetable and modifiable mechanisms that could influence the phenotypes we observed in older mice. The experiments in P42 versus P21 mice were originally two independent lines of investigation that converged in the current study.

eLife Assessment

This valuable study confirms the association between the human leukocyte antigen (HLA)-II region and tuberculosis (TB) susceptibility in genetically admixed South African populations, specifically identifying a near-genome-wide significant association in the HLA-DPB1 gene, which originates from KhoeSan ancestry. The evidence supporting the association between the HLA-II region and TB susceptibility is solid, and the work will be of interest to those studying the genetic basis of tuberculosis susceptibility/infection resistance.

Reviewer #2 (Public review):

Summary:

This manuscript is about using different analytical approaches to allow ancestry adjustments to GWAS analyses amongst admixed populations. This work is a follow-on from the recently published ITHGC multi-population GWAS (https://doi.org/10.7554/eLife.84394), with the focus on the admixed South African populations. Ancestry adjustment models detected a peak of SNPs in the class II HLA DPB1, distinct from the class II HLA DQA1 loci signficant in the ITHGC analysis.

Strengths:

Excellent demonstration of GWAS analytical pipelines in highly admixed populations. Particularly the utility of ancestry adjustment to improve study power to detect novel associations. Further confirmation of the importance of the HLA class II locus in genetic susceptibility to TB.

Weaknesses:

Limited novelty compared to the group's previous existing publications and the body of work linking HLA class II alleles with TB susceptibility in South Africa or other African populations. This work includes only ~100 new cases and controls from what has already been published. High resolution HLA typing has detected significant signals in both the DQA1 and DPB1 regions identified by the larger ITHGC and in this GWAS analysis respectively (Chihab L et al. HLA. 2023 Feb; 101(2): 124-137).<br /> Despite the availability of strong methods for imputing HLA from GWAS data (Karnes J et Plos One 2017), the authors did not confirm with HLA typing the importance of their SNP peak in the class II region. This would have supported the importance of this ancestry adjustment versus prior ITHGC analysis.

The populations consider active TB and healthy controls (from high-burden presumed exposed communities) and do not provide QFT or other data to identify latent TB infection.

Important methodological points for clarification and for readers to be aware of when reading this paper:

(1) One of the reasons cited for the lack of African ancestry-specific associations or suggestive peaks in the ITHGC study was the small African sample size. The current association test includes a larger African cohort and yields a near-genome-wide significant threshold in the HLA-DPB1 gene originating from the KhoeSan ancestry. Investigation is needed as to whether the increase in power is due to increased African samples and not necessarily the use of the LAAA model as stated on lines 295 and 296?

Authors response - The Manhattan plot in Figure 3 includes the results for all four models: the traditional GWAS model (GAO), the admixture mapping model (LAO), the ancestry plus allelic (APA) model and the LAAA model. In this figure, it is evident that only the LAAA model identified the association peak on chromosome 6, which lends support the argument that the increase in power is due to the use of the LAAA model and not solely due to the increase in sample size.<br /> Reviewer comment - This data supports the authors conclusions that increase power is related to the LAAA model application rather than simply increase sample size.

(2) In line 256, the number of SNPs included in the LAAA analysis was 784,557 autosomal markers; the number of SNPs after quality control of the imputed dataset was 7,510,051 SNPs (line 142). It is not clear how or why ~90% of the SNPs were removed. This needs clarification.

Authors response:<br /> In our manuscript (line 194), we mention that "...variants with minor allele frequency (MAF) < 1% were removed to improve the stability of the association tests." A large proportion of imputed variants fell below this MAF threshold and were subsequently excluded from this analysis.

Reviewers additional comment: The authors should specify the number of SNPs in the dataset before imputation and indicate what proportion of the 784,557 remaining SNPs were imputed. Providing this information might help the reader better understand the rationale behind the imputation process.

(3) The authors have used the significance threshold estimated by the STEAM p-value < 2.5x10-6 in the LAAA analysis. Grinde et al. (2019 implemented their significance threshold estimation approach tailored to admixture mapping (local ancestry (LA) model), where there is a reduction in testing burden. The authors should justify why this threshold would apply to the LAAA model (a joint genotype and ancestry approach).

Authors response: We describe in the methods (line 189 onwards) that the LAAA model is an extension of the APA model. Since the APA model itself simultaneously performs the null global ancestry only model and the local ancestry model (utilised in admixture mapping), we thus considered the use of a threshold tailored to admixture mapping appropriate for the LAAA model.

Reviewers additional comment: While the LAAA model is an extension of the APA model, the authors describe the LAAA test as 'models the combination of the minor allele and the ancestry of the minor allele at a specific locus, along with the effect of this interaction,' thus a joint allele and ancestry effects model. Grinde et al. (2019) proposed the significance threshold estimation approach, STEAM, specifically for the LA approach, which tests for ancestry effects alone and benefits from the reduced testing burden. However, it remains unclear why the authors found it appropriate to apply STEAM to the LAAA model, a joint test for both allele and ancestry effects, which does not benefit from the same reduction in testing burden.

(4) Batch effect screening and correction (line 174) is a quality control check. This section is discussed after global and local ancestry inferences in the methods. Was this QC step conducted after the inferencing? If so, the authors should justify how the removed SNPs due to the batch effect did not affect the global and local ancestry inferences or should order the methods section correctly to avoid confusion.

Authors response: The batch effect correction method utilised a pseudo-case-control comparison which included global ancestry proportions. Thus, batch effect correction was conducted after ancestry inference. We excluded 36 627 SNPs that were believed to have been affected by the batch effect. We have amended line 186 to include the exact number of SNPs excluded due to batch effect.<br /> The ancestry inference by RFMix utilised the entire merged dataset of 7 510 051 SNPs. Thus, the SNPs removed due to the batch effect make up a very small proportion of the SNPs used to conduct global and local ancestry inferences (less than 0.5%). As a result, we do not believe that the removed SNPs would have significantly affected the global and local ancestry inferences. However, we did conduct global ancestry inference with RFMix on each separate dataset as a sanity check. In the tables below, we show the average global ancestry proportions inferred for each separate dataset, the average global ancestry proportions across all datasets and the average global ancestry proportions inferred using the merged dataset. The SAC and Xhosa cohorts are shown in two separate tables due to the different number of contributing ancestral populations to each cohort. The differences between the combined average global ancestry proportions across the separate cohorts does not differ significantly to the global ancestry proportions inferred using the merged dataset.

This is an excellent response and should remain accessible to readers for clarifying this issue.

Comments on revisions:

Thank you for addressing my other recommendations to authors. These have all been satisfactorily addressed.

Author response:

The following is the authors’ response to the previous reviews

Recommendations for the authors:

Reviewer #1:

First, I thank the authors for clarifying some of the confusion I had in the previous comment and I appreciate the efforts the authors put into improving the quality of the manuscript. However, my concerns about the lack of novelty of the key findings are not perfectly addressed and there is no additional analysis done in this revision. Currently in this version of the manuscript, asserting that a p-value of 10-6 is close to genome-wide significance may be considered an overstatement. Further analysis focusing on finding novel and additional discovery is very necessary.

We thank the reviewer for their comments. Reviewer #2 also made a comment regarding the genomewide threshold, “However, it remains unclear why the authors found it appropriate to apply STEAM to the LAAA model, a joint test for both allele and ancestry effects, which does not benefit from the same reduction in testing burden.” The reviewers’ have correctly identified our oversight - we have amended the manuscript as follows:

(1) The abstract, “We identified a suggestive association peak (rs3117230, p-value = 5.292 x10-6, OR = 0.437, SE = 0.182) in the HLA-DPB1 gene originating from KhoeSan ancestry.”

(2) From line 233 to 239: “The R package STEAM (Significance Threshold Estimation for Admixture Mapping) (Grinde et al., 2019) was used to determine the admixture mapping significance threshold given the global ancestral proportions of each individual and the number of generations since admixture (g = 15). For the LA model, a genome-wide significance threshold of pvalue < 2.5 x 10-6 was deemed significant by STEAM. The traditional genome-wide significance threshold of 5 x 10-8 was used for the GA, APA and LAAA models, as recommended by the authors of the LAAA model (Duan et al., 2018).” 

(3) We excluded the results for the signal on chromosome 20, since this also did not reach the LAAA model genome-wide significance threshold.  

(4) From line 296 to 308: “LAAA models were successfully applied for all five contributing ancestries (KhoeSan, Bantu-speaking African, European, East Asian and Southeast Asian). However, no variants passed the threshold for statistical significance. Although no variants reached genome-wide significance, a suggestive peak was identified in the HLA-II region of chromosome 6 when using the LAAA model and adjusting for KhoeSan ancestry (Figure 3). The QQ-plot suggested minimal genomic inflation, which was verified by calculating the genomic inflation factor ( = 1.05289) (Supplementary Figure 1). The lead variants identified using the LAAA model whilst adjusting for KhoeSan ancestry in this region on chromosome 6 are summarised in Table 3. The suggestive peak encompasses the HLA-DPA1/B1 (major histocompatibility complex, class II, DP alpha 1/beta 1) genes (Figure 4). It is noteworthy that without the LAAA model, this suggestive peak would not have been observed for this cohort. This highlights the importance of utilising the LAAA model in future association studies when investigating disease susceptibility loci in admixed individuals, such as the SAC population.”

We acknowledge that our results are not statistically significant. However, our study advances this area of research by identifying suggestive African-specific ancestry associations with TB in the HLA-II region. These findings build upon the work of the ITHGC, which did not identify any significant associations or suggestive peaks in their African-specific analyses. We have included this argument in our manuscript (from lines 425 to 432):

“The ITHGC did not identify any significant associations or suggestive peaks in their African ancestryspecific analyses.  Notably, the suggestive peak in the HLA-DPB1 region was only captured in our cohort using the LAAA model whilst adjusting for KhoeSan local ancestry. This underscores the importance of incorporating global and local ancestry in association studies investigating complex multi-way admixed individuals, as the genetic heterogeneity present in admixed individuals (produced as a result of admixtureinduced and ancestral LD patterns) may cause association signals to be missed when using traditional association models (Duan et al., 2018; Swart, van Eeden, et al., 2022).”

We appreciate the comment regarding additional analyses. We acknowledge that we did not validate our SNP peak in the HLA-II region through fine-mapping due to the lack of a suitable reference panel (see lines 490 to 500). Our long-term goal is to develop a HLA-imputation reference panel incorporating KhoeSan ancestry; however, this is beyond the scope and funding allowances of this study.

Reviewer #2 (Recommendations for the authors):

The authors we think have done an excellent job with their responses and the manuscript has been substantially improved.

Thank you for taking the time to help us improve our manuscript.

eLife Assessment

This study exploring the role of TRPV1 signaling in recruiting macrophages and promoting angiogenesis during tympanic membrane wound healing presents useful findings. However, the strength of evidence supporting the central claims is incomplete, as the mechanistic links between TRPV1 activation and immune cell recruitment remain largely correlative and rely heavily on previously published datasets without sufficient functional validation. The work will be of interest to researchers studying wound healing and sensory-immune interactions, though substantial revisions are needed to support its broader significance.

Reviewer #1 (Public review):

Summary:

This study reveals that TRPV1 signaling plays a key role in tympanic membrane (TM) healing by promoting macrophage recruitment and angiogenesis. Using a mouse TM perforation model, researchers found that blood-derived macrophages accumulated near the wound, driving angiogenesis and repair. TRPV1-expressing nerve fibers triggered neuroinflammatory responses, facilitating macrophage recruitment. Genetic Trpv1 mutation reduced macrophage infiltration, angiogenesis, and delayed healing. These findings suggest that targeting TRPV1 or stimulating sensory nerve fibers could enhance TM repair, improve blood flow, and prevent infections. This offers new therapeutic strategies for TM perforations and otitis media in clinical settings. This is an excellent and high-quality study that provides valuable insights into the mechanisms underlying TM wound healing.

Strengths:

The work is particularly important for elucidating the cellular and molecular processes involved in TM repair. However, there are several concerns about the current version.

Weaknesses:

Major concerns

(1) The method of administration will be a critical factor when considering potential therapeutic strategies to promote TM healing. It would be beneficial if the authors could discuss possible delivery methods, such as topical application, transtympanic injection, or systemic administration, and their respective advantages and limitations for targeting TRPV1 signaling. For example, Dr. Kanemaru and his colleagues have proposed the use of Trafermin and Spongel to regenerate the eardrum.

(2) The authors appear to have used surface imaging techniques to observe the TM. However, the TM consists of three distinct layers: the epithelial layer, the fibrous middle layer, and the inner mucosal layer. The authors should clarify whether the proposed mechanism involving TRPV1-mediated macrophage recruitment and angiogenesis is limited to the epithelial layer or if it extends to the deeper layers of the TM.

Minor concerns

In Figure 8, the schematic illustration presents a coronal section of the TM. However, based on the data provided in the manuscript, it is unclear whether the authors directly obtained coronal images in their study. To enhance the clarity and impact of the schematic, it would be helpful to include representative images of coronal sections showing macrophage infiltration, angiogenesis, and nerve fiber distribution in the TM.

Reviewer #2 (Public review):

Summary:

This study examines the role of TRPV1 signaling in the recruitment of monocyte-derived macrophages and the promotion of angiogenesis during tympanic membrane (TM) wound healing. The authors use a combination of genetic mouse models, macrophage depletion, and transcriptomic approaches to suggest that neuronal TRPV1 activity contributes to macrophage-driven vascular responses necessary for tissue repair.

Strengths:

(1) The topic of neuroimmune interactions in tissue regeneration is of interest and underexplored in the context of the TM, which presents a unique model due to its anatomical features.

(2) The use of reporter mice and bone marrow chimeras allows for some dissection of immune cell origin.

(3) The authors incorporate transcriptomic data to contextualize inflammatory and angiogenic processes during wound healing.

Weaknesses:

(1) The primary claims of the manuscript are not convincingly supported by the evidence presented. Most of the data are correlative in nature, and no direct mechanistic experiments are included to establish causality between TRPV1 signaling and macrophage recruitment or function.

(2) Functional validation of key molecular players (such as Tac1 or Spp1) is lacking, and their roles are inferred primarily from gene expression data rather than experimentally tested.

(3) The reuse of publicly available scRNA-seq data is not sufficiently integrated or extended to yield new biological insights, and it remains largely descriptive.

(4) The macrophage depletion model (CX3CR1CreER; iDTR) lacks specificity, and possible off-target or systemic effects are not addressed.

(5) Several interpretations of the data appear overstated, particularly regarding the necessity of TRPV1 for monocyte recruitment and wound healing.

(6) Overall, the study appears to apply known concepts - namely, TRPV1-mediated neurogenic inflammation and macrophage-driven angiogenesis - to a new anatomical site without providing new mechanistic insight or advancing the field substantially.

Overall:

While the study addresses an interesting topic, the current version does not provide sufficiently strong or novel evidence to support its major conclusions. Additional mechanistic experiments and more rigorous validation would be necessary to substantiate the proposed model and clarify the relevance of the findings beyond this specific tissue context.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study reveals that TRPV1 signaling plays a key role in tympanic membrane (TM) healing by promoting macrophage recruitment and angiogenesis. Using a mouse TM perforation model, researchers found that blood-derived macrophages accumulated near the wound, driving angiogenesis and repair. TRPV1-expressing nerve fibers triggered neuroinflammatory responses, facilitating macrophage recruitment. Genetic Trpv1 mutation reduced macrophage infiltration, angiogenesis, and delayed healing. These findings suggest that targeting TRPV1 or stimulating sensory nerve fibers could enhance TM repair, improve blood flow, and prevent infections. This offers new therapeutic strategies for TM perforations and otitis media in clinical settings. This is an excellent and high-quality study that provides valuable insights into the mechanisms underlying TM wound healing.

Strengths:

The work is particularly important for elucidating the cellular and molecular processes involved in TM repair. However, there are several concerns about the current version.

We sincerely thank Reviewer #1 for their time and effort in evaluating and improving our study. Below, we are pleased to address the Reviewer's concerns point by point.

Weaknesses:

Major concerns

(1) The method of administration will be a critical factor when considering potential therapeutic strategies to promote TM healing. It would be beneficial if the authors could discuss possible delivery methods, such as topical application, transtympanic injection, or systemic administration, and their respective advantages and limitations for targeting TRPV1 signaling. For example, Dr. Kanemaru and his colleagues have proposed the use of Trafermin and Spongel to regenerate the eardrum.

We are grateful to the reviewer for raising this important point. While the present study primarily focuses on the mechanistic role of TRPV1 in TM repair, we agree that the mode of therapeutic delivery will be pivotal in translating these findings into clinical practice. In response, we will expand the discussion to explore possible delivery methods—such as topical application, transtympanic injection, and systemic routes—along with their respective benefits and challenges. We will also cite the work by Dr. Kanemaru and colleagues as an example of how local delivery systems may facilitate TM regeneration.

(2) The authors appear to have used surface imaging techniques to observe the TM. However, the TM consists of three distinct layers: the epithelial layer, the fibrous middle layer, and the inner mucosal layer. The authors should clarify whether the proposed mechanism involving TRPV1-mediated macrophage recruitment and angiogenesis is limited to the epithelial layer or if it extends to the deeper layers of the TM.

We apologize for any confusion caused by our previous description. In our study, we utilized Z-stack confocal imaging to capture the full thickness of the TM, as illustrated in Author response image 1 (reconstructed from the acquired Z-sections). This imaging technique allowed us to encompass all three layers of the TM entirely. Each sample was imaged using a 10X objective on an Olympus fluorescence microscope. Given the conical shape and size of the TM, we imaged it in four quadrants, acquiring approximately 30 optical sections (with a 3 µm step) per region. Each acquired images were projected and exported using FV10ASW 4.2 Viewer, then stitched together using Photoshop. The resulting Z-stack projections enabled us to visualize the distribution of macrophages, angiogenesis, and the localization of nerve fibers throughout the TM. We will include this detailed methodology in our revision to clarify any potential confusion.

Author response image 1.

Representative confocal images showing one quadrant of the TM collected from collected from CSR1F^EGFP bone marrow transplanted mouse at day 7 post-perforation. (A-B) 3D-rendered views from different angles reveal the close spatial relationship between CSF1R^EGFP cells (green) and blood vessels (red) within the TM. (C) Cross-sectional view highlights the depth-wise distribution of CSF1R^EGFP cells (green) and blood vessels (red) across the layered TM architecture. All images were processed using Imaris Viewer x64 (version 10.2.0).

Minor concerns

In Figure 8, the schematic illustration presents a coronal section of the TM. However, based on the data provided in the manuscript, it is unclear whether the authors directly obtained coronal images in their study. To enhance the clarity and impact of the schematic, it would be helpful to include representative images of coronal sections showing macrophage infiltration, angiogenesis, and nerve fiber distribution in the TM.

As noted above, we utilized Z-stack confocal imaging to capture the full thickness of the TM, enabling us to visualize structures across all three layers. This approach ensured that all layers were included in our analysis. Due to the thin and curved nature of the TM, traditional cross-sectional imaging often struggles to clearly depict the spatial relationships between macrophages, blood vessels, and nerve fibers, especially at low magnification as shown in Author response image 2. In response to the reviewer's suggestion, we will include representative coronal images in the revised manuscript to better illustrate the distribution of these structures at higher magnification.

Author response image 2.

Confocal images of eardrum cross-sections collected at day 1 (A), 3 (B), and 7 (C) post perforation to demonstrate the wound healing processes.

Reviewer #2 (Public review):

Summary:

This study examines the role of TRPV1 signaling in the recruitment of monocyte-derived macrophages and the promotion of angiogenesis during tympanic membrane (TM) wound healing. The authors use a combination of genetic mouse models, macrophage depletion, and transcriptomic approaches to suggest that neuronal TRPV1 activity contributes to macrophage-driven vascular responses necessary for tissue repair.

Strengths:

(1) The topic of neuroimmune interactions in tissue regeneration is of interest and underexplored in the context of the TM, which presents a unique model due to its anatomical features.

(2) The use of reporter mice and bone marrow chimeras allows for some dissection of immune cell origin.

(3) The authors incorporate transcriptomic data to contextualize inflammatory and angiogenic processes during wound healing.

We sincerely thank Reviewer #2 for their time and effort in improving our study and recognizing its strengths. Below, we are pleased to address the reviewer's concerns point by point.

Weaknesses:

(1) The primary claims of the manuscript are not convincingly supported by the evidence presented. Most of the data are correlative in nature, and no direct mechanistic experiments are included to establish causality between TRPV1 signaling and macrophage recruitment or function.

We appreciate Reviewer #2's perspective on the lack of molecular mechanisms linking TRPV1 signaling and macrophages. However, our data demonstrates that TRPV1 mutations significantly affect macrophage recruitment and angiogenesis. This initial study primarily focuses on the intriguing phenomenon of how sensory nerve fibers are involved in eardrum immunity and wound healing, an area that has not been clearly reported in the literature before. We believe that further research is necessary to explore this topic in greater depth.

(2) Functional validation of key molecular players (such as Tac1 or Spp1) is lacking, and their roles are inferred primarily from gene expression data rather than experimentally tested.

Although we have identified the TAC1 and SPP1 signals as potentially important for TM wound healing for the first time, we agree with the Reviewer's view regarding the lack of molecular mechanisms explored in this study. We have not yet tested the downstream signaling pathways, but we plan to investigate them in a series of future studies. As this is an early report, we will continue to explore these signals and their potential clinical applications based on our initial findings moving forward.

(3) The reuse of publicly available scRNA-seq data is not sufficiently integrated or extended to yield new biological insights, and it remains largely descriptive.

