Reviewer #3 (Public review):

Summary:

The manuscript by Dearborn et al investigates the kinetics of intron splicing in inflammation-associated transcripts after TNF-stimulation of macrophages, using targeted sequencing of chromatin-associated RNA to obtain high coverage across a focused set of induced genes. The authors' main conclusion is that splicing kinetics are heterogeneous across these transcripts, and that delayed introns (which they term "bottleneck introns") are associated with weak donor sequences. Using a deep learning approach, they have also identified additional sequence features that might contribute to intron splicing kinetics.

Overall, I think the findings in the manuscript are very intriguing and will be of interest to readers working on RNA biology. The changes the authors have made to the manuscript in response to some very valid comments from reviewers have strengthened the manuscript. While the existing data might not be sufficient to directly address some of the broader mechanistic claims made by the authors, I think the findings are nonetheless very interesting and should contribute towards a better understanding of the post-transcriptional regulation of gene expression.

Strengths:

A strength of the manuscript is the experimental design. The targeted capture approach is innovative and well-suited to the goal of measuring intron-specific splicing behaviour across time. The inclusion of experimental validation in minigene assays of some of the computational predictions also strengthens the claims made by the authors.

The authors have made a constructive effort to address some of the concerns raised in a previous round of review. The revised manuscript reads as a balanced text.

Weaknesses:

The study still does not fully resolve the downstream consequences of delayed splicing. In particular, it remains unclear whether the bottleneck introns lead primarily to delayed production of mature transcripts, reduced productive transcript output, or some combination of the two.

On a related point, the minigene reporter assays measure a steady-state level of the transcript and don't provide insights into the kinetics directly.

Lastly, given that the detailed analyses were performed on a selected subset of (inflammation-induced) transcripts, a broader evolutionary interpretation needs to be restrained given the current data.

Author response:

We thank the Reviewing Editor and reviewers for their thoughtful and constructive evaluation of our manuscript, Programmed Delayed Splicing: A Mechanism for Timed Inflammatory Gene Expression. We are encouraged that the reviewers found the study valuable, the experimental design strong for the core findings. We appreciate the reviewers’ careful attention to the limits of inference in several parts of the manuscript, and will address these points in a revised version. We especially want to acknowledge that this paper has benefited from the abiding interest in splicing regulation by the editors and reviewers who have meticulously improved nearly every aspect of this multifaceted work in its present state.

Our planned revisions will focus on five areas. First, we will more carefully evaluate and discuss the extent to which the hybrid-capture strategy may impose position-dependent constraints on apparent splicing behavior, particularly across 5′ and 3′ introns. Second, we will clarify the use of the term “bottleneck introns,” distinguishing descriptive use in the main text from the ranked subsets used in downstream analyses. Third, we will revise the framing of the reporter assays to make explicit that these measure steady-state reporter output and do not, on their own, resolve all downstream kinetic consequences of delayed splicing. Fourth, we will clarify the interpretation of the actinomycin D experiments as providing estimates of intron excision behavior under transcriptional arrest rather than a complete time-resolved model of splicing during TNF induction. Fifth, we will substantially revise the scope and stated limitations of the deep learning-aided interpretations of data in this work.

Reviewer #1

We thank Reviewer #1 for the positive assessment of the hybrid-capture strategy, the splice-site reporter experiments, and the potential value of the neural-network-based analysis. We appreciate the reviewer’s view that these approaches help extend a well-established system for studying temporal gene expression in TNF-stimulated macrophages. We address the main concerns raised in the public review below.

(1) While evidence is provided that these introns are distinct from previously published splicing kinetics studies, “bottleneck” introns are not adequately placed in context for assessment of how they are similar or different.

We appreciate this point and agree that the current manuscript does not yet place these introns in sufficiently clear context relative to prior literature. Our study builds on foundational work describing regulated changes in splicing kinetics, widespread intron retention, and detained introns as biologically meaningful modes of gene regulation, including transcript-specific regulation of splicing in response to stress (Pleiss, Mol Cell., 2007), widespread functional intron retention in mammals (Braunschweig, Genome Res., 2014), and the definition of detained introns as a distinct class of post-transcriptionally spliced introns (Boutz, Genes Dev., 2015). In revision, we will expand the comparison to previously described classes of delayed or retained introns and clarify more explicitly how the introns studied here are defined in the setting of inducible inflammatory transcripts and their temporal resolution over the course of stimulation. We will also revise the relevant Results and Discussion text so that the distinction is made directly in the manuscript rather than relying on inference from the broader presentation.

(2) Splicing reporters are a good approach, but the complexities of post-transcriptional gene expression regulation are not adequately addressed.

We agree that the interpretive limits of the reporter assays should be stated more clearly and consistently. In revision, we will revise the presentation of the minigene experiments to make explicit that these are steady-state reporter assays and therefore do not, on their own, resolve all downstream kinetic consequences of delayed splicing in the endogenous context. At the same time, we believe the assay remains informative because it provides a controlled system in which the contribution of splice donor sequence can be tested directly in matched reporter constructs. In that sense, the reporter experiments are valuable as a reductionist test of whether weak donor sequences are sufficient to alter reporter output, even if they do not fully recapitulate the broader endogenous post-transcriptional environment. We will emphasize that these data support an association between weak donor sites and altered reporter output, while moderating any broader mechanistic claims that extend beyond what the assay directly measures.

(3) Deep learning models are a potentially powerful tool for identifying novel regulatory sequences; however, their use here is underdeveloped.

We appreciate this concern and agree that the deep-learning section should be revised substantially. In a revised manuscript, we will clarify the training setup, the definition of the slow-intron subsets used in downstream analyses, and the interpretation of the attribution and motif analyses. Alongside, we believe the assay remains informative because it provides a controlled system in which the contribution of splice donor sequence can be tested directly in matched reporter constructs. In that respect, the reporter experiments are valuable as a reductionist test of whether weak donor sequences are sufficient to alter reporter output, even if they do not fully recapitulate the broader endogenous post-transcriptional environment. We will revise the framing of these results so that they are presented more explicitly as identifying candidate sequence features associated with delayed splicing, rather than as direct evidence of specific causal regulatory mechanisms.

Reviewer #2

We thank Reviewer #2 for the thoughtful and detailed comments, and for recognizing the strengths of the measurement strategy and the clarity of the manuscript. We appreciate the reviewer’s view that the study will be of interest to a broad audience, and we agree that several conclusions will be strengthened by additional analysis and clearer explanation. We address the main concerns raised in the public review below.

(1) Concern regarding possible bias of the hybrid-capture strategy toward introns closer to the 3′ end, and whether 5′ introns should be treated separately in some analyses.

We thank the reviewer for this careful and important point. We agree that this is a potential limitation of the approach and that it should be addressed more explicitly in the manuscript. Our assay begins with poly(A)-selected RNA and then enriches transcripts of interest through terminal-exon capture, so the molecules analyzed are completed, polyadenylated transcripts rather than nascent partial transcripts. This feature is important for reducing ambiguity arising from incomplete transcription, particularly in the chromatin-associated fraction. At the same time, we agree that for introns near the 5′ end, the assay may have limited power to distinguish very rapid splicing from moderately rapid splicing if excision is largely complete by the time the transcript is fully synthesized and polyadenylated.

In revision, we will address this concern directly in two ways. First, we will revise the Results and Discussion to clarify that the assay provides a population-level measure of splice completion in completed transcripts and that interpretation is strongest for introns whose excision is not already fully resolved before transcript completion. Second, we will more systematically evaluate whether apparent slow splicing covaries with transcript position, distance from the 3′ end, and intron length, and we will perform sensitivity analyses with and without the most 5′ introns to determine which conclusions are robust to these positional constraints. We will also examine transcript coverage patterns in greater detail to better assess the extent to which library construction and  cDNA generation may contribute to apparent positional bias. Our preliminary inspection suggests that transcript position is not the sole determinant of the observed heterogeneity, but we agree that a more explicit treatment of this issue is warranted in the revised manuscript.

(2) Request for more detailed discussion of alternative library-construction choices.

We appreciate this suggestion and agree that the revised manuscript would benefit from a fuller discussion of the strengths and limitations of the current enrichment strategy. We chose poly(A) selection followed by terminal-exon capture because this design enriches completed transcripts of interest and reduces ambiguity from nascent partial transcripts, which is particularly important in the chromatin-associated fraction. This approach also provides greater read depth over the selected inflammatory transcripts, enabling more informative intron-level comparisons within the targeted dataset. In revision, we will clarify this rationale more explicitly in the manuscript. We will also discuss the tradeoffs of this design relative to alternative exon-targeting strategies and how those alternatives might provide different, but complementary, views of splicing kinetics.

(3) Questions regarding biological replicates, error bars, and statistical analysis in Figure 1C and other plots.

We agree that the replicate structure and intended interpretation of these plots should be clarified more explicitly. In revision, we will revise the figure legends and Methods to distinguish panels that display a single bulk RNA-seq time course (for example, Figure 1C) from panels that summarize distributions across many introns (for example, Figure 2 and Supplementary Figure 6). We will also add statistical comparisons where they are most appropriate and informative, such as in sequence-feature comparisons like Supplementary Figure 4C, while making clear that some CoSI panels are intended as descriptive summaries of intron-level heterogeneity rather than replicate-based inferential plots.

(4) Concern that intron half-lives may be time-dependent during TNF induction, and that the logic of the actinomycin D measurements is therefore unclear.

We appreciate this point and agree that the manuscript should distinguish more clearly between two related but non-identical quantities: the CoSI trajectories observed during ongoing TNF induction, and the interruption-based half-life estimates derived from actinomycin D treatment. The actinomycin D experiments were performed using multiple post-treatment timepoints, but they were designed to estimate intron excision behavior after transcriptional arrest under a defined set of conditions, rather than to measure whether an individual intron’s effective splicing rate changes across all phases of the TNF response. We agree that these estimates should therefore be interpreted as constrained measurements under the assay conditions used, rather than as a complete time-resolved model of splicing kinetics during induction. In revision, we will clarify this point in the Results, Methods, and Discussion, and we will more explicitly acknowledge that effective splicing behavior could vary across the induction time course.

(5) Concern that the interpretation of Supplementary Figure 6 is unclear, particularly why delayed splicing in non-immediate groups appears to peak later rather than at the earliest time points.

We appreciate this point and agree that the current presentation of Supplementary Figure 6 does not explain this behavior clearly enough. Our interpretation is not that delayed splicing is the sole determinant of early versus later induction classes. Rather, the earliest time points reflect a combination of transcriptional induction timing and RNA processing state. In this framework, the dip in CoSI shortly after stimulation reflects the appearance of newly induced, incompletely spliced transcripts, and the later kinetic groups appear to recover from this dip more slowly than the immediate-early group. Thus, the strongest signal of delayed splicing may become most apparent only after sufficient transcript accumulation, rather than necessarily at the very earliest time point. In revision, we will revise the text to make this logic clearer and will consider a more intuitive visualization of these group-specific CoSI trajectories.

(6) Concern that the deep-learning setup does not make clear whether the model input and output are time-dependent.

We appreciate this concern and agree that the current manuscript does not explain the model setup clearly enough. Briefly, we will clarify the role of the three TNF timepoints in model training, including the fact that these outputs were modeled jointly and that time itself was not provided as an explicit input to the model. We will also revise the Results and Methods so that the scope and interpretation of the resulting analyses are more explicit.

Reviewer #3

We thank Reviewer #3 for the positive assessment of the targeted capture design, the evaluation of overall interest of the findings, and the improvements in the current version. We appreciate the reviewer’s view that the study is intriguing and that the manuscript has been strengthened in revision. We agree, however, that the manuscript should more clearly distinguish what is directly demonstrated from what remains mechanistically unresolved. We address the main concerns raised in the public review below.

(1) The study still does not fully resolve the downstream consequences of delayed splicing, including whether bottleneck introns lead primarily to delayed production of mature transcripts, reduced productive transcript output, or some combination of the two.

We agree with this assessment. The current data do not fully resolve whether delayed splicing primarily delays mature transcript production, reduces productive transcript output, or reflects some combination of the two. In revision, we will further moderate the framing of the downstream consequences of delayed splicing and will revise the Abstract, Results, and Discussion to make clear that the present data do not fully distinguish among delayed mature transcript production, reduced productive transcript output, or a combination of both. We will ensure that the manuscript consistently presents these possibilities as alternatives not fully resolved by the current data.

(2) The minigene reporter assays measure a steady-state level of the transcript and do not provide direct insight into kinetics.

We agree and will revise the manuscript to make this limitation explicit throughout. In particular, we will ensure that the reporter assays are described consistently as steady-state reporter assays that support an association between splice donor strength and altered reporter output, while avoiding stronger claims that they directly resolve endogenous splicing kinetics or downstream transcript fate.

(3) Given that the detailed analyses were performed on a selected subset of inflammation-induced transcripts, a broader evolutionary interpretation should be restrained.

We agree that the broader evolutionary and mechanistic framing should be more carefully defined. In revision, we will restrain these interpretations so that they remain closely aligned with the inflammation-focused and targeted-transcript scope of the current study, and we will moderate language that extends beyond what is directly supported by the present dataset.

Closing Remarks

We again thank the reviewers for their constructive comments. We believe that the planned revisions will strengthen the manuscript by clarifying the scope of the mechanistic conclusions, sharpening the interpretation of the experimental approaches, and more carefully defining the role of the computational analyses. We appreciate the opportunity to revise the work and to provide this provisional response to accompany the Reviewed Preprint.

eLife Assessment

In this important manuscript, Matsuda and colleagues present a model describing the regulation of tracheal tubulogenesis in Drosophila melanogaster embryos. The authors support this model using convincing approaches that combine novel experimental results with previously published work from their group. While some conclusions are consistent with earlier studies, the present manuscript introduces distinct molecular markers not previously reported, which reinforce the authors' prior findings. In addition, the manuscript analyses, using experimental strategies, the requirement of the Dpp and EGFR signalling pathways for the maintenance of trachealess (trh), one of the key transcription factors governing tracheal development.

Reviewer #1 (Public review):

Summary:

In this manuscript, Matsuda and collaborators present a model of how tracheal tubulogenesis is controlled in Drosophila embryos. Some of the results backing the model are new, but others are based on information already published by the authors. However, the results in this manuscript present different molecular markers not published before, which agree with previous conclusions. The manuscript also analyses the requirement of the dpp and EGFR signalling pathways for trachealess (trh) maintenance, one of the main tracheal transcription factors.

Strengths:

The two most interesting novel points of the manuscript are:

(1) Its contribution to the analysis of how the dpp and EGFR pathways contribute to the maintenance of trh expression.

(2) The experimental evidence showing that mechanical invagination is not a requirement for trh maintenance in the tracheal cells, an intriguing hypothesis previously suggested by (Kondo Hayashi 2019 eLife 8:e45145) that can now be discarded by the data presented in this work.

Weaknesses:

Because of the mixture of new and already published data, this manuscript can be considered as a review/experimental paper.

Already known data:<br /> - The results showing that hh and vvl drive tracheal invaginaton independently of trh are reported in Figure 5 of (Matsuda et al. 2015 eLife 4:e09646).<br /> - The results showing dpp requirement for trh maintenance are partially reported in Figure 6 of (Matsuda 2015 eLife 4:e09646).

Reviewer #2 (Public review):

Summary:

Matsuda et al. investigate the regulatory mechanisms controlling gene expression and morphogenesis in the Drosophila embryonic trachea. Building on previous findings that tracheal invagination can occur independently of trh, they identify extrinsic hh and intrinsic vvl as key regulators that cooperatively promote this process. The study also integrates major signaling pathways (Dpp/BMP and EGFR) in defining tracheal cell identity and demonstrates that Ras activation can upregulate trh. Overall, the work supports a model in which multiple transcription factors and signaling inputs coordinate airway progenitor specification.

Strengths:

This study uses genetic analysis of various mutants to dissect regulatory relationships underlying tracheal development. While the uncoupling of tracheal invagination from trh function has been previously recognized, this work advances the field by identifying hh and vvl as key regulators of invagination independent of trh. The study also integrates multiple signaling pathways, such as Dpp/BMP and EGFR, into a coherent framework for tracheal cell specification. In addition, the demonstration that Ras activation can upregulate trh provides a clear mechanistic link between RTK signaling and transcriptional regulation. Overall, the work offers important and broadly relevant insights into how gene expression and morphogenesis are coordinated during development.

Weaknesses:

Data presentation and clarity of interpretation could be improved. Many images primarily show lateral views of whole embryos, which can make it difficult to fully assess some phenotypes; higher-magnification or sectional views would enhance clarity. There are also some minor inconsistencies in the description of invagination phenotypes, particularly regarding whether all trh+ cells remain in a 2D plane versus indications of partial invagination in hh vvl double mutants blocking apoptosis, which would benefit from further clarification. Finally, some statements in the abstract, especially regarding the role of grn, are not directly supported by data in this study and could be better aligned with the scope of the presented results.

Author response:

Reviewer #1<br /> - The results showing that hh and vvl drive tracheal invaginaton independently of trh are reported in Figure 5 of (Matsuda et al. 2015 eLife 4:e09646).

Reviewer #2

Many images primarily show lateral views of whole embryos, which can make it difficult to fully assess some phenotypes; higher-magnification or sectional views would enhance clarity. There are also some minor inconsistencies in the description of invagination phenotypes, particularly regarding whether all trh+ cells remain in a 2D plane versus indications of partial invagination in hh vvl double mutants blocking apoptosis, which would benefit from further clarification.

The data in our previous eLife publication (DOI: 10.7554/eLife.09646)1 were mostly projection views. Therefore, it is hard to conclude if the airway progenitors of hh vvl double mutants failed to invaginate or they invaginated to form sacs. We will provide magnified views of the progenitor invagination in hh vvl double mutants and describe the degrees of their invagination phenotypes.

Reviewer #1

The results showing dpp requirement for trh maintenance are partially reported in Figure 6 of (Matsuda 2015 eLife 4:e09646).

Reviewer #2

Finally, some statements in the abstract, especially regarding the role of grn, are not directly supported by data in this study and could be better aligned with the scope of the presented results.

trh-lacZ (1-eve-1) has been used as the earliest and the strongest enhancer trap line to mark the airway primordia and the airway progenitors. Perdurance of beta-galactocidase proteins makes it difficult to conclude if the marker signals result from the active transcriptional state of the trh locus. In our previous eLife publication we showed that Trh proteins and trh_transcripts are not detectable in _H99 grn hh vvl quadruple mutants and in grn hh vvl triple mutants (Figure 5H and Figure 5-figure supplement 2A of DOI: 10.7554/eLife.09646, respectively)1, although trh-LacZ signals are detected in grn hh vvl triple mutants.

Similarly, although we previously showed trh-LacZ expression in dpp mutant combinations, Figure 2 in the current manuscript, shows that even strong trh-LacZ signals do not always correlate with trh transcripts in dpp mutants. Therefore, in the current manuscript we included the data of dpp-driven positive regulation of trh transcripts at later stages since they have not been shown before.

Assessments and advices of the Editors and the Reviewers are indispensable for improving the manuscript. We will address all the Reviewers comments (Weakness of Public review, major and minor issues of Recommendations for the authors) both experimentally and in the text.

Sincerely yours,

Christos Samakovlis on behalf of all authors

• (1) Matsuda, R., Hosono, C., Samakovlis, C. & Saigo, K. Multipotent versus differentiated cell fate selection in the developing Drosophila airways. eLife 4 (2015).

eLife Assessment

This important work advances our understanding of intraflagellar transport, ciliogenesis, and ciliary-based signaling, by identifying the interactions of IFT172 with IFT-A components, ubiquitin-binding, and ubiquitination, mediated by IFT172 C-terminus and its role in ciliogenesis and ciliary signaling. The evidence supporting the findings is convincing. This paper will be of interest to cell biologists and biochemists, especially those working on cilia and signaling.

Reviewer #1 (Public review):

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

Summary:

Zacharia and colleagues investigate the role of the C-terminus of IFT172 (IFT172c), a component of the IFT-B subcomplex. IFT172 is required for proper ciliary trafficking and mutations in its C-terminus are associated with skeletal ciliopathies. The authors begin by performing a pull-down to identify binding partners of His-tagged CrIFT172968-C in Chlamydomonas reinhardtii flagella. Interactions with three candidates (IFT140, IFT144, and a UBX-domain containing protein) are validated by AlphaFold Multimer with the IFT140 and IFT144 predictions in agreement with published cryo-ET structures of anterograde and retrograde IFT trains. They present a crystal structure of IFT172c and find that a part of the C-terminal domain of IFT172 resembles the fold of a non-canonical U-box domain. As U-box domains typically function to bind ubiquitin-loaded E2 enzymes, this discovery stimulates the authors to investigate the ubiquitin-binding and ubiquitination properties of IFT172c. Using in vitro ubiquitination assays with truncated IFT172c constructs, the authors demonstrate partial ubiquitination of IFT172c in the presence of the E2 enzyme UBCH5A. The authors also show a direct interaction of IFT172c with ubiquitin chains in vitro. Finally, the authors demonstrate that deletion of the U-box-like subdomain of IFT172 impairs ciliogenesis and TGFbeta signaling in RPE1 cells.

However, some of the conclusions of this paper are only partially supported by the data, and presented analyses are potentially governed by in vitro artifacts. In particular, the data supporting autoubiquitination and ubiquitin-binding are inconclusive. Without further evidence supporting a ubiquitin-binding role for the C-terminus, the title is potentially misleading.

Strengths:

(1) The pull-down with IFT172 C-terminus from C. reinhardtii cilia lysates is well performed and provides valuable insights into its potential roles.

(2) The crystal structure of the IFT172 C-terminus is of high quality.

(3) The presented AlphaFold-multimer predictions of IFT172c:IFT140 and IFT172c:IFT144 are convincing and agree with experimental cryo-ET data.

Reviewer #2 (Public review):

Summary:

Cilia are antenna-like extensions projecting from the surface of most vertebrate cells. Protein transport along the ciliary axoneme is enabled by motor protein complexes with multimeric so-called IFT-A and IFT-B complexes attached. While the components of these IFT complexes have been known for a while, precise interactions between different complex members, especially how IFT-A and IFT-B subcomplexes interact, are still not entirely clear. Likewise, the precise underlying molecular mechanism in human ciliopathies resulting from IFT dysfunction has remained elusive.

Here, the authors investigated the structure and putative function of the to-date poorly characterised C-terminus of IFT-B complex member IFT172 using alpha-fold predictions, crystallography and biochemical analyses including proteomics analyses followed by mass spectrometry, pull-down assays, and TGFbeta signalling analyses using chlamydomonas flagellae and RPE cells. The authors hereby provide novel insights into the crystal structure of IFT172 and identify novel interaction sites between IFT172 and the IFT-A complex members IFT140/IFT144. They suggest a U-box-like domain within the IFT172 C-terminus could play a role in IFT172 auto-ubiquitination as well as for TGFbeta signalling regulation.

As a number of disease-causing IFT72 sequence variants resulting in mammalian ciliopathy phenotypes in IFT172 have been previously identified in the IFT172 C-terminus, the authors also investigate the effects of such variants on auto-ubiquitination. This revealed no mutational effect on mono-ubiquitination which the authors suggest could be independent of the U-box-like domain but reduced overall IFT172 ubiquitination.

Strengths:

The manuscript is clear and well written and experimental data is of high quality. The findings provide novel insights into IFT172 function, IFT complex-A and B interactions, and they offer novel potential mechanisms that could contribute to the phenotypes associated with IFT172 C-terminal ciliopathy variants.

Reviewer #3 (Public review):

Summary:

Zacharia et al report on the molecular function of the C-terminal domain of the intraflagellar transport IFT-B complex component IFT172 by structure determination and biochemical in vitro and cell culture-based assays. The authors identify an IFT-A binding site that mediates a mutually exclusive interaction to two different IFT-A subunits, IFT144 and IFT140, consistent with interactions suggested in anterograde and retrograde IFT trains by previous cryo-electron tomography studies. Additionally, the authors identify a U-box-like domain that binds ubiquitin and conveys ubiquitin conjugation activity in the presence of the UbcH5a E2 enzyme in vitro. RPE1 cell lines that lack the U-box domain show a reduction in ciliation rate with shorter cilia, and heterozygous cells manifest TGF-beta signaling defects, suggesting an involvement of the U-box domain in cilium-dependent signaling.

Strengths:

(1) The structural analyses of the C-terminal domain of IFT172 combine crystallography with structure prediction using state-of-the-art algorithms, which gives high confidence in the presented protein structures. The structure-based predictions of protein interactions are validated by further biochemical experiments to assess the specific binding of the IFT172 C-terminal domains with other proteins.

(2) The finding that the IFT172 C-terminus interactions with the IFT-A components IFT140 and IFT144 appear mutually exclusive confirm a suggested role in mediating the binding of IFT-B to IFT-A in anterograde and retrograde IFT trains, which is of very high scientific value.

(3) The suggested molecular mechanism of IFT train coordination explains previous findings in Chlamydomonas IFT172 mutants, in particular an IFT172 mutant that appeared defective in retrograde IFT, as well as mutations identified in ciliopathy patients.

(4) The identification of other IFT172 interactors by unbiased mass spectrometry-based proteomics is very exciting. Analysis of stoichiometries between IFT components suggests that these interactors could be part of IFT trains, either as cargos or additional components that may fulfill interesting functions in cilia and flagella.

(5) The authors unexpectedly identify a U-box-like fold in the IFT172 C-terminus and thoroughly dissect it by sequence and mutational analyses to reveal unexpected ubiquitin binding and potential intrinsic ubiquitination activity.

(6) The overall data quality is very high. The use of IFT172 proteins from different organisms suggests a conserved function.

Overall, the authors achieved to characterize an understudied protein domain of the ciliary intraflagellar transport machinery and gained important molecular insights into its role in primary cilia biology, beyond IFT. By identifying an unexpected functional protein domain and novel interaction partners the work makes an important contribution to further our understanding of how ciliary processes might be regulated by ubiquitination on a molecular level. Based on this work it will be important for future studies in the cilia community to consider direct ubiquitin binding by IFT complexes.

Conceptually, the study highlights that protein transport complexes can exhibit additional intrinsic structural features for potential auto-regulatory processes. Moreover, the study adds to the functional diversity of small U-box and ubiquitin-binding domains, which will be of interest to a broader cell biology and structural biology audience.

Author response:

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

Public Reviews:

Reviewer #1 (Public review):

Summary:

Zacharia and colleagues investigate the role of the C-terminus of IFT172 (IFT172c), a component of the IFT-B subcomplex. IFT172 is required for proper ciliary trafficking and mutations in its C-terminus are associated with skeletal ciliopathies. The authors begin by performing a pull-down to identify binding partners of His-tagged CrIFT172968-C in Chlamydomonas reinhardtii flagella. Interactions with three candidates (IFT140, IFT144, and a UBX-domain containing protein) are validated by AlphaFold Multimer with the IFT140 and IFT144 predictions in agreement with published cryo-ET structures of anterograde and retrograde IFT trains. They present a crystal structure of IFT172c and find that a part of the C-terminal domain of IFT172 resembles the fold of a non-canonical U-box domain. As U-box domains typically function to bind ubiquitin-loaded E2 enzymes, this discovery stimulates the authors to investigate the ubiquitin-binding and ubiquitination properties of IFT172c. Using in vitro ubiquitination assays with truncated IFT172c constructs, the authors demonstrate partial ubiquitination of IFT172c in the presence of the E2 enzyme UBCH5A. The authors also show a direct interaction of IFT172c with ubiquitin chains in vitro. Finally, the authors demonstrate that deletion of the U-box-like subdomain of IFT172 impairs ciliogenesis and TGFbeta signaling in RPE1 cells.

However, some of the conclusions of this paper are only partially supported by the data, and presented analyses are potentially governed by in vitro artifacts. In particular, the data supporting autoubiquitination and ubiquitin-binding are inconclusive. Without further evidence supporting a ubiquitin-binding role for the C-terminus, the title is potentially misleading.

Strengths:

(1) The pull-down with IFT172 C-terminus from C. reinhardtii cilia lysates is well performed and provides valuable insights into its potential roles.

(2) The crystal structure of the IFT172 C-terminus is of high quality.

(3) The presented AlphaFold-multimer predictions of IFT172c:IFT140 and IFT172c:IFT144 are convincing and agree with experimental cryo-ET data.

Weaknesses:

(1) The crystal structure of HsIFT172c reveals a single globular domain formed by the last three TPR repeats and C-terminal residues of IFT172. However, the authors subdivide this globular domain into TPR, linker, and U-box-like regions that they treat as separate entities throughout the manuscript. This is potentially misleading as the U-box surface that is proposed to bind ubiquitin or E2 is not surface accessible but instead interacts with the TPR motifs. They justify this approach by speculating that the presented IFT172c structure represents an autoinhibited state and that the U-box-like domain can become accessible following phosphorylation. However, additional evidence supporting the proposed autoinhibited state and the potential accessibility of the U-box surface following phosphorylation is needed, as it is not tested or supported by the current data.

We thank the reviewer for this comment. IFT172C contains TPR region and Ubox-like region, which are admittedly tightly bound to each other. While there is a possibility that this region functions and exists as one domain, below are the reasons why we chose to classify these regions as two different domains.

(1) TPR and Ubox-like regions are two different structural classes

(2) TPR region is linked to Ubox-like region via a long linker which seems poised to regulate the relative movement between these regions.

(3) Many ciliopathy mutations are mapped to the interface of TPR region and the Ubox region hinting at a regulatory mechanism governed by this interface.

That said, we agree that the proposed autoinhibited state and its potential relief by phosphorylation remains a hypothesis that requires experimental validation. We have revised the manuscript to present this more clearly as a speculative model rather than an established mechanism. We clearly acknowledge this limitation on pg. 16-17 of the revised discussion: ‘The IFT172 U-box domain appears to be in an auto-inhibited state in our crystal structure of HsIFT172C2 (Fig. 2E), potentially explaining the absence of a robust auto-ubiquitination activity in-vitro. This structural inhibition is reminiscent of the RING ubiquitin ligase CBL [59], where phosphorylation and substrate binding trigger a conformational change that activates ligase activity [59,75]. Intriguingly, the phosphosite database [76] lists four residues (T1533, S1549, T1689, Y1691) at the U-box/TPR interface as phosphorylation sites (Fig. S2D). Phosphorylation of these residues could potentially alleviate the auto-inhibited state, suggesting a possible regulatory mechanism. Furthermore, a 30-residue linker connects the U-box domain to the last TPR of IFT172, likely providing significant conformational flexibility (Fig. 2A-B). This flexibility may be functionally crucial for the U-box domain, allowing it to adopt different conformations as needed for its various roles. However, we note that the proposed autoinhibition model and its potential regulation by phosphorylation remain hypothetical and require future experimental validation.

(2) While in vitro ubiquitination of IFT172 has been demonstrated, in vivo evidence of this process is necessary to support its physiological relevance.

We thank the reviewer for this important point. We agree that in vivo evidence of IFT172 ubiquitination would strengthen the physiological relevance of our findings. While our current study focuses on the in vitro characterization of this activity, we have revised the manuscript to more clearly state that demonstration of IFT172 ubiquitination activity in cells, including identification of bona fide substrates, is required to establish its physiological significance (p. 16). We consider this an important direction for future studies.

(3) The authors describe IFT172 as being autoubiquitinated. However, the identified E2 enzymes UBCH5A and UBCH5B can both function in E3-independent ubiquitination (as pointed out by the authors) and mediate ubiquitin chain formation in an E3-independent manner in vitro (see ubiquitin chain ladder formation in Figure 3A). In addition, point mutation of known E3-binding sites in UBCH5A or TPR/U-box interface residues in IFT172 has no effect on the mono-ubiquitination of IFT172c1. Together, these data suggest that IFT172 is an E3-independent substrate of UBCH5A in vitro. The authors should state this possibility more clearly and avoid terminology such as "autoubiquitination" as it implies that IFT172 is an E3 ligase, which is misleading. Similarly, statements on page 10 and elsewhere are not supported by the data (e.g. "the low in vitro ubiquitination activity exhibited by IFT172" and "ubiquitin conjugation occurring on HsIFT172C1 in the presence of UBCH5A, possibly in coordination with the IFT172 U-box domain").

We now consider this possibility and tone down our statements about the autoubiquitination activity of IFT172 in both the abstract and results/discussion parts of the revised version of the manuscript. We no longer refer to IFT172 as having auto-ubiquitination activity in the manuscript.

(4) Related to the above point, the conclusion on page 11, that mono-ubiquitination of IFT172 is U-box-independent while polyubiquitination of IFT172 is U-box-dependent appears implausible. The authors should consider that UBCH5A is known to form free ubiquitin chains in vitro and structural rearrangements in F1715A/C1725R variants could render additional ubiquitination sites or the monoubiquitinated form of IFT172 inaccessible/unfavorable for further processing by UBCH5A.

We agree and the conclusion on pg. 11 has now been changed to: Therefore, while mutations in the IFT172 U-box domain affect the formation of higher molecular weight ubiquitin conjugates, the prominent mono-ubiquitination of IFT172 is likely attributable to the E3-independent activity of UbcH5a, as this event is not impacted by these U-box mutations, rather than indicating an intrinsic auto-ubiquitination capacity of IFT172 itself.

(5) Identification of the specific ubiquitination site(s) within IFT172 would be valuable as it would allow targeted mutation to determine whether the ubiquitination of IFT172 is physiologically relevant. Ubiquitination of the C1 but not the C2 or C3 constructs suggests that the ubiquitination site is located in TPRs ranging from residues 969-1470. Could this region of TPR repeats (lacking the IFT172C3 part) suffice as a substrate for UBCH5A in ubiquitination assays?

We thank the reviewer for raising this important point about ubiquitination site identification. While not included in our manuscript, we did perform mass spectrometry analysis of ubiquitination sites using wild-type IFT172 and several mutants (P1725A, C1727R, and F1715A). As shown in Author response image 1, we detected multiple ubiquitination sites across these constructs. The wild-type protein showed ubiquitination at positions K1022, K1237, K1271, and K1551, while the mutants displayed slightly different patterns of modification. However, we should note that the MS intensity signals for these ubiquitinated peptides were relatively low compared to unmodified peptides, making it difficult to draw strong conclusions about site specificity or physiological relevance.

Author response image 1.

Consistent with the reviewer's suggestion, all detected ubiquitination sites fall within the TPR-containing region (residues 1022-1551), which is present in the C1 construct but absent from C2 and C3, explaining the construct-dependent ubiquitination pattern. We did not test the TPR region alone as a UBCH5A substrate, but this would be an informative experiment for future studies.

(6) The discrepancy between the molecular weight shifts observed in anti-ubiquitin Western blots and Coomassie-stained gels is noteworthy. The authors show the appearance of a mono-ubiquitinated protein of ~108 kDa in anti-ubiquitin Western blots. However, this molecular weight shift is not observed for total IFT172 in the corresponding Coomassie-stained gels (Figures 3B, D, F). Surprisingly, this MW shift is visible in an anti-His Western blot of a ubiquitination assay (Fig 3C). Together, this raises the concern that only a small fraction of IFT172 is being modified with ubiquitin. Quantification of the percentage of ubiquitinated IFT172 in the in vitro experiments could provide helpful context.

We acknowledge that the ubiquitin conjugation of IFT172 in vitro is weak, as stated in the manuscript (p. 16). The discrepancy between anti-ubiquitin Western blots and Coomassie-stained gels is consistent with only a small fraction of IFT172 being modified, which is expected given that the reaction likely reflects E3-independent ubiquitination by UBCH5A rather than a robust enzymatic activity of IFT172 itself. The anti-His Western blot (Fig. 3C) is more sensitive than Coomassie staining, explaining why the shift is visible there but not on Coomassie. We have not performed formal quantification of the ubiquitinated fraction, but based on the Coomassie data, we estimate it to be a minor proportion of total IFT172, consistent with the toned-down conclusions in our revised manuscript. The identification of physiological substrates and in vivo validation will be important future directions to establish the biological relevance of these observations.

(7) The authors propose that IFT172 binds ubiquitin and demonstrate that GST-tagged HsIFT172C2 or HsIFT172C3 can pull down tetra-ubiquitin chains. However, ubiquitin is known to be "sticky" and to have a tendency for weak, nonspecific interactions with exposed hydrophobic surfaces. Given that only a small proportion of the ubiquitin chains bind in the pull-down, specific point mutations that identify the ubiquitin-binding site are required to convincingly show the ubiquitin binding of IFT172.

We appreciate the reviewer's point regarding the potential for non-specific ubiquitin interactions and the value of mutational analysis for confirming specificity. While further mutagenesis of the predicted ubiquitin-binding interface was not performed for this revision, we note that our data show comparable tetra-ubiquitin pull-down by both the larger HsIFT172C2 construct and, importantly, the isolated HsIFT172C3 U-box domain itself (Fig. 4D). This localization of binding to the smaller U-box domain, coupled with our AlphaFold model predicting a specific interface with ubiquitin (Fig. 4E-F) and the observation that a mutation elsewhere (D1605R, Fig. 4C) does not abrogate this binding, collectively suggest a degree of specificity. We have revised the manuscript to more cautiously present these findings and acknowledge the need for future studies to definitively map the binding site. Specifically, we have now toned down the conclusion in the section on pg. 12-13 of the revised manuscript including a toned down heading: “IFT172 U-box domain pulls down ubiquitin in vitro”.

(8) The authors generated structure-guided mutations based on the predicted Ub-interface and on the TPR/U-box interface and used these for the ubiquitination assays in Fig 3. These same mutations could provide valuable insights into ubiquitin binding assays as they may disrupt or enhance ubiquitin binding (by relieving "autoinhibition"), respectively. Surprisingly, two of these sites are highlighted in the predicted ubiquitin-binding interface (F1715, I1688; Figure 4E) but not analyzed in the accompanying ubiquitin-binding assays in Figure 4.

We thank the reviewer for emphasizing the importance of mutational analysis to confirm the specificity of ubiquitin binding and for specifically inquiring about residues like F1715 and I1688 at the predicted ubiquitin interface. We tested purified HsIFT172C1 constructs containing the F1715A mutation (along with P1725A and C1727R variants) in pull-down assays with GST-Ubiquitin, see Author response image 2.

Author response image 2.

However, these experiments did not reveal a conclusive difference in ubiquitin binding for any of the tested variants compared to wild-type IFT172. The I1688A mutant, unfortunately, yielded insoluble protein and could not be evaluated. It is conceivable that the F1715A mutation was not disruptive enough to significantly alter binding, and future studies with different substitutions might be more informative. Nevertheless, our observations that the isolated HsIFT172C3 U-box domain itself pulls down tetra-ubiquitin (Fig. 4D), that our AlphaFold model predicts a specific interface (Fig. 4E-F), and that a mutation elsewhere (D1605R, Fig. 4C) does not abrogate this binding, collectively suggest a degree of specificity. We have revised the manuscript to present these ubiquitin binding findings cautiously, acknowledging the need for further investigation to definitively map the binding site and its functional relevance.

(9) If IFT172 is a ubiquitin-binding protein, it might be expected that the pull-down experiments in Figure S1 would identify ubiquitin, ubiquitinated proteins, or E2 enzymes. These were not observed, raising doubt that IFT172 is a ubiquitin-binding protein.

We acknowledge that the absence of ubiquitin or ubiquitinated proteins in our pull-down/MS experiment (Fig. S1) could raise questions about the ubiquitin-binding capacity of IFT172. However, several technical factors likely explain this. First, IFT172 appears to bind ubiquitin with low affinity, as indicated by our in vitro pull-downs and the AF-predicted interface. Second, we used extensive washes to remove non-specific interactors, which would also remove weak but potentially genuine ubiquitin interactions. Third, we did not include ubiquitination-preserving reagents such as NEM in our pull-down buffers, exposing ubiquitinated proteins to DUB-mediated deubiquitination during the experiment. These factors combined would strongly select against the detection of ubiquitin-related interactors under our experimental conditions.

(10) The cell-based experiments demonstrate that the U-box-like region is important for the stability of IFT172 but does not demonstrate that the effect on the TGFb pathway is due to the loss of ubiquitin-binding or ubiquitination activity of IFT172.

We acknowledge that our current data cannot definitively distinguish whether the TGFβ pathway defects arise from reduced IFT172 protein stability or from specific loss of ubiquitin-related functions of the U-box domain. Our experiments demonstrate that the U-box region is required for both IFT172 stability and proper TGFβ signaling, but we agree that establishing a direct mechanistic link between ubiquitin-binding/conjugation and signaling would require additional experiments such as point mutations that selectively disrupt ubiquitin-related activity without affecting protein stability. We have revised the discussion (p. 18-19) to more clearly acknowledge this limitation. Addition to text: “However, we note that our current experiments cannot distinguish whether these signaling effects result specifically from loss of ubiquitin-related functions of the U-box domain or from the reduced levels of functional IFT172 protein in the heterozygous U-box deleted cells. Targeted point mutations that selectively disrupt ubiquitin binding without affecting protein stability would be required to resolve this question.”

(11) The challenges in experimentally validating the interaction between IFT172 and the UBX-domain-containing protein are understandable. Alternative approaches, such as using single domains from the UBX protein, implementing solubilizing tags, or disrupting the predicted binding interface in Chlamydomonas flagella pull-downs, could be considered. In this context, the conclusion on page 7 that "The uncharacterized UBX-domain-containing protein was validated by AF-M as a direct IFT172 interactor" is incorrect as a prediction of an interaction interface with AF-M does not validate a direct interaction per se.

We agree with the reviewer that our AlphaFold-Multimer (AF-M) predictions alone do not constitute experimental validation of a direct interaction. We appreciate the reviewer's understanding of the technical challenges in validating this interaction experimentally. We have revised our text (p. 7) to state that "The uncharacterized UBX-domain-containing protein was predicted by AF-M as a potential direct IFT172 interactor" and discuss the AF-M predictions as computational evidence that suggests, but does not prove, a direct interaction.

Reviewer #2 (Public review):

Summary:

Cilia are antenna-like extensions projecting from the surface of most vertebrate cells. Protein transport along the ciliary axoneme is enabled by motor protein complexes with multimeric so-called IFT-A and IFT-B complexes attached. While the components of these IFT complexes have been known for a while, precise interactions between different complex members, especially how IFT-A and IFT-B subcomplexes interact, are still not entirely clear. Likewise, the precise underlying molecular mechanism in human ciliopathies resulting from IFT dysfunction has remained elusive.

Here, the authors investigated the structure and putative function of the to-date poorly characterised C-terminus of IFT-B complex member IFT172 using alpha-fold predictions, crystallography and biochemical analyses including proteomics analyses followed by mass spectrometry, pull-down assays, and TGFbeta signalling analyses using chlamydomonas flagellae and RPE cells. The authors hereby provide novel insights into the crystal structure of IFT172 and identify novel interaction sites between IFT172 and the IFT-A complex members IFT140/IFT144. They suggest a U-box-like domain within the IFT172 C-terminus could play a role in IFT172 auto-ubiquitination as well as for TGFbeta signalling regulation.

As a number of disease-causing IFT72 sequence variants resulting in mammalian ciliopathy phenotypes in IFT172 have been previously identified in the IFT172 C-terminus, the authors also investigate the effects of such variants on auto-ubiquitination. This revealed no mutational effect on mono-ubiquitination which the authors suggest could be independent of the U-box-like domain but reduced overall IFT172 ubiquitination.

Strengths:

The manuscript is clear and well written and experimental data is of high quality. The findings provide novel insights into IFT172 function, IFT complex-A and B interactions, and they offer novel potential mechanisms that could contribute to the phenotypes associated with IFT172 C-terminal ciliopathy variants.

Weaknesses:

Some suggestions/questions are included in the comments to the authors below.

Reviewer #3 (Public review):

Summary:

Zacharia et al report on the molecular function of the C-terminal domain of the intraflagellar transport IFT-B complex component IFT172 by structure determination and biochemical in vitro and cell culture-based assays. The authors identify an IFT-A binding site that mediates a mutually exclusive interaction to two different IFT-A subunits, IFT144 and IFT140, consistent with interactions suggested in anterograde and retrograde IFT trains by previous cryo-electron tomography studies. Additionally, the authors identify a U-box-like domain that binds ubiquitin and conveys ubiquitin conjugation activity in the presence of the UbcH5a E2 enzyme in vitro. RPE1 cell lines that lack the U-box domain show a reduction in ciliation rate with shorter cilia, and heterozygous cells manifest TGF-beta signaling defects, suggesting an involvement of the U-box domain in cilium-dependent signaling.

Strengths:

(1) The structural analyses of the C-terminal domain of IFT172 combine crystallography with structure prediction using state-of-the-art algorithms, which gives high confidence in the presented protein structures. The structure-based predictions of protein interactions are validated by further biochemical experiments to assess the specific binding of the IFT172 C-terminal domains with other proteins.

(2) The finding that the IFT172 C-terminus interactions with the IFT-A components IFT140 and IFT144 appear mutually exclusive confirm a suggested role in mediating the binding of IFT-B to IFT-A in anterograde and retrograde IFT trains, which is of very high scientific value.

(3) The suggested molecular mechanism of IFT train coordination explains previous findings in Chlamydomonas IFT172 mutants, in particular an IFT172 mutant that appeared defective in retrograde IFT, as well as mutations identified in ciliopathy patients.

(4) The identification of other IFT172 interactors by unbiased mass spectrometry-based proteomics is very exciting. Analysis of stoichiometries between IFT components suggests that these interactors could be part of IFT trains, either as cargos or additional components that may fulfill interesting functions in cilia and flagella.

(5) The authors unexpectedly identify a U-box-like fold in the IFT172 C-terminus and thoroughly dissect it by sequence and mutational analyses to reveal unexpected ubiquitin binding and potential intrinsic ubiquitination activity.

(6) The overall data quality is very high. The use of IFT172 proteins from different organisms suggests a conserved function.

Weaknesses:

(1) Interaction studies were carried out by pulldown experiments, which identified more IFT172 interaction partners. Whether these interactions can be seen in living cells remains to be elucidated in subsequent studies.

We agree with the reviewer that validation of protein-protein interactions in living cells provides important physiological context. While our pulldown experiments have identified several promising interaction partners and the AF-M predictions provide computational support for these interactions, we acknowledge that demonstrating these interactions in vivo would strengthen our findings. However, we believe our current biochemical and structural analyses provide valuable insights into the molecular basis of IFT172's interactions, laying important groundwork for future cell-based studies.

(2) The cell culture-based experiments in the IFT172 mutants are exciting and show that the U-box domain is important for protein stability and point towards involvement of the U-box domain in cellular signaling processes. However, the characterization of the generated cell lines falls behind the very rigorous analysis of other aspects of this work.

We thank the reviewer for noting that the characterization of our cell lines could be more rigorous. In the revised version of the manuscript, we have addressed this by providing additional validation data for all four engineered RPE1 cell lines. First, we performed Sanger sequencing to confirm precise in-frame integration of the GFP tag at the targeted loci and to exclude unintended insertions or deletions (indels), both for the full-length IFT172-eGFP lines (Fig. S6) and for the IFT172∆U-box-eGFP lines (Fig. S7). Second, we performed anti-IFT172 immunoblotting on all four cell lines alongside parental RPE1 cells, confirming expression of both the full-length and U-box-truncated IFT172 proteins (Fig. S8). Notably, the immunoblot revealed reduced steady-state levels of the IFT172∆U-box protein compared to full-length IFT172, providing direct biochemical evidence that loss of the U-box domain compromises IFT172 protein stability consistent with the ciliogenesis phenotype described in the main text. Together, these data verify the integrity of the edited loci at both the genomic and protein levels, and strengthen the validation of the cellular models used in this study.

Overall, the authors achieved to characterize an understudied protein domain of the ciliary intraflagellar transport machinery and gained important molecular insights into its role in primary cilia biology, beyond IFT. By identifying an unexpected functional protein domain and novel interaction partners the work makes an important contribution to further our understanding of how ciliary processes might be regulated by ubiquitination on a molecular level. Based on this work it will be important for future studies in the cilia community to consider direct ubiquitin binding by IFT complexes.

Conceptually, the study highlights that protein transport complexes can exhibit additional intrinsic structural features for potential auto-regulatory processes. Moreover, the study adds to the functional diversity of small U-box and ubiquitin-binding domains, which will be of interest to a broader cell biology and structural biology audience.

Additional comments:

The authors investigate the consequences of the U-box deletion on ciliary TGF-beta signaling. While a cilium-dependent effect of TGF-beta signaling on the phosphorylation of SMAD2 has been demonstrated, the precise function of cilia in AKT signaling has not been fully established in the field. Therefore, the relevance of this finding is somewhat unclear. It may help to discuss relevant literature on the topic, such as Shim et al., PNAS, 2020.

We appreciate the reviewer's comment highlighting that the role of primary cilia in AKT signaling is not as well established as for SMAD2/3. However, we note that a direct functional link between AKT signaling and ciliogenesis has been demonstrated, showing that AKT regulates ciliogenesis initiation through a Rab11-effector switch mechanism (Walia et al., 2019; PMID: 31204173, co-authored by the corresponding author of this study). Furthermore, Shim et al. (PMID: 33753495) demonstrated a cilia-dependent reciprocal activation of AKT1 and SMAD2/3. In the revised manuscript (p. 19, ref. 97), we have expanded the discussion to cite these studies and provide a clearer literature context for the cilia-AKT connection, while acknowledging that the precise mechanism by which the IFT172 U-box domain influences AKT activation requires further investigation.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Points for the discussion:

(1) The discussion should mention that IFT-A subunits IFT121, IFT122 and IFT144 share a similar domain organization to IFT172 (TPRs terminating in Zn-finger-like domains). Do the authors consider these as potential ubiquitin-binding proteins with E3 ligase activity? The possibility that these Zn-finger-like regions share a common origin, and function to stabilize the proteins or mediate IFT subunit interactions without a role in ubiquitin biology should be considered.

We appreciate this important point. We agree that the shared domain architecture across IFT121, IFT122, IFT144, and IFT172 raises the question of whether these C-terminal domains primarily serve structural rather than ubiquitin-related roles. We have added a discussion paragraph (p. 16) acknowledging that a structural/stabilizing function is the more parsimonious explanation, while noting that whether IFT172's U-box-like domain has additionally acquired ubiquitin-related activity remains an open question.

(2) From their modeling data, do the authors have an explanation for why a substitution as conservative as D1605E would cause disease?

The D1605E substitution maps to the IFT172-IFT-A interaction interface (Fig. 1F). While this is a conservative change, D1605 is located at a tightly packed protein-protein interface where even the addition of a single methylene group (the difference between aspartate and glutamate) could introduce steric clashes with residues of IFT140 or IFT144, or alter the precise geometry of hydrogen bonds or salt bridges critical for the interaction. Unfortunately, this level of detail is beyond the resolution of AlphaFold models. However, the fact that this residue is positioned directly at the binding interface provides a plausible structural rationale for its pathogenicity.

(3) The authors speculate that the L1615P mutation in the Chlamydomonas fla11 strain causes a faulty switch to retrograde IFT and this provides a molecular basis for the retrograde IFT phenotype. However, because the mutation is also within the IFT144 binding site, why is anterograde IFT also not affected?

The fla11 L1615P mutation resides in helix αA, which participates in both IFT144 (anterograde) and IFT140 (retrograde) interactions. The predominantly retrograde phenotype can be rationalized by the fundamentally different structural roles of the IFT172 C-terminus in anterograde versus retrograde trains. In anterograde trains, the IFT172 C-terminus acts as a flexible tether in stoichiometric excess (2:1 IFT-B:IFT-A ratio), providing an avidity effect that likely compensates for reduced binding affinity caused by L1615P (Lacey et al., 2023). Additional lateral interactions between IFT-B subunits further stabilize the anterograde polymer independently of the IFT172-IFT144 link. In contrast, the retrograde train requires the IFT172 C-terminus to adopt a rigid, resolved conformation that is integral to the IFT-A dimeric interface, with no redundant lateral interactions to compensate (Lacey et al., 2024). The helix-breaking L1615P mutation would specifically disrupt this precise structural requirement, explaining the selective retrograde IFT defect in fla11. We have added this discussion to the revised manuscript (p. 16).

Minor:

(1) On page 5, the authors describe the fla11 phenotypes including accumulation of IFT particles at the tip and accumulation of ubiquitinated proteins in the cilium. Could the authors please expand on how this suggests that IFT172 could be involved in ciliary ubiquitination events and discuss an alternative scenario of impaired assembly of functional retrograde IFT in this strain leading to accumulation of ubiquitinated proteins?

In the revised manuscript (p. 16), we have expanded the discussion of the fla11 phenotype to address this point. We now discuss how the distinct structural roles of the IFT172 C-terminus in anterograde versus retrograde trains explain the selective retrograde IFT defect in fla11, and explicitly note that the accumulation of ubiquitinated proteins in fla11 cilia may reflect impaired retrograde IFT-mediated clearance rather than a direct role of IFT172 in ciliary ubiquitination.

(2) The authors should also expand on the literature of known UBX-IFT interactions in their manuscript (e.g. Raman et al. PMID 26389662).

We have expanded the discussion of UBX-IFT interactions in the revised manuscript (p. 7) by citing the work of Raman et al. (PMID 26389662), who identified a direct interaction between the UBX-domain protein UBXN10 and IFT-B via CLUAP1/IFT38 for VCP-mediated regulation of IFT complex integrity. This provides important context for our identification of a UBX-domain protein as an IFT172 interactor.

(3) On page 11, I1688 is incorrectly referred to as I688.

Fixed.

Reviewer #2 (Recommendations for the authors):

(1) The finding that the interaction with IFT140/144 is mutually exclusive is very interesting. Could you speculate on or do you have any data regarding the effects to the overall IFT-complex conformation and downstream biological effects depending on which partner is bound?

I am not a structural biologist so this may be an irrelevant/impossible-to-answer question: I was also wondering as Ref 46 has shown that the dynein-2 motor complex binds to the edge of IFT-B2 (for assembled trains): Could the IFT172 C-terminus be involved here or somehow influence this interaction? In your mass spec data from Cr cilia using CrIFT172_968-C you don`t mention pulling down dynein-2 components so there doesn`t seem to be a direct interaction, but could the IFT-B2 conformation depend on if IFT172 has bound IFT-140 or IFT144 and hence this interaction influence the dynein-2 binding?

We thank the reviewer for this insightful question. Based on recent cryo-ET structures of anterograde and retrograde IFT trains (Lacey et al., 2023; 2024), the switch from IFT144 to IFT140 binding fundamentally changes IFT172's structural role. In anterograde trains, the IFT172 C-terminus acts as a flexible tether tolerating the 2:1 IFT-B:IFT-A stoichiometry and permitting long polymer formation. In retrograde trains, it adopts a rigid conformation integral to the IFT-A dimeric interface, driving the formation of discrete retrograde units with distinct architecture.

Regarding Dynein-2: while IFT172 does not directly bind Dynein-2 (consistent with our MS data), the reviewer's intuition is correct that IFT172's binding partner influences Dynein-2 association. In anterograde trains, autoinhibited Dynein-2 binds a composite surface formed between adjacent IFT-B2 repeats. When IFT172 switches to IFT140 at the ciliary tip, the resulting train depolymerization destroys this composite binding site, releasing Dynein-2 from its cargo mode to function as an active retrograde motor. The IFT172 binding switch may thus indirectly acts as a structural checkpoint for Dynein-2 activation.

(2) The data provided regarding TGFbeta signalling effects in cells with heterozygous U-box-like domain deletions is interesting. While secondary effects of impaired ciliogenesis due to homozygous deletion of the U-box-like domain can cause difficulties to analysing cell signalling effects, it would still be interesting to check the effects of bi-allelic human IFT172 disease variants in this region as well (the human disease phenotype is recessive and human mutations are likely hypomorphic variants still allowing for ciliogenesis).

Also, while there may be secondary effects, it would still be interesting to check homozygous U-box deleted cells as an aggravated effect would further support the data from the het cells.

We agree that testing bi-allelic human disease variants would strengthen the physiological relevance of our findings. While generating knock-in RPE1 lines was beyond the scope of this revision, we have obtained preliminary data from patient-derived fibroblasts carrying bi-allelic IFT172 missense variants in the U-box region (NPH2161). TGF-β1 stimulation time courses in these fibroblasts show altered p-SMAD2 kinetics compared to control fibroblasts, consistent with the phenotype observed in our heterozygous U-box deleted RPE1 cells (see Author response image 3).

Author response image 3.

While these results are preliminary and require further replication, they support the involvement of the IFT172 U-box domain in TGF-β signaling regulation in a disease-relevant context. Regarding homozygous U-box deleted cells, the severe reduction in IFT172 protein levels and ciliogenesis defects (Fig. 5B,D) make it difficult to separate U-box-specific effects from secondary consequences of impaired cilia formation, as the reviewer notes. We consider this an important direction for future studies using targeted point mutations rather than domain deletions.

(3) Figure 5 E-G: Overall, the effects upon TGFB1 addition are rather small compared to previously published data eg Clement et al Cell reports 2013 where one of the authors is the senior. Are RPE cells less responsive or do you have another theory? Did you check TGFB receptor levels to ensure the differences are not due to different levels of receptor expression? I feel it could be interesting to also check ciliary phopsho-SMAD localisation by IF. In Clement et al, loss of IFT88 results in reduced phospho-SMAD2 levels, do you have any theory why these opposite effects compared to the IFT172 loss of function could occur?

We thank the reviewer for this insightful comment. The Tg737orpk fibroblasts used in Clement et al. (2013), which harbor a hypomorphic mutation in IFT88, exhibit severely stunted cilia. This defect broadly disrupts cilium-dependent signaling pathways, including R-SMAD activation, and is therefore expected to produce more pronounced signaling phenotypes. In contrast, our study utilizes RPE-1 cells with structurally intact cilia, enabling us to investigate more specific alterations in ciliary signaling associated with IFT172 function rather than the global effects of cilia loss. Consequently, the more modest effects observed in our system are consistent with the less severe structural and functional perturbation. Both fibroblasts and RPE-1 cells are known to express TGF-β receptors and to respond robustly to TGF-β stimulation, making it unlikely that differences in receptor abundance alone account for the observed discrepancies. We also note that increasing evidence supports a role for the primary cilium in fine-tuning TGF-β signaling output by coordinating both canonical (R-SMAD-mediated) and non-canonical (e.g., AKT/ERK-mediated) pathways. Our data raise the possibility that loss of the IFT172 U-box domain, or reduced IFT172 levels, may differentially affect this balance, rather than simply attenuating signaling uniformly, as seen with more severe ciliary defects such as IFT88 disruption in Tg737orpk cells. We agree that the current dataset does not fully resolve the underlying mechanism. We therefore consider it an important direction for future work to examine, in greater detail, the localization and phosphorylation status of key canonical and non-canonical signaling components in context of the primary cilium by IF analyses.

(4) In the summary conclusion at the end of the discussions, the authors propose that IFT72 could directly influence the fate of ubiquitinated TGFB receptors. Do you have any data supporting the theory that TGFB ubiquitination is influenced by IFT172 ?

We acknowledge that our current data are insufficient to establish a direct link between IFT172-dependent ubiquitination events and TGF-β receptor regulation. Accordingly, we have revised the Discussion (page 19) to remove our previous hypothesis proposing a role for IFT172 in modulating TGF-β receptor ubiquitination.

While our experiments demonstrate that the U-box region is required for both IFT172 stability and proper TGF-β signaling, we agree that establishing a direct mechanistic connection between ubiquitin-related activity of IFT172 and signaling outcomes would require additional approaches such as targeted point mutations that selectively disrupt ubiquitin-binding or conjugation functions.

Furthermore, we note that our current data do not allow us to distinguish whether the observed signaling phenotypes arise specifically from the loss of ubiquitin-related functions of the U-box domain or from reduced levels of functional IFT172 protein in the heterozygous U-box–deleted cells.

(5) Wording:

Abstract

"IFT72..is associated with several disease variants causing ciliopathies". I would change this to "..and several disease-causing IFT172 variants have been identified in ciliopathy patients".

Corrected.

Introduction

"Another cohort of patients with milder ciliopathy resembling BBS also presented with ...". I would reword this to "Another cohort of patients with phenotypically slightly different ciliopathy features resembling BBS also presented with ...". It`s not necessarily less severe (they may die of cardiovascular complications in their early thirties for example due to metabolic syndrome, they are intellectually impaired, become blind...), but rather different.

Changed according to the reviewer’s recommendations.

Reviewer #3 (Recommendations for the authors):

(1) Recommended modifications:

(a) The RPE lines generated should be described better, i.e. sequencing information should be provided, or some kind of evidence that the lines are what they are supposed to be.

As also noted above, we acknowledge that the characterization presented for the RPE cell lines was insufficient in the initial version of the manuscript. In the revised version, we have addressed this limitation by including detailed sequencing analyses to validate the modifications introduced. Specifically, we provide sequencing data confirming both the integration of the GFP tag and the successful deletion of the U-box domain in all four engineered RPE cell lines. These data verify the integrity of the edited loci and exclude the presence of unintended insertions or deletions at the targeted regions. The corresponding results are presented in Figures S6 and S7 of the revised manuscript, thereby strengthening the validation of the cellular models used in this study.

(b) It would be more convincing if more than one clone of the RPE lines were presented, as this could rule out possible clonal effects.

We acknowledge that only a single clone was characterized for each of the four genotypes (IFT172-FL homozygous, IFT172-FL heterozygous, IFT172∆U-box homozygous, IFT172∆U-box heterozygous), and we agree that independent clones would provide stronger protection against clonal artifacts. Generating and validating additional clones was not feasible within the scope of this revision. However, several features of our data mitigate this concern. First, the phenotypes scale with allele dosage: the homozygous ∆U-box line shows the strongest reduction in IFT172 protein level, ciliation, and cilium length, while the heterozygous line shows intermediate defects (Fig. 5B, D and Fig. S8). A clonal off-target effect would not be expected to produce this dose-dependent pattern across two independently isolated lines. Second, the reduced steady-state IFT172 level in the ∆U-box lines (Fig. S8) is consistent with our in vitro observation that the U-box/TPR interface is required for protein stability, providing an independent biochemical rationale for the cellular phenotype. Third, Sanger sequencing of all four lines confirmed precise in-frame integration with no indels at the targeted locus (Figs. S6, S7). We have added a sentence to the Discussion (p. 20) acknowledging that confirmation in additional independent clones remains an important goal for follow-up work.

(c) Figure 5C: distribution of the GFP-tagged IFT172∆U-box protein could be quantified to support the statement.

In the revised version of the manuscript, we have included additional quantification of GFP fluorescence across all four cell lines to support our conclusions regarding IFT172 ciliary localization. The corresponding data for each cell line are presented in Figure S5C–F.

(d) The final sentences include quite bold statements about a general function of IFT172 in signal regulation. Yet, the evidence is the weakest part of the work. It is only shown in i) one cell line, ii) in one cell clone that is not extensively characterized, and iii) for one signaling pathway that is not the best-studied cilia signaling pathway. Therefore, I recommend a more moderate statement.

Abstract last sentence has now been toned down and reads: Our findings suggest that IFT172, beyond its structural role in bridging IFT-A and IFT-B complexes within IFT trains, harbors a conserved U-box-like domain with potential involvement in ciliary ubiquitination processes and signaling, providing new insights into the molecular mechanisms underlying IFT172-related ciliopathies.

(e) The order of the figures is not followed in the main text, which is distracting.

The order of figures is now consecutive in the revised manuscript.

(2) Questions and comments to consider:

(a) It is unclear why tetra-ubiquitin chains have been used.

We thank the reviewer for this question. Recent evidence suggests that ubiquitin chains, rather than monomeric ubiquitin, act as sorting and signaling cues at the primary cilium (Shinde et al., 2020). To probe the ubiquitin-binding activity of IFT172, we therefore used a tetrameric ubiquitin chain as a model substrate, which better reflects the multivalent nature and binding avidity expected for physiological polyubiquitin signals than a ubiquitin monomer. Specifically, we used a recombinantly expressed linear (Met1-linked) tetra-ubiquitin chain, generated as a genetically encoded fusion. Linear ubiquitin chains are well-established non-degradative signaling chains recognized by a dedicated class of ubiquitin-binding domains, making them a suitable probe for detecting ubiquitin-binding activity outside the canonical proteasomal pathway. In addition, monomeric ubiquitin (~8 kDa) is poorly retained during membrane transfer in Western blotting, which further precluded its reliable use as a probe in our pull-down assays. Together, these considerations motivated the use of tetrameric ubiquitin as a biologically and technically appropriate substrate for assessing IFT172's ubiquitin-binding activity.

(b) Figure 4D: described in the text as "pulldown tetraubiquitin at comparable levels", which is not obvious from the figure presented, it appears reduced by at least 30%.

We thank the reviewer for this observation. As described on page 10 of the manuscript and evident from Figure 4D, the purified GST–HsIFT172C3 construct underwent substantial proteolytic cleavage during purification. This degradation limited our ability to include amounts of intact GST–HsIFT172C3 comparable to those of the full-length GST–HsIFT172C2 construct in the pull-down assays. Importantly, when accounting for the reduced proportion of full-length GST–HsIFT172C3 present in the assay, the observed differences in tetra-ubiquitin pull-down efficiency between the two constructs are expected to be comparable. This is supported by the Coomassie staining shown in Figure 4D, which reflects the relative abundance of the intact protein species used in the experiment.

(c) With the proposed model, why would the fla11 mutant only affect retrograde IFT?

We have revised our manuscript in page 16 of the discussion section providing a plausible explanation of why only retrograde IFT is affected in the fla11 mutant.

(3) Minor copy-editing:

(a) Page 3, first paragraph: led := leads.

(b) Kinesin-2 and Dynein-2 should be hyphenated.

(c) Page 4: wwp1 should be WWP1.

(d) Bonafide should be italicized: bona fide.

(e) Some abbreviations appear uncommon and therefore somewhat distracting: TGFB instead of TGF-beta, Cr in instances where specifically referred to the organism.

(f) Unprecise lab jargon: "very C-terminal".

(g) Lab jargon: "purified a C-terminal construct".

(h) Lab jargon: "pull-downs".

(i) Page 8: "DALI" only abbreviated.

(j) Page 9: "Appearance ... were observed" should be "was".

(k) Page 11: "I688" should be "I1688".

(l) Page 12: "PDs" unclear.

These minor points have been corrected.

We have revised the text and figures to ensure using the widely accepted nomenclature, using TGF-β to refer to the signaling pathway and TGF-β1 specifically when referring to the ligand.

We further revised the text to reflect the use “Chlamydomonas reinhardtii” in instances when referring to the organism and “Cr” when referring to the protein.

We have removed the informal phrases "very C-terminal" and "purified a C-terminal construct" from the revised manuscript. We have retained the term "pull-down," as this is well-established and widely used terminology in the biochemistry literature to describe the affinity-based co-isolation assays used here. PD has been replaced with pull-down.

The grammatical error on page 9 ("Appearance... “were observed") has been corrected to "was observed”.

eLife Assessment

This valuable study reports the architectural reorganization of the uterine luminal epithelium during the implantation period. The data presented are solid, although improvements are needed. This work is of interest to reproductive biologists and physicians practicing reproductive medicine.

Reviewer #1 (Public review):

Summary:

This manuscript asks how the uterine lumen is remodeled across the peri-implantation window and whether this remodeling is functionally linked to embryo attachment and subsequent pregnancy establishment. The authors combine whole-organ three-dimensional imaging of optically cleared mouse uteri with single-cell and spatial transcriptomic profiling, conditional deletion of p38α at the uterine-wide versus epithelial-restricted level, and rescue experiments using progesterone and leukemia inhibitory factor. Based on these datasets, the authors propose that the luminal epithelium undergoes a previously underappreciated phase of organ-scale architectural reorganization before attachment, and that a p38α-dependent stress-responsive program coordinates epithelial remodeling together with epithelial-stromal communication required for implantation competence.

Strengths:

By moving beyond local attachment-site morphology to a horn-level representation of luminal topology, the work provides anatomical context that is difficult to reconstruct from conventional section-based approaches and should be broadly useful to the implantation community. The integration of organ-scale morphology with single-cell and spatial transcriptomic datasets, together with genetic perturbation and rescue experiments, adds breadth and increases the potential long-term utility of the dataset for investigators interested in uterine receptivity and embryo-uterine interactions.

Weaknesses:

(1) The whole-uterus analysis of luminal folds and creases requires stronger methodological support. Given the mechanical compliance of the uterine lumen, it is difficult to evaluate from the current description whether dissection, fixation, clearing, and/or mounting could influence the observed luminal topography. This issue is particularly important because several key conclusions depend on the spatial distribution of folds across the uterine horn. A fuller account of tissue handling and reconstruction, together with validation that the preparation preserves native morphology, would substantially strengthen confidence in the organ-scale conclusions.

(2) Several of the central morphological claims are supported primarily only by representative reconstructions. This includes the proposed flattening/creasing dynamics, alternating stretched and shrunken regions, persistence of abnormal folding in the mutant uterus, and the extent of structural rescue following progesterone supplementation. The authors could extract objective measures from the reconstructed luminal surface and provide more statistical analysis to demonstrate the reproducibility of the results.

(3) The manuscript appears to over-reach in concluding that luminal remodeling zones embryos before attachment from day 4 to 5. As presented, the data support a correlation between luminal architecture and embryo position, but do not discriminate between (i) luminal remodeling directing embryo positioning, (ii) embryos locally shaping the lumen, or (iii) parallel regulation of both. The evidence is based on observations of the uterus and the inside blastocysts at certain time points around implantation. Without the time-lapse analysis within the uterus, the dynamic interactions between embryos and the uterus couldn't be determined.

(4) The key conclusion of the manuscript is that uterine p38α regulates luminal epithelial remodeling required for embryo attachment, as shown in the title. Against this background, the finding that epithelial-restricted loss of p38α does not overtly impair fertility is notable, as it suggests that the major function of p38α may not be epithelial cell-autonomous but instead may arise through other uterine compartments that secondarily influence the epithelium. At present, however, this conclusion remains insufficiently supported: the epithelial-specific model is not characterized in sufficient depth during the peri-implantation period, and the transcriptomic evidence for altered epithelial-stromal communication does not by itself explain the phenotypic difference between uterine-wide and epithelial-specific deletion. If stromal p38α is proposed as the critical upstream regulator, more direct testing, such as stromal-specific deletion, would be needed.

Reviewer #2 (Public review):

Summary:

In this study, the authors aimed to characterize the architectural reorganization of the uterine luminal epithelium during the implantation period. Using 3D histological reconstruction, single-cell RNA sequencing, and spatial transcriptomics, the authors characterize luminal remodeling during the peri-implantation period and employ a mouse model to explore the role of p38α in regulating luminal flattening.

Strengths:

This study clearly described the changes in luminal architecture during implantation. Moreover, they also used integration of multiple advanced techniques, including 3D tissue reconstruction, single-cell transcriptomics, and spatial transcriptomics, which together provide a detailed description of the molecular characteristics of the uterine architecture during implantation.

Weaknesses:

The authors used PR-Cre to generate uterine p38α knockout mice. This Cre driver deletes p38α not only in epithelial cells but also in stromal compartments. Therefore, it remains unclear whether the observed phenotype arises from epithelial cells, stromal cells, or a combination of both. Previous studies have shown that p38α regulates epithelial polarity, cytoskeletal organization, and E-cadherin localization. However, the current study does not examine changes in cell adhesion or epithelial junction integrity. Previous studies have reported that uterine fluid absorption during implantation is closely associated with luminal closure and remodeling. It would be important to determine whether epithelial transport-related genes are altered in the mutant uterus. Could dysregulated fluid homeostasis contribute to the implantation defects observed in the p38α-deficient mice?

eLife Assessment

This manuscript provides valuable high-resolution structural insights into the interaction between vaccine-elicited antibodies and SARS‑CoV‑2 evolution. The evidence is solid; however, the conclusions could be strengthened with further experimentation and analysis.

Reviewer #1 (Public review):

Summary:

The authors provide high-resolution cryoEM structures to map and functionally characterize human antibodies against SARS-CoV-2 elicited by a standard mRNA vaccine. Here, they report high-resolution structural information on seven previously documented neutralizing antibodies from this response, which were produced from early plasmablasts and which engage diverse targets on the viral spike glycoprotein. This structural information is then integrated with functional assays to define how antibody epitope specificity, geometry, and conformational dynamics may shape neutralization outcomes.

Strengths:

A core strength of the study is a technically-well executed analysis of multiple 'ectopically balanced' mAbs elicited by early B cell plasmablast responses. These antibodies engage different neutralizing targets on the S-trimer of SARS-CoV-2, including the RBD and NTD domains. This has resolved a core distinction in terms of how nAbs engaging these features (and subfeatures, e.g., more conserved hydrophobic pocket within NTD) neutralize the virus.

Weaknesses:

A general weakness is that these antibody classes have been structurally characterized already (albeit individually), and much of this work has been done in the context of understanding susceptibility to escape mutations (delta, omicron, and subvariants therein; class I-IV antibody crossreactivity on Wuhan SARS-CoV-2 to present). It is exceptionally fine technical work presenting the antibodies in a collection like this, but perhaps the new predictive power of this analysis is somewhat overstated.

The early plasmablast angle seems like it could be better fleshed out. Many of the known SARS-CoV-2 nAbs are from the plasmablast pool, but how does this predict the antibody profile at latter stages, as per the stated goal and claim of the current study? Does the paratope pattern of plasmablast antibodies then change within the immune sera at later time points? New or existing cryoEMPEM data could shed light on this.

Reviewer #2 (Public review):

Summary:

This manuscript provides important insights into the interaction between early vaccine-elicited antibodies and SARS‑CoV‑2 evolution. The work will be of broad interest to researchers in structural virology, immunology, and vaccine development. However, several conclusions-particularly those involving neutralization breadth and spike destabilization-require additional functional and biophysical validation.

Strengths:

The manuscript provides an unusually comprehensive structural dataset, resolving all neutralizing antibodies in complex with the SARS‑CoV‑2 spike and enabling direct mechanistic comparison across epitope classes. Its integration of cryo‑EM structures with variant binding, sequence analysis, and fusion‑inhibition assays offers a coherent, multidimensional explanation for antibody breadth and escape. Notably, the identification of a conserved NTD hydrophobic pocket targeted by broad-reactive antibodies represents a conceptually important advance with clear implications for future vaccine design.

Weaknesses:

The study lacks variant-specific neutralization assays, limiting the ability to directly correlate binding breadth with functional viral inhibition. It also omits kinetic affinity measurements, leaving important mechanistic questions, such as why certain antibodies retain breadth, only partially resolved. Additionally, reliance on HEK293T-based spike display raises concerns about glycosylation-related artifacts, especially for NTD loop-dependent antibodies.

Reviewer #3 (Public review):

Summary:

In this manuscript by Jaiswal et al., the authors used structural biology combined with cellular assays to determine the molecular basis underlying the neutralizing ability of the SARS-CoV-2 antibodies. The authors compared the binding mode of the neutralizing antibodies that have two distinct binding interfaces and identified key sites that determine their vulnerability to virus evolution. Interestingly, the author also demonstrated that the trimer-disrupting antibody has the broadest activity as the variations at the trimer interface are limited in evolution.

Strengths:

This manuscript reported a large collection of structures and covered a broad range of binding modes and mechanisms of action. Many of the cryo-EM structures are of good quality. The authors' hypothesis regarding the molecular determinants of evolution vulnerability is solid.

Weaknesses:

However, in my opinion, several points listed below need to be addressed.

(1) At the beginning of the results section, the authors started by determining the structures of Fab PVI.V3-9 and Fab PVI.V6-4 in complex with the ancestral SARS-CoV-2 spike. However, the readers could benefit from a brief introduction of the Fabs PVI.V3-9 and PVI.V6-4. The same applies to the anti-NTD Fabs.

(2) In Figure 1A and E, the spike protein is shown with two different views. It is best to show the same view for comparison.

(3) Throughout the manuscript, the map quality of some Fabs (e.g., V6-11, V6-7, V6-2) is suboptimal. Does the map support the claims on the residues that form the interface? The authors should provide a figure showing the cryo-EM density for all side-chain residues involved at the interface.

(4) Line 152, the terminology "NTD top binders" could be ambiguous, as it could mean those Fabs have the strongest binding affinity. Maybe the authors can change the "top" to "tip".

(5) The authors described the interface between the spike protein and the Fabs in great detail. However, it would be nice if the authors could summarize the common binding strategy for each group of antibodies that utilize the same binding surface.

(6) Line 275, the authors should define what strain of Omicron is in Figure 4. The authors should also explain that the strains in Figure 4A are ordered by evolutionary age.

(7) Lines 286-287, isn't this conclusion already made from the cell-based flow cytometry binding assay? This sentence could be deleted.

(8) In both Figures S10 and S11, the readers could benefit from an additional row highlighting the residues interacting with ACE2.

(9) Lines 298-301, based on Figure S11, no contact is made between the N2 loop and the Fab. The authors may elaborate on why the mutations observed in the N2 loop indirectly influenced Fab recognition.

(10) Lines 321-323, even though this is a well-established assay, it is probably better to clearly explain that one pool of cells expresses the spike and the other pool of cells expresses ACE2.

eLife Assessment

This valuable study compares orthogonal approaches for detecting RNA chemical modifications and provides a helpful framework for improving the reliability of direct RNA sequencing-based identification of RNA modifications. The evidence supporting the technical benchmarking claims is solid. However, support for the broader biological conclusions is not as strong, and the quantitative interpretation of the results, as well as the limitations of the underlying models, would benefit from further clarification.

Reviewer #1 (Public review):

Summary:

The authors set out to evaluate how accurately direct sequencing of RNA can identify and quantify several chemical modifications on RNA molecules, focusing primarily on m6A. A central goal of the work is to compare this approach with an independent chemical-based method (glyoxal and nitrite-mediated deamination of unmethylated adenosines (GLORI), using the same RNA samples, in order to assess reproducibility, false-positive signals, and sensitivity across a range of detection strategies. The authors further aim to demonstrate the biological utility of this approach by applying it to two human cell types, primary human fibroblasts and HD10.6 neurons. While the manuscript also reports detection of additional RNA modifications (pseudouridine and m5C, the depth of analysis and strength of controls are greatest for m6A, which forms the primary focus of the study

Strengths:

A strength of this work is the direct comparison of two distinct measurement approaches performed on the same RNA input material; this has not been done in other recently published benchmarking studies evaluating the utility of the recent direct RNA sequencing for calling m6A. The authors systematically test multiple analysis models and show that, when appropriate filtering is applied, detection of modified sites is reproducible across software versions. The use of synthetic RNA standards and METTL3 inhibitors as negative controls helps to reinforce the overall results.

The data show good agreement between the two methods at higher m6A modification levels, supporting the conclusion that direct RNA sequencing can reliably detect high-confidence modification sites. The authors also demonstrate that this approach can, in principle, provide information at the level of individual RNA variants (although only one example was provided), which is difficult to achieve with short-read methods. The methodology described here is likely to be useful to others seeking to apply similar approaches to identify and quantify m6A. The study also explores the detection of other RNA modifications, which highlights the broader potential of the approach, although these analyses are necessarily more exploratory given the more limited controls and data available.

Weaknesses:

Despite these strengths, several issues limit the interpretation of the results and should be clarified for readers.

First, the authors appropriately address false-positive signals by estimating expected false-positive rates and by quantitatively comparing sequence motif enrichment before and after filtering. These analyses provide important support for the use of stoichiometry-based thresholds and demonstrate that filtering substantially improves specificity. However, even after filtering, a subset of detected sites remains outside the expected sequence context. It therefore remains unclear to what extent these non-canonical sites reflect genuine biology versus residual technical artifacts.

Second, claims regarding the ability of direct RNA sequencing to resolve modification patterns across different RNA variants are supported by very limited evidence. The conclusion that this approach provides superior isoform-level quantification relative to short-read methods is based largely on a single gene example. While this case is interesting, it does not establish how widespread or general this advantage is. A broader analysis indicating how many genes show isoform-specific modification patterns detectable by this method, and how often these are missed by the comparison approach, would be necessary to support a general claim.

Third, the biological interpretation of cell type-specific differences in modification levels remains underdeveloped. Although differences in modification stoichiometry are reported between fibroblasts and neuron-derived cells, the functional consequences of these differences are not addressed. It is unclear whether changes in modification levels are associated with differences in RNA abundance, stability, or translation. As a result, statements suggesting that these modifications fine-tune core cellular pathways are speculative and should either be supported with additional analyses or framed more cautiously.

Related to this point, differences in gene expression between the two cell types are a potential confounding factor. The pathway enrichment patterns presented appear biased toward particular functional categories, but without clear control for differential gene expression, it is difficult to determine whether the observed enrichment reflects cell type-specific regulation of RNA modification or simply differences in which genes are expressed. Clarifying how background gene sets were defined for these analyses would help readers interpret the results.

The manuscript also suggests broader differences in overall modification levels between cell types, but this is not validated using an independent global assay. An orthogonal measurement of total modification levels on polyadenylated RNA (for example, dot blot) would help place site-specific stoichiometry differences in a clearer biological context.

Finally, the effects of the METTL3 inhibitor on these cell types are not fully characterized. While changes in m6A modification patterns are reported following treatment, the manuscript does not address whether the treatment affects cell growth or viability.

Appraisal of conclusions and impact:

Overall, the study provides an informative technical assessment of direct RNA sequencing for modification detection and establishes clear conditions under which the method performs well. The evidence strongly supports conclusions related to technical benchmarking, reproducibility, and the importance of filtering and controls, particularly for m6A. In contrast, conclusions regarding isoform-specific regulation and cell type-specific biological roles of RNA modification are less well supported by the data currently presented, and would benefit from either additional analysis or more restrained interpretation.

The work is likely to have a meaningful impact as a practical reference for researchers using direct RNA sequencing, particularly by clarifying sources of false positives and the value of appropriate controls. With clearer limits placed on biological interpretation or more data presented in support of the biological interpretation, the study would serve as a valuable reference for users seeking to apply these technologies reliably.

Reviewer #2 (Public review):

Summary:

In this study, the authors aim to establish a calibrated framework for detecting RNA modifications using long-read sequencing and apply it to compare modification patterns between fibroblasts and neuron-like cells. The work combines long-read sequencing, in vitro transcribed controls, methyltransferase inhibition, and comparison to an orthogonal sequencing-based method in an attempt to derive filtering strategies that reduce false positive modification calls. The authors further apply this framework to explore differences in modification levels between the two cell types.

The resulting dataset may be of interest to researchers working on RNA modification detection using long-read sequencing technologies. Independent datasets across additional cellular systems can be useful for benchmarking computational methods and evaluating the behavior of modification detection models. However, the conceptual advance of the analytical framework presented here remains somewhat unclear, as many aspects of the analysis closely resemble strategies that have already been described in recent benchmarking studies.

Strengths:

A clear strength of the study is the generation of a relatively large long-read sequencing dataset together with several useful experimental controls, including in vitro transcribed RNA and pharmacological inhibition of the methyltransferase enzyme responsible for installing this modification. These controls are helpful for illustrating the challenges associated with distinguishing high-confidence modification sites from background signals. The inclusion of two different human cellular systems also provides an additional dataset that may be useful for benchmarking and cross-validation in the field. The study addresses a practically relevant question for the community, namely, how to reduce false positive calls in long-read sequencing-based RNA modification analyses.

Weaknesses:

The main weakness of the manuscript is its limited methodological novelty. Much of the analytical framework presented here closely follows benchmarking strategies that have already been described in recent studies of RNA modification detection using long-read sequencing. Several previous studies have evaluated modification-aware basecalling approaches, discussed the need for stringent filtering strategies, and compared long-read sequencing-based predictions with orthogonal mapping approaches. The manuscript would therefore benefit from a deeper engagement with the recent benchmarking literature and a clearer explanation of what conceptual or methodological advance the present study provides beyond these earlier analyses.

