Reviewer #2 (Public review):

Summary:

Sphingosine-1-phosphate (S1P) metabolic and signaling genes are expressed highly in retinal Müller glia (MG) cells. This study tested how S1P signaling regulates glial phenotype, dedifferentiation of, reprogramming into proliferating MG-derived progenitor cells (MGPCs), and neuronal differentiation of the progeny of MGPCs using in vivo chick retina. Major techniques used are Sc-RNASeq and immunohistochemistry to determine the gene expression and proliferation of MG cells that co-label with signaling antibodies or mRNA FISH following treating the in vivo eyes with various S1P signaling antagonists, agonists, and signal modulators. The major conclusions drawn are supported by the results presented. However, the methodology they have used to modulate the S1P pathway using various chemical drugs raises questions about the outcomes and whether those are the real effects of S1P receptor modulation or S1P synthesis inhibition.

Strengths:

- Use of elaborated single-cell RNAseq expression data.<br /> - Use of FISH for S1P receptors and kinase as a good quality antibody is not available.<br /> - Use of EdU assay in combination with IHC<br /> - Comparison with human and Zebrafish Sc-RNA data

Weaknesses:

The methodology is not very clean. A number of drugs (inhibitors/ antagonists/agonists signal modulators) are used to modulate S1P expression or signaling in the retina without evidence that these drugs are reaching the target cells. No alternative evaluation if the drugs, in fact, are effective. The drug solubility in the vehicle and in the vitreous is not provided, and how did they decide on using a single dose of each drug to have the optimal expected effect on the S1P pathway?

eLife Assessment

This is a useful contribution to our understanding of taste perception. The idea that specific receptors function in the pharynx to mediate responses to carboxylic acids is interesting, although the expression analysis is incomplete. Reviewers also have a number of other suggestions for improvement, including the request that authors provide more details about the methodology used. In general, the claims are supported by solid evidence and add to a growing body of literature on this topic.

Reviewer #1 (Public review):

Summary:

Shrestha et al report an investigation of mechanisms underlying gustatory preference for carboxylic acids in Drosophila. They begin with a screen of selected IR mutants, identifying 5 candidates - 2 IR co-receptors and 3 other IRs - whose loss of function causes defects in feeding preference for one or more of the three tested carboxylic acids. The requirement for IR51b, IR94a, and IR94h in carboxylic acid responses is evaluated in more detail using behavior, electrophysiology (labellar sensilla), and calcium imaging (pharyngeal neurons). The behavioral valence of IR94a and IR94h neurons is assessed using optogenetics. Overall the study uses a variety of approaches to test and validate the requirement of IRs in pharyngeal carboxylic acid taste.

Strengths:

The involvement of the identified IRs in gustatory responses to carboxylic acids is very clear from this study. The authors use mutants and transgenic rescue experiments and evaluate outcomes using electrophysiology, behavior, and imaging. Complementary approaches of loss-of-function and artificial activation support the main conclusion that the identified pharyngeal neurons sense carboxylic acids and convey a positive behavioral valence.

Weaknesses:

Some aspects of expression analysis and calcium imaging need to be clarified to better support the conclusions.

(1) The conclusion of two parallel IR-mediated pathways rests on expression analysis of Ir94a-GAL4 and Ir94h-GAL4 lines and the observation that Ir51b expression driven by either can rescue the Ir51b mutant phenotype. However, the expression analysis is not as rigorous as it needs to be for such a conclusion. Prior work found co-expression of Ir94a and Ir94h in the LSO. Here, the co-expression of the two drivers has not been examined, and Ir94a-GAL4 does not appear to be expressed in the LSO. Given the challenges in validating expression patterns in pharyngeal organs, the possibility that the drivers do not entirely capture endogenous expression cannot be ruled out. Rescue experiments using feeding preference or single-cell imaging don't suffice as validation. Plus, the expression of Ir51b could not be defined.

(2) The description of methods and results for the ex vivo calcium imaging is not satisfactory. Details about which cells are being analyzed, and in which organs are not included. No solvent stimulus is tested. The temporal dynamics of the responses are not presented. Movies of the imaging are not included as supplementary information - it would be important to visualize those with what was considered modest movement.

(3) The observed differences in phenotypes of Ir25a and Ir76b mutants are intriguing, as are those between the co-receptor mutants and Ir51b, Ir94a, and Ir94h, but have not been sufficiently considered. Prior studies have also found roles for other response modes (OFF response), other IRs and GRs, and other organs (labellum, tarsi) in behavioral responses to carboxylic acids. Overall, the authors' model may be overly simplistic, and the discussion does not do justice to how their model reconciles with the body of work that already exists.

Reviewer #2 (Public review):

Shrestha et al investigated the role of IR receptors in the detection of 3 carboxylic acids in adult Drosophila. A low concentration of either of these carboxylic acids added to 2 mM sucrose (1% lactic acid (LA), citric acid (CA), or glycolic acid (GA)) stimulates the consumption of adult flies in choice conditions. The authors use this behavioral test to screen the impact of mutations within 33 receptors belonging to the IR family, a large family of receptors derived from glutamate receptors and expressed both in the olfactory and gustatory sensilla of insects. Within the panel of mutants tested, they observed that 3 receptors (IR25a, IR51b, and IR76b) impaired the detection of LA, CA, and GA, and that 2 others impacted the detection of CA and GA (IR94a and IR94h). Interestingly, impairing IR51b, IR94a, and IR94h did not affect the electrophysiological responses of external gustatory sensilla to LA, CA, and GA. Thanks to the use of GAL4 strains associated with these receptors and thanks to the use of poxn mutants (which do not develop external gustatory sensilla but still have functional internal receptors), they show evidence that IR94a and IR94h are only expressed in two clusters of gustatory neurons of the pharynx, respectively in the VCSO (ventral cibarial sense organ) and in the VCSO + LSO (labral sense organ). As for IR51b, the GAL4 approach was not successful but RT-PCR made on different parts of the insect showed an expression both in the pharyngeal organs and in peripheral receptors. These main findings are then complemented by a host of additional experiments meant to better understand the respective roles of IR94a and IR94h, by using optogenetics and brain calcium imaging using GCamp6. They also report a failed attempt to co-express IR51b, IR94a, and IR94h into external receptors, a co-expression which did not confer the capability of bitter-sensitive cells (expressing GR33a-GAL4) to detect either of the carboxylic acids. These data complete and expand previous observations made on this group and others, and dot to 2 new IR receptors which show an unsuspected specific expression, into organs that still remain difficult to study.

The conclusions of this paper are supported by the data presented, but it remains difficult to make general conclusions as concerns the mechanisms by which carboxylic acids are detected.

(1) All experiments were done with 1% of carboxylic acids. What is the dose dependency of the behavioral responses to these acids, and is it conceivable that other receptors are involved at other concentrations?

(2) One result needs to be better discussed and hypotheses proposed - which is why the mutations of most receptors lead to a loss of detection (mutant flies become incapable of detecting the acid) while mutations in IR94a and IR94h make CA and GA potent deterrents. Does it mean that CA and GA are detected by another set of receptors that, when activated, make flies actively avoid CA and GA? In that case, do the authors think that testing receptors one by one is enough to uncover all the receptors participating in the detection of these substances?

(3) The paper needs to be updated with a recent paper published by Guillemin et al (2024), indicating that LA is detected externally by a combination of IR94e, IR76b and IR25a. IR25a might help to form a fully functional receptor in GR33a neurons (a former study from Chen et al (2017) indicate that IR25a is expressed in all gustatory neurons of the pharynx).

(4) Although it was not the main focus of the paper, it would have been most interesting if the cells expressing IR94a and IR94h were identified, and placed on the functional map proposed by the group of Dahanukar (Chen et al 2017 Cell Reports, Chen et al 2019 Cell Reports).

Reviewer #3 (Public review):

Summary:

In this work, the authors investigated the molecular and cellular basis of sour taste perception in Drosophila melanogaster, focusing on identifying receptors that mediate attractive responses to certain carboxylic acids. It builds on previous work from the same group that had identified the IR co-receptors IR25a and IR76b for this sensory process, screening a set of mutants in IRs to identify three, IR51b, IR94a, and IR94h, required for feeding preference responses to some or all of the tested acids.

Strengths:

The work is of interest because it assigns sensory roles to IRs of previously unknown function, in particular IR94a and IR94h, and points to pharyngeal neurons in which these receptors are expressed as the relevant sensory neurons (potentially with different roles for IR94a- and IR94h-expressing neurons). The work combines elegant genetics, simple but effective feeding and taste assays, chemo-/opto-genetic activation, and some calcium imaging. Overall the presented data look solid and well-controlled.

Weaknesses:

The in situ expression analysis relies entirely on transgenic driver lines for IR94a and IR94h (which had been previously described, though not fully cited in this work). Importantly, given that many of the behavioral experiments (genetic rescue, physiology, artificial activation) use the IR94a and IR94h GAL4 driver lines, it would be helpful to validate that these faithfully reflect IR94a and IR94h expression (as far as I can tell, such validation wasn't done in the original papers describing these lines as part of a large collection of IR drivers). For IR51b, pharyngeal expression is concluded indirectly from non-quantitative RT-PCR analysis (genetic reporters did not work). The lack of direct detection of gene/protein expression (for example, through RNA FISH, immunofluorescence, or protein tagging) would have made for a more complete characterization of these receptors (for example, there is no direct evidence that they also express IR25a and IR76b, as one might expect). Finally, the relationship of IR94a and IR94h neurons to other types of pharyngeal neurons remains unclear, as are their projection patterns in the SEZ.

Conceptually, the work is of interest mostly to those in the immediate field; there have been a very large number of studies in the past decade (several from this lab) characterizing the contributions of different IRs to various chemosensory processes. The current work doesn't lend much insight into the nature of the minimal functional unit of gustatory IRs (reconstitution of a functional IR in a heterologous neuron/cell has not been achieved here, but this is a limitation of many other previous studies), nor to how different pharyngeal sensory pathways might collaborate to control behavior. Nevertheless, the findings provide a useful contribution to the literature.

Author response:

Reviewer #1 (Public review):

Summary:

Shrestha et al report an investigation of mechanisms underlying gustatory preference for carboxylic acids in Drosophila. They begin with a screen of selected IR mutants, identifying 5 candidates - 2 IR co-receptors and 3 other IRs - whose loss of function causes defects in feeding preference for one or more of the three tested carboxylic acids. The requirement for IR51b, IR94a, and IR94h in carboxylic acid responses is evaluated in more detail using behavior, electrophysiology (labellar sensilla), and calcium imaging (pharyngeal neurons). The behavioral valence of IR94a and IR94h neurons is assessed using optogenetics. Overall the study uses a variety of approaches to test and validate the requirement of IRs in pharyngeal carboxylic acid taste.

Strengths:

The involvement of the identified IRs in gustatory responses to carboxylic acids is very clear from this study. The authors use mutants and transgenic rescue experiments and evaluate outcomes using electrophysiology, behavior, and imaging. Complementary approaches of loss-of-function and artificial activation support the main conclusion that the identified pharyngeal neurons sense carboxylic acids and convey a positive behavioral valence.

Weaknesses:

Some aspects of expression analysis and calcium imaging need to be clarified to better support the conclusions.

(1) The conclusion of two parallel IR-mediated pathways rests on expression analysis of Ir94a-GAL4 and Ir94h-GAL4 lines and the observation that Ir51b expression driven by either can rescue the Ir51b mutant phenotype. However, the expression analysis is not as rigorous as it needs to be for such a conclusion. Prior work found co-expression of Ir94a and Ir94h in the LSO. Here, the co-expression of the two drivers has not been examined, and Ir94a-GAL4 does not appear to be expressed in the LSO. Given the challenges in validating expression patterns in pharyngeal organs, the possibility that the drivers do not entirely capture endogenous expression cannot be ruled out. Rescue experiments using feeding preference or single-cell imaging don't suffice as validation. Plus, the expression of Ir51b could not be defined.

Based on current literature, Ir94a and Ir94h exhibit distinct expression patterns localized to different sensory regions. Specifically, Ir94a is primarily expressed in the V5 region of the VCSO, where it co-localizes with Ir94c-GAL4 (Chen et al., 2017). Conversely, Ir94h is found in the L7-7 sensilla of the LSO, where it co-expresses with Ir94f, and also within the V2 cells of the VCSO. Notably, the projections of Ir94a and Ir94h into the dorso-anterior subesophageal ganglion suggest divergent expression patterns rather than co-expression in the pharyngeal regions (Koh et al., 2014). Regarding co-expression of Ir94a and Ir94h in the LSO, we did not find any evidence to support this claim. Our data reinforce this view, showing that Ir94a-GAL4 expression is limited to the VCSO, while Ir94h-GAL4 is present in both the LSO and VCSO. Thus, the notion of co-expression of Ir94a and Ir94h in the LSO is not substantiated by current evidence.

As a reviewer suggested, it is possible that the GAL4 drivers utilized may not fully reflect the endogenous expression of these receptors. Despite this limitation, our behavioral, expression, and physiological analyses strongly suggest that Ir94a and Ir94h are located in distinct regions, supporting a model of two parallel IR-mediated pathways operating within the sensory system.

In addition, RT-PCR analysis confirmed the presence of Ir51b. However, due to methodological constraints, we were unable to conduct cell-type-specific expression studies using Ir51b-GAL4. This limitation, which we have acknowledged in the manuscript, does not detract from our core findings but highlights an area for future research. Further studies utilizing cell-specific expression analysis and co-expression studies with additional drivers could offer more definitive insights into IR51b’s functional role and its interactions within broader IR-mediated pathways.

(2) The description of methods and results for the ex vivo calcium imaging is not satisfactory. Details about which cells are being analyzed, and in which organs are not included. No solvent stimulus is tested. The temporal dynamics of the responses are not presented. Movies of the imaging are not included as supplementary information - it would be important to visualize those with what was considered modest movement.

We appreciate this valuable feedback. As discussed above, Ir94h is specifically expressed in the L7-7 sensilla of the LSO, while Ir94a is expressed in the V2 cells of the VCSO. This evidence led us to focus specifically on these cells in our calcium imaging study to ensure accuracy and relevance. In our experiments, Adult hemolymph solution (AHL) (108 mM NaCl, 5 mM KCl, 8.2 mM MgCl2, 2 mM CaCl2, 4 mM NaHCO3, 1 mM NaH2PO4, 5 mM HEPES, pH 7.5) was used as the solvent and employed as a pre-stimulus (as mentioned in the Methods section). During this phase, we observed no changes in fluorescence, indicating that AHL itself did not influence the responses. Fluorescence changes occurred only when the test chemical, dissolved in AHL, was introduced. To further confirm that AHL had no impact on the results, we conducted continuous recordings with AHL alone before beginning our main experiments, and these trials confirmed the absence of fluorescence alterations. We have included the temporal dynamics and supplementary video recordings to provide a more comprehensive understanding of our findings.

(3) The observed differences in phenotypes of Ir25a and Ir76b mutants are intriguing, as are those between the co-receptor mutants and Ir51b, Ir94a, and Ir94h, but have not been sufficiently considered. Prior studies have also found roles for other response modes (OFF response), other IRs and GRs, and other organs (labellum, tarsi) in behavioral responses to carboxylic acids. Overall, the authors' model may be overly simplistic, and the discussion does not do justice to how their model reconciles with the body of work that already exists.

Stanley et al. (2021) reported that the gustatory detection of lactic acid requires both IRs and GRs functioning together. Specifically, they found that IR25a mediates the onset peak response (ON response) to lactic acid, while GRs dampen this response and contribute to a removal peak (OFF response). Interestingly, in Ir25a mutants, a small onset peak still occurred, while Gr64a-f mutants showed an enhanced onset, suggesting that IRs and GRs interact dynamically to modulate taste responses.

In our previous work, we also observed the role of sweet GRs, in addition to Ir25a and Ir76b, in detecting carboxylic acids in the labellum (Shrestha et al., 2021). This raises the possibility of a similar interplay with carboxylic acids in our current study, where different IRs may contribute to distinct aspects of sensory responses in the pharynx, leading to the phenotypic differences we observed. Moreover, Chen et al. (2017) demonstrated that sour-sensing neurons in the tarsi express both IR76b and IR25a and specifically respond to carboxylic and inorganic acids without reacting to sweet or bitter compounds. This finding points to a specialized role for these receptors in sour detection and suggests a coordinated response involving multiple sensory organs—such as the labellum, tarsi, and pharynx.

The phenotypic differences observed in our mutants align with a more integrated model of carboxylic acid detection, in which multiple receptors and sensory organs contribute to the overall behavioral response. This supports the idea that our current model offers a more detailed understanding of how different carboxylic acids are detected and processed by the gustatory system.

Reviewer #2 (Public review):

Shrestha et al investigated the role of IR receptors in the detection of 3 carboxylic acids in adult Drosophila. A low concentration of either of these carboxylic acids added to 2 mM sucrose (1% lactic acid (LA), citric acid (CA), or glycolic acid (GA)) stimulates the consumption of adult flies in choice conditions. The authors use this behavioral test to screen the impact of mutations within 33 receptors belonging to the IR family, a large family of receptors derived from glutamate receptors and expressed both in the olfactory and gustatory sensilla of insects. Within the panel of mutants tested, they observed that 3 receptors (IR25a, IR51b, and IR76b) impaired the detection of LA, CA, and GA, and that 2 others impacted the detection of CA and GA (IR94a and IR94h). Interestingly, impairing IR51b, IR94a, and IR94h did not affect the electrophysiological responses of external gustatory sensilla to LA, CA, and GA. Thanks to the use of GAL4 strains associated with these receptors and thanks to the use of poxn mutants (which do not develop external gustatory sensilla but still have functional internal receptors), they show evidence that IR94a and IR94h are only expressed in two clusters of gustatory neurons of the pharynx, respectively in the VCSO (ventral cibarial sense organ) and in the VCSO + LSO (labral sense organ). As for IR51b, the GAL4 approach was not successful but RT-PCR made on different parts of the insect showed an expression both in the pharyngeal organs and in peripheral receptors. These main findings are then complemented by a host of additional experiments meant to better understand the respective roles of IR94a and IR94h, by using optogenetics and brain calcium imaging using GCamp6. They also report a failed attempt to co-express IR51b, IR94a, and IR94h into external receptors, a co-expression which did not confer the capability of bitter-sensitive cells (expressing GR33a-GAL4) to detect either of the carboxylic acids. These data complete and expand previous observations made on this group and others, and dot to 2 new IR receptors which show an unsuspected specific expression, into organs that still remain difficult to study.

The conclusions of this paper are supported by the data presented, but it remains difficult to make general conclusions as concerns the mechanisms by which carboxylic acids are detected.

(1) All experiments were done with 1% of carboxylic acids. What is the dose dependency of the behavioral responses to these acids, and is it conceivable that other receptors are involved at other concentrations?

In our study, we conducted experiments to examine the dose dependency of behavioral responses to carboxylic acids, with results presented in Supplementary Figure 1. We found that lower concentrations of carboxylic acids are perceived as attractive, while higher concentrations are aversive. This differential response suggests that the receptors identified in our study are primarily tuned to detect low concentrations of these acids. Since higher concentrations elicited aversive responses, it is plausible that additional receptors, beyond the scope of our study, may be involved in sensing these higher concentrations. These receptors could be part of other gustatory receptor neurons that respond specifically to increased acid levels, as fruit flies tend to avoid higher concentrations. We propose that future research could investigate these alternative pathways to gain a complete understanding of the behavioral responses to carboxylic acids. In summary, our findings suggest that specific receptors are involved in detecting low concentrations, while distinct receptor pathways—possibly mediated by other GRNs—may regulate responses to higher concentrations.

(2) One result needs to be better discussed and hypotheses proposed - which is why the mutations of most receptors lead to a loss of detection (mutant flies become incapable of detecting the acid) while mutations in IR94a and IR94h make CA and GA potent deterrents. Does it mean that CA and GA are detected by another set of receptors that, when activated, make flies actively avoid CA and GA? In that case, do the authors think that testing receptors one by one is enough to uncover all the receptors participating in the detection of these substances?

As we mentioned above, it is possible that distinct receptor pathways mediate avoidance of GA and CA. This suggests that CA and GA might activate different sets of receptors that trigger avoidance behavior, pointing to a more complex interplay of receptor activity than we initially considered. Certain acids may indeed be detected by multiple receptors, with each receptor contributing uniquely to the behavioral response. Regarding the sufficiency of testing receptors individually, we recognize the limitations of this approach. Examining receptors one by one may not reveal the full spectrum of receptors involved, especially due to potential interactions or compensatory mechanisms that only emerge when certain receptors are inactive. Therefore, a more holistic approach—such as genetic screens for behavioral responses or using complex genetic models to disrupt multiple receptors simultaneously—could provide deeper insights. Moving forward, incorporating receptor interactions that modulate each other, along with more comprehensive assays, could help explain these discrepancies by uncovering previously overlooked receptor functions.

(3) The paper needs to be updated with a recent paper published by Guillemin et al (2024), indicating that LA is detected externally by a combination of IR94e, IR76b and IR25a. IR25a might help to form a fully functional receptor in GR33a neurons (a former study from Chen et al (2017) indicate that IR25a is expressed in all gustatory neurons of the pharynx).

According to Guillemin et al. (2024), the combination of IR94e, IR76b, and IR25a is required for amino acid detection but not for detecting lactic acid (LA). In their calcium imaging experiments, 100 mM LA elicited a response similar to the vehicle control, suggesting that these receptors do not play a role in LA detection.

(4) Although it was not the main focus of the paper, it would have been most interesting if the cells expressing IR94a and IR94h were identified, and placed on the functional map proposed by the group of Dahanukar (Chen et al 2017 Cell Reports, Chen et al 2019 Cell Reports).

The expression patterns of IR94a and IR94h were previously detailed by Chen et al. (2017), showing that IR94h is expressed in the labial sense organ (LSO, specifically in L7-7) and the ventral cibarial sense organ (VCSO, V2), while IR94a is expressed in the VCSO (V5). Given this established information, we referenced these known expression patterns without replicating the mapping in our study. Our primary focus was to investigate the functional role of these neurons within the pharynx, and we believe we have successfully highlighted their specific contributions. However, we recognize that integrating the functional mapping of these neurons in alignment with the work of Dahanukar’s group would have strengthened our findings and provided a more comprehensive understanding. We acknowledge this as a limitation of our study and appreciate your suggestion, as it points to a valuable direction for future research.

Reviewer #3 (Public review):

Summary:

In this work, the authors investigated the molecular and cellular basis of sour taste perception in Drosophila melanogaster, focusing on identifying receptors that mediate attractive responses to certain carboxylic acids. It builds on previous work from the same group that had identified the IR co-receptors IR25a and IR76b for this sensory process, screening a set of mutants in IRs to identify three, IR51b, IR94a, and IR94h, required for feeding preference responses to some or all of the tested acids.

Strengths:

The work is of interest because it assigns sensory roles to IRs of previously unknown function, in particular IR94a and IR94h, and points to pharyngeal neurons in which these receptors are expressed as the relevant sensory neurons (potentially with different roles for IR94a- and IR94h-expressing neurons). The work combines elegant genetics, simple but effective feeding and taste assays, chemo-/opto-genetic activation, and some calcium imaging. Overall the presented data look solid and well-controlled.

Weaknesses:

The in situ expression analysis relies entirely on transgenic driver lines for IR94a and IR94h (which had been previously described, though not fully cited in this work). Importantly, given that many of the behavioral experiments (genetic rescue, physiology, artificial activation) use the IR94a and IR94h GAL4 driver lines, it would be helpful to validate that these faithfully reflect IR94a and IR94h expression (as far as I can tell, such validation wasn't done in the original papers describing these lines as part of a large collection of IR drivers). For IR51b, pharyngeal expression is concluded indirectly from non-quantitative RT-PCR analysis (genetic reporters did not work). The lack of direct detection of gene/protein expression (for example, through RNA FISH, immunofluorescence, or protein tagging) would have made for a more complete characterization of these receptors (for example, there is no direct evidence that they also express IR25a and IR76b, as one might expect). Finally, the relationship of IR94a and IR94h neurons to other types of pharyngeal neurons remains unclear, as are their projection patterns in the SEZ.

Conceptually, the work is of interest mostly to those in the immediate field; there have been a very large number of studies in the past decade (several from this lab) characterizing the contributions of different IRs to various chemosensory processes. The current work doesn't lend much insight into the nature of the minimal functional unit of gustatory IRs (reconstitution of a functional IR in a heterologous neuron/cell has not been achieved here, but this is a limitation of many other previous studies), nor to how different pharyngeal sensory pathways might collaborate to control behavior. Nevertheless, the findings provide a useful contribution to the literature.

We appreciate your thoughtful feedback. As noted in our response, our primary objective was to investigate the sensory functions of IR94a and IR94h. To this end, we conducted behavioral assays, which we validated with additional approaches including genetic rescue, physiological tests, and artificial activation. Throughout these experiments, we extensively utilized Ir94a- and Ir94h-GAL4 driver lines. To ensure these lines accurately reflect the expression of IR94a and IR94h, we verified their expression patterns using immunohistochemistry across various body parts. Our results align with previous findings that show both receptors are exclusively expressed in the pharynx. Regarding IR51b, we employed RT-PCR due to its high sensitivity and specificity, which supported our hypothesis. Nonetheless, we agree that more direct detection methods would have provided a stronger validation of IR51b expression. Our previous study (Sang et al., 2024) also demonstrated the pharyngeal expression of co-expressed receptors, specifically IR25a and IR76b. However, we recognize that the lack of direct evidence for their co-expression with IR51b remains a significant gap. This limitation primarily stems from the unavailability of specific reagents needed for direct assays targeting IR51b, which restricted our experimental approach.

You also raised the potential relationship between IR94a and IR94h neurons and other pharyngeal neuron types, including their projection patterns in the subesophageal zone. This is indeed an important area for future research that could clarify neural connectivity and further our understanding of sensory mechanisms. However, our study was focused on exploring sensory mechanisms in peripheral regions rather than detailed neural mapping in the SEZ. Investigating these connections would undoubtedly provide valuable insights into the neural circuitry involved and represents an intriguing direction for future research.

Reviewer #2 (Public review):

Summary:

Salt stress is a significant and growing concern for agriculture in some parts of the world. While the effects of sodium excess have been studied in Arabidopsis and (many) crop species, most studies have focused on Na uptake, toxicity and overall effects on yield, rather than on developmental responses to excess Na, per se. The work by Ishka and colleagues aims to fill this gap.

Working from an existing dataset that exposed a diverse panel of A. thaliana accessions to control, moderate, and severe salt stress, the authors identify candidate loci associated with altering the root:shoot ratio under salt stress. Following a series of molecular assays, they characterize a DUF247 protein which they dub SR3G, which appears to be a negative regulator of root growth under salt stress.

Overall, this is a well-executed study which demonstrates the functional role played by a single gene in plant response to salt stress in Arabidopsis.

Review of revised manuscript:

The authors have addressed my point-by-point comments to my satisfaction. In the cases where they have changed their manuscript language, clarified figures, or added analyses I have no further comment. In some cases, there is a fruitful back-and-forth discussion of methodology which I think will be of interest to readers.

I have nothing to add during this round of review. I think that the paper and associated discussion will make a nice contribution to the field

Author response:

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

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The authors aim to assess the effect of salt stress on root:shoot ratio, identify the underlying genetic mechanisms, and evaluate their contribution to salt tolerance. To this end, the authors systematically quantified natural variations in salt-induced changes in root:shoot ratio. This innovative approach considers the coordination of root and shoot growth rather than exploring biomass and the development of each organ separately. Using this approach, the authors identified a gene cluster encoding eight paralog genes with a domain-of-unknown-function 247 (DUF247), with the majority of SNPs clustering into SR3G (At3g50160). In the manuscript, the authors utilized an integrative approach that includes genomic, genetic, evolutionary, histological, and physiological assays to functionally assess the contribution of their genes of interest to salt tolerance and root development.

Strengths:

The holistic approach and integrative methodologies presented in the manuscript are essential for gaining a mechanistic understanding of a complex trait such as salt tolerance. The authors focused on At3g50160 but included in their analyses additional DUF247 paralogs, which further contributes to the strength of their approach. In addition, the authors considered the developmental stage (young seedlings, early or late vegetative stages) and growth conditions of the plants (agar plates or soil) when investigating the role of SR3G in salt tolerance and root or shoot development.

Weaknesses:

The authors' claims and interpretation of the results are not fully supported by the data and analyses. In several cases, the authors report differences that are not statistically significant (e.g., Figures 4A, 7C, 8B, S14, S16B, S17C), use inappropriate statistical tests (e.g., t-test instead of Dunnett Test/ANOVA as in Figures 10B-C, S19-23), present standard errors that do not seem to be consistent with the post-hoc Tukey HSD Test (e.g., Figures 4, 9B-C, S16B), or lack controls (e.g., Figure 5C-E, staining of the truncated versions with FM4-64 is missing).

We thank the reviewer for their critical thoughts on the presented data. We have revised our data interpretation in the main text to more accurately reflect the results. Given the nature of our experimental setup, where we trace the roots of individual Arabidopsis seedlings grown on plates, there is considerable biological variation, which makes achieving strong statistical significance between samples or genotypes challenging. However, we think that the representation of the data as transparently as possible is necessary to provide the readers and reviewers a true picture of the variability that we are observing.  Consequently, we have centered our data interpretation around observable trends that facilitate drawing conclusions.

The choice of statistical test is closely tied to the specific biological question being addressed. In Figures 10A-C, as in Figures 6A-B, we compared all genotypes to the wild-type Col-0 within each condition, and thus ANOVA analysis, testing the general effect of the genotype across both mutants and Col-0 wild-type is not appropriate. Similarly, in Figures S19-S23, we compared each mutant line to the wild-type Col-0 under each condition.

We repeated the post-hoc Tukey HSD Test for Figures 4, 9B-C, and S16B and made adjustments where necessary (see tracked changes manuscript).

The truncated versions do not localize to the plasma membrane; instead, they are targeted to the nucleus and cytosol, mimicking the localization pattern of free GFP, which was used as a control in Panel F. Therefore, we believe that having FM4-64 as a control for these specific images is not informative, but instead using free GFP is serving as a better control in that particular construct.

In other cases, traits of root system architecture and expression patterns are inconsistent between different assays despite similar growth conditions (e.g., Figures S17A-B vs. 10A-C vs. 6A, and Figures S16B vs. 4A/9B), or T-DNA insertion alleles of WRKY75 that are claimed to be loss-of-function show comparable expression of WRKY75 as WT plants. Additionally, several supplemental figures are mislabeled (Figures S6-9), and some figure panels are missing (e.g., Figures S16C and S17E).

We thank the reviewer for raising these points and noticing the inconsistency between different assays (e.g., Figures S17A-B vs. 10A-C vs. 6A, and Figures S16B vs. 4A/9B). As mentioned above, considerable biological variation makes achieving strong statistical significance between samples, genotypes, or experiments challenging. Thus, we have centered our data interpretation around observable “trends” between experiments to facilitate drawing conclusions. Considering Figures S17A-B, 10A-C, and 6A, we acknowledge the reviewer's concern about inconsistencies in root system architecture across experiments. Initially, we observed that the sr3g mutant had reduced lateral root length compared to Col-0 under salt stress. This led us to focus on this specific phenotypic trait rather than the overall root system architecture. Despite some variation, the sr3g mutant consistently showed a similar trend/phenotype when compared to Col-0 under salt stress. We believe the variation in main root length and lateral root number between experiments is due to inherent differences between biological replicates.

Regarding gene expression patterns between Figures S16B and 4A/9B, we included part of Figure 9B (SR3G gene expression in Col-0) in Figure 4A. Figure S16B represents a completely different assay. Despite variations between assays, the overall message remains consistent: SR3G gene expression is induced under salt stress in the root but not in the shoot.

Both SR3G and WRKY75 are expressed at very low levels, even under the 75 mM salt stress condition we tested. When gene expression is so low, detecting changes is challenging due to inherent variations. Nonetheless, we observed a reduction in WRKY75 expression in the mutant lines compared to wild-type Col-0, though this reduction was not statistically significant. More importantly, we observed a similar phenotype in the wrky75 mutant, specifically reduced main root length under salt stress, consistent with the findings of the published paper in The Plant Cell by Lu et al. (2023) “Lu, K.K., Song, R.F., Guo, J.X., Zhang, Y., Zuo, J.X., Chen, H.H., Liao, C.Y., Hu, X.Y., Ren, F., Lu, Y.T. and Liu, W.C., 2023. CycC1; 1–WRKY75 complex-mediated transcriptional regulation of SOS1 controls salt stress tolerance in Arabidopsis. The Plant Cell, 35(7), pp.2570-2591”.

We appreciate the reviewer for spotting the missing labels for Figures S6-9. We corrected them at the main text, figures, and legends. We added panel C to Figure S16 and removed panel E from Figure S17 legend,  now they match to actual figures and legends.

Consequently, the authors' decisions regarding subsequent functional assays, as well as major conclusions about gene function, including SR3G function in root system architecture, involvement in root suberization, and regulation of cellular damage are incomplete.

We greatly appreciate the reviewer's thorough review of our manuscript and their critical comments. We have carefully addressed all comments and concerns.

Reviewer #2 (Public Review):

Salt stress is a significant and growing concern for agriculture in some parts of the world. While the effects of sodium excess have been studied in Arabidopsis and (many) crop species, most studies have focused on Na uptake, toxicity, and overall effects on yield, rather than on developmental responses to excess Na, per se. The work by Ishka and colleagues aims to fill this gap.

Working from an existing dataset that exposed a diverse panel of A. thaliana accessions to control, moderate, and severe salt stress, the authors identify candidate loci associated with altering the root:shoot ratio under salt stress. Following a series of molecular assays, they characterize a DUF247 protein which they dub SR3G, which appears to be a negative regulator of root growth under salt stress.

Overall, this is a well-executed study that demonstrates the functional role played by a single gene in plant response to salt stress in Arabidopsis.

The abstract and beginning of the Discussion section highlight the "new tool" developed here for measuring biomass accumulation. I feel that this distracts from the central aims of the study, which is really about the role of a specific gene in root development under salt stress. I would suggest moving the tool description to less prominent parts of the manuscript.

We appreciate the reviewer's suggestion. We believe that the innovative tool used to extract shoot-to-root ratio data from previous experiments underscores the value of reutilizing previously acquired data for new discoveries and demonstrates how reanalyzing the same data can provide fresh insights, such as identification of new allelic variation. Therefore, we decided to retain this section, as our discovery of the SR3G gene originated from this innovative tool.

Recommendations for the authors:

Reviewer #3 (Recommendations For The Authors):

Line 58 (opening sentence) - salt accumulation in the soil is not caused by evaporation exceeding input; that scenario results in soil water deficit. The issue is when the input water has dissolved ions.

We thank the reviewer for raising this important point. While this point is theoretically true, all of the water that is found in natural environments contains some dissolved ions. Therefore, drought conditions will lead, over time, to increased soil salinization. We have amended this sentence to represent our point better.

“Salt stress is predominant in the dryland areas where evaporation rate exceeds water input. As all water contains dissolved ions, the prolonged exposure to drought stress results in increased accumulation of salts in the upper soil layers 1–3.”

I feel that it would be helpful, for replication and for interpretation, if the authors could provide water potentials for the growing media used throughout. What water potentials are the plants experiencing when grown in 1/2 MS + agar at 0, 75, and 150mM NaCl? Juenger and Verslues present a great recent discussion of the importance of reporting these values (Juenger, T. E. and P. E. Verslues (2023). "Time for a drought experiment: Do you know your plants' water status?" Plant Cell 35(1): 10-23.)

Critically, how do the water potentials experienced by agar-grown plants compare to those experienced in soil-grown plants? As a stated aim of this study is to allow translation to crops these data are very important to convince physiologists of the relevance of the results.

We thank the reviewer for raising this important point. We completely agree that growing plants on agar plates is an artificial setup and knowing the water potential of the plants within this setup would be highly informative. However, as indicated in review by Juenger and Verslues 2023, the agar plate setup is much more reproducible compared to various soil conditions, and we report the media composition in sufficient detail for it to be reproduced in other laboratory conditions.

Furthermore, while investigating the water status of plants and soil is indeed intriguing, it is beyond the scope of this study and would require us to redo the experiments with specific tools listed within the Juennger and Verslues review, which are currently not within our laboratory equipment list.

Importantly, any changes reported in this manuscript apply equally to both wild-type and mutant lines under all conditions. We provide extensive report on the soil type used, as well as soil quantity. We are using the gravimetric method to determine the water content, and salt stress application, as described in previous works from our lab (Yu and Sussman et al., 2024 Plant Physiology and Awlia et al., 2016 Frontiers in Plant Science). 

Nonetheless, we have now included water content measurements for soil-grown plants under different conditions, calculated by subtracting dry weight from fresh weight (new Fig. S24). Although plant water content may not fully capture the water status of the media or soil, our measurements did not reveal any significant differences in water content between genotypes across the various conditions tested.

Line 69- missing an "and" after "(ABA)."

Thanks. We added the missing “and”.

Line 79 - I think the association being made is between natural variation in root and shoot growth and genetic variants, not "underlying genes."

We thank the reviewer for this suggestion. The cause for the identified association indeed relies on allelic variation within the genetic region. We have re-phrased this sentence within the manuscript.

“Many forward genetic studies were highly successful in associating natural variation in root and shoot growth with allelic variation in gene coding and promoter regions, thereby identifying potential new target traits for improved stress resilience 18,20,21.”

Figure 1 - what do "seGF" and "reGF" stand for? Shoot and root growth rate, respectively, but there are extra letters in there…

The abbreviations stand for shoot exponential Growth Factor and root exponential Growth factor. An explanation of the acronym has been added to the text.

“The increase in the projected area of shoot and root (Fig. S2) was used to estimate (A) shoot and (B) root exponential growth rate (seGR and reGR respectively).”

Figure 1 legend - there's an "s" missing in "across." And two "additionally" in the penultimate sentence.

Thanks for spotting the errors. We fixed these errors.

Line 109 - how was the white balance estimated for the images on the flatbed scanner?

Within the developed tool, we have not adjusted or controlled for white balance in any way, as the white balance from the flatbed scanner is kept at one value. The tool transforms the imaged pixels into bins consisting of white (root), green (shoot), and blue (place) pixels based on the closest distance in the RGB scale to the particular color, which makes correcting for white balance obsolete. We have provided an additional explanation for this within the M&M section.

“A Matlab-based tool was developed to simplify and speed up the segmentation and analysis pipeline. For automatic segmentation, the tool uses a combination of image operations (histogram equalization), thresholding on different color spaces (e.g., RGB, YCbCr, Lab, HSV), and binary image processing (boundary and islands removal). As the tool is digitalizing various color scales and classifies pixels into either white (root), green (shoot) or blue (background) categories, the adjustment for white balance is obsolete. ”

GWAS was performed separately on traits measured at control, 75mM, and 150mM NaCl treatments. Would it also be informative to map the STI measurement (i.e. plasticity) introduced here?

We thank the reviewer for this important point. We have performed GWAS on both “raw” and STI traits, however, we found that the identified associations were not as abundant as the ones identified with “raw traits”. This makes sense, as we are compounding the root or shoot growth under both conditions, and plastic responses to the environment are expected to be genetically more complex, as they involve more genetic regulators compared to phenotypes that have low plasticity. We have added this as a part of the result description, as we acknowledge that this might be an interesting observation for the field to build upon, and might provide fodder for new methods to deconvolute the complexity in mapping the plastic traits. 

“To identify genetic components underlying salt-induced changes in root:shoot ratio, we used the collected data as an input for GWAS. The associations were evaluated based on the p-value, the number of SNPs within the locus, and the number of traits associated with individual loci. As Bonferroni threshold differs depending on the minor allele count (MAC) considered, we identified significant associations based on a Bonferroni threshold for each subpopulation of SNPs based on MAC (Table S3). While we conducted a GWAS on directly measured traits, as well as their Salt Tolerance Index (STI) values, however the amount of associations with STI was much lower compared to directly measured traits (Table S3). This observation aligns with the understanding that plastic responses to environmental conditions tend to be genetically more complex. This complexity likely stems from the involvement of more genetic regulators compared to low-plasticity phenotypes.”

Line 167 - how was LD incorporated into this analysis? Did you use a genome average? Or was LD allowed to vary (as it does) across the genome?

Initially, we have used genome average LD for this purpose (10 kbp for Arabidopsis), and extended the region of interest based on the number of coding genes within the window. We have added this as a part of description to our manuscript.

“For the most promising candidate loci (Table S4), we have identified the gene open reading frames that were located within the genome-wide linkage-disequilibrium (LD) of the associated SNPs. The LD was expanded if multiple SNPs were identified within the region, and the region of interest was expanded based on the number of coding genes within the LD window. ”

Line 291 - I think the water potentials are essential, here. What does 50% of soil water holding capacity equal in these soils? In the substrate that we use in our lab, that would represent a considerable soil water deficit even without any salts in the soil.

We thank the reviewer for this comment. As Arabidopsis is occurring naturally in low soil water holding capacity soils (i.e. sandy soils), it is typically growing better in soils that are not very saturated with the water. Throughout many experiments, performed within this study, and other studies performed in our lab (results reported in Awlia et al., 2016 Frontiers in Plant Science and Yu & Sussman et al., 2024 Plant Physiology), we have not observed any drought like symptoms at 50% soil water holding capacity. The fact that this is reproducible across similar soil types across two laboratories (one in Saudi Arabia and one in the USA) is not to be dismissed. Again - we are currently not equipped to measure water potentials for these plants, as this is not a standard practice (yet) for stress experiments, but we are taking these comments on board for all of our future experiments.

Moreover, our control plants are also “dried down” to 50% of SWHC, and soaked in non-saline water during the “salt stress treatment” to make sure that the soil water saturation is accounted for within the experimental setup. This “dry down” of soil is necessary to ensure equal and effective salt penetration into the soil particles. More details on this method can be found in Awlia et al., 2016.

Again - We have added a new dataset measuring water content in individually soil-grown plants under different conditions as a proxy for soil water status (see new Fig. S24). While we did not observe any significant differences in water content between genotypes under the various conditions, the sr3g mutant showed a slightly higher, though non-significant, water content compared to wild-type Col-0 under control conditions.

We have provided additional information and comments to warn the readers about this method:

“The seeds were germinated in ½ MS media for one week, as described for the agar-based plate experiments. One week after germination, the seedlings were transplanted to the pot (12 x 4 cm insert) containing the Cornell Mix soil (per batch combine: 0.16 m3 of peat moss, 20.84 kg of vermiculite, 0.59 kg of Uni-Mix fertilizer, and 2.27 kg of lime) watered to 100% water holding capacity and placed in the walk-in growth chamber with the 16 h light / 8 h dark period, 22°C and 60% relative humidity throughout the growth period. When all of the pots dried down to the weight corresponding to 50% of their water holding capacity, they were soaked for 1 h in tap water or a 200 mM NaCl solution, resulting in an effective concentration of 100 mM NaCl based on the 50% soil water holding capacity, which corresponded to a moderate level of salt stress (Awlia et al., 2016). The control pots were soaked for the same length of time in 0 mM NaCl solution, to account for the soil saturation effect. We then allowed the pots to be drained for 2-3 h to eliminate excess moisture. The pots were placed under phenotyping rigs equipped with an automated imaging system (Yu et al., 2023) and the pot weight was measured daily to maintain the reference weight corresponding to 50% of the soil water holding capacity throughout the experiment. We would like to note that this gravimetric based method for application of salt stress has been developed for soils typically used for pot-grown plants, with relatively high water holding capacity (Awlia et al. 2016). Within these specific conditions, no drought stress symptoms were observed.”

Lines 415-416 - are these contrasts significant? Figure S3 likewise does not have any notation for significant differences in the means.

We have previously not tested the stronger effect of 125 mM vs 75 mM on relative root and shoot growth, and thus these test results were initially not included in Fig. S3. We have now added the tests and included them within Fig. S3, and added description of their significance into the main body of the manuscript:

“In comparison, the growth rates of the shoot were significantly reduced to 0.71 and 0.43 of the control in 75 and 125 mM NaCl treatments, respectively (Fig. S3). While the mean value of root:shoot growth rate did not change upon salt stress treatment, the variance in the root:shoot ratio significantly expanded with the increasing concentrations of salt (Fig. 1C). These results suggest that while root and shoot growth are well coordinated under non-stress conditions, salt stress exposure results in loss of coordination of organ growth across Arabidopsis accessions.”

Line 418 - same comment as preceding. Is this change in variance significant?

We have previously not tested this. We have now added the ANOVA tests and included them within each figure, and added description of their significance into the main body of the manuscript. (see text above)

Line 421 - why would we expect there to be a correlation between root:shoot growth ratio and seedling size?

We were trying to use the seedling size as a proxy for “fitness” - or how well the plants can survive under these specific conditions. We were testing here whether any simple and directional strategy - such as increase or decrease in root:shoot ratio under salt stress - is resulting in better salt tolerance - which would translate into larger overall seedlings. We have rephrased this within the manuscript, to better explain the hypothesis being tested within this specific figure:

“To test whether there is a clear directional correlation between the change in root:shoot ratio and overall salt stress tolerance, we have used the overall seedling size as a proxy for plant salt tolerance (Fig. S4, S5). No significant correlation was found between the root:shoot growth ratio and total seedling size (Fig. S4, S5), indicating that the relationship between coordination of root and shoot growth and salt tolerance during the early seedling establishment is complex.”

Line 438 - I think a stable web link would be more appropriate than listing Dr. Nordborg's email address.

Sorry about this. There is a glitch with our reference citing software. We agree, and thank the reviewer for noticing this! We assigned reference number 43 to it.

Line 439 - I expect that many of your readers may not be experienced with GWAS. Can you provide an explanation as to why only one locus was detected with both the 250K SNP panel and the 4M SNP panel?

We thank the reviewer for raising this point. We have added additional explanation to this observation:

“Increased SNP density can provide more potential associations, highlighting the associated loci with more confidence, due to more SNPs being detected within specific region. The different panels could capture different LD blocks across the genome. If the locus detected by both panels is in a region of strong LD or under selection, it could be detected consistently. In contrast, other loci may not be captured well by the lower-density 250K SNP panel. The new GWAS revealed 32 additional loci, with only one significantly associated locus being picked up by both 250k and 4M SNPs GWAS (locus 30, Table S3). The detection of only one common locus between the two SNP panels is likely due to differences in resolution, statistical power, and how well each panel captures the genomic regions associated with the trait. ”

Figure 2A and B - I suggest adding the p-value cutoff to the y-axis of the Manhattan Plots

We thank the reviewer for this suggestion, however this is not appropriate. The genome wide p-value cutoffs for GWAS studies are arbitrary, and we have not used a genome-wide cutoff for our SNPs, but rather used cutoffs depending on the minor allele frequency. Therefore, we think adding a straight line to the graphs in Fig. 2A-B representing the overall cutoff, would be misleading. Please see below the text where we explain how the threshold was calculated for individual groups of SNPs with varying MAF:

“The GWAS associations were evaluated for minor allele count (MAC) and association strength above the Bonferroni threshold with -log10(p-value/#SNPs), calculated for each sub-population of SNPs above threshold MAC (Table S3, Bonf.threshold.MAC.specific)”

Line 490-492 - Presents the results of the gene tree to support a model in which SR3G diverged from AT3G50150 prior to the speciation events leading to Capsella and Arabidopsis. But this topology requires at least two independent losses of SR3G - can you rule out the hypothesis that the position of SR3G on the gene tree is a result of long branch attraction? Given the syntenic orientation of AT3G50150 and SR3G, and apparent directional selection experienced by the latter lineage, it seems more parsimonious that AT3G50150 and SR3G arose from a very recent duplication event.

We agree with the reviewer that it seemed most parsimonious for AT3G50160 (SR3G) to be a recent tandem duplication of AT3G50150 – and this was certainly our expectation given the other tandem duplications that have occurred in this genomic region. However, irrespective of the type of alignment from which we built the phylogeny (nucleotide vs AA; sometimes nucleotide is noisier but provides more information) we were never able to recapitulate a tree where AT3G50160 was immediately sister to AT3G50150 – even with a long branch for AT3G50160 indicating a rapid pace of nucleotide/AA change relative to AT3G50150. In regards to long branch attraction, it is our interpretation that long branch attraction typically requires multiple long branches that get placed together at a poorly supported node where sampling is sparse (https://www.nature.com/articles/s41576-020-0233-0), whereas we have the single long branch for AT3G50160, and all other A/C clade (Arabidopsis/Camelina/Capsella) members forming a lineage with a much shorter branch. To test the possibility of long branch attraction we subtracted out individual members of the AT3G50150/160 clade to see if there was algorithmic uncertainty in the placement of AT3G50160. We did not observe this in any of the branch subtractions that we performed (see below). Thus, it appears that we must stick with our original interpretation. If the reviewer would like us to soften this interpretation, we would be more than happy to do so, as it does not impact the overall conclusions for AT3G50160 being a rapidly evolving member of this clade.

Author response image 1.

Line 494 (and throughout) - I expect that all of the genes being studied herein are "experiencing selection," even if it's boring-old purifying selection on functionally conserved proteins. I think you mean to say "directional selection."

We thank the reviewer for this comment and completely agree that we lacked precision on our statement. We have corrected this throughout the manuscript.

Line 497 - state the background and foreground values of omega, here.

We apologize for not including these values and have added them at this point in the manuscript (new Table S6).

Line 511 and Line 673 - Inspection of Figure S13B suggests that SR3G is not "predominantly" expressed nor does it have the "highest enrichment" in the root stele. Certainly, among root cell types, this is predominant. But it appears to be quite highly expressed in late-stage seeds and some floral organs, as well.

We appreciate the reviewer for recognizing that SR3G is not a highly expressed gene. In root cell types, its expression is enriched in the root stele. Overall, SR3G is expressed at both early and later developmental stages. Our investigation of later developmental stages related to seed production did not reveal any significant phenotypic differences in fertility.

Line 514 - "54-folds" should be "54-fold."

Thanks. We made corrections.

Figure 7 - For symmetry, I suggest adding the "Beginning of salt stress" arrow to the "Early Stress" panel as well (even if it's right at day 0).

Thanks. We added the arrow to Early Stress in both Panels A and B.

Figure S2 - both graphs should have the same scale on the y-axis

Thanks - we have now re-plotted the graph with the matching y-axis scales.

Line 531 - I feel that this is a significant overstatement. The strongest statement supported by the results presented here is that SR3G is the most prominent DUF247 studied herein in root development under salt stress.

Thanks for the comments. We rephrase the statement.

“These results suggest that SR3G is the most prominent DUF247 studied within our study to affect root development under salt stress.”

Lines 583-605 - These data seem to me to be tangential to the central aims of the study. I suggest removing them for clarity/brevity.

We greatly appreciate the reviewer's suggestion. Our study primarily focused on characterizing the main GWAS candidate, SR3G. Since SR3G is located within a cluster of other DUF247 genes on chromosome 3, we believe that screening the neighboring DUF247 genes could provide further insights into SR3G’s role in root development. Additionally, we believe that the generated data and lines will serve as a valuable resource for other researchers interested in studying these genes. For these reasons, we have decided to retain these datasets in the manuscript.

Lines 650-652 - these sections 1-3 differences in suberization between SR3G and Col-0 under control conditions are not significant. At best, this may be described as a "trend" and not "higher levels." In section 4, it is VERY marginally significant (and probably not at all after the large number of tests performed, here.)

We appreciate the reviewer's feedback and have revised the wording accordingly.

Line 660 - this statement is only true for Section 1. I suggest adding this caveat.

We appreciate the reviewer's comments on this matter. We quantified four suberin monomers in whole root seedlings rather than in individual root sections due to the technical challenges of separating the sections without microscopy and the limited availability of samples for GS-MS analysis.

eLife Assessment

Through cellular, developmental, and physiological analysis, this valuable study identifies a gene that regulates the relative growth of roots and shoots under salt stress. The holistic approach taken provides solid evidence that this member of a larger tandemly duplicated gene family together with an upstream regulator contributes to salt tolerance, although the statistical or biological support for some conclusions could be more robust. The manuscript will be of interest to plant biologists studying mechanisms of abiotic stress tolerance and gene family evolution.

Reviewer #1 (Public review):

Summary:

The authors aim to assess the effect of salt stress on root:shoot ratio, identify the underlying genetic mechanisms, and evaluate their contribution to salt tolerance. To this end, the authors systematically quantified natural variations in salt-induced changes in root:shoot ratio. This innovative approach considers the coordination of root and shoot growth rather than exploring biomass and the development of each organ separately. Using this approach, the authors identified a gene cluster encoding eight paralog genes with a domain-of-unknown-function 247 (DUF247), with the majority of SNPs clustering into SR3G (At3g50160). In the manuscript, the authors utilized an integrative approach that includes genomic, genetic, evolutionary, histological, and physiological assays to functionally assess the contribution of their genes of interest to salt tolerance and root development.

Comments on revisions:

As the authors correctly noted, variations across samples, genotypes, or experiments make achieving statistical significance challenging. Should the authors choose to emphasize trends across experiments to draw biological conclusions, careful revisions of the text, including titles and figure legends, will be necessary to address some of the inconsistencies between figures (see examples below). However, I would caution that this approach may dilute the overall impact of the work on SR3G function and regulation. Therefore, I strongly recommend pursuing additional experimental evidence wherever possible to strengthen the conclusions.

(1) Given the phenotypic differences shown in Figures S17A-B, 10A-C, and 6A, the statement that "SR3G does not play a role in plant development under non-stress conditions" (lines 680-681) requires revision to better reflect the observed data.<br /> (2) I agree with the authors that detecting expression differences in lowly expressed genes can be challenging. However, as demonstrated in the reference provided (Lu et al., 2023), a significant reduction in WRKY75 expression is observed in T-DNA insertion mutant alleles of WRKY75. In contrast, Fig. 9B in the current manuscript shows no reduction in WRKY75 expression in the two mutant alleles selected by the authors, which suggests that these alleles cannot be classified as loss-of-function mutants (line 745). Additionally, the authors note that the wrky75 mutant exhibits reduced main root length under salt stress, consistent with the phenotype reported by Lu et al. (2023). However, other phenotypic discrepancies exist between the two studies. For example, 1) Lu et al. (2023) report that w¬rky75 root length is comparable to WT under control conditions, whereas the current manuscript shows that wrky75 root growth is significantly lower than WT; 2) under salt stress, Lu et al. (2023) show that wrky75 accumulates higher levels of Na+, whereas the current study finds Na+ levels in wrky75 indistinguishable from WT. To confirm the loss of WRKY75 function in these T-DNA insertion alleles the authors should provide additional evidence (e.g., Western blot analysis).

eLife Assessment

This important work advances our understanding of the impact of malnutrition on hematopoiesis and subsequently infection susceptibility. Support for the overall claims is convincing in some respects and incomplete in others as highlighted by reviewers. This work will be of general interest to those in the fields of hematopoiesis, malnutrition, and dietary influence on immunity.

Reviewer #1 (Public review):

Summary:

In this study, the authors used a chronic murine dietary restriction model to study the effects of chronic malnutrition on controls of bacterial infection and overall immunity, including cellularity and functions of different immune cell types. They further attempted to determine whether refeeding can revert the infection susceptibility and immunodeficiency. Although refeeding here improves anthropometric deficits, the authors of this study show that this is insufficient to recover the impairments across the immune cell compartments.

Strengths:

The manuscript is well-written and conceived around a valid scientific question. The data supports the idea that malnutrition contributes to infection susceptibility and causes some immunological changes. The malnourished mouse model also displayed growth and development delays. The work's significance is well justified. Immunological studies in the malnourished cohort (human and mice) are scarce, so this could add valuable information.

Weaknesses:

The assays on myeloid cells are limited, and the study is descriptive and overstated. The authors claim that "this work identifies a novel cellular link between prior nutritional state and immunocompetency, highlighting dysregulated myelopoiesis as a major." However, after reviewing the entire manuscript, I found no cellular mechanism defining the link between nutritional state and immunocompetency.

Reviewer #2 (Public review):

Summary:

Sukhina et al. use a chronic murine dietary restriction model to investigate the cellular mechanisms underlying nutritionally acquired immunodeficiency as well as the consequences of a refeeding intervention. The authors report a substantial impact of undernutrition on the myeloid compartment, which is not rescued by refeeding despite rescue of other phenotypes including lymphocyte levels, and which is associated with maintained partial susceptibility to bacterial infection.

Strengths:

Overall, this is a nicely executed study with appropriate numbers of mice, robust phenotypes, and interesting conclusions, and the text is very well-written. The authors' conclusions are generally well-supported by their data.

Weaknesses:

There is little evaluation of known critical drivers of myelopoiesis (e.g. PMID 20535209, 26072330, 29218601) over the course of the 40% diet, which would be of interest with regard to comparing this chronic model to other more short-term models of undernutrition.

Further, the microbiota, which is well-established to be regulated by undernutrition (e.g. PMID 22674549, 27339978, etc.), and also well-established to be a critical regulator of hematopoiesis/myelopoiesis (e.g. PMID 27879260, 27799160, etc.), is completely ignored here.

Reviewer #3 (Public review):

Summary:

Sukhina et al are trying to understand the impacts of malnutrition on immunity. They model malnutrition with a diet switch from ad libitum to 40% caloric restriction (CR) in post-weaned mice. They test impacts on immune function with listeriosis. They then test whether re-feeding corrects these defects and find aspects of emergency myelopoiesis that remain defective after a precedent period of 40% CR. Overall, this is a very interesting observational study on the impacts of sudden prolonged exposure to less caloric intake.

Strengths:

The study is rigorously done. The observation of lasting defects after a bout of 40% CR is quite interesting. Overall, I think the topic and findings are of interest.

Weaknesses:

While the observations are interesting, in this reviewer's opinion, there is both a lack of mechanistic understanding of the phenomena and also some lack of resolution/detail about the phenomena itself. Addressing the following major issues would be helpful towards aspects of both:

(1) Is it calories, per se, or macro/micronutrients that drive these phenotypes observed with 40% CR. At the least, I would want to see isocaloric diets (primarily protein, fat, or carbs) and then some of the same readouts after 40% CR. Ie does low energy with relatively more eg protein prevent immunosuppression (as is commonly suggested)? Micronutrients would be harder to test experimentally and may be out of the scope of this study. However, it is worth noting that many of the malnutrition-associated diseases are micronutrient deficiencies.

(2) Is immunosuppression a function of a certain weight loss threshold? Or something else? Some idea of either the tempo of immunosuppression (happens at 1, in which weight loss is detected; vs 2-3, when body length and condition appear to diverge; or 5 weeks), or grade of CR (40% vs 60% vs 80%) would be helpful since the mechanism of immunosuppression overall is unclear (but nailing it may be beyond the scope of this communication).

(3) Does an obese mouse that gets 40% CR also become immunodeficient? As it stands, this ad libitum --> 40% CR model perhaps best models problems in the industrial world (as opposed to always being 40% CR from weaning, as might be more common in the developing world), and so modeling an obese person losing a lot of weight from CR (like would be achieved with GLP-1 drugs now) would be valuable to understanding generalizability.

(4) Generalizing this phenomenon as "bacterial" with listeriosis, which is more like a virus in many ways (intracellular phase, requires type I IFN, etc.) and cannot be given by the natural route of infection in mice, may not be most accurate. I would want to see an experiment with E.Coli, or some other bacteria, to test the statement of generalizability (ie is it bacteria, or type I IFN-pathway dominant infections, like viruses). If this is unique listeriosis, it doesn't undermine the story as it is at all, but it would just require some word-smithing.

(5) Previous reports (which the authors cite) implicate Leptin, the levels of which scale with fat mass, as "permissive" of a larger immune compartment (immune compartment as "luxury function" idea). Is their phenotype also leptin-mediated (ie leptin AAV)?

(6) The inability of re-feeding to "rescue" the myeloid compartment is really interesting. Can the authors do a bone marrow transplantation (CR-->ad libitum) to test if this effect is intrinsic to the CR-experienced bone marrow?

(7) Is the defect in emergency myelopoiesis a defect in G-CSF? Ie if the authors injected G-CSF in CR animals, do they equivalently mobilize neutrophils? Does G-CSF supplementation (as one does in humans) rescue host defense against Listeria in the CR or re-feeding paradigms?

eLife Assessment

This study provides a valuable new resource to investigate the molecular basis of the particular features characterizing the pipefish embryo. The authors found both unique and shared gene expression patterns in pipefish organs compared with other teleost fishes. The solid data collected in this unconventional model organism will give new insights into understanding the extraordinary adaptations of the Syngnathidae family and will be of interest in the domain of evolution of fish development.

Reviewer #1 (Public review):

Syngnathid fishes (seahorses, pipefishes, and seadragons) present very particular and elaborated features among teleosts and a major challenge is to understand the cellular and molecular mechanisms that permitted such innovations and adaptations. The study provides a valuable new resource to investigate the morphogenetic basis of four main traits characterizing syngnathids, including the elongated snout, toothlessness, dermal armor and male pregnancy. More particularly, the authors have focused on a late stage of pipefish organogenesis to perform single-cell RNA-sequencing (scRNA-seq) completed by in situ hybridization analyses to identify molecular pathways implicated in the formation of the different specific traits.<br /> The first set of data explores the scRNA-seq atlas composed of 35,785 cells from two samples of gulf pipefish embryos that authors have been able to classify into major cell types characterizing vertebrate organogenesis, including epithelial, connective, neural and muscle progenitors. To affirm identities and discover potential properties of clusters, authors primarily use KEGG analysis that reveals enriched genetic pathways in each cell types. After revisions, the authors have provided extended supplementary files to well interpret the dataset and some statements have been clarified. I thank the authors for the revisions/completions of ISH results compared to initial submission.

To conclude, the scRNA-seq dataset in this unconventional model organism will be useful for the community and will provide clues for future research to understand the extraordinary evolution of the Syngnathidae family.

Reviewer #2 (Public review):

Summary:

The authors present the first single-cell atlas for syngathid fishes, providing a resource for future evolution & development studies in this group.

Strengths:

The concept here is simple and I find the manuscript to be well written. I like the in situ hybridization of marker genes >> this is really nice. I also appreciate the gene co-expression analysis to identify modules of expression. There are no explicit hypotheses tested in the manuscript, but the discovery of these cell types should have value in this organism and in the determination of morphological novelties in seahorses and their relatives.

Weaknesses:

I think there are a few computational analyses that might improve the generality of the results.

(1) The cell types: The authors use marker gene analysis and KEGG pathways to identify cell types. I'd suggest a tool like SAMap (https://elifesciences.org/articles/66747) which compares single cell data sets from distinct organisms to identify 'homologous' cell types -- I imagine the zebrafish developmental atlases could serve as a reasonable comparative reference.

(2) Trajectory analyses: Authors suggest that their analyses might identify progenitor cell states and perhaps related differentiated states. They might explore cytoTRACE and/or pseudotime-based trajectory analyses to more fully delineate these ideas.

(3) Cell-cell communication: I think it's very difficult to identify 'tooth primordium' cell types, because cell types won't be defined by organ in this way. for instance dental glia will cluster with other glia, dental mesenchyme will likely cluster with other mesenchymal cell types. so the histology and ISH in most convincing in this regard. having said this, given the known signaling interactions in the developing tooth (and in development generally) the authors might explore cell-cell communication analysis (e.g., CellChat) to identify cell types that may be interacting.

Comments on revisions:

I feel essentially the same about this manuscript. it's a useful resource for future experimental forays into this unique system. The team made improvements to deal with comments from other reviewers related to quality of confirmatory in situ hybridization. This is good.

Regarding their response that one can't use CellChat if you're not working in mice or human, this is inaccurate. the assumption one makes is that ligand-receptor pairs and signaling pathways have conserved functions across animals (vertebrates). It's the same assumption the authors make when using the KEGG pathway to score enrichment of pathways in clusters. CellChat used in fishes in Johnson et al 2023 Nature Communications | ( 2023) 14:4891.

Reviewer #3 (Public review):

Summary:

This study established a single-cell RNA sequencing atlas of pipefish embryos. The results obtained identified unique gene expression patterns for pipefish-specific characteristics, such as fgf22 in the tip of the palatoquadrate and Meckel's cartilage, broadly informing the genetic mechanisms underlying morphological novelty in teleost fishes. The data obtained are unique and novel, potentially important in understanding fish diversity. Thus, I would enthusiastically support this manuscript if the authors improve it to generate stronger and more convincing conclusions than the current forms.

Weakness:

Regarding the expression of sfrp1a and bmp4 dorsal to the elongating ethmoid plate and surrounding the ceratohyal: Are their expression patterns spatially extended or broader compared to the pipefish ancestor? Is there a much closer species available to compare gene expression patterns with pipefish? Did the authors consider using other species closely related to pipefish for ISH? Sfrp1a and bmp4 may be expressed in the same regions of much more closely related species without face elongation. I understand that embryos of such species are not always accessible, but it is also hard to argue responsible genes for a specific phenotype by only comparing gene expression patterns between distantly related species (e.g., pipefish vs. zebrafish). Due to the same reason, I would not directly compare/argue gene expression patterns between pipefish and mice, although I should admit that mice gene expression patterns are sometimes helpful to make a hypothesis of fish evolution. Alternatively, can the authors conduct ISH in other species of pipefish? If the expression patterns of sfrp1a and bmp4 are common among fishes with face elongation, the conclusion would become more solid. If these embryos are not available, is it possible to reduce the amount of Wnt and BMP signal using Crispr/Cas, MO, or chemical inhibitor? I do think that there are several ways to test the Wnt and/or BMP hypothesis in face elongation.

Author response:

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

Public Reviews: 

Reviewer #1 (Public Review): 

Syngnathid fishes (seahorses, pipefishes, and seadragons) present very particular and elaborated features among teleosts and a major challenge is to understand the cellular and molecular mechanisms that permitted such innovations and adaptations. The study provides a valuable new resource to investigate the morphogenetic basis of four main traits characterizing syngnathids, including the elongated snout, toothlessness, dermal armor, and male pregnancy. More particularly, the authors have focused on a late stage of pipefish organogenesis to perform single-cell RNA-sequencing (scRNA-seq) completed by in situ hybridization analyses to identify molecular pathways implicated in the formation of the different specific traits. 

The first set of data explores the scRNA-seq atlas composed of 35,785 cells from two samples of gulf pipefish embryos that authors have been able to classify into major cell types characterizing vertebrate organogenesis, including epithelial, connective, neural, and muscle progenitors. To affirm identities and discover potential properties of clusters, authors primarily use KEGG analysis that reveals enriched genetic pathways in each cell types. While the analysis is informative and could be useful for the community, some interpretations appear superficial and data must be completed to confirm identities and properties. Notably, supplementary information should be provided to show quality control data corresponding to the final cell atlas including the UMAP showing the sample source of the cells, violin plots of gene count, UMI count, and mitochondrial fraction for the overall

dataset and by cluster, and expression profiles on UMAP of selected markers characterizing cluster identities. 

We thank the reviewer for these suggestions, and have added several figures and supplemental files in response. We added a supplemental UMAP showing the sample that each cell originated (S1). We also added supplemental violin plots for each sample showing the gene count, unique molecular identifier (UMI) count, mitochondrial fraction, and the doublet scores (S2). We added feature plots of zebrafish marker genes for these major cell types and marker genes identified from our dataset to the supplement (S3:S57). We also provided two supplemental files with marker genes. These changes should clarify the work that went into labeling the clusters. Although some of the cluster labels are general, we decided it would be unwise to label clusters with speculated specific annotations. We only gave specific annotations to clusters with concrete markers and/or in situ hybridization (ISH) results that cemented an annotation.  As shown in the new supplemental figures and files, certain clusters had clear, specific markers while others did not. Therefore, we used caution when we annotated clusters without distinct markers. 

The second set of data aims to correlate the scRNA-seq analysis with in situ hybridizations (ISH) in two different pipefish (gulf and bay) species to identify and characterize markers spatially, and validate cell types and signaling pathways active in them. While the approach is rational, the authors must complete the data and optimize labeling protocols to support their statements. One major concern is the quality of ISH stainings and images; embryos show a high degree of pigmentation that could hide part of the expression profile, and only subparts and hardly detectable tissues/stainings are presented. The authors should provide clear and good-quality images of ISH labeling on whole-mount specimens, highlighting the magnification regions and all other organs/structures (positive controls) expressing the marker of interest along the axis. Moreover, ISH probes have been designed and produced on gulf pipefish genome and cDNA respectively, while ISH labeling has been performed indifferently on bay or gulf pipefish embryos and larvae. The authors should specify stages and species on figure panels and should ensure sequence alignment of the probe-targeted sequences in the two species to validate ISH stainings in the bay pipefish. Moreover, spatiotemporal gene expression being a very dynamic process during embryogenesis, interpretations based on undefined embryonic and larval stages of pipefish development and compared to 3dpf zebrafish are insufficient to hypothesize on developmental specificities of pipefish features, such as on the absence of tooth primordia that could represent a very discrete and transient cell population. The ISH analyses would require a clean and precise spatiotemporal expression comparison of markers at the level of the entire pipefish and zebrafish specimens at well-defined stages, otherwise, the arguments proposed on teleost innovations and adaptations turn out to be very speculative. 

We are appreciative of the reviewer’s feedback. We primarily used the in situ hybridization (ISH) data as supplementary to the scRNAseq library and we are aware that further evidence is necessary to identify origins of syngnathid’s evolutionary novelties. Our goal was to provide clues for the developmental genetic basis of syngnathid derived features.  We hope that our study will inspire future investigations and are excited for the prospect that future research could include this reviewer’s ideas. 

All of the developmental stages and species information for the embryos used were in the figure captions as well as in supplemental file 6. Because we primarily used wild caught embryos, we did not have specific ages of most embryos. Syngnathid species are challenging to culture in the laboratory, and extracting embryos requires euthanizing the father which makes it difficult to obtain enough embryos for ISH. In addition, embryos do not survive long when removed from the brood pouch prematurely. We supplemented our ISH with bay pipefish caught off the Oregon coast because these fish have large broods. Wild caught pregnant male bay pipefish were immediately euthanized, and their broods were fixed. Because we did not have their age, we classified them based on developmental markers such as presence of somites and the extent of craniofacial elongation. Although these classification methods are not ideal, they are consistent with the syngnathid literature (Sommer et al. 2012). Since the embryos used for the ISH were primarily wild caught, we had a few different developmental stages represented in our ISH data. For our tooth primordia search, we used embryos from the same brood (therefore, same stage) for these experiments.

We understand the concern for the degree of pigmentation in the samples. We completed numerous bleach trials before embarking on the in situ hybridization experiments. After completing a bleach trial with a probe created from the gene tnmd for ISH_,_ we noticed that the bleached embryos were missing expression domains found in the unbleached embryos. We were, therefore, concerned that using bleached embryos for our experiments would result incorrect conclusions about the expression domains of these genes. We sparingly used bleaching at older stages, hatched larvae, where it was fundamentally necessary to see staining. As stated above, the primary goal of this manuscript was to generate and annotate the first scRNA-seq atlas in a syngnathid, and the ISHs were utilized to support inferred cluster annotations only through a positive identification of marker gene expression in expected tissues/cells. Therefore, the obscuring of gene expression by pigmentation would have resulted in the absence of evidence for a possible cluster annotation, not an incorrect annotation.

For the ease of viewing the ISHs, we improved annotations and clarity. We increased the brightness and contrast of images. In the original submission, we had to lower the image resolution to make the submission file smaller. We hope that these improvements plus the true image quality improves clarity of ISH results. We also included alignments in our supplementary files of bay pipefish sequences to the Gulf pipefish probes to showcase the high degree of sequence similarity. 

Sommer, S., Whittington, C. M., & Wilson, A. B. (2012). Standardised classification of pre-release development in male-brooding pipefish, seahorses, and seadragons (Family Syngnathidae). BMC Developmental Biology, 12, 12–15. 

To conclude, whereas the scRNA-seq dataset in this unconventional model organism will be useful for the community, the spatiotemporal and comparative expression analyses have to be thoroughly pushed forward to support the claims. Addressing these points is absolutely necessary to validate the data and to give new insights to understand the extraordinary evolution of the Syngnathidae family. 

We really appreciate the reviewer’s enthusiasm for syngnathid research, and hope that the additional files and explanation of the supporting role of the ISHs have adequately addressed their concerns. We share the reviewer’s enthusiasm and are excited for future work that can extend this study. 

Reviewer #2 (Public Review):

Summary: 

The authors present the first single-cell atlas for syngnathid fishes, providing a resource for future evolution & development studies in this group. 

Strengths: 

The concept here is simple and I find the manuscript to be well written. I like the in situ hybridization of marker genes - this is really nice. I also appreciate the gene co-expression analysis to identify modules of expression. There are no explicit hypotheses tested in the manuscript, but the discovery of these cell types should have value in this organism and in the determination of morphological novelties in seahorses and their relatives.  

We are grateful for this reviewer’s appreciation of the huge amount of work that went into this study, and we agree that the in situ hybridizations (ISHs) support the scRNAseq study as we intended. We appreciate that the reviewer thinks that this work will add value to the syngnathid field.

Weaknesses: 

I think there are a few computational analyses that might improve the generality of the results. 

(1) The cell types: The authors use marker gene analysis and KEGG pathways to identify cell types. I'd suggest a tool like SAMap (https://elifesciences.org/articles/66747) which compares single-cell data sets from distinct organisms to identify 'homologous' cell types - I imagine the zebrafish developmental atlases could serve as a reasonable comparative reference. 

We appreciate the reviewer’s request, and in fact we would have loved to integrate our dataset with zebrafish. However, syngnathid’s unique craniofacial development makes it challenging to determine the appropriate stage for comparison. While 3 days post fertilization (dpf) zebrafish data were appropriate for comparisons of certain cell types (e.g. epidermal cells), it would have been problematic for other cell types (e.g. osteoblasts) that are not easily detectable until older zebrafish stages. Therefore, determining equivalent stages between these species is difficult and contains potential for error. Future research should focus on trying to better match stages across syngnathids and zebrafish (and other fish species such as stickleback). Studies of this nature promise to uncover the role of heterochrony in the evo-devo of syngnathid’s unique snouts.

(2) Trajectory analyses: The authors suggest that their analyses might identify progenitor cell states and perhaps related differentiated states. They might explore cytoTRACE and/or pseudotime-based trajectory analyses to more fully delineate these ideas.

We thank the reviewer for this suggestion! We added a trajectory analysis using cytoTRACE to the manuscript. It complemented our KEGG analysis well (L172-175; S73) and has improved the manuscript.

(3) Cell-cell communication: I think it's very difficult to identify 'tooth primordium' cell types, because cell types won't be defined by an organ in this way. For instance, dental glia will cluster with other glia, and dental mesenchyme will likely cluster with other mesenchymal cell types. So the histology and ISH is most convincing in this regard. Having said this, given the known signaling interactions in the developing tooth (and in development generally) the authors might explore cell-cell communication analysis (e.g., CellChat) to identify cell types that may be interacting. 

We agree! It would have been a wonderful addition to the paper to include a cell-cell communication analysis. One limitation of CellChat is that it only includes mouse and human orthologs. Given concerns of reviewer #3 for mouse-syngnathid comparisons, we decided to not pursue CellChat for this study. We are looking forward to future cell communication resources that include teleost fishes.

Reviewer #3 (Public Review): 

Summary: 

This study established a single-cell RNA sequencing atlas of pipefish embryos. The results obtained identified unique gene expression patterns for pipefish-specific characteristics, such as fgf22 in the tip of the palatoquadrate and Meckel's cartilage, broadly informing the genetic mechanisms underlying morphological novelty in teleost fishes. The data obtained are unique and novel, potentially important in understanding fish diversity. Thus, I would enthusiastically support this manuscript if the authors improve it to generate stronger and more convincing conclusions than the current forms. 

Thank you, we appreciate the reviewer’s enthusiasm!

Weaknesses: 

Regarding the expression of sfrp1a and bmp4 dorsal to the elongating ethmoid plate and surrounding the ceratohyal: are their expression patterns spatially extended or broader compared to the pipefish ancestor? Is there a much closer species available to compare gene expression patterns with pipefish? Did the authors consider using other species closely related to pipefish for ISH? Sfrp1a and bmp4 may be expressed in the same regions of much more closely related species without face elongation. I understand that embryos of such species are not always accessible, but it is also hard to argue responsible genes for a specific phenotype by only comparing gene expression patterns between distantly related species (e.g., pipefish vs. zebrafish). Due to the same reason, I would not directly compare/argue gene expression patterns between pipefish and mice, although I should admit that mice gene expression patterns are sometimes helpful to make a hypothesis of fish evolution. Alternatively, can the authors conduct ISH in other species of pipefish? If the expression patterns of sfrp1a and bmp4 are common among fishes with face elongation, the conclusion would become more solid. If these embryos are not available, is it possible to reduce the amount of Wnt and BMP signal using Crispr/Cas, MO, or chemical inhibitor? I do think that there are several ways to test the Wnt and/or BMP hypothesis in face elongation. 