We appreciate Reviewer #2 for highlighting this point. Leveraging publicly available scRNA-seq databases and established analysis pipelines not only saves time and resources—my lab recently collected macrophages from the eardrums of postnatal P15 mice, with each trial requiring 20 eardrums from 10 animals to obtain a sufficient number of cells—but also allows researchers to build on previous work and focus on new biological questions without the need to repeat experiments. A prior study conducted by Dr. Tward and his team utilized scRNA-seq data to make initial discoveries related to eardrum wound healing, primarily focusing on epithelial cells rather than macrophages. We are building on their raw data to uncover new biological insights regarding macrophages, even though we have not yet tested the unidentified signals, which we believe will be valuable to our peers.

(4) The macrophage depletion model (CX3CR1CreER; iDTR) lacks specificity, and possible off-target or systemic effects are not addressed.

We agree with reviewer #2, although macrophage depletion model used in our study is a standard and well-used animal model (Shi, Hua et al. 2018), which has been used by many other laboratories, it is important to note that any macrophage depletion model may have potential issues. We will discuss this in our revision.

(5) Several interpretations of the data appear overstated, particularly regarding the necessity of TRPV1 for monocyte recruitment and wound healing.

We thank the reviewer for pointing this out. We will revise our manuscript where it is overstated accordingly.

(6) Overall, the study appears to apply known concepts - namely, TRPV1-mediated neurogenic inflammation and macrophage-driven angiogenesis - to a new anatomical site without providing new mechanistic insight or advancing the field substantially.

Although our study may not seem highly innovative at first glance, it reveals a previously unknown role of the TRPV1 pain signaling pathway in promoting eardrum healing for the first time. This healing process includes the recruitment of monocyte-derived macrophages and the formation of new blood vessels (angiogenesis). While this process has been documented in other organs, most research on macrophage-driven angiogenesis has been conducted using in vitro models, with very few studies demonstrating this process in vivo. Our findings could lead to new translational opportunities, especially considering that tympanic membrane perforation, along with damage-induced otitis media and conductive hearing loss, are common clinical issues affecting millions of people worldwide. Targeting TRPV1 signaling could enhance tympanic membrane immunity, improve blood circulation, promote the repair of damaged tympanic membranes, and ultimately prevent middle ear infections—an idea that has not been previously proposed.

Overall:

While the study addresses an interesting topic, the current version does not provide sufficiently strong or novel evidence to support its major conclusions. Additional mechanistic experiments and more rigorous validation would be necessary to substantiate the proposed model and clarify the relevance of the findings beyond this specific tissue context.

We greatly thank the two reviewers for their helpful critiques to improve our study. We especially thank the Section Editors for their insightful and constructive comments on this initial study.

References:

Shi, J., L. Hua, D. Harmer, P. Li and G. Ren (2018). "Cre Driver Mice Targeting Macrophages." Methods Mol Biol 1784: 263-275.

eLife Assessment

This high-N, multi-task study offers a comprehensive examination of rhythmicity in behavioral performance during listening. It presents a valuable set of findings that reveal task- and ear-specific effects, challenging the notion of a universal rhythmicity in auditory perception. While the evidence is solid, the study would benefit from a stronger conceptual framework to contextualize and explain the observed patterns. Nonetheless, the work is likely to be of significant interest to behavioral and cognitive scientists focused on perception and neural oscillations.

Reviewer #1 (Public review):

Summary:

This paper presents results from four independent experiments, each of which tests for rhythmicity in auditory perception. The authors report rhythmic fluctuations in discrimination performance at frequencies between 2 and 6 Hz. The exact frequency depends on the ear and experimental paradigm, although some frequencies seem to be more common than others.

Strengths:

The first sentence in the abstract describes the state of the art perfectly: "Numerous studies advocate for a rhythmic mode of perception; however, the evidence in the context of auditory perception remains inconsistent". This is precisely why the data from the present study is so valuable. This is probably the study with the highest sample size (total of > 100 in 4 experiments) in the field. The analysis is very thorough and transparent, due to the comparison of several statistical approaches and simulations of their sensitivity. Each of the experiments differs from the others in a clearly defined experimental parameter, and the authors test how this impacts auditory rhythmicity, measured in pitch discrimination performance (accuracy, sensitivity, bias) of a target presented at various delays after noise onset.

Weaknesses:

(1) The authors find that the frequency of auditory perception changes between experiments. I think they could exploit differences between experiments better to interpret and understand the obtained results. These differences are very well described in the Introduction, but don't seem to be used for the interpretation of results. For instance, what does it mean if perceptual frequency changes from between- to within-trial pitch discrimination? Why did the authors choose this experimental manipulation? Based on differences between experiments, is there any systematic pattern in the results that allows conclusions about the roles of different frequencies? I think the Discussion would benefit from an extension to cover this aspect.

(2) The Results give the impression of clear-cut differences in relevant frequencies between experiments (e.g., 2 Hz in Experiment 1, 6 Hz in Exp 2, etc), but they might not be so different. For instance, a 6 Hz effect is also visible in Experiment 1, but it just does not reach conventional significance. The average across the three experiments is therefore very useful, and also seems to suggest that differences between experiments are not very pronounced (otherwise the average would not produce clear peaks in the spectrum). I suggest making this point clearer in the text.

(3) I struggle to understand the hypothesis that rhythmic sampling differs between ears. In most everyday scenarios, the same sounds arrive at both ears, and the time difference between the two is too small to play a role for the frequencies tested. If both ears operate at different frequencies, the effects of the rhythm on overall perception would then often cancel out. But if this is the case, why would the two ears have different rhythms to begin with? This could be described in more detail.

Reviewer #2 (Public review):

Summary:

The current study aims to shed light on why previous work on perceptual rhythmicity has led to inconsistent results. They propose that the differences may stem from conceptual and methodological issues. In a series of experiments, the current study reports perceptual rhythmicity in different frequency bands that differ between different ear stimulations and behavioral measures. The study suggests challenges regarding the idea of universal perceptual rhythmicity in hearing.

Strengths:

The study aims to address differences observed in previous studies about perceptual rhythmicity. This is important and timely because the existing literature provides quite inconsistent findings. Several experiments were conducted to assess perceptual rhythmicity in hearing from different angles. The authors use sophisticated approaches to address the research questions.

Weaknesses:

(1) Conceptional concerns:

The authors place their research in the context of a rhythmic mode of perception. They also discuss continuous vs rhythmic mode processing. Their study further follows a design that seems to be based on paradigms that assume a recent phase in neural oscillations that subsequently influence perception (e.g., Fiebelkorn et al.; Landau & Fries). In my view, these are different facets in the neural oscillation research space that require a bit more nuanced separation. Continuous mode processing is associated with vigilance tasks (work by Schroeder and Lakatos; reduction of low frequency oscillations and sustained gamma activity), whereas the authors of this study seem to link it to hearing tasks specifically (e.g., line 694). Rhythmic mode processing is associated with rhythmic stimulation by which neural oscillations entrain and influence perception (also, Schroeder and Lakatos; greater low-frequency fluctuations and more rhythmic gamma activity). The current study mirrors the continuous rather than the rhythmic mode (i.e., there was no rhythmic stimulation), but even the former seems not fully fitting, because trials are 1.8 s short and do not really reflect a vigilance task. Finally, previous paradigms on phase-resetting reflect more closely the design of the current study (i.e., different times of a target stimulus relative to the reset of an oscillation). This is the work by Fiebelkorn et al., Landau & Fries, and others, which do not seem to be cited here, which I find surprising. Moreover, the authors would want to discuss the role of the background noise in resetting the phase of an oscillation, and the role of the fixation cross also possibly resetting the phase of an oscillation. Regardless, the conceptional mixture of all these facets makes interpretations really challenging. The phase-reset nature of the paradigm is not (or not well) explained, and the discussion mixes the different concepts and approaches. I recommend that the authors frame their work more clearly in the context of these different concepts (affecting large portions of the manuscript).

(2) Methodological concerns:

The authors use a relatively unorthodox approach to statistical testing. I understand that they try to capture and characterize the sensitivity of the different analysis approaches to rhythmic behavioral effects. However, it is a bit unclear what meaningful effects are in the study. For example, the bootstrapping approach that identifies the percentage of significant variations of sample selections is rather descriptive (Figures 5-7). The authors seem to suggest that 50% of the samples are meaningful (given the dashed line in the figure), even though this is rarely reached in any of the analyses. Perhaps >80% of samples should show a significant effect to be meaningful (at least to my subjective mind). To me, the low percentage rather suggests that there is not too much meaningful rhythmicity present. I suggest that the authors also present more traditional, perhaps multi-level, analyses: Calculation of spectra, binning, or single-trial analysis for each participant and condition, and the respective calculation of the surrogate data analysis, and then comparison of the surrogate data to the original data on the second (participant) level using t-tests. I also thought the statistical approach undertaken here could have been a bit more clearly/didactically described as well.

The authors used an adaptive procedure during the experimental blocks such that the stimulus intensity was adjusted throughout. In practice, this can be a disadvantage relative to keeping the intensity constant throughout, because, on average, correct trials will be associated with a higher intensity than incorrect trials, potentially making observations of perceptual rhythmicity more challenging. The authors would want to discuss this potential issue. Intensity adjustments could perhaps contribute to the observed rhythmicity effects. Perhaps the rhythmicity of the stimulus intensity could be analyzed as well. In any case, the adaptive procedure may add variance to the data.

Additional methodological concerns relate to Figure 8. Figures 8A and C seem to indicate that a baseline correction for a very short time window was calculated (I could not find anything about this in the methods section). The data seem very variable and artificially constrained in the baseline time window. It was unclear what the reader might take from Figure 8.

Motivation and discussion of eye-movement/pupillometry and motor activity: The dual task paradigm of Experiment 4 and the reasons for assessing eye metrics in the current study could have been better motivated. The experiment somehow does not fit in very well. There is recent evidence that eye movements decrease during effortful tasks (e.g., Contadini-Wright et al. 2023 J Neurosci; Herrmann & Ryan 2024 J Cog Neurosci), which appears to contradict the results presented in the current study. Moreover, by appealing to active sensing frameworks, the authors suggest that active movements can facilitate listening outcomes (line 677; they should provide a reference for this claim), but it is unclear how this would relate to eye movements. Certainly, a person may move their head closer to a sound source in the presence of competing sound to increase the signal-to-noise ratio, but this is not really the active movements that are measured here. A more detailed discussion may be important. The authors further frame the difference between Experiments 1 and 2 as being related to participants' motor activity. However, there are other factors that could explain differences between experiments. Self-paced trials give participants the opportunity to rest more (inter-trial durations were likely longer in Experiment 2), perhaps affecting attentional engagement. I think a more nuanced discussion may be warranted.

Discussion:

The main data in Figure 3 showed little rhythmicity. The authors seem to glance over this fact by simply stating that the same phase is not necessary for their statistical analysis. Previous work, however, showed rhythmicity in the across-participant average (e.g., Fiebelkorn's and similar work). Moreover, one would expect that some of the effects in the low-frequency band (e.g., 2-4 Hz) are somewhat similar across participants. Conduction delays in the auditory system are much smaller than the 0.25-0.5 s associated with 2-4 Hz. The authors would want to discuss why different participants would express so vastly different phases that the across-participant average does not show any rhythmicity, and what this would mean neurophysiologically.

An additional point that may require more nuanced discussion is related to the rhythmicity of response bias versus sensitivity. The authors could discuss what the rhythmicity of these different measures in different frequency bands means, with respect to underlying neural oscillations.

Figures:

Much of the text in the figures seems really small. Perhaps the authors would want to ensure it is readable even for those with low vision abilities. Moreover, Figure 1A is not as intuitive as it could be and may perhaps be made clearer. I also suggest the authors discuss a bit more the potential monoaural vs binaural issues, because the perceptual rhythmicity is much slower than any conduction delays in the auditory system that could lead to interference.

Reviewer #3 (Public review):

Summary:

The finding of rhythmic activity in the brain has, for a long time, engendered the theory of rhythmic modes of perception, that humans might oscillate between improved and worse perception depending on states of our internal systems. However, experiments looking for such modes have resulted in conflicting findings, particularly in those where the stimulus itself is not rhythmic. This paper seeks to take a comprehensive look at the effect and various experimental parameters which might generate these competing findings: in particular, the presentation of the stimulus to one ear or the other, the relevance of motor involvement, attentional demands, and memory: each of which are revealed to effect the consistency of this rhythmicity.

The need the paper attempts to resolve is a critical one for the field. However, as presented, I remain unconvinced that the data would not be better interpreted as showing no consistent rhythmic mode effect. It lacks a conceptual framework to understand why effects might be consistent in each ear but at different frequencies and only for some tasks with slight variants, some affecting sensitivity and some affecting bias.

Strengths:

The paper is strong in its experimental protocol and its comprehensive analysis, which seeks to compare effects across several analysis types and slight experiment changes to investigate which parameters could affect the presence or absence of an effect of rhythmicity. The prescribed nature of its hypotheses and its manner of setting out to test them is very clear, which allows for a straightforward assessment of its results

Weaknesses:

There is a weakness throughout the paper in terms of establishing a conceptual framework both for the source of "rhythmic modes" and for the interpretation of the results. Before understanding the data on this matter, it would be useful to discuss why one would posit such a theory to begin with. From a perceptual side, rhythmic modes of processing in the absence of rhythmic stimuli would not appear to provide any benefit to processing. From a biological or homeostatic argument, it's unclear why we would expect such fluctuations to occur in such a narrow-band way when neither the stimulus nor the neurobiological circuits require it.

Secondly, for the analysis to detect a "rhythmic mode", it must assume that the phase of fluctuations across an experiment (i.e., whether fluctuations are in an up-state or down-state at onset) is constant at stimulus onset, whereas most oscillations do not have such a total phase-reset as a result of input. Therefore, some theoretical positing of what kind of mechanism could generate this fluctuation is critical toward understanding whether the analysis is well-suited to the studied mechanism.

Thirdly, an interpretation of why we should expect left and right ears to have distinct frequency ranges of fluctuations is required. There are a large number of statistical tests in this paper, and it's not clear how multiple comparisons are controlled for, apart from experiment 4 (which specifies B&H false discovery rate). As such, one critical method to identify whether the results are not the result of noise or sample-specific biases is the plausibility of the finding. On its face, maintaining distinct frequencies of perception in each ear does not fit an obvious conceptual framework.

Author response:

We are grateful to the reviewers for their extensive and constructive feedback. In large the three reviewers noted the following main points:

(1) The overall evidence for any rhythmicity in this data is not ‘very strong’.

We do agree and will tone down the conclusions accordingly. However, as one of the reviewers noted, a qualitative interpretation of the specific statistical results remains somewhat vague and speculative by necessity.

(2) The differences between the results for the individual experiments are generally small. Yet, the same reviewer also asks for speculations as to how differences between experiments can be interpreted.

We will consider these, but also note that a clear demonstration of the robustness of specific effects requires the replication of individual experiments in a separate experiment.

(3) A clear-cut interpretation of the current experimental design in the context of continuous listening and true vigilance tasks remains difficult. This makes the interpretation and generalization of the results difficult.

We do agree in principle, but also note that task designs very widely in previous work, which may be one reason for why there is no clear consensus on the existence or absence of a rhythmic mode of listening. We will consider specific suggestions for future work to be included in the revision.

(4) The adjustment of task difficulty in the present task design may pose a challenge. Reviewers also suggest analyzing potential rhythmicity in this task difficulty parameter.

We will consider this for the revision.

(5) A more clear-cut interpretation of what potential differences in the rhythmicity of sensitivity and bias would mean should be included.

We will provide this in the revision.

(6) The study should provide a stronger conceptual framework both for the source of "rhythmic modes" and why one may expect differences between ears.

In large this has been put forward by many previous studies testing and reporting rhythmicity in auditory tasks.  Rhythmicity is pervasive in neural activity, but whether and how this relates to behavioral data remains less clear. These points will be clarified in a revision.

(7) Parallels to work in the visual domain by Fiebelkorn, Landau & Fries should be included.

We will discuss similarities and differences between studies on perceptual rhythmicity in the visual and auditory domains.

eLife Assessment

This study provides valuable insights into the evolution of pesticide resistance, demonstrating that resistance can arise rapidly and repeatedly, which complements prior work on parallel evolution across species. The combination of extensive temporal sampling in the field, experimental evolution, and genomics makes for convincing findings. The authors are to be commended for acknowledging the main limitations of their study in the Discussion. Framing the work in a broader context of resistance beyond arthropod pests would further increase the appeal of the study, which is of relevance for both agronomic practitioners and evolutionary biologists.

Reviewer #1 (Public review):

Summary:

The study by Cao et al. provides a compelling investigation into the role of mutational input in the rapid evolution of pesticide resistance, focusing on the two-spotted spider mite's response to the recent introduction of the acaricide cyetpyrafen. This well-documented introduction of the pesticide - and thus a clearly defined history of selection - offers a powerful framework for studying the temporal dynamics of rapid adaptation. The authors combine resistance phenotyping across multiple populations, extensive resequencing to track the frequency of resistance alleles, and genomic analyses of selection in both contemporary and historical samples. These approaches are further complemented by laboratory-based experimental evolution, which serves as a baseline for understanding the genetic architecture of resistance across mite populations in China. Their analyses identify two key resistance-associated genes, sdhB and sdhD, within which they detect 15 mutations in wild-collected samples. Protein modeling reveals that these mutations cluster around the pesticide's binding site, suggesting a direct functional role in resistance. The authors further examine signatures of selective sweeps and their distribution across populations to infer the mechanisms - such as de novo mutation or gene flow-driving the spread of resistance, a crucial consideration for predicting evolutionary responses to extreme selection pressure. Overall, this is a well-rounded, thoughtfully designed, and well-written manuscript. It shows significant novelty, as it is relatively rare to integrate broad-scale evolutionary inference from natural populations with experimentally informed bioassays, however, some aspects of the methods and discussion have an opportunity to be clarified and strengthened.

Strengths:

One of the most compelling aspects of this study is its integration of genomic time-series data in natural populations with controlled experimental evolution. By coupling genome sequencing of resistant field populations with laboratory selection experiments, the authors tease apart the individual effects of resistance alleles along with regions of the genome where selection is expected to occur, and compare that to the observed frequency in the wild populations over space and time. Their temporal data clearly demonstrates the pace at which evolution can occur in response to extreme selection. This type of approach is a powerful roadmap for the rest of the field of rapid adaptation.

The study effectively links specific genetic changes to resistance phenotypes. The identification of sdhB and sdhD mutations as major drivers of cyetpyrafen resistance is well-supported by allele frequency shifts in both field and experimental populations. The scope of their sampling clearly facilitated the remarkable number of observed mutations within these target genes, and the authors provide a careful discussion of the likelihood of these mutations from de novo or standing variation. Furthermore, the discovered cross-resistance that these mutations confer to other mitochondrial complex II inhibitors highlights the potential for broader resistance management and evolution.

Weaknesses:

(1) Experimental Evolution:

- Additional information about the lab experimental evolution would be useful in the main text. Specifically, the dose of cyetpyrafen used should be clarified, especially with respect to the LD50 values. How does it compare to recommended field doses? This is expected to influence the architecture of resistance evolution. What was the sample size? This will help readers contextualize how the experimental design could influence the role of standing variation.

- The finding that lab-evolved strains show cross-resistance is interesting, but potentially complicates the story. It would help to know more about the other mitochondrial complex II inhibitors used across China and their impact on adaptive dynamics at these loci, particularly regarding pre-existing resistance alleles. For example, a comparison of usage data from 2013, 2017, and 2019 could help explain whether cyetpyrafen was the main driver of resistance or if previous pesticides played a role. What happened in 2020 that caused such rapid evolution 3 years after launch?

(2) Evolutionary history of resistance alleles:

- It would be beneficial to examine the population structure of the sampled populations, especially regarding the role of migration. Though resistance evolution appears to have had minimal impact on genome-wide diversity (as shown in Supplementary Figure 2), could admixture be influencing the results? An explicit multivariate regression framework could help to understand factors influencing diversity across populations, as right now much is left to the readers' visual acuity.

- It is unclear why lab populations were included in the migration/treemix analysis. We might suggest redoing the analysis without including the laboratory populations to reveal biologically plausible patterns of resistance evolution.

- Can the authors explore isolation by distance (IBD) in the frequency of resistance alleles?

- Given the claim regarding the novelty of the number of pesticide resistance mutations, it is important to acknowledge the evolution of resistance to all pesticides (antibiotics, herbicides, etc.). ALS-inhibiting herbicides have driven remarkable repeatability across species based on numerous SNPs within the target gene.

- Figure 5 A-B. Why not run a multivariate regression with status at each resistance mutation encoded as a separate predictor? It is interesting that focusing on the predominant mutation gives the strongest r2, but it is somewhat unintuitive and masks some interesting variation among populations.

(3) Haplotype Reconstruction (Line 271-):

- We are a bit sceptical of the methods taken to reconstruct these haplotypes. It seems as though the authors did so with Sanger sequencing (this should be mentioned in the text), focusing only on homozygous SNPs. How many such SNPs were used to reconstruct haplotypes, along what length of sequence? For how many individuals were haplotypes reconstructed? Nonetheless, I appreciated that the authors looked into the extent to which the reconstructed haplotypes could be driven by recombination. Can the authors elaborate on the calculations in line 296? Is that the census population size estimate or effective?

(4) Single Mutations and Their Effect (line 312-):

- It's not entirely clear how the breeding scheme resulted in near-isogenic lines. Could the authors provide a clearer explanation of the process and its biological implications?