A second concern relates to the filtering strategy that forms the core of the proposed workflow. The manuscript applies several thresholds, including modification probability, stoichiometry, and read coverage cutoffs, but it is not clearly explained how these thresholds were determined. It remains unclear whether these cutoffs were derived from statistical calibration, empirical optimization using the presented dataset, or adopted from previous studies. Because the downstream conclusions depend strongly on these filtering choices, a clearer methodological justification would strengthen the work and help readers assess the robustness of the proposed framework.

The interpretation of the comparison between the two modification detection approaches also appears somewhat overstated. Differences between the methods are frequently interpreted as evidence that one approach produces large numbers of false positive calls, but the analyses presented do not fully exclude alternative explanations such as differences in sensitivity, sequencing depth, or methodological biases. A more cautious interpretation of these discrepancies would therefore be appropriate.

Some discussion points also appear speculative. In particular, certain interpretations propose mechanistic explanations without presenting analyses that would allow these possibilities to be distinguished. Such interpretations would benefit from either additional supporting analyses or more cautious phrasing.

From a methodological perspective, the statistical robustness of the thresholds used throughout the analysis could also be discussed in more detail. Given the relatively modest read coverage cutoff applied in the study, low stoichiometry estimates may be strongly influenced by sampling noise, and fixed stoichiometry thresholds may therefore not correspond to a consistent level of confidence across sites. In addition, the manuscript relies heavily on fixed modification probability cutoffs to define high-confidence calls, but it does not discuss whether these scores are statistically calibrated or how they relate to expected error rates. Neural network outputs are often not well-calibrated probabilities, and interpreting these values as direct confidence estimates can therefore be problematic. Finally, modification detection models trained on known modification sites may capture sequence-context patterns present in the training data, meaning that motif enrichment or positional distributions along transcripts may partly reflect model biases rather than purely biological signals. A brief discussion of these limitations would help readers better interpret the robustness of the proposed filtering strategy and the downstream biological conclusions.

Overall, while the dataset may be of interest to the community, the extent to which the study advances current methodological understanding beyond recent benchmarking efforts remains limited.

Minor comments:

The discussion of the "DRACH" versus "all-context" outputs would benefit from greater technical precision. The statement that the number of sites within DRACH motifs identified by the all-context approach was nearly identical to the number reported by the DRACH model may suggest that these outputs derive from fundamentally different predictive models. As I understand it, the underlying neural network is the same, whereas the distinction lies primarily in the classification context. Clarifying this explicitly in the manuscript would improve interpretability and avoid potential confusion for readers.

The manuscript compares results obtained with different basecalling and modification settings but refers primarily to Dorado software versions. This may be misleading, as software version and model version are not necessarily equivalent. Different basecalling or modification models can be used with the same software release, and newer software versions may still use older models. For clarity and reproducibility, the authors should report the exact basecalling and modification model names used in the analyses rather than referring only to the Dorado software version.

Reviewer #3 (Public review):

In this study, the authors aim to establish a calibrated framework for identifying RNA chemical marks from direct RNA sequencing data using a modification-aware basecalling workflow, with a particular focus on N6-methyladenosine. By combining native RNA sequencing with an unmodified control transcriptome, enzyme inhibition, comparison across multiple software versions, and orthogonal validation using an independent mapping approach, the authors seek to define a best-practice pipeline for reducing false-positive calls and improving confidence in quantitative interpretation across cell types.

A major strength of the work is the rigor of the benchmarking strategy. In particular, the inclusion of an unmodified control transcriptome is both important and useful, and the study provides compelling evidence that this control remains necessary for robust interpretation, despite being omitted in many current workflows. The comparison across software versions and the matched analysis with an independent sequencing-based approach also substantially strengthen the evidence presented. The work therefore makes a valuable contribution to the community by offering a more stringent analytical framework that will likely be broadly useful to groups applying native RNA sequencing to study RNA chemical marks.

The evidence supporting the main conclusions is solid overall. The authors convincingly show that stringent filtering substantially reduces false-positive calls and improves agreement with orthogonal approaches, particularly at highly modified sites. The observation that many sites are conserved across cell types, while showing differences in relative modification levels, is also supported by the presented analyses.

At the same time, several conceptual issues limit the strength of some downstream interpretations. Most importantly, the manuscript repeatedly refers to the reported values as "stoichiometry," whereas the underlying software output is more appropriately interpreted as a statistical estimate of the proportion of aligned reads classified as modified. This distinction is important because the conclusions regarding cell-type differences rely on quantitative comparisons of these values. In addition, the current calling framework depends on successful canonical base assignment before modification calling, which raises an important limitation: sites with the strongest signal deviations may be underrepresented if they are more likely to be miscalled during basecalling. This issue may be especially relevant for RNA marks that induce stronger mismatch signatures than N6-methyladenosine and should be more explicitly discussed.

Overall, the authors largely achieve their primary aim of establishing a more rigorous and broadly applicable analytical framework for direct RNA sequencing-based modification detection. The work is likely to have a meaningful impact on the field, particularly by reinforcing the importance of appropriate negative controls and benchmarking standards. With clearer framing of the quantitative outputs and explicit discussion of current software limitations, this study will serve as a highly useful resource for the community.

eLife Assessment

This well-designed study offers important insights into the development of infants' responses to music based on the exploration of EEG neural auditory responses and video-based movement analysis. The compelling results revealed that evoked responses emerge between 3 and 12 months of age, but no age group demonstrated evidence of coordinated movements to music. This study will be of significant interest to developmental psychologists and neuroscientists, as well as researchers interested in music processing and in the translation of perception into action.

Reviewer #1 (Public review):

Summary:

This study aims to investigate the development of infants' responses to music by examining neural activity via EEG and spontaneous body kinematics using video-based analysis. The authors also explore the role of musical pitch in eliciting neural and motor responses, comparing infants at 3, 6, and 12 months of age.

Strengths:

A key strength of the study lies in its analysis of body kinematics and modeling of stimulus-motor coupling, demonstrating how the amplitude envelope of music predicts infant movement, and how higher musical pitch may enhance auditory-motor synchronization.

EEG data provide evidence for enhanced neural responses to music compared to shuffled auditory sequences. These findings ecourage further investigation of the proposed developmental trajectory of neural responses to music and their link to musical behavior in infants.

Comments on revisions:

The authors have addressed my questions in their revision. I have no other questions. Thank you for the opportunity to read and evaluate this interesting study and also for all the work carried out in response to the comments.

Author response:

The following is the authors’ response to the previous reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study aims to investigate the development of infants' responses to music by examining neural activity via EEG and spontaneous body kinematics using video-based analysis. The authors also explore the role of musical pitch in eliciting neural and motor responses, comparing infants at 3, 6, and 12 months of age.

Strengths:

A key strength of the study lies in its analysis of body kinematics and modeling of stimulus-motor coupling, demonstrating how the amplitude envelope of music predicts infant movement, and how higher musical pitch may enhance auditory-motor synchronization.

EEG data provide evidence for enhanced neural responses to music compared to shuffled auditory sequences. These findings ecourage further investigation of the proposed developmental trajectory of neural responses to music and their link to musical behavior in infants.

Comments on revisions:

I thank the authors for the considerable effort devoted to revising the manuscript and addressing the raised questions and comments. I particularly appreciate the additional analyses and the extended arguments included in the discussion. I believe that this paper represents a valuable contribution to the literature on music development.

One remaining comment concerns the evoked response observed in the shuffled condition, which I still find intriguing. Considering that the auditory events in the shuffled condition display a clear rise time, particularly for those events that were selected based on being preceded and followed by longer periods of silence, one would expect to observe an evoked response emerging from baseline. However, this pattern is not evident in the presented curves. The authors may further examine and discuss the shape and characteristics of these response patterns.

We thank the Reviewer for highlighting this intriguing aspect of our data. We entirely agree that from a purely bottom-up, acoustic perspective, one would expect a clear onset-locked evoked response, such as an P1/P2 complex in adults or its developmental equivalent, given the prominent acoustic rise times and the surrounding periods of silence (such as those accounted for in the control analyses)

The fact that these responses are not present in the curves for the shuffled condition was striking to us as well. We interpret this severe attenuation not as a failure of sensory perception, but potentially as a consequence of higher-level cognitive modulation. Specifically, because the shuffled condition completely lacks structural regularities, the brain might be unable to build reliable temporal and/or melodic expectations. In the absence of a learnable structure, the auditory system likely down-weights the processing of these random sequences to conserve cognitive resources, leading participants to attentionally disengage.

This phenomenon aligns with both developmental and adult models of auditory processing. For instance, the "Goldilocks effect" demonstrates that infants systematically withdraw attention from auditory sequences that are entirely unpredictable (Kidd et al., 2014). Similarly, adult auditory literature suggests that while predictable patterns automatically capture attention, random and unpredictable acoustic streams could be actively tuned out (Dayan et al., 2000; Esber & Haselgrove, 2011).

Following the Reviewer’s helpful suggestion to further discuss the characteristics of these response patterns, we have expanded our description and interpretation of the shuffled condition curves in the revised manuscript. We added the following text to the Methods and Discussion to explicitly address the dampened shape of these responses:

p. 9: “Importantly, and in line with the adults’ data, all infant groups exhibited enhanced P1 amplitudes in response to music compared to shuffled music. Actually, across all groups, shuffled music did not elicit clear ERPs as the ones elicited by music”.

p.20: “This process was markedly dampened or interrupted by shuffled music (Bianco et al., 2024, 2025; Lense et al., 2022), a finding that could be interpreted as evidence of disengagement from such highly unpredictable sequences (Dayan et al., 2000; Esber & Haselgrove, 2011; Kidd et al., 2014).”

Reviewer #2 (Public review):

Summary:

Infants' auditory brain responses reveal processing of music (clearly different from shuffled music patterns) from the age of 3 months; however, they do not show related increase in spontaneous movement activity to music until the age of 12 months.

Strengths:

This is a nice paper, well designed, with sophisticated analyses and presenting clear results filling an important gap about early infant sensitivity, detection, and differentiation of musical sounds. The addition of EEG recordings (specifically ERPs) in response to music presentations at 3 different infant ages in the first postnatal year is important, and the manipulation of the music stimuli into shuffled, high and low pitch to capture differences in brain response processing and spontaneous movements is interesting. Further, the movement analysis based on Quantity of Movements (QoM) and movement subdivision into 10 distinct Principal Movements (PMs) is novel and creative.

Overall, results show that ERPs responses to music occurs earlier than QoM in early development, and that even at 12 months, motor responses to music remain coarse and not rhythmically aligned with the music tempo. This work increases our fundamental understanding of infants' early music perception in relation to auditory processing and motor response.

Comments on revisions:

The authors have addressed my questions in their revision. I have no other questions. Thanks again for the opportunity to read and evaluate this interesting work.

We thank the Reviewer for their time, their positive evaluation of our revised manuscript, and their constructive feedback throughout the review process, which has greatly helped us to strengthen this paper.

Reviewer #3 (Public review):

Summary

This study provides a detailed investigation of neural auditory responses and spontaneous movements in infants listening to music. Analyses of EEG data (event-related potentials and steady-state responses) first highlighted that infants at 3, 6 and 12 months of age and adults showed enhanced auditory responses to music than shuffled music. 6-month-olds also exhibited enhanced P1 response to high-pitch vs low-pitch stimuli, but not the other groups. Besides, whole body spontaneous movements of infants were decomposed into 10 principal components. Kinematic analyses revealed that the quantity of movement was higher in response to music than shuffled music only at 12 months of age. Although Granger causality analysis suggested that infants' movement was related to the music intensity changes, particularly in the high-pitch condition, infants did not exhibit phase-locked movement responses to musical events, and the low movement periodicity was not coordinated with music.

Strengths

This study investigates an important topic on the development of music perception and translation to action and danse. It targets a crucial developmental period that is difficult to explore. It evaluates two modalities by measuring neural auditory responses and kinematics, while cross-modal development is rarely evaluated. Overall, the study fills a clear gap in the literature.

Besides, the study uses state-of-the-art analyses. Detailed investigations were performed, as well as exploratory analyses in supplementary information. The discussion is rich in neurodevelopmental interpretations and comparisons with the literature. All steps are clearly detailed. The manuscript is very clear, well-written and pleasant to read. Figures are well-designed and informative. The authors' responses to previous reviews are also detailed and informative.

Comments on revisions:

The authors answered all my questions.

Thank you very much for your positive evaluation and for taking the time to review our revisions. We deeply appreciate your insightful comments across the review rounds, which have helped us improve the clarity and rigor of our paper.

eLife Assessment

This study provides important insights into how tumorous germline stem cells (GSCs) in the Drosophila melanogaster ovary can mimic niche function and suppress the differentiation of neighboring cells. The findings that GSC tumors can incorporate non-mutant cells and inhibit their differentiation are compelling and extend current understanding of stem cell-niche interactions. However, the evidence supporting the conclusion that GSC tumors produce BMP ligands to mediate this effect remains incomplete, due to concerns regarding the quality and interpretation of the HCR-FISH data.

Reviewer #1 (Public review):

Summary:

This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche. This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche.

Strengths:

(1) Use of an excellent and established model for tumorous cells in a stem cell microenvironment

(2) Powerful genetics allow them to test various factors in the tumorous vs non-tumorous cells

(3) Appropriate use of quantification and statistics

Weaknesses:

(1) What is the frequency of SGCs in nos>flp; bam-mutant tumors? For example, are they seen in every germarium, or in some germaria, etc or in a few germaria.

This concern was addressed in the rebuttal. The line number is 106, not line 103.

(2) Does the breakdown in clonality vary when they induce hs-flp clones in adults as opposed to in larvae/pupae?

This concern was addressed in the rebuttal. However, these statements are no on lines 331-335 but instead starting on line 339. Please be accurate about the line numbers cited in the rebuttal. They need to match the line numbers in the revised manuscript.

(3) Approximately 20-25% of SGCs are bam+, dad-LacZ+. Firstly, how do the authors explain this? Secondly, of the 70-75% of SGCs that have no/low BMP signaling, the authors should perform additional characterization using markers that are expressed in GSCs (i.e., Sex lethal and nanos).

The authors did not perform additional staining for GSC-enriched protein like Sex lethal and nanos.

(4) All experiments except Fig. 1I (where a single germarium with no quantification) were performed with nos-Gal4, UASp-flp. Have the authors performed any of the phenotypic characterizations (i.e., figures other than figure 1) with hs-flp?

In the rebuttal, the authors stated that they used nos>flp for all figures except for Fig. 1I. It would be more convincing for them to prove in Fig. 1 than there is not phenoytpic difference between the two methods and then switch to the nos>FLP method for the rest of the paper.

(5) Does the number of SGCs change with the age of the female? The experiments were all performed in 14-day old adult females. What happens when they look at young female (like 2-day old). I assume that the nos>flp is working in larval and pupal stages and so the phenotype should be present in young females. Why did the authors choose this later age? For example, is the phenotype more robust in older females? or do you see more SGCs at later time points?

The authors did not supply any data to prove that the clones were larger in 14-day-old flies than in younger flies. Additionally, the age of "younger" flies was not specified. Therefore, the authors did not satisfactorily answer my concern.

(6) Can the authors distinguish one copy of GFP versus 2 copies of GFP in germ cells of the ovary? This is not possible in the Drosophila testis. I ask because this could impact on the clonal analyses diagrammed in Fig. 4A and 4G and in 6A and B. Additionally, in most of the figures, the GFP is saturated so it is not possible to discern one vs two copies of GFP.

In the rebuttal, the authors stated that they cannot differential one vs two copies of GFP. They used other clone labeling methods in Fig. 4 and 6. I think that the authors should make a statement in the manuscript that they cannot distinguish one vs two copies of GFP for the record.

(7) More evidence is needed to support the claim of elevated Dpp levels in bam or bgcn mutant tumors. The current results with dpp-lacZ enhancer trap in Fig 5A,B are not convincing. First, why is the dpp-lacZ so much brighter in the mosaic analysis (A) than in the no-clone analysis (B); it is expected that the level of dpp-lacZ in cap cells should be invariant between ovaries and yet LacZ is very faint in Fig. 5B. I think that if the settings in A matched those in B, the apparent expression of dpp-lacZ in the tumor would be much lower and likely not statistically significantly. Second, they should use RNA in situ hybridization with a sensitive technique like hybridization chain reactions (HCR) - an approach that has worked well in numerous Drosophila tissues including the ovary.

The HCR FISH in Fig.5 of the revised manuscript needs an explanation for how the mRNA puncta were quantified. Currently, there is no information in the methods. What is meant but relative dpp levels. I think that the authors should report in and unbiased manner "number" of dpp or gbb puncta in TFs. For the germaria, I think that they should report the number of puncta of dpp or gbb divide by the total area in square pixels counted. Additionally, the background fluorescence is noticeably much higher in bamBG/delta86 germaria, which would (falsely) increase the relative intensity of dpp and gbb in bam mutants. Although, I commend the authors for performing HCR FISH, these data are still not convincing to me.

(8) In Fig 6, the authors report results obtained with the bamBG allele. Do they obtain similar data with another bam allele (i.e., bamdelta86)?

The authors did not try any experiments with the bamdelta86 allele, despite this allele being molecularly defined, where the bamBG allele is not defined.

Comments on second revision:

The authors have adequately addressed several points. However, there is still no information in the material and methods for how they measured and quantified the HCR-FISH probe signal. They have the same size region that they use for each genotype, but they do not control for the number of nuclei in each square. I would also be helpful if they provided a different image for the gbb probe stained in the mutant background. It is the only panel that does not have other germaria in very close proximity. I am still not fully convinced of the HCR data, esp for gbb.

Reviewer #2 (Public review):

In the current version, Zhang et al. have made substantial improvements to the manuscript. It is now easier to read, and the data are more solid compared with the previous version, supporting their conclusion that tumor GSCs secrete stemness factors (BMPs and Dpp) to suppress the differentiation of neighboring wild-type GSCs. This study should benefit a broad readership across developmental biology, germ cell biology, stem cell biology, and cancer biology.

Comments on revision:

If the exact number of germaria was not recorded (as described), an approximate number can be provided in the Materials and Methods; for example, stating that more than 10 germaria were analyzed per biological replicate.

Reviewer #3 (Public review):

Zhang et al. investigated how germline tumors influence the development of neighboring wild-type (WT) germline stem cells (GSC) in the Drosophila ovary. They report that germline tumors generated by differentiation-arrested mutations (bam and bgcn) inhibit the differentiation of neighboring WT GSCs by arresting them in an undifferentiated state, resulting from reduced expression of the differentiation-promoting factor Bam. They find that these tumor cells produce low levels of the niche-associated signaling molecules Dpp and Gbb, which suppress bam expression and consequently inhibit the differentiation of neighboring WT GSCs non-cell-autonomously. Based on these findings, the authors propose that germline tumors mimic the niche to suppress the differentiation of the neighboring wild-type germline stem cells.

Strengths:

The study uses a well-established in vivo model to addresses an important biological question concerning the interaction between germline tumor cells and wild-type (WT) germline stem cells in the Drosophila ovary. If the findings are substantiated, this study could provide valuable insights that are applicable to other stem cell systems.

Weaknesses:

The authors have addressed some of my concerns in the revised submission. However, the data presented do not allow the authors to distinguish whether the failed differentiation of WT stem cells/germline cells results from "arrested differentiation due to the loss of the differentiation niche" or from "direct inhibition by tumor-derived expression of niche-associated molecules Dpp and Gbb". The critical supporting data, HCR in situ results, are not sufficiently convincing.

Author response:

The following is the authors’ response to the previous reviews

eLife Assessment

This study presents results supporting a model that tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the stem cell niche and inhibit the differentiation of neighboring cells. The valuable findings show that GSC tumors often contain non-mutant cells whose differentiation is suppressed by the GSC tumorous cells. However, the evidence showing that the GSC tumors produce BMP ligands to suppress differentiation of non-mutant cells is incomplete due to concerns about the new HCR data.

Thanks for this assessment. All concerns raised by the reviewers regarding the HCR data and others are followed by our responses below.

Public Reviews:

Reviewer #1 (Public review):

Summary:

This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche. This preprint from Shaowei Zhao and colleagues presents results that suggest tumorous germline stem cells (GSCs) in the Drosophila ovary mimic the ovarian stem cell niche and inhibit the differentiation of neighboring non-mutant GSC-like cells. The authors use FRT-mediated clonal analysis driven by a germline-specific gene (nos-Gal4, UASp-flp) to induce GSC-like cells mutant for bam or bam's co-factor bgcn. Bam-mutant or bgcn-mutant germ cells produce tumors in the stem cell compartment (the germarium) of the ovary (Fig. 1). These tumors contain non-mutant cells - termed SGC for single-germ cells. 75% of SGCs do not exhibit signs of differentiation (as assessed by bamP-GFP) (Fig. 2). The authors demonstrate that block in differentiation in SGC is a result of suppression of bam expression (Fig. 2). They present data suggesting that in 73% of SGCs BMP signaling is low (assessed by dad-lacZ) (Fig. 3) and proliferation is less in SGCs vs GSCs. They present genetic evidence that mutations in BMP pathway receptors and transcription factors suppress some of the non-autonomous effects exhibited by SGCs within bam-mutant tumors (Fig. 4). They show data that bam-mutant cells secrete Dpp, but this data is not compelling (see below) (Fig. 5). They provide genetic data that loss of BMP ligands (dpp and gbb) suppresses the appearance of SGCs in bam-mutant tumors (Fig. 6). Taken together, their data support a model in which bam-mutant GSC-like cells produce BMPs that act on non-mutant cells (i.e., SGCs) to prevent their differentiation, similar to what in seen in the ovarian stem cell niche.

Strengths:

(1) Use of an excellent and established model for tumorous cells in a stem cell microenvironment

(2) Powerful genetics allow them to test various factors in the tumorous vs non-tumorous cells

(3) Appropriate use of quantification and statistics

Thank you for your valuable comments, and we greatly appreciate them.

Weaknesses:

(1) What is the frequency of SGCs in nos>flp; bam-mutant tumors? For example, are they seen in every germarium, or in some germaria, etc or in a few germaria.

This concern was addressed in the rebuttal. The line number is 106, not line 103.

(2) Does the breakdown in clonality vary when they induce hs-flp clones in adults as opposed to in larvae/pupae?

This concern was addressed in the rebuttal. However, these statements are no on lines 331-335 but instead starting on line 339. Please be accurate about the line numbers cited in the rebuttal. They need to match the line numbers in the revised manuscript.

We have rechecked the line numbers and confirmed that the mismatch arose from the Word-to-PDF conversion process on the eLife website. As this issue has recurred and reviewers’ file-format preferences are unknown to us, we have added a clarifying note at the beginning of each response letter: “Please note that the line numbers cited refer to the revised manuscript in the Microsoft Word format”.

(3) Approximately 20-25% of SGCs are bam+, dad-LacZ+. Firstly, how do the authors explain this? Secondly, of the 70-75% of SGCs that have no/low BMP signaling, the authors should perform additional characterization using markers that are expressed in GSCs (i.e., Sex lethal and nanos).

The authors did not perform additional staining for GSC-enriched protein like Sex lethal and nanos.

The 70-75% of SGCs that have low BMP signaling display the following characteristics: 1) dot-like spectrosomes, 2) positivity for Dad-lacZ, and 3) absence of bamP-GFP expression. This combination of traits is sufficient to classify them as GSC-like cells. Neither Sex lethal nor Nanos is expressed exclusively in GSCs (Chau et al., 2009; Li et al., 2009), rendering them unsuitable for distinguishing GSC-like from cystoblast-like cells.

(4) All experiments except Fig. 1I (where a single germarium with no quantification) were performed with nos-Gal4, UASp-flp. Have the authors performed any of the phenotypic characterizations (i.e., figures other than figure 1) with hs-flp?

In the rebuttal, the authors stated that they used nos>flp for all figures except for Fig. 1I. It would be more convincing for them to prove in Fig. 1 than there is not phenoytpic difference between the two methods and then switch to the nos>FLP method for the rest of the paper.

We appreciate this suggestion. These data are included in Figure 1-figure supplement 3 in the revised manuscript.

(5) Does the number of SGCs change with the age of the female? The experiments were all performed in 14-day old adult females. What happens when they look at young female (like 2-day old). I assume that the nos>flp is working in larval and pupal stages and so the phenotype should be present in young females. Why did the authors choose this later age? For example, is the phenotype more robust in older females? or do you see more SGCs at later time points?

The authors did not supply any data to prove that the clones were larger in 14-day-old flies than in younger flies. Additionally, the age of "younger" flies was not specified. Therefore, the authors did not satisfactorily answer my concern.

We appreciate this critical comment. Figure 1J includes the SGC phenotype data from 1-, 7-, and 14-day-old flies. Both 1- and 7-day-old flies are younger flies in our analyses. The evidence that germline clones were larger in 14-day-old flies than in younger flies was provided in Figure 1-figure supplement 2 in the revised manuscript.

(6) Can the authors distinguish one copy of GFP versus 2 copies of GFP in germ cells of the ovary? This is not possible in the Drosophila testis. I ask because this could impact on the clonal analyses diagrammed in Fig. 4A and 4G and in 6A and B. Additionally, in most of the figures, the GFP is saturated so it is not possible to discern one vs two copies of GFP.

In the rebuttal, the authors stated that they cannot differential one vs two copies of GFP. They used other clone labeling methods in Fig. 4 and 6. I think that the authors should make a statement in the manuscript that they cannot distinguish one vs two copies of GFP for the record.

Thank you for this suggestion. Such statement has been added in the revised manuscript (Lines 177-178).

(7) More evidence is needed to support the claim of elevated Dpp levels in bam or bgcn mutant tumors. The current results with dpp-lacZ enhancer trap in Fig 5A,B are not convincing. First, why is the dpp-lacZ so much brighter in the mosaic analysis (A) than in the no-clone analysis (B); it is expected that the level of dpp-lacZ in cap cells should be invariant between ovaries and yet LacZ is very faint in Fig. 5B. I think that if the settings in A matched those in B, the apparent expression of dpp-lacZ in the tumor would be much lower and likely not statistically significantly. Second, they should use RNA in situ hybridization with a sensitive technique like hybridization chain reactions (HCR) - an approach that has worked well in numerous Drosophila tissues including the ovary.

The HCR FISH in Fig.5 of the revised manuscript needs an explanation for how the mRNA puncta were quantified. Currently, there is no information in the methods. What is meant but relative dpp levels. I think that the authors should report in and unbiased manner "number" of dpp or gbb puncta in TFs. For the germaria, I think that they should report the number of puncta of dpp or gbb divide by the total area in square pixels counted. Additionally, the background fluorescence is noticeably much higher in bamBG/delta86 germaria, which would (falsely) increase the relative intensity of dpp and gbb in bam mutants. Although, I commend the authors for performing HCR FISH, these data are still not convincing to me.

We appreciate these critical comments. Due to variable puncta sizes and frequent clustering in TF and cap cells (see Figure 5A, C), direct quantification of puncta number was unreliable. Therefore, we quantified mean fluorescence intensity instead, as described in the revised figure legend of Figure 5 (Lines 603-604). In Author response image 1 1A, B (modified from Figure 5A, C) , magenta ovals indicate empty background fluorescence areas, which appear similar between w¹¹¹⁸ (wild-type control) and bam^-/- germaria. In Author response image 1, the yellow oval outlines a neighboring germarium, not an empty area (see the DAPI channel).

Author response image 1.

In situ-HCR results of dpp and gbb in wild-type and bam mutant germaria. Magenta ovals indicate empty areas displaying only background fluorescence. In panel (B), the yellow oval outlines a neighboring germarium, not an empty area (see the DAPI channel below).

(8) In Fig 6, the authors report results obtained with the bamBG allele. Do they obtain similar data with another bam allele (i.e., bamdelta86)?

The authors did not try any experiments with the bamdelta86 allele, despite this allele being molecularly defined, where the bamBG allele is not defined.

While we agree that repeating the experiments in Figure 6 with bam^Δ86 would be helpful, our mosaic analysis strategy for two genes on different chromosome arms is technically complex (see genotypes in Source data 1). Switching from bam^BG to bam^Δ86 would necessitate extensive and time-consuming genetic recombination. Given that both alleles induce the SGC phenotype indistinguishably (Figure 1J), we believe that repeating these experiments with bam^Δ86 would not alter our key conclusion. We appreciate your understanding regarding this technical complexity.

Reviewer #2 (Public review):

In the current version, Zhang et al. have made substantial improvements to the manuscript. It is now easier to read, and the data are more solid compared with the previous version, supporting their conclusion that tumor GSCs secrete stemness factors (BMPs and Dpp) to suppress the differentiation of neighboring wild-type GSCs. This study should benefit a broad readership across developmental biology, germ cell biology, stem cell biology, and cancer biology.

Thank you for your valuable comments, and we greatly appreciate them.

However, the following suggestions may further improve the clarity and rigor of the research content:

(1) Clarification of sample size (n).

Each germarium can contain highly variable numbers of SGCs, sometimes reaching 50-100. When reporting "n" values, the authors are encouraged to also indicate the number of germaria analyzed. For example, in lines 126-128:

"Notably, 74% of SGCs (n = 132) were GFP-negative, while the remaining 26% were GFP-positive (Figure 2B, C). This suggests that SGCs can be categorized into two distinct groups: those resembling GSCs (GSC-like) and those resembling cystoblasts (cystoblast-like)." Please clarify how many germaria were examined to obtain n = 132.

We appreciate this comment. In 14-day-old fly ovaries, each germarium that met our criterion for quantifying the SGC phenotype contains approximately 1.5 SGCs (see Figure 1K). For the specific analysis of the “132” SGCs presented in Figure 2C, we did not record the number of germaria from which they originated.

In addition, it is unclear whether the authors intend to suggest that the GFP-negative SGCs are GSC-like or cystoblast-like; this point should be clarified.

Thank you for this suggestion. We intend to suggest that the bamP-GFP-negative SGCs are GSC-like, which information has been added in the revised manuscript (Line 129).

(2) Improvement of Fig. 6 in situ hybridization images.

The in situ hybridization images in Fig. 6 are not fully convincing. The control images, in particular, would benefit from higher resolution and enlarged views of the germarium region.

Thank you for this valuable suggestion. The enlarged views of both the control and bam^-/- germarium regions were included in Figure 5A, C in the revised manuscript.

In panel C, abundant signals are also present outside the germarium, which may complicate interpretation and should be clarified or controlled for.

In the right panel of Figure 5C, the abundant signals noted by the reviewer originate from neighboring germaria (see the DAPI channel), not from empty areas, which would be expected to show only background fluorescence. For more details, please refer to our response to Question (7) raised by Reviewer #1.

Alternatively, the authors could strengthen the in situ analysis by using bam mutants or bam dpp / bam gbb double mutants as controls to better define signal specificity.

We appreciate this comment. Homozygous dpp or gbb mutants are lethal, precluding the generation of dpp bam or gbb bam double-mutant flies. Additionally, the GFP signal was drastically reduced during our HCR processing, preventing mosaic clone analysis.

Reviewer #3 (Public review):

Zhang et al. investigated how germline tumors influence the development of neighboring wild-type (WT) germline stem cells (GSC) in the Drosophila ovary. They report that germline tumors generated by differentiation-arrested mutations (bam and bgcn) inhibit the differentiation of neighboring WT GSCs by arresting them in an undifferentiated state, resulting from reduced expression of the differentiation-promoting factor Bam. They find that these tumor cells produce low levels of the niche-associated signaling molecules Dpp and Gbb, which suppress bam expression and consequently inhibit the differentiation of neighboring WT GSCs non-cell-autonomously. Based on these findings, the authors propose that germline tumors mimic the niche to suppress the differentiation of the neighboring wild-type germline stem cells.

Strengths:

The study uses a well-established in vivo model to address an important biological question concerning the interaction between germline tumor cells and wild-type (WT) germline stem cells in the Drosophila ovary. If the findings are substantiated, this study could provide valuable insights that are applicable to other stem cell systems.

Thank you for your valuable comments, and we greatly appreciate them.

Weaknesses:

The authors have addressed some of my concerns in the revised submission. However, the data presented do not allow the authors to distinguish whether the failed differentiation of WT stem cells/germline cells results from "arrested differentiation due to the loss of the differentiation niche" or from "direct inhibition by tumor-derived expression of niche-associated molecules Dpp and Gbb".

Blocking Dpp or Gbb secretion specifically from germline tumor cells promoted differentiation of neighboring wild-type germ cells (Figure 6). This indicates that BMP ligands secreted by germline tumors are required to inhibit this differentiation. However, we cannot rule out the possibility that disruption of the differentiation niche also contributes to the SGC phenotype, a point highlighted in the manuscript (Line 204).

The critical supporting data, HCR in situ results, are not sufficiently convincing.

Below, we provide a point-by-point reply addressing each of your specific recommendations.

Recommendations for the authors:

Reviewer #3 (Recommendations for the authors):

It's a surprising that the authors failed to induce germline tumors at the adult stage, as this has been reported by many labs and would allow for time course analysis of SGC phenotype. As a result, the data in this manuscript address only events occurring after the germline tumor formation (with clonal induction at larval stage) and and focus on the already presene "arrested wild-type germ cells", without providing insight into the process of by which these arrested germ cells are formed.

In our hands, inducing germline clones by the hs-FLP method at the adult stage was efficient in males but not in females, despite subjecting adult flies to intensive heat-shock at 37°C.

The HCR in situ data exhibit a high background.

Regarding the background issue, please see our response to Reviewer #1’s Question (7).

First, the signal appears stronger in TF cells than in cap cells.

As demonstrated by Li et al. (Li et al., 2016), dpp-lacZ (P4-lacZ) signals are also stronger in TF cells than in cap cells (see their Figure 4D').

Second, both dpp and gbb are detected broadly in somatic cells including escort cells. These observations are inconsistent with published data.

As shown in Figure 5A and C, dpp and gbb were detected broadly in somatic cells of bam^-/- germaria, but not in those of w¹¹¹⁸ (wild-type) controls. To our knowledge, no previous study has reported the expression pattern of these ligands in a bam mutant background.

To demonstrate the tumor-derived dpp and gbb, the HCR in situ analysis could be performed in the germarium with mosaic clones. If these niche-associated molecules are indeed expressed in tumor cells, the authors should observe a mosaic expression pattern of these molecules, with signal "ON" in tumor cells and "OFF" in neighbouring arrested germ cells.

This is a great idea and was indeed our original approach. However, GFP signal was drastically reduced during our HCR processing, ultimately precluding mosaic clone analysis.

References

Chau, J., Kulnane, L.S., and Salz, H.K. (2009). Sex-lethal facilitates the transition from germline stem cell to committed daughter cell in the Drosophila ovary. Genetics 182, 121-132.

Li, X., Yang, F., Chen, H., Deng, B., Li, X., and Xi, R. (2016). Control of germline stem cell differentiation by Polycomb and Trithorax group genes in the niche microenvironment. Development 143, 3449-3458.

Li, Y., Minor, N.T., Park, J.K., McKearin, D.M., and Maines, J.Z. (2009). Bam and Bgcn antagonize Nanos-dependent germ-line stem cell maintenance. Proc Natl Acad Sci U S A 106, 9304-9309.

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potentially unique contribution to local field potentials (LFPs). The biophysical foundations of transmembrane electric dipoles, and the associated argument points, are generally compelling. Experimental constraints on gap junctions and strictly quantitative matching between chemical vs. junctional inputs have been hard to achieve. This computational investigation thus offers a specific way to enhance conceptual understanding and provides interesting testable predictions, which would be of great interest to experimental neurophysiologists who interpret relevant recordings.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

Strengths:

Novelty and Scope:

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience.<br /> It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

Methodological Rigor:

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance.<br /> Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

Biological Relevance:

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses.<br /> The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

Clarity and Depth:

The manuscript is well-structured, with logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

Comments on revised version:

The authors have addressed all my concerns in the revised version of the manuscript.

Reviewer #2 (Public review):

Summary:

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to differences in the amplitudes and time courses of the input currents). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

Strengths:

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

Weaknesses:

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known, and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S4, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies, although it is closely related to models used in earlier modeling work from the same laboratory. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

Comments on revised version:

The authors made mainly cosmetic changes in the manuscript (primarily by adding more discussion), and most of these do not affect my earlier assessment. I have updated my Public Review in a few places to reflect those few changes that substantially address my previous concerns.

Author response:

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

eLife Assessment:

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence can be more convincing if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

We gratefully thank the editors and the reviewers for the time and effort in rigorously assessing our manuscript, for the constructive review process, for their enthusiastic responses to our study, and for the encouraging and thoughtful comments. We especially thank you for deeming our study to be a valuable exploration on the differential contributions of active dendritic gap junctions vs. chemical synapses to local field potentials. We thank you for your appreciation of the quantitative biophysical demonstration on the differences in electric dipoles that appear in extracellular potentials with gap junctions vs. chemical synapses.

However, we are surprised by aspects of the assessment that resulted in deeming the approach incomplete, especially given the following with specific reference to the points raised:

(1) Testing against experiments: With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

In addition, the complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

Together, we emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials.

(2) Model validation: The model used in this study was adopted from a physiologically validated model from our laboratory (Roy & Narayanan, 2021). Please note that the original model was validated against several physiological measurements along the somatodendritic axis. We sincerely regret our oversight in not mentioning clearly that we have used an existing, thoroughly physiologically-validated model from our laboratory in this study.

(3) Comparisons between chemical vs. junctional inputs: We had taken elaborate precautions in our experimental design to match the intracellular electrophysiological signatures with reference to synchronous as well as oscillatory inputs, irrespective of whether inputs arrived through gap junctions or chemical synapses. A new Supplementary Figure S3 has been added to address this concern raised by the reviewers.

In the revised manuscript, we have addressed all the concerns raised by the reviewers in detail. We have provided point-by-point responses to reviewers’ helpful and constructive comments below. We thank the editors and the reviewers for this constructive review process, which helped us in improving our manuscript with specific reference to emphasizing the novelty of our approach and conclusions. The specific changes incorporated into the revised manuscript are detailed below.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

Novelty and Scope

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience. It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

We thank you for the positive comments on the novelty of our approach and how our study addresses an underexplored area in neuroscience. The assumptions about the passive nature of dendritic structures had indeed resulted in an underestimation of the contributions of gap junctions to extracellular potentials. Once the realities of active structures are accounted for, the contributions of gap junctions increases by several orders of magnitude compared to passive structures (Fig. 1D).

Methodological Rigor

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance. Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

We thank you for your encouraging comments on the experimental design and methodological rigor of our approach.

Biological Relevance

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses. The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

We thank you for your positive comments on the biological relevance of our approach. We also gratefully thank you for emphasizing the two striking novelties unveiling the dichotomy between gap junctions and chemical synapses in their contributions to field potentials: polarity differences and spectral characteristics.

Clarity and Depth

The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

We sincerely thank you for the positive comments on the structure and comprehensive coverage of our manuscript encompassing different types of inputs that neurons typically receive.

Weaknesses and Areas for Improvement

Generality and Validation

The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings. Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

We thank you for raising this important point. The prime novelty and the principal conclusion of this study is that gap junctional contributions to extracellular field potentials are orders of magnitude higher when the active nature of cellular compartments are accounted for. The lacuna in the literature has been consequent to the assumption that cellular compartments are passive, resulting in the dogma that gap junctional contributions to field potentials are negligible. Despite knowledge about active dendritic structures for decades now, this assumption has kept studies from understanding or even exploring the contributions of gap junctions to field potentials. The rationale behind the choice of a computational approach to address the lacuna were as follows:

(1) The complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

(2) With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non-specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). 'The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

We highlight the novelty of our approach and of the conclusions about differences in extracellular signatures associated with active-dendritic chemical synapses and gap junctions, against these experimental difficulties. We emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials. Our analyses clearly demonstrates that gap junctions do contribute to extracellular potentials if the active nature of the cellular compartments is explicitly accounted for (Fig. 1D). We also show theoretically well-grounded and mechanistically elucidated differences in polarity (Figs. 1–3) as well as in spectral signatures (Figs. 5–8) of extracellular potentials associated with gap junctional vs. chemical synaptic inputs. Together, our fundamental demonstration in this study is the critical need to account for the active nature of cellular compartments in studying gap junctional contributions of extracellular potentials, with CA1 pyramidal neuronal dendrites used as an exemplar.

In the revised version of the manuscript, we have emphasized the motivations for the approach we took, highlighting the specific novelties both in methodological and conceptual aspects, finally emphasizing the need to account for other cell types and gap junctional contributions therein. Importantly, we have emphasized the non-specificities associated with gap-junctional blockers as the reason why experimental delineation of gap junctional vs. chemical synaptic contributions to LFP becomes tedious. We believe that these points underscore the need for the computational approach that we took to address this important question, apart from the novelties of the study.

In response to your constructive comments, we have added the following to the revised version of the manuscript, in the Introduction section as motivation for the specific route we took:

“Given the complexity arising from the concurrent activity of chemical synapses, gap junctions, and active dendritic conductances across multiple neuronal populations, experimentally isolating the contributions of individual components to extracellular potentials remains highly challenging. To address this limitation, we employed a computational modeling approach, which provides a quantitative framework for systematically dissecting the distinct roles of specific cellular and synaptic elements. This strategy is consistent with previous studies that have successfully used computational methods to elucidate the contributions of active dendritic mechanisms to LFPs (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to LFPs (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). In addition, experimentally isolating the contribution of gap junctions is complicated by non-specific effects of available pharmacological modulators targeting these connections (Behrens et al., 2011; Rouach et al., 2003). Most genetic knockouts of gap junctional proteins are either lethal or trigger functional compensatory mechanisms (Bedner et al., 2012; Lo, 1999), thereby rendering causal attribution of specific gap junctional contributions infeasible with currently available experimental approaches. Consequently, biophysically and morphologically detailed computational modeling provides a crucial means to evaluate the impact of individual neuronal components on extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).”

We thank you for raising this point as this allowed us to expand on the specific motivations for the approach we took, and to present the specific novelties of our study to the analyses of extracellular field potentials. Thank you.

Role of Active Dendritic Currents

The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

We thank you for this constructive suggestion. We agree that it is important to consider the implications for broader network dynamics of the outward HCN currents that are observed with synchronous inputs. In the revised manuscript, we have elaborated on the implications of the outward HCN current to network dynamics in detail. The following paragraph has been added to Discussion subsection on “Outward HCN currents regulate extracellular potentials”:

“HCN channels play a critical role in shaping hippocampal network dynamics by modulating neuronal excitability, oscillatory behavior, and susceptibility to pathological states (Kessi et al., 2022; Magee, 1998; Mishra & Narayanan, 2025; Nolan et al., 2004). The outward-like properties of the HCN current we observed may have specific functional implications at different scales. At the cellular scale, the manifestation of outward current during action potentials or plateau potentials could contribute to after hyperpolarization thereby regulating firing properties. In cortical and hippocampal pyramidal neurons, most single-neuron processing occurs in their elaborate dendritic branches, where there is spatiotemporal summation of different synaptic potentials, plateau potentials, back propagating action potentials, and dendritic spikes (Johnston & Narayanan, 2008; Major et al., 2013; Stuart & Spruston, 2015). Considering the heavy expression of HCN channels in the dendrites of hippocampal and cortical pyramidal neurons (Kole et al., 2006; Lorincz et al., 2002; Magee, 1998; Williams & Stuart, 2000), the back propagating action potentials, plateau potentials, or dendritic spikes at dendritic location could yield outward currents. These outward currents could act as a hyperpolarizing mechanism that suppresses spatiotemporal summation of the different dendritic potentials.

At the network scale, such regulation of dendritic potentials and somatic firing could contribute to overall reduction in firing rates of different neurons in the network. For instance, as inhibitory neurons typically elicit action potentials at higher frequencies, somatic outward HCN currents would occur more frequently in inhibitory neurons that express HCN channels compared to excitatory neurons. However, the heavy expression of HCN channels in the dendrites and the higher prevalence of dendritic spikes and plateau potentials in dendrites (Basak & Narayanan, 2018; Larkum et al., 2022; Moore et al., 2017) imply that the impact on outward HCN currents might be higher. Thus, the presence of outward HCN currents would regulate network balance of excitation inhibition in an activity-dependent manner. Additionally, the outward component of the current through HCN channels could contribute to stabilization of network synchrony by promoting spike phase coherence and to modulation of spike-LFP phase relationships (Das et al., 2017; Ness et al., 2016, 2018; Seenivasan & Narayanan, 2020; Sinha & Narayanan, 2015, 2022).

Together, the outward HCN current could play critical roles in regulating several cellular and network functions including spatiotemporal summation within single neurons, amplitude and phase of different oscillations, excitatory-inhibitory interactions, and rate and temporal coding involved in spatial navigation (Hussaini et al., 2011; Nolan et al., 2004; O'Keefe & Recce, 1993). In the context of brain rhythms, future investigations are needed to explore ripple-frequency oscillations, specifically to assess whether high-frequency network interactions are modulated by HCN outward currents. Importantly, future studies could specifically focus on delineating the prevalence and specific contributions of outward currents through HCN channels to single-neuron and network physiology.”

We thank you for highlighting this point, as it allowed us to elaborate the broader roles of HCN channels to single-cell computation, network dynamics, and field potentials. Thank you.

Analysis of Plasticity

While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

We thank you for this constructive suggestion. Please note that we have presented consistent results for both fewer and more gap junctions in our analyses (Figure 1 with 217 gap junctions and Supplementary Figure 1 with 99 gap junctions). Thus, our fundamentally novel result that gap junctions onto active dendrites differentially shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron. Thus, these results demonstrate that the conclusions about their contributions to LFP are invariant to plasticity in their gap junctional numerosity.

We had only briefly mentioned plasticity in the Introduction to highlight the different modes of synaptic transmission and to emphasize that plasticity has been studied in both chemical synapses and gap junctions, playing a role in learning and adaptation. However, it seems that this wording inadvertently suggested that our study includes plasticity simulations. Therefore, we have removed that sentence from Introduction in the revised manuscript to ensure clarity.

In the ‘Limitations of analyses and future studies’ section in Discussion, we suggested investigating the impact of plasticity mechanisms—specifically, activity-dependent plasticity of ion channels—on synaptic receptors vs. gap junctions and their effects on extracellular field potentials under various input conditions and plasticity combinations across different structures. We fully agree with the reviewer that such studies would offer valuable insights and further enhance the broader relevance of our findings. However, while our study implies this direction, it was not the primary focus of our investigation.

In the revised manuscript, we have also expanded on intrinsic/synaptic plasticity and how they could contribute to LFPs (Sinha & Narayanan, 2015, 2022), while also pointing to simulations with different numbers of gap junction in this context. The following specific changes have been incorporated to the revised manuscript:

Discussion subsection “Limitations of analyses and future directions”

“We demonstrated that the contribution of gap junctions to extracellular field potentials remains consistent regardless of the number of gap junctions. Specifically, we showed that the distinct positive LFP deflections persisted irrespective of their relative density on neurons (Fig. 1 with 217 gap junctions and Supplementary Fig. 1 with 99 gap junctions). Previous studies have quantitatively demonstrated that intrinsic and synaptic plasticity modulate hippocampal LFPs and phase coding (Sinha & Narayanan, 2015, 2022). Future analyses should also assess the impact of activity-dependent plasticity in ion channels (on dendrites, axonal initial segments, and other compartments), in synaptic receptors, and in gap junctions (Andersen et al., 2006; Coulon & Landisman, 2017; Johnston & Narayanan, 2008; Magee & Grienberger, 2020; Mishra & Narayanan, 2021; Neves et al., 2008; O'Brien, 2014; Pereda, 2014; Vaughn & Haas, 2022) on extracellular potentials with various kinds of gap junctional inputs and different combinations of plasticity in various structures. Interactions among different forms of plasticity and how co-dependent plasticity in different components alters extracellular field potentials could provide deeper insights about physiological changes during learning and pathological changes observed in different neurological disorders (Sinha & Narayanan, 2022).”

We thank you for highlighting this as this allowed us to improve on the specific focus of the manuscript and the study. Thank you.

Frequency-Dependent Effects

The study demonstrates that gap junctional inputs suppress highfrequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

We sincerely thank you for these insightful comments that we totally agree with. As it so happens, this manuscript forms the first part of a broader study where we explore the implications of gap junctions to ripple frequency oscillations. The ripple oscillations part of the work was presented as a poster in the Society for Neuroscience (SfN) annual meeting 2024 (Sirmaur & Narayanan, 2024). There, we simulate a neuropil made of hundreds of morphologically realistic neurons to assess the role of different synaptic inputs excitatory, inhibitory, and gap junctional and active dendrites to ripple frequency oscillations. We demonstrate there that the conclusions from single-neuron simulations in this current manuscript extend to a neuropil with several neurons, each receiving excitatory, inhibitory and gap-junctional inputs, especially with reference to high-frequency oscillations. Our network based analyses unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions in conjunction with return-current contributions from active dendrites playing regulatory roles in determining ripple characteristics (Sirmaur & Narayanan, 2024).

Our principal goal in this study, therefore, was to lay the single-neuron foundation for network analyses of the impact of gap junctions on LFPs. We are preparing the network part of the study, with a strong focus on ripple-frequency oscillations, for submission for peer review separately. Please see abstract of our poster presented at the Society for Neuroscience annual meeting 2024 on the topic here: https://tinyurl.com/57ehvsep).

In the revised manuscript, we have mentioned the results from our SfN abstract with reference to network simulations and high-frequency oscillations, while also presenting discussions from other studies on the role of gap junctions in synchrony and LFP oscillations. The following has been added to the revised manuscript under the Discussion subsection “High-frequency LFP power was suppressed with gap junctional inputs”:

“In this context, our analyses lay the foundation for network analyses of the impact of gap junctions on LFPs. The conclusions from the single-neuron simulations in this study extend to a neuropil with several neurons, each receiving synaptic and gap junctional inputs, especially with reference to high-frequency ripple oscillations (Sirmaur & Narayanan, 2024). A neuropil made of hundreds of morphologically realistic pyramidal neurons was used to assess the role of different synaptic inputs excitatory, inhibitory, and gap junctional with different patterns of stimulation and active dendritic contributions to ripple-frequency oscillations. Network-based analyses have unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions, in conjunction with return-current contributions from active dendrites, playing modulatory roles in governing ripple characteristics (Sirmaur & Narayanan, 2024). Future studies could expand on these conclusions to explore the implications of frequency-dependent filtering (with reference to gap junctional coupling) on high-frequency extracellular oscillations.”

We thank you for highlighting this point as it allowed us to expand on the implications for our analyses to brain rhythms, especially with reference to high-frequency oscillations. Thank you.

Visualization

Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

We thank you for this constructive suggestion. We used the specific visualization throughout, where we place the outcomes associated with chemical synapses and gap junctions in the same figure, adjacent to each other. We believe that this offers visually intuitive distinction between the outcomes for chemical synapses and gap junctions, rather than placing them in different figures. Splitting them would place the outcomes in different figures and requires turning pages or placing two different figures adjacent to each other for quantitative comparison. We respectfully request that we be allowed to retain this form of visualization in the figures. Thank you.

Contextual Relevance

The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

We sincerely appreciate your valuable suggestion and acknowledge the importance of integrating our results into established neural network dynamics, particularly their implications for gamma rhythms. We have addressed this aspect in the revised version of our manuscript. We have added this to the Discussion subsection on “High-frequency LFP power was suppressed with gap junctional inputs” of the revised manuscript:

“In the context of oscillations and gap-junctional coupling, electrical synapses have been shown to regulate the emergence and stability of the network interactions underlying rhythms of different frequencies, especially gamma-frequency oscillations (Bocian et al., 2009; Buhl et al., 2003; Draguhn et al., 1998; Hormuzdi et al., 2001; Konopacki et al., 2004; LeBeau et al., 2003; Posluszny, 2014; Traub et al., 2003). Specifically, both genetic and pharmacological manipulations of gap junctions have been shown to disrupt gamma rhythms. Genetic deletion of connexin-36 impairs the gamma oscillations associated with awake, active behavioral states (Buhl et al., 2003; Hormuzdi et al., 2001). High-frequency oscillations in the hippocampus have been shown to be sensitive to pharmacological agents like carbenoxolone and octanol that are known to inhibit gap junctions. Carbenoxolone has been known to reduce the transient gamma-frequency oscillations while octanol abolishes the persistent gamma rhythm (Draguhn et al., 1998; Hormuzdi et al., 2001; Posluszny, 2014; Traub et al., 2003). In the context of our results, where we demonstrate that the relative contributions of gap-junctional coupling to high-frequency extracellular potentials is low (Figs. 6–7), how do gap junctions contribute to enhanced extracellular gamma oscillations in these circuits?

It should be noted that in hippocampal circuits, gamma oscillations emerge predominantly due to interactions between inhibitory interneurons through GABAA103046 receptors (Buzsaki & Wang, 2012; Colgin, 2016; Colgin & Moser, 2010; Wang, 2010; Wang & Buzsaki, 1996; Whittington et al., 1995). Thus, the presence of additional gap junctional coupling between these inhibitory neurons allows for tighter synchrony between these reciprocally inhibition-coupled neurons. In other words, the presence of gap junctions increases the probability of action potential generation in other neurons that are electrically coupled to them, together increasing the population of inhibitory neurons that elicit synchronous action potentials. When these synchronous action potentials act on the adjacent cells, both excitatory and inhibitory, the transmembrane GABAA receptor currents yield stronger gamma-frequency oscillations in the extracellular potentials (Draguhn et al., 1998; Hormuzdi et al., 2001; Posluszny, 2014; Traub et al., 2003). Thus, the stronger high-frequency oscillations observed in these scenarios is owing to the enhanced synchrony that is brought about the gap-junctional coupling, which translates to stronger transmembrane inhibitory receptor currents.

These observations also strongly emphasize the utility of the computational approach we took in this study towards discerning the specific roles of gap junctions. Gap junctional coupling have strong physiological roles in terms of enhancing synchronous activity across the neurons that they couple and often express along with other receptors that connect the sets of neurons. Thus, the specific contributions of different neuronal components need to be studied with reference to how they contribute to physiological characteristics vs. their contributions to extracellular potentials. Thus, computational modeling offers an ideal route to understand the specific contributions of different neural-circuit components to extracellular field potentials and rhythms therein (Buzsaki et al., 2012; Einevoll et al., 2019; Einevoll et al., 2013; Sinha & Narayanan, 2022).”

We thank you for highlighting this point as this allowed us to delineate the impact of gap junctions to regulating synchrony across connected neurons vs. modulating field potentials. Thank you.

Reviewer #2 (Public review):

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

We gratefully thank you for the positive comments and the encouraging words about the novel contributions of our study. We are particularly thankful to you for your comment on the generality of our conclusions that hold for different cell types and neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

We sincerely thank you for the positive comments about the readability of the paper.

Weaknesses

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

We thank you for raising this important point. We would like to emphasize that our experimental design and analyses quantitatively account for the spatial distribution and temporal pattern of specific kinds of inputs that arrive through gap junctions and chemical synapses. We submit that our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. We elucidate these points below:

(1) Spatial distribution: The inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were of the same nature: excitatory.

(2) Different numbers of inputs: We have presented consistent results for both fewer and more gap junctions or chemical synapses in our analyses (see Figure 1 with 217 gap junctions or 245 chemical synapses and Supplementary Figure 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

(3) Synchronous inputs (Figs. 1–3): For chemical synapses, the waveforms are in the shape of postsynaptic potentials. For gap junctional inputs, the waveforms are in the shape of postsynaptic potentials or dendritic spikes (to respect the active nature of inputs from the other cell). Here, the electrical response of the postsynaptic cell is identical irrespective of whether inputs arrive through gap junctions or chemical synapses: an action potential. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic and gap junctional inputs. We mechanistically analyzed the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Figs. 2–3). In the revised manuscript, we have shown the intracellular responses to demonstrate that they are electrically matched (new Supplementary Figure 3).

(4) Random inputs (Fig. 4): For random inputs, we did not account for the number of action potentials that arrived, as the only observation we made here was with reference to the biphasic nature of the extracellular potentials with gap junctional inputs in the “No Sodium” scenario. We note that in the “No Sodium” scenario, the time-domain amplitudes were comparable for the field potentials (Fig. 4B, Fig. 4D).

(5) Rhythmic inputs (Fig. 5–8): For rhythmic inputs, please note that the intracellular and extracellular waveforms for every frequency are provided in supplementary figures S5– S11. It may be noted that the intracellular responses are comparable. In simulations for assessing spike-LFP comparison, we tuned the strengths to produce a single spike per cycle, ensuring fair comparison of LFPs with gap junctions vs. chemical synapses.

Taken together, we demonstrate through explicit sets of simulations and analyses that the differences in LFPs were not driven by the strength or patterns of the inputs but rather by the differences in direct transmembrane currents, which are subsequently reflected in the LFPs. In the revised manuscript, we have emphasized these points in the Discussion section, apart from providing intracellular traces for cases where they were not provided before (new Supplementary Figure 3):

Discussion subsection “Dominance of active dendritic currents with LFP associated with gap junctions”

“Our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. A multitude of factors suggests that the observed LFP differences result not from variations in input strength or patterns but rather from differences in direct transmembrane currents, which are subsequently reflected in the LFP signals.

First, the inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were exclusively excitatory in nature.

Second, the results remained consistent regardless of the number of gap junctions or chemical synapses. (Fig. 1 with 217 gap junctions or 245 chemical synapses and Supplementary Fig. 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

Third, for synchronous chemical synaptic inputs, the waveforms resembled typical postsynaptic potentials. Whereas, for gap junctional inputs, the waveforms showed characteristics of postsynaptic potentials or dendritic spikes (accounting the active nature of inputs from the potential presynaptic cells). Electrical response of postsynaptic cell remains identical, producing an action potential regardless of whether inputs arrive via gap junctions or chemical synapses. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic or gap junctional inputs. We mechanistically analyzed the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Fig. 23).

Fourth, for random inputs, the models were not specifically tuned to generate a single action potential. Here, the inputs served as a proxy for asynchronous inputs arriving from other subregions at random times.

Finally, the intracellular responses were comparable for chemical synaptic and gap junctional rhythmic inputs (Supplementary Fig. S5-S11). Here, the model was tuned to elicit a single spike per cycle in simulations evaluating spike-LFP interactions, ensuring a fair comparison between LFPs from gap junctional and chemical synaptic inputs.”

We have added a new Supplementary Figure 3 to the revised manuscript and have referred to this figure in the Results subsection. We thank you for raising these points as it allowed to elaborate on the several novelties and implications of our methodology and conclusions. Thank you.

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

We thank you for raising this important question. Leak channels were among the several contributors to the positive deflection observed in LFPs associated with gap junctions. This effect was present not only in gap junctional models with intact sodium conductance but also in the no-sodium model, where the amplitude of the positive deflection was reduced across other models as well (Fig. 2F, I). Furthermore, even in the absence of leak conductance, a small positive deflection was still observed (Fig. 2F), leading us to further investigate other transmembrane currents over time and across spatial locations, from the proximal to the distal dendritic ends relative to the soma (Fig. 3D). We had observed that the dominant contributor in the case of chemical synapses was the inward synaptic current (Fig. 3A), whereas for gap junctions, the primary contributors were leak conductance along with other outward currents, such as potassium and HCN currents (Fig. 3D). Together, the direct transmembrane component of chemical synapses provides a dominant contribution to extracellular potentials. This dominance translates to differences in the relative contributions of indirect currents (including leak currents) to extracellular potentials associated chemical synaptic vs. gap junctional inputs. Our analyses of the exact ionic mechanisms (Fig. 3) demonstrates the involvement of several ion channels contributing to the indirect component in either scenario.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

We thank you for raising this important point. We agree with the analyses presented by the reviewer on the importance of network simulations and bidirectional gap junctions that respect the voltages in both neurons. However, the complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses “post-synaptic” currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a pointneuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; Martinez-Canada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications becomes essential.

Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations. Thus, the inputs were not pre-determined by “pre” neurons. Instead, the recorded voltages from potential synaptic partner neurons were randomized across locations and scaled using factors at the dendrites before being injected into the target neuron (Supplementary Fig. S1). While incorporating a network of interconnected neurons is indeed important, we utilized biophysical, morphologically realistic CA1 neuron model with different sets of input patterns to model LFPs, which were derived from the total transmembrane currents across all compartments of the multi-compartmental neuron model. Given the complexity of this approach, adding further network-level interactions or pre-post connections would have been computationally demanding.

In the revised manuscript, we have elaborated on the general methodology used in LFP modeling studies to introduce synaptic currents. We have emphasized that our study extends this approach to modeling gap junctional inputs, while also highlighting randomization of locations and the scaling process in assigning gap junctional synaptic strengths. The following paragraphs were specifically added to the revised version of the manuscript:

Methods subsection “Chemical synaptic and gap junctional inputs: Characteristics and temporal dynamics”:

“The complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses post-synaptic currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a point-neuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; MartinezCanada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications become essential.”

“Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations from potential presynaptic sources.”

We thank for you highlighting these points as it allowed us to place our methodology in the specific context of the literature. Thank you.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

We thank you for raising this point and agree with you on every point. Please note that we do not assert that the outward HCN currents are exclusively associated with gap junctional inputs. Rather, our results show that synchronous inputs generate outward HCN currents in both chemical synapses (Fig. 3B; positive/outward HCN currents, except in the no sodium or leak model) and gap junctions (Fig. 3D; positive/outward HCN currents). We emphasized this in the case of gap junctions because, in the absence of inward synaptic currents, HCN (acting as outward currents with synchronous inputs) contributed to the positive deflection observed in the LFPs. While HCN would also contribute in the case of chemical synapses, its effect was negligible due to the presence of large inward synaptic currents. Since LFPs reflect the collective total transmembrane currents, the dominant contributors differ between these two scenarios, which we aimed to highlight. Since HCN exhibited outward currents in our synchronous input simulations, we have elaborated on this mechanism in the supplementary figure (Fig. S3). Our intention was not to emphasize this effect for only one synaptic mode but rather to highlight HCN's contribution to the positive deflection as one of the contributing factors.

We agree that HCN currents are relatively small in magnitude; therefore, our conclusions were based on HCN being one of the several contributing factors. Leak conductance and other outward conductances, including HCN currents (Fig. 3D), collectively contribute to the positive deflections observed in the case of gap junctional synchronous inputs.

In the revised manuscript, we have provided the following clarifications in the Results subsection on” Synchronous inputs: Outward transmembrane currents from active dendrites contribute to positive deflection in extracellular potentials associated with gap junctional inputs”:

“It is important to note that despite their relatively small magnitude, the outward HCN currents (Fig. 3D) substantially contribute to positive extracellular potential deflections associated with gap junctional inputs (Fig. 2), together with leak and other outward conductances.”

“While outward HCN currents (Fig. 3B) are also expected to influence LFPs under chemical synaptic input, their impact was minimal due to the predominance of large inward synaptic currents (Fig. 3A). As LFPs reflect the summation of all transmembrane currents, the dominant contributors vary across different modes of synaptic transmission.”

We thank you for emphasizing this point. This allowed us to expand on the specific roles of HCN channels and potential contributions of the outward nature of the HCN current. We have also expanded the Discussion subsection on “Outward HCN currents regulate extracellular potentials” to elaborate on this aspect as well. Thank you.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

We thank you for raising this important point. The CA1 pyramidal neuronal model used in this study is built with ion-channel models derived from biophysical and electrophysiological recordings from these cells. As mentioned in the Methods section “Dynamics and distribution of active channels” and Supplementary Table S1, models for individual channels, their gating kinetics, and channel distributions across the somatodendritic arbor (wherever known) are all derived from their physiological equivalents. Importantly, these values were derived from previously validated models from the laboratory, which contain these very ion channel models and the exact same morphology (Roy & Narayanan, 2021). Please compare Supplementary Table S1 with Table 1 from (Roy & Narayanan, 2021). Please note that this model was validated against several physiological measurements along the somatodendritic axis (Fig. 1 of (Roy & Narayanan, 2021)).

In the revised manuscript, we have explicitly mentioned this while also mentioning the different physiological properties that were used for the validation process from (Roy & Narayanan, 2021):

Methods subsection “Pyramidal neuron model”

“All parameters and their corresponding values for the neuronal model were derived from previously validated models (Roy & Narayanan, 2021). These CA1 models were validated against several physiological measurements along the somato dendritic axis (Roy & Narayanan, 2021).”

“These channel distributions and the associated parametric values (Supplementary Table S1) were demonstrated to satisfy 22 different somato-dendritic measurements (Roy & Narayanan, 2021). Specifically, six physiological measurements input resistance, maximal impedance amplitude, resonance frequency, resonance strength, total inductive phase, and back-propagating action potential were validated with respective electrophysiological ranges at three somato-dendritic locations (Soma, ~150 µm dendrite, and ~300 µm dendrite) each (6×3=18 measurements). In addition, action potential firing frequency at each of 100 pA, 150 pA, 200 pA, and 250 pA (4 measurements) were also matched in the model to fall within the respective ranges of corresponding electrophysiological measurements. The electrophysiological ranges of intrinsic measurements were derived from respective somato-dendritic recordings (Malik et al., 2016; Narayanan et al., 2010; Narayanan & Johnston, 2007, 2008; Spruston et al., 1995). Together, the CA1 pyramidal model neuron used in this study was tuned to match several electrophysiological characteristics and ion-channel distributions (Roy & Narayanan, 2021).”

We thank you for pointing us to this slip in elaborating on how the model was validated. We have now rectified this. Thank you.

References

Andersen, P., Morris, R., Amaral, D., Bliss, T., & O'Keefe, J. (2006). The hippocampus book. Oxford University Press.

Basak, R., & Narayanan, R. (2018). Spatially dispersed synapses yield sharply-tuned place cell responses through dendritic spike initiation. Journal of Physiology, 596(17), 4173-4205. https://doi.org/10.1113/JP275310

Bedner, P., Steinhauser, C., & Theis, M. (2012). Functional redundancy and compensation among members of gap junction protein families? Biochim Biophys Acta, 1818(8), 1971-1984. https://doi.org/10.1016/j.bbamem.2011.10.016

Behrens, C. J., Ul Haq, R., Liotta, A., Anderson, M. L., & Heinemann, U. (2011). Nonspecific effects of the gap junction blocker mefloquine on fast hippocampal network oscillations in the adult rat in vitro. Neuroscience, 192, 11-19. https://doi.org/10.1016/j.neuroscience.2011.07.015

Bocian, R., Posluszny, A., Kowalczyk, T., Golebiewski, H., & Konopacki, J. (2009). The effect of carbenoxolone on hippocampal formation theta rhythm in rats: in vitro and in vivo approaches. Brain Res Bull, 78(6), 290-298. https://doi.org/10.1016/j.brainresbull.2008.10.005

Buhl, D. L., Harris, K. D., Hormuzdi, S. G., Monyer, H., & Buzsaki, G. (2003). Selective impairment of hippocampal gamma oscillations in connexin-36 knock-out mouse in vivo. J Neurosci, 23(3), 1013-1018. https://doi.org/10.1523/jneurosci.23-03-01013.2003

Buzsaki, G., Anastassiou, C. A., & Koch, C. (2012). The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nat Rev Neurosci, 13(6), 407-420. https://doi.org/10.1038/nrn3241

Buzsaki, G., & Wang, X. J. (2012). Mechanisms of gamma oscillations. Annual Review of Neuroscience, Vol 36, 35, 203-225. https://doi.org/10.1146/annurev-neuro-062111150444

Colgin, L. L. (2016). Rhythms of the hippocampal network. Nat Rev Neurosci, 17(4), 239249. https://doi.org/10.1038/nrn.2016.21

Colgin, L. L., & Moser, E. I. (2010). Gamma oscillations in the hippocampus. Physiology (Bethesda), 25(5), 319-329. https://doi.org/10.1152/physiol.00021.2010

Coulon, P., & Landisman, C. E. (2017). The Potential Role of Gap Junctional Plasticity in the Regulation of State. Neuron, 93(6), 1275-1295. https://doi.org/10.1016/j.neuron.2017.02.041

Das, A., Rathour, R. K., & Narayanan, R. (2017). Strings on a Violin: Location Dependence of Frequency Tuning in Active Dendrites. Front Cell Neurosci, 11, 72. https://doi.org/10.3389/fncel.2017.00072

Draguhn, A., Traub, R. D., Schmitz, D., & Jefferys, J. G. (1998). Electrical coupling underlies high-frequency oscillations in the hippocampus in vitro. Nature, 394(6689), 189-192. https://doi.org/10.1038/28184

Einevoll, G. T., Destexhe, A., Diesmann, M., Grun, S., Jirsa, V., de Kamps, M., Migliore, M., Ness, T. V., Plesser, H. E., & Schurmann, F. (2019). The Scientific Case for Brain Simulations. Neuron, 102(4), 735-744. https://doi.org/10.1016/j.neuron.2019.03.027

Einevoll, G. T., Kayser, C., Logothetis, N. K., & Panzeri, S. (2013). Modelling and analysis of local field potentials for studying the function of cortical circuits. Nat Rev Neurosci, 14(11), 770-785. https://doi.org/10.1038/nrn3599

Gold, C., Henze, D. A., Koch, C., & Buzsaki, G. (2006). On the origin of the extracellular action potential waveform: A modeling study. J Neurophysiol, 95(5), 3113-3128. https://doi.org/10.1152/jn.00979.2005

Hagen, E., Dahmen, D., Stavrinou, M. L., Linden, H., Tetzlaff, T., van Albada, S. J., Grun, S., Diesmann, M., & Einevoll, G. T. (2016). Hybrid Scheme for Modeling Local Field Potentials from Point-Neuron Networks. Cereb Cortex, 26(12), 4461-4496. https://doi.org/10.1093/cercor/bhw237

Halnes, G., Ness, T. V., Næss, S., Hagen, E., Pettersen, K. H., & Einevoll, G. T. (2024). Electric Brain Signals: Foundations and Applications of Biophysical Modeling. Cambridge University Press. https://doi.org/10.1017/9781009039826

Hormuzdi, S. G., Pais, I., LeBeau, F. E., Towers, S. K., Rozov, A., Buhl, E. H., Whittington, M. A., & Monyer, H. (2001). Impaired electrical signaling disrupts gamma frequency oscillations in connexin 36-deficient mice. Neuron, 31(3), 487-495. https://doi.org/10.1016/s0896-6273(01)00387-7

Hussaini, S. A., Kempadoo, K. A., Thuault, S. J., Siegelbaum, S. A., & Kandel, E. R. (2011). Increased size and stability of CA1 and CA3 place fields in HCN1 knockout mice. Neuron, 72(4), 643-653. https://doi.org/10.1016/j.neuron.2011.09.007

Johnston, D., & Narayanan, R. (2008). Active dendrites: colorful wings of the mysterious butterflies. Trends Neurosci, 31(6), 309-316. https://doi.org/10.1016/j.tins.2008.03.004

Kessi, M., Peng, J., Duan, H., He, H., Chen, B., Xiong, J., Wang, Y., Yang, L., Wang, G., Kiprotich, K., Bamgbade, O. A., He, F., & Yin, F. (2022). The Contribution of HCN Channelopathies in Different Epileptic Syndromes, Mechanisms, Modulators, and Potential Treatment Targets: A Systematic Review. Front Mol Neurosci, 15, 807202. https://doi.org/10.3389/fnmol.2022.807202

Kole, M. H., Hallermann, S., & Stuart, G. J. (2006). Single Ih channels in pyramidal neuron dendrites: properties, distribution, and impact on action potential output [Research Support, Non-U.S. Gov't]. J Neurosci, 26(6), 1677-1687. https://doi.org/10.1523/JNEUROSCI.3664-05.2006

Konopacki, J., Kowalczyk, T., & Golebiewski, H. (2004). Electrical coupling underlies theta oscillations recorded in hippocampal formation slices. Brain Res, 1019(1-2), 270-274. https://doi.org/10.1016/j.brainres.2004.05.083

Larkum, M. E., Wu, J., Duverdin, S. A., & Gidon, A. (2022). The Guide to Dendritic Spikes of the Mammalian Cortex In Vitro and In Vivo. Neuroscience, 489, 15-33. https://doi.org/10.1016/j.neuroscience.2022.02.009

LeBeau, F. E., Traub, R. D., Monyer, H., Whittington, M. A., & Buhl, E. H. (2003). The role of electrical signaling via gap junctions in the generation of fast network oscillations. Brain Res Bull, 62(1), 3-13. https://doi.org/10.1016/j.brainresbull.2003.07.004

Lo, C. W. (1999). Genes, gene knockouts, and mutations in the analysis of gap junctions. Dev Genet, 24(1-2), 1-4. https://doi.org/10.1002/(SICI)1520-6408(1999)24:1/2%3C1::AID-DVG1%3E3.0.CO;2-U

Lorincz, A., Notomi, T., Tamas, G., Shigemoto, R., & Nusser, Z. (2002). Polarized and compartment-dependent distribution of HCN1 in pyramidal cell dendrites. Nat Neurosci, 5(11), 1185-1193. https://doi.org/10.1038/nn962

Magee, J. C. (1998). Dendritic hyperpolarization-activated currents modify the integrative properties of hippocampal CA1 pyramidal neurons. J Neurosci, 18(19), 7613-7624. https://doi.org/10.1523/jneurosci.18-19-07613.1998

Magee, J. C., & Grienberger, C. (2020). Synaptic Plasticity Forms and Functions. Annual Review of Neuroscience, Vol 36, 43, 95-117. https://doi.org/10.1146/annurev-neuro090919-022842

Major, G., Larkum, M. E., & Schiller, J. (2013). Active properties of neocortical pyramidal neuron dendrites [Review]. Annual Review of Neuroscience, Vol 36, 36, 1-24. https://doi.org/10.1146/annurev-neuro-062111-150343

Malik, R., Dougherty, K. A., Parikh, K., Byrne, C., & Johnston, D. (2016). Mapping the electrophysiological and morphological properties of CA1 pyramidal neurons along the longitudinal hippocampal axis. Hippocampus, 26(3), 341-361. https://doi.org/10.1002/hipo.22526

Martinez-Canada, P., Ness, T. V., Einevoll, G. T., Fellin, T., & Panzeri, S. (2021). Computation of the electroencephalogram (EEG) from network models of point neurons. PLoS Comput Biol, 17(4), e1008893. https://doi.org/10.1371/journal.pcbi.1008893

Mazzoni, A., Linden, H., Cuntz, H., Lansner, A., Panzeri, S., & Einevoll, G. T. (2015). Computing the Local Field Potential (LFP) from Integrate-and-Fire Network Models. PLoS Comput Biol, 11(12), e1004584. https://doi.org/10.1371/journal.pcbi.1004584

Mishra, P., & Narayanan, R. (2021). Stable continual learning through structured multiscale plasticity manifolds. Curr Opin Neurobiol, 70, 51-63. https://doi.org/10.1016/j.conb.2021.07.009

Mishra, P., & Narayanan, R. (2025). The enigmatic HCN channels: A cellular neurophysiology perspective. Proteins, 93(1), 72-92. https://doi.org/10.1002/prot.26643

Moore, J. J., Ravassard, P. M., Ho, D., Acharya, L., Kees, A. L., Vuong, C., & Mehta, M. R. (2017). Dynamics of cortical dendritic membrane potential and spikes in freely behaving rats. Science, 355(6331). https://doi.org/10.1126/science.aaj1497

Narayanan, R., Dougherty, K. J., & Johnston, D. (2010). Calcium Store Depletion Induces Persistent Perisomatic Increases in the Functional Density of h Channels in Hippocampal Pyramidal Neurons. Neuron, 68(5), 921-935. https://doi.org/10.1016/j.neuron.2010.11.033

Narayanan, R., & Johnston, D. (2007). Long-term potentiation in rat hippocampal neurons is accompanied by spatially widespread changes in intrinsic oscillatory dynamics and excitability. Neuron, 56(6), 1061-1075. https://doi.org/10.1016/j.neuron.2007.10.033

Narayanan, R., & Johnston, D. (2008). The h channel mediates location dependence and plasticity of intrinsic phase response in rat hippocampal neurons. J Neurosci, 28(22), 5846-5860. https://doi.org/10.1523/JNEUROSCI.0835-08.2008

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2016). Active subthreshold dendritic conductances shape the local field potential. Journal of Physiology, 594(13), 38093825. https://doi.org/10.1113/JP272022

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2018). h-Type Membrane Current Shapes the Local Field Potential from Populations of Pyramidal Neurons. J Neurosci, 38(26), 6011-6024. https://doi.org/10.1523/jneurosci.3278-17.2018

Neves, G., Cooke, S. F., & Bliss, T. V. (2008). Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. Nat Rev Neurosci, 9(1), 65-75. https://doi.org/10.1038/nrn2303

Nolan, M. F., Malleret, G., Dudman, J. T., Buhl, D. L., Santoro, B., Gibbs, E., Vronskaya, S., Buzsaki, G., Siegelbaum, S. A., Kandel, E. R., & Morozov, A. (2004). A behavioral role for dendritic integration: HCN1 channels constrain spatial memory and plasticity at inputs to distal dendrites of CA1 pyramidal neurons. Cell, 119(5), 719-732. https://doi.org/10.1016/j.cell.2004.11.020

O'Brien, J. (2014). The ever-changing electrical synapse. Curr Opin Neurobiol, 29, 64-72. https://doi.org/10.1016/j.conb.2014.05.011

O'Keefe, J., & Recce, M. L. (1993). Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus, 3(3), 317-330. https://doi.org/10.1002/hipo.450030307

Pereda, A. E. (2014). Electrical synapses and their functional interactions with chemical synapses. Nat Rev Neurosci, 15(4), 250-263. https://doi.org/10.1038/nrn3708

Posluszny, A. (2014). The contribution of electrical synapses to field potential oscillations in the hippocampal formation. Front Neural Circuits, 8, 32. https://doi.org/10.3389/fncir.2014.00032

Reimann, M. W., Anastassiou, C. A., Perin, R., Hill, S. L., Markram, H., & Koch, C. (2013). A biophysically detailed model of neocortical local field potentials predicts the critical role of active membrane currents. Neuron, 79(2), 375-390. https://doi.org/10.1016/j.neuron.2013.05.023

Rouach, N., Segal, M., Koulakoff, A., Giaume, C., & Avignone, E. (2003). Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions. Journal of Physiology, 553(Pt 3), 729-745. https://doi.org/10.1113/jphysiol.2003.053439

Roy, A., & Narayanan, R. (2021). Spatial information transfer in hippocampal place cells depends on trial-to-trial variability, symmetry of place-field firing, and biophysical heterogeneities. Neural Netw, 142, 636-660. https://doi.org/10.1016/j.neunet.2021.07.026

Schomburg, E. W., Anastassiou, C. A., Buzsaki, G., & Koch, C. (2012). The spiking component of oscillatory extracellular potentials in the rat hippocampus. J Neurosci, 32(34), 11798-11811. https://doi.org/10.1523/JNEUROSCI.0656-12.2012

Seenivasan, P., & Narayanan, R. (2020). Efficient phase coding in hippocampal place cells. Physical Review Research, 2(3), 033393. https://doi.org/10.1103/PhysRevResearch.2.033393

Sinha, M., & Narayanan, R. (2015). HCN channels enhance spike phase coherence and regulate the phase of spikes and LFPs in the theta-frequency range. Proc Natl Acad Sci U S A, 112(17), E2207-2216. https://doi.org/10.1073/pnas.1419017112

Sinha, M., & Narayanan, R. (2022). Active Dendrites and Local Field Potentials: Biophysical Mechanisms and Computational Explorations. Neuroscience, 489, 111-142. https://doi.org/10.1016/j.neuroscience.2021.08.035

Sirmaur, R., & Narayanan, R. (2024). Distinct extracellular signatures of chemical and electrical synapses impinging on active dendrites differentially contribute to ripplefrequency oscillations. Society for Neuroscience annual meeting, Chicago, USA.

Spruston, N., Schiller, Y., Stuart, G., & Sakmann, B. (1995). Activity-dependent action potential invasion and calcium influx into hippocampal CA1 dendrites [Research Support, Non-U.S. Gov't]. Science, 268(5208), 297-300. https://doi.org/10.1126/science.7716524

Stuart, G. J., & Spruston, N. (2015). Dendritic integration: 60 years of progress. Nat Neurosci, 18(12), 1713-1721. https://doi.org/10.1038/nn.4157

Szarka, G., Balogh, M., Tengolics, A. J., Ganczer, A., Volgyi, B., & Kovacs-Oller, T. (2021). The role of gap junctions in cell death and neuromodulation in the retina. Neural Regen Res, 16(10), 1911-1920. https://doi.org/10.4103/1673-5374.308069

Traub, R. D., Cunningham, M. O., Gloveli, T., LeBeau, F. E., Bibbig, A., Buhl, E. H., & Whittington, M. A. (2003). GABA-enhanced collective behavior in neuronal axons underlies persistent gamma-frequency oscillations. Proc Natl Acad Sci U S A, 100(19), 11047-11052. https://doi.org/10.1073/pnas.1934854100

Vaughn, M. J., & Haas, J. S. (2022). On the Diverse Functions of Electrical Synapses. Front Cell Neurosci, 16, 910015. https://doi.org/10.3389/fncel.2022.910015

Wang, X. J. (2010). Neurophysiological and computational principles of cortical rhythms in cognition. Physiol Rev, 90(3), 1195-1268. https://doi.org/10.1152/physrev.00035.2008

Wang, X. J., & Buzsaki, G. (1996). Gamma oscillation by synaptic inhibition in a hippocampal interneuronal network model. J Neurosci, 16(20), 6402-6413. https://doi.org/10.1523/jneurosci.16-20-06402.1996

Whittington, M. A., Traub, R. D., & Jefferys, J. G. (1995). Synchronized oscillations in interneuron networks driven by metabotropic glutamate receptor activation. Nature, 373(6515), 612-615. https://doi.org/10.1038/373612a0

Williams, S. R., & Stuart, G. J. (2000). Site independence of EPSP time course is mediated by dendritic I(h) in neocortical pyramidal neurons [In Vitro]. J Neurophysiol, 83(5), 3177-3182. https://doi.org/10.1152/jn.2000.83.5.3177

eLife Assessment

In this useful paper, the authors present a comprehensive method for the purification of recombinant Snake Venom Metalloproteinases (SVMPs) using the MultiBac expression system, explain the self-activation of the enzymes by Zn2+ incubation, and establish high-throughput screening (HTS) techniques. The authors addressed a key problem: producing a substantial amount of pure and enzymatically active SVMPs required for structural and functional studies. Altogether, this work builds a solid foundation for the large-scale production of active SVMPs for future biochemical and structural characterization as well as for drug discovery, albeit leaving certain caveats about the universal applicability of the described methodology for the production of any recombinant SVMPs.

Reviewer #1 (Public review):

Summary:

The authors Hall et al. establish a purification method for snake venom metalloproteinases (SVMPs). By generating a generic approach to purify this divergent class of recombinant proteins, they enhance the field's accessibility to larger quantity SVMPs with confirmed activity and, for some, characterized kinetics. In some cases, the recombinant protein displayed comparable substrate specificity and substrate recognition compared to the native enzyme, providing convincing evidence of the authors' successful recombinant expression strategy. Beyond describing their route towards protein purification, they further provide evidence for self-activation upon Zn2+ incubation. They further provide initial insights on how to design high throughput screening (HTS) methods for drug discovery and outline future perspectives for the in-depth characterization of these enzyme classes to enable the development of novel biomedical applications.