We appreciate the reviewer’s suggestion, and their recognition for challenges within this system. In response to this comment, we completed further in situ hybridization experiments in threespine stickleback, a short snouted fish that is much more closely related to syngnathids than is zebrafish, to make comparisons with pipefish craniofacial expression patterns (S76-S79). We added ISH data for the signaling genes (fgf22, bmp4, and sfrp1a) as well as prdm16. Through adding this additional ISH results, we speculated that craniofacial expression of bmp4, sfrp1a, and prdm16 is conserved across species. However, compared to the specific ceratohyal/ethmoid staining seen in pipefish, stickleback had broad staining throughout the jaws and gills. These data suggest that pipefish have co-opted existing developmental gene networks in the development of their derived snouts. We added this interpretation to the results and discussion of the manuscript (L244-L248; L262-277; L444-470).

Recommendations for the authors:  

Reviewing Editor (Recommendations for the Authors): 

We hope that the eLife assessment, as well as the revisions specified here, prove helpful to you for further revisions of your manuscript. 

Revisions considered essential: 

(1) Marker genes and single-cell dataset analyses. While these analyses have been performed to a good standard in broad terms, there is a majority view here that cell type annotations and trajectory analyses can be improved. In particular, there is question about the choice of marker genes for the current annotation. For one it can depend on the use of single marker genes (see tnnti1 example for clusters 17 and 31). Here, we recommend incorporating results from SAMap and trajectory analysis (e.g., cytoTRACE or standard pseudotime).

Because of the reviewer comments, we became aware that we insufficiently communicated how cell clusters were annotated. We did mention in the manuscript that we did not use single marker genes to annotate clusters, but instead we used multiple marker genes for each cluster for the annotation process. We used both marker genes derived from our dataset and marker genes identified from zebrafish resources for cluster annotation. We chose single marker genes for each cluster for visualization purposes and for in situ hybridizations. However, it is clear from the reviewers’ comments that we needed to make more clear how the annotations were performed. To make this effort more clear in our revision, we included two new supplementary files – one with Seurat derived marker genes and one with marker genes derived from our DotPlot method. We also included extensive supplementary figures highlighting different markers. Using Daniocell, we identified 6 zebrafish markers per major cell type and showed their expression patterns in our atlas with FeaturePlots. We also included feature plots of the top 6 marker genes for each cluster. We hope that the addition of these 40+ plots (S3:S57) to the supplement fully addresses these concerns. 

We appreciated the suggestion of cytotrace from reviewer #2! We ran cytotrace on three major cell lineages (neural, muscle, and connective; S73) which complemented our KEGG analysis in suggesting an undifferentiated fate for clusters 8, 10, and 16. We chose to not run SAMap because it is a scRNA-seq library integration tool. Although we compared our lectin epidermal findings to 3 dpf zebrafish scRNA-seq data, we did not integrate the datasets out of concern that we could draw erroneous conclusions for other cell types.  Future work that explores this technical challenge may uncover the role of heterochrony in syngnathid craniofacial development. We detail these changes more fully in our responses to reviewers.

(2) The claims regarding evolutionary novelty and/or the genes involved are considered speculative. In part, this comes from relying too heavily on comparisons against zebrafish, as opposed to more closely related species. For example, the discussion regarding C-type lectin expression in the epidermis and KEGG enrichment (lines 358 - 364) seems confusing. Another good example here is the discussion on sfrp1a (lines 258 - 261). Here, the text seems to suggest craniofacial sfrp1a expression (or specifically ethmoid expression?) is connected to the development of the elongated snout in pipefish. However, craniofacial expression of sfrp1a is also reported in the arctic charr, which the authors grouped into fishes with derived craniofacial structures. Separately, sfrp2 expression was also reported in stickleback fish, for example. Do these different discussions truly support the notion that sfrp1a expression is all that unique in pipefish, rather than that pipefish and zebrafish are only distantly related and that sfrp1a was a marker gene first, and co-opted gene second? The authors should respond to the comments in the public review related to this aspect, and include more informative comparison and discussion. 

A much more nuanced discussion with appropriate comparisons and caveats would be strongly recommended here.  

We appreciate this insight and used it as a motivator to complete and add select comparative ISH data to this manuscript. We added in situ hybridization experiments from stickleback fish for craniofacial development genes (sfrp_1a, prdm16, bmp4_, and fgf22; S76-S79).  After adding stickleback ISH to the manuscript, we were able to make comparisons between pipefish and stickleback patterns and draw more informed conclusions (L244-L248; L262-277; L444-470). We added additional nuance to the discussion of the head, tooth (L485-489), and male pregnancy (L358-L391) sections to address concerns of study limitations. We describe in more detail these additional data in response to reviewers.

(3) In situ hybridization results: as already included above, there is generally weak labeling of species, developmental stages, and other markings that can provide context. The collective feeling here is that as it is currently presented, the ISH results do not go too far beyond simply illustrative purposes. To take these results further, more detailed comparison may be needed. At a minimum, far better labeling can help avoid making the wrong impression. 

Based on the reviewers’ comments, we made changes to improve ISH clarity and add select comparative ISH findings. ISH was used to further interpretation of the scRNAseq atlas. All the developmental stages and species information for the embryos used were in the figure captions as well as in supplemental file 4. Since we primarily used wild caught embryos, we did not have specific ages of most embryos. The technical challenges of acquiring and staging Syngnathus embryos are detailed above. Because we did not have their age, we classified them based on developmental markers (such as presence of somites and the extent of craniofacial elongation). Although these classification methods are not ideal, they are consistent with the syngnathid literature (Sommer et al. 2012).  

We followed reviewer #1’s recommendations by adding an annotated graphic of a pipefish head, aligning bay and Gulf pipefish sequences for the probe regions, expanding out our supplemental figures for ISH into a figure for each probe, and improving labeling. These changes improved the description of the ISH experiments and have increased the quality of the manuscript.

We would have loved to complete detailed comparative studies as suggested, but doing such a complete analysis was not feasible for this study. Therefore, we completed an additional focused analysis. We followed reviewer #3’s idea and added ISHs from threespine stickleback, a short snouted fish, for 4 genes (sfrp1a, prdm16, fgf22, and bmp4). While more extensive ISHs tracking all marker genes through a variety of developmental stages in pipefish and stickleback would have provided crucial insights, we feel that it is beyond the scope of this study and would require a significant amount of additional work. We, thus, primarily interpreted the ISH results as illustrative data points in our discussion. As we state in the response to reviewer 1, the generation and annotation of the first scRNA-seq atlas in a syngnathid is the primary goal of this manuscript.  The ISHs were utilized primarily to support inferred cluster annotations if a positive identification of marker gene expression in expected tissues/cells occurred. 

Reviewer #1 (Recommendations For The Authors): 

While the scRNA-seq dataset offers a valuable resource for evo-devo analyses in fish and the hypotheses are of interest, critical aspects should be strengthened to support the claims of the study. 

Concerning the scRNA-seq dataset, the major points to be addressed are listed below: 

- Supplementary file 3 reports the single markers used to validate cluster annotations. To confirm cluster identities, more markers specific to each cluster should be highlighted and presented on the UMAP. 

We recognize the reviewer’s concern and had in reality used numerous markers to annotate the clusters. Based upon the reviewer’s comment we decided to make this clear by creating feature plots for every cluster with the top 6 marker genes. These plots showcase gene specificity in UMAP space. We also added feature plots for zebrafish marker genes for key cell types. Through these changes and the addition of 54 supplementary figures (S3:S57), we hope that it is clear that numerous markers validated cluster identity.

For example, as clusters 17 and 37 share the same tnnti1 marker, which other markers permit to differentiate their respective identity. 

This is a fair point. Cluster 17 and 37 both are marked by a tnni1 ortholog.

Different paralogous co-orthologs mark each cluster (cluster 17: LOC125989146; cluster 37: LOC125970863). In our revision to the above comment, additional (6) markers per cluster were highlighted which should remedy this concern. 

- L146: the low number of identified cartilaginous cells (only 2% of total connective tissue cells) appears aberrant compared to bone cell number, while Figure 1 presents a welldeveloped cartilaginous skeleton with poor or no signs of ossification. Please discuss this point. 

We also found this to be interesting and added a brief discussion on this subject to the results section (L147-L149). Single cell dissociations can have variable success for certain cell types. It is possible that the cartilaginous cells were more difficult to dissociate than the osteoblast cells.

- L162: pax3a/b are not specific to muscle progenitors as the genes are also expressed in the neural tube and neural crest derivatives during organogenesis. Please confirm cluster 10 identity.  

Thank you for the reminder, we added numerous feature plots that explored zebrafish (from Daniocell) and pipefish markers (identified in our dataset). Examining zebrafish satellite muscle markers (myog, pabpc4, and jam2a) shows a strong correspondence with cluster #10.

- L198: please specify in the text the pigment cell cluster number. 

We completed this change.

- L199: it is not clear why considering module 38 correlated to cluster 20 while modules 2/24 appear more correlated according to the p-value color code. 

We thank the reviewer for pointing this confusing element out! Although the t-statistic value for module 38 (3.75) is lower than the t-statistics for modules 2 and 24 (5.6 and 5.2, respectively), we chose to highlight module 38 for its ‘connectivity dependence’ score. In our connectivity test, we examined whether removing cells from a specific cell cluster reduced the connectivity of a gene network. We found that removing cluster 20 led to a decrease in module 38’s connectivity (-.13, p=0) while it led to an increase in modules 2 and 24’s connectivity (.145, p=1; .145, p=9.14; our original supplemental files 9-10). Therefore, the connectivity analysis showed that module 38’s structure was more dependent on cluster 20 than in comparison with modules 2 and 24. Although you highlighted an interesting quandary, we decided that this is tangential to the paper and did not add this discussion to the manuscript. 

- Please describe in the text Figure 4A. 

Completed, we thank the reviewer for catching this! 

Concerning embryo stainings, the major points to be addressed are listed below: 

- Figure 1: please enhance the light/contrast of figures to highlight or show the absence of alcian/alizarin staining. Mineralized structures are hardly detectable in the head and slight differences can be seen between the two samples. The developmental stage should be added. Please homogenize the scale bar format (remove the unit on panels E and, G as the information is already in the text legend). It would be useful to illustrate the data with a schematic view of the structures presented in panels B, and E, and please annotate structures in the other panels.  

We thank the reviewer for these suggestions to improve our figure. We increased the brightness and contrast for all our images. We also added an illustration of the head with labels of elements. As discussed, we used wild caught pregnant males and, therefore, do not know the exact age of the specimens. However, we described the developmental stage based on morphological observations. Slight differences in morphology between samples is expected. We and others have noticed that

developmental rate varies, even within the same brood pouch, for syngnathid embryos. We observed several mineralization zones including in the embryos including the upper and lower jaws, the mes(ethmoid), and the pectoral fin. We recognize the cartilage staining is more apparent than the bone staining, though increasing image brightness and contrast did improve the visibility of the mineralization front.

- All ISH stainings and images presented in Figures 4-6/ Figures S2-3 should be revised according to comments provided in the public review. 

We thank the reviewer for providing thorough comments, we provided an in-depth response to the public review. We made several improvements to the manuscript to address their concerns. 

- Figure 4: Figure 4B should be described before 4C in the text or inverse panels / L222 the Meckel's cartilage is not shown on Figure 4C. The schematic views in H should be annotated and the color code described / the ISH data must be completed to correlate spatially clusters to head structures. 

We thank the reviewer for pointing this out, we fixed the issues with this figure and added annotations to the head schematics.

- Figure 5: typo on panels 'alician' = alcian. 

We completed this change. 

- Figures S2-3: data must be better presented, polished / typo in captions 'relavant'= relevant. 

Thank you for this critique, we created new supplementary figures to enhance interpretation of the data (S59-S71). In these new figures, we included a feature plot for each gene and respective ISHs.

- Figure S3: soat2 = no evidence of muscle marker neither by ISH presented nor in the literature. 

We realized this staining was not clear with the previous S2/S3 figures. Our new changes in these supplementary figures based on the reviewer’s ideas made these ISH results clearer. We observed soat2 staining in the sternohyoideus muscle (panel B in S71).

Other points: 

- The cartilage/bone developmental state (Alcian/alizarin staining) and/or ISH for classical markers of muscle development (such as pax3/myf5) could be used to clarify the This could permit the completion of a comparative analysis between the two species and the interpretation of novel and adaptative characters.  

We appreciate this idea! We thought deeply about a well characterized comparative analysis between pipefish and zebrafish for this study. We discussed our concerns in our public response to reviewer 2. We found that it was challenging to stage match all cell types, and were concerned that we could make erroneous conclusions. For example, our pipefish samples were still inside the male brood pouch and possessed yolk sacs. However, we found osteoblast cells in our scRNAseq atlas, and in alizarin staining. Although zebrafish literature notes that the first zebrafish bone appears at 3 dpf (Kimmel et al. 1995), osteoblasts were not recognized until 5 dpf in two scRNAseq datasets (Fabian et al. 2022; Lange et al. 2023). A 5dpf zebrafish is considered larval and has begun hunting. Therefore, we chose to not integrate our data out of concern that osteoblast development may occur at different timelines between the fishes. 

Fabian, P., Tseng, K.-C., Thiruppathy, M., Arata, C., Chen, H.-J., Smeeton, J., Nelson, N., & Crump, J. G. (2022). Lifelong single-cell profiling of cranial neural crest diversification in zebrafish. Nature Communications 2022 13:1, 13(1), 1–13. 

Lange, M., Granados, A., VijayKumar, S., Bragantini, J., Ancheta, S., Santhosh, S., Borja, M., Kobayashi, H., McGeever, E., Solak, A. C., Yang, B., Zhao, X., Liu, Y., Detweiler, A. M., Paul,

S., Mekonen, H., Lao, T., Banks, R., Kim, Y.-J., … Royer, L. A. (2023). Zebrahub – Multimodal Zebrafish Developmental Atlas Reveals the State-Transition Dynamics of Late-Vertebrate Pluripotent Axial Progenitors. BioRxiv, 2023.03.06.531398. 

Kimmel, C., Ballard, S., Kimmel, S., Ullmann, B., Schilling, T. (1995). Stages of Embryonic Development of the Zebrafish. Developmental Dynamics 203:253:-310.

'in situs' in the text should be replaced by 'in situ experiments'.  

We made this change (L395, L663, L666, L762).

- Lines 562-565: information on samples should be added at the start of the result section to better apprehend the following scRNA-seq data.

We thank the reviewer for pointing out this issue. Although we had a few sentences on the samples in the first paragraph of the result section, we understand that it was missing some critical pieces of information. Therefore, we added these additional details to the beginning of the results section (L126-L132). 

- Lines 629-665: PCR with primers designed on gulf pipefish genome could be performed in parallel on bay and gulf cDNA libraries, and amplification products could be sequenced to analyze alignment and validate the use of gulf pipefish ISH probes in bay pipefish embryos. Probe production could also be performed using gulf primers on bay pipefish cDNA pools. 

After the submission of this manuscript, a bay pipefish genome was prepared by our laboratory. We used this genome to align our probes, these alignments demonstrate strong sequence conservation between the species. We included these alignments in our supplemental files.

- L663: the bleaching step must be optimized on pipefish embryos. 

We understand this concern and had completed several bleach optimization experiments prior to publication. Although we found that bleaching improved visibility of staining, we noticed with the probe tnmd that bleached embryos did not have complete staining of tendons and ligaments. The unbleached embryos had more extensive staining than the bleached embryos. We were concerned that bleaching would lead to failures to detect expression domains (false negatives) important for our analysis. Therefore, we did not use bleaching with our in situs experiments (except with hatched fish with a high degree of pigmentation). 

- Indicate the number of specimens analyzed for each labeling condition.  

We thank the reviewer for noticing this issue. We added this information to the methods (L766-767).

- Describe the fixation and pre-treatment methods previous to ISH and skeleton stainings

We thank the reviewer for pointing out this issue, we added these descriptions (L765-766; L772-774). 

Reviewer #3 (Recommendations For The Authors): 

(1) If sfrp1a expression is observed also in other fish species with derived craniofacial structures, it's important to discuss this more in the Discussion. This could be a common mechanism to modify craniofacial structures, although functional tests are ultimately required (but not in this paper, for sure). Can lines 421-428 involve the statement "a prolonged period of chondrocyte differentiation" underlies craniofacial diversity?

This is a great idea, and we added a sentence that captures this ethos (L451-452).

(2) Lines 334-346 need to be rephrased. It's hard to understand which genes are expressed or not in pipefish and zebrafish. Did "23 endocytosis genes" show significant enrichment in zebrafish epidermis, or are they expressed in zebrafish epidermis? 

We thank the reviewer for this comment, we re-phrased this section for clarity (L365-368).

(3) Figure 4 is missing the "D" panel and two "E" panels. 

We thank the reviewer for noticing this, we fixed this figure.

(4) Line 302: "whole-mount" or "whole mount"

We thank the reviewer for the catch!

eLife Assessment

This important study investigates how working memory load influences the Stroop effect from a temporal dynamics perspective. Solid evidence is provided that the working memory load influences the Stroop effect in the late-stage stimulus-response mapping instead of the early sensory stage. This study will be of interest to both neuroscientists and psychologists who work on cognitive control.

Reviewer #1 (Public review):

Summary:

This study investigates an intriguing question in cognitive control from a temporal dynamics perspective: why does concurrent verbal working memory load eliminate the color-word Stroop effect? Through a series of thorough data analyses, the authors propose that verbal working memory load occupies the stimulus-response mapping resources represented by theta-band activity, thereby disrupting the mapping process for task-irrelevant distractors. This reduces the response tendency to the distractors, ultimately leading to the elimination of the Stroop effect.

Strengths:

The behavioral and neural evidence presented in the manuscript is solid, and the findings have valuable theoretical implications for research on Stroop conflict processing.

Weaknesses:

There are several areas where the manuscript could be improved.

Major Comments:

(1) In the Results section, the rationale behind selecting the beta band for the central (C3, CP3, Cz, CP4, C4) regions and the theta band for the fronto-central (Fz, FCz, Cz) regions is not clearly explained in the main text. This information is only mentioned in the figure captions. Additionally, why was the beta band chosen for the S-ROI fronto-central region and the theta band for the S-ROI central region? Was this choice influenced by the MVPA results?

(2) In the Data Analysis section, line 424 states: "Only trials that were correct in both the memory task and the Stroop task were included in all subsequent analyses. In addition, trials in which response times (RTs) deviated by more than three standard deviations from the condition mean were excluded from behavioral analyses." The percentage of excluded trials should be reported. Also, for the EEG-related analyses, were the same trials excluded, or were different criteria applied?

(3) In the Methods section, line 493 mentions: "A 400-200 ms pre-stimulus time window was selected as the baseline time window." What is the justification in the literature for choosing the 400-200 ms pre-stimulus window as the baseline? Why was the 200-0 ms pre-stimulus period not considered?

(4) Is the primary innovation of this study limited to the methodology, such as employing MVPA and RSA to establish the relationship between late theta activity and behavior?

(5) On page 14, lines 280-287, the authors discuss a specific pattern observed in the alpha band. However, the manuscript does not provide the corresponding results to substantiate this discussion. It is recommended to include these results as supplementary material.

(6) On page 16, lines 323-328, the authors provide a generalized explanation of the findings. According to load theory, stimuli compete for resources only when represented in the same form. Since the pre-memorized Chinese characters are represented semantically in working memory, this explanation lacks a critical premise: that semantic-response mapping is also represented semantically during processing.

(7) The classic Stroop task includes both a manual and a vocal version. Since stimulus-response mapping in the vocal version is more automatic than in the manual version, it is unclear whether the findings of this study would generalize to the impact of working memory load on the Stroop effect in the vocal version.

(8) While the discussion section provides a comprehensive analysis of the study's results, the authors could further elaborate on the theoretical and practical contributions of this work.

Reviewer #2 (Public review):

Summary:

Li et al. explored which stage of Stroop conflict processing was influenced by working memory loads. Participants completed a single task (Stroop task) and a dual task (the Sternberg working memory task combined with the Stroop task) while their EEG data was recorded. They adopted the event-related potential (ERP), and multivariate pattern analyses (MVPA) to investigate the interaction effect of task (single/dual) and congruency (congruent/incongruent). The results showed that the interaction effect was significant on the sustained potential (SP; 650-950 ms), the late theta (740-820 ms), and beta (920-1040 ms) power but not significant on the early P1 potential (110-150 ms). They used the representational similarity analyses (RSA) method to explore the correlation between behavioral and neural data, and the results revealed a significant contribution of late theta activity.

Strengths:

(1) The experiment is well-designed.

(2) The data were analyzed in depth from both time and frequency domain perspectives by combining several methods.

Weaknesses:

(1) As the researchers mentioned, a previous study reported a diminished Stroop effect with concurrent working memory tasks to memorize meaningless visual shapes rather than memorize Chinese characters as in the study. My main concern is that lower-level graphic processing when memorizing visual shapes also influences the Stroop effect. The stage of Stroop conflict processing affected by the working memory load may depend on the specific content of the concurrent working memory task. If that's the case, I sense that the generalization of this finding may be limited.

(2) The P1 and N450 components are sensitive to congruency in previous studies as mentioned by the researchers, but the results in the present study did not replicate them. This raised concerns about data quality and needs to be explained.

Author response:

Reviewer #1 (Public review):

Comment 1: In the Results section, the rationale behind selecting the beta band for the central (C3, CP3, Cz, CP4, C4) regions and the theta band for the fronto-central (Fz, FCz, Cz) regions is not clearly explained in the main text. This information is only mentioned in the figure captions. Additionally, why was the beta band chosen for the S-ROI central region and the theta band for the S-ROI fronto-central region? Was this choice influenced by the MVPA results?

We thank the reviewer for the question regarding the rationale for the S-ROI selection in our study. The beta band was chosen for the central region due to its established relevance in motor control (Engel & Fries, 2010), movement planning (Little et al., 2019) and motor inhibition (Duque et al., 2017). The fronto-central theta band (or frontal midline theta) was a widely recognized indicator in cognitive control research (Cavanagh & Frank, 2014), associated with conflict detection and resolution processes. Moreover, recent empirical evidence suggested that the fronto-central theta reflected the coordination and integration between stimuli and responses (Senoussi et al., 2022). Although we have described the cognitive processes linked to these different frequencies in the introduction and discussion sections, along with the potential patterns of results observed in Stroop-related studies, we did not specify the involved cortical areas. Therefore, we have specified these areas in the introduction to enhance the clarity of the revised version (in the fourth paragraph of the Introduction section).

Regarding whether the selection of S-ROIs was influenced by the MVPA results, we would like to clarify here that we selected the S-ROIs based on prior research and then conducted the decoding analysis. Specifically, we first extracted the data representing different frequency indicators (three F-ROIs and three S-ROIs) as features, followed by decoding to obtain the MVPA results. Subsequently, the time-frequency analysis, combined with the specific time windows during which each frequency was decoded, provided detailed interaction patterns among the variables for each indicator. The specifics of feature selection are described in the revised version (in the first paragraph of the Multivariate Pattern Analysis section).

Comment 2: In the Data Analysis section, line 424 states: “Only trials that were correct in both the memory task and the Stroop task were included in all subsequent analyses. In addition, trials in which response times (RTs) deviated by more than three standard deviations from the condition mean were excluded from behavioral analyses.” The percentage of excluded trials should be reported. Also, for the EEG-related analyses, were the same trials excluded, or were different criteria applied?

We thank the reviewer for this suggestion. Beyond the behavioral exclusion criteria, trials with EEG artifacts were also excluded from the data for the EEG-related analyses. We have now reported the percentage of excluded trials for both behavioral and EEG data analyses in the revised version (in the second paragraph of the EEG Recording and Preprocessing section and the first paragraph of the Behavioral Analysis section).

Comment 3: In the Methods section, line 493 mentions: “A 400-200 ms pre-stimulus time window was selected as the baseline time window.” What is the justification in the literature for choosing the 400-200 ms pre-stimulus window as the baseline? Why was the 200-0 ms pre-stimulus period not considered?

We thank the reviewer for this question and would like to provide the following justification. First, although a baseline ending at 0 ms is common in ERP analyses, it may not be suitable for time-frequency analysis. Due to the inherent temporal smoothing characteristic of wavelet convolution in time-frequency decomposition, task-related early activities can leak into the pre-stimulus period (before 0 ms) (Cohen, 2014). This means that extending the baseline to 0 ms will include some post-stimulus activity in the baseline window, thereby increasing baseline power and compromising the accuracy of the results. Second, an ideal baseline duration is recommended to be around 10-20% of the entire trial of interest (Morales & Bowers, 2022). In our study, the epoch duration was 2000 ms, making 200-400 ms an appropriate baseline length. Third, given that the minimum duration of the fixation point before the stimulus in our experiment was 400 ms, we chose the 400 ms before the stimulus as the baseline point to ensure its purity. In summary, considering edge effects, duration requirements, and the need to exclude other influences, we selected a baseline correction window of -400 to -200 ms. To enhance the clarity of the revised version, we have provided the rationale for the selected time windows along with relevant references (in the first paragraph of the Time-frequency analysis section).

Comment 4: Is the primary innovation of this study limited to the methodology, such as employing MVPA and RSA to establish the relationship between late theta activity and behavior?

We thank the reviewer for this insightful question and would like to clarify that our research extends beyond mere methodological innovation; rather, it utilized new methods to explore novel theoretical perspectives. Specifically, our research presents three levels of innovation: methodological, empirical, and theoretical. First, methodologically, MVPA overcame the drawbacks of traditional EEG analyses based on specific averaged voltage intensities, providing new perspectives on how the brain dynamically encoded particular neural representations over time. Furthermore, RSA aimed to identify which indicators among the decoded were directly related to behavioral representation patterns. Second, in terms of empirical results, using these two methods, we have identified for the first time three EEG markers that modulate the Stroop effect under verbal working memory load: SP, late theta, and beta, with late theta being directly linked to the elimination of the behavioral Stroop effect. Lastly, from a theoretical perspective, we proposed the novel idea that working memory played a crucial role in the late stages of conflict processing, specifically in the stimulus-response mapping stage (the specific theoretical contributions are detailed in the second-to-last paragraph of the Discussion section).

Comment 5: On page 14, lines 280-287, the authors discuss a specific pattern observed in the alpha band. However, the manuscript does not provide the corresponding results to substantiate this discussion. It is recommended to include these results as supplementary material.

We thank the reviewer for this suggestion. We added a new figure along with the corresponding statistical results that displayed the specific result patterns for the alpha band (Supplementary Figure 1).

Comment 6: On page 16, lines 323-328, the authors provide a generalized explanation of the findings. According to load theory, stimuli compete for resources only when represented in the same form. Since the pre-memorized Chinese characters are represented semantically in working memory, this explanation lacks a critical premise: that semantic-response mapping is also represented semantically during processing.

We thank the reviewer for this insightful suggestion. We fully agree with the reviewer’s perspective. As stated in our revised version, load theory suggests that cognitive resources are limited and dependent on a specific type (in the second paragraph of the Discussion section). The previously memorized Chinese characters are stored in working memory in the form of semantic representations; meanwhile the stimulus-response mapping should also be represented semantically, leading to resource occupancy. We have included this logical premise in the revised version (in the third-to-last paragraph of the Discussion section).

Comment 7: The classic Stroop task includes both a manual and a vocal version. Since stimulus-response mapping in the vocal version is more automatic than in the manual version, it is unclear whether the findings of this study would generalize to the impact of working memory load on the Stroop effect in the vocal version.

We fully agree with the reviewer’s point that the verbal version of the Stroop task differs from the manual version in terms of the degree of automation in the stimulus-response mapping. Specifically, the verbal version relies on mappings that are established through daily language use, while the manual version involves arbitrary mappings created in the laboratory. Therefore, the stimulus-response mapping in the verbal response version is more automated and less likely to be suppressed. However, our previous research indicated that the degree of automation in the stimulus-response mapping was influenced by practice (Chen et al., 2013). After approximately 128 practice trials, semantic conflict almost disappears, suggesting that the level of automation in stimulus-response mapping for the verbal Stroop task is comparable to that of the manual version (Chen et al., 2010). Given that participants in our study completed 144 practice trials (in the Procedure section), we believe these findings can be generalized to the verbal version.

Comment 8: While the discussion section provides a comprehensive analysis of the study’s results, the authors could further elaborate on the theoretical and practical contributions of this work.

We thank the reviewer for the constructive suggestions. We recognize that the theoretical and practical contributions of the study were not thoroughly elaborated in the original manuscript. Therefore, we have now provided a more detailed discussion. Specifically, the theoretical contributions focus on advancing load theory and highlighting the critical role of working memory in conflict processing. The practical contributions emphasize the application of load theory and the development of intervention strategies for enhancing inhibitory control. A more detailed discussion can be found in the revised version (in the second-to-last paragraph of the Discussion section).

Reviewer #2 (Public review):

Comment 1: As the researchers mentioned, a previous study reported a diminished Stroop effect with concurrent working memory tasks to memorize meaningless visual shapes rather than memorize Chinese characters as in the study. My main concern is that lower-level graphic processing when memorizing visual shapes also influences the Stroop effect. The stage of Stroop conflict processing affected by the working memory load may depend on the specific content of the concurrent working memory task. If that’s the case, I sense that the generalization of this finding may be limited.

We thank the reviewer for this insightful concern. As mentioned in the manuscript, this may be attributed to the inherent characteristics of Chinese characters. In contrast to English words, the processing of Chinese characters relies more on graphemic encoding and memory (Chen, 1993). Therefore, the processing of line patterns essentially occupies some of the resources needed for character processing, which aligns with our study’s hypothesis based on dimensional overlap. Additionally, regarding the results, even though the previous study presents lower-level line patterns, the results still showed that the working memory load modulated the later theta band. We hypothesize that, regardless of the specific content of the pre-presented working memory load, once the stimulus disappears from view, these loads are maintained as representations in the working memory platform. Therefore, they do not influence early perceptual processing, and resource competition only occurs once the distractors reach the working memory platform. Lastly, previous study has shown that spatial loads, which do not overlap with either the target or distractor dimensions, do not influence conflict effect (Zhao et al., 2010). Taken together, we believe that regardless of the specific content of the concurrent working memory tasks, as long as they occupy resources related to irrelevant stimulus dimensions, they can influence the late-stage processing of conflict effect. Perhaps our original manuscript did not convey this clearly, so we have rephrased it in a more straightforward manner (in the second paragraph of the Discussion section).

Comment 2: The P1 and N450 components are sensitive to congruency in previous studies as mentioned by the researchers, but the results in the present study did not replicate them. This raised concerns about data quality and needs to be explained.

We thank the reviewer for this insightful concern. For P1, we aimed to convey that the early perceptual processing represented by P1 is part of the conflict processing process. Therefore, we included it in our analysis. Additionally, as mentioned in the discussion, most studies find P1 to be insensitive to congruency. However, we inappropriately cited a study in the introduction that suggested P1 shows differences in congruency, which is among the few studies that hold this perspective. To prevent confusion for readers, we have removed this citation from the introduction.

As for N450, most studies have indeed found it to be influenced by congruency. In our manuscript, we did not observe a congruency effect at our chosen electrodes and time window. However, significant congruency effects were detected at other central-parietal electrodes (CP3, CP4, P5, P6) during the 350-500 ms interval. The interaction between task type and consistency remained non-significant, consistent with previous results. Furthermore, with respect to the location of the electrodes chosen, existing studies on N450 vary widely, including central-parietal electrodes and frontal-central electrodes (for a review, see Heidlmayr et al., 2020). We speculate that this phenomenon may be related to the extent of practice. With fewer total trials, the task may involve more stimulus conflicts, engaging more frontal brain areas. On the other hand, with more total trials, the task may involve more response conflicts, engaging more central-parietal brain areas (Chen et al., 2013; van Veen & Carter, 2005). Due to the extensive practice required in our study, we identified a congruency N450 effect in the central-parietal region. We apologize for not thoroughly exploring other potential electrodes in the previous manuscript, and we have revised the results and interpretations regarding N450 accordingly in the revised version (in the N450 section of the ERP results and the third paragraph of the Discussion section).

Reference

Cavanagh, J. F., & Frank, M. J. (2014). Frontal theta as a mechanism for cognitive control. Trends in Cognitive Sciences, 18(8), 414–421. https://doi.org/10.1016/j.tics.2014.04.012

Chen, M. J. (1993). A Comparison of Chinese and English Language Processing. In Advances in Psychology (Vol. 103, pp. 97–117). North-Holland. https://doi.org/10.1016/S0166-4115(08)61659-3

Chen, X. F., Jiang, J., Zhao, X., & Chen, A. (2010). Effects of practice on semantic conflict and response conflict in the Stroop task. Psychol. Sci., 33, 869–871.

Chen, Z., Lei, X., Ding, C., Li, H., & Chen, A. (2013). The neural mechanisms of semantic and response conflicts: An fMRI study of practice-related effects in the Stroop task. NeuroImage, 66, 577–584. https://doi.org/10.1016/j.neuroimage.2012.10.028

Cohen, M. X. (2014). Analyzing Neural Time Series Data: Theory and Practice. The MIT Press. https://doi.org/10.7551/mitpress/9609.001.0001

Duprez, J., Gulbinaite, R., & Cohen, M. X. (2020). Midfrontal theta phase coordinates behaviorally relevant brain computations during cognitive control. NeuroImage, 207, 116340. https://doi.org/10.1016/j.neuroimage.2019.116340

Duque, J., Greenhouse, I., Labruna, L., & Ivry, R. B. (2017). Physiological Markers of Motor Inhibition during Human Behavior. Trends in Neurosciences, 40(4), 219–236. https://doi.org/10.1016/j.tins.2017.02.006

Engel, A. K., & Fries, P. (2010). Beta-band oscillations—Signalling the status quo? Current Opinion in Neurobiology, 20(2), 156–165. https://doi.org/10.1016/j.conb.2010.02.015

Heidlmayr, K., Kihlstedt, M., & Isel, F. (2020). A review on the electroencephalography markers of Stroop executive control processes. Brain and Cognition, 146, 105637. https://doi.org/10.1016/j.bandc.2020.105637

Little, S., Bonaiuto, J., Barnes, G., & Bestmann, S. (2019). Human motor cortical beta bursts relate to movement planning and response errors. PLOS Biology, 17(10), e3000479. https://doi.org/10.1371/journal.pbio.3000479

Morales, S., & Bowers, M. E. (2022). Time-frequency analysis methods and their application in developmental EEG data. Developmental Cognitive Neuroscience, 54, 101067. https://doi.org/10.1016/j.dcn.2022.101067

Senoussi, M., Verbeke, P., Desender, K., De Loof, E., Talsma, D., & Verguts, T. (2022). Theta oscillations shift towards optimal frequency for cognitive control. Nature Human Behaviour, 6(7), Article 7. https://doi.org/10.1038/s41562-022-01335-5

van Veen, V., & Carter, C. S. (2005). Separating semantic conflict and response conflict in the Stroop task: A functional MRI study. NeuroImage, 27(3), 497–504. https://doi.org/10.1016/j.neuroimage.2005.04.042

Zhao, X., Chen, A., & West, R. (2010). The influence of working memory load on the Simon effect. Psychonomic Bulletin & Review, 17(5), 687–692. https://doi.org/10.3758/PBR.17.5.687

eLife Assessment

This study presents useful albeit preliminary findings on transcriptome changes in cardiac lymphatic cells after myocardial infarction in mice. Despite revision, the conclusions of the authors remain uncertain as sample sizes in general are very low, and even sometimes too low to allow for valid statistical comparisons. Accordingly, there are concerns regarding statistical robustness, raised by both the editors and the reviewers. While the single-cell transcriptomic data were analyzed using solid advanced methodology, too few cells were included in the scRNA-seq data set and the spatial transcriptomics analyses. Thus, this study rather represents more a collection of preliminary transcriptomic data than a full scientific report that would definitively advance the field.

Reviewer #1 (Public review):

Summary:

Assessment of cardiac LEC transcriptomes post-MI may yield new targets to improve lymphatic function. scRNAseq is a valid approach as cardiac LECs are rare compared to blood vessel endothelial cells.

Strengths:

Extensive bioinformatics approaches employed by the group

Weaknesses:

Too few cells included in scRNAseq data set and the spatial transcriptomics data that was exploited has little relevance, or rather specificity, for cardiac lymphatics. This study seems more a collection of preliminary transcriptomic data than a true scientific report to help advance the field.

Comments on revisions:

Thank you for the revision that helps clarify some outstanding questions.

(1) I still have questions relating to the relevance of the spatial maps generated and shown in fig 3C. They are supposedly generated using a 'molecular finger print' specific to each sub-cluster of LECs. However, given that at early stages postMI most populations are exceedingly rare in your analyses, could you please explain or comment on the relevance of the spatial maps?

(2) Fig 3 s1 would indicate that the population CaII is the majoritarian one in healthy hearts, while quantifications in Fig 3A show that rather the LEC Co subpopulation is majoritarian. Further, in mouse hearts histological analyses have demonstrated that cardiac lymphatics are restricted to the outer layers of the heart. This is not seen in your spatial maps. This seems to be the case only for the LEc Co population in healthy hearts, but not for other subpopulation signatures. Please explain.

(3) Further, the population of CaI, with 1 cell analysed in d3, but appears very prevalent in the spatial maps at d3. Please explain.

(4) In your list of 12 genes used as matrix anchors to identify LEC subpopulations in your screens, it is not apparent how LEC CaI, II and III differ so much as to allow selective detection of subpopulations. This similitude of profiles is supported by Fig 2F, and further explanations are needed to explain how the spatial maps of LEC ca subpopulations appear as distinct as shown in fig 3 S1 and Fig 3C.

Reviewer #2 (Public review):

Summary:

This study integrated single-cell sequencing and spatial transcriptome data from mouse heart tissue at different time points post-MI. They identified four transcriptionally distinct subtypes of lymphatic endothelial cells and localized them in space. They observed that LECs subgroups are localized in different zones of infarcted heart with functions. Specifically, they demonstrated that LEC ca III may be involved in directly regulating myocardial injuries in the infarcted zone concerning metabolic stress, while LEC ca II may be related to the rapid immune inflammatory responses of the border zone in the early stage of MI. LEC ca I and LEC collection mainly participate in regulating myocardial tissue edema resolution in the middle and late stages post-MI. Finally, cell trajectory and Cell-Chat analyses further identified that LECs may regulate myocardial edema through Aqp1, and likely affect macrophage infiltration through the galectin9-CD44 pathway. The authors concluded that their study revealed the dynamic transcriptional heterogeneity distribution of LECs in different regions of the infarcted heart and that LECs formed different functional subgroups that may exert different bioeffects in myocardial tissue post-MI.

Strengths:

The study addresses a significant clinical challenge, and the results are of great translational value. All experiments were carefully performed, and their data support the conclusion.

Editors' comments (Public review):

Weaknesses:

(1) Figure 7C, 7E, 7I, and "Figure7-figure supplement 1 ": All data in these data panels are based on only n=3, which is insufficient. Sample sizes of n=3 are too low to correctly assess normality of distribution and, as a consequence, do not allow to select the appropriate parametric/non-parametric tests. Accordingly, no statistical comparison can be performed and all p values and symbols currently indicating statistically significant differences between groups must be removed.

(2) Figure 3A, 3B, or 3C: No information about n numbers per group. Should n numbers per group be n=4 or less, no statistical comparison can be performed and all p values and symbols indicating statistically significant differences between groups must be removed.

(3) Figure 4 E and 4F: No information about n numbers per group. Should n numbers per group be n=4 or less, no statistical comparison can be performed and all p values and symbols indicating statistically significant differences between groups must be removed.

(4) Figure 5: No information about n numbers per group is provided. Should n numbers per group be n=4 or less, no statistical comparison can be performed and all p values and symbols indicating statistically significant differences between groups must be removed.

Author response:

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

Reviewer #1 (Recommendations for The Authors):

Q1: In response to reviewers you noted totally 292 sequenced LECs, however in reviewer figure 3 B the numbers seem to add up to 221. Please include mention of the total number of LEC sequences. Please mention line 119, page 4 the total number of explored LEC transcriptomes

Thank you for your carefully review. We have updated Fig 2A, 2C and 2E. It was 242 (not 292) LECs included in our initial analysis, which contains the sample of d5 post MI in raw data (E-MTAB-7895). We dropped d5 in our subsequent analysis because the change in d5 did not significant differ from d3. Therefore, we included 221 LECs in our final analysis as we updated in Fig2A, 2C and 2E.

Q2-1: Figure 3A supposedly shows % of LEC subpopulations relative to their numbers found in day 0 samples. However, there seem to be some errors, because for example the subpop LEC Cap I include 13 cells day 1 and 6 cells day 1, which corresponds to 46% of initial numbers. However, from your graph 3B the blue population seems to occupy 10%. Please revise or explain how these relative % were calculated.

Thank you for your question. In the Figure 3A, each column was calculated by dn/d0*100%, that is d0=57/57*100%=100%, and d1= 21/57*100%=36.84%, d3=9/57*100%=15.79%, d7, d14, d28...Therefor, Cap I in d0 (13 cells) is 13/57*100%=22.81%, and Cap I in d1(6 cells) is 6/57*100%= 10.53%.

Q2-2: Further, based on the relative % of LEC subpopulations, using the numbers mentioned in Fig 3B, it would appear that the relative frequency LEC cap II population is actually stable at around 20-30% of all LECs per time point throughout the study (except day 1 drop). This contrasts with line 136 p. 4 statement. I would also urge caution for interpreting too much into the variation of relative levels of LEC co, as these represent exceeding rare cells in your samples, and could reflect technical issues rather than true biological variation (total LEC co numbers analyzed ranging from 1-24 cells/ time point). The same could be said of LEC cap II and cap III.

We strongly agree with your comment on the proportion of LEC cell subtypes post MI. As you pointed out, we have revised the result description on Page 4, line 137-143 as followed.

“In the early stages of myocardial infarction (D1 and D3), the quantity of LECs decreased sharply. The number of LECs gradually increasing from day 7 and returning to normal levels by day 14 after MI. Moreover, from day 14 onwards, the number and proportion of Ca I type LECs significantly increased.”

Q3: Please list in supplement the gene features used to identify in spatial transcriptomics the different LEC subpopulations, as their profiles (notably for capillary LECs) don't appear to be very different based on data in Fig 2F.

We have supplied gene features in supplementary materials.

Q4: In section 2.7 you refer to Gal9 secretion. Please replace with expression as no measure of protein levels from LECs has been described in your study.

Thank you for your suggestion, we have replaced secretion with expression.

Q5: The updated method to exclude non-lymphatic cells from lymphatic vessel analyses by incorporating pdpn as an additional marker ('present costained areas wherever possible' line 350 p 10)

Thank you for your correction. We have updated the description as follows and lighted them in the manuscript: rabbit anti-Lyve1 (1:300, ab14917, Abcam, UK), [Syrian hamster anti-Podoplanin (1:100, 53-5381-82, Thermo, USA), rabbit anti-Prox1(1:300, ab199359, Abcam, UK), both anti-podoplain and anti-prox1 are additional markers co-stained with Lyve1 to exclude non-lymphatic cells from lymphatic vessel].

Q6: Fig 1B, it is highly surprising to see the lymphatic density in the BZ go from 25 um² at day 3 to more than 1000 um² only four days later (day 7). Is it possible that your day 3 measurements were in the infarct area, and not BZ area? The H&E image shown in Fig1a for d3 sample would seem to indicate the analysis was done in a dead area, rather than BZ. Please revise (perhaps select similar zone as shown for d1 in fig 6D, adjusted for subepicardial region and not mid-myocardial as seems to be the case currently), and also provide lymphatic area measures in healthy myocardium for day 0 samples. The unit used (um²) also would depend on the size of the area examined. Is this unit per image? If so please report total imaged area as a reference.

A6: Thank you for your reminding and advises. We have labeled each zone on H&E and IF images in Fig1-supplementary Fig2B, and updated a clearer histological photo taken at 3 days post MI in Fig1A. Furthermore, we recalculated the lymphatic vessel area ratio as you suggested by calculating the ratio of LEC co-stained area to total imaged area under 100-fold magnification.

Q7: The mention that CD68 antibody isn't compatible with lyve1 antibody could easily have been bridged by using other macrophage markers, such as F4/80, which is readily available and often used marker for macs in mice and comes notably as a rat anti-mouse F4-80. It would have added much more relevant information to exclude Lyve1-/F4/80+ cells as compared to the current analysis, which may indeed include in area measures Lyve1+ /Pdpn- single cells erroneously spotted as 'lymphatic vessels'

Thank you for your excellent suggestion. We co-stained the sample with F4/80 and LYVE1 and supplied in the Fig1-supplementary Figure 1E, as shown in Author response image 1.

Author response image 1.

Immunofluorescence (IF) co-staining of tissue section with F4/80 and LYVE1 in sham and MI mice model at d3, d7, d14, and d28 post-MI. LYVE1: lymphatic vessel endothelial hyaluronan receptor 1; DAPI: 4’6-diamidino-2-phenylindole; scale bar in 10×-100 μm, 40×-25μm.

Reviewer 2 (Recommendations for The Authors):

Q1: Language expression must be improved. Many incomplete sentences exist throughout the manuscript. A few examples: Line 70-71: In order to further elucidate the effects and regulatory mechanisms of the lymphatic vessels in the repair process of myocardial injury following MI. Line 71-73. This study, integrated single-cell sequencing and spatial transcriptome data from mouse heart tissue at different timepoints after MI from publicly available data (E-MTAB-7895, GSE214611) in the ArrayExpress and gene expression omnibus (GEO) databases. Line 88-89: Since the membrane protein LYVE1 can present lymphatic vessel morphology more clearly than PROX1.

Thank you for your correction. We have carefully inspected and corrected the whole manuscript.

Q2: The type of animal models (i.e., permanent MI or MI plus reperfusion) included in Array Express and gene expression omnibus (GEO) databases must be clearly defined as these two models may have completely different effects on lymphatic vessel development during post-MI remodeling.

Thank you for your excellent suggestion. The animal models used in both E-MTAB-7895 and GSE214611 are permanent MI. We have modified the model information in the methodology section (page 12, line 400-401).

Q3: Line 119-120: Caution must be taken regarding Cav1 as a lymphocyte marker because Cav1 is expressed in all endothelial cells, not limited to LEC.

Thanks for your reminding. Cav 1 used in our clustering is one of the marker gene for its different expression in sub-types of LECs, referred in article PMID: 31402260

Q4: Figure 1 legend needs to be improved. RZ, BZ, and IZ need to be labeled in all IF images. Day 0 images suggest that RZ is the tissue section from the right ventricle.

Thank you for your suggestion. We have labeled and updated the regions of RZ, BZ, and IZ in H&E and IF image in Figure1-Figure supplement 2B.

Q5: The discussion section needs to be improved and better focused on the findings from the current study.

Thank you for your good comment. Based on your suggestion, we have revised the first paragraph of the discussion from lines 250-256 (Page 7) as followed:

Cardiac lymphatics play an important role in myocardial edema and inflammation. This study, for the first time, integrated single-cell sequencing data and spatial transcriptome data from mouse heart tissue at different time points of post-MI, and identified four transcriptionally distinct subtypes of LECs and their dynamic transcriptional heterogeneity distribution in different regions of myocardial tissue post-MI. These subgroups of LECs were shown to form different function involved in the inflammation, apoptosis, ferroptosis, and water absorption related regulation of vasopressin during the process of myocardial repair after MI.

eLife Assessment

This important study presents a series of results aimed at uncovering the involvement of the endosomal sorting protein SNX4 in neurotransmitter release. While the evidence supporting the conclusions is solid, the molecular mechanisms remain unclear. This paper will be of interest to cell biologists and neurobiologists.

Reviewer #1 (Public review):

Summary:

In the work Josse Poppinga and collaborators addressed the synaptic function of Sortin-Nexin 4 (SNX4). Employing a newly-developed in vitro KO model, with live imaging experiments, electrophysiological recordings and ultrastructural analysis, the authors evaluate modifications in synaptic morphology and function upon loss of SNX4. The data demonstrate increased neurotransmitter release and alteration in synapse ultrastructure with higher number of docked vesicles and shorter AZ. The evaluation of presynaptic function of SNX4 is of relevance and tackles an open and yet unresolved question in the field of presynaptic physiology.

Strengths:

The sequential characterization of the cellular model is nicely conducted, and the different techniques employed are appropriate for the morpho-functional analysis of the synaptic phenotype and the derived conclusions on SNX4 function at presynaptic site. The authors succeeded in presenting a novel in vitro model that results in chronic deletion of SNX4 in neurons. A convincing sequence of experimental techniques are applied to the model to unravel the role of SNX4, whose functions in neuronal cells and at synapses are largely unknown. The understanding of the role of endosomal sorting at presynaptic site is relevant and of high interest in the field of synaptic physiology and on the pathophysiology of the many described synaptopathies that broadly result in loss of synaptic fidelity and quality control at release sites.

Weaknesses:

The flow of the data presentation is mostly descriptive with several consistent morphological and functional modifications upon SNX loss. The paper would benefit from a wider characterization that would allow to address the physiological roles of SNX4 at synaptic site and speculate on the underlying molecular mechanisms. The novel experiments on autophagy progression as well as spontaneous neurotransmission are well conducted, although do not assist for the explanation of the molecular mechanism underneath.

Comments on revisions:

Other implementations in the revised version are quite limited and would benefit from a more detailed presentation and description. i.e.: Sholl analysis in the new figure 1h, is presented with no definition of number of cells employed and standard deviations of the replication. The "simil" Sholl analysis performed on VAMP2 is still puzzling and some explanations on the reason for the constant value of VAMP2 fluorescent signal from less than 0 to 160 µm from the cell body is to be added. How is the increased number of active synapses explained? How is this related to shorter AZ and higher number of docked vesicles?

Reviewer #2 (Public review):

Summary:

SNX4 is thought to mediate recycling from endosomes back to the plasma membrane in cells. In this study, the authors demonstrate the increases in the amounts of transmitter release and the number of docked vesicles by combining genetics, electrophysiology and EM. They failed to find evidence for its role in synaptic vesicle cycling and endocytosis, which may be intuitively closer to the endosome function.

Strengths:

The electrophysiological data and EM data are in principle, convincing, though there are several issues in the study.

Weaknesses:

It is unclear why the increase in the amounts of transmitter release and docked vesicles happened in the SNX4 KO mice. In other words, it is unclear how the endosomal sorting proteins in the end regulate or are connected to presynaptic, particularly the active zone function.

Comments on revisions:

I am fine with revision in principle. the authors have addressed my concerns.

Reviewer #3 (Public review):

Summary:

The study aims to determine whether the endosomal protein SNX4 performs a role in neurotransmitter release and synaptic vesicle recycling. The authors exploited a newly generated conditional knockout mouse to allow them to interrogate SNX4 function. A series of basic parameters were assessed, with an observed impact on neurotransmitter release and active zone morphology. The work is interesting, however as things currently stand, the work is descriptive with little mechanistic insight. There are a number of places where some of the conclusions require further validation.

Strengths:

The strengths of the work are the state-of-the-art methods to monitor presynaptic function.

Weaknesses:

The weaknesses are the fact that the work is largely descriptive, with no mechanistic insight into the role of SNX4.

Comments on revisions:

The authors have addressed a couple of the more major concerns with the manuscript, however many of the original weaknesses remain. The primary weakness being the lack of mechanism. It is disappointing that real-time VAMP2 trafficking was not investigated, and the authors justification as to why the experiment was not performed was not convincing (especially since this is the approach that all other groups employ to examine SV cargo trafficking). In a number of instances "contractual constraints" are referred to as an explanation for not performing additional experiments. It was unclear whether this refers to licencing issues with the mouse line or the lack of personnel to perform the work. Regardless it still leaves this work as somewhat incomplete.

Author response:

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

Public Reviews:

Reviewer #1 (Public Review):

Summary:

In the work: "Endosomal sorting protein SNX4 limits synaptic vesicle docking and release" Josse Poppinga and collaborators addressed the synaptic function of Sortin-Nexin 4 (SNX4). Employing a newly developed in vitro KO model, with live imaging experiments, electrophysiological recordings, and ultrastructural analysis, the authors evaluate modifications in synaptic morphology and function upon loss of SNX4. The data demonstrate increased neurotransmitter release and alteration in synapse ultrastructure with a higher number of docked vesicles and shorter AZ. The evaluation of the presynaptic function of SNX4 is of relevance and tackles an open and yet unresolved question in the field of presynaptic physiology.

Strengths:

The sequential characterization of the cellular model is nicely conducted and the different techniques employed are appropriate for the morpho-functional analysis of the synaptic phenotype and the derived conclusions on SNX4 function at presynaptic site. The authors succeeded in presenting a novel in vitro model that resulted in chronical deletion of SNX4 in neurons. A convincing sequence of experimental techniques is applied to the model to unravel the role of SNX4, whose functions in neuronal cells and at synapses are largely unknown. The understanding of the role of endosomal sorting at the presynaptic site is relevant and of high interest in the field of synaptic physiology and in the pathophysiology of the many described synaptopathies that broadly result in loss of synaptic fidelity and quality control at release sites.

We thank the reviewer for their positive evaluation of our manuscript.

Weaknesses:

The flow of the data presentation is mostly descriptive with several consistent morphological and functional modifications upon SNX loss. The paper would benefit from a wider characterization that would allow us to address the physiological roles of SNX4 at the synaptic site and speculate on the underlying molecular mechanisms. In addition, due to the described role of SNX4 in autophagy and the high interest in the regulation of synaptic autophagy in the field of synaptic physiology, an initial evaluation of the autophagy phenotype in the neuronal SNX4KO model is important, and not to be only restricted to the discussion section.

We thank the reviewer for their suggestions and agree that broader characterization would help us speculate on the underlying mechanism. To address this, we have conducted additional independent experiments investigating the role of SNX4 in neuronal autophagy, as suggested by this reviewer. These experiments are now included in the main figures and are no longer limited to the discussion section. Please see the detailed responses to this reviewer's recommendations below.

Reviewer #2 (Public Review):

Summary:

SNX4 is thought to mediate recycling from endosomes back to the plasma membrane in cells. In this study, the authors demonstrate the increases in the amounts of transmitter release and the number of docked vesicles by combining genetics, electrophysiology, and EM. They failed to find evidence for its role in synaptic vesicle cycling and endocytosis, which may be intuitively closer to the endosome function.

Strengths:

The electrophysiological data and EM data are in principle, convincing, though there are several issues in the study.

We thank the reviewer for their positive evaluation of our manuscript.

Weaknesses:

It is unclear why the increase in the amounts of transmitter release and docked vesicles happened in the SNX4 KO mice. In other words, it is unclear how the endosomal sorting proteins in the end regulate or are connected to presynaptic, particularly the active zone function.

We thank the reviewer for their suggestions and agree that further characterization would help to understand how endosomal sorting proteins regulate presynaptic neurotransmission. We have now added extra data on electrophysiological recordings clarifying SNX4’s role in the synapse. Please see the detailed responses to this reviewer's recommendations below.

Reviewer #3 (Public Review):

Summary:

The study aims to determine whether the endosomal protein SNX4 performs a role in neurotransmitter release and synaptic vesicle recycling. The authors exploited a newly generated conditional knockout mouse to allow them to interrogate the SNX4 function. A series of basic parameters were assessed, with an observed impact on neurotransmitter release and active zone morphology. The work is interesting, however as things currently stand, the work is descriptive with little mechanistic insight. There are a number of places where the data appear to be a little preliminary, and some of the conclusions require further validation.

Strengths:

The strengths of the work are the state-of-the-art methods to monitor presynaptic function.

We thank the reviewers for their positive evaluation of our manuscript.

Weaknesses:

The weaknesses are the fact that the work is largely descriptive, with no mechanistic insight into the role of SNX4. Further weaknesses are the absence of controls in some experiments and the design of specific experiments.

We thank the reviewer for their suggestions and agree that addition of extra control groups and experiments would strengthen interpretation of the observed phenotype. To address this, we have now performed experiments to investigate the miniature excitatory postsynaptic currents and added extra control groups such as overexpression of SNX4 on control background. In addition, we assessed SNX4-mediated neuronal autophagy as a potential molecular mechanism by which SNX4 affects synaptic output. Please see the detailed responses to this reviewers’ recommendations below.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

(1) The characterization of the neurite outgrowth presented in Figure 1 is a necessary starting point for the characterization of the model and the interpretation of the following data. Being the analysis conducted at 21 DIV, a significant portion of the neurite tree is out of the analyzed field. Adding sholl analysis will better indicate the complexity of the that appears to be influenced by SNX4 loss in the representative images shown in Figure 1f.

We fully agree and have now performed a Sholl analysis of dendrite branches to investigate dendritic complexity. (Figure 1(i), page 2-3, line 86-88). SNX4 depletion does not affect dendrite length or dendrite branching.

(2) Analogously, the characterization of synapse number is of relevance for the interpretation of the data. For a better flow of the data, Figure 4 might be presented as Figure 2 (without the repetition of panel h in Figure 1). An explanation of how VAMP2 puncta are processed is necessary in the method section. A double labelling with a postsynaptic marker would allow trafficking organelles to be distinguished from mature synaptic contacts. Indeed, the analysis of VAMP2 intensity along neurite in mature 21DIV neurons should reveal peaks in the intensity profile that represent synaptic contacts. For unexplained reasons, the profile is rather flat in the two experimental groups. Focusing on axonal branches will surely result in a peaked profile for VAMP2 labelling.

We fully agree that the characterization of synapses is relevant for the interpretation of the data. We have now added a section in our Material and Methods how the VAMP2 puncta are processed (p14 line 517-520). Instead of labeling mature synapses using double labeling of VAMP2 and PSD95, we analyzed the number of active synapses in live neurons using SypHy (Fig. 3g). The reviewer is correct that the VAMP2 data presented in Fig 1I and Fig 4 is part of the same dataset and we have clarified this in the figure legend. In Fig 1I only the total number of VAMP2 puncta is plotted as a marker for synapse number, while in Fig 4 we assess VAMP2 as potential SNX4 sorting cargo (Ma et al., 2017). Because of these different aims, we prefer to keep the figures separate. The analysis of VAMP2 intensity along the distance of the soma is a Sholl analysis (Fig. 4d), represents the average VAMP2 intensity over distance from the soma of 35-41 neurons per group. In contrast to a line scan of a single neurite, this average profile lacks the peaks of individual synapses.

(3) Miniature excitatory postsynaptic currents recordings would strengthen the synaptic characterization and complement the electrophysiological recordings shown in Figure 2. Analyzing frequency and amplitude parameters would complement the data on the number of synaptic connections defined by the pre and postsynaptic colocalization puncta as suggested above and may support the data shown in Figure 3 g that suggests a decreased number of active synapses in SNX4-KO cells.

We fully agree that the characterization of miniature excitatory postsynaptic currents would strengthen the synaptic characterization and complement the other electrophysiological data. Therefore, we have now added additional experiments showing the mEPSCs (Fig. 2k-m, page 4) in SNX4 cKO neurons versus control. This data shows that the amplitude and frequency of spontaneous miniature EPSCs (mEPSCs) were not affected upon SNX4 depletion, consistent with a normal first evoked EPSC and RRP estimate. Furthermore, these data suggest that it is unlikely that the observed increase in neurotransmission is due to post-synaptic effects.

(4) Recordings on the first evoked response shown in Figure 2 b and quantified in Figures c and d suggest that SNX4 overexpression per se exerts some effect on the Amplitude and the Charge of the first evoked response. This is also evident in the supplementary Figure 2 with lower frequency trains. An additional experimental group, namely control+SNX4 is needed for the correct interpretation of the observed phenotype. The possibility that SNX4 per se exerts an effect on evoked transmission could be discussed in terms of putative mechanisms and interactions.

We thank the reviewer for their suggestion and agree that an additional experimental group (control + SNX4) would strengthen interpretation of the observed phenotype. We have now added a new experimental condition with overexpression of SNX4 on a control background (Supplementary Fig. 3, page 20). This data shows that the amplitude and charge of the first evoked response were not affected in control + SNX4 neurons compared to control, and no differences were detected in the response to the 40 Hz stimulation train (Supplementary Fig. 3a-e).  Together, these data suggest that SNX4 overexpression in itself does not affect the neurotransmission protocols studied in SNX4 cKO experiments.

(5) To correctly interpret the SyPhy experiments and exclude an effect of SNX silencing on SV recycling, it is suggested to repeat the experiments shown in Figure 3 in the absence and in the presence of bafilomycin. Indeed, the quantifications shown in Figure 3 d and f do not represent "release fraction" as stated (lines 139/140) but they rather refer to an average difference between release fraction and recovered fraction. With the use of bafilomycin, the comparison of the deltaFmax/deltaFNH4Cl with and without bafilomycin would enable the release fraction to be correctly evaluated and compared.

We appreciate the reviewer’s suggestion and agree on the importance of considering the impact of SV recycling when evaluating the released fraction. We agree that the presence of bafilomycin is critical to isolate the released component during stimulation. We have now rephrased this conclusion. To assess synaptic recycling in these assays, bafilomycin in not critically required and we show by multiple independent experiments, including SypHy and FM64 dye assays, that SV recycling is either not affected or the effect is too small to be detected by these methods.

(6) In the ultrastructural analysis, additional quantifications are needed to exclude the accumulation of endosome-like structures. It is not clear if, in the evaluation of total SV number (Figure 5e), the authors counted all vesicles or vesicles < 50nm. This has to be explained and additional quantification of # of SV < 50nm and # SV > 50nm is informative, taking into account the endosomal nature of SNX4. Indeed, although the average size of SV is not changed (fig. 5 d), the density of "bigger vesicle" may result from endosomal-like structure accumulation. An additional suggested quantification is on vesicle # SV > 80nm as previously reported in the cited references dealing with endosomal proteins and presynaptic morphology.

We fully agree that the characterization of vesicle size is important and that it was not clearly stated which vesicles were included in the total number of SV (Fig. 5e). We have now added this to the figure description. We have also added a histogram that contains the vesicle numbers of different bin sizes for SNX4 cKO synapses and control synapses (Supplementary Fig. 4, page 21) including # SVs > 80nm. (Whilst it seems that there are more “bigger” vesicles in the KO, further analysis revealed that this is mostly driven by one experiment and this effect is not consistent.)

(7) Due to the high scientific interest in presynaptic autophagy for SV recycling and degradation, and the paucity of experimental work assessing the proteins involved, an initial evaluation of the neuronal autophagy process (by western blot analysis and immunocytochemistry) for the characterization of the model will better support the paragraph in the discussion (lines 314-322) and contribute to future work in the field. Although very rare, autophagosomes quantification at presynaptic sites can also be performed from the already acquired images. A double membrane structure with the material inside is evident in the representative control image presented!

We appreciate the reviewer’s suggestion and agree that presynaptic autophagy is an interesting potential mechanism that would elaborate our current working model. To address the reviewers’ suggestion, we added multiple independent experiments to investigate basal autophagy markers such as ATG5 using western blot analysis, characterization of p62 levels using immunohistochemistry and performed additional morphometric analysis on the electron microscopy data (Supplementary Fig. 5). In SNX4 cKO neurons, there was no significant difference in P62 puncta numbers or P62 somatic intensity under basal conditions or after blocking autophagic P62 degradation by bafilomycin treatment, suggesting that autophagic flux remains normal. Also, no changes in total ATG5 protein levels were observed and ultrastructural analysis revealed no differences in the total number of autophagosomes. Collectively, these data indicate that SNX4 depletion does not impact the basal autophagic flux, ATG5 protein levels, or the number of autophagosomes.

Minor points:

(1) Dorrbaun et al. 2018 is missing from the reference list. In the legend to figure 1 there is an incorrect reference to Figure 6, rather than Figure 4.

We have now adjusted the figure legend and added the reference (page 16, line 604).

(2) Information on the construct employed for the rescue is missing. Is it a fluorescent tag construct? Representative images of the three autaptic neurons (control, KO, KO+SNX4) would nicely complement data presentation in Figure 2. 

We have now elaborated on this in material and methods section (p12, line 418-421). Unfortunately, we did not obtain pictures of autaptic neurons used for electrophysiology experiments.

Reviewer #2 (Recommendations For The Authors):

(1) Figure 2d and f are somewhat inconsistent. Total charges for the 1st EPSCs differ almost 2-fold in the same condition.

We appreciate the reviewer’s concern. The average EPSCs charge of the first evoked was 89, 122 and 57 pC for control, KO and rescued neurons respectfully. The average charge of the first pulse of 40Hz train was 41,58 and 32 pC for control, KO and rescued neurons respectfully, which is roughly 50% of the naïve response of the same cells. These trains were recorded after 2 or 3 other stimulation paradigms, which can have affected the total charge released in the 40Hz train. That said, the proportional difference between groups is high comparable, with a 37% increased average charge released in SNX4 cKO compared to control in the naïve response and 41% increased response in the first response of the 40 Hz train, and rescued cells show a 53% reduction in average released charge compared to control in the naïve response compared to a 44% reduction in the first response of the 40 Hz train. Although the absolute values differ between these readouts, we conclude that the biological comparison between groups is consistent.

(2) Figure 2h. This type of analysis has a drawback. See Neher (2015) for the problems associated with this analysis.

We fully agree with the reviewer’s comment. As noted in our discussion (page 9 line 285), while this analysis has its limitations, it can still provide an indication of the ready releasable pool.   

(3) The EPSC phenotype may be due to postsynaptic effects. This should be excluded by additional experiments (mEPSC analysis) or further clarification.

We fully agree that the characterization of miniature excitatory postsynaptic currents recording would strengthen the synaptic characterization and complement the electrophysiological recordings. Therefore, we have now added additional experiments showing the mEPSCs (Fig. 2k-m) in SNX4 cKO neurons versus control. This data shows that the amplitude and frequency of spontaneous miniature EPSCs (mEPSCs) were not affected upon SNX4 depletion, suggesting that it is unlikely that the observed increase in neurotransmission is due to post-synaptic effects.

(4) The increased number of docked vesicles observed in EM and the increased slope (vesicle recruitment, Figure 2h) are not consistent with each other. Maybe the definition of docked vesicles is unclear in this version of the manuscript.

As noted in our material & methods (page 15, line 547-548), SVs were defined as docked if there was no distance visible between the SV membrane and the active zone membrane. We have added the pixel size for clarification. Indeed, we do not observe an increase in release probability or first evoked response, which would correspond with an increased docked pool. However, we think that the increase in docked vesicles might contribute to an enhanced SV recruitment (see discussion).

(5) Figure 3: Vesicle cycling was monitored in only a limited condition. It is known that there are multiple pathways of vesicle cycling. Ideally, these pathways should be dissected. At least, the authors mention the possibility that they have missed some "positive" conditions.

We fully agree with the reviewer’s comment that vesicle recycling is complex with several parallel pathways involved. While we did not study individual endocytosis pathways, we used different assays covering various recycling pathways. The SypHy assay (Fig. 3c & f) combined with the 100 AP stimulation paradigm at room temperature predominantly addresses clathrin-mediated endocytosis. Additionally, the FM-64 dye assay at 37 degrees Celsius covers ultrafast endocytosis pathways as well as bulk endocytosis routes. Since neither assay showed major effects, we decided not to pursue further experiments focusing on different endocytosis pathways.

Reviewer #3 (Recommendations For The Authors):

Major points:

(1) Since all of the work here is culture-focussed, the in vivo phenotype is not as relevant, however the in vitro properties are. The incomplete Cre-dependent removal of SNX4 is concerning (especially axonal SNX4 levels identified via immunofluorescence), however, the main concern is that there was no profiling of the other molecular changes within these cultures. This is important, since there may be considerable alterations in the expression of a number of presynaptic proteins which may explain the observed phenotypes. Ideally, these cultures could have been profiled in an unbiased manner via mass spectrometry to identify potential changes in the presynaptic proteome, or at the very least the levels of key fusion molecules would have been assessed via Western blotting.

We thank the reviewer for their suggestion and agree that mass spectrometry would strengthen the interpretation of the observed phenotype. However, due to contractual constraints, we are unable to pursue a mass spectrometry follow-up experiment. We agree that characterizing key fusion molecules is of potential interest. Therefore, based on literature, we selected a likely candidate, VAMP2, which did not show any alterations in expression levels when knocking out SNX4. Given the previously described role of SNX4 in the degradation pathway, one would expect increased degradation of key fusion molecules if they are recycled by SNX4. Other literature indicates that reduced levels of key fusion molecules, such as synaptotagmin or SNAP-25 (Broadie et al., 1994; Washbourne et al., 2001) , do not mimic our phenotype.

(2) The experiments reported in Figure 2, in particular those in 2c and 2d, suggest that overexpression of SNX4 has a dominant-negative effect on neurotransmitter release. This is strongly supported by the supplementary data during a stimulus train (particularly the start point of the 5 Hz train in Supplementary Figure 2). Therefore, the perceived rescue of EPSC charge in Figure 2f, 2g may be a result of SNX4 inhibiting neurotransmitter release. A determination of the impact of SNX4 overexpression (and level of overexpression) in WT neurons is essential to show that this is a bonefide rescue, rather than a direct inhibition by SNX4 overexpression.

We thank the reviewer for their suggestion and agree that an additional experimental group (control + SNX4) would strengthen interpretation of the observed phenotype. We have now added a new experiment with an extra experimental condition with overexpression of SNX4 on a control background (Supplementary Fig. 3 page 21). This data shows that the amplitude and charge of the first evoked response were not affected in control + SNX4 neurons compared to control, and no differences were detected in the response to the 40 Hz stimulation train (Supplementary Fig. 3a-e).  Together, these data suggest that SNX4 overexpression in itself does not affect the neurotransmission protocols studied in SNX4 cKO experiments.

(3) The experiments in Figure 3 clearly reveal a lack of effect of SNX4 depletion on synaptic vesicle endocytosis. However, the assumption that synaptic vesicle recycling is unaffected is a little premature. The fact that the second evoked SypHy peak is significantly larger than the first (Figures 3c-e) suggests that more vesicles may be recycling in KO neurons. Furthermore, the FM dye experiments do not aid interpretation, since there may be insufficient time (10 min) for new vesicles to be generated from endosomal intermediates experiments. Therefore, to confirm an absence of effect on recycling, the authors could either 1) perform the same experiment as 3c, but with 4 stimulation trains (to drive the system harder to reveal any phenotype) or 2) repeat the FM dye experiment but increase the time between loading and unloading to 30 min.

We fully agree with the reviewers' comment that vesicle recycling is an important component to consider and is complex with several parallel pathways involved. We conducted multiple independent experiments covering the most significant recycling pathways. The SypHy assay (Fig. 3c & f) combined with the 100 AP stimulation paradigm at room temperature predominantly addresses clathrin-mediated endocytosis. Additionally, the FM-64 dye assay at 37 degrees Celsius covers ultrafast endocytosis pathways as well as bulk endocytosis routes. To further challenge the system and reveal recycling phenotypes, we included a second 100 AP stimulation in our SypHy assay. While only the increase of the second SypHy peak is significant, the absolute numbers do not differ much from the first peak (0,17 for control and 0,21 for KO second peak and 0,19 for control and 0,22 for KO first peak, Supplementary table1). We nevertheless do not see any effects on recycling after the second peak (mean decay time is 27 for control and 26 for KO Supplementary Table 1). A single 100 AP 40 Hz train depletes all the synchronous release (not shown) and most of the evoked charge (see Fig 2f), hence two of these trains with one minute recovery is already a very demanding protocol. Although increasing the time between loading and unloading to 30 minutes might uncover other recycling components, it has been shown that ultrafast endocytosis occurs within 30 seconds (Watanabe et al., 2013), suggesting that 10 minutes should provide enough time for synaptic vesicle recycling. This is also evident from the fact that we can significantly destain synapses loaded with FM dye by electrical stimulation (Fig 3j), indicating that synaptic vesicle recycling took place. Since neither assay showed major effects, we concluded that under these circumstances, synaptic recycling is not significantly affected. However, we cannot exclude the possibility that recycling deficits in SNX4 cKO neurons could be detected in other paradigms,

(4) There is no obvious effect on VAMP2 levels or location in SNX4 KO neurons (Figure 4). However, when one considers that SNX4 is proposed to have a role in VAMP2 trafficking, it is surprising that an experiment examining the live trafficking of VAMP2-SypHy was not performed. This would have revealed activity-dependent alterations that would have been missed by simply measuring VAMP2 expression and localization, and potentially provided a molecular explanation for the enhanced neurotransmitter release during a stimulus train.

We appreciate the reviewer’s suggestion and agree that it could be a valuable experiment However, overexpressing a VAMP2-pHluorin construct might obscure potential phenotypes related to VAMP2 trafficking. SNX4 is expected to be involved in VAMP2 recycling, even with activity-dependent changes. Mis-sorted VAMP2 would accumulate in acidic vesicles, which could be masked by the VAMP2-pHluorin construct. Similarly, mis-sorting of other SNX4 cargo, such as the transferrin receptor, has been identified through lysosomal degradation, as shown by Western blot analysis of expression levels of the endogenous protein. We did not detect any differences in endogenous levels of VAMP2 within 21 days of SNX4 deletion (Fig 4), indicating that SNX4-dependent endosome sorting is not essential for VAMP2 recycling.

(5) The morphological data in Figure 5 report a series of small changes in docked vesicles and active zone length. In many cases, significance is obtained due to synapses being used as the experimental n, and thus inflating the statistical power. When one considers that no significant effect was observed on evoked release (apart from during a stimulus train), it suggests that the number of docked vesicles does not alter release probability in this system (which the authors point out). Instead, they suggest that an increased supply of vesicles is responsible, via increased recruitment to RRP/releasable pool (but not via increased recycling). If this is the case, it should have been reflected as an increase in the evoked SypHy response in Fig 2c,d (which is borderline significant). What may help is to determine the morphological landscape immediately after a stimulus strain, since this is the only condition where enhanced release is observed, and thus provide a morphological correlate to the physiological data.

We fully agree with the reviewer’s suggestion that an ultrastructural characterization immediately after a stimulus train would be informative. Unfortunately, contract constraints prevent us from performing this experiment. For our ultrastructural morphological data, we treated synapses as individual experimental n since it is not possible to determine whether synapses in a micronetwork on one sapphire originate from the same neuron. We used 18 independent sapphires from 3 independent pups to ensure the technical and biological replication of our data and measuring independent neurons. We fully agree with the reviewers comment to be careful with ‘inflating the statistical power’ due to potential nesting effects when using synapses as experimental n. To mitigate the potential nesting effect of analyzing multiple synapses per neuron, the intracluster correlation (ICC) is calculated per variable and per nesting effect. If ICC was close to 0.1, indicating that a considerable portion of the total variance can be attributed to e.g. synapse or sapphire, multilevel analysis was performed to accommodate nested data (Aarts et al., 2014).

Minor points

(1) When a new mouse model is generated, it is usually accompanied by a thorough characterization of its properties. However, in this case, there was no information provided about the conditional SNX4 knockout mouse. This is surprising and at a minimum, the following should be provided a) the background strain, b) method of generation, c) the number of animals used to establish the colony, d) breeding strategy, e) backcrossing strategy, f) genotyping protocol.

We apologize that a thorough characterization of our novel mouse model was lacking and therefore added this to our material & methods section (page 11, line 377-391).

(2) There is a noticeable difference between WT and KO neurons during train stimulation in Figure 2f, however, this appears to be due to the fact that there is a far higher EPSC charge to begin with in KO neurons. Why is there such a disparity when there is no difference in response to single pulses (Figures 2b-d) or presynaptic plasticity (Figure 2e)?

We understand the reviewer’s concern. We excluded an outlier (3x SD) in the KO dataset that drove the initial far higher EPSC charge in the graph (was already excluded for the statistics, Supplementary table 1). The average charge of the first pulse of 40Hz train is 41 pC and for KO neurons 58 pC, which did not differ significantly.  These trains of Fig. 2f were recorded after 2 or 3 other stimulation paradigms, which can have affected the total charge released in the 40Hz train. That said, the proportional difference between groups is high comparable between Fig 2b-d and 2f, with a 37% increased average charge released in SNX4 cKO compared to control in the naïve response (Fig. 2d) and 41% increased response in the first response of the 40 Hz train (Fig. 2f), and rescued cells show a 53% reduction in average released charge compared to control in the naïve response compared to a 44% reduction in the first response of the 40 Hz train. Although the absolute values differ between these readouts, we conclude that the biological comparison between groups is consistent.

(3) Line 343-344 - "(Supplementary Figure 1a)" should be "(Figure 1a)".

We thank the reviewer for this comment and adjusted this in the manuscript.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

The study by McKim et al seeks to provide a comprehensive description of the connectivity of neurosecretory cells (NSCs) using a high-resolution electron microscopy dataset of the fly brain and several single-cell RNA seq transcriptomic datasets from the brain and peripheral tissues of the fly. They use connectomic analyses to identify discrete functional subgroups of NSCs and describe both the broad architecture of the synaptic inputs to these subgroups as well as some of the specific inputs including from chemosensory pathways. They then demonstrate that NSCs have very few traditional presynapses consistent with their known function as providing paracrine release of neuropeptides. Acknowledging that EM datasets can't account for paracrine release, the authors use several scRNAseq datasets to explore signaling between NSCs and characterize widespread patterns of neuropeptide receptor expression across the brain and several body tissues. The thoroughness of this study allows it to largely achieve it's goal and provides a useful resource for anyone studying neurohormonal signaling.