- If they are indeed isogenic, it's interesting that individual resistance mutations have effects on resistance that vary considerably among lines. Could the authors run a multivariate analysis including all potential resistance SNPs to account for interactions between them? Given the variable effects of the D116G substitution (ranging from 4-25%), could polygenic or epistatic factors be influencing the evolution of resistance?

- Why are there some populations that segregate for resistance mutations but have no survival to pesticides (i.e., the green points in Figure 5)? Some discussion of this heterogeneity seems required in the absence of validation of the effects of these particular mutations. Could it be dominance playing a role, or do the authors have some other explanation?

- The authors mention that all resistance mutations co-localized to the Q-site. Is this where the pesticide binds? This seems like an important point to follow their argument for these being resistance-related.

(5) Statistical Considerations for Allele Frequency Changes (Figure 3):

- It might be helpful to use a logistic regression model to assess the rate of allele frequency changes and determine the strength of selection acting on these alleles (e.g., Kreiner et al. 2022; Patel et al. 2024). This approach could refine the interpretation of selection dynamics over time.

Reviewer #2 (Public review):

Summary:

This paper investigates the evolution of pesticide resistance in the two-spotted spider mite following the introduction of an SDHI acaricide, cyatpyrafen, in China. The authors make use of cyatpyrafen-naive populations collected before that pesticide was first used, as well as more recent populations (both sensitive and resistant) to conduct comparative population genomics. They report 15 different mutations in the insecticide target site from resistant populations, many reported here for the first time, and look at the mutation and selection processes underlying the evolution of resistance, through GWAS, haplotype mapping, and testing for loss of diversity indicating selective sweeps. None of the target site mutations found in resistant populations was found in pre-exposure populations, suggesting that the mutations may have arisen de novo rather than being present as standing variation, unless initially present at very low frequencies; a de novo origin is also supported by evidence of selective sweeps in some resistant populations. Furthermore, there is no significant evidence of migration of resistant genotypes between the sampled field populations, indicating multiple origins of common mutations. Overall, this indicates a very high mutation rate and a wide range of mutational pathways to resistance for this target site in this pest species. The series of population genomic analyses carried out here, in addition to the evolutionary processes that appear to underlie resistance development in this case, could have implications for the study of resistance evolution more widely.

Strengths:

This paper combines phenotypic characterisation with extensive comparative population genomics, made possible by the availability of multiple population samples (each with hundreds of individuals) collected before as well as after the introduction of the pesticide cyatpyrafen, as well as lab-evolved lines. This results in findings of mutation and selection processes that can be related back to the pesticide resistance trait of concern. Large numbers of mites were tested phenotypically to show the levels of resistance present, and the authors also made near-isogenic lines to confirm the phenotypic effects of key mutations. The population genomic analyses consider a range of alternative hypotheses, including mutations arising by de novo mutation or selection from standing genetic variation, and mutations in different populations arising independently or arriving by migration. The claim that mutations most likley arose by multiple repeated de novo mutations is therefore supported by multiple lines of evidence: the direct evidence of none of the mutations being found in over 2000 individuals from naive populations, and the indirect evidence from population genomics showing evidence of selective sweeps but not of significant migration between the sampled populations.

Weaknesses:

As acknowledged within the discussion, whilst evidence supports a de novo origin of the resistance-associated mutations, this cannot be proven definitively as mutations may have been present at a very low frequency and therefore not found within the tested pesticide-naive population samples.

Near-isofemale lines were made to confirm the resistance levels associated with five of the 15 mutations, but otherwise, the genotype-phenotype associations are correlative, as confirmation by functional genetics was beyond the scope of this study.

eLife Assessment

This important study demonstrates that lipid binding can regulate the dimerization state of the SARS-CoV2 Orf9b protein. The data from biophysical and cellular experiments, along with mathematical modeling, are convincing. However, this study can further benefit from more rigorous quantitative analyses and from resolving the role of dimerization in viral infection and host innate responses. This paper is broadly relevant to those studying coupled equilibria across all aspects of biology.

Reviewer #1 (Public review):

Summary:

Felipe and colleagues try to answer an important question in Sarbecovirus Orf9b-mediated interferon signaling suppression, given that this small viral protein adopts two distinct conformations, a dimeric β-sheet-rich fold and a helix-rich monomeric fold when bound by Tom70 protein. Two Orf9b structures determined by X-ray crystallography and Cryo-EM suggest an equilibrium between the two Orf9b conformations, and it is important to understand how this equilibrium relates to its functions. To answer these questions, the authors developed a series of ordinary differential equations (ODE) describing the Orf9b conformation equilibrium between homodimers and monomers binding to Tom70. They used SPR and a fluorescent polarization (FP) peptide displacement assay to identify parameters for the equilibrium and create a theoretical model. They then used the model to characterize the effect of lipid-binding and the effects of Orf9b mutations in homodimer stability, lipid binding, and dimer-monomer equilibrium. They used their model to further analyze dimerization, lipid binding, and Orf9b-Tom70 interactions for truncated Orf9b, Orf9b fusion mutant S53E (blocking Tom70 binding), and Orf9b from a set of Sars-CoV-2 VOCs. They evaluated the ability of different Orf9b variants for binding Tom70 using Co-IP experiments and assessed their activity in suppressing IFN signaling in cells.

Overall, this work is well designed, the results are of high quality and well-presented; the results support their conclusions.

Strengths:

(1) They developed a working biophysical model for analyzing Orf9b monomer-dimer equilibrium and Tom70 binding based on SPR and FP experiments; this is an important tool for future investigation.

(2) They prepared lipid-free Orf9b homodimer and determined its crystal structure.

(3) They designed and purified obligate Orf9b monomer, fused-dimer, etc., a very important Orf9b variant for further investigations.

(4) They identified the lipid bound by Orf9b homodimer using mass spectra data.

(5) They proposed a working model of Orf9b-Tom70 equilibrium.

Weaknesses:

(1) It is difficult to understand why the obligate Orf9b dimer has similar IFN inhibition activity as the WT protein and obligate Orf9b monomer truncations.

(2) The role of Orf9b homodimer and the role of Orf9b-bound lipid in virus infection, remains unknown.

Reviewer #2 (Public review):

Summary:

This study focuses on Orf9b, a SARS-COV1/2 protein that regulates innate signaling through interaction with Tom70. San Felipe et al use a combination of biophysical methods to characterize the coupling between lipid-binding, dimerization, conformational change, and protein-protein-interaction equilibria for the Orf9b-Tom70 system. Their analysis provides a detailed explanation for previous observations of Orf9b function. In a cellular context, they find other factors may also be important for the biological functioning of Orf9b.

Strengths:

San Felipe et al elegantly combine structural biology, biophysics, kinetic modelling, and cellular assays, allowing detailed analysis of the Orf9b-Tom70 system. Such complex systems involving coupled equilibria are prevalent in various aspects of biology, and a quantitative description of them, while challenging, provides a detailed understanding and prediction of biological outcomes. Using SPR to guide initial estimates of the rate constants for solution measurements is an interesting approach.

Weaknesses:

This study would benefit from a more quantitative description of uncertainties in the numerous rate constants of the models, either through a detailed presentation of the sensitivity analysis or another approach such as MCMC. Quantitative uncertainty analysis, such as MCMC is not trivial for ODEs, particularly when they involve many parameters and are to be fitted to numerous data points, as is the case for this study. The authors use sensitivity analysis as an alternative, however, the results of the sensitivity analysis are not presented in detail, and I believe the authors should consider whether there is a way to present this analysis more quantitatively. For example, could the residuals for each +/-10% parameter change for the peptide model be presented as a supplementary figure, and similarly for the more complex models? Further details of the range of rate constants tested would be useful, particularly for the ka and kB parameters.

The authors build a model that incorporates an α-helix-β-sheet conformational change, but the rate constant for the conversion to the α-helix conformation is required to be second order. Although the authors provide some rationale, I do not find this satisfactorily convincing given the large number of adjustable parameters in the model and the use of manual model fitting. The authors should discuss whether there is any precedence for second-order rate constants for conformational changes in the literature. On page 14, the authors state this rate constant "had to be non-linear in the monomer β-sheet concentration" - how many other models did the authors explore? For example, would αT↔α↔αα↔ββ (i.e., conformational change before dimer dissociation) or α↔βαT↔ββ (i.e., Tom70 binding driving dimer dissociation) be other plausible models for the conformational change that do not require assumptions of second-order rate constants for the conformational change?

Overall, this study progresses the analysis of coupled equilibria and provides insights into Orf9b function.

eLife Assessment

Despite the conserved anti-inflammatory activity in birds, whether IL-10 also controls avian intestinal homeostasis remains unclear. Generating genetic knockouts, Meunier et al. firmly established that a complete lack of IL-10 strengthened immunity against enteric bacteria in chickens, while also aggravating infection-inflicted tissue damage upon parasite infection. The findings presented in this manuscript are valuable, and the strength of evidence is convincing; however, it is advised that the deficiencies and weaknesses pointed out by all the reviewers are meticulously addressed.

Reviewer #1 (Public review):

Summary:

In this study, Meunier et al. investigated the functional role of IL-10 in avian mucosal immunity. While the anti-inflammatory role of IL-10 is well established in mammals, and several confirmatory knockout models are available in mice, IL-10's role in avian mucosal immunity is so far correlative. In this study, the authors generated two different models of IL-10 ablation in Chickens. A whole body knock-out model and an enhancer KO model leading to reduced IL10 expression. The authors first performed in vitro LPS stimulation-based experiments, and then in vivo two different infection models employing C. jejuni and E. tenella, to demonstrate that complete ablation of IL10 leads to enhanced inflammation-related pathology and gene expression, and enhanced pathogen clearance. At a steady-state level, however, IL-10 ablation did not lead to spontaneous colitis.

Strengths:

Overall, the study is well executed and establishes an anti-inflammatory role of IL-10 in birds. While the results are expected and not surprising, this appears to be the first report to conclusively demonstrate IL-10's anti-inflammatory role upon its genetic ablation in the avian model. Provided this information is applicable in combating pathogen infection in livestock species in sustainable industries like poultry, the study will be of interest to the field.

Weaknesses:

The study is primarily a confirmation of the already established anti-inflammatory role of IL-10.

Reviewer #2 (Public review):

Summary:

The authors were to investigate the functional role of IL10 on mucosal immunity in chickens. CRISPR technology was employed to generate IL10 knock-out chickens in both exon and putative enhancer regions. IL10 expressions were either abolished (knockout in exon) or reduced (enhancer knock-out). IL-10 plays an important role in the composition of the caecal microbiome. Through various enteric pathogen challenges, deficient IL10 expression was associated with enhanced pathogen clearance, but with more severe lesion scores and body weight loss.

Strengths:

Both in vitro and in vivo knock-out abolished and reduced IL10 expression, and broad enteric pathogens were challenged in vivo, and various parameters were examined to evaluate the functional role of IL10 on mucosal immunity.

Weaknesses:

Overexpression of IL-10 either in vitro or in vivo may further support the findings from this study.

eLife Assessment

This work provides fundamental findings on how the mouse barrel cortex connects to the dorsolateral striatum, uncovering that inputs from discrete whisker cortical columns are convergent and SPN-specific, but topographically organized at the population level. The evidence supporting this claim is compelling, demonstrating that SPNs uniquely integrate sparse input from variable stretches across the barrel cortex. The study would be of interest to basal ganglia and sensory-motor integration researchers.

Reviewer #1 (Public review):

Summary:

By applying a laser scanning photostimulation (LSPS) approach to a novel slice preparation, the authors aimed to study the degree of convergence and divergence of cortical inputs to individual striatal projection neurons (SPNs).

Strengths:

The experiments were well-designed and conducted, and data analysis was thorough. The manuscript was well written, and related work in the literature was properly discussed. This work has the potential to advance our understanding of how sensory inputs are integrated into the striatal circuits.

Weaknesses:

This work focuses on the connection strength of the corticostriatal projections, without considering the involvement of synaptic plasticity in sensory integration.

Reviewer #2 (Public review):

Summary:

How corticostriatal synaptic connectivity gives rise to SPN encoding of sensory information is an important and currently unanswered question. The authors utilize a clever slice preparation in combination with electrophysiology and glutamate uncaging to dissect the synaptic connectivity between barrel cortex and individual striatal SPNs. In addition to mapping connectivity across major anatomical axes and cortical layers, the authors provide data showing that SPNs uniquely integrate sparse input from variable stretches across barrel cortex.

Strengths:

The methodology shows impressive rigor, and the data robustly support the authors' conclusions. Overall, the manuscript addresses its core question, provides valuable insights into corticostriatal architecture, and is a welcome addition to the field.

Weaknesses:

A few minor changes to the figures and text could be made to improve clarity.

Reviewer #3 (Public review):

Summary:

The authors explored how individual dorsolateral striatum (DLS) spiny projection neurons (SPNs) receive functional input from whisker-related cortical columns. The authors developed and validated a novel slice preparation and method to which they applied rigorous functional mapping and thorough analysis. They found that individual SPNs were driven by sparse, scattered cortical clusters. Interestingly, while the cortical input fields of nearby SPNs had some degree of overlap, connectivity per SPN was largely distinct. Despite sparse, heterogeneous connectivity, topographical organization was identified. The authors lastly compared direct (D1) vs. indirect (D2) pathway cells, concluding that overall connectivity patterns were the same, but D1 cells received stronger input from L6 and D2 cells from L2/3. The paper thoughtfully addresses the question of whether barrel cortex broadly or selectively innervates SPNs. Their results indicate selective input that is loosely topographic. Their work deepens the understanding of how whisker-related somatosensory signals can drive striatal neurons.

Strengths:

Overall, this is a carefully conducted study, and the major claims are well-supported. The use of a novel ex vivo slice prep that keeps relevant corticostriatal projections intact allows for careful mapping of the barrel cortex to dorsolateral striatum SPNs. Careful reporting of both columnar and layer position, as well as postsynaptic SPN type (D1 or D2), allows the authors to uncover novel details about how the dorsolateral striatum represents whisker-related sensory information.

Weaknesses:

(1) Several factors may contribute to an underestimation of barrel cortex inputs to SPNs (and thus an overestimate of the input heterogeneity among SPNs). First, by virtue of the experiments being performed in an acute slice prep, it is probable that portions of recorded SPN dendritic trees have been dissected (in an operationally consistent anatomical orientation). If afferents happen to systematically target the rostral/caudal projections of SPN dendritic fields, these inputs could be missed. Similarly, the dendritic locations of presynaptic cortical inputs remain unknown (e.g., do some inputs preferentially target distal vs proximal dendritic positions?). As synaptic connectivity was inferred from somatic recordings, it's likely that inputs targeting the proximal dendritic arbor are the ones most efficiently detected. Mapping the dendritic organization of synapses is beyond the scope of this work, but these points could be broached in the text.

(2) In general, how specific (or generalizable) is the observed SPN-specific convergence of cortical barrel cortex projections in the dorsolateral striatum? In other words, does a similar cortical stimulation protocol targeted to a non-barrel sensory (or motor) cortex region produce similar SPN-specific innervation patterns in the dorsolateral striatum?

(3) In general, some of the figure legends are extremely brief, making many details difficult to infer. Similarly, some statistical analyses were either not carried out or not consistently reported.

eLife Assessment

In this study, the authors investigated whether HIV-1 cell-to-cell transmission activates the CARD8 inflammasome in macrophages. The data convincingly support the idea that CARD8 is activated by the viral protease, promoting inflammation. The study's significance is further enhanced by including time-course analyses in primary T cells and macrophages and provides valuable insights into the role of CARD8 in HIV-induced inflammation.

Joint Public Review:

Following up on their previous work, the authors investigated whether HIV-1 cell-to-cell transmission activates the CARD8 inflammasome in macrophages, a key question given that inflammasome activation in myeloid cells triggers proinflammatory cytokine release. Co-cultures of HIV-infected T cells with macrophages led to viral spreading, resulting in IL1β release and cell death, with CARD8 playing a crucial role in this inflammasome response, triggered by HIV protease. The authors also found that HIV isolates resistant to protease inhibitors showed differences in CARD8 activation and IL1β production, highlighting the clinical relevance of their findings. Overall, this well-written study provides strong evidence for the role of CARD8 in protease-dependent sensing of viral spread, with implications for understanding chronic inflammation in HIV infections and its potential contribution to systemic immune activation, especially under ART. The authors have addressed initial weaknesses and verified effects in cocultures of primary T cells and macrophages. They now also provide evidence that CARD8 is activated by protease from incoming viral particles. Further studies are needed to clarify how much this mechanism contributes to systemic immune activation in untreated infections and whether this mechanism drives chronic inflammation under ART.

Author response:

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

We again thank you for the positive and constructive feedback on our manuscript, and for highlighting its contributions to understanding the role of CARD8 in viral protease-triggered sensing of viral spread, and the potential impact of our findings on chronic inflammation and immune activation. We agree that it will be important for future work to address whether or not HIV-1 protease-triggered CARD8 inflammasome activation contributes to chronic inflammation in PLWH who are receiving ART.

In response to the question about the baseline level of IL-1β in Fig. 4D, the figure below shows the mock condition for the CD4+ T cell:MDM coculture. We had done this control in parallel with the data presented in the submitted figure. Levels of IL-1β during HIV-1 infection are increased over background (i.e., mock infection). We note that for donor G the IL-1β concentration is below the limit of detection for this assay. Thus, it remains possible that other inflammasomes contribute modestly during cell-to-cell transmission of HIV-1; however, incomplete knockout of CARD8 in a minority of cells may also contribute to the observed levels of IL-1β in response to HIV-1 infection. Nonetheless, collectively, our data strongly supports the role for CARD8 in HIV-1 protease-triggered inflammasome activation.

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

Joint Public Review:

Following up on their previous work, the authors investigated whether cell-to-cell transmission of HIV-1 activates the CARD8 inflammasome in macrophages, an important question given that inflammasome activation in myeloid cells triggers proinflammatory cytokine release. The data support the idea that CARD8 is activated by the viral protease and promotes inflammation. However, time-course analyses in primary T cells and macrophages and further information on the specific inflammasome involved would further increase the significance of the study.

Strengths:

The manuscript is well-written and the data is of good quality. The evidence that CARD8 senses the HIV-1 protease in the context of cell-to-cell transmission is important since cell-to-cell transmission is thought to play a key role in viral spread in vivo, and inflammation is a major driver of disease progression. Clean knockout experiments in primary macrophages are a notable strength and the results clearly support the role of CARD8 in protease-dependent sensing of viral spread and the induction of IL1β release and cell death. The finding that HIV-1 strains are resistant to protease inhibitors differ in CARD8 activation and IL1β production is interesting and underscores the potential clinical relevance of these results.

Weaknesses:

One weakness is that the authors used T cell lines which might not faithfully reflect the efficiency of HIV-1 production and cell-cell transfer by primary T cells. To assess whether CARD8 is also activated by protease from incoming viral particles earlier time points should be analyzed. Finally, while the authors exclude the role of NLRP3 in IL-1b and the death of macrophages it would be interesting to know whether the effect is still Gasdermin D dependent.

Recommendations for the authors

(1) Co-culture assay should also be done between primary CD4 cells and primary MDMs, because T-cell lines produce much more viruses, and the efficiency of cell-tocell transmission might be dramatically different in primary cells compared to cell lines.

We have now added data from experiments using infected primary CD4 cells as the donor cells in cell-to-cell HIV-1 transmission to MDMs in new Figure 4. The results largely phenocopy the SUPT1:MDM coculture in that we observe inflammasome activation after co-culture of HIV-infected primary T cells with primary MDMs. We find that this inflammasome activity induced by the CD4:MDM cell-to-cell transmission is abrogated by knockout of CARD8 in the MDMs or treatment of HIV protease inhibitor lopinavir (LPV) or caspase 1 inhibitor VX765, suggesting that this activation is dependent on CARD8, HIV protease, and caspase 1. Additionally, the signal persists in the presence of reverse transcriptase inhibitor nevirapine (NVP), suggesting that the incoming protease is driving activation.

(2) For all co-culture experiments, supernatants were collected at 48 or 72 hours. Since CARD8 activation is expected to be driven by incoming viral particles without RT, they should measure cytokine production at much earlier time points. 2-3 days co-culture raises concerns. Ideally, the authors can provide a time-course.

We have now added a time course of the SUPT1:MDM coculture from 3 unique donors taken at 4, 24, 48, and 72 hours post coculture in the presence or absence of reverse transcriptase inhibitor (see new Figure 3B) as well as for the primary CD4 cells to MDM co-culture (see new Figure 4B). We detect IL-1β at the 24hour time point (and later), but not at the 4-hour time point which is slower than what was detected by direct cell-free infection (Kulsuptrakul et al., 2023). However, we still hypothesize that this is driven by active incoming viral protease because the signal is not abrogated by a reverse transcriptase inhibitor, which indicates that de novo protease production is not necessary. We also observed that IL-1β levels do not increase after plateauing 24h after establishing the co-culture, suggesting that secondary infection does not further amplify inflammasome activation. We now speculate on this in the Discussion.

(3) A potential confounder in the data in Figure 4 is that despite rightly including the cognate adaptations in the Gag cleavage sites with the PI-R protease mutants, some of these viruses still display Gag processing defects. Can the authors disentangle the potency of PR mutant cleavage with either reduced cell entry or reduced protease availability due to processing defects in the incoming virions?