Strengths:

The study is well presented and structured in a compelling way and the universal applicability of the approach is nicely presented.<br /> The purification strategy results in highly pure protein products, well characterized by size exclusion chromatography, SDS page as well as confirmed by mass spectrometry analysis. Further, a significant portion of the manuscript focuses on enzyme activity, thereby validating function. Particularly convincing is the comparability between recombinant vs. native enzymes; this is successfully exemplified by insulin B digestion. By testing the fluorogenic substrate, the authors provide evidence that their production method of recombinant protein can open up possibilities in HTS. Since their purification method can be applied to three structurally variable SVMP classes, this demonstrates the robust nature of the approach.

Weakness

The product obtained from the purification protocol appears to be a heterogenous mixture of self-activated and intact protein species. The protocol would benefit from improved control over the self-activation process. The authors explain well why they cannot deplete Zn2+ in cell culture or increase the pH to prevent autoactivation during the current purification steps. However, this leads me to the suggestion, if the His tag could be exchanged to a different tag that is less pH sensitive and not dependent on divalent ions (Strep-Tactin XT?) to allow for removal of divalent ions and low pH during purification steps. Another suggestion would be if they could replace the endogenous protease cleavage site in their expression construct design to a TEV protease recognition site, for example, to have more control over activation of the recombinant proteins.

The graphic to explain the universal applicability of the approach, Figure S1, has some mistakes, like duplication of text, an arrow without a meaning and should be revised.

Overall, the authors successfully purified active SVMP proteins of all three structurally diverse classes in high quality and provided convincing evidence throughout the manuscript to support their claims. The described method will be of use for a broader community working with self-activating and cytotoxic proteases.

Comment on the revised version:

I find that the clarity and overall structure of the manuscript have improved. However, the weakness I previously highlighted has neither been addressed experimentally nor convincingly explained. Therefore, the assessment stayed unchanged from my side.

Reviewer #2 (Public review):

Summary:

The aim of the study by Hall et al. was to establish a generic method for production of Snake Venom Metalloproteases (SVMPs). These have been difficult to purify in the mg quantities required for mechanistic biochemical and structural studies.

Strengths:

The authors have successfully applied the MultiBac system and describe with a high level of details, the downstream purification methods applied to purify the SVMP PI, PII and PIII. The paper carefully presents the non-successful approaches taken (such as expression of mature proteins, the use of protease inhibitors, prodomain segments and co-expression of disulfide-isomerases) before establishing the construct and expression conditions required. The authors finally convincingly describe various activity assays to demonstrate the activity of the purified enzymes in a variety of established SVMP assays.

Weaknesses:

Some experiments are difficult to perform with relevant controls (i.e. native SVMP from the venome), but authors have explained this and provided the best possible assessment.

Overall, the data presented demonstrates a very credible path for production of active SVMP for further downstream characterization. The generality of the approach to all SVMP from different snakes remains to be demonstrated by the community, but if generally applicable, the method will enable numerous studies with the aim of either utilizing SVMPS as therapeutic agents or to enable generation of specific anti-venom reagents such as antibodies or small molecule inhibitors.

Comment on the revised version:

I think the manuscript has benefited from the review and the revised version provides more clarity, is more concise and reads significantly better with the preliminary data/experiments moved to the supplements. My overall assessment of the manuscript remains unchanged.

Author response:

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

Reviewer #1 (Public review):

Summary:

The authors Hall et al. establish a purification method for snake venom metalloproteinases (SVMPs). By generating a generic approach to purify this divergent class of recombinant proteins, they enhance the field's accessibility to larger quantities of SVMPs with confirmed activity and, for some, characterized kinetics. In some cases, the recombinant protein displayed comparable substrate specificity and substrate recognition compared to the native enzyme, providing convincing evidence of the authors' successful recombinant expression strategy. Beyond describing their route towards protein purification, they further provide evidence for self-activation upon Zn2+ incubation. They further provide insights on how to design high-throughput screening (HTS) methods for drug discovery and outline future perspectives for the in-depth characterization of these enzyme classes to enable the development of novel biomedical applications.

Strengths:

The study is well-presented and structured in a compelling way. The purification strategy results in highly pure protein products, well characterized by size exclusion chromatography, SDS page as well as confirmed by mass spectrometry analysis. Further, a significant portion of the manuscript focuses on enzyme activity, thereby validating function. Particularly convincing is the comparability between recombinant vs. native enzymes; this is successfully exemplified by insulin B digestion. By testing the fluorogenic substrate, the authors provide evidence that their production method of recombinant protein can open up possibilities in HTS. Since their purification method can be applied to three structurally variable SVMP classes, this demonstrates the robust nature of the approach.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The universal applicability of the approach could be emphasized more clearly. The potential for this generic protocol for recombinant SVMP zymogen production to be adapted to other SVMPs is somewhat obscured by the detailed optimization steps. A general schematic overview would strengthen the manuscript, presented as a final model, to illustrate how this strategy can be extended to other targets with similar features. Such a schematic might, for example, outline the propeptide fusion design, including its tags, relevant optimizations during expression, lysis, purification (e.g., strategies for metal ion removal and maintenance of protease inactivity), as well as the controllable auto-activation.

In the revised version of the manuscript, we moved the detailed description of the optimisation of SVMP expression, including mature SVMP expression, Marimastat addition, active site mutations and fusion of propeptides, into the supplement as supplementary text. We hope this improves the clarity and flow. As suggested, we now include a new figure outlining the SVMP production strategy and optimisation steps in the revised manuscript (new Figure S1).

The product obtained from the purification protocol appears to be a heterogeneous mixture of selfactivated and intact protein species. The protocol would benefit from improved control over the selfactivation process. The Methods section does not indicate whether residual metal ions were attempted to be removed during the purification, which could influence premature activation.

We agree that improved control of self-activation would be desirable. However, there is an issue: Previous studies reported that (1) SVMP zymogens are processed within secretory cells of the venom gland (Portes-Junior et al., 2014), and (2) mature SVMPs accumulate in secretory vesicles during venom production (Carneiro et al., 2002). Accordingly, preventing the auto-processing of SVMP zymogens is difficult to achieve because this would require Zn²⁺ depletion within the insect cells during production which would result in cytotoxicity. We have included this information in the updated Discussion section of the revised manuscript.

Additionally, it has not been discussed whether the shift to pH 8 in the purification process is necessary from the initial steps onwards, given that a lower pH would be expected to maintain enzyme latency.

The shift to pH 8 is required for the affinity purification of the SVMP zymogens from the medium, involving the poly-histidine-tag and immobilized metal affinity chromatography (IMAC). At lower pH, the histidines would become protonated, preventing binding of the His-tag to the column. Thus, with the His-tag the shift to pH 7.5 or pH 8 is necessary.

The characterization of PIII activity using the fluorogenic peptide effectively links the project to its broader implications for drug design. However, the absence of comparable solutions for PI and PII classes limits the overall scope and impact of the finding.

We agree that such assays would be extremely useful. However, the development of fluorescence based high-throughput assays to test for PI and PII SVMP activity is beyond the scope of this study. Here, our overarching objective is to report a broadly applicable production method for PI, PII and PIII SVMPs.

Overall, the authors successfully purified active SVMP proteins of all three structurally diverse classes in high quality and provided convincing evidence throughout the manuscript to support their claims. The described method will be of use for a broader community working with self-activating and cytotoxic proteases.

Thank you.

Reviewer #2 (Public review):

Summary:

The aim of the study by Hall et al. was to establish a generic method for the production of Snake Venom Metalloproteases (SVMPs). These have been difficult to purify in the mg quantities required for mechanistic, biochemical, and structural studies.

Strengths:

The authors have successfully applied the MultiBac system and describe with a high level of detail the downstream purification methods applied to purify the SVMP PI, PII, and PIII. The paper carefully presents the non-successful approaches taken (such as expression of mature proteins, the use of protease inhibitors, prodomain segments, and co-expression of disulfide-isomerases) before establishing the construct and expression conditions required. The authors finally convincingly describe various activity assays to demonstrate the activity of the purified enzymes in a variety of established SVMP assays.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The manuscript suffers from a lack of bottoming out and stringent scientific procedures in the methodology and the characterization of the generated enzymes.

As an example, a further characterization of the generated protein fragments in Figure 3 by intact mass spectroscopy would have aided in accurate mass determination rather than relying on SEC elution volumes against a standard. Protein shape and charge can affect migration in SEC.

We agree that intact MS would be useful to determine the mass of the produced SVMPs. In this manuscript, we performed SEC as a purification step, removing aggregates. Furthermore, SEC allowed determining if the SVMPs form monomers or dimers. MS characterisation of intact SVMPs (and their PTMs) is not trivial and beyond the scope of this manuscript (see below).

Also, the analysis of N-linked glycosylation demonstrates some reactivity of PIII to PNGase F, but fails to conclude whether one or more sites are occupied, or whether other types of glycosylation is present. Again, intact mass experiments would have resolved such issues.

We concur that glycosylation of SVMPs is an important question. However, analysing the glycosylation of the SVMPs is beyond the scope of this manuscript; it is actually a project on its own: Intact MS can indeed provide information on glycosylation but is not very precise. Unambiguous assignment of the number and occupancy of glycosylation sites is more challenging, especially for large, glycosylated proteins such as our PIII SVMP zymogen. In practice, confident mapping of glycosylation sites would require peptide-level mass spectrometry following enzymatic digestion (Trypsin and Multi-Enzymatic Limited Digestion, ideally). Sample preparation, method optimization, MS acquisition, and data analysis together would require a significant investment. Moreover, we do not have access to the native PIII SVMP from Echis carinatus sochureki venom - this is the main point of our manuscript: we describe a protocol to produce SVMPs which could not be purified from venom. Therefore, a comparison of the glycosylation of the recombinant SVMP and the native SVMP cannot be performed unfortunately (see below).

The activity assays in Figure 4 are not performed consistently with kinetic assays and degradation assays performed for some, but not all, enzymes, and there is no Echis ocellatus comparison in Figure 4h.

This is correct. The suggested control experiment is not possible for the PII SVMP and PIII SVMP because we cannot purify the native PII and PIII SVMPs from Echis venom. We have highlighted this information in the revised manuscript in the insulin B degradation section.

Overall, whilst not affecting the main conclusion, this leaves the reader with an impression of preliminary data being presented. For consistency, application of the same assays to all enzymes (high-grade purified) would have provided the reader with a fuller picture.

In the revised manuscript, we included new data showing the requested characterisations of all three SVMPs.

We have included the respective assays in Figure 5 and Supplementary Figure S11. In the original manuscript, we had omitted these assays as the data show no enzymatic activity in the respective assays. Specifically, we show that (1) PII does not cause insulin B degradation (Fig. S11b), (2) that the PI and PII SVMPs do not degrade the fluorogenic peptide which is prototypic for PIII SVMPs and MMPs (Fig. S11a), (3) PI and PIII do not cause platelet aggregation because they lack the entire disintegrin domain (PI) or the RGD motif (PIII) (Fig. 5a), and (4) that the PI and PII SVMPs, like the PIII SVMP, are not pro-coagulant and do not cause blood clotting (Fig. 5d,5e and Fig. S11c). We also included this new information in the main text of our revised manuscript.

Overall, the data presented demonstrates a very credible path for the production of active SVMP for further downstream characterization. The generality of the approach to all SVMP from different snakes remains to be demonstrated by the community, but if generally applicable, the method will enable numerous studies with the aim of either utilizing SVMPS as therapeutic agents or to enable the generation of specific anti-venom reagents, such as antibodies or small molecule inhibitors.

Thank you.

Reviewer #3 (Public review):

Summary:

The presented study describes the long journey towards the expression of members' SVMP toxins from snake venom, which are toxins of major importance in a snakebite scenario. As in the past, their functional analysis relied on challenging isolation; the toxins' heterologous expression offers a potential solution to some major obstacles hindering a better understanding of toxin pathophysiology. Through a series of laborious and elegantly crafted experiments, including the reporting of various failed attempts, the authors establish the expression of all three SVMP subtypes and prove their activity in bioassays. The expression is carried out as naturally occurring zymogens that autocleave upon exposure to zinc, which is a novel modus operandi for yielding fusion proteins and sheds also some new light on the potential mechanism that snakes use to activate enzymatic toxins from zymogenic preforms.

Strengths:

The manuscript draws from an extensive portfolio of well-reasoned and hypothesis-driven experiments that lead to a stepwise solution. The wetlands data generated is outstanding, although not all experiments along this rocky road to victory were successful. A major strength of the paper is that, translationally speaking, it opens up novel routes for biodiscovery since a first reliable platform for expression of an understudied, yet potent toxin class is established. The discovered strategy to pursue expression as zymogens could see broad application in venom biotechnology, where several toxin types are pending successful expression. The work further provides better insights into how snake toxins are processed.

We thank the reviewer for their positive assessment of our work.

Weaknesses:

The manuscript contains several chapters reporting failed experiments, which makes it difficult to follow in places.

Based on a similar comment of Reviewer 1, we now moved the ‘failed’ experiments reporting on SVMP expression optimisation to the supplement as new supplementary text. We hope that the revisions have improved the clarity and overall readability of our manuscript.

The reporting of experimental details, especially sample sizes and replicates, could be optimised.

The number of replicates has now been added to the figure legends in the revised manuscript. Detailed experimental information is found in the revised Methods part.

At the time of writing, it remains unclear whether the glycosilations detected at a pIII SVMP could have an impact on the bioactivities measured, which is a major aspect, and future follow-ups should clarify this.

A detailed analysis of glycosylation of the PIII SVMP is beyond the scope of our manuscript (see above, response to Reviewer 2). Our manuscript describes a generic protocol to produce active SVMPs. Importantly, we cannot purify the native PIII SVMP from Echis carinatus sochureki venom. Therefore, it is not possible to compare our PIII SVMP with the native PIII SVMP.

We agree that this is an important question, and we will aim in the future to perform such a comparison of a different insect cell-produced PIII with a native PIII SVMP that can be readily purified from venom.

Finally, the work, albeit of critical importance, would benefit from a more down-to-earth evaluation of its findings, as still various persistent obstacles that need to be overcome.

We consider cytotoxicity to be the principal bottleneck in SVMP production. In this study, we present a strategy to overcome this bottleneck.

Major comments to the manuscript:

(1) Lines 148-149: "indicating that expressing inactivated SVMPs could be a viable, although inefficient, approach". I think this text serves a good purpose to express some thoughts on the nature of how the current draft is set up. It is quite established that various proteases cause extreme viability losses to their expression host (whether due to toxicity, but surely also because of metabolic burden), which is why their expression as inactive fusion proteins is the default strategy in all cases I have thus far seen. I believe that, especially in venom studies, this is of importance given the increased toxicity often targeting cellular integrity, and especially here, because Echis are known to feed on arthropods at younger life history stages, making it very likely that some venom components are especially active against insects and other invertebrates. With that in mind, I would argue that exploring their production in inactive form is the obvious strategy one would come up with and not really the conclusion of a series of (well-conducted and scientifically sound!) experiments. For me, the insight of inactive expression is largely confirmatory of what is established, unless I miss something in the authors' rationale. If yes, it would be important to clarify that in the online version.

We agree that producing zymogens represents a straightforward strategy and now, in hindsight, would have wished we had tested this first thing, it would have saved us and apparently many others significant effort. However, realising this, and implementing this approach took us considerable time and insight as we described in this manuscript. The alternative strategies we describe in the manuscript, in particular the use of inhibitors and active-site mutation, have been successfully applied for recombinant production of diverse enzymes before, including enzymes that are toxic to host cells.

We have revised the manuscript as requested and moved the optimisation of SVMP expression to the Supplement. We hope this improved the clarity, overall readability of the text and thus addressed the reviewer’s comment.

(2) Line 173: Here, Alphafold 3 was used, whereas in previous sections (e.g., line 153, line 210), it was Alphafold 2. I suggest using one release across the manuscript.

Thank you for bringing this to our attention. In the revised version of the manuscript, we clarified that all models were generated using AlphaFold 3.

(3) Line 252-254: I fully agree, the PIII SVMP is glycosylated. Glycosylation is an important mediator of snake venom activity, and several works have described their importance in the field. This raises the question, which glycosylations have been introduced here in the SVMP, and to verify that these are glycosylations that belong to those found in snakes. This is important as insects facilitate thousands of N- and O- O-glycosylations to modulate the activity of their proteome, of which many are specific to insects. If some of these were integrated into the SVMP, this could have an impact on downstream produced bioassays and also antigenicity (the surface would be somewhat different from natural toxins, causing different selection).

We agree that glycosylation is important and warrants a follow-up in the future.

However, most publications we found reported that de-glycosylation has a negative effect on stability and solubility of SVMPs, which is expected to have a knock-on effect on toxin activity (e.g. AndradeSilva et al., 2025; DOI: 10.1021/acs.jproteome.5c00249). It will be difficult to separate the two effects from each other. We found only a few examples where SVMP glycosylation (sialylation and Nglycosylation) modulated proteolytic and haemorrhagic functions, including interaction with substrates such as e.g. fibrinogen (Schluga et al., 2024; https://doi.org/10.3390/toxins16110486; Chen et al., 2008; 10.1111/j.1742-4658.2008.06540.x; Nikai et al., 2000; DOI: 10.1006/abbi.2000.1795. PMID: 10871038). In our manuscript, we show that our PIII SVMP is very cytotoxic and highly active in casein, fibrinogen and ESO10 degradation assays, with a K_M and k_cat/K_M comparing favourably with other SVMPs and MMPs. We are not aware of a specific substrate for this particular PIII SVMP that depends on a distinct glycosylation pattern. Recombinant production of such SVMPs with specific glycosylation pattern requirement would be a challenge in all commonly used expression systems (yeast, plant, insect cells and mammalian cells). In fact, insect cell expression systems could be advantageous in this respect because the Sf21 and High Five (Hi5) lepidopteran cell lines we utilised are well-characterized for their ability to perform posttranslational modifications on complex secreted proteins:

(1) N-Glycan conservation: Both Sf21 and Hi5 cells typically produce N-glycans that are trimmed to a core 'paucimannose' structure (Man3GlcNAc2), often with an alpha1,6-fucosylation. While snakes can produce more complex, sialylated N-glycans, glycomic studies of native venoms (e.g., Bothrops venom) have demonstrated that high-mannose and paucimannose structures are also prevalent in native SVMPs. Therefore, the recombinant glycoforms produced in our system are not 'unnatural' in the snake venom context but rather represent a subset of the native glycan microheterogeneity.

(2) Occupancy vs structure: The critical function of glycosylation in PIII SVMPs is thought to be often structural, facilitating correct folding and protecting the large metalloprotease and disintegrin-like domains from proteolytic degradation. Because Sf21 and Hi5 cells recognize the same Nglycosylation sequon (Asn-X-Ser/Thr) as reptilian cells, the site-occupancy remains consistent with the native protein, preserving the overall topography of the toxin.

(3) Activity and authentic self-processing: We acknowledge that insect-specific alpha1,3-fucosylation can occur in Hi5 cells and is potentially antigenic. As the recombinant SVMPs will be used for binder selections and for testing in silico designed binders, useful binders will be selected based on neutralising activity against venom toxins. Here, our assays focused on auto-activation and proteolytic activity, which is primarily driven by the catalytic Zn²⁺-site and the protein backbone.

As stated above, analysis of glycosylation pattern of the PIII SVMP is a project on its own and beyond the scope of this manuscript.

We have incorporated some of the above information into the discussion section of the revised manuscript to clarify that insect cell glycosylation does not recapitulate the full diversity of SVMP glycosylation observed in native venoms.

(4) General comment for the bioassays: It would be good to specify the replicates again and report the data, including standard deviations.

We included this information in the figure legends.

Discussion:

I think the data generated in the study is very valuable and will be instrumental for pushing the frontiers in SVMP research, but still I would like to see a bit of modesty in their discussion. As I have pointed out above, it is unclear which effect the glycosilations may have (i.e., are the glycosilations found reminiscent of natural ones?), despite their being functionally important. Also, yes, isolation of SVMPs is challenging, but the reality is that their expression is equally challenging, as evidenced by the heaps of presented negative data (with which I have no problems, I think reporting such is actually important). So far, the "generic" protocol has been used to express one member per structural class of Echis SVMP, but no evidence is provided that it would work equally well on other members from taxonomically more distant snakes (e.g., the pIII known from Naja oxiana). It is very likely, but at the time of writing, purely speculative.

We have expressed additional PIII SVMPs from Echis and Daboia species and will report their production and characterisation in due course.

Lastly, the reality is also that the expression in insect cells can only be carried out by highly specialized labs (even in the expression world, as most laboratories work with bacterial or fungal hosts), whereas the isolation can be attempted in most venom labs. That said, production in insect cells also has economic repercussions as it will be very challenging to generate yields that are economically viable versus other systems, which is pivotal because the authors talk about bioprospecting and the toxins used in snakebite agent research.

We thank the reviewer for this perspective on the practicalities of protein expression. However, we respectfully disagree with the characterization of insect cell expression as an inaccessible or economically non-viable platform for toxin research. We offer the following points:

(1) Prevalence and accessibility: Contrary to the suggestion that insect cell expression is restricted to highly specialized labs, the Baculovirus Expression Vector System (BEVS) has become a cornerstone of modern biologics production, structural biology and biochemistry. For instance, our MultiBac system (which is but one of several systems currently widely in use) is utilised by over 1,000 laboratories and institutions, academic and pharma/biotech, worldwide. The maturation of commercially available kits, automated platforms, and standardized protocols has moved this technology into the mainstream, making it a standard tool for any lab requiring high-quality eukaryotic proteins.

(2) Biological necessity: Bacterial (E. coli) and fungal (P. pastoris) systems are widely accessible, however, they appear to be fundamentally incapable of producing functional SVMPs. SVMPs require complex disulfide-bond formation, intricate folding, and N-glycosylation for stability and solubility. Bacterial systems have been widely tried by us and others but typically result in very low expression or misfolded inclusion bodies. Of note, originally, we had invested significant effort to adapt P. pastoris to the production of eukaryotic proteins we are interested in, without success, before moving on to the MultiBac system. The SVMPs that we analysed here are highly cytotoxic, rendering the baculovirus/insect cell system in a way a logical choice given that the cells are no longer 'living' after infection with the baculovirus (but more akin membrane-enveloped bioreactors). Thus, one can make the argument that insect cells represent the most accessible middle ground that provides folding apparatus and necessary post-translational modifications (PTMs) required for biological relevance, and it is possible to produce mg amounts of SVMP proteins per litre cell culture as reported here in our manuscript.

(3) Economic viability and bioprospecting: Regarding the economic argument, we contend that viability in bioprospecting is defined by functional yield rather than simple volume. Producing large quantities of non-functional or misfolded protein in a cheaper system is economically inefficient. Furthermore, for snakebite research, the ability to produce specific, pure isoforms recombinantly without the contamination of other toxic venom components found in native isolations is essential for high-throughput screening and drug design.

(4) Scalability: Historically, insect cell production was seen as expensive, but current bioreactor technology and reduction in consumables and media costs allow for significant scaling. Many therapeutic reagents (vaccines, viral vectors, protein biologics) are produced routinely in baculovirus/insect cells. For the purposes of bioprospecting and lead identification, the yields provided by our Hi5/Sf21 system are sufficient for rigorous downstream bioassays and structural characterization.

Again, I believe the paper is highly important and excellently crafted, but I think especially the discussion should see some refinement to address the drawbacks and to evaluate the paper's findings with more modesty.

Thank you. We included the discussion about glycosylation patterns.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) It is not entirely clear to me if the final constructs are indeed "fusion-proteins" (line 172, 974), in the sense of chimeric proteins. From the current description, it appears that the prodomain is encoded in the same gene rather than fused as a separate domain. Thus, referring to these constructs as fusion proteins may overstate the degree of protein engineering involved in the study.

This is correct. In the revised manuscript, ‘fusion protein’ is only used in the context of the propeptide SVMP fusion construct to avoid confusion.

(2) Figure 2J: It is difficult to assess how much protein is secreted relative to the intracellular amounts. The blot is surely misleading, as the effective protein dilution differs substantially between intracellularly vs. extracellularly. Providing an estimate of the relative dilution of extracellular protein would help clarify the extent of secretion.

We estimate that the SNP and SN fractions are at least 10-times more concentrated than the media fraction. The blot is analytical and not quantitative.

(3) The manuscript appears to use both alphafold 2 and alphafold 3 for structural predictions. Clarification on the choice of the version and its impact on results would improve consistency.

In the revised version of the manuscript, we clarify that all structural models were generated using AlphaFold 3.

(4) Figure S3b and others: a clear description of the antibodies used in the Western blots would be appreciated (including in the methods).

We included this information in the figure legends and a paragraph in the methods section for Western blots in the revised manuscript.

(5) MTT cytotoxicity testing would be more convincing if done in a concentration-dependent manner.

We repeated this assay using different concentrations of SVMPs and show the results as a new Figure 5f in the revised manuscript.

(6) Figure S3c: It could be interesting to show the sequence coverage to get an impression of what part of the protein is there.

We have included this information as Supplementary Figure S4d in the revised manuscript.

Reviewer #2 (Recommendations for the authors):

Overall, the study is presented in a step-by-step manner, and its conclusions are valid.

(1) As suggested in the public review, further characterization of the purified material would be good, for example, by intact mass-spectroscopy to characterize the enzymes in further detail.

Preliminary MALDI-MS analysis (performed in Loic Quinton’s laboratory) of our PIII SVMP revealed a broad and heterogeneous mass distribution, consistent with heterogeneity caused by the presence of multiple glycoforms (which is not unlike the microheterogeneity in native snake venom). However, owing to the inherent limitations of MALDI-MS for the analysis of glycoproteins, our data do not allow determination of the number of occupied N-glycosylation sites or the identification of additional types of glycosylation.

Moreover, the relatively large molecular mass of these proteins (zymogen 70.2 kDa protein only, mature PIII 50.6 kDa protein only) makes analysis by electrospray ionisation mass spectrometry technically challenging.

An MS-based deep analysis of the glycosylation patterns would therefore be a project on its own, and beyond the scope of the present manuscript.

(2) The studies involving PII appear challenging due to low yields and stability of the enzyme and the mentioned self-degradation. Some studies, such as the casein-degradation, would benefit from working with a well-characterized batch of enzymes to ensure, it is not auto-degrading during the experiment.

We believe that the finding that the PII SVMP degrades itself after incubation with Zn²⁺ is an important observation. It is novel to the best of our knowledge. Moreover, the key message of our manuscript is that we can produce and characterise novel SVMPs that cannot be readily purified from venom (and thus are not well characterised).

Besides, there are very few intact PII SVMPs in venom (e.g. Suntravat et al. BMC Molecular Biol 2016); the vast majority cleaves itself into a PI and a disintegrin.

(3) Figure 4h. Degradation of insulin is only shown for recombinant PIII, not the native enzyme, and therefore doesn't convey any information with respect to how well they compare.

We do not have available any native PII and PIII SVMPs for a comparison with the recombinant SVMPs (in our manuscript we show expression of new, uncharacterised SVMPs). We have included the PIII SVMP in the original manuscript to show that the enzyme is active and has a different specificity compared to PI SVMP. In the revised manuscript, we also included the PII SVMP insulin B degradation assay in Supplementary Figure S11b.

(4) Figure 5a. Inconsistent use of enzymes - data for PII is presented (both as mature protein and Zymogen) and compared to PIII, but not PI, as both zymogen and mature protein. The current data presentation is confusing and gives the idea of the manuscript assembled with figures produced during the exploratory phase of the study, and not from subsequent experiments systematically conducted for the purposes of clarity and completeness.

In the revised manuscript, we included the missing enzymatic characterisations in Figure 5 (panel a and e) and Supplementary Figure S11a-c. These data were initially not included because the respective enzymes are inactive in these assays.

(5) The manuscript would benefit from editing to make it more concise. For an early-career reader, it is of interest and utility to follow the thought and experimental processes that led to the successful solution, but there is a risk of losing the reader's interest along the way by going through expression experiments that did not "work" in the typical sense of the word. To this reviewer, there is no added value in a full paragraph around co-expression with disulfide isomerase, as it did not improve the protein yield. A single sentence, "co-expression with PDI did not improve yields," with a reference to a supplemental figure would convey that message.

We have moved the optimisation of SVMP expression to the Supplementary Information, which we hope has improved the clarity and flow of the main text.

We note that the hypothesis that co-expression of protein disulfide isomerases (PDIs) enhances yields of functional SVMPs, given the high expression of PDIs in snake venom gland cells, is well established in the field. While we consider PDIs (and other chaperones) likely to play an important role in SVMP expression, we were unable to demonstrate this effect using the baculovirus-insect cell expression system and hypothesize that efficient insect and/or baculoviral PDIs are already present.

(6) Similarly with N-linked glycosylation, the section needs a headline (line 241) and firming up of a sentence like "and possibly not all of the glycosylation..." which is vague and appears to state that it was not really of interest to pursue this further. My view is that either an experiment is done properly with a stated aim and purpose, interpreted, and then, based on whether the results are of interest to the main story or not, they are included. If N-linked glycosylation is to be included in the manuscript, it should be with a purpose (e.g., N-linked glycosylation affects enzyme activity). As it stands, the message is "there is some N-linked glycosylation" without further explanation, and this generates information without justifying the inclusion hereof.

Please see our reply above regarding an in-depth characterisation of insect cell glycosylation of the recombinant PIII SVMP without access to the native enzyme for comparison. In our revised manuscript, we confirm that the PIII SVMP is glycosylated and that this at least partly accounts for the apparent discrepancy in molecular weight observed in SEC and SDS PAGE. We have modified the text to clarify the purpose of the PNGase deglycosylation experiment.

(7) The manuscript, in its current form, appears to have been copied from a Thesis with very detailed step-by-step logic and description. While this is useful in a scholarly context, a scientific manuscript should be presented more compactly, assuming the readers know basic biochemistry.

We trust that this Reviewer finds the revised version of our manuscript more compact and concise. 

Reviewer #3 (Recommendations for the authors):

(1) Material and Methods plus Figures:

Please report the number of replicates per experiment and how data is presented (means/ medians/ standard deviation/ others), and add error bars to the plots where needed.

In the revised manuscript we have included the number of repeats in the figure legends.

(2) Abstract

Line 4: I would not say that SVMPs are the most potent viper toxins. This place is probably taken by some of the highly neurotoxic PLA2, such as Crotoxin. Nevertheless, SVMPs are surely some of the most important toxins responsible for pathophysiological effects stemming from viper envenoming, but I would suggest rephrasing for accuracy.

In the revised manuscript, we have modified this sentence.

(3) Introduction

Lines 27-31: I would like to see a reference supporting the existence of all SVMP types across vipers.

We have included references supporting the existence of PI, PII and PIII SVMPs in viper venom. We also rewrote the sentence to state that “representatives of all three sub-classes are present in different viper venoms.” This clarifies that we do not say that all classes are present in all venoms.

Lines 59-60: I am not sure if this should be considered such an important impediment. Essentially, many vipers yield double- to triple-digit mg amounts of crude venom per specimen from only a single milking.

We have rewritten this text in the revised manuscript.

Currently, it is not possible to purify any given SVMP of interest from venom; in particular for E. ocellatus SVMP isoform mixtures are typically purified rather than individual enzymes (see also introduction section of our manuscript line 57ff). Also, many SVMPs are not present in sufficient amounts in the venom. Here, we provide an approach to recombinantly produce any SVMP of interest, independent of its abundance in the venom.

(4) Results

Line 102: The army-fallworms name is Spodoptera, not Spotoptera. Please correct the typo.

Done. Apologies for our oversight.

Line 311: Please provide the data at least as a supplement.

In the revised manuscript, we have included this experiment in Supplementary Figure S6c.

Line 432- 433: It would be useful to clarify whether the protein should have a pro-coagulant activity (or not).

We have changed this sentence as follows in the revised manuscript: This shows that our recombinantly produced SVMPs have no pro-coagulant activity, which was unknown before.

eLife Assessment

This study provides valuable data on the role of Hsd17b7, a gene involved in cholesterol biosynthesis, as a potential regulator of mechanosensory hair cell function. The authors used both zebrafish and the HEI cell line to examine the effects of deletion of Hsd17b7 on hair cell function and survival. While the study presents convincing evidence, the effect sizes observed across several experiments, including functional readouts such as the acoustic startle response, are modest, which raises questions about the biological significance of the proposed mechanism.

Reviewer #1 (Public review):

Summary:

This study identifies HSD17B7 as a cholesterol biosynthesis gene enriched in sensory hair cells, with demonstrated importance for auditory behavior and potential involvement in mechanotransduction. Using zebrafish knockdown and rescue experiments, the authors show that loss of hsd17b7 reduces cholesterol levels and impairs hearing behavior. They also report a heterozygous nonsense variant in a patient with hearing loss. The gene mutation has a complex and somewhat inconsistent phenotype, appearing to mislocalize, reduce mRNA and protein levels, and alter cholesterol distribution, supporting HSD17B7 as a potential deafness gene.

The study presents an interesting deafness candidate with a complex mechanism, and highlights an underexplored role for cholesterol (and lipids) in hair cell function.

The authors were very responsive to the initial reviews, and the manuscript is now significantly stronger.

Strengths:

- HSD17B7 is a new candidate deafness gene with plausible biological relevance.

- Cross-species RNAseq convincingly shows hair-cell enrichment.

- Lipid metabolism, particularly cholesterol homeostasis, is an emerging area of interest in auditory function.

- The connection between cholesterol levels and MET is potentially impactful and, if substantiated, would represent a significant advance.

- The localization of HSD17B7 is reasonably convincing, despite the lack of a KO control: In HEI-OC1 cells, HSD17B7 localizes to the ER, as expected. In mouse hair cells, the staining pattern is cytosolic. The developmental increase between P1 and P4, and the higher expression in OHCs aligns nicely with RNAseq data.

Weaknesses:

- The pathogenic mechanism of the E182STOP variant is unclear: The mutant protein presumably does not affect WT protein localization, arguing against a dominant-negative effect. Yet, overexpression of HSD17B7-E182* alone causes toxicity in zebrafish and it binds and mislocalizes cholesterol in HEI-OC1 cells, suggesting some gain-of-function or toxic effect. In addition, the mRNA of the variant has low expression level, suggesting nonsense-mediated decay. The mechanistic conclusions of the study are therefore not as clear cut as one would had hoped, but it might just be a reflection of real biological complexity.

- The link to human deafness is based on a single heterozygous patient with no syndromic features. Given that nearly all known cholesterol metabolism disorders are syndromic, this raises concerns about causality or specificity. HSD17B7 is therefore, at this point, a candidate deafness gene, and not a fully established "novel deafness gene". This is acknowledged by the authors.

- This study does not directly investigate how reduced cholesterol levels affect MET. However, this is not a significant limitation given the study's scope, and it is reasonable that such detailed functional analyses are left to specialists in hair cell physiology.

Reviewer #2 (Public review):

A summary of what the authors were trying to achieve.

The authors aim to determine whether the gene Hsb17b7 is essential for hair cell function and, if so, to elucidate the underlying mechanism, specifically the HSB17B7 metabolic role in cholesterol biogenesis. They use animal, tissue, or data from zebrafish, mouse, and human patients.

Strengths:

(1) This is the first study of Hsb17b7 in the zebrafish (a previous report identified this gene as a hair cell marker in the mouse utricle).

(2) The authors demonstrate that Hsb17b7 is expressed in hair cells of zebrafish and the mouse cochlea.

(3) In zebrafish larvae, a likely KO of the Hsb17b7 gene causes a mild phenotype in an acoustic/vibrational assay, which also involves a motor response.

(4) In zebrafish larvae, a likely KO of the Hsb17b7 gene causes a mild reduction in lateral line neuromast hair cell number and a mild decrease in the overall mechanotransduction activity of hair cells, assayed with a fluorescent dye entering the mechanotransduction channels.

(5) When HSB17B7 is overexpressed in a cell line, it goes to the ER, and an increase in Cholesterol cytoplasmic puncta is detected. Instead, when a truncated version of HSB17B7 is overexpressed, HSB17B7 forms aggregates that co-localize with cholesterol.

(6) It seems that the level of cholesterol in crista and neuromast hair cells decreases when Hsb17b7 is defective

Comments on the revised version:

Overall, the paper has been improved, but it still needs to be moderated regarding the observed effects and their qualification. I suggest expressing each effect as % {plus minus} SD and indicating it in the main text to inform the reader.

- The title " HSD17B7 is required for the function of sensory hair cells by regulating cholesterol Synthesis" should be moderated: "affects" instead of "required" would be better.

- In the abstract "conserved and essential role for HSD17B7-mediated cholesterol biosynthesis", the term essential seems overstated and premature

- In the discussion: "Collectively, these results suggest that the heterozygous c.544G>T (p.E182*) variant contributes to auditory dysfunction through potential pathogenic mechanisms: haploinsufficiency caused by reduced"...; "could contribute" would be safer.

- In the discussion: "In summary, our study identifies HSD17B7 as a critical regulator of cholesterol synthesis in sensory hair cells and as an essential factor in normal MET and sound-evoked sensory responses. "This part is still an overstatement. The effect in zebrafish is not directly shown to affect hearing, and startle reflex impairment is mild. It is not essential.

Author response:

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

Public Reviews:

Reviewer #1 (Public review):

(1) The pathogenic mechanism of the E182STOP variant is unclear. The mutant protein does not appear to affect WT protein localization, arguing against a dominant-negative effect. Yet, overexpression of HSD17B7-E182* alone causes toxicity in zebrafish and mislocalizes cholesterol in HEI-OC1 cells, suggesting a gain-of-function or toxic effect. In addition, the variant mRNA is expressed at a low level, consistent with nonsense-mediated decay. This apparent complexity and inconsistency need clearer explanation.

We appreciate the reviewer’s careful evaluation of this mechanistic complexity. Based on our combined molecular, cellular, and in vivo data, we propose that the pathogenic effect of the HSD17B7-E182* variant reflects a composite mechanism, rather than a classical dominant-negative effect.

At the transcript level, the E182* variant introduces a premature termination codon and shows markedly reduced mRNA abundance, consistent with partial degradation by nonsense-mediated mRNA decay. This reduction is expected to decrease overall HSD17B7 dosage, contributing a loss-of-function component. Unlike HSD17B7, the truncated HSD17B7^E182* mislocalizes cholesterol in HEI-OC1 cells, and overexpression alone reduces hair cell MET function and startle response in zebrafish embryos. We therefore propose that the truncated protein disturbing local cholesterol homeostasis, thereby exerts a toxic or ectopic gain-of-function.

We have revised the manuscript to clarify the dual-mechanism model.

(2) The link to human deafness is based on a single heterozygous patient with no syndromic features. Given that nearly all known cholesterol metabolism disorders are syndromic, this raises concerns about causality or specificity. The term "novel deafness gene" is premature without additional cases or segregation data.

We thank the reviewer for this important point. We fully agree that, based on a single heterozygous case without segregation data, it is premature to designate HSD17B7 as a novel deafness gene. Therefore, we have revised the manuscript to use the description of "candidate deafness genes".

(3) The localization of HSD17B7 should be clarified better: In HEI-OC1 cells, HSD17B7 localizes to the ER, as expected. In mouse hair cells, the staining pattern is cytosolic and almost perfectly overlaps with the hair cell marker used, Myo7a. This needs to be discussed. Without KO tissue, HSD17B7 antibody specificity remains uncertain.

We thank the reviewer for the constructive comments regarding HSD17B7 localization and antibody specificity.

Regarding subcellular localization, the original Figure 1K was intended to demonstrate the expression of HSD17B7 in mouse hair cells. To address this concern, we performed additional immunostaining on dissected organ of Corti sections at P1, P4, and P7 using higher magnification. Using parvalbumin as a hair cell marker, HSD17B7 displayed a partially punctate intracellular pattern in hair cells (revised Figure 1K). This pattern is consistent with localization to membrane-associated compartments, including the endoplasmic reticulum, and agrees with the ER-associated localization observed in HEI-OC1 cells and zebrafish hair cells. In mature hair cells, ER-associated signals may appear cytosolic and overlap with general hair cell markers such as Myo7a.

Regarding antibody specificity, although HSD17B7 knockout tissue was not available, we performed complementary validation experiments in HEI-OC1 cells. Cells were transfected with pCMV-Flag, pCMV-Flag-hHSD17B7WT, or pCMV-hHSD17B7WT-EGFP constructs and stained with anti-Flag, anti-EGFP, and anti-HSD17B7 antibodies. The HSD17B7 antibody signal showed strong co-localization with both FLAG- and EGFP-tagged HSD17B7 (revised Figure S1A and B), supporting its specificity.

Reviewer #2 (Public review):

(1) The statement that HSD17B7 is "highly" expressed in sensory hair cells in mice and zebrafish seems incorrect for zebrafish:

(a) The data do not support the notion that HSB17B7 is "highly expressed" in zebrafish. Compared to other genes (TMC1, TMIE, and others), the HSB17B7 level of expression in neuromast hair cells is low (Figure 1F), and by extension (Figure 1C), also in all hair cells. This interpretation is in line with the weak detection of an mRNA signal by ISH (Figure 1G I"). On this note, the staining reported in I" does not seem to label the cytoplasm of neuromast hair cells. An antisense probe control, along with a positive control (such as TMC1 or another), is necessary to interpret the ISH signal in the neuromast.

We thank the reviewer for this detailed evaluation and agree that the description of HSD17B7 expression in zebrafish hair cells requires clarification.

To address this, we performed a quantitative comparison of average expression levels within neuromast hair cells using log-normalized single-cell RNA-seq data. This analysis shows that hsd17b7 is expressed at a level comparable to several known MET-associated genes (e.g., tmc1 and lhfpl5a) (revised Figure 1D). Regarding the pseudotime heatmap (Figure 1F), we now state that this analysis illustrates temporal expression dynamics within neuromast hair cell development.

In addition, we have clarified the interpretation of the whole-mount in situ hybridization data by emphasizing that the signal indicates spatial enrichment rather than high transcript abundance.

We have updated the figure panels, legends, and corresponding text in the Results section to reflect these changes.

(b) However, this is correct for mouse cochlear hair cells, based on single-cell RNA-seq published databases and immunostaining performed in the study. However, the specificity of the anti-HSD17B7 antibody used in the study (in immunostaining and western blot) is not demonstrated. Additionally, it stains some supporting cells or nerve terminals. Was that expression expected?

To assess antibody specificity, we performed validation experiments using distinct epitopes. In HEI-OC1 cells transfected with pCMV-Flag-HSD17B7, or pCMV-HSD17B7-EGFP constructs, immunostaining with anti-HSD17B7 showed strong co-localization with both FLAG- and EGFP-tag (revised Figure S1B). In addition, western blot analyses using the same constructs confirmed the specific detection of HSD17B7 protein (revised Figure S1B). These validation data have now been included as supplementary figures in the revised manuscript and provide independent supporting evidence for the specificity of the anti-HSD17B7 antibody.

(2) A previous report showed that HSD17B7 is expressed in mouse vestibular hair cells by single-cell RNAseq and immunostaining in mice, but it is not cited: Spatiotemporal dynamics of inner ear sensory and non-sensory cells revealed by single-cell transcriptomics. Jan TA, Eltawil Y, Ling AH, Chen L, Ellwanger DC, Heller S, Cheng AG. Cell Rep. 2021 Jul 13;36(2):109358. doi: 10.1016/j.celrep.2021.109358.

We have now cited this reference in the revised manuscript.

(3) Overexpressed HSD17B7-EGFP C-terminal fusion in zebrafish hair cells shows a punctiform signal in the soma but apparently does not stain the hair bundles. One limitation is the consequence of the C-terminal EGFP fusion to HSD17B7 on its function, which is not discussed.

We thank the reviewer for raising this important technical point. The apparent absence of an HSD17B7-EGFP signal in hair bundles is primarily due to the imaging strategy and the selection of representative images. In zebrafish hair cells, the EGFP signal within hair bundles is extremely strong. To better visualize the intracellular distribution of HSD17B7 within the hair cell soma, we selected representative confocal optical sections that were focused on the cell body rather than on the apical hair bundle plane. As a result, the hair bundle signal is not visible in the images shown.

Importantly, we agree that C-terminal EGFP fusion may potentially influence protein localization or function. We have therefore revised the Discussion to discuss this limitation and to clarify that our central conclusions regarding HSD17B7 function are primarily supported by loss-of-function analyses, rescue experiments using untagged mRNA, and cholesterol perturbation phenotypes, rather than relying solely on EGFP-tagged overexpression constructs.

(4) A mutant Zebrafish CRISPR was generated, leading to a truncation after the first 96 aa out of the 340 aa total. It is unclear why the gene editing was not done closer to the ATG. This allele may conserve some function, which is not discussed.

Targeting regions close to the ATG is indeed a commonly used strategy for CRISPR-mediated gene disruption. In this study, sgRNA selection was guided by online CRISPR design tools (CRISPRscan), prioritizing predicted cutting efficiency and specificity. This strategy resulted in a frameshift mutation introducing a premature stop codon after amino acid 96 of the 340-aa Hsd17b7 protein.

Importantly, this truncation removes most of the conserved catalytic core required for 17β-hydroxysteroid dehydrogenase activity, including key motifs involved in NAD(P)-binding and substrate recognition. Therefore, although the mutation does not occur immediately adjacent to the ATG, the resulting allele is predicted to lack enzymatic function. We have clarified this rationale and discussed the functional consequences of the truncation in the revised manuscript.

(5) The hsd17b7 mutant allele has a slightly reduced number of genetically labeled hair cells (quantified as a 16% reduction, estimated at 1-2 HC of the 9 HC present per neuromast). On a note, it is unclear what criteria were used to select HC in the picture. Some Brn3C:mGFP positive cells are apparently not included in the quantifications (Figure 2F, Figure 5A).

Upon re-evaluation, we recognized that the original figure annotations were not sufficiently clear and may have led to confusion regarding hair cell selection. In the original images, the absence of dashed outlines around some Brn3c:mGFP⁺ cells may have been misinterpreted as their exclusion from analysis. To address this issue, we have revised Figures 2F and 5A by updating the annotations to ensure that all Brn3c:mGFP⁺ hair cells within each neuromast are clearly visible and unambiguously included (revised Figures 2F and 6A). Corresponding figure legends have also been revised to clarify the criteria used for hair cell identification and quantification.

(6) The authors used FM4-64 staining to evaluate the hair cell mechanotransduction activity indirectly. They found a 40% reduction in labeling intensity in the HCs of the lateral line neuromast. Because the reduction of hair cell number (16%) is inferior to the reduction of FM4-64 staining, the authors argue that it indicates that the defect is primarily affecting the mechanotransduction function rather than the number of HCs. This argument is insufficient. Indeed, a scenario could be that some HC cells died and have been eliminated, while others are also engaged in this path and no longer perform the MET function. The numbers would then match. If single-cell staining can be resolved, one could determine the FM4-64 intensity per cell. It would also be informative to evaluate the potential occurrence of cell death in this mutant. On another note, the current quantification of the FM4-64 fluorescence intensity and its normalization are not described in the methods. More importantly, an independent and more direct experimental assay is needed to confirm this point. For example, using a GCaMP6-T2A-RFP allele for Ca2+ imaging and signal normalization. 

We have revised the FM4-64 quantification strategy. Instead of measuring fluorescence intensity at the neuromast level, FM4-64 uptake was re-quantified at the single hair cell level. Hair cells within each neuromast were identified based on mGFP labeling, and the mean FM4-64 fluorescence intensity was measured for each individual hair cell. The average FM4-64 intensity per hair cell was then calculated for each neuromast and used for group comparisons (revised Figures 2F, 6B, and 8B, Figure S5B). The updated quantification method, normalization procedure, and analysis pipeline have now been described in the revised Methods section.

As supportive evidence, we further analyzed single-cell RNA-seq data from control and hsd17b7 mutant hair cells (revised Figure 3). This analysis revealed dysregulation of multiple genes involved in the MET machinery, including reduced expression of tip-link–associated components and altered expression of other MET-related genes. While these transcriptional changes do not constitute a direct functional assay, they are consistent with perturbation of MET-associated pathways and complement the FM4-64 findings.

(7) The authors used an acoustic startle response to elicit a behavioral response from the larvae and evaluate the "auditory response". They found a significative decrease in the response (movement trajectory, swimming velocity, distance) in the hsd17b7 mutant. The authors conclude that this gene is crucial for the "auditory function in zebrafish".

This is an overstatement:

(a) First, this test is adequate as a screening tool to identify animals that have lost completely the behavioral response to this acoustic and vibrational stimulation, which also involves a motor response. However, additional tests are required to confirm an auditory origin of the defect, such as Auditory Evoked Potential recordings, or for the vestibular function, the Vestibulo-Ocular Reflex. 

We thank the reviewer for highlighting the limitations in interpreting the acoustic startle assay. We have revised the manuscript to avoid overstatement and now describe the observed phenotype as a reduction in the behavioral response to acoustic and vibrational stimulation, rather than concluding a specific impairment of auditory function.

(b) Secondly, the behavioral defects observed in the mutant compared to the control are significantly different, but the differences are slight, contained within the Standard Deviation (20% for velocity, 25% for distance). To this point, the Figure 2 B and C plots are misleading because their y-axis do not start at 0.

We have corrected Figures 2B and 2C so that the y-axes start at zero, thereby providing a more transparent visualization of the behavioral differences. The figure legends have also been revised to clarify the presentation of the data.

(8) Overexpression of HSD17B7 in cell line HEI-OC1 apparently "significantly increases" the intensity of cholesterol-related signal using a genetically encoded fluorescent sensor (D4H-mCherry). However, the description of this quantification (per cell or per surface area) and the normalization of the fluorescent signal are not provided. 

The quantification of the D4H-mCherry signal in HEI-OC1 cells was performed at the single-cell level. Specifically, individual cells were segmented based on morphology, and the mean fluorescence intensity of D4H-mCherry per cell was measured. To account for variability in cell size and imaging conditions, fluorescence intensity was normalized to the background signal measured from cell-free regions in the same field of view. We have now clarified the quantification strategy and normalization procedure in the revised Methods and Results sections.

(9) When this experiment is conducted in vivo in zebrafish, a reduction in the "DH4 relative intensity" is detected (same issue with the absence of a detailed method description). However, as the difference is smaller than the standard deviation, this raises questions about the biological relevance of this result.

We have now clarified the quantification strategy and normalization procedure in the revised Methods and Results sections.

(10) The authors identified a deaf child as a carrier of a nonsense mutation in HSB17B7, which is predicted to terminate the HSB17B7 protein before the transmembrane domain. However, as no genetic linkage is possible, the causality is not demonstrated.

We thank the reviewer for raising this important point. Unfortunately, we were unable to obtain the parents' genetic testing data to perform formal genetic and linkage analysis. To address this limitation, we have revised the manuscript to avoid causal overstatement and now describe the HSD17B7 E182* variant as a candidate pathogenic variant associated with hearing loss. Importantly, our functional analyses in zebrafish and cell-based systems demonstrate that the E182* truncation abolishes key biological activities of HSD17B7, including subcellular localization, cholesterol regulation, mechanotransduction-related activity, and behavioral responses. These convergent functional data provide biological support for the potential pathogenic relevance of this variant.

(11) Previous results obtained from mouse HSD17B7-KO (citation below) are not described in sufficient detail. This is critical because, in this paper, the mouse loss-of-function of HSD17B7 is embryonically lethal, whereas no apparent phenotype was reported in heterozygotes, which are viable and fertile. Therefore, it seems unlikely that heterozygous mice exhibit hearing loss or vestibular defects; however, it would be essential to verify this to support the notion that the truncated allele found in one patient is causal.

Hydroxysteroid (17beta) dehydrogenase 7 activity is essential for fetal de novo cholesterol synthesis and for neuroectodermal survival and cardiovascular differentiation in early mouse embryos.

Jokela H, Rantakari P, Lamminen T, Strauss L, Ola R, Mutka AL, Gylling H, Miettinen T,

Pakarinen P, Sainio K, Poutanen M. Endocrinology. 2010 Apr;151(4):1884-92. doi: 10.1210/en.2009-0928. Epub 2010 Feb 25.

We thank the reviewer for raising this important point. We acknowledge that previous work has shown that complete loss of Hsd17b7 in mice is embryonically lethal, whereas heterozygous animals are viable and fertile (Jokela et al., 2010). Notably, this study primarily focused on embryonic development, cholesterol metabolism, and cardiovascular and neuroectodermal survival, and auditory or vestibular functions were not specifically examined. Therefore, subtle or sensory organ–specific phenotypes in heterozygous mice cannot be excluded.

The human variant identified in this study (E182*) is a nonsense mutation predicted to truncate the HSD17B7 protein prior to the transmembrane and cytoplasmic domains. We therefore present it as a candidate loss-of-function variant, providing supportive human genetic evidence that is consistent with our functional analyses in zebrafish hair cells, rather than as definitive proof of causality. We have revised the manuscript to clarify these points and to acknowledge this limitation.

(12) The authors used this truncated protein in their startle response and FM4-64 assays. First, they show that contrary to the WT version, this truncated form cannot rescue their phenotypes when overexpressed. Secondly, they tested whether this truncated protein could recapitulate the startle reflex and FM4-64 phenotypes of the mutant allele. At the homozygous level (not mentioned by the way), it can apparently do so to a lesser degree than the previous mutant. Again, the differences are within the Standard Deviation of the averages. The authors conclude that this mutation found in humans has a "negative effect" on hearing, which is again not supported by the data. 

We thank the reviewer for this important comment. We agree that the overexpression strategy employed in this study does not fully replicate the endogenous heterozygous state observed in patients, and that the magnitude of the observed effects varies across samples. Accordingly, our experiments were not intended to demonstrate a definitive causal role of the HSD17B7 ^E182* variant in hearing loss.

Instead, the overexpression assays were designed to assess whether the truncated HSD17B7 protein displays abnormal cellular properties and whether its presence can interfere with processes relevant to hair cell function. Under these conditions, HSD17B7^E182* exhibited aberrant subcellular localization, altered intracellular cholesterol distribution, and was associated with reduced FM4-64 uptake and changes in startle-associated behaviors, whereas the wild-type protein did not.

We revised the manuscript to moderate our conclusions. Rather than claim that the E182* mutation has a definitive “negative effect on auditory function,” we now describe it as a functionally compromised allele that disrupts cholesterol distribution and MET-related activity under overexpression conditions, providing mechanistic support consistent with our zebrafish loss-of-function data and the identification of this variant in a patient with hearing loss. In addition, the "negative effect" statement was based on the result that overexpression of the E182* mutation in wild-type embryos caused the compromised MET function and startle response defect.

(13) The authors looked at the distribution of the HSB17B7 in a cell line. The WT version goes to the ER, while the truncated one forms aggregates. An interesting experiment consisted of co-expressing both constructs (Figure S6) to see whether the truncated version would mislocalize the WT version, which could be a mechanism for a dominant phenotype. However, this is not the case.

We thank the reviewer for raising this important point regarding a potential dominant-negative mechanism. Consistent with the reviewer’s interpretation, we found that HSD17B7^WT predominantly localizes to the endoplasmic reticulum, whereas the truncated HSD17B7^E182* protein forms intracellular aggregates. Importantly, we further observed that the E182* mutation markedly reduces the stability of both HSD17B7 mRNA and protein, resulting in substantially decreased abundance of the truncated protein (Figure S6B–E). As a consequence, the cellular levels of HSD17B7^E182* are abnormally low.

Based on these findings, we consider it unlikely that the E182* variant exerts its effect through interference with the wild-type protein. Our results suggest that the heterozygous c.544G>T (p.E182*) variant contributes to auditory dysfunction through potential pathogenic mechanisms: 1, haploinsufficiency caused by reduced HSD17B7 expression, 2, functional impairment due to altered protein subcellular localization and cholesterol distribution.

We have revised the Results and Discussion sections. Our conclusions now emphasize that the functional impact of this variant is attributable to decreased effective HSD17B7 dosage, consistent with the observed defects in cholesterol synthesis, MET-related activity, and auditory-associated phenotypes in our model.

(14) Through mass spectrometry of HSB17B7 proteins in the cell line, they identified a protein involved in ER retention, RER1. By biochemistry and in a cell line, they show that truncated HSB17B7 prevents the interaction with RER1, which would explain the subcellular localization.

Consistent with the reviewer’s interpretation, wild-type HSD17B7 interacts with RER1, a protein known to participate in ER retention, whereas this interaction is lost in the truncated HSD17B7 variant. We propose that RER1 is an interacting partner of HSD17B7, providing a mechanistic explanation for the protein's subcellular localization.

(15) Information and specificity validation of the HSB17B7 antibody are not presented. It seems that it is the same used on mice by IF and on zebrafish by Western. If so, the antibody could be used on zebrafish by IF to localize the endogenous protein (not overexpression as done here). Secondly, the specificity of the antibody should be verified on the mutant allele. That would bring confidence that the staining on the mouse is likely specific.

We thank the reviewer for raising this important point regarding antibody specificity and validation. Information on the HSD17B7 antibody and its validation has been provided in our response to comment 1, where we described the use of antibodies recognizing different epitopes and the experimental strategies employed to assess specificity (revised Figure S1A and B).

Although the same antibody was used for Western blot analysis in zebrafish samples, its performance in immunofluorescence staining of zebrafish tissues was suboptimal, with relatively high background. For this reason, we did not rely on this antibody for endogenous Hsd17b7 localization in zebrafish by immunofluorescence and instead employed tagged constructs for subcellular localization analyses. This approach provides more reliable and interpretable localization information under the current experimental conditions.

Recommendations for the authors:

Reviewing Editor Comments:

Suggested revisions to help improve the study and the eLife Assessment:

(1) FM4-64 uptake: Isolate the effect of hair cell loss and MET reduction.

(2) Clarify the mechanistic model: Is the mutant protein pathogenic due to toxicity, lack of expression or function, or both? Come up with a clearer causal chain of events.

(3) Mouse immunostaining: Validate the HSD17B7 antibody, and since mouse RNAseq data (gEAR database) suggest that HSD17B7 expression increases dramatically between P0-P5, show this developmental progression by immunostaining of the mouse organ of Corti at P0, P3, and P5.

(4) The HSD17B7-E182* expression disrupts cholesterol (D4H staining) in OC1 cells. This should also be demonstrated in the mutant zebrafish.

(5) Structural modeling of E182* is uninformative; half the protein is absent. This kind of analysis is better suited for missense variants. Suggest removing this analysis.

We thank the Reviewing Editor for these constructive suggestions. The major points raised here substantially overlap with the concerns raised in the public reviews. In response, we have:

(1) revised FM4-64 quantification and interpretation to better distinguish hair cell loss from MET impairment;

(2) Clarify the mechanistic mode. Mechanistically, the mutation decreases mRNA abundance and significantly reduces protein levels. Moreover, expression of the p.E182* mutation disrupted the interaction between HSD17B7 and the ER retention receptor RER1, leading to aberrant subcellular localization and altered cholesterol distribution, thereby exacerbating HC dysfunction.

(3) provided additional validation of the HSD17B7 antibody using antibodies targeting distinct epitopes, and extended mouse organ of Corti immunostaining to postnatal stages P1, P4, and P7 to demonstrate the developmental upregulation of HSD17B7 expression;

(4) added in vivo zebrafish experiments demonstrating that expression of HSD17B7^E182* disrupts cholesterol distribution in hair cells, consistent with the effects observed in HEI-OC1 cells using D4H staining;

(5) removed the structural modeling of the E182* variant.

Recommendations for the authors:

The recommendations from Reviewer #1 and Reviewer #2 were carefully considered and addressed. Most of these points overlap with the public reviews and the Reviewing Editor's comments and have been addressed through a revised mechanistic interpretation, additional clarifications in the Methods, more moderate claims regarding auditory function and human genetics, and the removal or revision of potentially misleading analyses. In addition, a number of minor issues were corrected, including missing or incorrect references, repetitive or unclear statements in the Introduction, insufficient methodological details, imprecise terminology, and typographical or formatting errors. Collectively, these revisions improve the clarity, rigor, and transparency of the study without altering its central conclusions.

eLife Assessment

This important study describes computationally designed proteins that bind to the chemokine CCL25. The authors present evidence that some binders simply prevent chemokine binding to the CCR9 receptor, while one binder changes the downstream signaling triggered by chemokine binding. The evidence is solid overall, but some uncertainty remains with respect to functional selectivity due to sensitivity differences between functional assays and the degree of binder selectivity between the large family of chemokine ligands.

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors describe the use of BindCraft computational protein design to create a series of binders to the chemokine CCL25. This chemokine normally mediates CCR9-dependent trafficking of immune cells to the gut, making it a potential target for the treatment of inflammatory bowel disease and related conditions. Importantly, CCL25 also binds a scavenging receptor, ACKR4. The computational protein design approach used does not involve defining particular epitopes to be targeted, allowing a free search for any potential interaction surface.

Among four designs tested, three were predicted to interact at a similar site on the chemokine, while a fourth clone, VUP25111, was predicted to bind to a different site. All four designs showed binding to CCL25, with similar high-nM KD values in all cases. The first three clones showed evidence of direct competition with the receptor for CCL25 binding, while VUP25111 showed incomplete inhibition of binding. In functional assays, all clones acted as antagonists except for VUP25111, which inhibited arrestin recruitment by CCR9, but did not affect G protein activation by CCR9 or arrestin recruitment by ACKR4 (which signals exclusively through arrestin and not G protein).

Strengths:

The work is completed to a high technical standard, and the functional diversity of the clones is intriguing. It is exciting to see pathway-selective modulation of signaling, and this basic paradigm is likely to generalize to other chemokine/receptor systems. The exceptional complexity of chemokine signaling makes this an excellent area to explore the development of modulators that can restrict signaling to a specific subset of receptors.

Weaknesses:

No major weaknesses were noted by this reviewer.

Reviewer #2 (Public review):

This study from de Boer, Lamme, Verdwaald and Schafer describes the de novo AI-guided design of miniproteins that target the chemokine CCL25, with the aim to modulate the activation and signalling of the chemokine receptors CCR9 and ACKR4. The study focuses on characterising four miniproteins that all bind CCL25 with good affinity. Three designs appear to prevent CCL25 binding to both CCR9 and ACKR4, with increasing concentrations of miniproteins resulting in decreased arrestin (both receptors) and mini G protein recruitment (CCR9), less inhibition of forskolin-stimulated cAMP (CCR9), and decreased GRK3 recruitment and receptor internalisation (CCR9). One miniprotein, VUP25111, changes the properties of CCL25 rather than preventing ligand/receptor interactions, resulting in greater selectivity for CCR9 over ACKR4 and a G protein-biased signalling profile (maintenance of mini G protein recruitment, GRK3 recruitment, inhibition of cAMP and receptor internalisation, but loss of arrestin recruitment). VUP25111 also maintained chemotactic migration in MOLT-4 T lymphoblast cells, whereas this response was lost in the presence of the other three miniproteins.

Overall, this is a very interesting application of AI-designed de novo miniproteins to modulate GPCR responses by directly binding the ligand rather than the receptor. This is a conceptually very intriguing approach that could, in principle, be extended to other GPCR systems beyond the chemokine family. The authors deploy an impressive array of assays spanning multiple signalling endpoints, providing a thorough picture of how each miniprotein influences receptor activation and downstream signalling. The presentation of concentration-response relationships for CCL25 alone and in the presence of each miniprotein is particularly informative, and the figures are very well constructed throughout. The inclusion of clear cartoons illustrating the basis of each assay is a nice touch that will help readers from outside the immediate field follow the logic of each experiment.

There are two main conclusions that are not currently as well-supported by the evidence as they might be, and that would benefit from some qualification. The first concerns the selectivity of the miniproteins for CCL25. Testing the impact of the miniproteins on CXCL12 activation of CXCR4 is an important and welcome experiment, but it addresses selectivity against only one other chemokine system, and the current claim of specificity is therefore stronger than the data allow. Additionally, at the highest concentration tested (10 µM), the more potent miniproteins (VUP25101, VUP25107) appear to show some inhibition of arrestin recruitment to CXCR4 - perhaps unsurprising given the degree of structural conservation among chemokines. The statements regarding selectivity and the lack of effect on the CXCL12/CXCR4 system would benefit from revision to more accurately reflect these observations.

The second concern relates to the interpretation of the preserved GRK3 recruitment, but the complete loss of arrestin recruitment observed with VUP25111. In the GRK3 recruitment experiments, 20 nM CCL25 was used, representing an EC40 concentration in this assay. VUP25111 causes a concentration-dependent reduction in CCL25-induced GRK3 recruitment, down to approximately 15% of the maximal response. It is worth considering whether this degree of reduction in GRK3 recruitment could itself be sufficient to disrupt patterns of receptor phosphorylation and thereby prevent observable arrestin recruitment. Both interpretations are complicated by the fact that the GRK3 recruitment and arrestin recruitment assays likely differ in their sensitivity and dynamic windows, making direct quantitative comparisons between them difficult. In the absence of direct measurements of CCR9 phosphorylation in the presence of VUP25111, the alternative interpretation remains open and would benefit from acknowledgement. Given recent work from the same group demonstrating that receptor internalisation is only partially dependent on arrestins (Lamme et al., 2025, J Biol Chem), further discussion of the relationship between GRK and arrestin recruitment and CCR9 internalisation would be of value to the broader GPCR audience this work is likely to attract.

Finally, some additional justification for the use of 20 nM CCL25 across all assays would strengthen the study, as this concentration represents different points on the concentration-response curve depending on the assay and receptor in question. It ranges from an EC40 for CCR9 GRK3 recruitment and internalisation, to an EC50 for CCR9 arrestin and mini-Gi recruitment, an EC80 for CCR9 cAMP inhibition, and an EMax for ACKR4 arrestin recruitment. This has potential consequences for the interpretation and cross-assay comparison of miniprotein potency, and the authors are encouraged to acknowledge and discuss this in the context of their conclusions.

Reviewer #3 (Public review):

Summary:

The authors employed the BindCraft platform to develop binders targeting the chemokine CCL25, a selective activator of the chemokine receptor CCR9. They successfully generated two classes of binders: one that inhibits all CCL25-mediated CCR9 activation, and another that permits CCR9 G protein signaling while simultaneously preventing arrestin recruitment. These tools will enable the dissection of arrestin involvement in regulating cell migration.

My comments, in the order of reading:

(1) Title: I strongly recommend removing the term "biasing" from the title. In this context, it does not convey a specific mechanistic concept. The term "biased signaling" has been used for a very broad range of mechanistically distinct pharmacological phenomena, and without a precise definition, it adds more confusion than clarity. I therefore suggest refraining from using it in the title.

(2) Abstract, line 34: The term "bias" should be replaced. As currently used, it appears to suggest a dichotomy between G protein signaling and arrestin recruitment. However, arrestin recruitment is a consequence of G protein signaling, and it is not conceptually sound to compare a signaling event mediated by one protein family with a protein-protein interaction involving another protein family. A meaningful comparison requires experimental paradigms that differ by a single variable; in this case, there are two - distinct protein families and fundamentally different types of readouts (signaling versus protein-protein interaction).

(3) Abstract, line 34: The term "balanced agonist" should be removed. Any chosen reference ligand is, by definition, the "balanced" agonist for that analysis, regardless of its actual signaling profile. Consequently, the expression "balanced agonist" adds no mechanistic information beyond "the agonist used as reference in a particular bias calculation" and is potentially misleading, as it implies that this ligand possesses a uniquely unbiased, system‑independent signaling profile, which is not the case.

(4) Abstract, line 36: I also recommend removing the term "bias" at this point. The concept of bias typically arises from experiments that quantitatively compare more than one variable. As currently written, the phrasing suggests a dichotomy between G protein- and arrestin-mediated signaling, yet the study does not assess arrestin signaling, only arrestin recruitment. Under these conditions, the use of "bias" is not appropriate. The data are clear and compelling on their own without the need for this potentially misleading terminology.

(5) Introduction: This is interesting to read and generally well written, though certain statements would benefit from improved semantic precision. For example, in lines 110-111, the phrase "G protein-biased complex" should be reconsidered, as it relies on the notion of G protein- versus arrestin-mediated signaling. Arrestins themselves do not signal; what is measured here is their recruitment. Comparing G protein signaling with arrestin recruitment is therefore conceptually unsound, since arrestin engagement is a downstream consequence of G protein activation. Comparisons become meaningful only when designed to differentiate between G protein-dependent and G protein-independent arrestin recruitment, which is not the case in this study.

(6) Results, 122,123: The authors should consider being more precise; possibly, the truncated CCL25 is somewhat less potent on CCR9. The authors should make a statistical test and then decide whether to rephrase or not for enhanced precision.

(7) Figure S5: This figure is currently confusing and needs clarification. The authors state in the main text that CXCR4 is stimulated with CXCL12, yet the figure legend refers to CCL25; this discrepancy should be corrected to ensure consistency. In addition, inhibition of CXCR4 by the miniprotein binders should be analyzed and presented with normalization to CXCR4 responses, not to CCL25-stimulated CCR9. To avoid misinterpretation, inhibition by the miniproteins should be quantified separately for CCR9 and CXCR4, each normalized to its own receptor-specific and functionally equivalent stimulation condition, rather than to the "other" receptor.

(8) Results, lines 211-213: The authors should be more semantically precise. They state that no binder has any effect on arrestin recruitment to CXCR4. If I see the data, this is not really true, as 25101 and 25107 inhibit arrestin recruitment by about 50 % or more at the highest applied concentrations; only 111 and 112 are completely inactive. As already commented, normalization should be done to arrestin recruitment of CXCR4 and not CCR9.

Author response:

We thank the editors and reviewers for thoroughly reviewing our manuscript and offering thoughtful and constructive feedback. We appreciate the positive reception of our work and welcome the opportunity to address the lingering concerns. In the coming revisions, we will be directly addressing the question of the miniprotein’s specificity and increase the precision in the language used to discuss our findings.

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence would be stronger if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

Strengths:

Novelty and Scope:<br /> The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience.<br /> It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

Methodological Rigor:<br /> The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance.<br /> Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

Biological Relevance:<br /> The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses.<br /> The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

Clarity and Depth:<br /> The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

Weaknesses and Areas for Improvement:

Generality and Validation:<br /> The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings.<br /> Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

Role of Active Dendritic Currents:<br /> The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

Analysis of Plasticity:<br /> While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

Frequency-Dependent Effects:<br /> The study demonstrates that gap junctional inputs suppress high-frequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

Visualization:<br /> Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

Contextual Relevance:<br /> The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

Suggestions for Improvement:

Broader Application:<br /> Simulate EFPs in multi-neuron networks to assess how the findings extend to network-level interactions, particularly in regions with mixed synaptic connectivity.

Experimental Correlation:<br /> Collaborate with experimental groups to validate the computational predictions using in vivo or in vitro recordings.

Mechanistic Insights:<br /> Provide a more detailed mechanistic explanation of how specific ionic currents (e.g., HCN, sodium, leak) interact during gap junctional vs. chemical synaptic inputs.

Implications for Neural Coding:<br /> Discuss how the observed differences in EFP signatures might influence neural coding, especially in circuits with heavy gap junctional connectivity.

Reviewer #2 (Public review):

Summary:

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

Strengths:

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

Weaknesses:

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

Author response:

eLife Assessment

This study presents a valuable theoretical exploration on the electrophysiological mechanisms of ionic currents via gap junctions in hippocampal CA1 pyramidal-cell models, and their potential contribution to local field potentials (LFPs) that is different from the contribution of chemical synapses. The biophysical argument regarding electric dipoles appears solid, but the evidence can be more convincing if their predictions are tested against experiments. A shortage of model validation and strictly comparable parameters used in the comparisons between chemical vs. junctional inputs makes the modeling approach incomplete; once strengthened, the finding can be of broad interest to electrophysiologists, who often make recordings from regions of neurons interconnected with gap junctions.

We gratefully thank the editors and the reviewers for the time and effort in rigorously assessing our manuscript, for the constructive review process, for their enthusiastic responses to our study, and for the encouraging and thoughtful comments. We especially thank you for deeming our study to be a valuable exploration on the differential contributions of active dendritic gap junctions vs. chemical synapses to local field potentials. We thank you for your appreciation of the quantitative biophysical demonstration on the differences in electric dipoles that appear in extracellular potentials with gap junctions vs. chemical synapses.

However, we are surprised by aspects of the assessment that resulted in deeming the approach incomplete, especially given the following with specific reference to the points raised:

(1) Testing against experiments: With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established nonspecificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021), reproduced below. In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

In addition, the complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

Together, we emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials.

(2) Model validation: The model used in this study was adopted from a physiologically validated model from our laboratory (Roy & Narayanan, 2021). Please note that the original model was validated against several physiological measurements along the somatodendritic axis. We sincerely regret our oversight in not mentioning clearly that we have used an existing, thoroughly physiologically-validated model from our laboratory in this study.

(3) Comparisons between chemical vs. junctional inputs: We had taken elaborate precautions in our experimental design to match the intracellular electrophysiological signatures with reference to synchronous as well as oscillatory inputs, irrespective of whether inputs arrived through gap junctions or chemical synapses.

In a revised manuscript, we will address all the concerns raised by the reviewers in detail. We have provided point-by-point responses to reviewers’ helpful and constructive comments below. We thank the editors and the reviewers for this constructive review process, which we believe will help us in improving our manuscript with specific reference to emphasizing the novelty of our approach and conclusions.

Reviewer #1 (Public review):

This manuscript makes a significant contribution to the field by exploring the dichotomy between chemical synaptic and gap junctional contributions to extracellular potentials. While the study is comprehensive in its computational approach, adding experimental validation, network-level simulations, and expanded discussion on implications would elevate its impact further.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

Novelty and Scope

The manuscript provides a detailed investigation into the contrasting extracellular field potential (EFP) signatures arising from chemical synapses and gap junctions, an underexplored area in neuroscience. It highlights the critical role of active dendritic processes in shaping EFPs, pushing forward our understanding of how electrical and chemical synapses contribute differently to extracellular signals.

We thank you for the positive comments on the novelty of our approach and how our study addresses an underexplored area in neuroscience. The assumptions about the passive nature of dendritic structures had indeed resulted in an underestimation of the contributions of gap junctions to extracellular potentials. Once the realities of active structures are accounted for, the contributions of gap junctions increases by several orders of magnitude compared to passive structures (Fig. 1D).

Methodological Rigor

The use of morphologically and biophysically realistic computational models for CA1 pyramidal neurons ensures that the findings are grounded in physiological relevance. Systematic analysis of various factors, including the presence of sodium, leak, and HCN channels, offers a clear dissection of how transmembrane currents shape EFPs.

We thank you for your encouraging comments on the experimental design and methodological rigor of our approach.

Biological Relevance

The findings emphasize the importance of incorporating gap junctional inputs in analyses of extracellular signals, which have traditionally focused on chemical synapses. The observed polarity differences and spectral characteristics provide novel insights into how neural computations may differ based on the mode of synaptic input.

We thank you for your positive comments on the biological relevance of our approach. We also gratefully thank you for emphasizing the two striking novelties unveiling the dichotomy between gap junctions and chemical synapses in their contributions to field potentials: polarity differences and spectral characteristics.

Clarity and Depth

The manuscript is well-structured, with a logical progression from synchronous input analyses to asynchronous and rhythmic inputs, ensuring comprehensive coverage of the topic.

We sincerely thank you for the positive comments on the structure and comprehensive coverage of our manuscript encompassing different types of inputs that neurons typically receive.

Weaknesses and Areas for Improvement

Generality and Validation

The study focuses exclusively on CA1 pyramidal neurons. Expanding the analysis to other cell types, such as interneurons or glial cells, would enhance the generalizability of the findings. Experimental validation of the computational predictions is entirely absent. Empirical data correlating the modeled EFPs with actual recordings would strengthen the claims.

We thank you for raising this important point. The prime novelty and the principal conclusion of this study is that gap junctional contributions to extracellular field potentials are orders of magnitude higher when the active nature of cellular compartments are accounted for. The lacuna in the literature has been consequent to the assumption that cellular compartments are passive, resulting in the dogma that gap junctional contributions to field potentials are negligible. Despite knowledge about active dendritic structures for decades now, this assumption has kept studies from understanding or even exploring the contributions of gap junctions to field potentials. The rationale behind the choice of a computational approach to address the lacuna were as follows:

(1) The complex interactions between co-existing chemical synaptic, gap junctional, and active dendritic contributions from several cell-types make the delineation of the contributions of specific components infeasible with experimental approaches. A computational approach is the only quantitative route to specifically delineate the contributions of individual components to extracellular potentials, as seen from studies that have addressed the question of active dendritic contributions to field potentials (Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Sinha & Narayanan, 2015, 2022) or spiking contributions to local field potentials (Buzsaki et al., 2012; Gold et al., 2006; Schomburg et al., 2012). The biophysically and morphologically realistic computational modeling route is therefore invaluable in assessing the impact of individual components to extracellular field potentials (Einevoll et al., 2019; Halnes et al., 2024).

(2) With specific reference to gap junctions, quantitative experimental verification becomes extremely difficult because of the well-established non-specificities associated with gap junctional modulators (Behrens et al., 2011; Rouach et al., 2003). The non-specific actions of gap junctions are tabulated in Table 2 of (Szarka et al., 2021). In addition, genetic knockouts of gap junctional proteins are either lethal or involve functional compensation (Bedner et al., 2012; Lo, 1999), together making causal links to specific gap junctional contributions with currently available techniques infeasible.

We highlight the novelty of our approach and of the conclusions about differences in extracellular signatures associated with active-dendritic chemical synapses and gap junctions, against these experimental difficulties. We emphasize that the computational modeling route is currently the only quantitative methodology to delineate the contributions of gap junctions vs. chemical synapses to extracellular potentials. Our analyses clearly demonstrates that gap junctions do contribute to extracellular potentials if the active nature of the cellular compartments is explicitly accounted for (Fig. 1D). We also show theoretically well-grounded and mechanistically elucidated differences in polarity (Figs. 1–3) as well as in spectral signatures (Figs. 5–8) of extracellular potentials associated with gap junctional vs. chemical synaptic inputs. Together, our fundamental demonstration in this study is the critical need to account for the active nature of cellular compartments in studying gap junctional contributions of extracellular potentials, with CA1 pyramidal neuronal dendrites used as an exemplar.

In a revised version of the manuscript, we will emphasize the motivations for the approach we took, highlighting the specific novelties both in methodological and conceptual aspects, finally emphasizing the need to account for other cell types and gap junctional contributions therein. Importantly, we will emphasize the non-specificities associated with gap-junctional blockers as the reason why experimental delineation of gap junctional vs. chemical synaptic contributions to LFP becomes tedious. We hope that these points will underscore the need for the computational approach that we took to address this important question, apart from the novelties of the manuscript.

Role of Active Dendritic Currents

The paper emphasizes active dendritic currents, particularly the role of HCN channels in generating outward currents under certain conditions. However, further discussion of how this mechanism integrates into broader network dynamics is warranted.

We thank you for this constructive suggestion. We agree that it is important to consider the implications for broader network dynamics of the outward HCN currents that are observed with synchronous inputs. In a revised manuscript, we will elaborate on the implications of the outward HCN current to network dynamics in detail.

Analysis of Plasticity

While the manuscript mentions plasticity in the discussion, there are no simulations that account for activity-dependent changes in synaptic or gap junctional properties. Including such analyses could significantly enhance the relevance of the findings.

We thank you for this constructive suggestion. Please note that we have presented consistent results for both fewer and more gap junctions in our analyses (Figure 1 with 217 gap junctions and Supplementary Figure 1 with 99 gap junctions). Thus, our fundamentally novel result that gap junctions onto active dendrites differentially shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron. Thus, these results demonstrate that the conclusions about their contributions to LFP are invariant to plasticity in their gap junctional numerosity.

We had only briefly mentioned plasticity in the Introduction to highlight the different modes of synaptic transmission and to emphasize that plasticity has been studied in both chemical synapses and gap junctions, playing a role in learning and adaptation. However, if this wording inadvertently suggests that our study includes plasticity simulations, we would remove it from Introduction in the updated manuscript to ensure clarity.

In the ‘Limitations of analyses and future studies’ section in Discussion, we suggested investigating the impact of plasticity mechanisms—specifically, activity-dependent plasticity of ion channels—on synaptic receptors vs. gap junctions and their effects on extracellular field potentials under various input conditions and plasticity combinations across different structures. We fully agree with the reviewer that such studies would offer valuable insights and further enhance the broader relevance of our findings. However, while our study implies this direction, it was not the primary focus of our investigation.

In the revised manuscript, we will expand on intrinsic/synaptic plasticity and how they could contribute to LFPs (Sinha & Narayanan, 2015, 2022), while also pointing to simulations with different numbers of gap junction in this context.

Frequency-Dependent Effects

The study demonstrates that gap junctional inputs suppress highfrequency EFP power due to membrane filtering. However, it could delve deeper into the implications of this for different brain rhythms, such as gamma or ripple oscillations.

We sincerely thank you for these insightful comments that we totally agree with. As it so happens, this manuscript forms the first part of a broader study where we explore the implications of gap junctions to ripple frequency oscillations. The ripple oscillations part of the work was presented as a poster in the Society for Neuroscience (SfN) annual meeting 2024 (Sirmaur & Narayanan, 2024). There, we simulate a neuropil made of hundreds of morphologically realistic neurons to assess the role of different synaptic inputs — excitatory, inhibitory, and gap junctional — and active dendrites to ripple frequency oscillations. We demonstrate there that the conclusions from single-neuron simulations in this current manuscript extend to a neuropil with several neurons, each receiving excitatory, inhibitory and gap-junctional inputs, especially with reference to high-frequency oscillations. Our networkbased analyses unveiled a dominant mediatory role of patterned inhibition in ripple generation, with recurrent excitations through chemical synapses and gap junctions in conjunction with return-current contributions from active dendrites playing regulatory roles in determining ripple characteristics (Sirmaur & Narayanan, 2024).

Our principal goal in this study, therefore, was to lay the single-neuron foundation for network analyses of the impact of gap junctions on LFPs. We are preparing the network part of the study, with a strong focus on ripple-frequency oscillations, for submission for peer review separately.

In a revised manuscript, we will mention the results from our SfN abstract with reference to network simulations and high-frequency oscillations, while also presenting discussions from other studies on the role of gap junctions in synchrony and LFP oscillations.

Visualization

Figures are dense and could benefit from more intuitive labeling and focused presentations. For example, isolating key differences between chemical and gap junctional inputs in distinct panels would improve clarity.

We thank you for this constructive suggestion. In the revised manuscript, we will enhance the visualization of the figures to ensure a clearer and more intuitive distinction between chemical synapses and gap junctions.

Contextual Relevance

The manuscript touches on how these findings relate to known physiological roles of gap junctions (e.g., in gamma rhythms) but does not explore this in depth. Stronger integration of the results into known neural network dynamics would enhance its impact.

We sincerely appreciate your valuable suggestion and acknowledge the importance of integrating our results into established neural network dynamics, particularly their implications for gamma rhythms. We will address this aspect more comprehensively in the revised version of our manuscript.

Reviewer #2 (Public review):

This computational work examines whether the inputs that neurons receive through electrical synapses (gap junctions) have different signatures in the extracellular local field potential (LFP) compared to inputs via chemical synapses. The authors present the results of a series of model simulations where either electric or chemical synapses targeting a single hippocampal pyramidal neuron are activated in various spatio-temporal patterns, and the resulting LFP in the vicinity of the cell is calculated and analyzed. The authors find several notable qualitative differences between the LFP patterns evoked by gap junctions vs. chemical synapses. For some of these findings, the authors demonstrate convincingly that the observed differences are explained by the electric vs. chemical nature of the input, and these results likely generalize to other cell types. However, in other cases, it remains plausible (or even likely) that the differences are caused, at least partly, by other factors (such as different intracellular voltage responses due to, e.g., the unequal strengths of the inputs). Furthermore, it was not immediately clear to me how the results could be applied to analyze more realistic situations where neurons receive partially synchronized excitatory and inhibitory inputs via chemical and electric synapses.