Strengths:

The strengths of this study are the thorough nature of the approach and the integration of several large-scale datasets to address short-comings of individual datasets. The study also acknowledges the limitations that are inherent to studying hormonal signaling and provides interpretations within the the context of these limitations.

Weaknesses:

Overall, the framing of this paper needs to be shifted from statements of what was done to what was found. Each subsection, and the narrative within each, is framed on topics such as "synaptic output pathways from NSC" when there are clear and impactful findings such as "NSCs have sparse synaptic output". Framing the manuscript in this way allows the reader to identify broad takeaways that are applicable to other model system. Otherwise, the manuscript risks being encyclopedic in nature. An overall synthesis of the results would help provide the larger context within which this study falls.

We agree with the reviewer and will replace all the subsection titles as suggested.

The cartoon schematic in Figure 5A (which is adapted from a 2020 review) has an error. This schematic depicts uniglomerular projection neurons of the antennal lobe projecting directly to the lateral horn (without synapsing in the mushroom bodies) and multiglomerular projection neurons projecting to the mushroom bodies and then lateral horn. This should be reversed (uniglomerular PNs synapse in the calyx and then further project to the LH and multiglomerular PNs project along the mlACT directly to the LH) and is nicely depicted in a Strutz et al 2014 publication in eLife.

We thank the reviewer for spotting this error. We will modify the schematic as suggested.

Reviewer #2 (Public review):

Summary:

The authors aim to provide a comprehensive description of the neurosecretory network in the adult Drosophila brain. They sought to assign and verify the types of 80 neurosecretory cells (NSCs) found in the publicly available FlyWire female brain connectome. They then describe the organization of synaptic inputs and outputs across NSC types and outline circuits by which olfaction may regulate NSCs, and by which Corazon-producing NSCs may regulate flight behavior. Leveraging existing transcriptomic data, they also describe the hormone and receptor expressions in the NSCs and suggest putative paracrine signaling between NSCs. Taken together, these analyses provide a framework for future experiments, which may demonstrate whether and how NSCs, and the circuits to which they belong, may shape physiological function or animal behavior.

Strengths:

This study uses the FlyWire female brain connectome (Dorkenwald et al. 2023) to assign putative cell types to the 80 neurosecretory cells (NSCs) based on clustering of synaptic connectivity and morphological features. The authors then verify type assignments for selected populations by matching cluster sizes to anatomical localization and cell counts using immunohistochemistry of neuropeptide expression and markers with known co-expression.

The authors compare their findings to previous work describing the synaptic connectivity of the neurosecretory network in larval Drosophila (Huckesfeld et al., 2021), finding that there are some differences between these developmental stages. Direct comparisons between adults and larvae are made possible through direct comparison in Table 1, as well as the authors' choice to adopt similar (or equivalent) analyses and data visualizations in the present paper's figures.

The authors extract core themes in NSC synaptic connectivity that speak to their function: different NSC types are downstream of shared presynaptic outputs, suggesting the possibility of joint or coordinated activation, depending on upstream activity. NSCs receive some but not all modalities of sensory input. NSCs have more synaptic inputs than outputs, suggesting they predominantly influence neuronal and whole-body physiology through paracrine and endocrine signaling.

The authors outline synaptic pathways by which olfactory inputs may influence NSC activity and by which Corazon-releasing NSCs may regulate flight. These analyses provide a basis for future experiments, which may demonstrate whether and how such circuits shape physiological function or animal behavior.

The authors extract expression patterns of neuropeptides and receptors across NSC cell types from existing transcriptomic data (Davie et al., 2018) and present the hypothesis that NSCs could be interconnected via paracrine signaling. The authors also catalog hormone receptor expression across tissues, drawing from the Fly Cell Atlas (Li et al., 2022).

Weaknesses:

The clustering of NSCs by their presynaptic inputs and morphological features, along with corroboration with their anatomical locations, distinguished some, but not all cell types. The authors attempt to distinguish cell types using additional methodologies: immunohistochemistry (Figure 2), retrograde trans-synaptic labeling, and characterization of dense core vesicle characteristics in the FlyWire dataset (Figure 1, Supplement 1). However, these corroborating experiments often lacked experimental replicates, were not rigorously quantified, and/or were presented as singular images from individual animals or even individual cells of interest. The assignments of DH44 and DMS types remain particularly unconvincing.

We thank the reviewer for this comment. We would like to clarify that the images presented in Figure 2 and Figure 1 Supplement 1 are representative images based on at least 5 independent samples. We will clarify this in the figure caption and methods. The electron micrographs showing dense core vesicle (DCV) characteristics (Figure 1 Supplement E-G) are also representative images based on examination of multiple neurons. However, we agree with the reviewer that a rigorous quantification would be useful to showcase the differences between DCVs from NSC subtypes. Therefore, we have now performed a quantitative analysis of the DCVs in putative m-NSC^DH44 (n=6), putative m-NSC^DMS (n=6) and descending neurons (n=4) known to express DMS. For consistency, we examined the cross section of each cell where the diameter of nuclei was the largest. We quantified the mean gray value of at least 50 DCV per cell. Our analysis shows that mean gray values of putative m-NSC^DMS and DMS descending neurons are not significantly different, whereas the mean gray values of m-NSC^DH44 are significantly larger. This analysis is in agreement with our initial conclusion.

Author response image 1.

The authors present connectivity diagrams for visualization of putative paracrine signaling between NSCs based on their peptide and receptor expression patterns. These transcriptomic data alone are inadequate for drawing these conclusions, and these connectivity diagrams are untested hypotheses rather than results. The authors do discuss this in the Discussion section.

We fully agree with the reviewer and will further elaborate on the limitations of our approach in the revised manuscript. However, there is a very high-likelihood that a given NSC subtype can signal to another NSC subtype using a neuropeptide if its receptor is expressed in the target NSC. This is due to the fact that all NSC axons are part of the same nerve bundle (nervi corpora cardiaca) which exits the brain. The axons of different NSCs form release sites that are extremely close to each other. Neuropeptides from these release sites can easily diffuse via the hemolymph to peripheral tissues that (e.g. fat body and ovaries) that are much further away from the release sites on neighboring NSCs. We believe that neuropeptide receptors are expressed in NSCs near these release sites where they can receive inputs not just from the adjacent NSCs but also from other sources such as the gut enteroendocrine cells. Hence, neuropeptide diffusion is not a limiting factor preventing paracrine signaling between NSCs and receptor expression is a good indicator for putative paracrine signaling.

Reviewer #3 (Public review):

Summary:

The manuscript presents an ambitious and comprehensive synaptic connectome of neurosecretory cells (NSC) in the Drosophila brain, which highlights the neural circuits underlying hormonal regulation of physiology and behaviour. The authors use EM-based connectomics, retrograde tracing, and previously characterised single-cell transcriptomic data. The goal was to map the inputs to and outputs from NSCs, revealing novel interactions between sensory, motor, and neurosecretory systems. The results are of great value for the field of neuroendocrinology, with implications for understanding how hormonal signals integrate with brain function to coordinate physiology.

The manuscript is well-written and provides novel insights into the neurosecretory connectome in the adult Drosophila brain. Some, additional behavioural experiments will significantly strengthen the conclusions.

Strengths:

(1) Rigorous anatomical analysis

(2) Novel insights on the wiring logic of the neurosecretory cells.

Weaknesses:

(1) Functional validation of findings would greatly improve the manuscript.

We agree with this reviewer that assessing the functional output from NSCs would improve the manuscript. Given that we currently lack genetic tools to measure hormone levels and that behaviors and physiology are modulated by NSCs on slow timescales, it is difficult to assess the immediate functional impact of the sensory inputs to NSC using approaches such as optogenetics. However, since l-NSC^CRZ are the only known cell type that provide output to descending neurons, we will functionally test this output pathway using different behavioral assays recommended by this reviewer.

eLife Assessment

This manuscript provides fundamental studies to help us better understand the effects of mutations in the presenilin-1 (PSEN1) gene on proteolytic processing of the amyloid precursor protein (APP). The authors provide compelling evidence using mutations in PSEN to understand what drives alternative substrate turnover with conclusive data and rigorous analysis. This deep mechanistic study provides a framework towards the development of small molecule inhibitors to treat Alzheimer's disease.

Reviewer #1 (Public review):

Summary:

Arafi et al. present results of studies designed to better understand the effects of mutations in the presenilin-1 (PSEN1) gene on proteolytic processing of the amyloid precursor protein (APP). This is important because APP processing can result in the production of the amyloid β-protein (Aβ), a key pathologic protein in Alzheimer's disease (AD). Aβ exists in various forms that differ in amino acid sequence and assembly state. The predominant forms of Aβ are Aβ40 and Aβ42, which are 40 and 42 amino acids in length, respectively. Shorter and longer forms derive from processive proteolysis of the Aβ region of APP by the heterotetramer β-secretase, within which presenilin 1 possesses the active site of the enzyme. Each form may become toxic if it assembles into non-natively folded, oligomeric, or fibrillar structures. A deep mechanistic understanding of enzyme-substrate interactions is a first step toward the design and successful use of small-molecule therapeutics for AD.

The key finding of Arafi et al. is that three PSEN mutations display unusual profiles of effects on Aβ production that have novel implications for the stalled E-S complex hypothesis. PSEN1 F386S is unique in that initial ε cleavage is not reduced compared with WT PSEN1; only certain trimming steps are deficient, results consistent with FLIM experiments that reveal stabilized E-S complexes only in Aβ-rich regions in the cell. In contrast, PSEN1 A431E and A434T display very little ε cleavage and therefore very little overall Aβ production, suggesting a limited role of Aβ in the pathogenesis of these two mutants and pointing to stalled E-S complexes as the common factor. For the biochemist, this may not be surprising, but in the context of understanding and treating AD, it is immense because it shifts the paradigm from targeting the results of γ-secretase action, viz., Aβ oligomers and fibrils, to targeting initial Aβ production at the molecular level. It is the equivalent of taking cancer treatment from simple removal of tumorous tissue to prevention of tumor formation and growth. Arafi et al. have provided us with a blueprint for the design of small-molecule inhibitors of γ-secretase. The significance of this achievement cannot be overstated.

Strengths and weaknesses:

The comprehensiveness and rigor of the study are notable. Rarely have I reviewed a manuscript reporting the results of so many orthogonal experiments, all of which support the authors' hypotheses, and of so many excellent controls. In addition, as found in clinical trial reports, the limitations of the study were discussed explicitly. None of these significantly affected the conclusions of the study.

Reviewer #2 (Public review):

Summary:

The work by Arafi et al. shows the effect of Familial Alzheimer's Disease presenilin-1 mutants on endoproteinase and carboxylase activity. They have elegantly demonstrated how some mutants alter each step of processing. Together with FLIM experiments, this study provides additional evidence to support their 'stalled complex hypotheses'.

Strengths:

This is a beautiful biochemical work. The approach is comprehensive.

Weaknesses:

(1) It appears that the purified g-secretase complex generates the same amount of Ab40 and Ab42, which is quite different in cellular and biochemical studies. Is there any explanation for this?

(2) It has been reported the Ab production lines from Ab49 and Ab48 can be crossed with various combinations (PMID: 23291095 and PMID: 38843321). How does the production line crossing impact the interpretation of this work?

(3) In Figure 5, did the authors look at the protein levels of PS1 mutations and C99-720, as well as secreted Ab species? Do the different amounts of PS1 full-length and PS1-NTF/CTF influence FILM results?

(4) It is interesting that both Ab40 and Ab42 Elisa kits detect Ab43. Have the authors tested other kits in the market? It might change the interpretation of some published work.

Author response:

Reviewer 2:

(1) It appears that the purified γ-secretase complex generates the same amount of Aβ40 and Aβ42, which is quite different in cellular and biochemical studies. Is there any explanation for this?

Roughly equal production of Aβ40 and Aβ42 is a phenomenon seen with purified enzyme assays, and the reason for this has not been identified. However, we suggest that what is meaningful in our studies is the relative difference between the effects of FAD-mutant vs. WT PSEN1 on each proteolytic processing step. All FAD mutations are deficient in multiple cleavage steps in γ-secretase processing of APP substrate, and these deficiencies correlate with stabilization of E-S complexes.

(2) It has been reported the Aβ production lines from Aβ49 and Aβ48 can be crossed with various combinations (PMID: 23291095 and PMID: 38843321). How does the production line crossing impact the interpretation of this work?

In the cited reports, such crossover was observed when using synthetic Aβ intermediates as substrate. In PMID 2391095 (Okochi M et al, Cell Rep, 2013), Aβ43 is primarily converted to Aβ40, but also to some extent to Aβ38. In PMID: 38843321 (Guo X et al, Science, 2024), Aβ48 is ultimately converted to Aβ42, but also to a minor degree to Aβ40. We have likewise reported such product line “crossover” with synthetic Aβ intermediates (PMID: 25239621; Fernandez MA et al, JBC, 2014). However, when using APP C99-based substrate, we did not detect any noncanonical tri- and tetrapeptide co-products of Aβ trimming events in the LC-MS/MS analyses (PMID: 33450230; Devkota S et al, JBC, 2021). In the original report on identification of the small peptide coproducts for C99 processing by γ-secretase using LC-MS/MS (PMID: 19828817; Takami M et al, J Neurosci, 2009), only very low levels of noncanonical peptides were observed. In the present study, we did not search for such noncanonical trimming coproducts, so we cannot rule out some degree of product line crossover.

(3) In Figure 5, did the authors look at the protein levels of PS1 mutations and C99-720, as well as secreted Aβ species? Do the different amounts of PS1 full-length and PS1-NTF/CTF influence FILM results?

This is a good question. Our preliminary investigation by Western Blot shows no correlation between C99 and PSEN1 expressions and FLIM results, but we will fully address the concern in our point-by-point responses submitted with a revised manuscript. 

(4) It is interesting that both Aβ40 and Aβ42 Elisa kits detect Aβ43. Have the authors tested other kits in the market? It might change the interpretation of some published work.

We have not tested other ELISA kits. In light of our findings, it would be a good idea for other investigators to test whatever ELISAs they use for specificity vis-à-vis Aβ43.

eLife Assessment

This valuable study provides a novel method to detect sleep cycles based on variations in the slope of the power spectrum from electroencephalography signals. The method, dispensing with time-consuming and potentially subjective manual identification of sleep cycles, is supported by solid evidence and analyses. This study will be of interest to researchers and clinicians working on sleep and brain dynamics.

Reviewer #1 (Public review):

In this study, Rosenblum et al introduce a novel and automatic way of calculating sleep cycles from human EEG. Previous results have shown that the slope of the non-oscillatory component of the power spectrum (called the aperiodic or fractal component) changes with sleep stage. Building on this, the authors present an algorithm that extracts the continuous-time fluctuations in the fractal slope and propose that peaks in this variable can be used to identify sleep cycle limits. Cycles defined in this way are termed "fractal cycles". The main focus of the article is a comparison of "fractal" and "classical" (ie defined manually based on the hypnogram) sleep cycles in numerous datasets.

The manuscript amply illustrates through examples the strong overlap between fractal and classical cycle identification. Accordingly, a high percentage (81%) can be matched one-to-one between methods and sleep cycle duration is well correlated (around R = 0.5). Moreover, the methods track certain global changes in sleep structure in different populations: shorter cycles in children and longer cycles in patients medicated with REM-suppressing anti-depressants. Finally, a major strength of the results is that they show similar agreement between fractal and classical sleep cycle length in 5 different data sets, showing that it is robust to changes in recording settings and methods.

The match between fractal and classical cycles is not one-to-one. For example, the fractal method identifies a correlation between age and cycle duration in adults that is not apparent with the classical method.<br /> The difference between the fractal and classical methods appear to be linked to the uncertain definition of sleep cycles since they are tied to when exactly the cycle begins/ends and whether or not to count cycles during fractured sleep architecture at sleep onset. Moreover, the discrepancies between the two are on the order of that found between classical cycles defined manually or via an automatic algorithm.

Overall the fractal cycle is an attractive method to study sleep architecture since it dispenses with time-consuming and potentially subjective manual identification of sleep cycles. However, given its difference with the classical method, it is unlikely that fractal scoring will be able to replace classical scoring directly. By providing a complementary quantification, it will likely contribute to refining the definition of sleep cycles that is currently ambiguous in certain cases. Moreover, it has the potential to be applied on animal studies which rarely deal with sleep cycle structure.

Reviewer #2 (Public review):

Summary:

This study focused on using strictly the slope of the power spectral density (PSD) to perform automated sleep scoring and evaluation of the durations of sleep cycles. The method appears to work well because the slope of the PSD is highest during slow-wave sleep, and lowest during waking and REM sleep. Therefore, when smoothed and analyzed across time, there are cyclical variations in the slope of the PSD, fit using an IRASA (Irregularly resampled auto-spectral analysis) algorithm proposed by Wen & Liu (2016).

Strengths:

The main novelty of the study is that the non-fractal (oscillatory) components of the PSD that are more typically used during sleep scoring can be essentially ignored because the key information is already contained within the fractal (slope) component. The authors show that for the most part, results are fairly consistent between this and conventional sleep scoring, but in some cases show disagreements that may be scientifically interesting.

Weaknesses:

The previous weaknesses were well-addressed by the authors in the revised manuscript. I will note that from the fractal cycle perspective, waking and REM sleep are not very dissimilar. Combining these states underlies some of the key results of this study.

Author response:

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

Reviewer 1:

Weaknesses:

The match between fractal and classical cycles is not one-to-one. For example, the fractal method identifies a correlation between age and cycle duration in adults that is not apparent with the classical method. This raises the question as to whether differences are due to one method being more reliable than another or whether they are also identifying different underlying biological differences. It is not clear for example whether the agreement between the two methods is better or worse than between two human scorers, which generally serve as a gold standard to validate novel methods. The authors provide some insight into differences between the methods that could account for differences in results. However, given that the fractal method is automatic it would be important to clearly identify criteria for recordings in which it will produce similar results to the classical method.

We thank the reviewer for the insightful suggestions. In the revised Manuscript, we have added a number of additional analyses that provide a quantitative comparison between the classical and fractal cycle approaches aiming to identify the source of the discrepancies between classical and fractal cycle durations. Likewise, we assessed the intra-fractal and intra-classical method reliability.

Reviewer 2:

One weakness of the study, from my perspective, was that the IRASA fits to the data (e.g. the PSD, such as in Figure 1B), were not illustrated. One cannot get a sense of whether or not the algorithm is based entirely on the fractal component or whether the oscillatory component of the PSD also influences the slope calculations. This should be better illustrated, but I assume the fits are quite good.

Thank you for this suggestion. In the revised Manuscript, we have added a new figure (Fig.S1 E, Supplementary Material 2), illustrating the goodness of fit of the data as assessed by the IRASA method.

The cycles detected using IRASA are called fractal cycles. I appreciate the use of a simple term for this, but I am also concerned whether it could be potentially misleading? The term suggests there is something fractal about the cycle, whereas it's really just that the fractal component of the PSD is used to detect the cycle. A more appropriate term could be "fractal-detected cycles" or "fractal-based cycle" perhaps?

We agree that these cycles are not fractal per se. In the Introduction, when we mention them for the first time, we name them “fractal activity-based cycles of sleep” and immediately after that add “or fractal cycles for short”. In the revised version, we renewed this abbreviation with each new major section and in Abstract. Nevertheless, given that the term “fractal cycles” is used 88 times, after those “reminders”, we used the short name again to facilitate readability. We hope that this will highlight that the cycles are not fractal per se and thus reduce the possible confusion while keeping the manuscript short.

The study performs various comparisons of the durations of sleep cycles evaluated by the IRASA-based algorithm vs. conventional sleep scoring. One concern I had was that it appears cycles were simply identified by their order (first, second, etc.) but were not otherwise matched. This is problematic because, as evident from examples such as Figure 3B, sometimes one cycle conventionally scored is matched onto two fractal-based cycles. In the case of the Figure 3B example, it would be more appropriate to compare the duration of conventional cycle 5 vs. fractal cycle 7, rather than 5 vs. 5, as it appears is currently being performed.

In cases where the number of fractal cycles differed from the number of classical cycles (from 34 to 55% in different datasets as in the case of Fig.3B), we did not perform one-to-one matching of cycles. Instead, we averaged the duration of the fractal and classical cycles over each participant and only then correlated between them (Fig.2C). For a subset of the participants (45 – 66% of the participants in different datasets) with a one-to-one match between the fractal and classical cycles, we performed an additional correlation without averaging, i.e., we correlated the durations of individual fractal and classical cycles (Fig.4S of Supplementary Material 2). This is stated in the Methods, section Statistical analysis, paragraph 2.

There are a few statements in the discussion that I felt were either not well-supported. L629: about the "little biological foundation" of categorical definitions, e.g. for REM sleep or wake? I cannot agree with this statement as written. Also about "the gradual nature of typical biological processes". Surely the action potential is not gradual and there are many other examples of all-or-none biological events.

In the revised Manuscript, we have removed these statements from both Introduction and Discussion.

The authors appear to acknowledge a key point, which is that their methods do not discriminate between awake and REM periods. Thus their algorithm essentially detected cycles of slow-wave sleep alternating with wake/REM. Judging by the examples provided this appears to account for both the correspondence between fractal-based and conventional cycles, as well as their disagreements during the early part of the sleep cycle. While this point is acknowledged in the discussion section around L686. I am surprised that the authors then argue against this correspondence on L695. I did not find the "not-a-number" controls to be convincing. No examples were provided of such cycles, and it's hard to understand how positive z-values of the slopes are possible without the presence of some wake unless N1 stages are sufficient to provide a detected cycle (in which case, then the argument still holds except that its alterations between slow-wave sleep and N1 that could be what drives the detection).

In the revised Manuscript, we have removed the “NaN analysis” from both Results and Discussion. We have replaced it with the correlation between the difference between the durations of the classical and fractal cycles and proportion of wake after sleep onset. The finding is as follows:

“A larger difference between the durations of the classical and fractal cycles was associated with a higher proportion of wake after sleep onset in 3/5 datasets as well as in the merged dataset (Supplementary Material 2, Table S10).” Results, section “Fractal cycles and wake after sleep onset”, last two sentences. This is also discussed in Discussion, section “Fractal cycles and age”, paragraph 1, last sentence. 

To me, it seems important to make clear whether the paper is proposing a different definition of cycles that could be easily detected without considering fractals or spectral slopes, but simply adjusting what one calls the onset/offset of a cycle, or whether there is something fundamentally important about measuring the PSD slope. The paper seems to be suggesting the latter but my sense from the results is that it's rather the former.

Thank you for this important comment. Overall, our paper suggests that the fractal approach might reflect the cycling nature of sleep in a more precise and sensitive way than classical hypnograms. Importantly, neither fractal nor classical methods can shed light on the mechanism underlying sleep cycle generation due to their correlational approach. Despite this, the advantages of fractal over classical methods mentioned in our Manuscript are as follows:

(1) Fractal cycles are based on a real-valued metric with known neurophysiological functional significance, which introduces a biological foundation and a more gradual impression of nocturnal changes compared to the abrupt changes that are inherent to hypnograms that use a rather arbitrary assigned categorical value (e.g., wake=0, REM=-1, N1=-2, N2=-3 and SWS=-4, Fig.2 A).

(2) Fractal cycle computation is automatic and thus objective, whereas classical sleep cycle detection is usually based on the visual inspection of hypnograms, which is time-consuming, subjective and error-prone. Few automatic algorithms are available for sleep cycle detection, which only moderately correlated with classical cycles detected by human raters (r’s = 0.3 – 0.7 in different datasets here).

(3) Defining the precise end of a classical sleep cycle with skipped REM sleep that is common in children, adolescents and young adults using a hypnogram is often difficult and arbitrary.   The fractal cycle algorithm could detect such cycles in 93% of cases while the hypnogram-based agreement on the presence/absence of skipped cycles between two independent human raters was 61% only; thus, 32% lower.

(4) The fractal analysis showed a stronger effect size, higher F-value and R-squared than the classical analysis for the cycle duration comparison in children and adolescents vs young adults. The first and second fractal cycles were significantly shorter in the pediatric compared to the adult group, whereas the classical approach could not detect this difference.

(5) Fractal – but not classical – cycle durations correlated with the age of adult participants.

These bullets are now summarized in Table 5 that has been added to the Discussion of the revised manuscript.

Reviewer #1 (Recommendations for the authors):

The authors have added a lot of quantifications to provide a more complete comparison of classical and fractal cycles that address the points I raised.

Regarding, the question of skipped REM cycles: I am not sure the comparison of skipped cycle accuracies between fractal and manual methods makes sense. To make a fair comparison fractal and 2nd scorer classifications should be compared to the same baseline dataset which doesn't seem to be the case since the number of skipped cycles is not the same. Moreover, it's not indicated whether the fractal method identifies any false positive skipped cycles.

Thank you for this comment. In the revised Manuscript, we have reported the number of false positive skipped cycles identified by the fractal algorithm. Likewise, we have added the comparison between the fractal algorithm and the second scorer detection of cycles with skipped REM sleep (Results, the section “Skipped cycles”, last paragraph). The text has been revised as follows:

“Visual inspection of the hypnograms from Datasets 1 – 6 was performed by two independent researchers. Scorer 1 and Scorer 2 detected that out of 226 first sleep cycles 58 (26%) and 64 (28%), respectively, lacked REM episodes. The agreement on the presence of skipped cycles between two human raters equaled 91% (58 cycles detected by both raters out of 64 cycles detected by either one or two scorers). The fractal cycle algorithm detected skipped cycles in 57 out of 58 (98%) cases detected by Scorer 1 with one false positive (which, however, was tagged as a skipped cycle by Scorer2), and in 58 out of 64 (91%) cases detected by Scorer 2 with no false positives.”

Minor points

I suggest reporting the values of inter-method / inter-scorer correlations with the classical method in the main text since otherwise interpreting the value for fractal vs classical is impossible.

Thank you for this comment. In the revised Manuscript, we have moved this section to the main text (Table 3).

Table 5 + text of discussion: cycle identification based on hypnograms is claimed to be. "based on arbitrary assigned categorical values" the categories are not arbitrary since they correspond to well-validate sleep states, only the number associated it and this does not seem to be very important since it's only for visualization purposes.

Thank you for this comment. In the revised Manuscript, we have removed the phrase “arbitrary assigned“.

eLife Assessment

This important study investigates how working memory load influences the Stroop effect from a temporal dynamics perspective. Convincing evidence is provided that the working memory load influences the Stroop effect in the late-stage stimulus-response mapping instead of the early sensory stage. This study will be of interest to both neuroscientists and psychologists who work on cognitive control.

Reviewer #1 (Public review):

Summary:

This study investigates an intriguing question in cognitive control from a temporal dynamics perspective: why does concurrent verbal working memory load eliminate the color-word Stroop effect? Through a series of thorough data analyses, the authors propose that verbal working memory load occupies the stimulus-response mapping resources represented by theta-band activity, thereby disrupting the mapping process for task-irrelevant distractors. This reduces the response tendency to the distractors, ultimately leading to the elimination of the Stroop effect.

Strengths:

The behavioral and neural evidence presented in the manuscript is solid, and the findings have valuable theoretical implications for research on Stroop conflict processing.

Comments on revisions:

The authors have addressed all concerns

Reviewer #2 (Public review):

Summary

Li et al. explored which stage of Stroop conflict processing was influenced by working memory loads. Participants completed a single task (Stroop task) and a dual task (the Sternberg working memory task combined with the Stroop task) while their EEG data was recorded. They adopted the event-related potential (ERP), and multivariate pattern analyses (MVPA) to investigate the interaction effect of task (single/dual) and congruency (congruent/incongruent). The results showed that the interaction effect was significant on the sustained potential (SP; 650-950 ms), the late theta (740-820 ms), and beta (920-1040 ms) power but not significant on the early P1 potential (110-150 ms). They used the representational similarity analyses (RSA) method to explore the correlation between behavioral and neural data, and the results revealed a significant contribution of late theta activity.

Strength

The experiment is well designed.<br /> The data were analyzed in depth from both time and frequency domain perspectives by combining several methods.

Comments on revisions:

All my concerns have been properly addressed, no further comments.

Author response:

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

Reviewer #1 (Public review):

Comment 1: In the Results section, the rationale behind selecting the beta band for the central (C3, CP3, Cz, CP4, C4) regions and the theta band for the fronto-central (Fz, FCz, Cz) regions is not clearly explained in the main text. This information is only mentioned in the figure captions. Additionally, why was the beta band chosen for the S-ROI central region and the theta band for the S-ROI fronto-central region? Was this choice influenced by the MVPA results?

We thank the reviewer for the question regarding the rationale for the S-ROI selection in our study. The beta band was chosen for the central region due to its established relevance in motor control (Engel & Fries, 2010), movement planning (Little et al., 2019) and motor inhibition (Duque et al., 2017). The fronto-central theta band (or frontal midline theta) was a widely recognized indicator in cognitive control research (Cavanagh & Frank, 2014), associated with conflict detection and resolution processes. Moreover, recent empirical evidence suggested that the fronto-central theta reflected the coordination and integration between stimuli and responses (Senoussi et al., 2022). Although we have described the cognitive processes linked to these different frequencies in the introduction and discussion sections, along with the potential patterns of results observed in Stroop-related studies, we did not specify the involved cortical areas. Therefore, we have specified these areas in the introduction to enhance the clarity of the revised version (in the fourth paragraph of the Introduction section).

Regarding whether the selection of S-ROIs was influenced by the MVPA results, we would like to clarify here that we selected the S-ROIs based on prior research and then conducted the decoding analysis. Specifically, we first extracted the data representing different frequency indicators (three F-ROIs and three S-ROIs) as features, followed by decoding to obtain the MVPA results. Subsequently, the time-frequency analysis, combined with the specific time windows during which each frequency was decoded, provided detailed interaction patterns among the variables for each indicator. The specifics of feature selection are described in the revised version (in the first paragraph of the Multivariate Pattern Analysis section).

Comment 2: In the Data Analysis section, line 424 states: “Only trials that were correct in both the memory task and the Stroop task were included in all subsequent analyses. In addition, trials in which response times (RTs) deviated by more than three standard deviations from the condition mean were excluded from behavioral analyses.” The percentage of excluded trials should be reported. Also, for the EEG-related analyses, were the same trials excluded, or were different criteria applied?

We thank the reviewer for this suggestion. Beyond the behavioral exclusion criteria, trials with EEG artifacts were also excluded from the data for the EEG-related analyses. We have now reported the percentage of excluded trials for both behavioral and EEG data analyses in the revised version (in the second paragraph of the EEG Recording and Preprocessing section and the first paragraph of the Behavioral Analysis section).

Comment 3: In the Methods section, line 493 mentions: “A 400-200 ms pre-stimulus time window was selected as the baseline time window.” What is the justification in the literature for choosing the 400-200 ms pre-stimulus window as the baseline? Why was the 200-0 ms pre-stimulus period not considered?

We thank the reviewer for this question and would like to provide the following justification. First, although a baseline ending at 0 ms is common in ERP analyses, it may not be suitable for time-frequency analysis. Due to the inherent temporal smoothing characteristic of wavelet convolution in time-frequency decomposition, task-related early activities can leak into the pre-stimulus period (before 0 ms) (Cohen, 2014). This means that extending the baseline to 0 ms will include some post-stimulus activity in the baseline window, thereby increasing baseline power and compromising the accuracy of the results. Second, an ideal baseline duration is recommended to be around 10-20% of the entire trial of interest (Morales & Bowers, 2022). In our study, the epoch duration was 2000 ms, making 200-400 ms an appropriate baseline length. Third, given that the minimum duration of the fixation point before the stimulus in our experiment was 400 ms, we chose the 400 ms before the stimulus as the baseline point to ensure its purity. In summary, considering edge effects, duration requirements, and the need to exclude other influences, we selected a baseline correction window of -400 to -200 ms. To enhance the clarity of the revised version, we have provided the rationale for the selected time windows along with relevant references (in the first paragraph of the Time-frequency analysis section).

Comment 4: Is the primary innovation of this study limited to the methodology, such as employing MVPA and RSA to establish the relationship between late theta activity and behavior?

We thank the reviewer for this insightful question and would like to clarify that our research extends beyond mere methodological innovation; rather, it utilized new methods to explore novel theoretical perspectives. Specifically, our research presents three levels of innovation: methodological, empirical, and theoretical. First, methodologically, MVPA overcame the drawbacks of traditional EEG analyses based on specific averaged voltage intensities, providing new perspectives on how the brain dynamically encoded particular neural representations over time. Furthermore, RSA aimed to identify which indicators among the decoded were directly related to behavioral representation patterns. Second, in terms of empirical results, using these two methods, we have identified for the first time three EEG markers that modulate the Stroop effect under verbal working memory load: SP, late theta, and beta, with late theta being directly linked to the elimination of the behavioral Stroop effect. Lastly, from a theoretical perspective, we proposed the novel idea that working memory played a crucial role in the late stages of conflict processing, specifically in the stimulus-response mapping stage (the specific theoretical contributions are detailed in the second-to-last paragraph of the Discussion section).

Comment 5: On page 14, lines 280-287, the authors discuss a specific pattern observed in the alpha band. However, the manuscript does not provide the corresponding results to substantiate this discussion. It is recommended to include these results as supplementary material.

We thank the reviewer for this suggestion. We added a new figure along with the corresponding statistical results that displayed the specific result patterns for the alpha band (Supplementary Figure 1).

Comment 6: On page 16, lines 323-328, the authors provide a generalized explanation of the findings. According to load theory, stimuli compete for resources only when represented in the same form. Since the pre-memorized Chinese characters are represented semantically in working memory, this explanation lacks a critical premise: that semantic-response mapping is also represented semantically during processing.

We thank the reviewer for this insightful suggestion. We fully agree with the reviewer’s perspective. As stated in our revised version, load theory suggests that cognitive resources are limited and dependent on a specific type (in the second paragraph of the Discussion section). The previously memorized Chinese characters are stored in working memory in the form of semantic representations; meanwhile the stimulus-response mapping should also be represented semantically, leading to resource occupancy. We have included this logical premise in the revised version (in the third-to-last paragraph of the Discussion section).

Comment 7: The classic Stroop task includes both a manual and a vocal version. Since stimulus-response mapping in the vocal version is more automatic than in the manual version, it is unclear whether the findings of this study would generalize to the impact of working memory load on the Stroop effect in the vocal version.

We fully agree with the reviewer’s point that the verbal version of the Stroop task differs from the manual version in terms of the degree of automation in the stimulus-response mapping. Specifically, the verbal version relies on mappings that are established through daily language use, while the manual version involves arbitrary mappings created in the laboratory. Therefore, the stimulus-response mapping in the verbal response version is more automated and less likely to be suppressed. However, our previous research indicated that the degree of automation in the stimulus-response mapping was influenced by practice (Chen et al., 2013). After approximately 128 practice trials, semantic conflict almost disappears, suggesting that the level of automation in stimulus-response mapping for the verbal Stroop task is comparable to that of the manual version (Chen et al., 2010). Given that participants in our study completed 144 practice trials (in the Procedure section), we believe these findings can be generalized to the verbal version.

Comment 8: While the discussion section provides a comprehensive analysis of the study’s results, the authors could further elaborate on the theoretical and practical contributions of this work.

We thank the reviewer for the constructive suggestions. We recognize that the theoretical and practical contributions of the study were not thoroughly elaborated in the original manuscript. Therefore, we have now provided a more detailed discussion. Specifically, the theoretical contributions focus on advancing load theory and highlighting the critical role of working memory in conflict processing. The practical contributions emphasize the application of load theory and the development of intervention strategies for enhancing inhibitory control. A more detailed discussion can be found in the revised version (in the second-to-last paragraph of the Discussion section).

Reviewer #2 (Public review):

Comment 1: As the researchers mentioned, a previous study reported a diminished Stroop effect with concurrent working memory tasks to memorize meaningless visual shapes rather than memorize Chinese characters as in the study. My main concern is that lower-level graphic processing when memorizing visual shapes also influences the Stroop effect. The stage of Stroop conflict processing affected by the working memory load may depend on the specific content of the concurrent working memory task. If that’s the case, I sense that the generalization of this finding may be limited.

We thank the reviewer for this insightful concern. As mentioned in the manuscript, this may be attributed to the inherent characteristics of Chinese characters. In contrast to English words, the processing of Chinese characters relies more on graphemic encoding and memory (Chen, 1993). Therefore, the processing of line patterns essentially occupies some of the resources needed for character processing, which aligns with our study’s hypothesis based on dimensional overlap. Additionally, regarding the results, even though the previous study presents lower-level line patterns, the results still showed that the working memory load modulated the later theta band. We hypothesize that, regardless of the specific content of the pre-presented working memory load, once the stimulus disappears from view, these loads are maintained as representations in the working memory platform. Therefore, they do not influence early perceptual processing, and resource competition only occurs once the distractors reach the working memory platform. Lastly, previous study has shown that spatial loads, which do not overlap with either the target or distractor dimensions, do not influence conflict effect (Zhao et al., 2010). Taken together, we believe that regardless of the specific content of the concurrent working memory tasks, as long as they occupy resources related to irrelevant stimulus dimensions, they can influence the late-stage processing of conflict effect. Perhaps our original manuscript did not convey this clearly, so we have rephrased it in a more straightforward manner (in the second paragraph of the Discussion section).

Comment 2: The P1 and N450 components are sensitive to congruency in previous studies as mentioned by the researchers, but the results in the present study did not replicate them. This raised concerns about data quality and needs to be explained.

We thank the reviewer for this insightful concern. For P1, we aimed to convey that the early perceptual processing represented by P1 is part of the conflict processing process. Therefore, we included it in our analysis. Additionally, as mentioned in the discussion, most studies find P1 to be insensitive to congruency. However, we inappropriately cited a study in the introduction that suggested P1 shows differences in congruency, which is among the few studies that hold this perspective. To prevent confusion for readers, we have removed this citation from the introduction.

As for N450, most studies have indeed found it to be influenced by congruency. In our manuscript, we did not observe a congruency effect at our chosen electrodes and time window. However, significant congruency effects were detected at other central-parietal electrodes (CP3, CP4, P5, P6) during the 350-500 ms interval. The interaction between task type and consistency remained non-significant, consistent with previous results. Furthermore, with respect to the location of the electrodes chosen, existing studies on N450 vary widely, including central-parietal electrodes and frontal-central electrodes (for a review, see Heidlmayr et al., 2020). We speculate that this phenomenon may be related to the extent of practice. With fewer total trials, the task may involve more stimulus conflicts, engaging more frontal brain areas. On the other hand, with more total trials, the task may involve more response conflicts, engaging more central-parietal brain areas (Chen et al., 2013; van Veen & Carter, 2005). Due to the extensive practice required in our study, we identified a congruency N450 effect in the central-parietal region. We apologize for not thoroughly exploring other potential electrodes in the previous manuscript, and we have revised the results and interpretations regarding N450 accordingly in the revised version (in the N450 section of the ERP results and the third paragraph of the Discussion section).

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Comment 1: In the Introduction, line 108 states: “Second, alpha oscillations (8-13 Hz) can serve as a neural inverse index of mental activity or alertness, while a decrease in alpha power reflects increased alertness or enhanced attentional inhibition of distractors (Arakaki et al., 2022; Tafuro et al., 2019; Zhou et al., 2023; Zhu et al., 2023).” Please clarify which specific psychological process related to conflict processing is reflected by alpha oscillations.

We appreciate your suggestion and we have clearly highlighted the role of alpha oscillations in attentional engagement during conflict processing in the revised version (in the third-to-last paragraph of the introduction).

Comment 2: In Figures 3C and 3E, a space is needed between “amplitude” and the preceding parenthesis. Similar adjustments are required in Figures 4A, 4B, 4C, 5C, and 6C. Additionally, in Figures 3B and 3D, a space should be added between the numbers and “ms.” This issue also appears in Figure 8. Please review all figures for these formatting inconsistencies.

We apologize for the inconsistency in formatting and have corrected them throughout the revised version.

Comment 3: There are some clerical errors in the manuscript that need correction. For instance, on page 19, line 403: “Participants were asked to answer by pressing one of two response buttons (“S with the left ring finger and “L” with the left ring finger).” This should be corrected to: “L” with the right ring finger. I recommend that the authors carefully proofread the manuscript to identify and correct such errors.

We sincerely apologize for the errors present in the manuscript and have now carefully proofread it (in the Procedure section).

Comment 4: On page 13, line 254, the elimination of the Stroop effect should not be interpreted as an improvement in processing.

We greatly appreciate your suggestion. We agree that the elimination of the Stroop effect should not be confused with improvements in processing. We have corrected this in the revised version (the second paragraph of the Discussion section).

Reviewer #3 (Recommendations for the authors):

Comment 1: In the introduction section, the N450 was introduced as “a frontal-central negative deflection”, but in the methods part the N450 was computed using central-parietal electrodes. This inconsistency is confusing and needs to be clarified.

We apologize for this confusion. We have provided a detailed explanation regarding the differences in electrodes and the rationale behind choosing central-parietal electrodes in our response to Reviewer 2’s second comment. To clarify, we have updated the introduction to consistently label them as central-parietal deflections (in the third paragraph of the Introduction section).

Comment 2: I speculate the “beta” was mistakenly written as “theta” in line 212.

We sincerely apologize for this mistake. We have corrected this error (in the RSA results section).

Comment 3: The speculation that “changes in beta bands may be influenced by theta bands, thereby indirectly influencing the behavioral Stroop effect” needs to be rationalized.

We appreciate your suggestion. What we intended to convey is that we found an interaction effect in the beta bands; however, the RSA results did not show a correlation with the behavioral interaction effect. We speculate that beta activity might be influenced by the theta bands. On the one hand, we realize that the idea of beta bands indirectly influencing the behavioral Stroop effect was inappropriate, and we have removed this point in the revised version. On the other hand, we have provided rational evidence for the idea that beta bands may be influenced by theta bands. This is based on the biological properties of theta oscillations, which support communication between different cortical neural signals, and their functional role in integrating and transmitting task-relevant information to response execution (in the third-to-last paragraph of the Discussion section).

Comment 4: Typo in line 479: [10,10].

We sincerely apologize for this mistake. We have corrected this error: [-10,10] (in the Multivariate pattern analysis section).

Reference

Cavanagh, J. F., & Frank, M. J. (2014). Frontal theta as a mechanism for cognitive control. Trends in Cognitive Sciences, 18(8), 414–421. https://doi.org/10.1016/j.tics.2014.04.012

Chen, M. J. (1993). A Comparison of Chinese and English Language Processing. In Advances in Psychology (Vol. 103, pp. 97–117). North-Holland. https://doi.org/10.1016/S0166-4115(08)61659-3

Chen, X. F., Jiang, J., Zhao, X., & Chen, A. (2010). Effects of practice on semantic conflict and response conflict in the Stroop task. Psychol. Sci., 33, 869–871.

Chen, Z., Lei, X., Ding, C., Li, H., & Chen, A. (2013). The neural mechanisms of semantic and response conflicts: An fMRI study of practice-related effects in the Stroop task. NeuroImage, 66, 577–584. https://doi.org/10.1016/j.neuroimage.2012.10.028

Cohen, M. X. (2014). Analyzing Neural Time Series Data: Theory and Practice. The MIT Press. https://doi.org/10.7551/mitpress/9609.001.0001

Duprez, J., Gulbinaite, R., & Cohen, M. X. (2020). Midfrontal theta phase coordinates behaviorally relevant brain computations during cognitive control. NeuroImage, 207, 116340. https://doi.org/10.1016/j.neuroimage.2019.116340

Duque, J., Greenhouse, I., Labruna, L., & Ivry, R. B. (2017). Physiological Markers of Motor Inhibition during Human Behavior. Trends in Neurosciences, 40(4), 219–236. https://doi.org/10.1016/j.tins.2017.02.006

Engel, A. K., & Fries, P. (2010). Beta-band oscillations—Signalling the status quo? Current Opinion in Neurobiology, 20(2), 156–165. https://doi.org/10.1016/j.conb.2010.02.015

Heidlmayr, K., Kihlstedt, M., & Isel, F. (2020). A review on the electroencephalography markers of Stroop executive control processes. Brain and Cognition, 146, 105637. https://doi.org/10.1016/j.bandc.2020.105637

Little, S., Bonaiuto, J., Barnes, G., & Bestmann, S. (2019). Human motor cortical beta bursts relate to movement planning and response errors. PLOS Biology, 17(10), e3000479. https://doi.org/10.1371/journal.pbio.3000479

Morales, S., & Bowers, M. E. (2022). Time-frequency analysis methods and their application in developmental EEG data. Developmental Cognitive Neuroscience, 54, 101067. https://doi.org/10.1016/j.dcn.2022.101067

Senoussi, M., Verbeke, P., Desender, K., De Loof, E., Talsma, D., & Verguts, T. (2022). Theta oscillations shift towards optimal frequency for cognitive control. Nature Human Behaviour, 6(7), Article 7. https://doi.org/10.1038/s41562-022-01335-5

van Veen, V., & Carter, C. S. (2005). Separating semantic conflict and response conflict in the Stroop task: A functional MRI study. NeuroImage, 27(3), 497–504. https://doi.org/10.1016/j.neuroimage.2005.04.042

Zhao, X., Chen, A., & West, R. (2010). The influence of working memory load on the Simon effect. Psychonomic Bulletin & Review, 17(5), 687–692. https://doi.org/10.3758/PBR.17.5.687

eLife Assessment

This study uses carefully designed experiments to generate a useful behavioural and neuroimaging dataset on visual cognition. The results provide solid evidence for the involvement of higher-order visual cortex in processing visual oddballs and asymmetry. However, the evidence provided for the very strong claims of homogeneity as a novel concept in vision science, separable from existing concepts such as target saliency, is incomplete. The authors and the reviewers do not agree on several points, which are explained in the reviews and author response.

Reviewer #1 (Public review):

Summary:

The authors define a new metric for visual displays, derived from psychophysical response times, called visual homogeneity (VH). They attempt to show that VH is explanatory of response times across multiple visual tasks. They use fMRI to find visual cortex regions with VH-correlated activity. On this basis, they declare a new visual region in human brain, area VH, whose purpose is to represent VH for the purpose of visual search and symmetry tasks.

Link to original review: https://elifesciences.org/reviewed-preprints/93033v2/reviews#peer-review-0

Comments on latest version:

Authors rebuttal: We agree that visual homogeneity is similar to existing concepts such as target saliency, memorability etc. We have proposed it as a separate concept because visual homogeneity has an independent empirical measure (the reciprocal of target-absent search time in oddball search, or the reciprocal of same response time in a same-different task, etc) that may or may not be the same as other empirical measures such as saliency and memorability. Investigating these possibilities is beyond the scope of our study but would be interesting for future work. We have now clarified this in the revised manuscript (Discussion, p. 42).

Reviewer response to rebuttal: Neither the original ms nor the comments on that ms pretended that "visual homogeneity" was entirely separate from target saliency etc. So this is a response to a criticism that was never made. What the authors do claim, and what the comments question, is that they have successfully subsumed long-recognized psychophysical concepts like target saliency etc. under a new, uber-concept, "visual homogeneity" that explains psychophysical experimental results in a more unified and satisfying way. This subsumption of several well-established psychophysical concepts under a new, unified category is what reviewers objected to.

Authors rebuttal: However, we'd like to emphasize that the question of whether visual homogeneity is novel or related to existing concepts misses entirely the key contribution of our study.

Reviewer response to rebuttal: Sorry, but the claim of a new uber-concept in psychophysics, "visual homogeneity", is a major claim of the paper. The fact that it is not the only claim made does not absolve the authors from having to prove it satisfactorily.

"Authors rebuttal: "In addition, the large regions of VH correlations identified in Experiments 1 and 2 vs. Experiments 3 and 4 are barely overlapping. This undermines the claim that VH is a universal quantity, represented in a newly discovered area of visual cortex, that underlies a wide variety of visual tasks and functions."<br /> • We respectfully disagree with your assertion. First of all, there is partial overlap between the VH regions, for which there are several other obvious explanations that must be considered first before dismissing VH outright as a flawed construct. We acknowledge these alternatives in the Results (p. 27), and the relevant text is reproduced below.

"We note that it is not straightforward to interpret the overlap between the VH regions identified in Experiments 2 & 4. The lack of overlap could be due to stimulus differences (natural images in Experiment 2 vs silhouettes in Experiment 4), visual field differences (items in the periphery in Experiment 2 vs items at the fovea in Experiment 4) and even due to different participants in the two experiments. There is evidence supporting all these possibilities: stimulus differences (Yue et al., 2014), visual field differences (Kravitz et al., 2013) as well as individual differences can all change the locus of neural activations in object-selective cortex (Weiner and Grill-Spector, 2012a; Glezer and Riesenhuber, 2013). We speculate that testing the same participants on search and symmetry tasks using similar stimuli and display properties would reveal even larger overlap in the VH regions that drive behavior."

Reviewer response to rebuttal: The authors are saying that their results merely look unconvincing (weak overlap between VH regions defined in different experiments) because there were confounding differences between their experiments, in subject population, stimuli, etc. That is possible, but in that case it is up to the authors to show that their definition of a new "area VH" is convincing when the confounding differences are resolved, e.g. by using the same stimuli in the different experiments they attempt to agglomerate here. That would require new experiments, and none are offered in this revision.

Authors rebuttal: • Thank you for carefully thinking through our logic. We agree that a distance-to-centre calculation is entirely unnecessary as an explanation for target-present visual search. The similarity between target and distractor, so there is nothing new to explain here. However, this is a narrow and selective interpretation of our findings because you are focusing only on our results on target-present searches, which are only half of all our data. The other half is the target-absent responses which previously have had no clear explanation. You are also missing the fact that we are explaining same-different and symmetry tasks as well using the same visual homogeneity computation. We urge you to think more deeply about the problem of how to decide whether an oddball is present or not in the first place. How do we actually solve this task?

Reviewer response to rebuttal: It is the role of the authors to think deeply about their paper and on that basis present a clear and compelling case that readers can understand quickly and agree with. That is not done here.

Authors rebuttal: There must be some underlying representation and decision process. Our study shows that a distance-to-centre computation can actually serve as a decision variable to solve disparate property-based visual tasks. These tasks pose a major challenge to standard models of decision-making because the underlying representation and decision variable have been unclear. Our study resolves this challenge by proposing a novel computation that can be used by the brain to solve all these disparate tasks, and bring these tasks into the ambit of standard theories of decision-making.

Reviewer response to rebuttal: There is only a "challenge" if you accept the authors' a priori assumption that all of these tasks must have a common explanation and rely on a single neural mechanism. I do not accept that assumption, and I don't think the authors provide evidence to support the assumption. There is nothing "unclear" about how search, oddball, etc. have been thoroughly explained, separately, in the psychophysical literature that spans more than a century.

Authors rebuttal: • You are indeed correct in noting that both Experiment 1 & 2 involve oddball search, and so at the superficial level, it looks circular that the oddball search data of Experiment 1 is being used to explain the oddball search data of Experiment 2.<br /> However a deeper scrutiny reveals more fundamental differences: Experiment 1 consisted of only oddball search with the target appearing on the left or right, whereas Experiment 2 consisted of oddball search with the target either present or completely absent. In fact, we were merely using the search dissimilarities from Experiment 1 to reconstruct the underlying object representation, because it is well-known that neural dissimilarities are predicted well by search dissimilarities (Sripati & Olson, 2009; Zhivago et al, 2014).

Reviewer response to rebuttal: Here again the authors cite differences between their multiple experiments as a virtue that supports their conclusions. Instead, the experiments should have been designed for maximum similarity if the authors intended to explain them with the same theory.

Authors rebuttal: To thoroughly refute any lingering concern about circularity, we reasoned that the model predictions for Experiment 2 could have been obtained by a distance-to-center computation on any brain like object representation. To this end, we used object representations from deep neural networks pretrained on object categorization, whose representations are known to match well with the brain, and asked if a distance-to-centre computation on these representations could predict the search data in Experiment 2. This was indeed the case, and these results are now included an additional section in Supplementary Material (Section S1).

Reviewer response to rebuttal: The authors' claims are about human performance and how it is based on the human brain. Their claims are not well supported by the human experiments that they performed. It serves no purpose to redo the same experiments in silico, which cannot provide stronger evidence that compensates for what was lacking in the human data.

Authors rebuttal: "Confirming the generality of visual homogeneity<br /> We performed several additional analyses to confirm the generality of our results, and to reject alternate explanations.

First, it could be argued that our results are circular because they involve taking oddball search times from Experiment 1 and using them to explain search response times in Experiment 2. This is a superficial concern since we are using the search dissimilarities from Experiment 1 only as a proxy for the underlying neural representation, based on previous reports that neural dissimilarities closely match oddball search dissimilarities (Sripati and Olson, 2010; Zhivago and Arun, 2014). Nonetheless, to thoroughly refute this possibility, we reasoned that we would get similar predictions of the target present/absent responses in Experiment using any other brain-like object representation. To confirm this, we replaced the object representations derived from Experiment 1 with object representations derived from deep neural networks pretrained for object categorization, and asked if distance-to-center computations could predict the target present/absent responses in Experiment 2. This was indeed the case (Section S1).

Second, we wondered whether the nonlinear optimization process of finding the best-fitting center could be yielding disparate optimal centres each time. To investigate this, we repeated the optimization procedure with many randomly initialized starting points, and obtained the same best-fitting center each time (see Methods).

Third, to confirm that the above model fits are not due to overfitting, we performed a leave-one-out cross validation analysis. We left out all target-present and target-absent searches involving a particular image, and then predicted these searches by calculating visual homogeneity estimated from all other images. This too yielded similar positive and negative correlations (r = 0.63, p < 0.0001 for target-present, r = -0.63, p < 0.001 for target-absent).

Fourth, if heterogeneous displays indeed elicit similar neural responses due to mixing, then their average distance to other objects must be related to their visual homogeneity. We confirmed that this was indeed the case, suggesting that the average distance of an object from all other objects in visual search can predict visual homogeneity (Section S1).

Fifth, the above results are based on taking the neural response to oddball arrays to be the average of the target and distractor responses. To confirm that averaging was indeed the optimal choice, we repeated the above analysis by assuming a range of relative weights between the target and distractor. The best correlation was obtained for almost equal weights in the lateral occipital (LO) region, consistent with averaging and its role in the underlying perceptual representation (Section S1).

Finally, we performed several additional experiments on a larger set of natural objects as well as on silhouette shapes. In all cases, present/absent responses were explained using visual homogeneity (Section S2)."

Reviewer response to rebuttal: The authors can experiment on side questions for as long as they please, but none of the results described above answer the concern about how center-fitting undercuts the evidentiary value of their main results.

Authors rebuttal: • While it is true that the optimal center needs to be found by fitting to the data, there no particular mystery to the algorithm: we are simply performing a standard gradient-descent to maximize the fit to the data. We have described the algorithm clearly and are making our codes public. We find the algorithm to yield stable optimal centers despite many randomly initialized starting points. We find the optimal center to be able to predict responses to entirely novel images that were excluded during model training. We are making no assumption about the location of centre with respect to individual points. Therefore, we see no cause for concern regarding the center-finding algorithm.

Reviewer response to rebuttal: The point of the original comment was that center-fitting should not be done in the first place because it introduces unknowable effects.

•Authors rebuttal: Most visual tasks, such as finding an animal, are thought to involve building a decision boundary on some underlying neural representation. Even visual search has been portrayed as a signal-detection problem where a particular target is to be discriminated from a distractor. However none of these formulations work in the case of property-based visual tasks, where there is no unique feature to look for.<br /> We are proposing that, when we view a search array, the neural response to the search array can be deduced from the neural responses to the individual elements using well-known rules, and that decisions about an oddball target being present or absent can be made by computing the distance of this neural response from some canonical mean firing rate of a population of neurons. This distance to center computation is what we denote as visual homogeneity. We have revised our manuscript throughout to make this clearer and we hope that this helps you understand the logic better.<br /> • You are absolutely correct that the stimulus complexity should matter, but there are no good empirically derived measures for stimulus complexity, other than subjective ratings which are complex on their own and could be based on any number of other cognitive and semantic factors. But considering what factors are correlated with target-absent response times is entirely different from asking what decision variable or template is being used by participants to solve the task.

Reviewer response to rebuttal: If stimulus complexity is what matters, as the authors agree here, then it is incumbent on them to measure stimulus complexity. The difficulty of measuring stimulus complexity does not justify avoiding the problem with an analysis that ignores complexity.

Authors rebuttal: • We have provided empirical proof for our claims, by showing that target-present response times in a visual search task are correlated with "different" responses in the same-different task, and that target-absent response times in the visual search task are correlated with "same" responses in the same-different task (Section S4).

Reviewer response to rebuttal: Sorry, but there is still no reason to think that same-different judgments are based on a mythical boundary halfway between the two. If there is a boundary, it will be close to the same end of the continuum, where subjects might conceivably miss some tiny difference between two stimuli. The vast majority of "different" stimuli will be entirely different from the same stimulus, producing no confusability, and certainly not a decision boundary halfway between two extremes.

Authors rebuttal: • Again, the opposite correlations between target present/absent search times with VH are the crucial empirical validation of our claims that a distance-to-center calculation explain how we perform these property-based tasks. The VH predictions do not fully explain the data. We have explicitly acknowledged this shortcoming, so we are hardly dismissing it as a problem.

Reviewer response to rebuttal: The authors' acknowledgement of flaws in the ms does not argue in favor of publication, but rather just the opposite.

Authors rebuttal: • Finding an oddball, deciding if two items are same or different and symmetry tasks are disparate visual tasks that do not fit neatly into standard models of decision-making. The key conceptual advance of our study is that we propose a plausible neural representation and decision variable that allows all three property-based visual tasks to be reconciled with standard models of decision-making.

Reviewer response to rebuttal: The original comment stands as written. Same/different will have a boundary very close to the "same" end of the continuum. The boundary is only halfway between two choices if the stimulus design forces the boundary to be there, as in the motion and cat/dog experiments.

Authors rebuttal: "There is no inherent middle point boundary between target present and target absent. Instead, in both types of trial, maximum information is present when target and distractors are most dissimilar, and minimum information is present when target and distractors are most similar. The point of greatest similarity occurs at then limit of any metric for similarity. Correspondingly, there is no middle point dip in information that would produce greater difficulty and higher response times. Instead, task difficulty and response times increase monotonically with similarity between targets and distractors, for both target present and target absent decisions. Thus, in Figs. 2F and 2G, response times appear to be highest for animals, which share the largest numbers of closely similar distractors."<br /> • Your alternative explanation rests on vague factors like "maximum information" which cannot be quantified. By contrast we are proposing a concrete, falsifiable model for three property-based tasks - same/different, oddball present/absent and object symmetry. Any argument based solely on item similarity to explain visual search or symmetry responses cannot explain systematic variations observed for target-absent arrays and for symmetric objects, for the reasons explained earlier.

Reviewer response to rebuttal: There is nothing vague about this comment. The authors use an analysis that assumes a decision boundary at the centerpoint of their arbitrarily defined stimulus space. This assumption is not supported, and it is unlikely, considering that subjects are likely to notice all but the smallest variations between same and different stimuli, putting the boundary nearly at the same end of the continuum, not the very middle.

Authors rebuttal: "(1) The area VH boundaries from different experiments are nearly completely non-overlapping.

In line with their theory that VH is a single continuum with a decision boundary somewhere in the middle, the authors use fMRI searchlight to find an area whose responses positively correlate with homogeneity, as calculated across all of their target present and target absent arrays. They report VH-correlated activity in regions anterior to LO. However, the VH defined by symmetry Experiments 3 and 4 (VHsymmetry) is substantially anterior to LO, while the VH defined by target detection Experiments 1 and 2 (VHdetection) is almost immediately adjacent to LO. Fig. S13 shows that VHsymmetry and VHdetection are nearly non-overlapping. This is a fundamental problem with the claim of discovering a new area that represents a new quantity that explains response times across multiple visual tasks. In addition, it is hard to understand why VHsymmetry does not show up in a straightforward subtraction between symmetric and asymmetric objects, which should show a clear difference in homogeneity."

• We respectfully disagree. The partial overlap between the VH regions identified in Experiments 1 & 2 can hardly be taken as evidence against the quantity VH itself, because there are several other obvious alternate explanations for this partial overlap, as summarized earlier as well. The VH region does show up in a straightforward subtraction between symmetric and asymmetric objects (Section S7), so we are not sure what the Reviewer is referring to here.

Reviewer response to rebuttal: In disagreeing with the comment quoted above, the authors are maintaining that a new functional area of cerebral cortex can be declared even if that area changes location on the cortical map from one experiment to another. That position is patently absurd.

Authors rebuttal: "(3) Definition of the boundaries and purpose of a new visual area in the brain requires circumspection, abundant and convergent evidence, and careful controls.

Even if the VH metric, as defined and calculated by the authors here, is a meaningful quantity, it is a bold claim that a large cortical area just anterior to LO is devoted to calculating this metric as its major task. Vision involves much more than target detection and symmetry detection. Cortex anterior to LO is bound to perform a much wider range of visual functionalities. If the reported correlations can be clarified and supported, it would be more circumspect to treat them as one byproduct of unknown visual processing in cortex anterior to LO, rather than treating them as the defining purpose for a large area of visual cortex."

• We totally agree with you that reporting a new brain region would require careful interpretation and abundant and converging evidence. However, this requires many studies worth of work, and historically category-selective regions like the FFA have achieved consensus only after they were replicated and confirmed across many studies. We believe our proposal for the computation of a quantity like visual homogeneity is conceptually novel, and our study represents a first step that provides some converging evidence (through replicable results across different experiments) for such a region. We have reworked our manuscript to make this point clearer (Discussion, p 32).

Reviewer response to rebuttal: Indeed, declaring a new brain area depends on much more work than is done here. Thus, the appropriate course here is to wait before claiming to have identified a new cortical area.

Reviewer #2 (Public review):

Summary:

This study proposes visual homogeneity as a novel visual property that enables observers perform to several seemingly disparate visual tasks, such as finding an odd item, deciding if two items are same, or judging if an object is symmetric. In Exp 1, the reaction times on several objects were measured in human subjects. In Exp 2, visual homogeneity of each object was calculated based on the reaction time data. The visual homogeneity scores predicted reaction times. This value was also correlated with the BOLD signals in a specific region anterior to LO. Similar methods were used to analyze reaction time and fMRI data in a symmetry detection task. It is concluded that visual homogeneity is an important feature that enables observers to solve these two tasks.

Strengths:

(1) The writing is very clear. The presentation of the study is informative.

(2) This study includes several behavioral and fMRI experiments. I appreciate the scientific rigor of the authors.

Weaknesses:

Before addressing the manuscript itself, I would like to comment the review process first. Having read the lasted revised manuscript, I shared many of the concerns raised by the two reviewers in the last two rounds of review. It appears that the authors have disagreed with the majority of comments made by the two reviewers. If so, I strongly recommend that the authors proceed to make this revision as a Version of Record and conclude this review process. According to eLife's policy that the authors have the right to make a Version of Record at any time during the review process, and I fully respect that right. However, I also ask that the authors respect the reviewer's right to retain the comments regarding this paper.

Beside that, I still have several further questions about this study.

(1) My main concern with this paper is the way visual homogeneity is computed. On page 10, lines 188-192, it says: "we then asked if there is any point in this multidimensional representation such that distances from this point to the target-present and target-absent response vectors can accurately predict the target-present and target-absent response times with a positive and negative correlation respectively (see Methods)". This is also true for the symmetry detection task. If I understand correctly, the reference point in this perceptual space was found by deliberating satisfying the negative and positive correlations in response times. And then on page 10, lines 200-205, it shows that the positive and negative correlations actually exist. This logic is confusing. The positive and negative correlations emerge only because this method is optimized to do so. It seems more reasonable to identify the reference point of this perceptual space independently, without using the reaction time data. Otherwise, the inference process sounds circular. A simple way is to just use the mean point of all objects in Exp 1, without any optimization towards reaction time data.<br /> I raised this question in my initial review. However, the authors did not address whether the positive and negative correlations still hold if the mean point is defined as the reference point without any optimization. The authors also argue that it is similar to a case of fitting a straight line. It is fine that the authors insist on the straight line (e.g., correlation). However, I would not call "straight line correlations" a "quantitative model" as a high-profile journals like eLife. Please remove all related arguments of a novel quantitative model.

(2) Visual homogeneity (at least given the current form) is an unnecessary term. It is similar to distractor heterogeneity/distractor variability/distractor saliency in literature. However, the authors attempt to claim it as a novel concept. Both R1 and me raised this question in the very first review. However, the authors refused to revise the manuscript. In the last review, I mentioned this and provided some example sentences claiming novelty. The authors only revised the last sentence of the abstract, and even did not bother to revise the last sentence of significance: "we show that these tasks can be solved using a simple property WE DEFINE as visual homogeneity". Also, lines 851 still shows "we have defined a NOVEL image property, visual homogeneity...". I am confused about whether the authors agree or disagree that "visual homogeneity is an unnecessary term". If the authors agree, they should completely remove the related phrase throughout the paper. If not, they should keep all these and state the reasons. I don't think this is a correct approach to revising a manuscript.

(3) If the authors agree that visual homogeneity is not new, I suggest a complete rewrite of the title, abstract, significance, and introduction. Let me ask a simple question, can we remove "visual homogeneity" and use some more well-established term like "image feature similarity"? If yes, visual homogeneity is unnecessary.

(4) If I understand it correctly, one of the key findings of this paper is "the response times for target-present searches were positively correlated with visual homogeneity. By contrast, the response times for target-absent searches were negatively correlated with visual homogeneity" (lines 204-207). I think the authors have already acknowledged that this positive correlation is not surprising at all because it reflects the classic target-distractor similarity effect. If this is the case, please completely remove the positive correlation as a novel prediction and finding.

(5) In my last review, I mentioned the seminal paper by Duncan and Humphreys (1989) has clearly stated that "difficulty increases with increased similarity of targets to nontargets and decreased similarity between nontargets" (the sentence in their abstract). Here, "similarity between nontargets" is the same as the visual homogeneity defined here. Similar effects have been shown in Duncan (1989) and Nagy, Neriani, and Young (2005). See also the inconsistent results in Nagy& Thomas, 2003, Vicent, Baddeley, Troscianko&Gilchrist, 2009. More recently, Wei Ji Ma has systematically investigated the effects of heterogeneous distractors in visual search. I think the introduction part of Wei Ji Ma's paper (2020) provides a nice summary of this line of research.

Thanks to the authors' revision, I now better understand the negative correlation. The between-distrator similarity mentioned above describes the heterogeneity of distractors WITHIN an image. However, if I understand it correctly, this study aims to address the negative correlation of reaction time and target-absent stimuli ACROSS images. In other words, why do humans show a shorter reaction time to an image of four pigeons than to an image of four dogs (as shown in Figure 2C), simply because the later image is closer to the reference point of the image space. In this sense, this negative correlation is indeed not the same as distractor heterogeneity. However, this is known as the saliency effect or oddball effects. For example, it seems quite natural to me that humans respond faster to a fish image if the image set contains many images of four-leg dogs that look very different from fish. If this is indeed a saliency effect, why should we define a new term "visual homogeneity"?

(6) The section "key predictions" is quite straightforward. I understand the logic of positive and negative correlations. However, what is the physical meaning of "decision boundary" (Fig. 1G) here? How does the "decision boundary" map on the image space?

(7) In my opinion, one of the advantages of this study is the fMRI dataset, which is valuable because previous studies did not collect fMRI data. The key contribution may be the novel brain region associated with display heterogeneity. If this is the case, I would suggest using a more parametric way to measure this region. For example, one can use Gabor stimuli and systematically manipulate the variations of multiple Gabor stimuli, the same logic also applies to motion direction. If this study uses static Gabor, random dot motion, object images that span from low-level to high-level visual stimuli, and consistently shows that the stimulus heterogeneity is encoded in one brain region, I would say this finding is valuable. But this sounds another experiment. In other words, it is insufficient to claim a new brain region given the current form of the manuscript.

References:

* Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus similarity. Psychological Review, 96(3), 433-458. doi: 10.1037/0033-295x.96.3.433<br /> * Duncan, J. (1989). Boundary conditions on parallel processing in human vision. Perception, 18(4), 457-469. doi: 10.1068/p180457<br /> * Nagy, A. L., Neriani, K. E., & Young, T. L. (2005). Effects of target and distractor heterogeneity on search for a color target. Vision Research, 45(14), 1885-1899. doi: 10.1016/j.visres.2005.01.007<br /> * Nagy, A. L., & Thomas, G. (2003). Distractor heterogeneity, attention, and color in visual search. Vision Research, 43(14), 1541-1552. doi: 10.1016/s0042-6989(03)00234-7<br /> * Vincent, B., Baddeley, R., Troscianko, T., & Gilchrist, I. (2009). Optimal feature integration in visual search. Journal of Vision, 9(5), 15-15. doi: 10.1167/9.5.15<br /> * Singh, A., Mihali, A., Chou, W. C., & Ma, W. J. (2023). A Computational Approach to Search in Visual Working Memory.<br /> * Mihali, A., & Ma, W. J. (2020). The psychophysics of visual search with heterogeneous distractors. BioRxiv, 2020-08.<br /> * Calder-Travis, J., & Ma, W. J. (2020). Explaining the effects of distractor statistics in visual search. Journal of Vision, 20(13), 11-11.

Reviewer #3 (Public review):

Summary of the review process from the Reviewing Editor:

The authors and the reviewers did not agree on several important points made in this paper. The reviewers were critical of the operationalisation of the concept of visual homogeneity (VH), and questioned its validity. For instance, they found it unsatisfying that VH was not calculated on the basis of images themselves, but on the basis of reaction times instead. The authors responded by providing further explanation and argumentation for the importance of this novel concept, but the reviewers were not persuaded. The reviewers also pointed out some data features that did not fit the theory (e.g., overlapping VH between present and absent stimuli), which the authors acknowledge as a point that needs further refining. Finally, the reviewers pointed out that the new so-called visual homogeneity brain region does not overlap very much in the two studies, to which the authors have responded that it is remarkable that there is even partial overlap, given the many confounding differences between the two studies. Altogether, the authors have greatly elaborated their case for VH as an important concept, but the reviewers were not persuaded, and we conclude that the current evidence does not yet meet the high bar for declaring that a novel image property, visual homogeneity, is computed in a localised brain region.

Author response:

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

We are grateful to the editors and reviewers for their careful reading and constructive comments. We have now done our best to respond to them fully through additional analyses and text revisions. In the sections below, the original reviewer comments are in black, and our responses are in red.

To summarize, the major changes in this round of review are as follows:

(1) We have included a new introductory figure (Figure 1) to explain the distinction between feature-based tasks and property-based tasks.

(2) We have included a section on “key predictions” and a section on “overview of this study” in the Introduction to clearly delineate our key predictions and provide a overview of our study.

(3) We have included additional analyses to address the reviewers’ concerns about circularity in Experiments 1 & 2. We show that distance-to-center or visual homogeneity computations performed on object representations obtained from deep networks (instead of the perceptual dissimilarities from Experiment 1) also yields comparable predictions of target-present and target-absent responses in Experiment 2. 

(4) We have extensively reworked the manuscript wherever possible to address the specific concerns raised by the reviewers.

We hope that the revised manuscript adequately addresses the concerns raised in this round of review, and we look forward to a positive assessment.

eLife Assessment

This study uses carefully designed experiments to generate a useful behavioural and neuroimaging dataset on visual cognition. The results provide solid evidence for the involvement of higher-order visual cortex in processing visual oddballs and asymmetry. However, the evidence provided for the very strong claims of homogeneity as a novel concept in vision science, separable from existing concepts such as target saliency, is inadequate.

Thank you for your positive assessment. We agree that visual homogeneity is similar to existing concepts such as target saliency, memorability etc. We have proposed it as a separate concept because visual homogeneity has an independent empirical measure (the reciprocal of target-absent search time in oddball search, or the reciprocal of same response time in a same-different task, etc) that may or may not be the same as other empirical measures such as saliency and memorability. Investigating these possibilities is beyond the scope of our study but would be interesting for future work. We have now clarified this in the revised manuscript (Discussion, p. 42).

However, we’d like to emphasize that the question of whether visual homogeneity is novel or related to existing concepts misses entirely the key contribution of our study.

Our key contribution is a quantitative, falsifiable model for how the brain could be solving property-based tasks like same-different, oddball or symmetry. Most theories of decision making consider feature-based tasks where there is a well-defined feature space and decision variable. Property-based tasks pose a significant challenge to standard theories since it is not clear how these tasks could be solved. In fact, oddball search, same-different and symmetry tasks have been considered so different that they are rarely even mentioned in the same study. Our study represents a unifying framework showing that all three tasks can be understood as solving the same underlying fundamental problem, and presents evidence in favor of this solution.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The authors define a new metric for visual displays, derived from psychophysical response times, called visual homogeneity (VH). They attempt to show that VH is explanatory of response times across multiple visual tasks. They use fMRI to find visual cortex regions with VH-correlated activity. On this basis, they declare a new visual region in human brain, area VH, whose purpose is to represent VH for the purpose of visual search and symmetry tasks.

Thank you for your accurate and positive assessment.

Strengths:

The authors present carefully designed experiments, combining multiple types of visual judgments and multiple types of visual stimuli with concurrent fMRI measurements. This is a rich dataset with many possibilities for analysis and interpretation.

Thank you for your accurate and positive assessment.

Weaknesses:

The datasets presented here should provide a rich basis for analysis. However, in this version of the manuscript, I believe that there are major problems with the logic underlying the authors' new theory of visual homogeneity (VH), with the specific methods they used to calculate VH, and with their interpretation of psychophysical results using these methods. These problems with the coherency of VH as a theoretical construct and metric value make it hard to interpret the fMRI results based on searchlight analysis of neural activity correlated with VH.

We respectfully disagree with your concerns, and have done our best to respond to them fully below.

In addition, the large regions of VH correlations identified in Experiments 1 and 2 vs. Experiments 3 and 4 are barely overlapping. This undermines the claim that VH is a universal quantity, represented in a newly discovered area of visual cortex, that underlies a wide variety of visual tasks and functions.

We respectfully disagree with your assertion. First of all, there is partial overlap between the VH regions, for which there are several other obvious explanations that must be considered first before dismissing VH outright as a flawed construct. We acknowledge these alternatives in the Results (p. 27), and the relevant text is reproduced below.

“We note that it is not straightforward to interpret the overlap between the VH regions identified in Experiments 2 & 4. The lack of overlap could be due to stimulus differences (natural images in Experiment 2 vs silhouettes in Experiment 4), visual field differences (items in the periphery in Experiment 2 vs items at the fovea in Experiment 4) and even due to different participants in the two experiments. There is evidence supporting all these possibilities: stimulus differences (Yue et al., 2014), visual field differences (Kravitz et al., 2013) as well as individual differences can all change the locus of neural activations in object-selective cortex (Weiner and Grill-Spector, 2012a; Glezer and Riesenhuber, 2013). We speculate that testing the same participants on search and symmetry tasks using similar stimuli and display properties would reveal even larger overlap in the VH regions that drive behavior.”

Maybe I have missed something, or there is some flaw in my logic. But, absent that, I think the authors should radically reconsider their theory, analyses, and interpretations, in light of detailed comments below, in order to make the best use of their extensive and valuable datasets combining behavior and fMRI. I think doing so could lead to a much more coherent and convincing paper, albeit possibly supporting less novel conclusions.

We respectfully disagree with your assessment, and we hope that our detailed responses below will convince you of the merit of our claims.

THEORY AND ANALYSIS OF VH

(1) VH is an unnecessary, complex proxy for response time and target-distractor similarity.<br /> VH is defined as a novel visual quality, calculable for both arrays of objects (as studied in Experiments 1-3) and individual objects (as studied in Experiment 4). It is derived from a center-to-distance calculation in a perceptual space. That space in turn is derived from multi-dimensional scaling of response times for target-distractor pairs in an oddball detection task (Experiments 1 and 2) or in a same different task (Experiments 3 and 4).  Proximity of objects in the space is inversely proportional to response times for arrays in which they were paired. These response times are higher for more similar objects. Hence, proximity is proportional to similarity. This is visible in Fig. 2B as the close clustering of complex, confusable animal shapes.

VH, i.e. distance-to-center, for target-present arrays is calculated as shown in Fig. 1C, based on a point on the line connecting target and distractors. The authors justify this idea with previous findings that responses to multiple stimuli are an average of responses to the constituent individual stimuli. The distance of the connecting line to the center is inversely proportional to the distance between the two stimuli in the pair, as shown in Fig. 2D. As a result, VH is inversely proportional to distance between the stimuli and thus to stimulus similarity and response times. But this just makes VH a highly derived, unnecessarily complex proxy for target-distractor similarity and response time. The original response times on which the perceptual space is based are far more simple and direct measures of similarity for predicting response times.

Thank you for carefully thinking through our logic. We agree that a distance-to-centre calculation is entirely unnecessary as an explanation for target-present visual search. The difficulty of target-present search is already known to be directly proportional to the similarity between target and distractor, so there is nothing new to explain here.