The reviewer is correct that although the western blot with the p24^gag antibody suggests that Gag is processed, we cannot rule out that other variables do not contribute to the observed difference in CARD8 inflammasome activation. For example, PI-R clones relative to the LAI strain may have distinct protease substrate specificity, variable efficiency/kinetics in viral assembly, gag dimerization, and other factors may ultimately influence CARD8 inflammasome activation. We have updated the text to reflect these possibilities. Nonetheless, this argument does not change the conclusion that CARD8 inflammasome activation is affected by protease mutations acquired during drug resistance.

(4) There is considerable donor variation in the macrophages (unsurprising) but can the authors correlate this with CARD8 expression and are there any off-target effects on macrophage permissivity to HIV-1 infection?

We have now considerably increased the number of primary cell donors from the first submission (see Author response table 1 below). We find that the non-responsive donor presented in the first submission is aberrant since all others do respond to a greater or lesser degree (Figure 3, Figure 4). However, the reviewer may be correct that the particular aberrant donor MDMs were poorly infected. We also note that despite donor variability in the degree of activation (IL-1β secretion) from cocultures with HIV_BaL-infected SUPT1 cells, HIV-induced activation is comparable to the activation induced by VbP (see new Figure 3–figure supplement 1B). We do not see a notable difference in CARD8 expression between donors. Nonetheless, with the added number of primary cell donors, the data are consistent with a role of primary MDMs from nearly all donors in supporting a CARD8-dependent, HIV-protease dependent inflammasome response after co-culture with infected T cells. We have left in data from all of the donors so that readers can appreciate the variability among primary cells.

Author response table 1.

In addition, to address the reviewer concerns about off-target effects of the sgRNAs on macrophage permissivity, we assessed our CD4:MDM cocultures for percent infectivity via intracellular p24^gag in AAVS1 vs CARD8 KO MDMs and we observed no significant difference in infectivity in AAVS1 vs CARD8 KO MDMs (see Author response image 1 of MDMs after co-culture with T cells that is not affected any potential off-target effects of the sgRNAs.

Author response image 1.

Equivalent infection in AAVS1 vs CARD8 KOMDMs. AAVS1 or CARD8 KO from donor 12 were cocultured with mock or HIV infected CD4 T cells as described in Figure 4D for 72 hours then assessed for HIV infection of the MDMs by washing away CD4 T cells, harvesting MDMs, and staining attached MDMs for intracellular p24^gag for flow cytometry analysis. Datasets represent mean ± SD (n=2 technical replicates from one donor). One-way ANOVA with Dunnett’s test using GraphPad Prism 10. ns = not significant, *p<0.05,**p<0.01, ***p<0.001, ****p<0.0001.

(5) The authors suggest that NLRP3 is unlikely to be the mediator of IL-1b and cell death in the macrophages. Is this death still GSDMDdependent, what other NLRs are expressed in this system and does it make a difference what PAMP you use to prime the response?

We have now added additional data in support of the conclusion that NLRP3 is not a mediator of the IL-1β secretion in the infected SUPT1 cells to primary MDMs coculture. In addition to using an NLRP3 inhibitor, we have now also made NLRP3 KOs MDMs and used these in the coculture experiments which show that the IL-1β secretion after coculture of infected SUPT1 cells and primary MDMs is mediated by CARD8 and not NLRP3 because the signal is abrogated by CARD8 knockout, but not by NLRP3 knockout. This new data is shown in Figure 3C and D.

To assess the role of GSDMD, we treated SUPT1:MDM cocultures with disulfiram, a GSDMD inhibitor (Hu et al., 2020). Disulfiram treatment abrogated IL-1β secretion, suggesting that this activation is indeed GSDMD-mediated (see Author response image 2 below). We choose not to include the disulfiram result in the final manuscript since we have not ruled out cytotoxic effects of the drug.

There are likely other NLRs expressed in primary MDMs; however, since inflammasome activation is completely absent in the CARD8 KO MDMs, we infer that CARD8 is the main inflammasome-forming sensor in this system. However, we cannot rule out the possibility of other innate sensors being activated downstream of CARD8 or under different differentiation conditions.

To address the concern that alternative priming affects CARD8 activation, we compared pre-treatment of cells with Pam3CSK4 or lipopolysaccharide (LPS) in the presence or absence of HIV protease inhibitor and reverse transcriptase inhibitor. Regardless of the priming agent used, we observed HIV protease-dependent activation that persisted in the presence of reverse transcriptase inhibitor, suggesting that CARD8 is the main sensor under LPS and Pam3CSK4 priming (new Figure 3–figure supplement 1A).

Author response image 2.

Inflammasome activation following cell-to-cell HIV infection is mediated by GSDMD. SUPT1-CCR5 cells were either mock-infected or infected with HIV-1_NL4.3BaL for 20 hours before coculturing with MDMs in either the presence or absence of GSDMD inhibitor disulfarim (25μM). Cocultures were harvested 24 hours later to assess (left) IL-1β secretion via IL-1 reporter assay and (right) cell viability via CellTiter-Glo® assay. Viability was calculated by normalizing to relative luminescence units in the mock untreated control. Dotted line indicates limit of detection (LoD). Dashed line indicates 100% viability as determined by untreated mock control. Datasets represent mean ± SD (n=2 technical replicates for one donor). Two-way ANOVA with Sidak’s test (using GraphPad Prism 10. ns = not significant, *p<0.05,**p<0.01, ***p<0.001, ****p<0.0001.

Minor points

(1) In Figure 1, the authors should clarify whether LAI or LAI-VSV-G was used.

Wild-type virus (LAI strain) was used in Figure 1. This has now been clarified in the figure legend.

(2) In Figure 1, the fraction of infected cells without DEAE was ~20% in both WT and CARD8 KO THP-1, suggesting somewhat efficient viral entry even in the absence of DEAE. How do the authors reconcile this with the lack of IL-1β production? The increase in infection observed in WT THP-1 +DEAE was overall modest (from ~20% to 25-30%) compared to the dramatic difference in IL-1β production. Can they provide more evidence or discuss how DEAE might be impacting cytokine production? If differences in viral entry are the explanation for differences in inflammasome activation, then they should be able to overcome this by using virus at a higher MOI in the absence of DEAE. Experiments proposed in Figure 1 +/- DEAE should be repeated using a range of MOI for LAI and showing the corresponding percent infection in THP-1 cells (which is not shown in Figure S2 for LAI-VSVG).

We hypothesize that the lack of IL-1β production without DEAE is likely due to an insufficient amount of incoming viral protease to induce CARD8 activation. Though the increase in infection with DEAE is modest by intracellular p24^gag at 24 hours post infection, we infer that intracellular p24^gag may be largely underestimating the actual increase in viral efficiency achieved with DEAE (now in Supplemental Note). We have also updated Figure S2 (now Figure 2–figure supplement 1) legend to include the percent infection for HIV-1_LAI and HIV-1_LAI-VSVG infections. We agree that activation in the absence of DEAE could be overcome by infecting with a more concentrated viral stock to increase the MOI. Indeed, our decision to use the cell-to-cell transmission model achieves this in a more physiologic context.

(3) In Figure S1, the authors point out that RT-activity in the supernatants was similar in the cell-free vs. cell-to-cell model. While in the transwell system THP-1 cells are the only cells capable of producing new virions, how are they able to differentiate viral production from sup-T1 vs. THP-1 in the cell-to-cell system? At a minimum, they should provide some data on the observed RT activity in matching wells containing the same number of infected sup-T1 cells utilized in coculture experiments.

We think this may have been a misinterpretation. In Figure S1 (now Figure 1B, right), we compare the amount of virus available in the lower chamber of the transwell versus the cell-to-cell condition. We are not comparing cell-free to cell-to-cell infection. We have changed the text and figure title to clarify this point.

(4) Can the authors provide additional comments on the lack of IL-1β release in donor C in Figure 3? The donor did not produce IL-1β in response to VbP or HIV, although the WB for CARD8 appears similar to the other two donors.

We have now tested MDMs from additional donors and continue to find a range of IL-1β secretion after the coculture. However, donor C is aberrant since each of the other donors had detectable IL-1β secretion in response to VbP and HIV-1 to greater or lesser extents. Nonetheless, we have included additional donors summarized in the table above corresponding to major comment #4.

(5) For Figure 3, can the authors provide information on the fraction of MDMs that were infected after coculture with sup-T1 cells? Why didn't the authors measure cell death in MDMs?

It is difficult to measure the fraction of MDMs infected or dying in the cocultures since it is hard to separate signal from the T cells. Although it would be possible to do so, in this manuscript, we instead prefer to focus on the potential contribution of CARD8 inflammasome activation in exacerbating chronic inflammation in response to HIV rather than the depletion of macrophages.

(6) In Figure 4, did the authors introduce the mutations associated with PI resistance into the same LAI backbone? If not, this is not a fair comparison, as viral protein expression levels were not at the same level, indicated in Figure 4A. Additionally, such comparison will be further strengthened by using cells other than 293T cells for the coculture assay.

No, we did not introduce these mutations into LAI, since they were already in an NL4.3 backbone and NL4.3 and LAI differ by only 1 amino acid in protease. We have updated Table S1 to report this amino acid difference. We also note that in our previous manuscript we tested much more diverse proteases such as a clade A HIV-1, HIV-2, and SIVs and find comparable CARD8 cleavage to LAI.

Additions not requested by Reviewers:

THP-1 characterization

In our previous work, we noticed that different “wildtype” THP-1 lines behaved uniquely in response to DEAE-dextran. In particular, we observed inflammasome activation in response to DEAE-dextran alone at the concentration used for spinoculations (20μg/mL), whereas the other THP-1 line did not. Thus, we performed STR profiling on each THP-1 cell line and determined that the THP-1 cells used in our studies (JK THP1s) are distinct from THP-1 cells from ATCC at 3 different loci. This data is now included in the Supplemental Note (Figure A1). Please note that all data in this and the accompanying manuscript were performed in JK THP-1 cells.

Whole plasmid sequencing of the PI-resistant HIV clones

Since preprint submission, we have done whole plasmid Oxford Nanopore sequencing on the PI-resistant HIV clones obtained from the NIAID HIV/AIDS Specimen Repository Program. Of note, there were a handful of previously unreported mutations included in these plasmid stocks within protease. We have updated Table S1 to include an additional column titled “Additional amino acid changes in HIV^PR relative to NL4.3.”

References

Hu JJ, Liu X, Xia S, Zhang Z, Zhang Y, Zhao J, Ruan J, Luo X, Lou X, Bai Y, Wang J, Hollingsworth LR, Magupalli VG, Zhao L, Luo HR, Kim J, Lieberman J, Wu H. 2020. FDA-approved disulfiram inhibits pyroptosis by blocking gasdermin D pore formation. Nat Immunol 21:736–745. doi:10.1038/s41590-020-0669-6

Kulsuptrakul J, Turcotte EA, Emerman M, Mitchell PS. 2023. A human-specific motif facilitates CARD8 inflammasome activation after HIV-1 infection. eLife 12:e84108. doi:10.7554/eLife.84108

eLife Assessment

This paper addresses a valuable research question on the heritability of the brain's response to movie watching, given various parameters such as regional spatial hyperalignment and BOLD frequency bands. The topic of this paper would be of interest to fMRI methodological experts, and potentially to a broader cognitive neuroscience audience, and those with an interest in understanding the heritable sources of individual differences in brain function. However, the current findings provide incomplete support for the conclusions, since several key methodological concerns need to be addressed to ensure the validity of the analyses and results.

Reviewer #1 (Public review):

Summary:

Gruskin and colleagues use twin data from a movie-watching fMRI paradigm to show how genetic control of cortical function intersects with the processing of naturalistic audiovisual stimuli. They use hyperalignment to dissect heritability into the components that can be explained by local differences in cortical-functional topography and those that cannot. They show that heritability is strongest at slower-evolving neural time scales and is more evident in functional connectivity estimates than in response time series.

Strengths:

This is a very thorough paper that tackles this question from several different angles. I very much appreciate the use of hyperalignment to factor out topographic differences, and I found the relationship between heritability and neural time scales very interesting. The writing is clear, and the results are compelling.

Weaknesses:

The only "weaknesses" I identified were some points where I think the methods, interpretation, or visualization could be clarified.

(1) On page 16, the authors compare heritability in functional connectivity (FC) and response time series, and find that the heritability effect is larger in FC. In general, I agree with your diagnosis that this is in large part due to the fact that FC captures the covariance structure across parcels, whereas response time series only diverge in terms of univariate time-point-by-time-point differences. Another important factor here is that (within-subject) FC can be driven by intrinsic fluctuations that occur with idiosyncratic timing across subjects and are unrelated to the stimulus (whereas time-locked metrics like ISC and time-series differences cannot, by definition). This makes me wonder how this connectivity result would change if the authors used intersubject functional connectivity (ISFC) analysis to specifically isolate the stimulus-driven components of functional connectivity (Simony et al., 2016). This, to me, would provide a closer comparison to the ISC and response time series results, and could allow the authors to quantify how much of the heritability in FC is intrinsic versus stimulus-driven. I'm not asking that the authors actually perform this analysis, as I don't think it's critical for the message of the manuscript, but it could be an interesting future direction. As the authors discuss on page 17, I also suspect there's something fundamentally shared between response time series and connectivity as they relate to functional topography (Busch et al., 2021) that drives part of the heritability effect.

(2) The observation that regions with intermediate ISC have the largest differences between MZ, DZ, and UR is very interesting, but it's kind of hard to see in Figure 1B. Is there any other way to plot this that might make the effect more obvious? For example, I could imagine three scatter plots where the x- and y-axes are, e.g., MZ ISC and UR ISC, and each data point is a parcel. In this kind of plot, I would expect to see the middle values lifted visibly off the diagonal/unity line toward MZ. The authors could even color the data points according to networks, like in Figure 3C. (They also might not need to scale the ISC axis all the way to r = 1, which would make the differences more visible.)

(3) On page 9, if I understand correctly, the authors regress the vector of ISC values across parcels out of the vector of heritability values across parcels, and then plot the residual heritability values. Do they center the heritability values (or include some kind of intercept) in the process? I'm trying to understand why the heritability values go from all positive (Figure 2A) to roughly balanced between positive and negative (Figure 2B). Important question for me: How should we interpret negative values in this plot? Can the authors explain this explicitly in the text? (I also wonder if there's a more intuitive way to control for ISC. For example, instead of regressing out ISC at the parcel/map level, could they go into a single parcel and then regress the subject-level pairwise ISC values out when computing the heritability score?).

(4) On page 4 (line 155), the authors say "we shuffled dyad labels"- is this equivalent to shuffling rows and columns of the pairwise subject-by-subject matrix combined across groups? I'm trying to make sure their approach here is consistent with recommendations by Chen et al., 2016. Is this the same kind of shuffling used for the kinship matrix mentioned in line 189?

(5) I found panel A in Figure 4 to be a little bit misleading because their parcel-wise approach to hyperalignment won't actually resolve topographic idiosyncrasies across a large cortical distance like what's depicted in the illustration (at the scale of the parcels they are performing hyperalignment within). Maybe just move the green and purple brain areas a bit closer to each other so they could feasibly be "aligned" within a large parcel. Worth keeping in mind when writing that hyperalignment is also not actually going to yield a one-to-one mapping of functionally homologous voxels across individuals: it's effectively going to model any given voxel time series as a linear combination of time series across other voxels in the parcel.

(6) I believe the subjects watched all different movies across the two days, however, for a moment I was wondering "are Day 1 and Day 2 repetitions of the same movies?" Given that Day 1 and Day 2 are an organizational feature of several figures, it might be worth making this very explicit in the Methods and reminding the reader in the Results section.

References:

Busch, E. L., Slipski, L., Feilong, M., Guntupalli, J. S., di Oleggio Castello, M. V., Huckins, J. F., Nastase, S. A., Gobbini, M. I., Wager, T. D., & Haxby, J. V. (2021). Hybrid hyperalignment: a single high-dimensional model of shared information embedded in cortical patterns of response and functional connectivity. NeuroImage, 233, 117975. https://doi.org/10.1016/j.neuroimage.2021.117975

Chen, G., Shin, Y. W., Taylor, P. A., Glen, D. R., Reynolds, R. C., Israel, R. B., & Cox, R. W. (2016). Untangling the relatedness among correlations, part I: nonparametric approaches to inter-subject correlation analysis at the group level. NeuroImage, 142, 248-259. https://doi.org/10.1016/j.neuroimage.2016.05.023

Simony, E., Honey, C. J., Chen, J., Lositsky, O., Yeshurun, Y., Wiesel, A., & Hasson, U. (2016). Dynamic reconfiguration of the default mode network during narrative comprehension. Nature Communications, 7, 12141. https://doi.org/10.1038/ncomms12141

Reviewer #2 (Public review):

Summary:

The authors attempt to estimate the heritability of brain activity evoked from a naturalistic fMRI paradigm. No new data were collected; the authors analyzed the publicly available and well-known data from the Human Connectome Project. The paper has 3 main pieces, as described in the Abstract:

(1) Heritability of movie-evoked brain activity and connectivity patterns across the cortex.

(2) Decomposition of this heritability into genetic similarity in "where" vs. "how" sensory information is processed.

(3) Heritability of brain activity patterns, as partially explained by the heritability of neural timescales.

Strengths:

The authors investigate a very relevant topic that concerns how heritable patterns of brain activity among individuals subjected to the same kind of naturalistic stimulation are. Notably, the authors complement their analysis of movie-watching data with resting-state data.

Weaknesses:

The paper has numerous problems, most of which stem from the statistical analyses. I also note the lack of mapping between the subsections within the Methods section and the subsections within the Results section. We can only assess results after understanding and confirming the methods are valid; here, however, Methods and Results, as written, are not aligned, so we can't always be sure which results are coming from which analysis.

(A) Intersubject correlation (ISC) (section that starts from line 143): "We used non-parametric permutation testing to quantify average differences in ISC for each parcel in the Schaefer 400 atlas for each day of data collection across three groups: MZ dyads, DZ dyads, and unrelated (UR) dyads, where all UR dyads were matched for gender and age in years." ... "some participants contributed to ISC values for multiple dyads (thus violating independence assumptions)"

This is an indirect attempt to demonstrate heritability. And it's also incorrect since, as the authors themselves point out, some subjects contribute to more than one dyad.

Permutation tests don't quantify "average differences", they provide a measure of evidence about whether differences observed are sufficient to reject a hypothesis of no difference.

Matching subjects is also incorrect as it artificially alters the sample; covarying for age and sex, as done in standard analyses of heritability, would have been appropriate.

It isn't clear why the authors went through the trouble of implementing their own non-parametric test if HCP recommends using PALM, which already contains the validated and documented methods for permutation tests developed precisely for HCP data.

The results from this analysis, in their current form, are likely incorrect.

(B) Functional connectivity (FC) (section that starts from line 159): Here the authors compute two 400x400 FC matrix for each subject, one for rest, one for movie-watching, then correlate the correlations within each dyad, then compared the average correlation of correlations for MZ, DZ, and UR. In addition to the same problems as the previous analysis, here it is not clear what is meant by "averaging correlations [...] within a network combination". What is a "network combination"? Further, to average correlations, they need to be r-to-z transformed first. As with the above, the results from this analysis in its current form are likely incorrect.

(C) ISC and FC profile heritability analyses (section that starts from line 175): Here, the authors use first a valid method remarkably similar to the old Haseman-Elston approach to compute heritability, complemented by a permutation test. That is fine. But then they proceed with two novel, ill-described, and likely invalid methods to (1) "compare the heritability of movie and rest FC profiles" and (2) to "determine the sample size necessary for stable multidimensional heritability results". For (1), they permute, seemingly under the alternative, rest and movie-watching timeseries, and (2), by dropping subjects and estimating changes in the distribution.

The (1) might be correct, but there are items that are not clearly described, so the reader cannot be sure of what was done. What are the "153 unique network combinations"? Why do the authors separate by day here, whereas the previous analyses concatenated both days? Were the correlations r-to-z transformed before averaging?

The (2) is also not well described, and in any case, power can be computed analytically; it isn't clear why the authors needed to resort to this ad hoc approach, the validity of which is unknown. If the issue is the possibility that the multidimensional phenotypic correlation matrix is rank-deficient, it suffices that there are more independent measurements per subject than the number of subjects.

(D) Frequency-dependent ISC heritability analysis (from line 216): Here, the authors decompose the timeseries into frequency bands, then repeat earlier analyses, thus bringing here the same earlier problems and questions of non-exchangability in the permutations given the dyads pattern, r-z transforms, and sex/age covariates.

(E) FC strength heritability analysis (from line 236): Here, the authors use the univariate FC to compute heritability using valid and well-established methods as implemented in SOLAR. There is no "linkage" being done here (thus, the statement in line 238 is incorrect in this application. SOLAR already produces SEs, so it's unclear why the authors went out of their way to obtain jackknife estimates. If the issue is non-normality, I note that the assumption of normality is present already at the stage in which parameters themselves are estimated, not just the standard errors; for non-normal data, a rank-based inverse-normal transformation could have been used. Moreover, typically, r-to-z transformed values tend to be fairly normally distributed. So, while the heritabilities might be correct, the standard errors may not be (the authors don't demonstrate that their jackknife SE estimator is valid). The comparison of h2 between dyads raises the same questions about permutations, age/sex covariates, and r-z transforms as above.