We gratefully thank you for your time and effort in rigorously assessing our manuscript, for the enthusiastic response, and the encouraging and thoughtful comments on our study. In what follows, we have provided point-by-point responses to the specific comments.

Strengths

The main strength of the paper is that it draws attention to the fact that inputs to a neuron via gap junctions are expected to give rise to a different extracellular electric field compared to inputs via chemical synapses, even if the intracellular effects of the two types of input are similar. This is because, unlike chemical synaptic inputs, inputs via gap junctions are not directly associated with transmembrane currents. This is a general result that holds independent of many details such as the cell types or neurotransmitters involved.

We gratefully thank you for the positive comments and the encouraging words about the novel contributions of our study. We are particularly thankful to you for your comment on the generality of our conclusions that hold for different cell types and neurotransmitters involved.

Another strength of the article is that the authors attempt to provide intuitive, non-technical explanations of most of their findings, which should make the paper readable also for non-expert audiences (including experimentalists).

We sincerely thank you for the positive comments about the readability of the paper.

Weaknesses

The most problematic aspect of the paper relates to the methodology for comparing the effects of electric vs. chemical synaptic inputs on the LFP. The authors seem to suggest that the primary cause of all the differences seen in the various simulation experiments is the different nature of the input, and particularly the difference between the transmembrane current evoked by chemical synapses and the gap junctional current that does not involve the extracellular space. However, this is clearly an oversimplification: since no real attempt is made to quantitatively match the two conditions that are compared (e.g., regarding the strength and temporal profile of the inputs), the differences seen can be due to factors other than the electric vs. chemical nature of synapses. In fact, if inputs were identical in all parameters other than the transmembrane vs. directly injected nature of the current, the intracellular voltage responses and, consequently, the currents through voltage-gated and leak currents would also be the same, and the LFPs would differ exactly by the contribution of the transmembrane current evoked by the chemical synapse. This is evidently not the case for any of the simulated comparisons presented, and the differences in the membrane potential response are rather striking in several cases (e.g., in the case of random inputs, there is only one action potential with gap junctions, but multiple action potentials with chemical synapses). Consequently, it remains unclear which observed differences are fundamental in the sense that they are directly related to the electric vs. chemical nature of the input, and which differences can be attributed to other factors such as differences in the strength and pattern of the inputs (and the resulting difference in the neuronal electric response).

We thank you for raising this important point. We would like to emphasize that our experimental design and analyses quantitatively account for the spatial distribution and temporal pattern of specific kinds of inputs that arrive through gap junctions and chemical synapses. We submit that our analyses quantitatively demonstrates that the fundamental difference between the gap junctional and chemical synaptic contributions to extracellular potentials is the absence of the direct transmembrane component from gap junctional inputs. We elucidate these points below:

(1) Spatial distribution: The inputs were distributed randomly across the basal dendrites, irrespective of whether they were through gap junctions or chemical synapses. For both chemical synapses and gap junctions, the inputs were of the same nature: excitatory.

(2) Different numbers of inputs: We have presented consistent results for both fewer and more gap junctions or chemical synapses in our analyses (see Figure 1 with 217 gap junctions or 245 chemical synapses and Supplementary Figure 2 with 99 gap junctions or 30 chemical synapses). Our fundamentally novel result that gap junctions onto active dendrites shape LFPs holds true irrespective of the relative density of gap junctions onto the neuron.

(3) Synchronous inputs (Figs. 1–3): For chemical synapses, the waveforms are in the shape of postsynaptic potentials. For gap junctional inputs, the waveforms are in the shape of postsynaptic potentials or dendritic spikes (to respect the active nature of inputs from the other cell). Here, the electrical response of the postsynaptic cell is identical irrespective of whether inputs arrive through gap junctions or chemical synapses: an action potential. We quantitatively matched the strengths such that the model generated a single action potential in response to synchronous inputs, irrespective of whether they arrived through chemical synaptic and gap junctional inputs. We mechanistically analyze the contributions of different cellular components and show that the direct transmembrane current in chemical synapses is the distinguishing factor that determines the dichotomy between the contributions of gap junctions vs. chemical synapses to extracellular potentials (Figs. 2–3). In a revised manuscript, we will show the intracellular responses to demonstrate that they are electrically matched.

(4) Random inputs (Fig. 4): For random inputs, we did not account for the number of action potentials that arrived, as the only observation we made here was with reference to the biphasic nature of the extracellular potentials with gap junctional inputs in the “No Sodium” scenario. We note that in the “No Sodium” scenario, the time-domain amplitudes were comparable for the field potentials (Fig. 4B, Fig. 4D).

(5) Rhythmic inputs (Fig. 5–8): For rhythmic inputs, please note that the intracellular and extracellular waveforms for every frequency are provided in supplementary figures S5– S11. It may be noted that the intracellular responses are comparable. In simulations for assessing spike-LFP comparison, we tuned the strengths to produce a single spike per cycle, ensuring fair comparison of LFPs with gap junctions vs. chemical synapses.

Taken together, we demonstrate through explicit sets of simulations and analyses that the differences in LFPs were not driven by the strength or patterns of the inputs but rather by the differences in direct transmembrane currents, which are subsequently reflected in the LFPs. In a revised manuscript, we will add a section to emphasize these points apart from providing intracellular traces for cases where they are not provided.

Some of the explanations offered for the effects of cellular manipulations on the LFP appear to be incomplete. More specifically, the authors observed that blocking leak channels significantly changed the shape of the LFP response to synchronous synaptic inputs - but only when electric inputs were used, and when sodium channels were intact. The authors seemed to attribute this phenomenon to a direct effect of leak currents on the extracellular potential - however, this appears unlikely both because it does not explain why blocking the leak conductance had no effect in the other cases, and because the leak current is several orders of magnitude smaller than the spike-generating currents that make the largest contributions to the LFP. An indirect effect mediated by interactions of the leak current with some voltage-gated currents appears to be the most likely explanation, but identifying the exact mechanism would require further simulation experiments and/or a detailed analysis of intracellular currents and the membrane potential in time and space.

We thank you for raising this important question. Leak channels were among the several contributors to the positive deflection observed in LFPs associated with gap junctions. This effect was present not only in gap junctional models with intact sodium conductance but also in the no-sodium model, where the amplitude of the positive deflection was reduced across other models as well (Fig. 2F, I). Furthermore, even in the absence of leak conductance, a small positive deflection was still observed (Fig. 2F), leading us to further investigate other transmembrane currents over time and across spatial locations, from the proximal to the distal dendritic ends relative to the soma (Fig. 3D). We had observed that the dominant contributor in the case of chemical synapses was the inward synaptic current (Fig. 3A), whereas for gap junctions, the primary contributors were leak conductance along with other outward currents, such as potassium and HCN currents (Fig. 3D). Together, the direct transmembrane component of chemical synapses provides a dominant contribution to extracellular potentials. This dominance translates to differences in the relative contributions of indirect currents (including leak currents) to extracellular potentials associated chemical synaptic vs. gap junctional inputs. Our analyses of the exact ionic mechanisms (Fig. 3) demonstrates the involvement of several ion channels contributing to the indirect component in either scenario.

In every simulation experiment in this study, inputs through electric synapses are modeled as intracellular current injections of pre-determined amplitude and time course based on the sampled dendritic voltage of potential synaptic partners. This is a major simplification that may have a significant impact on the results. First, the current through gap junctions depends on the voltage difference between the two connected cellular compartments and is thus sensitive to the membrane potential of the cell that is treated as the neuron "receiving" the input in this study (although, strictly speaking, there is no pre- or postsynaptic neuron in interactions mediated by gap junctions). This dependence on the membrane potential of the target neuron is completely missing here. A related second point is that gap junctions also change the apparent membrane resistance of the neurons they connect, effectively acting as additional shunting (or leak) conductance in the relevant compartments. This effect is completely missed by treating gap junctions as pure current sources.

We thank you for raising this important point. We agree with the analyses presented by the reviewer on the importance of network simulations and bidirectional gap junctions that respect the voltages in both neurons. However, the complexities of LFP modeling precludes modeling of networks of morphologically realistic models with patterns of stimulations occurring across the dendritic tree. LFP modeling studies predominantly uses “post-synaptic” currents to analyze the impact of different patterns of inputs arriving on to a neuron, even when chemical synaptic inputs are considered. Explicitly, individual neurons are separately simulated with different patterns of synaptic inputs, the transmembrane current at different locations recorded, and the extracellular potential is then computed using line source approximation (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). Even in scenarios where a network is analyzed, a hybrid approach involving the outputs of a pointneuron-based network being coupled to an independent morphologically realistic neuronal model is employed (Hagen et al., 2016; Martinez-Canada et al., 2021; Mazzoni et al., 2015). Given the complexities associated with the computation of electrode potentials arising as a distance-weighted summation of several transmembrane currents, these simplifications becomes essential.

Our approach models gap junctional currents in a similar way as the other model incorporate synaptic currents in LFP modeling (Buzsaki et al., 2012; Gold et al., 2006; Halnes et al., 2024; Ness et al., 2018; Reimann et al., 2013; Schomburg et al., 2012; Sinha & Narayanan, 2015, 2022). As gap junctions are typically implemented as resistors from the other neuronal compartment, we accounted for gap-junctional variability in our model by randomizing the scaling-factors and the exact waveforms that arrive through individual gap junctions at specific locations. Thus, the inputs were not pre-determined by “pre” neurons. Instead, the recorded voltages from potential synaptic partner neurons were randomized across locations and scaled using factors at the dendrites before being injected into the target neuron (Supplementary Fig. S1). While incorporating a network of interconnected neurons is indeed important, we utilized biophysical, morphologically realistic CA1 neuron model with different sets of input patterns to model LFPs, which were derived from the total transmembrane currents across all compartments of the multi-compartmental neuron model. Given the complexity of this approach, adding further network-level interactions or pre-post connections would have been computationally demanding.

In a revised manuscript, we will introduce the general methodology used in LFP modeling studies to introduce synaptic currents. We will emphasize that our study extends this approach to modeling gap junctional inputs, while also highlighting randomization of locations and the scaling process in assigning gap junctional synaptic strengths.

One prominent claim of the article that is emphasized even in the abstract is that HCN channels mediate an outward current in certain cases. Although this statement is technically correct, there are two reasons why I do not consider this a major finding of the paper. First, as the authors acknowledge, this is a trivial consequence of the relatively slow kinetics of HCN channels: when at least some of the channels are open, any input that is sufficiently fast and strong to take the membrane potential across the reversal potential of the channel will lead to the reversal of the polarity of the current. This effect is quite generic and well-known and is by no means specific to gap junctional inputs or even HCN channels. Second, and perhaps more importantly, the functional consequence of this reversed current through HCN channels is likely to be negligible. As clearly shown in Supplementary Figure S3, the HCN current becomes outward only for an extremely short time period during the action potential, which is also a period when several other currents are also active and likely dominant due to their much higher conductances. I also note that several of these relevant facts remain hidden in Figure 3, both because of its focus on peak values, and because of the radically different units on the vertical axes of the current plots.

We thank you for raising this point and agree with you on every point. Please note that we do not assert that the outward HCN currents are exclusively associated with gap junctional inputs. Rather, our results show that synchronous inputs generate outward HCN currents in both chemical synapses (Fig. 3B; positive/outward HCN currents, except in the no sodium or leak model) and gap junctions (Fig. 3D; positive/outward HCN currents). We emphasized this in the case of gap junctions because, in the absence of inward synaptic currents, HCN (acting as outward currents with synchronous inputs) contributed to the positive deflection observed in the LFPs. While HCN would also contribute in the case of chemical synapses, its effect was negligible due to the presence of large inward synaptic currents. Since LFPs reflect the collective total transmembrane currents, the dominant contributors differ between these two scenarios, which we aimed to highlight. Since HCN exhibited outward currents in our synchronous input simulations, we have elaborated on this mechanism in the supplementary figure (Fig. S3). Our intention was not to emphasize this effect for only one synaptic mode but rather to highlight HCN's contribution to the positive deflection as one of the contributing factors.

We agree that HCN currents are relatively small in magnitude; therefore, our conclusions were based on HCN being one of the several contributing factors. Leak conductance and other outward conductances, including HCN currents (Fig. 3D), collectively contribute to the positive deflections observed in the case of gap junctional synchronous inputs.

We will ensure that we will account for all the points appropriately in a revised manuscript.

Finally, I missed an appropriate validation of the neuronal model used, and also the characterization of the effects of the in silico manipulations used on the basic behavior of the model. As far as I understand, the model in its current form has not been used in other studies. If this is the case, it would be important to demonstrate convincingly through (preferably quantitative) comparisons with experimental data using different protocols that the model captures the physiological behavior of at least the relevant compartments (in this case, the dendrites and the soma) of hippocampal pyramidal neurons sufficiently well that the results of the modeling study are relevant to the real biological system. In addition, the correct interpretation of various manipulations of the model would be strongly facilitated by investigating and discussing how the physiological properties of the model neuron are affected by these alterations.

We thank you for raising this important point. The CA1 pyramidal neuronal model used in this study is built with ion-channel models derived from biophysical and electrophysiological recordings from these cells. As mentioned in the Methods section “Dynamics and distribution of active channels” and Supplementary Table S1, models for individual channels, their gating kinetics, and channel distributions across the somatodendritic arbor (wherever known) are all derived from their physiological equivalents. Importantly, these values were derived from previously validated models from the laboratory, which contain these very ion channel models and the exact same morphology (Roy & Narayanan, 2021). Please compare Supplementary Table S1 with the Table 1 from (Roy & Narayanan, 2021). Please note that this model was validated against several physiological measurements along the somatodendritic axis (Fig. 1 of (Roy & Narayanan, 2021)).

In a revised manuscript, we will explicitly mention this while also mentioning the different physiological properties that were used for the validation process from (Roy & Narayanan, 2021). We sincerely regret not mentioning these details in the current version of our manuscript.

We will fix these in a revised version of the manuscript.

References

Bedner, P., Steinhauser, C., & Theis, M. (2012). Functional redundancy and compensation among members of gap junction protein families? Biochim Biophys Acta, 1818(8), 1971-1984. https://doi.org/10.1016/j.bbamem.2011.10.016

Behrens, C. J., Ul Haq, R., Liotta, A., Anderson, M. L., & Heinemann, U. (2011). Nonspecific effects of the gap junction blocker mefloquine on fast hippocampal network oscillations in the adult rat in vitro. Neuroscience, 192, 11-19. https://doi.org/10.1016/j.neuroscience.2011.07.015

Buzsaki, G., Anastassiou, C. A., & Koch, C. (2012). The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nat Rev Neurosci, 13(6), 407-420. https://doi.org/10.1038/nrn3241

Einevoll, G. T., Destexhe, A., Diesmann, M., Grun, S., Jirsa, V., de Kamps, M., Migliore, M., Ness, T. V., Plesser, H. E., & Schurmann, F. (2019). The Scientific Case for Brain Simulations. Neuron, 102(4), 735-744. https://doi.org/10.1016/j.neuron.2019.03.027

Gold, C., Henze, D. A., Koch, C., & Buzsaki, G. (2006). On the origin of the extracellular action potential waveform: A modeling study. J Neurophysiol, 95(5), 3113-3128. https://doi.org/10.1152/jn.00979.2005

Hagen, E., Dahmen, D., Stavrinou, M. L., Linden, H., Tetzlaff, T., van Albada, S. J., Grun, S., Diesmann, M., & Einevoll, G. T. (2016). Hybrid Scheme for Modeling Local Field Potentials from Point-Neuron Networks. Cereb Cortex, 26(12), 4461-4496. https://doi.org/10.1093/cercor/bhw237

Halnes, G., Ness, T. V., Næss, S., Hagen, E., Pettersen, K. H., & Einevoll, G. T. (2024). Electric Brain Signals: Foundations and Applications of Biophysical Modeling. Cambridge University Press. https://doi.org/DOI: 10.1017/9781009039826

Lo, C. W. (1999). Genes, gene knockouts, and mutations in the analysis of gap junctions. Dev Genet, 24(1-2), 1-4. https://doi.org/10.1002/(SICI)1520-6408(1999)24:1/2%3C1::AID-DVG1%3E3.0.CO;2-U

Martinez-Canada, P., Ness, T. V., Einevoll, G. T., Fellin, T., & Panzeri, S. (2021). Computation of the electroencephalogram (EEG) from network models of point neurons. PLoS Comput Biol, 17(4), e1008893. https://doi.org/10.1371/journal.pcbi.1008893

Mazzoni, A., Linden, H., Cuntz, H., Lansner, A., Panzeri, S., & Einevoll, G. T. (2015). Computing the Local Field Potential (LFP) from Integrate-and-Fire Network Models. PLoS Comput Biol, 11(12), e1004584. https://doi.org/10.1371/journal.pcbi.1004584

Ness, T. V., Remme, M. W. H., & Einevoll, G. T. (2018). h-Type Membrane Current Shapes the Local Field Potential from Populations of Pyramidal Neurons. J Neurosci, 38(26), 6011-6024. https://doi.org/10.1523/jneurosci.3278-17.2018

Reimann, M. W., Anastassiou, C. A., Perin, R., Hill, S. L., Markram, H., & Koch, C. (2013). A biophysically detailed model of neocortical local field potentials predicts the critical role of active membrane currents. Neuron, 79(2), 375-390. https://doi.org/10.1016/j.neuron.2013.05.023

Rouach, N., Segal, M., Koulakoff, A., Giaume, C., & Avignone, E. (2003). Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions. Journal of Physiology, 553(Pt 3), 729-745. https://doi.org/10.1113/jphysiol.2003.053439

Roy, A., & Narayanan, R. (2021). Spatial information transfer in hippocampal place cells depends on trial-to-trial variability, symmetry of place-field firing, and biophysical heterogeneities. Neural Netw, 142, 636-660. https://doi.org/10.1016/j.neunet.2021.07.026

Schomburg, E. W., Anastassiou, C. A., Buzsaki, G., & Koch, C. (2012). The spiking component of oscillatory extracellular potentials in the rat hippocampus. J Neurosci, 32(34), 11798-11811. https://doi.org/10.1523/JNEUROSCI.0656-12.2012

Sinha, M., & Narayanan, R. (2015). HCN channels enhance spike phase coherence and regulate the phase of spikes and LFPs in the theta-frequency range. Proc Natl Acad Sci U S A, 112(17), E2207-2216. https://doi.org/10.1073/pnas.1419017112

Sinha, M., & Narayanan, R. (2022). Active Dendrites and Local Field Potentials: Biophysical Mechanisms and Computational Explorations. Neuroscience, 489, 111-142. https://doi.org/10.1016/j.neuroscience.2021.08.035

Sirmaur, R., & Narayanan, R. (2024). Distinct extracellular signatures of chemical and electrical synapses impinging on active dendrites differentially contribute to ripple-frequency oscillations. Society for Neuroscience annual meeting (https://www.abstractsonline.com/pp8/?_gl=1*1bxo7m*_gcl_au*MTc5MTQ0NjE0NC4xNzI3MDcwOTMw*_ga*MTMxMTE5OTcyMy4xNzI3MDcwOTMx*_ga_T09K 3Q2WDN*MTcyNzA3MDkzMS4xLjEuMTcyNzA3MDkzNy41NC4wLjA.#!/20433/ presentation/13949), Chicago, USA.

Szarka, G., Balogh, M., Tengolics, A. J., Ganczer, A., Volgyi, B., & Kovacs-Oller, T. (2021). The role of gap junctions in cell death and neuromodulation in the retina. Neural Regen Res, 16(10), 1911-1920. https://doi.org/10.4103/1673-5374.308069

eLife Assessment

This study resolves a cryo-EM structure of the GPCR, human GPR30, which responds to bicarbonate and regulates cellular responses to pH and ion homeostasis. Understanding the ligand and the mechanism of activation is important to the field of receptor signaling and potentially facilitates drug development targeting this receptor. Structures and functional assays provide solid evidence for a potential bicarbonate binding site.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers.]

Summary:

This study resolves a cryo-EM structure of the GPCR, GPR30, in the presence of bicarbonate, which the author's lab recently identified as the physiological ligand. Understanding the ligand and the mechanism of activation is of fundamental importance to the field of receptor signaling. This solid study provides important insight into the overall structure and suggests a possible bicarbonate binding site.

Strengths:

The overall structure, and proposed mechanism of G-protein coupling are solid. Based on the structure, the authors identify a binding pocket that might accommodate bicarbonate. Although assignment of the binding pocket is speculative, extensive mutagenesis of residues in this pocket identifies several that are important to G-protein signaling. The structure shows some conformational differences with a previous structure of this protein determined in the absence of bicarbonate (PMC11217264). To my knowledge, bicarbonate is the only physiological ligand that has been identified for GPR30, making this study an important contribution to the field. However, the current study provides novel and important circumstantial evidence for the bicarbonate binding site based on mutagenesis and functional assays.

Weaknesses:

Bicarbonate is a challenging ligand for structural and biochemical studies, and because of experimental limitations, this study does not elucidate the exact binding site. Higher resolution structures would be required for structural identification of bicarbonate. The functional assay monitors activation of GPR30, and thus reports on not only bicarbonate binding, but also the integrity of the allosteric network that transduces the binding signal across the membrane. However, biochemical binding assays are challenging because the binding constant is weak, in the mM range.

The authors appropriately acknowledge the limitations of these experimental approaches, and they build a solid circumstantial case for the bicarbonate binding pocket based on extensive mutagenesis and functional analysis. However, the study does fall short of establishing the bicarbonate binding site.

Reviewer #2 (Public review):

Summary:

In this manuscript, "Cryo-EM structure of the bicarbonate receptor GPR30," the authors aimed to enrich our understanding of the role of GPR30 in pH homeostasis by combining structural analysis with a receptor function assay. This work is a natural development and extension of their previous work on Nature Communications (PMID: 38413581). In the current body of work, they solved the cryo-EM structure of the human GPR30-G-protein (mini-Gsqi) complex in the presence of bicarbonate ions at 3.15 Å resolution. From the atomic model built based on this map, they observed the overall canonical architecture of class A GPCR and also identified 3 extracellular pockets created by ECLs (Pockets A-C). Based on the polarity, location, size, and charge of each pocket, the authors hypothesized that pocket A is a good candidate for the bicarbonate binding site. To identify the bicarbonate binding site, the authors performed an exhaustive mutant analysis of the hydrophilic residues in Pocket A and analyzed receptor reactivity via calcium assay. In addition, the human GPR30-G-protein complex model also enabled the authors to elucidate the G-protein coupling mechanism of this special class A GPCR, which plays a crucial role in pH homeostasis.

Strengths:

As a continuation of their recent Nature Communications publication, the authors used cryo-EM coupled with mutagenesis and functional studies to elucidate bicarbonate-GPR30 interaction. This work provided atomic-resolution structural observations for the receptor in complex with G-protein, allowing us to explore its mechanism of action, and will further facilitate drug development targeting GPR30. There were 3 extracellular pockets created by ECLs (Pockets A-C). The authors were able to filter out 2 of them and hypothesized that pocket A was a good candidate for the bicarbonate binding site based on the polarity, location, and charge of each pocket. From there, the authors identified the key residues on GPR30 for its interaction with the substrate, bicarbonate. Together with their previous work, they mapped out amino acids that are critical for receptor reactivity.

Weaknesses:

When we see a reduction of a GPCR-mediated downstream signaling, several factors could potentially contribute to this observation: 1) a reduced total expression of this receptor due to the mutation (transcription and translation issue); 2) a reduced surface expression of this receptor due to the mutation (trafficking issue); and 3) a dysfunctional receptor that doesn't signal due to the mutation.

Altogether, the wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

Reviewer #3 (Public review):

Summary

GPR30 responds to bicarbonate and plays a role in regulating cellular pH and ion homeostasis. However, the molecular basis of bicarbonate recognition by GPR30 remains unresolved. This study reports the cryo-EM structure of GPR30 bound to a chimeric mini-Gq in the presence of bicarbonate, revealing mechanistic insights into its G-protein coupling. Nonetheless, the study does not identify the bicarbonate-binding site within GPR30.

Strengths

The work provides strong structural evidence clarifying how GPR30 engages and couples with Gq.

Weaknesses

Several GPR30 mutants exhibited diminished responses to bicarbonate, but their expression levels were also reduced. As a result, the mechanism by which GPR30 recognizes bicarbonate remains uncertain.

Author Response:

The following is the authors’ response to the previous reviews

Reviewer #1 (Public review):

Summary:

This study resolves a cryo-EM structure of the GPCR, GPR30, in the presence of bicarbonate, which the author's lab recently identified as the physiological ligand. Understanding the ligand and the mechanism of activation is of fundamental importance to the field of receptor signaling. This solid study provides important insight into the overall structure and suggests a possible bicarbonate binding site.

Strengths:

The overall structure, and proposed mechanism of G-protein coupling are solid. Based on the structure, the authors identify a binding pocket that might accommodate bicarbonate. Although assignment of the binding pocket is speculative, extensive mutagenesis of residues in this pocket identifies several that are important to G-protein signaling. The structure shows some conformational differences with a previous structure of this protein determined in the absence of bicarbonate (PMC11217264). To my knowledge, bicarbonate is the only physiological ligand that has been identified for GPR30, making this study an important contribution to the field. However, the current study provides novel and important circumstantial evidence for the bicarbonate binding site based on mutagenesis and functional assays.

Weaknesses:

Bicarbonate is a challenging ligand for structural and biochemical studies, and because of experimental limitations, this study does not elucidate the exact binding site. Higher resolution structures would be required for structural identification of bicarbonate. The functional assay monitors activation of GPR30, and thus reports on not only bicarbonate binding, but also the integrity of the allosteric network that transduces the binding signal across the membrane. However, biochemical binding assays are challenging because the binding constant is weak, in the mM range.

The authors appropriately acknowledge the limitations of these experimental approaches, and they build a solid circumstantial case for the bicarbonate binding pocket based on extensive mutagenesis and functional analysis. However, the study does fall short of establishing the bicarbonate binding site.

We thank the reviewer for this thoughtful and constructive assessment of our revised manuscript. We are grateful for the recognition of the overall quality of the cryo-EM structure and the proposed mechanism of G-protein coupling, as well as for highlighting the importance of identifying bicarbonate as a physiological ligand for GPR30 and the contribution this work makes to the receptor signaling field. We also appreciate the reviewer’s careful and balanced discussion of the inherent challenges posed by bicarbonate as a low-affinity, small, negatively charged ligand, and we fully agree that, given current experimental limitations, our data provide circumstantial—rather than definitive—evidence for the binding site and that higher-resolution structures would be required for direct visualization. Importantly, we value the reviewer’s acknowledgement that we transparently describe these limitations and that our extensive mutagenesis and functional analyses nonetheless build a solid case for the proposed bicarbonate-binding pocket, which we believe will serve as a useful framework for future biochemical and structural investigation

Reviewer #1 (Recommendations for the authors):

Overall, the authors do a good job responding to the previous review, with updated structures and experimental data. I have two comments on the current version:

(1) When the authors compare their structure to a previously published structure of the same receptor, they say that the previous structure came out while the current manuscript was in revision (line 255). This is not correct. The previous manuscript was published May 14, 2024, and the current manuscript was received by eLife on May 20, 2024. This sentence should be corrected to "During the preparation of this manuscript..."

We corrected the sentence accordingly (line 259).

(2) Line 173: what other structures are the authors referring to? Citations should be included here.

Is Line 193 correct? We added citations (line 190).

Reviewer #2 (Public review):

Summary:

In this manuscript, "Cryo-EM structure of the bicarbonate receptor GPR30," the authors aimed to enrich our understanding of the role of GPR30 in pH homeostasis by combining structural analysis with a receptor function assay. This work is a natural development and extension of their previous work on Nature Communications (PMID: 38413581). In the current body of work, they solved the cryo-EM structure of the human GPR30-G-protein (mini-Gsqi) complex in the presence of bicarbonate ions at 3.15 Å resolution. From the atomic model built based on this map, they observed the overall canonical architecture of class A GPCR and also identified 3 extracellular pockets created by ECLs (Pockets A-C). Based on the polarity, location, size, and charge of each pocket, the authors hypothesized that pocket A is a good candidate for the bicarbonate binding site. To identify the bicarbonate binding site, the authors performed an exhaustive mutant analysis of the hydrophilic residues in Pocket A and analyzed receptor reactivity via calcium assay. In addition, the human GPR30-G-protein complex model also enabled the authors to elucidate the G-protein coupling mechanism of this special class A GPCR, which plays a crucial role in pH homeostasis.

Strengths:

As a continuation of their recent Nature Communications publication, the authors used cryo-EM coupled with mutagenesis and functional studies to elucidate bicarbonate-GPR30 interaction. This work provided atomic-resolution structural observations for the receptor in complex with G-protein, allowing us to explore its mechanism of action, and will further facilitate drug development targeting GPR30. There were 3 extracellular pockets created by ECLs (Pockets A-C). The authors were able to filter out 2 of them and hypothesized that pocket A was a good candidate for the bicarbonate binding site based on the polarity, location, and charge of each pocket. From there, the authors identified the key residues on GPR30 for its interaction with the substrate, bicarbonate. Together with their previous work, they mapped out amino acids that are critical for receptor reactivity.

Weaknesses:

When we see a reduction of a GPCR-mediated downstream signaling, several factors could potentially contribute to this observation: 1) a reduced total expression of this receptor due to the mutation (transcription and translation issue); 2) a reduced surface expression of this receptor due to the mutation (trafficking issue); and 3) a dysfunctional receptor that doesn't signal due to the mutation. In the current revision, based on the gating strategy, the surface expression of the HA-positive WT GPR30-expressing cells is only 10.6% of the total population, while the surface expression levels of the mutants range from 1.89% (P71A) to 64.4% (D111A). Combining this information with the functional readout in Figure 3F and G, as well as their previous work, the authors concluded that mutations at P71, E115, D125, Q138, C207, D210, and H307 would decrease bicarbonate responses. Among those sites,

E115, Q138, and H307 were from their previous Nature Comm paper.

Authors claim P71 and C207 make a structural-stability contribution, as their mutations result in a significant reduction in surface expression: P71A (1.89%) and C207A (2.71%). However, compared to 10.6% of the total population in the WT, (P71A is 17.8% of the WT, and C207A is 25.6% of the WT), this doesn't rule out the possibility that the mutated receptor is also dysfunctional: at 10 mM NaHCO3, RFU of WT is ~500, RFU of P71 and C207 are ~0.

The authors also interpret "The D125ECL1A mutant has lost its activity but is located on the surface" and only mention "D125 is unlikely to be a bicarbonate binding site, and the mutational effect could be explained due to the decreased surface expression". Again, compared to 10.6% of the total population in the WT, D125A (3.94%) is 37.2% of the WT. At 10 mM NaHCO3, the RFU of the WT is ~500, the RFU of D125 is ~0. This doesn't rule out the possibility that the mutated receptor is also dysfunctional. It is not clear why D125A didn't make it to the surface.

Other mutants that the authors didn't mention much in their text: D111A (64.4%, 607.5% of WT surface expression), E121A (50.4%, 475.5% of WT surface expression), R122 (41.0%, 386.8% of WT surface expression), N276A (38.9%, 367.0% of WT surface expression) and E218A (24.6%, 232.1% of WT surface expression) all have similar RFU as WT, although the surface expression is about 2-6 times more. On the other hand, Q215A (3.18%, 30% of WT surface expression) has similar RFU as WT, with only a third of the receptor on the surface.

Altogether, the wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

We sincerely thank the reviewer for their careful reading and thoughtful evaluation of our manuscript on the cryo-EM structure of the bicarbonate receptor GPR30. We greatly appreciate the reviewer’s positive assessment of the overall significance of combining structural determination with extensive mutagenesis and functional assays to advance understanding of bicarbonate–GPR30 interactions and G-protein coupling, as well as their recognition that these atomic-level insights will be valuable for future mechanistic studies and drug-development efforts. We are also grateful for the reviewer’s constructive critique regarding the interpretation of reduced signaling in the context of variable surface expression across mutants, which highlights an important point about disentangling effects of expression/trafficking from intrinsic receptor dysfunction; these comments are highly insightful and will help us strengthen the clarity and rigor of our presentation and conclusions in the revised manuscript.

Reviewer #2 (Recommendations for the authors):

In this revision, the authors have made a significant effort to improve and validate the structural observations, as well as address the comments in the previous submission. They updated the functional assays and evaluated the receptor function by measuring intracellular calcium mobilization, which is a more direct measurement for the downstream signaling of hGPR30-Gq signaling. They also used flow cytometry with an HA-antibody for a more direct measurement of the surface expression of the receptor, replacing their previous assay that normalized to the housekeeping gene Na-K-ATPase.

I appreciate the effort the authors made to address the previous comments made by the reviewers. However, there are still some concerns about the current data.

(1) The authors have addressed my previous comment on untangling the mixture of their previous and new data in the "insights into bicarbonate binding" section. They have made it clear that the importance of E115, Q138, and H307 in the receptor-bicarbonate interaction was shown in their Nature Communications paper.

(2) The authors have addressed my previous comment on adding some content about the physiological concentration of HCO3, or referring more to their previous work about the rationale to select the bicarbonate dose in their functional assay.

(3) The authors have updated Figure 3

(4) The authors have updated Supplemental Figure 1 to show the full gel with molecular weight markers in the supplemental data to demonstrate the sample purity.

(5) The authors have updated the predicted model using AF3

(6) The authors added E218A as suggested before.

Some new suggestions for this R1:

(1) The wide range of surface expression across the different cell lines, combined with the different receptor function readouts, makes the cell functional data only partially support their structural observations.

We acknowledge this limitation. The wide range of surface expression among cell lines, together with differences in assay modalities, may introduce variability that complicates direct quantitative comparisons and therefore only partially supports the structural observations. Future work using more standardized expression systems and matched functional readouts will be important to strengthen the structure–function linkage.

(2) Line 101, "ICL1 and ECL1 contain short α helices", no α helix of ICL1 is shown in Figure 2C

We removed the word “ICL1” (line 98).

(3) For the unsolved region of ECL2, could the author put a dashed line connecting ECL2 with TM4? In the current Figure 2B, it looks like ECL2 connects TM3 and TM5.

According to the suggestion, we corrected Figure 2B.

(4) I appreciate that the authors updated the predicted model with AF3, but they didn't make it clear why they had the comparison between their cryo-EM structure (bicarbonate-activated G-protein-incorporated GPR30) and the predicted AF3 model (inactive GPR30)

We wish to assert the usefulness of experimental structures, not merely predictions. These include structures independent of receptor activation, such as SS bonds.

(5) I appreciate that the authors have addressed my previous comment on adding some content about the physiological concentration of HCO3, but it was still not clear to me why they picked 11 mM in Figure 3G for the bar graph. Also, since a dose-response curve was made in Figure 3F, why not just calculate and report the EC50 of NaHCO3 for each mutant?

Thank you for your comment. Thank you for the comment. We’ve calculated the EC50 of the calcium response and assessed its correlation with receptors’ cell surface expression. We chose 11 mM in Fig .3G since our previous paper in Nature Communications showed the EC50 value of IPs assay was around 11 mM. However, the calcium response was more sensitive and gave a lower value than expected. Therefore, according to your advice, we deleted the bar graph with 11 mM responses, calculated EC50, and drew pictures of the correlation among cell surface expression, EC50, and maximum responses (Figure 3F-I, Supplementary File 1). Moreover, we revised the explanation about this mutagenesis study (lines139-154 and 217-230).

(6) In the previous submission and comments, E218 was in close contact with bicarbonate in the previous Figure 4D (the bicarbonate is deleted in the new structure). I thank the authors for making an E218A mutant and performing the functional assay. As mentioned above, E218A (24.6%, 232.1% of WT surface expression) has a similar functional readout as WT. Doesn't this also indicate that E218A is partially broken, so you will need twice as much as WT to have the same downstream signal?

Thank you for your comment. In our revised manuscript, we described the correlation between cell surface expression and EC50 and found that cell surface expression and the response to bicarbonate are not correlated, which you mentioned in your review comment (Figure 3F-I, Supplementary File 1). There are many possibilities that could explain this: GPR30 localization in specific spots on the plasma membrane might limit the response stoichiometry, GPR30 might also work intracellularly to blunt the increased response because of more GPR30 expression on PM, redundant GPR30 on PM might be broken, or E118A might be less functional and need twice as much as WT. We will examine cell surface expression of GPR30 and its response to bicarbonate in a future study.

I would suggest that the authors in future studies consider using the Tet-on inducible cell lines, such as HEK293 Flp-In Trex. These cell lines will allow the authors to fine-tune the surface expression of their mutants to the same level with different doses of Tetracycline in their stable cell lines.

We appreciate your advice. We’ll introduce Tet-on inducible cell lines for future research.

Reviewer #3 (Public review):

Summary

GPR30 responds to bicarbonate and plays a role in regulating cellular pH and ion homeostasis. However, the molecular basis of bicarbonate recognition by GPR30 remains unresolved. This study reports the cryo-EM structure of GPR30 bound to a chimeric mini-Gq in the presence of bicarbonate, revealing mechanistic insights into its G-protein coupling. Nonetheless, the study does not identify the bicarbonate-binding site within GPR30.

Strengths

The work provides strong structural evidence clarifying how GPR30 engages and couples with Gq.

Weaknesses

Several GPR30 mutants exhibited diminished responses to bicarbonate, but their expression levels were also reduced. As a result, the mechanism by which GPR30 recognizes bicarbonate remains uncertain, leaving this aspect of the study incomplete.

We sincerely thank the reviewer for this thoughtful and balanced assessment of our manuscript, including the clear summary of the central advance and the constructive identification of remaining limitations. We particularly appreciate the recognition that our cryo-EM analysis provides strong structural evidence for how GPR30 engages and couples with Gq, and we agree that pinpointing the bicarbonate-binding site remains a critical open question. In the revised manuscript, we will make this point more explicit, clarify the interpretation of the mutagenesis results in light of reduced receptor expression for some variants, and further strengthen the presentation and discussion of what our current data do—and do not—allow us to conclude regarding bicarbonate recognition by GPR30

Reviewer #3 (Recommendations for the authors):

The authors have removed the bicarbonate assignment from their model and have addressed all of my concerns. In this study, or in future work, it would be advisable for the authors to explore the use of bicarbonate mimetics with higher binding affinity to facilitate more definitive structural characterization.

Thank you for this constructive suggestion. We agree that exploring bicarbonate mimetics with higher binding affinity would be an important next step to enable more definitive structural characterization of GPR30 and to strengthen mechanistic conclusions. In future work, we plan to pursue the identification and/or design of such mimetics, guided by the architecture and mutational landscape of the extracellular pocket described here, and to combine these ligands with optimized cryo-EM sample preparation and complementary functional assays to better stabilize and visualize the bound state.

eLife Assessment

This is an important study on the role of the neurokinin-2 receptor (NK2R) as a regulatory node connecting intestinal lipid metabolism, mucosal immunity, and the gut microbiome, bidirectionally regulating enterocyte lipid uptake, lipid droplet storage, chylomicron output, and systemic metabolic parameters in DIO mice. The authors present solid evidence linking Tacr2 deletion to reprogrammed epithelial lineage allocation, dampened immune gene expression, and male-biased protection from DSS colitis, despite dysbiotic microbiota. However, the causal evidence for some mechanistic and pro-inflammatory NK2R claims remains incomplete and potentially confounding, requiring additional cell-type-specific and functional experiments.

Reviewer #1 (Public review):

Summary:

This study identifies NK2R as an intestinal GPCR that tunes enterocyte lipid uptake, lipid droplet storage, and chylomicron output, with loss or antagonism enhancing post‑prandial triglyceridemia and epithelial lipid stores, and agonism reducing adiposity and improving glycemia in DIO mice. Through bulk RNA‑seq, deconvolution, DSS colitis, and 16S profiling, the authors link Tacr2 deletion to coordinated induction of epithelial lipid‑metabolic programs, dampened immune gene expression, sex‑specific remodeling of secretory lineages, and male‑biased protection from experimental colitis despite dysbiotic microbiota. This is an overall important and thorough paper on an emerging obesity drug target, but it should temper some interpretations, and the following points would be needed to strengthen the claims in the manuscript.

Strengths:


The study uses an impressive combination of genetic loss‑of‑function, pharmacological agonism/antagonism, transcriptomics, and in vivo physiology to establish NK2R as a bidirectional regulator of epithelial lipid handling. The integration of RNA‑seq, epithelial cell‑type deconvolution, DSS colitis, and microbiome profiling provides a rich, systems‑level view of how Tacr2 deletion reshapes epithelial metabolism, lineage allocation, and inflammatory responsiveness in a sex‑specific manner. The gain- and loss‑of‑function data particularly support a model in which NK2R acts as an epithelial metabolic rheostat that restrains lipid absorption and chylomicron export, with downstream consequences for barrier fitness and immune tone.

Weaknesses:

Major points

While the data convincingly establish NK2R's role in epithelial lipid handling, the manuscript arguably overstates a primary "pro‑inflammatory" function for NK2R, given that Tacr2‑/‑ mice show enhanced enterocyte lipid uptake and storage, higher post‑prandial triglycerides, and a dysbiotic microbiota yet reduced mucosal immune gene expression and, in males, protection from DSS colitis. It remains equally plausible that the apparent "protection" reflects a mucosa that is less reactive to unfavorable microbiota rather than genuinely protected, and that NK2R's main function is metabolic, with immune changes emerging secondarily. Such a model would actually help reconcile the long-standing question as to why NK2R antagonism has not translated into clear benefit in clinical trials for GI inflammation over the past several decades.

Without temporal resolution, it is equally plausible that antagonists primarily perturb epithelial lipid homeostasis rather than directly and beneficially modulating immune tone. To discriminate between these possibilities and strengthen the potential direct inflammatory claims, the authors should:

(1) generate epithelial‑specific, immune‑cell-specific, and nociceptor‑specific Tacr2 deletions in the DSS model

(2) test gut‑restricted NK2R agonism versus antagonism under controlled dietary fat conditions for effects on LD load, barrier integrity, and colitis severity

(3) perform ex vivo tachykinin/NK2R stimulation of isolated epithelial versus immune compartments with functional readouts

(4) assess whether microbiota transfer from Tacr2‑/‑ versus WT donors into germ‑free or antibiotic‑treated recipients can recapitulate protection or susceptibility independently of epithelial NK2R status.

Minor points

Additional clarifications on Tac1 and tachykinin receptor expression in male/female colitis models, and validation of the NK2R antibody in KO tissue (or in situ hybridization), would also be needed to strengthen key mechanistic and localization claims.

Reviewer #2 (Public review):

Summary:

This manuscript- "NK2R signaling governs intestinal lipid mobilization and mucosal inflammation" by Perez et al investigates the role of the neurokinin-2 receptor (NK2R) as a regulatory node connecting intestinal lipid metabolism, mucosal immunity, and the gut microbiome. The authors utilized a ubiquitous deleted Tacr2 mouse model alongside targeted pharmacological treatments to demonstrate that NK2R limits luminal lipid uptake and chylomicron secretion. Additionally, the study uncovers that Tacr2 deficiency promotes male-biased protection against DSS-induced colitis and drives distinct diet- and genotype-dependent shifts in the fecal microbiota.

Strengths:

(1) The authors successfully utilized both a genetic whole-body knockout model (Tacr2-/-) and targeted pharmacological agents, such as the antagonist GR159897 and the agonist EB1002. This dual approach effectively corroborates the core phenotypic findings.

(2) The study provides a compelling case for targeting the tachykinin-NK2R axis therapeutically. The remarks that NK2R agonists could be leveraged to treat obesity, while antagonists might be used for inflammatory bowel disease, will be an exciting clinical outcome if further validated.

(3) The integration of RNAseq for epithelial lineage analysis, combined with in vivo gut permeability assays, lipid tolerance assays, and 16S microbiome sequencing, provides a robust and highly detailed physiological picture.

Weaknesses:

This manuscript has some notable limitations. While the transcriptomic data show an upregulation of the enterocyte lipid droplet program in Tacr2-/- mice, the manuscript lacks biochemical experiments to conclude the downstream signaling mechanism driving such changes. The reliance on a global whole-body knockout model confounds the ability to definitively conclude that the observed metabolic and inflammatory phenotypes are linked to the intestinal epithelium. The authors discuss a male-biased protection against DSS-induced colitis, but they rely on speculation regarding sex hormones rather than providing experimental data to explain this dimorphism.

eLife Assessment

This useful study identified XAP5 as an ancient transcriptional regulator critical for primary ciliogenesis. The evidence supporting the conceptual framework linking evolutionary conservation to functional specialization in primary ciliogenesis remains incomplete. This work will be of interest to developmental biologists and to those studying diseases caused by ciliopathies.

Reviewer #1 (Public review):

Summary:

The authors have attempted to establish a role for XAP5, a transcriptional regulator they have previously identified for flagellar biogenesis in Chlamydomonas and mice, in primary cilia differentiation.

Strengths:

Genetic and biochemical analysis using a cultured mouse cell line, NIH3T3.

Weaknesses:

(1) The authors have ignored established data that, like in C. elegans and Drosophila, there is in vivo genetic evidence that primary cilia formation is regulated by the RFX transcriptional module (for example, PMID 19887680, PMID 29510665).

(2) The analysis with one mammalian cell line, NIH3T3, while done quite rigorously, is not sufficient. Also, the effect on cilia differentiation is very modest - a shortening of cilia length on XAP5, NONO and SOX5 knockout - which can happen for a variety of reasons, especially in culture conditions. In my view, this relatively mild phenotype does not establish that the XAP5/NONO and SOX5 axis is an important regulator of primary cilia differentiation.

(3) The lack of any data that validates the findings in the model vertebrate is a major weakness of this paper. Validation using clean genetics (whole body knockouts or tissue-specific conditional knockouts) is absolutely essential for these data to be acceptable.

Reviewer #2 (Public review):

Summary:

This study investigates how evolutionarily conserved transcription factors are repurposed to regulate the functional diversification of cilia. Building on previous work identifying Xap5 as a regulator of motile ciliogenesis during spermatogenesis, the authors now propose a broader role for Xap5 as a master regulator of primary ciliogenesis. Through extensive mechanistic analyses, they identify an Xap5-NONO-SOX transcriptional axis and suggest that this module contributes to ciliary diversity and may be implicated in ciliopathies.

Overall, the work addresses an important and timely question regarding the transcriptional control of primary ciliogenesis. However, additional evidence is required to fully support the proposed conceptual framework linking evolutionary conservation to functional specialization.

Strengths:

(1) Addresses a timely and fundamental question in cilia biology.

(2) Extends Xap5 function beyond motile ciliogenesis.

(3) Identifies a novel regulatory axis (Xap5-NONO-SOX).

(4) Combines multiple well-designed mechanistic approaches.

(5) Proposes an interesting conceptual framework linking evolution and ciliogenesis.

Weaknesses:

(1) Specificity for primary ciliogenesis not demonstrated.

(2) No data on motile ciliogenesis in somatic MCCs.

(3) Conclusions drawn from NIH/3T3 cells (murine stromal cells).

(4) GC-rich motif identified but underexplored.

(5) Link to ciliopathies is speculative.

eLife Assessment

These findings are important because they suggest that more selective JAK inhibition, particularly targeting JAK1 or JAK2, can effectively reduce organ pathology and pathogenic IFN-γ-producing immune cells in AIRE deficiency, refining therapeutic strategies beyond broad JAK inhibition. The work highlights JAK2 inhibition as a promising and potentially more targeted clinical approach for treating autoimmunity in this setting. The evidence is solid and moderately strong, building on the prior efficacy of ruxolitinib and supported by comparative studies in Aire-deficient models, though further validation in human systems would strengthen translational confidence.

Reviewer #1 (Public review):

Summary:

Heller et al use a murine model of AIRE deficiency, a disease that leads to systemic autoimmune disease, to demonstrate differential effects of selective JAK inhibitors. This group and others have previously demonstrated the efficacy of the JAK1/2 inhibitor ruxolitinib in patients with AIRE deficiency. Here, they focus on the ability of ruxolitinib versus drugs inhibiting either JAK1, JAK2, or JAK3 to alter organ pathology and accumulation of interferon-gamma producing immune cells in the lungs, which are important mediators of inflammation in patients with this disease. The current study provides evidence that selective JAK2 or JAK1 both reduce disease in this mouse model. There is potentially a more beneficial effect of selective JAK2 inhibition, although these differences are minor, and it is uncertain whether this is clinically relevant for patients. They demonstrate that inhibition of JAK3 alone in the mouse was clearly not beneficial for disease. Overall, this study provides evidence for consideration of more selective JAK inhibition in patients with AIRE deficiency.

Strengths:

(1) Robust model for investigating AIRE deficiency.

(2) They combine cellular studies (immune cell production of IFN-g) and robust organ pathology scoring to evaluate the effects of the drugs tested here.

(3) Data clearly demonstrates that JAK3 inhibition, at least as used here, may increase IFN-g production and does not reduce organ pathology.

Weaknesses:

(1) There is no direct comparison of the effects of JAK2 vs. JAK1 inhibition to support that JAK2 inhibition is clearly superior.

(2) They were not able to perform pharmacokinetic studies or measure the efficacy of JAK inhibition in their model, and it is uncertain how the doses of drug used here will translate to the treatment of patients.

(3) It is uncertain whether this study, performed in a murine model, will correspond to tissue/cell specificity of JAK inhibition in patients.

Reviewer #2 (Public review):

Summary:

This work from Heller et al. examines the differential responses of treatment with selective JAK inhibitors in Aire knockout mice, which develop several autoimmune diseases. The authors had previously shown efficacious responses in both mice and humans with a broader JAK-I, Ruxolitinib, that had Aire-deficiency. Because of the side effect profile, it may be better to determine if selective JAK-I therapy could continue to work with less of the side effects of Ruxolitinib. Here, they develop a protocol of treating mice for four weeks with JAK1,2, and 3 inhibitors and then examining tissues for infiltration of T cells and gamma-interferon-producing T cells. They also perform analyses of infiltration of the tissues versus intravascular localization of T cells. They find that JAK2 inhibition provided the most robust results for decreasing infiltrates and gamma interferon-producing T cells. All JAK-I's resulted in decreased T cell infiltration of tissues, and somewhat paradoxically, the JAK3 inhibitor caused an increased accumulation of gamma-interferon-producing T cells in tissues.

Strengths:

This is a nice set of studies that makes some inroads on a more refined approach to treating autoimmunity in the Aire knockout model. The work here will be important for developing the next clinical trial for patients with APS1 and represents an advance for efforts in that space.

Weaknesses:

The increase in gamma-interferon-producing cells in tissues with JAK3 inhibition is interesting, but essentially remains unanswered in any way. There is a minimal assessment of the broad STAT pathways that the selective JAK-i's could be hitting, and perhaps that could be assessed more systematically. Finally, there is no pharmacokinetic data, which makes comparisons between the treatments a bit limited.

eLife Assessment

This is a valuable study that investigates the role of the long non-coding RNA Dreg1 for the development, differentiation, or maintenance of group 2 ILC (ILC2). The authors generate Dreg1-/- mice and show solid evidence for a reduction of group 2 innate lymphoid cells (ILC2). However, the strength of evidence supporting and analysing the impact of Dreg1 on Gata3 expression, a transcription factor required for ILC2 cell fate decisions, remains incomplete. This study will be of interest to immunologists.

Reviewer #1 (Public review):

Summary:

This study examines the role of the long non-coding RNA Dreg1 in regulating Gata3 expression and ILC2 development. Using Dreg1 deficient mice, the authors show a selective loss of ILC2s but not T or NK cells, suggesting a lineage-specific requirement for Dreg1. By integrating public chromatin and TF-binding datasets, they propose a Tcf1-Dreg1-Gata3 regulatory axis. The topic is relevant for understanding epigenetic regulation of ILC differentiation.

Strengths:

(1) Clear in vivo evidence for a lineage-specific role of Dreg1.

(2) Comprehensive integration of genomic datasets.

(3) Cross-species comparison linking mouse and human regulatory regions.

Weaknesses:

(1) Mechanistic conclusions remain correlative, relying on public data.

(2) Lack of direct chromatin or transcriptional validation of Tcf1-mediated regulation.

(3) Human enhancer function is not experimentally confirmed.

(4) Insufficient methodological detail and limited mechanistic discussion.

Comments on revisions:

The authors have provided clear evidence that Dreg1 is necessary for ILC2 development, but their refusal to perform any mechanistic experiment remains a significant weakness. While their appeal to the 3Rs and the use of public datasets is noted, re-analyzing external data from heterogeneous sources cannot substitute for direct, internal validation of the Tcf1-Dreg1-Gata3 axis in their specific knockout model. This is particularly problematic because ILC2 progenitors, though rare, can be isolated from bone marrow, especially since assays like CUT&Tag and others are specifically designed for low cell numbers. By relying on public T-cell CRISPR screens to justify human ILC2 functions, the authors are substituting cross-cell-type correlation for definitive functional proof. Consequently, the manuscript currently describes a discovery of necessity without providing a verified molecular mechanism, which should be more explicitly reflected in the title and conclusions.

Reviewer #2 (Public review):

The authors investigate the role of the long non-coding RNA Dreg1 for the development, differentiation or maintenance of group 2 ILC (ILC2). Dreg1 is encoded close to the Gata3 locus, a transcription factor implicated in the differentiation of T cells and ILC, and in particular of type 2 immune cells (i.e., Th2 cells and ILC2). The center of the paper is the generation of a Dreg1-deficient mouse. The role of Dreg1 in ILC2 was documented by mixed bone marrow experiments. While Dreg1-/- mice did not show any profound ab T or gd T cell, ILC1, ILC3 and NK cell phenotypes, ILC2 frequencies were reduced in various organs tested (small intestine, lung, visceral adipose tissue). In the bone marrow, immature ILC2 or ILC2 progenitors were reduced whereas a common ILC progenitor was overrepresented suggesting a differentiation block. Using ATAC-seq, the authors find the promoter of Dreg1 is open in early lymphoid progenitors and the acquisition of chromatin accessibility downstream correlates with increased Dreg1 expression in ILC2 progenitors. Examining publicly available Tcf1 CUT&Run data, they find that Tcf1 was specifically bound to the accessible sites of the Dreg1 locus in early innate lymphoid progenitors. Finally, the syntenic region in the human genome contains two non-coding RNA genes with an expression pattern resembling mouse Dreg1.

The topic of the manuscript is interesting. The article is focused on the first description of the Dreg1 knockout mouse and the specific effect of Dreg1 deficiency on ILC2 development.

(1) The data of how Dreg1 contributes to the differentiation and or maintenance of ILC2 is not addressed at a very definitive level. Does Dreg1 affect Gata3 expression, mRNA stability or turnover in ILC2? Previous work of the authors indicated that knock-down of Dreg1 does not affect Gata3 expression (PMID: 32970351). The current data (Figure 2H) showed small differences in Gata3 expression in CHILP which were, however, not statistically significant. No differences were found in ILCP and ILC2P.

(2) How Dreg1 exactly affects ILC2 differentiation remains unclear.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study examines the role of the long non-coding RNA Dreg1 in regulating Gata3 expression and ILC2 development. Using Dreg1-deficient mice, the authors show a selective loss of ILC2s but not T or NK cells, suggesting a lineage-specific requirement for Dreg1. By integrating public chromatin and TF-binding datasets, they propose a Tcf1-Dreg1-Gata3 regulatory axis. The topic is relevant for understanding epigenetic regulation of ILC differentiation.

Strengths:

(1) Clear in vivo evidence for a lineage-specific role of Dreg1.

(2) Comprehensive integration of genomic datasets.

(3) Cross-species comparison linking mouse and human regulatory regions.

Weaknesses:

(1) Mechanistic conclusions remain correlative, relying on public data.

We agree that the mechanistic conclusions are of our study are indeed correlative and we mention this in the discussion. The primary work of the study is the discovery of Dreg1's necessity for ILC2 development via the new knockout mouse model. Re-analysing good quality publicly available data on rare cell populations is an appropriate approach and in line with DORA guidelines for ethical research.

(2) Lack of direct chromatin or transcriptional validation of Tcf1-mediated regulation.

The most appropriate way to examine direct Tcf1 target genes in primary cells is to examine the association of Tcf1 binding with the changes that occur in Tcf1-bound genes after Tcf7 knockout. By analysing publicly available data on ILC progenitors we indeed did this. We revealed that Tcf1 bound to Dreg1 and that Dreg1 was not expressed when Tcf1 was knocked out in ILC progenitors. In addition we examined H3K27ac at the Dreg1 locus in the same ILC progenitors to demonstrate that Tcf1 appears to be important for decorating the Dreg1 gene with this histone modification. We believe that this analysis is sufficient to conclude that Tcf1 is required for the expression of Dreg1 in ILC progenitors.

(3) Human enhancer function is not experimentally confirmed.

We agree that the potential human enhancer of GATA3 we identified has not been confirmed in human ILC. However, a previous study showed clear evidence that this region has GATA3 enhancer activity in human T cells. Therefore, while not specific to ILC2s the region where the DREG1 homologues lie does indeed harbour enhancer activity.

(4) Insufficient methodological detail and limited mechanistic discussion.

We have now made the changes suggested by the reviewer to both the methods/figure legends and also the discussion.

Reviewer #1 (Recommendations for the authors):

The authors generated Dreg1-deficient mice and demonstrated that loss of this locus selectively reduces ILC2s but not T or NK cells, indicating a lineage-specific requirement for Dreg1 in ILC development. By analyzing publicly available chromatin accessibility and transcription factor-binding datasets, they link Dreg1 expression to Tcf1-dependent chromatin activation and extend their findings to human data by identifying a syntenic GATA3 enhancer that produces homologous Dreg lncRNAs in ILC2s. While the study addresses an interesting question, most of the mechanistic interpretations rely heavily on publicly available datasets rather than the authors' own functional evidence. To establish causality and reinforce the overall conclusions, I provide below some comments and suggestions for additional experiments and clarifications that would considerably strengthen the manuscript.

(1) In Figure 3, the authors use public datasets to argue that Tcf1 regulates Dreg1 expression by modulating chromatin accessibility and H3K27ac at its locus. However, since these data are derived from heterogeneous external sources, the conclusions remain associative. To better support causality, the authors should generate matched datasets from their own sorted progenitor populations and perform CUT&Tag for Tcf1 and H3K27ac in wild-type and Tcf7 knockout progenitors to directly test whether Tcf1 binding establishes an active chromatin state at Dreg1. Also, complementing this with nascent RNA or pre-mRNA quantification would link chromatin activation to transcriptional output. These experiments are technically feasible in progenitors and would substantially strengthen the claim that Tcf1 directly drives Dreg1 activation during ILC development.

We believe that utilising publicly available data sufficiently answers this question while also adhering to ethical considerations. The ILC populations used to produce the publicly available data were akin to those we examined in our analyses, and the data was of sufficient quality. Moreover, they enable us to access data from Tcf1-deficient mice. Redoing large-scale chromatin profiling on rare cell types would require hundreds of mice to achieve sufficient cell numbers. Repeating this solely for “originality” contradicts the 3Rs principles (replacement, reduction, refinement) if high quality public data already exists and we feel will require years of redundant work. In addition, we believe the fact that the data derive from heterogenous external sources, yet align well, only strengthen our conclusions. We have now added mention to our use of publicly available data in the discussion.

(2) In Figure 4, the authors provide correlative evidence from public datasets suggesting that the human region syntenic to the murine Dreg1 locus acts as a distal enhancer of GATA3 and gives rise to two ILC2-specific lncRNAs. To substantiate this claim, the authors should perform CUT&Tag for H3K27ac in human ILC2s to confirm enhancer activation and use 3C or HiChIP to demonstrate physical interaction with the GATA3 promoter. These experiments should be doable by fusing pooled ILC2 samples and would provide more direct evidence that this region actively regulates GATA3 expression.

Assessing the activity of a distal enhancer region on its target gene in primary human cells is extremely difficult, due to a number of technical and biological complications such as enhancer redundancy. This is why we chose to reanalyse an extensive enhancer deletion screen performed in human T cells by Chen et al., AJHG 2023. This analysis clearly showed deletion of the region we identified as harbouring Dreg1 homologues affected GATA3 expression, thus confirming its enhancer activity. While we agree with the reviewer that specific profiling of human ILC populations for H3K27ac and 3D genome architecture would provide further correlative evidence this will be a time-consuming and costly endevour with human material and ultimately the definitive proof in ILCs would require specific deletion of this region in ILC2s. We have mentioned this caveat in the discussion.

(3) Several figure legends lack essential methodological details. Figure 1 should specify how NK and ILC populations were gated, including intermediate steps and markers used. The same applies to Supplementary Figure 1, and particularly to Supplementary Figure 2, where gating strategies for progenitors are shown but not explained. Figure 2 should also indicate that these analyses were performed in bone marrow. Clearer legends are crucial for interpreting and reproducing the data.

We have made the suggested changes.

(4) It is also unclear throughout the manuscript whether the authors performed any ATACseq experiments themselves or relied entirely on public datasets. This information should be stated explicitly in the main text and figure legends, not only in the Methods section. Similarly, the source of the ChIPseq or CUT&Run datasets should be clearly indicated alongside the relevant figures.

We apologise for not making this clearer and have now clearly articulated if the data was public in the text.

(5) As the authors themselves suggest, performing experiments that selectively suppress Dreg1 transcription using antisense oligonucleotides or CRISPR interference at the Dreg1 promoter would provide more valuable mechanistic insights. Conducting these experiments in their own system would allow them to determine whether Dreg1 functions through its RNA product or as a DNA enhancer element, thereby strengthening the causal link between Dreg1 activity and Gata3 regulation.

We agree with the reviewer, however, this, in our opinion is beyond the scope of this manuscript. The strength of this manuscript lies in the findings from the novel Dreg1 knockout mouse strain. Future studies will focus on understanding how Dreg1 influences Gata3 expression.

(6) The discussion would benefit from a clearer and more integrated explanation of how Dreg1 fits into the transcriptional network that controls ILC2 differentiation. The authors could elaborate on whether Dreg1 fine-tunes Gata3 expression or functions as part of a regulatory loop with Tcf1, and better explain how this mechanism might be conserved in humans. In addition, the authors should explicitly acknowledge the limitations of relying on publicly available datasets and emphasize the need for direct experimental validation to support their mechanistic interpretation.

We have now made these suggested inclusions.

Reviewer #2 (Public review):

The authors investigate the role of the long non-coding RNA Dreg1 for the development, differentiation, or maintenance of group 2 ILC (ILC2). Dreg1 is encoded close to the Gata3 locus, a transcription factor implicated in the differentiation of T cells and ILC, and in particular of type 2 immune cells (i.e., Th2 cells and ILC2). The center of the paper is the generation of a Dreg1-deficient mouse. While Dreg1-/- mice did not show any profound ab T or gd T cell, ILC1, ILC3, and NK cell phenotypes, ILC2 frequencies were reduced in various organs tested (small intestine, lung, visceral adipose tissue). In the bone marrow, immature ILC2 or ILC2 progenitors were reduced, whereas a common ILC progenitor was overrepresented, suggesting a differentiation block. Using ATAC-seq, the authors find that the promoter of Dreg1 is open in early lymphoid progenitors, and the acquisition of chromatin accessibility downstream correlates with increased Dreg1 expression in ILC2 progenitors. Examining publicly available Tcf1 CUT&Run data, they find that Tcf1 was specifically bound to the accessible sites of the Dreg1 locus in early innate lymphoid progenitors. Finally, the syntenic region in the human genome contains two non-coding RNA genes with an expression pattern resembling mouse Dreg1.

The topic of the manuscript is interesting. However, there are various limitations that are summarized below.

(1) The authors generated a new mouse model. The strategy should be better described, including the genetic background of the initially microinjected material. How many generations was the targeted offspring backcrossed to C57BL/6J?

The mice were backcrossed for at least 2 generations to C57BL/6. This information is now included in the methods section.

(2) The data is obtained from mice in which the Dreg1 gene is deleted in all cells. A cell-intrinsic role of Dreg1 in ILC2 has not been demonstrated. It should be shown that Dreg1 is required in ILC2 and their progenitors.

We now provide new mixed bone marrow irradiation chimera data that shows that the effect is intrinsic to Dreg1-deficient ILC2 cells (Figure 1F and Supplementary Figure 1E-G).

(3) The data on how Dreg1 contributes to the differentiation and or maintenance of ILC2 is not addressed at a very definitive level. Does Dreg1 affect Gata3 expression, mRNA stability, or turnover in ILC2? Previous work of the authors indicated that knockdown of Dreg1 does not affect Gata3 expression (PMID: 32970351).

We have indeed shown that Dreg1-deficient ILC2P have reduced levels of Gata3 (Figure 2H) however we have not determined the exact mechanisms by which Dreg1 controls ILC2 development.

(4) How Dreg1 exactly affects ILC2 differentiation remains unclear.

We agree with the reviewer, however, this article is focused on the first description of the Dreg1 knockout mice and the surprisingly specific effect on ILC2 development.

Reviewer #2 (Recommendations for the authors):

(1) Relating to point 2 of public review:

It should be shown that Dreg1 is required in ILC2 and their progenitors. Mixed bone marrow chimeras would be an adequate strategy.

We have now done this and clearly showed that the effect is intrinsic to Dreg1-deficient ILC2s.

(2) Relating to point 3 of public review:

Minimally, Gata3 expression should be analyzed in ILC2, ILC2P, and the ILC progenitors by qRT-PCR and antibody stain.

We have indeed shown reduced Gata3 levels by antibody stain in Figure 2H.

(3) Relating to point 4 of public review:

The manuscript would benefit from additional data studying ILC2 differentiation in (competitive) adoptive transfer experiments or using in vitro differentiation assays.

We have performed the mixed bone marrow chimera experiments which are testing the competitiveness of Dreg1-deficient bone barrow with control wildtype. In this case the WT ILC2s outcompeted the Dreg1-deficient ILC2s for the same niche.

eLife Assessment

This valuable study reports a spatiotemporal atlas of mouse placental development and explores the role of glycogen trophoblast cells in fetal viability. Solid data are presented to support the main conclusion. This work will be of great interest to developmental DNA reproductive biologists.

Reviewer #1 (Public review):

In this manuscript, the authors combine single-nucleus RNA sequencing with spatial transcriptomics to generate a spatiotemporal atlas of mouse placental development and explore the role of glycogen trophoblast cells in fetal viability. The study integrates several computational approaches, including trajectory analysis, regulatory network inference, and spatial mapping, together with histology and glycogen measurements. Based on these analyses, the authors propose that glycogen trophoblast cells provide metabolic support that is important for maintaining placental function and fetal survival.

One of the main strengths of the study is the quality and scope of the dataset. The integration of snRNA-seq with Stereo-seq spatial transcriptomics provides a detailed view of placental organization across regions and developmental stages. This type of combined spatial and transcriptional analysis is still relatively rare in placental biology and represents an important contribution to the field. The atlas itself will likely be a valuable resource for future studies.

Another strength is the effort to connect transcriptional findings with tissue-level validation. The glycogen staining and biochemical measurements support the interpretation that glycogen trophoblast cells contribute to placental metabolic function. The spatial analyses identifying macrophage accumulation in the labyrinth region of mutant placentas are also interesting and illustrate how spatial approaches can reveal microenvironmental changes that are difficult to detect otherwise.

The main limitation of the study is that the conclusion that glycogen cells are essential mediators of metabolic support for fetal viability remains partly indirect. The transcriptomic and spatial data strongly suggest a role for these cells, but it is still difficult to determine whether glycogen cell dysfunction is the primary cause of fetal lethality or a consequence of broader placental abnormalities. Clarifying this point would strengthen the central message of the paper.

Similarly, the macrophage accumulation observed in the labyrinth appears consistent with a response to tissue stress or injury, but its relationship to glycogen cell function is not fully explained. A clearer discussion of whether this represents a primary mechanism or a secondary effect would improve the interpretation.

Overall, this is a strong dataset and a useful spatial atlas of placental development. The study provides convincing descriptive insight into glycogen trophoblast biology, and with some clarification of the mechanistic conclusions, the manuscript will be even stronger.

Reviewer #2 (Public review):

This manuscript constructs a spatiotemporal transcriptomic atlas (STAMP) of the mouse placenta from E9.5-E18.5 by integrating Stereo-seq and snRNA-seq, and identifies two glycogen trophoblast cell (GC) subtypes (GC-1 and GC-2), a spatial transition from the junctional zone (JZ) to the decidua, and metabolic defects in Ano6-null placentas including GC persistence, glycogen accumulation, reduced glycogenolysis metabolites, and partial rescue by maternal glucose supplementation. The breadth of the dataset and the integration of atlas construction with PAS/TEM/LC-MS analyses are impressive, and the study has the potential to provide a valuable resource for the placental biology community.

However, in its current form, the central claim that "GC-mediated metabolic support is essential/indispensable for fetal viability" is not sufficiently disentangled from the complex phenotype of a global Ano6 knockout model. In addition, the stage-level biological replication in the atlas and the claim of "single-cell resolution" require more careful presentation. Therefore, while the study is interesting and potentially impactful, substantial revisions are required, particularly to recalibrate the strength of the conclusions and causal interpretations.

Major comments

(1) The most significant concern is that the manuscript overinterprets the phenotype observed in a global Ano6 knockout as direct evidence that GC glycogen metabolism is essential for fetal viability. The authors themselves report multiple severe placental abnormalities in the knockout, including reduced placental size and weight, structural defects in the labyrinth, impaired vascularization, and accumulation of abnormal regions. Previous studies cited in the manuscript also indicate that Ano6 deficiency leads to defects in syncytiotrophoblast formation, impaired maternofetal exchange, and perinatal lethality.

In this context, the current data support an association between GC metabolic defects and fetal lethality, but do not establish that GC glycogen metabolism is the primary causal driver. The conclusion should therefore be moderated (e.g., "contributes to" rather than "is essential for"), unless additional placenta-specific or GC-specific functional validation is provided.

(2) Maternal glucose supplementation is an interesting functional experiment, but in its current form, it provides supportive rather than definitive mechanistic evidence. While survival improves (from ~3% to ~10%), the rescue remains partial. Moreover, the readouts are largely limited to metabolite restoration (glucose, G1P, G6P) in the placenta and fetal liver.

To support a stronger causal claim, the authors should assess whether glucose supplementation also rescues: placental morphology (especially labyrinth structure), GC number and PAS staining, ultrastructural glycogen features (TEM), fetal growth and developmental outcomes.

(3) The atlas is constructed from nine placentas across developmental stages, suggesting limited biological replication per stage. It remains unclear how robust the observed temporal trends are to litter effects, sex differences, or sectioning variability.

Furthermore, the "single-cell resolution" is not directly measured but inferred via image segmentation and reference-based mapping (e.g., TACCO). This should be more explicitly stated, as it represents computational inference rather than direct single-cell measurement.

The authors should:<br /> - clearly report biological replicates per stage (including litter and sex),<br /> - demonstrate reproducibility of key patterns across independent samples,<br /> - refine the wording to reflect segmentation- and reference-based single-cell inference.

(4) The proposed developmental trajectory (JZ progenitor → GC precursor → GC-1 → GC-2) and the claim of GC migration from JZ to decidua are based on spatial distribution and computational trajectory analyses (Monocle, CytoTRACE).

While this is a compelling model, it remains inferential. The language throughout the manuscript should be softened (e.g., "consistent with spatial transition" rather than "migration"). Ideally, additional experimental validation, such as stage-resolved RNAscope/immunostaining quantification or lineage tracing, would strengthen this claim.

(5) The manuscript concludes that ANO6 deficiency leads to impaired glycogen utilization, based primarily on the observation that differentiation markers and glycogenolytic enzyme transcripts are unchanged.

However, this demonstrates what is not altered rather than what is mechanistically responsible for the defect. A more direct mechanistic link is needed, such as changes in enzyme activity, altered intracellular localization, effects on ion homeostasis or membrane biology.

(6) The statistical framework requires clarification. Several analyses use n = 4-8 placentas or "independent experiments," but it is unclear whether these represent independent litters or multiple samples from the same dam.

Given the risk of pseudoreplication in placental studies, the authors should define whether n refers to placentas or litters, report the number of dams per genotype, and ensure appropriate statistical treatment (e.g., litter-based analysis or mixed-effects models).

Author response:

eLife Assessment

This valuable study reports a spatiotemporal atlas of mouse placental development and explores the role of glycogen trophoblast cells in fetal viability. Solid data are presented to support the main conclusion. This work will be of great interest to developmental DNA reproductive biologists.

We thank the editors for this positive and balanced assessment of our study. We are encouraged that the spatiotemporal mouse placental atlas and the functional analysis of glycogen trophoblast cells were considered valuable, and that the data were viewed as providing solid support for the main conclusions.

In the revised manuscript, we will further clarify the scope of these conclusions, particularly regarding the contribution of GC-associated glycogen metabolism to fetal viability in the global Ano6 knockout model. We will also refine the wording where needed to ensure that the mechanistic interpretation accurately reflects the strength of the available evidence.

Public Reviews:

Reviewer #1 (Public review):

In this manuscript, the authors combine single-nucleus RNA sequencing with spatial transcriptomics to generate a spatiotemporal atlas of mouse placental development and explore the role of glycogen trophoblast cells in fetal viability. The study integrates several computational approaches, including trajectory analysis, regulatory network inference, and spatial mapping, together with histology and glycogen measurements. Based on these analyses, the authors propose that glycogen trophoblast cells provide metabolic support that is important for maintaining placental function and fetal survival.

One of the main strengths of the study is the quality and scope of the dataset. The integration of snRNA-seq with Stereo-seq spatial transcriptomics provides a detailed view of placental organization across regions and developmental stages. This type of combined spatial and transcriptional analysis is still relatively rare in placental biology and represents an important contribution to the field. The atlas itself will likely be a valuable resource for future studies.

Another strength is the effort to connect transcriptional findings with tissue-level validation. The glycogen staining and biochemical measurements support the interpretation that glycogen trophoblast cells contribute to placental metabolic function. The spatial analyses identifying macrophage accumulation in the labyrinth region of mutant placentas are also interesting and illustrate how spatial approaches can reveal microenvironmental changes that are difficult to detect otherwise.

The main limitation of the study is that the conclusion that glycogen cells are essential mediators of metabolic support for fetal viability remains partly indirect. The transcriptomic and spatial data strongly suggest a role for these cells, but it is still difficult to determine whether glycogen cell dysfunction is the primary cause of fetal lethality or a consequence of broader placental abnormalities. Clarifying this point would strengthen the central message of the paper.

Similarly, the macrophage accumulation observed in the labyrinth appears consistent with a response to tissue stress or injury, but its relationship to glycogen cell function is not fully explained. A clearer discussion of whether this represents a primary mechanism or a secondary effect would improve the interpretation.

Overall, this is a strong dataset and a useful spatial atlas of placental development. The study provides convincing descriptive insight into glycogen trophoblast biology, and with some clarification of the mechanistic conclusions, the manuscript will be even stronger.

We thank the reviewer for this constructive assessment of our manuscript. We are pleased that the reviewer recognized the quality and scope of the dataset, particularly the integration of snRNA sequencing with Stereo-seq spatial transcriptomics to generate a spatiotemporal atlas of mouse placental development. We also appreciate the reviewer’s view that this atlas represents a valuable resource for the placental biology and developmental biology communities. We also appreciate the reviewer’s important point that the causal relationship between glycogen trophoblast cell dysfunction, placental metabolic impairment, and fetal viability should be presented with appropriate caution. In the revised manuscript, we will clarify that our data support a strong association between impaired glycogen trophoblast cell function, altered placental glycogen metabolism, and fetal lethality in the global Ano6 knockout model, but do not by themselves establish glycogen trophoblast dysfunction as the sole or primary cause of fetal loss. We will revise the relevant sections to avoid overstatement and to distinguish more clearly between direct experimental evidence, correlative spatial-transcriptomic observations, and mechanistic interpretation. Similarly, we agree that the macrophage accumulation observed in the labyrinth region is most appropriately interpreted as a spatially localized immune or tissue-stress response in the mutant placenta. In the revised manuscript, we will expand the discussion to clarify that, while this observation may reflect downstream consequences of placental dysfunction and altered tissue homeostasis, the current data do not establish macrophage accumulation as a primary mechanism linking glycogen trophoblast defects to fetal lethality. We will therefore frame this finding as an important microenvironmental alteration revealed by the spatial atlas, rather than as definitive evidence of a direct causal pathway.

Reviewer #2 (Public review):

This manuscript constructs a spatiotemporal transcriptomic atlas (STAMP) of the mouse placenta from E9.5-E18.5 by integrating Stereo-seq and snRNA-seq, and identifies two glycogen trophoblast cell (GC) subtypes (GC-1 and GC-2), a spatial transition from the junctional zone (JZ) to the decidua, and metabolic defects in Ano6-null placentas including GC persistence, glycogen accumulation, reduced glycogenolysis metabolites, and partial rescue by maternal glucose supplementation. The breadth of the dataset and the integration of atlas construction with PAS/TEM/LC-MS analyses are impressive, and the study has the potential to provide a valuable resource for the placental biology community.

However, in its current form, the central claim that "GC-mediated metabolic support is essential/indispensable for fetal viability" is not sufficiently disentangled from the complex phenotype of a global Ano6 knockout model. In addition, the stage-level biological replication in the atlas and the claim of "single-cell resolution" require more careful presentation. Therefore, while the study is interesting and potentially impactful, substantial revisions are required, particularly to recalibrate the strength of the conclusions and causal interpretations.

Major comments

(1) The most significant concern is that the manuscript overinterprets the phenotype observed in a global Ano6 knockout as direct evidence that GC glycogen metabolism is essential for fetal viability. The authors themselves report multiple severe placental abnormalities in the knockout, including reduced placental size and weight, structural defects in the labyrinth, impaired vascularization, and accumulation of abnormal regions. Previous studies cited in the manuscript also indicate that Ano6 deficiency leads to defects in syncytiotrophoblast formation, impaired maternofetal exchange, and perinatal lethality.

In this context, the current data support an association between GC metabolic defects and fetal lethality, but do not establish that GC glycogen metabolism is the primary causal driver. The conclusion should therefore be moderated (e.g., "contributes to" rather than "is essential for"), unless additional placenta-specific or GC-specific functional validation is provided.

(2) Maternal glucose supplementation is an interesting functional experiment, but in its current form, it provides supportive rather than definitive mechanistic evidence. While survival improves (from ~3% to ~10%), the rescue remains partial. Moreover, the readouts are largely limited to metabolite restoration (glucose, G1P, G6P) in the placenta and fetal liver.

To support a stronger causal claim, the authors should assess whether glucose supplementation also rescues: placental morphology (especially labyrinth structure), GC number and PAS staining, ultrastructural glycogen features (TEM), fetal growth and developmental outcomes.

(3) The atlas is constructed from nine placentas across developmental stages, suggesting limited biological replication per stage. It remains unclear how robust the observed temporal trends are to litter effects, sex differences, or sectioning variability.

Furthermore, the "single-cell resolution" is not directly measured but inferred via image segmentation and reference-based mapping (e.g., TACCO). This should be more explicitly stated, as it represents computational inference rather than direct single-cell measurement.

The authors should:

- clearly report biological replicates per stage (including litter and sex),

- demonstrate reproducibility of key patterns across independent samples,

- refine the wording to reflect segmentation- and reference-based single-cell inference.

(4) The proposed developmental trajectory (JZ progenitor → GC precursor → GC-1 → GC-2) and the claim of GC migration from JZ to decidua are based on spatial distribution and computational trajectory analyses (Monocle, CytoTRACE).

While this is a compelling model, it remains inferential. The language throughout the manuscript should be softened (e.g., "consistent with spatial transition" rather than "migration"). Ideally, additional experimental validation, such as stage-resolved RNAscope/immunostaining quantification or lineage tracing, would strengthen this claim.

(5) The manuscript concludes that ANO6 deficiency leads to impaired glycogen utilization, based primarily on the observation that differentiation markers and glycogenolytic enzyme transcripts are unchanged.

However, this demonstrates what is not altered rather than what is mechanistically responsible for the defect. A more direct mechanistic link is needed, such as changes in enzyme activity, altered intracellular localization, effects on ion homeostasis or membrane biology.

(6) The statistical framework requires clarification. Several analyses use n = 4-8 placentas or "independent experiments," but it is unclear whether these represent independent litters or multiple samples from the same dam.

Given the risk of pseudoreplication in placental studies, the authors should define whether n refers to placentas or litters, report the number of dams per genotype, and ensure appropriate statistical treatment (e.g., litter-based analysis or mixed-effects models).

We thank the Reviewer for the careful evaluation of our manuscript and for recognizing the breadth of the STAMP dataset and the value of integrating spatial transcriptomics, snRNA-seq, PAS, TEM and LC-MS analyses.

We agree that the current manuscript overstates some mechanistic conclusions. In the revision, we will moderate the central claim and more clearly acknowledge that the global Ano6 knockout model has complex placental defects.

Comment 1: Causality in the global Ano6 knockout model

We agree that our current data do not prove that GC glycogen metabolism is the primary cause of fetal lethality in the global Ano6 knockout model. In the revised manuscript, we will avoid presenting GC dysfunction as the sole causal mechanism. We will replace stronger terms such as “essential” or “indispensable” with more measured wording such as “contributes to” or “supports.” We will frame impaired GC-associated glycogen metabolism as one important component of Ano6-null placental dysfunction.

Comment 2: Maternal glucose supplementation

We agree that maternal glucose supplementation provides supportive, but not definitive, mechanistic evidence. In the revision, we will describe the partial survival rescue more cautiously and will not use it as proof of GC-specific causality. Where possible, we will also assess whether glucose supplementation affects additional phenotypes, including fetal growth, placental morphology, GC abundance and PAS/glycogen readouts.

Comment 3: Biological replication and single-cell resolution

We agree that the replication structure and the wording of “single-cell resolution” need clarification. We will report the number of placentas, litters and available sex information for each stage. We will also revise the wording to make clear that the spatial single-cell annotation is based on image segmentation and snRNA-seq reference mapping, rather than direct single-cell measurement by Stereo-seq alone.

Comment 4: GC trajectory and spatial transition

We agree that the proposed GC trajectory and JZ-to-decidua transition remain inferential. We will soften the language throughout the manuscript, using terms such as “spatial transition,” “redistribution,” or “consistent with migration” rather than stating that migration has been directly proven.

Comment 5: Mechanism of impaired glycogen utilization

We agree that unchanged GC markers and glycogenolytic enzyme transcripts do not reveal the direct mechanism. In the revision, we will state more clearly that these data argue against gross GC differentiation defects or transcriptional loss of glycogenolytic enzymes, but that the direct mechanism may involve enzyme activity, localization, ion homeostasis or ANO6-dependent membrane biology.

Comment 6: Statistical framework

We agree that the statistical framework needs clearer reporting. We will define what each n represents, including placenta, section, litter, dam or independent experiment, and will revise the analysis or description where needed to minimize concerns about pseudoreplication.

Overall, we appreciate these comments and will use them to make the revised manuscript more precise, transparent and appropriately cautious.

eLife Assessment

This important study introduces LUNA, a new autofocusing method that achieves nanoscale precision and robustly corrects focus drift during time-lapse microscopy, improving imaging under temperature shifts. The authors exploit this technical advance to investigate the bacterial cold shock response, providing convincing evidence that individual cells continue to grow and divide in a highly coordinated process that cannot be observed in population-level measurements. This work offers a technical and conceptual framework for reconciling discrepancies between bulk and single-cell growth measurements, with broad relevance for cell biology and microbiology.