However, this is a narrow and selective interpretation of our findings because you are focusing only on our results on target-present searches, which are only half of all our data. The other half is the target-absent responses which previously have had no clear explanation. You are also missing the fact that we are explaining same-different and symmetry tasks as well using the same visual homogeneity computation.

We urge you to think more deeply about the problem of how to decide whether an oddball is present or not in the first place. How do we actually solve this task? There must be some underlying representation and decision process. Our study shows that a distance-to-centre computation can actually serve as a decision variable to solve disparate property-based visual tasks. These tasks pose a major challenge to standard models of decision making, because the underlying representation and decision variable have been unclear. Our study resolves this challenge by proposing a novel computation that can be used by the brain to solve all these disparate tasks, and bring these tasks into the ambit of standard theories of decision making.  

Our results also explain several interesting puzzles in the literature. If oddball search was driven only by target-distractor similarity, the time taken to respond when a target is absent should not vary at all, and should actually take longer than all target-present searches. But in fact, systematic variations in target-absent times have been observed always in the literature, but have never been explained using any theoretical models. Our results explain why target-absent times vary systematically – it is due to visual homogeneity.

Similarly, in same-different tasks, participants are known to take longer to make a “different” response when the two items differ only slightly. By this logic, they should take the longest to make a “same” response, but in fact, paradoxically, participants are actually faster to make “same” responses. This fast-same effect has been noted several times, but never explained using any models. Our results provide an explanation of why “same” responses to an image vary systematically – it is due to visual homogeneity. 

Finally, in symmetry tasks, symmetric objects evoke fast responses, and this has always been taken as evidence for special symmetry computations in the brain. But we show that the same distance-to-center computation can explain both responses to symmetric and asymmetric objects. Thus there is no need for a special symmetry computation in the brain.

(2) The use of VH derived from Experiment 1 to predict response times in Experiment 2 is circular and does not validate the VH theory.<br /> The use of VH, a response time proxy, to predict response times in other, similar tasks, using the same stimuli, is circular. In effect, response times are being used to predict response times across two similar experiments using the same stimuli. Experiment 1 and the target present condition of Experiment 2 involve the same essential task of oddball detection. The results of Experiment 1 are converted into VH values as described above, and these are used to predict response times in experiment 2 (Fig. 2F). Since VH is a derived proxy for response values in Experiment 1, this prediction is circular, and the observed correlation shows only consistency between two oddball detection tasks in two experiments using the same stimuli.

You are indeed correct in noting that both Experiment 1 & 2 involve oddball search, and so at the superficial level, it looks circular that the oddball search data of Experiment 1 is being used to explain the oddball search data of Experiment 2.

However a deeper scrutiny reveals more fundamental differences: Experiment 1 consisted of only oddball search with the target appearing on the left or right, whereas Experiment 2 consisted of oddball search with the target either present or completely absent. In fact, we were merely using the search dissimilarities from Experiment 1 to reconstruct the underlying object representation, because it is well known that neural dissimilarities are predicted well by search dissimilarities (Sripati & Olson, 2009; Zhivago et al, 2014).

To thoroughly refute any lingering concern about circularity, we reasoned that the model predictions for Experiment 2 could have been obtained by a distance-to-center computation on any brain like object representation. To this end, we used object representations from deep neural networks pretrained on object categorization, whose representations are known to match well with the brain, and asked if a distance-to-centre computation on these representations could predict the search data in Experiment 2. This was indeed the case, and these results are now included an additional section in Supplementary Material (Section S1).

(3) The negative correlation of target-absent response times with VH as it is defined for target-absent arrays, based on distance of a single stimulus from center, is uninterpretable without understanding the effects of center-fitting. Most likely, center-fitting and the different VH metric for target-absent trials produce an inverse correlation of VH with target-distractor similarity.

Unfortunately, as we have mentioned above, target-distractor similarity cannot explain how target-absent searches behave, since there is no distractor in such searches.

We do understand your broader concern about the center-fitting algorithm itself. We performed a number of additional analyses to confirm the generality of our results and reject alternate explanations – these are summarized in a new section titled “Confirming the generality of visual homogeneity” (p. 12), and the section is reproduced below for your convenience.   

“Confirming the generality of visual homogeneity

We performed several additional analyses to confirm the generality of our results, and to reject alternate explanations.

First, it could be argued that our results are circular because they involve taking oddball search times from Experiment 1 and using them to explain search response times in Experiment 2. This is a superficial concern since we are using the search dissimilarities from Experiment 1 only as a proxy for the underlying neural representation, based on previous reports that neural dissimilarities closely match oddball search dissimilarities (Sripati and Olson, 2010; Zhivago and Arun, 2014). Nonetheless, to thoroughly refute this possibility, we reasoned that we would get similar predictions of the target present/absent responses in Experiment using any other brain-like object representation. To confirm this, we replaced the object representations derived from Experiment 1 with object representations derived from deep neural networks pretrained for object categorization, and asked if distance-to-center computations could predict the target present/absent responses in Experiment 2. This was indeed the case (Section S1). 

Second, we wondered whether the nonlinear optimization process of finding the best-fitting center could be yielding disparate optimal centres each time. To investigate this, we repeated the optimization procedure with many randomly initialized starting points, and obtained the same best-fitting center each time (see Methods).

Third, to confirm that the above model fits are not due to overfitting, we performed a leave-one-out cross validation analysis. We left out all target-present and target-absent searches involving a particular image, and then predicted these searches by calculating visual homogeneity estimated from all other images. This too yielded similar positive and negative correlations (r = 0.63, p < 0.0001 for target-present, r = -0.63, p < 0.001  for target-absent).

Fourth, if heterogeneous displays indeed elicit similar neural responses due to mixing, then their average distance to other objects must be related to their visual homogeneity. We confirmed that this was indeed the case, suggesting that the average distance of an object from all other objects in visual search can predict visual homogeneity (Section S1).

Fifth, the above results are based on taking the neural response to oddball arrays to be the average of the target and distractor responses. To confirm that averaging was indeed the optimal choice, we repeated the above analysis by assuming a range of relative weights between the target and distractor. The best correlation was obtained for almost equal weights in the lateral occipital (LO) region, consistent with averaging and its role in the underlying perceptual representation (Section S1).

Finally, we performed several additional experiments on a larger set of natural objects as well as on silhouette shapes. In all cases, present/absent responses were explained using visual homogeneity (Section S2).”

The construction of the VH perceptual space also involves fitting a "center" point such that distances to center predict response times as closely as possible. The effect of this fitting process on distance-to-center values for individual objects or clusters of objects is unknowable from what is presented here. These effects would depend on the residual errors after fitting response times with the connecting line distances. The center point location and its effects on distance-to-center of single objects and object clusters are not discussed or reported here.

While it is true that the optimal center needs to be found by fitting to the data, there no particular mystery to the algorithm: we are simply performing a standard gradient-descent to maximize the fit to the data. We have described the algorithm clearly and are making our codes public. We find the algorithm to yield stable optimal centers despite many randomly initialized starting points. We find the optimal center to be able to predict responses to entirely novel images that were excluded during model training. We are making no assumption about the location of centre with respect to individual points. Therefore, we see no cause for concern regarding the center-finding algorithm. 

Yet, this uninterpretable distance-to-center of single objects is chosen as the metric for VH of target-absent displays (VHabsent). This is justified by the idea that arrays of a single stimulus will produce an average response equal to one stimulus of the same kind. But it is not logically clear why response strength to a stimulus should be a metric for homogeneity of arrays constructed from that stimulus, or even what homogeneity could mean for a single stimulus from this set. And it is not clear how this VHabsent metric based on single stimuli can be equated to the connecting line VH metric for stimulus pairs, i.e. VHpresent, or how both could be plotted on a single continuum.

Most visual tasks, such as finding an animal, are thought to involve building a decision boundary on some underlying neural representation. Even visual search has been portrayed as a signal-detection problem where a particular target is to be discriminated from a distractor. However none of these formulations work in the case of property-based visual tasks, where there is no unique feature to look for.

We are proposing that, when we view a search array, the neural response to the search array can be deduced from the neural responses to the individual elements using well known rules, and that decisions about an oddball target being present or absent can be made by computing the distance of this neural response from some canonical mean firing rate of a population of neurons. This distance to center computation is what we denote as visual homogeneity. We have revised our manuscript throughout to make this clearer and we hope that this helps you understand the logic better. 

It is clear, however, what *should* be correlated with difficulty and response time in the target-absent trials, and that is the complexity of the stimuli and the numerosity of similar distractors in the overall stimulus set. Complexity of the target, similarity with potential distractors, and number of such similar distractors all make ruling out distractor presence more difficult. The correlation seen in Fig. 2G must reflect these kinds of effects, with higher response times for complex animal shapes with lots of similar distractors and lower response times for simpler round shapes with fewer similar distractors.

You are absolutely correct that the stimulus complexity should matter, but there are no good empirically derived measures for stimulus complexity, other than subjective ratings which are complex on their own and could be based on any number of other cognitive and semantic factors. But considering what factors are correlated with target-absent response times is entirely different from asking what decision variable or template is being used by participants to solve the task.

The example points in Fig. 2G seem to bear this out, with higher response times for the deer stimulus (complex, many close distractors in the Fig. 2B perceptual space) and lower response times for the coffee cup (simple, few close distractors in the perceptual space). While the meaning of the VH scale in Fig. 2G, and its relationship to the scale in Fig. 2F, are unknown, it seems like the Fig. 2G scale has an inverse relationship to stimulus complexity, in contrast to the expected positive relationship for Fig. 2F. This is presumably what creates the observed negative correlation in Fig. 2G.

Taken together, points 1-3 suggest that VHpresent and VHabsent are complex, unnecessary, and disconnected metrics for understanding target detection response times. The standard, simple explanation should stand. Task difficulty and response time in target detection tasks, in both present and absent trials, are positively correlated with target-distractor similarity.

We strongly disagree. Your assessment seems to be based on only considering target-present searches, which are of course driven by target-distractor similarity. Your  argument is flawed because systematic variations in target-absent trials cannot be linked to any target-distractor similarity since there are no targets in the first place in such trials.

We have shown that target-absent response times are in fact, independent of experimental context, which means that they index an image property that is independent of any reference target (Results, p. 15; Section S4). This property is what we define as visual homogeneity.

I think my interpretations apply to Experiments 3 and 4 as well, although I find the analysis in Fig. 4 especially hard to understand. The VH space in this case is based on Experiment 3 oddball detection in a stimulus set that included both symmetric and asymmetric objects. But the response times for a very different task in Experiment 4, a symmetric/asymmetric judgment, are plotted against the axes derived from Experiment 3 (Fig. 4F and 4G). It is not clear to me why a measure based on oddball detection that requires no use of symmetry information should be predictive of within-stimulus symmetry detection response times. If it is, that requires a theoretical explanation not provided here.

We were simply using an oddball detection task to construct the underlying object representation, on the basis of observations that search dissimilarities are strongly correlated with neural   dissimilarities. In Section S1, we show that similar results could have been obtained using other object representations such as deep networks, as long as the representation is brain-like.

(4) Contrary to the VH theory, same/different tasks are unlikely to depend on a decision boundary in the middle of a similarity or homogeneity continuum.

We have provided empirical proof for our claims, by showing that target-present response times in a visual search task are correlated with “different” responses in the same-different task, and that target-absent response times in the visual search task are correlated with “same” responses in the same-different task (Section S4).

The authors interpret the inverse relationship of response times with VHpresent and VHabsent, described above, as evidence for their theory. They hypothesize, in Fig. 1G, that VHpresent and VHabsent occupy a single scale, with maximum VHpresent falling at the same point as minimum VHabsent. This is not borne out by their analysis, since the VHpresent and VHabsent value scales are mainly overlapping, not only in Experiments 1 and 2 but also in Experiments 3 and 4. The authors dismiss this problem by saying that their analyses are a first pass that will require future refinement. Instead, the failure to conform to this basic part of the theory should be a red flag calling for revision of the theory.

Again, the opposite correlations between target present/absent search times with VH are the crucial empirical validation of our claims that a distance-to-center calculation explain how we perform these property-based tasks. The VH predictions do not fully explain the data. We have explicitly acknowledged this shortcoming, so we are hardly dismissing it as a problem. 

The reason for this single scale is that the authors think of target detection as a boundary decision task, along a single scale, with a decision boundary somewhere in the middle, separating present and absent. This model makes sense for decision dimensions or spaces where there are two categories (right/left motion; cats vs. dogs), separated by an inherent boundary (equal left/right motion; training-defined cat/dog boundary). In these cases, there is less information near the boundary, leading to reduced speed/accuracy and producing a pattern like that shown in Fig. 1G.

Finding an oddball, deciding if two items are same or different and symmetry tasks are disparate visual tasks that do not fit neatly into standard models of decision making. The key conceptual advance of our study is that we propose a plausible neural representation and decision variable that allow all three property-based visual tasks to be reconciled with standard models of decision making.

This logic does not hold for target detection tasks. There is no inherent middle point boundary between target present and target absent. Instead, in both types of trial, maximum information is present when target and distractors are most dissimilar, and minimum information is present when target and distractors are most similar. The point of greatest similarity occurs at then limit of any metric for similarity. Correspondingly, there is no middle point dip in information that would produce greater difficulty and higher response times. Instead, task difficulty and response times increase monotonically with similarity between targets and distractors, for both target present and target absent decisions. Thus, in Figs. 2F and 2G, response times appear to be highest for animals, which share the largest numbers of closely similar distractors.        

Your alternative explanation rests on vague factors like “maximum information” which cannot be quantified. By contrast we are proposing a concrete, falsifiable model for three property-based tasks – same/different, oddball present/absent and object symmetry. Any argument based solely on item similarity to explain visual search or symmetry responses cannot explain systematic variations observed for target-absent arrays and for symmetric objects, for the reasons explained earlier.

DEFINITION OF AREA VH USING fMRI

(1) The area VH boundaries from different experiments are nearly completely non-overlapping.

In line with their theory that VH is a single continuum with a decision boundary somewhere in the middle, the authors use fMRI searchlight to find an area whose responses positively correlate with homogeneity, as calculated across all of their target present and target absent arrays. They report VH-correlated activity in regions anterior to LO. However, the VH defined by symmetry Experiments 3 and 4 (VHsymmetry) is substantially anterior to LO, while the VH defined by target detection Experiments 1 and 2 (VHdetection) is almost immediately adjacent to LO. Fig. S13 shows that VHsymmetry and VHdetection are nearly non-overlapping. This is a fundamental problem with the claim of discovering a new area that represents a new quantity that explains response times across multiple visual tasks. In addition, it is hard to understand why VHsymmetry does not show up in a straightforward subtraction between symmetric and asymmetric objects, which should show a clear difference in homogeneity.

We respectfully disagree. The partial overlap between the VH regions identified in Experiments 1 & 2 can hardly be taken as evidence against the quantity VH itself, because there are several other obvious alternate explanations for this partial overlap, as summarized earlier as well. The VH region does show up in a straightforward subtraction  between symmetric and asymmetric objects (Section S7), so we are not sure what the Reviewer is referring to here.

(2) It is hard to understand how neural responses can be correlated with both VHpresent and VHabsent.

The main paper results for VHdetection are based on both target-present and target-absent trials, considered together. It is hard to interpret the observed correlations, since the VHpresent and VHabsent metrics are calculated in such different ways and have opposite correlations with target similarity, task difficulty, and response times (see above). It may be that one or the other dominates the observed correlations. It would be clarifying to analyze correlations for target-present and target-absent trials separately, to see if they are both positive and correlated with each other.

Thanks for raising this point. We have now confirmed that the positive correlation between VH and neural response holds even when we do the analysis separately for target-present and -absent searches (correlation between neural response in VH region and visual homogeneity (n = 32, r = 0.66, p < 0.0005 for target-present searches & n = 32, r = 0.56, p < 0.005 for target-absent searches).

(3) Definition of the boundaries and purpose of a new visual area in the brain requires circumspection, abundant and convergent evidence, and careful controls.

Even if the VH metric, as defined and calculated by the authors here, is a meaningful quantity, it is a bold claim that a large cortical area just anterior to LO is devoted to calculating this metric as its major task. Vision involves much more than target detection and symmetry detection. Cortex anterior to LO is bound to perform a much wider range of visual functionalities. If the reported correlations can be clarified and supported, it would be more circumspect to treat them as one byproduct of unknown visual processing in cortex anterior to LO, rather than treating them as the defining purpose for a large area of visual cortex.

We totally agree with you that reporting a new brain region would require careful interpretation and abundant and converging evidence. However, this requires many studies worth of work, and historically category-selective regions like the FFA have achieved consensus only after they were replicated and confirmed across many studies. We believe our proposal for the computation of a quantity like visual homogeneity is conceptually novel, and our study represents a first step that provides some converging evidence (through replicable results across different experiments) for such a region. We have reworked our manuscript to make this point clearer (Discussion, p 32).

Reviewer #3 (Public Review):

Summary:

This study proposes visual homogeneity as a novel visual property that enables observers perform to several seemingly disparate visual tasks, such as finding an odd item, deciding if two items are same, or judging if an object is symmetric. In Exp 1, the reaction times on several objects were measured in human subjects. In Exp 2, visual homogeneity of each object was calculated based on the reaction time data. The visual homogeneity scores predicted reaction times. This value was also correlated with the BOLD signals in a specific region anterior to LO. Similar methods were used to analyze reaction time and fMRI data in a symmetry detection task. It is concluded that visual homogeneity is an important feature that enables observers to solve these two tasks.

Thank you for your accurate and positive assessment.

Strengths:

(1) The writing is very clear. The presentation of the study is informative.

(2) This study includes several behavioral and fMRI experiments. I appreciate the scientific rigor of the authors.

We are grateful to you for your balanced assessment and constructive comments.

Weaknesses:

(1) My main concern with this paper is the way visual homogeneity is computed. On page 10, lines 188-192, it says: "we then asked if there is any point in this multidimensional representation such that distances from this point to the target-present and target-absent response vectors can accurately predict the target-present and target-absent response times with a positive and negative correlation respectively (see Methods)". This is also true for the symmetry detection task. If I understand correctly, the reference point in this perceptual space was found by deliberating satisfying the negative and positive correlations in response times. And then on page 10, lines 200-205, it shows that the positive and negative correlations actually exist. This logic is confusing. The positive and negative correlations emerge only because this method is optimized to do so. It seems more reasonable to identify the reference point of this perceptual space independently, without using the reaction time data. Otherwise, the inference process sounds circular. A simple way is to just use the mean point of all objects in Exp 1, without any optimization towards reaction time data.

We disagree with you since the same logic applies to any curve-fitting procedure. When we fit data to a straight line, we are finding the slope and intercept that minimizes the error between the data and the straight line, but we would hardly consider the process circular when a good fit is achieved – in fact we take it as a confirmation that the data can be fit linearly. In the same vein, we would not have observed a good fit to the data, if there did not exist any good reference point relative to which the distances of the target-present and target-absent search arrays predicted these response times.

In Section S2, we show that the visual homogeneity estimates for each object is strongly correlated with the average distance of each object to all other objects (r = 0.84, p<0.0005, Figure S1).

We have performed several additional analyses to confirm the generality of our results and to reject alternate explanations (see Results, p. 12, Section titled “Confirming the generality of visual homogeneity”). In particular, to confirm that the results we obtained are not due to overfitting, we performed a cross-validation analysis, where we removed all searches involving a particular image and predicted these response times using visual homogeneity. This too revealed a significant model correlation confirming that our results are not due to overfitting.

(2) Visual homogeneity (at least given the current from) is an unnecessary term. It is similar to distractor heterogeneity/distractor variability/distractor statics in literature. However, the authors attempt to claim it as a novel concept. The title is "visual homogeneity computations in the brain enable solving generic visual tasks". The last sentence of the abstract is "a NOVEL IMAGE PROPERTY, visual homogeneity, is encoded in a localized brain region, to solve generic visual tasks". In the significance, it is mentioned that "we show that these tasks can be solved using a simple property WE DEFINE as visual homogeneity". If the authors agree that visual homogeneity is not new, I suggest a complete rewrite of the title, abstract, significance, and introduction.

We respectfully disagree that visual homogeneity is an unnecessary term. Please see our comments to Reviewer 1 above. Just like saliency and memorability can be measured empirically, we propose that visual homogeneity can be empirically measured as the reciprocal of the target-absent search time in a search task, or as the reciprocal of the “same” response time in a same-different task. Understanding how these three quantities interact will require measuring them empirically for an identical set of images, which is beyond the scope of this study but an interesting possibility for future work.

(3) Also, "solving generic tasks" is another overstatement. The oddball search tasks, same-different tasks, and symmetric tasks are only a small subset of many visual tasks. Can this "quantitative model" solve motion direction judgment tasks, visual working memory tasks? Perhaps so, but at least this manuscript provides no such evidence. On line 291, it says "we have proposed that visual homogeneity can be used to solve any task that requires discriminating between homogeneous and heterogeneous displays". I think this is a good statement. A title that says "XXXX enable solving discrimination tasks with multi-component displays" is more acceptable. The phrase "generic tasks" is certainly an exaggeration.

Thank you for your suggestion. We have now replaced the term “generic tasks” with the term property-based tasks, which we feel is more appropriate and reflect the fact that oddball search, same-different and symmetry tasks all involve looking for a specific image property.

(4) If I understand it correctly, one of the key findings of this paper is "the response times for target-present searches were positively correlated with visual homogeneity. By contrast, the response times for target-absent searches were negatively correlated with visual homogeneity" (lines 204-207). I think the authors have already acknowledged that the positive correlation is not surprising at all because it reflects the classic target-distractor similarity effect. But the authors claim that the negative correlations in target-absent searches is the true novel finding.

(5) I would like to make it clear that this negative correlation is not new either. The seminal paper by Duncan and Humphreys (1989) has clearly stated that "difficulty increases with increased similarity of targets to nontargets and decreased similarity between nontargets" (the sentence in their abstract). Here, "similarity between nontargets" is the same as the visual homogeneity defined here. Similar effects have been shown in Duncan (1989) and Nagy, Neriani, and Young (2005). See also the inconsistent results in Nagy & Thomas, 2003, Vicent, Baddeley, Troscianko & Gilchrist, 2009. More recently, Wei Ji Ma has systematically investigated the effects of heterogeneous distractors in visual search. I think the introduction part of Wei Ji Ma's paper (2020) provides a nice summary of this line of research. I am surprised that these references are not mentioned at all in this manuscript (except Duncan and Humphreys, 1989).

You are right in noting that Duncan and Humphreys (1989) propose that searches are more difficult when nontargets are dissimilar. However, since our searches have identical distractors, the similarity between nontargets is always constant across target-absent searches, and therefore this cannot predict any systematic variation in target-absent search that is observed in our data. By contrast, our results explain both target-absent searches and target-present searches.

Thank you for pointing us to previous work. These studies show that it is not just the average distractor similarity but the statistics of the distractor similarity that drive visual search. However these studies do not explain why target-absent searches should vary systematically. 

(6) If the key contribution is the quantitative model, the study should be organized in a different way. Although the findings of positive and negative correlations are not novel, it is still good to propose new models to explain classic phenomena. I would like to mention the three studies by Wei Ji Ma (see below). In these studies, Bayesian observer models were established to account for trial-by-trial behavioral responses. These computational models can also account for the set-size effect, behavior in both localization and detection tasks. I see much more scientific rigor in their studies. Going back to the quantitative model in this paper, I am wondering whether the model can provide any qualitative prediction beyond the positive and negative correlations? Can the model make qualitative predictions that differ from those of Wei Ji's model? If not, can the authors show that the model can quantitatively better account for the data than existing Bayesian models? We should evaluate a model either qualitatively or quantitatively.

Thank you for pointing us to prior work by Wei Ji Ma. These studies systematically examined visual search for a target among heterogeneous distractors using simple parametric stimuli and a Bayesian modeling framework. By contrast, our experiments involve searching for single oddball targets among multiple identical distractors, so it is not clear to us that the Wei Ji Ma models can be easily used to generate predictions about these searches used in our study. 

We are not sure what you mean by offering quantitative predictions beyond positive and negative correlations. We have tried to explain systematic variation in target-present and target-absent response times using a model of how these decisions are being made. Our model explains a lot of systematic variation in the data for both types of decisions.

(7) In my opinion, one of the advantages of this study is the fMRI dataset, which is valuable because previous studies did not collect fMRI data. The key contribution may be the novel brain region associated with display heterogeneity. If this is the case, I would suggest using a more parametric way to measure this region. For example, one can use Gabor stimuli and systematically manipulate the variations of multiple Gabor stimuli, the same logic also applies to motion direction. If this study uses static Gabor, random dot motion, object images that span from low-level to high-level visual stimuli, and consistently shows that the stimulus heterogeneity is encoded in one brain region, I would say this finding is valuable. But this sounds like another experiment. In other words, it is insufficient to claim a new brain region given the current form of the manuscript.

We agree that parametric stimulus manipulations are important for studying early visual areas where stimulus dimensions are known (e.g. orientation, spatial frequency). Using parametric stimulus manipulations for more complex stimuli is fraught with issues because the underlying representation may not be encoding the dimensions being manipulated. This is the reason why we attempted to recover the underlying neural representation using dissimilarities measured using visual search, and then asked whether a decision making process operating on this underlying representation can explain how decisions are made. Therefore we disagree that parametric stimulus manipulations are the only way to obtain insight into such tasks.

We have proposed a quantitative model that explains how decisions about target present and absent can be made through distance-to-center computations on an underlying object representation. We feel that the behavioural and the brain imaging results strongly point to a novel computation that is being performed in a localized region in the brain. These results represent an important first step in understanding how complex, property-based tasks are performed by the brain. We have revised our manuscript to make this point clearer.

REFERENCES

- Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus similarity. Psychological Review, 96(3), 433-458. doi: 10.1037/0033-295x.96.3.433

- Duncan, J. (1989). Boundary conditions on parallel processing in human vision. Perception, 18(4), 457-469. doi: 10.1068/p180457

- Nagy, A. L., Neriani, K. E., & Young, T. L. (2005). Effects of target and distractor heterogeneity on search for a color target. Vision Research, 45(14), 1885-1899. doi: 10.1016/j.visres.2005.01.007

- Nagy, A. L., & Thomas, G. (2003). Distractor heterogeneity, attention, and color in visual search. Vision Research, 43(14), 1541-1552. doi: 10.1016/s0042-6989(03)00234-7

- Vincent, B., Baddeley, R., Troscianko, T., & Gilchrist, I. (2009). Optimal feature integration in visual search. Journal of Vision, 9(5), 15-15. doi: 10.1167/9.5.15

- Singh, A., Mihali, A., Chou, W. C., & Ma, W. J. (2023). A Computational Approach to Search in Visual Working Memory.

- Mihali, A., & Ma, W. J. (2020). The psychophysics of visual search with heterogeneous distractors. BioRxiv, 2020-08.

- Calder-Travis, J., & Ma, W. J. (2020). Explaining the effects of distractor statistics in visual search. Journal of Vision, 20(13), 11-11.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

The authors have not made substantive changes to address my major concerns. Instead, they have responded with arguments about why their original manuscript was good as written. I did not find these arguments persuasive. Given that, I've left my public review the same, since it still represents my opinions about the paper. Readers can judge which viewpoints are more persuasive.

We respectfully disagree: we have tried our best to address your concerns with additional analysis wherever feasible, and by acknowledging any limitations.

Reviewer #3 (Recommendations For The Authors):

(1) As I mentioned above, please consider rewriting title, abstract, introduction, and significance. Please remove the word "visual homogeneity" and instead use distractor heterogeneity/distractor variability/distractor statistics as often used in literature.

To clarify, visual homogeneity is NOT the same as distractor homogeneity. Visual homogeneity refers to a distance-to-center computation and represents an image-computable property that can vary systematically even when all distractors are identical. By contrast distractor heterogeneity varies only when distractors are different from each other.

(2) Better to remove the phrase "generic tasks".

Thanks for your suggestions. We now refer to these tasks as property-based tasks. 

(3) Better to explicitly specify the predictions made by the quantitative model beyond positive and negative correlations.

The predictions of the quantitative model are to explain systematic variation in the response times. We are not sure what else is there to predict in the response times.

(4) If the quantitative model is the key contribution, better to highlight the details and algorithmic contribution of the model, and show the advantage of this model either qualitatively and quantitatively.

Please see our responses above. Our quantitative model explains behavior and brain imaging data on three disparate tasks – the same/different, oddball visual search and symmetry tasks. 

(5) If the new brain region is the key contribution, better to downplay the quantitative model.

Please see our responses above. Our quantitative model explains behavior and brain imaging data on three disparate tasks – the same/different, oddball visual search and symmetry tasks.

eLife Assessment

This important study enhances our understanding of ephaptic interactions by utilizing earthworm recordings to refine a general model and use it to predict ephaptic influences across various synaptic configurations. The integration of experimental evidence, a robust mathematical framework and computer simulations convincingly demonstrates the effects of action potential propagation and collision properties on nearby membranes. The study will interest both computational neuroscientists and physiologists.

Reviewer #1 (Public review):

The authors explain that an action potential that reach an axon terminal emits a small electrical field as it "annihilates". This happens even though there is no gap junction, at chemical synapses. The generated electrical field is simulated to show that it can affect a nearby, disconnected target membrane by tens of microvolts for tenths of a microsecond. Longer effects are simulated for target locations a few microns away.

To simulate action potentials (APs), the paper does not use the standard Hodgkin-Huxley formalism because it fails to explain AP collision. Instead it uses the Tasaki and Matsumoto (TM) model which is simplified to only models APs with three parameters and as a membrane transition between two states of resting versus excited. The authors expand the strictly binary, discrete TM method to a Relaxing Tasaki Model (RTM) that models the relaxation of the membrane potential after an AP. They find that the membrane leak can be neglected in determining AP propagation and that the capacitive currents dominate the process.

The strength of the work is that authors identified an important interaction between neurons that is neglected by the standard models. A weakness of the proposed approach is the assumptions that it makes. For instance, the external medium is modeled as a homogeneous conductive medium, which may be further explored to properly account for biological processes. To the authors' credit, the external medium can be largely varying and could be left out from the general model, only to be modeled specific instances.

The authors provide convincing evidence by performing experiments to record action potential propagation and collision properties and then developing a theoretical framework to simulate effect of their annihilation on nearby membranes. They provide both experimental evidence and rigorous mathematical and computer simulation findings to support their claims. The work has a potential of explaining significant electrical interaction between nerve centers that are connected via a large number of parallel fibers.

Comments on revisions:

The authors responded to all of my previous concerns and significantly improved the manuscript.

Reviewer #2 (Public review):

In this study, the authors measured extracellular electrical features of colliding APs travelling in different directions down an isolated earthworm axon. They then used these features to build a model of the potential ephaptic effects of AP annihilation, i.e. the electrical signals produced by colliding/annihilating APs that may influence neighbouring tissue. The model was then applied to some different hypothetical scenarios involving synaptic connections. In a revised version of the manuscript, it was also applied, with success, to published experimental data on the cerebellar basket cell-to-Purkinje cell pinceau connection. The conclusion is that an annihilating AP at a presynaptic terminal can emphatically influence the voltage of a postsynaptic cell (this is, presumably, the 'electrical coupling between neurons' of the title), and that the nature of this influence depends on the physical configuration of the synapse.

As an experimental neuroscientist who has never used computational approaches, I am unable to comment on the rigour of the analytical approaches that form the bulk of this paper. The experimental approaches appear very well carried out, and the data showing equal conduction velocity of anti- and orthodromically propagating APs in every preparation is now convincing.

The conclusions drawn from the synaptic modelling have been considerably strengthened by the new Figure 5. Here, the authors' model - including AP annihilation at a synaptic terminal - is used to predict the amplitude and direction of experimentally observed effects at the cerebellar basket cell-to-Purkinje cell synapse (Blot & Barbour 2014). One particular form of the model (RTM with tau=0.5ms and realistic non-excitability of the terminal) matches the experimental data extremely well. This is a much more convincing demonstration that the authors' model of ephaptic effects can quantitatively explain key features of experimental data pertaining to synaptic function. As such, the implications for the relevance of ephaptic coupling at different synaptic contacts may be widespread and important.

However, it appears that all of the models in the new Fig5 involve annihilating APs, yet only one fits the data closely. A key question, which should be addressed if at all possible, is what happens to the predictive power of the best-fitting model in Fig5 if the annihilation, and only the annihilation, is removed? In other words, can the authors show that it is specifically the ephaptic effects of AP annihilation, rather than other ephaptic effects of, say AP waveform/amplitude/propagation, that explain the synaptic effects measured in Blot & Barbour (2014)? This would appear to be a necessary demonstration to fully support the claims of the title.

Author response:

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

Public Reviews:

Reviewer 1 (Public Review):

The authors explain that an action potential that reaches an axon terminal emits a small electrical field as it ”annihilates”. This happens even though there is no gap junction, at chemical synapses. The generated electrical field is simulated to show that it can affect a nearby, disconnected target membrane by tens of microvolts for tenths of a microsecond. Longer effects are simulated for target locations a few microns away.

To simulate action potentials (APs), the paper does not use the standard Hodgkin-Huxley formalism because it fails to explain AP collision. Instead, it uses the Tasaki and Matsumoto (TM) model which is simplified to only model APs with three parameters and as a membrane transition between two states of resting versus excited. The authors expand the strictly binary, discrete TM method to a Relaxing Tasaki Model (RTM) that models the relaxation of the membrane potential after an AP. They find that the membrane leak can be neglected in determining AP propagation and that the capacitive currents dominate the process.

The strength of the work is that the authors identified an important interaction between neurons that is neglected by the standard models. A weakness of the proposed approach is the assumptions that it makes. For instance, the external medium is modeled as a homogeneous conductive medium, which may be further explored to properly account for biological processes.

The authors provide convincing evidence by performing experiments to record action potential propagation and collision properties and then developing a theoretical framework to simulate the effect of their annihilation on nearby membranes. They provide both experimental evidence and rigorous mathematical and computer simulation findings to support their claims. The work has the potential of explaining significant electrical interaction between nerve centers that are connected via a large number of parallel fibers.

We thank the reviewer for the distinct analysis of our work and the assessment that we ’identified an important interaction between neurons that is neglected by standard models’.

Indeed, we modeled the external (extracellular) medium as homogeneous conductive medium and, compared to real biological systems, this is a simplification. Our intention is to keep our formal model as general as possible, however, it can be extended to account for specific properties. Accessory structures at axon terminals (such as the pinceau at Purkinje cells) most likely evolved to shape ephaptic coupling. In addition, the extracellular medium is neither homogeneous nor isotropic, and to fully mimic a particular neural connection this has to be implemented in a model as well. We agree and look forward to see how specific modification of the external medium in biological systems will affect ephaptic coupling. We hope to facilitate progress on this question by providing our source code for further exploration. Using the tools that have been developed by the BRIAN community one can generate or import arbitrary complex cell morphologies (e.g. NeuroML files). Our source code adds the TM- and RTM model, which allows exploring the direct impact of extracellular properties on target neurons.

Reviewer 2 (Public Review):

In this study, the authors measured extracellular electrical features of colliding APs travelling in different directions down an isolated earthworm axon. They then used these features to build a model of the potential ephaptic effects of AP annihilation, i.e. the electrical signals produced by colliding/annihilating APs that may influence neighbouring tissue. The model was then applied to some different hypothetical scenarios involving synaptic connections. The conclusion was that an annihilating AP at a presynaptic terminal can ephaptically influence the voltage of a postsynaptic cell (this is, presumably, the ’electrical coupling between neurons’ of the title), and that the nature of this influence depends on the physical configuration of the synapse.

As an experimental neuroscientist who has never used computational approaches, I am unable to comment on the rigour of the analytical approaches that form the bulk of this paper. The experimental approaches appear very well carried out, and here I just have one query - an important assumption made is that the conduction velocity of anti- and orthodromically propagating APs is identical in every preparation, but this is never empirically/statistically demonstrated.

My major concern is with the conclusions drawn from the synaptic modelling, which, disappointingly, is never benchmarked against any synaptic data. The authors state in their Introduction that a ’quantitative physical description’ of ephaptic coupling is ’missing’, however, they do not provide such a description in this manuscript. Instead, modelled predictions are presented of possible ephaptic interactions at different types of synapses, and these are then partially and qualitatively compared to previous published results in the Discussion. To support the authors’ assertion that AP annihilation induces electrical coupling between neurons, I think they need to show that their model of ephaptic effects can quantitatively explain key features of experimental data pertaining to synaptic function. Without this, the paper contains some useful high-precision quantitative measurements of axonal AP collisions, some (I assume) high-quality modelling of these collisions, and some interesting theoretical predictions pertaining to synaptic interactions, but it does not support the highly significant implications suggested for synaptic function.

We thank the reviewer for highlighting the potential and the limitation of our model. We demonstrated with empirical data that measured conduction velocities of anti- and orthodromic propagating APs are indeed very similar and values are provided in Appendix 3 – table 1.

In order to address how our model ’of ephaptic effects can quantitatively explain key features of experimental data’, we used the measured modulation of AP rates in Purkinje fibers by Blot and Babour (2014) and our results are now included in the manuscript. In our model, we implemented the ephaptic coupling of the Basket cell (with an annihilating AP) and predicted the modulation of AP rate in the Purkinje cell. Our model predictions are compared to the measured modulation of AP-rates in Purkinje cells and is added as Fig. 5 to the main manuscript (line 264 to 284 ). With this example, we show that ephaptic coupling as described with our RTM model can quantitatively describe key features of experimental data. Both, the rapid inhibition and the rebound activity is described by our model with implementation of non-excitable parts at the pinceau of the Basket cell. Future, experimental research can use the provided formalism to investigate in more detail the ephaptic coupling in systems like the Mauthner cell and the Purkinje cell by exploring how accessory structures and concomitant physical parameters, e.g. the extracellular properties impact ephaptic coupling.

Reviewer 3 (Public Review):

This manuscript aims to exploit experimental measurements of the extracellular voltages produced by colliding action potentials to adjust a simplified model of action potential propagation that is then used to predict the extracellular fields at axon terminals. The overall rationale is that when solving the cable equation (which forms the substrate for models of action potential propagation in axons), the solution for a cable with a closed end can be obtained by a technique of superposition: a spatially reflected solution is added to that for an infinite cable and this ensures by symmetry that no axial current flows at the closed boundary. By this method, the authors calculate the expected extracellular fields for axon terminals in different situations. These fields are of potential interest because, according to the authors, their magnitude can be larger than that of a propagating action potential and may be involved in ephaptic signalling. The authors perform direct measurements of colliding action potentials, in the earthworm giant axon, to parameterise and test their model.

Although simplified models can be useful and the trick of exploiting the collision condition is interesting, I believe there are several significant problems with the rationale, presentation, and application, such that the validity and potential utility of the approach is not established.

Simplified model vs. Hogdkin and Huxley

The authors employ a simplified model that incorporates a two-state membrane (in essence resting and excited states) and adds a recovery mechanism. This generates a propagating wave of excitation and key observables such as propagation speed and action potential width (in space) can be adjusted using a small number of parameters. However, even if a Hodgkin-Huxley model does contain a much larger number of parameters that may be less easy to adjust directly, the basic formalism is known to be accurate and typical modifications of the kinetic parameters are very well understood, even if no direct characterisations already exist or cannot be obtained. I am therefore unconvinced by the utility of abandoning the HodgkinHuxley version.

In several places in the manuscript, the simplified model fits the data well whereas the Hodgkin-Huxley model deviates strongly (e.g. Fig. 3CD). This is unsatisfying because it seems unlikely that the phenomenon could not be modelled accurately using the HH formulation. If the authors really wish to assert that it is ”not suitable to predict the effects caused by AP [collision]” (p9) they need to provide a good deal more analysis to establish the mechanism of failure.

We are not as convinced as the reviewer that, at the current state of parameter estimation, the HH model is suited for predicting ephaptic coupling after ’adjusting’ parameters. There are strong arguments against such an approach. A major function of a model is to make testable predictions rather than to just mimic a biological phenomenon. The predictive power of a model heavily depends on how reasonable model parameters can be estimated or measured. As the reviewer correctly points out in the specific comments (”... the parameters adjusted to fit the model are the membrane capacitance and intracellular resistance. These have a physical reality and could easily be measured or estimated quite accurately...”), our model contains only parameters that can be assessed experimentally, thus it has a better predictive power compared to the HH model with a multitude of parameters for which ”no direct characterisations already exist or cannot be obtained” (citing reviewer from above).

Already the founders of the HH model were well aware of the limitations, as stated by Hodgkin and Huxley in 1952 (J Physiol 117:500–544):

An equally satisfactory description of the voltage clamp data could no doubt have been achieved with equations of very different form ... The success of the equations is no evidence in favour of the mechanism of permeability change that we tentatively had in mind when formulating them.

A catchy but sloppy description for the problem of overfitting with too many parameters is given by the quote of John von Neumann: With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

We do not rule out the possibility that the HH model eventually can be used to predict ephaptic coupling. However, at the moment, parameter estimation for the HH model prevents its usability for predicting ephaptic coupling.

(In)applicability of the superposition principle

The reflecting boundary at the terminal is implemented using the symmetry of the collision of action potentials. However, at a closed cable there is no reflecting boundary in the extracellular space and this implied assumption is particularly inappropriate where the extracellular field is one objective of the modelling, as here. I believe this assumption is not problematic for the calculation of the intracellular voltage, because extracellular voltage gradients can usually be neglected1, but the authors need to explain how the issue was dealt with for the calculation of the extracellular fields of terminals. I assume they were calculated from the membrane currents of one-half of the collision solution, but this does not seem to be explained. It might be worth showing a spatial profile of the calculated field.

We disagree with the reviewer’s statement ’...at a closed cable there is no reflecting boundary in the extracellular space and this implied assumption is particularly inappropriate...’. We do not imply this assumption in our model! We do not assume any symmetry or boundary condition in the extracellular space. Instead, the extracellular field is calculated for an infinite homogeneous volume conductor (Eq.

6).

We conduct separate calculations for (1) source membrane current, (2) resulting extracellular field, and (3) impact upon a target neuron. The boundary condition used for our calculations only refers to the axial current being zero at the axon terminal. Consequently all the internal current that enters the last compartment must leave the last compartment as membrane current and contributes to the extracellular current and field.

The extracellular field around the axon terminal is not symmetric, as can be seen by it’s impact upon a target in Figure 4—figure supplement 1 which is also not symmetric. The symmetry of the extracellular field when APs are colliding (Cf. symmetry in Fig 1C) is merly the result of the symmetric stimulation and counterpropagation of two APs. We now are describing more specifically the bounday condition for colliding and terminating APs already in the introduction: ’A suitable boundary condition (intracellular, axial current equals zero) can be generated experimentally by a collision of two counter-propagating APs ... Within any cable model, the very same boundary condition also exists within the axon at the synaptic terminal due to the broken translation symmetry for the current loops ...’ Later, at the result section (Discharge of colliding APs), we continue with ’AP propagation is blocked when the axial current is shut down at a boundary condition, e.g. by reaching the axon terminal or by AP collision....’ and implement this condition in our calculations for the axon terminals.

Missing demonstrations

Central analytical results are stated rather brusquely, notably equations (3) and (4) and the relation between them. These merit an expanded explanation at the least. A better explanation of the need for the collision measurements in parameterising the models should also be provided.

We thank the reviewer for pointing out the insufficient explanation of the equations 3 and 4. We rephrased the paragraph ’Discharge of colliding APs’ in order to clarify the origin and the function of the two equations (eq. 3: how much charge is expelled and eq. 4: the resulting extracellular potential that is used for model validation).

Later, in the Discussion, we rephrased the paragraph where we describe the annihilation process and explain further that one term of eq. 4 sometimes is refered to ’activating function’ when using microelectrodes for stimulation.

With respect to the ’explanation of the need for the collision measurement’, we think that the explanations we give at several locations in the manuscript are sufficient as is. We explain and elaborate in the introduction: ’We explore the behaviour of APs at boundaries ... In this study, we first focus on collisions of APs. Our experimental observation of colliding APs provides unique access to the spatial profile of the extracellular potential around APs that are blocked by collisions and thus annihilate..... Recording propagating APs allows to determine both the propagation velocity and the amplitude of the extracellular electric potentials. The collision experiment provides additional information ... In the results we recall: ’The width of the collision is a measure of the characteristic length λ⋆ of the AP and is uniquely revealed by a collision sweep experiment.’

Adjusted parameters

I am uncomfortable that the parameters adjusted to fit the model are the membrane capacitance and intracellular resistance. These have a physical reality and could easily be measured or estimated quite accurately. With a variation of more than 20-fold reported between the different models in Appendix 2 we can be sure that some of the models are based upon quite unrealistic physical assumptions, which in turn undermines confidence in their generality.

The fact that the parameters of our model have physical realities is clearly in favor of our models. We rephrased the legend of the table, now explaining the procedure for the model fitting and the rational behind. Although the values of g⋆ can differ by a factor of 15 and the resulting amplitude is very different, the relationship ri cm \= vpλ⋆ is very similar, independently of the model used and this confirms our analytical framework.

p8 - the values of both the extracellular (100 Ohm m) and intracellular resistivity (1 Ohm m) appear to be in error, especially the former.

We have the following justification for the resistivity values we used. For the intracellular resistivity, literature values range from 0.4 - 1.5 Ohm m, and therefore we selected 1 Ohm m. See: Carpenter et al (1975) doi: 10.1085/jgp.66.2.139; Cole et al (1975) doi: 10.1085/jgp.66.2.133; Bekkers (2014) doi: 10.1007/978-1-46147320-6 35-2.

Estimating extracellular resistivity is less straight forward, since it depends crucially on the structure around the synapse which consists of conducting saline and insulating fatty tissue. Ranges from 3 to 600 Ohm m are reported (Linden et al (2011) doi: 10.1016/j.neuron.2011.11.006) and Bakiri et al (2011) doi: 10.1113/jphysiol.2010.201376). Weiss et al (2008; doi: 10.1073/pnas.0806145105) report extracellular resistivities in the Mauthner Cap between 50-600 Ohm m in SI. Since the pinceau is structurally similar to the Mauthner cells axon cap, we argue that a value of 100 Ohm m is a reasonable choice for our calculations. Additionally, we derived a value from Blot and Barbour (doi:c10.1038/nn.3624), rephrased the paragraph in the main text and added our calculation to the supplementary material (Appendix 1).

(In)applicability to axon terminals

The rationale of the application of the collision formalism to axon terminals is somewhat undermined by the fact that they tend not to be excitable. There is experimental evidence for this in the Calyx of Held and the cerebellar pinceau.

The solution found via collision is therefore not directly applicable in these cases.

We do not agree with the reviewer’s statement that ’the solution found via collision is (therefore) not directly applicable...’. Our model is well suited for application on axon terminals that are not excitable, e.g. the pinceau of the basket cell, as the reviewer points out. We have included a calculation for this case and present the results in the new Fig. 5 (main text line 264 to 284 ).

Comparison with experimental data

More effort should be made to compare the modelling with the extracellular terminal fields that have been reported in the literature.

As outlined above (see: Reponse to reviewer 2), we now compare directly the predictions of our models with measured modulation of AP rates in Purkinje fibers (Blot and Babour 2014) and our results are included in the manuscript (Fig. 5 and main text line 264 to 284). See also our response to reviewer 2 in which we address how our model ’of ephaptic effects can quantitatively explain key features of experimental data’.

Choice of term ”annihilation”

The term annihilation does not seem wholly appropriate to me. The dictionary definitions are something along the lines of complete destruction by an external force or mutual destruction, for example of an electron and a positron. I don’t think either applies exactly here. I suggest retaining the notion of collision which is well understood in this context.

Experimentally, we generated a collision of APs and showed that colliding APs dissapear and do not pass each other. For this process the term annihilation is used in our and in other studies (see e.g. Berg et al (2017) doi: 10.1103/PhysRevX.7.028001; Johnson et al (2018) doi: 10.3389/fphys.2018.00779; Follmann (2015) doi: 10.1103/PhysRevE.92.032707; Shrivastava et al (2018) doi: 10.1098/rsif.2017.0803). The physical processes involved in the termination of an AP at a closed end are essentially identical to those of two colliding APs. This we think justifies using the term annihilation for those processes.

Recommendations for the authors:

We believe the work is of high quality and should motivate future experimental work. We are including the review comments here for your information. The main piece of feedback we are offering is that the broad claims need to be adjusted to the strength of evidence provided: as is, the manuscript provides compelling predictions but the claim that these predictions are in full agreement with data remains to be substantiated. A technical concern raised by the reviewers is that the reflecting boundary condition may need further justification. The authors may wish to respond to this issue in a rebuttal and/or adjust the manuscript as necessary.

We substantiated our claim that our predictions are in full agreement with experimental data. We added to the manuscript a section in which we compare our models’ predictions to published, experimental data. To this aim, we extracted date from the publication of Blot and Babour (2014), we elaborated on the parameters used and run our model accordingly. We added to the Results/Model of ephaptic coupling a paragraph on ’The modulation of activity in Purkinje cells...’ (line 264), where we describe our results and we also included another figure to the main text for illustration (Fig. 5).

We clarified the term ’boundary condition’ by rephrasing parts of the introduction and we explain the rational behind in ’Discharge of colliding APs (...AP propagation is blocked when axial current is shut down...) and in ’Model of ephaptic coupling (Within any cable model, the same boundary...). See also our response to the general comments of reviewer 3 above.

Reviewer 1 (Recommendations For The Authors):

Major:

Accessing data and code requires signing in, which should not be required. The link provided also seems to be not accessible yet - could be pending review.

The repository is now publicly availible. We did provide an access code within the letter to the editor, this code is no longer required.

Line 74: how about morphology? Authors should clarify and emphasize in the introduction that the TM model is a spatially continuous model with partial differential equations as opposed to discrete morphological models to simulate HH equations.

The reviewer is correct that the TM model is continous. However, so is the HH model. The difference between HH and TM is only that the TM model can be solved analytically, which yields a spatially homogeneous analytical solution. It should be noted that this analytical solution can only be valid for a homogeneous (therefore infinite) nerve. Every numerical computation, be it HH or TM, requires a finite number of discrete compartments. In our calculations, we used identical compartment models for HH, TM and RTM model. In each compartment, the differential equations are solved numerically. Since there is no fundamental difference between these models, we obstain from changing the text.

Minor:

Major typo: ventral nerve cord, not ”chord”. Repeated in several places.

Thank you for indicating this typo to us.

Line 25: inhibition, excitation, and modulation?

We changed the line to: ... leads to modulation, e.g. excitation or inhibition

Line 70: better term for ”length” of AP would be ”duration”. Also, the sentence could be simplified to use either ”its” or ”of the AP”

Space and time are not interchangable. Thus, the term lenght can not be replaced by duration. We simplified the structure of the sentence as suggested.

Fig 1A/B: it’s strange that panel B precedes panel A.

Exchanged

Fig 1C: don’t see the ”horizontal line”; also regarding ”The recording was at a medial position”, the caption is not clear until one reads the main text.

We changed the legend to: ... The collision is captured in the recording line at y-position 0 mm, while orthodromic propagation is at the top and antidromic propagation is at the bottom. (D) The peak amplitude as a function of the distance to the collision. Examples of four sweeps at three positions along the nerve cord....

Line 127: the per distance measures could be named as ”specific” conductivity, etc.

We explicitly provide the units thereby defining the quantities unambigously.

Line 176: typo ”ad-hoc”.

Thank you.

Fig 4B: should clarify that the circle in the schematic is not the soma but a synaptic bouton.

We rephrased to ’...(B,C) when the AP is annihilating at a bouton of a neuron terminal (upper neuron in end-to-shaft geometry, similar to the Basket cell–Purkinje cell synapse)...’, and we added a label to Fig 4B.

Reviewer 2 (Recommendations For The Authors):

Can the authors’ model be quantitatively compared with experimental data of ephaptic interactions at synapses (e.g. the Blot & Barbour study described in the Discussion)?

We did so as outlined in our response to the reviewer above.

Can statistical evidence be provided that the velocities of anti- and orthodromic APs are indeed identical in the earthworm nerve recordings?

These data and statistics are available in Appendix 2, now 3 – table 1

Why not reorder ABCD in Fig1 so the subpanels run from left to right?

We adjusted the labels accordingly.

eLife Assessment

This paper represents a "classic" approach towards evaluating a novel taste stimulus in an animal model, including standard behavioral tests (some with nerve transections), taste nerve physiology and immunocytochemistry of the tongue. The stimulus being tested is ornithine, from a class of stimuli called "kokumi", which enhance other canonical tastes, increasing their hedonic attributes; the mechanism for ornithine detection is thought to be GPRC6A receptors expressed in taste cells. The authors showed evidence for this in an earlier paper with mice; this paper evaluates ornithine taste in a rat model. This work is valuable but incomplete.

Reviewer #1 (Public review):

Summary:

This paper contains what could be described as a "classic" approach towards evaluating a novel taste stimuli in an animal model, including standard behavioral tests (some with nerve transections), taste nerve physiology, and immunocytochemistry of the tongue. The stimulus being tested is ornithine, from a class of stimuli called "kokumi", which are stimuli that enhance other canonical tastes, increasing essentially the hedonic attributes of these other stimuli; the mechanism for ornithine detection is thought to be GPRC6A receptors expressed in taste cells. The authors showed evidence for this in an earlier paper with mice; this paper evaluates ornithine taste in a rat model.

Strengths:

The data show the effects of ornithine on taste: in two-bottle and briefer intake tests, adding ornithine results in a higher intake of most, but not all, stimuli tests. Bilateral nerve cuts or the addition of GPRC6A antagonists decrease this effect. Small effects of ornithine are shown in whole-nerve recordings.

Weaknesses:

The conclusion seems to be that the authors have found evidence for ornithine acting as a taste modifier through the GPRC6A receptor expressed on the anterior tongue. It is hard to separate their conclusions from the possibility that any effects are additive rather than modulatory. Animals did prefer ornithine to water when presented by itself. Additionally, the authors refer to evidence that ornithine is activating the T1R1-T1R3 amino acid taste receptor, possibly at higher concentrations than they use for most of the study, although this seems speculative. It is striking that the largest effects on taste are found with the other amino acid (umami) stimuli, leading to the possibility that these are largely synergistic effects taking place at the tas1r receptor heterodimer.

Reviewer #2 (Public review):

Summary:

The authors used rats to determine the receptor for a food-related perception (kokumi) that has been characterized in humans. They employ a combination of behavioral, electrophysiological, and immunohistochemical results to support their conclusion that ornithine-mediated kokumi effects are mediated by the GPRC6A receptor. They complemented the rat data with some human psychophysical data. I find the results intriguing, but believe that the authors overinterpret their data.

Strengths:

The authors examined a new and exciting taste enhancer (ornithine). They used a variety of experimental approaches in rats to document the impact of ornithine on taste preference and peripheral taste nerve recordings. Further, they provided evidence pointing to a potential receptor for ornithine.

Weaknesses:

The authors have not established that the rat is an appropriate model system for studying kokumi. Their measurements do not provide insight into any of the established effects of kokumi on human flavor perception. The small study on humans is difficult to compare to the rat study because the authors made completely different types of measurements. Thus, I think that the authors need to substantially scale back the scope of their interpretations. These weaknesses diminish the likely impact of the work on the field of flavor perception.

Reviewer #3 (Public review):

Summary:

In this study, the authors set out to investigate whether GPRC6A mediates kokumi taste initiated by the amino acid L-ornithine. They used Wistar rats, a standard laboratory strain, as the primary model and also performed an informative taste test in humans, in which miso soup was supplemented with various concentrations of L-ornithine. The findings are valuable and overall the evidence is solid. L-Ornithine should be considered to be a useful test substance in future studies of kokumi taste and the class C G protein-coupled receptor known as GPRC6A (C6A) along with its homolog, the calcium-sensing receptor (CaSR) should be considered candidate mediators of kokumi taste.

Strengths:

The overall experimental design is solid based on two bottle preference tests in rats. After determining the optimal concentration for L-Ornithine (1 mM) in the presence of MSG, it was added to various tastants, including inosine 5'-monophosphate; monosodium glutamate (MSG); mono-potassium glutamate (MPG); intralipos (a soybean oil emulsion); sucrose; sodium chloride (NaCl); citric acid and quinine hydrochloride. Robust effects of ornithine were observed in the cases of IMP, MSG, MPG, and sucrose, and little or no effects were observed in the cases of sodium chloride, citric acid, and quinine HCl. The researchers then focused on the preference for Ornithine-containing MSG solutions. The inclusion of the C6A inhibitors Calindol (0.3 mM but not 0.06 mM) or the gallate derivative EGCG (0.1 mM but not 0.03 mM) eliminated the preference for solutions that contained Ornithine in addition to MSG. The researchers next performed transections of the chord tympani nerves (with sham operation controls) in anesthetized rats to identify the role of the chorda tympani branches of the facial nerves (cranial nerve VII) in the preference for Ornithine-containing MSG solutions. This finding implicates the anterior half-two thirds of the tongue in ornithine-induced kokumi taste. They then used electrical recordings from intact chorda tympani nerves in anesthetized rats to demonstrate that ornithine enhanced MSG-induced responses following the application of tastants to the anterior surface of the tongue. They went on to show that this enhanced response was insensitive to amiloride, selected to inhibit 'salt tastant' responses mediated by the epithelial Na+ channel, but eliminated by Calindol. Finally, they performed immunohistochemistry on sections of rat tongue demonstrating C6A positive spindle-shaped cells in fungiform papillae that partially overlapped in its distribution with the IP3 type-3 receptor, used as a marker of Type-II cells, but not with (i) gustducin, the G protein partner of Tas1 receptors (T1Rs), used as a marker of a subset of type-II cells; or (ii) 5-HT (serotonin) and Synaptosome-associated protein 25 kDa (SNAP-25) used as markers of Type-III cells.

Weaknesses:

The researchers undertook what turned out to be largely confirmatory studies in rats with respect to their previously published work on Ornithine and C6A in mice (Mizuta et al Nutrients 2021).

The authors point out that animal models pose some difficulties of interpretation in studies of taste and raise the possibility in the Discussion that umami substances may enhance the taste response to ornithine (Line 271, Page 9). They miss an opportunity to outline the experimental results from the study that favor their preferred interpretation that ornithine is a taste enhancer rather than a tastant.

At least two other receptors in addition to C6A might mediate taste responses to ornithine: (i) the CaSR, which binds and responds to multiple L-amino acids (Conigrave et al, PNAS 2000), and which has been previously reported to mediate kokumi taste (Ohsu et al., JBC 2010) as well as responses to Ornithine (Shin et al., Cell Signaling 2020); and (ii) T1R1/T1R3 heterodimers which also respond to L-amino acids and exhibit enhanced responses to IMP (Nelson et al., Nature 2001). While the experimental results as a whole favor the authors' interpretation that C6A mediates the Ornithine responses, they do not make clear either the nature of the 'receptor identification problem' in the Introduction or the way in which they approached that problem in the Results and Discussion sections. It would be helpful to show that a specific inhibitor of the CaSR failed to block the ornithine response. In addition, while they showed that C6A-positive cells were clearly distinct from gustducin-positive, and thus T1R-positive cells, they missed an opportunity to clearly differentiate C6A-expressing taste cells and CaSR-expressing taste cells in the rat tongue sections.

It would have been helpful to include a positive control kokumi substance in the two-bottle preference experiment (e.g., one of the known gamma-glutamyl peptides such as gamma-glu-Val-Gly or glutathione), to compare the relative potencies of the control kokumi compound and Ornithine, and to compare the sensitivities of the two responses to C6A and CaSR inhibitors.

The results demonstrate that enhancement of the chorda tympani nerve response to MSG occurs at substantially greater Ornithine concentrations (10 and 30 mM) than were required to observe differences in the two bottle preference experiments (1.0 mM; Figure 2). The discrepancy requires careful discussion and if necessary further experiments using the two-bottle preference format.

Author response:

Reviewer #1 (Public review):

Summary:

This paper contains what could be described as a "classic" approach towards evaluating a novel taste stimuli in an animal model, including standard behavioral tests (some with nerve transections), taste nerve physiology, and immunocytochemistry of the tongue. The stimulus being tested is ornithine, from a class of stimuli called "kokumi", which are stimuli that enhance other canonical tastes, increasing essentially the hedonic attributes of these other stimuli; the mechanism for ornithine detection is thought to be GPRC6A receptors expressed in taste cells. The authors showed evidence for this in an earlier paper with mice; this paper evaluates ornithine taste in a rat model.

Strengths:

The data show the effects of ornithine on taste: in two-bottle and briefer intake tests, adding ornithine results in a higher intake of most, but not all, stimuli tests. Bilateral nerve cuts or the addition of GPRC6A antagonists decrease this effect. Small effects of ornithine are shown in whole-nerve recordings.

Weaknesses:

The conclusion seems to be that the authors have found evidence for ornithine acting as a taste modifier through the GPRC6A receptor expressed on the anterior tongue. It is hard to separate their conclusions from the possibility that any effects are additive rather than modulatory. Animals did prefer ornithine to water when presented by itself. Additionally, the authors refer to evidence that ornithine is activating the T1R1-T1R3 amino acid taste receptor, possibly at higher concentrations than they use for most of the study, although this seems speculative. It is striking that the largest effects on taste are found with the other amino acid (umami) stimuli, leading to the possibility that these are largely synergistic effects taking place at the tas1r receptor heterodimer.

We would like to thank Reviewer #1 for the valuable comments. Our basis for considering ornithine as a taste modifier stems from our observation that a low concentration of ornithine (1 mM), which does not elicit a preference on its own, enhances the preference for umami substances, sucrose, and soybean oil through the activation of the GPRC6A receptor. Notably, this receptor is not typically considered a taste receptor. The reviewer suggested that the enhancement of umami taste might be due to potentiation occurring at the TAS1R receptor heterodimer. However, we propose that a different mechanism may be at play, as an antagonist of GPRC6A almost completely abolished this enhancement. In the revised manuscript, we will endeavor to provide additional information on the role of ornithine as a taste modifier acting through the GPRC6A receptor.

Reviewer #2 (Public review):

Summary:

The authors used rats to determine the receptor for a food-related perception (kokumi) that has been characterized in humans. They employ a combination of behavioral, electrophysiological, and immunohistochemical results to support their conclusion that ornithine-mediated kokumi effects are mediated by the GPRC6A receptor. They complemented the rat data with some human psychophysical data. I find the results intriguing, but believe that the authors overinterpret their data.

Strengths:

The authors examined a new and exciting taste enhancer (ornithine). They used a variety of experimental approaches in rats to document the impact of ornithine on taste preference and peripheral taste nerve recordings. Further, they provided evidence pointing to a potential receptor for ornithine.

Weaknesses:

The authors have not established that the rat is an appropriate model system for studying kokumi. Their measurements do not provide insight into any of the established effects of kokumi on human flavor perception. The small study on humans is difficult to compare to the rat study because the authors made completely different types of measurements. Thus, I think that the authors need to substantially scale back the scope of their interpretations. These weaknesses diminish the likely impact of the work on the field of flavor perception.

We would like to thank Reviewer #2 for the valuable comments and suggestions. Regarding the question of whether the rat is an appropriate model system for studying kokumi, we have chosen this species for several reasons: it is readily available as a conventional experimental model for gustatory research; the calcium-sensing receptor (CaSR), known as the kokumi receptor, is expressed in taste bud cells; and prior research has demonstrated the use of rats in kokumi studies involving gamma Glu-Val-Gly (Yamamoto and Mizuta, Chem. Senses, 2022). We acknowledge that fundamentally different types of measurements were conducted in the human psychophysical study and the rat study. Kokumi can indeed be assessed and expressed in humans; however, we do not currently have the means to confirm that animals experience kokumi in the same way that humans do. Therefore, human studies are necessary to evaluate kokumi, a conceptual term denoting enhanced flavor, while animal studies are needed to explore the potential underlying mechanisms of kokumi. We believe that a combination of both human and animal studies is essential, as is the case with research on sugars. While sugars are known to elicit sweetness, it is unclear whether animals perceive sweetness identically to humans, even though they exhibit a strong preference for sugars. In the revised manuscript, we will incorporate additional information to address the comments raised by the reviewer. We will also carefully review and revise our previous statements to ensure accuracy and clarity.

Reviewer #3 (Public review):

Summary:

In this study, the authors set out to investigate whether GPRC6A mediates kokumi taste initiated by the amino acid L-ornithine. They used Wistar rats, a standard laboratory strain, as the primary model and also performed an informative taste test in humans, in which miso soup was supplemented with various concentrations of L-ornithine. The findings are valuable and overall the evidence is solid. L-Ornithine should be considered to be a useful test substance in future studies of kokumi taste and the class C G protein-coupled receptor known as GPRC6A (C6A) along with its homolog, the calcium-sensing receptor (CaSR) should be considered candidate mediators of kokumi taste.

Strengths:

The overall experimental design is solid based on two bottle preference tests in rats. After determining the optimal concentration for L-Ornithine (1 mM) in the presence of MSG, it was added to various tastants, including inosine 5'-monophosphate; monosodium glutamate (MSG); mono-potassium glutamate (MPG); intralipos (a soybean oil emulsion); sucrose; sodium chloride (NaCl); citric acid and quinine hydrochloride. Robust effects of ornithine were observed in the cases of IMP, MSG, MPG, and sucrose, and little or no effects were observed in the cases of sodium chloride, citric acid, and quinine HCl. The researchers then focused on the preference for Ornithine-containing MSG solutions. The inclusion of the C6A inhibitors Calindol (0.3 mM but not 0.06 mM) or the gallate derivative EGCG (0.1 mM but not 0.03 mM) eliminated the preference for solutions that contained Ornithine in addition to MSG. The researchers next performed transections of the chord tympani nerves (with sham operation controls) in anesthetized rats to identify the role of the chorda tympani branches of the facial nerves (cranial nerve VII) in the preference for Ornithine-containing MSG solutions. This finding implicates the anterior half-two thirds of the tongue in ornithine-induced kokumi taste. They then used electrical recordings from intact chorda tympani nerves in anesthetized rats to demonstrate that ornithine enhanced MSG-induced responses following the application of tastants to the anterior surface of the tongue. They went on to show that this enhanced response was insensitive to amiloride, selected to inhibit 'salt tastant' responses mediated by the epithelial Na+ channel, but eliminated by Calindol. Finally, they performed immunohistochemistry on sections of rat tongue demonstrating C6A positive spindle-shaped cells in fungiform papillae that partially overlapped in its distribution with the IP3 type-3 receptor, used as a marker of Type-II cells, but not with (i) gustducin, the G protein partner of Tas1 receptors (T1Rs), used as a marker of a subset of type-II cells; or (ii) 5-HT (serotonin) and Synaptosome-associated protein 25 kDa (SNAP-25) used as markers of Type-III cells.

Weaknesses:

The researchers undertook what turned out to be largely confirmatory studies in rats with respect to their previously published work on Ornithine and C6A in mice (Mizuta et al Nutrients 2021).

The authors point out that animal models pose some difficulties of interpretation in studies of taste and raise the possibility in the Discussion that umami substances may enhance the taste response to ornithine (Line 271, Page 9). They miss an opportunity to outline the experimental results from the study that favor their preferred interpretation that ornithine is a taste enhancer rather than a tastant.

At least two other receptors in addition to C6A might mediate taste responses to ornithine: (i) the CaSR, which binds and responds to multiple L-amino acids (Conigrave et al, PNAS 2000), and which has been previously reported to mediate kokumi taste (Ohsu et al., JBC 2010) as well as responses to Ornithine (Shin et al., Cell Signaling 2020); and (ii) T1R1/T1R3 heterodimers which also respond to L-amino acids and exhibit enhanced responses to IMP (Nelson et al., Nature 2001). While the experimental results as a whole favor the authors' interpretation that C6A mediates the Ornithine responses, they do not make clear either the nature of the 'receptor identification problem' in the Introduction or the way in which they approached that problem in the Results and Discussion sections. It would be helpful to show that a specific inhibitor of the CaSR failed to block the ornithine response. In addition, while they showed that C6A-positive cells were clearly distinct from gustducin-positive, and thus T1R-positive cells, they missed an opportunity to clearly differentiate C6A-expressing taste cells and CaSR-expressing taste cells in the rat tongue sections.

It would have been helpful to include a positive control kokumi substance in the two-bottle preference experiment (e.g., one of the known gamma-glutamyl peptides such as gamma-glu-Val-Gly or glutathione), to compare the relative potencies of the control kokumi compound and Ornithine, and to compare the sensitivities of the two responses to C6A and CaSR inhibitors.

The results demonstrate that enhancement of the chorda tympani nerve response to MSG occurs at substantially greater Ornithine concentrations (10 and 30 mM) than were required to observe differences in the two bottle preference experiments (1.0 mM; Figure 2). The discrepancy requires careful discussion and if necessary further experiments using the two-bottle preference format.

We would like to thank Reviewer #3 for the valuable comments and helpful suggestions. We propose that ornithine has two stimulatory actions: one acting on GPRC6A, particularly at lower concentrations, and another on amino acid receptors such as T1R1/T1R3 at higher concentrations. Consequently, ornithine is not preferable at lower concentrations but becomes preferable at higher concentrations. For our study on kokumi, we used a low concentration (1 mM) of ornithine. The possibility mentioned in the Discussion that 'the umami substances may enhance the taste response to ornithine' is entirely speculative. We will reconsider including this description in the revised version. As the reviewer suggested, in addition to GPRC6A, ornithine may bind to CaSR and/or T1R1/T1R3 heterodimers. However, we believe that ornithine mainly binds to GPRC6A, as a specific inhibitor of this receptor almost completely abolished the enhanced response to umami substances, and our immunohistochemical study indicated that GPRC6A-expressing taste cells are distinct from CaSR-expressing taste cells (see Supplemental Fig. 3). We conducted essentially the same experiments using gamma-Glu-Val-Gly in Wistar rats (Yamamoto and Mizuta, Chem. Senses, 2022) and compared the results in the Discussion. The reviewer may have misunderstood the chorda tympani results: we added the same concentration (1 mM) used in the two-bottle preference test to MSG (Fig. 5-B). Fig. 5-A shows nerve responses to five concentrations of plain ornithine. In the revised manuscript, we will strive to provide more precise information reflecting the reviewer’s comments.

eLife Assessment

This study proposes an important new approach to analyzing cell-count data that are often undersampled and cannot be correctly assessed with traditional statistical analyses. The presented case studies provide convincing evidence of the superiority of the proposed methodology to existing approaches, which could promote the use of Bayesian statistics among neuroscientists. However, the generalizability of the methodology to other data types is not fully evidenced.

Reviewer #1 (Public review):

Summary:

This work proposes a new approach to analyse cell-count data from multiple brain regions. Collecting such data can be expensive and time-intensive, so, more often than not, the dimensionality of the data is larger than the number of samples. The authors argue that Bayesian methods are much better suited to correctly analyse such data compared to classical (frequentist) statistical methods. They define a hierarchical structure, partial pooling, in which each observation contributes to the population estimate to more accurately explain the variance in the data. They present two case studies in which their method proves more sensitive in identifying regions where there are significant differences between conditions, which otherwise would be hidden.

Strengths:

The model is presented clearly, and the advantages of the hierarchical structure are strongly justified. Two alternative ways are presented to account for the presence of zero counts. The first involves the use of a horseshoe prior, which is the more flexible option, while the second involves a modified Poisson likelihood, which is better suited to datasets with a large number of zero counts, perhaps due to experimental artifacts. The results show a clear advantage of the Bayesian method for both case studies.

The code is freely available, and it does not require a high-performance cluster to execute for smaller datasets. As Bayesian statistical methods become more accessible in various scientific fields, the whole scientific community will benefit from the transition away from p-values. Hierarchical Bayesian models are an especially useful tool that can be applied to many different experimental designs. However, while conceptually intuitive, their implementation can be difficult. The authors provide a good framework with room for improvement.

Weaknesses:

Alternative possibilities are discussed regarding the prior and likelihood of the model. Given that the second case study inspired the introduction of the zero-inflation likelihood, it is not clear how applicable the general methodology is to various datasets. If every unique dataset requires a tailored prior or likelihood to produce the best results, the methodology will not easily replace more traditional statistical analyses that can be applied in a straightforward manner. Furthermore, the differences between the results produced by the two Bayesian models in case study 2 are not discussed. In specific regions, the models provide conflicting results (e.g., regions MH, VPMpc, RCH, SCH, etc.), which are not addressed by the authors. A third case study would have provided further evidence for the generalizability of the methodology.

Reviewer #2 (Public review):

Summary:

This is a well-written methodology paper applying a Bayesian framework to the statistics of cell counts in brain slices. A sharpening of the bounds on measured quantities is demonstrated over existing frequentist methods and therefore the work is a contribution to the field.

Strengths:

As well as a mathematical description of the approach, the code used is provided in a linked repository.

Weaknesses:

A clearer link between the experimental data and model-structure terminology would be a benefit to the non-expert reader.

Author response:

We thank both reviewers for their considerate reviews. In this provisional response we would like to make a few key points.

Given that we introduced a bespoke likelihood model for the second dataset, Reviewer 1 asks whether "every unique dataset requires a tailored prior or likelihood to produce the best results". Our intention is to advocate for the horseshoe prior model as a 'standard' first analysis for any cell count dataset. If extra knowledge about the data is available, or if any data artefacts are detected, more elaborate likelihoods could be introduced as needed in a follow-up analysis. Our introduction of the zero-inflated Poisson likelihood for the second dataset was one such example, but many alternatives could exist. This iterative approach to model building, sometimes referred to as a `Bayesian workflow' is seen as good practise in Bayesian data analysis literature. In the revised version of the paper, we will try to explain the recommendations and modelling philosophy behind this method while emphasising that tailoring or bespoke modelling is not required for our `standard analysis', what we would regard as the Bayesian replacement for a t-test on counts.

Reviewer 1 notes that "the differences between the results produced by the two Bayesian models in case study 2 are not discussed". We agree that this discrepancy, arising from the specific assumptions of each model is an interesting issue which we should better explore in the paper. In Figure 6 we plotted the actual data values alongside posterior and confidence intervals to explain how the results from the ZIP likelihood and Horseshoe prior compare with those from a t-test. However, our example regions did not highlight cases where differences could be noted between the the two Bayesian models. In the revised version of the paper, we will extend Figure 6 to include further brain regions, such as those mentioned by the referee, and will use that as an opportunity to discuss the broader issue of what to do when the Bayesian models give conflicting results.

We agree with reviewer 2's point that the model description terminology could be made clearer for the target eLife audience. We tried to strike a balance between introducing the reader to the conventional technical terminology used in the Bayesian data analysis necessary for understanding the model while avoiding exhaustive statistical terminology. We erred too much on the side of the latter instead of providing clear links between the model construction and experimental data. In the revised version of the paper, we will augment any technical terms with more biological language and provide a Glossary for reader reference.

eLife Assessment

This study utilizes an elegant approach to examine valence encoding of the mesolimbic dopamine system. The findings are valuable, demonstrating differential responses of dopamine to the same taste stimulus according to its valence (i.e., appetitive or aversive) and in alignment with distinct behavioral responses. The evidence supporting the claims is convincing, resulting from a well-controlled experimental design with minimal confounds and thorough reporting of the data.

Reviewer #1 (Public review):

Summary:

Loh and colleagues investigate valence encoding in the mesolimbic dopamine system. Using an elegant approach, they show that sucrose, which normally evokes strong dopamine neuron activity and release in the nucleus accumbens, is made aversive via conditioned taste aversion, the same sucrose stimulus later evokes much less dopamine neuron activity and release. Thus, dopamine activity can dynamically track the changing valence of an unconditioned stimulus. These results are important for helping clarify valence and value related questions that are the matter of ongoing debate regarding dopamine functions in the field.

Strengths:

This is an elegant way to ask this question, the within subject's design and the continuity of the stimulus is a strong way to remove a lot of the common confounds that make it difficult to interpret valence-related questions. I think these are valuable studies that help tie up questions in the field while also setting up a number of interesting future directions. There are number of control experiments and tweaks to the design that help eliminate a number of competing hypotheses regarding the results. The data are clearly presented and contextualized.

Weaknesses for consideration:

The focus on one relatively understudied region of the rat striatum for dopamine recordings could potentially limit generalization of the findings. While this can be determined in future studies, the implications should be further discussed in the current manuscript.

Reviewer #2 (Public review):

Summary:

Koh et al. report an interesting manuscript studying dopamine binding in the lateral accumbens shell of rats across the course of conditioned taste aversion. The question being asked here is how does the dopamine system respond to aversion? The authors take advantage of unique properties of taste aversion learning (notably, within-subjects remapping of valence to the same physical stimulus) to address this.

They combine a well controlled behavioural design (including key, unpaired controls) with fibre photometry of dopamine binding via GrabDA and of dopamine neuron activity by gCaMP, careful analyses of behaviour (e.g., head movements; home cage ingestion), the authors show that, 1) conditioned taste aversion of sucrose suppresses the activity of VTA dopamine neurons and lateral shell dopamine binding to subsequent presentations of the sucrose tastant; 2) this pattern of activity was similar to the innately aversive tastant quinine; 3) dopamine responses were negatively correlated with behavioural (inferred taste reactivity) reactivity; and 4) dopamine responses tracked the contingency of between sucrose and illness because these responses recovered across extinction of the conditioned taste aversion.