(F) Hyperalignment (from line 245): It isn't clear at this point in the manuscript in what way hyperalignment would help to decompose heritability in "where vs. how" (from the Abstract). That information and references are only described much later, from around line 459. The description itself provides no references, and one cannot even try to reproduce what is described here in the Methods section. Regardless, it isn't entirely clear why this analysis was done: by matching functional areas, all heritabilities are going to be reduced because there will be less variance between subjects. Perhaps studying the parameters that drive the alignment (akin to what is done in tensor-based and deformation-based morphometry) could have been more informative. Plus, the alignment process itself may introduce errors, which could also reduce heritability. This could be an alternative explanation for the reduced heritability after hyperalignment and should be discussed. An investigation of hyperaligment parameters, their heritability, and their co-heritability with the BOLD-phenotypes can inform on this.

(G) Relationships between parcel area and heritability (from line 270): As under F), how much the results are distorted likely depends on the accuracy of the alignment, and the error variance (vs heritable variance) introduced by this.

(H) Neural timescale analyses (from line 280): Here, a valid phenotype (NT) is assessed with statistical methods with the same limitations as those previously (exchangability of dyads, age/sex covariates, and r-z transforms). NT values are combined across space and used as covariates in "some multivariate analyses". As a reader, I really wanted to see the results related to NT, something as simple as its heritability, but these aren't clearly shown, only differences between types of dyads.

(I) Significance testing for autocorrelated brain maps and FC matrices (from line 310): Here, the authors suddenly bring up something entirely different: reliability of heritability maps, and then never return to the topic of reliability again. As a reader, I find this confusing. In any case, analyses with BrainSMASH with well-behaved, normally distributed data are ok. Whether their data is well behaved or whether they ensured that the data would be well behaved so that BrainSMASH is valid is not described. As to why Spearman correlations are needed here, Mantel tests, or whether the 1000 "surrogate" maps are valid realizations of the data under the null, remains undemonstrated.

(J) Global signal was removed, and the authors do not acknowledge that this could be a limitation in their analyses, nor offer a side analysis in which the global signal is preserved.

(K) FDR is used to control the error rate, but in many cases, as it's applied to multiple sets of p-values, the amount of false discoveries is only controlled across all tests, but not within each set. The number of errors within any set remains unknown.

(L) Generally, when studying the heritability of a trait, the trait must be defined first. Here, multiple traits are investigated, but are never rigorously defined. Worse, the trait being analyzed changes at every turn.

Reviewer #3 (Public review):

Strengths:

It's sort of novel to study the heritability of movie-watching fMRI data. The methodology the authors used in the paper is also supportive of their findings. Figures are nicely organized and plotted. They finally found that sensory processing in the human brain is under genetic control over stable aspects of brain function (here referring to neural timescale and resting state connectivity).

Weaknesses:

What I am worried about most is the sample size and interpretation of heritability.

(1) Figure 1. I assumed that the authors just calculated the ISC within each group (MZ, DZ, and UR). Of course, you can get different variations between each group. Therefore, there is heritability. Why not calculate ISC across the whole sample, then separate MZ, DZ, and UR?

(2) Heritability scores in the paper are sort of small. If the sample size is small, please consider p-values, which will tell more about the trustworthiness of your heritability.

(3) I don't understand the high-frequency signals in fMRI data. It's always regarded as noise, the band 1 here in particular.

(4) The statement "we show that the heritability of brain activity patterns can be partially explained by the heritability of the neural timescale" should come from Figure 5. However, after controlling for NT, the heritability decreased max. 0.025 in temporal areas. I am not sure this change supports the statement. If the visual cortex is outlined, and combining ISC changes in the visual cortex, I think this would somehow be answered. Instead of delta h2, adding a new model h2 would be obvious to the readers.

(5) Figures 7 and 8, when getting the difference of heritability, please also consider the standard errors of the heritability estimates. Then you can compare across networks/regions.

(6) I think movie VS resting state is a really important result in this paper. However, there is almost no discussion. Discussing this part would be more beneficial for understanding the genetic control over the neuron arousal and excitation circuits.

Author Response:

eLife assessment

This is a valuable initial study of cell type and spatially resolved gene expression in and around the locus coeruleus, the primary source of the neuromodulator norepinephrine in the human brain. The data are generated with cutting-edge techniques, and the work lays the foundation for future descriptive and experimental approaches to understand the contribution of the locus coeruleus to healthy brain function and disease. However, due to small sample size and the need for additional confirmatory data, the data only incompletely support the main conclusions presented here. With the strengthening of the analyses, this paper, and the associated web application, will be of great interest to neuroscientists working on arousal-based behaviors and neurological and neuropsychiatric phenotypes.

Thank you for the assessment and comments. Overall, the majority of the issues raised by the reviewers relate either directly or indirectly to limitations of the sample size that precluded further optimization of protocols and expansion of the dataset. We fully acknowledge the limited sample size in this dataset and aim to be transparent about the limitations of the study. This is the first report of snRNA-seq and spatially-resolved transcriptomics in the human locus coeruleus (LC). The LC is a very small nucleus, located deep within the brainstem, which is extremely challenging to study due to its small size, difficult to access location, and the very small number of norepinephrine (NE) neurons located within the nucleus, which were of prime interest for this study. We note that this study represents our initial attempt to molecularly and spatially characterize cell types within the human LC. We note that we did not have significant, established funding from extramural sources dedicated to this study, and tissue resources for the LC are difficult to ascertain, contributing to the small sample size in this initial study. We acknowledge that there are limitations in sample size as well as data quality. Findings from this study will be used to inform, improve, and optimize future and ongoing experimental design, as well as technical and analytical workflows for larger-scale studies. As brought up by one of the reviewers, this field is still in its infancy -- pilot experimentation in new brain regions is labor-intensive and these sequencing approaches remain costly. Moreover, due to the small size and difficulties in dissecting, tissue resources from the human brain in this area are a highly limited resource. Hence, notwithstanding limitations, in our view it is important to release the data for community access at this time. Specific responses to the reviewers’ comments are provided point-by-point in the following sections.

Reviewer #1 (Public Review):

Weber et al. collect locus coeruleus (LC) tissue blocks from 5 neurotypical European men, dissect the dorsal pons around the LC and prepare 2-3 tissue sections from each donor on a slide for 10X spatial transcriptomics. […] The authors transparently present limitations of their work in the discussion, but some points discussed below warrant further attention.

Specific comments:

1) snRNAseq:

a. Major concerns with the snRNAseq dataset are A) the low recovery rate of putative LC-neurons in the snRNAseq dataset, B) the fact that the LC neuron cluster is contaminated with mitochondrial RNA, and C) that a large fraction of the nuclei cannot be assigned to a clear cell type (presumably due to contamination or damaged nuclei). The authors chose to enrich for neurons using NeuN antibody staining and FACS. But it is difficult to assess the efficacy of this enrichment without images of the nuclear suspension obtained before FACS, and of the FACS results. As this field is in its infancy, more detail on preliminary experiments would help the reader to understand why the authors processed the tissue the way they did. It would be nice to know whether omitting the FACS procedure might in fact result in higher relative recovery of LC-neurons, or if the authors tried this and discovered other technical issues that prompted them to use FACS.

Thank you for these comments. We agree these are valid concerns in assessing the data quality and validity of the findings from the snRNA-seq dataset. We will respond to these concerns here to the best of our ability, but in some cases, we do not have definitive answers since comparison data are not yet available for this region. In particular, we were limited in resources for this initial study -- some of the results of the study and issues that we identified in attempting to molecularly profile cells in the human LC were surprising to us, and we intend to generate additional samples and troubleshoot these issues to improve data quality and increase recovery in future work. However, these experiments are (i) expensive, (ii) time- and labor-intensive, and (iii) the tissue for this region is limited and difficult to ascertain. Given the extremely small size of the LC, the tissue resource is quickly depleted. For this study, we had fixed resources and made best-guess decisions on how to proceed with the experimental design, based on our experience with snRNA-seq in other human brain regions (Tran and Maynard et al. 2021). However, the LC is a unique region, and our experiences with this dataset will guide us to make technical adjustments in future studies. Due to the limitations in the tissue resources and the lack of data currently available to the community, we wanted to share these results immediately while acknowledging the limitations of the study as we work to increase our resource availability to expand molecular and spatial profiling studies in this region of the human brain.

Regarding the reviewer’s concern that our choice to use FANS to enrich for neurons could have potentially led to more damage and contributed to the low recovery rate of LC-NE neurons and the mitochondrial contamination -- we do not have a definitive answer to this question, since we did not perform a direct comparison with non-sorted data. As noted above, our limited tissue resource dictated that we could not do both. We made the decision to enrich for neurons based on our previous experience with identifying relatively rare populations in other brain regions (e.g. nucleus accumbens and amygdala; Tran and Maynard et al. 2021). Based on this previous work, our rationale was that without neuronal enrichment, we could potentially miss the LC-NE population, given the relative scarcity of this neuronal population. The low recovery rate and relatively lower quality / contamination issues may be due to technical issues that lead to LC-NE neurons being more susceptible to damage during nuclear preparation and sorting. We agree that directly comparing to data prepared without NeuN labeling and sorting is reasonable, as the additional perturbations may indeed contribute to cell damage. As mentioned in the discussion, we do not have a definitive answer to the reasons for increased mitochondrial contamination and we suspect that multiple technical factors may contribute -- including the relatively large size and increased fragility of LC-NE neurons. We agree that systematically optimizing the preparation to attempt to increase recovery rate and decrease mitochondrial contamination are important avenues for future work.

b. It is unclear what percentage of cells that make up each cluster.

We will add this information in the clustering heatmaps or as a supplementary plot in a revised version of the manuscript.

c. The number of subjects used in each analysis was not always clear. Only 3 subjects were used for snRNAseq, and one of them only yielded 4 LC-nuclei. This means the results are essentially based on n=2. The authors report these numbers in the corresponding section, but the first sentence of the results section (and Figure 1C specifically!) create the impression that n=5 for all analyses. Even for spatial transcriptomics, if I understood it correctly, 1 sample had to be excluded (n=4).

This is correct. We will update the figures and text in a revised version of the manuscript to make this limitation (small sample size) more clear, and to further emphasize that the intention of this study is to provide initial data to help determine next steps and best practices for a larger scale and more comprehensive study on this region, especially given the limited availability of tissue resources and currently limited data resources available for this region.

2) Spatial transcriptomics:

a. It is not clear to me what the spatial transcriptomics provides beyond what can be shown with snRNAseq, nor how these two sets of results compare to each other. It would be more intuitive to start the story with snRNAseq and then try to provide spatial detail using spatial transcriptomics. The LC is not a homogeneous structure but can be divided into ensembles based on projection specificity. Spatial transcriptomics could - in theory - offer much-needed insights into the spatial variation of mRNA profiles across different ensembles, or as a first step across the spatial (rostral/caudal, ventral/dorsal) extent of the LC. The current analyses, however, cannot address this issue, as the orientation of the LC cannot be deduced from the slices analyzed.

We understand the point of the reviewer. However, we structured the manuscript in this format due to our aims of creating a data resource for the community as well as being transparent about the limitations of our study. Our experiments began with the spatial experiments on the tissue blocks because this (i) helped orient ourselves to the region, and (ii) provided guidance for how best to score the tissue blocks for the snRNA-seq experiments to maximize recovery of LC-NE neurons. Therefore, we also decided to present the results in this sequence.

The spatial data also provides more information in that the measurements are from nuclei, cytoplasm, and cell processes (instead of nuclei only). This is one of the main differences / advantages between the platforms at this level of spatial resolution. As noted above, we were also working with a finite tissue resource -- if we ran snRNA-seq first and captured no neurons, the tissue block would be depleted. Due to the logistics / thickness of the required tissue sections for Visium and snRNA-seq respectively, running Visium first allowed us to ensure that we could collect data from both assays.

Regarding a point raised below on why we only ran snRNA-seq on a subset of the donors -- this was due to resource depletion and not enough available tissue remaining on the tissue blocks to run the assay. We have conducted extensive piloting in other brain regions on the amount (mg) of tissue that is needed from various sized cryosections, and the LC is particularly difficult since these are small tissue blocks and the extent of the structure is small. Hence, in some of the subjects, we did not have sufficient tissue available for the snRNA-seq assay.

We agree with the reviewer that spatial studies could, in future work, offer needed and important information about expression profiles across the spatial axes (rostral/caudal, ventral/dorsal) of the LC. Our study provides us with insight about optimizing the dissections for spatial assays, as well as bringing to light a number of technical and logistical issues that we had not initially foreseen. For example, during the course of this study and parallel, ongoing work in other small, challenging brain regions, we have now developed a number of specialized technical and logistical strategies for keeping track of orientation and mounting serial sections from the same tissue block onto a single spatial array, which is extremely technically challenging. We are now well-prepared for addressing these issues in future studies with larger numbers of donors and samples, e.g. spaced serial sections across the extent of the LC to make these types of insights. Due to the rarity of the tissue, limited availability of information in this region, and high expense of conducting these studies, we want to share this initial data with the community immediately. We also note that in addition to the 10x Genomics Visium platform, which lacks cellular and sub-cellular resolution, many new and exciting spatial platforms are entering the market, which may be able to address questions in very small regions such as the LC at higher spatial resolution.

b. Unfortunately, spatial transcriptomics itself is plagued by sampling variability to a point where the RNAscope analyses the authors performed prove more powerful in addressing direct questions about gene expression patterns. Given that the authors compare their results to published datasets from rodent studies, it is surprising that a direct comparison of genes identified with spatial transcriptomics vs snRNAseq is lacking (unless this reviewer missed this comparison). Supplementary Figure 17 seems to be a first step in that direction, but this is not a gene-by-gene comparison of which analysis identifies which LC-enriched genes. Such an analysis should not compare numbers of enriched genes using artificial cutoffs for significance/fold-change, but rather use correlations to get a feeling for which genes appear to be enriched in the LC using both methods. This would result in one list of genes that can serve as a reference point for future work.

We agree this is a good suggestion, and will add additional computational analyses to address this point in a revised version of the manuscript.

c. Maybe the spatial transcriptomics could be useful to look at the peri-LC region, which has generated some excitement in rodent work recently, but remains largely unexplored in humans.

We agree this is an excellent suggestion -- assessing cross-species comparisons related to convergence, especially, of GABAergic cell populations in the human LC is of high interest. We note that these types of extensions are exactly the reason why we have provided the publicly accessible web app (R/Shiny app, which includes the ability to annotate regions). We hope that others will use these apps for specialized topics they are interested in. As discussed above, we note that our initial dissections precluded the ability to keep track of the exact orientation of our tissue sections on the Visium arrays with respect to their location within the brainstem, so definitive localization of this region across subjects is difficult in our current study. However, it is possible, for example, to investigate whether there is a putative peri-LC region that is densely GABAergic that is homologous with the GABAergic peri-LC region in rodents. We also raise attention to a recent preprint by Luskin and Li et al. (2022), who apply snRNA-seq and spatially-resolved transcriptomics to molecularly define both LC and peri-LC cell types in mice -- in a revised version of our manuscript, we will extend our computational analyses of inhibitory neuronal subtypes in our data (Supplementary Figures 13, 16) to directly compare with those identified in this study in more detail. As noted above, we we have now developed a number of specialized technical and logistical strategies for keeping track of orientation of sections from the tissue block onto a single spatial array, and we feel that combined with optimized dissection strategies for this region and the guide of RNAscope for GABAergic markers on serial sections, that annotating the peri-LC region on spatial arrays in future studies will be possible.

3) The comparison of snRNAseq data to published literature is laudable. Although the authors mention considerable methodological differences between the chosen rodent work and their own analyses, this needs to be further explained. The mouse dataset uses TRAPseq, which looks at translating mRNAs associated with ribosomes, very different from the nuclear RNA pool analyzed in the current work. The rat dataset used single-cell LC laser microdissection followed by microarray analyses, leading to major technical differences in terms of tissue processing and downstream analyses. The authors mention and reference a recent 10x mouse LC dataset (Luskin et al, 2022), however they only pick some neuropeptides from this study for their analysis of interneuron subtypes (Figure S13). Although this is a very interesting part of the manuscript, a more in-depth analysis of these two datasets would be very useful. It would likely allow for a better comparison between mouse and human, given that the technical approach is more similar (albeit without FACS), and Luskin et al have indicated that they are willing to share their data.

As noted above, we plan to extend our comparisons with the dataset from Luskin and Li et al. (2022) in a revised version of the manuscript, which will provide a more in-depth cross-species comparison. In addition, we also note that there are some additional recent studies using TRAPseq of LC-NE neurons in a functional context, i.e. treatment vs. control experiments or in model systems (e.g. Iannitelli et al. 2023), which provide new opportunities for understanding disease context using in-depth cross-species comparisons. By providing our dataset and reproducible code, we will enable others to adapt and extend these types of comparisons (i.e. TRAPseq of LC-NE neurons or LC snRNA-seq following functional manipulations or in the context of disease or behavioral models) in the future.

4) Statements in the manuscript about the unexpected identification of a 5-HT (serotonin) cell-cluster seem somewhat contradictory. Figure S14 suggests that 5-HT markers are expressed in the LC-regions just as much as anywhere else, but the RNAscope image in Figure S15 suggests spatial separation between these two populations. And Figure S17 again suggests almost perfect overlap between the LC and 5HT clusters. Maybe I misunderstood, in which case the authors should better clarify/explain these results.

In our view, the most likely scenario is that the 5-HT neurons come from contamination from the dorsal raphe nucleus based on spatial separation from the RNAscope images, which we agree are more definitive. As mentioned above, since we do not have definitive documentation for the tissue sections in terms of orientation, it is difficult to say with clarity that the regions are the dorsal raphe and which sub-portion of the dorsal raphe they are. This initial study has now allowed us to optimize and improve our dissection strategy and approaches for retaining documentation of the orientation of the tissue sections from their intact position within the brainstem as they move from cryosection to placement on the array, which will enable us to better annotate regions with definitive anatomical information with respect to the rostral/caudal and dorsal/ventral axes in future experiments. Given that there are reports in the rodent that 5-HT markers have been identified in LC-NE neurons (Iijima 1993; Iijima 1989), and taking into account the technical limitations in our study, we felt that it was premature to definitively conclude in the manuscript that we were sure these signals arose from the dorsal raphe. We will update this language in a revised version of the manuscript to ensure that these limitations are clear (referring to Supplementary Figures S14-15, S17).

Reviewer #2 (Public Review):

The data generated for this paper provides an important resource for the neuroscience community. The locus coeruleus (LC) is the known seed of noradrenergic cells in the brain. Due to its location and size, it remains scarcely profiled in humans. Despite the physically minute structure containing these cells, its impact is wide-reaching due to the known neuromodulatory function of norepinephrine (NE) in processes like attention and mood. As such, profiling NE cells has important implications for most neurological and neuropsychiatric disorders. This paper generates transcriptomic profiles that are not only cell-specific but which also maintain their spatial context, providing the field with a map for the cells within the region.

Strengths:

Using spatial transcriptomics in a morphologically distinct region is a very attractive way to generate a map. Overlaying macroscopic information, i.e. a region with greater pigmentation, with its corresponding molecular profile in an unbiased manner is an extremely powerful way to understand the specific cellular and molecular composition of that brain structure.

The technologies were used with an astute awareness of their limitations, as such, multiple technologies were leveraged to paint a more complete and resolved picture of the cellular composition of the region. For example, the lack of resolution in the spatial transcriptomic platform was compensated by complementary snRNA-seq and single molecule FISH.

This work has been made publicly available and accessible through a user-friendly application such that any interested researcher can investigate the level of expression of their gene of interest within this region.

Two important implications from this work are 1) the potential that the gene regulatory profiles of these cells are only partially conserved across species, humans, and rodents, and 2) that there may be other neuromodulatory cell types within the region that were otherwise not previously localized to the LC

Weaknesses:

Given that the markers used to identify cells are not as specific as they need to be to definitively qualify the desired cell type, the results may be over-interpreted. Specifically, TH is the primary marker used to qualify cells as noradrenergic, however, TH catalyzes the synthesis of L-DOPA, a precursor to dopamine, which in turn is a precursor for epinephrine and norepinephrine suggesting some of the cells in the region may be dopaminergic and not NE cells. Indeed, there are publications to support the presence of dopaminergic cells in the LC (see Kempadoo et al. 2016, Takeuchi et al., 2016, Devoto et al. 2005). This discrepancy is further highlighted by the apparent lack of overlap per given Visium spots with TH, SCL6A2, or DBH. While the single-nucleus FISH confirms that some of the cells in the region are noradrenergic, others very possibly represent a different catecholamine. As such it is suggested that the nomenclature for the cells be reconsidered.

We appreciate the reviewer’s comment, and are aware of the reports suggesting the potential presence of dopaminergic cells in the LC. We initially had the same thought as the reviewer when we observed Visium spots in the spatial data with lack of overlap between TH, SLC6A2, and DBH as well as single nuclei in the snRNA-seq data with lack of overlap between TH, SLC6A2, and DBH. This surprising result was exactly why we performed the smFISH/RNAscope experiment with these three marker genes. Given known issues with read depth and coverage in the 10x Genomics assays, we wanted to better understand if this was a technical limitation in the sequencing coverage, or rather a true biological finding. The RNAscope data showed very clearly that nearly every cell body we looked at had co-localization of these three marker genes. We included an image from a single capture array of one tissue section in Supplementary Figure 11, but could, in a revised version of the manuscript, provide additional examples to illustrate how conclusive the images were by visualization. As such, we were quite convinced that the lack of overlap on Visium spots and in single nuclei in the snRNA-seq data was more likely related to technical issues with sequencing coverage, rather than a biological finding. We also note that we checked for the presence of the dopamine transporter, SLC6A3, and as can be appreciated in the iSEE web app for the snRNA-seq data or the R/Shiny web app for the Visium data, there is virtually no expression of SLC6A3 in the dataset, which in our view provides additional evidence against the possibility that there are substantial quantities of dopaminergic cells in this human LC dataset. We will include supplementary plots showing the lack of SLC6A3 expression in a revised version of the manuscript.