Reviewer #1 (Public review):

Summary:

The authors present a new autofocusing method, LUNA (Locking Under Nanoscale Accuracy), designed to overcome severe focus drift, a major challenge in long-term time-lapse microscopy. Using this method, they address a fundamental question in bacterial cold shock response: whether cells halt growth and division following an abrupt temperature downshift. Through single-cell analysis, the authors uncover a multi-phase adaptation process with distinct growth deceleration dynamics, and show that bacterial cells adapt to cold shock in a largely uniform manner across the population. Overall, this work provides new insights into the bacterial cold shock response at the single-cell level, extending beyond what can be inferred from population-level measurements.

Strengths:

(1) The LUNA method shows improved performance compared to existing autofocusing systems, achieving nanoscale precision over a large focusing range. Its focusing speed is sufficient for the experiments presented, with potential for further improvement through faster motors and optimized control algorithms, suggesting broad applicability. Theoretical simulations and experimental validation together provide strong support for the method's robustness.

(2) Using LUNA, the authors address a long-standing question in bacterial physiology: whether cells arrest growth and division during the acclimation phase following cold shock. Single-cell analyses across the full course of cold adaptation reveal features that are obscured in bulk-culture studies. Cells continue to grow and divide at reduced rates while maintaining cell size regulation, and exhibit a three-phase adaptation program with distinct growth dynamics. This response appears uniform across the population, with no evidence for bet-hedging. Overall, the experiments are well designed, and the analyses are solid and support the authors' conclusions.

(3) The authors further propose a model describing how population-level optical density (OD) depends on cell dry mass density, volume, and concentration. Following cold shock, cells grow more slowly and exhibit smaller sizes, explaining the apparently unchanged OD. This model provides a valuable conceptual framework for interpreting OD-based growth measurements, a widely used method in microbiology, and will be of broad interest to the field.

Weaknesses:

No major weaknesses identified.

Comments on revisions:

The authors have thoroughly addressed all of my questions. I thank them for their clear clarifications and thoughtful revisions, and I greatly appreciate their efforts in improving the manuscript.

Reviewer #2 (Public review):

Summary:

This study presents LUNA, an autofocus method that compensates for focus drift during rapid temperature changes. Using this approach, the authors show that E. coli cells continue to grow and divide during cold shock, revealing a coordinated, multi-phase adaptation process that could not be deduced from traditional population measurements. They propose a scattering-theory-based model that reconciles the paradox between growth differences of the bacteria at the single-cell level vs population level.

Strengths:

(1) The LUNA approach is pretty creative, turning coma aberration from what is normally a nuisance into an exploit. LUNA enabled long-term single-cell imaging during rapid temperature downshifts.

(2) The authors show that the long-assumed growth arrest during cold shock from population-level measurements is misleading. At the single-cell level, bacteria do not stop growing or dividing but undergo a continuous, three-phase adaptation process. Importantly, this behavior is highly synchronized across the population and not based on bet-hedging.

(3) Finally, the authors propose a model to resolve a long-standing paradox between single-cell vs population behavior: if cells keep growing, why does optical density (OD) of the culture stop increasing? Using light-scattering theory, they show that OD depends not only on cell number but also on cell volume, which decreases after cold shock. As a result, OD can remain flat, or even decrease, despite continued biomass accumulation. This demonstrates that OD is not a reliable proxy for growth under non-steady conditions.

Weaknesses:

(1) While the authors theoretically explain the advantages of LUNA over existing autofocus methods, it is unclear whether practical head-to-head comparisons have been performed, apart from the comparison to Nikon PFS shown in Video S1. As written, the manuscript gives the impression that only LUNA can solve this problem, but such a claim would require more systematic and rigorous benchmarking against alternative approaches.

(2) No mutants/inhibitors used to test and challenge the proposed model.

(3) Cells display a high degree of synchronization, but they are grown in confined microfluidic channels under highly uniform conditions. It is unclear to what extent this synchrony reflects intrinsic biology versus effects imposed by the microfluidic environment.

(4) To further test and generalize the model, it would be informative to also examine bacterial responses at intermediate temperatures rather than focusing primarily on a single cold-shock condition.

Comments on revisions:

The authors have addressed my comments in their response, but have chosen not to incorporate most of them into the manuscript. Readers may refer to the peer review section for further details.

Author response:

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

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors present a new autofocusing method, LUNA (Locking Under Nanoscale Accuracy), designed to overcome severe focus drift, a major challenge in long-term time-lapse microscopy. Using this method, they address a fundamental question in bacterial cold shock response: whether cells halt growth and division following an abrupt temperature downshift. Through single-cell analysis, the authors uncover a multi-phase adaptation process with distinct growth deceleration dynamics, and show that bacterial cells adapt to cold shock in a largely uniform manner across the population. Overall, this work provides new insights into the bacterial cold shock response at the single-cell level, extending beyond what can be inferred from population-level measurements.

Strengths:

(1) The LUNA method shows improved performance compared to existing autofocusing systems, achieving nanoscale precision over a large focusing range. Its focusing speed is sufficient for the experiments presented, with potential for further improvement through faster motors and optimized control algorithms, suggesting broad applicability. Theoretical simulations and experimental validation together provide strong support for the method's robustness.

(2) Using LUNA, the authors address a long-standing question in bacterial physiology: whether cells arrest growth and division during the acclimation phase following cold shock. Single-cell analyses across the full course of cold adaptation reveal features that are obscured in bulk-culture studies. Cells continue to grow and divide at reduced rates while maintaining cell size regulation, and exhibit a three-phase adaptation program with distinct growth dynamics. This response appears uniform across the population, with no evidence for bet-hedging. Overall, the experiments are well designed, and the analyses are solid and support the authors' conclusions.

(3) The authors further propose a model describing how population-level optical density (OD) depends on cell dry mass density, volume, and concentration. Following cold shock, cells grow more slowly and exhibit smaller sizes, explaining the apparently unchanged OD. This model provides a valuable conceptual framework for interpreting OD-based growth measurements, a widely used method in microbiology, and will be of broad interest to the field.

Weaknesses:

No major weaknesses identified.

Comments on revisions:

The authors have thoroughly addressed all of my questions. I thank them for their clear clarifications and thoughtful revisions, and I greatly appreciate their efforts in improving the manuscript.

We sincerely thank the reviewer’s for the encouraging comments and positive assessment. We greatly appreciate the reviewer’s constructive feedback during the review process, which helped us improve the manuscript.

Reviewer #2 (Public review):

Summary:

This study presents LUNA, an autofocus method that compensates for focus drift during rapid temperature changes. Using this approach, the authors show that E. coli cells continue to grow and divide during cold shock, revealing a coordinated, multi-phase adaptation process that could not be deduced from traditional population measurements. They propose a scattering-theory-based model that reconciles the paradox between growth differences of the bacteria at the single-cell level vs population level.

Strengths:

(1) The LUNA approach is pretty creative, turning coma aberration from what is normally a nuisance into an exploit. LUNA enabled long-term single-cell imaging during rapid temperature downshifts.

(2) The authors show that the long-assumed growth arrest during cold shock from population-level measurements is misleading. At the single-cell level, bacteria do not stop growing or dividing but undergo a continuous, three-phase adaptation process. Importantly, this behavior is highly synchronized across the population and not based on bet-hedging.

(3) Finally, the authors propose a model to resolve a long-standing paradox between single-cell vs population behavior: if cells keep growing, why does optical density (OD) of the culture stop increasing? Using light-scattering theory, they show that OD depends not only on cell number but also on cell volume, which decreases after cold shock. As a result, OD can remain flat, or even decrease, despite continued biomass accumulation. This demonstrates that OD is not a reliable proxy for growth under non-steady conditions.

Weaknesses:

(1) While the authors theoretically explain the advantages of LUNA over existing autofocus methods, it is unclear whether practical head-to-head comparisons have been performed, apart from the comparison to Nikon PFS shown in Video S1. As written, the manuscript gives the impression that only LUNA can solve this problem, but such a claim would require more systematic and rigorous benchmarking against alternative approaches.

(2) No mutants/inhibitors used to test and challenge the proposed model.

(3) Cells display a high degree of synchronization, but they are grown in confined microfluidic channels under highly uniform conditions. It is unclear to what extent this synchrony reflects intrinsic biology versus effects imposed by the microfluidic environment.

(4) To further test and generalize the model, it would be informative to also examine bacterial responses at intermediate temperatures rather than focusing primarily on a single cold-shock condition.

Comments on revisions:

The authors have addressed my comments in their response, but have chosen not to incorporate most of them into the manuscript. Readers may refer to the peer review section for further details.

We thank the reviewer for this additional comments and for the careful suggestions, and we appreciate that the raised points are valuable for a broader discussion of the topic. In the revised manuscript, we have incorporated the comments most directly relevant to the scope and central conclusions of the study, and have clarified these points in the text where appropriate. Specifically, we have clarified several key issues, including the interpretation of the OD lag as a “combined effect,” the performance and application scope of LUNA, the alignment of cell-cycle progression after cold shock, and relevant methodological details.

For the remaining contextual issues, we have kept the detailed discussion in the response to reviewers rather than expanding the manuscript extensively, so as to preserve the focus and readability of the main text. We hope that the revisions now better acknowledge the reviewer’s concerns while maintaining a concise presentation of the central findings.

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

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors developed a new autofocusing method, LUNA (Locking Under Nanoscale Accuracy), to address severe focus drift-a major challenge in time-lapse microscopy. Using this method, they tackle a fundamental question in bacterial cold shock whether cells halt growth and division following an abrupt temperature downshift. Overall, the experimental design, modeling, and data analysis are solid and well executed. However, several points require clarification or further support to fully substantiate the authors' conclusions.

Strengths:

(1) The LUNA method outperforms existing autofocusing systems with nanoscale precision over a large focusing range. The focusing time is reasonable for the presented experiments, and the authors note potential improvements by using faster motors and optimized control algorithms, suggesting broad applicability. The theoretical simulations and experimental validation provide solid support for the robustness of the method.

(2) Using LUNA, the authors address a long-standing question in bacterial physiology: whether cells arrest growth and division after an abrupt cold shock. Single-cell analyses monitoring the entire course of cold adaptation and steady-state growth reveal features that are obscured in bulk-culture studies: cells continue to grow at reduced rates with smaller cell sizes, resulting in an apparently unchanged population-level OD. The experiments are well designed and analyses are generally solid and largely support the authors' conclusions.

(3) The authors also propose a model describing how population-level OD measurements depend on cell dry mass density, volume, and concentration. This provides a valuable conceptual contribution to the interpretation of OD-based growth measurements, which remain a gold-standard method in microbiology.

We thank the reviewer for acknowledging the strengths of our study.

Weaknesses:

(1) It is unclear whether the author's model explaining the population-level OD during acclimation is broadly applicable. Most analyses focus on a shift from 37˚C to 14˚C, where the model agrees well with experimental data. However, in the 37˚C to 12˚C experiment, OD600 decreases after cold shock (Fig. 5e), and the computed OD does not match the experimental measurements (Fig. S16a). Although the authors attribute this discrepancy to a "complicated interplay," no further explanation is provided, which limits confidence in the model's general applicability.

Thank you for this careful evaluation regarding the model generality. In the experiment with a temperature shift from 37°C to 12°C, the measured OD600 values were 0.243 at 0 hours and 0.242 at 5 hours. In comparison, our model-computed OD600 values were 0.243 at 0 hours and 0.271 at 5 hours. The absolute difference between the measured and computed values at 5 hours is therefore 0.028.

Given the typical experimental variability in OD600 measurements and the limited linear range of the OD-to-biomass approximation (generally considered reliable below ~0.5), this deviation is quantitatively modest. We appreciate your valuable feedback and are happy to provide further clarification if needed.

(2) The manuscript proposes that cell-cycle progression becomes synchronized across the population after cold shock, but the supporting evidence is not fully convincing. If synchronization refers primarily to the uniform reduction in growth rate following cold shock, this could plausibly arise from global translation inhibition affecting all cells. However, the additional claim that "cells encountering a relatively late CSR will accelerate division to maintain synchronization" is not strongly supported by the presented data.

We appreciate your critical reading, which has helped us identify ambiguities in our terminology and strengthen the clarity of our work. Regarding the term “synchronization”, we would like to clarify that it refers to two different scenarios: (i) the synchrony in the timing of growth rate changes after cold shock. The cells initiate the slowdown in growth almost simultaneously, suggesting a highly coordinated, non-stochastic population-level response to cold shock; (ii) the synchrony in division cycle progression.

In the sentence you referenced “cells encountering a relatively late CSR will accelerate divisions to maintain synchronization”, we intended to describe that cells maintain consistent progression of the division cycle after cold shock, meaning that after the same number of elapsed cycles, different cells are at a similar stage in their division timing (Figure 4f, 4g, Figure S14). The term “accelerate” refers to our observation that cells which complete a given cycle later than others tend to have shorter subsequent inter-division intervals, thereby “catching up” to maintain alignment in cycle number across the population. We acknowledge that using “synchronization” in this scenario may be ambiguous, and we will replace it with more precise phrasing “progression of division cycle” to accurately convey this finding.

(3) Several technical terms used in the method development section are not clearly defined and may be unfamiliar to a broad readership, which makes it difficult to fully understand the methodology and evaluate its performance. Examples include depth of focus, focusing precision, focusing time, focusing frequency, and drift threshold value. In addition, the reported average focusing time per location (~0.6 s) lacks sufficient context, limiting the reader's ability to assess its significance relative to existing autofocusing methods.

Thank you for your valuable comments and suggestions. In response, we have added more detailed descriptions in the Methods section of the revised version.

The reviewer noted that the reported average focusing time (~0.6 s) lacks sufficient context, which may limit readers’ ability to assess its significance relative to existing autofocusing methods. We would like to clarify that the core innovation of this work lies in the proposed theoretical framework for autofocusing, which offers advantages over existing methods in terms of focusing precision and range. While focusing time is a practically relevant performance metric, it is primarily presented here as an implementation-dependent parameter rather than a central theoretical contribution of this study. In our experimental setup, an average focusing time of 0.6 s proved sufficient for routine timelapse imaging in microscopy, thereby demonstrating the practical usability of LUNA.

Reviewer #2 (Public review):

Summary:

This study presents LUNA, an autofocus method that compensates for focus drift during rapid temperature changes. Using this approach, the authors show that E. coli cells continue to grow and divide during cold shock, revealing a coordinated, multi-phase adaptation process that could not be deduced from traditional population measurements. They propose a scattering-theory-based model that reconciles the paradox between growth differences of the bacteria at the single-cell level vs population level.

Strengths:

(1) The LUNA approach is pretty creative, turning coma aberration from what is normally a nuisance into an exploit. LUNA enabled long-term single-cell imaging during rapid temperature downshifts.

(2) The authors show that the long-assumed growth arrest during cold shock from population-level measurements is misleading. At the single-cell level, bacteria do not stop growing or dividing but undergo a continuous, three-phase adaptation process. Importantly, this behavior is highly synchronized across the population and not based on bet-hedging.

(3) Finally, the authors propose a model to resolve a long-standing paradox between single-cell vs population behavior: if cells keep growing, why does optical density (OD) of the culture stop increasing? Using light-scattering theory, they show that OD depends not only on cell number but also on cell volume, which decreases after cold shock. As a result, OD can remain flat, or even decrease, despite continued biomass accumulation. This demonstrates that OD is not a reliable proxy for growth under non-steady conditions.

We thank the reviewer for acknowledging the strengths of our study.

Weaknesses:

(1) While the authors theoretically explain the advantages of LUNA over existing autofocus methods, it is unclear whether practical head-to-head comparisons have been performed, apart from the comparison to Nikon PFS shown in Video S1. As written, the manuscript gives the impression that only LUNA can solve this problem, but such a claim would require more systematic and rigorous benchmarking against alternative approaches.

Thank you for your insightful comment regarding the comparison of LUNA with other autofocus methods.

In our study, we primarily compared LUNA with the Nikon PFS system (as shown in Video S1) because Nikon PFS is one of the most widely used commercial autofocus systems in single-cell time-lapse imaging, and its manufacturer provides well-defined performance parameters (e.g., focusing precision within 1/3 depth-of-focus, response time <0.7 s), which facilitates a quantitative comparison. For other commercial systems, such as Olympus ZDC, Zeiss Definite Focus, Leica AFC, and ASI CRISP, the publicly available specifications are often less clearly defined, or are measured under inconsistent conditions, making a direct head-to-head comparison challenging and potentially misleading. Additionally, in our preliminary experiments, we also tested an Olympus microscope and observed severe focus drift during slow cooling processes. From a physical perspective, LUNA is specifically designed to meet the demanding requirements of single-cell experiments, including a wide focusing range and high precision, while existing commercial systems may not physically achieve the combination of range and accuracy needed for such extreme conditions.

(2) No mutants/inhibitors used to test and challenge the proposed model.

We agree that such approaches would provide valuable mechanistic insights and further strengthen the validation of the model presented in this study. In the current work, our primary goal was to introduce LUNA autofocusing method and demonstrate its capability to resolve bacterial cold shock response at the single-cell level with unprecedented precision. As such, we focused on characterizing the wild-type physiological dynamics under cold shock, which already revealed several previously unreported phenomena. We acknowledge that the use of genetic mutants or chemical inhibitors targeting specific cold shock proteins or regulatory pathways would be a logical and powerful next step to dissect the underlying molecular mechanisms and test the causality of the observed growth dynamics. We plan to address this in future work by incorporating such perturbations to further test and refine the model.

(3) Cells display a high degree of synchronization, but they are grown in confined microfluidic channels under highly uniform conditions. It is unclear to what extent this synchrony reflects intrinsic biology versus effects imposed by the microfluidic environment.

The reviewer raises a pertinent question regarding whether the observed high degree of cell synchronization represents an intrinsic biological phenomenon or an artifact induced by the microfluidic environment.

Over the past decade, microfluidic chips, including the specific design used in our work, have become a widely accepted and powerful tool in microbial physiology research. A broad consensus has emerged within the community that the microenvironment within these microchannels does not significantly interfere with or perturb the natural physiological behavior of microorganisms (Dusny, C. & Grünberger, Curr Opin Biotechnol. 63, 26-33 (2020)). This understanding is also supported by the fact that key findings obtained with microfluidic single-cell technologies are reproducible by other methods. For example, the adder model of cell-size homeostasis in E. coli firstly observed in microfluidic chips has been repeatedly validated by different methods (Taheri-Araghi, S. et al. Curr. Biol. 25, 385-391 (2015)). Therefore, while we acknowledge the importance of considering environmental effects, we are confident that the synchronization we report reflects the genuine biological dynamics of E. coli cells.

(4) To further test and generalize the model, it would be informative to also examine bacterial responses at intermediate temperatures rather than focusing primarily on a single cold-shock condition.

We thank the reviewer for this thoughtful suggestion. In designing our experiments, we aimed to study the bacterial cold shock response at the single-cell level. A key feature of this response is that it is typically triggered only when the temperature drops below a certain threshold within a short time duration. We therefore chose to lower the temperature from 37 °C to 14 °C as rapidly as possible. This approach allowed us to leverage the unique capabilities of LUNA while also providing an opportunity to explore this biological process in greater detail.

We agree that investigating bacterial responses across intermediate temperatures would be highly informative for understanding how temperature changes affect cellular physiology. However, this direction addresses a distinct scientific question that lies beyond the scope of the current work. We fully acknowledge its value and do have the intention to explore it in future studies.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Major points:

(1) To strengthen the generality of the conclusions regarding cold shock response, it would be helpful to include a similar single-cell analysis of growth and division (cell size and concentration) for the 37˚C to 12˚C temperature shift. In this case, the experimental acclimation lasts ~5 hours, whereas the model predicts ~2 hours (Fig. S16a). Examining whether the model still holds or whether additional factors (e.g., further reductions in cell size) contribute to the observed OD decrease would clarify this discrepancy.

We thank the reviewer for this valuable suggestion. Our model for explaining the population-level OD dynamics during acclimation does not depend on single-cell time-lapse microscopy data. Instead, the single-cell inputs used for parameterization were obtained from flow cytometry measurements, which quantify population-wide single-cell distributions. Therefore, the model is not intrinsically restricted to a specific imaging-based experimental setup or to a particular temperature shift.

Most of the quantitative analysis presented in the manuscript focuses on the 37°C to 14°C transition, where the model shows strong agreement with experimental OD measurements. We selected this condition because it provides high-quality, internally consistent datasets at both the single-cell and population levels. However, the modeling framework itself is mechanistic and parameter-based, rather than temperature-specific. In principle, it can be applied to other temperature shifts, provided that the corresponding single-cell growth and state-transition parameters are experimentally determined.

Regarding the temperature shift from 37°C to 12°C, the model demonstrates good agreement with the experimental observation that acclimation lasts approximately 5 hours. The minor deviations in several data points during the acclimation period can be attributed to systematic errors in the measurement of cell concentration and volume, as illustrated in the lower panel of Figure S16a. We are open to extend our analysis to additional temperature shifts in future work to further validate the model’s generality.

(2) Related to weakness #2, it would be helpful for the authors to clarify their definition of "synchronization" and to provide additional explanation or evidence supporting this claim. In particular, further discussion of the data in Fig. 4f, 4g, and S14 could help strengthen the proposed hypothesis.

We thank the reviewer for this constructive suggestion. In previous response (public review weakness #2), we clarified the definition of “synchronization” in the revised manuscript by explicitly distinguishing between two types of synchrony: (i) the synchrony in the timing of growth rate changes after cold shock, and (ii) the synchrony in division cycle progression. For the latter, we now use the more precise term “progression of division cycle” to avoid ambiguity. Furthermore, we have expanded the discussion of the data in Figures 4f, 4g, and S14 to better support the claim that cells actively maintain alignment in cycle progression. We hope these revisions address the reviewer’s concern and strengthen the evidence for our hypothesis.

Minor points:

(1) Line 78: "... and concluded that the OD lag is actually the outcome of the synergy of changes in bacterial concentration and volume, ..." The term synergy usually implies a combined effect greater than the sum of individual effects. Are the changes in bacterial concentration and volume synergistic here?

We agree with your observation that the term "synergy" in scientific contexts typically implies an interaction effect that is greater than the sum of individual effects. In our original phrasing, we intended to convey that the observed OD lag is a result of the combined contributions from both changes in bacterial concentration and changes in cell volume, rather than being dominated by a single factor. We did not mean to imply a super-additive interaction between these two variables.

We acknowledge that the relationship between bacterial concentration and cell volume can be complex and may even exhibit interdependence under certain conditions (e.g., under nutrient limitation at high OD). However, using "synergy" could indeed be misleading. To ensure terminological precision and avoid any potential misinterpretation, we will revise the text in the revised manuscript. We will replace "synergy" with a more neutral and accurate phrase "combined effect".

(2) Figure 2d: Why does the focusing time increase even after temperature stabilizes following the downshift? Does focus drift depend not only on rapid cooling but also on the lower steady-state temperature? Additional explanation would be helpful.

As noted in the Methods section ("Time-lapse imaging of bacteria under CS"), when the temperature was lowered, the objective lens heater was stopped, which caused a slightly longer focusing time. This is because prior to the temperature downshift, the objective heater maintained the objective at a temperature close to that of the sample (37°C), minimizing any thermal gradient between them. After the temperature decrease to 14°C, while the sample chamber was precisely controlled at the target low temperature, the objective lens now without active heating gradually equilibrated to ambient room temperature (approximately 22–25°C). This created a stable temperature mismatch between the relatively warmer objective and the colder sample. Such a temperature gradient can cause minor thermal expansion or contraction of the objective lens barrel, leading to a small but persistent shift in the focal plane. Consequently, the focusing time remained slightly elevated (∼0.6 s) compared to the 37°C condition (∼0.3 s), even after the sample temperature had stabilized. This offset reflects the steady-state thermal disequilibrium between the objective and the sample, rather than a transient cooling effect. We hope this explanation clarifies the reviewer’s concern.

(3) Line 234: "Reanalysis of the protein synthesis dynamics after CS revealed increase in CSPs synthesis (Figure 3e)." A citation is needed here. Additionally, the dataset referenced here was generated using a 37˚C to 10˚C cold shock.

We thank the reviewer for the insightful comments and the careful reading of our manuscript. We have now added the appropriate citation in the main text (Zhang, Y. et al. Molecular Cell 70, 274–286 (2018)). The dataset used in this reanalysis was generated under a 37°C to 10°C cold shock, rather than 12°C, and we have clarified this in the Methods section to avoid any ambiguity.

We would also like to clarify our rationale for using this published dataset in the present context. To our knowledge, no published dataset exists with comparable protein synthesis dynamics specifically at 12°C. Our intention here was to reference a well-characterized cold-shock dataset to support the qualitative point that CSP synthesis increases and ribosome synthesis decreases after cold shock. In cold shock studies, many qualitative conclusions are broadly consistent across low-temperature conditions (e.g., below ~15°C, and in some cases more broadly below ~20°C), including the observation that the ribosomal protein fraction is relatively insensitive to temperature change (Herendeen, S. L. et al. Journal of Bacteriology. 139, 185–194 (1979), Knapp, B. D. & Huang, K. C. Annual Review of Biophysics. 51, 499–526 (2022)). We appreciate the reviewer’s valuable feedback, which has helped us improve the clarity and accuracy of our work.

(4) Figure 3f and 3g: How is growth rate defined here, and why do the elongation rate and growth rate yield different results? My understanding is that, during steady-state growth, cell elongation rate increases as cells progress through a single cell cycle prior to division, whereas G0 cells exhibit reduced elongation rate following cold shock. Is this correct? More explanation is also needed for "linear growth in growth mode" (Line 267).

Thank you for this important comment. In our manuscript, we use:

Elongation rate = dL/dt (the absolute rate of increase in cell length; y-axis in Figure 3f)

Growth rate = (dL/dt)/L (i.e., λ, y-axis in Figure 3g; also referred to in some studies as the instantaneous growth rate)

Because these are different quantities, they do not necessarily follow the same trend across the cell cycle. To clarify the logic behind our “growth mode” classification (also see Willis & Huang, Nat Rev Microbiol 2017):

For a rod-shaped cell growing in length L,

(1) Exponential growth means the elongation rate is proportional to cell size, i.e.,

𝑑𝐿/𝑑𝑡 ∝ 𝐿

or equivalently,

(𝑑𝐿/𝑑𝑡)/𝐿) = constant

(2) Linear growth means the elongation rate is constant throughout the cell cycle, i.e.,

𝑑𝐿/𝑑𝑡 = constant

which implies that

(𝑑𝐿/𝑑𝑡)/𝐿)

decreases as the cell elongates.

Based on these two basic cases, additional growth modes (e.g., super-exponential, sub-exponential, sub-linear) can also be defined, as illustrated in the Author response image 1.

Author response image 1.

With this definition, our interpretation of Figure 3f and 3g is as follows: before cold shock, cells are consistent with approximately exponential growth (red line in Figure 3g), whereas after cold shock, the G0 cells are better described as undergoing approximately linear growth (yellow line in Figure 3f).

(5) Figure S12: Why are the curves not continuous across GN, G0, G1, and G2?

In this figure, we present two different metrics: elongation rate (𝑑𝐿/𝑑𝑡) in panel (a) and growth rate (𝜆 = (𝑑𝐿/𝑑𝑡)/𝐿) in panel (b). During bacterial division, the cell length approximately halves while the growth rate remains constant under steady-state conditions. As a result, elongation rate, which is proportional to the instantaneous length, also halves at each division event, leading to the observed discontinuities at the time points corresponding to divisions (GN, G0, G1, and G2). In contrast, growth rate is inherently continuous across divisions, as shown in panel (b), although minor apparent discontinuities may appear due to the finite temporal resolution of our measurements. We hope this explanation clarifies the figure.

(6) Figure 4d: X-axis labels are missing.

Thank you for your insightful comment. The six panels share identical axes in Figure 4d. To enhance the visual focus on the data trends across different generations, we intentionally displayed the X-axis label and numerical tick labels only on the first panel. The subsequent panels show only the tick marks without the numerical labels, as their scale is identical to that of the first panel.

(7) Line 285 and Figure 4e: "The changes in λ are highly synchronized in time, with the exact time lag between any pair of ξ not exceeding 2 min ..." What is the definition of time lag?

In our study, the term "time lag" refers to the absolute difference in time at which a large sudden drop of the λ curve occurs between any two pairs of ξ. Essentially, it quantifies how closely the dynamic changes in λ are aligned across different groups. A time lag of zero would indicate perfect synchrony, while a value within 2 minutes implies that the variations in λ for any pair of ξ occur nearly simultaneously.

(8) Figure S14: Why can the elapsed cycles take negative values?

In Figure S14, we plotted the centered values. Specifically, at each time point, we calculated the mean elapsed cycle number across all lineages, and then subtracted this mean from each group’s value. The resulting values are presented in the figure as “Elapsed cycles (zero-centered)”. Thus, negative values are expected and meaningful they represent lineages that are progressing more slowly than the average at that time point. This transformation helps to highlight the relative differences among groups over time, while removing the overall temporal trend (which is already shown in Figure 4g).

(9) Figure 5 legend: Fitting for the acclimation has a R2 of -0.263 (Pearson correlation coefficient -0.00). R^2 should not be negative, and it doesn't agree with the calculated Pearson correlation coefficient.

Thank you for this important observation. Indeed, R² should normally fall within the range [0, 1]. This discrepancy arises because the fitting model used differs from the default linear regression, and we did not specify this in the original figure legend. In the revised manuscript, this has been corrected. The explanation why R² is negative here is as follows:

The linear fit used is y = a·x (i.e., no-intercept, forced through the origin). This is based on the physical principle that when OD is zero (no bacteria), the total bacterial mass must also be zero. For ordinary linear models with an intercept, R² ranges from 0 to 1. However, for no-intercept models, the calculation of total sum of squares (SS_tot) differs (typically relative to zero rather than the mean of y), and R² can become negative if the fit performs worse than the baseline y = 0. Here, R² = -0.263 simply indicates that for these specific data points, the origin-constrained linear fit does not outperform the trivial y=0 model. Regarding the Pearson correlation: The near-zero coefficient (-0.00) suggests no significant linear trend between X and Y, which is consistent with the poor fit performance.

(10) Language and typos: The manuscript contains grammatical errors and typos that require careful proofreading (one example: Line 56 "..., and reflection-based approaches ...").

We thank the reviewer for the careful reading and for drawing our attention to the language and typographical issues in the manuscript. In the revised version, we will carefully proofread the entire text and correct any errors and inconsistencies, including the example pointed out in line 56.

Reviewer #2 (Recommendations for the authors):

(1) The LUNA section is extremely technical and advanced for most biologists - it might be useful to include a few sentences in simple language why LUNA helps solve the biology question.

We thank the reviewer for the valuable suggestion. We have now added a concise, plain-language overview at the end of the LUNA section (Performance Analysis of LUNA):

“In brief, LUNA locks the focal plane with nanometer-scale precision over an ultra-large range rapidly, ensuring stable focus during long-term imaging for reliable observation of fine subcellular structures and dynamics.”

(2) The suggestions I included in the weakness section are not mandatory to perform, but will be helpful to at least discuss in the paper.

We thank the reviewer for the thoughtful comment and for acknowledging that the suggestions in the weakness section are not mandatory. We have carefully considered each point raised and have provided detailed responses in the point-by-point reply. While we recognize the potential value of these suggestions for further expanding the study, we respectfully believe that incorporating them into the current manuscript would go beyond the intended scope of this work.

Thanks

Otherwise, great job with the paper!

We are truly grateful to the reviewer for the encouraging feedback and appreciate the time and effort invested in improving our manuscript.

eLife Assessment

The authors proposed two hypotheses: first, that methamphetamine induces neuroinflammation, and second, that it alters neuronal stem cell differentiation. These are valuable hypotheses, and the authors provided in vivo observations of the methamphetamine response in mice. However, concerns about data interpretation, and the current evidence is incomplete, requiring further experimental validation.

Reviewer #1 (Public review):

Summary:

The manuscript, titled Hippocampal Single-Cell RNA Atlas of Chronic Methamphetamine Abuse-Induced Cognitive Decline in Mice, focuses on single-cell RNA sequencing (scRNA-seq) analysis following chronic methamphetamine (METH) treatment in mice. The authors propose two hypotheses: (1) METH induces neuroinflammation involving T and NKT cells, and (2) METH alters neuronal stem cell differentiation.

Strengths:

The authors provide a substantial dataset with numerous replicates, offering valuable resources to the research community.

Weaknesses:

Concerns remain regarding the interpretation of the data and the appropriateness of the statistical analyses.

Although the authors provided detailed responses to the reviewer's concerns, I am still concerned that several key issues have not yet been fully addressed in the revised manuscript.

First, in Figure 5, the authors state that neural stem cells (NSCs) preferentially differentiate into astrocytes rather than neuroblasts following METH treatment. However, based on the presented trajectories, it is difficult to visually confirm differences in the relative proportions of astrocyte versus neuroblast differentiation between the control and METH-treated conditions. The current figures do not provide a quantitative or clearly interpretable comparison of lineage allocation that would support this conclusion.

Moreover, in Figures 5C and 5F, the inferred pseudotime trajectories differ both the starting cell populations and the intermediate and terminal cell identities. As a result, the trajectories are not directly comparable between the control and METH conditions. Under these circumstances, it is inappropriate to interpret gene expression changes as occurring along equivalent differentiation paths, and the current analysis does not convincingly support the stated conclusions regarding altered NSC differentiation.

If the authors intend to claim differential gene expression associated with altered differentiation trajectories, the analysis should at minimum present the expression of the same set of genes (e.g., Bsg, Ccl4, Fos, Sox11, Flt1, Hspb1, Igfbp7, and Tmsb10) plotted along a matched trajectory (for example, NSC-to-astrocyte or NSC-to-neuroblast lineages) in both control and METH-treated samples, so that readers can directly compare expression dynamics across conditions.

In addition, several statements throughout the manuscript describing changes in cell-type proportions are not supported by corresponding statistical analyses. For example, in Figure 2C (around line 430), the authors report changes in cell proportions of ~0.1% or 2-3%. Without appropriate statistical testing, it is unclear whether such marginal differences are biologically meaningful or reproducible. The authors should either provide statistical testing (e.g., sample-level proportion analysis with p-values or confidence intervals) or revise the text to describe these findings as descriptive rather than significant changes.

Finally, the reported decrease in astrocyte proportion following METH exposure (from 6.6% to 5.5%), together with the lack of reported changes in neuroblast proportions, appears inconsistent with the trajectory-based conclusion that NSCs preferentially differentiate into astrocytes in METH-treated mice. This apparent discrepancy should be clarified or the conclusions appropriately tempered.

Author response:

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

Public Reviews:

Reviewer #1 (Public review):

(1) Concerns persist regarding the interpretation of data and the validation of experiments. First, the presence of T cells, NKT cells, and neutrophils in both the control and METH-treated hippocampi suggests that blood contamination rather than immune cell infiltration is the cause. Since the authors claim that METH disrupts the blood-brain barrier, increasing the infiltration of these immune cells, identifying the source of these immune cells is critical.

We sincerely appreciate the valuable suggestions you have provided. Your professional perspective impresses us. Based on your suggestion, we conducted a systematic review and in-depth analysis of the experimental process.

As you have pointed out, we believe that the T cells, NK cells and neutrophils detected in the single-cell sequencing of the mouse hippocampus may have a blood-derived origin. However, this does not mean that the presence of these cell types in the control group is abnormal, because in many literature, these cells can also be found in the hippocampus of control mice. Nevertheless, clarifying the origin and location of these cells will help to further strengthen the persuasiveness of the research hypothesis. Although there is currently no systematic discussion on the role of such cells in the field of methamphetamine neurotoxicity research, we believe that the relevant findings still have certain reference value for subsequent research in this field.

Our response is based on the following description:

(1) Insufficient perfusion during the extraction of the hippocampus may lead to a certain degree of blood contamination.

Given that the single-cell sequencing technique employed in this study can detect all the mRNA of the entire cell, in order to ensure that the cells are in the optimal physiological state and to minimize the stress response caused by the experimental operation on the cells, we perfused the anesthetized mice with cold PBS for approximately 3 min (this has been supplemented in the Materials and methods Line165-166), and completed the rapid dissection and collection of the mouse hippocampus on the ice surface within 2 min, and immediately placed it in an appropriate amount of tissue preservation solution for storage. The time of tissue perfusion might be insufficient or the perfusion volume might not be adequate, resulting in the incomplete expulsion of all the blood. Subsequently, the decomposition operations of the tissue samples were all carried out in the preservation solution or PBS buffer, which to some extent reduced the potential interference of blood components on the experimental results. Additionally, T cells, NKT cells and neutrophils in the capillary perivascular spaces of the hippocampal tissue might still remain and be successfully captured, and were reflected in the final sequencing data.

(2) The presence of T cells, NKT cells, and neutrophils in the brain tissue of normal mice has been supported by existing literature. Moreover, several studies have specifically described the localization of these immune cell types within the brain parenchyma.

Contemporary studies have completely changed the view of brain immunity from envisioning the brain as isolated and inaccessible to peripheral immune cells to an organ in close physical and functional communication with the immune system for its maintenance, function, and repair. Circulating immune cells reside in special niches in the brain’s borders, the choroid plexus, meninges, and perivascular spaces, from which they patrol and sense the brain in a remote manner [1].

A large-scale mouse brain cell atlas study also reported that approximately 8% of non-neuronal cells are immune cells, including microglia, boundary-associated macrophages, lymphocytes, dendritic cells, and monocytes [2].

Hang Yao et al. demonstrated through flow cytometry that neutrophils were present in the hippocampal tissues of both healthy control mice and depressed mice (Fig.2 H) [3]. Wei Su et al. identified through single-cell sequencing that dendritic cells, neutrophils, macrophages, T cells, and NKT cells were present in the brain tissues of non-transgenic (Non-Tg) control mice (Fig.1a-b), and the localization of these cells was explicitly characterized as brain parenchyma in the study [4]. Tomomi M Yoshida et al. discovered through immunohistochemistry (IHC) and single-cell sequencing techniques that there were a certain number of CD3+ and CD4+ T cells in the hippocampus and other regions of the brain, and they observed that these cells were located outside the blood vessels. (Fig.1a-c, g) [5].

(3) Both the analysis of immune cells within blood vessels and those in the brain parenchyma contribute to elucidating the immune effects in the hippocampal microenvironment under chronic METH exposure, as well as their interactions with other cell types. At present, the understanding of the neurotoxicity of methylphenidate and the immune system is still limited to the central resident immune cells, such as microglia, astrocytes and oligodendrocytes [6]. Adaptive immune cells and myeloid cells recruited from the circulation have also been implicated in brain development, function, and aging. Their depletion during developmental stages can disrupt critical neural processes, including glial cell maturation, neuronal activity, and myelinogenesis. However, the precise developmental stage at which lymphocyte infiltration into the central nervous system occurs remains to be elucidated [7].

Our data results indicate that during chronic METH abuse, T cells are more active and participate in the regulation of cytokines through complement signaling. At the same time, the frequency of cell communication between endothelial cells and epithelial cells is increased. Moreover, microglia upregulated the processes of cell chemotaxis and migration, as well as the communication with immune cells such as T cells, and to some extent, this also suggests an enhanced infiltration of T cells. However, we also recognize that the current conclusions regarding immune cell infiltration based on sequencing data and literature reports lack the support of experimental data. Currently, we are conducting morphological analysis using the same batch of brain tissue samples to further validate the relevant findings.

Immune fluorescence staining and flow cytometry can be utilized to further determine the locations of these immune cells in the hippocampus. The classical pathways through which peripheral immune cells enter the brain mainly include the BBB and the choroid plexus. In June 2025, Kim N. Green et al. published a study in Neuron, further revealing that during the developmental stage and in cases of inflammatory diseases, immune cells can also infiltrate the brain parenchyma through a newly identified channel - the medial ventricle, thereby further confirming that these cells have the ability to migrate to the central nervous system under specific physiological or pathological conditions [8].

(1) Castellani G, Croese T, Peralta Ramos JM, Schwartz M. Transforming the understanding of brain immunity. Science. 2023 Apr 7;380(6640):eabo7649. doi: 10.1126/science.abo7649.

(2) Zhang M, Pan X, Jung W, Halpern AR, Eichhorn SW, Lei Z, Cohen L, Smith KA, Tasic B, Yao Z, Zeng H, Zhuang X. Molecularly defined and spatially resolved cell atlas of the whole mouse brain. Nature. 2023 Dec;624(7991):343-354. doi: 10.1038/s41586-023-06808-9.

(3) Yao H, Jiang SY, Jiao YY, Zhou ZY, Zhu Z, Wang C, Zhang KZ, Ma TF, Hu G, Du RH, Lu M. Astrocyte-derived CCL5-mediated CCR5+ neutrophil infiltration drives depression pathogenesis. Sci Adv. 2025 May 23;11(21):eadt6632. doi: 10.1126/sciadv.adt6632.

(4) Su W, Saravia J, Risch I, Rankin S, Guy C, Chapman NM, Shi H, Sun Y, Kc A, Li W, Huang H, Lim SA, Hu H, Wang Y, Liu D, Jiao Y, Chen PC, Soliman H, Yan KK, Zhang J, Vogel P, Liu X, Serrano GE, Beach TG, Yu J, Peng J, Chi H. CXCR6 orchestrates brain CD8+ T cell residency and limits mouse Alzheimer's disease pathology. Nat Immunol. 2023 Oct;24(10):1735-1747. doi: 10.1038/s41590-023-01604-z.

(5) Yoshida TM, Nguyen M, Zhang L, Lu BY, Zhu B, Murray KN, Mineur YS, Zhang C, Xu D, Lin E, Luchsinger J, Bhatta S, Waizman DA, Coden ME, Ma Y, Israni-Winger K, Russo A, Wang H, Song W, Al Souz J, Zhao H, Craft JE, Picciotto MR, Grutzendler J, Distasio M, Palm NW, Hafler DA, Wang A. The subfornical organ is a nucleus for gut-derived T cells that regulate behaviour. Nature. 2025 Jul;643(8071):499-508. doi: 10.1038/s41586-025-09050-7.

(6) Shi S, Sun Y, Zan G, Zhao M. The interaction between central and peripheral immune systems in methamphetamine use disorder: current status and future directions. J Neuroinflammation. 2025 Feb 15;22(1):40. doi: 10.1186/s12974-025-03372-z.

(7) Castellani G, Croese T, Peralta Ramos JM, Schwartz M. Transforming the understanding of brain immunity. Science. 2023 Apr 7;380(6640):eabo7649. doi: 10.1126/science.abo7649.

(8) Hohsfield LA, Kim SJ, Barahona RA, Henningfield CM, Mansour K, Vallejo KD, Tsourmas KI, Kwang NE, Ghorbanian Y, Angulo JAA, Gao P, Pachow C, Inlay MA, Walsh CM, Xu X, Lane TE, Green KN. Identification of the velum interpositum as a meningeal-CNS route for myeloid cell trafficking into the brain. Neuron. 2025 May 28:S0896-6273(25)00351-4. doi: 10.1016/j.neuron.2025.05.004.

(2) Secondly, the pseudotime analysis, which suggests altered neural stem cell (NSC) differentiation, is not conclusively supported by the current data and requires further validation.

We sincerely appreciate your valuable feedback, which we find highly relevant and constructive. It is important to acknowledge that the sequencing data presented in our study currently lacks experimental validation. Nevertheless, considering that existing research on the effects of METH on neural stem cell differentiation predominantly emphasizes observational phenomena and remains limited in terms of in vivo experimental evidence and mechanistic investigations, we aim to contribute our analytical findings as a reference for further scholarly exploration in this field.

Our study utilized pseudotime analysis (powered by Monocle2) to reconstruct an "imaginary timeline" (pseudo-time) based on intercellular gene expression similarities, thereby modeling the dynamic state transitions of cells during continuous biological processes. Drawing upon single-cell RNA sequencing data captured as "snapshots" from hippocampal astrocytes, neural stem cells, and neuroblasts in mice four weeks after METH exposure, we applied computational algorithms to integrate the originally discrete cellular states into a continuous pseudo-time trajectory. This approach was employed to elucidate the differentiation stages of these cell populations, identify potential branching points in their developmental pathways, and uncover the key regulatory genes driving the differentiation process. Pseudotime analysis, as a computational approach grounded in mathematical modeling, yields inferences that are contingent upon the underlying assumptions of the algorithms employed. Consequently, experimental validation through methodologies such as time-series sampling and lineage tracing is essential to substantiate the derived biological interpretations. In light of the insufficiency of such empirical verification to date, our conclusions concerning alterations in the dynamic behavior of neural stem cell differentiation remain preliminary and require further experimental support.

In Figures 5C and 5F, we present the expression profiles of the four genes exhibiting the most statistically significant differences across the differentiation trajectory. In Figures 5B and 5E, we conducted GO and KEGG functional enrichment analyses on the genes that showed significant differential expression at different differentiation stages. While no studies within the current METH research domain have reported on the potential effects of these genes on neural stem cell differentiation, emerging evidence from related fields provides preliminary insights into their functional roles. For instance, the Flt1 gene (also known as VEGFR1), referred to as the vascular endothelial growth factor receptor, has been demonstrated to play a critical role in the conversion of Müller glial cells into neurons within the zebrafish retina [1], serves as a critical regulator in promoting definitive neural stem cell survival [2]. Furthermore, it substantiates the intricate interconnection between neurons, neural stem cells, and vascular cells, as identified in our cell communication analysis. Hsp1b gene plays a significant role in ferroptosis and autophagy processes of nerve cells[3, 4], and may be closely related to the self-renewal ability of neural stem cell, while METH may impair neural stem cell function by disrupting autophagy, leading to reduced self-renewal capacity and altered differentiation potential [5]. In METH group, Sox11 has been shown to play a critical role in early differentiation and neuronal growth, both during perinatal development and in adult neurogenesis [6] Fos gene plays a critical regulatory role in the differentiation of neural stem cells into neurons and in modulating neuronal functional activities [7]; Alterations in Ccl5 expression levels may indicate astrocyte-mediated inflammatory responses, which could represent one of the underlying mechanisms through which METH promotes the differentiation of neural stem cells into astrocytes.

Thank you very much for your thoughtful questions and valuable suggestions. These suggestions have helped us gain a deeper understanding of the areas where we can improve, and have guided us toward more meaningful directions for future research.

(1) Mitra S, Devi S, Lee MS, Jui J, Sahu A, Goldman D. Vegf signaling between Müller glia and vascular endothelial cells is regulated by immune cells and stimulates retina regeneration. Proc Natl Acad Sci U S A. 2022 Dec 13;119(50):e2211690119. doi: 10.1073/pnas.2211690119.

(2) Wada T, Haigh JJ, Ema M, Hitoshi S, Chaddah R, Rossant J, Nagy A, van der Kooy D. Vascular endothelial growth factor directly inhibits primitive neural stem cell survival but promotes definitive neural stem cell survival. J Neurosci. 2006 Jun 21;26(25):6803-12. doi: 10.1523/JNEUROSCI.0526-06.2006.

(3) Meng J, Fang J, Bao Y, Chen H, Hu X, Wang Z, Li M, Cheng Q, Dong Y, Yang X, Zou Y, Zhao D, Tang J, Zhang W, Chen C. The biphasic role of Hspb1 on ferroptotic cell death in Parkinson's disease. Theranostics. 2024 Aug 1;14(12):4643-4666. doi: 10.7150/thno.98457.

(4) Sisto A, van Wermeskerken T, Pancher M, Gatto P, Asselbergh B, Assunção Carreira ÁS, De Winter V, Adami V, Provenzani A, Timmerman V. Autophagy induction by piplartine ameliorates axonal degeneration caused by mutant HSPB1 and HSPB8 in Charcot-Marie-Tooth type 2 neuropathies. Autophagy. 2025 May;21(5):1116-1143. doi: 10.1080/15548627.2024.2439649.

(5) Gu C, Wang Z, Luo W, Ling H, Cui X, Deng T, Li K, Huang W, Xie Q, Tao B, Qi X, Peng X, Ding J, Qiu P. Impaired olfactory bulb neurogenesis mediated by Notch1 contributes to olfactory dysfunction in mice chronically exposed to methamphetamine. Cell Biol Toxicol. 2025 Feb 20;41(1):46. doi: 10.1007/s10565-025-10004-y.

(6) Rasetto NB, Giacomini D, Berardino AA, Waichman TV, Beckel MS, Di Bella DJ, Brown J, Davies-Sala MG, Gerhardinger C, Lie DC, Arlotta P, Chernomoretz A, Schinder AF. Transcriptional dynamics orchestrating the development and integration of neurons born in the adult hippocampus. Sci Adv. 2024 Jul 19;10(29):eadp6039. doi: 10.1126/sciadv.adp6039.

(7) Pagin M, Pernebrink M, Pitasi M, Malighetti F, Ngan CY, Ottolenghi S, Pavesi G, Cantù C, Nicolis SK. FOS Rescues Neuronal Differentiation of Sox2-Deleted Neural Stem Cells by Genome-Wide Regulation of Common SOX2 and AP1(FOS-JUN) Target Genes. Cells. 2021 Jul 12;10(7):1757. doi: 10.3390/cells10071757.

Reviewer #2 (Public review):

(1) Despite this potential novelty, the study has numerous weaknesses. Notably, single-cell RNA sequencing was unable to capture an adequate number of neuronal populations. Neurons accounted for only approximately 0.6% of the total nuclei, representing a significant underrepresentation compared to their actual physiological proportion. Given that the behavioral effects of METH are likely mediated by neuronal dysfunction, readers would reasonably expect to see transcriptional changes in neurons. The authors should explain why they were unable to capture a sufficient number of neurons and justify how this incomplete dataset can still provide meaningful scientific insights for researchers studying METH-induced hippocampal damage and behavioral alterations.

Thank you sincerely for bringing this important issue to our attention.

Firstly, this represents an unavoidable technical bottleneck. The single-cell sequencing (scRNA-seq) we perform involves the detection of mRNA at the whole-cell level, a process that necessitates cells with high structural integrity, robust viability, and minimal exposure to external stimuli. During the preparation of single-cell suspensions, mature neurons due to their highly differentiated state, morphological rigidity, and excessively long axons often fail to maintain structural integrity. These cells typically undergo death during the dissociation process, lose viability, and are subsequently excluded prior to sequencing. To retain a substantial amount of neuron-related data, an alternative technique single-cell nuclear sequencing (snRNA-seq) should be employed. This method does not necessitate cell viability and focuses exclusively on the nuclei of individual cells, thereby capturing mRNA information solely from the nuclear compartment. Consequently, mRNA data originating from the cytoplasm and organelles will not be represented.

Secondly, numerous studies have shown that the neurological damage caused by chronic exposure to methamphetamine exhibits a high degree of similarity in clinical manifestations and pathogenesis to neurodegenerative diseases (such as Alzheimer's disease, Parkinson's disease, etc.) [1-4].

We fully acknowledge the central role of neurons in cognitive functions and the pathogenesis of cognitive disorders. However, despite decades of neuron-centric research that has yielded significant advancements, major challenges remain in elucidating disease origins, identifying early pathological events, and developing effective therapeutic strategies. For example, current models fail to adequately explain early disease events. Many pathological hallmarks of cognitive disorders such as amyloid plaques, neurofibrillary tangles, and α-synuclein aggregation emerge in the extracellular space long before overt neuronal loss or dysfunction occurs, and are increasingly recognized to be initiated or modulated by non-neuronal cells, including astrocytes and microglia [5]. Furthermore, the critical contribution of the neural microenvironment is often overlooked. Neuronal function and survival are highly dependent on this microenvironment, which is predominantly established and maintained by non-neuronal cell types such as astrocytes, oligodendrocytes, microglia, vascular endothelial cells, pericytes, and interstitial cells and matrix [6-10]. Additionally, systemic factors such as metabolic dysregulation, peripheral inflammation, and vascular pathology are closely associated with cognitive disorders. These factors often initially impact non-neuronal cells, particularly those forming the blood-brain barrier (e.g., endothelial cells) or mediating immune responses (e.g., microglia), before exerting downstream effects on neurons [11,12]. Finally, current therapeutic approaches for neuron face significant limitations, highlighting an urgent need for novel intervention strategies.

During the development of neurodegenerative chronic diseases, although the structural or functional abnormalities of neurons are the direct factors leading to clinical symptoms (such as cognitive decline), this process is often regulated by various auxiliary cell types such as glial cells, immune cells, and stromal cells, and constitutes a complex pathological mechanism network. It is worth noting that the chronic and persistent progression of the disease usually results from the failure of these auxiliary cells to effectively provide support and nutrition to neurons, and even in some pathological states, they transform into effector cells that promote neuronal damage [13,14]. In recent years, a growing number of evidence has demonstrated that glial cells, immune cells, and stromal cells exert critical regulatory functions in the pathogenesis of neurodegenerative diseases. These cell types not only contribute to the maintenance of neural microenvironmental homeostasis during the early stages of disease progression but also display substantial functional heterogeneity in modulating inflammatory responses, synaptic plasticity, the repair of neuronal injury, linking genetic risks with environmental factors and the pathogenic mechanism of pathological protein propagation [15-19]. These research results indicate that they have the potential to become key therapeutic targets in clinical interventions: 1. compared to neurons themselves, they are more susceptible to being targeted by drugs or biological agents (such as antibodies), and have higher accessibility; 2. Non-neuronal cells (especially glial cells) exhibit high plasticity and reactivity during the course of diseases, providing an opportunity window for intervening in their functional states (such as inhibiting harmful activation and promoting protective functions); 3. they can serve as early intervention targets before irreversible damage occurs to neurons, helping to prevent or delay the progression of the disease;4. intervention methods targeting these targets are diverse, including immunomodulation, anti-inflammatory, vascular protection, and metabolic regulation strategies, which are usually more feasible in practical applications than directly protecting the fragile neurons.

Early pharmacological studies have extensively characterized the neurotoxic effects of METH, including the induction of autophagy, apoptosis, oxidative stress, endoplasmic reticulum stress, and dopaminergic neurotoxicity [20]. However, therapeutic options and pharmacological interventions for METH abuse remain limited [21]. In recent years, increasing attention has been directed toward the impact of METH on non-neuronal cells. Research into mechanisms such as neuroinflammatory responses, blood-brain barrier disruption, and immune modulation is progressively contributing to a more comprehensive understanding of METH-induced neural injury [22-24]. Moreover, METH is a substance that induces widespread damage across multiple organ systems and diverse cell types throughout the body. Beyond its effects on neurons, various cell types exhibit distinct responses to METH exposure, which differ significantly depending on the duration of exposure. Our research dataset encompasses high-quality whole-cell mRNA sequencing data from multiple cell types within the hippocampus of mice subjected to chronic METH exposure, offering substantial data support and a robust foundation for in-depth investigation into the pathological mechanisms underlying METH-induced neurodamage.

Thirdly, the selection of scRNA-seq was guided by our experimental objectives and prior research experience. Our earlier investigations have primarily centered on astrocytes, endothelial cells, and microglia. This single-cell sequencing study is intended to enhance our understanding of these neural support cells, comprehensively explore their underlying mechanisms and cellular interactions, and ultimately provide a solid foundation and reference for future research. However, our experience and infrastructure in the field of neuronal research remain relatively limited. To ensure the generation of high-quality data and to systematically advance the experimental objectives, we have prioritized the analysis of the neural microenvironment as the central focus of this study.

Fourthly, the hippocampal region is a brain area with highly specialized and collaborative characteristics, which can be further divided into the ventral hippocampus, the dorsal hippocampus, and multiple subregions such as DG, CA1, CA2, and CA3. The neurons in these subregions exhibit strong heterogeneity, and the experimental methods we currently adopt are still unable to precisely distinguish the neurons in these different regions, which may to some extent affect the accuracy of data interpretation. To address the impact of neuronal heterogeneity, we believe that single-cell spatial transcriptomics technology can be adopted for in-depth research. However, due to the high cost of this technology, it is currently difficult to apply it in our research group.

(1) Lappin JM. Rare but relevant: Methamphetamine and Parkinson's disease. Addiction. 2025 Apr;120(4):797-800. doi: 10.1111/add.16695. Epub 2024 Oct 22. PMID: 39434702.

(2) Lappin JM, Darke S. Methamphetamine and heightened risk for early-onset stroke and Parkinson's disease: A review. Exp Neurol. 2021 Sep;343:113793. doi: 10.1016/j.expneurol.2021.113793. Epub 2021 Jun 21. PMID: 34166684.

(3) Shukla M, Vincent B. The multi-faceted impact of methamphetamine on Alzheimer's disease: From a triggering role to a possible therapeutic use. Ageing Res Rev. 2020 Jul;60:101062. doi: 10.1016/j.arr.2020.101062.

(4) Shrestha P, Katila N, Lee S, Seo JH, Jeong JH, Yook S. Methamphetamine induced neurotoxic diseases, molecular mechanism, and current treatment strategies. Biomed Pharmacother. 2022 Oct;154:113591. doi: 10.1016/j.biopha.2022.113591.

(5) Gabitto MI, et al.. Integrated multimodal cell atlas of Alzheimer's disease. Nat Neurosci. 2024 Dec;27(12):2366-2383. doi: 10.1038/s41593-024-01774-5.

(6) Stogsdill JA, Harwell CC, Goldman SA. Astrocytes as master modulators of neural networks: Synaptic functions and disease-associated dysfunction of astrocytes. Ann N Y Acad Sci. 2023 Jul;1525(1):41-60. doi: 10.1111/nyas.15004.

(7) Terreros-Roncal J, et al.. Impact of neurodegenerative diseases on human adult hippocampal neurogenesis. Science. 2021 Nov 26;374(6571):1106-1113. doi: 10.1126/science.abl5163.

(8) Zhu K, Fu Y, Zhao Y, Niu B, Lu H. Perineuronal nets: Role in normal brain physiology and aging, and pathology of various diseases. Ageing Res Rev. 2025 Jun;108:102756. doi: 10.1016/j.arr.2025.102756.

(9) Depp C, Doman JL, Hingerl M, Xia J, Stevens B. Microglia transcriptional states and their functional significance: Context drives diversity. Immunity. 2025 May 13;58(5):1052-1067. doi: 10.1016/j.immuni.2025.04.009.

(10) Sweeney MD, Zhao Z, Montagne A, Nelson AR, Zlokovic BV. Blood-Brain Barrier: From Physiology to Disease and Back. Physiol Rev. 2019 Jan 1;99(1):21-78. doi: 10.1152/physrev.00050.2017.

(11) Nation DA, et al.. Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. Nat Med. 2019 Feb;25(2):270-276. doi: 10.1038/s41591-018-0297-y.

(12) Montagne A, Zhao Z, Zlokovic BV. Alzheimer's disease: A matter of blood-brain barrier dysfunction? J Exp Med. 2017 Nov 6;214(11):3151-3169. doi: 10.1084/jem.20171406. Epub 2017 Oct 23.

(13) Huang Q, Wang Y, Chen S, Liang F. Glycometabolic Reprogramming of Microglia in Neurodegenerative Diseases: Insights from Neuroinflammation. Aging Dis. 2024 May 7;15(3):1155-1175. doi: 10.14336/AD.2023.0807.

(14) Shi FD, Yong VW. Neuroinflammation across neurological diseases. Science. 2025 Jun 19;388(6753):eadx0043. doi: 10.1126/science.adx0043.

(15) Xu X, Mei B, Yang Y, Li J, Weng J, Yang Y, Zhu Q, Zhang H, Liu X. Astrocytes Lingering at a Crossroads: Neuroprotection and Neurodegeneration in Neurocognitive Dysfunction. Int J Biol Sci. 2025 Apr 28;21(7):3122-3143. doi: 10.7150/ijbs.109315.

(16) Bedolla A, et al.. Adult microglial TGFβ1 is required for microglia homeostasis via an autocrine mechanism to maintain cognitive function in mice. Nat Commun. 2024 Jun 21;15(1):5306. doi: 10.1038/s41467-024-49596-0.

(17) Castellani G, Croese T, Peralta Ramos JM, Schwartz M. Transforming the understanding of brain immunity. Science. 2023 Apr 7;380(6640):eabo7649. doi: 10.1126/science.abo7649.

(18) Chen YH, Jin SY, Yang JM, Gao TM. The Memory Orchestra: Contribution of Astrocytes. Neurosci Bull. 2023 Mar;39(3):409-424. doi: 10.1007/s12264-023-01024-x.

(19) Deng Q, Wu C, Parker E, Liu TC, Duan R, Yang L. Microglia and Astrocytes in Alzheimer's Disease: Significance and Summary of Recent Advances. Aging Dis. 2024 Aug 1;15(4):1537-1564. doi: 10.14336/AD.2023.0907.

(20) Jayanthi S, Daiwile AP, Cadet JL. Neurotoxicity of methamphetamine: Main effects and mechanisms. Exp Neurol. 2021 Oct;344:113795. doi: 10.1016/j.expneurol.2021.113795.

(21) Paulus MP, Stewart JL. Neurobiology, Clinical Presentation, and Treatment of Methamphetamine Use Disorder: A Review. JAMA Psychiatry. 2020 Sep 1;77(9):959-966. doi: 10.1001/jamapsychiatry.2020.0246.

(22) Shi S, Sun Y, Zan G, Zhao M. The interaction between central and peripheral immune systems in methamphetamine use disorder: current status and future directions. J Neuroinflammation. 2025 Feb 15;22(1):40. doi: 10.1186/s12974-025-03372-z.

(23) Pang L, Wang Y. Overview of blood-brain barrier dysfunction in methamphetamine abuse. Biomed Pharmacother. 2023 May;161:114478. doi: 10.1016/j.biopha.2023.114478.

(24) Shaerzadeh F, Streit WJ, Heysieattalab S, Khoshbouei H. Methamphetamine neurotoxicity, microglia, and neuroinflammation. J Neuroinflammation. 2018 Dec 12;15(1):341. doi: 10.1186/s12974-018-1385-0.

(2) Another significant weakness of this study is the lack of a cohesive hypothesis or overarching conclusion regarding how METH impacts neural populations. The authors provide a largely descriptive account of transcriptional alterations across various cell types, but the manuscript lacks clear, biologically meaningful conclusions. This descriptive approach makes it difficult for readers to identify the key findings or take-home messages. To improve clarity and impact, the authors should focus on developing and presenting a few plausible hypotheses or mechanistic scenarios regarding METH-induced neurotoxicity, grounded in their scRNA-seq data. Including schematic figures to illustrate these hypotheses would also help readers better understand and interpret the study.

We sincerely appreciate your valuable comments on our article. As you pointed out, the current research lacks experimental verification to further support our conclusions. To enhance the clarity and readability of the mechanism explanation, we have added several hypothetical diagrams (such as Figures.7, 8, and 9) in the discussion section to present the biological mechanisms reflected by the data more intuitively. Additionally, relevant verification work is underway, such as marking specific cell types with marker proteins. Author response image 1 shows some of our preliminary experimental results that have not been published yet, and their trends are consistent with the conclusions of this article. However, since the complete verification still requires a certain period of time, to ensure the rigor of the data, these results have not been included in the current manuscript for the time being. Finally, we would like to thank you again for your constructive suggestions.

Author response image 1.

(3) The final major weakness of this study is its poor readability. It appears that the authors did not adequately proofread the manuscript, as there are numerous typographical errors (e.g., line 333: trisulting; line 756: essencial), unsupported scientific claims lacking citations (e.g., lines 485, 503, 749-753), and grammatically incorrect sentences (e.g., lines 470-472, 540-543, 749-753). In addition, many paragraphs are unorganized and overly descriptive, which further hinders clarity. Some figures are also problematic - too small in size and overcrowded with text in fonts that are difficult to read. It is recommended that the authors carry out quality control. There are too many typographical and grammatical errors to list individually; the authors should carefully review and revise the entire manuscript to address all of these issues.

We truly appreciate your thoughtful feedback and sincerely apologize for any inconvenience experienced by you and other readers.

The text of this research manuscript was manually entered, which unfortunately resulted in some spelling and grammatical errors. In response, we have carefully revised the entire manuscript using word processing tools in the second version. Meanwhile, we have restructured and organized some lengthy paragraphs to enhance the clarity and readability of the content.

Regarding the issue you raised about certain viewpoints lacking citation support, we have added the necessary references to those sections and reviewed the entire text to ensure all scientific claims are properly supported. 

As for the image clarity, we made sure the submitted images met the 600dpi resolution requirement. However, we acknowledge that there were clarity issues in the final published version. We have since re-adjusted and re-uploaded the images to improve their quality.

We are committed to continuously improving the manuscript and enhancing the overall quality of our academic presentation. Thank you sincerely for your kind attention to our work, your careful review, and the valuable suggestions you provided.

Reviewer #3 (Public review):

(1) While the bioinformatics analyses are extensive, the study is primarily descriptive at the molecular level. The absence of experimental validation, such as targeted mRNA/protein quantification and gene knockdown/overexpression to confirm the causal relationship between these identified genes and METH-induced cognitive deficits, is a notable limitation.

We sincerely appreciate your valuable comments and suggestions. Indeed, there are still certain limitations in our manuscript in some aspects. It may not be able to systematically answer specific questions, and it is also difficult to fully clarify the functional roles of certain genes or specific cell types through experimental evidence.

Although our manuscript still has certain limitations, we believe that the publication of this research is expected to provide new perspectives and theoretical support for the in-depth exploration of METH toxicity damage-related fields, thereby promoting the progress of research in this direction：

(1) At present, the single-cell sequencing datasets on chronic damage caused by METH are still relatively limited, especially in terms of studies at the whole-cell level. Our dataset is expected to fill the research gap in this field to some extent, providing reference and support for subsequent related research.

(2) During the sampling process of the sequencing experiment, we ensured high cell viability and sequencing quality. The experiment exhibited good reproducibility (each group consisted of 10 mice, and 2 mice from each group were selected to mix their hippocampal tissues into one sample), and the obtained data had high credibility.

(3) The effects of METH have a wide distribution pattern across various organs and tissues. Through single-cell sequencing data, the common and differential expression patterns of related genes under different conditions can be systematically analyzed, which is helpful for future targeted knockout studies of these genes and provides a predictive basis for the evaluation of intervention measures, thereby enabling precise regulation of gene functions.

(4) This is conducive to the orderly implementation of our subsequent research plans. Our subsequent research plan can be further developed based on a specific aspect of this study. We are indeed planning to do exactly that. During our earlier research on astrocytes, we discovered that astrocytes have two phenotypes (protective and inflammatory) in neuroinflammation. Given that astrocytes in the hippocampus show great variability depending on their location, the cells they come into contact with, and the stimuli they receive, we aim to investigate the changes in the function of astrocyte subpopulations in chronic METH-induced cognitive impairment. We focused on the role of the cAMP signaling pathway in the transformation of astrocyte phenotypes and attempted to link changes in astrocyte energy metabolism to their inflammatory phenotype. In addition, we found that endothelial cells can be easily distinguished into many subpopulations, which are related to their specific functions in immune responses, material transport, vascular growth regulation, energy metabolism, and other processes. We believe that single-cell technology can help us find the key mechanisms and intervention targets of chronic METH abuse-induced damage with greater precision.

(2) While the discussion extensively covers the functional implications of specific molecular pathways and cell types, it would greatly benefit from a comparison of these findings with existing RNA sequencing data from other METH models in hippocampal tissue.

We are very grateful for your professional suggestions, which have been of great help in improving the quality of our manuscript. We agree that comparing our findings with existing RNA sequencing data from other METH models in hippocampal tissue would strengthen the discussion. In response to your suggestion, we have actively reviewed relevant literature and databases, and attempted to request the database administrators and original authors for the download and use of the relevant data. However, as data integration still requires some time, we may not be able to conduct a detailed analysis of the data in this revised version. We can only discuss the conclusions of some authors.

Palsamy Periyasamy et al. published a scRNA sequencing (live-cell) study on chronic METH exposure almost at the same time as us. They also adopted a similar gradual incremental 4-week METH exposure model and conducted sequencing analysis on glial cells in the cerebral cortex of mice [1]. The changes they observed in the circadian rhythm, adherens junctions, Rap1 signaling pathway, and cAMP signaling pathway (Disscusion, Lines 892-897) in the cortical astrocytes were also similar in the astrocytes of the hippocampal region that we studied. Similarly, in oligodendrocytes, we observed an upregulation trend of key genes regulating the circadian rhythm, such as Per2, Per3, and Nr1d1 (Disscusion, Lines 916-939). This result is consistent with their research findings. Non etheless, we believe that the changes in oligodendrocytes in terms of metabolic regulation and axonal function homeostasis are more significant.

Pingming Qiu et al. further confirmed the correlation between the NF-κB signaling pathway in hippocampal astrocytes under METH action and neuroinflammation, neuroinjury, and learning and memory impairments in mice by integrating the GEO dataset [2]. This conclusion is also consistent with the sequencing results and analysis conclusions we obtained (Results, Lines 473-476).

In terms of the neuro-immune system disorder caused by chronic METH exposure, our research findings are consistent with those of Biao Wang et al [3]. We both observed that METH exposure may involve the participation of related immune cells (such as T cells, monocytes) and may be related to the regulation of the innate immune response and the homeostasis of myeloid cells, etc. Through the identification and analysis of cell subtypes, we further revealed that these signals may be closely related to the interaction between microglia and other immune cells mediated by MHC molecules (Disscusion, Lines 870-894).

Currently, the research results related to METH are still scattered and lack systematicness. There are differences among the research models, and there are relatively few studies on chronic exposure and in vivo experiments. Sequencing data sets with strong correlations are also scarce. We hope that this dataset can comprehensively and elaborately depict the molecular map of the hippocampus of mice after chronic METH exposure (although due to technical limitations, mature neurons die during dissociation, thus making it impossible to obtain the relevant data). In addition, we also hope to integrate the single-cell sequencing data and spatial transcriptome data of the hippocampus of mice after chronic METH exposure, providing a reliable data foundation and theoretical support for subsequent research in this field.

Finally, we would like to express our sincere gratitude for your valuable suggestions and support. Although we still need some time to further refine the manuscript based on your opinions, we sincerely hope that more readers will provide us with constructive feedback to promote the continuous improvement and deepening of this research.

(1) Oladapo A, Deshetty UM, Callen S, Buch S, Periyasamy P. Single-Cell RNA-Seq Uncovers Robust Glial Cell Transcriptional Changes in Methamphetamine-Administered Mice. Int J Mol Sci. 2025 Jan 14;26(2):649. doi: 10.3390/ijms26020649.

(2) Li K, Ling H, Wang X, Xie Q, Gu C, Luo W, Qiu P. The role of NF-κB signaling pathway in reactive astrocytes among neurodegeneration after methamphetamine exposure by integrated bioinformatics. Prog Neuropsychopharmacol Biol Psychiatry. 2024 Feb 8;129:110909. doi: 10.1016/j.pnpbp.2023.110909.

(3) Wu L, Liu X, Jiang Q, Li M, Liang M, Wang S, Wang R, Su L, Ni T, Dong N, Zhu L, Guan F, Zhu J, Zhang W, Wu M, Chen Y, Chen T, Wang B. Methamphetamine-induced impairment of memory and fleeting neuroinflammation: Profiling mRNA changes in mouse hippocampus following short-term and long-term exposure. Neuropharmacology. 2024 Dec 15;261:110175. doi: 10.1016/j.neuropharm.2024.110175.

(3) The conclusion that "prolonged METH use may progressively impair cognitive function" may not be uniformly supported by the behavioral data: Figures 1C and F (discrimination and preference indexes) exhibited that the 4-week test further declined in the METH group compared to the 2-week. In contrast, Figure 1E and H present a contradictory pattern.

Thank you very much for pointing this out. Your observation is very detailed and constructive. Regarding the conclusion "prolonged use of METH may progressively impair cognitive function", our main basis is the discrimination index and preference index shown in Figures 1C and 1F. These two indicators are usually calculated based on the total exploration time of new and old objects by mice. They are widely adopted as important references for cognitive function assessment in many relevant literature [1-3], thus providing strong support for our conclusion. The exploration frequency data we provided can, on the one hand, reflect the curiosity of mice towards new things, and on the other hand, can be calculated as the average time of each exploration by "total exploration time / exploration frequency", thereby evaluating their learning interest and the degree of their focus during exploration. We believe this is also of certain significance for reflecting the effect of METH on learning. As for the fact that there is no statistically significant difference in the exploration frequency of new and old objects in the 4-week-old mice in Figure 1H, we are also regretful about this. This might be due to the fact that our tests allow mice to freely explore in a stress-free environment, and there are significant differences among individual mice within the group. However, the mean values still show certain differences between the two groups. Compared to the mice at 2 weeks, the mice at 4 weeks have undergone a NOR test once and may have formed memories, which were retained in the subsequent assessment after four weeks. Moreover, we believe that injecting normal saline to the control group mice for a long time may affect their emotional state, because they cannot obtain the same pleasure as that brought by METH from the injection behavior.

(1) Riva M, Moriceau S, Morabito A, Dossi E, Sanchez-Bellot C, Azzam P, Navas-Olive A, Gal B, Dori F, Cid E, Ledonne F, David S, Trovero F, Bartolomucci M, Coppola E, Rebola N, Depaulis A, Rouach N, de la Prida LM, Oury F, Pierani A. Aberrant survival of hippocampal Cajal-Retzius cells leads to memory deficits, gamma rhythmopathies and susceptibility to seizures in adult mice. Nat Commun. 2023 Mar 18;14(1):1531. doi: 10.1038/s41467-023-37249-7.

(2) Lu Y, Chen X, Liu X, Shi Y, Wei Z, Feng L, Jiang Q, Ye W, Sasaki T, Fukunaga K, Ji Y, Han F, Lu YM. Endothelial TFEB signaling-mediated autophagic disturbance initiates microglial activation and cognitive dysfunction. Autophagy. 2023 Jun;19(6):1803-1820. doi: 10.1080/15548627.2022.2162244.

(3) Arroyo-García LE, Tendilla-Beltrán H, Vázquez-Roque RA, Jurado-Tapia EE, Díaz A, Aguilar-Alonso P, Brambila E, Monjaraz E, De La Cruz F, Rodríguez-Moreno A, Flores G. Amphetamine sensitization alters hippocampal neuronal morphology and memory and learning behaviors. Mol Psychiatry. 2021 Sep;26(9):4784-4794. doi: 10.1038/s41380-020-0809-2.

eLife Assessment

This important cross-species study tests whether the corpus callosum contains parallel, segregated pathways for ipsilateral and contralateral visual-field information, rather than mixed inputs from the two hemispheres. A major strength is its use of a combination of high-field functional magnetic resonance inaging and Bayesian population receptive field (pRF) modelling in humans with viral tracing in mice to offer complementary evidence for pathway segregation. At present, the evidence supporting the authors' claims is incomplete and would benefit from ruling out potential confounds that could mimic tract segregation in the human white-matter pRF data and the mouse anatomical tracing results, and from sharpening claims about laminar specificity.

Reviewer #1 (Public review):

Summary:

This study combined high-field fMRI with computational modelling (including a Bayesian population receptive field [pRF] model and functional gradient analysis) in humans to demonstrate that the architecture of the corpus callosum (CC) and its interhemispheric connections is organized into parallel ipsilateral and contralateral streams, rather than functioning as a mixed integration of inputs from both hemispheres. The human findings were validated through preclinical experiments in mice using viral axonal tracing, which revealed a non-overlapping laminar arrangement of axons carrying left and right visual field information.

These results suggest that the CC operates as a set of parallel, segregated pathways, with each stream independently conveying information from one side of the visual field. This organization preserves the spatial origin of visual signals within the white matter. Although the overall concept of interhemispheric parallel pathways is not entirely unexpected, this refined understanding of callosal organization provides important scientific and clinical insights in relation to pathway-specific perturbations and in neurological disorders.

Strengths:

The manuscript is well written, the methodology is sound, and the analyses are carefully conducted. I particularly appreciate the effort to integrate functional and structural approaches and to validate the human neuroimaging findings with more sensitive preclinical techniques, such as viral tracing.

Weaknesses:

Several points require clarification to allow a more complete interpretation of the results. In addition, some further analyses are necessary to fully substantiate the claims made in the manuscript. These are detailed below

Comment 1:

BOLD signals in white matter remain a matter of debate, although this is not the central focus of the present study. Nevertheless, it is important to establish whether the underlying data have sufficient tSNR to support robust pRF estimation in white matter. In Figure 1, the EV appears relatively robust; however, it seems that only the best-fitting examples are shown. In contrast, the group-average EV reported in Figure 2, and the individual maps in the Supplementary Information indicate very low EV values, typically below 5%. In conventional fMRI analyses, thresholds of approximately 15-20% EV are often applied to exclude poor fits that may bias pRF parameter estimates. It appears that no such threshold was applied here. Interestingly, in Figure S6, the average EV for dual pRF models appears to be approximately 17%. Do dual and triple pRF models systematically produce higher EV compared to single pRF models? Additionally, Figure 2 suggests the presence of baseline activation that is captured by the model. Could this be related to a delayed or altered hemodynamic response function (HRF) in white matter? Clarification would be helpful. To better assess the robustness of the reported findings, the authors should provide quantitative measures of tSNR within the white matter tracts where the pRF model was fitted. Furthermore, a plot showing the average BOLD signal during visual stimulation versus baseline in those tracts would greatly strengthen confidence in the signal quality.

Although the reported linear relationship between pRF size and eccentricity, as well as the test-retest reliability analyses, suggest the presence of consistent receptive field estimates, these analyses are based on distributions and may lack the sensitivity required to differentiate single, dual, and triple pRF models. Moreover, the pRF estimates within the FMA appear noisy, particularly at the individual level (Figure S4), making it difficult to clearly dissociate information originating from the left and right hemifields.

Comment 2.1:

The Bayesian modelling approach is interesting and robust. However, as I understand it, the authors must specify a priori the number of pRFs to be estimated. This introduces a strong assumption about the expected underlying receptive field structure. An alternative Bayesian approach, such as micro-probing (Carvalho et al., 2020), does not require prior assumptions regarding the number or shape of pRFs. Instead, it estimates receptive field profiles in a more data-driven manner and provides a direct visualization of the pRF structure. Implementing such an approach, or at least comparing it with the current modelling strategy, could yield more reliable and potentially less biased estimates of multiple pRFs, particularly in white matter where signal quality is limited.

Comment 2.2:

Some clarifications regarding the pRF model are needed: in the Methods section, the authors mention the use of a Difference-of-Gaussians (DoG) model. However, it appears from the Results that the analyses were performed using a single-Gaussian model. Additionally, in Section 5.6, the authors state that six different pRF models were tested. Which specific models were included in this comparison? A clear description of each model, along with justification for the final model selection criteria, would help better understand the study

Comment 3:

Throughout the manuscript, the authors repeatedly refer to laminar-specific findings. However, the reported functional resolution of 1.6 mm isotropic is insufficient to reliably resolve cortical layers. Given this limitation, the laminar interpretations appear overstated. For example, in the Discussion section titled "Integrating White Matter with Laminar-Resolved Function", the authors state: "The combination of anatomically segregated white matter pathways with functionally specific cortical laminae presents a powerful synergy for human brain circuit research." Given the spatial resolution of the functional data, how are laminar-specific functional claims justified?

Similarly, the authors suggest that: "It becomes possible to assess not just if the CC is damaged, but precisely which directional pathways are compromised-either the pathways projecting from the lesioned hemisphere, or those projecting to the other, or both." It is unclear to me how the current methodology uniquely enables this level of directional specificity, and whether this was not already feasible using existing structural and diffusion-based approaches. The authors should clarify what is genuinely novel in this study.

Comment 4:

In the Discussion, the authors state: "These findings fundamentally reframe our understanding of interhemispheric communication, moving beyond static connectivity to reveal a dynamic, directionally specific highway where spatial location encodes the origin of information. This framework provides a novel blueprint for decoding directional information flow in the living human brain." Based on the analyses presented, it is unclear how the findings of this study demonstrate dynamic connectivity or true directional specificity. The reported results appear to characterize spatial organization and segregation of callosal pathways, but they do not measure the directionality of information flow, temporal dynamics, or causal directionality between hemispheres. To substantiate claims regarding dynamic or directional communication, additional analyses, such as connective field model (Haak et al.2013), effective connectivity modelling, time-resolved approaches, or perturbation-based methods (neuromodulation) would be required. As currently presented, the findings seem to support structural and functional segregation rather than dynamic or directionally resolved interhemispheric information transfer. The authors should either provide stronger evidence for these claims or moderate them.

Comment 5:

I agree with the authors that pooling of information across hemispheres represents a plausible explanation for the presence of dual pRFs. As discussed in the manuscript, such an effect would be expected to predominantly affect pRFs located near the vertical meridian. However, Figures S6C and S6D do not appear to demonstrate that bilateral pRFs are preferentially located along the vertical meridian.

Reviewer #2 (Public review):

Summary:

The manuscript proposes a "parallel wires" architecture for the visual corpus callosum, suggesting that contralateral and ipsilateral visual streams remain spatially segregated into distinct anatomical channels. The authors use a cross-species approach, combining Bayesian population receptive field (pRF) modeling in humans with dual-color viral tracing in mice. The analysis of the publicly available human fMRI dataset indicates a 92% probability of single-hemifield representation, arguing for functional segregation. The mouse mesoscale tracing data support the idea of anatomical parallel wires by displaying dorso-ventral segregation of callosal axons post-midline crossing.

Strengths:

The primary strength of this study is its cross-species integration. Observing that functional segregation in humans is mirrored by specific anatomical pathways in the mouse provides a convincing, multimodal argument for the "parallel wires" hypothesis. The data is generally well-presented, and the Bayesian modeling of the human data is a robust methodological choice.

Weaknesses:

There are weaknesses in the description, presentation, and methodological details of the mouse tracing data. First, the authors must provide detailed information regarding spectral unmixing, intensity normalization, and threshold-sensitivity analyses. These factors are critical as they directly influence the Dice and Jaccard overlap estimates that underpin the study's primary conclusions. Second, it is unclear which cortical layers have been virally labelled as there is no quantification of the spatial extent of the injection site, and there is ambiguity regarding the dorso-ventral stereotaxic coordinates.

Reviewer #3 (Public review):

Summary:

This manuscript describes a study into the functional organization of the forceps major (FMA). The authors present a Bayesian population receptive field (pRF) analysis of group-averaged HCP fMRI retinotopic mapping data, focusing on voxels within the FMA. This is unconventional because pRF modelling is usually limited to gray matter voxels, where synaptic activity underlying neural computation is the highest. Nevertheless, some previous work suggests that meaningful fMRI signals can also be gleaned from white matter voxels, where the signals are thought to reflect metabolic activity from action potentials that travel along axons. However, these signals are generally much noisier, and possible confounding effects due to partial voluming, draining veins, and different hemodynamics must be carefully ruled out. Based on the Bayesian pRF analysis, the authors claim evidence of segregated contralateral and ipsilateral representations of the visual field in the FMA. Anatomical tract tracing based on HCP diffusion MRI data from seeds identified using the pRF analysis further suggests that these representations are underpinned by separate fiber bundles, which also appear to be consistent with the results of viral tracing in mice. The results of this study could mean an important step forward in understanding transcallosal signaling.

Strengths:

The study treads uncharted territory, leveraging multiple data modalities across species and advanced analytical approaches.