Strengths:

There are important strengths here. The use of a well-controlled design, the measurement of both dopamine binding and VTA dopamine neuron activity, the inclusion of an extinction manipulation; and the thorough reporting of the data. I was not especially surprised by these results, but these data are a potentially important piece of the dopamine puzzle (e.g., as the authors note, salience-based argument struggles to explain these data).

Weaknesses for consideration:

(1) The focus here is on the lateral shell. This is a poorly investigated region in the context of the questions being asked here. Indeed, I suspect many readers might expect a focus on the medial shell. So, I think this focus is important. But, I think it does warrant greater attention in both the introduction and discussion. We do know from past work that there can be extensive compartmentalisation of dopamine responses to appetitive and aversive events and many of the inconsistent findings in the literature can be reconciled by careful examination of where dopamine is assessed. I do think readers would benefit from acknowledgement this - for example it is entirely reasonable to suppose that the findings here may be specific to the lateral shell.

(2) Relatedly, I think readers would benefit from an explicit rationale for studying the lateral shell as well as consideration of this in the discussion. We know that there are anatomical (PMID: 17574681), functional (PMID: 10357457), and cellular (PMID: 7906426) differences between the lateral shell and the rest of the ventral striatum. Critically, we know that profiles of dopamine binding during ingestive behaviours there can be highly dissimilar to the rest of ventral striatum (PMID: 32669355). I do think these points are worth considering.

(3) I found the data to be very thoughtfully analysed. But in places I was somewhat unsure:<br /> (a) Please indicate clearly in the text when photometry data show averages across trials versus when they show averages across animals.<br /> (b) I did struggle with the correlation analyses, for two reasons.<br /> (i) First, the key finding here is that the dopamine response to intraoral sucrose is suppressed by taste aversion. So, this will significantly restrict the range of dopamine transients, making interpretation of the correlations difficult.

(ii) Second, the authors report correlations by combining data across groups/conditions. I understand why the authors have done this, but it does risk obscuring differences between the groups. So, my question is: what happens to this trend when the correlations are computed separately for each group? I suspect other readers will share the same question. I think reporting these separate correlations would be very helpful for the field - regardless of the outcome.

(4) Figure 1A is not as helpful as it might be. I do think readers would expect a more precise reporting of GCaMP expression in TH+ and TH- neurons. I also note that many of the nuances in terms of compartmentalisation of dopamine signalling discussed above apply to ventral tegmental area dopamine neurons (e.g. medial v lateral) and this is worth acknowledging when interpreting.

Reviewer #3 (Public review):

Summary:

This study helps to clarify the mixed literature on dopamine responses to aversive stimuli. While it is well accepted that dopamine in the ventral striatum increases in response to various rewarding and appetitive stimuli, aversive stimuli have been shown to evoke phasic increases or decreasing depending on the exact aversive stimuli, behavioral paradigm, and/or dopamine recording method and location examined. Here the authors use a well-designed set of experiments to show differential responses to an appetitive primary reward (sucrose) that later becomes a conditioned aversive stimulus (sucrose previously paired with lithium chloride in a conditioned taste aversion paradigm). The results are interesting and add valuable data to the question of how the mesolimbic dopamine system encodes aversive stimuli, however, the conclusions are strongly stated given that the current data do not necessarily align with prior conflicting data in terms of recording location, and it is not clear exactly how to interpret the generally biphasic dopamine response to the CTA-sucrose which also evolves over exposures within a single session.

Strengths:

• The authors nicely demonstrate that their two aversive stimuli examined, quinine and sucrose following CTA, evoked aversive facial expressions and paw movements that differed from those following rewarding sucrose to support that the stimuli experienced by the rats differ in valence.

• Examined dopamine responses to the exact same sensory stimuli conditioned to have opposing valences, avoiding standard confounds of appetitive and aversive stimuli being sensed by different sensory modalities (i.e., sweet taste vs. electric shock).

• The authors examined multiple measurements of dopamine activity - cell body calcium (GCaMP6f) in midbrain and release in NAc (Grab-DA2h), which is useful as the prior mixed literature on aversive dopamine responses comes from a variety of recording methods.

• Correlations between sucrose preference and dopamine signals demonstrate behavioral relevance of the differential dopamine signals.

• The delayed testing experiment in Figure 7 nicely controls for the effect of time to demonstrate that the "rewarding" dopamine response to sucrose only recovers after multiple extinction sucrose exposures to extinguish the CTA.

Weaknesses for consideration:

• Regional differences in dopamine signaling to aversive stimuli are mentioned in the introduction and discussion. For instance, the idea that dopamine encodes salience is strongly argued against in the discussion, but the paper cited as arguing for that (Kutlu et al. 2021) is recording from the medial core in mice. Given other papers cited in the text about the regional differences in dopamine signaling in the NAc and from different populations of dopamine neurons in midbrain, it's important to mention this distinction wrt to salience signaling. Relatedly, the text says that the lateral NAc shell was targeted for accumbens recordings, but the histology figure looks like the majority of fibers were in the anterior lateral core of NAc. For the current paper to be a convincing last word on the issue, it would be extremely helpful to have similar recordings done in other parts of the NAc to do a more thorough comparison against other studies.

• Dopamine release in the NAc never dips below baseline for the conditioned sucrose. Is it possible to really consider this as a signal for valence per se, as opposed to it being a weaker response relative to the original sucrose response?

• Related to this, the main measure of the dopamine signal here, "mean z-score," obscures the temporal dynamics of the aversive dopamine response across a trial. This measure is used to claim that sucrose after CTA is "suppressing" dopamine neuron activity and release, which is true relative to the positive valence sucrose response. However, both GRAB-DA and cell-body GCaMP measurements show clear increases after onset of sucrose infusion before dipping back to baseline or slightly below in the average of all example experiments displayed. One could point to these data to argue either that aversive stimuli cause phasic increases in dopamine (due to the initial increase) or decreases (due to the delayed dip below baseline) depending on the measurement window. Some discussion of the dynamics of the response and how it relates to the prior literature would be useful.<br /> - Would this delayed below-baseline dip be visible with a shorter infusion time?<br /> - Does the max of the increase or the dip of the decrease better correlate with the behavioral measures of aversion (orofacial, paw movements) or sucrose preference than "mean z-score" measure used here?<br /> - The authors argue strongly in the discussion against the idea that dopamine is encoding "salience." Could this initial peak (also seen in the first few trials of quinine delivery, fig 1c color plot) be a "salience" response?

• Related to this, the color plots showing individual trials show a reduction in the increases to positive valence sucrose across conditioning day trials and a flip from infusion-onset increase to delayed increases across test day trials. This evolution across days makes it appear that the last few conditioning day trials would be impossible to discriminate from the first few test day trials in the CTA-paired. Presumably, from strength of CTA as a paradigm, the sucrose is already aversive to the animals at the first trial of test day. Why do the authors think the response evolves across this session?

• Given that most of the work is using a conditioned aversive stimulus, the comparison to a primary aversive tastant quinine is useful. However, the authors saw basically no dopamine response to a primary aversive tastant quinine (measured only with GRAB-DA) and saw less noticeable decreases following CTA for NAc recordings with GRAB-DA2h than with cell body GCaMP. Given that they are using the high-affinity version of the GRAB sensor, this calls into question whether this is a true difference in release vs. soma activity or issue of high affinity release sensor making decreases in dopamine levels more difficult to observe.

Author response:

Reviewer #1 (Public review):

Summary:

Loh and colleagues investigate valence encoding in the mesolimbic dopamine system. Using an elegant approach, they show that sucrose, which normally evokes strong dopamine neuron activity and release in the nucleus accumbens, is made aversive via conditioned taste aversion, the same sucrose stimulus later evokes much less dopamine neuron activity and release. Thus, dopamine activity can dynamically track the changing valence of an unconditioned stimulus. These results are important for helping clarify valence and value related questions that are the matter of ongoing debate regarding dopamine functions in the field.

Strengths:

This is an elegant way to ask this question, the within subject's design and the continuity of the stimulus is a strong way to remove a lot of the common confounds that make it difficult to interpret valence-related questions. I think these are valuable studies that help tie up questions in the field while also setting up a number of interesting future directions. There are number of control experiments and tweaks to the design that help eliminate a number of competing hypotheses regarding the results. The data are clearly presented and contextualized.

Weaknesses for consideration:

The focus on one relatively understudied region of the rat striatum for dopamine recordings could potentially limit generalization of the findings. While this can be determined in future studies, the implications should be further discussed in the current manuscript.

We agree that the manuscript would benefit from providing a stronger rationale for our recording sites and acknowledging the potential for regional differences in dopamine signaling. We have made the following additions to the manuscript:

Added to the Discussion: “Recordings were targeted to the lateral VTA and the corresponding approximate terminal site in the NAc lateral shell (Lammel et al., 2008). Subregional differences in dopamine activity likely contribute to mixed findings on dopamine and affect. For example, dopamine in the NAc lateral shell differentially encodes cues predictive of rewarding sucrose and aversive footshock, which is distinct from NAc medial shell dopamine responses (de Jong et al., 2019). Our findings are similar to prior work from our group targeting recordings to the NAc dorsomedial shell (Hsu et al., 2020; McCutcheon et al., 2012; Roitman et al., 2008): there, intraoral sucrose increased NAc dopamine release while the response in the same rats to quinine was significantly lower.”

Reviewer #2 (Public review):

Summary:

Koh et al. report an interesting manuscript studying dopamine binding in the lateral accumbens shell of rats across the course of conditioned taste aversion. The question being asked here is how does the dopamine system respond to aversion? The authors take advantage of unique properties of taste aversion learning (notably, within-subjects remapping of valence to the same physical stimulus) to address this.

They combine a well controlled behavioural design (including key, unpaired controls) with fibre photometry of dopamine binding via GrabDA and of dopamine neuron activity by gCaMP, careful analyses of behaviour (e.g., head movements; home cage ingestion), the authors show that, 1) conditioned taste aversion of sucrose suppresses the activity of VTA dopamine neurons and lateral shell dopamine binding to subsequent presentations of the sucrose tastant; 2) this pattern of activity was similar to the innately aversive tastant quinine; 3) dopamine responses were negatively correlated with behavioural (inferred taste reactivity) reactivity; and 4) dopamine responses tracked the contingency of between sucrose and illness because these responses recovered across extinction of the conditioned taste aversion.

Strengths:

There are important strengths here. The use of a well-controlled design, the measurement of both dopamine binding and VTA dopamine neuron activity, the inclusion of an extinction manipulation; and the thorough reporting of the data. I was not especially surprised by these results, but these data are a potentially important piece of the dopamine puzzle (e.g., as the authors note, salience-based argument struggles to explain these data).

Weaknesses for consideration:

(1) The focus here is on the lateral shell. This is a poorly investigated region in the context of the questions being asked here. Indeed, I suspect many readers might expect a focus on the medial shell. So, I think this focus is important. But, I think it does warrant greater attention in both the introduction and discussion. We do know from past work that there can be extensive compartmentalisation of dopamine responses to appetitive and aversive events and many of the inconsistent findings in the literature can be reconciled by careful examination of where dopamine is assessed. I do think readers would benefit from acknowledgement this - for example it is entirely reasonable to suppose that the findings here may be specific to the lateral shell.

As with our response to Reviewer 1, we agree that we should provide further rationale for focusing our recordings on the lateral shell and acknowledge potential differences in dopamine dynamics across NAc subregions. In addition to the changes in the Discussion detailed in our response to Reviewer 1, we have made the following additions to the Introduction:

Added to the Introduction: “NAc lateral shell dopamine differentially encodes cues predictive of rewarding (i.e., sipper spout with sucrose) and aversive stimuli (i.e., footshock), which is distinct from other subregions (de Jong et al., 2019). It is important to note that other regions of the NAc may serve as hedonic hotspots (e.g. dorsomedial shell; or may more closely align with the signaling of salience (e.g. ventromedial shell; (Yuan et al., 2021)).”

(2) Relatedly, I think readers would benefit from an explicit rationale for studying the lateral shell as well as consideration of this in the discussion. We know that there are anatomical (PMID: 17574681), functional (PMID: 10357457), and cellular (PMID: 7906426) differences between the lateral shell and the rest of the ventral striatum. Critically, we know that profiles of dopamine binding during ingestive behaviours there can be highly dissimilar to the rest of ventral striatum (PMID: 32669355). I do think these points are worth considering.

There are several reasons why dopamine dynamics were recorded in the NAc lateral shell:

(1) Dopamine neurons in more medial aspects of the VTA preferentially target the NAc medial shell and core whereas dopamine neurons in the lateral VTA – our target for VTA DA recordings – project to the lateral shell of the NAc (Lammel et al., 2008). Thus, our goal was to sample NAc release dynamics in areas that receive projections from our cell body recording sites.

(2) Cues predictive of reward availability (i.e., sipper spout with sucrose) and aversive stimuli (i.e., footshock) are differentially encoded by NAc lateral shell dopamine, which is distinct from NAc ventromedial shell dopamine responses (de Jong et al., 2019). These findings suggest a role for NAc lateral shell dopamine in the encoding of a stimulus’s valence, which made the subregion an area of interest for further examination.

(3) With respect to the medial NAc shell specifically, extensive literature had already shown it to be a ‘hedonic hotspot’ (Morales and Berridge, 2020; Yuan et al., 2021) whereas the ventral portion is more mixed with respect to valence (Yuan et al., 2021). We had previously shown that intraoral infusions of primary taste stimuli of opposing valence (i.e., sucrose and quinine) evoke differential responses in dopamine release within the NAc dorsomedial shell (Roitman et al., 2008). We more recently replicated differential dopamine responses from dopamine cell bodies in the lateral VTA (Hsu et al., 2020) and thus endeavored to the possibility of changing dopamine responses in the lateral VTA to the same stimulus as its valence changes. As a result of these choices, measuring dopamine release in the lateral shell was a logical choice. The field would greatly benefit from continued future work surveying the entirety of the VTA DA projection terminus. 

We have included these points of justification in the Introduction and Discussion sections.

(3) I found the data to be very thoughtfully analysed. But in places I was somewhat unsure:

(a) Please indicate clearly in the text when photometry data show averages across trials versus when they show averages across animals.

We have now explicitly indicated in the figure legends of Figures 1, 3, 5, 7, and 8:

(1) In heat maps, each row represents the averaged (across rats) response on that trial.

(2) Traces below heat maps represent the response to infusion averaged first across trials for each rat and then across all rats.

(3) Insets represent the average z-score across the infusion period averaged first across all trials for each rat and then across all rats.

(b) I did struggle with the correlation analyses, for two reasons.

(i) First, the key finding here is that the dopamine response to intraoral sucrose is suppressed by taste aversion. So, this will significantly restrict the range of dopamine transients, making interpretation of the correlations difficult.

The overall hypothesis is that the dopamine response would correlate with the valence of a taste stimulus – even and especially when the stimulus remained constant but its valence changed. We inferred valence from the behavioral reactivity to the stimulus – reasoning that an appetitive taste will evoke minimal movement of the nose and paws (presumably because the animals are primarily engaging in small mouth movements associated with ingestion as shown by the seminal work of Grill and Norgren (1978) and the many studies published by the K.C. Berridge group) whereas an aversive taste will evoke significantly more movement as the rats engage in rejection responses (e.g. forelimb flails, chin rubs, etc.). When we conducted our regression analyses we endeavored to be as transparent as possible and labeled each symbol based on group (Unpaired vs Paired) and day (Conditioning vs Test). Both behavioral reactivity and dopamine responses change – but only for the Paired rats across days. In this sense, we believe the interpretation is clear. However, the Reviewer raises an important criticism that there would essentially be a floor effect with dopamine responses. We believe this is mitigated by data acquired across extinction and especially in Figure 9B. Here, the observations that dopamine responses fall to near zero but return to pre-conditioning levels in the Paired group with strong correlation between dopamine and behavioral reactivity throughout would hopefully partially allay the Reviewer’s concerns. See Part ii below for further support.

(ii) Second, the authors report correlations by combining data across groups/conditions. I understand why the authors have done this, but it does risk obscuring differences between the groups. So, my question is: what happens to this trend when the correlations are computed separately for each group? I suspect other readers will share the same question. I think reporting these separate correlations would be very helpful for the field -

regardless of the outcome.

To address this concern, we performed separate regression analyses for Paired and Unpaired rats and provide the table below to detail results where data were combined across groups or separated. Expectedly, all analyses in Paired rats indicated a significant inverse relationship between dopamine and behavioral reactivity. Afterall, it is only in this group where behavioral reactivity to the taste stimulus changes as function of conditioning. Perhaps even more striking is that in almost all comparisons, even when restricting the regression analysis to Unpaired rats, we still observed a significant inverse relationship between dopamine and behavioral reactivity in most experiments. We have outlined the separated correlations below (asterisks denote slopes significantly different from 0; * p<0.05; ** p<0.01; *** p<0.005; **** p<0.001):

Author response table 1.

(4) Figure 1A is not as helpful as it might be. I do think readers would expect a more precise reporting of GCaMP expression in TH+ and TH- neurons. I also note that many of the nuances in terms of compartmentalisation of dopamine signalling discussed above apply to ventral tegmental area dopamine neurons (e.g. medial v lateral) and this is worth acknowledging when interpreting t

Others have reported (Choi et al., 2020) and quantified (Hsu et al., 2020) GCaMP6f expression in TH+ neurons. While we didn’t report these quantifications, our observations were very much in line with previous quantifications from our laboratory (Hsu et al. 2020).

We agree that we should elaborate on VTA subregional differences and have answered this response above (See responses to Reviewer 1 Weakness #1 and Reviewer 2 Weakness #2).

Reviewer #3 (Public review):

Summary:

This study helps to clarify the mixed literature on dopamine responses to aversive stimuli. While it is well accepted that dopamine in the ventral striatum increases in response to various rewarding and appetitive stimuli, aversive stimuli have been shown to evoke phasic increases or decreasing depending on the exact aversive stimuli, behavioral paradigm, and/or dopamine recording method and location examined. Here the authors use a well-designed set of experiments to show differential responses to an appetitive primary reward (sucrose) that later becomes a conditioned aversive stimulus (sucrose previously paired with lithium chloride in a conditioned taste aversion paradigm). The results are interesting and add valuable data to the question of how the mesolimbic dopamine system encodes aversive stimuli, however, the conclusions are strongly stated given that the current data do not necessarily align with prior conflicting data in terms of recording location, and it is not clear exactly how to interpret the generally biphasic dopamine response to the CTA-sucrose which also evolves over exposures within a single session.

Strengths:

• The authors nicely demonstrate that their two aversive stimuli examined, quinine and sucrose following CTA, evoked aversive facial expressions and paw movements that differed from those following rewarding sucrose to support that the stimuli experienced by the rats differ in valence.

• Examined dopamine responses to the exact same sensory stimuli conditioned to have opposing valences, avoiding standard confounds of appetitive and aversive stimuli being sensed by different sensory modalities (i.e., sweet taste vs. electric shock)

• The authors examined multiple measurements of dopamine activity - cell body calcium (GCaMP6f) in midbrain and release in NAc (Grab-DA2h), which is useful as the prior mixed literature on aversive dopamine responses comes from a variety of recording methods.

• Correlations between sucrose preference and dopamine signals demonstrate behavioral relevance of the differential dopamine signals.

• The delayed testing experiment in Figure 7 nicely controls for the effect of time to demonstrate that the "rewarding" dopamine response to sucrose only recovers after multiple extinction sucrose exposures to extinguish the CTA.

Weaknesses for consideration:

(1) Regional differences in dopamine signaling to aversive stimuli are mentioned in the introduction and discussion. For instance, the idea that dopamine encodes salience is strongly argued against in the discussion, but the paper cited as arguing for that (Kutlu et al. 2021) is recording from the medial core in mice. Given other papers cited in the text about the regional differences in dopamine signaling in the NAc and from different populations of dopamine neurons in midbrain, it's important to mention this distinction wrt to salience signaling. Relatedly, the text says that the lateral NAc shell was targeted for accumbens recordings, but the histology figure looks like the majority of fibers were in the anterior lateral core of NAc. For the current paper to be a convincing last word on the issue, it would be extremely helpful to have similar recordings done in other parts of the NAc to do a more thorough comparison against other studies.

As the Reviewer notes, NAc dopamine recordings were aimed at the lateral NAc shell. It is possible that some dopamine neurons lying within the anterior lateral core were recorded. Fiber photometry and the size of the fiber optics cannot definitively identify the precise location and number of dopamine neurons from which we recorded. Still, recording sites did not systematically differ between groups. Further, the within-subjects design helps to mitigate any potential biases for one subregion over another. The results presented in the manuscript strongly support a valence code. It is difficult to be the ‘last word’ on this topic and we suspect debate will continue. We used taste stimuli for appetitive and aversive stimuli – whereas many in the field will continue to use other noxious stimuli (e.g. foot shock) that likely recruit different circuits en route to the VTA. And there may very well be a different regional profile for dopamine signaling with different noxious stimuli. Moreover, we used intraoral infusion to avoid confounds of stimulus avoidance and competing motivations (e.g. food or fluid deprivation). We believe that this is one of the most important and unique features of our report. Recent work supports a role for phasic increases in dopamine in avoidance of noxious stimuli (Jung et al., 2024) and it will be critical for the field to reflect on the differences between avoidance and aversion. Moreover, in ongoing studies we aspire to fully survey dopamine signaling in conditioned taste aversion across the medial-lateral and dorsal-ventral axes of the VTA and NAc.

(2) Dopamine release in the NAc never dips below baseline for the conditioned sucrose. Is it possible to really consider this as a signal for valence per se, as opposed to it being a weaker response relative to the original sucrose response?

Indeed, NAc dopamine release to intraoral quinine nor aversive sucrose doesn’t dip below baseline but rather dopamine binding doesn’t change from pre-infusion baseline levels. It should be noted that VTA dopamine cell body activity does indeed dip below baseline in response to aversive sucrose. Moreover, using fast-scan cyclic voltammetry, we showed that dopamine release dips below baseline in the NAc dorsomedial shell in response to intraoral quinine (Roitman et al., 2008). The differences across recording sites may reflect regional differences but they may also reflect differences in recording approaches. GrabDA2h, used here, has relatively slow kinetics that may obscure dips below baseline (see response Weakness# 8 below).

(3) Related to this, the main measure of the dopamine signal here, "mean z-score," obscures the temporal dynamics of the aversive dopamine response across a trial. This measure is used to claim that sucrose after CTA is "suppressing" dopamine neuron activity and release, which is true relative to the positive valence sucrose response. However, both GRAB-DA and cell-body GCaMP measurements show clear increases after onset of sucrose infusion before dipping back to baseline or slightly below in the average of all example experiments displayed. One could point to these data to argue either that aversive stimuli cause phasic increases in dopamine (due to the initial increase) or decreases (due to the delayed dip below baseline) depending on the measurement window. Some discussion of the dynamics of the response and how it relates to the prior literature would be useful.

We have used mean z-score to do much of our quantitative analyses but the Reviewer raises the intriguing possibility that we are masking an initial increase in dopamine release and VTA DA activity evoked by aversive taste by doing so. We included the heat maps in the manuscript to be as transparent as possible about the time course of dopamine responses – both within a trial and across trials. The Reviewer’s point prompted us to reflect further on the heat maps and recognize that trials early in the session often showed a brief increase in dopamine for aversive sucrose but this response dissipated (NAc dopamine release) or flipped (VTA DA cell body activity) over trials. We now quantitatively characterize this feature by looking at the timecourse of dopamine responses in each third of the trials (1-10, 11-20, 21-30; see Author response images 1,2 and 3). As we infer the valence of the stimulus from nose and paw movements (behavioral reactivity), it is especially striking that we a similar timecourse for changes in behavior. Collectively, the data may reflect an updating process that is relatively slow and requires experience of the stimulus in a new (aversive) state – that is, a model-free process. While our experiments were not designed to test the updating of dopamine responses and discern their participation in model-based versus model-free learning processes – another debate in the dopamine field (Cone et al., 2016; Deserno et al., 2021)– the data reflect a model-free process. This is further supported in the experiment involving multiple conditioning sessions, where dopamine ‘dips’ are observed in trials 1-10 on Conditioning Day 3 and Extinction Day 1 when the new value of sucrose has been established. Finally, the relatively slow updating of the value of sucrose is reflected in older literature using a continuous intraoral infusion. Using this approach, rats began rejecting the saccharin infusion only after ~2min rather than immediately (Schafe et al., 1998; Schafe and Bernstein, 1996; Wilkins and Bernstein, 2006).   

Author response image 1.

Author response image 2.

Author response image 3.

(4) Would this delayed below-baseline dip be visible with a shorter infusion time?

While our experiments did not explore this parameter, it would be interesting to parametrically vary infusion duration times and examine differences in dopamine responses. However, we believe the most parsimonious explanation is that the ‘dip’ in VTA cell body activity develops as a function of the slow updating of the value of sucrose reflective of a model-free process. We recognize that this is mere speculation.

(5) Does the max of the increase or the dip of the decrease better correlate with the behavioral measures of aversion (orofacial, paw movements) or sucrose preference than "mean z-score" measure used here?

It seems plausible that finding the most extreme value from baseline could better correlate to behavioral measures. Time courses to max increase and max decrease are different. Moreover, with appetitive sucrose, there are often multiple transients that occur throughout a single intraoral infusion. Coupled with a noisy time course for individual components of behavioral reactivity, we determined that averaging data across the whole infusion period (i.e. mean z-score) was the most objective way we could analyze the dopamine and behavioral responses to taste stimuli.

(6) The authors argue strongly in the discussion against the idea that dopamine is encoding "salience." Could this initial peak (also seen in the first few trials of quinine delivery, fig 1c color plot) be a "salience" response?

Our response above to the potential for ‘mixed’ dopamine responses to aversive sucrose led to additional analyses that support a slow updating of both behavior and dopamine to the new, aversive value of sucrose. Quinine is innately aversive and thus the Reviewer rightly points out that even here we observe an increase in dopamine release evoked by quinine on the first few trials (as observed in the heat map). We’d like to note, though, that the order of stimulus exposure was counterbalanced across rats. In those rats first receiving a sucrose session, quinine initially caused a modest increase in dopamine release during the first 10 trials (which is more pronounced in the first 2 trials). In the subsequent 2 blocks of 10 trials, no such increase was observed. Interestingly, in rats for which quinine was their first stimulus, we did not see an increase in dopamine release on the first few trials (see Author response image 4). We speculate that the initial sucrose session required the value of intraoral infusions to be updated when quinine was delivered to these rats and that, once more, the updating process may be slow and akin to a model-free process. This analysis, at present, is underpowered but will direct future attention in follow-up work.

Author response image 4.

(7) Related to this, the color plots showing individual trials show a reduction in the increases to positive valence sucrose across conditioning day trials and a flip from infusion-onset increase to delayed increases across test day trials. This evolution across days makes it appear that the last few conditioning day trials would be impossible to discriminate from the first few test day trials in the CTA-paired. Presumably, from strength of CTA as a paradigm, the sucrose is already aversive to the animals at the first trial of test day. Why do the authors think the response evolves across this session?

As the Reviewer noted, Points 3-7 are related. We have speculated that the evolving dopamine response in Paired rats across test day trials reflects a model-free process. Importantly, as in the manuscript, our additional analyses once again show a tight relationship between behavioral reactivity and the dopamine response across the test session trials. It is important to note, though, that these experiments were not designed to test if responses reflect model-free or model-based processes.

(8) Given that most of the work is using a conditioned aversive stimulus, the comparison to a primary aversive tastant quinine is useful. However, the authors saw basically no dopamine response to a primary aversive tastant quinine (measured only with GRAB-DA) and saw less noticeable decreases following CTA for NAc recordings with GRAB-DA2h than with cell body GCaMP. Given that they are using the high-affinity version of the GRAB sensor, this calls into question whether this is a true difference in release vs. soma activity or issue of high affinity release sensor making decreases in dopamine levels more difficult to observe.

We share the same speculation as the Reviewer. Using fast-scan cyclic voltammetry, albeit measuring dopamine concentration in the dorsomedial shell, we observed a clear decrease from baseline with intraoral infusions of quinine (Roitman et al., 2008). Using fiber photometry here, the Reviewer and we note that GRAB_DA2h is a high-affinity (i.e., EC50: 7nM) dopamine sensor with relatively long off-kinetics (i.e., t1/2 decay time: 7300ms) (Labouesse et al., 2020). It may therefore be much more difficult to observe decreases (below baseline) using this sensor. The publication of new dopamine sensors - with lower affinity, faster kinetics, and greater dynamic range (Zhuo et al., 2024) – introduces opportunities for comparison and the greater potential for capturing decreases below baseline. Due to the poorer kinetics associated with GRAB_DA2h, we would not assert that direct comparisons between the GCaMP- and GRAB-based signals observed here represent true differences between somatic and terminal activity.

References

Choi JY, Jang HJ, Ornelas S, Fleming WT, Fürth D, Au J, Bandi A, Engel EA, Witten IB. 2020. A Comparison of Dopaminergic and Cholinergic Populations Reveals Unique Contributions of VTA Dopamine Neurons to Short-Term Memory. Cell Rep 33. doi:10.1016/j.celrep.2020.108492

Cone JJ, Fortin SM, McHenry JA, Stuber GD, McCutcheon JE, Roitman MF. 2016. Physiological state gates acquisition and expression of mesolimbic reward prediction signals. Proc Natl Acad Sci U S A 113. doi:10.1073/pnas.1519643113

de Jong JW, Afjei SA, Pollak Dorocic I, Peck JR, Liu C, Kim CK, Tian L, Deisseroth K, Lammel S. 2019. A Neural Circuit Mechanism for Encoding Aversive Stimuli in the Mesolimbic Dopamine System. Neuron 101. doi:10.1016/j.neuron.2018.11.005

Deserno L, Moran R, Michely J, Lee Y, Dayan P, Dolan RJ. 2021. Dopamine enhances model-free credit assignment through boosting of retrospective model-based inference. Elife 10. doi:10.7554/eLife.67778

Hsu TM, Bazzino P, Hurh SJ, Konanur VR, Roitman JD, Roitman MF. 2020. Thirst recruits phasic dopamine signaling through subfornical organ neurons. Proc Natl Acad Sci U S A 117:30744–30754. doi:10.1073/PNAS.2009233117/-/DCSUPPLEMENTAL

Jung K, Krüssel S, Yoo S, An M, Burke B, Schappaugh N, Choi Y, Gu Z, Blackshaw S, Costa RM, Kwon HB. 2024. Dopamine-mediated formation of a memory module in the nucleus accumbens for goal-directed navigation. Nat Neurosci. doi:10.1038/s41593-024-01770-9

Labouesse MA, Cola RB, Patriarchi T. 2020. GPCR-based dopamine sensors—A detailed guide to inform sensor choice for in vivo imaging. Int J Mol Sci. doi:10.3390/ijms21218048

Lammel S, Hetzel A, Häckel O, Jones I, Liss B, Roeper J. 2008. Unique Properties of Mesoprefrontal Neurons within a Dual Mesocorticolimbic Dopamine System. Neuron 57. doi:10.1016/j.neuron.2008.01.022

McCutcheon JE, Ebner SR, Loriaux AL, Roitman MF, Tobler PN. 2012. Encoding of aversion by dopamine and the nucleus accumbens. Front Neurosci 6. doi:10.3389/fnins.2012.00137

Morales I, Berridge KC. 2020. ‘Liking’ and ‘wanting’ in eating and food reward: Brain mechanisms and clinical implications. Physiol Behav. doi:10.1016/j.physbeh.2020.113152

Roitman MF, Wheeler RA, Wightman RM, Carelli RM. 2008. Real-time chemical responses in the nucleus accumbens differentiate rewarding and aversive stimuli. Nature Neuroscience 2008 11:12 11:1376–1377. doi:10.1038/nn.2219

Schafe GE, Bernstein IL. 1996. Forebrain contribution to the induction of a brainstem correlate of conditioned taste aversion: I. The amygdala. Brain Res 741. doi:10.1016/S0006-8993(96)00906-7

Schafe GE, Thiele TE, Bernstein IL. 1998. Conditioning method dramatically alters the role of amygdala in taste aversion learning. Learning and Memory 5. doi:10.1101/lm.5.6.481

Wilkins EE, Bernstein IL. 2006. Conditioning method determines patterns of c-fos expression following novel taste-illness pairing. Behavioural Brain Research 169. doi:10.1016/j.bbr.2005.12.006

Yuan L, Dou YN, Sun YG. 2021. Topography of reward and aversion encoding in the mesolimbic dopaminergic system. Journal of Neuroscience 39. doi:10.1523/JNEUROSCI.0271-19.2019

Zhuo Y, Luo B, Yi X, Dong H, Miao X, Wan J, Williams JT, Campbell MG, Cai R, Qian T, Li F, Weber SJ, Wang L, Li B, Wei Y, Li G, Wang H, Zheng Y, Zhao Y, Wolf ME, Zhu Y, Watabe-Uchida M, Li Y. 2024. Improved green and red GRAB sensors for monitoring dopaminergic activity in vivo. Nat Methods 21. doi:10.1038/s41592-023-02100-w

Author response:

Reviewer #1:

We agree with Reviewer 1 that the flexibility of SPRAWL also makes it difficult to interpret its outputs. We consider SPRAWL to be a hypothesis-generation tool to answer simple questions of subcellular localization in a statistically robust manner. In this paper we include examples of how it can be incorporated with other tools and wetlab experimentation to build biological intuition. Our hope is that the SPRAWL software, or even the underlying simple statistical ideas are of use to others in the field.

Reviewer #2:

We agree with Reviewer #2 that this manuscript does not demonstrate biological significance of the observed results of applying SPRAWL to massively multiplexed FISH datasets. We agree it would require additional wetlab experiments such as cell-type specific and isoform-resolved fluorescence in-situ hybridization, which we consider beyond the scope of this paper. We believe that the observed correlations of subcellular localization detected by SPRAWL and the differential 3’ UTR usage detected by ReadZS are compelling, although not conclusive, as are the Timp3 experimental studies.

Our understanding is that Baysor is primarily a cell-segmentation algorithm, which is not what SPRAWL attempts to achieve. Baysor states that it identifies “cells of a distinct type will give rise to small molecular neighborhoods with stereotypical transcriptional composition, making it possible to interpret such neighborhoods without performing explicit cell segmentation” which we understand to mean that Baysor identifies spatial groupings of cells with “stereotypical transcriptional composition” rather than subcellular RNA localization. We do not think that SPRAWL and Baysor are comparable, but instead Baysor could be used as an upstream step to SPRAWL to potentially improve cell segmentation.

Reviewer #3:

We thank Reviewer #3 for identifying discrepancies in the paper which we addressed to the best of our abilities.

Author response:

Reviewer 1:

Many thanks for your positive review and clear overview of our paper. We also agree with your interpretation of our results that ‘the information that is decodable and the information that is task-relevant may relate in very different ways’ and we could have emphasised this point more in the paper.

With regards to the qualitative similarities between our models and our data, we agree that due to the fact that one can achieve any desired level of activity, decoding accuracy, performance, etc in a model, we focussed on changes over learning of key metrics that are commonly used in the field. Although this can appear qualitative at times because the raw values can differ between the data and our models, our main results are ultimately strongly quantitative (e.g., Fig. 3c,d, and Fig. 5f). We note that we could have fine tuned the models to have similar activity levels, decoding accuracies etc to our data, and on the face of it this may have made the results appear more convincing, but we felt that such trivial fine tuning does not change any of our key results in any fundamental way and is not the aim of computational modelling. The model one chooses to analyse will always be abstracted from biology in some way, by definition.

Reviewer 2:

Thank you very much for your kind comments and clear overview of our paper. We also hope that our paper ‘provides a valuable analysis of the effect of two parameters on representations of irrelevant stimuli in trained RNNs.’

With regards to our suggested mechanism of suppressing dynamically irrelevant stimuli, we are sorry that we did not provide a sufficient enough explanation of suppressing color representations when they are irrelevant. We hopefully provide a longer explanation here. Our mechanism of suppression of dynamically irrelevant stimuli does not suggest that it becomes un-suppressed later, only the behaviourally relevant variable should be decodable when it is needed (i.e., XOR). Although color decodability did increase slightly in the data and some of the models from the color period to the shape period, it was typically not significant and was therefore not a result that we emphasise in the paper (although this could be analysed further to see if additional mechanisms might explain it). We emphasise throughout that color decoding is typically similar between color and shape periods (either high or low) and either decreases or increases over time in both periods. We also focus on whether color decodability increases or decreases over learning during the color period when it is irrelevant (which we call ‘early color decoding’). Importantly, decoding of color or shape is not needed to perform the task, only decoding of XOR is needed to perform the task. For example, in our two-neuron networks, we observe perfect XOR decoding and only 75% decoding of color and shape, and decoding during the shape period is the same as the network at initialisation before any training. The mechanism we suggest of suppressing dynamically irrelevant stimuli does not predict that that stimulus should be un-suppressed later, only the behaviourally relevant variable should be decodable (i.e., XOR). Instead, what we try to explain is that color inputs can generate 0 firing rate during the color period, when that input does not need to be used and is therefore irrelevant (and color decoding decreases during the color period over learning), but these inputs can be combined with shape inputs later to create a perfectly decodable XOR response.

With regards to interpretation of our results based on metabolic cost constraints, we feel that this is an unnecessarily strong criticism to say that it ‘is not backed up by the presented data/analyses.’ All of our models were trained with only a metabolic cost constraint, a noise strength, and a task performance term. Therefore, the results of the models are directly attributable to the strength of metabolic cost that we use. Additionally, although one could in principle pick any of infinitely many different parameters to change and measure the response in an optimized network, varying metabolic cost and noise are two of the most fundamental phenomena that neural circuits must contend with, and many studies have analysed the impact they have on neural circuit dynamics. Furthermore, in line with previous studies (Yang et al., 2019, Whittington et al., 2022, Sussillo et al., 2015, Orhan et al., 2019, Kao et al., 2021, Cueva et al., 2020, Driscoll et al., 2022, Song et al., 2016, Masse et al., 2019, Schimel et al., 2023), we operationalized metabolic cost in our models through L2 firing rate regularization. This cost penalizes high overall firing rates. (Such an operationalization of metabolic cost also makes sense for our models because network performance is based on firing rates rather than subthreshold activities.) There are however alternative conceivable ways to operationalize a metabolic cost; for example L1 firing rate regularization has been used previously when optimizing neural networks and promotes more sparse neural firing. Interestingly, although our L2 is generally conceived to be weaker than L1 regularization, we still found that it encouraged the network to use purely sub-threshold activity in our task. The regularization of synaptic weights may also be biologically relevant because synaptic transmission uses the most energy in the brain compared to other processes (Faria-Pereira et al., 2022, Harris et al., 2012). Additionally, even subthreshold activity could be regularized as it also consumes energy (although orders of magnitude less than spiking (Zhu et al., 2019)). Therefore, future work will be needed to examine how different metabolic costs affect the dynamics of task-optimized networks.

With regards to color representations in PFC only qualitatively matching those in our models, in line with the comment from Reviewer 1, we agree that due to the fact that one can achieve any desired level of activity, decoding accuracy, performance, etc in a model, we focussed on changes over learning of key metrics that are commonly used in the field. Although this can appear qualitative at times because the raw values can differ between the data and our models, our main results are ultimately strongly quantitative (e.g., Fig. 3c,d, and Fig. 5f). We note that we could have fine tuned the models to have similar activity levels, decoding accuracies etc to our data, and on the face of it this may have made the results appear more convincing, but we felt that such trivial fine tuning does not change any of our key results in any fundamental way and is not the aim of computational modelling. The model one chooses to analyse will always be abstracted from biology in some way, by definition. Finally, of course we note that changes in color decoding could result from other causes, but we focussed on two key phenomena that neural circuits must contend with: noise and metabolic costs. Therefore, it is likely that these two variables play a strong role in stimulus representations in neural circuits

Reviewer 3:

Thank you very much for your thorough and clear overview of our paper and we agree that it is important to investigate phenomena and manipulations in computational models that are almost impossible to do in vivo and we are pleased you found our mathematical analyses rigorous and nicely documented.

Although we agree that it can be useful to study the responses of individual neurons, we focussed on population analyses of all available neurons without omitting or specifically selecting neurons based on their dynamics. We are also not suggesting that the activities of individual ‘neurons’ in the models and data should be similar since our models are highly abstract firing rate models. But rather, the overall computational strategy, which one can access through population decoding and cross-generalised decoding, was what we were interested in comparing between the models and the data and is arguably the correct level of analysis of such models (an data) given our key questions (Vyas et al., 2020, Churchland et al., 2012, Mante et al., 2013, Ebitz et al., 2021).

We also certainly agree and are more than open to the fact that suppression of irrelevant stimuli may already be happening on the inputs arriving in PFC. Indeed, we actually suggest this as the mechanism in Fig. 5 (together with recurrent circuit dynamics that make use of these inputs).

With regards to the dynamics of the two-neuron networks not being ‘informative of what happens in brain networks’, we agree that these models are very simplified and may only contain very fundamental similarities with biological neurons. However, we only used them to illustrate the fundamental mechanism of generating 0 firing rate during the color epoch so that it is more easily understandable for readers as they can see the entire 2-dimensional state space and the entire computational strategy can be seen (Fig. 5a-d). We also note that we did this for both rectified linear and tanh networks, thus showing that such a mechanism is preserved across fundamentally different firing rate nonlinearities. Additionally, after illustrating this fundamental mechanism of networks receiving color information but generating 0 firing rate, we show that the exact same mechanism is at play in the large networks we use throughout the paper (Fig. 5e). We also only compare the large networks to our neural recordings. We do agree though that it would be interesting to further compare fundamental similarities and differences between our models and our neural recordings (always at the right level of analysis that makes sense for our chosen models) to show that the mechanisms we uncover in our models are also strongly relevant for our data.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors have used full-length single-cell sequencing on a sorted population of human fetal retina to delineate expression patterns associated with the progression of progenitors to rod and cone photoreceptors. They find that rod and cone precursors contain a mix of rod/cone determinants, with a bias in both amounts and isoform balance likely deciding the ultimate cell fate. Markers of early rod/cone hybrids are clarified, and a gradient of lncRNAs is uncovered in maturing cones. Comparison of early rods and cones exposes an enriched MYCN regulon, as well as expression of SYK, which may contribute to tumor initiation in RB1 deficient cone precursors.

Strengths:

(1) The insight into how cone and rod transcripts are mixed together at first is important and clarifies a long-standing notion in the field.

(2) The discovery of distinct active vs inactive mRNA isoforms for rod and cone determinants is crucial to understanding how cells make the decision to form one or the other cell type. This is only really possible with full-length scRNAseq analysis.

(3) New markers of subpopulations are also uncovered, such as CHRNA1 in rod/cone hybrids that seem to give rise to either rods or cones.

(4) Regulon analyses provide insight into key transcription factor programs linked to rod or cone fates.

(5) The gradient of lncRNAs in maturing cones is novel, and while the functional significance is unclear, it opens up a new line of questioning around photoreceptor maturation.

(6) The finding that SYK mRNA is naturally expressed in cone precursors is novel, as previously it was assumed that SYK expression required epigenetic rewiring in tumors.

Weaknesses:

(1) The writing is very difficult to follow. The nomenclature is confusing and there are contradictory statements that need to be clarified.

(2) The drug data is not enough to conclude that SYK inhibition is sufficient to prevent the division of RB1 null cone precursors. Drugs are never completely specific so validation is critical to make the conclusion drawn in the paper.

We thank the reviewer for describing the study’s strengths and weaknesses.  In the upcoming revision, we will:

(1) improve the writing and clarify the nomenclature and contradictory statements, particularly those noted in the Reviewer’s Recommendations for Authors; and

(2) scale back the claims related to the role of SYK in the cone precursor response to RB1 loss; we agree that genetic perturbation of SYK is required to prove it’s role and will perform such analyses in a separate study.

Reviewer #2 (Public review):

Summary:

The authors used deep full-length single-cell sequencing to study human photoreceptor development, with a particular emphasis on the characteristics of photoreceptors that may contribute to retinoblastoma.

Strengths:

This single-cell study captures gene regulation in photoreceptors across different developmental stages, defining post-mitotic cone and rod populations by highlighting their unique gene expression profiles through analyses such as RNA velocity and SCENIC. By leveraging full-length sequencing data, the study identifies differentially expressed isoforms of NRL and THRB in L/M cone and rod precursors, illustrating the dynamic gene regulation involved in photoreceptor fate commitment. Additionally, the authors performed high-resolution clustering to explore markers defining developing photoreceptors across the fovea and peripheral retina, particularly characterizing SYK's role in the proliferative response of cones in the RB loss background. The study provides an in-depth analysis of developing human photoreceptors, with the authors conducting thorough analyses using full-length single-cell RNA sequencing. The strength of the study lies in its design, which integrates single-cell full-length RNA-seq, long-read RNA-seq, and follow-up histological and functional experiments to provide compelling evidence supporting their conclusions. The model of cell type-dependent splicing for NRL and THRB is particularly intriguing. Moreover, the potential involvement of the SYK and MYC pathways with RB in cone progenitor cells aligns with previous literature, offering additional insights into RB development.

Weaknesses:

The manuscript feels somewhat unfocused, with a lack of a strong connection between the analysis of developing photoreceptors, which constitutes the bulk of the manuscript, and the discussion on retinoblastoma. Additionally, given the recent publication of several single-cell studies on the developing human retina, it is important for the authors to cross-validate their findings and adjust their statements where appropriate.

We thank the reviewer for summarizing the main findings and for noting the compelling support for the conclusions, the intriguing cell type-dependent splicing of rod and cone lineage factors, and the insights into retinoblastoma development. 

We concur that some studies of developing photoreceptors were not well connected to retinoblastoma, which diminished the focus.  However, we suggest that it was valuable to highlight how deep, long read sequencing provided new insights into retinoblastoma. For example, our demonstration of similar rod- and cone-related gene expression in early cones and RB cells addressed concerns with the proposed cone cell-of-origin, adding disease relevance.

We will address the Reviewer’s request to cross-validate our findings with those of other single-cell studies of developing human retina and to adjust the related statements in our upcoming revision.

Reviewer #3 (Public review):

Summary:

The authors use high-depth, full-length scRNA-Seq analysis of fetal human retina to identify novel regulators of photoreceptor specification and retinoblastoma progression.

Strengths:

The use of high-depth, full-length scRNA-Seq to identify functionally important alternatively spliced variants of transcription factors controlling photoreceptor subtype specification, and identification of SYK as a potential mediator of RB1-dependent cell cycle reentry in immature cone photoreceptors.

Human developing fetal retinal tissue samples were collected between 13-19 gestational weeks and this provides a substantially higher depth of sequencing coverage, thereby identifying both rare transcripts and alternative splice forms, and thereby representing an important advance over previous droplet-based scRNA-Seq studies of human retinal development.

Weaknesses:

The weaknesses identified are relatively minor. This is a technically strong and thorough study, that is broadly useful to investigators studying retinal development and retinoblastoma.

We thank the reviewer for describing the strengths of the study. Our upcoming revision will address the minor concerns that were raised separately in the Reviewer’s Recommendations for Authors.

eLife Assessment

In this paper, the authors use single-cell RNA sequencing to understand post-mitotic cone and rod developmental states and identify cone-specific features that contribute to retinoblastoma genesis. The work is important and the evidence is generally convincing. The findings of rod/cone fate determination at a very early stage are intriguing.

Reviewer #1 (Public review):

Summary:

The authors have used full-length single-cell sequencing on a sorted population of human fetal retina to delineate expression patterns associated with the progression of progenitors to rod and cone photoreceptors. They find that rod and cone precursors contain a mix of rod/cone determinants, with a bias in both amounts and isoform balance likely deciding the ultimate cell fate. Markers of early rod/cone hybrids are clarified, and a gradient of lncRNAs is uncovered in maturing cones. Comparison of early rods and cones exposes an enriched MYCN regulon, as well as expression of SYK, which may contribute to tumor initiation in RB1 deficient cone precursors.

Strengths:

(1) The insight into how cone and rod transcripts are mixed together at first is important and clarifies a long-standing notion in the field.

(2) The discovery of distinct active vs inactive mRNA isoforms for rod and cone determinants is crucial to understanding how cells make the decision to form one or the other cell type. This is only really possible with full-length scRNAseq analysis.

(3) New markers of subpopulations are also uncovered, such as CHRNA1 in rod/cone hybrids that seem to give rise to either rods or cones.

(4) Regulon analyses provide insight into key transcription factor programs linked to rod or cone fates.

(5) The gradient of lncRNAs in maturing cones is novel, and while the functional significance is unclear, it opens up a new line of questioning around photoreceptor maturation.

(6) The finding that SYK mRNA is naturally expressed in cone precursors is novel, as previously it was assumed that SYK expression required epigenetic rewiring in tumors.

Weaknesses:

(1) The writing is very difficult to follow. The nomenclature is confusing and there are contradictory statements that need to be clarified.

(2) The drug data is not enough to conclude that SYK inhibition is sufficient to prevent the division of RB1 null cone precursors. Drugs are never completely specific so validation is critical to make the conclusion drawn in the paper.

Reviewer #2 (Public review):

Summary:

The authors used deep full-length single-cell sequencing to study human photoreceptor development, with a particular emphasis on the characteristics of photoreceptors that may contribute to retinoblastoma.

Strengths:

This single-cell study captures gene regulation in photoreceptors across different developmental stages, defining post-mitotic cone and rod populations by highlighting their unique gene expression profiles through analyses such as RNA velocity and SCENIC. By leveraging full-length sequencing data, the study identifies differentially expressed isoforms of NRL and THRB in L/M cone and rod precursors, illustrating the dynamic gene regulation involved in photoreceptor fate commitment. Additionally, the authors performed high-resolution clustering to explore markers defining developing photoreceptors across the fovea and peripheral retina, particularly characterizing SYK's role in the proliferative response of cones in the RB loss background. The study provides an in-depth analysis of developing human photoreceptors, with the authors conducting thorough analyses using full-length single-cell RNA sequencing. The strength of the study lies in its design, which integrates single-cell full-length RNA-seq, long-read RNA-seq, and follow-up histological and functional experiments to provide compelling evidence supporting their conclusions. The model of cell type-dependent splicing for NRL and THRB is particularly intriguing. Moreover, the potential involvement of the SYK and MYC pathways with RB in cone progenitor cells aligns with previous literature, offering additional insights into RB development.

Weaknesses:

The manuscript feels somewhat unfocused, with a lack of a strong connection between the analysis of developing photoreceptors, which constitutes the bulk of the manuscript, and the discussion on retinoblastoma. Additionally, given the recent publication of several single-cell studies on the developing human retina, it is important for the authors to cross-validate their findings and adjust their statements where appropriate.

Reviewer #3 (Public review):

Summary:

The authors use high-depth, full-length scRNA-Seq analysis of fetal human retina to identify novel regulators of photoreceptor specification and retinoblastoma progression.

Strengths:

The use of high-depth, full-length scRNA-Seq to identify functionally important alternatively spliced variants of transcription factors controlling photoreceptor subtype specification, and identification of SYK as a potential mediator of RB1-dependent cell cycle reentry in immature cone photoreceptors.

Human developing fetal retinal tissue samples were collected between 13-19 gestational weeks and this provides a substantially higher depth of sequencing coverage, thereby identifying both rare transcripts and alternative splice forms, and thereby representing an important advance over previous droplet-based scRNA-Seq studies of human retinal development.

Weaknesses:

The weaknesses identified are relatively minor. This is a technically strong and thorough study, that is broadly useful to investigators studying retinal development and retinoblastoma.

eLife Assessment

This important study addresses how DNA replication restarts in Escherichia coli in the absence of a functional replication initiator protein DnaA. The authors show that helicase DnaB loading at the replication origin oriC can be executed by PriC under sub-optimal initiation conditions. While the genetic and biochemical evidence is solid, there is so far no direct evidence for PriC acting at oriC in vivo.

Reviewer #1 (Public review):

Summary:

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

Major Questions/Comments:

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

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

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

Reviewer #2 (Public review):

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

This is a major finding because it clearly demonstrates that the two roles of DnaA in the initiation process can be separated into initially forming an open complex at the DUE region by binding/nucleation onto DnaA-boxes and second by loading of the helicase. Whereas these two functions are normally assumed to be coupled, the present data clearly show that they can be separated and that PriC can perform at least part of the helicase loading provided that an area of duplex opening is formed by DnaA.

This puts into question the interpretation of a large body of previous work on mutagenesis of oriC and dnaA to find a minimal oriC/DnaA complex in many bacteria. In other words, mutants in which oriC is truncated/mutated may support the initiation of replication and cell viability only in the presence of PriC. Such mutants are capable of generating single-strand openings but may fail to load the helicase in the absence of PriC. Similarly, dnaA mutants may generate an aberrant complex on oriC that trigger strand opening but are incapable of loading DnaB unless PriC is present.

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

Reviewer #3 (Public review):

Summary:

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

Strengths:

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

Weaknesses:

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

Figure 2<br /> The reduced number of two oriC copies per cell in the dnaA46priC-deficient strain was considered moderate inhibition. When combined with the data suggested by the dnaAC2priC-deficient strain containing two origins in cells with or without PriC (indicating no inhibition)-the conclusion was drawn that PriC rescue blocked replication via assisting DnaC-dependent DnaB loading step at oriC ssDNA.

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

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

Figure 3<br /> One should consider the fact that dnA46 is present in these cells. Overexpressing pdnaAFH could produce mixed multimers containing subunits of DnaA46 (reduced ATP binding) and DnaAFH (reduced DnaB binding). Both have intact DnaA-DnaA oligomerization ability. The cooperativity between the two functions by a subpopulation of two DnaA variants may compensate for the individual deficiencies, making a population of an active protein, which in the presence of PriC could lead to the promotion of the stable DnaA: DnaBC complexes, able to initiate replication. In the light of results presented in Hayashi et al. and J Biol Chem. 2020 Aug 7;295(32):11131-11143, where mutant DnaBL160A identified was shown to be impaired in DnaA binding but contained an active helicase function and still inhibited for growth; how one could explain the hypothesis presented in this manuscript. If PriC-assisted helicase loading could bypass DnaA interaction, then how growth inhibition in a strain carrying DnaBL160A should be described. However, seeing the results in light of the alternative possibility that PriC assists in stabilizing the DnaA: DnaBC complex is more compatible with the previously published data.

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

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

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

eLife Assessment

This is a useful report of a spatially-extended model to study the complex interactions between immune cells, fibroblasts, and cancer cells, providing insights into how fibroblast activation can influence tumor progression. The model opens up new possibilities for studying fibroblast-driven effects in diverse settings, which is crucial for understanding potential tumor microenvironment manipulations that could enhance immunotherapy efficacy. While the results presented are solid and follow logically from the model's assumptions, some of these assumptions may require further validation, as they appear to oversimplify certain aspects in light of complex experimental findings, system geometry, and general principles of active matter research.

Reviewer #1 (Public review):

The authors present an important work where they model some of the complex interactions between immune cells, fibroblasts and cancer cells. The model takes into account the increased ECM production of cancer-associated fibroblasts. These fibres trap the cancer but also protect it from immune system cells. In this way, these fibroblasts' actions both promote and hinder cancer growth. By exploring different scenarios, the authors can model different cancer fates depending on the parameters regulating cancer cells, immune system cells and fibroblasts. In this way, the model explores non-trivial scenarios. An important weakness of this study is that, though it is inspired by NSCLC tumors, it is restricted to modelling circular tumor lesions and does not explore the formation of ramified tumors, as in NSCLC. In this way, is only a general model and it is not clear how it can be adapted to simulate more realistic tumor morphologies.

Reviewer #2 (Public review):

Summary:

The authors develop a computational model (and a simplified version thereof) to treat an extremely important issue regarding tumor growth. Specifically, it has been argued that fibroblasts have the ability to support tumor growth by creating physical conditions in the tumor microenvironment that prevent the relevant immune cells from entering into contact with, and ultimately killing, the cancer cells. This inhibition is referred to as immune exclusion. The computational approach follows standard procedures in the formulation of models for mixtures of different material species, adapted to the problem at hand by making a variety of assumptions as to the activity of different types of fibroblasts, namely "normal" versus "cancer-associated". The model itself is relatively complex, but the authors do a convincing job of analyzing possible behaviors and attempting to relate these to experimental observations.

Strengths:

As mentioned, the authors do an excellent job of analyzing the behavior of their model both in its full form (which includes spatial variation of the concentrations of the different cellular species) and in its simplified mean field form. The model itself is formulated based on established physical principles, although the extent to which some of these principles apply to active biological systems is not clear (see Weaknesses). The results of the model do offer some significant insights into the critical factors which determine how fibroblasts might affect tumor growth; these insights could lead to new experimental ways of unraveling these complex sets of issues and enhancing immunotherapy.

Weaknesses:

Models of the form being studied here rely on a large number of assumptions regarding cellular behavior. Some of these seemed questionable, based on what we have learned about active systems. The problem of T cell infiltration as well as the patterning of the extracellular matrix (ECM) by fibroblasts necessarily involve understanding cell motion and cell interactions due e.g. to cell signaling. Adopting an approach based purely on physical systems driven by free energies alone does not consider the special role that active processes can play, both in motility itself and in the type of self-organization that can occur due to these cell-cell interactions. This to me is the primary weakness of this paper.

A separate weakness concerns the assumption that fibroblasts affect T cell behavior primarily by just making a more dense ECM. There are a number of papers in the cancer literature (see, for some examples, Carstens, J., Correa de Sampaio, P., Yang, D. et al. Spatial computation of intratumoral T cells correlates with survival of patients with pancreatic cancer. Nat Commun 8, 15095 (2017); Sun, Xiujie, Bogang Wu, Huai-Chin Chiang, Hui Deng, Xiaowen Zhang, Wei Xiong, Junquan Liu et al. "Tumour DDR1 promotes collagen fibre alignment to instigate immune exclusion." Nature 599, no. 7886 (2021): 673-678) that seem to indicate that density alone is not a sufficient indicator of T cell behavior. Instead, the organization of the ECM (for example, its anisotropy) could be playing a much more essential role than is given credit for here. This possibility is hinted at in the Discussion section but deserves much more emphasis.

Finally, the mixed version of the model is, from a general perspective, not very different from many other published models treating the ecology of the tumor microenvironment (for a survey, see Arabameri A, Asemani D, Hadjati J (2018), A structural methodology for modeling immune-tumor interactions including pro-and anti-tumor factors for clinical applications. Math Biosci 304:48-61). There are even papers in this literature that specifically investigate effects due to allowing cancer cells to instigate changes in other cells from being tumor-inhibiting to tumor-promoting. This feature occurs not only for fibroblasts but also for example for macrophages which can change their polarization from M1 to M2. There needed to be some more detailed comparison with this existing literature.

eLife Assessment

This manuscript presents important information as to how adolescent alcohol exposure (AIE) alters pain behavior and relevant neurocircuits, with compelling data. The manuscript focuses on how AIE alters the basolateral amygdala, to the PFC (PV-interneurons), to the periaquaductal gray circuit, resulting in feed-forward inhibition. The manuscript is a detailed study of the role of alcohol exposure in regulating the circuit and reflexive pain, however, the role of the PV interneurons in mechanistically modulating this feed-forward circuit could be more strongly supported.

Reviewer #1 (Public review):

Summary:

In this manuscript by Obray et al., the authors show that adolescent ethanol exposure increases mechanical allodynia in adulthood. Additionally, they show that BLA-mediated inhibition of the prelimbic cortex is reduced, resulting in increased excitability in neurons that then project to vlPAG. This effect was mediated by BLA inputs onto PV interneurons. The primary finding of the manuscript is that these AIE-induced changes further impact acute pain processing in the BLA-PrL-vlPAG circuit, albeit behavioral readouts after inducing acute pain were not different between AIE rats and controls. These results provide novel insights into how AIE can have long-lasting effects on pain-related behaviors and neurophysiology. In this manuscript by Obray et al., the authors show that adolescent ethanol exposure increases mechanical allodynia in adulthood. Additionally, they show that BLA-mediated inhibition of the prelimbic cortex is reduced, resulting in increased excitability in neurons that then project to vlPAG. This effect was mediated by BLA inputs onto PV interneurons. The primary finding of the manuscript is that these AIE-induced changes further impact acute pain processing in the BLA-PrL-vlPAG circuit, albeit behavioral readouts after inducing acute pain were not different between AIE rats and controls. These results provide novel insights into how AIE can have long-lasting effects on pain-related behaviors and neurophysiology.

Strengths:

The manuscript was very well written and the experiments were rigorously conducted. The inclusion of both behavioral and neurophysiological circuit recordings was appropriate and compelling. The attention to SABV and appropriate controls was well thought out. The Discussion provided novel ideas for how to think about AIE and chronic pain and proposed several interesting mechanisms. This was a very well-executed set of experiments.

Weaknesses:

There is a mild disconnect between behavioral readout (reflexive pain) and neural circuits of interest (emotional). Considering that this circuit is likely engaged in the aversiveness of pain, it would have been interesting to see how carrageenan and/or AIE impacted non-reflexive pain measures. Perhaps this would reveal a potentiated or dysregulated phenotype that matches the neurophysiological changes reported. However, this critique does not take away from the value of the paper or its conclusions.

Reviewer #2 (Public review):

Summary:

The study by Obray et al. entitled "Adolescent alcohol exposure promotes mechanical allodynia and alters synaptic function at inputs from the basolateral amygdala to the prelimbic cortex" investigated how adolescent intermittent ethanol exposure (AIE) affects the BLA -> PL circuit, with an emphasis on PAG projecting PL neurons, and how AIE changes mechanical and thermal nociception. The authors found that AIE increased mechanical, but not thermal nociception, and an injection of an inflammatory agent did not produce changes in an ethanol-dependent manner. Physiologically, a variety of AIE-specific effects were found in PL neuron firing at BLA synapses, suggestive of AIE-induced alterations in neurotransmission at BLA-PVIN synapses.

Strengths:

This was a comprehensive examination of the effects of AIE on this neural circuit, with an in-depth dissection of the various neuronal connections within the PL.

Sex was included as a biological variable, yet there were little to no sex differences in AIE's effects, suggestive of similar adaptations in males and females.

Reviewer #3 (Public review):

Summary:

Obray et al. investigate the long-lasting effects of adolescent intermittent ethanol (AIE) in rats, a model of alcohol dependence, on a neural circuit within the prefrontal cortex. The studies are focused on inputs from the basolateral amygdala (BLA) onto parvalbumin (PV) interneurons and pyramidal cells that project to the periaqueductal gray (PAG). The authors found that AIE increased BLA excitatory drive onto parvalbumin interneurons and increased BLA feedforward inhibition onto PAG-projecting neurons.

Strengths:

Fully powered cohorts of male and female rodents are used, and the design incorporates both AIE and an acute pain model. The authors used several electrophysiological techniques to assess synaptic strength and excitability from a few complimentary angles. The design and statistical analysis are sound, and the strength of evidence supporting synaptic changes following AIE results is solid.

Weaknesses:

(1) There is incomplete evidence supporting some of the conclusions drawn in this manuscript. The authors claim that the changes in feedforward inhibition onto pyramidal cells are due to the changes in parvalbumin interneurons, but evidence is not provided to support that idea. PV cells do not spontaneously fire action potentials spontaneously in slices (nor do they receive high levels of BLA activity while at rest in slices). It is possible that spontaneous GABA release from PV cells is increased after AIE but the authors did not report sIPSC frequency. Second, the authors did not determine that PV cells mediate the feedforward BLA op-IPSCs and changes following AIE (this would require manipulation to reduce/block PV-IN activity). This limitation in results and interpretation is important because prior work shows BLA-PFC feedforward IPSCs can be driven by somatostatin cells. Cholecystokinin cells are also abundant basket cells in PFC and have been recently shown to mediate feedforward inhibition from the thalamus and ventral hippocampus, so it's also possible that CCK cells are involved in the effects observed here.

(2) The authors conclude that the changes in this circuit likely mediate long-lasting hyperalgesia, but this is not addressed experimentally. In some ways, the focused nature of the study is a benefit in this regard, as there is extensive prior literature linking this circuit with pain behaviors in alternative models (e.g., SNI), but it should be noted that these studies have not assessed hyperalgesia stemming from prior alcohol exposure. While the current studies do not include a causative behavioral manipulation, the strength of the association between BLA-PL-PAG function and hyperalgesia could be bolstered by current data if there were relationships detected between electrophysiological properties and hyperalgesia. Have the authors assessed this? In addition, this study is limited by not addressing the specificity of synaptic adaptations to the BLA-PL-PAG circuit. For instance, PL neurons send reciprocal projections to BLA and send direct projections to the locus coeruleus (which the authors note is an important downstream node of the PAG for regulating pain).

(3) I have some concerns about methodology. First, 5-ms is a long light pulse for optogenetics and might induce action-potential independent release. Does TTX alone block op-EPSCs under these conditions? Second, PV cells express a high degree of calcium-permeable AMPA receptors, which display inward rectification at positive holding potentials due to blockade from intracellular polyamines. Typically, this is controlled/promoted by including spermine in the internal solution, but I do not believe the authors did that. Nonetheless, the relatively low A/N ratios for this cell type suggest that CP-AMPA receptors were not sampled with the +40/+40 design of this experiment, raising concerns that the majority of AMPA receptors in these cells were not sampled during this experiment. Finally, it should be noted that asEPSC frequency can also reflect changes in a number of functional/detectable synapses. This measurement is also fairly susceptible to differences in inter-animal differences in ChR2 expression. There are other techniques for assessing presynaptic release probability (e.g., PPR, MK-801 sensitivity) that would improve the interpretation of these studies if that is intended to be a point of emphasis.

(4) In a few places in the manuscript, results following voluntary drinking experiments (especially Salling et al. and Sicher et al.) are discussed without clear distinction from prior work in vapor models of dependence

(5) Discussion (lines 416-420). The authors describe some differing results with the literature and mention that the maximum current injection might be a factor. To me, this does not seem like the most important factor and potentially undercuts the relevance of the findings. Are the cells undergoing a depolarization block? Did the authors observe any changes in the rheobase or AP threshold? On the other hand, a more likely difference between this and previous work is that the proportion of PAG-projecting cells is relatively low, so previous work in L5 likely sampled many types of pyramidal cells that project to other areas. This is a key example where additional studies by the current group assessing a distinct or parallel set of pyramidal cells would aid in the interpretation of these results and help to place them within the existing literature. Along these lines, PAG-projecting neurons are Type A cells with significant hyperpolarization sag. Previous studies showed that adolescent binge drinking stunts the development of HCN channel function and ensuing hyperpolarization sag. Have the authors observed this in PAG-projecting cells? Another interesting membrane property worth exploring with the existing data set is the afterhyperpolarization / SK channel function.

eLife Assessment

This study provides valuable advances in our understanding of how inputs from multiple sources can impact the physiology of motor neurons during the process of multisensory integration. Specifically, the authors show how streams of auditory and principally visual information modulate the physiology of Mauthner neurons in goldfish, thus allowing the different senses to influence escape behavior. Supporting evidence is generally convincing, although material reporting the direct control of behavior is less representative of the data.

Reviewer #1 (Public Review):

Otero-Coronel et al. address an important question for neuroscience - how does a premotor neuron capable of directly controlling behavior integrate multiple sources of sensory inputs to inform action selection? For this, they focused on the teleost Mauthner cell, long known to be at the core of a fast escape circuit. What is particularly interesting in this work is the naturalistic approach they took. Classically, the M-cell was characterized, both behaviorally and physiologically, using an unimodal sensory space. Here the authors make the effort (substantial!) to study the physiology of the M-cell taking into account both the visual and auditory inputs. They performed well-informed electrophysiological approaches to decipher how the M-cell integrates the information of two sensory modalities depending on the strength and temporal relation between them.

The empirical results are convincing and well-supported. The manuscript is well-written and organized. The experimental approaches and the selection of stimulus parameters are clear and informed by the bibliography. The major finding is that multisensory integration increases the certainty of environmental information in an inherently noisy environment.

Reviewer #2 (Public Review):

In this manuscript, Otero-Coronel and colleagues use a combination of acoustic stimuli and electrical stimulation of the tectum to study MSI in the M-cells of adult goldfish. They first perform a necessary piece of groundwork in calibrating tectal stimulation for maximal M-cell MSI, and then characterize this MSI with slightly varying tectal and acoustic inputs. Next, they quantify the magnitude and timing of FFI that each type of input has on the M-cell, finding that both the tectum and the auditory system drive FFI, but that FFI decays more slowly for auditory signals. These are novel results that would be of interest to a broader sensory neuroscience community. By then providing pairs of stimuli separated by 50ms, they assess the ability of the first stimulus to suppress responses to the second, finding that acoustic stimuli strongly suppress subsequent acoustic responses in the M-cell, that they weakly suppress subsequent tectal stimulation, and that tectal stimulation does not appreciably inhibit subsequent stimuli of either type. Finally, they show that M-cell physiology mirrors previously reported behavioural data in which stronger stimuli underwent less integration.

The manuscript is generally well-written and clear. The discussion of results is appropriately broad and open-ended. It's a good document. Our major concerns regarding the study's validity are captured in the individual comments below. In terms of impact, the most compelling new observation is the quantification of the FFI from the two sources and the logical extension of these FFI dynamics to M-cell physiology during MSI. It is also nice, but unsurprising, to see that the relationship between stimulus strength that MSI is similar for M-cell physiology to what has previously been shown for behavior. While we find the results interesting, we think that they will be of greatest interest to those specifically interested in M-cell physiology and function.

Author response:

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

Minor Concern (Original Comment 1):

“We think that this is sufficient to address our concern. Some citations may be in order to underpin the new text.”

We appreciate the reviewer’s assessment that the revised text clarifies the complexity of the upstream circuitry beyond the retina, including inputs from the thalamus. As recommended, we have now included additional citations in the revised manuscript to support these points.

Major Concern (Original Comment 5):

“We do not feel that this important concern has been addressed. The stats are definitively negative. There is no statistical evidence from these data that multisensory integration is occurring in this assay. The anesthesia, paralysis, and low n may provide explanations for this negative result, but it is still a negative result (p=0.5269). To show two examples of multisensory integration for subthreshold stimuli fits the narrative, but this result is not supported. Examples where individual stimuli caused APs (and combined stimuli did not) also occurred, presumably, and at a rate that is statistically indistinguishable to the examples shown in Figure 5. As such, if results from this assay are going to be in the manuscript, acoustic-only and tectum-only examples should be shown as well, although they would not fit the narrative. To be meaningful, this experiment would have to show that multisensory integration is happening in this circuit. Frustrating though it must be, the experiment has given a negative result to that question.”

We understand the reviewer’s concern regarding Figure 5C and the firing of action potentials (APs) in response to multisensory stimuli. We acknowledge that our assay is not suited to answer this question definitively and that our results do not provide statistical support for this hypothesis. In response, we have removed the examples previously shown in Figure 5C, along with the related description in the Results section (lines 420–426), to avoid implying unsupported integration in suprathreshold conditions.

Author response:

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

Reviewer #1:

Point 1: While the manuscript is methodologically sound, the following aspects of image acquisition and data analysis need to be clarified to ensure replicability and reproducibility. The authors state that the sample is a "population-derived adult lifespan sample", the lack of demographic information makes it impossible to know if the sample is truly representative. Though this may seem inconsequential, education may impact both cognitive performance and functional activation patterns. Moreover, the authors do not report race/ethnicity in the manuscript. This information is essential to ensure representativeness in the sample. It is imperative that barriers to study participation within minoritized groups are addressed to ensure rigor and reproducibility of findings.

First, the section Methods-Participants has been updated to refer readers to a prior article where the sample’s demographics are broken down into nine decile age groups (see Wu et al. 2023 Table 1), including information about their education levels. Secondly, we have updated the Data Availability section text to indicate that all Cam-CAN IDs are included in the available OSF datasets, allowing anyone to verify additional participant demographics described in the Cam-CAN protocol article (Shafto et al., 2014). Third, we have updated the Participants section text to refer to another prior study that reported on the representativeness of the Cam-CAN sample indicating that at least some elements of the sample have been independently deemed as representative (e.g., Sex).

Page-24

“A healthy population-derived adult lifespan human sample (N = 223; ages approximately uniformly distributed from 19 - 87 years; females = 112; 50.2%) was collected as part of the Cam-CAN study (Stage 3 cohort; Shafto et al., 2014). Participants were fluent English speakers in good physical and mental health, based on the Cam-CAN cohort’s exclusion criteria which includes poor mini mental state examination, ineligibility for MRI and medical, psychiatric, hearing or visual problems. Throughout analyses, age is defined at the Home Interview (Stage 1; Shafto et al., 2014). The study was approved by the Cambridgeshire 2 (now East of England–Cambridge Central) Research Ethics Committee and participants provided informed written consent. Further demographic information of the sample is reported in Wu et al. (2023) and is openly available (see section Data Availability) with a recent report indicating the representativeness of the sample across sexes (Green et al., 2018).”

Page-30

“Raw and minimally pre-processed MRI (i.e., from automatic analysis; Taylor et al., 2017) and behavioural data are available by submitting a data request to Cam-CAN (https://camcan-archive.mrc-cbu.cam.ac.uk/dataaccess/). The univariate and multivariate ROI data, and behavioural data, can be downloaded from the Open Science Framework, which includes Cam-CAN participant identifiers allowing the retrieval of any additional demographic data (https://osf.io/v7kmh), while the analysis code is available on GitHub.”

Point 2: For the whole-brain analysis in which the ROIs were derived, the authors used a threshold-free cluster enhancement (TFCE; Smith & Nichols 2009). The methodological paper cited suggests that individuals' TCFE image should still be corrected for multiple comparisons using the following: "to correct for multiple comparisons, one [...] has to build up the null distribution (across permutations of the input data) of the maximum (across voxels) TFCE score, and then test the actual TFCE image against that. Once the 95th percentile in the null distribution is found then the TFCE image is simply thresholded at this level to give inference at the p < 0.05 (corrected) level." (Smith & Nichols, 2009). Although the authors mention that clusters were estimated using 2000 permutations, there is no mention of the TFCE image itself being thresholded. While this would impact the overall size of the ROIs used in the study, the remaining analyses are methodologically sound.

We have updated the text to detail the t=1.97 (i.e., p = .05) threshold we applied before interpretation of the resultant TFCE images to the section: Experimental Design & Statistical Analysis. This threshold value can also be verified in the analytics code that is referenced on GitHub from the section Data Availability within the requisite toolbox functions: https://github.com/kamentsvetanov/CommonalityAnalysis/blob/main/code/ca_vba_tfce_threshold.m#L24 and https://github.com/kamentsvetanov/CommonalityAnalysis/blob/main/code/external/ca_matlab_tfce_transform.m

Page-30

“For whole-brain voxelwise analyses, clusters were estimated using threshold-free cluster enhancement (TFCE; Smith & Nichols 2009) with 2000 permutations and the resulting images were thresholded at a t-statistic of 1.97 before interpretation.”

Point 3: The authors should consider moving the ROI section to results. The way the manuscript currently reads, the ROIs seem to be derived a priori as opposed to being derived from activation maps in the current study.

After consideration of this point, we have decided to leave the methodological details regarding the definition of ROIs in the methods, to maintain the focus of the Results section. However, we have improved signposting in the results section to highlight that the ROIs were derived from the overlapped activation maps.

Page-8

“Crucially, two areas of the brain showed spatially-overlapping positive effects of age and performance, which is suggestive of an age-related compensatory response (Figure 2A yellow intersection). These were in bilateral cuneal cortex (Figure 2B magenta) and bilateral frontal cortex (Figure 2B brown), the latter incorporating parts of the middle frontal gyri and anterior cingulate. Therefore, based on traditional univariate analyses, these are two candidate regions for age-related functional compensation (Cabeza et al. 2013; 2018). Accordingly, we defined regions of interest within these two regions using the overlap activation maps (see section: ROIs) to be used for subsequent univariate and multivariate analysis.”

Point 4: The manuscript can be strengthened by explaining why the authors chose a greedy search algorithm over a dynamic Bayesian model.

The text is updated to refer to appropriateness of the computationally efficient greedy search implementation, due to the size of the fMRI cohort dataset.

Page-28

“The pattern weights specifying the mapping of data features to the target variable are optimized with a greedy search algorithm using a standard variational scheme (Friston et al., 2007) which was particularly appropriate given the large dataset.”

Reviewer #2:

Point 1: However, it might have been nice to see an analysis of a more crystallised intelligence task included too, as a contrast since this is an area that does not demonstrate such a decline (and perhaps continues to improve over aging).