The authors are unable to successfully implement unsupervised clustering with the spatial data, this greatly reduces the impact of the spatial technology as it implies that the transcriptomic data generated in the study did not have enough resolution to identify individual cell types.

The reviewer is correct -- this is a fundamental limitation of the 10x Genomics Visium platform, i.e. the spatial resolution captures multiple cells per spot (e.g. around 1-10 cells per spot in human brain tissue). We note that new spatial platforms now provide cellular resolution (e.g. Vizgen MERSCOPE, 10x Genomics Xenium, 10x Genomics Visium HD), which will help address this in future work. However, many of these cellular-resolution in situ sequencing platforms have the limitation that they do not quantify genome-wide expression, and instead require users to select a priori gene panels to investigate. This is a problem if no genome-wide reference datasets are available. Hence, despite the limited spatial resolution of the Visium platform, this dataset is useful precisely for helping investigators choose gene panels for higher-resolution platforms or higher-order smFISH multiplexing.

We also applied spatial clustering (using BayesSpace; Zhao et al. 2021) to attempt to segment the LC regions within the Visium samples in a data-driven manner as an alternative to the manual annotations, which was unsuccessful (and hence we relied on the manually annotated regions for downstream analyses) (Supplementary Figure S5). However, this is a different application of unsupervised clustering, which is separate from the task of identifying cell types.

The sample contribution to the results is highly unbalanced, which consequently, may result in ungeneralizable findings in terms of regional cellular composition, limiting the usefulness of the publicly available data.

We acknowledge the limitations of the work due to the small/unbalanced sample sizes. As mentioned above for Reviewer 1, this was an initial study in this region -- results of which will inform our (and hopefully others’) experimental design and approach to molecular profiling in this difficult to access brain region. Overall, this study was executed with finite tissue and financial resources and was intended to uncover limitations and help develop best practices and design workflows for future studies with larger numbers of donors and samples. Given the limited data availability for this brain region, we wanted to make this dataset available for the research community immediately. In addition, we note that making this genome-wide dataset available will help inform targeted gene panel design for higher-resolution platforms (e.g. 10x Genomics Xenium).

This study aimed to deeply profile the LC in humans and provide a resource to the community. The combination of data types (snRNA-seq, SRT, smFISH) does in fact represent this resource for the community. However, due to the limitations, of which, some were described in the manuscript, we should be cautious in the use of the data for secondary analysis. For example, some of the cellular annotations may lack precision, the cellular composition also may not reflect the general population, and the presence of unexpected cell types may represent the accidental inclusion of adjacent regions, in this case, serotonergic cells from the Raphe nucleus.

We agree, and have attempted to explain these limitations in the manuscript. We will clarify the language regarding the interpretation of the annotated cell populations and unexpected cell types, and the limited sample sizes, in a revised version of the manuscript.

Nonetheless having a well-developed app to query and visualize these data will be an enormous asset to the community especially given the lack of information regarding the region in general.

Reviewer #3 (Public Review):

[…] This study has many strengths. It is the first reported comprehensive map of the human LC transcriptome, and uses two independent but complementary approaches (spatial transcriptomics and snRNA-seq). Some of the key findings confirmed what has been described in the rodent LC, as well as some intriguing potential genes and modules identified that may be unique to humans and have the potential to explain LC-related disease states. The main limitations of the study were acknowledged by the authors and include the spatial resolution probably not being at the single cell level and the relatively small number of samples (and questionable quality) for the snRNA-seq data. Overall, the strengths greatly outweigh the limitations. This dataset will be a valuable resource for the neuroscience community, both in terms of methodology development and results that will no doubt enable important comparisons and follow-up studies.

Major comments:

Overall, the discovery of some cells in the LC region that express serotonergic markers is intriguing. However, no evidence is presented that these neurons actually produce 5-HT.

The reviewer is correct that we did not provide any additional evidence to show that these neurons actually produce 5-HT. As noted above in the response to Reviewer 1, in our view, the most likely explanation is that these neurons are from dorsal raphe contamination on the tissue section. However, due to technical and logistical limitations in this study, we could not definitively say this because we did not clearly track the orientation of the tissue sections, and we did not have remaining tissue sections from all donor tissue blocks to repeat RNAscope experiments. For some of the donors, where we had remaining tissue sections to go back to repeat RNAscope experiments after completion of the snRNA-seq and Visium assays, we could see clear separation of the LC region / LC-NE neuron core from where putative 5-HT neurons were located (Supplementary Figure 15). However, we did not have sufficient tissue resources to map this definitively in all donors, and the orientation and anatomy of each tissue block were not fully annotated.

Due to the lack of clarity, and the fact that there have been reports that LC-NE neurons express serotonergic markers (Iijima 1993; Iijima 1989), we felt that it was premature to definitively declare that these putative 5-HT neurons that we identified were definitively from the raphe. We will clarify the language around this discrepancy in a revised version of the manuscript to ensure that these limitations are clearly described.

Concerning the snRNA-seq experiments, it is unclear why only 3 of the 5 donors were used, particularly given the low number of LC-NE nuclear transcriptomes obtained, why those 3 were chosen, and how many 100 um sections were used from each donor. It is also unclear if the 295 nuclei obtained truly representative of the LC population or whether they are just the most "resilient" LC nuclei that survive the process.

As discussed above for Reviewer 1, the reason we included only 3 of the 5 donors for the snRNA-seq assays was due to the tissue availability on the tissue blocks. We will clarify the language in a revised version of the manuscript to make this limitation more clear. We will also include additional details in the Methods section on the number of 100 μm sections used for each donor (which varied between 10-15, approximating 60-80 mg of tissue).

The LC displays rostral/caudal and dorsal/ventral differences, including where they project, which functions they regulate, and which parts are vulnerable in neurodegenerative disease (e.g. Loughlin et al., Neuroscience 18:291-306, 1986; Dahl et al., Nat Hum Behav 3:1203-14, 2019; Beardmore et al., J Alzheimer's Dis 83:5-22, 2021; Gilvesy et al., Acta Neuropathol 144:651-76, 2022; Madelung et al., Mov Disord 37:479-89, 2022). It was not clear which part(s) of the LC was captured for the SRT and snRNAseq experiments.

As discussed above for Reviewer 1, a limitation of this study was that we did not record the orientation of the anatomy of the tissue sections, precluding our ability to annotate the tissue sections with the rostral/caudal and dorsal/ventral axis labels. We agree with the reviewer that additional spatial studies, in future work, could offer needed and important information about expression profiles across the spatial axes (rostral/caudal, ventral/dorsal) of the LC. Our study provides us with insight about optimizing the dissections for spatial assays, as well as bringing to light a number of technical and logistical issues that we had not initially foreseen. For example, during the course of this study and parallel, ongoing work in other, small, challenging regions, we have now developed a number of specialized technical and logistical strategies for keeping track of orientation and mounting serial sections from the same tissue block onto a single spatial array, which is extremely technically challenging. We are now well-prepared for addressing these issues in future studies with larger numbers of donors and samples in order to make these types of insights.

The authors mention that in other human SRT studies, there are typically between 1-10 cells per expression spot. I imagine that this depends heavily on the part of the brain being studied and neuronal density, but it was unclear how many LC cells were contained in each expression spot.

The reviewer is correct that we did not include this information in the manuscript. We attempted to apply a computational method to count nuclei contained in each gene expression spot based on analyzing the histological H&E images (VistoSeg; Tippani et al. 2022), which we have developed and previously applied in data from the dorsolateral prefrontal cortex (DLPFC) (Maynard and Collado-Torres et al. 2021). Based on the segmentation using this workflow we observe that the counts in this region are similar to what we observed in the DLPFC, i.e., typically between 1-10 LC cells per expression spot, with approximately 1-2 LC-NE neurons (which are characterized by their large size) per expression spot. However, these analyses had several technical issues related to the images themselves, the relatively large size and pigmentation of LC-NE neurons, and parameter settings that had been optimized for different brain regions. We are currently optimizing this analysis workflow for these images to provide more accurate estimates of cell counts per spot to give readers additional context on the number of nuclei per spot in the annotated LC regions and outside the LC regions in a revised version of the manuscript.

Regarding comparison of human LC-associated genes with rat or mouse LC-associated genes (Fig. 2D-F), the authors speculate that the modest degree of overlap may be due to species differences between rodents and human and/or methodological differences (SRT vs microarray vs TRAP). Was there greater overlap between mouse and rat than between mouse/rat and human? If so, that is evidence for the former. If not, that is evidence for the latter. Also would be useful for more in-depth comparison with snRNA-seq data from mouse LC: https://www.biorxiv.org/content/10.1101/2022.06.30.498327v1.

We will investigate this question and discuss this in updated results in a revised version of the manuscript.

The finding of ACHE expression in LC neurons is intriguing, especially in light of work from Susan Greenfield suggesting that ACHE has functions independent of ACH metabolism that contributes to cellular vulnerability in neurodegenerative disease.

We thank the reviewer for pointing this out. We were very surprised too by the observed expression of SLC5A7 and ACHE in the LC regions (Visium data) and within the LC-NE neuron cluster (snRNA-seq data), coupled with absence of other typical cholinergic marker genes (e.g. CHAT, SLC18A3), and we do not have a compelling explanation or theory for this. Hence, the work of Susan Greenfield and colleagues suggesting non-cholinergic actions of ACHE, particularly in other catecholaminergic neurons (e.g. dopaminergic neurons in the substantia nigra) is very interesting. We will include references to this work and how it could inform interpretation of this expression in a revised version of the manuscript (Greenfield 1991; Halliday and Greenfield 2012).

High mitochondrial reads from snRNA-seq can indicate lower quality. It was not clear why, given the mitochondrial read count, the authors are confident in the snRNA-seq data from presumptive LC-NE neurons.

We will include additional analyses to further investigate and/or confirm this finding (e.g. comparing sum of UMI counts / number of detected genes and mitochondrial percentage per nucleus for this population to confirm data quality) in additional supplementary figures in a revised version of the manuscript.

References

• Greenfield (1991), A noncholinergic action of acetylcholinesterase (AChE) in the brain: from neuronal secretion to the generation of movement, Cellular and Molecular Neurobiology, 11, 1, 55-77.

• Halliday and Greenfield (2012), From protein to peptides: a spectrum of non-hydrolytic functions of acetylcholinesterase, Protein & Peptide Letters, 19, 2, 165-172.

• Iannitelli et al. (2023), The neurotoxin DSP-4 dysregulates the locus coeruleus-norepinephrine system and recapitulates molecular and behavioral aspects of prodromal neurodegenerative disease, eNeuro, 10, 1, ENEURO.0483-22.2022.

• Iijima K. (1989), An immunocytochemical study on the GABA-ergic and serotonin-ergic neurons in rat locus ceruleus with special reference to possible existence of the masked indoleamine cells. Acta Histochema, 87, 1, 43-57.

• Iijima K. (1993), Chemocytoarchitecture of the rat locus ceruleus, Histology and Histopathology, 8, 3, 581-591.

• Luskin A.T., Li L. et al. (2022), A diverse network of pericoerulear neurons control arousal states, bioRxiv (preprint).

• Maynard and Collado-Torres et al. (2021), Transcriptome-scale spatial gene expression in the human dorsolateral prefrontal cortex, Nature Neuroscience, 24, 425-436.

• Tippani et al. (2022), VistoSeg: processing utilities for high-resolution Visium/Visium-IF images for spatial transcriptomics data, bioRxiv (preprint).

• Tran M.N., Maynard K.R. et al. (2021), Single-nucleus transcriptome analysis reveals cell-type-specific molecular signatures across reward circuitry in the human brain, Neuron, 109, 3088-3103.

• Zhao E. et al. (2021), Spatial transcriptomics at subspot resolution with BayesSpace, Nature Biotechnology, 39, 1375-1384.

eLife Assessment

The authors report that chemogenetic methods targeting the ventral cervical spinal cord can be used to increase phrenic inspiratory motor output and subsequent diaphragm EMG activity and ventilation in rodents. These findings are important because they are a necessary first step towards using chemogenetic methods to drive inspiratory activity in disorders in which motor neurons are compromised, such as spinal injury and degenerative disease. The data are convincing, with rigorous assessments of phrenic inspiratory activity and its ability to drive the diaphragm and subsequent ventilation, as well as assessments of DREADD expression.

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors report that activation of excitatory DREADDs in the mid-cervical spinal cord can increase inspiratory activity in mice and rats. This is an important first step toward an ultimate goal of using this, or similar, technology to drive breathing in disorders associated with decreased respiratory motor output, such as spinal injury or neurodegenerative disease. Strengths to this study include a comparison of non-specific DREADD expression in the mid-cervical spinal cord versus specific targeting to ChAT-positive neurons, and the measurement of multiple respiratory-related outcomes, including phrenic inspiratory output, diaphragm EMG activity and ventilation. The data show convincingly that DREADDs can be used to drive phrenic inspiratory activity, which in turn increases diaphragm EMG activity and ventilation.

Comments on revisions: All of my prior comments have been sufficiently addressed.

Reviewer #2 (Public review):

Summary:

This study shows that when excitatory DREADD receptors are expressed in the ventral area of the cervical spinal cord containing phrenic motoneurons, systemic administration of the DREADD ligand J60 increases diaphragm EMG activity without altering respiratory rate. The authors took a non-selective expression approach in wild-type mice, as well as a more selective Cre-dependent approach in Chat-Cre mice and Chat-Cre rats to stimulate cervical motoneurons in the spinal cord. This is a proof of principle study that supports the use of DREADD technology to stimulate the motor output to the diaphragm.

Strengths:

The strengths of the study lie in the use of both mice and rats to test whether the chomogenetic activation of phrenic motoneurons with multiple experimental approaches increases diaphragm EMG activity (both tonic and phasic) and tidal volume.

Comments on revisions:

Thanks for addressing my comments. One last comment that could be discussed or addressed is :

Line 295- was the time post-infection, which varies considerably between groups and across samples, taken into consideration when comparison of response was made between ChatCre mice (4-9 weeks post-infection) and WT mice (four to five weeks post-infection)?

Author response:

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

Reviewer #2:

Line 295 – was the time post-infection, which varies considerably between groups and across samples, taken into consideration when comparison of response was between ChatCre mice (4-9 weeks post-infection) and WT mice (four to five weeks post-infection)?

Thank you for your comment. We did not originally assess the effects of time post-injection on DREADD response. Generally, AAV transgene expression has been demonstrated to be long-term and stable in the CNS of mice.[1] However, there is some variation in the reporting time of peak transgene expression[2], and this may potentially impact our results.

In investigating this issue further, we discovered an error in our reporting as we did have n = 1 wild-type mouse that underwent EMG recordings 62 days (~9 weeks) post-AAV injection. This has been corrected in the manuscript (lines 87-88).

Addressing this question is challenging due to the uneven distribution of time points within the 4–9-week windows for each group. Essentially, there were two groups per cohort, one studied at 4-5 weeks and one at 8-9 weeks. More specifically:

- Wild-type cohort: n = 10 animals were studied 28–33 days post-injection, and n = 1 at 62 days.

- ChAT-Cre cohort: n = 4 animals were studied 28–30 days post-injection, and n = 5 at 56–59 days.

We performed Pearson correlation analyses between time post-injection and diaphragm EMG response to DREADD activation (peak amplitude and area under the curve, AUC) for both cohorts (Author response image 1):

- ChAT-Cre: No significant correlations were found (peak amplitude: r² = -0.117, r = -0.1492, p = 0.702, Figure 1a-b; AUC:r² = -0.0883, r = 0.2184, p = 0.572, Figure 1c-d).

- Wild type: Initial analysis of all data showed significant correlations (peak amplitude:r² = 0.362, r = 0.6523, p = 0.0296, Figure 1a; AUC: r² = 0.347, r = 0.6424, p = 0.033, Figure 1c), suggesting a moderate positive correlation between time post-injection and EMG response. However, when the single 8–9-week wild-type mouse was excluded, these correlations were no longer significant (peak amplitude: r² = 0.172, r = 0.5142, p = 0.128, Figure 1b; AUC: r² = 0.23, r = 0.5614, p = 0.0913, Figure1d).

Comparing wild-type and ChAT-Cre groups directly was unreliable due to the single wild-type mouse studied at the later time point. We attempted to model time post-injection as a continuous variable (i.e., exact days post-injection) using a restricted maximum likelihood mixed linear model in JMP; however, the analysis could not be performed because there were not sufficient overlapping time points between the two cohorts (i.e., not all days post-injection were represented in both groups). To mitigate this, we binned animals into two groups: 4–5 weeks and 8–9 weeks post-injection. This analysis returned a significant interaction between cohort and time post-injection (p = 0.0391), however there were no significant multiple comparisons upon Tukey post hoc test (i.e., p > 0.05).

Based on these findings, we feel confident that time post-injection is unlikely to have a significant impact on diaphragm EMG response to DREADD activation in the ChAT-Cre cohort. However, in the wild-type cohort, it is difficult to draw definitive conclusions, as only one animal was studied at the 8–9-week time point. For similar reasons, it remains unclear whether the relationship between time post-AAV transduction and DREADD response differs between cohorts. Given the inconclusive nature of these results, we have elected not to include this analysis in the manuscript. Nevertheless, to ensure transparency, we have provided Author response image 1 below of peak amplitude and AUC plotted against time, allowing readers to evaluate the data independently.

Author response image 1.

Plots of diaphragm EMG peak amplitude (a-b) and area under the curve (c-d) vs. days post-AAV injection for wild-type (blue) and ChAT-Cre (orange) mice. Pearson correlation analyses were performed to assess the relationship between time post-AAV injection and diaphragm EMG DREADD response in wild-type and ChAT-Cre mouse cohorts. r², r, and p-values are shown in each panel for both cohorts. Panels a and c display peak amplitude and AUC, respectively, including all animals. Panels b and d present the same variables with the n = 1 wild-type mouse at the 9-week time point excluded; ChAT-Cre data is unchanged between corresponding panels. Scatter points represent data from individual animals. Polynomial trendlines are displayed for each cohort with wild-type in blue and ChAT-Cre in orange.

REFERENCES

(1) Kim, J. Y., Grunke, S. D., Levites, Y., Golde, T. E. & Jankowsky, J. L. Intracerebroventricular viral injection of the neonatal mouse brain for persistent and widespread neuronal transduction. J Vis Exp, 51863 (2014). https://doi.org/10.3791/51863

(2) Hollidge, B. S. et al. Kinetics and durability of transgene expression after intrastriatal injection of AAV9 vectors. Front Neurol 13, 1051559 (2022). https://doi.org/10.3389/fneur.2022.1051559

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

Response to reviewer’s public reviews:

We chose the dose of J60 based on a prior publication that established that off-target effects were possible at relatively high doses[1]. The dose that we used (0.1 mg/kg) was 30-fold less than the dose that was reported in that paper to potentially have off-target responses (3 mg/kg). Further, Author response image 1 shows the results of experiments in which J60 was given to animals that did not have the excitatory DREADD expressed in the spinal cord. This includes a sample of mice (n = 2) and rats (n = 3), recorded from using the same diaphragm EMG procedure described in the manuscript. The figure shows that there was no consistent response to the J60 at 0.1 mg/kg in the “control experiment” in which the DREADD was not expressed in the spinal cord.

Author response image 1.

Diaphragm EMG response to J60 administrated to naïve rats and mice. Panel a-b show raw EMG values at baseline, following vehicle (saline) and J60 administration for the left and right hemidiaphragm. Panel c-d shows EMG values normalized to baseline. Neither One-way RM ANOVA (panel a-b) nor paired t-test (panel c-d) returned significant p values (p < 0.05).

Response to specific reviewer comments:

Reviewer #1:

How old were the animals at the time of AAV injection, and in subsequent experiments?

The wildtype cohort of mice were 7-9 weeks old at time of AAV injection and DREADD experiments took place 4-5 weeks after AAV injection. ChAT-Cre mice were 6-10 weeks old at time of AAV injection and DREADD experiments took place 4-9 weeks after AAV injection. ChAT-Cre rats were 2-5 months old at time of AAV spinal injection. These animals underwent plethysmography recordings 3-4 months post-AAV injection and subsequently phrenic nerve recording 3-8 weeks later. These details have been added to the Method section.

How many mice were excluded from electrophysiology experiments due to deteriorating electrode contact?

No mice were excluded from electrophysiology experiments due to deteriorating electrode contact. If you are referring to the n = 1 excluded ChAT-Cre mouse (line 368) this animal was excluded because it showed no histological evidence of DREADD expression (lines 200-206).

What was the urethane dose?

The urethane dose for phrenic nerve recordings was 2.1 g/kg. See methods section line 395.

A graphical timeline of the experimental progression for plethysmography and electrophysiology studies would enhance clarity.

A graphical timeline has been added. See Figure S6.

Significance indicators in the figures would greatly enhance clarity. It is a little awkward to have to refer to supplemental tables to figure out statistical differences.

Significance indicators have been added. See Figures 1, 2, 4, and 5

In Figures 1, 2, and 5, individual data points should be shown, as in Fig 4.

Thank you for this suggestion. We agree that, in general, it is best practice to scatter individual data points. However, when we drafted the new figures, it was apparent that including individual scatter points, in this case, created very “cluttered” figures that were very difficult to interpret.