Weaknesses:

The study does not address potential confounds related to BOLD imaging in white matter structures. If the fMRI results can be explained based on neighboring grey matter responses, the evidence that remains is limited to an apparent anatomical segregation of white matter bundles that appear to be present in both mice and humans.

Further details are also missing regarding the Bayesian pRF approach, including the priors used for the pRF model. These are important as they will dominate the estimates when the data are very noisy, and the authors have adopted unconventional, more complex pRF models compared with earlier work employing Bayesian pRF analyses.

It appears that the authors have not applied any statistical thresholding to ensure that only good-quality model fits are entered into subsequent analyses (i.e., the reported probabilities pertain to model comparisons, not goodness of fit). From Figure 2, it appears that the majority of the FMA voxels, barring those adjacent to visual gray matter, do not exhibit more than a few percentage points of explained variance (EV). In fact, a common threshold is >15% EV, but it looks like none of the FMA voxels exceed this threshold.

eLife Assessment

This study presents valuable findings regarding cardiac and autonomic effects of seizures and epilepsy, with relevance to sudden unexpected death in epilepsy (SUDEP). They present solid evidence that genetic deletion of the potassium-chloride co-transporter in hypothalamic corticotropin-releasing hormone (CRH) neurons exacerbates bradycardia and enhances autonomic disturbances in a mouse model of temporal lobe epilepsy. However, the evidence that this deletion produces chronic hyperexcitability of the hypothalamic-pituitary-adrenal axis was incomplete, leaving a mechanistic gap. This work will be of interest to neuroscientists working on epilepsy, the HPA axis, and autonomic control.

Reviewer #1 (Public review):

Summary:

The manuscript entitled "Autonomic reflex plasticity associates with time-dependent SUDEP susceptibility in a murine model with hyperactive stress circuits" by Dr. Saunders and colleagues combined a traditional mouse model of SUDEP, ventral intrahippocampal kainite (vIHKA) injection, with a genetic model of chronic hyperactivity of central corticotropin-releasing hormone (CRH) neurons (Kcc2/Crh) that further increases the risk of SUDEP in the weeks following seizure.

Strengths:

Their results show during spontaneous seizures Kcc2/Crh mice had more pronounced reflex-like ictal bradycardias compared to WT controls that notably occurred prior (~10 sec) to seizure termination and had greater autonomic disturbances compared to WT controls, including a pronounced serotonin-mediated Bezold Jarisch reflex. These results show chronic hyperactivity of central corticotropin-releasing hormone (CRH) neurons (Kcc2/Crh) increased autonomic disturbances and risk of SUDEP in a kainic acid model of epilepsy.

Weaknesses:

This study could be improved with a more thorough assessment of heart rate, blood pressure and breathing during and following the seizures, and in particular the fatal event. It is unclear if the bradycardias were spontaneous or a result of preceding central or obstructive apneas, oxygen desaturations, hypercapnia, arrhythmias, or other possible triggers.

Considerable prior work in the literature suggests SUDEP could be mediated, in some patients, by a burst of parasympathetic activity to the heart. Were the heart rate changes in these animals during seizures inhibited or blocked by atropine or atenolol?<br /> The injection of the 5HT agonist phenylbiguanide into the right jugular is not a selective approach for activating the Bezold Jarisch Reflex (BJR), which is caused by increased activity of intracardiac sensory neurons (generally activated with ischemia or a combination of low preload with high contractility). The results should be interpreted more cautiously, as a response to systemic administration of phenylbiguanide only.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors set out to evaluate the role of hypothalamic pituitary axis hyperactivity on cardiac and autonomic changes during epileptogenesis and following seizures in a mouse model of temporal lobe epilepsy. Epilepsy is very common. It can frequently result in death from sudden unexpected death in epilepsy, or SUDEP. SUDEP is thought to be at least in part due to seizure-related cardiac and autonomic instability. Increased stress states are well known to be comorbid with epilepsy. This comorbidity is thought to increase the risk of SUDEP. Here, the authors hypothesized that a mouse model of heightened stress in which there is hyperactivity of the CRH neurons in the hypothalamus would demonstrate exaggerated cardiac and autonomic effects of seizures and epilepsy.

Strengths:

For the chronic stress model, they employed the Kcc2/Crh mice that have a genetic deletion of the potassium chloride cotransporter in CRH neurons. They treated these mice and their wild-type littermates with intra-hippocampal kainic acid or saline, as epileptic and sham-treated animals, respectively. The assessed cardiac activity, blood pressure, baroreflex, and the Bezold-Jerisch reflex during epileptogenesis. This, in general, is an interesting study. They make some interesting and potentially important observations regarding heart rate and blood pressure in seizures and epilepsy.

Weaknesses:

Some of the conclusions may be a bit overstated as is and would benefit from more discussion and perhaps additional data.

eLife Assessment

This is a useful study, bolstering our understanding of spatial reference frames of visual perception. The high-resolution data and sophisticated analyses confirm and enhance earlier findings that visual representations operate in a predominantly retinotopic reference frame throughout the visual hierarchy in the human cortex. However, these analyses are currently incomplete, leaving open the possibility that eye-position gain and or spatiotopic representations may also be present.

Reviewer #1 (Public review):

In this study, Szinte et al. measured the spatial selectivity of fMRI BOLD responses while subjects viewed dynamic noise stimuli vignetted by a moving bar aperture. Subjects viewed these moving bar stimuli as they fixated at one of three screen locations. This design enabled the authors to test whether fMRI responses are better explained by a model in which stimulus location is encoded relative to the retina or relative to the screen (in other words, 'retintopic' vs. 'spatiotopic' encoding). In retinotopic encoding, the pRFs should move with the eyes. In spatiotopic encoding, the pRFs should be locked to particular screen locations, regardless of eye position. The results are unambiguous: the retinotopic model wins.

A number of prior human fMRI studies have addressed this issue, and there is an overwhelming consensus in the field that spatial encoding throughout human visual cortex (and high-level cortex) is retinotopic (during fixation). All of the results shown in the present manuscript are consistent with these earlier observations. Szinte et al. also find that the degree of retinotopic selectivity is not affected by the task or locus of spatial attention. This too has been observed in multiple prior studies.

So, while this manuscript is primarily confirmatory, the study does nonetheless provide valuable measurements at 7T with a higher signal-to-noise ratio and high spatial resolution than previous studies. The authors also apply an innovative Bayesian decoding analysis (which is beautifully documented on their webpage, with a step-by-step tutorial and ample examples). So, a major strength of this paper is the methods; this study does set a high standard and is an ideal example for a rigorous, replicable analysis pipeline and cutting-edge statistical inference.

The results focus on the spatial profile of pRFs with different eye positions. However, the main idea behind eye-position gain fields is that the amplitude of the visual responses changes with eye position. I could not find any analysis testing response amplitude as a function of eye position. In the Discussion, the authors assert: "We did not find an influence of gaze position at the level of individual voxels nor at the level of visual areas." The authors speculate that this might be because gain fields have a salt-and-pepper organization in the cortex that cancel out when pooled across a voxel. While the salt-and-pepper explanation seems like perfectly fine speculation, here they are discussing a result that isn't shown in the Results!

Several prior human fMRI studies have reported eye position gain fields in humans, suggesting that the salt-and-pepper explanation is not correct. Rather, it is likely the case that the authors did not test a sufficiently wide range of eye positions to detect a gain modulation. For example, a study from Merriam et al. (J. Neurosci, 2013), which is mysteriously not cited here, measured both the spatial selectivity of visual receptive fields AND the response amplitude at 8 different eye positions that were spaced by as much as 24 degrees of visual angle (including both vertical and horizontal changes in eye position). Under these conditions, Merriam et al. did find reliable modulation in response amplitude with changes in eye position, even though the spatial selectivity of the responses did not change. Importantly, Merriam et al. found that visual response selectivity was consistent with a retinotopic reference frame (not a spatiotopic reference frame) and that this selectivity was invariant to the attention task. Consideration of these issues suggests that the experimental design used in the current experiment may have precluded the detection of eye position gain fields. The current manuscript would be much improved by a careful consideration of this prior literature, which is so closely related to what the authors report here.

Reviewer #2 (Public review):

Summary:

This manuscript describes a study using fMRI voxel-wise receptive field modeling and Bayesian decoding to assess the reference frame (spatiotopic vs retinotopic) of visual information. Participants viewed sequences of visual stimuli that moved across different screen locations. Across different conditions, participants either fixated at the screen center and viewed stimuli drifting across the full screen (full-screen condition), or fixated at a central, left, or right fixation position while stimuli drifted across a 4-deg aperture centered on that fixation (gaze-center, gaze-left, gaze-right conditions). Within each of those conditions, participants either attended to visual changes around fixation (attend-fix) or in the stimulus bar (attend-bar). First, standard population receptive field mapping was conducted on the full-screen conditions to obtain fiducial maps for each subject. Then, a variety of different analyses were performed, testing retinotopic vs spatiotopic predictions for the gaze-left and gaze-right conditions. Across the extensive set of analyses performed, and across all ROIs tested, the results always best matched the retinotopic predictions. This was the case for both attend-fix and attend-bar conditions. The authors conclude that visual representations operate in a retinotopic reference frame throughout the visual hierarchy, necessitating a "re-orienting" of the search for visual stability mechanisms.

Strengths:

The analyses are sophisticated and thorough, and the results are convincingly in favor of retinotopic representations. The attention manipulation is carefully done. And the finding that the most informative/reliable voxels are the most retinotopic is an important novel contribution.

Weaknesses:

(1) The theoretical advance of this work is unclear, because the finding that visual representations operate in a retinotopic reference frame throughout the visual hierarchy, and regardless of the deployment of spatial attention, has already been demonstrated with fMRI pattern analysis almost 15 years ago (Golomb & Kanwisher, 2012). To be clear, the techniques used in this current study are considerably more modern and sophisticated, and the attention manipulation is much better, but the finding is the same. More importantly, it is never really explained why, from a theoretical perspective, the results might have been expected to differ. Referring to this as an open question feels like a copout. The manuscript needs to engage more with the prior findings and explain the motivation for the current study. Was there something about the prior findings that caused them to doubt the retinotopic conclusion? Did they think that the 7T resolution or alternative decoding approaches might uncover something different? Was this intended as a replication test with more sophisticated techniques?

(2) I think there are definitely some new and useful things this study has to offer, but the overall theoretical contribution needs to be better clarified and contextualized within the prior literature. I would strongly recommend revisiting things like the title (not a novel contribution of this study) and the implication that the current findings "reframe" or "reorient" the search for visual stability mechanisms away from static spatiotopic maps (the field has arguably been "reoriented" in that way for some time now, and this study is certainly not the first to suggest a reframing along these lines). The discussion section, in particular, has little to no acknowledgement that these findings and ideas have been shown before.

(3) The analyses always pit retinotopic vs spatiotopic predictions. But what if both types co-existed, just with retinotopic more predominant? I think this general idea needs some discussion, if not additional analyses. Would the analyses be sensitive enough to pick up sparse spatiotopic coding if present?

Additional questions/critiques/suggestions:

(4) For the out-of-sample predictions analysis (Figure 2):

a) The spatiotopic predictions are much worse for earlier visual regions, but don't seem so different from gaze-center or retinotopic in later areas. How much might this be driven by the fact that pRF size increases along the hierarchy, and for large pRF sizes, the retinotopic and spatiotopic predictions might not be very differentiable? Is there a way to quantify this or include a control model that is neither retinotopic nor spatiotopic?

b) It looks like in some of the regions, the retinotopic (and maybe even spatiotopic) R2 change compared to the gaze center is reliably positive. Why would this be? Is there a reason the fit should be better for the gaze right or gaze left conditions compared to the gaze center?

(5) For the fitting retinotopic and spatiotopic pRF models (Figure 3) and other voxel-specific analyses:

a) For many of the statistics, results are averaged across voxels. This makes sense. But it also seems to me that taking a simple average might obscure some of the potential advantages of this voxel-wise approach. For example, what if there are sparse spatiotopic effects that are washed out by the averaging? Perhaps some way of looking at the statistical distribution of voxels' RFIs could be worth considering?

b) Are there some spatiotopic areas in the searchlight maps? It looks like there may be some blue clusters, but these cortical map figures are really hard to resolve.

(6) For the RFI as a function of model overlap and explained variance (Figure 4):

a) I like this analysis; I find it convincing and novel. Could it be further quantified by correlating on a voxelwise basis the reliability (e.g., explained variance) vs RFI?

b) I'm intrigued by the seemingly reliable blueish (spatiotopic) cells at the bottom of the V1-V3 grids. These seem to suggest that for the voxels with less spatial relevance (overlap), there might be something spatiotopic, even for relatively informative voxels (high explained variance)?

c) On a related note, is the "spatial relevance" measure the same as, or correlated with, eccentricity? It sounds like voxels with high spatial relevance (overlap with the central 4deg aperture) are the more foveal voxels. Intuitively, foveal voxels might be expected to be more retinotopic, right? In addition to clarifying this measure, it'd be nice to see a similar plot with eccentricity on the y-axis.

(7) For the Bayesian decoding (Figure 5):

a) A benefit of the Bayesian decoding (e.g., over the earlier studies using non-Bayesian decoding of retinotopic vs spatiotopic) is the uncertainty estimates. I think these analyses are interesting and should be in the main text figures, not a supplement.

b) Instead of line plots showing the decoded (best) position using the posterior distribution STD as the error shading, could you show the actual posterior distribution as heat maps (like the cartoon in B)? Is it possible there could be a second peak (or clear absence of one) at the spatiotopic prediction location?

(8) Also note that Golomb & Kanwisher also calculated the RFI measure for similar ROIs for both of their attention conditions. It may be worth comparing.

(9) Methods:

a) Is it true that 2 of the authors were actually naïve as to the purpose of the study? Regardless, given the small number of subjects and high ratio of authors as subjects, it might be nice to confirm that the results are not driven by the author-participants.

b) I think 44ms TR is a typo?

c) Why was the order of the bar movement directions always the same? Wouldn't this make the stimuli very predictable for the subjects, which could be potentially problematic?

d) I'm also curious why the gaze conditions were all presented in separate runs, as opposed to different blocks within a run.

e) The eccentricity maps for the fiducial maps (Figure 1G) seem a bit strange to me. Shouldn't the foveal representation be centered at the occipital pole, not the lateral surface?

eLife Assessment

This fundamental study reports convincing evidence for early verbal episodic memory formation. The findings demonstrate that speaker identity is a crucial feature, enabling episodic-like memories from birth, and will be of interest to cognitive neuroscientists working on brain development, memory, language learning and social cognition.

Reviewer #1 (Public review):

Summary:

This manuscript investigates whether newborns can use speaker identity to separate verbal memories, aiming to shed light on the earliest mechanisms of language learning and memory formation. The authors employ a well-designed experimental paradigm using functional near-infrared spectroscopy (fNIRS) to measure neural responses in newborns exposed to familiar and novel words, with careful counterbalancing and acoustic controls. Their main finding is that newborns show differential neural activation to novel versus familiar words, particularly when speaker identity changes, suggesting that even at birth, infants can use indexical cues to support memory.

Strengths:

Major strengths of the work include its innovative approach to a longstanding question in developmental science, the use of appropriate and state-of-the-art neuroimaging methods for this age group, and a thoughtful experimental design that attempts to control for order and acoustic confounds. The study addresses a significant gap in our understanding of how infants process and remember speech, and the data are presented transparently, with clear reporting of both significant and non-significant results.

A previous concern was that the recognition effect appeared restricted to a subgroup of participants. The authors clarify that the bilateral STG and left IFG effects were present in both groups - it was only the right IFG modulation that was group-dependent. This is an important distinction and is now clearer in the revised manuscript. The timing of the effect emerging in a specific testing window also appears less arbitrary given the authors' explanation that prior work guided the analytical approach, and that task difficulty was expected to determine whether recognition would appear in earlier or later test blocks.

The sample size question is handled honestly. A power analysis based on a related ANOVA study produced an implausibly small estimate of N=5-7, which the authors rightly set aside. Aligning with fNIRS neonate studies - where mean sample sizes around N=24 are standard - is defensible, and the within-subject design with mixed-model analysis does improve sensitivity relative to simpler approaches. This is now explained in the manuscript.

The episodic memory framing has been scaled back appropriately. The revised discussion is clear that the study demonstrates what-who binding - an early component of episodic-like processing - rather than mature episodic memory in the Tulvingian sense. This is a more honest characterization of what the paradigm can show, and it opens a reasonable developmental question about how the remaining components (where, when) come online over the first months and years of life.

Weaknesses

The weaknesses are largely interpretive rather than fatal to the core findings. The absence of a same-speaker interference control within the current paradigm means the causal role of speaker change cannot be established entirely from internal evidence alone - the inference relies partly on comparison with Benavides-Varela et al. (2011), which used a somewhat different design. This is a reasonable approach given the ethical and practical constraints of testing newborns, and the authors are transparent about it, but readers should keep in mind that the conclusion about speaker change as the critical variable is supported by converging evidence across studies rather than a direct within-study manipulation.

Overall, the study contributes new and meaningful data on an underexplored aspect of early speech processing: the role of the speaker as a contextual dimension in word memory. The findings, taken together with the prior literature, tell a coherent story and have real implications for theories of early language acquisition and the developmental origins of episodic-like memory. The paradigm is sound and the results are worth pursuing in larger and more controlled follow-up studies.

Reviewer #2 (Public review):

Summary

Previous studies by some of the same authors of the actual manuscript showed that healthy human newborns memorize recently learned nonsense words. They exposed neonates to a familiarization period (several minutes) when multiple repetitions of a bisyllabic word were presented, uttered by the same speaker. Then they exposed neonates to an "interference period" when newborns listened to music or the same speaker uttering a different pseudoword. Finally, neonates were exposed to a test period when infants hear the familiarized word again. Interestingly, when the interference was music, the recognition of the word remained. The word recognition of the word was measured by using the NIRS technique, which estimates the regional brain oxygenation at the scalp level. Specifically, the brain response to the word in the test was reduced, unveiling a familiarity effect, while an increase in regional brain oxygenation corresponds to the detection of a "new word" due to a novelty effect. In previous studies, music does not erase the memory traces for a word (familiarity effect), while a different word uttered by the same speaker does.

The current study aims at exploring whether and how word memory is interfered with by other speech properties, specifically the changes in the speaker, while young children can distinguish speakers by processing the speech. The author's main hypothesis anticipates that new speaker recognition would produce less interference in the familiarized word because somehow neonates "separate" the processing of both words (familiarized uttered by one speaker, and interfering word, uttered by a different speaker), memorizing both words as different auditory events.

From my point of view, this hypothesis is interesting since the results would contribute to estimate the role of the speaker in word learning and speech processing early in life.

Major strengths:

(1) New data from neonates. Exploring neonates' cognitive abilities is a big challenge, and we need more data to enrich the knowledge of the early steps of language acquisition.

(2) The study contributes new data showing the role of speaker (recognition) on word learning (word memory), a quite unexplored factor. The idea that neonates include speakers in speech processing is not new, but its role in word memory has not been evaluated before. The possible interpretation is that neonates integrate the process of the linguistic and communicative aspects of speech at this early age.

(3) The study proposes a quite novel analytic approach. The new mixed models allow exploring the brain response considering an unbalanced design. More than the loss of data, which is frequent in infants' studies, the familiarization, interference and learning processes may take place at different moments of the experiment (e.g. related to changes in behavioural states along the experiment) or expressed in different regions (e.g. related to individual variations in optodes' locations and brain anatomy).

Main weaknesses:

I did not find major weaknesses. However, I would like to have more discussion or explanation in the following points.

(1) It would be fine to report the contribution of each infant to the analysis, i.e. how many good blocks, 1 to 5 in sequence 1 and 2, were provided by each infant.

(2) Why did the factor "blocknumber" range from 0 to 4? The authors should explain what block zero means and why not 1 to 5.

(3) I may suggest intending to integrate the changes in brain activity across the 3 phases. That is, whether changes in familiarization relate to changes in the test and interference phases. For instance, in Figure 2, the brain response distinguishes between same and novel words that occurred over IFG and STG in both hemispheres. However, in the right STG there was no initial increase in the brain response, and the response for the same was higher than the one for novels in the 5th block.

(4) Similarly, it is quite amazing that the brain did not increase the activity with respect to the familiarization during the interference phase, mainly over the left hemisphere, even if both the word and speaker changed. Although the discussion considers these findings, an integrated discussion of the detection of novel words and the detection of a novel speaker over time may benefit from a greater integration of the results.

Appraisal

The authors achieved their aims, because the design and analytic approaches showed significant differences. The conclusions are based on these results. Specifically, the hypothesis that neonates would memorize words after interference, when interfered speech is pronounced by a different speaker was supported by the data, in block 2 and 5 and discussed the potential mechanisms underlying these findings, such as separate processing for different speakers, likely related to the recognition of speaker identity.

I think the discussion is well structured, although I may suggest integrating the changes into the three phases of the study. Maybe comparing with other regions, not related to speech processing.

Evaluating neonates is a challenge. Because physiology is constantly changing. For instance, in 9 minutes newborns may transit from different behavioral states and experience different physiological needs.

This study offers the opportunity to inspire looking for commonalities and individual differences when investigating early memory capacities of newborns.

Comments on revisions:

The authors provided satisfactory answers to my concerns.

I recognize that, because of technical and ethical reasons, the studies with neonates are particularly challenging, however, with a well-balanced design as the one the authors applied, even with small samples the data constitute valuable sources to advance in the field.

Neonate brain works in a particularly state of intense metabolic, functional and structural changes, which we are far to understand. Current data contribute to fill this gap in knowledge.

Author response:

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

Reviewer #1 (Public review)

Summary:

This manuscript investigates whether newborns can use speaker identity to separate verbal memories, aiming to shed light on the earliest mechanisms of language learning and memory formation. The authors employ a well-designed experimental paradigm using functional nearinfrared spectroscopy (fNIRS) to measure neural responses in newborns exposed to familiar and novel words, with careful counterbalancing and acoustic controls. Their main finding is that newborns show differential neural activation to novel versus familiar words, particularly when speaker identity changes, suggesting that even at birth, infants can use indexical cues to support memory.

Strengths:

Major strengths of the work include its innovative approach to a longstanding question in developmental science, the use of appropriate and state-of-the-art neuroimaging methods for this age group, and a thoughtful experimental design that attempts to control for order and acoustic confounds. The study addresses a significant gap in our understanding of how infants process and remember speech, and the data are presented transparently, with clear reporting of both significant and non-significant results.

Weaknesses:

However, there are notable weaknesses that limit the strength of the conclusions. The main recognition effect is restricted to a specific subgroup of participants and emerges only during a particular testing window, raising questions about the robustness and generalizability of the findings. The sample size, while typical for infant neuroimaging, is modest, and the statistical power is further reduced by missing data and group-dependent effects. Additionally, the claims regarding episodic memory and evolutionary implications are somewhat overstated, as the paradigm primarily demonstrates memory retention over a few minutes without evidence of the rich, contextually bound recall characteristic of fully developed episodic memory.

Overall, the authors have achieved their primary aim of demonstrating that speaker identity can facilitate memory separation in newborns, providing valuable preliminary evidence for early indexical processing in language learning. The results are intriguing and likely to stimulate further research, but the limitations in effect robustness and theoretical interpretation mean that the findings should be viewed as an important step forward rather than a definitive answer. The methods and data will be of interest to researchers studying infant cognition, memory, and language, and the study highlights both the promise and the challenges of probing complex cognitive processes in the earliest stages of life.

We thank the reviewer for their thoughtful and positive assessment of our work, and for giving us the opportunity to clarify points that may have been unclear in the original manuscript.

First, considering that the recognition response was quite consistent in previous studies, we expected the effect to emerge within a specific testing window, in either the first or the second block, depending on task difficulty. Accordingly, our analytical approach was designed to reflect this expectation, which was subsequently confirmed by the results. Second, the main recognition effect is not restricted to a specific subgroup of participants. Recognition responses were observed in both groups in the left IFG and bilateral STG. The only group-specific modulation was found in the right IFG, where the effect was primarily driven by Group A. This suggests that activity in this specific region may be influenced by contextual factors such as the nature and amount of recently processed stimuli. We have clarified these points in the revised manuscript to avoid the impression that the core effect is limited to a subset of participants or not generalizable across studies. 

Regarding the sample size, a formal calculation was initially attempted based on the effect size reported in a closely related ANOVA-based study (Benavides-Varela et al., 2011; Study 2: Word recognition after intervening melodies, main effect for the comparison same vs novel word [F(1,26) = 19.318; p<0.0001 effect size f =.87). However, inputting this information into a dedicated software (G*power; α = 0.05; number of groups =1; number of measurements = 2) leads to an estimated sample size of N = 5 to 7 (depending on the desired power, range = 0.800.95). This sample size is unrealistically small and not representative of current research standards in the field. A proper formal power analysis for the LMM is otherwise hard to perform, as we lack information about the expected variance and random-effects structure. We therefore aligned our sample size with prior newborn studies using similar stimuli and experimental designs, and with fNIRS studies in newborns and infants (for recent metanalysis see De Roever et al., 2018; Boek et al., 2023; Gemignani et al., 2023; which examined studies with mean N =24; N range= 186 and sample sizes often including various conditions and groups). Note also that our design includes a within-subject comparison, our analytical approach models subject-level variance and handles unbalanced datasets and missing data (which are common in infant studies), thereby improving statistical sensitivity. We have now explicitly clarified this choice in the Introduction.

Finally, we revised the discussion to ensure that interpretations are aligned with our findings, by including a limitations section and a more explicit note regarding theories of memory.

Episodic memory is a multifaceted construct that matures over time through the integration of the what–who-where–when information. The present study does not aim to demonstrate the presence of a fully developed episodic memory system at birth; rather, it shows that specific features of episodic-like processing (i.e., what–who) are already bound from the first days of life. Future studies may track the progressive integration of additional episodic-related components leading to a mature episodic memory system.

Reviewer #1 (Recommendations for the authors):

(1) I wonder why a control condition with same-speaker interference was not included. Adding such a control would allow you to directly test whether the observed effects are truly due to speaker changes, rather than other acoustic or procedural factors. If it is not feasible to add this condition, please discuss its absence explicitly and clarify how it impacts the interpretation of your findings.

We thank the reviewer for raising the issue of a same-speaker interference control. A similar control has been tested previously using a closely related paradigm, showing that recognition does not persist when neonates hear another word produced by the same speaker during the retention period (Benavides-Varela et al., 2011). As noted in the manuscript, there were some methodological differences between that study and the current one. Most importantly, in the present study familiarization was reduced (from ten to five blocks) and the retention interval increased (two to three minutes), making the current paradigm more demanding. We reasoned that, if newborns forgot the word under the prior (less challenging) study, they would also forget it here if a same-speaker interference control would have been implemented. With the current manipulation, despite the difficulty of the paradigm, the recognition response was observed. This pattern suggests that speaker change, rather than general procedural factors, is central to the observed effect. Given these prior findings and the ethical constraints of testing newborns, we believe that adding a new same-speaker control is not essential. We have now made this rationale more explicit in the manuscript (discussion section, limitations, p. 16), hoping that this clarification will make our methodological choices clearer.

(2) It wasn't clear if Group A and Group B have the same number of infants, and whether they were randomly assigned. Please specify.

Participants were initially assigned to Group A or Group B in a counterbalanced way to maintain comparable group sizes. Due to attrition and subsequent exclusion for various reasons (e.g., low signal quality, fussiness, technical issues), the final sample consisted of 17 infants in Group A and 15 infants in Group B. We have now specified this information in the revised manuscript (p. 20).

(3) Please specify the exact number of fNIRS channels assigned to each region of interest (ROI), as it is currently difficult to map the channel numbers in Supplementary Table 2 to the optode montage shown in Figure 2. Additionally, report the percentage of usable channels after quality control.

The inferior frontal gyrus left and right ROIs comprised 4 channels each, the superior temporal gyrus left and right ROIs 5 channels each, and the parietal lobe left and right ROIs 7 channels each. This information has been added to the methods section, along with the average number of channels contributing to each ROI after data rejection and the percentage of channels rejected throughout the recording (p. 23).

(4) Also, a formal power analysis to justify your sample size would be helpful for evaluating the reliability of your findings and is increasingly expected in developmental neuroimaging research.

Thanks for this suggestion. As stated in the public response, we agree that power analyses constitute an important component of methodological rigor in the field. In our case, a formal calculation was initially attempted based on the effect size reported in a closely related ANOVAbased study (Benavides-Varela et al., 2011; Study. 2: Word recognition after intervening melodies, main effect for the comparison same vs novel word [F(1,26) = 19.318; p<0.0001 effect size f =.87).

However, inputting this information into a dedicated software (G’power; α = 0.05; power range = 0.80-0.95; number of groups =1; number of measurements = 2) leads to an estimated sample size of N = 5 to 7, which is unrealistically small and not representative of current research standards in the field. A proper formal power analysis for the LMM is otherwise hard to perform, as we lack information about the expected variance and random-effects structure. We therefore aligned our sample size with prior newborn studies using similar stimuli and experimental designs, and with fNIRS studies in newborns and infants (for recent metanalysis see De Roever et al., 2018; Boek et al., 2023; Gemignani et al., 2023; which examined studies with mean N =24; N range= 1-86 and sample sizes often including various conditions and groups. Note also that our design includes a within-subject comparison, and our analytical approach models subject-level variance and handles unbalanced datasets and missing data (which are common in infant studies), thereby improving statistical sensitivity.

(5) The manuscript references episodic memory explicitly in the abstract and introduction, emphasizing the role of speaker identity in enabling episodic-like memory from birth. However, this concept is not sufficiently addressed or delineated in the discussion. Episodic memory is generally understood as recalling events with contextual details, involving complex integrative processes that extend beyond simple recognition of auditory stimuli. Your paradigm demonstrates memory retention over a few minutes but does not provide strong evidence for the hallmark features of episodic memory, such as contextual binding or autobiographical recollection. Moreover, infant speech recognition and memory formation in early life are influenced by the immediacy and complexity of sensory input, which may not necessarily engage fully developed episodic systems. Clarifying these distinctions and making sure your interpretations and claims are consistent with them would enhance the conceptual clarity of the manuscript.

We agree that episodic memory is a multifaceted construct that, in its mature form, entails the ability to retrieve past events with contextual detail, typically involving autobiographical recollection and the integration of what–-who-where–when information (Tulving, 1993). Our study does not aim to demonstrate the presence of a fully developed episodic memory system at birth, nor do we claim that newborns’ performance satisfies all hallmark criteria of mature episodic memory. 

Here, we focused on sensitivity to speaker identity as a contextual dimension relevant to memory formation. Within this narrower sense, both, the patterns of activation and the localization of the response provide evidence for early source–content binding (i.e., what–who), which can be considered a foundational aspect of episodic-like processing. Following up on this foundational step, future studies may track the gradual integration of additional aspects (where-when), ultimately leading to the maturation of a fully functional human episodic memory system.

We have now clarified this point in the revised manuscript (p. 17)

(6) Please add a dedicated limitations section. This should address the group-dependent nature of your main effects, the timing-specific recognition response, and any other methodological constraints that may impact the generalizability of your results.

We thank the reviewer for this comment. We have made our best to expose the limitations of our study in the text (p.16), specifically regarding the reasons for the lack of a control condition and the effects of frequent changes in sleeping states in newborns. 

(7) Consider revising sections where claims may be overstated, particularly regarding episodic memory and evolutionary implications.

These sections have now been revised in the abstract and throughout the manuscript to ensure that interpretations remain proportionate to the data and consistent with current theoretical frameworks.

Reviewer #2 (Public review):

Summary:

Previous studies by some of the same authors of the actual manuscript showed that healthy human newborns memorize recently learned nonsense words. They exposed neonates to a familiarization period (several minutes) when multiple repetitions of a bisyllabic word were presented, uttered by the same speaker. Then they exposed neonates to an "interference period" when newborns listened to music or the same speaker uttering a different pseudoword. Finally, neonates were exposed to a test period when infants hear the familiarized word again. Interestingly, when the interference was music, the recognition of the word remained. The word recognition of the word was measured by using the NIRS technique, which estimates the regional brain oxygenation at the scalp level. Specifically, the brain response to the word in the test was reduced, unveiling a familiarity effect, while an increase in regional brain oxygenation corresponds to the detection of a "new word" due to a novelty effect. In previous studies, music does not erase the memory traces for a word (familiarity effect), while a different word uttered by the same speaker does.

The current study aims at exploring whether and how word memory is interfered with by other speech properties, specifically the changes in the speaker, while young children can distinguish speakers by processing the speech. The author's main hypothesis anticipates that new speaker recognition would produce less interference in the familiarized word because somehow neonates "separate" the processing of both words (familiarized uttered by one speaker, and interfering word, uttered by a different speaker), memorizing both words as different auditory events.

From my point of view, this hypothesis is interesting, since the results would contribute to estimating the role of the speaker in word learning and speech processing early in life.

Strengths:

(1) New data from neonates. Exploring neonates' cognitive abilities is a big challenge, and we need more data to enrich the knowledge of the early steps of language acquisition.

(2) The study contributes new data showing the role of speaker (recognition) on word learning (word memory), a quite unexplored factor. The idea that neonates include speakers in speech processing is not new, but its role in word memory has not been evaluated before. The possible interpretation is that neonates integrate the process of the linguistic and communicative aspects of speech at this early age.

(3) The study proposes a quite novel analytic approach. The new mixed models allow exploring the brain response considering an unbalanced design. More than the loss of data, which is frequent in infants' studies, the familiarization, interference and learning processes may take place at different moments of the experiment (e.g. related to changes in behavioural states along the experiment) or expressed in different regions (e.g. related to individual variations in optodes' locations and brain anatomy).

Weaknesses:

I did not find major weaknesses. However, I would like to have more discussion or explanation on the following points.

(1) It would be fine to report the contribution of each infant to the analysis, i.e. how many good blocks, 1 to 5 in sequence 1 and 2, were provided by each infant.

(2) Why did the factor "blocknumber" range from 0 to 4? The authors should explain what block zero means and why not 1 to 5.

(3) I may suggest intending to integrate the changes in brain activity across the 3 phases. That is, whether changes in familiarization relate to changes in the test and interference phases. For instance, in Figure 2, the brain response distinguishes between same and novel words that occurred over IFG and STG in both hemispheres. However, in the right STG there was no initial increase in the brain response, and the response for the same was higher than the one for novels in the 5th block.

(4) Similarly, it is quite amazing that the brain did not increase the activity with respect to the familiarization during the interference phase, mainly over the left hemisphere, even if both the word and speaker changed. Although the discussion considers these findings, an integrated discussion of the detection of novel words and the detection of a novel speaker over time may benefit from a greater integration of the results.

Appraisal:

The authors achieved their aims because the design and analytic approaches showed significant differences. The conclusions are based on these results. Specifically, the hypothesis that neonates would memorize words after interference, when interfered speech is pronounced by a different speaker, was supported by the data in blocks 2 and 5, and the potential mechanisms underlying these findings were discussed, such as separate processing for different speakers, likely related to the recognition of speaker identity.

I think the discussion is well-structured, although I may suggest integrating the changes into the three phases of the study. Maybe comparing with other regions, not related to speech processing.

Evaluating neonates is a challenge. Because physiology is constantly changing. For instance, in 9 minutes, newborns may transit from different behavioral states and experience different physiological needs.

We thank the reviewer for their constructive and positive appraisal of our work and for drawing attention to points that benefited from further clarification or discussion in the manuscript.

In the following, we address each point in turn, using the numbering of the reviewer’s identified concerns.

(1) In the Methods section (“Data Processing and Analysis”, p. 22), we have added detailed information about the number of data points contributed by each infant to the analyses.

(2) The factor “blocknumber” ranged from 0 to 4 for statistical purposes, allowing Block 0 to serve as the reference (intercept) in the model. This coding facilitated the interpretation of parameter estimates. We now clarify this in the revised manuscript (p. 7).

(3) Thanks for this relevant suggestion. In the Discussion, we now explicitly discuss the relationship across phases. We also acknowledged that a thorough examination of these issues lies beyond the scope of the present study as it will require future work based on multivariate and connectivity analyses.

(4) We thank the reviewer for this comment. In the revised manuscript, we have expanded the Discussion to clarify the absence of a strong novelty response during interference. The discussion highlights how the temporal properties of the hemodynamic response and the functional demands of each phase jointly shape the observable fNIRS signal in newborns, with purely sensory novelty effects likely increasing with maturation.

Finally, we agree that evaluating the transitions of sleeping states can further strengthen and clarify the results obtained in the present study. This has now been added as one of the limitations of this study.

eLife Assessment

This valuable study presents a framework for a shareable data analysis pipeline aimed at improving reproducibility in neuroscience. The evidence for robustness and inter-laboratory operability is convincing. Overall, this work will be of interest to neuroscientists engaged in the analysis of large-scale neuronal recordings.

Reviewer #1 (Public review):

Summary

The manuscript by K.H. Lee et al. presents Spyglass, a new open-source framework for building reproducible pipelines in systems neuroscience. The framework integrates the NWB (Neurodata Without Borders) data standard with the DataJoint relational database system to organize and manage analysis workflows. It enables the construction of complete pipelines, from raw data acquisition to final figures. The authors demonstrate their capabilities through examples, including spike sorting, LFP filtering, and sharp-wave ripple (SWR) detection. Additionally, the framework supports interactive visualizations via integration with Figurl, a platform for sharing neuroscience figures online.

Strengths:

Reproducibility in data analysis remains a significant challenge within the neuroscience community, posing a barrier to scientific progress. While many journals now require authors to share their data and code upon publication, this alone does not ensure that the code will execute properly or reproduce the original results. Recognizing this gap, the authors aim to address the community's need for a robust tool to build reproducible pipelines in systems neuroscience.

Comments on revisions:

In this revised version, the authors have addressed the majority of the concerns raised in the initial review. The manuscript is clearer, the documentation and explanations have been strengthened, and several important practical issues-particularly regarding usability, terminology, and deployment-have been meaningfully improved. While the framework continues to position itself both as a flexible analysis environment and as a mechanism for freezing and preserving reproducible pipelines, the authors have clarified their rationale for maintaining this dual role. I have no additional comments at this stage.

Reviewer #2 (Public review):

Summary:

Lee et al. introduce Spyglass, an open-source Python framework designed to tackle the reproducibility crisis in systems neuroscience by integrating the Neurodata Without Borders (NWB) standard with DataJoint relational databases. The framework aims to standardize data ingestion, preprocessing, analysis pipelines, and data sharing for complex electrophysiological and behavioral experiments.

Strengths:

(1) Handling of Complex Workflows: The architectural design is pragmatic and robust. Features such as the "cyclic iteration" motif for spike-sorting curation and the "merge" motif for consolidating multiple data streams effectively handle the iterative nature of data processing without incurring database bloat.

(2) Ecosystem Integration: The revised manuscript clarifies that Spyglass acts as a community hub, explicitly detailing its integration with established tools like SpikeInterface, DeepLabCut, GhostiPy, MoSeq, and Pynapple.

(3) Pipeline Clarity & Practical Demonstration: The addition of Supplementary Figure 1, in conjunction with Figure 5, successfully maps out the complex, multi-step decoding workflow for both the UCSF and NYU datasets. Together, these figures tell a complete and compelling story of how this pipeline can be used in practice, providing much-needed visual clarity on how raw data moves through the database to generate final results.

Appraisal:

The authors have successfully achieved their aims. Spyglass is a highly functional system capable of handling the heavy lifting of data management. The revisions have significantly improved transparency regarding the tool's limitations and its onboarding process, making it a highly attractive blueprint for labs aiming to adhere to FAIR principles.

Author response:

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

Reviewer #1 (Public review):

Summary

The manuscript by K.H. Lee et al. presents Spyglass, a new open-source framework for building reproducible pipelines in systems neuroscience. The framework integrates the NWB (Neurodata Without Borders) data standard with the DataJoint relational database system to organize and manage analysis workflows. It enables the construction of complete pipelines, from raw data acquisition to final figures. The authors demonstrate their capabilities through examples, including spike sorting, LFP filtering, and sharpwave ripple (SWR) detection. Additionally, the framework supports interactive visualizations via integration with Figurl, a platform for sharing neuroscience figures online.

Strengths:

Reproducibility in data analysis remains a significant challenge within the neuroscience community, posing a barrier to scientific progress. While many journals now require authors to share their data and code upon publication, this alone does not ensure that the code will execute properly or reproduce the original results. Recognizing this gap, the authors aim to address the community's need for a robust tool to build reproducible pipelines in systems neuroscience.

We appreciate the summary and the recognition of the key need for maximally reproducible scientific workflows.

Weaknesses:

The issues identified here may serve as a foundation for future development efforts.

(1) User-friendliness:

The primary concern is usability. The manuscript does not clearly define the intended user base within a modern systems neuroscience lab. Improving user experience and lowering the barrier to entry would significantly enhance the framework's potential for broad adoption. The authors provide an online example notebook and a local setup notebook. However, the local setup process is overly complex, with many restrictive steps that could discourage new users. A more streamlined and clearly documented onboarding process is essential. Additionally, the lack of Windows support represents a practical limitation, particularly if the goal is widespread adoption across diverse research environments.

We agree that usability is critical, and we now clarify that Spyglass

“… is designed to be used by everyone in a laboratory who works with the data, both as a general-purpose tool to enable the development of new analysis pipelines and a tool that allows those pipelines and associated results to be frozen and packaged to enable reproducibility…”

To address the local setup issue, we have now created an interactive quick start program to guide new users through the setup (scripts/install.py). It now leads the user through a few prompts with sensible defaults to reduce the complexity of the setup. It aids the user in installing the Spyglass dependencies and creating the Data joint configuration file. We also validate the configuration to make sure the set up was successful (scripts/validate.py). Combined, these should reduce the complexity and set up time for most users while allowing expert users to configure Spyglass as they need. We thank the reviewer for the suggestion.

We also agree that the lack of support for Windows is a key issue, and that is something we plan to address in the coming years. We note that it may be possible to run Spyglass under the Windows Subsystem for Linux (WSL 2), which allows users to run Linux programs on a Windows machine without the need for a virtual machine or dual boot setup.

(2) Dependency management and long-term sustainability:

The framework depends on numerous external libraries and tools for data processing. This raises concerns about long-term maintainability, especially given the short lifespan of many academic software projects and the instability often associated with Python's backward compatibility. It would be helpful for the authors to clarify how flexible and modular the pipeline is, and whether it can remain functional if upstream dependencies become deprecated or change substantially.

This is a very good point that reflects a broad challenge to maintainability and reproducibility. We now explicitly raise this point in our Limitations section, and note that

“…even in cases where reproducing a result would require installing older versions of software, the results themselves remain accessible within NWB files referenced in Spyglass, ensuring that previous results can be built on even as packages evolve.”

The merge table pattern also allows us to update (version) our pipelines as software changes. For example, we have already done so for changes in SpikeInterface versions for the version 1 pipeline for spike sorting. New and older versions of the pipeline (v0 and v1) are accessed through the merge table SpikeSortingOutput. This allows the user to have consistent results despite the version change.

(3) Extensibility for custom pipelines:

A further limitation is the insufficient documentation regarding the creation of custom pipelines. It is unclear how a user could adapt Spyglass to implement their own analysis workflows, especially if these differ from the provided examples (e.g., spike sorting, LFP analysis that are very specific to the hippocampal field). A clearer explanation or example of how to extend the framework for unrelated or novel analyses would greatly improve its utility and encourage community contributions.

Here we failed to provide the required links to the documentation. We now explicitly refer to documentation on Custom Pipeline, which include a link to a YouTube video walking users through the creation of such a pipeline:

Specifically, Spyglass uses DataJoint syntax to define tables as Python classes (see online documentation on Custom Pipelines and this video for examples).

(4) Flexibility vs. Standardization:

The authors may benefit from more explicitly defining the intended role of the framework: is Spyglass designed as a flexible, general-purpose tool for developing custom data analysis pipelines, or is its primary goal to provide a standardized framework for freezing and preserving pipelines post-publication to ensure reproducibility? While both goals are valuable, attempting to fully support both may introduce unnecessary complexity and result in a tool that is not well-suited for either purpose. The manuscript briefly touches on this tradeoff in the introduction, and the latter-pipeline preservation-may be the more natural fit for the package. If so, this intended use should be clearly communicated in the documentation to help users understand its scope and strengths.

We appreciate this point, and have now clarified in the beginning of the Results section that

It is both a general-purpose tool to enable the development of new analysis pipelines and a tool that allows those pipelines and associated results to be frozen and packaged to enable reproducibility.

In practice, our lab uses Spyglass to systematize analyses to enable rapid application across many datasets. Then, once a paper has been finalized, we can export the data and the code in a package that enables reproduction. Being able to do both things is, in our view, a key strength of Spyglass. More broadly, we feel it is critical that there be a clear path for users to take their analysis code and make it reproducible. That process normally involves a very substantial amount of work, and our goal was to reduce the burden on users and make this a straightforward extension of how analyses are carried out.

Impact:

This work represents a significant milestone in advancing reproducible data analysis pipelines in neuroscience. Beyond reproducibility, the integration of cloud-based execution and shareable, interactive figures has the potential to transform how scientific collaboration and data dissemination are conducted. The authors are at the forefront of this shift, contributing valuable tools that push the field toward more transparent and accessible research practices.

We appreciate this positive assessment.

Reviewer #1 (Recommendations for the authors):

(1) "The authors write: ‘the relational database, a well-known data structure that uses tables to organize data.’ This phrasing may be misleading… It would be more accurate to describe them as ‘well-established’ rather than ‘well-known.’"

We have made this change.

(2) The statement "It makes it easy to apply the same analysis to multiple datasets, as users need to specify only the data and parameters for computation ("what") rather than the execution details ("how")." would benefit from further elaboration. Specifically, how does this approach compare in practice to using a simple configuration file (e.g., YAML or JSON) to manage parameters and execution logic? A comparison or example would help ground the claim.?"

We agree one could in principle do something similar with configuration files, but this is a discipline that the user must impose on themselves, as configuration files in general have no constraint on how they are to be used. On the other hand, a system like Spyglass enforces the separation of data from parameters by design. We have now added a brief comment on this point in the Results:

“It provides a structure to organize and systematize the analysis parameters, data, and outputs into different tables. This contrasts with user-generated configuration files where each user could adopt their own idiosyncratic approach to specifying parameters and data.”

We also come back to this point in the Discussion:

Other approaches do away with the relational database altogether. For example, DataLad uses version control tools such as git and git-annex to manage both code and data as files [39]. This enables the creation of a data analysis environment and decentralized data sharing. For building analysis pipelines, it may be combined with other tools for managing the sequential execution of scripts. For example, Snakemakeb[40] (and related projects such as Cobrawap [41]) allows the users to gather and define the input, output, and the associated scripts to execute for each analysis step, thereby tracking the dependency between steps. But because these tools do not provide any formal structure for data analysis or parameter specification, they lack the advantages of the relational database that we discussed, such as being able to easily organize or search for the records of previous analysis based on specific parameters, efficient data sharing and access management to multiple users, and built-in data integrity checks based on constraints native to the database (e.g. primary keys).

(3) The sentence ‘It enables easy access to multiple datasets via queries’ may overstate the benefit… clarify what specific advantages database queries offer.

We agree that this is an important feature and we added the following as an example of the advantage of being able to query the database:

It enables easy access to multiple datasets via queries (e.g. to find all datasets with recordings from a particular brain region or that used a particular behavioral paradigm)

(4) Specifically, Spyglass uses DataJoint syntax to define tables as Python classes’ lacks clarity… Expanding this explanation with a brief, concrete example would

We agree that this sentence does not provide information on how to use DataJoint syntax to define a table. We carefully considered adding that syntax to the manuscript, but we are concerned that doing so here and in other places where syntax examples could be used would decrease the readability of the document. We also noted that other papers that present analysis frameworks typically provide much less information.

Nevertheless, it is clear that users would benefit from a concrete example, and as we mentioned above, we have added a link to the documentation describing how to make custom schema and pipelines, as well as a YouTube video that we created to walk users through this process.

(5) The authors write: "Selection tables associate parameter entries with data object entries." This terminology is confusing. From a naming perspective, it is not immediately obvious what a "selection table" is or how it differs from other components. Moreover, shouldn't parameter entries be associated with a specific pipeline rather than directly with data objects? Further clarification is needed. "

We appreciate that our terminology was not clear. The idea behind a selection table is that there are many data entries and many potential sets of parameters that can be used to analyze each of those entries. We have now revised this section of the text and added an explanatory paragraph:

An analysis pipeline consists of sets of tables downstream of the Common tables. In each step in the analysis, the user populates one of four table types (Figure 2A):

Data tables contain pointers to data objects in either the original NWB file or ones generated by an upstream analysis.

Parameter tables contain a list of the parameters needed to fully specify the desired analysis.

Selection tables allow users to select and pair a data entry and a parameter entry, defining the input to the Compute table.

Compute tables execute the computations to carry out the analysis using the Data and Parameters specified in the Selection table entry. These results are then stored and can serve as Data for downstream analysis.

This design has multiple features that we have found to be beneficial. First, Parameter tables store the full set of parameters needed to specify a given analysis. For example, a Parameter table entry for a firing rate analysis of a single neuron might specify the bin size and smoothing to be used for that analysis. Multiple such entries can be defined, allowing a user to select the most appropriate one for the question being addressed. Second, because Selection tables specify which Parameter table entry was used for a given analysis on the associated Data table entry, they provide the key information needed to know which parameters were used to generate the entry in the downstream Compute table. Third, it is simple to associate a given Data table entry with multiple Parameter table entries and then re-run the analysis on those pairs. This enables a user to understand how their choice of parameters impacts their results, something that is otherwise difficult to manage and track.

(6) Including ‘fitting state-space models’ as a standard example may be misleading… Presenting it as a routine task might set unrealistic expectations."

We agree and have changed “standard” to “a diverse range of”.

(7) Figure 2 would benefit from clearer sequential logic. For example, the object ‘LFPSelection’ appears after a method call referencing it."

We agree that the figure was not explained adequately. We now make it clear in the caption that the method call creates the entry in the LFPSelection table, and is thus upstream of the picture of the table entry that was created.

(8) Example 3 would be strengthened by a comparison to SpikeInterface, a framework increasingly adopted by the community."

Here we clearly did not explain the spike sorting pipeline sufficiently thoroughly. As we now clarify in the text:

This pipeline uses SpikeInterface [19] to perform the operations critical for spike sorting, but also tracks all of the parameters used and provides a system for tracking multiple sorting curations.

Thus, Spyglass takes advantage of the special purpose routines within SpikeInterface, but also provides an organizational framework for the outputs, and, equally critically, allows direct use of the outputs of sorting in downstream analyses with the ability to go back and know which sorting parameters were used for that analysis.

(9) The authors state: "These are saved as Docker containers and optionally uploaded to DANDI." However, it is unclear how end users are expected to interact with these containers. Additional guidance or an example interaction would be valuable.

We agree that this interaction was not described in the text, and we have now added the following to explain how a user might interact with these containers:

...This can be done by (i) hosting the database on the cloud and granting access to users outside the lab; or (ii) exporting and sharing parts of the database that were used by the project. Spyglass facilitates the second option by providing functions that automatically log the table entries and NWB files used for creating figures of a manuscript in a Python environment (Table 1, 05_Export). The dependencies of these entries are traced through the database to compile the complete set of raw, intermediate, and plotted NWB files and their corresponding database entries. These are stored in the `Export` table, which also generates a bash script to create SQL dumps of the identified database entries.

To upload these files to DANDI, users must first register a new dandiset for their project and record their API and dandiset ID. With this information, they can then use the method `DandiPath.compile_dandiset()` to automatically validate, organize, and upload all project files to the DANDI archive. Additionally, this process stores the archive information for each file in the `DandiPath` table, allowing `fetch_nwb` to automatically stream data from the DANDI cloud storage when not available locally.

To create a sharable docker image of the project, we provide a template repository spyglass-export-docker. Users first download a local copy of this repo and copy the SQL dump file, environment yaml, and figure-generating notebooks generated during spyglass export into the appropriate folders. Running the provided docker compose scripts then generates two linked docker containers: one running the reconstructed spyglass SQL database, and a second connected to this database and running a jupyter hub with a python environment matching that used when generating the figures. These can be readily shared with new users to provide them immediate access to all steps of the analysis process and the corresponding data through DANDI streaming

(10) The phrase "not requiring a central location to track available files and providing a user-friendly Python API" is somewhat vague. Does this imply that multiple sources can exist for the same NWB file? How does the system handle potential version conflicts, such as when an NWB file is modified locally? A clearer explanation would help users understand the system's behavior in collaborative scenarios. "

This is an important point that we now explain in the manuscript:

Critically, the downloaded files are never modified locally within Spyglass and attempt to access a modified file would result in a DataJoint error. This ensures that each user is working on the same underlying data even if they are at different sites.

To provide interested readers with more details, we also now point them to the repo for more information:

We point interested readers to the Kachery GitHub repo (https://github.com/magland/kachery) for further descriptions.

(11) "The concept of a ‘kachery zone’ in Figure 4 is ambiguous. Is this storage local or in the cloud? If a third-party storage system is involved, it should be explicitly labeled and described in the diagram."

We agree that the depiction of a Kachery zone in Figure 4 is hard to understand. For the reviewer’s reference, a Kachery zone defines a list of users that have permissions to upload and download a particular set of files that have been linked to that zone. This is a explained in the tutorials, and to simplify the figure we have replaced the Kachery zone with a remote computer.

(12) If one of the manuscript's goals is to showcase the functionality of the pipeline, Figure 5 would be more informative if it also illustrated the workflow or steps involved in generating the displayed figures.

We have added a supplementary figure (Supplementary Figure 1) related to figure 5 that illustrates the main data workflow used in generating the figure. In addition, we note that the code for generating the figure 5 and supplemental are included in the code repository for the paper (https://github.com/LorenFrankLab/spyglass-paper/).

(13) In the conclusion, the authors write: "By contrast, Spyglass begins with a shared data format that includes the raw data and offers both transparent data management and reproducible analysis pipelines using a formal data structure." However, the tools discussed in the previous paragraph seem to offer similar capabilities. The real challenge in transparent data management often lies in the technical overhead associated with setting up and maintaining a database, particularly when collaborating across labs.

Here we may not have explained the differences between Spyglass and these other approaches sufficiently clearly. The various tools mentioned in the paragraph above this one do not begin with a shared format nor do they include a formal data structure. That said, we agree that maintaining a database accessible across labs is a key challenge. We note here that we provide tutorials to ease this process, which are linked and described in the manuscript (e.g. Table 1).

(14) Specifying a preferred IDE… may not be necessary. This recommendation could be made optional or omitted."

We agree that it may not be necessary, but we have also noted that users come to Spyglass with a very wide range of expertise, and in our lab it has been helpful to specify the IDE.

Reviewer #2 (Public review):

Summary:

This valuable paper presents Spyglass, a comprehensive software framework designed to address the critical challenges of reproducibility and data sharing in neuroscience.

The authors have developed a robust ecosystem built on community standards such as NWB and DataJoint, and demonstrate its utility by applying it to datasets from two independent labs, successfully validating the framework's ability to reproduce and extend published findings. While the framework offers a powerful blueprint for modern, reproducible research, its immediate broad impact may be tempered by the significant upfront investment required for adoption and its current focus on electrophysiological data. Nevertheless, Spyglass stands as an important and practical contribution, providing a well-documented and thoughtfully designed path toward more transparent and collaborative science.

Strengths:

(1) Principled solution to a foundational challenge:

The work offers a concrete and comprehensive framework for reproducibility in neuroscience, moving beyond abstract principles to provide an implemented, end-to-end ecosystem.

(2) Pragmatic and robust architectural design:

Features such as the "cyclic iteration" motif for spike-sorting curation and the "merge" motif for pipeline consolidation demonstrate deep, practical experience with neurophysiological analysis and address real-world challenges.

(3) Cross-laboratory validation:

The successful replication and extension of published hippocampal decoding findings across independent datasets strongly support the framework's utility and underscore its potential for enabling reproducible science.

(4) Accessibility through documentation and demos:

Extensive tutorials and the availability of a public demo environment lower some of the barriers to adoption.

We appreciate the Reviewer’s recognition of these strengths.

Weaknesses:

(1) High barrier to adoption:

The requirement to convert all data into NWB, maintain a relational database, and train users in structured workflows is a significant hurdle, particularly for smaller labs.

We agree that this is a significant hurdle, but we also believe that it comes with many advantages. It is also increasingly easy to do given the many community-supported tools, regardless of how much resource the lab has. These points are discussed in detail in “Why NWB?” section.

We also note that, to our knowledge, there is no simpler alternative that provides the key features of Spyglass.

(2) Limited tool integration:

The current pipelines, while useful, still resemble proof-of-principle demonstrations.

Closer integration with established analysis libraries such as Pynapple and others could broaden the toolkit and reduce duplication of effort.

Here we clearly failed to explain that we have integrated other libraries, including Pynapple. We now make this clear in the Results section:

Our goal was take advantage of other open source packages, and we have therefore integrated support for Pynapple [21], a general purpose neural data analysis package. We also built our pipelines to take advantage of other community-developed, open-source packages, like GhostiPy [20], SpikeInterface [19], DeepLabCut [2] and Moseq [29].

We also have added a specific reference to the relevant function call in the Practical use cases and extensions section:

For example, the user can conveniently read specific data types from the NWB file by first ingesting it into Spyglass and accessing database tables with Spyglass functions (e.g. fetch_nwb) or even load those objects in a format compatible with Pynapple [21] (fetch_pynapple).

Pynapple support is actually aided by our design choice of relying on NWB. Because NWB files can be loaded by Pynapple, any analysis that uses a NWB file that can be read by Pynapple can be loaded as a Pynapple object. We have provided methods to do so.

(3) Experimental metadata support:

While NWB provides a solid foundation for storing neurophysiology data streams, it still lacks broad and standardized support for experimental metadata, including descriptions of conditions, subject details, and procedures, as well as links across datasets. This limitation constrains one of Spyglass's key promises: enabling reproducible, crosslaboratory science. The authors should clarify how Spyglass plans to address or mitigate this gap - for example, by adopting or contributing to metadata extensions, providing templates for experimental conditions, or integrating with complementary systems that manage metadata across datasets.

This is an important point. First, NWB provides methods for creating new metadata extensions, and our laboratory has contributed to multiple such extensions and have adopted metadata extensions as they come to exist (for example, we are currently integrating the ndx-pose extension, which has broader support for pose estimation algorithms such as DLC and SLEAP, enabling us to capture relationships between body parts). These extensions, once incorporated into NWB, make it easy to create parallel Spyglass tables that read in the associated metadata. Second, we note that by storing the metadata from the NWB file in a database, Spyglass naturally supports searches across datasets where the metadata is the same (e.g. all the datasets from a given subject or using a given behavioral apparatus).

That said, for these searches to be easy, the underlying NWB files need to use the same ontologies (naming systems). Creating shared naming systems within and across labs is very challenging, but even here having a database helps greatly, as it provides a way to find all the names used for a given field and to thereby make an effort to standardize them.

Finally, while Spyglass aims to enable reproducibility, it will not be possible to solve all standardization issues of the field. We believe that Spyglass is an important step forward in standardization and reproducibility in that it encourages users to use the same data format and processing. To our knowledge, there is no software like it in the field of systems neuroscience. Limitations of the field and of current progress does not invalidate the contribution of Spyglass as a framework.

We now mention all these issues in the Limitations section of the Discussion.

(4) Cross-laboratory interoperability:

While demonstrated across two datasets, the manuscript does not fully address how Spyglass will handle the diversity of metadata standards, acquisition systems, and labspecific practices that remain major obstacles to reproducibility.

We agree that the current version of Spyglass does not fully address this diversity. Neverless, we note that the NWB standard is increasingly widely adopted in our field, and that by building on this standard, it is much similar to create structures that store relevant data across labs.

(5) Visualization limitations:

Beyond the export system and Figurl, NWB offers relatively few options for interactive data exploration. The ability to explore data flexibly and discover new phenomena remains limited, which constrains one of the potential strengths of standardized pipelines.

We agree that there are many other tools, and we have considered additional integrations. We have chosen not to proceed in this direction because the various visualization tools are well constructed, and therefore already easy to use with data retrieved from Spyglass. Thus, users can choose to use Matplotlib, Seaborn, or any of many other visualization tools and apply thos to data accessed through Spyglass without the need for more explicit integration.

Spyglass is well-positioned to become a community framework for reproducible neuroscience workflows, with the potential to set new standards for transparency and data sharing. With expanded modality coverage, tighter integration of existing community tools, stronger solutions for cross-lab interoperability, and richer visualization capabilities, it could have a transformative impact on the field.

We appreciate this summary and will continue to try to make Spyglass more powerful, generalizable, and accessible to the community.

Reviewer #2 (Recommendations for the authors):

(1) Documentation/User onboarding:

While extensive documentation exists, new users may feel overwhelmed. A single Quickstart or "golden path" guide and a one-command validation script would substantially improve usability.

As mentioned in the response to reviewer 1, we have added an interactive quickstart program to walk users through installation and setup (scripts/install.py) and validate the install (scripts/validate.py). This should greatly reduce the complexity of the set-up process and allow new users to use Spyglass quickly and confidently. We thank the reviewer for the suggestion.

(2) Permission handling and multi-user scaling:

Current ad hoc solutions (like cautious deletes) may not scale well in large collaborations. This should be acknowledged, but it is not a fatal weakness given the framework's early stage.

This is a fair point and we now mention this when cautious delete is introduced in the Methods:

Though this is not a formal permission-management system, it serves to prevent accidental deletions. We note that this system does incur additional overhead, and while that has not been an issue for us, it is possible that this would become problematic in use for much larger cross-laboratory collaborations.

(3) Benchmarking and performance evaluation:

"More systematic testing (e.g., reproducibility across independent users, computational efficiency) would be reassuring, but the lack of it does not invalidate the proof-of-principle demonstration. "

We agree. So far at least two other labs have adopted this system and we are working with a consortium funded by the Simons Foundation to use Spyglass as a data sharing system across a larger number of labs.

(4) Support for Cloud solution:

To lower the barrier to adoption, the authors should consider cloud integration, such as preconfigured Docker/Cloud templates or hosted options, so end-users do not need to maintain databases and storage locally.

We agree that cloud-based solutions could be a good option for some labs, although we note that the cost of cloud-based computing can be very high. There is also the burden of moving and storing the data to where it needs to be processed, which can be particularly time intensive with the large-scale data being generated by many laboratories.

At the reviewer’s suggestion, we have added a docker-compose support to lower the barrier to adoption. This includes:

docker-compose.yml with health checks and persistent storage

.env.example configuration template

This allows one-command database setup: `docker compose up –d`

(5) Integration of greater modalities:

The authors should consider expanding support to other major data types, particularly calcium imaging, photometry, and other optical physiology data.

We entirely agree that pipelines to ingest and process these datatypes would be very valuable, and we would welcome collaborations with experts and the general community to build these pipelines. We are, for example, working with a collaborating lab on a photometry pipeline. However, we only have so many people to build and maintain Spyglass, so we are limited by the capacity and expertise of our developers.

(6) Integrate more community tools:

Closer integration with community tools such as Pynapple, Neurosift, and SpikeInterface would broaden functionality and position Spyglass as a hub rather than a parallel ecosystem.

As we mentioned in our responses to Reviewer 1, we entirely agree, and in fact we have already integrated Pynapple support into Spyglass. Because we store files in the NWB format and Pynapple supports NWB, it was easy for us to convert any data we have into the Pynapple format upon request, thus making it easily analyzable by the Pynapple package. Moreover, we use SpikeInterface for the SpikeSorting pipline, and similarly provide pipelines built on other open source projects. As we now clarify in the text:

Spyglass includes pipelines for a diverse range of analysis tasks in systems neuroscience, such as the analysis of LFP, spike sorting, video and position processing, and fitting state-space models for decoding neural data. Tutorials for all pipelines are available on the Spyglass documentation website (Table 1). Our goal was take advantage of other open source packages, and we have therefore integrated support for Pynapple [21], a general purpose neural data analysis package. We also built our pipelines to take advantage of other community-developed, open-source packages, like GhostiPy [20], SpikeInterface [19], DeepLabCut [2] and Moseq [29].

(7) Direct Dandi archive upload functionality:

Scripts and tutorials for uploading data directly from Spyglass to DANDI, with validation of metadata completeness, would provide users with a direct pipeline from raw data to a public archive.

The tutorials for DANDI upload are included as part of the export tutorial notebook (https://lorenfranklab.github.io/spyglass/latest/notebooks/05_Export/). We agree that this was not apparent from the manuscript before and have noted this within the Manuscript table describing these notebooks.

eLife Assessment

This study provides fundamental insight by identifying C. elegans SET-19 as a key enzyme that deposits H3K23me to somatic chromatin. The evidence is compelling, using a broad and modern toolkit of biochemical, genetic, and genome-wide analyses that consistently support the main claims. The significance of the study is further strengthened by the fact that H3K23me is an understudied histone modification, which is also conserved in mammals.

Reviewer #1 (Public review):

Summary:

The authors wanted to determine whether the set-19 gene, one of 38 SET-domain containing genes in C elegans, has a clear function in vivo with respect to lysine methylation. The question is not only whether it can modify this histone tail residue, but also what the impact of a loss of this locus is on the inheritance of repressive chromatin states.

Strengths:

The authors clearly achieved their goal, and it is convincingly shown that SET_19 is indeed a somatic cell histone methyltransferase with a striking specificity for H3K23. There is both recombinant protein work, quantitative mapping in vivo, of histone marks and transcriptional changes, and the authors rule out some other hypotheses that have been in the literature. Overall, this provides a compelling argument that SET-19 is indeed the major somatic cell HMT for this residue. Interestingly, the phenotypes are rather minimal, consistent with redundancy in the physiological roles of histone methylation, and redundancy as well in HMT function. For the most part, the data are not over-interpreted. The genetic alleles used, assuming they are confirmed, were revealing and well-documented.

Weaknesses:

The major weaknesses are easily fixed. The major weaknesses mainly reflect a slight overstatement of certain data (claiming insignificance, when it is not clear how that was determined) and claiming a bit too much about SET-32, which was independently claimed to be an H3K23 HMT. Clearly, the two SET domain enzymes are not redundant, nor is the claim that SET-32 has no role in H3K23 methylation completely convincing. Especially in germline or embryonic conditions. Finally, the imaging is not of very high quality, nor are the images fully quantitated. These points can be easily remedied.

Reviewer #2 (Public review):

Summary:

This manuscript identifies SET-19 as a somatic H3K23 methyltransferase in C. elegans, building on previous genetic evidence for a role of set-19 in H3K23me3 regulation. The authors combine quantitative mass spectrometry, western blotting, in vitro methyltransferase assays, ChIP-seq, and RNA-seq to show that loss of set-19 causes a strong reduction of H3K23me3, particularly in somatic tissues, and is associated with derepression of a subset of genes enriched for H3K23me3. They further conclude that SET-19 is dispensable for canonical feeding RNAi and for transgenerational or intergenerational inheritance of RNAi, distinguishing its function from other heterochromatin-associated methyltransferases such as SET-25, SET-32, and the H3K27 HMTs. Overall, the work adds an important piece to the H3K23 methylation pathway and tissue-specific chromatin regulation in C. elegans.

Strengths:

Very strong genetic and biochemical evidence for SET-19 as the major H3K23me3 HMT.

The mass spectrometry and western blot data convincingly demonstrate a strong reduction of H3K23me3 in two independent set-19 alleles and rescue by GFP::SET-19, which is a major strength (Figure 1, including Figure 1f).

The in vitro methyltransferase assays (Figure 2) showing robust H3K23me1/2/3 activity for SET-19 SET+CC and only modest H3K23me activity for SET-32, together with the SAM titration experiment in Figure 2C, are very informative and nicely support the conclusion that SET-19 is a high-activity H3K23 methyltransferase compared to SET-32.

The ChIP-seq analysis is central to the conclusions that H3K23me3 is enriched on chromosome arms, co-localizes with H3K9me3/H3K27me3, and is strongly reduced in set-19 mutants.

Weaknesses:

(1) The global reduction of H3K23me3 in Figure 3b,c and Figure S4c is convincing, but the correlation analysis between H3K23me3 loss and mRNA changes in Figure 3g could be strengthened. Currently, the analysis appears to focus on broad categories; it would be helpful to provide:

Representative genome browser tracks (e.g., exemplary gene coverage plots) for several genes that show clear H3K23me3 peaks in wild type, reduction in set-19, and concomitant upregulation of mRNA levels, and for a few genes that retain H3K23me3 and do not change expression. This would make the link between chromatin changes and transcriptional output more concrete.

(2) In Figure S4C, the authors note a pronounced reduction of H3K23me3 mainly on chromosome arms, but in the current data, it appears that the impact might be arm-specific (i.e., stronger reduction in one arm than the other in a chromosome), with a notable pattern at the X chromosome tip where H3K23me3 seems increased. This is potentially interesting and should be briefly commented on in the Results or Discussion, for example, whether this reflects compensatory activity of another HMT, changes in chromatin organization, or could be a technical artifact.

(3) Figure 3d suggests that some actively expressed genes can also display relatively high H3K23me3 levels, which complicates a simple model of H3K23me3 as exclusively repressive. If feasible, a limited additional analysis stratifying genes by both H3K23me3 and H3K9me3/H3K27me3 status might clarify whether these highly expressed, H3K23me3‑marked genes differ in other chromatin features.

(4) The authors argue that SET-19 primarily affects H3K23me3 and not other canonical repressive marks, based largely on mass spectrometry. It would significantly strengthen the mechanistic conclusions if the authors could assess H3K9me3 and H3K27me3 profiles in set-19 mutants, ideally by ChIP-seq or at least by focused ChIP-qPCR at a subset of loci that lose H3K23me3 and are derepressed at the RNA level. This would address whether H3K23me3 loss occurs independently of changes in other heterochromatin marks, or whether there is crosstalk.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors wanted to determine whether the set-19 gene, one of 38 SET-domain containing genes in C elegans, has a clear function in vivo with respect to lysine methylation. The question is not only whether it can modify this histone tail residue, but also what the impact of a loss of this locus is on the inheritance of repressive chromatin states.

Strengths:

The authors clearly achieved their goal, and it is convincingly shown that SET_19 is indeed a somatic cell histone methyltransferase with a striking specificity for H3K23. There is both recombinant protein work, quantitative mapping in vivo, of histone marks and transcriptional changes, and the authors rule out some other hypotheses that have been in the literature. Overall, this provides a compelling argument that SET-19 is indeed the major somatic cell HMT for this residue. Interestingly, the phenotypes are rather minimal, consistent with redundancy in the physiological roles of histone methylation, and redundancy as well in HMT function. For the most part, the data are not over-interpreted. The genetic alleles used, assuming they are confirmed, were revealing and well-documented.

Thanks very much for the positive comments on our work.

The alleles used in this study were confirmed by PCR and Sanger sequencing, and the sequence information will be added in the revised manuscript.

Weaknesses:

The major weaknesses are easily fixed. The major weaknesses mainly reflect a slight overstatement of certain data (claiming insignificance, when it is not clear how that was determined) and claiming a bit too much about SET-32, which was independently claimed to be an H3K23 HMT. Clearly, the two SET domain enzymes are not redundant, nor is the claim that SET-32 has no role in H3K23 methylation completely convincing. Especially in germline or embryonic conditions. Finally, the imaging is not of very high quality, nor are the images fully quantitated. These points can be easily remedied.

Thanks very much for the comments.

We agree that some interpretations in the original manuscript were too strong, particularly regarding the negative results and the role of SET-32. Our in vitro assays show that SET-32 exhibits H3K23me1 activity and, at higher SAM concentrations, activity toward H3K23me2/3. These findings indicate that SET-32 does have a role in H3K23 methylation. SET-32 is expressed in germ cells, oocytes, and embryos. It is quite likely that redundancy of H3K23 methyltransferase activity exists in these tissues. In the revised manuscript, we will tone down the interpretations and expand the Discussion section to include this possibility. We will also replace the relevant images with higher-quality versions and provide quantitative analyses for Figures 6a and 6b.

Reviewer #2 (Public review):

Summary:

This manuscript identifies SET-19 as a somatic H3K23 methyltransferase in C. elegans, building on previous genetic evidence for a role of set-19 in H3K23me3 regulation. The authors combine quantitative mass spectrometry, western blotting, in vitro methyltransferase assays, ChIP-seq, and RNA-seq to show that loss of set-19 causes a strong reduction of H3K23me3, particularly in somatic tissues, and is associated with derepression of a subset of genes enriched for H3K23me3. They further conclude that SET-19 is dispensable for canonical feeding RNAi and for transgenerational or intergenerational inheritance of RNAi, distinguishing its function from other heterochromatin-associated methyltransferases such as SET-25, SET-32, and the H3K27 HMTs. Overall, the work adds an important piece to the H3K23 methylation pathway and tissue-specific chromatin regulation in C. elegans.

Strengths:

Very strong genetic and biochemical evidence for SET-19 as the major H3K23me3 HMT.

The mass spectrometry and western blot data convincingly demonstrate a strong reduction of H3K23me3 in two independent set-19 alleles and rescue by GFP::SET-19, which is a major strength (Figure 1, including Figure 1f).

The in vitro methyltransferase assays (Figure 2) showing robust H3K23me1/2/3 activity for SET-19 SET+CC and only modest H3K23me activity for SET-32, together with the SAM titration experiment in Figure 2C, are very informative and nicely support the conclusion that SET-19 is a high-activity H3K23 methyltransferase compared to SET-32.

The ChIP-seq analysis is central to the conclusions that H3K23me3 is enriched on chromosome arms, co-localizes with H3K9me3/H3K27me3, and is strongly reduced in set-19 mutants.

Thanks very much for the positive comments on our work.

Weaknesses:

(1) The global reduction of H3K23me3 in Figure 3b,c and Figure S4c is convincing, but the correlation analysis between H3K23me3 loss and mRNA changes in Figure 3g could be strengthened. Currently, the analysis appears to focus on broad categories; it would be helpful to provide:

Representative genome browser tracks (e.g., exemplary gene coverage plots) for several genes that show clear H3K23me3 peaks in wild type, reduction in set-19, and concomitant upregulation of mRNA levels, and for a few genes that retain H3K23me3 and do not change expression. This would make the link between chromatin changes and transcriptional output more concrete.

Thanks very much for the suggestion.

To address this point, we will include representative genome browser tracks for selected genes in the revised manuscript. These examples will help better illustrate the relationship between H3K23me3 loss and mRNA expression changes.

(2) In Figure S4C, the authors note a pronounced reduction of H3K23me3 mainly on chromosome arms, but in the current data, it appears that the impact might be arm-specific (i.e., stronger reduction in one arm than the other in a chromosome), with a notable pattern at the X chromosome tip where H3K23me3 seems increased. This is potentially interesting and should be briefly commented on in the Results or Discussion, for example, whether this reflects compensatory activity of another HMT, changes in chromatin organization, or could be a technical artifact.

Thanks very much for bringing up this point.

As shown in Figure S4C, the overall chromosomal distribution pattern of H3K23me3 is broadly similar between wild type and set-19 mutants, with pronounced enrichment over one chromosomal arm, whereas the center and the opposite arm show relatively lower signal. In set-19 mutants, this asymmetry becomes more pronounced, with a larger difference between the highly enriched arm and the lower-signal regions. This pattern is particularly evident on chromosomes I, II, V, and X. These observations suggest that the effect of set-19 loss on H3K23me3 is not uniform across chromosomal regions.

Substantial H3K23me3 signal remains in specific regions in set-19 mutants, suggesting that additional enzyme(s) also contribute to H3K23me3 methylation. For example, SET-19 appears to function predominantly in somatic tissues, yet the ChIP-seq assays were performed using whole animals, including the germline. Alternatively, there might be compensatory activity of another HMT. In the revised manuscript, we will state these points more explicitly in the Results section and discuss the residual and locally increased H3K23me3 signals.

(3) Figure 3d suggests that some actively expressed genes can also display relatively high H3K23me3 levels, which complicates a simple model of H3K23me3 as exclusively repressive. If feasible, a limited additional analysis stratifying genes by both H3K23me3 and H3K9me3/H3K27me3 status might clarify whether these highly expressed, H3K23me3 marked genes differ in other chromatin features.

Thanks very much for the suggestion.

To address this point, we will perform additional stratified analyses of H3K23me3-marked genes according to their H3K9me3 and/or H3K27me3 status. We will also compare highly and weakly expressed H3K23me3-marked genes to examine whether they differ in other chromatin features, including H3K9me3, H3K27me3, and, if feasible, H3K4me3 and H3K36me3.

(4) The authors argue that SET-19 primarily affects H3K23me3 and not other canonical repressive marks, based largely on mass spectrometry. It would significantly strengthen the mechanistic conclusions if the authors could assess H3K9me3 and H3K27me3 profiles in set-19 mutants, ideally by ChIP-seq or at least by focused ChIP-qPCR at a subset of loci that lose H3K23me3 and are derepressed at the RNA level. This would address whether H3K23me3 loss occurs independently of changes in other heterochromatin marks, or whether there is crosstalk.

Thanks very much for the suggestions.

As suggested, H3K9me3 and H3K27me3 ChIP-seq in wild-type and set-19 mutants will be performed. We will compare their genome-wide distributions and identify loci with significantly altered H3K9me3 and/or H3K27me3 enrichment. These analyses should help clarify whether H3K23me3 loss occurs largely independently of H3K9me3/H3K27me3 changes or reflects potential crosstalk among these repressive chromatin marks. In addition, we will examine H3K9me3 and H3K27me3 enrichment at genes showing both H3K23me3 loss and increased mRNA expression in set-19 mutants to assess whether derepression at these loci is accompanied by changes in other canonical repressive marks.

eLife Assessment

This valuable study examines the subcellular dynamics of the mammalian circadian clock proteins PER2, CRY1, and CK1, providing solid evidence that CK1 modulates the PER2-CRY1 interaction and drives the cytoplasmic localization of PER2 complexes. This could play a key role in modulating transcriptional repression by PER2, CRY1, and CK that contributes to the molecular circadian clock. There are minor concerns regarding the overexpression of the clock proteins in this study.

[Editors' note: this paper was previously reviewed by another journal.]

Reviewer #2 (Public review):

Summary:

This study aims to examine the effects of the subcellular localization of the mammalian clock protein PER2 and its dedicated binding partners CRY1 and the kinase CK1. Using a combination of transient transfection and a Dox-inducible expression system, they show that CRY1 promotes nuclear retention of PER2, and that phosphorylation of PER2 by CK1 promotes cytoplasmic localization and release of CRY1. Changes in complex assembly and subcellular localization could impact the transcriptional repressive function of the CK1-PER2-CRY1 complex in the molecular clock.

Strengths:

The study establishes a system of transient transfection and Dox-inducible expression that allows for strict temporal control of the presence of fluorescently-tagged clock proteins. This is essential to conduct time-lapse microscopy studies that determine changes in the apparent subcellular localization and stability of associated clock proteins. With the potential caveats of overexpression set aside, the authors make use of good controls and supplement cell-based work with in vitro experiments where possible. The discovery that phosphorylation of PER2 by CK1 in the nucleus leads to cytoplasmic localization of PER2 and PER2-CRY1 complexes is a new finding. Moreover, the apparent dissociation of CRY1 from PER2 after CK1 phosphorylation provides a potentially new mechanism by which the repressive activity of this complex could be regulated.

Weaknesses:

Overexpression of circadian clock components, normally expressed at low levels, could disrupt the stoichiometry of native interactions, Although the authors provide a reasonable rationale for the Dox-inducible approach and use appropriate controls throughout the experiments, there is still concern that overexpression of the components of this transcriptional repressive complex far exceed the concentration of the transcription factor they regulate, and this has not been taken into consideration here. In addition, the interesting discovery that CK1 phosphorylation of PER2 leads to dissociation of CRY1 has not identified the phosphorylation site(s) responsible for this, so the mechanism by which this occurs is still unknown. Still, this study provides some interesting hypotheses regarding CK1 regulation of PER2 and CRY1 that could drive future work in the field.

Comments on latest version:

This manuscript has already undergone two rounds of review at a reputable journal, and we have been provided with the previous reviewers' comments and the authors' responses. I am satisfied with the responses and changes to the manuscript made in these previous rounds of review and don't have any further experiments to suggest that wouldn't represent significant additional work.

Author response:

[These author responses are to reviews from another journal.]

Reviewer #1:

This manuscript investigates the behaviour of a variety of clock proteins in cultured cells when epitope tagged and transiently expressed and try to draw general implications for endogenous function of circadian clock proteins.

Clock proteins are expressed at low levels in most cells, and so the clock interacting proteins (other kinases, phosphatases, ubiquitin-conjugated enzymes, etc.) are likewise probably at low abundance. Over-expression of one or two or even three components of a multicomponent system is going to produce odd and obscure non-physiological imbalances. The authors do not extend detailed study of these imbalances to more physiologic levels so the importance of their observations to clock function is not clear, and importantly, they are not tested in more biologically relevant models.

To study the function of components within a system, the steady state must be perturbed in one way or another. This can be achieved through pharmacological treatment, mutagenesis, downregulation, or overexpression. Such interventions are inherently non-physiological, and the relevance of the resulting observations must therefore be carefully validated.

In our study, the purpose of PER2 overexpression was to investigate its subcellular dynamics in the absence and presence of CRYs, specifically CRY1. This is far less trivial than it might appear at first glance, because our data clearly show that PER2 overexpression triggers, within 24 h, the accumulation of endogenous CRY1 (Fig. 1A), due to PER2-mediated stabilization of CRY1 (Fig. 4). PER2 overexpression also induces the accumulation of endogenous PER1, CK1, and BMAL1 (Fig. 2).

This effect was not considered in previous studies, such as Yagita et al. (2002), in which PER2 subcellular localization was assessed at a single time point following transient transfection. Yagita et al. found roughly equal proportions of cells with PER2 exclusively in the nucleus, exclusively in the cytoplasm, or distributed between both compartments. Such extreme cell-to-cell variability cannot be explained solely by PER2’s shuttling dynamics, as that would imply synchronous export in one cell and synchronous import in another.

Our time-resolved analysis of DOX-induced PER2 expression strongly suggests that the variability reported by Yagita et al. reflects a heterogeneous population of unsynchronized cells at different temporal stages along a trajectory from cytoplasmic PER2 (unbound) to nuclear PER2 fully saturated with CRYs (bound), owing to stabilization of endogenous CRYs. Similarly, Öllinger et al. (2014) analyzed PER2 nuclear export in cells constitutively expressing PER2-Dendra. Under such steady-state conditions, PER2-Dendra is already in complex with endogenous CRYs. The slow export rate and lack of dependence on additional CRY1 expression therefore likely reflect export of the complex, which is intrinsically slow.

Thus, prior to our work, no data on the true shuttling dynamics of PER2 were available.

Importantly, our results show not only that CRY1 promotes nuclear accumulation of PER2 (as reported by Öllinger et al.) but also that, conversely, PER2 promotes cytosolic accumulation of CRY1, depending on their expression ratio. Since CRY1 is predominantly nuclear and PER2 predominantly cytosolic, and because a PER2 dimer can bind one or two CRY1 molecules, our data suggest that the shuttling equilibrium depends on PER2 saturation state: a PER2 dimer bound to one CRY1 remains cytosolic, whereas a dimer bound to two CRY1 is nuclear.

These observations are novel and have not been reported previously. They were only possible through time-resolved analysis of overexpressed proteins.

A number of the findings are confirmatory rather than novel - the phosphorylation-regulated nuclear-cytoplasmic shuttling of CK1 and PER proteins is long known, and it's not clearly stated what is novel here. 