We (Samu et al., 2017) have previously investigated, but failed to find, univariate evidence for functional compensation in this cohort’s performance on a sentence comprehension task that is more closely aligned to a measure of crystallised intelligence. Based on the additional previous studies where we have applied these types of univariate and multivariate criteria of functional compensation (Morcom & Henson, 2018; Knights et al., 2021), we have consistently observed that the uni-/multivariate effects are in the same direction. Therefore, we would not initially expect a different conclusion here, where the univariate and multivariate effects suggest different outcomes. Notably, the univariate analysis approach in Samu et al. (2017) did differ from focusing on the age x behaviour interaction term here, so it could still be worth future investigation, but it does seem less likely that evidence of compensation would be observed than for fluid intelligence. However, as the Reviewer suggests, such a task may make another good contrast to show evidence against the existence of functional compensation (as in Morcom & Henson, 2018; Knights et al., 2021).

Point 2: Figure 1B: Consider adding coefficients describing relationships to plots.

Annotations of the coefficients have been added to Figure 1B:

Point 3: Figure 2C. The scale of the axis for RSFA-Scales cuneal cortex ROI activations should be the same as the other 3 plots.

Figure axes are updated such that ROIs are on matching scales, according to whether data were RSFA-scaled or not.

Point 4: Figure 2C. Adding in the age ranges for each of the three groups following the tertile split may be informative to the reader.

The age group tertile definition used for Figure 2C visualisations is now added to the Figure description.

Page-10

“Figure 2. Univariate analysis. (A) Whole-brain effects of age and performance. Age (green) and performance (red) positively predicted unique aspects of increased task activation, with their spatial overlap (yellow) being overlaid on a template MNI brain, using p < 0.05 TFCE. (B) Intersection ROIs. A bilateral cuneal (magenta) and frontal cortex (brown) ROI were defined from voxels that showed a positive and unique effect of both age and performance (yellow map in Figure 2A). (C) ROI Activation. Activation (raw = left; RSFA-scaled = right) is plotted against behavioural performance based on a tertile split between three age groups (19-44, 45-63 & 64-87 years).”

Reviewer #3:

Point 1: [Public Review] 1) I don't quite follow the argumentation that compensatory recruitment would need to show via non-redundant information carried by any given non-MDN region (cf. p14). Wouldn't the fact that a non-MDN region carries task-related information be sufficient to infer that it is involved in the task and, if activated increasingly with increasing age, that its stronger recruitment reflects compensation, rather than inefficiency or dedifferentiation? Put differently, wouldn't "more of the same" in an additional region suffice to qualify as compensation, as compared to the "additional information in an additional region" requirement set by the authors? As a consequence, in my honest opinion, showing that decoding task difficulty from non-MDN ROIs works better with higher age would already count as evidence for compensation, rather than asking for age-related increases in decoding boosts obtained from adding such ROIs. It would be interesting to see whether the arguably redundant frontal ROI would satisfy this less demanding criterion. At any rate, it seems useful to show whether the difference in log evidence for the real vs. shuffled models is also related to age.

We agree with the logic for conducting a weaker assessment of functional compensation whereby a brain region does not necessarily have to provide a unique contribution beyond that of the ordinarily activated task-relevant network. However, although non-unique recruitment is predicted by a compensation theory, it can also be explained by a nonspecific mechanism that recruits multiple regions in tandem. In contrast, unique additional recruitment is compatible with compensation but not with nonspecific recruitment. In this article, and those prior (Morcom & Henson, 2018; Knights et al. 2021), we have also deliberately avoided using the specific kind of analysis proposed (i.e., testing for an effect of age on differential log evidence) because these would involve applying statistical tests directly to the log evidence, a variable that is already a statistical test output.

Nevertheless, temporarily putting these caveats aside, we did run the suggested test. Results from multiple regression showed that using log evidence from frontal cortex models still did not meet this less demanding criterion for functional compensation as there was an effect of age in the opposite direction to that expected by functional compensation: there was a significant negative effect of age (t(218) = -7.95, p = < .001) indicating that as age increased, the difference in log evidence decreased. This effect is visualised below for transparency, but we preferred not to add this information to the article because we do not wish to encourage using this kind of analysis for the reason mentioned above. Thus, although our main multivariate test of interest is stringent, the additional step of mapping log evidence back to the boost-likelihood categories (e.g., boost vs. no difference to model performance) lends itself to the more appropriate logistic regression statistical approach.

Author response image 1.

Negative effect of age on MVB log evidence model outcomes for frontal cortex.

A different approach that could be taken to assess a more lenient definition of functional compensation would be to analyse the effects of age on the spread of multivariate responses predicting task difficulty (i.e., standard deviation of fitted MVB voxel weights; also see Morcom & Henson, 2018; Knights et al., 2021) specifically from models that only include the candidate ‘compensation’ ROIs.

Accordingly, these analyses and their discussion have been added to the article. To summarise, these analyses showed that (1) the frontal cortex still did not show evidence of functional compensation (i.e., a negative effect of age like in Morcom & Henson, 2018) and (2) no effect of age on the cuneal ROI, implying that the original model comparison approach (i.e., Figure 2C in the manuscript now) can provide more sensitivity for detecting evidence of functional compensation (perhaps because of the importance of including task-relevant network responses when building decoding models).

Page-15

“As a final analysis, we also tested a more lenient definition of functional compensation, whereby the multivariate contribution from the “compensation ROI” does not necessarily need to be above and beyond that of the task-relevant network (Morcom & Henson, 2018; Knights et al., 2021). To do this, we again assessed whether age was associated with an increase in the spread (standard deviation) of the weights over voxels, for smaller models containing only the cuneal or frontal ROI. This tested whether increased age led to more voxels carrying substantial information about task difficulty, a pattern predicted by functional compensation (but also consistent with non-specific additional recruitment). In this case, the results of this test did not support functional compensation, as there was no effect detected for the cuneal cortex and even a negative effect of age for the frontal cortex where the spread of the information across voxels was lower for older age (Figure 3C; Table 2).”

Page-21

“The age- and performance-related activation in our frontal region satisfied the traditional univariate criteria for functional compensation, but our multivariate (MVB) model comparison analysis showed that additional multivariate information beyond that in the MDN was absent in this region, which is inconsistent with the strongest definition of compensation. In fact, the results from the spread analysis showed that as age increased, this frontal area processed less, rather than more, multivariate information about the cognitive outcome (Figure 3C) as previously observed in two (memory) tasks for a comparable ROI within the same Cam-CAN cohort (Morcom & Henson, 2018).”

Page-24

“This said, univariate criteria for functional compensation will continue to play a role in hypothesis testing. For instance, the over-additive interaction observed in the cuneal cortex - where the increase in activity with better performance is more pronounced in older adults - offers stronger evidence of compensation compared to the simple additive effect of age and performance observed in the frontal cortex (Figure 2C). So far, the two studies that have combined these rigorous univariate, behavioral and multivariate approaches to assess functional compensation (i.e., Knights et al., 2021; the present study) have generally found converging evidence regardless of the method used. However, it is important to note that the MVB approach uniquely shifts the focus from individual differences to the specific task-related information that compensatory neural activations are assumed to carry and provides a specific test of region- (or network-) unique information. With further studies, it may also be that multivariate approaches prove more sensitive for detecting compensation effects than when using mean responses over voxels (e.g., Friston et al., 1995) particularly since over-additive effects are challenging to observe because compensatory effects are typically ‘partial’ and do not fully restore function (for review see Scheller et al., 2014; Morcom & Johnson, 2015). Within the multivariate analysis options themselves, it is also interesting to highlight that the stringent MVB boost likelihood analysis could detect functional compensation unlike the more lenient analysis focusing on the spread of MVB voxel weights. This suggests the importance of including task-relevant network responses when building decoding models to assess compensation.”

Page-32

“Alongside the MVB boost analysis, we also included an additional measure using the spread (standard deviation) of voxel classification weights (Morcom & Henson, 2018). This measure indexes the absolute amplitude of voxel contributions to the task, reflecting the degree to which multiple voxels carry substantial task-related information. When related to age this can serve as a multivariate index of information distribution, unlike univariate analyses. However, it is worth highlighting that even if an ROI shows an effect of age on this spread measure, such an effect could instead be explained by a non-specific mechanism that represents the same information in tandem across multiple regions (rather than reflecting compensation) as seen previously (Knights et al., 2021; also see Morcom & Johnson, 2015). Thus, it is the MVB boost analysis that is the most compelling assessment of functional compensation because it can directly detect novel information representation.”

Point 2: [Public Review] 2) Relatedly, does the observed boost in decoding by adding the cuneal ROI (in older adults) really reflect "additional, non-redundant" information carried by this ROI? Or could it be that this boost is just a statistical phenomenon that is obtained because the cuneus just happens to show a more clear-cut, less noisy difference in hard vs. easy task activation patterns than does the MDN (which itself may suffer from increased neural inefficiency in older age), and thus the cuneaus improves decoding performance without containing additional (novel) pieces of information (but just more reliable ones)? If so, the compensation account could still be maintained by reference to the less demanding rationale for what constitutes compensation laid out above.

We agree that this is a possibility and have added this as an additional explanation to the Discussion. We have also discussed why we think it is a less likely possibility, but do concede that it cannot be ruled out currently.

Page-20

“Another possibility is that the age-related increases in fMRI activations (for hard versus easy) in one or both of our ROIs do not reflect greater fMRI signal for hard problems in older than younger people, but rather lower fMRI signal for easy problems in the older. Without a third baseline condition, we cannot distinguish these two possibilities in our data. However, a reduced “baseline” level of fMRI signal (e.g., for easy problems) in older people is consistent with other studies showing an age-related decline in baseline perfusion levels, coupled with preserved capacity of cerebrovascular reactivity to meet metabolic demands of neuronal activity at higher cognitive load  (Calautti et al., 2001; Jennings et al., 2005). Though age-related decline in baseline perfusion occurs in the cuneal cortex (Tsvetanov et al., 2021), the brain regions showing modulation of behaviourally-relevant Cattell fMRI activity by perfusion levels did not include the cuneal cortex (Wu et al., 2023). This suggests that the compensatory effects in the cuneus are unlikely to be explained by age-related hypo-perfusion, consistent with the minimal effect here of adjusting for RSFA (Figure 2C).

One final possibility is whether the observed boost in decoding from adding the cuneal ROI simply reflects less noisy task-related information (i.e., a better signal-to-noise ratio (SNR)) than the MDN and, consequently, the boosted decoding is the result of more resilient patterns of information (rather than the representation of additional information) based on a steeper age-related decline of SNR in the MDN. Overall then, as none of the explanations above agree with all aspects of the results, to functionally explain the role of the cuneal cortex in this task would require further investigation.”

Point 3: [Public Review] 3) On page 21, the authors state that "...traditional univariate criteria alone are not sufficient for identifying functional compensation." To me, this conclusion is quite bold as I'd think that this depends on the unvariate criterion used. For instance, it could be argued that compensation should be more clearly indicated by an over additive interaction as observed for the relationship of cuneal activity with age and performance (i.e., the activity increase with better performance becomes stronger with age), rather than by an additive effect of age and performance as observed for the prefrontal ROI (see Fig. 2C). In any case, I'd appreciate it if the authors discussed this issue and the relationship between univariate and multivariate results in more detail (e.g. how many differences in sensitivity between the two approaches have contributed), in particular since the sophisticated multivariate approach used here is not widely established in the field yet.

We have now considered this point further in a section of the Discussion (which is merged with points 1 & 2 above) about the relevance and distinction of univariate / multivariate criteria for functional compensation. As described in text below, whilst we agree that univariate / behavioural approaches have a role in testing functional compensation, we still view the MVB boost analysis to be a particularly compelling approach for assessing this theory.

Page-22

“This said, univariate criteria for functional compensation will continue to play a role in hypothesis testing. For instance, the over-additive interaction observed in the cuneal cortex - where the increase in activity with better performance is more pronounced in older adults - offers evidence of compensation compared to the simple additive effect of age and performance observed in the frontal cortex (Figure 2C). However, the conclusions that can be drawn from age-related differences in cross-sectional associations of brain and behaviour are limited, mainly because individual performance differences are largely lifespan-stable (see Lindenberger et al., 2011; Morcom & Johnson, 2015). So far, the two studies that have combined these univariate-behavioral and multivariate approaches to assess functional compensation (i.e., Knights et al., 2021; the present study) have generally found converging evidence regardless of the method used. However, it is important to note that the MVB approach uniquely shifts the focus from individual differences to the specific task-related information that compensatory neural activations are assumed to carry. With further studies, it may also be that multivariate approaches prove more sensitive for detecting compensation effects than when using mean responses over voxels (e.g., Friston et al., 1995) particularly since over-additive effects are challenging to observe because compensatory effects are typically ‘partial’ and do not fully restore function. Within the multivariate analysis options themselves, it is also interesting to highlight that the stringent MVB boost likelihood analysis could detect functional compensation unlike the more lenient analysis focusing on the spread of MVB voxel weights. This suggests the importance of including task-relevant network responses when building decoding models to asses compensation.”

Point 4: [Public Review] 4) As to the exclusion of poorly performing participants (see p24): If only based on the absolute number of errors, wouldn't you miss those who worked (overly) slowly but made few errors (possibly because of adjusting their speed-accuracy tradeoff)? Wouldn't it be reasonable to define a criterion based on the same performance measure (correct - incorrect) as used in the main behavioural analyses?

This is a good point, though if we were to exclude participants using a chance level exclusion rate based on the formulae used for measuring behavioural performance, this removes identical subjects to those originally excluded. Based on this, the text has been updated to reflect this more parsimonious approach for defining exclusion criteria.

Page-25

“In a block design, participants completed eight 30-second blocks which contained a series of puzzles from one of two difficulty levels (i.e., four hard and four easy blocks completed in an alternating block order; Figure 1A). The fixed block time allowed participants to attempt as many trials as possible. Therefore, to balance speed and accuracy, behavioural performance was measured by subtracting the number of incorrect from correct trials and averaging over the hard and easy blocks independently (i.e., ((hard correct - hard incorrect) + (easy correct - easy incorrect))/2; Samu et al., 2017). For assessing reliability and validity, behavioural performance (total number of puzzles correct) was also collected from the same participants during a full version of the Cattell task (Scale 2 Form A) administered outside the scanner at Stage 2 of the Cam-CAN study (Shafto et al., 2014). Both the in- and out-of-scanner measures were z-scored. We excluded participants (N = 28; 17 females) who performed at chance level ((correct + incorrect) / incorrect < 0.5) on the fMRI task, leading to the same subset as reported in Samu et al. (2017).”

Point 5: [Public Review] 5) Did the authors consider testing for negative relationships between performance and brain activity, given that there is some literature arguing that neural efficiency (i.e. less activation) is the hallmark of high intelligence (i.e. high performance levels in the Cattell task)? If that were true, at least for some regions, the set of ROIs putatively carrying task-related information could be expanded beyond that examined here. If no such regions were found, it would provide some evidence bearing on the neural efficiency hypothesis.

No, we did not test for negative relationships between performance and brain activity in this study. However, In Wu et al. (2023) we did specifically test for this and neither of the relevant results reported in section 3.3.1 (i.e., unique relationship between activity and performance) nor section 3.3.2 (i.e., age-related relationship between activity and performance) showed the queried direction of effects. Note that the negative effect in section 3.3.2 (Age U Performance) is a more unique suppression effect representing a positive relationship between performance and activity where this becomes stronger as age is added to the model.

Point 6: [Recommendations for the authors] 1) Page 26: It is not quite clear how the authors made sure their age and performance covariates functioned as independent regressors in the univariate group-level GLM, given the correlation between age and performance (i.e. shared variance).

We included age and performance as covariates (of the age x performance effect of interest) by simply including these as independent regressors in the group-level GLM design matrix in addition to the interaction term (i.e., activity ~ age*performance + covariates equivalent to activity ~ age:performance + age + performance + covariates; Wilkinson & Roger 1973 notation), allowing us to examine the unique variance explained by each predictor (Table 1 and Table 2) and to control for their shared variance.

We should note that while the GLM approach we used accounts for unique and shared effects, it does not explicitly report shared effects in its standard output. To directly examine shared variance, one would need to employ commonality analysis. For reference, results from a commonality analysis on this task have been previously reported in Wu et al. (2023).

Prompted by this point, we have made some further minor improvements to help ensure our methodological steps are reproducible, as highlighted below.

Page-30

“Continuous age and behavioural performance variables were standardised and treated as linear predictors in multiple regression throughout the behavioural (Figure 1B), wholebrain voxelwise (Figure 1C/2A), univariate (Table 1; Figure 1B/2B) and MVB (Table 2; Figure 3) analyses. Throughout, sex was included as a covariate. The models, including interaction terms, can be described, according to Wilkinson & Roger’s (1973) notation, as activity ~ age * performance + covariates (which is equivalent to activity ~ age:performance + age + performance + covariates), allowing us to examine the unique variance explained by each predictor (Table 1) and to control for their shared variance. For whole-brain voxelwise analyses, clusters were estimated using threshold-free cluster enhancement (TFCE; Smith & Nichols 2009) with 2000 permutations and the resulting images were thresholded at a t-statistic of 1.97 before interpretation. Bonferroni correction was applied to a standard alpha = 0.05 based on the two ROIs (cuneal and frontal) that were examined. For Bayes Factors, interpretation criteria norms were drawn from Jarosz & Wiley (2014).”

Point 7: [Recommendations for the authors] 2) Figure 3: I suggest changing the subheading in panel B to "Joint vs. MDN-only Model," in line with the wording in the main text.

The subheading of Figure 3B is updated as suggested to `Joint vs. MDN-only Model`.

Point 8: [Recommendations for the authors] 3) In Figures 1C and 2A, MNI z coordinates should be added to the section views. The appreciation of Figure 2B could be enhanced by adding some rendering with a saggital (medial and/or lateral) view.

The slice mosaics in Figure 1C and 2A are now updated with each slice’s MNI Z coordinates and mentioned in the figure descriptions.

Point 9: [Recommendations for the authors] 4) Page 7 (l. 135): What exactly is meant by "lateral occipital temporal cortex"?

The text is updated to specify the anatomical landmarks that were used for guidance when referring to activation within the lateral occipital temporal cortex, based on ROI criteria definitions used in Knights, Mansfield et al. (2021):

Page-7 Line-135:

“Additional activation was observed bilaterally in the inferior/ventral and lateral occipital temporal cortex (i.e., a cluster around the lateral occipital sulcus that extended anteriorly beyond the anterior occipital sulcus), likely due to the visual nature of the task.”

Point 10: [Recommendations for the authors] 5) On p18ff. (ll. 259-318) the authors discuss in quite some detail how the age-related decoding boost seen with the cuneus ROI can be functionally explained, but it seems like none of the explanations agrees with all aspects of the results. While this is not a major problem for the paper, it may be advisable if this part of the discussion ends with a clearer statement that this issue is not fully solved yet and provides material for future research.

A more direct sentence has been added to make it clear that future investigation will be needed to explain the role of the cuneal cortex here.

Page-20 Line-322:

“Another possibility is that the age-related increases in fMRI activations (for hard versus easy) in one or both of our ROIs do not reflect greater fMRI signal for hard problems in older than younger people, but rather lower fMRI signal for easy problems in the older. Without a third baseline condition, we cannot distinguish these two possibilities in our data. However, a reduced “baseline” level of fMRI signal (e.g., for easy problems) in older people is consistent with other studies showing an age-related decline in baseline perfusion levels, coupled with preserved capacity of cerebrovascular reactivity to meet metabolic demands of neuronal activity at higher cognitive load  (Calautti et al., 2001; Jennings et al., 2005). Though age-related decline in baseline perfusion occurs in the cuneal cortex (Tsvetanov et al., 2021), the brain regions showing modulation of behaviourally-relevant Cattell fMRI activity by perfusion levels did not include the cuneal cortex (Wu et al., 2021). This suggests that the compensatory effects in the cuneus are unlikely to be explained by age-related hypo-perfusion, consistent with the minimal effect here of adjusting for RSFA (Figure 2C). Overall then, as none of the explanations above agree with all aspects of the results, to functionally explain the role of the cuneal cortex in this task will require further investigation.”

Point 11: [Recommendations for the authors] 6) The threshold choice for Bayesian log evidence (> 3) should be motivated in some more detail, rather than just pointing to a book reference, as there is no established convention in the field, the choice may depend on the type of data and/or analysis, and a sizeable part of the readership may not be deeply familiar with the particular Bayesian approach used here.

Text is updated to further clarify our motivation for using the log evidence BF>3 criterion:

Page-29

“The outcome measure was the log evidence for each model (Morcom & Henson, 2018; Knights et al., 2021). To test whether activity from an ROI is compensatory, we used an ordinal boost measure (Morcom & Henson, 2018; Knights et al., 2021) to assess the contribution of that ROI for the decoding of task-relevant information (Figure 3B). Specifically, Bayesian model comparison assessed whether a model that contains activity patterns from a compensatory ROI and the MDN (i.e., a joint model) boosted the prediction of task-relevant information relative to a model containing the MDN only. The compensatory hypothesis predicts that the likelihood of a boost to model decoding will increase with older age. The dependent measure, for each participant, was a categorical recoding of the relative model evidence to indicate the outcome of the model comparison. The three possible outcomes were: a boost to model evidence for the joint vs. MDN-only model (difference in log evidence > 3), ambiguous evidence for the two models (difference in log evidence between -3 to 3), or a reduction in evidence for the joint vs. MDN-only model (difference in log evidence < -3).These values were selected because a log difference of three corresponds to a Bayes Factor of 20, which is generally considered strong evidence (Lee & Wagenmakers, 2014). Further, with uniform priors, this chosen criterion (Bayes Factor > 3) corresponds to a p-value of p<~.05 (since the natural logarithm of 20 equals three, as evidence for the alternative hypothesis).”

Point 12: [Recommendations for the authors] 7) Adding page numbers would be helpful.

Page numbers have been added to the manuscript file – apologies for this oversight.

References

Green, E., Bennett, H., Brayne, C., & Matthews, F. E. (2018). Exploring patterns of response across the lifespan: The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) study. BMC Public Health, 18, 1-7.

Knights, E., Mansfield, C., Tonin, D., Saada, J., Smith, F. W., & Rossit, S. (2021). Hand-selective visual regions represent how to grasp 3D tools: brain decoding during real actions. Journal of Neuroscience, 41(24), 5263-5273.

Samu, D., Campbell, K. L., Tsvetanov, K. A., Shafto, M. A., & Tyler, L. K. (2017). Preserved cognitive functions with age are determined by domain-dependent shifts in network responsivity. Nature communications, 8(1), 14743.

Shafto, M. A., Tyler, L. K., Dixon, M., Taylor, J. R., Rowe, J. B., Cusack, R., ... & Cam-CAN. (2014). The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) study protocol: a cross-sectional, lifespan, multidisciplinary examination of healthy cognitive ageing. BMC neurology, 14, 1-25.

Wu, S., Tyler, L. K., Henson, R. N., Rowe, J. B., & Tsvetanov, K. A. (2023). Cerebral blood flow predicts multiple demand network activity and fluid intelligence across the adult lifespan. Neurobiology of aging, 121, 1-14.

eLife Assessment

This study provides an important advancement of knowledge by showing neural functional compensation in the brains of healthy older adults completing a fluid-intelligence task. Validated whole-brain voxel-wide analyses and multivariate Bayesian approaches provide compelling evidence that supports the claims of the authors. The work delivers methods for quantifying reserve and compensation in future studies and will be of interest to researchers in the field of the neuroscience of healthy aging.

Reviewer #1 (Public Review):

This work addresses how to quantify functional compensation throughout the aging process and identifies brain regions that engage in compensatory mechanisms during the Cattell task, a measure of fluid cognition. The authors find that regions of the frontal cortex and cuneus showed unique effects of both age and performance. Interestingly, these two regions demonstrated differential activation patterns taking into account both age and performance. Specifically, the researchers found that the relationship between performance and activation in the cuneal ROI was strongest in older adults, however, this was not found in younger adults. These findings suggest that specifically within the cuneus, greater activation is needed by older adults to maintain performance, suggestive of functional compensation.

The conclusions derived from the study are well supported by the data. The authors validated the use of the in-scanner Cattell task by demonstrating high reliability in the same sample with the standard out-of-scanner version. Some strengths of the study include the large sample size and wide age range of participants. The authors use a stringent Bayes factor of 20 to assess the strength of evidence. The authors used a whole-brain approach to define regions of interest (ROIs) based on activation patterns that were jointly related to age and performance. Overall, the methods are technically sound and support the authors' conclusions.

Comment from Reviewing Editor: In the revised manuscript, the authors have addressed the weaknesses previously identified by reviewer 1.

Reviewer #2 (Public Review):

This work by Knights et al., makes use of the Cam-CAN dataset to investigate functional compensation during a fluid processing task in older adults, in a fairly large sample of approximately 200 healthy adults ranging from 19 to 87. Using univariate methods, the authors identify two brain regions in which activity increases as a function of both age and performance and conduct further investigations to assess whether the activity of these regions provides information regarding task difficulty. The authors conclude that the cuneal cortex - a region of the brain previously implicated in visual attention - shows evidence of compensation in older adults.

The conclusions of the paper are well supported by the data, and the authors use appropriate statistical analyses. The use of multivariate methods over the last 20 years has demonstrated many effects that would have been missed using more traditional univariate analysis techniques. The data set is also of an appropriate size, and as the authors note, fluid processing is an extremely important domain in the field of cognition in aging, due to its steep decline over aging.

Comment from Reviewing Editor: It would have been nice to see an analysis of a more crystallised intelligence task included too, as a contrast since this is an area that does not demonstrate such a decline (and perhaps continues to improve over aging). This comment does not take away the important contributions of the manuscript.

Reviewer #3 (Public Review):

This neuroimaging study investigated how brain activity related to visual pattern-based reasoning changes over the adult lifespan, addressing the topic of functional compensation in older age. To this end, the authors employed a version of the Cattell task, which probes visual pattern recognition for identifying commonalities and differences within sets of abstract objects in order to infer the odd object among a given set. Using a state-of-the-art univariate analysis approach on fMRI data from a large lifespan sample, the authors identified brain regions in which the activation contrast between hard and easy Cattell task conditions was modulated by both age and performance. Regions identified comprised prefrontal areas and bilateral cuneus. Applying a multivariate decoding approach to activity in these regions, the authors went on to show that only in older adults, the cuneus, but not the prefrontal regions, carried information about the task condition (hard vs. easy) beyond that already provided by activity patterns of voxels that showed a univariate main effect of task difficulty. This was taken as compelling evidence for task-specific compensatory activity in the cuneus in advanced age.

The study is well-motivated and well-written. The authors used appropriate, rigorous methods that allowed them to control for a range of possible confounds or alternative explanations. Laudable aspects include the large sample with a wide and even age distribution, the validation of the in-scanner task performance against previous results obtained with a more standard version outside the scanner, and the control for vascular age-related differences in hemodynamic activity via a BOLD signal amplitude measure obtained from a separate resting-state fMRI scan. Overall, the conclusions are well-supported by the data.

Comment from Reviewing Editor: The revised manuscript has addressed the points raised during the review of the original submission.

eLife Assessment

This manuscript reports important findings that the methyltransferase METTL3 is involved in the repair of abasic sites and uracil in DNA, mediating resistance to floxuridine-driven cytotoxicity. Convincing evidence shows the involvement of m6A in DNA based on single cell imaging and mass spec data. The authors present evidence that the m6A signal does not result from bacterial contamination or RNA, but the text does not make this overly clear.

Reviewer #1 (Public review):

Summary:

The authors sought to identify unknown factors involved in the repair of uracil in DNA through a CRISPR knockout screen.

Strengths:

The screen identified both known and unknown proteins involved in DNA repair resulting from uracil or modified uracil base incorporation into DNA. The conclusion is that the protein activity of METTL3, which converts A nucleotides to 6mA nucleotides, plays a role in the DNA damage/repair response. The importance of METTL3 in DNA repair, and its colocalization with a known DNA repair enzyme, UNG2, is well characterized.

Weaknesses:

This reviewer identified no major weaknesses in this study. The manuscript could be improved by tightening the text throughout, and more accurate and consistent word choice around the origin of U and 6mA in DNA. The dUTP nucleotide is misincorporated into DNA, and 6mA is formed by methylation of the A base present in DNA. Using words like 6mA "deposition in DNA" seems to imply it results from incorporation of a methylated dATP nucleotide during DNA synthesis.

Reviewer #2 (Public review):

Summary:

In this work, the authors performed a CRISPR knockout screen in the presence of floxuridine, a chemotherapeutic agent that incorporates uracil and fluoro-uracil into DNA, and identified unexpected factors, such as the RNA m6A methyltransferase METTL3, as required to overcome floxuridine-driven cytotoxicity in mammalian cells. Interestingly, the observed N6-methyladenosine was embedded in DNA, which has been reported as DNA 6mA in mammalian genomes and is currently confirmed with mass spectrometry in this model. Therefore, this work consolidated the functional role of mammalian genomic DNA 6mA, and supported with solid evidence to uncover the METTL3-6mA-UNG2 axis in response to DNA base damage.

Strengths:

In this work, the authors took an unbiased, genome-wide CRISPR approach to identify novel factors involved in uracil repair with potential clinical interest.

The authors designed elegant experiments to confirm the METTL3 works through genomic DNA, adding the methylation into DNA (6mA) but not the RNA (m6A), in this base damage repair context. The authors employ different enzymes, such as RNase A, RNase H, DNase, and liquid chromatography coupled to tandem mass spectrometry to validate that METTL3 deposits 6mA in DNA in response to agents that increase genomic uracil.

They also have the Mettl3-KO and the METTL3 inhibition results to support their conclusion.

Weaknesses:

Although this study demonstrates that METTL3-dependent 6mA deposition in DNA is functionally relevant to DNA damage repair in mammalian cells, there are still several concerns and issues that need to be improved to strengthen this research.

First, in the whole paper, the authors never claim or mention the mammalian cell lines contamination testing result, which is the fundamental assay that has to be done for the mammalian cell lines DNA 6mA study.

Second, in the whole work, the authors have not supplied any genomic sequencing data to support their conclusions. Although the sequencing of DNA 6mA in mammalian models is challenging, recent breakthroughs in sequencing techniques, such as DR-Seq or NT/NAME-seq, have lowered the bar and improved a lot in the 6mA sequencing assay. Therefore, the authors should consider employing the sequencing methods to further confirm the functional role of 6mA in base repair.

Third, the authors used the METTL3 inhibitor and Mettl3-KO to validate the METTL3-6mA-UNG2 functional roles. However, the catalytic mutant and rescue of Mettl3 may be the further experiments to confirm the conclusion.

Reviewer #3 (Public review):

Summary:

The authors are showing evidence that they claim establishes the controversial epigenetic mark, DNA 6mA, as promoting genome stability.

Strengths:

The identification of a poorly understood protein, METTL3, and its subsequent characterization in DDR is of high quality and interesting.

Weaknesses:

(1) The very presence of 6mA (DNA) in mammalian DNA is still highly controversial and numerous studies have been conclusively shown to have reported the presence of 6mA due to technical artifacts and bacterial contamination. Thus, to my knowledge there is no clear evidence for 6mA as an epigenetic mark in mammals, and consequently, no evidence of writers and readers of 6mA. None of this is mentioned in the introduction. Much of the introduction can be reduced, but a paragraph clearly stating the controversy and lack of evidence for 6mA in mammals needs to be added, otherwise, the reader is given an entirely distorted view of the field.

These concerns must also be clearly in the limitations section and even in the results section which fails to nuance the authors' findings.

(2) What is the motivation for using HT-29 cells? Moreover, the materials and methods do not state how the authors controlled for bacterial contamination, which has been the most common cause of erroneous 6mA signals to date. Did the authors routinely check for mycoplasma?

(3) The single cell imaging of 6mA in various cells is nice. The results are confirmed by mass spec as an orthogonal approach. Another orthogonal and quantitative approach to assessing 6mA levels would be PacBio. Similarly, it is unclear why the authors have not performed dot-blots of 6mA for genomic DNA from the given cell lines.

(4) The results of Figure 3 need further investigation and validation. If the results are correct the authors are suggesting that the majority of 6mA in their cell lines is present in the DNA, and not the RNA, which is completely contrary to every other study of 6mA in mammalian cells that I am aware of. This could suggest that the antibody is not, in fact, binding to 6mA, but to unmodified adenine, which would explain why the signal disappears after DNAse treatment. Indeed, binding of 6mA to unmethylated DNA is a commonly known problem with most 6mA antibodies and is well described elsewhere.

(5) Given the lack of orthologous validation of the observed DNA 6mA and the lack of evidence supporting the presence of 6mA in mammalian DNA and consequently any functional role for 6mA in mammalian biology, the manuscript's conclusions need to be toned down significantly, and the inherent difficulty in assessing 6mA accurately in mammals acknowledged throughout.

Author response:

eLife Assessment <br /> This manuscript reports important findings that the methyltransferase METTL3 is involved in the repair of abasic sites and uracil in DNA, mediating resistance to floxuridine-driven cytotoxicity. The presented evidence for the involvement of m6A in DNA is incomplete and requires further validation with orthogonal approaches to conclusively show the presence of 6mA in the DNA and exclude that the source is RNA or bacterial contamination. 

We thank the editors for recognizing the importance of our work and the relevance of METTL3 in DNA repair. However, we wholly disagree with the second sentence in the eLife assessment, and we want to clarify why our evidence for the involvement of 6mA in DNA is complete.  

The identification of 6mA in DNA, upon DNA damage, is based first on immunofluorescence observations using an anti-m6A antibody. In this setting, removal of RNA with RNase treatment fails to reduce the 6mA signal, excluding the possibility that the source of signal is RNA. In contrast, removal of DNA with DNase treatment removes all 6mA signal, strongly suggesting that the species carrying the N6-methyladenosine modification is DNA (Figure 3D, E). Importantly, in Figure 3F, we provide orthogonal, quantitative mass spectrometry data that independently confirm this finding. Mass spectrometry-liquid chromatography of DNA analytes, conclusively shows the presence of 6mA in DNA upon treatment with DNA damaging agents and excludes that the source is RNA, based on exact mass. Reviewer #2 recognized the strengths of this approach to generate solid evidence for 6mA in DNA.

Cells only show the 6mA signal when treated with DNA damaging agents, and the 6mA is absent from untreated cells (Figure 3D, E, F). This provides strong evidence that the 6mA signal is not a result of bacterial contamination in our cell lines. Moreover, our cell lines are routinely tested for mycoplasma contamination. It could be possible that stock solutions of DNA damaging agents may be contaminated, but this would need to be true for all individual drugs and stocks tested. The data showing 6mA signal is not significantly different from untreated cells when a DNA damaging agent is combined with a METTL3 inhibitor (Figure 3G, H) provides strong evidence against bacterial contamination in our stocks.  

In summary, we provide conclusive evidence, based on orthogonal methods, that the METTL3-dependent N6-methyladenosine modification is deposited in DNA, not RNA, in response to DNA damage. 

Public Reviews: <br /> Reviewer #1 (Public review): <br /> Summary: 

The authors sought to identify unknown factors involved in the repair of uracil in DNA through a CRISPER knockout screen. 

Typo above: “CRISPER” should be “CRISPR”.

Strengths: 

The screen identified both known and unknown proteins involved in DNA repair resulting from uracil or modified uracil base incorporation into DNA. The conclusion is that the protein activity of METTL3, which converts A nucleotides to 5mA nucleotides, plays a role in the DNA damage/repair response. The importance of METTL3 in DNA repair, and its colocalization with a known DNA repair enzyme, UNG2, is well characterized. 

Typo above: “5mA” should be “6mA”.

Weaknesses: <br /> This reviewer identified no major weaknesses in this study. The manuscript could be improved by tightening the text throughout, and more accurate and consistent word choice around the origin of U and 6mA in DNA. The dUTP nucleotide is misincorporated into DNA, and 6mA is formed by methylation of the A base present in DNA. Using words like 6mA "deposition in DNA" seems to imply it results from incorporation of a methylated dATP nucleotide during DNA synthesis.

The increased presence of 6mA during DNA damage could result from methylation at the A base itself (within DNA) or from incorporation of pre-modified 6mA during DNA synthesis. Our data do not directly discriminate between these two mechanisms, and we will clarify this point in the discussion.

Reviewer #2 (Public review): <br /> Summary: <br /> In this work, the authors performed a CRISPR knockout screen in the presence of floxuridine, a chemotherapeutic agent that incorporates uracil and fluoro-uracil into DNA, and identified unexpected factors, such as the RNA m6A methyltransferase METTL3, as required to overcome floxuridine-driven cytotoxicity in mammalian cells. Interestingly, the observed N6-methyladenosine was embedded in DNA, which has been reported as DNA 6mA in mammalian genomes and is currently confirmed with mass spectrometry in this model. Therefore, this work consolidated the functional role of mammalian genomic DNA 6mA, and supported with solid evidence to uncover the METTL3-6mA-UNG2 axis in response to DNA base damage. <br /> Strengths: <br /> In this work, the authors took an unbiased, genome-wide CRISPR approach to identify novel factors involved in uracil repair with potential clinical interest. 

The authors designed elegant experiments to confirm the METTL3 works through genomic DNA, adding the methylation into DNA (6mA) but not the RNA (m6A), in this base damage repair context. The authors employ different enzymes, such as RNase A, RNase H, DNase, and liquid chromatography coupled to tandem mass spectrometry to validate that METTL3 deposits 6mA in DNA in response to agents that increase genomic uracil. <br /> They also have the Mettl3-KO and the METTL3 inhibition results to support their conclusion. <br /> Weaknesses:<br /> Although this study demonstrates that METTL3-dependent 6mA deposition in DNA is functionally relevant to DNA damage repair in mammalian cells, there are still several concerns and issues that need to be improved to strengthen this research.

First, in the whole paper, the authors never claim or mention the mammalian cell lines contamination testing result, which is the fundamental assay that has to be done for the mammalian cell lines DNA 6mA study.

Our cell lines are routinely tested for bacterial contamination, specifically mycoplasma, and we plan to state this information in a revised version of the manuscript.

Importantly, we do not observe 6mA in untreated cells, strongly suggesting that the 6mA signal observed is dependent on the presence of DNA damage and not caused by contamination in the cell lines (Figure 3D, E, F). While it could be possible that stock solutions of DNA damaging agents may be contaminated, this would need to be the case for all individual drugs and stocks tested that induce 6mA, which seems very unlikely. Finally, the data showing 6mA signal is not significantly different from untreated cells when a DNA damaging agent is combined with a METTL3 inhibitor (Figure 3 G, H) provides strong evidence against bacterial contamination in our drug stocks.

Second, in the whole work, the authors have not supplied any genomic sequencing data to support their conclusions. Although the sequencing of DNA 6mA in mammalian models is challenging, recent breakthroughs in sequencing techniques, such as DR-Seq or NT/NAME-seq, have lowered the bar and improved a lot in the 6mA sequencing assay. Therefore, the authors should consider employing the sequencing methods to further confirm the functional role of 6mA in base repair.

While we agree that it could be important to understand the precise genomic location of 6mA in relation to DNA damage, this is outside the scope of the current study. Moreover, this exercise may prove unproductive. If 6mA is enriched in DNA at damage sites or as DNA is replicated, the genomic mapping of 6mA is likely to be stochastic. If stochastic, it would be impossible to obtain the read depth necessary to map 6mA accurately.

Third, the authors used the METTL3 inhibitor and Mettl3-KO to validate the METTL3-6mA-UNG2 functional roles. However, the catalytic mutant and rescue of Mettl3 may be the further experiments to confirm the conclusion. 

We believe this to be an excellent suggestion from Reviewer #2 but we are unable to perform the proposed experiment at this time. We encourage future studies to explore the rescue experiment.

Reviewer #3 (Public review):

Summary:

The authors are showing evidence that they claim establishes the controversial epigenetic mark, DNA 6mA, as promoting genome stability.

Strengths:

The identification of a poorly understood protein, METTL3, and its subsequent characterization in DDR is of high quality and interesting.

Weaknesses:

(1) The very presence of 6mA (DNA) in mammalian DNA is still highly controversial and numerous studies have been conclusively shown to have reported the presence of 6mA due to technical artifacts and bacterial contamination. Thus, to my knowledge there is no clear evidence for 6mA as an epigenetic mark in mammals, and consequently, no evidence of writers and readers of 6mA. None of this is mentioned in the introduction. Much of the introduction can be reduced, but a paragraph clearly stating the controversy and lack of evidence for 6mA in mammals needs to be added, otherwise, the reader is given an entirely distorted view of the field.

These concerns must also be clearly in the limitations section and even in the results section which fails to nuance the authors' findings.

We agree with the reviewer that the presence and potential function of 6mA in mammalian DNA has been debated. Importantly, the debate regarding the presence and quantity of 6mA in DNA has been previously restricted to undamaged, baseline conditions. In complete agreement with this notion, we do not detect appreciable levels of 6mA in untreated cells. We will revise the introduction to introduce the debate about 6mA in DNA. We, however, want to highlight that our study provides for the first time, convincing evidence (based on orthogonal methods) that 6mA is present in DNA in response to a stimulus, DNA damage.

(2) What is the motivation for using HT-29 cells? Moreover, the materials and methods do not state how the authors controlled for bacterial contamination, which has been the most common cause of erroneous 6mA signals to date. Did the authors routinely check for mycoplasma?

HT-29 is a cell line of colorectal origin and chemotherapeutic agents that introduce uracil and uracil derivatives in DNA, as those used in this study, are relevant for the treatment of colorectal cancer. As indicated above, we do not observe 6mA in untreated cells, strongly suggesting that the 6mA signal observed is dependent on DNA damage and not caused by a potential bacterial contamination (Figure 3D, E, F). Additionally, our cell lines are routinely tested for bacterial contamination, specifically mycoplasma.

(3) The single-cell imaging of 6mA in various cells is nice but must be confirmed by orthogonal approaches. PacBio would provide an alternative and quantitative approach to assessing 6mA levels. Similarly, it is unclear why the authors have not performed dot-blots of 6mA for genomic DNA from the given cell lines.

We are confused by this point since an orthogonal approach to detect 6mA, mass spectrometry-liquid chromatography, was employed. This method does not use an antibody and confirms the increase of 6mA in DNA when cells were treated with DNA damaging agents. This data is presented in Figure 3F.

It is sensible to hypothesize that the localization of 6mA is consistent with DNA replication (like uracil deposition). In this event, the genomic mapping of 6mA is likely to be stochastic. This would make quantification with PacBio sequencing difficult because it would be very challenging to achieve the appropriate read depth to call a modified base.

Dot blots rely on an antibody and thus are not truly orthogonal to our immunofluorescence-based measurements. We preferred the mass spectrometry-liquid chromatography approach we took as a true orthogonal approach.

(4) The results of Figure 3 need further investigation and validation. If the results are correct the authors are suggesting that the majority of 6mA in their cell lines is present in the DNA, and not the RNA, which is completely contrary to every other study of 6mA in mammalian cells that I am aware of. This could suggest that the antibody is not, in fact, binding to 6mA, but to unmodified adenine, which would explain why the signal disappears after DNAse treatment. Indeed, binding of 6mA to unmethylated DNA is a commonly known problem with most 6mA antibodies and is well described elsewhere.

Based on this and the following comment, we are convinced that Reviewer #3 has overlooked two critical elements of our study:

First, the immunofluorescence work presented in Figure 3, showing 6mA signal in response to DNA damage, uses cells that were pre-extracted to remove excess cytoplasmic RNA. This method is often used in immunofluorescence experiments of this kind. The pre-extraction method removes most of the cytoplasmic content, and the majority of the cytoplasmic m6A RNA signal. Supplementary Figure 3D shows cells that have not been pre-extracted prior to staining. These images show the cytoplasmic m6A signal is abundant if we do not perform the pre-extraction step.

If the antibody used to label 6mA significantly reacted with unmodified adenine, we would expect a large signal in untreated or untreated and denatured conditions. In contrast, an increase in 6mA is not observed in either case.

Second, the orthogonal approach we employed, mass spectrometry coupled with liquid chromatography, measures 6mA DNA analytes specifically by exact mass. This approach does not depend on an antibody and yields results consistent with those from the immunofluorescence experiments.

(5) Given the lack of orthologous validation of the observed DNA 6mA and the lack of evidence supporting the presence of 6mA in mammalian DNA and consequently any functional role for 6mA in mammalian biology, the manuscript's conclusions need to be toned down significantly, and the inherent difficultly in assessing 6mA accurately in mammals acknowledged throughout.

Typo above: “difficultly” should be “difficulty”.

As discussed in response to prior comments, Figure 3 does provide two independent and orthologous methods that demonstrate 6mA presence in DNA specifically, and not RNA, in response to DNA damage. Complementary and orthogonal datasets are presented using either immunofluorescence microscopy or mass spectrometry-liquid chromatography of extracted DNA. The latter method does not rely on an antibody and can discriminate 6mA DNA versus RNA based on exact mass. We will revise the text to clarify that Figure 3F is a completely orthogonal approach.

Author response:

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

Public Reviews:

Reviewer #1 (Public Review):

Summary:

In this study, the authors describe the construction of an extremely large-scale anatomical model of juvenile rat somatosensory cortex (excluding the barrel region), which extends earlier iterations of these models by expanding across multiple interconnected cortical areas. The models are constructed in such a way as to maintain biological detail from a granular scale - for example, individual cell morphologies are maintained, and synaptic connectivity is founded on anatomical contacts. The authors use this model to investigate a variety of properties, from cell-type specific targeting (where the model results are compared to findings from recent large-scale electron microscopy studies) to network metrics. The model is also intended to serve as a platform and resource for the community by being a foundation for simulations of neuronal circuit activity and for additional anatomical studies that rely on the detailed knowledge of cellular identity and connectivity.

Strengths:

As the authors point out, the combination of scale and granularity of their model is what makes this study valuable and unique. The comparisons with recent electron microscopy findings are some of the most compelling results presented in the study, showing that certain connectivity patterns can arise directly from the anatomical configuration, while other discrepancies highlight where more selective targeting rules (perhaps based on molecular cues) are likely employed. They also describe intriguing effects of cortical thickness and curvature on circuit connectivity and characterize the magnitude of those effects on different cortical layers.

The detailed construction of the model is drawn on a wide range of data sources (cellular and synaptic density measures, neuronal morphologies, cellular composition measures, brain geometry, etc.) that are integrated together; other data sources are used for comparison and validation. This consolidation and comparison also represent a valuable contribution to the overall understanding of the modeled system.

We thank the reviewer for the kind comments.

Weaknesses:

The scale of the model, which is a primary strength, also can carry some drawbacks. In order to integrate all the diverse data sources together, many specific decisions must be made about, for example, translating findings from different species or regions to the modeled system, or deciding which aspects of the system can be assumed to be the same and which should vary. All these decisions will have effects on the predicted results from the model, which could limit the types of conclusions that can be made (both by the others and by others in the community who may wish to use the model for their own work).

We agree that this is a downside of the principle of biophysically detailed modeling that is best addressed by continuous refinement in collaboration with the community. We would like to once again invite any interested party to participate in this process.

As an example, while it is interesting that broad brain geometry has effects on network structure (Figure 7), it is not clear how those effects are actually manifested. I am not sure if some of the effects could be due to the way the model is constructed - perhaps there may be limited sets of morphologies that fit into columns of particular thicknesses, and those morphologies may have certain idiosyncrasies that could produce different statistics of connectivities where they are heavily used. That may be true to biology, but it may also be somewhat artifactual if, for example, the only neurons in the library that fit into that particular part of the cortex differ from the typical neurons that are actually found in that region (but may not have been part of the morphological sampling).

We agree that the limited pool of morphological reconstructions can lead to artifactual results in the way the reviewer pointed out. To investigate that hypothesis, we added a supplementary figure (S14) where we characterize (1): to what degree the morphological composition of a columnar subvolume reflects the overall composition of the model; and (2): The level of morphological diversity in each columnar subvolume. We discuss the results at the end of section 2.6. Briefly, while we cannot fully rule out the possibility of an artificial result, we found a high and virtually uniform level of morphological diversity in all columns and layers. This makes it unlikely that individual idiosyncratic morphologies strongly affect the local connectivity. However, we acknowledge that the minimum level of morphological diversity required is unknown. We believe that at this stage all we can do is characterize this and leave final interpretation to the reader.

I also wonder how much the assumption that the layers have the same relative thicknesses everywhere in the cortex affects these findings, since layer thicknesses do in fact vary across the cortex.

We agree that layer thickness variation would affect circuit properties. Variability of layer thickness can be split into two components: variability stemming from differences in total thickness, which our model covers, and variability of relative, i.e., normalized layer thickness, which we miss. In this region of cortex, though, data on the relative thickness of cortical layers is sparse. The Waxholm Atlas does not distinguish somatosensory cortical layers in its labels [Kleven et al, 2023]. Yusufoğulları (2015) compares layer thicknesses of rat hindlimb and barrel field regions. After normalization against total thickness, the relative difference increased towards the superficial layers from 0 in L6 to 33% in L1. Variability of normalized thicknesses within developed rat barrel cortex, based on layer boundaries reported in Narayanan et al. (2017) vary by 2% to 5% over approximately 2 mm. One major effect of such variability would be to scale the number of neurons in a given layer locally by the corresponding factors. For comparison, the resulting variability in neuron counts due to differences in conicality (Fig. 7D1) was around +-25%. A further effect of variable relative layer thickness would be its impact on the selection of suitable morphologies to be placed in the volume.

In summary, adjustment of layer thickness is a refinement which should be done in future versions of the model, once more data is available. The discussion section has been updated to acknowledge this limitation. However, as outlined at the beginning of this point-by-point reply, we will not conduct such updates to the model in the context of this manuscript, as it describes the version of the model used for a number of follow-up studies.

In addition, the complexity of the model means that some complicated analyses and decisions are only presented in this manuscript with perhaps a single panel and not much textual explanation. I find, for example, that the panels of Figure S2 seem to abstract or simplify many details to the point where I am not clear about what they are actually illustrating - how does Figure S2D represent the results of "the process illustrated in B"? Why are there abrupt changes in connectivity at region borders (shown as discontinuous colors), when dendrites and axons span those borders and so would imply interconnectivity across the borders? What do the histograms in E1 and E2 portray, and how are they related to each other?

We apologize for the confusion. We have updated the figure caption of Figure S2 to better explain its contents.

Overall, the model presented in this study represents an enormous amount of work and stands as a unique resource for the community, but also is made somewhat unwieldy for the community to employ due to the weight of its manifold specific construction decisions, size, and complexity.

Reviewer #2 (Public Review):

Summary:

The authors build a colossal anatomical model of juvenile rat non-barrel primary somatosensory cortex, including inputs from the thalamus. This enhances past models by incorporating information on the shape of the cortex and estimated densities of various types of excitatory and inhibitory neurons across layers. This is intended to enable an analysis of the micro- and mesoscopic organisation of cortical connectivity and to be a base anatomical model for large-scale simulations of physiology.

Strengths:

• The authors incorporate many diverse data sources on morphology and connectivity.

• This paper takes on the challenging task of linking micro- and mesoscale connectivity.

• By building in the shape of the cortex, the authors were able to link cortical geometry to connectivity. In particular, they make an unexpected prediction that cortical conicality affects the modularity of local connectivity, which should be testable.

• The author's analysis of the model led to the interesting prediction that layer 5 neurons connect local modules, which may be testable in the future, and provide a basis to link from detailed anatomy to functional computations.

• The visualisation of the anatomy in various forms is excellent.

• A subnetwork of the model is openly shared (but see question below).

We thank the reviewer for their kind comments.

Weaknesses:

• Why was non-barrel S1 of the juvenile rat cortex selected as the target for this huge modelling effort? This is not explained.

We have added an explanation of this decision to the third paragraph of the introduction.

• There is no effort to determine how specific or generalisable the findings here are to other parts of the cortex. Although there is a link to physiological modelling in another paper, there is no clear pathway to go from this type of model to understand how the specific function of the modelled areas may emerge here (and not in other cortical areas).

With respect to generality against specific findings, our philosophy is as follows: Despite the fact that most of our source data comes from juvenile rat somatosensory cortex, we also had to generalize many data sources across organisms, ages or regions. Hence, in this iteration we focused on investigating the general features of the (multi-region) mammalian cortex, e.g., high-order motifs, connected by L5 neurons across subregions or the effect of curvature on the connectivity. In the future, more specific data sources can be used to build diverging versions of the model, e.g. one for adult vs. juvenile rat. They can then be used to contrast the ages and focus on more specific findings. We already defined a number of structural metrics that can be used to contrast more specific versions of the model quantitatively.

We now clarify this pathway to understanding more specific function in the last paragraph of the discussion.

• In a few places the manuscript could be improved by being more specific in the language, for example:

- "our anatomy-based approach has been shown to be powerful", I would prefer instead to read about specific contributions of past papers to the field, and how this builds on them.

- similarly: "ensuring that the total number of synapses in a region-to-region pathway matches biology." Biology here is a loose term and implies too much confidence in the matching to some ground truth. Please instead describe the source of the data, including the type of experiment.

We have removed or rewritten the mentioned parts. We now clarify that we work based on biological estimates from experiments and cite the experiment sources. We also provide brief descriptions of the types of data and how they were derived.

• Some of the decisions seem a little ad-hoc, and the means to assess those decisions are not always available to the reader e.g.

- pg. 10. "Based on these results, we decided that the local connectome sufficed to model connectivity within a region.". What is the basis for this decision? Can it be formalised?

- "In the remaining layers the results of the objective classification were used to validate the class assignments of individual pyramidal cells. We found the objective classification to match the expert classification closely (i.e., for 80-90% of the morphologies). Consequently, we considered the expert classification to be sufficiently accurate to build the model." The description of the validation is a little informal. How many experts were there? What are their initials? Was inter-rater or intra-rater reliability assessed? What are these numbers? The match with Kanari's classification accuracy should be reported exactly. There are clearly experts among the author list, but we are all fallible without good controls in place, and they should be more explicit about those controls here, in my opinion.

- "Morphology selection was then performed as previously (Markram et al., 2015), that is, a morphology was selected randomly from the top 10% scorers for a given position." A lot of the decisions seem a little ad-hoc, without justification other than this group had previously done the same thing. For example, why 10% here? Shouldn't this be based on selecting from all of the reasonable morphologies?

We have clarified that the density of local connectivity is verified against the validation datasets by comparing the diagonals in Figure 4B, in addition to the quantification of Figure 4C.

For the classification, we have now published a detailed preprint describing the objective confirmation of expert classification by a variety of methods (see Kanari et al. 2024 https://www.biorxiv.org/content/10.1101/2024.09.13.612635v1). We cannot include the full methodology in the current paper, due to its large extent. For the benefit of the reader, we have included the appropriate citation and extended the short description of the methodology. As described in this paper, the classification accuracy varies per layer, cell type, etc. We have now described in more details these results, that can be accessed in details in out preprint.

• I would like to know if one of the key results relating to modularity and cortical geometry can be further explored. In particular, there seem to be sharp changes in the data at the end of the modelled cortical regions, which need to be explored or explained further.

We now explore these results further in supplementary figure S15, which we discuss in the results Section 2.6.

• The shape of the juvenile cortex - a key novelty of this work - was based on merely a scalar reduction of the adult cortex. This is very surprising, and surely an oversimplification. Huge efforts have gone into modelling the complex nonlinear development of the cortex, by teams including the developing Human Connectome Project. For such a fundamental aspect of this work, why isn't it possible to reconstruct the shape of this relatively small part of the juvenile rat cortex?

We agree that a more complex approach should be used in the future. However, as outlined at the beginning of this point-by-point reply, we will not conduct such updates to the model in the context of this manuscript, as it describes the version of the model used for a number of follow-up studies.

• The same relative laminar depths are used for all subregions. This will have a large impact on the model. However, relative laminar depths can change drastically across the cortex (see e.g. many papers by Palomero-Gallagher, Zilles, and colleagues). The authors should incorporate the real laminar depths, or, failing that, show evidence to show that the laminar depth differences across the subregions included in the model are negligible.

This point has also been raised by reviewer #1 above. For convenience, we repeat our reply below.

We agree that layer thickness variation would affect circuit properties. Variability of layer thickness can be split into two components: variability stemming from differences in total thickness, which our model covers, and variability of relative, i.e., normalized layer thickness, which we miss. In this region of cortex, though, data on the relative thickness of cortical layers is sparse. The Waxholm Atlas does not distinguish somatosensory cortical layers in its labels [Kleven et al, 2023]. Yusufoğulları (2015) compares layer thicknesses of rat hindlimb and barrel field regions. After normalization against total thickness, the relative difference increased towards the superficial layers from 0 in L6 to 33% in L1. Variability of normalized thicknesses within developed rat barrel cortex, based on layer boundaries reported in Narayanan et al. (2017) vary by 2% to 5% over approximately 2 mm. One major effect of such variability would be to scale the number of neurons in a given layer locally by the corresponding factors. For comparison, the resulting variability in neuron counts due to differences in conicality (Fig. 7D1) was around +-25%. A further effect of variable relative layer thickness would be its impact on the selection of suitable morphologies to be placed in the volume.

In summary, adjustment of layer thickness is a refinement which should be done in future versions of the model, once more data is available. The discussion section has been updated to acknowledge this limitation. However, as outlined at the beginning of this point-by-point reply, we will not conduct such updates to the model in the context of this manuscript, as it describes the version of the model used for a number of follow-up studies.

• The authors perform an affine mapping between mouse and rat cortex. This is again surprising. In human imaging, affine mappings are insufficient to map between two individual brains of the same species and nonlinear transformations are instead used. That an affine transformation should be considered sufficient to map between two different species is then very surprising. For some models, this may be fine, but there is a supposed emphasis here on biological precision in terms of anatomical location.

We agree that this is a weakness that we will address in future revisions of the model.

• One of the most interesting conclusions, that the connectivity pattern observed is in part due to cooperative synapse formation, is based on analyses that are unfortunately not shown.

We originally decided not to show this part as we underestimated the interest in this particular result. We have now included the result in supplementary figure S10 and discuss the figure in the results.

• Open code:

- Why is only a subvolume available to the community?

We have now made the entire model available under doi.org/10.7910/DVN/HISHXN. The Data and Code availability section has been updated to clarify this.

- Live nature of the model. This is such a colossal model, and effort, that I worry that it may be quite difficult to update in light of new data. For example, how much person and computer time would it take to update the model to account for different layer sizes across subregions? Or to more precisely account for the shape of the juvenile rat cortex?

To provide more information to people interested in participating in model refinements, we have added a new Figure 9. We discuss potential opportunities for refinement at the end of the discussion section.

Reviewer #3 (Public Review):

This manuscript reports a detailed model of the rat non-barrel somatosensory cortex, consisting of 4.2 million morphologically and biophysically detailed neuron models, arranged in space and connected according to highly sophisticated rules informed by diverse experimental data. Due to its breadth and sophistication, the model will undoubtedly be of interest to the community, and the reporting of anatomical details of modeling in this paper is important for understanding all the assumptions and procedures involved in constructing the model. While a useful contribution to this field, the model and the manuscript could be improved by employing data more directly and comparing simple features of the model's connectivity - in particular, connection probabilities - with relevant experimental data.

The manuscript is well-written overall but contains a substantial number of confusing or unclear statements, and some important information is not provided.

Below, major concerns are listed, followed by more specific but still important issues.

Major issues

(1) Cortical connectivity.

Section 2.3, "Local, mid-range and extrinsic connectivity modeled separately", and Figure 4: I am confused about what is done here and why. The authors have target data for connectivity (Figure 4B1). But then they use an apposition-based algorithm that results in connectivity that is quite different from the data (Figure 4B2, C). They then use a correction based on the data (Figure 4E) to arrive at a more realistic connectivity. Why not set the connectivity based on the data right away then? That would seem like a more straightforward approach.

We have completely re-written our description and discussion of connectivity in the model. We now more explicitly motivate our connectivity modeling choices in the first paragraph of section 2.3 of the results and in the second paragraph of the discussion.

The same comment applies to Section 2.4., "Specificity of axonal targeting": the distributions of synapses on different types of target cell compartments were not well captured by the original model based on axon-dendrite overlap and pruning, so the authors introduced further pruning to match data specificity. While details of this process and what worked and what didn't may be interesting to some, overall it is not surprising, as it has been well known that cell types exhibit connectivity that is much more specific than "Peters rule" or its simple variations. The question is, since one has the data, why not use the data in the first place to set up the connectivity, instead of using the convoluted process of employing axon-dendrite overlap followed by multiple corrections?

We would like to point out that we are not employing “Peters rule”, we now make this explicit in the revision in the first paragraph of section 2.3 of the results. Furthermore, we would argue that the match to the Motta et al. data indicates that our approach is more than just a “simple variation”. Finally, we believe that there is important insight in: 1. The specific ways in which the algorithm had to be changed to match the Schneider-Mizell data, e.g. that the connectivity of SST positive neurons did not have to be adapted at all. 2. That the specificity of the other two types could still be matched by a selection of a subset of axonal appositions (i.e., of potential synapses).

Most importantly, what is missing from the whole paper is the characterization of connection probabilities, at least for the local circuit within one area. Such connection probabilities can be obtained from the data that the authors already use here, such as the MICRONS dataset. Another good source of such data is Campagnola et al., Science, 2022. Both datasets are for mouse V1, but they provide a comprehensive characterization across all cortical layers, thus offering a good benchmark for comparison of the model with the data. It would be important for the authors to show how connection probabilities realized in their model for different cell types compared to these data.

We now report connection probabilities in the reworked figure 4 and compare them to reported connection probabilities from many different sources and labs in supplementary figure S8. We prefer a comparison to a wide range of sources to relying on a single report.

(2) Section 2.5, "Structure of thalamic inputs" and Figure 6.

The text in section 2.5 should provide more details on what was done - namely, that the thalamic axons were generated based on the axon density profiles and then synapses were established based on their overall with cortical dendrites. Figure S10 where the target axon densities from data and the model axon densities are compared is not even mentioned here. Now, Figure S10 only shows that the axon densities were generated in a way that matches the data reasonably well. However, how can we know that it results in connectivity that agrees with data? Are there data sources that can be used for that purpose? For example, the authors show that in their model "the peaks of the mean number of thalamic inputs per neuron occur at lower depths than the peaks of the synaptic density". Is this prediction of the model consistent with any available data?

Most importantly, the authors should show how the different cell types in their model are targeted by the thalamic inputs in each layer. Experimental studies have been done suggesting specificity in targeting of interneuron types by thalamic axons, such as PV cells being targeted strongly whereas SST and VIP cells being targeted less.

We have updated the Results section to provide context for the thalamic axon placement, and referred the reader to the methods for more detail. A reference to Figure S10 has now been added to this section as well.

As for validations of the structure of the thalamo-cortical inputs: We found that the existing literature on the topic, such as Cruikshank et al., 2007, 2010 and more recently Sermet et al., 2019, is predominately on the physiological strengths of the pathways. We acknowledge that the authors provide compelling arguments that their findings are likely partially due to differences in the anatomical innervation strengths. On the other hand, Sporns, 2013 cautioned against mixing up structural and functional connectivity. Overall, we believe that it is simply cleaner to perform this validation in the accompanying manuscript (“Part II: Physiology and Experimentation”), using the full physiological model. Note that we have actually performed that validation in the manuscript (see preprint under the following doi: 10.1101/2023.05.17.541168, Figure 3H1).

Note that a higher physiological strength onto PV+ neurons is observed.

(3) "We have therefore made not only the model but also most of our tool chain openly available to the public (Figure 1; step 7)."

In fact it is not the whole model that is made publicly available, but only about 5% of it (211,000 out of 4,200,000 neurons). Also, why is "most" of the tool chain made openly available, and not the whole tool chain?

We have now made the entire model available under doi.org/10.7910/DVN/HISHXN. This has also been added to the Key resource table.

With regard to the tool chain, everything is on our public github (https://github.com/BlueBrain/) except for the algorithm for detecting axonal appositions. For that tool there are currently unresolved potential copyright issues with former collaboration partners. We are working to resolve them.

Other issues

"At each soma location, a reconstruction of the corresponding m-type was chosen based on the size and shape of its dendritic and axonal trees (Figure S6). Additionally, it was rotated to according to the orientation towards the cortical surface at that point."

After this procedure, were cells additionally rotated around the white matter-pia axis? If yes, then how much and randomly or not? If not, then why not? Such rotations would seem important because otherwise additional order potentially not present in the real cortex is introduced in the model affecting connectivity and possibly also in vivo physiology (such as the dynamics of the extracellular electric field).

They are indeed additionally randomly rotated. We have clarified this in the revision.

The term "new in vivo reconstructions" for the 58 neurons used in this paper in addition to "in vitro reconstructions" is a misnomer. It is not straightforward to see where the procedure is described, but then one finds that the part of Methods that describes experimental manipulations is mostly about that (so, a clearer pointer to that part of Methods could be useful). However, the description in Methods makes it clear that it is only labeling that is done in vivo; the microscopy and reconstruction are done subsequently in vitro. I would recommend changing the terminology here, as it is confusing. Also, can the authors show reconstructions of these neurons in the supplementary figures? Is the reconstruction shown in Figure 4A representative?

The term is used because the staining is done in vivo. To the best of our knowledge, the reconstruction process cannot be performed in vivo. However, to avoid any confusion we modified the text to clarify this distinction to in-vivo stained.

With respect to the reconstruction in Figure 4: The intent of the panel is to demonstrate the concept of targeted long-range axons that our morphologies are missing, necessitating the use of a second algorithm for longer-range connectivity. As such, it is not one of the reconstructions we used, but one of Janelia MouseLight. While we mentioned MouseLight in the figure caption, we formulated it in a way that could be misunderstood to mean that we merely used the MouseLight browser to render one of our morphologies. We apologize for the confusion, and we have fixed the figure caption.

In this revision we have added exemplars of representative morphology reconstructions (in slice stained and in vivo stained) in a new supplementary figure, as requested (Figure S5). It is referenced in the last paragraph of section 2.1.

In the Discussion, "This was taken into account during the modeling of the anatomical composition, e.g. by using three-dimensional, layer-specific neuron density profiles that match biological measurements, and by ensuring the biologically correct orientation of model neurons with respect to the orientation towards the cortical surface. As local connectivity was derived from axo-dendritic appositions in the anatomical model, it was strongly affected by these aspects.

However, this approach alone was insufficient at the large spatial scale of the model, as it was limited to connections at distances below 1000μm."

As mentioned above, it is not clear that this approach was sufficient for local connectivity either. It would be great if the authors showed a systematic comparison of local connection probabilities between different cell types in their model with experimental data and commented here in the Discussion about how well the model agrees with the data.

As mentioned in the reply to a previous comment, we now report connection probabilities.

In the Discussion: "The combined connectome therefore captures important correlations at that level, such as slender-tufted layer 5 PCs sending strong non-local cortico-cortical connections, but thick-tufted layer 5 PCs not." (Also the corresponding findings in Results.)

If I understand this statement correctly, it may not agree with biological data. See analysis from MICRONS dataset in Bodor et al., https://www.biorxiv.org/content/10.1101/2023.10.18.562531v1.

Our statement was indeed misleading and formulated too strongly. While thick-tufted pyramidal cells do form long-range intra-cortical connections, the structural strength of these pathways is weaker than for slender-tufted PCs, which are associated with the IT (intra-telencephalic) projection type. We have made this clear in the revision.

Table 2 is confusing. What do pluses and minuses mean? What does it mean that some entries have two pluses? This table is not mentioned anywhere else in the text. If pluses mean some meaningful predictions of the model, then their distribution in the table seems quite liberal and arbitrary. It is not clear to me that the model makes that many predictions, especially for type-specificity and plasticity. Also, why is the hippocampus mentioned in this table? I don't see anything about the hippocampus anywhere else in the paper.

We have clarified the description of the table in its caption and removed references to hippocampus, which were left from an earlier draft of the paper.

In the Discussion, "Thus, we made the tools to improve our model also openly available (see Data and Code availability section)."

As mentioned before, the authors themselves write that they made "most of our tool chain openly available to the public", but not all of it.

With regard to the tool chain, everything is on our public github (https://github.com/BlueBrain/) except for the algorithm for detecting axonal appositions. For that tool there are currently unresolved potential copyright issues with former collaboration partners. We are working to resolve them.

Table S2 has multiple question marks. It is not clear whether the "predictions" listed in that table are truly well-thought-out and/or whether experimental confirmations are real.

Some of the citations in that table were broken due to technical difficulties with the citation manager used. We apologize and have fixed this in the revision.

Introduction: It would be quite appropriate to cite here Einevoll et al., Neuron, 2019 ("The Scientific Case for Brain Simulations").

We now reference this important work.

Recommendations for the authors:

Reviewing Editor's note:

Consultation with the reviewers highlighted three main issues: the integration of connection probability profiles, non-uniform cortical thickness, and the overall organization of the manuscript.

Reviewer #1 (Recommendations For The Authors):

Apart from the points discussed in the public review, my main concern is that the manuscript itself is not as tightly constructed as it should be, to the detriment of the reader's ability to understand the model itself and the conclusions from the presented analyses.

There are places where the text references seemingly incorrect figure panels or refers to panels that don't exist:

- Section 2.2, first paragraph - refers to Figure 2D, E but those panels do not exist in Figure 2.

- Section 2.2, second paragraph - refers to Figure 3D3 - perhaps it should be 3B3?

- Section 2.8, first paragraph - has no figure references but seems like it should be referring to parts of Figure 8 (perhaps Figure 8B1 specifically?)

- Is the reference to Figure S11A on page 16 supposed to be to S12A?

In other places, figure labels and descriptions are not clear, and terminology is not always well-defined or explained.

- Figure 8 and the associated section 2.8 are very difficult to draw conclusions from as presented - several of the terms used are opaque and not clearly defined in the text or legends. I could not easily infer how the normalization works for the "normalized node participation per layer", or what "position in simplex" means for "unique neurons in core", and what their "relative counts" are relative to.

- Are "targets" in Figure S12A the same as "sinks"? If so, it would be better to use a single term consistently throughout.

- Figure S12 - figures in part B do not have enough labels to interpret - what is the y-axis of the "rich-club analysis" graph? Also, the figures in part B bottom are labeled "long-range" rather than "mid-range" connections.

In general, I found the use of both letters and numbers for figure panels (e.g. Figure 7E1) more confusing than helpful - it didn't seem like panels with the same letter were visually grouped consistently, and it sometimes made it more difficult to follow the flow of a figure. I would recommend using only letters in nearly every case here.

We thank the reviewer for directing our attention to these issues. We have fixed them in the revision. However, we have decided to keep our original panel numbering scheme. Panels with the same letter are meant to be conceptually grouped as they address related or similar measures.

Other minor points:

- Section 2.4 - paragraph 2 - sentence 5 "inhbititory" -> "inhibitory".

- Figure 5B figure legend - references Schneider-Mizell et al. 2023 but probably should be Motta et al. 2019?

- Figure 5C - figure key "expcected" -> "expected".

- The lower part of Figure 7C looks like it belongs to panel D2 instead of panel C due to relative spacing.

We once again thank the reviewer, and we have fixed the listed issues.

Reviewer #2 (Recommendations For The Authors):

(1) Abstract:

- Is it really 'integrating whole brain-scale data'? This seems a bit misleading.

- "We delineated the limits of determining connectivity from anatomy" - here I think you mean determining connectivity from morphology, or dendrite/axon appositions. Electron microscopy is still anatomy and presumably would be much closer to function.

We originally used the term “anatomy” as connectivity depends on the correct placement of neurons in addition to their morphology. However, as the reviewer points out, this term is misleading as it would encompass electron microscopy, which can go beyond what we do with the model. We have updated the text to read “morphology and placement”.

(2) Introduction:

"Investigating the multi-scale interactions that shape perception requires a model of multiple cortical subregions with inter-region connectivity, but it also requires the subcellular resolution provided by a morphologically detailed model." - This statement, as written, is not true in my opinion. You can argue for the value of morphologically-detailed neuron models to the study of perception, but they are not required for the investigation of perception.

We have updated the text to be clearer: subcellular resolution is only required for certain aspects that are related to perception.

(3) Results:

- Pg. 9/10. There are three sentences in a row that are of the style: "ensuring that the total number of synapses in a region-to-region pathway matches biology." Biology here is a loose term and implies too much confidence in the matching to some ground truth. Please instead describe the source of the data, including the type of experiment here already. o Pg. 10. On the first read, I found it quite hard to follow what exactly was done in Figure 4.

What are the target values adapted from Reimann et al., 2019, for example?

- Pg. 10. "Based on these results, we decided that the local connectome sufficed to model connectivity within a region.". What is the basis for this decision? Can it be formalised? o Pg. 16, Figure 7 B-C. The apparent effect of geometry on modularity is potentially very interesting. However, are the sharp drop-offs in values for modularity (but also conicality and height) true, or are some artefacts due to columns at the edges of the sampled area?

We have discussed these points above in the general comments and strengths and weaknesses.

- Pg. 18. Simplicial cores define central subnetworks, tied together by mid-range connections. This work, in particular leading to the conclusion of the layer 5 highway hubs, stands out as being a successful attempt to simplify the highly detailed model to a degree that it generates useable new understanding.

We thank the reviewer for the kind comment.

(4) Figures:

Figure 2: The caption doesn't seem to match the Figure (e.g. there are no brain regions depicted in A). o Figure 4f. This is a key panel, but is squished into a small corner of Figure 4, and therefore hard-to-read.

We have fixed this in the revision.

Reviewer #3 (Recommendations For The Authors):

In Major comments, point (1) discusses the issue of connectivity known from data. For all the aspects of connectivity mentioned there, I would recommend the authors re-build their model using the connectivity data directly. It would be interesting to test whether a model constructed in such a way would have any difference in simulated neural activity relative to the model they have constructed.

This is indeed a very interesting avenue of research. However, we believe that it is best conducted in separate manuscripts. First, in Pokorny et al., 2024 (https://doi.org/10.1101/2024.05.24.593860) we conduct this investigation, comparing the emerging activity in the model to the one for simpler connectivity models. Additionally, in Egas-Santander et al., 2024 (https://www.biorxiv.org/content/10.1101/2024.03.15.585196v3) we found that simpler connectomes lead to less reliable spiking activity globally. Finally, in the accompanying manuscript (https://www.biorxiv.org/content/10.1101/2023.05.17.541168v5) we compare activity with and without the targeting specificity of Schneider-Mizell et al.

In Major comments, point (2) discusses thalamic inputs. I would recommend the authors to address the issues mentioned there.

We have replied to those comments above.

In addition, panels F and G of Figure 6 are mentioned in the caption but are not shown in the figure. In panel B, the choice of visualization is strange. It would make sense to show box plots for all the data instead of bars for mean values and points for randomly selected 50 cells. Panels E1 and E2 lack units.

We have removed mentions of panels F and G and changed the style of plot. Units for E1 and E2 are now explained in the figure caption.

In Major comments, point (3) touches upon model and tool sharing. I would recommend making such statements more accurate and reflecting what exactly is provided to the community since not everything is shared.

We have now made the entire model available under doi.org/10.7910/DVN/HISHXN.

With regard to the tool chain, everything is on our public github (https://github.com/BlueBrain/) except for the algorithm for detecting axonal appositions. For that tool there are currently unresolved potential copyright issues with former collaboration partners. We are working to resolve them.

I would recommend the authors address all the other points mentioned in the public review as well. In addition, below are some smaller issues that should be fixed.

Figure 2: the caption appears to be partially wrong and partially misassigned to the figure panels.

We fixed the issue.

Also, note that in L6 the types L6_TPC:A and L6_TPC:C are listed in the figure, but L6_TPC:B is not mentioned.

There is indeed no TPC:B type in layer 6. The distinction between TPC:A and TPC:B is based on early or late bifurcations of the apical dendrite and is only observed in layer 5.

Figure 3, panel B2: the caption refers to colors in panel (C), but the authors probably meant to refer to panel (A).

We fixed the issue.

"The placement of morphological reconstructions matched expectation, showing an appropriately layered structure with only small parts of neurites leaving the modeled volume (Figure 2D, E)."

Figure 2 does not have panels D and E.

"The volume was clearly dominated by dendrites, filling between 23% and 47% of the space, compared to 2% to 11% for axons (Figure 3D3)." There is no panel D or D3 in Figure 3.

"Recently, the MICrONS dataset (MICrONS-Consortium et al., 2021) has been analyzed with respect to the axonal targeting of inhibitory subtypes in a 100 x 100 μm subvolume spanning all layers (Schneider-Mizell et al., 2023)."

100 x 100 μm is an area (and should be 100 x 100 μm^2), not a volume.

Figure S11B requires a legend for the color map.

We fixed the issues.

Table S1: What is the difference between L6_BP and L6_BPC? They both are referred to as L6 bipolar cells.

We have changed the description of L6_BPC to “Layer 6 bitufted pyramidal cell”.

eLife Assessment

This manuscript reports a detailed model of juvenile rat somatosensory cortex, consisting of 4.2 million morphologically and biophysically detailed neuron models, arranged in space and connected according to diverse experimental data - a valuable tool for the field. The construction of the model is based on a methodology with solid supporting evidence. It should be noted that, by necessity, such a large-scale model development involves many assumptions, interpolations, and decisions that could have compounding downstream effects on further analyses that may be difficult to disambiguate.