More detail regarding the plethysmography studies is needed. Was saline/J60 infused via a tail vein catheter? Were animals handled during the infusion? How long is the "IV" period? What volume of fluid was delivered?

All IV infusions were delivered via a tail vein catheter. Animals were not handled during infusion nor at any point during the recording. An IV catheter was externalized via a port in the plethysmograph allowing for IV infusion without handling of the animal or opening the plethysmograph. The infusion period for both saline and J60 was standardized to 2 minutes. The volume of fluid of both saline and J60 was standardized to 0.6 mL. This information has been added to the methods section (lines 408-410, 415-16, 419-420).

Reviewer #2:

The abstract could be improved by briefly highlighting the rationale, scope, and novelty of the study - the intro does a great job of highlighting the scope of the study and the research questions.

A brief explanation of the rationale, scope, and novelty of the study has been added to the abstract. See lines 2-8.

Line 18, specifies that this was done under urethane anesthesia.

This detail has been added to the abstract (line 20).

The methods section should be moved to the end of the manuscript according to Journal policy.

The methods section has been moved to the end of the manuscript.

The authors mention the use of both female and male rats but it is not indicated if they tested for and observed any differences between sexes across experiments.

We included the use of both male and female animals in this study to improve the generalizability of the results. However, we were not adequately powered for sex comparisons and therefore did not perform any statistical analysis to assess differences between sexes across experiments. Text has been added to the methods section (lines 534-537) to clarify.

Line 40, since delivery of J60 was performed in both IV and IP, this general statement should be updated.

This detail has been revised to include both IV and IP. See line 43.

Line 42. "First, we determined if effective diaphragm activation requires focal DREADD expression targeting phrenic motor neurons, or if non-specific expression in the immediate vicinity of the phrenic motor nucleus would be sufficient...." I don't think that in the experiments with wild-type mice the authors can claim that they selectively targeted the cervical propriospinal network (in isolation from the motoneurons). Given the fact that the histological analysis did not quantify interneurons or motoneurons in the spinal cord, authors should be cautious in proposing which neuronal population is activated in the non-specific approach.

We agree, and this was a poorly worded statement in our original text. We agree that wild-type DREADD expression was not limited to the cervical propriospinal networks but likely a mix of interneurons and motoneurons. The text has been edited to reflect that (see lines 56-60).

AAV virus source is not described.

All AAVs were obtained from the UF Powell Gene Therapy Center. Details of virus source and production have been added to the methods section. See lines 336-347.

Line 108-125. Because the diaphragm EMG recordings are only described for mice here, I would suggest editing this methods section to clearly state mice instead of vaguely describing "animals" in the procedure.

“Animals” has been changed to “mice” to avoid ambiguity.

Line 120, add parenthesis.

Parenthesis has been added.

Line 126. Whole body plethysmography protocol. Three hypercapnic hypoxic challenges are a lot for a rat within a 3-hour recording session in freely behaving rats. Did the authors verify with control/ vehicle experiments that repeated challenges in the absence of J60 do not cause potentiation of the response? I understand that it is not possible to invert the order of the injections (due to likely long-term effects of J60) or it is too late to perform vehicle and J60 injections on different days, but controls for repeated challenges should be performed in this type of experiment, especially considering the great variability in the response observed in Figure 4 (in normoxic conditions).

We did not conduct control experiments to assess the impact of repeated hypercapnic hypoxic challenges on the naïve response (i.e., in the absence of J60). However, our experimental protocol was designed such that each experimental period (i.e., post-vehicle or post-J60 infusion) was normalized to baseline recordings taken immediately prior to the vehicle or J60 infusion. While repeated exposure to hypercapnic hypoxic challenges may have altered respiratory output, we are confident that normalizing each experimental period to its respective baseline effectively captures the impact of DREADD activation on ventilation, independent of any potential potentiation that may have occurred due to gas challenge exposure. We have included raw values for all plethysmography outcomes (see Figure 4, panels a-c) to ensure full data transparency. Still, we believe that the baseline-normalized values more accurately reflect the impact of DREADD activation on the components of ventilation.

Furthermore, why the response to the hypercapnic hypoxic challenges are not reported? These could be very interesting to determine the effects of DREADD stimulation on chemosensory responses and enhance the significance of the study.

Response to the hypercapnic hypoxic challenges has been added to the manuscript. See Figure S3 and results section lines 162-167. Briefly, there were no statistically significant (p < 0.05) differences in tidal volume, respiratory rate, or minute ventilation between J60 vs sham condition during hypercapnic-hypoxic ventilatory challenges.

Line 200 - what is the reason behind performing a qualitative analysis of mCherry in various quadrants? This limits the interpretation of the results. If the authors used Chat-cre rats, the virus should only be in Chat+ MN. Knowing how selective the virus is, and whether its expression was selective for Phrenic MN versus other MN pools, could address several technical questions.

We agree that detailed quantification of expression by motoneuron pool would be of value in future work.  However, for these initial proof-of-concept experiments, we performed the quadrant-based qualitative analysis of mCherry expression to provide a simple comparison of mCherry expression between groups (i.e., ChAT-Cre vs. wildtype mice). This analysis allowed us to: 1) show the reader that each animal included in the study showed evidence of mCherry expression and 2) give the reader an idea of patterns of mCherry expression throughout the mid-cervical spinal cord. Additionally, it is important to note that while ChAT is a marker of motoneurons some populations of interneurons also express ChAT(2-4).

Given the increased values of Dia EMG AUC and no changes in respiratory rate, did the authors determine if there was a change in the inspiratory time with J60 administration?

We did not assess inspiratory time.

High death rate in DREADD WT mice - was histological analysis performed on these mice? Could it be due to the large volume injected into the spinal cord that affects not only descending pathways but also ascending ones? Or caused by neuronal death due to the large volume of viral solution in injected in mice.

Histological analysis was performed on these animals to assess mCherry expression only (i.e., no staining for NeuN or other markers was performed). While the reviewer's speculations are reasonable, we feel these reasons are unlikely to explain the death rate in DREADD WT mice as ChAT-Cre mice received the same volume injected into their spine and lived up until and during diaphragm EMG recordings. Additionally, WT mice lived for 4-5 weeks post-injection which would be past the acute phase that a large immune response to the viral dose would have occurred.

Line 299-304. Can you please clarify whether these rats were tested under anesthesia?

These rats were assessed under anesthesia. This detail has been added (line 146).

Given some of the unexpected results on cardiovascular parameters in urethane anesthetized rats, did the authors test the effects of J60 in the absence of AAV construct infection?

A small cohort (n = 2) of urethane anesthetized naïve wildtype rats were given the J60 ligand (IV, 0.1 mg/kg dose). We did observe a sudden drop in blood pressure after J60 administration that was sustained for the duration of the recording. One animal showed a 12% decrease in mean arterial blood pressure following J60 administration while the other showed a 35% decrease. Thus, it does appear that in this preparation the J60 ligand is producing a drop in arterial blood pressure.

Line 393. I believe this comment is referred to the intrapleural and diaphragmatic injection. Maybe this should clarified in the sentence.

This sentence has been revised for clarity (see lines 248-250).

Figures 1 and 2. It would be informative to show raw traces of the Diaphragm EMG to demonstrate the increase in tonic EMG. It is not possible to determine that from the integrated traces in Figures 1A and B.

Thank you for bringing up this concern. While the mean data in Figures 1F and 2F do indicate that, on average, animals had tonic diaphragm EMG responses to DREADD activation, the examples given in Figures 1A and 2A show minimal responses. This makes it difficult to fully appreciate the tonic response from those particular traces. However, clear tonic activity can be appreciated from Figures 5A and S2. In these figures, tonic activity is evident from the integrated EMG signals, presenting as a sustained increase in baseline activity between bursts—essentially an upward shift from the zero point.

References

(1) Van Savage, J. & Avegno, E. M. High dose administration of DREADD agonist JHU37160 produces increases in anxiety-like behavior in male rats. Behav Brain Res 452, 114553 (2023). https://doi.org/10.1016/j.bbr.2023.114553

(2) Mesnage, B. et al. Morphological and functional characterization of cholinergic interneurons in the dorsal horn of the mouse spinal cord. J Comp Neurol 519, 3139-3158 (2011). https://doi.org/10.1002/cne.22668

(3) Gotts, J., Atkinson, L., Yanagawa, Y., Deuchars, J. & Deuchars, S. A. Co-expression of GAD67 and choline acetyltransferase in neurons in the mouse spinal cord: A focus on lamina X. Brain Res 1646, 570-579 (2016). https://doi.org/10.1016/j.brainres.2016.07.001

(4) Alkaslasi, M. R. et al. Single nucleus RNA-sequencing defines unexpected diversity of cholinergic neuron types in the adult mouse spinal cord. Nat Commun 12, 2471 (2021). https://doi.org/10.1038/s41467-021-22691-2

eLife Assessment

This important study uses single-cell transcriptomics to analyze syncytiotrophoblasts in two trophoblast organoid models compared to primary placental tissue, providing compelling insights into syncytialization and highlighting the utility of organoid models in placental research. It also serves as an invaluable resource for the field.

Reviewer #1 (Public review):

Summary:

This study provides an in-depth analysis of syncytiotrophoblast (STB) gene expression at the single-nucleus (SN) and single-cell (SC) levels, using both primary human placental tissues and two trophoblast organoid (TO) models. The authors compare the older TO model, where STB forms internally (STBin), with a newer model where STB forms externally (STBout). Through a series of comparative analyses, the study highlights the necessity of using both SN and SC techniques to fully understand placental biology. The findings demonstrate that the STBout model shows more differentiated STBs with higher expression of canonical markers and hormones compared to STBin. Additionally, the study identifies both conserved and distinct gene expression profiles between the TO models and human placenta, offering valuable insights for researchers using TOs to study STB and CTB differentiation.

Strengths:

The study offers a comprehensive SC- and SN-based characterization of trophoblast organoid models, providing a thorough validation of these models against human placental tissues. By comparing the older STBin and newer STBout models, the authors effectively demonstrate the improvements in the latter, particularly in the differentiation and gene expression profiles of STBs. This work serves as a critical resource for researchers, offering a clear delineation of the similarities and differences between TO-derived and primary STBs. The use of multiple advanced techniques, such as high-resolution sequencing and trajectory analysis, further enhances the study's contribution to the field.

Weaknesses were addressed during the revision.

The authors effectively addressed my critiques in the rebuttal letter and made corresponding changes in the manuscript. Specifically, they: 1) emphasized the importance of TO orientation in influencing STB nuclear subtype differentiation by adding text to the introduction; 2) clarified the differences in cluster numbers and names between primary tissue and TO data, explaining that each dataset was analyzed independently with separate clustering algorithms and adding clarifying text to the results section; 3) included additional rationale for using SN over SC sequencing, particularly for studying the multinucleated STB; 4) acknowledged that their original evidence was insufficient to definitively determine STBout nuclei differentiation status and removed language suggesting STB-3 as a terminally differentiated subtype, presenting alternative hypotheses in the discussion; and 5) incorporated new figures and clarifications, including RNA-FISH experiments, to validate subtype-specific marker gene expression. Overall, the authors' revisions strengthened the manuscript and aligned well with my critiques.

Reviewer #1 (Public review):

Summary:

This study provides an in-depth analysis of syncytiotrophoblast (STB) gene expression at the single-nucleus (SN) and single-cell (SC) levels, using both primary human placental tissues and two trophoblast organoid (TO) models. The authors compare the older TO model, where STB forms internally (STBin), with a newer model where STB forms externally (STBout). Through a series of comparative analyses, the study highlights the necessity of using both SN and SC techniques to fully understand placental biology. The findings demonstrate that the STBout model shows more differentiated STBs with higher expression of canonical markers and hormones compared to STBin. Additionally, the study identifies both conserved and distinct gene expression profiles between the TO models and human placenta, offering valuable insights for researchers using TOs to study STB and CTB differentiation.

Strengths:

The study offers a comprehensive SC- and SN-based characterization of trophoblast organoid models, providing a thorough validation of these models against human placental tissues. By comparing the older STBin and newer STBout models, the authors effectively demonstrate the improvements in the latter, particularly in the differentiation and gene expression profiles of STBs. This work serves as a critical resource for researchers, offering a clear delineation of the similarities and differences between TO-derived and primary STBs. The use of multiple advanced techniques, such as high-resolution sequencing and trajectory analysis, further enhances the study's contribution to the field.

Weaknesses were addressed during the revision.

The authors effectively addressed my critiques in the rebuttal letter and made corresponding changes in the manuscript. Specifically, they: 1) emphasized the importance of TO orientation in influencing STB nuclear subtype differentiation by adding text to the introduction; 2) clarified the differences in cluster numbers and names between primary tissue and TO data, explaining that each dataset was analyzed independently with separate clustering algorithms and adding clarifying text to the results section; 3) included additional rationale for using SN over SC sequencing, particularly for studying the multinucleated STB; 4) acknowledged that their original evidence was insufficient to definitively determine STBout nuclei differentiation status and removed language suggesting STB-3 as a terminally differentiated subtype, presenting alternative hypotheses in the discussion; and 5) incorporated new figures and clarifications, including RNA-FISH experiments, to validate subtype-specific marker gene expression. Overall, the authors' revisions strengthened the manuscript and aligned well with my critiques.

Reviewer #3 (Public review):

In this report, Keenen et al. present a thoroughly characterized platform for identifying potential molecular mechanisms regulating syncytiotrophoblast cell functions in placental biology. Application of single cell assessments to identify developmental trajectories of this lineage have been challenging due to the complex, multinucleated structure of the syncytium. The authors provide a comprehensive comparative assessment of term placental tissue and three independent trophoblast organoid models. They use single cell and single nucleus RNA sequencing followed by differential gene expression and pseudotime analyses to identify subpopulations and differentiation trajectories. They further compare the datasets generated in this study to publicly available datasets from first trimester placental tissue. The work is timely as optimization of trophoblast organoids is an evolving topic in placental research. And careful characterization of in vitro models has been noted as essential for model selection and result interpretation in the field.

The study elucidates syncytiotrophoblast nucleus subtypes and proportions in three different organoid models and compares subtypes and gene expression signatures to placental tissues. This work advances the field by demonstrating the utility of different trophoblast organoids to model syncytiotrophoblast differentiation. The in-depth characterization of cell types comprising the different organoid models and how they compare to placental tissue will help to inform model selection for future experimentation in the field. Defining cell composition and cell differentiation trajectories will also aid in data interpretation for data generated by these tissue and model sources. Overall, the conclusions presented in the manuscript are well supported by the data. The figures, as presented, are informative and striking.

The authors present outstanding progress toward their overall aim of identifying, "the underlying control of the syncytiotrophoblast". They identify the chromatin remodeler, RYBP, as well as other regulatory networks that they propose are critical to syncytiotrophoblast development.

The initial study was limited in fully addressing the aim, however, as functional evidence for the contributions of the factors/pathways to syncytiotrophoblast cell development was absent. In a revised version of the manuscript, the authors report the first application of CRISPR-mediated gene silencing in a TO model. They use CRISPR-Cas9-mediated gene targeting to generate RYBP and AFF1 knockout models. Deletion of either RYBP or AFF1 increased STB-2 marker gene expression, as determined using bulk RNA-seq. Future experimentation will assess the distribution of STB nuclear subtypes in the RYBP and AFF1 knockout models and explore the essentiality of RYBP, AFF1, and other identified factors to syncyiotrophoblast development and function.

Localization and validation of the identified factors within tissue and at the protein level will also provide further contextual evidence to address the hypotheses generated. In a revised version of the manuscript, the authors localize STB markers PAPPA2 and ADAMTS6 in TOs using RNA-FISH. Future work will aim to further validate the markers and hypotheses generated from this study.

Author response:

The following is the authors’ response to the original reviews

We thank the public reviewers and editors for their insightful comments on the manuscript. We have made the following changes to address their concerns and think the resulting manuscript is stronger as a result. Specifically, we have 1) added RNA FISH data of specific STB-2 and STB-3 RNA markers to confirm their distribution changes between STBⁱⁿ and STB^out TOs, 2) removed language throughout the text that refer to STB-3 as a terminally differentiated nuclear subtype, and 3) generated CRISPR-mediated knock-outs of two genes identified by network analysis and validated their rolse in mediating STB nuclear subtype gene expression.

Reviewer #1 (Public review): 

Strengths: 

The study offers a comprehensive SC- and SN-based characterization of trophoblast organoid models, providing a thorough validation of these models against human placental tissues. By comparing the older STBⁱⁿ and newer STB^out models, the authors effectively demonstrate the improvements in the latter, particularly in the differentiation and gene expression profiles of STBs. This work serves as a critical resource for researchers, offering a clear delineation of the similarities and differences between TO-derived and primary STBs. The use of multiple advanced techniques, such as high-resolution sequencing and trajectory analysis, further enhances the study's contribution to the field. 

Thank you for your thoughtful review—we appreciate your recognition of our efforts to comprehensively validate trophoblast organoid models and highlight key advancements in STB differentiation and gene expression.

Weaknesses: 

While the study is robust, some areas could benefit from further clarification. 

(1) The importance of the TO model's orientation and its impact on outcomes could be emphasized more in the introduction. 

We agree that TO orientation may significantly influence STB nuclear subtype differentiation. As the STB is critical for both barrier formation and molecular transport in vivo, lack of exposure to the surrounding media in STBⁱⁿ TOs in vitro could compromise these functions and the associated environmental cues that influence STB nuclear differentiation. We have added text to the introduction to highlight this point (lines 117-120).

(2) The differences in cluster numbers/names between primary tissue and TO data need a clearer explanation, and consistent annotation could aid in comparison. 

Thank you for highlighting that the comparisions and cluster annotations need clarification. In Figure 1, we did not aim to directly compare CTB and STB nuclear subtypes between TOs and tissue. Each dataset was analyzed independently, with clusters determined separately and with different resolutions decided via a clustering algorithm (Zappia and Oshlack, 2018). For example, for the STB, this approach identified seven subtypes in tissue but only two in TOs, making direct comparison challenging. To address this challenge, we integrated the SN datasets from TOs and tissue in Figure 6. This integration allowed us to directly compare gene expression between the sample types and examine the proportions within each STB subtype. Similarly, in Figure 2, direct comparison of individual CTB or STB clusters across the separate datasets is challenging (Figures 2A-C) due to differences in clustering. To overcome this, we integrated the datasets to compare cluster gene expression and relative proportions (Figures 2D-E). Nonetheless, to address the reviewers concern we have added text to the results section to clarify that subclusters of CTB and STB between datasets should not be directly compared until the datasets are integrated in Figure 2D-E and Figure 6 (lines 166-167).

(3) The rationale for using SN sequencing over SC sequencing for TO evaluations should be clarified, especially regarding the potential underrepresentation of certain trophoblast subsets. 

This is an important point as the challenges of studying a giant syncytial cell are often underappreciated by researchers that study mononucleated cells. We have added text to the introduction to clarify why traditional single cell RNA sequencing techniques were inadequate to collect  and characterize the STB (lines 91-93).

(4) Additionally, more evidence could be provided to support the claims about STB differentiation in the STB^out model and to determine whether its differentiation trajectory is unique or simply more advanced than in STBⁱⁿ. 

Our original conclusion that STB^out nuclei are more terminally differentiated than STBⁱⁿ was based on two observations: (1) STB^out TOs exhibit increased expression of STB-specific pregnancy hormones and many classic STB marker genes and (2) STB^out nuclei show an enrichment of the STB-3 nuclear subtype, which appears at the end of the slingshot pseudotime trajectory. However, upon consideration of the reviewer comments, we agree that this evidence is not sufficient to definitively distinguish if STB^out nuclei are more advanced or follow a unique differentiation trajectory dependent on new environmental cues. Pseudotime analyses provided only a predictive framework for lineage tracing, and these predictions must be experimentally validated. Real-time tracking of STB nuclear subtypes in TOs would require a suite of genetic tools beyond the scope of this study. Therefore, to address the reviewers' concerns we have removed language suggesting that STB-3 is a terminally differentiated subtype or that STB^out nuclei are more differentiated than STBⁱⁿ nuclei throughout the text until the discussion. Therein we present both our original hypothesis (that STB nuclei are further differentiated in STB^out) and alternative explanations like changing trajectories due to local environmental cues (lines 619-625).

Reviewer #2 (Public review): 

Strengths: 

(1) The use of SN and SC RNA sequencing provides a detailed analysis of STB formation and differentiation. 

(2) The identification of distinct STB subtypes and novel gene markers such as RYBP offers new insights into STB development. 

Thank you for highlighting these strengths—we appreciate your recognition of our use of SN and SC RNA sequencing to analyze STB differentiation and the discovery of distinct STB subtypes and novel gene markers like RYBP.

Weaknesses: 

(1) Inconsistencies in data presentation. 

We address the individual comments of reviewer 2 later in this response.

(2) Questionable interpretation of lncRNA signals: The use of long non-coding RNA (lncRNA) signals as cell type-specific markers may represent sequencing noise rather than true markers. 

We appreciate the reviewer’s attention to detail in noticing the lncRNA signature seen in many STB nuclear subtypes. However, we disagree that these molecules simply represent sequencing noise. In fact, may studies have rigorously demonstrated that lncRNAs have both cell and tissue specific gene expression (e.g., Zhao et al 2022, Isakova et al 2021, Zheng et al 2020). Further, they have been shown to be useful markers of unique cell types during development (e.g., Morales-Vicente et al 2022, Zhou et al 2019, Kim et al 2015) and can enhance clustering interpretability in breast cancer (Malagoli et al 2024). Many lncRNAs have also been demonstrated to play a functional role in the human placenta, including H19, MEG3, and MEG8 (Adu-Gyamfi et al 2023) and differences are even seen in nuclear subtypes in trophoblast stem cells (Khan et al 2021). Therefore, we prefer to keep these lncRNA signatures included and let future researchers test their functional role.