We acknowledge prior work by Milne et al. (2001), who showed that kinase-dead CK1 is predominantly nuclear and that prolonged treatment with leptomycin B (16 h) enhances its nuclear localization. We cite this study at the beginning of the relevant paragraph. While we confirm these earlier observations, our work extends them in several important and novel ways:

(1) Rapid dynamics of CK1 localization – We show that pharmacological inhibition of CK1 with PF670 induces rapid (within 1 h) depletion of CK1δ from the centrosome, accompanied by nuclear accumulation and elevated CK1δ levels. These kinetics have not previously been reported. We also show that proteasome inhibition with MG132 enhance centrosomal staining, indicating that centrosomal binding sites are not saturated. Together, the data show that CK1δ equilibrates rapidly between its binding partners. 

(2) Integration of localization with protein stability – We relate the known localization patterns of WT CK1 and the kinase-dead mutant K38R to CK1 degradation dynamics and further compare them to the tau-like kinase mutant CK1δ-R1178Q. This integration of subcellular localization data with turnover mechanisms provides new mechanistic insight.

(3) Comprehensive regulatory model – In the revised manuscript, we now include a schematic summarizing how CK1δ is posttranslationally regulated via subcellular shuttling, nuclear degradation, and dynamic interactions with binding partners (Figure EV5C). To our knowledge, such a comprehensive view of CK1δ regulation, linking localization, stability, and partner association, has not been presented before.

We believe these additions clearly distinguish our findings from prior reports and highlight the novel aspects of our study.

The formation of PER and CRY and CK1 complexes likewise is well established. The finding that formation of multiprotein complexes stabilize otherwise unstable over-expressed proteins is interesting but not novel.

We fully agree that the existence of PER–CRY–CK1 complexes is well established. It is also known that PER2 stabilizes CRY1 by occupying the FBXL3 binding site and that CRY1 promotes the nuclear accumulation of PER2. We do not present these established interactions as novel findings.

Our novel contribution, as outlined above, is the discovery that the shuttling and subcellular localization of PER2 and CRY1 are mutually dependent on their expression ratio. Specifically, we show for the first time that the steady-state shuttling distribution PER2 alone is cytosolic due to its rapid nuclear export wherease CRY1 is predominantly nuclear (known). Given that CRY1 facilitates the nuclear import of PER2 (known) and that a PER2 dimer can bind either one or two CRY1 molecules, our data showing that cytoplasmic PER2-CRY1 foci contain less CRY1 than nuclear foci lead us to conclude that cytoplasmic PER2 complexes contain one CRY1 molecule, while nuclear complexes contain two.

This model provides a mechanistic explanation for the distribution of PER2 between the cytosol and nucleus and for the relatively lower cytosolic CRY1 levels. Moost importantly, we further show (for the first time) that CK1-mediated phosphorylation of PER2 displaces CRY1. This phosphorylation event would produce PER2 dimers with one or no CRY1 bound, promoting their export to the cytosol. We believe this represents a novel and potentially important mechanism for regulating circadian clock function.

The results from many of the imaging assays are not quantitated, and the figures often show single cells. It's hard to draw statistical significance from these.

The phenotypes we report here are result of multiple technical and biological replicates (n >3). Image analysis and statistical analysis was performed when required. We show additional examples in the EVs.

There are a number of phenomena seen whose physiological relevance is unclear. In figure 1, forced over-expression of CRY1 and PER2 leads to formation of nuclear foci. It is unlikely these foci form at non-overexpressed levels, and so the general interest and relevance is not high nor investigated. This reduces the impact of the finding.

It has been shown that PERs and CRYs do not form thermodynamically stable, large (detectable) foci under physiological conditions, as we have stated in the manuscript. Whether these proteins have the propensity to form smaller, more dynamic structures of physiological relevance is an interesting question that could be explored elsewhere, but it is not relevant to our study. In our work, these foci are simply convenient markers for analyzing the interaction and subcellular (co)localization of clock proteins under investigation. In the revised version, we have kept the analysis of these foci and the discussion of their potential relevance to a minimum in order to avoid confusion and unnecessary discussions.

The finding that CK1δ is keep in the dephosphorylated state by binding to PER has been established previously by Johnson and colleagues and should perhaps be mentioned (Qin JBR 2015 (doi: 10.1177/0748730415582127).

There is clearly a misunderstanding here. Qin et al.’s data show that, in a cell-free system, CK1ε phosphorylates PER2 and also autophosphorylates its C-terminal tail (autoradiograph, Fig. 1E).  

However, because PER2 phosphorylation is carried out by CK1ε that is tightly anchored to PER2, there is competition between PER2 phosphorylation and tail autophosphorylation. As a result, the kinetics of tail phosphorylation are slower (Fig. 3B and quantification in C) than those observed with free CK1ε (as seen in the presence of the p53 substrate, Fig. 3A,C). We believe that his is also happening in the cell.

Author response image 1.

Our data, in contrast, address a different point. It has been known from the Virshup lab for decades that CK1δ/ε undergo futile cycles of (auto)phosphorylation and dephosphorylation, resulting in an active, dephosphorylated kinase in cells because cellular phosphatases are more efficient than CK1 autophosphorylation. We now show that CK1δ is also efficiently dephosphorylated when bound to PER2 (Fig. 3). Nevertheless, despite dephosphorylation of PER2-bound CK1δ, PER2 itself becomes hyperphosphorylated, indicating that cellular phosphatases act differently on these two substrates. To clarify this point, we inhibited phosphatases with calyculin A (CalA). Under these conditions, both PER2 and PER2-bound CK1δ became efficiently hyperphosphorylated (new Fig. 3).

The degradation of kinase-active but not inactive CK1 is only shown here with 50-fold overexpressed protein so it's interesting, but the relevance to circadian biology is not made clear. The fact that over-expressed CK1 is degraded primarily in the nucleus is interesting, but needs further characterization - is this affected by the epitope tag? Is it true of endogenous CK1 or only over-expressed CK1? Is this not seen with e.g. other forms of CK1, e.g. lacking the C-terminus?

The observation that unassembled kinase is rapidly degraded is most clearly demonstrated by overexpression experiments. However, Fig. 3 shows that overexpression of CRY1 and PER2 leads to the accumulation of elevated levels of endogenous CK1δ (untagged), indicating that endogenous kinase is likewise degraded in the absence of a stabilizing binding partner. In addition, we present data showing that overexpression of tagged CK1δ reduces the levels of endogenous, untagged CK1δ, further supporting the conclusion that unassembled endogenous CK1δ is unstable and subject to degradation.

Further characterization of the CK1 degradation pathway is of considerable interest and could form the basis of a separate study, particularly to identify the components that mediate activity-dependent nuclear export and activity-dependent nuclear degradation. The Δ-tail kinase is expressed at very low levels, although interpretation is complicated by the possibility that this reflects pleiotropic effects.

The final figure, showing that nuclear CK1 is the form responsible for shortening rhythms, is interesting. Is this because massive increases in nuclear CK1 alter PER, or BMAL/CLOCK, or proteasome activity?  

Our data show that cells expressing either nuclear or cytosolic CK1 are viable, proliferate normally, and maintain a functional circadian clock. Therefore, overexpression of the kinase does not produce pleiotropic effects.

To assume it's due to PER phosphorylation is in disagreement with the studies of Meng et al. Neuron 2008 DOI 10.1016/j.neuron.2008.01.019.

The data are not in disagreement with Meng et al.; in fact, they align quite well. Meng et al. showed that CK1ε-tau shortens the circadian period, which we had also previously reported for CK1δ-tau-like (Marzoll et al., 2022). We now demonstrate that CK1δtau-like is enriched in the nucleus, contributing to its period-shortening phenotype. Furthermore, we show that active CK1δ (but not CK1δ-K38R) promotes cytoplasmic accumulation of PER:CRY complexes, consistent with PER2 degradation in the cytosol as described by Meng et al.

Taken together, these findings suggest that PER proteins acquire their CK1 in the nucleus, and this interaction determines the circadian period length. Following a time delay—set by the kinetics of PER2 phosphorylation—PER2:CRY complexes are exported to the cytosol along with their bound CK1, where they are subsequently degraded.

Reviewer #2:

Interactions between the circadian clock proteins PER1/2 with CK1d/e and CRY1/2 influence each of their stability, subcellular localization, and activity, as countless studies over the last two decades have shown. However, many questions still remain, especially in light of newer models of the transcription-translation feedback loop (TTFL) in which the repression phase relies on two distinct mechanisms, a phosphorylation-dependent displacement of the transcription factor by CK1-PER-CRY complexes from DNA early in repression, and a CRY1dependent sequestration of the transcription factor activation domain later in repression. In particular, questions remain about mechanisms triggering nuclear entry/export and activity of these proteins in the cytoplasm and nucleus. 

Here, the authors utilize a system of induced and/or transient overexpression of proteins with or without with fluorophores to track subcellular localization, stability, and interactions. As the authors point out throughout the manuscript, the overexpression of these clock proteins often causes them to behave differently from the endogenous proteins. It looks as though the authors have done their best to account for these changes, and they have certainly been rigorous in pointing them out, but there is concern that some of the conclusions may be influenced by this overexpression. For example, the relevance of work related to the overexpression-dependent foci is unclear. 

Same answer as to Reviewer 1: It has been shown that PERs and CRYs do not form thermodynamically stable, large (detectable) foci under physiological conditions, as we have stated in the manuscript. Whether these proteins have the propensity to form smaller, more dynamic structures of physiological relevance is an interesting question that could be explored elsewhere, but it is not relevant to our study. In our work, these foci are simply convenient markers for analyzing the interaction and subcellular (co)localization of the clock proteins under investigation. In the revised version, we have kept the analysis of these foci and the discussion of their potential relevance to a minimum in order to avoid confusion.

The findings that the stability of the kinase depend on localization, its intrinsic activity, and interaction with PER2 are interesting and important. Use of the CKBD deletion to show that CK1 stabilization depends on its anchoring interaction with PER2 is a nice touch. The authors bring up an excellent point that most of the potential phosphorylation sites on PER1 and PER2 have not been functionally characterized aside from the phosphoswitch mechanism. Their observation that CK1 eventually induces cytoplasmic localization of the CK1-PER-CRY1 complex and the release of CRY1 is intriguing. In particular, the finding that pretreatment of PER2 with CK1 in vitro blocked its ability to interact with CRY1 is very interesting. However, the absence of mechanistic data to explore this in more detail limits the impact of this conclusion. Using the system they have established here to identify the site(s) on PER2 and/or CRY1 that lead to this would help to solidify this work and increase the impact of this work. Overall, there are some interesting findings here but the inclusion of some competing viewpoints and mechanistic data would strengthen the impact of the work.

Major

(1) The characterization of the tau-like CK1 mutant R178C as less active than the wild type enzyme is not entirely correct-it is less active on the FASP region as described, but it has increased activity on S478 in the phosphodegron that is independent of inhibition from the FASP region (Gallego et al. PNAS, 2007 and Philpott et al. eLife, 2020). It is still possible that some of the period shortening effects of the mutant could arise from enhanced nuclear accumulation, but the oversimplified description of the mutant as less active should be corrected.  

In the revised version, we discuss that the enhanced nuclear localization of the Tau-like kinase may contribute, at least in part, to period shortening, similar to how forced nuclear overexpression of wild-type kinase also shortens the period. We emphasize, however, that CK1 Tau is compromised in its priming-dependent activity, whereas its priming-independent activity is context-specific and enhanced toward the β-TrCP site.

(2) One of main conclusions from the paper, that CK1 induces cytoplasmic localization of the CK1-PER2-CRY1 complex and subsequent release of CRY1 would be strengthened significantly by identifying the phosphorylation site(s) responsible for the cytoplasmic localization of the complex and the release of CRY1. The system they have developed here seems ideal to identify these sites.

We fully agree with the reviewer. We substituted the known phosphorylation sites in PER2 surrounding the CRY-binding domain, but this had no effect on the phosphorylationdependent release of CRY1. Therefore, a more systematic analysis will be required, including the possibility that phosphorylations in CRY1 itself may contribute. To this end, we are generating PER2 and CRY1 variants in which all Ser/Thr residues are replaced by Ala. Using these constructs alongside the wild-type versions, we will by PCR systematically create hybrids in which specific regions containing phosphorylation sites are exchanged.

Nevertheless, this will require considerable time and effort, and we believe this investigation exceeds the scope of the present manuscript and will address it in future work.

(3) The concept of delayed release of CRY1 presented here is an interesting one. It's unclear why the authors have also not incorporated prior findings (Ukai-Tadenuma et al. Cell, 2012, Koike et al. Science, 2012) that peak levels of CRY1 are expressed in a later phase than CRY2, PER1, and PER2. It seems like figure EV6 should reflect the observation that CRY2 is the predominant cryptochrome present during early repression (Koike et al. Science, 2012).

The reviewer is absolutely right: the expression phases of CRY1, CRY2, PER1, and PER2 are important. I have recently discussed these issues in detail in a News & Views article in The EMBO Journal, commenting on a paper by Smyllie et al. In this News & Views article, I discuss that the presently available data suggest that CRY1 is always present throughout the circadian cycle and keeps circadian transcription partially repressed even at peak phases of expression. In the revised version, I refer to these publications, including those mentioned by the reviewer. However, I would like to keep the model presented in the supplementary figure as simple as possible and specifically focused on the work presented in this manuscript, rather than presenting a comprehensive conceptual model of the circadian clock.

(4) The model presented in figure EV6 and described throughout the text shows that PER-CRY complexes interact with CK1 in the nucleus, and not in the cytoplasm prior to nuclear entry. Prior work on endogenous protein complexes has shown that CK1-PER-CRY complexes exist in the cytoplasm very early on in the repression phase (Aryal et al. Mol Cell, 2017-ref. 14 in the manuscript). Work by Sancar and colleagues (Cao et al. PNAS, 2020) also shows with endogenous proteins that CK1d has a circadian pattern of nuclear entry (or possibly retention) concomitant with PER2 that is dependent on the presence of PERs and CRYs. Together, these data seem to be inconsistent with your model. 

We think the data are not inconsistent. The recent Smyllie et al. paper in EMBO Journal shows that PER2 is present in both the cytosol and the nucleus at all times when it is expressed, but cytosolic PER2 is not saturated with CRY, which is more nuclear. Our data demonstrate that PER2 shuttles between the cytosol and the nucleus depending on its occupancy with CRYs (see schematic Fig. 1). Occupancy, in turn, depends on expression levels and binding affinities, including those of CRY2 and PER1. Consequently, PER2 complexes could shuttle continuously throughout the circadian cycle—either because they are not saturated with CRYs due to the balance between expression levels, freely available CRY, and binding affinity, or later in the cycle because CRYs are displaced by phosphorylation. If PER2 acquires casein kinase in the nucleus early in the cycle, it will shuttle out to the cytosol together with the bound CK1. We believe this does occur, but early in the circadian cycle the saturation of PER2 with casein kinase is likely to be very low due to the limited availability of CK1 in the nucleus. I am aware that not everyone will share this interpretation point by point, but discussing it in greater length and detail exceeds the scope of the present manuscript.

Reviewer #3:

This manuscript by Serrano and co-workers is a tight body of work that provides much needed insights into the regulation of clock proteins by CK1D, and into the regulation of CK1D itself. While the whole paper relies on artificial overexpression of chimeric/tagged proteins that may have significant differences in the function, the stability and subcellular distribution of the endogenous proteins they are suppose to model, this limitation was been clearly stated by the authors, and nevertheless their study still provides important insights. 

While the authors have specified which Ck1d isoform (Ck1d1) they are overexpressing in their model cell lines, they may have thought to consider that the overexpression of one Ck1 homologue may affect the endogenous expression of the other homologues and their isoforms, e.g. ck1d1 overexpression may cause an increase in Ck1d2 or Ck1e, which would in turn affect the conclusions. 

We show in revised Fig. 3 that overexpression of CK1δ1 reduces the expression of endogenous CK1δ1/2. This is consistent with our prediction that overexpressed and endogenous CK1 (including CK1ε) compete for the same stabilizing binding partners, leading to rapid degradation of unassembled kinases.

Moreover, the antibody they used for endogenous Ck1d (which is ab85320, also mentioned as AF12G4 but that is the clone number, not the catalogue number) is discontinued and its specificity against Ck1d1, Ck1d2 or even the highly identical Ck1e, has not been clearly demonstrated. We know from Fig 3 that it can detect Ck1d1 but it would be great if the authors would provide additional evidence for the specificity of this antibody, for example by overexpressing Ck1d1/Ck1d2/Ck1e to see really which "endogenous" Ck1 we are seeing.

Are the three bands for example seen in Fig 4A corresponding to the different isoforms? This simple experiment would reinforce the conclusions. 

We show in the revised figure that the antibody recognizes CK1δ1 and CK1δ2, but not CK1ε. In U2OS cells, the antibody detects a single band (Figure); we do not know whether this represents predominantly one splice isoform or both, which are not resolved. However, this distinction is not relevant for our interpretation, because overexpression of tagged CK1δ1 reduces the expression of whichever endogenous kinase is present.

There are no minor comments, as the figures, the figure legends and main text are all of good quality and ready for publication.

Reviewers’ Responses to Point-by-Point Response to Peer Review 

Referee #1:

I appreciated the additional efforts by the authors to improve the manuscript. Unfortunately, the underlying approach of forced over-expression remains artifact-prone, and has been largely supplanted by readily available knockin and targeted mutagenesis methods. Over-expression may give clues, but I think more rigorous mechanistic validation is needed to make this compelling. I cannot support publication of this manuscript.

Referee #2:

In their response to reviewers, the authors make the valid point that the steady state of a system is usually perturbed to study it. In this study, they have used overexpression of the clock proteins PER2, CRY1 and CK1 to study their effects on subcellular dynamics and stability. In justifying this choice, they refer to several papers that similarly overexpressed at least one of these components, stating that their time-resolved approach brings novel insights. However, there is a missed opportunity here to translate any lessons learned from overexpression studies to a system where the proteins are expressed at physiological levels and stoichiometry.

The authors reply to reviewer 1 stating that they conclude PER proteins acquire CK1 in the nucleus, but this does not account for other studies showing an apparent PER-CK1 complex in the cytoplasm during the early phases of repression and/or a pattern of PER-dependent nuclear entry of CK1 (Lee et al. 2001, Cell; Aryal et al. 2017 Mol Cell; Cao et al. 2021 PNAS). Given that all 3 of these studies were done with native expression levels, it seems incumbent upon the authors to demonstrate that their conclusions from the overexpression study are physiologically relevant by translating them in some way to a more native system. This also addresses a point made by reviewer 2, major concern 4 that was not satisfactorily addressed by the authors. Perhaps they could validate their hypothesis of PER shuttling and interactions with CK1 or CRY1 that alter this in a native system similar to Aryal or Cao et al. with the use of nuclear export inhibitors?

The response to reviewer 2, major concern 1 is thoughtful and much appreciated. However, simplifying the effects of the tau mutation on CK1 as having a decreased rate on priming-dependent phosphorylation but not priming-independent is not quite true-the tau mutation also decreases the rate of priming-independent phosphorylation of S662 (in humans) (Philpott et al. 2020, eLife).

Other papers appearing in this journal seem to all include at least one major new mechanistic insight. Although the authors do a diligent job in characterizing the overexpressed proteins in this system, some of their conclusions are at odds with prior studies of the system in more native conditions, so the potential impact of this work is unclear. To verify these conclusions or test new ones (ie, that CK1 disrupts PER-CRY1 interactions), they should use their insights to generate mutations or make perturbations in a native system and demonstrate that they still hold.

Referee #3:

The authors have adequately addressed the reviewers' comments, and it is my opinion that the manuscript is ready for publication. It is true, as previously mentioned by other reviewers, that the evidence presented rely on overexpression, which for the other reviewers seem to preclude publication. However, I find this to be a too strict opinion.

If the authors had indeed provided evidence using crispr-cas9-mediated genetic manipulation and tagging/mutating endogenous genes for all their experiments, thereby providing more physiological evidence of how clock proteins interact, they would probably have submitted their manuscript to an alternative journal with a higher impact.

As it stands, it is my opinion that, considering the evidence and limitations of the study, this manuscript is a good match for the journal.

Author Rebuttal:

Apologies for the delayed reply regarding our manuscript. In the meantime, we have added several new experiments which address the comments of the reviewers and more. These are now included as Figures 1C, EV3, 4D, 6E, 6F, EV6D, and EV7.

Figure 1C reinforces our observations from Figure 1B showing that induction of stably-integrated PER2 also results in accumulation of endogenous CRY1 at a timescale that is compatible with the gradual localization of overexpressed PER2 into the nucleus.

Figure EV3 addresses several technical comments from Reviewers #3 and #1, respectively: Figure EV3A shows that our CK1δ antibody recognizes CK1δ1 and CK1δ2, but not CK1ε. Figures EV 3B and C clearly show how overexpression of our transgenic CK1δ results in decreased endogenous CK1δ which further demonstrates the rapid turnover of active kinase.

Figure 4D addresses the comment from Reviewer #2. We clearly show that CK1δ is not kept in a dephosphorylated state by binding to PER. In addition to our direct comment to this point, Figure 4D shows that CK1δ regardless if it is expressed alone or in complex with PER2 is phosphorylated to a similar extent when the cells are treated with the phosphatase inhibitor CalA. As indicated in our direct response, we are rather more interested in the observation that cellular phosphatases act differently on PER2 compared to CK1δ despite being in the same PER:CK1δ complex (as shown by the clear stabilization of overexpressed CK1δ by co-expression of PER2).

Figures 6E, 6F, and EV6D demonstrate that our observations from overexpression systems are also observed in a more physiological context, addressing comments from Reviewers #1 and #2. Figure 6E shows that dephosphorylation of PER2 leads to its relocalization from the cytosol to the nucleus, while Figure 6F analyzes the subcellular localization of PER2 in the context of a functional circadian clock in U2OS cells. The latter demonstrates that PER2 is predominantly nuclear early in the circadian cycle, but redistributes to the cytosol at later time points. We included these experiments in response to the reviewer’s request for a more physiological context. Since we are not a mouse lab, this cell-based system represents the most physiological model we can provide. Figure 6F show the dynamics of endogenous PER2 from DEX-synchronized cells. At early timepoints, PER2 is predominantly nuclear likely due to the incorporation of CRY1 forming the PER:CRY complex. At later timepoints PER2 is redistributed between the cytoplasm and nucleus due to PER2 phosphorylation. Importantly, these results are consistent with and recontextualize the results from Liu et al. (Xie et al., PNAS, 2023) showing the hypophosphorylated PER2 at early timepoints post-DEX is predominantly nuclear and hyperphosphoryated PER2, that appear later post-DEX is predominantly cytoplasmic.

Finally, Figure EV7 provides a model how the subcellular distribution of CK1δ affects its assembly into the PER:CRY complex emphasizing how nuclear kinase enacts its role in the circadian clock.

Response to Reviewers:

We were disappointed by the categorical rejection of overexpression experiments. Without a specific discussion of why they would be inappropriate or not sufficient in the context of the work presented here, the blanket assertion that overexpression inevitably produces artifacts functions more as a rhetorical device than as a substantiated scientific argument. The fact that the term ‘physiological’ generally carries a positive connotation, whereas ‘overexpression’ is often perceived negatively, does not in itself justify the categorical rejection of experiments.

While we appreciate that some reviewers may personally prefer alternative strategies, we believe that the suitability of any approach must be evaluated in light of the specific biological questions being addressed. I cannot see a single specific point in the reviewers’ responses indicating that any of our experiments yielded artificial results. It is true that targeted knock-in and mutagenesis methods are available, however, these approaches are simply not suited to the questions raised in this manuscript. We also fully agree that, whenever possible, insights from overexpression studies should be validated in systems with a functional clock where proteins are expressed at physiological levels, which we did using U2OS cells, and noting the compatibility of our results with those in the literature using endogenously-tagged constructs. We have cited several recent studies that have investigated the subcellular distribution and circadian dynamics of endogenous or endogenously-tagged clock proteins in mice (Cao et al, 2021; Smyllie et al, 2022, 2016, 2025) and U2OS cells (Öllinger et al, 2014; Gabriel et al, 2021; Xie et al, 2023). While we cannot substantially expand on these previous observations, we confirm them in the revised version by demonstrating the nuclear-to-cytoplasmic relocalization of PER2 in U2OS cells over the course of a circadian cycle. In addition, we show that this process is, in principle, reversible: when CK1 is inhibited with PF670, overexpressed hyperphosphorylated cytosolic PER2 becomes dephosphorylated and accumulates in the nucleus.

Overall, we consider our approach not only complementary but also essential, as it enables us to address two key questions that would otherwise be difficult or even impossible to resolve:

(1) Mutual impact of PER2 and CRY1 on subcellular dynamics and the role of PER2 phosphorylation

Evidence from mouse liver (Cao et al, 2021), mouse SCN (Smyllie et al, 2022, 2025), and U2OS cells (Xie et al, 2023) indicates that a substantial fraction of PER2 remains cytoplasmic throughout its expression cycle, even in the presence of CRY1, which promotes PER’s nuclear import. The mechanisms underlying this cytoplasmic retention remain unclear, and no circadian function has yet been attributed to the cytosolic PER2 pool. Our study addresses how PER2 abundance, phosphorylation state, and stoichiometry relative to CRY1 govern their interaction and subcellular dynamics. This is physiologically relevant because PER1/2 and CRY1/2 proteins oscillate in expression and degradation out of phase, such that their concentrations, stoichiometry, and phosphorylation state vary systematically over the circadian cycle. Transient transfection and inducible overexpression combined with time-lapse microscopy are essential here, as they uniquely allow modulation of protein ratios and CK1δ levels and to resolve their dynamics.

Previous work established that CRY1 is nuclear and promotes PER2 nuclear accumulation (Smyllie et al, 2022). Our data extend this by showing that subcellular distribution is determined by the CRY1:PER2 ratio. While CRY1 alone is nuclear we show that PER2 alone is cytoplasmic due to rapid nuclear export. Mixed conditions reveal ratio-dependent shifts: at low CRY1-to-PER2 ratios, CRY1 relocalizes to the cytoplasm, whereas at high ratios, PER2 is retained in the nucleus. We explain this behavior by PER2 dimerization: dimers bound to two CRY1 molecules remain nuclear, while dimers bound to a single CRY1 localize to the cytosol. Such species can be expected to form in a physiological context depending on binding affinities and rhythmic expression levels and ratios across circadian time. Importantly, we show that CK1δ-mediated phosphorylation destabilizes PER2 and CRY1 interactions. From this, we infer that PER2 dimers with only a single bound CRY1 transiently form and accumulate in the cytosol, consistent with the lower CRY1-to-PER2 ratio we observe in the cytosol and that has also been reported in the SCN (Smyllie et al, 2025). With continued phosphorylation, PER2 dimers lose CRY1 altogether, while the released CRY1 accumulates in the nucleus. We suggest that this mechanism supports and extends the late repressive phase of the circadian cycle. Recent data show that hypophosphorylated PER2 is predominantly nuclear, whereas hyperphosphorylated PER2 is largely cytoplasmic in mouse liver (Cao et al, 2021; Xie et al, 2023), linking our data to a physiological context.

Taken together, these findings suggest a mechanism whereby stoichiometry, subunit composition, and CK1δ phosphorylation determine PER:CRY complex composition and localization. Crucially, these complexes and their dynamic relocalization could only be observed using inducible overexpression; knock-in strategies at endogenous levels would not be able to capture such states.

(2) Posttranslational regulation and subcellular homeostasis of CK1δ and impact on the clock

Previous work has shown that nuclear export of CK1δ depends on its kinase activity (Milne et al, 2001). Here, we further demonstrate that unassembled CK1δ is subject to degradation, with nuclear turnover accelerated by its catalytic activity. Thus, when evaluating the impact of CK1δ mutants on the circadian clock, one must consider not only kinase activity but also protein stability and subcellular distribution. We find that CK1δ availability for PER2 differs between cytosol and nucleus. In particular, nuclear CK1δ is limiting, and its abundance directly determines circadian period length. This is significant because subcellular CK1δ availability and posttranslational regulation have not previously been examined or incorporated into circadian clock models, as the kinase has been assumed to be non-limiting given its constant expression throughout the circadian cycle. Complex formation between CK1δ and PER is a well-established determinant of circadian timing, with CK1δ overexpression known to shorten period length. Our data explain why: the binding equilibrium between CK1δ and PER must be finely tuned. Previous studies suggested that PER associates with CK1δ in the cytosol and enters the nucleus as a PER:CRY:CK1δ complex (Lee et al, 2001; Aryal et al, 2017). Our data suggest that nuclear PER is not saturated with CK1δ. This is because levels of free, active CK1δ in the nucleus are low, owing to its rapid export or degradation by the nuclear proteasome, which limits its availability for PER binding.

Our overexpression studies support this mechanism. NES-tagged CK1δ overexpression does not alter circadian period length, because it fails to increase nuclear CK1δ levels: Each PER molecule can coimport only one kinase, a process already occurring in wild-type cells, and the few co-imported molecules rapidly equilibrate with the nuclear pool, where they are subject to export or degradation. In contrast, NLS-tagged CK1δ overexpression directly increases nuclear kinase abundance by antagonizing export, thereby enhancing PER binding and shortening circadian period. This multilayered regulation of CK1δ stability and localization and its consequences for PER2 availability would not have been revealed without targeted overexpression. Our findings therefore fill a key knowledge gap and remain fully consistent with previous studies (Lee et al, 2001; Aryal et al, 2017; Cao et al, 2021).

Conclusion: In sum, our findings are novel and physiologically relevant, aligning with data from mouse liver and SCN. While studies at strictly endogenous protein levels are important and necessary, perturbation of steady state is a standard strategy to uncover and observe novel mechanisms. Endogenous-level experiments would demand technically unrealistic systems (for example, even the simplest case, analyzing the subcellular dynamics of PER2 alone, would require cells lacking PER1, CRY1/2, and CK1δ/ε). Moreover, adjustment of PER2-to-CRY1 ratios cannot be achieved with stably integrated genes and of course not at physiological expression levels. Thus, inducible overexpression is not merely practical but currently the most feasible approach to dissect these dynamics. We complement our findings with data from U2OS cells with a functional clock, showing that the availability of nuclear CK1δ directly determines circadian period length. Although specific aspects of our extended model require further experimental validation, no published evidence contradicts it to date. Mechanistic discussions of the circadian clock have so far focused primarily on PER protein degradation. Our model broadens this perspective by incorporating CK1δ homeostasis, PER:CRY complex composition, subcellular localization, and their regulation by phosphorylation. In doing so, it provides a detailed framework to be critically tested and refined in future studies.

eLife Assessment

This manuscript presents a valuable study of the activity and functional relevance of different circuits in the dentate gyrus of mice performing a pattern separation task. Convincing evidence is presented to support the paper's central conclusions. The study is likely to be of interest to those studying the subregional organization and cell type-specific functions of the dentate gyrus.

Reviewer #2 (Public review):

In this study, the authors investigate how increasing cognitive demand shapes activity patterns in the dorsal dentate gyrus (DG). Using a touchscreen-based TUNL task combined with TRAP/c-Fos tagging, birth-dating of adult-born granule cells (abDGCs), and chemogenetic inhibition, they show that higher task demand increases mature granule cell (mGC) recruitment and enhances suprapyramidal (SB) versus infrapyramidal (IB) blade bias. Functionally, mGC inhibition reduces overall activity and impairs performance without disrupting blade bias, whereas inhibition of {less than or equal to}7-week-old abDGCs increases mGC activity, abolishes blade bias, and impairs discrimination under high-demand conditions. These findings suggest that effective pattern separation depends not only on overall DG activity levels but also on the spatial organization of recruited ensembles.

The integration of touchscreen TUNL with temporally controlled activity tagging and birth-dated cohorts is technically strong. Quantification of SB-IB bias and radial/apical distributions adds anatomical precision beyond bulk activity measures. The comparison between mGC and abDGC inhibition is conceptually compelling and supports dissociable functional roles. Overall, the data convincingly demonstrate that increasing cognitive demand amplifies blade-biased DG recruitment and that mGCs and abDGCs differentially contribute to both behavioral performance and network organization.

However, how abDGCs are integrated into the mGC network under high cognitive demand remains unresolved. Additional experiments are needed to clarify how abDGCs shape spatial recruitment patterns and whether they directly inhibit or indirectly regulate mGC activity to maintain high performance.

Furthermore, the authors frame "high cognitive demand" as a multidimensional construct encompassing broad behavioral challenge. It would strengthen the work to delineate how local abDGC-mGC circuit interactions regulate specific task components in real time. This will require higher temporal resolution approaches, as TRAP and c-Fos labeling integrate activity over prolonged windows and primarily reflect sustained engagement rather than moment-to-moment computations.<br /> The central conclusion that dentate function depends on coordinated spatial recruitment rather than total activity magnitude is supported by the data, although mechanistic interpretations are tempered given methodological limitations.<br /> Overall, this work advances models of adult neurogenesis by emphasizing a critical-period modulatory role of abDGCs in organizing DG network activity during high-demand discrimination. The combined behavioral and circuit-level framework is likely to be influential in the field.

Comments on revisions:

None remaining.

Author response:

The following is the authors’ response to the previous reviews

Public Reviews:

Reviewer #1 (Public review):

This manuscript investigates how dentate gyrus (DG) granule cell subregions, specifically suprapyramidal (SB) and infrapyramidal (IB) blades, are differentially recruited during a high cognitive demand pattern separation task. The authors combine TRAP2 activity labeling, touchscreen-based TUNL behavior, and chemogenetic inhibition of adult-born dentate granule cells (abDGCs) or mature granule cells (mGCs) to dissect circuit contributions.

This manuscript presents an interesting and well-designed investigation into DG activity patterns under varying cognitive demands and the role of abDGCs in shaping mGC activity. The integration of TRAP2-based activity labeling, chemogenetic manipulation, and behavioral assays provides valuable insight into DG subregional organization and functional recruitment. However, several methodological and quantitative issues limit the interpretability of the findings. Addressing the concerns below will greatly strengthen the rigor and clarity of the study.

Major points:

(1) Quantification methods for TRAP+ cells are not applied consistently across panels in Figure 1, making interpretation difficult. Specifically, Figure 1F reports TRAP+ mGCs as density, whereas Figure 1G reports TRAP+ abDGCs as a percentage, hindering direct comparison. Additionally, Figure 1H presents reactivation analysis only for mGCs; a parallel analysis for abDGCs is needed for comparison across cell types.

In Figure 1G and 1H we report TRAP+ abDGCs as a percentage rather than density because we are analyzing colocalization of the two markers, which are very sparse in this population. Given the very low number of double-labeled abDGCs, calculating density would not be practical. In the revised manuscript we have clarified the rationale for using these measures. As noted in the current text, we did not observe abDGCs co-expressing TRAP and c-Fos; we have made this point more explicit to guide interpretation of these data.

(2) The anatomical distribution of TRAP+ cells is different between low- and high-cognitive demand conditions (Figure 2). Are these sections from dorsal or ventral DG? Is this specific to dorsal DG, as itis preferentially involved in cognitive function? What happens in ventral DG?

The sections shown in Figure 2 were obtained from the dorsal dentate gyrus (see Methods, “Histology and imaging”: stereotaxic coordinates −1.20 to −2.30 mm relative to bregma, Paxinos atlas). From a feasibility standpoint, it is not possible to analyze the entire longitudinal extent of the hippocampus with these low-throughput histological approaches. We therefore focused on the dorsal DG, for which there is a strong functional rationale. A large body of work indicates that the dorsal hippocampus, and specifically the dorsal DG, is preferentially involved in spatial memory and in the fine contextual discrimination that underlies pattern separation. The dorsal hippocampus is critical for encoding and distinguishing similar spatial representations, a core component of the high-cognitive demand task used here. In contrast, the ventral DG is more strongly associated with emotional regulation and affective memory processing and is less implicated in high-resolution spatial encoding. For these reasons, the present study was designed to assess TRAP+ cell distributions specifically in the dorsal DG.

(3) The activity manipulation using chemogenetic inhibition of abDGCs in AsclCreER; hM4 mice was performed; however, because tamoxifen chow was administered for 4 or 7 weeks, the labeled abDGC population was not properly birth-dated. Instead, it consisted of a heterogeneous cohort of cells ranging from 0 to 5-7 weeks old. Thus, caution should be taken when interpreting these results, and the limitations of this approach should be acknowledged.

We agree that prolonged tamoxifen administration results in labeling a heterogeneous population of abDGCs spanning approximately 0 to 5–7 weeks of age, rather than a precisely birth-dated cohort. This is a limitation of this approach and we have included discussion of this in more detail in the revised manuscript.

(4) There is a major issue related to the quantification of the DREADD experiments in Figure 4, Figure 5, Figure 6, and Figure 7. The hM4 mouse line used in this study should be quantified using HA, rather than mCitrine, to reliably identify cells derived from the Ascl lineage. mCitrine expression in this mouse line is not specific to adult-born neurons (off-targets), and its expression does not accurately reflect hM4 expression.

We agree that mCitrine is not a marker that allows localization of hM4Di as it is well known that the mCitrine can be independently expressed in a Cre independent manner in this mouse. As suggested, we have removed the figure that showed the mCitrine and have performed immunohistochemical localization of the DREADD with an antibody against the HA tag. This is now shown in Figure 5.

(5) Key markers needed to assess the maturation state of abDGCs are missing from the quantification. Incorporating DCX and NeuN into the analysis would provide essential information about the developmental stage of these cells.

The goal of this study was to examine activity patterns of adult-born versus mature granule cells, rather than to assess maturation state. The adult-born neurons analyzed were 25–39 days old, an age at which point most cells have progressed beyond the DCX⁺ stage and are expected to express NeuN based on prior work. We therefore do not think that including DCX or NeuN quantification would provide additional information relevant to the aims or interpretation of this study.

Minor points:

(1) The labeling (Distance from the hilus) in Figure 2B is misleading. Is that the same location as the subgranular zone (SGZ)? If so, it's better to use the term SGZ to avoid confusion.

We have updated Figure 2B, the Methods, and the main text to more explicitly localize this which it the boundary between the subgranular zone (SGZ) and the hilus.

(2) Cell number information is missing from Figures 2B and 2C; please include this data.

We have now added the cell number information to the figure legends. In Figures 2B and 2C, each point corresponds to a single cell, with an equal number of mice per group. The total number of TRAP⁺ cells per mouse is shown in Figure 1F, which reports TRAP⁺ cell densities by group.

(3) Sample DG images should clearly delineate the borders between the dentate gyrus and the hilus. In several images, this boundary is difficult to discern.

We made the DG-hilus boundaries clearer in the sample images to improve visualization and interpretation.

(4) In Figure 6, it is not clear how tamoxifen was administered to selectively inhibit the more mature 6-7-week-old abDGC population, nor how this paradigm differs from the chow-based approach. Please clarify the tamoxifen administration protocol and the rationale for its specificity.

We apologize for the confusion here. The protocol used in Figure 6 is the same tamoxifen chow–based approach as in Figure 5, differing only in the duration of tamoxifen exposure. Mice in Figure 5 received tamoxifen chow for 7 weeks, whereas mice in Figure 6 received it for 4 weeks, restricting labeling to a younger and narrower cohort of adult-born DGCs. Thus, the population targeted in Figure 6 is younger than that in Figure 5 and does not correspond to mature 6–7-week-old neurons. By contrast, the experiment in Figure 4 targets a more mature population, consisting predominantly of ~5-week-old adult-born neurons as well as mature granule cells, which are Dock10-positive and express Cre endogenously, allowing selective manipulation of this later-stage population.

We have corrected the paragraph accordingly and clarified the age range of the labeled populations in the revised manuscript.

Comments on revisions:

I appreciate the authors' careful and thorough revisions. They have addressed all of my previous concerns satisfactorily, and the manuscript is now significantly strengthened. I have no further concerns.

Reviewer #2 (Public review):

In this study, the authors investigate how increasing cognitive demand shapes activity patterns in the dorsal dentate gyrus (DG). Using a touchscreen-based TUNL task combined with TRAP/c-Fos tagging, birth-dating of adult-born granule cells (abDGCs), and chemogenetic inhibition, they show that higher task demand increases mature granule cell (mGC) recruitment and enhances suprapyramidal (SB) versus infrapyramidal (IB) blade bias. Functionally, mGC inhibition reduces overall activity and impairs performance without disrupting blade bias, whereas inhibition of {less than or equal to}7-week-old abDGCs increases mGC activity, abolishes blade bias, and impairs discrimination under high-demand conditions. These findings suggest that effective pattern separation depends not only on overall DG activity levels but also on the spatial organization of recruited ensembles.

The integration of touchscreen TUNL with temporally controlled activity tagging and birth-dated cohorts is technically strong. Quantification of SB-IB bias and radial/apical distributions adds anatomical precision beyond bulk activity measures. The comparison between mGC and abDGC inhibition is conceptually compelling and supports dissociable functional roles. Overall, the data convincingly demonstrate that increasing cognitive demand amplifies blade-biased DG recruitment and that mGCs and abDGCs differentially contribute to both behavioral performance and network organization.

However, how abDGCs are integrated into the mGC network under high cognitive demand remains unresolved. Additional experiments are needed to clarify how abDGCs shape spatial recruitment patterns and whether they directly inhibit or indirectly regulate mGC activity to maintain high performance.

Furthermore, the authors frame "high cognitive demand" as a multidimensional construct encompassing broad behavioral challenge. It would strengthen the work to delineate how local abDGC-mGC circuit interactions regulate specific task components in real time. This will require higher temporal resolution approaches, as TRAP and c-Fos labeling integrate activity over prolonged windows and primarily reflect sustained engagement rather than moment-to-moment computations.

The central conclusion that dentate function depends on coordinated spatial recruitment rather than total activity magnitude is supported by the data, although mechanistic interpretations should be tempered given methodological limitations.

Overall, this work advances models of adult neurogenesis by emphasizing a critical-period modulatory role of abDGCs in organizing DG network activity during high-demand discrimination. The combined behavioral and circuit-level framework is likely to be influential in the field.

Reviewer #3 (Public review):

This study examines the role of dentate gyrus neuronal populations, reflecting neurogenesis and anatomical location (suprapyramidal vs infrapyramidal blade), in a mnemonic discrimination task that taxes the pattern separation functions of the dentate. The authors measure dentate gyrus activity resulting from cognitive training and test whether adult neurogenesis is required for both the anatomical patterns of activity and performance in the cognitive task. The authors find that more cognitively challenging variants of the task evoked more dentate activity, but also distinct patterns of activity (more activity in the suprapyramidal blade, less in the infdrapyramidal blade). Using chemogenetic approaches they silence mature vs immature dentate gyrus neurons and find that only mature neurons (either the general population or specifically mature adult-born neurons), and not immature adult-born neurons, are required for the difficult version of the task. Inhibition of mature adult-born neurons furthermore increased overall activity in the dentate and reduced the biased pattern of activity across the blades, consistent with evidence that adult-born neurons broadly regulate dentate gyrus activity.

Comments on revisions:

I appreciate the efforts the authors have taken to revise this manuscript. I have only minor concerns with this revised version of the manuscript:

Methods state that significance is defined as P<0.05 but some results are interpreted as significant when P=0.05. Either the alpha value needs to change or the interpretation needs to change.

We have corrected the statement in the Methods section to define statistical significance as P ≤ 0.05, which aligns with how significance was interpreted throughout the manuscript.

I believe the statistical results for group and blade effects for the ANOVAs, in Figs 2,3 & 4, appear to be switched (blade should be significant, not group).

We thank the reviewer for pointing out this mistake. We have corrected the reported statistical results for the group and blade effects in the manuscript accordingly.

I appreciate that sometimes there is not a perfect overlap between immunohistochemical signals, but I continue to believe that the spatially-non-overlapping TRAP and EDU signals in Fig 3 is caused by these 2 markers being in different cells. A Z-stack or orthogonal projection could verify/disprove this concern.

We agree that limited overlap in single optical sections can raise the possibility that TRAP and EdU signals originate from different cells. However, based on our imaging conditions and inspection across focal planes, the signals are consistent with being present within the same cells, with partial spatial separation likely reflecting subcellular localization and/or sectioning effects.

eLife Assessment

This timely and fundamental study presents an innovative iPSC based co-culture system to model Kupffer cell-hepatocyte interactions and hepatotoxicity, demonstrating reciprocal acquisition of tissue identity and enhanced hepatocyte maturation. The work is convincing, supported by well-executed methodology and functional validation, including physiologically relevant, concentration-dependent hepatotoxic responses. The research approach is promising and of broad interest, further clarification of experimental design and interpretation may strengthen its impact.

Reviewer #1 (Public review):

The manuscript presents a compelling new in vitro system based on isogenic co-cultures of human iPSC-derived hepatocytes and macrophages, enabling the modelling of hepatic immune responses with unprecedented physiological relevance. The authors show that co-culture leads to enhanced maturation of hepatocytes and tissue-resident macrophage identity, which cannot be achieved through conditioned media alone. Using this system, they functionally validate immune-driven hepatotoxic responses to a panel of drugs and compare the system's predictive power to that of monocyte-derived macrophages. The results underscore the necessity of macrophage-hepatocyte crosstalk for accurate modelling of liver inflammation and drug toxicity in vitro. The manuscript is clearly written and addresses a key limitation in liver organoid systems: the lack of immune complexity and tissue-specific macrophage imprinting.

Strengths:

• Novelty and Relevance: The study presents a highly innovative co-culture system based on isogenic human iPSCs, addressing an unmet need in modelling immune-mediated hepatotoxicity.

• Mechanistic Insight: The reciprocal reprogramming between iHeps and iMacs, including induction of KC-specific pathways and hepatocyte maturation markers, is convincingly demonstrated.

• Functional Readouts: The application of the model to detect IL-6 responses to hepatotoxic compounds enhances its translational relevance.

Weaknesses:

The co-culture model with monocyte-derived macrophages is not fully characterised, making comparisons less informative.

Reviewer #3 (Public review):

Summary:

In this study, the authors establish a human in vitro liver model by co-culturing induced hepatocyte-like cells (iHEPs) with induced macrophages (iMACs). Through flow cytometry-based sorting of cell populations at days 3 and 7 of co-culture, followed by bulk RNA sequencing, they demonstrate that bidirectional interactions between these two cell types drive functional maturation. Specifically, the presence of iMACs accelerates the hepatic maturation program of iHEPs, while contact-dependent cues from iHEPs enhance the acquisition of Kupffer cell identity in iMACs, indicating that direct cell-cell interactions are critical for establishing tissue-resident macrophage characteristics.

Functionally, the authors show that iMAC-derived Kupffer-like cells respond to pathological stimuli by producing interleukin-6 (IL-6), a hallmark cytokine of hepatic immune activation. When exposed to a panel of clinically relevant hepatotoxic drugs, the co-culture system exhibited concentration-dependent modulation of IL-6 secretion consistent with reported drug-induced liver injury (DILI) phenotypes. Notably, this response was absent when hepatocytes were co-cultured with monocyte-derived macrophages from peripheral blood, underscoring the liver-specific phenotype and functional relevance of the iMAC-derived Kupffer-like cells. Collectively, the study proposes this co-culture platform as a more physiologically relevant model for interrogating macrophage-hepatocyte crosstalk and assessing immune-mediated hepatotoxicity in vitro.

Strengths:

A major strength of this study lies in its systematic dissection of cell-cell interactions within the co-culture system. By isolating each cell type following co-culture and performing comprehensive transcriptomic analyses, the authors provide direct evidence of bidirectional crosstalk between iMACs and iHEPs. The comparison with single-culture controls is particularly valuable, as it clearly demonstrates how co-culture enhances functional maturation and lineage-specific gene expression in both cell types. This approach allows for a more mechanistic understanding of how hepatocyte-macrophage interactions contribute to the acquisition of tissue-specific phenotypes

Weaknesses:

(1) Overreliance on bulk RNA-seq data:

The primary evidence supporting cell maturation is derived from bulk RNA sequencing, which has inherent limitations in resolving heterogeneous cellular states and functional maturation. The conclusions regarding hepatocyte maturation are based largely on increased expression of a subset of CYP genes and decreased AFP levels - markers that, while suggestive, are insufficient on their own to substantiate functional maturation. Additional phenotypic or functional assays (e.g., metabolic activity, protein-level validation) would significantly strengthen these claims.

(2) Insufficient characterization of input cell populations:

The manuscript lacks adequate validation of the cellular identities prior to co-culture. Although the authors reference previously published protocols for generating iHEPs and iMACs, it remains unclear whether the cells used in this study faithfully retain expected lineage characteristics. For example, hepatocyte preparations should be characterized by flow cytometry for ALB and AFP expression, while iMACs should be assessed for canonical macrophage markers such as CD45, CD11b, and CD14 before co-culture. Without these baseline data, it is difficult to interpret the magnitude or significance of any co-culture-induced changes.

(3) Quantitative assessment of IL-6 production is insufficient:

The analysis of drug-induced IL-6 responses is based primarily on relative changes compared to control conditions. However, percentage changes alone are inadequate to capture the biological relevance of these responses. Absolute cytokine production levels - particularly in response to LPS stimulation - should be reported and directly compared to PBMC-derived macrophages to determine whether iMAC-derived Kupffer-like cells exhibit enhanced cytokine output. Moreover, the Methods section should clearly describe how ELISA results were normalized or corrected to account for potential differences in cell number, viability, or culture conditions.

(4) Unclear mechanistic interpretation of IL-6 modulation:

The observed changes in IL-6 production upon drug treatment cannot be interpreted solely as evidence of Kupffer cell-specific functionality. For instance, IL-6 suppression by NSAIDs such as diclofenac is well known to result from altered prostaglandin synthesis due to COX inhibition, while leflunomide's effects are linked to metabolite-induced modulation of immune cell proliferation and broader cytokine networks. These mechanisms are distinct from Kupffer cell identity and may not directly reflect liver-specific macrophage function. Consequently, changes in IL-6 secretion alone - particularly without additional mechanistic evidence or analysis of other cytokines - are insufficient to conclude that co-culture with hepatocytes drives the acquisition of bona fide Kupffer cell maturity.

Reviewers comments to revised manuscript.

The authors successfully established an isogenic, iPSC-derived human liver co-culture model to investigate the role of hepatocyte-macrophage interactions in driving Kupffer cell (KC) identity and hepatocyte maturation. By utilizing a single genetic background, the authors effectively minimized the experimental variability often encountered in non-isogenic systems. A significant highlight of this work is the demonstration that direct co-culture-as opposed to conditioned media alone-is a primary driver for critical KC identity markers such as ID1 and ID3. Furthermore, the model's ability to recapitulate complex clinical IL-6 responses to known hepatotoxicants where standard models have failed underscores its potential utility for early-stage DILI screening. However, there are significant methodological concerns regarding the data analysis. While the study compares four or five distinct experimental groups (e.g., Day 0, Day 7, Day 3 co-culture, and Day 7 co-culture), the authors utilized Student's t-tests for these comparisons. This approach does not account for the multiple comparisons problem and increases the risk of Type I errors. Additionally, while IL-6 secretion is used as a primary functional readout, the individual mechanisms behind these drug responses were not explored experimentally. Finally, Pearson correlation analysis indicates that the iMacs remain poorly correlated with actual in vivo human embryonic liver macrophages, suggesting that the "imprinting" of true KC identity remains incomplete.

Author response:

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

Public Reviews:

Reviewer #1 (Public review):

The manuscript presents a compelling new in vitro system based on isogenic co-cultures of human iPSC-derived hepatocytes and macrophages, enabling the modelling of hepatic immune responses with unprecedented physiological relevance. The authors show that co-culture leads to enhanced maturation of hepatocytes and tissue-resident macrophage identity, which cannot be achieved through conditioned media alone. Using this system, they functionally validate immune-driven hepatotoxic responses to a panel of drugs and compare the system's predictive power to that of monocyte-derived macrophages. The results underscore the necessity of macrophage-hepatocyte crosstalk for accurate modelling of liver inflammation and drug toxicity in vitro.

The manuscript is clearly written and addresses a key limitation in liver organoid systems: the lack of immune complexity and tissue-specific macrophage imprinting. Nevertheless, several conclusions would benefit from a more careful interpretation of the data, and some important controls or explanations are missing, particularly in the flow cytometry gating strategies, stress marker validation, and cluster interpretations.

Strengths:

(1) Novelty and Relevance: The study presents a highly innovative co-culture system based on isogenic human iPSCs, addressing an unmet need in modelling immune-mediated hepatotoxicity.

(2) Mechanistic Insight: The reciprocal reprogramming between iHeps and iMacs, including induction of KC-specific pathways and hepatocyte maturation markers, is convincingly demonstrated.

(3) Functional Readouts: The application of the model to detect IL-6 responses to hepatotoxic compounds enhances its translational relevance.

Weaknesses:

(1) Several key claims, particularly those derived from PCA plots and DEG analyses, are overinterpreted and require more conservative language or further validation.

We agree that PCA does not allow for maturation trajectories and mentioned that it was a hypothesis that the co-culture was promoting maturation, which we later validated by looking at the expression of key hepatocyte markers as well as by pearson correlation comparison with fetal hepatocytes.

(2) The purity of sorted hepatocytes and macrophages is not convincingly demonstrated; contamination across gates may confound transcriptomic readouts.

We agree and have highlighted and addressed this limitation in our discussion. Unfortunately, this is a limitation of bulk sequencing that a small amount of contamination might be present, however the TPM values of ALB for example in the iMacs is extremely low especially when compared to the hepatocytes, indicating that the level of contamination is likely to be very low. Likewise, the expression of CSF1R in the co-cultured iHeps is also extremely low. This has been included in Supp Fig 1F and G.

(3) Stress response genes and ER stress/apoptosis signatures are not properly assessed, despite being potentially activated in the system.

This has been included in Supp Fig 2C, where we’ve included the expression of ATF4, CASP3 and CASP9. Although there’s a significant difference in ATF4 expression between Day 0 and Day 7 iHep only/Co-culture, there is no significant difference between the Day 7 iHep only and Day 7 iHep Co-culture. There are no significant differences in CASP3 and CASP9 expression across all the samples.

(4) Some figure panels and legends lack statistical annotations, and microscopy validation of morphological changes is missing.

Although we agree that the morphology changes would be interesting, we think that this question is unfortunately outside of the scope of our question. Although Kupffer cells are in direct contact with hepatocytes, they migrate from the liver parenchyma into the sinusoidal spaces where they primarily reside. We do not think that the morphology would add much to the paper, especially given that this is a 2D model as well.

(5) The co-culture model with monocyte-derived macrophages is not fully characterised, making comparisons less informative.

Although we agree that it would be interesting to look more closely at the monocyte-derived macrophage co-cultures as well, we think that this would be more suited to a future study as the transcriptomic analysis would likely include confounding effects of patient specific transcriptomic changes, and our primary focus was on developing an isogenic co-culture system.

Reviewer #2 (Public review):

Summary:

This study builds on work by Glass and Guilliams showing that mouse Kupffer cells depend on the surrounding cells, including endothelium, hepatocytes, and stellate cells, for their identity. Herein, the authors extend the work to human systems. It nicely highlights why taking monocyte-derived macrophages and pretending they are Kupffer cells is simply misleading.

Strengths:

Many, including human cells, difficult culture assays, and important new data.

Weaknesses:

This reviewer identified minor queries only, rather than 'weaknesses' as such.

Reviewer #3 (Public review):

Summary:

In this study, the authors establish a human in vitro liver model by co-culturing induced hepatocyte-like cells (iHEPs) with induced macrophages (iMACs). Through flow cytometry-based sorting of cell populations at days 3 and 7 of co-culture, followed by bulk RNA sequencing, they demonstrate that bidirectional interactions between these two cell types drive functional maturation. Specifically, the presence of iMACs accelerates the hepatic maturation program of iHEPs, while contact-dependent cues from iHEPs enhance the acquisition of Kupffer cell identity in iMACs, indicating that direct cell-cell interactions are critical for establishing tissue-resident macrophage characteristics.

Functionally, the authors show that iMAC-derived Kupffer-like cells respond to pathological stimuli by producing interleukin-6 (IL-6), a hallmark cytokine of hepatic immune activation. When exposed to a panel of clinically relevant hepatotoxic drugs, the co-culture system exhibited concentration-dependent modulation of IL-6 secretion consistent with reported drug-induced liver injury (DILI) phenotypes. Notably, this response was absent when hepatocytes were co-cultured with monocyte-derived macrophages from peripheral blood, underscoring the liver-specific phenotype and functional relevance of the iMAC-derived Kupffer-like cells. Collectively, the study proposes this co-culture platform as a more physiologically relevant model for interrogating macrophage-hepatocyte crosstalk and assessing immune-mediated hepatotoxicity in vitro.

Strengths:

A major strength of this study lies in its systematic dissection of cell-cell interactions within the co-culture system. By isolating each cell type following co-culture and performing comprehensive transcriptomic analyses, the authors provide direct evidence of bidirectional crosstalk between iMACs and iHEPs. The comparison with single-culture controls is particularly valuable, as it clearly demonstrates how co-culture enhances functional maturation and lineage-specific gene expression in both cell types. This approach allows for a more mechanistic understanding of how hepatocyte-macrophage interactions contribute to the acquisition of tissue-specific phenotypes.

Weaknesses:

(1) Overreliance on bulk RNA-seq data:

The primary evidence supporting cell maturation is derived from bulk RNA sequencing, which has inherent limitations in resolving heterogeneous cellular states and functional maturation. The conclusions regarding hepatocyte maturation are based largely on increased expression of a subset of CYP genes and decreased AFP levels - markers that, while suggestive, are insufficient on their own to substantiate functional maturation. Additional phenotypic or functional assays (e.g., metabolic activity, protein-level validation) would significantly strengthen these claims.

We have added a discussion on the limitations of our study.

(2) Insufficient characterization of input cell populations:

The manuscript lacks adequate validation of the cellular identities prior to co-culture. Although the authors reference previously published protocols for generating iHEPs and iMACs, it remains unclear whether the cells used in this study faithfully retain expected lineage characteristics. For example, hepatocyte preparations should be characterized by flow cytometry for ALB and AFP expression, while iMACs should be assessed for canonical macrophage markers such as CD45, CD11b, and CD14 before co-culture. Without these baseline data, it is difficult to interpret the magnitude or significance of any co-culture-induced changes.

We apologise for this oversight, some of the markers were used in determining the purity of the iMacs before co-culture, and we did not end up including these plots for brevity. We have added the purity plots in Supp Fig 2E now, showing that the iMacs were more than 90% pure before co-culture. We acknowledge the concern about cross-contamination for bulk sequencing, and have added in Supp Fig 2G and H the expression of ALB in the iMac fraction, as well as the expression of CSF1R in the iHep fraction, showing minimal contamination with our gating strategy.

(3) Quantitative assessment of IL-6 production is insufficient:

The analysis of drug-induced IL-6 responses is based primarily on relative changes compared to control conditions. However, percentage changes alone are inadequate to capture the biological relevance of these responses. Absolute cytokine production levels - particularly in response to LPS stimulation - should be reported and directly compared to PBMC-derived macrophages to determine whether iMAC-derived Kupffer-like cells exhibit enhanced cytokine output. Moreover, the Methods section should clearly describe how ELISA results were normalized or corrected to account for potential differences in cell number, viability, or culture conditions.

We apologise if this was unclear. The cytokine production from dosed cells was normalized based on the viability of cells measured from the same well.

(4) Unclear mechanistic interpretation of IL-6 modulation:

The observed changes in IL-6 production upon drug treatment cannot be interpreted solely as evidence of Kupffer cell-specific functionality. For instance, IL-6 suppression by NSAIDs such as diclofenac is well known to result from altered prostaglandin synthesis due to COX inhibition, while leflunomide's effects are linked to metabolite-induced modulation of immune cell proliferation and broader cytokine networks. These mechanisms are distinct from Kupffer cell identity and may not directly reflect liver-specific macrophage function. Consequently, changes in IL-6 secretion alone - particularly without additional mechanistic evidence or analysis of other cytokines - are insufficient to conclude that co-culture with hepatocytes drives the acquisition of bona fide Kupffer cell maturity.

We fully agree with the reviewer and have highlighted this in our discussion.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) GSE ID for RNA-seq data has not been provided.

This has been included.

(2) Line 291: Can the authors specify what they mean by "state-of-the-art"?

What we mean here is what others in the field have also recently described. We have rewritten this to be clearer.

(3) Lines 299-300: check sentence for grammar mistakes.

We have rewritten and clarified this.

(4) Figure 1B: The PCA does not really allow for following maturation trajectories. Also, all samples (day 3 Co-iHep, day 7 Co-iHep, day 7 iHep) look as if they cluster more or less together. Therefore, the conclusion drawn in lines 303-305 does not hold. Why is day 3 iHep not also shown here?

We agree that PCA does not allow for maturation trajectories and mentioned that it was a hypothesis that the co-culture was promoting maturation, which we later validated by looking at the expression of key hepatocyte markers as well as by pearson correlation comparison with fetal hepatocytes.

(5) Can the authors show that the cells that they are sorting in the double negative gate are indeed hepatocytes? Typically, these cells are big in cell size; therefore, showing the FSC/SSC gate would also be important.

We have added the FSC/SSC gate in supp fig. 1E to show that the populations have different sizes.

(6) Can the authors provide microscopy pictures of iHeps, iMacs, and the co-cultured cells for the reader to appreciate whether the morphology of cells already changes during the co-culture experiments?

Although we agree that the morphology changes would be interesting, we think that this question is unfortunately outside of the scope of our question. Although Kupffer cells are in direct contact with hepatocytes, they migrate from the liver parenchyma into the sinusoidal spaces where they primarily reside. We do not think that the morphology would add much to the paper, especially given that this is a 2D model as well.

(7) Please show expression of apoptotic and ER stress genes comparing Day7 iHeps and Co-iHeps, since genes such as c-Fos and Ppp2r3b can also be associated with cellular stress.

This has been included in Supp Fig 2C, where we’ve included the expression of ATF4, CASP3 and CASP9. Although there’s a significant difference in ATF4 expression between Day 0 and Day 7 iHep only/Co-culture, there is no significant difference between the Day 7 iHep only and Day 7 iHep Co-culture. There are no significant differences in CASP3 and CASP9 expression across all the samples.

(8) In addition to the genes shown in Figure 1E, could the authors extract a longer gene list of maturing hepatocytes and display them all in bar graphs or heatmaps, or similar? E.g., Albumin expression is shown later, but why not show it already here?

There are not many differences in the canonical hepatocyte markers, which is why we chose only to show the interesting genes that were different, as seen in the later ALB expression plot where there wasn’t a difference in ALB expression after 7 days of co-culture. Instead, we have included a new heatmap in Supp Fig 2B showing the top 40 genes that are contributing to the similarity by pearson correlation.

(9) Along these lines, how do the authors ensure that they are culturing only hepatocytes and do not have a mixture of cells that may "dilute" the hepatocyte signature?

Unfortunately, this is an limitation of our methodology, although the expression of key hepatic markers are routinely confirmed by qPCR to ensure that the majority of the cells are hepatocyte-like.

(10) Lines 347-350: similar to the interpretation of the PCA for hepatocytes, this is a completely random interpretation. The expression of ALB in the co-cultured iMacs indicates that there are some hepatocytes that ended up in the macrophage gate.

We agree and have highlighted and addressed this limitation in our discussion. Unfortunately, this is a limitation of bulk sequencing that a small amount of contamination might be present, however the TPM values of ALB for example in the iMacs is extremely low especially when compared to the hepatocytes, indicating that the level of contamination is likely to be very low. Likewise, the expression of CSF1R in the co-cultured iHeps is also extremely low. This has been included in Supp Fig 1F and G.

(11) Figure 2D: Among the pathways shown, there are also stress pathways (acute phase response, HMGB1). Also for these cells, control of apoptotic and ER stress signatures is necessary.

As mentioned, we have included some stress genes in Supp Fig 2C to address this.

(12) Lines 385-386: Why would FCGRA3 indicate tissue residency? Is there literature to support this statement?

CD16 is a marker often used to distinguish Kupffer cells from the surrounding cells, although it also expressed by non-classical monocytes, we have clarified the text here (Lines 356-357).

(13) Figure 3E: ALB and other genes were at the same or even lower levels expressed in D7 compared to D3. Why is that? Are the cells starting to de-differentiate after 7 days? Please discuss.

This is a very interesting question that we were wondering ourselves as well, although sadly we do not have an answer yet. We hypothesized that this might be due to the activation of cell proliferation/developmental programmes as the cells are kept longer together, as shown by the expression of morphogens like OSM and IGF-2 after co-culture. We have added some discussion for this (Lines 532-540)

(14) Line 459: Word "in" is double

We thank the reviewer for catching this, this has been corrected

(15) Figure 5: The findings are interesting, but the co-culture model remains somewhat unclear. Can the authors show, e.g., using qRT-PCR, how hepatocytes are developing in this culture system? If the development with monocyte-derived macrophages is altered, then one would expect that also the cellular response is different.

We agree with the reviewer, but we think that this question would be better answered in a follow-up study. We were looking to answer if the addition of isogenic iMacs would change the drug response of iHeps, and were using the PBMC-derived macrophages here as a control. A more complete study taking into account the genetic background of the donor PBMC-derived macrophages would be much more informative, but sadly outside of the scope of our present study.

(16) Lines 482-484: The authors talk about LPS-treated cultures and refer to Figure 4. However, there is no graph shown for LPS.

We apologise for being unclear here, but the co-cultures were co-treated with LPS during the drug stimulation assays, as it had been shown that LPS increases the sensitivity of the liver toward hepatotoxic drugs. We have clarified this in the main text (Lines 435-437).

Reviewer #2 (Recommendations for the authors):

(1) It would be nice to add some protein production by the hepatocytes. For example, can they produce albumin or some other protein that can be measured? Perhaps I missed this.

The protein expression of Albumin and Urea were assessed in the hepatocytes prior to co-culture in Supp Fig 1C; however we did not measure the protein level changes after co-culture as the co-culture would have a significant number of macrophages as well which we thought might affect the readout. Instead, after co-culture the primary analysis was done on the RNA levels of ALB and other cytochrome genes after sorting in Fig 3.

(2) Was there an increase in hepatocyte number? Did one cell outgrow the other, or did they maintain numbers?

The relative proportion of the iHeps remained the same, although we did see an expansion in the iMac population after 7 days by flow cytometry in Fig 1D.

(3) What happens if the iMACs and the iHeps are grown in Costar chambers with pore sizes too small to allow for cell contact, but allowing supernatant to be continuously exposed to both cell types?

We were primarily focused on the acquisition of KC-like phenotype in the iMacs with regards the question of direct contact, which was why we chose to use conditioned iHep media as part of the iMac experimental set up. However, it would be very interesting to see if the converse is also true, and whether secreted factors from the iMacs alone would be sufficient to drive the changes we observed in the iHeps after co-culture in a follow-up study.

(4) The discussion could use a brief paragraph on some limitations and what could be added to the co-culture system. For example, could stellate cells and sinusoidal endothelium also impart KC identity? Would growing KCs on endothelium provide a more natural substratum?

Once again, these are very interesting questions which are unfortunately outside of the scope of our study. However, we have included a short section discussing this in the paper, as we do think that it would be interesting to look at iMacs educated by hepatocyte vs stellate cells for example (Lines 530-536).

(5) The axonal guidance pathway in early iMACs is interesting. A recent report in vivo showed that macrophages migrate from the liver parenchyma into the sinusoids in neonates when they are still immature. The process could be chemotaxis, or it could be repulsion by parenchyma. Numerous axonal guidance molecules are repulsive, pushing axons away (robo/slit, etc). The migration of Kupffer cells into sinusoids could be a repulsive rather than a chemoattractant pathway. Did the RNA seq data provide any interesting molecules in this regard?

Reviewer #3 (Recommendations for the authors):

This manuscript presents a conceptually well-designed approach to modeling hepatocyte-macrophage crosstalk in vitro. The authors develop a co-culture system aimed at recapitulating key aspects of Kupffer cell (KC) identity and hepatocyte maturation. The data convincingly show that macrophages acquire KC-like features under co-culture conditions. However, several major issues limit the strength of the conclusions, the depth of mechanistic insight, and the translational impact of the work.

First, the study relies heavily on bulk RNA-seq data with minimal functional or protein-level validation - particularly for hepatocyte maturation. To substantiate claims of functional maturation, additional assays measuring albumin secretion, urea production, and CYP activity are essential. Furthermore, the omission of zonation-associated markers (e.g., GLUL, CPS1, CYP2E1) leaves a critical gap in assessing whether the iHEPs achieve physiologically relevant functional states.

Second, statistical interpretation and reporting are inconsistent. Significant and non-significant findings are frequently conflated, which risks overinterpretation. For instance, the reported reduction in HNF4A expression is not statistically significant, and AFP expression is only significantly reduced in Day 7 co-iHEPs - yet these distinctions are not clearly stated.

Third, although the authors emphasize the role of cell-cell contact in promoting KC identity, no experiments (e.g., transwell separation, adhesion-blocking assays) directly test this claim. As a result, the mechanistic basis for this conclusion remains speculative.

Finally, while the data support enhanced macrophage differentiation toward a KC-like phenotype, the evidence that co-culture significantly promotes hepatocyte maturation is far less convincing and requires additional functional, mechanistic, and statistical validation before firm conclusions can be drawn.

Minor comments:

(1) Methodology: The choice of a 2.5:1 iHEP:iMAC ratio is not justified. This proportion does not reflect physiological hepatocyte-to-KC ratios in vivo and should be either rationalized or benchmarked against native liver composition.

We admit that the ratio here is on the higher side of things, but it has been previously reported that there can be between 20 to 40 macrophages per 100 hepatocytes (1:5 to 1:2.5) in the adult mouse liver (Baratta et al., 2009), while admittedly in the developing mouse liver the ratio is closer to 1:4 (Lopez et al., 2011). We chose 1:2.5 as we anticipated that not all of the macrophages would be able to attach, and would thus be lost during media change, as evident by the flow cytometry of the co-culture on Day 3 of the co-culture, where only 20% of the cells had clear CD45 and CD14 expression. We have clarified our methodology in paper (Lines 141-143).

(2) Effect of iMAC on iHEP (Section 3.2, Supplementary Figure 1E):

(2.1) The authors should explain why Day 3 co-cultured iHEPs show stronger transcriptomic similarity to primary hepatocytes than Day 7 cells. Possible biological mechanisms (e.g., transient paracrine signaling or temporal changes in maturation dynamics) should be discussed.

We have added some discussion for this (Lines 309-311, 536-540).

(2.2) The figure legend refers to "fetal hepatocytes," while the correlation map states "hepatocytes." This discrepancy must be clarified. Moreover, if fetal hepatocytes are used as the reference, and the goal is to assess maturation, comparisons to adult hepatocytes are necessary. 

The comparison was done against fetal hepatocytes, and has been clarified in the figure. We chose to use fetal hepatocytes here as it would be unfair to compare iPSC-derived cells that are less than 3 weeks old to adult human tissue, and any similarity or differences between the mono/co-cultures to the adult tissue might be due to the shifting transcriptomic landscape during development. However, we do recognise the nuanced nature of using “maturation” here, and what we mean is that the iPSC-derived cells become more similar to their in-vivo counterparts.

(2.3) Baseline characterization of both cell types before co-culture is insufficient. For iHEPs, flow cytometry data on ALB and AFP positivity rates should be presented, along with post-co-culture changes. For iMACs, marker expression (CD45, CD11b, CD14) should be shown before and after co-culture. The methods mention CD163, CX3CR1, and CD11b, but these data are absent from the results. Additionally, the gating strategy for cell sorting prior to bulk RNA-seq must be clearly described - including how potential cross-contamination of cell fractions (e.g., macrophages in the hepatocyte population) was excluded.

We apologise for this oversight, some of the markers were used in determining the purity of the iMacs before co-culture, and we did not end up including these plots for brevity. We have added the purity plots in Supp Fig 2E now, showing that the iMacs were more than 90% pure before co-culture. We acknowledge the concern about cross-contamination for bulk sequencing, and have added in Supp Fig 2G and H the expression of ALB in the iMac fraction, as well as the expression of CSF1R in the iHep fraction, showing minimal contamination with our gating strategy.

(3) IGF2 Expression: The observed upregulation of IGF2, a fetal marker, contradicts the conclusion that co-culture promotes hepatocyte maturation. This inconsistency should be addressed, and possible explanations (e.g., transient fetal-like activation driven by macrophage-derived signals) discussed. The lack of statistical significance for this finding must also be explicitly noted.

We thank the reviewer for pointing this out. The expression of IGF2 was actually significantly different when comparing the Day 0 Hepatocyte only and Day 7 Hepatocyte only to the Day 3 Co-cultured Hepatocytes, but the significance is lost with the Day 7 co-cultured Hepatocytes. One possible explanation is as the reviewer suggested, that there is a transient program that is activated upon co-culture that is subsequently downregulated. We have updated the figure and text, and added some discussion to reflect this (Lines 309-311, 536-540).

(4) Effect of iHEP on iMAC: The reported upregulation of KC-related genes is overstated. Changes in LYVE1 and ID1 are not statistically significant (Figure 2G), yet they are presented as meaningful. Clear separation of statistically significant results from non-significant trends is critical to avoid overinterpretation.

We apologise for this, as it was never our intention to present these markers as significant, but rather we presented these markers because we thought that these markers would be of interest to the audience. We have clarified the text to reflect that these are trends and non-significant (Lines 367-369).

(5) Mimicking In Vivo Clinical Responses:

(5.1) The authors' conclusion that IL-6 responses are not recapitulated when iMACs are replaced by monocyte-derived macrophages (MoMs) is not fully supported by the data presented. In fact, the MoM co-cultures exhibit a noticeable trend toward increased IL-6 production (e.g., approximately 150% with LTG at 66.6 µM and 400 µM), suggesting that some degree of responsiveness is retained. To substantiate the claim that the observed cytokine modulation is unique to iKC-containing co-cultures, the authors should perform direct statistical comparisons of absolute IL-6 secretion levels between iKC and MoM co-cultures at each drug concentration. Such analyses are essential to determine whether the differences are statistically significant and biologically meaningful, and to clarify whether the observed effects truly reflect KC-specific functionality rather than general macrophage activation.

(5.2) The effects of drug exposure on hepatocytes themselves are not addressed. It is important to evaluate whether the co-culture remains viable under treatment, whether it recovers after drug withdrawal, and whether there is evidence of cytotoxicity or irreversible phenotypic loss.

(6) Interpretation of IL-6 Modulation and Model Specificity:

The authors show that IL-6 secretion in their co-culture system varies in response to multiple hepatotoxic drugs and parallels some reported clinical trends - notably, a concentration-dependent decrease with diclofenac (DIC) and leflunomide (LFM). They further report that this pattern is not observed in hepatocyte-PBMC-derived macrophage co-cultures, and they conclude that iMAC/iKC-like cells are essential for capturing immune-mediated hepatotoxic responses. However, the data presented do not fully justify such a conclusion. Several key mechanistic issues weaken the interpretation:

(6.1) Mechanistic ambiguity in the DIC response: The decrease in IL-6 following DIC exposure is most likely attributable to reduced prostaglandin E₂ (PGE₂) production via COX inhibition, which secondarily suppresses IL-6 signaling. This effect is a general pharmacological property of NSAIDs and is not necessarily reflective of Kupffer cell-specific pathways. Direct evidence - such as prostanoid quantification or PGE₂ rescue experiments - is required to establish that the observed effects are liver-specific rather than nonspecific NSAID responses.

(6.2) Pharmacogenetic complexity in the LFM response: LFM-induced hepatotoxicity is highly variable and largely dependent on CYP2C9 polymorphisms, which determine conversion to the active metabolite teriflunomide. Because hepatotoxicity and the associated cytokine responses are not universal among patients, a simplified co-culture model lacking metabolic diversity cannot be assumed to faithfully reproduce patient-specific immune responses. The observed IL-6 suppression could arise from differences in metabolic activation, intracellular exposure, or indirect signaling changes rather than from intrinsic KC-specific mechanisms.

These points significantly undermine the authors' claim that IL-6 modulation provides definitive evidence of model specificity or predictive value. At minimum, the manuscript should (i) explicitly acknowledge these mechanistic limitations, (ii) include supporting data such as prostanoid profiling, CYP2C9 modulation, or teriflunomide quantification, and (iii) temper its claims regarding the model's capacity to recapitulate immune-mediated hepatotoxicity. Without such evidence, the current interpretation risks overstating the functional significance and translational relevance of the co-culture system.

We fully agree with the reviewer and have highlighted this in our discussion (Lines 540 – 551).

eLife Assessment

The analysis of neural morphology across Heliconiini butterfly species revealed brain area-specific changes associated with new foraging behaviours. While the volume of the centre for learning and memory, the mushroom bodies, was known to vary widely across species, these new, valuable results show conservation of the volume of a center for navigation, the central complex, but with specific changes in neuropeptide expression in the noduli and in the numbers of ellipsoid body ring neurons. The presented evidence is convincing for both volumetric conservation in the central complex and fine neuroanatomical differences associated with pollen feeding, delivered by experimental approaches that are applicable to other insect species. This work will be of interest to evolutionary biologists, entomologists, and neuroscientists.

Reviewer #1 (Public review):

The authors previously reported that Heliconius, one genus of the Heliconiini butterflies, evolved to be efficient foragers to feed pollen of specific plants and have massively expanded mushroom bodies. Using the same image dataset, the authors segmented the central complex and associated brain regions and found that the volume of the central complex relative to the rest of brain are largely conserved across the Heliconiini butterflies. By performing immunostaining to label specific subset of neurons, the authors found several potential sites of evolutional divergence in the central complex neural circuits, including the numbers of GABAergic ellipsoid body ring neurons and the innervation patterns of Allatostatin A expressing neurons in the noduli. These neuroanatomical data will be helpful to guide the future studies to understand the evolution of the neural circuits for vector-based navigations.

Strength

The authors used sufficiently large scale of dataset from 307 individuals of 41 specifies of Heliconiini butterflies to solidify the quantitative conclusions, and present new microscopy data for fine neuroanatomical comparison of the central complex.

Weakness

(1) Although the figures display a concise summary of anatomical findings, it would be difficult for non-experts to learn from this manuscript to identify the same neuronal processes in the raw confocal stacks. It would be helpful to have instructive movies to show step by step guide for identifications of neurons of interests, segmentations and 3D visualizations (rotation) for several examples including ER neurons (to supplement texts in line 347-353) and Allatostatin A neurons.

(2) Related to (1), it was difficult for me to access if the data in Fig 7 support the author's conclusions that ER neuron number increased in Heliconius Melpomene. By my understanding, the resolution of this dataset isn't high enough to trace individual axons and therefore authors do not rule out that the portion of "ER ring neurons" in Heliconius may not innervate the ER, as stated in Line 635 "Importantly, we also found that some ER neurons bypass the ellipsoid body and give rise to dense branches within distinct layers in the fan-shaped body (ER-FB)". If they don't innervate the ellipsoid body, why are they named as "ER neurons"?

(3) Discussions around the line 577-584 requires the assumption that each ellipsoid body (EB) ring neuron typically arborise in a single microglomerulus to form largely one-to-one connection with TuBu neurons within the bulb (BU), and therefore the number of BU microglomeruli should provide an estimation of the number of ER neurons. Explain this key assumption or provide an alternative explanation.

(4) The details of antibody information are missing in the Key resource table. Instead of citing papers, list the catalogue numbers and identifier for commercially available antibodies, and describe the antigen and if they are monoclonal or polyclonal. Are antigens conserved across species?

(5) I did not understand why authors assume that foraging to feed on pollens is more difficult cognitive task than foraging to feed on nectars. Would it be possible that they are equality demanding tasks but pollen feeding allows Heliconius to pass more proteins and nucleic acids to their offsprings and therefore they can develop larger mushroom bodies?

Comments on revisions:

The authors fully addressed my concerns and significantly improved the accessibility of the manuscript.

Reviewer #2 (Public review):

Summary

In this study, Farnsworth et al. ask whether the previously established expansion of mushroom bodies in the pollen foraging Heliconius genus of Heliconiini butterflies co-evolved with adaptations in the central complex. Heliconius trap line foraging strategies to acquire pollen as a novel resource require advanced spatial memory mediated by larger mushroom bodies but the authors show that related navigation circuits in the central complex are highly conserved across the Heliconiini tribe, with a few interesting exceptions. Using general immunohistochemical stains and 3D reconstruction, the authors compared volumes of central complex regions and unlike the mushroom bodies, there was no evidence of expansion associated with pollen feeding. However, a second dataset of neuromodulator and neuropeptide antibody labeling reveal more subtle differences between pollen and non-pollen foragers and highlight sub-circuits that may mediate species-specific differences in behavior. Specifically, the authors found an expansion of GABAergic ER neurons projecting to the fan shaped body in Heliconius which may enhance their ability to path-integrate. They also found differences in Allatostatin A immunoreactivity, particularly increased expression in the noduli associated with pollen feeding. These differences warrant closer examination in future studies to determine their functional implication on navigation and foraging behaviors.

Strengths

The authors leveraged a large morphological data set from the Heliconiini to achieve excellent phylogenetic coverage across the tribe with 41 species represented. Their high quality histology resolves anatomical details to the level of specific, identifiable tracts and cell body clusters. They revealed differences at a circuit level, which would not be obvious from a volumetric comparison. The discussion of these adaptations in the context of central complex models is useful for generating new hypotheses for future studies on the function of ER-FB neurons and the role of Allatostatin A modulation in navigation.<br /> The conclusions drawn in this paper are measured and supported by rigorous statistics and evidence from micrographs.

Weaknesses

The majority of results in this study do not reveal adaptations in the central complex associated with pollen foraging. However, reporting conserved traits is useful and illustrates where developmental or functional constraints may be acting. The authors have now revised the introduction to set up two alternate hypotheses..

In the main text, the authors describe differences in GABAergic ER neurons between H. melpomene and an outgroup species, with additional images from other species in Figure S4. Quantification of ER cells in these other species would strengthen the claim that these are increased in Heliconius and not just the focal species, but this may hopefully be pursued in future studies.

Comments on revisions:

I am satisfied with the authors' revisions.

Author response:

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

eLife Assessment

The analysis of neural morphology across Heliconiini butterfly species revealed brain area specific changes associated with new foraging behaviours. While the volume of the centre for learning and memory, the mushroom bodies, was known to vary widely across species, new, valuable results show conservation of the volume of a center for navigation, the central complex. The presented evidence is convincing for both volumetric conservation in the central complex and fine neuroanatomical differences associated with pollen feeding, delivered by experimental approaches that are applicable to other insect species. This work will be of interest to evolutionary biologists, entomologists, and neuroscientists.

Many thanks for your assessment and time handling this manuscript. We value the constructive input of both reviewers and believe that the result is an improved publication.

Public Reviews:

Reviewer #1 (Public review):

The authors previously reported that Heliconius, one genus of the Heliconiini butterflies, evolved to be efficient foragers to feed pollen of specific plants and have massively expanded mushroom bodies. Using the same image dataset, the authors segmented the central complex and associated brain regions and found that the volume of the central complex relative to the rest of the brain is largely conserved across the Heliconiini butterflies. By performing immunostaining to label a specific subset of neurons, the authors found several potential sites of evolutionary divergence in the central complex neural circuits, including the number of GABAergic ellipsoid body ring neurons and the innervation patterns of Allatostatin A expressing neurons in the noduli. These neuroanatomical data will be helpful to guide future studies to understand the evolution of the neural circuits for vector-based navigation.

We thank Reviewer 1 for the constructive feedback and criticism, which will have strengthened this publication.