Reviewer #1 (Public review):

Summary:

In this study, the authors describe the construction of an extremely large-scale anatomical model of juvenile rat somatosensory cortex (excluding the barrel region), which extends earlier iterations of these models by expanding across multiple interconnected cortical areas. The models are constructed in a way to maintain biological detail from a granular scale - for example, individual cell morphologies are maintained, and synaptic connectivity is founded on anatomical contacts. The authors use this model to investigate a variety of properties, from cell-type specific targeting (where the model results are compared to findings from recent large-scale electron microscopy studies) to network metrics. The model is also intended to serve as a platform and resource for the community by being a foundation for simulations of neuronal circuit activity and for additional anatomical studies that rely on the detailed knowledge of cellular identity and connectivity.

Strengths:

As the authors point out, the combination of scale and granularity of their model are what make this study valuable and unique. The comparisons with recent electron microscopy findings are some of the most compelling results presented in the study, showing that certain connectivity patterns can arise directly from the anatomical configuration, while other discrepancies highlight where more selective targeting rules (perhaps based on molecular cues) are likely employed. They also describe intriguing effects of cortical thickness and curvature on circuit connectivity and characterize the magnitude of those effects on different cortical layers.

The detailed construction of the model is drawn on wide range of data sources (cellular and synaptic density measures, neuronal morphologies, cellular composition measures, brain geometry, etc.) that are integrated together; other data sources are used for comparison and validation. This consolidation and comparison also represents a valuable contribution to the overall understanding of the modeled system.

Weaknesses:

The scale of the model, which is a primary strength, also can carry some drawbacks. In order to integrate all the diverse data sources together, many specific decisions must be made about, for example, translating findings from different species or regions to the modeled system, or deciding which aspects of the system can be assumed to be same and which should vary. All these decisions will have effects on the predicted results from the model, which could limit the types of conclusions that can be made (both by the others and by others in the community who may wish to use the model for their own work). However, the public release of the models and most of the associated tools does provide others a somewhat easier path to modify and evaluate this iteration of the model for their own studies.

Overall, the model presented in this study represents an enormous amount of work and stands as the basis for other work by the same group as well as a unique resource for the community, even while acknowledging that it may be somewhat unwieldy for the community to employ due to the weight of its manifold specific construction decisions, size, and complexity.

Reviewer #2 (Public review):

Summary:

The authors build a colossal anatomical model of juvenile rat non-barrel primary somatosensory cortex, including inputs from the thalamus. This enhances past models by incorporating information on the shape of the cortex and estimated densities of various types of excitatory and inhibitory neuron across layers. This is intended to enable analysis of the micro- and mesoscopic organisation of cortical connectivity and to be a base anatomical model for large-scale simulations of physiology.

Strengths:

• The authors incorporate many diverse data sources on morphology and connectivity.<br /> • This paper takes on the challenging task of linking micro- and meso-scale connectivity<br /> • By building in the shape of the cortex, the authors were able to link cortical geometry to connectivity. In particular they make an unexpected prediction that cortical conicality affects the modularity of local connectivity, which should be testable.<br /> • The author's analysis of the model led to the interesting prediction that layer 5 neurons' connect local modules, which may be testable in the future, and provide a basis to link from detailed anatomy to functional computations.<br /> • The visualisation of the anatomy in various forms is excellent<br /> • The model is openly shared

Weaknesses:

• There is no effort to determine how specific or generalisable the findings here are to other parts of cortex.<br /> • Although there is a link to physiological modelling in another paper, there is no clear pathway to go from this type of model to understanding how the specific function of the modelled areas may emerge here (and not in other cortical areas).<br /> • Some of the decisions seem a little ad-hoc, and the means to assess those decisions is not always easily available to the reader<br /> • The shape of the juvenile cortex - a key novelty of this work - was based on merely a scalar reduction of the adult cortex. This is very surprising, and surely an oversimplification. Huge efforts have gone into modelling the complex nonlinear development of cortex, by teams including the developing Human Connectome Project. For such a fundamental aspect of this work, why isn't it possible to reconstruct the shape of this relatively small part of juvenile rat cortex?<br /> • The same relative laminar depths are used for all subregions. This will have a large impact on the model. However, relative laminar depths can change drastically across the cortex (see e.g. many papers by Palomero-Gallagher, Zilles and colleagues). The authors should incorporate the real laminar depths, or, failing that, show evidence to show that the laminar depth differences across the subregions included in the model are negligible.<br /> • The authors perform an affine mapping between mouse and rat cortex. This is again surprising. In human imaging, affine mappings are insufficient to map between two individual brains of the same species, and nonlinear transformations are instead used. That an affine transformation should be considered sufficient to map between two different species is then very surprising. For some models, this may be fine, but there is a supposed emphasis here on biological precision in terms of anatomical location.<br /> o Live nature of the model. This is such a colossal model, and effort, that I worry that it may be quite difficult to update in light of new data. For example, how much person and compute time would it take to update the model to account for different layer sizes across subregions? Or to more precisely account for the shape of juvenile rat cortex?

Reviewer #3 (Public review):

This manuscript reports a detailed model of the rat non-barrel somatosensory cortex, consisting of 4.2 million morphologically and biophysically detailed neuron models, arranged in space and connected according to highly sophisticated rules informed by diverse experimental data. Due to its breadth and sophistication the model will undoubtedly be of interest to the community, and the reporting of anatomical details of modeling in this paper is important for understanding all the assumptions and procedures involved in constructing the model. While a useful contribution to this field, the model and the manuscript could be improved by employing data more directly and comparing simple features of the model's connectivity - in particular, connection probabilities - with relevant experimental data.

The manuscript is overall well-written, but contains a substantial number of confusing or unclear statements, and some important information is not provided.

Comments on revisions:

The authors mostly addressed all my points and improved the paper substantially. I do not have further extensive comments except one general point below.

Regarding section 2.3 and metrics of connectivity like pairwise connection probabilities, it is great that the authors rewrote that section and added comparisons with experimental data in Figs. 4 and S9. Unfortunately, what one finds when direct comparisons are made is that the modeled pairwise connectivity is quite different from the data. Fig. S9 shows that the model's results do not agree with data in about half of the cases (purple and red arrows). Similarly large discrepancies can be seen for some other metrics, like in Fig. S10B and S10C1,C2. (And similar concerns apply to thalamocortical connections in section 2.5, where it looks like little to no data are available to verify the pairwise connectivity between the thalamic and cortical neurons via a direct comparison.)

This is concerning since this model forms the basis for multiple other studies of cortical dynamics and function by the same group and potentially others in the community, with multiple papers relying on it, whereas basic properties of connectivity are apparently not captured well.

On the other hand, this is also a "glass half full" situation, showing that the sophisticated algorithms for establishing connections, developed by the authors, are working well in at least half of the connection types explored. It is therefore imperative that the authors continue refining these algorithms to capture the remaining half in future iterations and producing improved models that the community can better rely on.

Please also note that Fig. S11 does not have a caption.

eLife Assessment

This work is potentially important and largely convincing given the state-of-the-art approaches used to unravel the mechanism underlying the release of Claudins via Rho-mediated activation of Matriptase during tight junction formation. However, there are a few concerns. Addressing the following two major concerns a) showing Matriptase is indeed activated and b) Matriptase inhibition does not interfere with keratinocyte specification, would significantly improve the strength of the evidence. In addition, including quantifications, missing methods, and improving the rigor of the analyses would be helpful.

Reviewer #1 (Public review):

Summary:

The manuscript "Rho-ROCK liberates sequestered claudin for rapid de novo tight junction formation" by Cho and colleagues investigates de novo tight junction formation during the differentiation of immortalized human HaCaT keratinocytes to granular-like cells, as well as during epithelial remodeling that occurs upon the apoptotic of individual cells in confluent monolayers of the representative epithelial cell line EpH4. The authors demonstrate the involvement of Rho-ROCK with well-conducted experiments and convincing images. Moreover, they unravel the underlying molecular mechanism, with Rho-ROCK activity activating the transmembrane serine protease Matriptase, which in turn leads to the cleavage of EpCAM and TROP2, respectively, releasing Claudins from EpCAM/TROP2/Claudin complexes at the cell membrane to become available for polymerization and de novo tight junction formation. These functional studies in the two different cell culture systems are complemented by localization studies of the according proteins in the stratified mouse epidermis in vivo.

In total, these are new and very intriguing and interesting findings that add important new insights into the molecular mechanisms of tight junction formation, identifying Matriptase as the "missing link" in the cascade of formerly described regulators. The involvement of TROP2/EpCAM/Claudin has been reported recently (Szabo et al., Biol. Open 2022; Bugge lab), and Matriptase had been formerly described to be required for in tight junction formation as well, again from the Bugge lab. Yet, the functional correlation/epistasis between them, and their relation to Rho signaling, had not been known thus far.

However, experiments addressing the role of Matriptase require a little more work.

Strengths:

Convincing functional studies in two different cell culture systems, complemented by supporting protein localization studies in vivo. The manuscript is clearly written and most data are convincingly demonstrated, with beautiful images and movies.

Weaknesses:

The central finding that Rho signaling leads to increased Matriptase activity needs to be more rigorously demonstrated (e.g. western blot specifically detecting the activated version or distinguishing between the full-length/inactive and processed/active version).

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors investigate how epithelia maintain intercellular barrier function despite and during cellular rearrangements upon e.g. apoptotic extrusion in simple epithelia or regenerative turnover in stratified epithelia like this epidermis. A fundamental question in epithelial biology. Previous literature has shown that Rho-mediated local regulation of actomyosin is essential not only for cellular rearrangement itself but also for directly controlling tight junction barrier function. The molecular mechanics however remained unclear. Here the authors use extensive fluorescent imaging of fixed and live cells together with genetic and drug-mediated interference to show that Rho activation is required and sufficient to form novo tight junctional strands at intercellular contacts in epidermal keratinocytes (HaCat) and mammary epithelial cells. After having confirmed previous literature they then show that Rho activation activates the transmembrane protease Matriptase which cleaves EpCAM and TROP2, two claudin-binding transmembrane proteins, to release claudins and enable claudin strand formation and therefore tight junction barrier function.

Strengths:

The presented mechanism is shown to be relevant for epithelial barriers being conserved in simple and stratifying epithelial cells and mainly differs due to tissue-specific expression of EpCAM and TROP2. The authors present careful state-of-the-art imaging and logical experiments that convincingly support the statements and conclusion. The manuscript is well-written and easy to follow.

Weaknesses:

Whereas the in vitro evidence of the presented mechanism is strongly supported by the data, the in vivo confirmation is mostly based on the predicted distribution of TROP2. Whereas the causality of Rho-mediated Matriptase activation has been nicely demonstrated it remains unclear how Rho activates Matriptase.

Author response:

Reviewer #1 (Public review):

Summary:

The manuscript "Rho-ROCK liberates sequestered claudin for rapid de novo tight junction formation" by Cho and colleagues investigates de novo tight junction formation during the differentiation of immortalized human HaCaT keratinocytes to granular-like cells, as well as during epithelial remodeling that occurs upon the apoptotic of individual cells in confluent monolayers of the representative epithelial cell line EpH4. The authors demonstrate the involvement of Rho-ROCK with well-conducted experiments and convincing images. Moreover, they unravel the underlying molecular mechanism, with Rho-ROCK activity activating the transmembrane serine protease Matriptase, which in turn leads to the cleavage of EpCAM and TROP2, respectively, releasing Claudins from EpCAM/TROP2/Claudin complexes at the cell membrane to become available for polymerization and de novo tight junction formation. These functional studies in the two different cell culture systems are complemented by localization studies of the according proteins in the stratified mouse epidermis in vivo.

In total, these are new and very intriguing and interesting findings that add important new insights into the molecular mechanisms of tight junction formation, identifying Matriptase as the "missing link" in the cascade of formerly described regulators. The involvement of TROP2/EpCAM/Claudin has been reported recently (Szabo et al., Biol. Open 2022; Bugge lab), and Matriptase had been formerly described to be required for in tight junction formation as well, again from the Bugge lab. Yet, the functional correlation/epistasis between them, and their relation to Rho signaling, had not been known thus far.

However, experiments addressing the role of Matriptase require a little more work.

Strengths:

Convincing functional studies in two different cell culture systems, complemented by supporting protein localization studies in vivo. The manuscript is clearly written and most data are convincingly demonstrated, with beautiful images and movies.

Weaknesses:

The central finding that Rho signaling leads to increased Matriptase activity needs to be more rigorously demonstrated (e.g. western blot specifically detecting the activated version or distinguishing between the full-length/inactive and processed/active version).

We plan to provide more direct evidence that matriptase activation is regulated by the Rho-ROCK pathway, utilizing antibodies that specifically recognize the activated form of matriptase.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors investigate how epithelia maintain intercellular barrier function despite and during cellular rearrangements upon e.g. apoptotic extrusion in simple epithelia or regenerative turnover in stratified epithelia like this epidermis. A fundamental question in epithelial biology. Previous literature has shown that Rho-mediated local regulation of actomyosin is essential not only for cellular rearrangement itself but also for directly controlling tight junction barrier function. The molecular mechanics however remained unclear. Here the authors use extensive fluorescent imaging of fixed and live cells together with genetic and drug-mediated interference to show that Rho activation is required and sufficient to form novo tight junctional strands at intercellular contacts in epidermal keratinocytes (HaCat) and mammary epithelial cells. After having confirmed previous literature they then show that Rho activation activates the transmembrane protease Matriptase which cleaves EpCAM and TROP2, two claudin-binding transmembrane proteins, to release claudins and enable claudin strand formation and therefore tight junction barrier function.

Strengths:

The presented mechanism is shown to be relevant for epithelial barriers being conserved in simple and stratifying epithelial cells and mainly differs due to tissue-specific expression of EpCAM and TROP2. The authors present careful state-of-the-art imaging and logical experiments that convincingly support the statements and conclusion. The manuscript is well-written and easy to follow.

Weaknesses:

Whereas the in vitro evidence of the presented mechanism is strongly supported by the data, the in vivo confirmation is mostly based on the predicted distribution of TROP2. Whereas the causality of Rho-mediated Matriptase activation has been nicely demonstrated it remains unclear how Rho activates Matriptase.

As noted, while we have demonstrated that Rho activation is both necessary and sufficient to induce matriptase activation, the precise mechanism by which Rho mediates this activation remains unclear. As discussed in the manuscript, several potential molecular mechanisms could underlie the contribution of Rho to matriptase activation. As part of our future work, we intend to systematically investigate each of these mechanisms.

eLife Assessment

The ingenious design in this study achieved the observation of 3D cell spheroids from additional lateral view and gained more comprehensive information than the traditional one angle of imaging. This extended the methods to investigate cell behaviors in the growth or migration of tumor organoids in a time-lapse manner and these extensions should be valuable to the field. The authors provide solid evidence that the methods work as described.

Reviewer #1 (Public review):

Summary:

The ingenious design in this study achieved the observation of 3D cell spheroids from an additional lateral view and gained more comprehensive information than the traditional one angle of imaging, which extensively extended the methods to investigate cell behaviors in the growth or migration of tumor organoids in the present study. I believe that this study opens an avenue and provides an opportunity to characterize the spheroid formation dynamics from different angles, in particular side-view with high resolution, in other organoids study in the future.

Reviewer #2 (Public review):

Summary:

The author developed a new device to overcome current limitations in the imaging process of 3D spheroidal structures. In particular, they created a system to follow in real-time tumour spheroid formation, fusion and cell migration without disrupting their integrity. The system has also been exploited to test the effects of a therapeutic agent (chemotherapy) and immune cells.

Strengths:

The system allows the in situ observation of the 3D structures along the 3 axes (x,y and z) without disrupting the integrity of the spheroids; in a time-lapse manner it is possible to follow the formation of the 3D structure and the spheroids fusion from multiple angles, allowing a better understanding of the cell aggregation/growth and kinetic of the cells.

Interestingly the system allows the analysis of cell migration/ escape from the 3D structure analysing not only the morphological changes in the periphery of the spheroids but also from the inner region demonstrating that the proliferating cells in the periphery of the structure are more involved in the migration and dissemination process. The application of the system in the study of the effects of doxorubicin and NK cells would give new insights in the description of the response of tumor 3D structure to killing agents.

Author response:

Reviewer #1:

We sincerely thank you for your thoughtful review and constructive comments on our work and we appreciate your positive assessment of our study’s innovative design, which allows for improved observation of 3D cell spheroids from an additional lateral view. Your comments underscore the importance of our approach in advancing methods for investigating cell behaviors in tumor organoid studies.

In response to your suggestions, we will first add a detailed image of the ‘First surface mirror’ in Fig. 1 to provide a reference for readers and other researchers, thereby facilitating broader use of this method in similar observations. Regarding the suitable sample sizes for this device, as the spheroid sizes are relatively small compared to the mirror and culture dish, we have been able to image samples up to 5 mm in height, which provides ample capacity for most spheroids under 1 mm. We will include additional experiments and explanations in the manuscript to clarify this further.

Concerning the ring-shaped seeding pattern of spheroids, we have conducted extensive culture experiments to optimize this method. The agarose microwells-based method has proven to be highly tolerant of variations. Within these microwells, cells have a propensity to self-aggregate, leading to the formation of spheroid structures. We will add a discussion in the revised manuscript to address this issue.

Lastly, this device can accommodate the fluorescence imaging of 3D spheroid samples. We will supplement the discussion with a schematic illustrating the principles of fluorescence imaging using this device, providing a foundation for future work in this area. We will also regarding language improvements to enhance the overall quality of the manuscript.

Thank you once again for your valuable insights, which have greatly contributed to the strengthening of our manuscript.

Reviewer #2:

We sincerely thank you for your detailed and supportive review of our manuscript. Your recognition of our system’s capabilities for in situ observation of 3D structures along multiple axes, as well as its potential applications in studying therapeutic effects, is highly encouraging. Your comments on the advantages of this system for analyzing cell migration, morphological changes, and responses to therapeutic agents are especially appreciated.

Thank you again for your thoughtful feedback and for highlighting the contributions of our work. Your insights have been invaluable in refining the focus and clarity of our study, and we hope that our revisions meet your expectations.

eLife Assessment

In this valuable study, the authors used an elegant genetic approach to delete EED at the post-neural crest induction stage. The usage of the single-cell RNA-seq analysis method is extremely suitable to determine changes in the cell type-specific gene expression during development. Results backed by solid evidence demonstrate that Eed is required for craniofacial osteoblast differentiation and mesenchymal proliferation after the induction of the neural crest.

Reviewer #1 (Public review):

Epigenetic regulation complex (PRC2) is essential for neural crest specification, and its misregulation has been shown to cause severe craniofacial defects. This study shows that Eed, a core PRC2 component, is critical for craniofacial osteoblast differentiation and mesenchymal proliferation after neural crest induction. Using mouse genetics and single-cell RNA sequencing, the researcher found that conditional knockout of Eed leads to significant craniofacial hypoplasia, impaired osteogenesis, and reduced proliferation of mesenchymal cells in post-migratory neural crest populations.

Overall, the study is superficial and descriptive. No in-depth mechanism was analyzed and the phenotype analysis is not comprehensive.

Reviewer #2 (Public review):

Summary:

The role of PRC2 in post-neural crest induction was not well understood. This work developed an elegant mouse genetic system to conditionally deplete EED upon SOX10 activation. Substantial developmental defects were identified for craniofacial and bone development. The authors also performed extensive single-cell RNA sequencing to analyze differentiation gene expression changes upon conditional EED disruption.

Strengths:

(1) Elegant genetic system to ablate EED post neural crest induction.

(2) Single-cell RNA-seq analysis is extremely suitable for studying the cell type-specific gene expression changes in developmental systems.

Weaknesses:

(1) Although this study is well designed and contains state-of-the-art single-cell RNA-seq analysis, it lacks the mechanistic depth in the EED/PRC2-mediated epigenetic repression. This is largely because no epigenomic data was shown.

(2) The mouse model of conditional loss of EZH2 in neural crest has been previously reported, as the authors pointed out in the discussion. What is novel in this study to disrupt EED? Perhaps a more detailed comparison of the two mouse models would be beneficial.

(3) The presentation of the single-cell RNA-seq data may need improvement. The complexity of the many cell types blurs the importance of which cell types are affected the most by EED disruption.

(4) While it's easy to identify PRC2/EED target genes using published epigenomic data, it would be nice to tease out the direct versus indirect effects in the gene expression changes (e.g Figure 4e).

Author response:

Public reviews:

Reviewer #1:

Epigenetic regulation complex (PRC2) is essential for neural crest specification, and its misregulation has been shown to cause severe craniofacial defects. This study shows that Eed, a core PRC2 component, is critical for craniofacial osteoblast differentiation and mesenchymal proliferation after neural crest induction. Using mouse genetics and single-cell RNA sequencing, the researcher found that conditional knockout of Eed leads to significant craniofacial hypoplasia, impaired osteogenesis, and reduced proliferation of mesenchymal cells in post-migratory neural crest populations.

Overall, the study is superficial and descriptive. No in-depth mechanism was analyzed and the phenotype analysis is not comprehensive.

We thank the reviewer for sharing their expertise and for taking the time to provide a helpful suggestion to improve our study. We are gratified that the striking phenotypes we report from Eed loss in post-migratory neural crest craniofacial tissues were appreciated. The breadth and depth of our phenotyping techniques, including skeletal staining, micro-CT, echocardiogram, immunofluorescence, histology, and unbiased single-cell gene expression analysis, provide comprehensive data in support our conclusion that PRC2 is required for craniofacial osteoblast differentiation. We hypothesize that epigenetic regulation of chromatin accessibility downstream of PRC2 activity is the molecular mechanism that underlies these phenotypes. To test this hypothesis in our revision, we are using CUT&Tag to profile H3K27me3 epigenetic modifications genome-wide and at the loci encoding the differentially expressed genes revealed by our single-cell transcriptomics in developing craniofacial structures. We anticipate that these experiments will reveal an epigenetic mechanism underlying the phenotypes we report from Eed loss in post-migratory neural crest craniofacial tissues.

Reviewer #2:

Summary:The role of PRC2 in post-neural crest induction was not well understood. This work developed an elegant mouse genetic system to conditionally deplete EED upon SOX10 activation. Substantial developmental defects were identified for craniofacial and bone development. The authors also performed extensive single-cell RNA sequencing to analyze differentiation gene expression changes upon conditional EED disruption.

Strengths:

(1) Elegant genetic system to ablate EED post neural crest induction.

(2) Single-cell RNA-seq analysis is extremely suitable for studying the cell type-specific gene expression changes in developmental systems.

We thank the reviewer for their generous and helpful comments on our study. We are pleased that our mouse genetic and single-cell RNA sequencing approaches were appropriate in pairing the craniofacial phenotypes we report with distinct gene expression changes in post-migratory neural crest tissues upon Eed deletion.

Weaknesses:

(1) Although this study is well designed and contains state-of-the-art single-cell RNA-seq analysis, it lacks the mechanistic depth in the EED/PRC2-mediated epigenetic repression. This is largely because no epigenomic data was shown.

Thank you for this suggestion. As described in response to Reviewer #1, we will include H2K27me3 CUT&Tag data in craniofacial tissue harvested from E12.5 and E16.5 Sox10-Cretg+ Eedfl/fl and Sox10-Cretg+ Eedfl/wt  embryos in our revision. Our analyses will including genome-wide and targeted metaplot visualizations across genotypes and developmental timepoints and assess how H3K27me3 occupancy relates to gene expression changes in our single-cell RNA sequencing data.

(2) The mouse model of conditional loss of EZH2 in neural crest has been previously reported, as the authors pointed out in the discussion. What is novel in this study to disrupt EED? Perhaps a more detailed comparison of the two mouse models would be beneficial.

We acknowledge the study the reviewer has indicated (Schwarz et al. Development 2014). This elegant investigation uses Wnt1-Cre to delete Ezh2 and found a similar phenotype to ours in the form of catastrophic craniofacial hypoplasia. We sought to add depth to the study of PRC2’s vital role in neural crest development by ablating Eed, which has a unique function in the PRC2 complex by binding to H3K27me3 and allosterically activating Ezh2. In this sense, we sought to test if phenotypes arising from deletion of Eed, the PRC2 “reader”, differ from phenotypes arising from deletion of Ezh2, the PRC2 “writer”, in neural crest derived tissues. Due to limitations associated with the Wnt1-Cre transgene (Lewis et al. Developmental Biology 2013), we used the Sox10-Cre allele which targets the migratory neural crest and is completely recombined by E10.5, instead of Wnt1-Cre which targets pre-migratory neural crest cells. A more detailed comparison of these mouse models will be included in the Discussion section of our revised manuscript, and we thank the reviewer for this thoughtful suggestion.

(3) The presentation of the single-cell RNA-seq data may need improvement. The complexity of the many cell types blurs the importance of which cell types are affected the most by EED disruption.

We agree with the reviewer’s critique of the scRNA-seq data presentation. Because Sox10+ cells were not sorted (via FACS, for example) from craniofacial tissues before single-cell RNA sequencing, we identified a breath of cell types in UMAP space unrelated to epigenetic disruption of neural crest derived tissues. We will include subcluster visualization plots in the figures of our revised manuscript to highlight specific changes in clusters, such as osteoblasts and mesenchymal stem cells, that arise from Eed loss in post-migratory neural crest craniofacial tissues.

(4) While it's easy to identify PRC2/EED target genes using published epigenomic data, it would be nice to tease out the direct versus indirect effects in the gene expression changes (e.g Figure 4e).

We agree with the reviewer that our single-cell RNA sequencing data do not provide insight into direct versus indirect changes in gene expression downstream of PRC2. We hope that the aforementioned CUT&Tag experiment will provide the necessary mechanistic insight into H3K27me3 occupancy and direct effects on gene expression resulting from PRC2 inactivation in our mouse model.

eLife Assessment

This study introduces an important approach using selection linked integration (SLI) to generate Plasmodium falciparum lines expressing single, specific surface adhesins PfEMP1 variants, enabling precise study of PfEMP1 trafficking, receptor binding, and cytoadhesion. By moving the system to different parasite strains and introducing an advanced SLI2 system for additional genomic edits, this work provides compelling evidence for an innovative and rigorous platform to explore PfEMP1 biology and identify novel proteins essential for malaria pathogenesis including immune evasion.

Reviewer #1 (Public review):

One of the roadblocks in PfEMP1 research has been the challenges in manipulating var genes to incorporate markers to allow the transport of this protein to be tracked and to investigate the interactions taking place within the infected erythrocyte. In addition, the ability of Plasmodium falciparum to switch to different PfEMP1 variants during in vitro culture has complicated studies due to parasite populations drifting from the original (manipulated) var gene expression. Cronshagen et al have provided a useful system with which they demonstrate the ability to integrate a selectable drug marker into several different var genes that allows the PfEMP1 variant expression to be 'fixed'. This on its own represents a useful addition to the molecular toolbox and the range of var genes that have been modified suggests that the system will have broad application. As well as incorporating a selectable marker, the authors have also used selective linked integration (SLI) to introduce markers to track the transport of PfEMP1, investigate the route of transport, and probe interactions with PfEMP1 proteins in the infected host cell.

What I particularly like about this paper is that the authors have not only put together what appears to be a largely robust system for further functional studies, but they have used it to produce a range of interesting findings including:

- Co-activation of rif and var genes when in a head-to-head orientation.

- The reduced control of expression of var genes in the 3D7-MEED parasite line.

- More support for the PTEX transport route for PfEMP1.

- Identification of new proteins involved in PfEMP1 interactions in the infected erythrocyte, including some required for cytoadherence.

In most cases the experimental evidence is straightforward, and the data support the conclusions strongly. The authors have been very careful in the depth of their investigation, and where unexpected results have been obtained, they have looked carefully at why these have occurred.

(1) In terms of incorporating a drug marker to drive mono-variant expression, the authors show that they can manipulate a range of var genes in two parasite lines (3D7 and IT4), producing around 90% expression of the targeted PfEMP1. Removal of drug selection produces the expected 'drift' in variant types being expressed. The exceptions to this are the 3D7-MEED line, which looks to be an interesting starting point to understand why this variant appears to have impaired mutually exclusive var gene expression and the EPCR-binding IT4var19 line. This latter finding was unexpected and the modified construct required several rounds of panning to produce parasites expressing the targeted PfEMP1 and bind to EPCR. The authors identified a PTP3 deficiency as the cause of the lack of PfEMP1 expression, which is an interesting finding in itself but potentially worrying for future studies. What was not clear was whether the selected IT4var19 line retained specific PfEMP1 expression once receptor panning was removed.

(2) The transport studies using the mDHFR constructs were quite complicated to understand but were explained very clearly in the text with good logical reasoning.

(3) By introducing a second SLI system, the authors have been able to alter other genes thought to be involved in PfEMP1 biology, particularly transport. An example of this is the inactivation of PTP1, which causes a loss of binding to CD36 and ICAM-1. It would have been helpful to have more insight into the interpretation of the IFAs as the anti-SBP1 staining in Figure 5D (PTP-TGD) looks similar to that shown in Figure 1C, which has PTP intact. The anti-EXP2 results are clearly different.

(4) It is good to see the validation of PfEMP1 expression includes binding to several relevant receptors. The data presented use CHO-GFP as a negative control, which is relevant, but it would have been good to also see the use of receptor mAbs to indicate specific adhesion patterns. The CHO system if fine for expression validation studies, but due to the high levels of receptor expression on these cells, moving to the use of microvascular endothelial cells would be advisable. This may explain the unexpected ICAM-1 binding seen with the panned IT4var19 line.

(5) The proxiome work is very interesting and has identified new leads for proteins interacting with PfEMP1, as well as suggesting that KAHRP is not one of these. The reduced expression seen with BirA* in position 3 is a little concerning but there appears to be sufficient expression to allow interactions to be identified with this construct. The quantitative impact of reduced expression for proxiome experiments will clearly require further work to define it.

(6) The reduced receptor binding results from the TryThrA and EMPIC3 knockouts were very interesting, particularly as both still display PfEMP1 on the surface of the infected erythrocyte. While care needs to be taken in cross-referencing adhesion work in P. berghei and whether the machinery truly is functionally orthologous, it is a fair point to make in the discussion. The suggestion that interacting proteins may influence the "correct presentation of PfEMP1" is intriguing and I look forward to further work on this.<br /> Overall, the authors have produced a useful and reasonably robust system to support functional studies on PfEMP1, which may provide a platform for future studies manipulating the domain content in the exon 1 portion of var genes. They have used this system to produce a range of interesting findings and to support its use by the research community.<br /> Finally, a small concern. Being able to select specific var gene switches using drug markers could provide some useful starting points to understand how switching happens in P. falciparum. However, our trypanosome colleagues might remind us that forcing switches may show us some mechanisms but perhaps not all.

Reviewer #2 (Public review):

Summary

Croshagen et al develop a range of tools based on selection-linked integration (SLI) to study PfEMP1 function in P. falciparum. PfEMP1 is encoded by a family of ~60 var genes subject to mutually exclusive expression. Switching expression between different family members can modify the binding properties of the infected erythrocyte while avoiding the adaptive immune response. Although critical to parasite survival and Malaria disease pathology, PfEMP1 proteins are difficult to study owing to their large size and variable expression between parasites within the same population. The SLI approach previously developed by this group for genetic modification of P. falciparum is employed here to selectively and stably activate the expression of target var genes at the population level. Using this strategy, the binding properties of specific PfEMP1 variants were measured for several distinct var genes with a novel semi-automated pipeline to increase throughput and reduce bias. Activation of similar var genes in both the common lab strain 3D7 and the cytoadhesion competent FCR3/IT4 strain revealed higher binding for several PfEMP1 IT4 variants with distinct receptors, indicating this strain provides a superior background for studying PfEMP1 binding. SLI also enables modifications to target var gene products to study PfEMP1 trafficking and identify interacting partners by proximity-labeling proteomics, revealing two novel exported proteins required for cytoadherence. Overall, the data demonstrate a range of SLI-based approaches for studying PfEMP1 that will be broadly useful for understanding the basis for cytoadhesion and parasite virulence.

Comments

(1) While the capability of SLI to actively select var gene expression was initially reported by Omelianczyk et al., the present study greatly expands the utility of this approach. Several distinct var genes are activated in two different P. falciparum strains and shown to modify the binding properties of infected RBCs to distinct endothelial receptors; development of SLI2 enables multiple SLI modifications in the same parasite line; SLI is used to modify target var genes to study PfEMP1 trafficking and determine PfEMP1 interactomes with BioID. Curiously, Omelianczyk et al activated a single var (Pf3D7_0421300) and observed elevated expression of an adjacent var arranged in a head-to-tail manner, possibly resulting from local chromatin modifications enabling expression of the neighboring gene. In contrast, the present study observed activation of neighboring genes with head-to-head but not head-to-tail arrangement, which may be the result of shared promoter regions. The reason for these differing results is unclear although it should be noted that the two studies examined different var loci.

(2) The IT4var19 panned line that became binding-competent showed increased expression of both paralogs of ptp3 (as well as a phista and gbp), suggesting that overexpression of PTP3 may improve PfEMP1 display and binding. Interestingly, IT4 appears to be the only known P. falciparum strain (only available in PlasmoDB) that encodes more than one ptp3 gene (PfIT_140083100 and PfIT_140084700). PfIT_140084700 is almost identical to the 3D7 PTP3 (except for a ~120 residue insertion in 3D7 beginning at residue 400). In contrast, while the C-terminal region of PfIT_140083100 shows near-perfect conservation with 3D7 PTP3 beginning at residue 450, the N-terminal regions between the PEXEL and residue 450 are quite different. This may indicate the generally stronger receptor binding observed in IT4 relative to 3D7 results from increased PTP3 activity due to multiple isoforms or that specialized trafficking machinery exists for some PfEMP1 proteins.

Reviewer #3 (Public review):

Summary:

The submission from Cronshagen and colleagues describes the application of a previously described method (selection linked integration) to the systematic study of PfEMP1 trafficking in the human malaria parasite Plasmodium falciparum. PfEMP1 is the primary virulence factor and surface antigen of infected red blood cells and is therefore a major focus of research into malaria pathogenesis. Since the discovery of the var gene family that encodes PfEMP1 in the late 1990s, there have been multiple hypotheses for how the protein is trafficked to the infected cell surface, crossing multiple membranes along the way. One difficulty in studying this process is the large size of the var gene family and the propensity of the parasites to switch which var gene is expressed, thus preventing straightforward gene modification-based strategies for tagging the expressed PfEMP1. Here the authors solve this problem by forcing the expression of a targeted var gene by fusing the PfEMP1 coding region with a drug-selectable marker separated by a skip peptide. This enabled them to generate relatively homogenous populations of parasites all expressing tagged (or otherwise modified) forms of PfEMP1 suitable for study. They then applied this method to study various aspects of PfEMP1 trafficking.

Strengths:

The study is very thorough, and the data are well presented. The authors used SLI to target multiple var genes, thus demonstrating the robustness of their strategy. They then perform experiments to investigate possible trafficking through PTEX, they knock out proteins thought to be involved in PfEMP1 trafficking and observe defects in cytoadherence, and they perform proximity labeling to further identify proteins potentially involved in PfEMP1 export. These are independent and complimentary approaches that together tell a very compelling story.

Weaknesses:

(1) When the authors targeted IT4var19, they were successful in transcriptionally activating the gene, however, they did not initially obtain cytoadherent parasites. To observe binding to ICAM-1 and EPCR, they had to perform selection using panning. This is an interesting observation and potentially provides insights into PfEMP1 surface display, folding, etc. However, it also raises questions about other instances in which cytoadherence was not observed. Would panning of these other lines have been successfully selected for cytoadherent infected cells? Did the authors attempt panning of their 3D7 lines? Given that these parasites do export PfEMP1 to the infected cell surface (Figure 1D), it is possible that panning would similarly rescue binding. Likewise, the authors knocked out PTP1, TryThrA, and EMPIC3 and detected a loss of cytoadhesion, but they did not attempt panning to see if this could rescue binding. To ensure that the lack of cytoadhesion in these cases is not serendipitous (as it was when they activated IT4var19), they should demonstrate that panning cannot rescue binding.

(2) The authors perform a series of trafficking experiments to help discern whether PfEMP1 is trafficked through PTEX. While the results were not entirely definitive, they make a strong case for PTEX in PfEMP1 export. The authors then used BioID to obtain a proxiome for PfEMP1 and identified proteins they suggest are involved in PfEMP1 trafficking. However, it seemed that components of PTEX were missing from the list of interacting proteins. Is this surprising and does this observation shed any additional light on the possibility of PfEMP1 trafficking through PTEX? This warrants a comment or discussion.

Author response:

Public Reviews:

Reviewer #1 (Public review):

One of the roadblocks in PfEMP1 research has been the challenges in manipulating var genes to incorporate markers to allow the transport of this protein to be tracked and to investigate the interactions taking place within the infected erythrocyte. In addition, the ability of Plasmodium falciparum to switch to different PfEMP1 variants during in vitro culture has complicated studies due to parasite populations drifting from the original (manipulated) var gene expression. Cronshagen et al have provided a useful system with which they demonstrate the ability to integrate a selectable drug marker into several different var genes that allows the PfEMP1 variant expression to be 'fixed'. This on its own represents a useful addition to the molecular toolbox and the range of var genes that have been modified suggests that the system will have broad application. As well as incorporating a selectable marker, the authors have also used selective linked integration (SLI) to introduce markers to track the transport of PfEMP1, investigate the route of transport, and probe interactions with PfEMP1 proteins in the infected host cell.

What I particularly like about this paper is that the authors have not only put together what appears to be a largely robust system for further functional studies, but they have used it to produce a range of interesting findings including:

- Co-activation of rif and var genes when in a head-to-head orientation.

- The reduced control of expression of var genes in the 3D7-MEED parasite line.

- More support for the PTEX transport route for PfEMP1.

- Identification of new proteins involved in PfEMP1 interactions in the infected erythrocyte, including some required for cytoadherence.

In most cases the experimental evidence is straightforward, and the data support the conclusions strongly. The authors have been very careful in the depth of their investigation, and where unexpected results have been obtained, they have looked carefully at why these have occurred.

(1) In terms of incorporating a drug marker to drive mono-variant expression, the authors show that they can manipulate a range of var genes in two parasite lines (3D7 and IT4), producing around 90% expression of the targeted PfEMP1. Removal of drug selection produces the expected 'drift' in variant types being expressed. The exceptions to this are the 3D7-MEED line, which looks to be an interesting starting point to understand why this variant appears to have impaired mutually exclusive var gene expression and the EPCR-binding IT4var19 line. This latter finding was unexpected and the modified construct required several rounds of panning to produce parasites expressing the targeted PfEMP1 and bind to EPCR. The authors identified a PTP3 deficiency as the cause of the lack of PfEMP1 expression, which is an interesting finding in itself but potentially worrying for future studies. What was not clear was whether the selected IT4var19 line retained specific PfEMP1 expression once receptor panning was removed.

This is a very interesting point. We do not have systematic long-term data for the Var19 line but medium-term data. After panning the Var19 line, the binding assays were done within 3 months without additional panning. The first binding assay was 2 months after the panning and the last binding assays three weeks later. While there is inherent variation in these assays that precludes detection of smaller changes, the last assay showed the highest level of binding, giving no indication for rapid loss of the binding phenotype. Hence, we can say that the binding phenotype appears to be stable for many weeks without panning the cells again and there was no indication for a rapid loss of binding in these parasites.

Systematic long-term experiments to assess how long the Var19 parasites retain binding would be interesting, but given that the binding-phenotype appears to remain stable over many weeks, this would only make sense if done for a much longer time (6 months or more). Due to the time needed to carry out such an experiment this would not be practical to still include into the present study. But this might be advisable if the Var19 line is used in future experiments that go over extended periods of time. We intend to include a statement in the discussion of the revised manuscript to highlight that if long-term work with this line is planned, monitoring the binding phenotype and potentially re-panning might be advisable.

(2) The transport studies using the mDHFR constructs were quite complicated to understand but were explained very clearly in the text with good logical reasoning.

We are aware of this being a complex issue and are glad this was nevertheless understandable.

(3) By introducing a second SLI system, the authors have been able to alter other genes thought to be involved in PfEMP1 biology, particularly transport. An example of this is the inactivation of PTP1, which causes a loss of binding to CD36 and ICAM-1. It would have been helpful to have more insight into the interpretation of the IFAs as the anti-SBP1 staining in Figure 5D (PTP-TGD) looks similar to that shown in Figure 1C, which has PTP intact. The anti-EXP2 results are clearly different.

We realize the description of the PTP1-TGD IFA data and that of the other TGDs was rather cursory. We intend to amend this in the revision.

(4) It is good to see the validation of PfEMP1 expression includes binding to several relevant receptors. The data presented use CHO-GFP as a negative control, which is relevant, but it would have been good to also see the use of receptor mAbs to indicate specific adhesion patterns. The CHO system if fine for expression validation studies, but due to the high levels of receptor expression on these cells, moving to the use of microvascular endothelial cells would be advisable. This may explain the unexpected ICAM-1 binding seen with the panned IT4var19 line.

We agree with the reviewer that it is desirable to have better binding systems for studying individual binding interactions. As the main purpose of this paper was to introduce the system and show binding, we did not move to more complicated binding systems. However, we would like to point out that the CSA binding was done on receptor alone in addition to the CSA-expressing HBEC-5i cells and was competed successfully with soluble CSA. In addition, apart from the additional ICAM1-binding of the Var19 line, all binding phenotypes were conform with expectations. We therefore hope the tools used for binding studies are acceptable at this stage of introducing the system while future work interested in specific PfEMP1 receptor interactions are advised to use better systems, ideally including also endothelial organoid models, inhibitory antibodies and possibly domain competition. We intend to add a sentence to the discussion highlighting that future work using this system to study individual receptor-interactions could benefit from using optimized binding systems.

(5) The proxiome work is very interesting and has identified new leads for proteins interacting with PfEMP1, as well as suggesting that KAHRP is not one of these. The reduced expression seen with BirA* in position 3 is a little concerning but there appears to be sufficient expression to allow interactions to be identified with this construct. The quantitative impact of reduced expression for proxiome experiments will clearly require further work to define it.

This is a valid point. Clearly there seems to be some impact on binding when BirA* is placed in the extracellular domain (either through reduced presentation or direct reduction of binding efficiency of the modified PfEMP1). The exact impact on the proxiome is indeed difficult to assess. However, we hope that the general coverage of proteins proximal to PfEMP1 with the 3 PfEMP1-BirA* constructs will aid in the identification of proteins involved in PfEMP1 transport and surface display as illustrated with two of the hits targeted here.

(6) The reduced receptor binding results from the TryThrA and EMPIC3 knockouts were very interesting, particularly as both still display PfEMP1 on the surface of the infected erythrocyte. While care needs to be taken in cross-referencing adhesion work in P. berghei and whether the machinery truly is functionally orthologous, it is a fair point to make in the discussion. The suggestion that interacting proteins may influence the "correct presentation of PfEMP1" is intriguing and I look forward to further work on this.

We hope we future work will be able to shed light on this.

Overall, the authors have produced a useful and reasonably robust system to support functional studies on PfEMP1, which may provide a platform for future studies manipulating the domain content in the exon 1 portion of var genes. They have used this system to produce a range of interesting findings and to support its use by the research community.<br /> Finally, a small concern. Being able to select specific var gene switches using drug markers could provide some useful starting points to understand how switching happens in P. falciparum. However, our trypanosome colleagues might remind us that forcing switches may show us some mechanisms but perhaps not all.

Point noted! From non-systematic data with the Var01 line that has been cultured for extended periods of time (several years), it seems other non-targeted vars remain silent in our SLI “activation” lines but how much SLI-based var-expression “fixing” tampers with the integrity of natural switching mechanisms is indeed very difficult to gage at this stage. We intend to add a statement to the manuscript that even if mutually exclusive expression is maintained, it is not certain the mechanisms controlling var expression all remain intact.

Reviewer #2 (Public review):

Summary

Croshagen et al develop a range of tools based on selection-linked integration (SLI) to study PfEMP1 function in P. falciparum. PfEMP1 is encoded by a family of ~60 var genes subject to mutually exclusive expression. Switching expression between different family members can modify the binding properties of the infected erythrocyte while avoiding the adaptive immune response. Although critical to parasite survival and Malaria disease pathology, PfEMP1 proteins are difficult to study owing to their large size and variable expression between parasites within the same population. The SLI approach previously developed by this group for genetic modification of P. falciparum is employed here to selectively and stably activate the expression of target var genes at the population level. Using this strategy, the binding properties of specific PfEMP1 variants were measured for several distinct var genes with a novel semi-automated pipeline to increase throughput and reduce bias. Activation of similar var genes in both the common lab strain 3D7 and the cytoadhesion competent FCR3/IT4 strain revealed higher binding for several PfEMP1 IT4 variants with distinct receptors, indicating this strain provides a superior background for studying PfEMP1 binding. SLI also enables modifications to target var gene products to study PfEMP1 trafficking and identify interacting partners by proximity-labeling proteomics, revealing two novel exported proteins required for cytoadherence. Overall, the data demonstrate a range of SLI-based approaches for studying PfEMP1 that will be broadly useful for understanding the basis for cytoadhesion and parasite virulence.

Comments

(1) While the capability of SLI to actively select var gene expression was initially reported by Omelianczyk et al., the present study greatly expands the utility of this approach. Several distinct var genes are activated in two different P. falciparum strains and shown to modify the binding properties of infected RBCs to distinct endothelial receptors; development of SLI2 enables multiple SLI modifications in the same parasite line; SLI is used to modify target var genes to study PfEMP1 trafficking and determine PfEMP1 interactomes with BioID. Curiously, Omelianczyk et al activated a single var (Pf3D7_0421300) and observed elevated expression of an adjacent var arranged in a head-to-tail manner, possibly resulting from local chromatin modifications enabling expression of the neighboring gene. In contrast, the present study observed activation of neighboring genes with head-to-head but not head-to-tail arrangement, which may be the result of shared promoter regions. The reason for these differing results is unclear although it should be noted that the two studies examined different var loci.

The point that we are looking at different loci is very valid and we realize this is not mentioned in the discussion. In the revision we intend to add this as a possible reason for this discrepancy. As stated in the discussion, the head-to-head scenario was observed before in lines obtained with panning. However, given the rather few examples where this was analyzed, it is well possible that this varies with gene locus and we will make sure that the revised version of the manuscript will be careful to highlight that it is not clear how much this observation in our work can be generalized.

(2) The IT4var19 panned line that became binding-competent showed increased expression of both paralogs of ptp3 (as well as a phista and gbp), suggesting that overexpression of PTP3 may improve PfEMP1 display and binding. Interestingly, IT4 appears to be the only known P. falciparum strain (only available in PlasmoDB) that encodes more than one ptp3 gene (PfIT_140083100 and PfIT_140084700). PfIT_140084700 is almost identical to the 3D7 PTP3 (except for a ~120 residue insertion in 3D7 beginning at residue 400). In contrast, while the C-terminal region of PfIT_140083100 shows near-perfect conservation with 3D7 PTP3 beginning at residue 450, the N-terminal regions between the PEXEL and residue 450 are quite different. This may indicate the generally stronger receptor binding observed in IT4 relative to 3D7 results from increased PTP3 activity due to multiple isoforms or that specialized trafficking machinery exists for some PfEMP1 proteins.

We thank the reviewer for pointing this out, it is an interesting idea that the PTP3 duplication could be a reason for the superior binding of IT4. We intend to add this point to the discussion of the revision.

So far it seems the PTP3 issue occurred only with Var19. The thought of an extra layer of control, particularly for PfEMP1 variants that might be associated with virulence such as Var19, is very attractive. At present, the manuscript alludes to the possibility of an extra layer of control in the discussion. As var-type specificity and existence of such mechanisms in vivo are so far not known we decided not to speculate on this.

Reviewer #3 (Public review):

Summary:

The submission from Cronshagen and colleagues describes the application of a previously described method (selection linked integration) to the systematic study of PfEMP1 trafficking in the human malaria parasite Plasmodium falciparum. PfEMP1 is the primary virulence factor and surface antigen of infected red blood cells and is therefore a major focus of research into malaria pathogenesis. Since the discovery of the var gene family that encodes PfEMP1 in the late 1990s, there have been multiple hypotheses for how the protein is trafficked to the infected cell surface, crossing multiple membranes along the way. One difficulty in studying this process is the large size of the var gene family and the propensity of the parasites to switch which var gene is expressed, thus preventing straightforward gene modification-based strategies for tagging the expressed PfEMP1. Here the authors solve this problem by forcing the expression of a targeted var gene by fusing the PfEMP1 coding region with a drug-selectable marker separated by a skip peptide. This enabled them to generate relatively homogenous populations of parasites all expressing tagged (or otherwise modified) forms of PfEMP1 suitable for study. They then applied this method to study various aspects of PfEMP1 trafficking.

Strengths:

The study is very thorough, and the data are well presented. The authors used SLI to target multiple var genes, thus demonstrating the robustness of their strategy. They then perform experiments to investigate possible trafficking through PTEX, they knock out proteins thought to be involved in PfEMP1 trafficking and observe defects in cytoadherence, and they perform proximity labeling to further identify proteins potentially involved in PfEMP1 export. These are independent and complimentary approaches that together tell a very compelling story.

Weaknesses:

(1) When the authors targeted IT4var19, they were successful in transcriptionally activating the gene, however, they did not initially obtain cytoadherent parasites. To observe binding to ICAM-1 and EPCR, they had to perform selection using panning. This is an interesting observation and potentially provides insights into PfEMP1 surface display, folding, etc. However, it also raises questions about other instances in which cytoadherence was not observed. Would panning of these other lines have been successfully selected for cytoadherent infected cells? Did the authors attempt panning of their 3D7 lines? Given that these parasites do export PfEMP1 to the infected cell surface (Figure 1D), it is possible that panning would similarly rescue binding. Likewise, the authors knocked out PTP1, TryThrA, and EMPIC3 and detected a loss of cytoadhesion, but they did not attempt panning to see if this could rescue binding. To ensure that the lack of cytoadhesion in these cases is not serendipitous (as it was when they activated IT4var19), they should demonstrate that panning cannot rescue binding.

These are very important points. Indeed, we had repeatedly attempted to pan 3D7 when we failed to get the SLI-generated 3D7 PfEMP1 expressor lines to bind, but this had not been successful. After the move to IT4 which readily bound we made no further efforts to understand why 3D7 does not bind but the fact that PfEMP1 is on the surface indicates this is not a PTP3 issue. Also, as the parent 3D7 could not be panned, we assumed it is not easily fixed.

Panning the TGD lines: we see the reasoning for conducting panning experiments with the TGD lines, but on second thought we are unsure this should be attempted. The outcome might not be easily interpretable if panning leads to increased binding and considerable follow up analyses would be needed to define what has happened. The reason for this is that at least two forces will contribute to the selection in panning experiments with TGD lines that lost binding. Firstly, panning would work against the SLI of the TGD, resulting in a tug of war between the TGD-SLI and binding: a very low frequency of parasites can be expected to loop out the TGD plasmid and would normally be eliminated during standard culturing due to the SLI drug used for the TGD. These revertant cells would bind and the panning would enrich them (hence, panning and SLI are opposed in the case of a TGD abolishing binding). It is unclear how strong such an effect can be, but this might lead to mixed populations that complicate interpretations. The second selecting force are possible compensatory changes to restore binding. These can come in two flavors: reversal of potential independent changes that may have occurred in the TGD parasites and that are in reality causing the binding loss (the concern of the reviewer) or new changes to compensate the loss of the TGD target (in case the TGD is the cause of the binding loss). As both of the TGDs in the paper show some residual binding and have VAR01 on the surface to at least some extent, it is possible that new compensatory changes might indeed occur that indirectly increase binding again. In summary, even if more binding after panning of the lines occurs, it is not clear whether this is due to a compensatory change ameliorating the TGD or reversal of an unrelated change. The impact of repeated panning against SLI is also unknown. To determine the cause, the panned TGD lines would need to be subjected to a complex and time-consuming analysis (WGS, RNASeq, possibly Maurer’s clefts IFA phenotype) to find out whether they had an unrelated chance change that was reverted or a new compensatory change that helps binding.

The detection of VAR01 on the surface of these TGDs speaks against a PTP3 effect. While we can’t fully exclude other changes in the TGDs that might affect binding, we conducted WGS which did not show any obvious alterations that could be responsible. To fully exclude loss of ptp3 expression as the reason as seen with Var19 (something we would not have seen in the WGS if it is only due to a transcriptional change), we intend to carry out RNASeq with the two TGD lines. The third TGD mentioned by the reviewer (targeting ptp1) was a positive control of a known PfEMP1 trafficking protein, so we assume this does not need to be further validated.

(2) The authors perform a series of trafficking experiments to help discern whether PfEMP1 is trafficked through PTEX. While the results were not entirely definitive, they make a strong case for PTEX in PfEMP1 export. The authors then used BioID to obtain a proxiome for PfEMP1 and identified proteins they suggest are involved in PfEMP1 trafficking. However, it seemed that components of PTEX were missing from the list of interacting proteins. Is this surprising and does this observation shed any additional light on the possibility of PfEMP1 trafficking through PTEX? This warrants a comment or discussion.

This is an interesting comment and we agree we should have discussed this. A likely reason why PTEX components are not picked up as interactors is that BirA* is expected to become unfolded when it passes through the channel and in that state can’t biotinylate. Labelling likely would only be possible if PfEMP1 lingered at the PTEX translocation step before BirA* became unfolded to go through the channel which we would not expect under physiological conditions. We intend to add a sentence to the discussion why we think PTEX components would not be detected in our BioIDs even if PfEMP1 passes through it but that this might also be an argument against it passing through PTEX.

eLife Assessment

This is a study that makes the important finding that pleiotropy is positively associated with parallelism of evolutionary responses in gene expression, while theory would predict the opposite. The analysis uses a state-of-the-art experimental evolution approach to study the genetic basis of adaptation of Drosophila simulans to a hot environment. The experimental data is relevant and its analysis is robust, however, this paper appears to conflate gene expression variation and its underlying causative variation, in both its data interpretation and theoretical framework. This leads to incomplete conclusions on the causal link between pleiotropy and genetic variation and their role during adaptation.

Reviewer #1 (Public review):

The results of this manuscript look at the interplay between pleiotropy, standing genetic variation, and parallelism (i.e. predictability of evolution) in gene expression. Ultimately, their results suggest that (a) pleiotropic genes typically have a smaller range in variation/expression, and (b) adaptation to similar environments tends to favor changes in pleiotropic genes, which leads to parallelism in mechanisms (though not dramatically). However, it is still uncertain how much parallelism is directly due to pleiotropy, instead of a complex interplay between them and ancestral variation.

I have a few things that I was uncertain about. It may be these things are easily answered but require more discussion or clarity in the manuscript.

(1) The variation being talked about in this manuscript is expression levels, and not SNPs within coding regions (or elsewhere). The cause of any specific gene having a change in expression can obviously be varied - transcription factors, repressors, promoter region variation, etc. Is this taken into account within the "network connectivity" measurement? I understand the network connectivity is a proxy for pleiotropy - what I'm asking is, conceptually, what can be said about how/why those highly pleiotropic genes have a change (or not) in expression. This might be a question for another project/paper, but it feels like a next step worth mentioning somewhere.

(2) The authors do have a passing statement in line 361 about cis-regulatory regions. Is the assumption that genetic variation in promoter regions is the ultimate "mechanism" driving any change in expression? In the same vein, the authors bring up a potential confounding factor, though they dismiss it based on a specific citation (lines 476-481; citation 65). I'm of the mindset that in order to more confidently disregard this "issue" based on previous evidence, it requires more than one citation. Especially since the one citation is a plant. That specific point jumps out to me as needing a more careful rebuttal.

(3) I feel like there isn't enough exploration of tissue specificity versus network connectivity. Tissue specificity was best explained by a model in which pleiotropy had both direct and indirect effects on parallelism; while network connectivity was best explained (by a small margin) via the model which was mostly pleiotropy having a direct effect on ancestral variation, that then had a direct effect on parallelism. When the strengths of either direct/indirect effects were quantified, tissue specificity showed a stronger direct effect, while network connectivity had none (i.e. not significant). My confusion is with the last point - if network connectivity is explained by a direct effect in the best-supported model, how does this work, since the direct effect isn't significant? Perhaps I am misunderstanding something.

Also, network connectivity might favor the most pleiotropic genes being transcription factor hubs (or master regulators for various homeostasis pathways); while the tissue specificity metric perhaps is a kind of a space/time element. I get that a gene having expression across multiple tissues does fit the definition of pleiotropy in the broad sense, but I'm wondering if some important details are getting lost - I'm just thinking about the relative importance of what tissue specificity measurements say versus the network connectivity measurement.

Reviewer #2 (Public review):

Summary:

Lai and collaborators use a previously published RNAseq dataset derived from an experimental evolution set up to compare the pleiotropic properties of genes whose expression evolved in response to fluctuating temperature for over 100 generations. The authors correlate gene pleiotropy with the degree of parallelisms in the experimental evolution set up to ask: are genes that evolved in multiple replicates more or less pleiotropic?

They find that, maybe counter to expectation, highly pleiotropic genes show more replicated evolution. Such an effect seems to be driven by direct effects (which the authors can only speculate on) and indirect effects through low variance in pleiotropic genes (which the authors indirectly link to genetic variation underlying gene expression variance).

Weaknesses:

The results offer new insights into the evolution of gene expression and into the parameters that constrain such evolution, i.e., pleiotropy. Although the conclusions are supported by the data, I find the interpretation of the results a little bit complicated.

Major comment:

The major point I ask the authors to address is whether the connection between polygenic adaptation and parallelism can indeed be used to interpret gene expression parallelism. If the answer is not, please rephrase the introduction and discussion, if the answer is yes, please make it explicit in the text why it is so.

The authors' argument: parallelism in gene expression is the same as parallelism in SNP allele frequency (AFC) (see L389-383 here they don't mention that this explanation is derived from SNP parallelism and not trait parallelism, and see Figure 1 b). In previous publications, the authors have explained the low level of AFC parallelism using a polygenic argument. Polygenic traits can reach a new trait optimum via multiple SNPs and therefore although the trait is parallel across replicates, the SNPs are not necessarily so.

In the current paper, they seem to be exchanging SNP AFC by gene expression, and to me, those are two levels that cannot be interchanged. Gene expression is a trait, not an SNP, and therefore the fact that a gene expression doesn't replicate cannot be explained by a polygenic basis, because again the trait is gene expression itself. And, actually, the results of the simulations show that high polygenicity = less trait parallelism (Figure 4).

Now, if the authors focus on high parallel genes (present in e.g. 7 or more replicates) and they show that the eQTLs for those genes are many (highly polygenic) and the AFC of those eQTLs are not parallel, then I would agree with the interpretation. But, given that here they just assess gene expression and not eQTL AFC, I do not think they can use the 'highly polygenic = low parallelism' explanation.

The interpretation of the results to me, should be limited to: genes with low variance and high pleiotropy tend to be more parallel, and the explanation might be synergistic pleiotropy.

Reviewer #3 (Public review):

The authors aim to understand how gene pleiotropy affects parallel evolutionary changes among independent replicates of adaptation to a new hot environment of a set of experimental lines of Drosophila simulans using experimental evolution. The flies were RNAsequenced after more than 100 generations of lab adaptation and the changes in average gene expression were obtained relative to ancestral expression levels from reconstructed ancestral lines. Parallelism of gene expression change among lines is evaluated as variance in differential gene expression among lines relative to error variance. Similarly, the authors ask how the standing variation in gene expression estimated from a handful of flies from a reconstructed outbred line affects parallelism. The main findings are that parallelism in gene expression responses is positively associated with pleiotropy and negatively associated with expression variation. Those results are in contradiction with theoretical predictions and empirical findings. To explain those seemingly contradictory results the authors invoke the role of synergistic pleiotropy and correlated selection, although they do not attempt to measure either.

Strengths:

(1) The study uses highly replicated outbred laboratory lines of Drosophila simulans evolved in the lab under a constant hot regime for over 100 generations. This allows for robust comparisons of evolutionary responses among lines.

(2) The manuscript is well written and the hypotheses are clearly delineated at the onset.

(3) The authors have run a causal analysis to understand the causal dependencies between pleiotropy and expression variation on parallelism.

(4) The use of whole-body RNA extraction to study gene expression variation is well justified.

Weaknesses:

(1) It is unclear how well phenotypic variation in gene expression of the evolved lines has been estimated by the sample of 20 males from a reconstructed outbred line not directly linked to the evolved lines under study. I see this as a general weakness of the experimental design.

(2) There are no estimates of standing genetic variation of expression levels of the genes under study, only phenotypic variation. I wished the authors had been clear about that limitation and had discussed the consequences of the analysis. This also constitutes a weakness of the study.

(3) Moreover, since the phenotype studied is gene expression, its genetic basis extends beyond expressed sequences. The phenotypic variation of a gene's expression may thus likely misrepresent the genetic variation available for its evolution. The genetic variation of gene expression phenotypes could be estimated from a cross or pedigree information but since individuals were pool-sequenced (by batches of 50 males), this type of analysis is not possible in this study.

(4) The authors have not attempted to estimate synergistic pleiotropy among genes, nor how selection acts on gene expression modules. It makes any conclusion regarding the role of synergistic pleiotropy highly speculative.

I don't understand the reason why the analysis would be restricted to significantly differentially expressed genes only. It is then unclear whether pleiotropy, parallelism, and expression variation do play a role in adaptation because the two groups of adaptive and non-adaptive genes have not been compared. I recommend performing those comparisons to help us better understand how "adaptive" genes differentially contribute to adaptation relative to "non-adaptive" genes relative to their difference in population and genetic properties.

There is a lack of theoretical groundings on the role of so-called synergistic pleiotropy for parallel genetic evolution. The Discussion does not address this particular prediction. It could be removed from the Introduction.

Author response:

Reviewer #1 (Public review):

The results of this manuscript look at the interplay between pleiotropy, standing genetic variation, and parallelism (i.e. predictability of evolution) in gene expression. Ultimately, their results suggest that (a) pleiotropic genes typically have a smaller range in variation/expression, and (b) adaptation to similar environments tends to favor changes in pleiotropic genes, which leads to parallelism in mechanisms (though not dramatically). However, it is still uncertain how much parallelism is directly due to pleiotropy, instead of a complex interplay between them and ancestral variation.

I have a few things that I was uncertain about. It may be these things are easily answered but require more discussion or clarity in the manuscript.

(1) The variation being talked about in this manuscript is expression levels, and not SNPs within coding regions (or elsewhere). The cause of any specific gene having a change in expression can obviously be varied - transcription factors, repressors, promoter region variation, etc. Is this taken into account within the "network connectivity" measurement? I understand the network connectivity is a proxy for pleiotropy - what I'm asking is, conceptually, what can be said about how/why those highly pleiotropic genes have a change (or not) in expression. This might be a question for another project/paper, but it feels like a next step worth mentioning somewhere.

In current study, we are only able to detect significant and repeatable expression changes but unable to identify the underlying causal variants. An eQTL study in the founder population in combination with genomic resequencing for both evolved and ancestral populations would be required to address this question.

(2) The authors do have a passing statement in line 361 about cis-regulatory regions. Is the assumption that genetic variation in promoter regions is the ultimate "mechanism" driving any change in expression? In the same vein, the authors bring up a potential confounding factor, though they dismiss it based on a specific citation (lines 476-481; citation 65). I'm of the mindset that in order to more confidently disregard this "issue" based on previous evidence, it requires more than one citation. Especially since the one citation is a plant. That specific point jumps out to me as needing a more careful rebuttal.

It was not our intention to claim that the expression changes in our experiment are caused by cis-regulatory variation only. We believe that the observed expression variation has both cis- and trans-genetic components, where as some studies tend to estimate much higher cisvariation for gene expression in Drosophila populations (e.g. [1, 2]). We mentioned the positive correlation between cis-regulatory polymorphism and expression variation to (1) highlight the genetic control of gene expression and (2) make the connection between polygenic adaptation and gene expression evolutionary parallelism.

(3) I feel like there isn't enough exploration of tissue specificity versus network connectivity. Tissue specificity was best explained by a model in which pleiotropy had both direct and indirect effects on parallelism; while network connectivity was best explained (by a small margin) via the model which was mostly pleiotropy having a direct effect on ancestral variation, that then had a direct effect on parallelism. When the strengths of either direct/indirect effects were quantified, tissue specificity showed a stronger direct effect, while network connectivity had none (i.e. not significant). My confusion is with the last point - if network connectivity is explained by a direct effect in the best-supported model, how does this work, since the direct effect isn't significant? Perhaps I am misunderstanding something.

To clarify, for network connectivity, there’s a significant “indirect” effect on parallelism (i.e. network connectivity affect ancestral gene expression and ancestral gene expression affect parallelism). Hence, in table 2, the direct effect of network connectivity on parallelism is weak and not significant while the indirect effect via ancestral variation is significant.

Also, network connectivity might favor the most pleiotropic genes being transcription factor hubs (or master regulators for various homeostasis pathways); while the tissue specificity metric perhaps is a kind of a space/time element. I get that a gene having expression across multiple tissues does fit the definition of pleiotropy in the broad sense, but I'm wondering if some important details are getting lost - I'm just thinking about the relative importance of what tissue specificity measurements say versus the network connectivity measurement.

We examined the statistical relationship between the two measures and found a moderate positive correlation on the basis of which we argued that the two measures may capture different aspects of pleiotropy. We appreciate the reviewer’s suggestions about the biological basis of the two estimates of pleiotropy, but we think that without further experimental insights, an extended discussion of this topic is too premature to provide meaningful insights to the readership.

Reviewer #2 (Public review):

Summary:

Lai and collaborators use a previously published RNAseq dataset derived from an experimental evolution set up to compare the pleiotropic properties of genes whose expression evolved in response to fluctuating temperature for over 100 generations. The authors correlate gene pleiotropy with the degree of parallelisms in the experimental evolution set up to ask: are genes that evolved in multiple replicates more or less pleiotropic?

They find that, maybe counter to expectation, highly pleiotropic genes show more replicated evolution. Such an effect seems to be driven by direct effects (which the authors can only speculate on) and indirect effects through low variance in pleiotropic genes (which the authors indirectly link to genetic variation underlying gene expression variance).

Weaknesses:

The results offer new insights into the evolution of gene expression and into the parameters that constrain such evolution, i.e., pleiotropy. Although the conclusions are supported by the data, I find the interpretation of the results a little bit complicated.

Major comment:

The major point I ask the authors to address is whether the connection between polygenic adaptation and parallelism can indeed be used to interpret gene expression parallelism. If the answer is not, please rephrase the introduction and discussion, if the answer is yes, please make it explicit in the text why it is so.

Our answer is yes, we interpreted gene expression parallelism (high ancestral variance -> less parallelism) using the same framework that links polygenic adaptation and parallelism (high polygenicity = less trait parallelism). We believe that our response covers several of the reviewer’s concerns.

The authors' argument: parallelism in gene expression is the same as parallelism in SNP allele frequency (AFC) (see L389-383 here they don't mention that this explanation is derived from SNP parallelism and not trait parallelism, and see Figure 1 b). In previous publications, the authors have explained the low level of AFC parallelism using a polygenic argument. Polygenic traits can reach a new trait optimum via multiple SNPs and therefore although the trait is parallel across replicates, the SNPs are not necessarily so.

Importantly, our rationale is based on the idea that gene expression is rarely the direct target of selection, but rather an intermediate trait [3]. Recently, we have specifically tested this assumption for gene expression and metabolite concentrations and our analysis showed that both traits were are redundant [4], as previously shown for DNA sequences [5]. The important implication for this manuscript is that gene expression is also redundant, so that adaptation can be achieved by distinct changes in gene expression in replicate populations adapting to the same selection pressure. This implies that we can use the same simulation framework for gene expression as for sequencing data. In our case different SNP frequencies correspond to different expression levels (averaged across individuals from a population), which in turn increases fitness by modifying the selected trait. Importantly, the selected trait in our simulations is not gene expression, but a not defined high level phenotype. A key insight from our simulations is that with increasing polygenicity the expression of a gene is more variable in the ancestral population.

In the current paper, they seem to be exchanging SNP AFC by gene expression, and to me, those are two levels that cannot be interchanged. Gene expression is a trait, not an SNP, and therefore the fact that a gene expression doesn't replicate cannot be explained by a polygenic basis, because again the trait is gene expression itself. And, actually, the results of the simulations show that high polygenicity = less trait parallelism (Figure 4).

As detailed above, because adaptation can be reached by changes in gene expression at different sets of genes, redundancy is also operating on the expression level not just on the level of SNPs. To clarify, the x-axis of Fig. 4 is the expression variation in the ancestral population.

Now, if the authors focus on high parallel genes (present in e.g. 7 or more replicates) and they show that the eQTLs for those genes are many (highly polygenic) and the AFC of those eQTLs are not parallel, then I would agree with the interpretation. But, given that here they just assess gene expression and not eQTL AFC, I do not think they can use the 'highly polygenic = low parallelism' explanation.

The interpretation of the results to me, should be limited to: genes with low variance and high pleiotropy tend to be more parallel, and the explanation might be synergistic pleiotropy.

While we understand the desire to model the full hierarchy from eQTLs to gene expression and adaptive traits, we raise caution that this would be a very challenging task. eQTLs very often underestimate the contribution of trans-acting factors, hence the understanding of gene expression evolution based on eQTLs is very likely incomplete and cannot explain the redundancy of gene expression during adaptation. Hence, we think that the focus on redundant gene expression is conceptually simpler and thus allows us to address the question of pleiotropy without the incorporation of allele frequency changes.  

Reviewer #3 (Public review):

The authors aim to understand how gene pleiotropy affects parallel evolutionary changes among independent replicates of adaptation to a new hot environment of a set of experimental lines of Drosophila simulans using experimental evolution. The flies were RNAsequenced after more than 100 generations of lab adaptation and the changes in average gene expression were obtained relative to ancestral expression levels from reconstructed ancestral lines. Parallelism of gene expression change among lines is evaluated as variance in differential gene expression among lines relative to error variance. Similarly, the authors ask how the standing variation in gene expression estimated from a handful of flies from a reconstructed outbred line affects parallelism. The main findings are that parallelism in gene expression responses is positively associated with pleiotropy and negatively associated with expression variation. Those results are in contradiction with theoretical predictions and empirical findings. To explain those seemingly contradictory results the authors invoke the role of synergistic pleiotropy and correlated selection, although they do not attempt to measure either.

Strengths:

(1) The study uses highly replicated outbred laboratory lines of Drosophila simulans evolved in the lab under a constant hot regime for over 100 generations. This allows for robust comparisons of evolutionary responses among lines.

(2) The manuscript is well written and the hypotheses are clearly delineated at the onset.

(3) The authors have run a causal analysis to understand the causal dependencies between pleiotropy and expression variation on parallelism.

(4) The use of whole-body RNA extraction to study gene expression variation is well justified.

Weaknesses:

(1) It is unclear how well phenotypic variation in gene expression of the evolved lines has been estimated by the sample of 20 males from a reconstructed outbred line not directly linked to the evolved lines under study. I see this as a general weakness of the experimental design.

Our intention was not to measure the phenotypic variance of the evolved lines, but rather to estimate the phenotypic variance at the beginning of the experiment. Hence, we measured and investigated the variation of gene expression in the ancestral population since this was the beginning of the replicated experimental evolution. Furthermore, since the ancestral population represents the natural population in Florida, the gene expression variation reflects the history of selection history acting on it.

(2) There are no estimates of standing genetic variation of expression levels of the genes under study, only phenotypic variation. I wished the authors had been clear about that limitation and had discussed the consequences of the analysis. This also constitutes a weakness of the study.

The reviewer is correct that we do not aim to estimate the standing genetic variation, which is responsible for differences in gene expression. While we agree that it could be an interesting research question to use eQTL mapping to identify the genetic basis of gene expression, we caution that trans-effects are difficult to estimate and therefore an important component of gene expression evolution will be difficult to estimate. Hence, we consider that our focus on variation in gene expression without explicit information about the genetic basis is simpler and sufficient to address the question about the role of pleiotropy.

(3) Moreover, since the phenotype studied is gene expression, its genetic basis extends beyond expressed sequences. The phenotypic variation of a gene's expression may thus likely misrepresent the genetic variation available for its evolution. The genetic variation of gene expression phenotypes could be estimated from a cross or pedigree information but since individuals were pool-sequenced (by batches of 50 males), this type of analysis is not possible in this study.

We agree with the reviewer that gene expression variation may also have a non-genetic basis, we discuss this in depth in the discussion of the manuscript.  

(4) The authors have not attempted to estimate synergistic pleiotropy among genes, nor how selection acts on gene expression modules. It makes any conclusion regarding the role of synergistic pleiotropy highly speculative.

We mentioned synergistic pleiotropy as a possible explanation for our results. A positive correlation between the fitness effect of gene expression variation would predict more replicable evolutionary changes. A similar argument has been made by [6]. 

I don't understand the reason why the analysis would be restricted to significantly differentially expressed genes only. It is then unclear whether pleiotropy, parallelism, and expression variation do play a role in adaptation because the two groups of adaptive and non-adaptive genes have not been compared. I recommend performing those comparisons to help us better understand how "adaptive" genes differentially contribute to adaptation relative to "nonadaptive" genes relative to their difference in population and genetic properties.

We agree with the reviewer that the comparison between the pleiotropy of adaptive and nonadaptive genes is interesting. We performed the analysis but omitted from the current manuscript for simplicity. Similar to the results in [6], non-adaptive genes are more pleiotropic than the adaptive genes. For adaptive genes we find a positive correlation between the level of pleiotropy and evolutionary parallelism. Thus, high pleiotropy limits the evolvability of a gene, but moderate and potentially synergistic pleiotropy increases the repeatability of adaptive evolution. We included this result in the revised manuscript and discuss it.

There is a lack of theoretical groundings on the role of so-called synergistic pleiotropy for parallel genetic evolution. The Discussion does not address this particular prediction. It could be removed from the Introduction.

We modestly disagree with the reviewer, synergistic pleiotropy is covered by theory and empirical results also support the importance of synergistic pleiotropy. 

References

(1) Genissel A, McIntyre LM, Wayne ML, Nuzhdin SV. Cis and trans regulatory effects contribute to natural variation in transcriptome of Drosophila melanogaster. Molecular biology and evolution. 2008;25(1):101-10. Epub 20071112. doi: 10.1093/molbev/msm247. PubMed PMID: 17998255.

(2) Osada N, Miyagi R, Takahashi A. Cis- and Trans-regulatory Effects on Gene Expression in a Natural Population of Drosophila melanogaster. Genetics. 2017;206(4):2139-48. Epub 20170614. doi: 10.1534/genetics.117.201459. PubMed PMID: 28615283; PubMed Central PMCID: PMCPMC5560811.

(3) Barghi N, Hermisson J, Schlötterer C. Polygenic adaptation: a unifying framework to understand positive selection. Nature reviews Genetics. 2020;21(12):769-81. Epub 2020/07/01. doi: 10.1038/s41576-020-0250-z. PubMed PMID: 32601318.

(4) Lai WY, Otte KA, Schlötterer C. Evolution of Metabolome and Transcriptome Supports a Hierarchical Organization of Adaptive Traits. Genome biology and evolution. 2023;15(6). Epub 2023/05/26. doi: 10.1093/gbe/evad098. PubMed PMID: 37232360; PubMed Central PMCID: PMCPMC10246829.

(5) Barghi N, Tobler R, Nolte V, Jaksic AM, Mallard F, Otte KA, et al. Genetic redundancy fuels polygenic adaptation in Drosophila. PLoS biology. 2019;17(2):e3000128. Epub 2019/02/05. doi: 10.1371/journal.pbio.3000128. PubMed PMID: 30716062.

(6) Rennison DJ, Peichel CL. Pleiotropy facilitates parallel adaptation in sticklebacks. Molecular ecology. 2022;31(5):1476-86. Epub 2022/01/09. doi: 10.1111/mec.16335. PubMed PMID: 34997980; PubMed Central PMCID: PMCPMC9306781.

eLife Assessment

Liang et al. have conducted a small pilot study investigating the feasibility and tolerability of a regimen of neoadjuvant chemo-immunotherapy for non-small cell lung cancer; with lower cumulative dose of chemotherapy and with the immunotherapy delivered on D8 of each cycle. The clinical data are interesting and novel, and overall the findings of the study are valuable. However, the translational data and analyses are incomplete and do not support key claims in the title.

Reviewer #1 (Public review):

Liang et al. have conducted a small-scale pilot study focusing on the feasibility and tolerability of Low-dose chemotherapy combined with delayed immunotherapy in the neoadjuvant treatment of non-small cell lung cancer. The design of delayed immunotherapy after chemotherapy is relatively novel, while the reduced chemotherapy, although somewhat lacking in innovation, still serves as an early clue for exploring future feasible strategies. Also, the dynamic ctDNA and TCR profiles could give some important hints of intrinsic tumor reaction.

However, as the author mentioned in the limitation part, due to the small sample size and lack of a control group, we cannot fully understand the advantages and disadvantages of this approach compared to standard treatment. Compared to standard immunotherapy, the treatment group in this study has three differences: (1) reduced chemotherapy, (2) the use of cisplatin instead of the commonly used carboplatin in neoadjuvant therapy trials, and (3) delayed immunotherapy. Generally, in the exploration of updated treatment strategies, the design should follow the principle of "controlling variables." If there are too many differences at once, it becomes difficult to determine which variable is responsible for the effects, leading to confusion in the interpretation of the results. Moreover, the therapeutic strategy may lack practical clinical operability due to the long treatment duration.