To improve the study's validity and significance, it is crucial to address the inconsistencies and to provide additional evidence for the claims. Supplementing with immunofluorescence staining for validating the distribution of STB_in, STB_out, and EVT_enrich in the organoid models is recommended to strengthen the results and conclusions. 

Each general trophoblast cell type (CTB, STB, EVT) has been visualized by immunofluorescence by the Coyne laboratory in their initial papers characterizing the STBⁱⁿ, STB^out, and EVT^enrich models (Yang et al, 2022 and 2023). We agree that it is important to validate the STB nuclear subtypes found in our genomic study. However, one challenge in studying a syncytia is that immunofluorescence may not be a definitive method when the nuclei share a common cytoplasm. This is because protein products from mRNAs transcribed in one nucleus are translated in the cytoplasm and could diffuse beyond sites of transcription. Therefore, RNA fluorescence in situ hybridization (RNA-FISH) is instead needed. While a systematic characterization of the spatial distribution of the many marker genes found each subtype is outside the scope of this study, we include RNA-FISH of one STB-2 marker (PAPPA2) and one STB-3 marker (ADAMTS6) in Figure 3F-G and Supplemental Figure 3.3. This demonstrates there is an increase in STB-2 marker gene expression in STBⁱⁿ TOs and an increase in STB-3 marker gene expression in STB^out TOs. 

Reviewer #3 (Public review):  

The authors present outstanding progress toward their aim of identifying, "the underlying control of the syncytiotrophoblast". They identify the chromatin remodeler, RYBP, as well as other regulatory networks that they propose are critical to syncytiotrophoblast development. This study is limited in fully addressing the aim, however, as functional evidence for the contributions of the factors/pathways to syncytiotrophoblast cell development is needed. Future experimentation testing the hypotheses generated by this work will define the essentiality of the identified factors to syncytiotrophoblast development and function. 

We thank the reviewer for their thoughtful assessment, constructive feedback, and encouraging comments. We acknowledge that the initial manuscript primarily presented analyses suggesting correlations between RYBP and other factors identified in the gene network analysis and STB function. Understanding how gene networks in the STB are formed and regulated is a long-term goal that will require many experiments with collaborative efforts across multiple research groups.

Nonetheless, to address this concern we have knocked out two key genes, RYBP and AFF1, in TOs using CRISPR-Cas9-mediated gene targeting. Bulk RNA sequencing of STBⁱⁿ TOs from both wild-type (WT) and knockout strains revealed that deletion of either gene caused a statistically significant decrease in the expression of the pregnancy hormone human placental lactogen and an increase in the expression of several genes characteristic of the oxygen-sensing STB-2 subtype, including FLT-1, PAPPA2, SPON2, and SFXN3. These findings demonstrate that knocking out RYBP or AFF1 results in an increase in STB-2 marker gene expression and therefore play a role in inhibiting their expression in WT TOs (Figure 5D-E and supplemental Figure 5.2). We also note that this is the first application of CRISPR-mediated gene silencing in a TO model.

Future work will visualize the distribution of STB nuclear subtypes in these mutants and explore the mechanistic role of RYBP and AFF1 in STB nuclear subtype formation and maintenance. However, these investigations fall outside the scope of the current study.

Localization and validation of the identified factors within tissue and at the protein level will also provide further contextual evidence to address the hypotheses generated. 

We agree that visualizing STB nuclear subtype distribution is essential for testing the many hypotheses generated by our analysis. To address this, we have included RNA-FISH experiments for two STB subtype markers (PAPPA2 for STB-2 and ADAMTS6 for STB-3) in TOs. These experiments reveal an increase in PAPPA2 expression in STBⁱⁿ TOs and an increase in ADAMTS6 expression in STB^out TOs (Figure 3F-G and Supplemental Figure 3.3). Genomic studies serve as powerful hypothesis generators, and we look forward to future work—both our own and that of other researchers—to validate the markers and hypotheses presented from our analysis.

Recommendations for the authors: 

Reviewing Editor Comments: 

We strongly encourage the authors to further strengthen the study by addressing all reviewers' comments and recommendations, with particular attention to the following key aspects:

(1) Clarifying the uniqueness of the STB differentiation trajectory between STBⁱⁿ and STB^out, and determining whether STB^out represents a more advanced stage of differentiation compared to STBⁱⁿ. It is also important to specify which developmental stage of placental villi the STB^out and STBⁱⁿ are simulating. 

We have revised the manuscript to remove definitive language claiming that STB-3 represents a terminally differentiated subtype or that STB^out nuclei are more differentiated than STBⁱⁿ nuclei. Instead, we now present our hypothesis and alternative explanations in the discussion (lines 619-625), and emphasize the need for experimental validation of pseudotime predictions to test these hypotheses.

(2) Utilizing immunofluorescence to validate the distribution of cell types in the organoid models. 

The Coyne lab has previously performed immunofluorescence of CTB and STB markers in STBⁱⁿ and STB^out TOs (Yang et al 2023). The syncytial nature of STBs complicates immunofluorescence-based validation of the STB nuclear subtypes due translating proteins all sharing a single common cytoplasm and therefore being able to diffuse and mix. Instead, we performed RNA-FISH for two STB subtype markers (PAPPA2, STB-2 and ADAMTS6, STB-3), which showed subtype-specific nuclear enrichment in STBⁱⁿ and STB^out TOs, respectively (Figure 3F-G and Supplemental Figure 3.3).  

(3) Addressing concerns regarding the use of lncRNA as cell marker genes. Employing canonical markers alongside critical TFs involved in differentiation pathways to perform a more robust cell-type analysis and validation is recommended.  

As discussed in detail above, we maintain that lncRNAs are valuable markers, supported by their demonstrated roles in cell and tissue specificity and placental function. These signatures provide important insights and hypotheses for future research, and we have clarified this rationale in the revised manuscript.

Reviewer #1 (Recommendations for the authors): 

(1) The authors have presented an extensive SC- and SN-based characterization of their improved trophoblast TO model, including a comparison to human placental tissues and the previous TO iteration. In this way, the authors' work represents an invaluable resource for investigators by providing thorough validation of the TO model and a clear description of the similarities and differences between primary and TO-derived STBs. I would suggest that the authors reshape the study to further highlight and emphasize this aspect of the study. 

We thank the reviewer for their thoughtful recommendation and agree that our datasets will serve as an invaluable resource for comparing in vitro models to in vivo gene expression. However, extensive validation is required to make definitive conclusions about the extent to which these systems mirror one another and where they diverge. For this reason, in this manuscript, we have focused on characterizing STB subtypes to provide a foundational understanding of the model and this poorly characterized subtype.

(2) Introduction, Paragraph 3: What is the importance of orientation for the trophoblast TO model? The authors may consider removing some of the less important methodologic details from this paragraph and including more emphasis on why their TO model is an improvement. 

Text has been added to this paragraph to highlight the importance of outward facing STB orientation, which is essential to mirror the STB’s transport function in vitro (lines 118-120).

(3) Results, Figure 1: In addition to the primary placental tissue plots showing all cell populations, it may be useful to have side-by-side versions of similar plots showing only the trophoblast subsets, so that the primary and TO data could be more easily compared visually. 

This has been implemented and added to the Supplemental Figure 1.4.

(4) Results, Figure 1: In simple terms, what is the reason for ending up with different cluster numbers/names from the primary tissue and TO? Would it be possible to apply the same annotation to each (at least for trophoblast types) and thus allow direct comparison between the two? 

As described above, each dataset was separately analyzed and clusters determined with an algorithm to determine the optimal clustering resolution. Therefore, the number of clusters between each dataset cannot be directly compared until the SN TO and tissue datasets are integrated together in Figure 6. We have added text to the manuscript to make it clear that they should not be compared except for in bulk number until this point (230-232).

(5) Results, Figure 2: For subsequent evaluation of different in vitro TO conditions, did the authors use only SN sequencing because they wanted to focus on STB? Based on Figure 1, it seems some CTB subsets would be underrepresented if using only SN. Given that the authors look at both STB and CTB in their different TOs, is this an issue? 

The CTB clusters that showed the greatest divergence between SC and SN datasets were those associated with mitosis and the cell cycle, likely due to nuclear envelope breakdown interfering with capture by the 10x microfluidics pipeline. While cytoplasmic gene expression provides valuable insights into CTB function, our manuscript focuses on the STB starting from Figure 2. Since the STB is captured exclusively by the SN dataset, we concentrated on this approach to streamline our analysis.

(6) Results, Figure 3: What do the authors consider to be the primary contributing factors for why the STB subsets display differential gene expression between STBⁱⁿ and STB^out? Is this due primarily to the cultural conditions and/or a result of the differing spatial arrangement with CTBs? 

This is an intriguing question that is challenging to disentangle because the culture conditions are integral to flipping the orientation. The two primary factors that differ between STBⁱⁿ and STB^out TOs are the presence of extracellular matrix in STBⁱⁿ and direct exposure to the surrounding media in STB^out. We believe these environmental cues play a significant role in shaping the gene expression of STB subsets. Fully disentangling this relationship would require a method to alter the TO orientation without changing the culture conditions. While this is an exciting direction for future research, it falls outside the scope of the present study.

(7) Results, Figure 4: The authors' analysis indicates that the STB nuclei from the STB^out TO are likely "more differentiated" than those in STBⁱⁿ TO. Could the authors provide some qualitative or quantitative support for this? Is the STB^out differentiated phenotype closer to what would be observed in a fully formed placenta? 

As discussed earlier, we agree with the reviewers that this claim should be removed from the text outside of the discussion.

(8) Results, Figure 5: Based on the trajectory analysis, do the authors consider that the STB from STB^out TO are simply further along the differentiation pathway compared to those from STBⁱⁿ TO, or do the STB from STB^out TO follow a differentiation pathway that is intrinsically distinct from STBⁱⁿ TO? 

We think the idea of an intrinsically distinct pathway is a fascinating alternative hypothesis and have added it into the discussion. We do not find the pseudotime currently allows us to answer this question without additional experiments, so we have removed claims that the STB^out STB nuclei are further along the differentiation pathway.

(9) Results, Figure 6: A notable difference between the STB^out TO and the term tissue is that the CTB subsets are much more prevalent. Is this simply a scale difference, i.e. due to the size of the human placenta compared to the limited STB nuclei available in the STB^out TO? Or are there other contributing factors? 

The proportion of CTB to STB nuclei in our term tissue (9:1) aligns with expectations based on stereological estimates. We believe the relatively low number of CTB nuclei in our dataset is due to the need for a larger sample size to capture more of this less abundant cell type. Since the primary focus of this paper is on STB, and we analyzed over 4,000 STB nuclei, we do not view this as a limitation. However, future studies utilizing SN to investigate term tissue should account for the abundance of STB nuclei and plan their sampling carefully to ensure sufficient representation of CTB nuclei if this is a desired focus.

Reviewer #2 (Recommendations for the authors): 

(1) The color annotations for cell types in Figure 2 are inconsistent between the different panels, and the term "Prolif" in Figure 2E is not explained by the authors. 

We chose colors to enhance visibility on the UMAP. We do not wish readers to make direct comparisons between the different CTB or STB subtypes of the sample types until the datasets are integrated in Figure 2D. This is because an algorithm for the clustering resolution has been chosen independently for each dataset. Cluster proportions are better compared in the integrated datasets in Figure 2D. We have added text to the results section to make this clear to the reader (lines 166-167).

(2) In Figure 3 and Supplementary Figures 1.3, the authors frequently present long non-coding RNA (lncRNA) signals as cell type-specific markers in the bubble plots. These signals are likely sequencing noise and may not accurately represent true markers for those cell types. It is recommended to revise this interpretation. 

As referenced above, there are many examples of lncRNAs that have biological and pathological significance in the placenta (H19, Meg3, Meg8) and lncRNAs often have cell type specific expression that can enhance clustering. We prefer to keep these signatures included and let future researchers determine their biological significance.

(3) In Figure 3C, the authors performed pathway enrichment analysis on the STB subtypes after integrating STB_in and STB_out organoids. The enrichment of the "transport across the blood-brain barrier" pathway in the STB-3 subtype does not align with the current understanding of STB cell function. Please provide corresponding supporting evidence. Additionally, please verify whether the other functional pathways represent functions specific to the STB subtypes. 

Interestingly, many of the genes categorized under “transport across the blood-brain barrier” are transporters shared with “vascular transport.” These include genes involved in the transport of amino acids (SLC7A1, SLC38A1, SLC38A3, SLC7A8), molecules essential for lipid metabolism (SLC27A4, SLC44A1), and small molecule exchange (SLC4A4, SLC5A6). Given that the vasculature, the STB, and the blood-brain barrier all perform critical barrier functions, it is unsurprising that molecules associated with these GO terms are enriched in the STB-3 subtype, which expresses numerous transporter proteins. Since the transport of materials across the STB is a well-established function, we have not included additional supporting evidence but have clarified the genes associated with this GO term in the text (lines 392-394 and supplemental Table 9).

(4) The pseudotime heatmap in Figure 4B is not properly arranged and is inconsistent with the differentiation relationships shown in Figure 4A. It is recommended to revise this. 

We are uncertain which aspect of the heatmap in Figure 4A is perceived as inconsistent with Figure 4B. One distinction is that pseudotime in Figure 4A is normalized from 0 to 100 to fit the blue-to-yellow-to-red color scale, whereas in Figure 4B, the color scale is not normalized and the color bar ranging from white to red. This difference reflects our intent to simplify Figure 4B-C, as the abundance of color between cell types and gene expression changes required a streamlined representation to ensure the figure remained clear and easy to interpret. This is classically done in the field and consistent with the default code in the slingshot package.

(5) In Figures 4C and 4D, although RYBP is highly expressed in STB, it is difficult to support the conclusion that RYBP shows the most significant expression changes. It is recommended to provide additional evidence. 

The claim that RYBP exhibits the most significant expression changes was based on p-value ordering of genes associated with pseudotime via the associationTest function in slingshot and not with immunofluorescence data. The text has been revised to make this distinction clear (lines 390-393).

(6) In Figure 4E, staining for CTB marker genes is missing, and in Figure 4F, CYTO is difficult to use as a classical STB marker. It is recommended to use the CGBs antibody from Figure 4E as a STB marker for staining to provide evidence.  

We have revised the Figure 5B-C to use e-Cadherin as a CTB marker gene in TOs and CGB antibody as a marker of STB.

In tissue, however, obtaining a good STB marker that does not overlap with the RYBP antibody (rabbit) in term tissue is difficult as the STB downregulates hCG expression closer to term to initiate contractions. SDC1 is often used but only labels the plasma membrane so does not help in distinguishing the STB cytoplasm. We have added an image of cytokeratin, e-Cadherin, and the STB marker ENDOU to validate that our current approach with e-Cadherin and cytokeratin allows us to accurately distinguish between CTB and STB cells.

(7) The velocity results in Figure 5A do not align with the differentiation relationships between cells and contradict the pseudotime results presented in Figure 4 by the authors. 

The reviewer raises an interesting observation regarding the velocity map in Figure 5A, which appears to show a bifurcation into two STB subtypes. This observation aligns with similar findings reported in tissue by our colleagues (Wang et al., 2024). However, given the low number of CTB cells in our tissue dataset, we were cautious about making definitive conclusions about pseudotime without a larger sample size. Notably, the RNA velocity map closely resembles the pseudotime trajectory in TOs, with CTB transitioning into the CTB-pf subtype and subsequently into the STB. One potential explanation for discrepancies between tissue and TOs is the difference in nuclear age: nuclei in tissue can be up to nine months old, whereas those in TOs are only hours or days old. It is possible that the lineage in TOs could bifurcate if cultured for longer than 48 hours, but our current dataset captures only the early stages of the STB differentiation process. While exploring these hypotheses is fascinating, they are beyond the scope of this current study.  

Reviewer #3 (Recommendations for the authors): 

Amazing work - I greatly enjoyed reading the manuscript. Here are a few questions and suggestions for consideration: 

Evidence presented throughout the results sections hints that the organoids may represent an earlier stage of placental development compared to the term. Increased hCG gene expression is observed, but as noted expression is decreased in term STB. STB:CTB ratios are also higher at term compared to the first trimester, etc. It was difficult to conclude definitively based on how data is presented in Fig 6 and discussed. Maybe there is no clear answer. Perhaps the altered cell type ratios in the organoid models (e.g., few STB in EVT enrich conditions) impact recapitulation of the in vivo local microenvironment signaling. As such, can the authors speculate on whether cell ratios could be strategically leveraged to model different gestational time points? 

Along these same lines, syncytiotrophoblast in early implantation (before proper villi development) is often described as invasive and later at the tertiary villi stage defined by hormone production, barrier function, and nutrient/gas exchange. Do the authors think the different STB subtypes captured in the organoid models represent different stages/functions of syncytiotrophoblast in placental development? 

Minor Comments 

(1) Please clarify what the third number represents in the STB:CTB ratio (e.g., 1:3:1 and 2:5:1). EVT? 

The first number is a decimal point and not a colon (ie 1.3 and 2.5). Therefore these numbers are to be read as the STB:CTB ratio is 1.3 to 1 or 2.5 to 1.

(2) Could consider co-localizing RYBP in term tissue with a syncytio-specific marker like CGB used for organoids (Fig 4F). 

We addressed this concern in comment 6 to reviewer 2.

(3) Recommend defining colors-which colors represent which module in Figure 5C in the legend and main body text. I see the labels surrounding the heatmap in 5B, but defining colors in text (e.g. cyan, magenta, etc.) would be helpful. Do the gray circles represent targets that don't belong to a specific module? Are the bolded factor names based on a certain statistical cutoff/defining criteria or were they manually selected? 

The text of both the results and figure legends has been revised to clarify these points.

(4) Data Availability: It would be helpful to provide supplemental table files for analyses (e.g., 5C to list the overlapping relationships in TGs for each TF/CR (5C) and 3E/6F to list DEG genes in comparisons). 

Supplemental files for each analysis have been added (Supplemental Table 8-14). In addition, the raw and processed data is available on GEO and we have created an interactive Shiny App so people without coding experience can interact with each dataset (lines 917-919).

(5) “...and found that each sample expressed these markers (Figure 6D), suggesting..." Consider clarifying "these". 

Text has been added to refer to a few of these marker genes within the text (line 540).

Citations

(1) Zappia L, Oshlack A. Clustering trees: a visualization for evaluating clusterings at multiple resolutions. GigaScience. 2018;7(7):giy083. PMCID: PMC6057528

(2) Zhou J, Xu J, Zhang L, Liu S, Ma Y, Wen X, Hao J, Li Z, Ni Y, Li X, Zhou F, Li Q, Wang F, Wang X, Si Y, Zhang P, Liu C, Bartolomei M, Tang F, Liu B, Yu J, Lan Y. Combined Single-Cell Profiling of lncRNAs and Functional Screening Reveals that H19 Is Pivotal for Embryonic Hematopoietic Stem Cell Development. Cell Stem Cell. 2019;24(2):285-298.e5. PMID: 30639035

(3) Malagoli G, Valle F, Barillot E, Caselle M, Martignetti L. Identification of Interpretable Clusters and Associated Signatures in Breast Cancer Single-Cell Data: A Topic Modeling Approach. Cancers. 2024;16(7):1350. PMCID: PMC11011054

(4) Adu-Gyamfi EA, Cheeran EA, Salamah J, Enabulele DB, Tahir A, Lee BK. Long non-coding RNAs: a summary of their roles in placenta development and pathology†. Biol Reprod. 2023;110(3):431–449. PMID: 38134961

(5) Zheng M, Hu Y, Gou R, Nie X, Li X, Liu J, Lin B. Identification three LncRNA prognostic signature of ovarian cancer based on genome-wide copy number variation. Biomed Pharmacother. 2020;124:109810. PMID: 32000042

(6) Khan T, Seetharam AS, Zhou J, Bivens NJ, Schust DJ, Ezashi T, Tuteja G, Roberts RM. Single Nucleus RNA Sequence (snRNAseq) Analysis of the Spectrum of Trophoblast Lineages Generated From Human Pluripotent Stem Cells in vitro. Front Cell Dev Biol. 2021;9:695248. PMCID: PMC8334858

(7) Isakova A, Neff N, Quake SR. Single-cell quantification of a broad RNA spectrum reveals unique noncoding patterns associated with cell types and states. Proc Natl Acad Sci United States Am. 2021;118(51):e2113568118. PMCID: PMC8713755

(8) Morales-Vicente DA, Zhao L, Silveira GO, Tahira AC, Amaral MS, Collins JJ, Verjovski-Almeida S. Singlecell RNA-seq analyses show that long non-coding RNAs are conspicuously expressed in Schistosoma mansoni gamete and tegument progenitor cell populations. Front Genet. 2022;13:924877. PMCID: PMC9531161

(9) Kim DH, Marinov GK, Pepke S, Singer ZS, He P, Williams B, Schroth GP, Elowitz MB, Wold BJ. Single-Cell

Transcriptome Analysis Reveals Dynamic Changes in lncRNA Expression during Reprogramming. Cell Stem Cell. 2015;16(1):88–101. PMCID: PMC4291542

(10) Yang L, Liang P, Yang H, Coyne CB. Trophoblast organoids with physiological polarity model placental structure and function. bioRxiv. 2023;2023.01.12.523752. PMCID: PMC9882188