Strengths:

The authors used a sufficiently large scale of dataset from 307 individuals of 41 species of Heliconiini butterflies to solidify the quantitative conclusions and present new microscopy data for fine neuroanatomical comparison of the central complex.

Weaknesses:

(1) Although the figures display a concise summary of anatomical findings, it would be difficult for non-experts to learn from this manuscript to identify the same neuronal processes in the raw confocal stacks. It would be helpful to have instructive movies to show a step-by-step guide for identification of neurons of interest, segmentations, and 3D visualizations (rotation) for several examples, including ER neurons (to supplement texts in line 347-353) and Allatostatin A neurons.

We approached this with the following logic:

All 3D segmentations were animated, to illustrate how they are generated from raw imaging data. This means we are providing a video file for each major species group (Heliconius/outgroup-Heliconiini) for Figure 4 (general CX anatomy), Figure 7 (ER neuron projections), Figure S5 (ER neuron/bulb anatomy). This visual connection should help the reader relate 3D segmentations to image stacks. We have also added a reference to these videos in the relevant Figure captions.

We also annotated image stacks, but did so selectively. We annotated key stacks of Figure 4 (general CX anatomy), Figure 7 (ER neuron projections), Figure S5 (ER neuron/bulb anatomy) and include a reference in figure caption to them.

We refrained from annotating stacks of Figures 5, 6, 8 and S4. This is because we believe that the annotations we have performed in the figure panels will be sufficient for readers interested in the finer detail of these anatomies who are familiar with general CX anatomy.

We believe that our approach will help the reader to gain a visual illustration of those parts of the manuscript which report key results and novel insights, such as ER neuronal variation, and that the data and figures collectively provide accessible information sufficient for this purpose.

Text changes in Figure captions 4, 7 and S5: “See animated 3D segmentations and annotated stacks in file repository.”

(2) Related to (1), it was difficult for me to assess if the data in Figure 7 support the author's conclusions that ER neuron number increased in Heliconius Melpomene. By my understanding, the resolution of this dataset isn't high enough to trace individual axons and therefore authors do not rule out that the portion of "ER ring neurons" in Heliconius may not innervate the ER, as stated in Line 635 "Importantly, we also found that some ER neurons bypass the ellipsoid body and give rise to dense branches within distinct layers in the fan-shaped body (ER-FB)". If they don't innervate the ellipsoid body, why are they named as "ER neurons"?

Thanks for pointing to this. We believe this is primarily a nomenclature issue but have tried to specify in the text.

Ultimately, neurons from this group that project to the EB forming the actual ring neurons and those that project to the FB with unclear function, thus far, emerge through the same lineage, DALv2 (as determined by Kandimalla et al 2023) and therefore have common developmental origin (also noted by Homberg et al 2018). To acknowledge their common developmental origin and to simplify nomenclature, and therefore also provide easier comprehension by non-experts, we specify which DALv2 progeny project to which areas, but refer to both adult neuron populations to “ER neurons”. We have changed the following text to acknowledge our definition specifically, which we hope mitigates the understandable confusion.

Lines 354-357: “Here, we refer to these neurons, as well as those neurons projecting to the fan-shaped body (GU neurons in [66]), as ER neurons due to their common developmental origin [45,66] and to simplify anatomical descriptions.”

Lines 386-387: “Whether these ER neurons solely branch in the fan-shaped body, as shown for GU neurons elsewhere [66] or have additional side branches entering the ellipsoid body is not clear.”

(3) Discussions around the lines 577-584 require the assumption that each ellipsoid body (EB) ring neuron typically arborises in a single microglomerulus to form a largely one-to-one connection with TuBu neurons within the bulb (BU), and therefore, the number of BU microglomeruli should provide an estimation of the number of ER neurons. Explain this key assumption or provide an alternative explanation.

Thanks for this. We do not think that our hypothesis necessarily requires any specific assumptions regarding the ratio of microglomerulus to ER or TuBu neurons. Even in Drosophila the ratio of ER to MG is only approximately 1:1, as some microglomeruli seem to combine into one. In other species this relationship might be very different. Indeed, our data suggests that in outgroup-Heliconiini the ratio is 4.4 microglomeruli to 1 ER neuron, and in Heliconius it is 3.4. However, as these MG numbers are extrapolated and cannot be precisely counted, they may be too imprecise to come to a definite conclusion, hence why we do not mention this in the text. Importantly, extrapolation in the current form is a valid additional way for us to describe overall bulb anatomy (next to bulb volume, average microglomerulus size).

In any case, the inference we make here is that a conserved bulb anatomy in volume, MG numbers and size supports our assumption that the additional neurons in the ER neuron group/DALv2 progeny do not arborize in the bulb, but do so in the SMP/SLP region and in the fanshaped body. We believe we have described this inference accurately in the current manuscript.

An additional point, not mentioned in the manuscript, but emerging through lineage annotations of connectome data, is that some DALv2 progeny have been identified as MBONs as well as being GABA-ergic, which could potentially be the ER-FB neurons that we describe (Schlegel et al 2024 Nature). We refrain from mentioning this here, as its too speculatory, but we thought the reviewer may be interested in this observation.

(4) The details of antibody information are missing in the Key resource table. Instead of citing papers, list the catalogue numbers and identifier for commercially available antibodies, and describe the antigen, and whether they are monoclonal or polyclonal. Are antigens conserved across species?

We have now added substantial information to Table 2, including research resource identifiers (RRIDs) and antigen descriptions, as well as information about specificity and conservation. In the text itself, in line 757, we already provide publications that have illustrated conservation very extensively.

We believe that with the additional information provided in Table 2, all necessary information is now provided.

(5) I did not understand why authors assume that foraging to feed on pollens is a more difficult cognitive task than foraging to feed on nectar. Would it be possible that they are equally demanding tasks, but pollen feeding allows Heliconius to pass more proteins and nucleic acids to their offspring and therefore they can develop larger mushroom bodies?

This is an excellent point. Our current understanding is that pollen feeding is a cognitively more demanding task, because, a) the density of pollen resources is lower than nectar resources, and b) the competition for pollen is higher (pollen is depleted quickly, and Heliconius compete with each other, and other taxa including hummingbirds). There is therefore a benefit to high foraging efficiency, which favours the evolution of learning. This is likely reinforced by the long lives of Heliconius which live up to a year, compared to ~4 weeks for most outgroups and the temporal stability of major pollen resources, resulting in a memorised location providing benefit for the long periods of time (Young and Montgomery 2020 Proc B).

We now refer to an additional publication (Young and Montgomery 2020 Proc B) in lines 103-104 for a fuller description of the ecology of pollen feeding, and in the current manuscript simply focus on the impact of mushroom body expansion on the CX.

Reviewer #2 (Public review):

Summary:

In this study, Farnsworth et al. ask whether the previously established expansion of mushroom bodies in the pollen foraging Heliconius genus of Heliconiini butterflies co-evolved with adaptations in the central complex. Heliconius trap line foraging strategies to acquire pollen as a novel resource require advanced spatial memory mediated by larger mushroom bodies, but the authors show that related navigation circuits in the central complex are highly conserved across the Heliconiini tribe, with a few interesting exceptions. Using general immunohistochemical stains and 3D reconstruction, the authors compared volumes of central complex regions, and unlike the mushroom bodies, there was no evidence of expansion associated with pollen feeding. However, a second dataset of neuromodulator and neuropeptide antibody labeling reveals more subtle differences between pollen and non-pollen foragers and highlights sub-circuits that may mediate species-specific differences in behavior. Specifically, the authors found an expansion of GABAergic ER neurons projecting to the fanshaped body in Heliconius, which may enhance their ability to path-integrate. They also found differences in Allatostatin A immunoreactivity, particularly increased expression in the noduli associated with pollen feeding. These differences warrant closer examination in future studies to determine their functional implication on navigation and foraging behaviors.

We thank Reviewer 2 for the constructive and thorough review. We believe that addressing these criticisms will have improved this publication.

Strengths:

The authors leveraged a large morphological data set from the Heliconiini to achieve excellent phylogenetic coverage across the tribe with 41 species represented. Their high-quality histology resolves anatomical details to the level of specific, identifiable tracts and cell body clusters. They revealed differences at a circuit level, which would not be obvious from a volumetric comparison. The discussion of these adaptations in the context of central complex models is useful for generating new hypotheses for future studies on the function of ER-FB neurons and the role of Allatostatin A modulation in navigation.

The conclusions drawn in this paper are measured and supported by rigorous statistics and evidence from micrographs.

Weaknesses:

The majority of results in this study do not reveal adaptations in the central complex associated with pollen foraging. However, reporting conserved traits is useful and illustrates where developmental or functional constraints may be acting. The implied hypothesis in the introduction is that expansion of mushroom bodies in Heliconius co-evolved with central complex adaptations, so it may be helpful to set up the alternate hypotheses in the beginning.

Thank you for this relevant comment. We have added to the text in lines 124-128, as follows

“Indeed, these circumstances permit us to test the hypotheses that modifications in the mushroom bodies either occurred in isolation from other integrative centres, or that they occurred in concert with specific changes in centres, such as the central complex. This provides insights into the functional flexibility of two interacting, integrative centres across evolutionary time.”

In the main text, the authors describe differences in GABAergic neurons "across several species" but only one Heliconius and one outgroup species seem to be represented in the figures. ER numbers in Figure 7H are only compared for these two species. If this data is available for other species, it would strengthen the paper to add them to the analysis, since this was one of the most intriguing findings in the study. I would want to know if the increased ER number is a trend in Heliconius or specific to H. melpomene.

This points to imprecise phrasing. We indeed have additional data in other species, but unfortunately not to an extent that would permit quantification of cell numbers, which is why we chose to put these data into the supplement, Fig. S4.

We modified the text to more directly point at the additional data in Fig S4, now reading in lines 362-368

“…, we noticed a pronounced difference in a portion of projections leading into the fan-shaped body and a strong difference in signal inside layer III in our two focal species H. Melpomene and D. iulia, as well as other representatives of the Heliconiini tribe (Figure S4A-B, Figure 7). To understand how these differences could have occurred, we quantified ER neuron numbers in our focal species, and identified a significant difference, reflecting a 35% increase in Heliconius (t = 4.221, P = 0.004; Figure 7H).”

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) Add a detailed description about each of the tiff files that were deposited at https://doi.org/10.5281/zenodo.15304965. It was hard for me to relate these raw images with the Figure panels. For instance, "Melp_GAD_26-F_detailed_conc.tif" in the Figure 7 folder seems to be used to make Figure 7L and N, but that information is cryptic.

We agree with the reviewer. We added further descriptions, and have created a detailed readme file which explains which original file refers to which figure. Together with the efforts for Reviewer 1’s first comment, we hope that this updated version of our repository is easier to understand.

In addition, we made additional changes in image orientation in some of the files supplied, and which were originally incorrect.

(2) Add descriptions about the dataset for large-scale volumetric analysis. With the current methods and texts, it is hard to understand what kinds of staining and microscopes were used. I initially thought that they could be micro-CT data.

We have made two improvements:

We have added an additional readme file to explain the different datasets, and which datasets were used for each figure, to relate them to the original data deposited at zenodo.org (see your previous comment).

We have added descriptions in several places in the manuscript file, i.e.

Lines 133-135, now reading “To assess evidence of volumetric changes in the central complex and associated neuropils, we drew data from a large dataset of immunostained brains from 307 individuals of 41 species, …”

Lines 144-149, now reading “We used a combination of phylogenetic comparative analysis across a large dataset of brains immunostained against the structural marker synapsin in 41 species and 307 individuals, and more targeted sampling of species that represent the behavioural and neuroanatomical diversity of Heliconiini for more fine-scale assessments of patterns of divergence in substructures of the CX with various antibodies (Figure 1A-B).”

(3) Line 275: Non-expert readers would need an explanation about what the gamma lobe is.

Agreed and added in line 273

“Some of the ventral projections seemed to directly originate from the γ lobe, a portion of the mushroom body, thus potentially labelling projections of mushroom body output neurons into the fan-shaped body (Figure 5a-c) [12,21].”

(4) Figures 4 I-L are missing.

We modified the figure caption accordingly, and address annotated differences more directly. This section now reads

“G/H: Labelling reveals two distinguishable layers in the fan-shaped body while additional staining elsewhere reveals further detail (arrows in G/H-2/3). Thicker tract conflations indicate the columnar architecture determined through the four columnar neuron bundles (arrowheads in G/H-3). Labelling in the EB reveals two pronounced layers (arrows in G/H-1/2), while obvious columns could not be indicated. PB protocerebral bridge, FB fan-shaped body, EB ellipsoid body. A anterior, P posterior. Scale bars are 50 μm.”

(5) In the current version of Figure 1B, AOTU is displayed with the mushroom body. The authors can emphasize its relation to the central complex by showing it on the right side of panels together with the central complex.

Great suggestion. We have done this now. We have kept the AOTU at the scale of the MB, indicated by the different scale bars of the bottom of the figure, as we’re showing the CX at a slightly larger scale.

(6) Figure 1C: What do the colors of the lines represent?

We now changed these colours so that they correspond to the colours chosen in Figures 2 and S2 as well as in a previous publication of the lab, added an asterisk next to Heliconius aoede, and added text to the figure legend:

“Colour indicates focal groups here and elsewhere [29]. The asterisk at the branch of H. aoede indicates a secondary loss of pollen feeding.”

(7) Figures 2A and B: What does the size of the circles represent? I guess that small ones are individuals, and larger ones are species averages. Plots with only species averages would be easier to see. It is difficult to distinguish Heliconius and Helicononius aoede in these panels. It would be easier if Heliconius circles were outlined with thin black lines. 

Thanks for this. We wanted to keep both the averages and individual data points in one figure, as to not overcrowd the manuscript with additional figures. We still hope that the changes we made address the confusion sufficiently. We made the following modifications to Figure 2 and S1 and S2:

(1) Added text in the figure legend clarifying what solid and transparent circles indicate (“Solid data points indicate species averages, while opaque circles indicate individual data points.”)

(2) Added, as suggested, additional contours, to all Heliconius data points, and added corresponding text to the legend (“Black contours indicate Heliconius sp. data points.”)

(3) Changed opacity settings of individual data points.

Reviewer #2 (Recommendations for the authors):

(1) Line 391 and Methods. It was unclear how the extrapolated microglomeruli numbers were calculated. Please clarify this in the methods.

Agreed. We substantially modified the text to address this.

Lines 392-396: “We generated high resolution images of the bulb to determine its size (Figure S5 C-F), and 3D segmented seven microglomeruli per individual with which we generated an extrapolated approximation of total microglomeruli number by dividing bulb volume with average microglomerulus volume. This was necessary as most microglomeruli were not discernible from each other (Figure S5 G-H).”

Lines 862-873: “To segment the bulb, we created high resolution images and were particularly careful to only segment the area of the bulb that comprised large synapses/glomeruli, excluding parts of the LEa/IT projection. This was essential, because we relied on extrapolating the total number of microglomeruli from a subset of segmented microglomeruli and the total volume that contained microglomeruli, which means any section containing tracts and not glomerular structures would skew the estimated total number of microglomeruli. Extrapolation was necessary, as not all microglomeruli were visually discernible. We achieved an unskewed bulb volume by leaving out dense pieces of tubulin-positive tract material. We segmented seven microglomeruli per individual from the posterior section of the bulb, where they were most clearly visible, to get the most comparable impression across individuals and species. We then calculated average microglomerulus size and divided this by bulb volume to determine an approximation of microglomeruli number.”

(2) Line 439. It would be helpful to add that Kaiser et al. studied honeybees.

Agreed! Now reads in lines 443-444

“Moreover, Kaiser et al. [75] identified Allatostatin A expression in three fan-shaped and two ellipsoid body layers in the honey bee brain, …”

(3) Line 492. "outcome" should be "outcomes".

We believe that this refers to original line 481. Corrected. Thank you.

(4) Figure 3B. If there is significance to the colors and triangle directions, please include a key/legend.

We have added:

“Cell type depictions are examples with localisation inside each neuropil being purely visual (as well as their colour), while triangles indicate approximate output sites.”

We also corrected the following issues that were noted during our revisions:

line 587, wrong reference.

We updated references 37 and 44, which are now respectively

Hodge, E. A. et al. Modality-specific long-term memory enhancement in Heliconius butterflies. Philos Trans R Soc Lond B Biol Sci 380, 20240119 (2025).

Hodge, E. A. et al. Conservation of sensory pathways implies a localised change in the mushroom bodies is associated with cognitive evolution in Heliconius butterflies. Evol qpag005 (2026) doi:10.1093/evolut/qpag005.

Figure S5 had an error in panels C and D, where the pictures in C were actually for H. Melpomene in D and the reverse; the other panels were correct. We have corrected this.

In the data submitted on Zenodo: we corrected a few inconsistencies in channel colours and orientation in the .tiff files for Fig 6, 8 and S4.

We added important bulb 3D segmentation files to the repository on Zenodo.

eLife Assessment

This valuable study introduces miRTarDS, a novel computational framework that predicts microRNA-target interactions based on a publicly available pretrained Sentence-BERT language model and downstream classification analysis. The strength of the evidence is incomplete, as the evaluation framework relies on unreliable ground-truth and false sets. Furthermore, the analysis fails to compare miRTarDS against existing state-of-the-art biomedical language models.

Reviewer #1 (Public review):

The author presents a new method for microRNA target prediction based on (1) a publicly available pretrained Sentence-BERT language model that the author fine-tunes using MeSH information and (2) downstream classification analysis for microRNA target prediction. In particular, the author's approach, named "miRTarDS", attempts to solve the microRNA target prediction problem by utilizing disease information (i.e., semantic similarity scores) from their language model. The author then compares the prediction performance with other sequence- and disease-based methods and attempts to show that miRTarDS is superior or at least comparable to existing methods. The author's general approach to this microRNA target prediction problem seems promising, but fails to demonstrate concrete computational evidence that miRTarDS outperforms other existing methods. The author's claim that disease information-based language models are sufficient is unfounded. The manuscript requires substantial rewriting and reorganization for readers with a strong background in biomedical research.

A major issue related to the author's claim of computational advance of miRTarDS: The author does not introduce existing biomedical-specific language models, and does not compare them against miRTarDS's fine-tuned model. The performance of miRTarDS is largely dependent on the semantic embedding of disease terms. The author shows in Figure 5 that MeSH-based fine-tuning leads to a substantial improvement in MeSH-based correlation compared to the publicly available pretrained SBERT model "multi-qa-MiniLM-L6-cos-v1" without sacrificing a large amount of BIOSSES-based correlation. However, the author does not compare the performance of MeSH- and BIOSSES-based correlation with existing language models such as ChatGPT, BioBERT, PubMedBERT, and more. Also, the substantial improvement in MeSH-based correlation is a mere indication that the MeSH-based fine-tuning strategy was reasonable and not that it's superior to the publicly available pretrained SBERT model "multi-qa-MiniLM-L6-cos-v1".

Another major issue is in the author's claim that disease-information from miRTarDS's language model is "sufficient" for accurate microRNA target prediction. Available microRNA targets with experimental evidence are largely biased for those with disease implications that have been reported in the biomedical literature. It's possible that their language model is biased by existing literature that has also been used to build microRNA target databases. Therefore, it is important that the author provides strong evidence that excludes the possibility of data leakage circularity. Similar concerns are prevalent across the manuscript, and so I highly recommend that the author reassess the evaluation frameworks and account for inflated performance, biased conclusions, and self-confirming results.

Last but not least, the manuscript requires a deeper and careful description and computational encoding of microRNA biology. I'd advise the author to include an expert in microRNA biology to improve the quality of this manuscript. For example, the author uses the pre-miRNA notation and replaces the mature miRNA notation to maintain computational encoding consistency across databases. However, the mature microRNA notation "the '-3p' or '-5p' is critical as the 3p and 5p mature microRNAs have different seed sequences and thus different mRNA targets. The 3p mature microRNA would most likely not target an mRNA targeted by the 5p mature microRNA.

Reviewer #2 (Public review):

Summary:

This study introduces a novel knowledge-driven approach, miRTarDS, which enables microRNA-Target Interaction (MTI) prediction by leveraging the disease association degree between a miRNA and its target gene. The core hypothesis is that this single feature is sufficient to distinguish experimentally validated functional MTIs from computationally predicted MTIs in a binary classification setting. To quantify the disease association, the authors fine-tuned a Sentence-BERT (SBERT) model to generate embeddings of disease descriptions and compute their semantic similarity. Using only this disease association feature, miRTarDS achieved an F1 score of 0.88 on the test set.

Strengths:

The primary strength is the innovative use of the disease association degree as an independent feature for MTI classification. In addition, this study successfully adapts and fine-tunes the Sentence-BERT (SBERT) model to quantify the semantic similarity between biomedical texts (disease descriptions). This approach establishes a critical pathway for integrating powerful language models and the vast growth in clinical/disease data into biochemical discovery, like MTI prediction.

Weaknesses:

The main weakness lies in its definition of the ground-truth dataset, which serves as a foundation for methodological evaluation. The study defines the Negative Set as computationally predicted MTIs that lack experimental evidence. However, the absence of experimental validation does not equate to non-functionality. Similarly, the miRAW sets are classified by whether the target and miRNA could form a stable duplex structure according to RNA structure prediction. This definition is biologically irrelevant, as duplex stability does not fully encapsulate the complex in vivo binding of miRNAs within the AGO protein complex.

Author response:

We would like to express our sincere gratitude to the editors and the two reviewers for providing their constructive and valuable comments that will greatly guide us in improving the manuscript. We will revise the manuscript according to their critiques and suggestions. The existing code for this study, along with preliminary code developed in response to the review comments, has been made publicly available at https://github.com/cbaiming/miRTarDS. We now provide detailed responses to each reviewer below.

Reviewer #1 (Public review):

The author presents a new method for microRNA target prediction based on (1) a publicly available pretrained Sentence-BERT language model that the author fine-tunes using MeSH information and (2) downstream classification analysis for microRNA target prediction. In particular, the author's approach, named "miRTarDS", attempts to solve the microRNA target prediction problem by utilizing disease information (i.e., semantic similarity scores) from their language model. The author then compares the prediction performance with other sequence- and disease-based methods and attempts to show that miRTarDS is superior or at least comparable to existing methods. The author's general approach to this microRNA target prediction problem seems promising, but fails to demonstrate concrete computational evidence that miRTarDS outperforms other existing methods. The author's claim that disease information-based language models are sufficient is unfounded. The manuscript requires substantial rewriting and reorganization for readers with a strong background in biomedical research.

We appreciate the reviewer’s careful examination of modeling, benchmarking, and interpretation, and we are particularly encouraged that they found the proposed method promising. We will make corresponding revisions to the manuscript based on the reviewer’s comments.

A major issue related to the author's claim of computational advance of miRTarDS: The author does not introduce existing biomedical-specific language models, and does not compare them against miRTarDS's fine-tuned model. The performance of miRTarDS is largely dependent on the semantic embedding of disease terms. The author shows in Figure 5 that MeSH-based fine-tuning leads to a substantial improvement in MeSH-based correlation compared to the publicly available pretrained SBERT model "multi-qa-MiniLM-L6-cos-v1" without sacrificing a large amount of BIOSSES-based correlation. However, the author does not compare the performance of MeSH- and BIOSSES-based correlation with existing language models such as ChatGPT, BioBERT, PubMedBERT, and more. Also, the substantial improvement in MeSH-based correlation is a mere indication that the MeSH-based fine-tuning strategy was reasonable and not that it's superior to the publicly available pretrained SBERT model "multi-qa-MiniLM-L6-cos-v1".

We thank the reviewer for the constructive suggestions regarding the benchmarking of language models. We acknowledge that the performance of miRTarDS largely depends on the semantic embeddings of disease terms. So, in the revisions, I will: 1) conduct a literature review to introduce existing biomedical-specific language models, and 2) perform a horizontal comparison between our fine-tuned model and these existing models, to more comprehensively evaluate the model’s capabilities.

Another major issue is in the author's claim that disease-information from miRTarDS's language model is "sufficient" for accurate microRNA target prediction. Available microRNA targets with experimental evidence are largely biased for those with disease implications that have been reported in the biomedical literature. It's possible that their language model is biased by existing literature that has also been used to build microRNA target databases. Therefore, it is important that the author provides strong evidence that excludes the possibility of data leakage circularity. Similar concerns are prevalent across the manuscript, and so I highly recommend that the author reassess the evaluation frameworks and account for inflated performance, biased conclusions, and self-confirming results.

We thank the reviewer for the comment. We recognize that existing experimentally validated microRNA targets may be biased toward those reported in biomedical literature as disease‑related. To mitigate this bias, we attempted to extract predicted microRNA targets that share a very similar number of miRNA- and gene‑ disease entries as the experimentally validated microRNA targets using the K‑Nearest Neighbors (KNN) method. Then applied Positive‑Unlabeled (PU) Learning to classify the two groups. PU‑Learning is designed to address scenarios where only a subset of the training data is explicitly labeled as positive, while the remaining data are unlabeled—with the unlabeled set containing both potential positives and true negatives—which is highly suitable for the application context of this manuscript [1]. Preliminary results show that after applying the new data extraction and classification approach, model performance drops to around F1=0.73 (the MISIM method also shows a decline, with F1 around 0.58; detailed code is available on GitHub). The specific reasons for this require further investigation.

Last but not least, the manuscript requires a deeper and careful description and computational encoding of microRNA biology. I'd advise the author to include an expert in microRNA biology to improve the quality of this manuscript. For example, the author uses the pre-miRNA notation and replaces the mature miRNA notation to maintain computational encoding consistency across databases. However, the mature microRNA notation "the '-3p' or '-5p' is critical as the 3p and 5p mature microRNAs have different seed sequences and thus different mRNA targets. The 3p mature microRNA would most likely not target an mRNA targeted by the 5p mature microRNA.

We thank the reviewer for the critique and suggestion. We fully agree with the reviewer that the distinction between the 3p and 5p mature strands is critical for determining mRNA targeting, as they possess distinct seed sequences. In our study, we relied on the miRNA–disease associations provided by the HMDD database, which annotates interactions at the pre-miRNA level: “… the enriched functions of each mature miRNA are aggregated to the corresponding miRNA precursor.” [2] Furthermore, existing literature suggests that the pre-miRNA level can be appropriate and informative for disease association analyses: “Compared with the mature miRNA method, the pre-miRNA method is more useful for studying disease association.” [3] We also find that, in some cases, both strands cooperate to regulate the same or complementary pathways [4]. We acknowledge the reviewer’s point as an important consideration for future revision. We plan to consult or collaborate with biologists to enhance the quality of the manuscript in biology.

Reviewer #2 (Public review):

This study introduces a novel knowledge-driven approach, miRTarDS, which enables microRNA-Target Interaction (MTI) prediction by leveraging the disease association degree between a miRNA and its target gene. The core hypothesis is that this single feature is sufficient to distinguish experimentally validated functional MTIs from computationally predicted MTIs in a binary classification setting. To quantify the disease association, the authors fine-tuned a Sentence-BERT (SBERT) model to generate embeddings of disease descriptions and compute their semantic similarity. Using only this disease association feature, miRTarDS achieved an F1 score of 0.88 on the test set.

We thank the reviewers for their positive feedback, especially for their recognition of the novelty of this manuscript.

Strengths:

The primary strength is the innovative use of the disease association degree as an independent feature for MTI classification. In addition, this study successfully adapts and fine-tunes the Sentence-BERT (SBERT) model to quantify the semantic similarity between biomedical texts (disease descriptions). This approach establishes a critical pathway for integrating powerful language models and the vast growth in clinical/disease data into biochemical discovery, like MTI prediction.

We would like to thank the reviewer again for their positive feedback. We appreciate their recognition of the novelty of our work, as well as their acknowledgment that the proposed method paves the way for integrating language models with clinical/disease data into biochemical discovery.

Weaknesses:

The main weakness lies in its definition of the ground-truth dataset, which serves as a foundation for methodological evaluation. The study defines the Negative Set as computationally predicted MTIs that lack experimental evidence. However, the absence of experimental validation does not equate to non-functionality. Similarly, the miRAW sets are classified by whether the target and miRNA could form a stable duplex structure according to RNA structure prediction. This definition is biologically irrelevant, as duplex stability does not fully encapsulate the complex in vivo binding of miRNAs within the AGO protein complex.

We thank the reviewers for their constructive feedback. We have realized that treating predicted MTI as a negative class may pose some issues. Therefore, we have decided to adopt Positive Unlabeled (PU) Learning in subsequent updates. This classification method can be applied to datasets such as ours, which contain only positive classes and lack negative ones [1]. We used the miRAW dataset to enable a horizontal comparison of our method with traditional sequence-based prediction approaches. We acknowledge that miRAW may overlook some biological insights, and we plan to optimize the construction of test datasets in the future. Some preliminary explorations have already been conducted, and the relevant code is available on GitHub.

Furthermore, we will make the following revisions: 1) We will clearly specify the version of miRBase and incorporate more miRNA-related databases. 2) Conduct a further literature review on miRNA biological mechanisms to enhance the quality of the manuscript in biology. 3) Perform a more comprehensive evaluation of the model’s performance. 4) Attempt to identify some representative MTIs that have been overlooked by existing prediction tools but can be predicted by our proposed method.

References

(1) Li, F., Dong, S., Leier, A., Han, M., Guo, X., Xu, J., ... & Song, J. (2022). Positive-unlabeled learning in bioinformatics and computational biology: a brief review. Briefings in Bioinformatics, 23(1), bbab461.

(2) Huang, Z., Shi, J., Gao, Y., Cui, C., Zhang, S., Li, J., ... & Cui, Q. (2019). HMDD v3. 0: a database for experimentally supported human microRNA–disease associations. Nucleic acids research, 47(D1), D1013-D1017.

(3) Wang, H., & Ho, C. (2023). The human pre-miRNA distance distribution for exploring disease association. International Journal of Molecular Sciences, 24(2), 1009.

(4) Mitra, R., Adams, C. M., Jiang, W., Greenawalt, E., & Eischen, C. M. (2020). Pan-cancer analysis reveals cooperativity of both strands of microRNA that regulate tumorigenesis and patient survival. Nature Communications, 11(1), 968.

eLife Assessment

This is an important paper that reports in vivo physiological abnormalities in the hippocampus of a rat model of traumatic brain injury (TBI). In this study, authors focused on changes in theta-gamma phase coupling and action potential entrainment to theta, phenomena hypothesized to be critical for cognition. The authors provide convincing evidence of deficits in both features post-TBI and contributes new understanding to how disruptions in oscillatory coordination and spike timing may relate to cognitive impairment.

Reviewer #1 (Public review):

Summary:

This study examines how traumatic brain injury (TBI) alters hippocampal network dynamics and single-unit activity in awake, behaving rats. Using laminar recordings, the authors report reductions in theta power, theta-gamma phase-amplitude coupling, and spike-field entrainment, alongside impairments in spatial memory performance.

Strengths of the study include the use of high-density laminar electrodes to localize activity across hippocampal layers and the integration of electrophysiological and behavioral measures. Analyses that consider behavioral state and account for broadband power changes improve confidence in the interpretation of oscillatory effects. Additional controls suggest that the observed differences are unlikely to be explained by gross motor or motivational deficits. The reported relationships between theta amplitude, phase-amplitude coupling, and spike entrainment provide useful insight into how network coordination is disrupted following injury.

There are a few minor weaknesses. The analyses of single-unit activity across environments are relatively limited, and more comprehensive approaches to characterizing spatial coding would strengthen conclusions about how TBI impacts hippocampal representations. The behavioral assessment relies primarily on a single task, which constrains the interpretation of the cognitive deficits. In addition, the relatively small number of animals is a limitation, although this is partially mitigated by the number of recorded units and the consistency of effects across measures.

Overall, this work provides a careful characterization of hippocampal circuit dysfunction following TBI and contributes to understanding how disruptions in oscillatory coordination and spike timing may relate to cognitive impairment.

Comments on revisions:

The authors have adequately addressed all of my concerns.

Reviewer #3 (Public review):

Summary:

In this study, authors studied the effects of traumatic brain injury created by LFPI procedure on the CA1 at network level. The major findings in this study seem to be that the TBI reduces theta and gamma powers in CA1, reduces phase amplitude coupling in between theta and gamma bands as well as disrupts the gamma entrainment of interneurons. I think the authors have made some important discoveries that could help advance the understanding of TBI effects at physiological level, however, more investigations into deciphering the relationship of the behavioral and brain states to the observed effects would help clarify the interpretations for the readers.

Strengths:

The authors in this study were able to combine behavioral verification of the TBI model with the laminar electrophysiological recordings of CA1 region to bring forward network level anomalies such as the temporal coordination of network level oscillations as well as in the firing of the interneurons. Indeed, it seems that the findings may serve future studies to functionally better understand and/or refine the therapies for the TBI.

Weaknesses:

Discoveries made in the paper and their broad interpretations can be helped with further characterization and comparison among the brain and behavioral states both during immobility and movement. The impact of brain injury in several parts of the brain can alter brain wide LFP and/or behavior. The altered behavior and/or LFP patterns might then lead to reduced spiking and unreliable LFP oscillations in the hippocampus. Hence, claims made in abstract such as "These results reveal deficits in information encoding and retrieval schemes essential to cognition that likely underlie TBI-associated learning and memory impairments, and elucidate potential targets for future neuromodulation therapies" does not have enough evidence in testing whether the disruptions were information encoding and retrieval related or due to sensory-motor and/or behavioral deficits that could also occur during TBI.

Movement velocity is already known to be correlated to the entrainment of spikes with the theta rhythm and also in some cases with the gamma oscillations. So, it is of importance to disentangle the differences in behavioral variables and the observed effects. As an example, the author's claims of disrupted temporal coding (as shown in the graphical abstract) might have suffered from these confounds. The observed results of reduced entrainment might on one hand be due to the decreased LFP power (induced by injury in different brain areas) resulting in altered behavior and/or the unreliable oscillations of the LFP bands such as theta and gamma, rather than memory encoding and retrieval related disruption of spikes synchrony to the rhythms, while on the other hand they may simply be due to reduced excitability in the neurons particularly in the behavioral and brain state in which the effects were observed, rather than disrupted temporal code. Hence, further investigations into dissociating these factors could help readers mechanistically understand the interesting results observed by the authors.

Comments on revisions:

The authors have substantially improved the manuscript in response to the previous reviews. In particular, the revisions addressing the issue of behavioral deficits that could be caused due to the TBI, which were surprisingly not present (if anything minimal) in the injured rats, have strengthened the study and improved the support for the main conclusions. Overall, the manuscript is now clearer and more rigorous. Authors have also addressed all the minor points raised in the study. As a result, the study is now solid, with the major findings broadly supported by the data.

Author response:

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

eLife Assessment

This is an important paper that reports in vivo physiological abnormalities in the hippocampus of a rat model of traumatic brain injury (TBI). In this study, authors focused on changes in theta-gamma phase coupling and action potential entrainment to theta, phenomena hypothesized to be critical for cognition. While the authors provide solid evidence of deficits in both features post-TBI, the study would have been stronger with a more hypothesis-driven approach and consideration of alterations of the animal's behavioral state or sensorimotor deficits beyond memory processes.

We would like to thank the reviewers for their comments on our manuscript. By incorporating their feedback, we were able to make our hypotheses more clear, expand our analyses to compare physiological processes across similar behavioral states, and address extra hippocampal input and potential sensorimotor confounds in our data.

Specifically, we have added new data in Figure 5 showing how theta amplitude correlates with theta-gamma PAC and entrainment strength. We have also added supplementary Figure 1 demonstrating that there are no differences in exploration or movement velocity in injured animals compared to shams. Supplementary Figures 2, 3, and 4 were added to compare oscillatory power while animals were still, moving at a higher velocity, and following a broadband power shift correction respectively. We also added Supplementary Figure 7 demonstrating that there were no differences in firing rates between sham and injured animals while they were still or moving and Supplementary Figure 8 showing no changes in pyramidal cell bursting. Finally, we added Supplementary Figure 10 showing that there was no difference in velocity or distance traveled during testing in the MWM between sham and injured animals and that learning curves were similar across groups before sham/injury surgery. We believe that the addition of this data significantly improves our manuscript by more strongly controlling for the animal’s behavioral state in our analyses and provides strong evidence that significant sensory/motor deficits were not present in injured animals at this injury level and time point post injury. Below we address specific points raised by the reviewers.

Reviewer #1 (Public review):

Summary:

This study investigated how traumatic brain injury affects oscillatory and single-unit hippocampal activity in awake-behaving rats.

Strengths:

The use of high-density laminar electrodes enabled precise localization of recording sites. To ensure an unbiased, rigorous approach, single-unit analysis was performed by a reviewer who was blind to experimental conditions. A proof of concept study was undertaken to characterize the pathology that resulted from the specific TBI model used in the main study. There was an effort to link abnormalities in hippocampal activity to memory disruption by running a cohort of rats on the Morris Water Maze task.

Weaknesses:

The paper is written as if the experiment was exploratory and not hypothesis-driven despite the fact that there is a wealth of experimental evidence about this TBI model that could have informed very specific predictions to test a hypothesis that is only hinted at in the discussion. The number of rats used for the spatial working memory experiment is not reported. Some of the statistics are not completely reported. It is also unclear what the rationale was for recording single units in a novel and familiar environment. Furthermore, this analysis comparing single-unit activity between familiar and novel environments is quite rudimentary. There are much more rigorous analyses to answer the question of how hippocampal single-unit firing patterns differ across changes in environments. There are details lacking about the number of units recorded per session and per rat, all of which are usually reported in studies that record single units. Spatial working memory assessment is delegated to a single panel of a supplementary figure. More importantly, there is no effort to dissociate between spatial working memory deficits and other motor, motivational, or sensory deficits that could have been driving the lower "memory score" in the experimental group.

In order to address these important concerns, we have made the following changes:

(1) We have updated the results section to include more rationale for the recordings and analyses used to clarify our hypotheses. In addition, we hope that our extensive characterization will lay the groundwork to inform future studies investigating circuit-specific disruptions following TBI and neuromodulatory therapies.

(2) The number of rats used for the spatial working memory experiment is reported in the text and figure legend.

(3) We have added supplemental Table 2 to include the requested statistical information (t-statistic, degrees of freedom, and 1 vs 2-tailed analyses).

(4) Unfortunately, we did not have adequate occupancy to robustly extract and compare place cell properties across groups and environments which obscured the rationale of our study design and limited us to more rudimentary analyses. While animals did actively explore the two environments, the relatively short recording time limited the spatial sampling of the two-dimensional environment. We were able to extract putative place cells and found some evidence that place cells in TBI rats had lower spatial information content than in shams (as has previously been described). However, we did not feel that place cell analyses were rigorous enough to include in this manuscript due to the limited spatial sampling. Future studies in the lab will assess how TBI affects place cell information content, stability, and phase precession with better occupancy.

(5) We have added Supplemental Table 1 that includes the total number of units recorded for each animal.

(6) The spatial working memory deficit we report in the MWM is not a novel finding in this model of TBI. However, we wanted to ensure that _LFPI in our hands at this injury level reproduced this known deficit. Importantly, the swim speed and distance traveled during testing did not differ between groups, suggesting that differences were not due to motor deficits. Additionally, the learning curves before sham/_LFPI surgery were the same across groups. This data has been added to the manuscript in Supplementary Figure 10. While we did not test animals in a version of the task where the platform was visibly marked, previous studies have demonstrated that sham and injured rats perform comparably in a version of the MWM where the platform is visible or when a constant start location is used. These citations have been added to the manuscript.

Reviewer #1 (Recommendations for the authors):

For a more rigorous way of analyzing changes in hippocampal firing patterns across environments, see Wills et al 2005 for example.

Addressed in point 4 above

Spatial working memory tasks should always be compared with a control task to rule out confounding performance variables. Examples would be to use a variant of the MWM task that does not require the hippocampus such as using a visible escape platform.

Addressed in point 6 above

Statistics are typically reported including a t-statistic and degrees of freedom, not just the p-value. In addition, the authors should indicate whether the t-test is one or two-tailed.

Addressed in point 3 above

Reviewer #2 (Public review):

Summary:

The authors investigate changes in theta-gamma phase amplitude coupling, and action potential entrainment to theta following traumatic brain injury (TBI). Both phenomena are widely hypothesized to be important for cognition, and the authors report deficits in both after TBI. The manuscript is well-written, the figures are well-constructed, and the author's use of high-level analysis methods for TBI EEG data collected from awake, behaving animals is welcome.

Major Comments:

The animal n's are small (4 sham and 5 injured). In Figure 3, for instance, one wonders if panels D and E might have shown significant differences if more animals had been recorded.

There are conflicting reports regarding the effect of _LFPI on single cell firing rates. This is likely due to differential task demands and variations in _LFPI severity across studies. We agree that the firing rates do appear to be trending; however, overall firing rate changes can be difficult to interpret. Because firing rates are influenced by behavior and brain state, we further separated firing rates into epochs when animals were moving or still and found similar trends that did not reach significance (data added in Supplementary Figure 7). We also assessed bursting in pyramidal cells to investigate whether potential changes in bursting influenced overall firing rates, and we found no differences between sham and injured animals across conditions (data added in Supplementary Figure 8). While the n’s are small when considered by animal, the number of units is actually fairly large, so if there were robust effects (as there were for the entrainment analyses), we would expect to see significant differences.

The text focuses on deficits in the theta and gamma bands, but the reduction in power appears to be broadband (see Figure 1F, especially Pyramidal cell layer panel). Therefore, the overall decrease in broadband (in the injured population) must be normalized between sham and injured animals before a selective comparison between sham and injured animals can be conducted. That is the only way that selective narrow bands i.e., theta and low gamma can be compared between the two cohorts. A brief discussion of the significance of a broadband decrease would be appreciated.

This is an excellent point that has now been addressed with the addition of Supplementary Figure 4. We used a well-established method (Donoghue et al 2020) to flatten power spectra in order to compare specific frequency bands in the context of a broadband shift. After applying this correction, we show that theta power is still reduced in injured rats compared to shams. While there is no difference in gamma power between groups in the corrected power spectra, this result should be interpreted with caution especially since there is not a large distinct peak in the gamma frequency range in the power spectrum of either sham or injured animals. However, if this is interpreted to mean that gamma power is not different between sham and injured animals, it makes the PAC data even more compelling. While there is clearly a broadband shift, the frequency range of this shift is still limited in the frequency domain to ~4-90Hz which contains physiologically relevant frequencies associated with synaptic currents. Importantly, the power spectra of sham and injured animals converge at low (<4Hz) and high (>100Hz) frequencies. This suggests that slow oscillations which could include delta and respiration-associated oscillations are not affected by TBI (though sleep recordings would be needed to properly address this). High-frequency activity can include ripples and HFOs which need to be separately extracted when comparing between groups due to their transient nature. However, overall spiking activity including the depolarizing spike and the after hyperpolarization significantly contribute to power in the high frequency range. Because this general high-frequency power is not different between groups, it suggests that the limited range of the broadband power reduction still contains important physiological signals. This broadband shift may result from a global reduction in or desynchronization of synaptic input to CA1. The specific mechanisms behind this broadband shift and the consequences it has on coding information in the hippocampus are fascinating questions that we hope will be specifically investigated in future studies. This point is now addressed in the Discussion.

Reviewer #2 (Recommendations for the authors):

Minor Comments:

Please define your reference waveform for theta - is it theta recorded on the channel containing the cell? Average theta for all electrodes in SP? SP + SO? Theta for the nominal "St. pyr." channel? Please define.

For all entrainment analyses, entrainment was measured referenced to the theta oscillation recorded from st. pyr. on the specific shank where the unit was detected. We added clarification in the results and methods sections regarding this point.

Similarly, even though the peak of the theta wave appears from the figures to be taken as 0 degrees, please explicitly state this in the text.

This has been added to the results and methods.

Did the authors check for any difference between interneurons in SP and interneurons in SO?

This is an excellent suggestion that we had hoped to investigate as it could inform whether specific interneuron populations were affected. However, we did not record enough units in st. ori to make this comparison.

On page 8, Figures 3E and 3F are incorrectly labeled 4E and 4F.

This has been fixed.

Figure 1, panel C: please add a numerical scale to the colored scale bar.

This has been added

Figure 1, panel F: how was the significance between the frequency bands calculated?

Statistics were done using a t-test at each frequency point with significance set at α=0.01 for multiple comparisons. This has been clarified in the figure legend and methods.

Figure 3, panel A legend: Please add "Spike at 0 ms omitted for clarity.”

This has been added

Figure 4, panel A, right side: please provide the MVL for this cell, so that readers have a benchmark for evaluating the MVL as a parameter. A sample poorly entrained cell, with MVL, would also be informative.

We added the MVL for this cell. We were unable to add a poorly entrained cell without making the figure more confusing.

Raw data must be provided for the Morris Water Maze experiments described in Supplementary Figure 3.

We added data showing no difference in the swim velocity or distance traveled between the sham and injured groups during memory testing as well as data showing that the two groups had similar learning curves during training before sham/injury surgery. See Supplementary Figure 10.

Antibody 22C11 for APP has been shown to be non-specific when used for immunocytochemistry (it may be fine for Westerns). In addition, using a biotinylated secondary with an ABC kit for visualization risks contamination by post-injury changes in biotin. Reviewed in Xiong et al., 2023, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10580020/.

As is standard practice in neuropathology, negative controls were run for all of these experiments (identical preparations minus the primary antibody.) No non-specific staining was present that could be mis-interpreted as APP-positive axonal profiles in either sham or injured tissue. While beyond the scope of this response, there are many reasons the authors of the cited paper may have had non-specific staining, including a concentration 450X that of the one utilized here and the absence of an antigen-retrieval technique in their protocol.

Tummala et al. used in vivo calcium-imaging after TBI and also investigated single-cell activity in familiar and novel environments, and when moving or still. The authors could consider discussing their work.

We have added a citation for this paper

Reviewer #3 (Public review):

Summary:

In this study, the authors studied the effects of traumatic brain injury created by LFPI procedure on the CA1 at the network level. The major findings in this study seem to be that the TBI reduces theta and gamma powers in CA1, reduces phase-amplitude coupling in between theta and gamma bands as well as disrupts the gamma entrainment of interneurons. I think the authors have made some important discoveries that could help advance the understanding of TBI effects at the physiological level, however, more investigations into deciphering the relationship of the behavioral and brain states to the observed effects would help clarify the interpretations for the readers.

Strengths:

The authors in this study were able to combine behavioral verification of the TBI model with the laminar electrophysiological recordings of the CA1 region to bring forward network-level anomalies such as the temporal coordination of network-level oscillations as well as in the firing of the interneurons. Indeed, it seems that the findings may serve future studies to functionally better understand and/or refine the therapies for the TBI.

Weaknesses:

Discoveries made in the paper and their broad interpretations can be helped with further characterization and comparison among the brain and behavioral states both during immobility and movement. The impact of brain injury in several parts of the brain can alter brain-wide LFP and/or behavior. The altered behavior and/or LFP patterns might then lead to reduced spiking and unreliable LFP oscillations in the hippocampus. Hence, claims made in the abstract such as "These results reveal deficits in information encoding and retrieval schemes essential to cognition that likely underlie TBI-associated learning and memory impairments, and elucidate potential targets for future neuromodulation therapies" do not have enough evidence to test whether the disruptions were information encoding and retrieval related or due to sensorymotor and/or behavioral deficits that could also occur during TBI.

Movement velocity is already known to be correlated to the entrainment of spikes with the theta rhythm and also in some cases with the gamma oscillations. So, it is important to disentangle the differences in behavioral variables and the observed effects. As an example, the author's claims of disrupted temporal coding (as shown in the graphical abstract) might have suffered from these confounds. The observed results of reduced entrainment might, on one hand, be due to the decreased LFP power (induced by injury in different brain areas) resulting in altered behavior and/or the unreliable oscillations of the LFP bands such as theta and gamma, rather than memory encoding and retrieval related disruption of spikes synchrony to the rhythms, while on the other hand, they may simply be due to reduced excitability in the neurons particularly in the behavioral and brain state in which the effects were observed, rather than disrupted temporal code. Hence, further investigations into dissociating these factors could help readers mechanistically understand the interesting results observed by the authors.

We appreciate the Reviewer’s insights into disentangling the complex interactions between power, entrainment, and excitability, and have attempted to dissociate these further in our analyses. Regarding the broad effects of TBI, we agree that TBI affects many brain regions outside of the hippocampus as well as white matter pathways containing axons from areas where pathology is not visible, which likely results in widespread changes to LFPs across regions and altered behavior. Here we report disrupted network activity in the hippocampus which is likely a consequence of numerous pathologies across multiple brain regions. In the discussion, we speculate that disrupted power and coupling comes from desynchronization of inputs (especially those from the mEC and MS) as well as changes to local circuits within the hippocampus which combine to disrupt temporal coding. While the disrupted processes we report in the hippocampus are implicated in computational processes thought to support learning and memory, we acknowledge that results from this study do not causally reveal a specific mechanism that is directly responsible for cognitive impairments. We have changed the language of the quoted sentence from the abstract to make our claim less causal as we agree that the direct effects of these results on cognition are difficult to quantify due to the fact that animals were not performing a spatial navigation task with measurable outcomes during recordings. We have also removed the graphical abstract as we believe it is an oversimplification of the results given new analyses.

Regarding the possible contribution of sensory and motor deficits or differences in behavioral states to the observed changes, we agree that it is essential to consider potential sensorimotor deficits as well as the animal’s behavioral state when comparing oscillations and single unit activity in the hippocampus, especially since these phenomena have been extensively liked to movement velocity and exploration. To address this, we have added Supplementary Figure 1 showing that there are no differences in movement velocity or exploration time between sham and injured animals. Because animals were simply foraging during electrophysiological experiments we do not expect there to be any major additional behavioral differences that would influence oscillations or spiking once locomotion is controlled for, though differences in attention or arousal cannot be ruled out. Additionally, analyses throughout the manuscript are performed independently during periods when animals were moving or still. Data in Figures 1 and 2 also only include data from the familiar environment to rule out any effects of novelty on hippocampal oscillations. Supplementary Figures 2 and 3 were added to demonstrate that TBI-associated reductions in power were consistent when animals were still and when a higher threshold for movement (>20 cm/sec) was used. Finally, supplementary Figure 10 was added showing no differences in swim velocity or distance traveled in the MWM between sham and injured animals, further suggesting that there are no significant sensorimotor deficits at this injury level and timepoint. Additionally, previous studies have demonstrated that sham and injured rats perform comparably in a version of the MWM where the platform is visible or when a constant start location is used, which provides further support that sensorimotor deficits are not responsible for memory deficits in this task (see above).

Regarding the contribution of neuronal excitability to the reported changes, we agree that changes in the excitability of neurons could have a strong effect on entrainment. Importantly, we show that the disrupted oscillations recorded in the injured hippocampus do not coincide with significant changes in neuronal firing rates between sham and injured animals. We have added Supplementary Figure 7 demonstrating this holds true both when animals are still and when they are moving. Additionally, we have added Supplementary Figure 8 showing no differences in pyramidal cell bursting between sham and injured animals. While this suggests that there are not major changes in excitability, homeostatic plasticity mechanisms may impact firing rates and bursting, and the extent of these effects and their role on entrainment are unclear. This point was added to the Discussion.

To address the effects of LFP power on entrainment strength, Figure 5 has been updated to show theta and gamma entrainment strength as well as theta-gamma PAC as a function of theta amplitude. We found that, during periods of comparable theta power, interneurons from sham and injured animals are similarly entrained to theta, but pyramidal cells from injured animals become significantly more entrained to theta than in shams. We address the potential implications of these results in the Discussion.

Reviewer #3 (Recommendations for the authors):

The authors have stated on page 7 and Figure 2E, "Taken together, injured rats show a decrease in the strength of theta-gamma PAC that is specific to st. pyr, and a shift in peak gamma amplitude to a later phase of theta in both st. pyr and st. rad". Is the shift in the peak position greater than expected by chance?

We are unaware of a rigorous method that would allow us to compare this shift statistically. We have reported the observed shift and avoided calling the shift significant for that reason.

The authors state on page 9 "cells (sham familiar=1.63{plus minus}0.23 Hz, n=51, injured familiar=2.11{plus minus}0.20 Hz, n=141, p=0.446; sham novel=1.84{plus minus}0.18 Hz, n=55, injured novel=2.23{plus minus}0.21 Hz, n=134, p=0.170; mean{plus minus}SEM; ks-test; Fig 4E) between sham and injured groups, but a higher percentage of pyramidal cells were active (firing rate >0.1Hz) in both the familiar and novel environment in injured rats compared to shams (sham=74%, injured=87%, p=0.025, Fisher's exact test; Fig 4F)." Do the authors mean Figures 3E and 3F respectively in place of Figures 4E and 4F?

This has been fixed.

Regarding the finding of similar firing rates and differences in the overlap of the neurons that were active in between injured and control animals, it is imperative to study the differences in behaviors of the animals. First of all, it seems appropriate to quantify and compare the immobility and mobile periods as well as the movement velocity of the animals in both groups. Then, it would be interesting to see if any behavioral variables correlate with the firing characteristics of the cells in both the sham and the injured animals. Since hippocampal cells have been known to have different levels of recruitment and firing rates according to different behavioral states such as movement velocity, some of the similarities or differences in neural findings might as well be attributed to the differences in behaviors in between the groups. However, some differences may be observed in the injured rats despite similar behavior and the LFP powers. In other words, studying the effects of injury during similar behavioral (e.g. firing rate as a function of movement velocity) and brain states (e.g. categorical effects of awake theta state, type two theta, and ripple states on firing rates and the entrainment) might help dissociate some effects that might only be due to difference in the behavior caused by the injury throughout the brain and might as well have less to do with specific injury induced local circuits level deficits in the hippocampus. The results in Figures 4, 5, and 6 reveal such interesting differences and hence, it becomes even more important to quantify and correlate behavioral states (movement velocity and theta/ripple) to the neuronal characteristics (LFP power, PAC, firing rates, and entrainment) presented in Figure 3.

These are excellent points, and we have addressed them in the following ways:

We added Supplementary Figure 1 demonstrating that there were no differences in movement velocity between sham and injured animals during electrophysiological recordings.

Power and PAC analyses were done exclusively when the animal was moving to compare across similar behavioral states. Additionally, these analyses were constrained to recordings from the familiar environment to rule out any effects of novelty. Because animals were simply foraging during recordings we do not expect other behavioral factors besides movement velocity to play a major role in these processes. We have also added Supplementary Figures 2 and 3 which demonstrate that TBI-associated differences in oscillatory power follow similar trends when animals are still (Sup. Fig 2) or when a higher movement threshold (>20cm/sec) is used (Sup Fig 3). We also added Supplementary Figures 7 and 8 showing that there were no significant differences in firing rates or bursting while animals were still or while they were moving.

The Discussion was expanded to discuss how TBI may disrupt circuits outside the hippocampus which may contribute to our findings. Additionally, we acknowledge the limitation that these recordings were not obtained while animals were doing a quantitatively measurable spatial navigation task which limits our ability to assess whether changes are truly behaviorally relevant.

We have also updated Figure 5 to show entrainment across different levels of theta power.

Elaborating on the abovementioned point, Figures 4B and 4E depict a finding that mean entrainment is reduced in the injured during immobility. The following factors may contribute to the results:

(1) Reduction in theta power during immobility (reduced attention and/or LFP profile due to brain-wide injury), which makes theta cycles unreliable, which can contribute to the results.

(2) Changes in neural firing properties during immobility, such as reduced burst rates or firing rates during immobility.

(3) As the authors claimed in the graphical abstract, there might be an actual disruption of temporal code associated with the memory encoding. It would be awesome if the temporal disruption could be investigated during the comparable theta power and behavioral states. This analysis would test whether there is an unconfounded disruption in the temporal code in the hippocampus due to the injury. In any case, it would be ideal to isolate the epochs during sleep in which animals were in theta state and exclude ripple states to make a definitive assessment of the aforementioned factors. These further investigations would also help the interpretations made by authors in the discussion section such as "This can disrupt type II theta which occurs when animals are not actively moving and exploring the environment. We found that single unit entrainment to theta was substantially decreased in injured rats when they were not moving, a phenomenon not seen in shams, which suggests a disruption in type II theta. This provides further evidence that cholinergic signaling may be dysfunctional following TBI."

(1) While theta power is reduced in injured animals, it can still be reliably detected even at rest. We added Supplementary Figure 2 showing power spectra while animals were not moving, and a distinct peak can be seen in the theta frequency range. Additionally, clear peaks in entrainment can be seen in the theta frequency band in Fig 4B while animals were still. This suggests that theta can still be reliably detected in injured animals even when they are not moving. However, we agree that reduced attention or arousal could contribute to these changes, and this point has been added to the Discussion.

(2) We added Supplementary Figures 7 and 8 showing no differences in firing rates or bursting parameters between groups during periods of immobility.

(3) We updated Figure 5 which now shows entrainment strength as a function of theta amplitude. We found that the theta entrainment strength of both pyramidal cells and interneurons increased with increasing theta amplitudes. We address potential implications of these changes in the Discussion.

On page 10 the authors state, "theta entrainment strength drastically increased when rats began moving in injured but not sham animals." It is unclear if the effect was confined to the periods when rats started movement. Also, it would be of interest to investigate whether movement epochs and velocity were affected in the periods when the effects were observed.

This was not confined to the exact points when the rats started moving. We removed the word “began” for clarity. See point regarding velocity above.

On page 12 the authors state, "On test day, injured rats had a lower memory score than shams (sham=114.8 {plus minus} 21.8, n=9; injured=51.5{plus minus}6.8, n=14; p=0.020; mean {plus minus} SEM; Welch's t-test) indicating poor spatial memory (Sup Fig 3A)." The result is the validation of the TBI injury on a hippocampal-dependent Morris water maze task. However, it would be nice to see the quantification of the movement velocity in the water maze and the trajectory length in each group to further dissect whether animals were constrained in the movement and hence, they could not get to the platform or they forgot where it was located. Also, it would help to compare the rats' performance after sham or TBI surgeries to their performance during the training before the surgeries (assuming the data during the training periods were recorded as well).

We have added Supplemental Figure 10 to include all of this information. Importantly, movement velocity and distance traveled were not different between groups on testing day, and the learning curves of both groups were the same before sham/injury surgery.

eLife Assessment

This important study details changes in the brain functional connectivity in a longitudinal cohort of Gambian children assessed outside a lab setup with functional near-infrared spectroscopy (fNIRS) from age 5 to 24 months, in relation to early physical growth and cognitive flexibility capacities at preschool age. Evidence supporting conclusions on the evolution of brain connectivity is convincing and highlights a different trajectory compared with populations from high-income countries. However, analyses linking connectivity trajectories with early adverse conditions such as undernutrition and later cognitive development are only partially supported due to insufficient longitudinal data and statistical power. This study will be of significant interest to neuroscientists, psychologists and neuroimaging researchers working on infant development in relation to environmental factors.

Reviewer #1 (Public review):

Summary:

This study utilizes fNIRS to investigate the effects of undernutrition on functional connectivity patterns in infants from a rural population in Gambia. fNIRS resting-state data recording spanned ages 5 to 24 months, while growth measures were collected from birth to 24 months. Additionally, executive functioning tasks were administered at 3 or 5 years of age. The results show an increase in left and right frontal-middle and right frontal-posterior connections with age and, contrary to previous findings in high-income countries, a decrease in frontal interhemispheric connectivity. Restricted growth during the first months of life was associated with stronger frontal interhemispheric connectivity and weaker right frontal-posterior connectivity at 24 months of age. Additionally, the study describes some connectivity patterns, including stronger frontal interhemispheric connectivity, which is associated with better cognitive flexibility at preschool age.

Strengths:

- The study analyses longitudinal data from a large cohort (n = 204) of infants living in a rural area of Gambia. This already represents a large sample for most infant studies, and it is impressive, considering it was collected outside the lab in a population that is underrepresented in the literature. The research question regarding the effect of early nutritional deficiency on brain development is highly relevant and may highlight the importance of early interventions. The study may also encourage further research on different underrepresented infant populations (i.e., infants not residing in Western high-income countries) or in settings where fMRI is not feasible.

- The preprocessing and analysis steps are carefully described, which is very welcome in the fNIRS field, where well-defined standards for preprocessing and analysis are still lacking.

Weaknesses:

- The study provides a solid description of the functional connectivity changes in the first two years of life at the group level and investigates how restricted growth influences connectivity patterns at 24 months. However, it does not explore the links between adverse situations and developmental trajectories for functional connectivity. Given the longitudinal nature of the dataset, future work should expand the analysis using more sophisticated tools to link undernutrition to specific developmental trajectories in functional connectivity, and eventually incorporate additional data to increase statistical power.

- Connectivity was assessed in 6 big ROIs to reduce variability due to head size and optode placement. Nevertheless, this also implies a significant reduction in spatial resolution. Individual digitalisation and co-registration of the optodes to a head model, followed by image reconstruction, could provide better spatial resolution. This is not a weakness specific to this study but rather a limitation common to most fNIRS studies, which typically analyse data at the channel level since digitalisation and co-registration can be challenging, especially in complex setups like this. The authors made an important effort to identify subjects with major optode displacement; however, future work might use tools to digitally record the positions of optodes and head markers.

Reviewer #2 (Public review):

Strengths:

The article addresses a topic of significant importance, focusing on early life growth faltering in low-income countries-a key marker of undernutrition-and its impact on brain functional connectivity (FC) and cognitive development. The study's strengths include the laborious data collection process, as well as the rigorous data preprocessing methods employed to ensure high data quality. The use of cutting-edge preprocessing techniques further enhances the reliability and validity of the findings, making this a valuable contribution to the field of developmental neuroscience and global health.

Weaknesses:

The study lacks specificity in identifying which specific brain networks are affected by growth faltering, as the current exploratory analyses mainly provide an overall conclusion that infant brain network development is impacted without pinpointing the precise neural mechanisms or networks involved.

Reviewer #3 (Public review):

Summary

This study aimed to investigate whether the development of functional connectivity (FC) is modulated by early physical growth, and whether these might impact cognitive development in childhood. This question was investigated by studying a large group of infants (N=204) assessed in Gambia with fNIRS at 5 visits between 5 and 24 months of age. Given the complexity of data acquisition at these ages and following data processing, data could be analyzed for 53 to 97 infants per age group. FC was analyzed considering 6 ensembles of brain regions and thus 21 types of connections. Results suggested that: i) compared to previously studied groups, this group of Gambian infants have different FC trajectory, in particular with a change in frontal inter-hemispheric FC with age from positive to null values; ii) early physical growth, measured through weight-for-length z-scores from birth onwards, is associated with FC at 24 months. Some relationships were further observed between FC during the first two years and cognitive flexibility, in different ways between 4- and 5-year-old preschoolers, but results did not survive corrections for multiple comparisons.

Strengths

The question investigated in this article is important for understanding the role of early growth and undernutrition on brain and behavioral development in infants and children. The longitudinal approach considered is highly relevant to investigate neurodevelopmental trajectories. Furthermore, this study targets a little studied population from a low-/middle-income country, which was made possible by the use of fNIRS outside the lab environment. The collected dataset is thus impressive and it opens up a wide range of analytical possibilities.

Weaknesses

Data analyses were constrained by the limited number of children with longitudinal data on NIRS functional connectivity. Applying advanced statistical modeling approaches such as structural equation modelling would provide further insights on neurodevelopmental trajectories and relationships with early growth and later cognitive development.

Author response:

The following is the authors’ response to the previous reviews

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study utilises fNIRS to investigate the effects of undernutrition on functional connectivity patterns in infants from a rural population in Gambia. fNIRS resting-state data recording spanned ages 5 to 24 months, while growth measures were collected from birth to 24 months. Additionally, executive functioning tasks were administered at 3 or 5 years of age. The results show an increase in left and right frontal-middle and right frontal-posterior connections with age and, contrary to previous findings in high-income countries, a decrease in frontal interhemispheric connectivity. Restricted growth during the first months of life was associated with stronger frontal interhemispheric connectivity and weaker right frontal-posterior connectivity at 24 months of age. Additionally, the study describes some connectivity patterns, including stronger frontal interhemispheric connectivity, which is associated with better cognitive flexibility at preschool age.

Strengths:

The study analyses longitudinal data from a large cohort (n = 204) of infants living in a rural area of Gambia. This already represents a large sample for most infant studies, and it is impressive, considering it was collected outside the lab in a population that is underrepresented in the literature. The research question regarding the effect of early nutritional deficiency on brain development is highly relevant and may highlight the importance of early interventions. The study may also encourage further research on different underrepresented infant populations (i.e., infants not residing in Western high-income countries) or in settings where fMRI is not feasible.

The preprocessing and analysis steps are carefully described, which is very welcome in the fNIRS field, where well-defined standards for preprocessing and analysis are still lacking.

We thank the reviewer for highlighting the strengths of this work.

Weaknesses:

While the study provides a solid description of the functional connectivity changes in the first two years of life at the group level and investigates how restricted growth influences connectivity patterns at 24 months, it does not explore the links between adverse situations and developmental trajectories for functional connectivity. Considering the longitudinal nature of the dataset, it would have been interesting to apply more sophisticated analytical tools to link undernutrition to specific developmental trajectories in functional connectivity. The authors mention that they lack the statistical power to separate infants into groups according to their growing profiles. However, I wonder if this aspect could not have been better explored using other modelling strategies and dimensional reduction techniques. I can think about methods such as partial least squares correlation, with age included as a numerical variable and measures of undernutrition.

We agree with the reviewer that this complex and rich longitudinal dataset would benefit from more sophisticated analytical approaches to characterise developmental trajectories in functional connectivity and to more directly link them to measures of undernutrition. However, conducting such analyses would require substantial additional methodological development, model validation, and careful interpretation, which fall beyond the scope and timeline of the present manuscript. Our aim here was to provide a clear and robust characterisation of functional connectivity changes during the first two years of life and to examine associations with growth outcomes at a specific developmental stage, while ensuring methodological transparency and statistical reliability. Importantly, these more advanced trajectory-based analyses are currently being pursued in the final phase of the BRIGHT project (BRIGHT IMPACT), in collaboration with expert statisticians and data scientists. This ongoing work aims specifically to leverage the longitudinal richness of the dataset to model developmental trajectories and their associations with early-life adversity and nutritional factors. We therefore see the present study as an important foundation for these forthcoming analyses.

Connectivity was assessed in 6 big ROIs. While the authors justify this choice to reduce variability due to head size and optodes placement, this also implies a significant reduction in spatial resolution. Individual digitalisation and co-registration of the optodes to the head model, followed by image reconstruction, could have provided better spatial resolution. This is not a weakness specific to this study but rather a limitation common to most fNIRS studies, which typically analyse data at the channel level since digitalisation and co-registration can be challenging, especially in complex setups like this. However, the BRIGHT project has demonstrated that it is possible and that differences in placement affect activation patterns, which become more localised when data is co-registered at the subject level (Collins-Jones et al., 2021). Could the co-registration of individual data have increased sensitivity, particularly given that longitudinal effects are being investigated?

We agree with the reviewer that the fNIRS community should work toward more precise methods for spatial registration of optodes, not only at the group level but also at the subject level, in order to make more precise inferences about the locations of activations. However, we followed a very thorough offline procedure to model headgear placement based on each participant’s photographs, which we believe complements the coregistration work performed by Collins-Jones in 2021. As reported in the fNIRS data acquisition section “Infants were excluded from further analysis if the band was excessively high over the front above the eyebrows” (line 409, methods section). Moreover channels displacement was measured from the photos, and if it was “equal or greater than 1.6 cm were renumbered, so that each channel was shifted either backward or forward one full channel location in space” (line 413, methods section). While these practices are thoroughly followed in the BRIGHT project, we are aware that they are not part of the standard procedure in many infant fNIRS studies. We hope that this work provides guidance for other researchers on how to coregister infant fNIRS data.

Considering the spatial resolution of fNIRS, which is on the order of centimetres, and the thorough procedure combining fNIRS–MRI coregistration with channel displacement assessment based on photographs, we do not think that individual-level coregistration would have significantly increased the sensitivity of the results.

I believe that a further discussion in the manuscript on the application of global signal regression and its effects could have been beneficial for future research and for readers to better understand the negative correlations described in the results. Since systemic physiological changes affect HbO/HbR concentrations, resulting in an overestimation of functional connectivity, regressing the global signal before connectivity computation is a common strategy in fNIRS and fMRI studies. However, the recommendation for this step remains controversial, likely depending on the case (Murphy & Fox, 2017). I understand that different reasons justify its application in the current study. In addition to systemic physiological changes originating from brain tissue, fNIRS recordings are contaminated by changes occurring in superficial layers (i.e., the scalp and skull). While having short-distance channels could have helped to quantify extracerebral changes, challenges exist in using them in infant populations, especially in a longitudinal study such as the one presented here. The optimal source-detector distance that minimises sensitivity to changes originating from the brain would increase with head size, and very young participants would require significantly shorter source-detector distances (Brigadoi & Cooper, 2015). Thus, having them would have been challenging. Under these circumstances (i.e., lack of short channels and external physiological measures), and considering that the amount the signal is affected by physiological noise (either coming from the brain or superficial tissue) might change through development, the choice of applying global signal regression is justified. Nevertheless, since the method introduces negative correlations in the data by forcing connectivity to average to zero, I believe a further discussion of these points would have enriched the interpretation of the results.

We added a paragraph discussing the choice of using GSR in our pipeline in the discussion of the manuscript as follows: “Importantly, these results remained significant even without GSR, indicating that our findings are not solely driven by preprocessing choices. While the use of GSR in FC studies remains debated (Murphy & Fox, 2017), in the absence of short channels (which are difficult to use reliably with infants (Emberson et al., 2016)) and external physiological measures, applying GSR represented the most appropriate preprocessing option. In fact, failure to correct for systemic physiological fluctuations can, in fact, lead to artificially elevated connectivity estimates in fNIRS data (Abdalmalak et al., 2022)” (line 250, discussion section).

Reviewer #2 (Public review):

Strengths:

The article addresses a topic of significant importance, focusing on early life growth faltering in low-income countries-a key marker of undernutrition-and its impact on brain functional connectivity (FC) and cognitive development. The study's strengths include the laborious data collection process, as well as the rigorous data preprocessing methods employed to ensure high data quality. The use of cutting-edge preprocessing techniques further enhances the reliability and validity of the findings, making this a valuable contribution to the field of developmental neuroscience and global health.

We thank the reviewer for highlighting the strengths of this work.

Weaknesses:

The study fails to fully leverage its longitudinal design to explore neurodevelopmental changes or trajectories, as highlighted by all three reviewers. The revised manuscript still primarily focuses on FC values at a single age stage (i.e., 24 months) rather than utilizing the longitudinal data to investigate how FC evolves over time or predicts cognitive development. Although the authors acknowledge that analyzing changes in FC (ΔFC) would reduce degrees of freedom (to ~30) and risk interpretability, they do not report or discuss these results, even as exploratory findings.

As suggested, we added the table reporting the results of the associations between changes in functional connectivity (DFC) between 5 and 24 months and cognitive flexibility in the supplementary materials (Table SI3). We additionally explored the relationship between changes in growth and cognitive flexibility as suggested by Reviewer #3 and we reported these additional analyses in the text as follows: “We also explored whether changes in growth and changes in functional connectivity between 5 and 24 months were associated with cognitive flexibility at preschool age, but we did not find any significant association (Table SI3 and Table SI4).” (line 213, results section).

Furthermore, the study lacks specificity in identifying which specific brain networks are affected by growth faltering, as the current exploratory analyses mainly provide an overall conclusion that infant brain network development is impacted without pinpointing the precise neural mechanisms or networks involved.

We added this limitation in the discussion as follows: “While the impact of undernutrition on brain development has been documented in LMICs (46), herein, we provided empirical evidence that growth faltering specifically in infants younger than five months of age impacts observable development of functional brain networks in the second year of life. Future studies may be needed to pinpoint which specific brain networks are impacted” (line 279, discussion section).

Reviewer #3 (Public review):

Summary

This study aimed to investigate whether the development of functional connectivity (FC) is modulated by early physical growth, and whether these might impact cognitive development in childhood. This question was investigated by studying a large group of infants (N=204) assessed in Gambia with fNIRS at 5 visits between 5 and 24 months of age. Given the complexity of data acquisition at these ages and following data processing, data could be analyzed for 53 to 97 infants per age group. FC was analyzed considering 6 ensembles of brain regions and thus 21 types of connections. Results suggested that: i) compared to previously studied groups, this group of Gambian infants have different FC trajectory, in particular with a change in frontal inter-hemispheric FC with age from positive to null values; ii) early physical growth, measured through weight-for-length z-scores from birth on, is associated with FC at 24 months. Some relationships were further observed between FC during the first two years and cognitive flexibility, in different ways between 4- and 5-year-old preschoolers, but results did not survive corrections for multiple comparisons.

Strengths

The question investigated in this article is important for understanding the role of early growth and undernutrition on brain and behavioral development in infants and children. The longitudinal approach considered is highly relevant to investigate neurodevelopmental trajectories. Furthermore, this study targets a little studied population from a low-/middle-income country, which was made possible by the use of fNIRS outside the lab environment. The collected dataset is thus impressive and it opens up a wide range of analytical possibilities.

We thank the reviewer for highlighting the strengths of this work.

Weaknesses

Data analyses were constrained by the limited number of children with longitudinal data on NIRS functional connectivity. Nevertheless, considering more advanced statistical modelling approaches would be relevant to further explore neurodevelopmental trajectories as well as relationships with early growth and later cognitive development.

While in this study we selected specific FC and outcome variables based on our hypothesis, the final phase of the BRIGHT project, known as BRIGHT IMPACT, aims to apply advanced statistical models to integrate a range of project variables into a single comprehensive analysis. We have acknowledged this in the discussion as follows: “Applying more advanced statistical modelling methods and structural equation modelling analyses may provide greater insight with further investigations in contexts of adversity and, in turn, establish which outcomes are predicted by FC” (line 309, discussion section).

The abstract and end of the discussion should make it clearer that the associations between FC and cognitive flexibility are results that need to be confirmed, insofar as they did not survive correction for multiple comparisons.

We have acknowledged this in the abstract as follows: “Our results highlight the measurable effects that poor growth in early infancy has on brain development and the possible subsequent impact on pre-school age cognitive development, underscoring the need for early life interventions throughout global settings of adversity”.

We have acknowledged this in the discussion as follows: “While our results are consistent with previous studies, we acknowledge that the significant associations between early FC and later cognitive flexibility do not withstand multiple comparisons. Therefore, we encourage future studies that may replicate these findings with a larger sample” (line 300, discussion section).

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) In Figure 1 B and C the authors should indicate that the results refer to HbO.

We have added the suggested specification in the caption of the figure as suggested.

(2) Figure SI2. Please indicate in the caption that these are the results when pre-processing did not include global signal regression.

We have added the suggested specification in the caption of the figure as suggested.

Reviewer #3 (Recommendations for the authors):

(1) The sentence l529-531 ("To investigate whether FC early in life predicted...") should be more explicit as it is not clear which of the two variables is regressed by the other: is it the measure of cognitive flexibility that is regressed by FC, as the hypothesis suggests? Were other variables considered in the regression model? (For linear regression with only one "prediction" variable, the square root of the coefficient of determination 𝑅2 is equal to the correlation between the two variables.)

Yes, it is the measure of cognitive flexibility that is regressed by FC. We have rephrased it in the text as follows: “we regressed later cognitive flexibility against FC that showed a significant change across the first two years of life”. There were no other variables in the regression model.

(2) A summary table of the statistical results for FC-cognitive flexibility associations should be included as for other analyses, in addition to Figure 3B.

We added a table of the results for the association between FC and cognitive flexibility in the supplementary materials (Table SI2, page 10), matching the same colours of Table 2. We referenced the table in the text in the main manuscript (line 211, result section).

(3) Figure 3B: The legend should precise that these results did not survive corrections for multiple comparisons.

We have specified this in the legend of Figure 3 as suggested.

(4) For the young pre-schooler group, it seems that the age is around 4 years (age mean +/- SD=47.96 +/- 2.77 months) and not 3 years as indicated at several places in the manuscript.

We found only once instance in which we erroneously said that the younger preschoolers were around 3 years. We replaced “Gambian infants from BRIGHT were cross-sectionally assessed at the age of 3 or 5 years for cognitive flexibility” with Gambian infants from BRIGHT were cross-sectionally assessed between the age of 3 and 5 years for cognitive flexibility (line 489, method section).

(5) The authors use the term "intra-hemispheric" connections for the ones within each of the 6 sections. This might be misleading since fronto-posterior connections are also intra-hemispheric ones. Specifying "short-range" or "within-section" connections might be clearer.

As suggested by the reviewer, we replaced “intra-hemispheric” with “intra-hemispheric within section” where appropriate through the whole manuscript.

(6) Abstract: what is the justification for using the term "optimal" for describing developmental trajectories of FC?

The term “optimal” refers to knowledge about typical developmental trajectories, coming especially from fMRI studies, as mentioned in the introduction: “Based on data from fMRI, current models hypothesize that FC patterns mature throughout early development (23–27), where in typically developing brains, adult-like networks emerge over the first years of life as long-range functional connections between pre-frontal, parietal, temporal, and occipital regions become stronger and more selective (28–31). [...]. Importantly, normative developmental patterns may be disrupted and even reversed in clinical conditions that impact development; e.g., increased short-range and reduced long-range FC have been observed in preterm infants (36) and in children with autism spectrum disorder (37, 38)” (line 93-106, introduction).

(7) The confidence interval should be added in Figure SI3.

As suggested, confidence intervals have been added in Figure SI3.

(8) Other scatterplot examples of associations might be added as supplementary information.

As suggested, we added several additional scatterplots to Figure SI3 (with confidence intervals as noted in the comment above) to show other associations between changes in growth and FC at 24 months.

(9) Figure SI6: % in x-axis is still indicated.

We apology for the oversight, all the percentage signs have now been removed from the x-axis tick labels.

(10) The authors might show the (even not significant) results of the associations between changes in growth and cognitive flexibility in supplementary information.

As suggested, we added the table reporting the results of the associations between changes in growth (DWLZ) and cognitive flexibility in the supplementary materials (Table SI3). We additionally explored the relationship between changes in functional connectivity and cognitive flexibility as suggested by Reviewer #2 and we reported these additional analyses in the text as follows: “We also explored whether changes in growth and changes in functional connectivity between 5 and 24 months were associated with cognitive flexibility at preschool age, but we did not find any significant association (Table SI3 and Table SI4).” (line 213, results section).

eLife Assessment

Hoverflies are known for their sexually dimorphic visual systems and exquisite flight behaviors. This valuable study reports how two types of visual descending neurons differ between males and females in their motion- and speed-dependent responses, yet surprisingly, the behavior they control lacks any sexual dimorphism. The results convincingly support these findings, which will be of interest for studies of visuomotor transformations and network-level brain organization.

Reviewer #1 (Public review):

Summary:

Hoverflies are renowned for their striking sexual dimorphism in eye morphology and early visual system physiology, as well as in sexually dimorphic behaviors. Surprisingly, male and female flight behaviors in response to optic flow exhibit only subtle differences. Nicholas et al. investigate the sensorimotor transformation of sexually dimorphic visual information into flight steering commands via descending neurons. Using a combination of intracellular and extracellular recordings, neuroanatomical analysis, and behavioral assays, the authors convincingly demonstrate that descending neurons-particularly at high optic flow velocities-exhibit pronounced sexual dimorphisms, while wing steering responses remain largely monomorphic. The study highlights a very interesting discrepancy between neuronal and behavioral response properties.

More specifically, the authors focused on two types of descending neurons that receive inputs from well-characterized wide-field sensitive tangential cells: OFS DN1 and OFS DN2. Their likely counterparts in Drosophila connect to neck, wing and haltere neuropils. The authors characterized the visual response properties of these two neuronal classes in both male and female hoverflies and identified several interesting differences. They then presented the same set of stimuli, tracked wing beat amplitude and analyzed the sum and the difference of right and left wing beat amplitude as a readout of lift or thrust, and yaw turning, respectively. Behavioral responses showed little to no sexual dimorphism, despite the observed neuronal differences.

Strengths:

I find the question very interesting and the results both convincing and intriguing. A fundamental goal in neuroscience is to link neuronal responses and behavior. The current study highlights that the transformations - even at the level of descending neurons to motoneurons - is complex and less straightforward than one might expect.

Weaknesses:

The authors investigated two types of descending neurons, but it was not clear to me how many other descending neurons are thought to be involved in wing steering responses to wide-field motion. I would suggest providing a more in-depth overview of what is known in hoverflies and Drosophila, since the conclusions drawn from the study would be different if these two types were the only descending neurons involved, as opposed to representing a subset of the neurons conveying visual information to the wing neuropil.

Both neuronal classes have counterparts in Drosophila that also innervate neck motor regions. The authors filled hoverfly DNs in intracellular recordings to characterize their arborization in the ventral nerve cord. In my opinion, these anatomical data could be further exploited and discussed a bit more: is the innervation in hoverflies also consistent with connecting to the neck and haltere motor regions? Are there any obvious differences and similarities to the Drosophila neurons mentioned by the authors? If the arborization also supports a role in neck movements, the authors could discuss whether they would expect any sexual dimorphism in head movements.

Revision comment:

I thank the authors for their detailed replies to my questions and the additional clarifications and analysis included in the paper. All my concerns have been addressed.

Reviewer #2 (Public review):

Summary:

Many fly species exhibit male-specific visual behaviors during courtship while little is known about the circuit underlying the dimorphic visuomotor transformations. Nicholas et al focus on two types of visual descending neurons (DNs) in hoverflies, a species in which only males exhibit high-speed pursuit of conspecifics. They combined electrophysiology and behavior analysis to identify these DNs and characterize their response to a variety of visual stimuli in both male and female flies. The results show that the neurons in both sexes have similar receptive fields but exhibit speed-dependent dimorphic responses to different optic flow stimuli.

Strengths:

Hoverflies, though not a common model system, show very interesting dimorphic behaviors and provide a unique and valuable entry point to explore the brain organization behind sexual dimorphism. The findings here are not only interesting on their own right but will also likely inspire those working in other systems, particularly Drosophila.

The authors employed rigorous morphology, electrophysiology, and behavior methods to deliver comprehensive characterization of the neurons in question. The precision of the measurements allowed for identifying a subtle and nuanced neuronal dimorphism and set a standard for future work in this area.

Weaknesses:

I'd like to thank the authors for the revised manuscript, especially the new analyses and figures. Most of my earlier concerns have been satisfactorily addressed by now. Interested readers are kindly referred to the authors' responses for the discussion of the limitations of this work.

Author response:

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

eLife Assessment

Hoverflies are known for their sexually dimorphic visual systems and exquisite flight behaviors. This valuable study reports how two types of visual descending neurons differ between males and females in their motion- and speed-dependent responses, yet surprisingly, the behavior they control lacks any sexual dimorphism. The results convincingly support these findings, which will be of interest for studies of visuomotor transformations and network-level brain organization.

This statement perfectly recapitulates our findings.

Public Reviews:

Reviewer #1 (Public review):

Summary:

Hoverflies are known for a striking sexual dimorphism in eye morphology and early visual system physiology. Surprisingly, the male and female flight behaviors show only subtle differences. Nicholas et al. investigate the sensori-motor transformation of sexually dimorphic visual information to flight steering commands via descending neurons. The authors combined intra- and extracellular recordings, neuroanatomy, and behavioral analysis. They convincingly demonstrate that descending neurons show sexual dimorphisms - in particular at high optic flow velocities - while wing steering responses seem relatively monomorphic. The study highlights a very interesting discrepancy between neuronal and behavioral response properties.