Furthermore, in the exploration of biomarkers, the authors emphasized the procedure of whole RNA sequencing in tumor tissues in the method section, and this was also noted in the flowchart in Figure 1. However, I didn't find any mention of RNA-related analyses in the Results section, which raises some concerns about the quality of this paper for me. If the authors have inadvertently omitted some results, they should supplement the RNA-related analyses so that I can re-evaluate the paper.

To sum up, this article exhibited a certain degree of innovation to some extent, However, due to its intrinsic design defects and data omissions, the quality of the research warranted further improvement.

Reviewer #2 (Public review):

Summary:

In this single center, single arm, open label non-randomised study the authors tested the use of paclitaxel at 180-220 mg/m2 and cisplatin at 60mg/m2 in patients with squamous NSCLC and pemetrexed at 500mg/m2 and cisplatin at 60mg/m2 in adenocarcinoma of lung origin in the neoadjuvant setting. The chemotherapy appears to have been given at a relatively standard dose; though the platin dose at 60mg/m2 is somewhat lower than has been used in the checkmate 816 trial (75mg/m2/dose), this is a well-established dose for NSCLC.

Key differences to currently approved neoadjuvant chemo-ICI treatment is that anti-PD1 antibody sintilimab (at 200mg/dose) was given on day 5 and that only 2 cycles of chemotherapy were given pre surgery, but then repeated on two occasions post surgery. Between May/2020 and Nov/2023 50 patients were screened, 38 went on to have this schedule of tx, 31 (~82%) went on to have surgery and 27 had the adjuvant treatment. The rate of surgery is entirely consistent with the checkmate 816 data.

Question to the authors:

It would be very helpful to understand why 7 (~18% of the population) patients did not make it to surgery and whether this is related to disease progression, toxicity or other reasons for withdrawal.

The key clinical endpoints were pCR and mPR rates. 2/38 patients are reported to have achieved a radiological pCR but only 31 patients underwent surgery with histological verification. Supp table2 suggests that 10/31 patients achieved a pCR, 6/31 additional patients achieved a major pathological response and that 13/31 did not achieve a major pathological response

It would be really helpful for understanding the clinical outcome to present the histopathological findings in the text in a bit more detail and to refer the outcome to the radiological findings. I note that the reference for pathological responses incorrectly is 38 patients as only 31 patients underwent surgery and were evaluated histologically.

The treatment was very well tolerated with only 1 grade 3 AE reported. The longer term outcome will need to be assessed over time as the cohort is very 'young'. It is not clear what the adjuvant chemo-ICI treatment would add and how this extra treatment would be evaluated for benefit - if all the benefit is in the neoadjuvant treatment then the extra post-operative tx would only add toxicity

Please consider what the two post-operative chemo-ICI cycles might add to the outcome and how the value of these cycles would be assessed. Would there be a case for a randomised assessment in the patients who have NOT achieved a mPR histologically?

While the clinical dataset identifies that the proposed reduced chemo-ICI therapy has clinical merit and should be assessed in a randomized study, the translational work is less informative.

The authors suggest that the treatment has a positive impact on T lymphocytes. Blood sampling was done at day 0 and day 5 of each of the four cycle of chemotherapy with an additional sample post cycle 4. The authors state that data were analysed at each stage.

The data in Figure 3B are reported for three sets of pairs: baseline to pre day 5 in cycle 1, day 5 to day 21 in cycle 1, baseline of cycle to to day 5. It remains unclear whether the datasets contain the same top 20 clones and it would be very helpful to show kinetic change for the individual 'top 20 clones' throughout the events in individual patients; as it stands the 'top20 clones' may vary widely from timepoint to timepoint. Of note, the figures do not demonstrate that the top 20 TCR clones were 'continuously increased'.

Instead, the data suggest that there are fluctuations in the relative distributions over time but that may simply be a reflection of shifts in T cell populations following chemotherapy rather than of immunological effects in the cancer tissue.<br /> Consistent with this the authors conclude (line 304/5): "No significant difference was observed in the diversity, evenness, and clonality of TCR clones across the whole treatment procedure" and this seems to be a more persuasive conclusion than the statement 'that a positive effect on T lymphocytes was observed' - where it is also not clear what 'positive' means.

The text needs a more balanced representation of the data: only a small subset of four patients appear to have been evaluated to generate the data for figure 3B and only three patients (P5, P6, P7) can have contributed to figure 3C if the sample collection is represented accurately in Figure 3A.

The text refers to flow cytometric results in SF3. However, no information is given on the flow cytometry in M&M, markers or gating strategy.

Please consider changing the terminology of the 'phases' into something that is easier to understand. One option would be to use a reference to a more standard unit (cycle 1-4 of chemotherapy and then d0/d5/d21).

Please make it explicit in the text that molecular analyses were undertaken for some patients only, and how many patients contribute to the data in figures 3B-F. Figure 3A suggests paired mRNA data were obtained in 2 patients (P2 and P5) but I cannot find the results on these analyses; four individual blood samples to assess TCR changes int PH1/PH2/PH3and PH4 were only available in four patients (P4,P5,P7,P9). Only three patients seem to have the right samples collected to allow the analysis for 'C3' in figure 3C.

Please display for each of the 'top 20 clones' at any one timepoint how these clones evolve throughout the study; I expect that a clone that is 'top 20' at a given timepoint may not be among the 'top twenty' at all timepoints.

Please also assess if the expanded clonotypes are present (and expanded) in the cancer tissue at resection, to link the effect in blood to the tumour. Given that tissue was collected for 31 patients, mRNA sequencing to generate TCR data should be possible to add to the blood analyses in the 12 patients in Figure 3A. Without this data no clear link can be made to events in the cancer.

Please provide in M&M the missing information on the flow cytometry methodology (instrument, antibody clones, gating strategy) and what markers were used to define T cell subsets (naïve, memory, central memory, effector memory).

The authors also describe that ctDNA reduces after chemo-ICI treatment. This is well documented in their data but ultimately irrelevant: if the cancer volume is reduced to the degree of a radiological or pathological response /complete response then the quantity of circulating DNA from the cancer cells must reduce. More interesting would be the question whether early changes predict clinical outcome and whether recurrent ct DNA elevations herald recurrence.

Please probe whether the molecular data identify good radiological or pathological outcomes before cycle 2 is started and whether the ctDNA levels identify patients who will have a poor response and/or who relapse early.

eLife Assessment

This valuable study is a detailed investigation of how chromatin structure influences replication origin function in yeast ribosomal DNA, with a focus on the role of the histone deacetylase Sir2 and the chromatin remodeler Fun30. The paper shows that Sir2 does not affect origin licensing but rather affects local transcription and nucleosome positioning which correlates with increased origin firing. Overall, the evidence is convincing and the model is plausible.

Reviewer #1 (Public review):

This paper presents a mechanistic study of rDNA origin regulation in yeast by SIR2. Each of the ~180 tandemly repeated rDNA gene copies contains a potential replication origin. Early-efficient initiation of these origins is suppressed by Sir2, reducing competition with origins distributed throughout the genome for rate-limiting initiation factors. Previous studies by these authors showed that SIR2 deletion advances replication timing of rDNA origins by a complex mechanism of transcriptional de-repression of a local PolII promoter causing licensed origin proteins (MCMcomplexes) to re-localize (slide along the DNA) to a different (and altered) chromatin environment. In this study, they identify a chromatin remodeler, FUN30, that suppresses the sir2∆ effect, and remarkably, results in a contraction of the rDNA to about one-quarter it's normal length/number of repeats, implicating replication defects of the rDNA. Through examination of replication timing, MCM occupancy and nucleosome occupancy on the chromatin in sir2, fun30, and double mutants, they propose a model where nucleosome position relative to the licensed origin (MCM complexes) intrinsically determines origin timing/efficiency. While their interpretations of the data are largely reasonable and can be interpreted to support their model, a key weakness is the connection between Mcm ChEC signal disappearance and origin firing. While the cyclical chromatin association-dissociation of MCM proteins with potential origin sequences may be generally interpreted as licensing followed by firing, dissociation may also result from passive replication and as shown here, displacement by transcription and/or chromatin remodeling. Moreover, linking its disappearance from chromatin in the ChEC method with such precise resolution needs to be validated against an independent method to determine the initiation site(s). Differences in rDNA copy number and relative transcription levels also are not directly accounted for, obscuring a clearer interpretation of the results. Nevertheless, this paper makes a valuable advance with the finding of Fun30 involvement, which substantially reduces rDNA repeat number in sir2∆ background. The model they develop is compelling and I am inclined to agree, but I think the evidence on this specific point is purely correlative and a better method is needed to address the initiation site question. The authors deserve credit for their efforts to elucidate our obscure understanding of the intricacies of chromatin regulation.

Overall, the paper is improved by providing additional data and improved analysis. The paper nicely characterizes the effect of Fun30. The model is reasonable but remains lacking in precise details of mechanism.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors follow up on their previous work showing that in the absence of the Sir2 deacetylase the MCM replicative helicase at the rDNA spacer region is repositioned to a region of low nucleosome occupancy. Here they show that the repositioned displaced MCMs have increased firing propensity relative to non-displaced MCMs. In addition, they show that activation of the repositioned MCMs and low nucleosome occupancy in the adjacent region depend on the chromatin remodeling activity of Fun30.

Strengths:

The paper provides new information on the role of a conserved chromatin remodeling protein in regulation of origin firing and in addition provides evidence that not all loaded MCMs fire and that origin firing is regulated at a step downstream of MCM loading.

Comments on revisions:

The authors have addressed my concerns with the addition of new experiments and analysis.

Reviewer #3 (Public review):

Summary:

Heterochromatin is characterized by low transcription activity and late replication timing, both dependent on the NAD-dependent protein deacetylase Sir2, the founding member of the sirtuins. This manuscript addresses the mechanism by which Sir2 delays replication timing at the rDNA in budding yeast. Previous work from the same laboratory (Foss et al. PLoS Genetics 15, e1008138) showed that Sir2 represses transcription-dependent displacement of the Mcm helicase in the rDNA. In this manuscript, the authors show convincingly that the repositioned Mcms fire earlier and that this early firing partly depends on the ATPase activity of the nucleosome remodeler Fun30. Using read-depth analysis of sorted G1/S cells, fun30 was the only chromatin remodeler mutant that somewhat delayed replication timing in sir2 mutants, while nhp10, chd1, isw1, htl1, swr1, isw2, and irc5 had no effect. The conclusion was corroborated with orthogonal assays including two-dimensional gel electrophoresis and analysis of EdU incorporation at early origins. Using an insightful analysis with an Mcm-MNase fusion (Mcm-ChEC), the authors show that the repositioned Mcms in sir2 mutants fire earlier than the Mcm at the normal position in wild type. This early firing at the repositioned Mcms is partially suppressed by Fun30. In addition, the authors show Fun30 affects nucleosome occupancy at the sites of the repositioned Mcm, providing a plausible mechanism for the effect of Fun30 on Mcm firing at that position. However, the results from the MNAse-seq and ChEC-seq assays are not fully congruent for the fun30 single mutant. Overall, the results support the conclusions providing a much better mechanistic understanding how Sir2 affects replication timing at rDNA,

Strengths:

(1) The data clearly show that the repositioned Mcm helicase fires earlier than the Mcm in the wild type position.

(2) The study identifies a specific role for Fun30 in replication timing and an effect on nucleosome occupancy around the newly positioned Mcm helicase in sir2 cells.

Comments on revisions:

In the previous revision the authors addressed my concerns and improved the manuscript and the presentation of the data. All my recommendations were implemented.

Author response:

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

eLife Assessment 

This valuable study is a detailed investigation of how chromatin structure influences replication origin function in yeast ribosomal DNA, with a focus on the role of the histone deacetylase Sir2 and the chromatin remodeler Fun30. Convincing evidence shows that Sir2 does not affect origin licensing but rather affects local transcription and nucleosome positioning which correlates with increased origin firing. Overall, the evidence is solid and the model plausible. However, the methods employed do not rigorously establish a key aspect of the mechanism where initiation precisely occurs or rigorously exclude alternative models and the effect of Sir2 on transcription is not re-examined in the fun30 context. 

Clarification on Sir2 Effect on Transcription in the fun30 Context

We appreciate the reviewers’ thorough assessment but would like to clarify that the effect of Sir2 on transcription in the fun30 context was addressed in both the original and revised manuscripts. However, we recognize that the presentation of the qPCR results may have been unclear, as we initially plotted absolute transcript levels without normalizing for rDNA array size differences among the genotypes. We have now corrected this.

After normalizing for copy number variations, the qPCR data show that the sir2 fun30 double mutant results in a ~40-fold increase in C-pro transcription relative to WT, compared to a 4-fold and 19-fold increase in fun30 and sir2 single mutants, respectively (Figure 5, figure supplement 6). These results have been discussed in the manuscript result section, where we note that "C-pro RNA levels were approximately twice as high in sir2 fun30 compared to sir2 cells when adjusted for rDNA size differences." This observation is critical for addressing both alternative models of MCM disappearance and for pinpointing transcription initiation sites, as detailed in the following sections.

Public Reviews: 

Reviewer #1 (Public review): 

Summary: 

This paper presents a mechanistic study of rDNA origin regulation in yeast by SIR2. Each of the ~180 tandemly repeated rDNA gene copies contains a potential replication origin. Earlyefficient initiation of these origins is suppressed by Sir2, reducing competition with origins distributed throughout the genome for rate-limiting initiation factors. Previous studies by these authors showed that SIR2 deletion advances replication timing of rDNA origins by a complex mechanism of transcriptional de-repression of a local PolII promoter causing licensed origin proteins (MCMcomplexes) to re-localize (slide along the DNA) to a different (and altered) chromatin environment. In this study, they identify a chromatin remodeler, FUN30, that suppresses the sir2∆ effect, and remarkably, results in a contraction of the rDNA to about onequarter it's normal length/number of repeats, implicating replication defects of the rDNA. Through examination of replication timing, MCM occupancy and nucleosome occupancy on the chromatin in sir2, fun30, and double mutants, they propose a model where nucleosome position relative to the licensed origin (MCM complexes) intrinsically determines origin timing/efficiency. While their interpretations of the data are largely reasonable and can be interpreted to support their model, a key weakness is the connection between Mcm ChEC signal disappearance and origin firing. While the cyclical chromatin association-dissociation of MCM proteins with potential origin sequences may be generally interpreted as licensing followed by firing, dissociation may also result from passive replication and as shown here, displacement by transcription and/or chromatin remodeling. Moreover, linking its disappearance from chromatin in the ChEC method with such precise resolution needs to be validated against an independent method to determine the initiation site(s). Differences in rDNA copy number and relative transcription levels also are not directly accounted for, obscuring a clearer interpretation of the results. Nevertheless, this paper makes a valuable advance with the finding of Fun30 involvement, which substantially reduces rDNA repeat number in sir2∆ background. The model they develop is compelling and I am inclined to agree, but I think the evidence on this specific point is purely correlative and a better method is needed to address the initiation site question. The authors deserve credit for their efforts to elucidate our obscure understanding of the intricacies of chromatin regulation. At a minimum, I suggest their conclusions on these points of concern should be softened and caveats discussed. Statistical analysis is lacking for some claims. 

Strengths are the identification of FUN30 as suppressor, examination of specific mutants of FUN30 to distinguish likely functional involvement. Use of multiple methods to analyze replication and protein occupancies on chromatin. Development of a coherent model. 

Weaknesses are failure to address copy number as a variable; insufficient validation of ChEC method relationship to exact initiation locus; lack of statistical analysis in some cases. 

Review of revised version and response letter: 

In the response, the authors make some improvements by better quantifying 2D gels, adding some missing statistical analyses, analyzing the effect of fun30 on rDNA replication in strains with reduced rDNA copy number, and using ChIP-seq of MCMs to support the ChEC-seq data. However, these additions do not address the main issue that is at the heart of their model: where initiation precisely occurs and whether the location is altered in the mutant(s). Thus, mechanistic insight is limited.

We discuss the issue regarding the initiation site below.

Under the section "Addressing Alternative Explanations", the authors claim that processes like transcription and passive replication cannot affect the displaced complex specifically. Why? They are not on same DNA (as mentioned in the Fig 1 legend). 

Premature origin activation, not transcription, drives the disappearance of repositioned MCM complexes in sir2 mutants in HU.

Indeed, the reviewer is correct in suggesting that C-pro transcription confined to rDNA units with repositioned MCM complexes could selectively displace those complexes, potentially explaining the selective disappearance of displaced MCMs in sir2 cells. However, our analysis of C-pro transcription and MCM occupancy in G1 versus HU across the genotypes allows us to rule out this possibility.

We show that the fraction of repositioned MCMs in G1 cells is proportional to the level of C-pro transcription (WT < fun30 << sir2 < sir2 fun30), consistent with the involvement of transcription in the repositioning process during MCM loading in G1. Accordingly, with approximately twice the transcription in sir2 fun30 compared to sir2, we observe more repositioned MCMs in sir2 fun30 cells than in sir2 cells in G1 (Fig 5C).

However, if the disappearance of repositioned MCMs in HU were solely due to C-pro transcription rather than origin activation, we would expect the repositioned MCMs to disappear more quickly in sir2 fun30 cells. Contrary to this expectation, our data show that repositioned MCM complexes are more stable in sir2 fun30 mutants compared to sir2 mutants, indicating that transcription is not the primary factor in the disappearance of displaced MCM complexes in HU; rather, rDNA origin activation appears to be the key factor.

Replication initiation site in sir2. Using multiple independent approaches, including 2D gels, ChIP-seq, and EdU incorporation, we have demonstrated that rDNA origins fire prematurely in sir2 mutants, a conclusion that the reviewer does not contest. Once an origin fires, the MCM signal disappears from the site of its initial deposition, as expected, and this is confirmed in our MCM ChIP and HU ChEC data, both at rDNA origins and across the genome.

Given that the majority of MCM complexes in sir2 mutants are repositioned, it is expected that these repositioned complexes disappear following premature origin activation. With less than half of the licensed origins (or <30% of total rDNA copies) retaining MCM at non-repositioned sites in sir2 mutants, if only these non-repositioned complexes were firing, and the repositioned MCM complexes were disappearing via mechanisms other than replication initiation (e.g., transcription), rDNA replication in sir2 mutants would be severely compromised rather than accelerated. Given this, and the strong experimental evidence that repositioned MCM complexes fire prematurely, continued focus on alternative explanations for MCM complex disappearance seems unwarranted.

We present this analysis in the results section as follows:

“Finally, although deletion of FUN30 could suppress replication initiation at the rDNA either by inhibiting the firing of the active, repositioned MCM complex or by preventing MCM repositioning to the "active location" in the first place, our results suggest that suppression occurs through the former mechanism. Consistent with previous reports that fun30 mutants are deficient in transcriptional silencing (Neves-Costa et al. 2009), C-pro RNA levels were approximately twice as high in sir2 fun30 cells compared to sir2 cells when adjusted for rDNA size (Figure 5—figure supplement 6).

Moreover, deletion of FUN30 shifts the distribution toward the repositioned MCM location over the non-repositioned one in G1 cells (Figure 5C), aligning with the increased C-pro transcription observed in fun30 mutants. This shift is evident in both sir2 and SIR2 cells. Despite the increased transcription-mediated repositioning in sir2 fun30 cells compared to sir2 cells during G1, repositioned MCM persists longer in sir2 fun30 cells than in sir2 cells after release into HU. Additionally, sir2 fun30 mutants exhibit reduced MCM accumulation at the RFB compared to sir2 mutants after release into HU, supporting the conclusion that MCM disappearance in HU reflects origin activation rather than transcription-mediated displacement.”

The model in Fig 7 implies that initiation sites are different in WT versus the mutants and this determines their timing/efficiency. But they also suggest that the same site might be used with different efficiencies in this response. I agree that both are possibilities and are not resolved. 

Adjustment of the model to account for repositioned MCMs in WT cells In Figure 5—figure supplement 5, we demonstrate that even in WT cells, a small fraction of repositioned MCMs (~5%) can be detected, and that these repositioned MCM complexes disappear prematurely. However, because this represents a very small fraction of MCMs in WT cells, we initially did not include it in our overall model in Figure 7. In light of the reviewer's comment, we have now revised the model to incorporate this detail.

Supporting their model requires better resolution to determine the actual replication initiation site. While this may be challenging, it should be feasible with methods to map nascent strands like DNAscent, or Okazaki fragment mapping.

The initiation site in sir2 mutants has been thoroughly analyzed and supported by extensive experimental data, as discussed above. While high-resolution techniques such as DNAscent or Okazaki fragment mapping could potentially offer another layer of validation, the likelihood of obtaining finer detail that would change the conclusions is minimal. The methods we employed provide sufficient resolution to pinpoint the initiation site, and our results align consistently with established replication models.

Further experimentation would not only be redundant but also unlikely to provide new insights beyond revalidation. Given the strength of our current data, we believe the conclusions regarding replication initiation are robust and well-supported, making additional experiments unnecessary at this stage. Our priority is to focus on advancing other aspects of the research that require deeper exploration.

The 2D gel analysis of strains with reduced rDNA copy numbers adequately addresses the copy number variable with regard to the replication effect. 

Overall, the paper is improved by providing additional data and improved analysis. The paper nicely characterizes the effect of Fun30. The model is reasonable but remains lacking in precise details of mechanism. 

Reviewer #2 (Public review): 

Summary: 

In this manuscript, the authors follow up on their previous work showing that in the absence of the Sir2 deacetylase the MCM replicative helicase at the rDNA spacer region is repositioned to a region of low nucleosome occupancy. Here they show that the repositioned displaced MCMs have increased firing propensity relative to non-displaced MCMs. In addition, they show that activation of the repositioned MCMs and low nucleosome occupancy in the adjacent region depend on the chromatin remodeling activity of Fun30. 

Strengths: 

The paper provides new information on the role of a conserved chromatin remodeling protein in regulation of origin firing and in addition provides evidence that not all loaded MCMs fire and that origin firing is regulated at a step downstream of MCM loading. 

Weaknesses: 

The relationship between the authors results and prior work on the role of Sir2 (and Fob1) in regulation of rDNA recombination and copy number maintenance is not explored, making it difficult to place the results in a broader context. Sir2 has previously been shown to be recruited by Fob1, which is also required for DSB formation and recombination-mediated changes in rDNA copy number. Are the changes that the authors observe specifically in fun30 sir2 cells related to this pathway? Is Fob1 required for the reduced rDNA copy number in fun30 sir2 double mutant cells? 

Reviewer #3 (Public review): 

Summary: 

Heterochromatin is characterized by low transcription activity and late replication timing, both dependent on the NAD-dependent protein deacetylase Sir2, the founding member of the sirtuins. This manuscript addresses the mechanism by which Sir2 delays replication timing at the rDNA in budding yeast. Previous work from the same laboratory (Foss et al. PLoS Genetics 15, e1008138) showed that Sir2 represses transcription-dependent displacement of the Mcm helicase in the rDNA. In this manuscript, the authors show convincingly that the repositioned Mcms fire earlier and that this early firing partly depends on the ATPase activity of the nucleosome remodeler Fun30. Using read-depth analysis of sorted G1/S cells, fun30 was the only chromatin remodeler mutant that somewhat delayed replication timing in sir2 mutants, while nhp10, chd1, isw1, htl1, swr1, isw2, and irc5 had no effect. The conclusion was corroborated with orthogonal assays including two-dimensional gel electrophoresis and analysis of EdU incorporation at early origins. Using an insightful analysis with an Mcm-MNase fusion (Mcm-ChEC), the authors

show that the repositioned Mcms in sir2 mutants fire earlier than the Mcm at the normal position in wild type. This early firing at the repositioned Mcms is partially suppressed by Fun30. In addition, the authors show Fun30 affects nucleosome occupancy at the sites of the repositioned Mcm, providing a plausible mechanism for the effect of Fun30 on Mcm firing at that position. However, the results from the MNAse-seq and ChEC-seq assays are not fully congruent for the fun30 single mutant. Overall, the results support the conclusions providing a much better mechanistic understanding how Sir2 affects replication timing at rDNA, 

Strengths 

(1) The data clearly show that the repositioned Mcm helicase fires earlier than the Mcm in the wild type position. 

(2) The study identifies a specific role for Fun30 in replication timing and an effect on nucleosome occupancy around the newly positioned Mcm helicase in sir2 cells. 

Weaknesses 

(1) It is unclear which strains were used in each experiment. 

(2) The relevance of the fun30 phospho-site mutant (S20AS28A) is unclear. 

(3) For some experiments (Figs. 3, 4, 6) it is unclear whether the data are reproducible and the differences significant. Information about the number of independent experiments and quantitation is lacking. This affects the interpretation, as fun30 seems to affect the +3 nucleosome much more than let on in the description. 

Recommendations for the authors:  

Reviewer #2 (Recommendations for the authors): 

The authors have addressed my concerns by the addition of new experiments and analysis. 

One point remains unclear regarding additional support for the Mcm-ChEC results using ChIP experiments to verify whether MCM redistributes in sir2D cells. In their rebuttal, the authors state that, "New supporting based evidence: ChIP at rDNA Origins. Our ChIP analysis also shows that the disappearance of the MCM signal at rDNA origins in sir2Δ cells released into HU is accompanied by signal accumulation at the replication fork barrier (RFB), indicative of stalled replication forks at this location (Figure 5 figure supplement 3)...." The ChIP data in Figure 5 supplement 3 show accumulation of the Mcm2 ChIP signal to the left of the RFB in sir2D cells but it doesn't look like there is any decrease in the MCM signal in sir2D relative to wild-type cells for the peak C-Pro. There is a new MCM peak suggesting perhaps a new MCM loading event. 

Figure 5 figure supplement 3 shows the relative abundance of the MCM ChIP signal across the ~2 kb rDNA region, spanning from the MCM loading site at the rDNA origin (on the left) to the replication fork barrier (RFB) on the right. The MCM-ChIP data are normalized to the highest signal within this rDNA region rather than across the entire genome, meaning that only the relative abundance of MCM within this region is represented, and not comparisons between different conditions. We have now presented the results with the same axes for both alpha factor and HU.

In wild-type (WT) cells, the MCM signal remains primarily at the initial loading site. However, in sir2 mutants, a significant portion of the MCM signal shifts rightward, consistent with rDNA origin activation and the movement of MCM along with the progressing replication fork. While some replication forks stall at the RFB, others are positioned between the MCM loading site and the RFB. The additional MCM peak observed does not represent a new MCM loading event, as the experiment was conducted during S-phase, when new MCM loading is not possible.

Reviewer #3 (Recommendations for the authors): 

In this revision the authors addressed my concerns and improved the manuscript and the presentation of the data. All my recommendations were implemented.

eLife Assessment

This paper aims to understand why prostate cancer with CDK12 loss does not respond to HRd-based therapeutics, such as PARP inhibitors. The work is felt to be fundamental given a thorough computational and genomic analysis, the generation of CDK12-adapted cell lines, and potential synthetic vulnerability to CDK13 loss with genetic knockdown or co-inhibition with a CDK12/13 inhibitor. The evidence is compelling given the authors' systematic testing of components of the CDK12/13 pathways in a number of prostate cancer models. Some weaknesses focused on the functional effect of the various mutations found at different CDK12 sites (loss vs. altered), more comprehensive characterization of CDK12 KO lines, and specificity of the CDK12/13 inhibitor and in vivo experimental schema.

Reviewer #1 (Public review):

Summary:

The authors were attempting to identify the molecular and cellular basis for why modulators of the HR pathway, specifically PARPi, are not effective in CDK12 deleted or mutant prostate cancers and they seek to identify new therapeutic agents to treat this subset of metastatic prostate cancer patients. Overall, this is an outstanding manuscript with a number of strengths and in my opinion represents a significant advance in the field of prostate cancer biology and experimental therapeutics.

Strengths:

The patient data cohort size and clinical annotation from Figure 1 are compelling and comprehensive in scope. The associations between tandem duplications and amplifications of oncogenes that have been well-credentialed to be drivers of cancer development and progression are fascinating and the authors identify that in those that have AR amplification for example, there is evidence for AR pathway activation. The association between CDK12 inactivation and various specific gene/pathway perturbations is fascinating and is consistent with previously published studies - it would be interesting to correlate these changes with cell line-based studies in which CDK12 is specifically deleted or inhibited with small molecules to see how many pathways/gene perturbations are shared between the clinical samples and cell and mouse models with CDK12 perturbation. The short-term inhibitor studies related to changes in HRD genes and protein expression with CDK12/13 inhibition are fascinating and suggest differential pathway effects between short inhibition of CDK12/13 and long-term loss of CDK12. The in vivo studies with the inhibitor of CDK12/13 are intriguing but not definitive

Weaknesses:

Given that there are different mutations identified at different CDK12 sites as illustrated in Figure 1B it would be nice to know which ones have been functionally classified as pathogenic and for which ones that the pathogenicity has not been determined. This would be especially interesting to perform in light of the differences in the LOH scores and WES data presented - specifically, are the pathogenic mutations vs the mutations for which true pathogenicity is unknown more likely to display LOH or TD? For the cell inhibition studies with the CDK12/13 inhibitor, more details characterizing the specificity of this molecule to these targets would be useful. Additionally, could the authors perform short-term depletion studies with a PROTAC to the target or short shRNA or non-selected pool CRISPR deletion studies of CDK12 in these same cell lines to complement their pharmacological studies with genetic depletion studies? Also perhaps performing these same inhibitor studies in CDK12/13 deleted cells to test the specificity of the molecule would be useful. Additionally, expanding these studies to additional prostate cancer cell lines or organdies models would strengthen the conclusions being made. More information should be provided about the dose and schedule chosen and the rationale for choosing those doses and schedules for the in vivo studies proposed should be presented and discussed. Was there evidence for maximal evidence of inhibition of the target CDK12/13 at the dose tested given the very modest tumor growth inhibition noted in these studies?

Reviewer #2 (Public review):

Summary:

The study explores the functional consequence of CDK12 loss in prostate cancer. While CDK12 loss has been shown to confer homologous recombination (HR) deficiency through premature intronic polyadenylation of HR genes, the response of PARPi monotherapy has failed. This study therefore performed an in-depth analysis of genomic sequencing data from mCRPC patient tumors, and showed that tumors with CDK12 loss lack pertinent HR signatures and scars. Furthermore, functional exploration in human prostate cancer cell lines showed that while the acute inhibition of CDK12 resulted in aberrant polyadenylation of HR genes like BRCA1/2, HR-specific effects were overall modest or absent in cell lines or xenografts adapted to chronic CDK12 loss. Instead, vulnerability to genetically targeting CDK13 resulted in a synthetic lethality in tumors with CDK12 loss, as shown in vivo with SR4825, a CDK12/13 inhibitor - thus serving as a potential therapeutic avenue.

The evidence supporting this study is based on in-depth genomic analyses of human patients, acute knockdown studies of CDK12 using a CDK12/13 inhibitors SR4835, adaptive knockout of CDK12 using LuCaP 189.4_CL and inducible re-expression of CDK12, CDK12 single clones in 22Rv1 (KO2 and KO5) and Skov3 (KO1), Tet-inducible knockdown of BRCA2 or CDK12 followed by ionizing radiation and measurement of RAD51 foci, lack of sensitivity generally to PARPi and platinum chemotherapy in cells adapted to CDK12 loss, loss of viability of CDK13 knockout in CDK12 knockout cells, and in vivo testing of SE4825 in LuCaP xenografts with intact and CDK12 loss.

Strengths:

Overall, this study is robust and of interest to the broader homologous recombination and CDK field. First, the topic is clinically relevant given the lack of PARPi response in CDK12 loss tumors. Second, the strength of the genomic analysis in CDK12 lost PCa tumors is robust with clear delineation that BRCA1/2 genes and maintenance of most genes regulating HR are intact. Specifically, the authors find that there is no mutational signature or genomic features suggestive of HR, such as those found in BRCA1/2 tumors. Lastly, novel lines are generated in this study, including de novo LuCaP 189.4_CL with CDK12 loss that can be profound for potential synthetic lethalities.

Weakness:

One caveat that continues to be unclear as presented, is the uncoupling of cell cycle/essentiality of CDK12/13 from HR-directed mechanisms. Is this purely a cell cycle arrest phenotype acutely with associated down-regulation of many genes?

While the RAD51 loading ssRNA experiments are informative, the Tet-inducible knockdown of BRCA2 and CDK12 is confusing as presented in Figure 5, shBRCA2 + and -dox are clearly shown. However, were the CDK12_K02 and K05 also knocked down using inducible shRNA or a stable knockout? The importance of this statement is the difference between acute and chronic deletion of CDK12. Previously, the authors showed that acute knockdown of CDK12 led to an HR phenotype, but here it is unclear whether CDK12-K02/05 are acute knockdowns of CDK12 or have been chronically adapted after single cell cloning from CRISPR-knockout.

Given the multitude of lines, including some single-cell clones with growth inhibitory phenotypes and ex-vivo derived xenografts, the variability of effects with SR4835, ATM, ATR, and WEE1 inhibitors in different models can be confusing to follow. Overall, the authors suggest that the cell lines differ in therapeutic susceptibility as they may have alternate and diverse susceptibilities. It may be possible that the team could present this more succinctly and move extraneous data to the supplement.

The in-vitro data suggests that SR4835 causes growth inhibition acutely in parental lines such as 22RV1. However, in vivo, tumor attenuation appears to be observed in both CDK12 intact and deficient xenografts, LuCAP136 and LuCaP 189.4 (albeit the latter is only nominally significant). Is there an effect of PARPi inhibition specifically in either model? What about the the 22RV1-K02/05? Do these engraft? Given the role of CDK12/13 in RNAP II, these data might suggest that the window of susceptibility in CDK12 tumors may not be that different from CDK12 intact tumors (or intact tissue) when using dual CDK12/13 inhibitors but rather represent more general canonical essential functions of CDK12 and CDK13 in transcription. From a therapeutic development strategy, the authors may want to comment in the discussion on the ability to target CDK13 specifically.

Reviewer #3 (Public review):

Significance:

About 5% of metastatic castration-resistant prostate cancers (mCRPC) display genomic alterations in the transcriptional kinase CDK12. The mechanisms by which CDK12 alterations drive tumorigenesis in this molecularly-defined subset of mCRPC have remained elusive. In particular, some studies have suggested that CDK12 loss confers a homologous recombination deficiency (HRd) phenotype, However, clinical studies have not borne out the benefit to PARP inhibitors in patients with CDK12 alterations, despite the fact that these agents are typically active against tumors with HRd.

In this study, Frank et al. reconcile these findings by showing that: (1) tumors with biallelic CDK12 alterations do not have genomic features of HRd; (2) in vitro, HR gene downregulation occurs with acute depletion of CDK12 but is far less pronounced with chronic CDK12 loss; (3) CDK12-altered cells are uniquely sensitive to genetic or pharmacologic inhibition of CDK13.

Strengths:

Overall, this is an important study that reconciles disparate experimental and clinical observations. The genomic analyses are comprehensive and conducted with a high degree of rigor and represent an important resource to the community regarding the features of this molecular subtype of mCRPC.

Weaknesses:

(1) It is generally assumed that CDK12 alterations are inactivating, but it is noteworthy that homozygous deletions are comparatively uncommon (Figure 1a). Instead many tumors show missense mutations on either one or both alleles, and many of these mutations are outside of the kinase domain (Figure 1b). It remains possible that the CDK12 alterations that occur in some tumors may retain residual CDK12 function, or may confer some other neomorphic function, and therefore may not be accurately modeled by CDK12 knockout or knockdown in vitro. This would also reconcile the observation that knockout of CDK12 is cell-essential while the human genetic data suggest that CDK12 functions as a tumor suppressor gene.

(2) It is not entirely clear whether CDK12 altered tumors may require a co-occurring mutation to prevent loss of fitness, either in vitro or in vivo (e.g. perhaps one or more of the alterations that occur as a result of the TDP may mitigate against the essentiality of CDK12 loss).

eLife Assessment

This is an important study showing that age-related gut microbiota modulate uric acid metabolism through the NLRP3 inflammasome pathway and thereby regulate susceptibility to age-related gout. Several experimental approaches (mechanistic insights) and methods (data quality) remain incomplete. This paper should be of interest to researchers working on gout and microbiota.

Reviewer #2 (Public review):

Summary:

In their manuscript, the authors report that fecal transplantation from young mice into old mice alleviates susceptibility to gout. The gut microbiota in young mice is found to inhibit activation of the NLRP3 inflammasome pathway and reduce uric acid levels in the blood in the gout model.

Strengths:

The authors focused on the butanoate metabolism pathway based on the results of metabolomics analysis after fecal transplantation and identified butyrate as the key factor in mitigating gout susceptibility. In general, this is a well-performed study.

Weaknesses:

The discussion on the current results and previous studies regarding the effect of butyrate on gout symptoms is insufficient.

Reviewer #3 (Public review):

The manuscript presents interesting findings on the role of gut microbiota in gout, focusing on the interplay between age-related changes, inflammation, and microbiota-derived metabolites, particularly butyrate. The study provides valuable insights into the therapeutic potential of microbiota interventions and metabolites for managing hyperuricemia and gout.

The manuscript has improved with the revisions made, particularly regarding clarifications on experimental design and the inclusion of supplementary data.

Comments on latest version:

The authors have addressed many previous concerns; however, some areas still require clarification and improvement to support more definitive conclusions.

(1) This study suggests that microbiota interventions, particularly butyrate, show promising therapeutic potential for hyperuricemia and gout. While the authors discuss the functions of certain butyrate-producing bacteria, I recommend further validating the gut microbiota-butyrate pathway by supplementing germ-free animal models with a single butyrate-producing strain, such as Clostridium butyricum. To strengthen the manuscript, I suggest the authors make further revisions to address these key issues.

(2) Additionally, I was unable to locate the full-length, uncropped Western blot images in the manuscript or supplementary materials. Could the authors please provide these?

Author response:

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

Reviewer #2 (Public review):

Summary:

In their manuscript the authors report that fecal transplantation from young mice into old mice alleviates susceptibility to gout. The gut microbiota in young mice is found to inhibit activation of the NLRP3 inflammasome pathway and reduce uric acid levels in the blood in the gout model.

Strengths:

They focused on the butanoate metabolism pathway based on the results of metabolomics analysis after fecal transplantation and identified butyrate as the key factor in mitigating gout susceptibility. In general, this is a well-performed study.

Weaknesses:

The discussion on the current results and previous studies regarding the effect of butyrate on gout symptoms is insufficient. The authors need to provide a more thorough discussion of other possible mechanisms and relevant literature.

Reviewer #2 (Recommendations for the authors):

General comments:

I appreciate the authors' efforts to answer the comments raised in my previous review (as Reviewer#2). However, I still detect some issues that need to be fully addressed, with inadequate or even no answers for several comments.

Thank you for your valuable feedback. Your previous suggestions have been incredibly helpful for our paper. Although we have strived to make the article as comprehensive as possible, there may still be some areas that are not perfectly refined.

The response to comment 1: The author's statement is not very convincing. What are the trends of inflammation factors? The data in Figure 1G-H suggest that butyrate may not be the only factor to explain this phenomenon. Authors should carefully interpret the data in Figure 1G-H.

Sorry for the inadequate clarification on your question. We utilize antibiotics for treatment in order to establish the relationship between gut microbiota, age, and gout. Our research findings indicate that there is a trend for serum uric acid levels to increase with age, and we also observe that the older the age, the more pronounced the stimulation to MSU. We found that after clearing the gut microbiota and then stimulating with MSU, the trend of inflammation factors and serum uric acid level changing with age disappears. Thus, we can preliminarily draw the conclusion that the gut microbiota is closely associated with age, gout, and hyperuricemia.

The response to comment 2: I understand the importance of evaluating a range of indicators, but food thickness is the most crucial clinical marker for diagnosing goats. Please move the data from Supplemental Figure 1A to the main figure.

Thank you for your suggestions. We have included the most significant results in the main figure, and the description of “foot thickness” has already been provided descriptively in the manuscript. Additionally, considering the layout and arrangement of the images, we have placed it in the supplementary figures 1.

The response to comment 3: The immunostaining for ZO-1 and Occludin is unclear. Please provide higher magnification images to confirm the specific staining.

Thank you for your valuable feedback. We have enhanced the clarity of the images. In addition to adding immunohistochemical images in Supplementary Material 4, we have also submitted independent images.

The response to comment 4: The authors still haven't directly addressed my comment.

Please accept our sincere apologies for not providing a clearer response to your question. The indicators related to uric acid-producing enzymes and uric acid transporters have been separately analyzed according to different age groups. The specific results are detailed in section " The expression of uric acid-producing enzymes activity and uric acid transporters at the mRNA level across different age groups" of Supplementary Material 4.

No response was given for comment 5. Please address it.

In a PCoA plot, the distance between samples reflects the similarity in the structure of the microbial communities: the closer the distance, the more similar the composition of the communities; the greater the distance, the more pronounced the differences. We judge based on the relative distances of each group in the plot, observing their degree of proximity.

The response to comment 6: I understand the author's statement, and I suggest incorporating it into the discussion section of the revised manuscript.

Thank you for your suggestions. We have incorporated the relevant content into our discussion.

The response to comment 7: Again, please incorporate this statement into the discussion section of the revised manuscript.

Thank you for your suggestions. We have incorporated the relevant content into our discussion.

Reviewer #3 (Public review):

Summary:

The revised manuscript presents interesting findings on the role of gut microbiota in gout, focusing on the interplay between age-related changes, inflammation, and microbiota-derived metabolites, particularly butyrate. The study provides valuable insights into the therapeutic potential of microbiota interventions and metabolites for managing hyperuricemia and gout. While the authors have addressed many of the previous concerns, a few areas still require clarification and improvements to strengthen the manuscript's clarity and overall impact.

(1) While the authors mention that outliers in the data do not affect the conclusions, there remains a concern about the reliability of some figures (e.g., Figure 2D-G). It is recommended to provide a more detailed explanation of the statistical analysis used to handle outliers. Additionally, the clarity of the Western blot images, particularly IL-1β in Figure 3C, should be improved to ensure clear and supportive evidence for the conclusions.

Thank you for your suggestion. We respond as follows: (1) Outliers can occasionally constitute intrinsic elements of the dataset, reflecting genuine occurrences within the experimental context. The elimination of such outliers has the potential to introduce bias into the results, thereby facilitating misconceptions regarding the underlying phenomenon under investigation. In order to maintain the transparency and integrity of the dataset, we have elected to retain the outliers within our analysis. This decision is based on the recognition that these values may represent genuine experimental observations or unique conditions that are inherently meaningful to the phenomenon under investigation. By preserving these data points, we aim to provide a comprehensive and unbiased representation of the experimental results, allowing for a more nuanced interpretation of the findings. (2) Due to the scarcity of samples, we are unable to fulfill your request in the short term. Furthermore, we have noted that the band for IL-1β in Figure 3C is indeed visible and we consider it suitable for subsequent analysis.

(2) The manuscript raises a key question about why butyrate supplementation and FMT have different effects on uric acid metabolism and excretion. While the authors have addressed this by highlighting the involvement of multiple bacterial genera, it is still recommended to expand on the differences between these interventions in the discussion, providing more mechanistic insights based on available literature.

Thank you for your suggestion. We have enriched the discussion in the manuscript and included additional comparisons

(3) It is noted that IL-6 and TNF-α results in foot tissue were requested and have been added to supplementary material. However, the main text should clearly reference these additions, and the supplementary figures should be thoroughly reviewed for consistency with the main findings. The use of abbreviations (e.g., ns for no significant difference) and labeling should also be carefully checked across all figures.

Thank you for your valuable feedback. We have revised the manuscript in accordance with your suggestions.

(4) The manuscript presents butyrate as a key molecule in gout therapy, yet there are lingering concerns about its central role, especially given that other short-chain fatty acids (e.g., acetic and propionic acids) also follow similar trends. The authors should consider further acknowledging these other SCFAs and discussing their potential contribution to gout management. Additionally, the rationale for focusing primarily on butyrate in subsequent research should be made clearer.

Thank you for your input. We have incorporated additional evidence into the discussion, explaining why we ultimately chose butyrate in subsequent research.

(5) The full-length uncropped Western blot images should be provided as requested, to ensure transparency and reproducibility of the data.

Thank you for your suggestion. We have already included the relevant explanations in the manuscript.

(6) Despite the authors' revisions, several references still lack page numbers. Please ensure that all references are properly formatted, including complete page ranges.

Thank you for your suggestions; we will make more detailed revisions to the references.

The manuscript has improved with the revisions made, particularly regarding clarifications on experimental design and the inclusion of supplementary data. However, some concerns about data quality, mechanistic insights, and clarity in the figures remain. Addressing these points will enhance the overall impact of the work and its potential contribution to the understanding of the gut microbiome in gout and hyperuricemia. A final revision, with careful attention to both major and minor points, is highly recommended before resubmission.

Once again, we are grateful for your suggestions and recognition. Your input has been of immense help to our manuscript and has also provided us with a valuable learning opportunity.

eLife Assessment

The aim of this important study is to identify novel genes involved in sleep regulation and memory consolidation. It combines transcriptomic approaches following memory induction with measurements of sleep and memory to discover molecular pathways underlying these interlinked behaviors. The authors explore transcriptional changes in specific mushroom body neurons and suggest roles for two genes involved in RNA processing, Polr1F and Regnase-1, in the regulation of sleep and memory. Their findings offer convincing evidence that the expression of RNA processing genes is modulated during sleep-dependent memory, with Polr1F potentially contributing to increased sleep.

Reviewer #2 (Public review):

Sleep and memory are intertwined processes, with sleep-deprivation having a negative impact on long-term memory in many species. Recently, the authors showed that fruit flies form sleep-dependent long-term appetitive memory only when fed. They showed that this context-dependent memory trace maps to the anterior posterior (ap) α'β' mushroom body neurons (MBNs) (Chouhan et al., (2021) Nature). However, the molecular cascades induced by during training that promote sleep and memory have remained enigmatic.

Here the authors investigate this issue by combining cell-specific transcriptomics, genetic perturbations, and measurements of sleep and memory. They identify an array of genes altered in expression following appetitive training. These genes are mainly downregulated, and predominantly encode regulators of transcription and RNA biosynthesis. This is a conceptually attractive finding given that long-term memory requires de novo protein translation.

The authors then screen these genes for novel regulators of sleep and memory. They show that one of these genes (Polr1F) acts in ap α'β' MBNs to promote wakefulness, while another (Regnase-1) promotes sleep. They also identify a specific role for Regnase-1 in ap α'β' MBNs in regulating short- and long-term memory formation - likely through effects on the development of ap α'β' MBNs - and demonstrate that Pol1rF inhibits translation throughout the fly brain.

The analyses of molecular alterations in ap α'β' MBNs are interesting and impressive. However, as noted by the authors, further experiments are required to clarify the precise contribution of reductions in Polr1F and Regnase-1 to training-induced changes in memory and sleep. Nonetheless, this study provides a useful platform for such studies, and provides a conceptual advance in linking acute changes in RNA processing pathways to the interconnected processes of sleep, memory, and protein translation.

Reviewer #3 (Public review):

Previous work (Chouhan et al., 2022) from the Sehgal group investigated the relationship between sleep and long-term memory formation by dissecting the role of mushroom body intrinsic neurons, extrinsic neurons, and output neurons during sleep-dependent and sleep-independent memory consolidation. In this manuscript, Li et al., profiled transcriptome in the anterior-posterior (ap) α'/β' neurons and identified genes that are differentially expressed after training in fed condition, which supports sleep-dependent memory formation. By knocking down candidate genes systematically, the authors identified Polr1F and Regnase-1 as two important hits that play potential roles in sleep and memory formation. What is the function of sleep and how to create a memory are two long-standing questions in science. The present study used a new approach to identify novel components that may link sleep and memory consolidation in a specific type of neuron. Importantly, these components implicated that RNA processing may play a role in these processes.

I am enthusiastic about the innovative approach employed to identify RNA processing genes involved in sleep regulation and memory consolidation. During the revision process, the authors fully addressed major concerns raised by reviewers. First, the author used the Gal80ts to restrict the knockdown of Regnase-1 in adult animals and concluded that Regnase-1 RNAi appears to affect sleep through development. Second, the author showed that Regnase-1 knockdown produced robust phenotypes for both sleep-dependent and sleep-independent memory, as well as a severe short-term memory phenotype. The author cautiously concluded that flies with constitutive Regnase-1 knockdown could be poor learners, thereby exhibiting a memory phenotype. Although we don't yet have a strong link between sleep and long-term memory consolidation, the interpretation presented in the manuscript is sufficiently justified by the data. This work presents a novel strategy to explore the link between sleep and memory consolidation.

Reviewer #4 (Public review):

Summary:

Li and Chouhan et al. follow up on a previous publication describing the role of anterior-posterior (ap) and medial (m) ɑ′/β′ Kenyon cells in mediating sleep-dependent and sleep-independent memory consolidation, respectively, based on feeding state in Drosophila melanogaster. The authors sequenced bulk RNA of ap ɑ′/β′ Kenyon cells 1h after flies were either trained-fed, trained-starved or untrained-fed and find a small number of genes (59) differentially expressed (3 upregulated, 56 downregulated) between trained-fed and trained-starved conditions. Many of these genes encode proteins involved in the regulation of gene expression. The authors then screened these differentially expressed genes for sleep phenotypes by expressing RNAi hairpins constitutively in ap ɑ′/β′ Kenyon cells and measuring sleep patterns. Two hits were selected for further analysis: Polr1F, which promoted sleep, and Regnase-1, which reduced sleep. The pan-neuronal expression of Polr1F and Regnase-1 RNAi constructs was then temporally restricted to adult flies using the GeneSwitch system. Polr1F sleep phenotypes were still observed, while Regnase-1 sleep phenotypes were not, indicating developmental defects. Appetitive memory was then assessed in flies with constitutive knockdown of Polr1F and Regnase-1 in ap ɑ′/β′ Kenyon cells. Polr1F knockdown did not affect sleep-dependent or sleep-independent memory, while Regnase-1 knockdown disrupted sleep-dependent memory, sleep-independent memory, as well as learning. Polr1F knockdown increased pre-ribosomal RNA transcripts in the brain, as measured by qPCR, in line with its predicted role as part of the RNA polymerase I complex. A puromycin incorporation assay to fluorescently label newly synthesized proteins also indicated higher levels of bulk translation upon Polr1F knockdown. Regnase-1 knockdown did not lead to observable changes in measurements of bulk translation.

Strengths:

The proposed involvement of RNA processing genes in regulating sleep and memory processes is interesting, and relatively unexplored. The methods are satisfactory.

Weaknesses:

The main weakness of previous versions of the paper was the over-interpretation of results, particularly relating to the proposed link between sleep and memory consolidation. This has now been appropriately addressed, as reflected in the change of title and incorporation of alternative interpretations of the data in the text.

Author response:

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

eLife Assessment

The aim of this valuable study is to identify novel genes involved in sleep regulation and memory consolidation. It combines transcriptomic approaches following memory induction with measurements of sleep and memory to discover molecular pathways underlying these interlinked behaviors. The authors explore transcriptional changes in specific mushroom body neurons and suggest roles for two genes involved in RNA processing, Polr1F and Regnase-1, in the regulation of sleep and memory. Although this work exploits convincing and validated methodology, the strength of the evidence is incomplete to support the main claim that these two genes establish a definitive link between sleep and memory consolidation.

We appreciate the reconsideration of our manuscript and recognize that we should have toned down the claims, especially with respect to the link between sleep and memory consolidation.  We have now changed the title, the abstract and the main text and also Figure 5 to essentially just state our findings.  While there is a little speculation in the Discussion, we point out that future work would be required to draw conclusions. We believe the manuscript still represents a considerable advance in showing the modulation of RNA processing genes during sleep-dependent memory consolidation in the relevant neurons, and also showing how one such gene affects sleep and translation and a second affects sleep and memory. 

Public Reviews:

Reviewer #2 (Public review):

Prior work by the Sehgal group has shown that a small group of neurons in the fly brain (anterior posterior (ap) α'β' mushroom body neurons (MBNs)) promote sleep and sleep-dependent appetitive memory specifically under fed conditions (Chouhan et al., (2021) Nature). Here, Li, Chouhan et al. combine cell-specific transcriptomics with measurements of sleep and memory to identify molecular processes underlying this phenomenon. They define transcriptional changes in ap α'β' MBNs and suggest a role for two genes downregulated following memory induction (Polr1F and Regnase-1) in regulating sleep and memory.

The transcriptional analyses in this manuscript are impressive. The authors have now included additional experiments that define acute and developmental roles for Polr1F and Regnase-1 respectively in regulating sleep. They have also provided additional data to strengthen their conclusion that Polr1F knockdown in α'β' mushroom body neurons enhances sleep.

The resubmitted work represents a convincing investigation of two novel sleep-regulatory proteins that may also play important roles in memory formation.

The authors have comprehensively addressed my comments, which I very much appreciate. I congratulate them on this excellent work.

We very much appreciate the reviewer’s positive feedback. Thank you!

Reviewer #3 (Public review):

Previous work (Chouhan et al., 2022) from the Sehgal group investigated the relationship between sleep and long-term memory formation by dissecting the role of mushroom body intrinsic neurons, extrinsic neurons, and output neurons during sleep-dependent and sleep-independent memory consolidation. In this manuscript, Li et al., profiled transcriptome in the anterior-posterior (ap) α'/β' neurons and identified genes that are differentially expressed after training in fed condition, which supports sleep-dependent memory formation. By knocking down candidate genes systematically, the authors identified Polr1F and Regnase-1 as two important hits that play potential roles in sleep and memory formation. What is the function of sleep and how to create a memory are two long-standing questions in science. The present study used a new approach to identify novel components that may link sleep and memory consolidation in a specific type of neuron. Importantly, these components implicated that RNA processing may play a role in these processes.

While I am enthusiastic about the innovative approach employed to identify RNA processing genes involved in sleep regulation and memory consolidation, I feel that the data presented in the manuscript is insufficient to support the claim that these two genes establish a definitive link between sleep and memory consolidation. First, the developmental role of Regnase-1 in reducing sleep remains unclear because knocking down Regnase-1 using the GeneSwitch system produced neither acute nor chronic sleep loss phenotype. In the revised manuscript, the author used the Gal80ts to restrict the knockdown of Regnase-1 in adult animals and concluded that Regnase-1 RNAi appears to affect sleep through development. Conducting overexpression experiments of Regnase-1 would lend some credibility to the phenotypes, however, this is not pursued in the revised manuscript. Second, while constitutive Regnase-1 knockdown produced robust phenotypes for both sleep-dependent and sleep-independent memory, it also led to a severe short-term memory phenotype. This raises the possibility that flies with constitutive Regnase-1 knockdown are poor learners, thereby having little memory to consolidate. The defect in learning could be simply caused by chronic sleep loss before training. Thus, this set of results does not substantiate a strong link between sleep and long-term memory consolidation. Lastly, the discussion on the sequential function of training, sleep, and RNA processing on memory consolidation appears speculative based on the present data.

We thank the reviewer for the enthusiasm about the approach. As noted above, we have now removed all claims about a link between sleep and memory, and instead just emphasize that we have identified RNA processing genes that affect sleep and memory.  We agree with the reviewer that the basis of the Regnase-1 memory phenotype is unclear as the flies may be poor learners.  Also, the learning/memory defects could be secondary to sleep loss or, as Reviewer 4 below suggests, all the behavioral deficits could be caused by impaired development/function of the relevant ap ɑ′/β′ cells. We have now included this possibility in the discussion of the manuscript.  And we have modified the discussion on training, RNA processing, sleep and memory to emphasize the need for future experiments to address the sequence and relationship of these different processes. 

Reviewer #4 (Public review):

Summary:

Li and Chouhan et al. follow up on a previous publication describing the role of anterior-posterior (ap) and medial (m) ɑ′/β′ Kenyon cells in mediating sleep-dependent and sleep-independent memory consolidation, respectively, based on feeding state in Drosophila melanogaster. The authors sequenced bulk RNA of ap ɑ′/β′ Kenyon cells 1h after flies were either trained-fed, trained-starved or untrained-fed and find a small number of genes (59) differentially expressed (3 upregulated, 56 downregulated) between trained-fed and trained-starved conditions. Many of these genes encode proteins involved in the regulation of gene expression. The authors then screened these differentially expressed genes for sleep phenotypes by expressing RNAi hairpins constitutively in ap ɑ′/β′ Kenyon cells and measuring sleep patterns. Two hits were selected for further analysis: Polr1F, which promoted sleep, and Regnase-1, which reduced sleep. The pan-neuronal expression of Polr1F and Regnase-1 RNAi constructs was then temporally restricted to adult flies using the GeneSwitch system. Polr1F sleep phenotypes were still observed, while Regnase-1 sleep phenotypes were not, indicating developmental defects. Appetitive memory was then assessed in flies with constitutive knockdown of Polr1F and Regnase-1 in ap ɑ′/β′ Kenyon cells. Polr1F knockdown did not affect sleep-dependent or sleep-independent memory, while Regnase-1 knockdown disrupted sleep-dependent memory, sleep-independent memory, as well as learning. Polr1F knockdown increased pre-ribosomal RNA transcripts in the brain, as measured by qPCR, in line with its predicted role as part of the RNA polymerase I complex. A puromycin incorporation assay to fluorescently label newly synthesized proteins also indicated higher levels of bulk translation upon Polr1F knockdown. Regnase-1 knockdown did not lead to observable changes in measurements of bulk translation.

Strengths:

The proposed involvement of RNA processing genes in regulating sleep and memory processes is interesting, and relatively unexplored. The methods are satisfactory.

Weaknesses:

The main weakness of the paper is in the overinterpretation of their results, particularly relating to the proposed link between sleep and memory consolidation, as stated in the title. Constitutive Polr1F knockdown in ap ɑ′/β′ Kenyon cells had no effect on appetitive long-term memory, while constitutive Regnase-1 knockdown affected both learning and memory. Since the effects of constitutive Regnase-1 knockdown on sleep could be attributed to developmental defects, it is quite plausible that these same developmental defects are what drive the observed learning and memory phenotypes. In this case, an alternative explanation of the authors' findings is that constitutive Regnase-1 knockdown disrupts the entire functioning of ap ɑ′/β′ Kenyon cells, and as a consequence behaviors involving these neurons (i.e. learning, memory and sleep) are disrupted. It will be important to provide further evidence of the function of RNA processing genes in memory in order to substantiate the memory link proposed by the authors.

As noted above, we have removed claims of a link between sleep and memory and instead focused the manuscript on our findings of RNA processing genes modulated during sleep-dependent memory. We concur that impaired development of ap ɑ′/β′neurons could account for the sleep and memory phenotype observed and have included this possibility in the manuscript.

Recommendations for the authors:

Reviewer #4 (Recommendations for the authors):

The title of the paper should be reconsidered to reflect the results. The evidence for a link between RNA processing genes and memory is weak.

We have changed the title.

Line 328. The term "central dogma" is misused. The central dogma refers to the unidirectional flow of information from DNA to protein. Instead the authors mean "gene expression".

Changed, thank you.

A couple of minor comments relating to the figures:

Figure 1b. It is not clear what the number 10570 in the bottom right corner refers to.

Fixed.

Figure 3b. RU- and RU+ annotation is missing (as shown in 3d).

Fixed.

Author response:

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

Reviewer #1 (Public Review)

(1) The identification of the proximal to distal degeneration of the tailgut within the human tail is difficult to distinguish with the current images present in Figure 3. A picture within a picture of the area containing the tail gut could be provided to prominently demonstrate the cellular architecture. Additionally, quantification of the localization of apoptosis would strongly support this observation, as well as provide a visualization of the tail's regression overall. For example, a graph plotting the number of apoptotic cells versus the rostral to caudal locations of the transverse sections while accounting for the CS stage of each analyzed embryo could be created; this could even be further broken down by region of tail, for example, tailgut, ventral ectodermal ridge, somite, etc.

To provide more information on apoptosis, we prepared serial sections from an additional 6 human tails, 5 of which were processed for fluorescence anti-caspase 3 immunohistochemistry with DAPI staining (Fig 4) and H&E (Fig 6). This confirmed our previous finding of apoptosis especially in the tailgut and ventral mesoderm. We have not quantified the apoptosis, given the difficulty of deciding whether anti-caspase signals represent single or multiple dying cells. Instead, we performed a tissue area analysis from caudal to rostral along the tail (new section on p 9). This shows a progressive enlargement of the neural tube, no change in the notochord and a striking reduction in tailgut area (Fig 4C,D). The smaller tailgut has fewer nuclei in cross section rostrally compared with more caudally (Fig 4E). Given that apoptosis is present in the tailgut at all rostro-caudal levels, this is consistent with a rostralto-caudal loss of the tailgut, as is also found in mouse and rat embryos.

(2) The identification of the mode of formation of the secondary neural tube is probably the most interesting question to be addressed, however, Figure 7's evidence is not completely satisfying in its current form. While I agree that it is unlikely that multiple polarization foci form within the most caudal part of the tail and coalesce more rostrally, I am equally unsure that a single polarization would form rostrally and then split and re-coalesce as it moves caudally, as is currently depicted by 7B. Multiple groups have recently shown the influence of geometric confinement on neuroectoderm and its ability to polarize and form a singular central lumen (Karzbrun 2021, Knight 2018), or the inverse situation of a lack of confinement resulting in the presence of multiple lumens. The tapering of the diameter of the tail and its shared perimeter and curvature with the polarization bears a striking resemblance to this controlled confinement. An interesting quantification to depict would include the number of lumens versus the transverse section diameter and CS stage to see if there is any correlation between embryo size and the number of multiple polarizations. Anecdotally, the fusion of multiple polarizations/lumens tends to occur often in these human organoid-type platforms, while splitting to multiple lumens as the tissues mature does not. Other supplements to Figure 7 could include 3D renderings of lumens of interest as depicted in Catala 2021, especially if it demonstrates the recoalescence as seen in 7B. The non-pathologic presence of multiple polarizations in human tails compared to the rodent pathogenic counterpart is interesting given that rodents obviously maintain this appendage while it is lost in humans.   

The additional 6 sectioned human embryo tails (as described above) provide further information in support of the original findings of the paper: (i) that the secondary neural tube formation initially involves a single lumen, and (ii) that neural tube duplication occurs in many tails at more rostral levels. Neural tube duplication was observed in 15/25 of our sectioned tails: hence, overall 60% of human tails exhibited neural tube duplication in this study. We have replaced all the cross sectional images in the original Fig 7 (now Fig 6) to better illustrate the findings of neural tube duplication at relatively rostral levels of the human tail. Additionally, the axial position of sections containing duplicated neural tube are indicated by arrows in the graph of neural tube areas (Fig 4C). From this analysis it appears that neural tube duplication is not contingent on an increasing tail diameter, as raised by the reviewer, because some tails show a transition to neural tube duplication, and then return to an single lumen morphology more rostrally. While the 3D renderings of lumens would be interesting, we consider it beyond the scope of the present study.

(3) Of potential interest is the process of junctional neurulation describing the mechanistic joining of the primary and secondary neural tube, which has recently been explored in chick embryos and demonstrated to have relevance to human disease (Dady 2014, Eibach 2017, Kim 2021). While it is clear this paper's goal does not center on the relationship between primary and secondary neurulation, such a mechanism may be relevant to the authors' interpretation of their observations of lumen coalescence. I wonder if the embryos studied provide any evidence to support junctional neurulation.  

We agree this is an important point to address in the paper, and a new section has been inserted in the Discussion: ‘Transition from primary to secondary neurulation’ (pp 13-14). In brief, we find no evidence for a specific mode of ‘junctional neurulation’ in the human embryos. In any event, its existence is hypothetical in humans, suggested largely as an ‘embryological explanation’ for the finding of rare interrupted spinal cord defects in neurosurgical patients (Eibach, 2017). In chick neurulation there is longitudinal dorso-ventral overlap between the primary and secondary neural tubes (Dryden, 1980), with the junctional zone derived from ingressing cells at the node-streak border (Dady, 2014), a known source of neuromesodermal progenitors (NMPs). However, this is a very different developmental situation from the human so-called ‘junctional neurulation’ defect (Eibach, 2017), in which the spinal cord is physically and functionally interrupted, with only a rudimentary filament connecting the rostral and caudal parts.

Reviewer #1 (Recommendations For The Authors):

(1) Figures 3, 4, and 7, would be easier to digest quickly with inclusions of labels that mark the rostral and caudal transverse sections. For example, "caudal" over 3G and "rostral" over 3F.  

Figures 3 and 4 have been combined to form revised Figure 3, and the rostral/caudal sections are no longer included, as these are superseded by the new Figure 4. Similarly Figure 7 has been replaced by new images in the revised Figure 6, with clear labelling of axial levels.

(2) The manuscript does a nice job of comparing and contrasting the human findings to mouse, however, there are several instances where it would be nice to continue this trend within the text, such as including the rate of somite formation for rodents in the sections that you state the quantified human and published organoid findings, as well as the total number of somite rodents' exhibit. Additionally, the last sentence of the "Morphology of human PNP closure" section correctly states that human PNP's seem to close via Mode 2 neurulation that is seen in the mouse. However, my read of the literature (published by Dr. Copp) demonstrates that the PNP in mice actually closes via Mode 3 at the most caudal portion. If this is the case, it would be pointed to explicitly state that regionally dependent morphogenetic difference between the two species.  

We agree these are important points to include. The additional somite data (for mouse) has been inserted in the Results section on ‘Somite formation’ (p 8), and the apparent absence of Mode 3 during human spinal neural tube closure is now included in the new Discussion section, ‘Transition from primary to secondary neurulation’ (pp 13-14).

(3) The introduction to secondary neural tube formation with the hypothesis diagrams in Figure 7 is slightly jarring. At the beginning of the Figure, a schematic depicting the morphogenetic differences between primary and secondary would be helpful in introducing the readership to these complex embryologic events. An example of this could be similar to Figure 1 in Dr. Copp's paper:

Nikolopoulou, E., et al. Neural tube closure: cellular, molecular and biomechanical mechanisms. Development 144, 552-566 (2017).  

We feel that a summary diagram of primary and secondary neurulation would simply reproduce diagrams that are already widespread in the literature. As noted by the Reviewer, our article in Development (Nikolopoulou, 2017) contains just such a summary diagram as Figure 1. Therefore, we prefer to explicitly cite this article/figure in our Introduction (see modified first sentence, third paragraph, p 3), so that readers can consult the freely accessible Nikolopoulou review for more detail. The diagram in Figure 7 (now revised Figure 6) has been completely redrawn to make much clearer the hypotheses being examined in the study of human secondary neural tube formation, and neural tube duplication.

(4) Finally, a matter of semantics, the second paragraph of the introduction describes myelomeningocele as a neurodegenerative defect, while it is true amniotic fluid further degrades exposed neural tissue while exposed, to me, the term neurodegenerative defect suggests a lifelong degeneration, which is not the case for human patients. Perhaps shortening to neurological defect is a compromise. Thank you for the important and interesting work.  

We agree that ‘neurodegenerative’ can mean different things to different people. Literally, it refers to degeneration of neural tissue, which of course includes neuroepithelial loss due to amniotic fluid action in the uterus. Nevertheless, to avoid confusion, the word has been removed and the sentence expanded to include a reference to the adverse effects of amniotic fluid on the exposed neuroepithelium (see Introduction, second paragraph, p 3).

Reviewer #2 (Public Review)

It is not clear how the gestational age of the specimens was determined or how that can be known with certainty. There is no information given in the methods on this. With this in mind, bunching the samples at 2-day intervals in Figure 1J will lead to inaccuracies in assessing the rate of somite formation. This is pointed out as a major difference between specimens and organoids in the abstract but a similar result in the results section. The data supporting either of these statements is not convincing.

Human embryos were assigned to Carnegie stages based on standard morphological criteria. This was stated, with references, in the first Results paragraph, and we have now also included this information in the Methods (first paragraph, p 19). We assigned the embryos to 2-day intervals based on the standard literature timing of these Carnegies stages, as described in O’Rahilly and Muller (1987). We have clarified both Carnegie staging and assignment of embryos to 2-day intervals in a new sentence within the Methods, first paragraph, p 19. “Embryos were assigned to Carnegie Stages (CS) using morphological criteria (O'Rahilly and Muller 1987; Bullen and Wilson 1997) and to 2-day post-conception intervals for regression analysis based on timings in Table 0-1 of O’Rahilly and Muller (1987).” This has also been inserted in the legend to Figure 1J.

The regression analysis of somite number against days post-conception (Figure 1J) allowed a conclusion to be drawn on the rate of somite formation in early human embryos. We have added 95% confidence intervals to our finding of a new somite formed every 7.1 h in humans. We consider this to be important for comparison with non-human species and organoid systems. On p 8, second paragraph, we simply state our finding of a 7.1 h somite periodicity in human embryos, compared with 5 h in the organoid system (and 2 h in mouse and rat – as suggested by Reviewer 1). We are careful not to say it is a ‘major difference’ or ‘similar result’ in different parts of the paper, as the Reviewer has drawn attention to.

Whenever possible, give the numbers of specimens that had the described findings. For example, in Figure 2C - how many embryos were examined with the massive rounded end at CS13? Apoptosis in Figures 3 and 4?  

Numbers of embryos analysed in Figures 2 and 3 (the latter now a combined version of the original Figures 3 and 4) are shown in Table 2. We have also created a new Supplementary Figure 1 to show additional examples of human embryonic tails, which illustrate the consistency of morphology through the stages from CS13 to CS18. Numbers of samples that contributed to Figures 4-6 are detailed in the legends.

For Figure 2I-K, it would be informative to superimpose the individual data points on the box plots distinguishing males from females, as in Figure 1I.  

This was attempted but the data points overlie the box plots and look confusing. Instead, we have created Supplementary Table 2 which gives the raw data on which Figure 2I-K are

based. We have also drawn attention to the fact that not all embryos yielded all types of measurement, especially tail lengths.

Is it possible to quantitate apoptosis and proliferation data?  

We have not quantified apoptosis, given the difficulty of deciding whether anti-caspase signals represent single or multiple dying cells. Instead, we performed a new tissue area analysis along the body axis, which has shed light on the possible direction (rostral to caudal) of tailgut loss in the human caudal region (see response to Reviewer 1 above). Since the cell proliferation data were limited in extent, and not a major focus of the paper, we have removed that analysis completely from the revised version.

The Tunel staining in Figure 3 is difficult to make out.  

We have extended our analysis of anti-caspase 3 immunohistochemisty and removed the TUNEL images.

Reviewer #2 (Recommendations For The Authors)

The anatomy of the sections in Figures 3, 4, and 7 is difficult to discern. Is it possible to insert adjacent panels tracing and labeling the structures in each panel? Also, drawings showing the axial level of each section would be helpful.

To clarify the axial levels of sections, we have inserted images of mouse and human embryos as parts A and B of the revised Figure 3. We have tried to clarify the morphology of sections by labelling all relevant structures in the sections themselves.   

High-magnification views of the tailbud in Figure 5 would be more informative. Staining is difficult to see after CS13. The low-magnification views can be shown in an insert. Figures 5 and 6 can be combined.

At the reviewer’s suggestion, we have merged Figures 5 and 6 into a revised Figure 5. Now, the sections provide higher magnification images of the areas of expression as shown in the lower magnification whole mount images. We feel this makes the gene expression findings much clearer than before.

Some of the writing in the abstract, introduction, and results is very descriptive, with a lack of summary and integration of information. For instance, the abstract could be rewritten to include an overall conclusion at the end and a better description of the longstanding questions addressed. Moreover, the abstract suggests multiple lumens are not found in human specimens. Another example is the second paragraph of the introduction lists various NTDs but doesn't provide an integrative conclusion of the information. The discussion is much better but lacks a conclusion at the end.

We agree that more concluding sentences should be used, as the Reviewer suggests. To this end, we have rewritten the Abstract (p 2) to emphasise the long-standing questions that our study addresses, and concluding sentences are now included in other places (e.g. somite results, p 8). A new ‘Conclusions’ section has been added at the end of the Discussion (pp 17-18).

ADDITIONAL CHANGES MADE TO REVISED MANUSCRIPT

Title. This has been amended to: “Spinal neural tube formation and tail development in human embryos” to reflect the greater focus on developmental events, and less on tail regression.

Additional studies have been added to Supplementary Table 1, to include the main transcriptomic studies of human embryos in the primary/secondary neurulation stage range. This takes the number of previous studies to 28 and the total number of embryos to 925. See p 4, top and p 12, first paragraph for corresponding changes to the text.

We added a sentence to the Discussion (p 13, first paragraph) to counter the claim that humans have undergone ‘tail-loss’, as included in Xia et al, 2024, “On the genetic basis of tail-loss evolution in humans and apes”. Nature 626:1042-8. Clearly, the human embryo is tailed, which undermines these authors’ statement.

eLife Assessment

This is a fundamental study into human spinal neurulation, which substantially advances our understanding of human neural tube closure. Crucial unanswered questions in the field currently rely on model systems, not faithful to human development. The evidence provided is compelling, with a large number of specimens and the rigorous use of state-of-the-art methodology providing robustness. The work will be of broad interest to developmental biologists, embryologists, and medical professionals working on neural tube defects, and will act as a precious reference resource for future studies.

Reviewer #1 (Public review):

Summary:

The authors analyzed 108 human embryos in order to address outstanding questions about human lower spinal development and secondary neural tube formation. Through whole embryo imaging and histologic analysis, they have provided exceptional quantification of the timing of posterior neuropore closure, rate of lower spinal somite formation, and formation and regression of the human tail. Their analysis has also provided convincing qualitative evidence of the cellular and molecular mechanisms at play during lower spinal development, by identifying the presence of caspase-dependent programmed cell death and the dynamic expression of FGF8/WNT3A within the elongating embryo. Interestingly, they identified multiple polarized lumens within the site of secondary neural tube formation, and added a solid argument for the mode of formation of this structure; however, the evidence for a conclusive morphogenetic mechanism remains elusive. Finally, the authors provided a substantial review of the existing publications related to human lower spinal development, creating an excellent reference and demonstrating the importance of continuing to use each of these precious samples to further advance our understanding of human development.

Strengths:

This manuscript provides an excellent window into the key morphogenetic events of human caudal neural tube formation. Figures 1 and 2 provide beautiful images and quantification of the developmental events, enabling comparison to models that are currently in use, including model organisms and the developing spinal organoid field. The characterization of somite development and later regression is particularly important.

In Figures 3 and 4, the authors use immunohistochemistry to examine the cellular death mechanisms and spatiotemporal organization of tissue regression within the tail. They demonstrate a proximal to distal tapering of the overall tail and neural tube areas that is not present for the notochord and reveal a proximal to distal degeneration of the tailgut, similar to what is observed in rodents. The identification of caspase-dependent cell death within the human tail provides an explanation for the mechanism of this regression, especially given the notable lack of presence of any gross necrosis.

Next, the authors have addressed current questions regarding the molecular pathways present during elongation of the embryo and later regression of the tail structure. The in situ hybridization experiments in Figure 5 show important evidence for a maintained neuromesodermal progenitor pool of stem cells that promote axial elongation. Additionally, the authors have conducted serial transverse sections of the tail to better understand the formation of the secondary neural tube in humans. They found a rodent-like formation involving a singular rosette caudally at the tailbud tip, and that multiple lumens, if present, were located more rostrally. This clearly differs from chick secondary neurulation. Finally, as mentioned above, the non-trivial collection and review of the existing human secondary neural tube and body formation literature is an important tool that organizes and synthesizes ~ 100 years of observations from precious human samples.

Weaknesses:

(1) The non-pathologic presence of multiple polarizations in human tails compared to the rodent pathogenic counterpart is interesting given that rodents obviously maintain this appendage that is lost in humans. A clear mechanism for how the secondary tube becomes continuous with the primary tube and how this relates to the presence of multiple polarizations in humans remains elusive.

Reviewer #2 (Public review):

This study utilizes an extensive series of neurulation human embryos to address several open questions about the similarities and differences between human primary and secondary neurulation in the tail. Results are compared to other model systems, such as the chicken and rodent. Histology, in situ hybridization, and apoptosis analysis provide molecular data about how the tail regresses in the human embryo. The number of embryos utilized for the analysis and the quality of the histological analysis provide robustness to the findings.

Comments on revised version:

The authors have meticulously addressed all the concerns raised by the reviewers, using new data and modifications to the text to further strengthen the quality of the manuscript.

This is a fabulous manuscript. I have nothing more scientifically to critique.

eLife Assessment

This article uncovers a new important role of STAMBPL1 in promoting angiogenesis in triple-negative breast cancer (TNBC) and elucidates the specific mechanisms by which it activates the GRHL3/HIF1α/VEGFA axis through interaction with FOXO1. The finding that STAMBPL1 mediates GRHL3 transcription through the interaction with FOXO1 is novel. These experimental results corroborate each other, forming a solid foundation of evidence that supports the main findings of the article.