Reviewer #1 (Public review):

The manuscript by Yang, Wang, and Cléry presents a pipeline for real-time identification of common marmosets in a laboratory setting. Models were trained and evaluated on data derived from a family of three closely related adults and a set of juvenile twins. Freely moving animals entered an enclosed space fixed to the housing cage door, which permitted the entry of individual animals for data acquisition. Utilizing YOLOv8-nano, identification was improved through the introduction of uniquely colored collar beads. Analyses of facial similarity showed close morphological relatedness amongst individuals and highlighted the need for highly discriminative classification. The authors demonstrate that combining facial detection with visual markers enables adequate identity assignment under controlled laboratory conditions with minimal cross-individual misclassification.

The main strengths are that the proposed pipeline offers a solution for real-time identity tracking in common marmosets. Its lightweight design enables deployment across a wide range of hardware configurations. Furthermore, if similar strategies are employed, this methodology is likely adaptable for other species with minimal modification. Additionally, evaluation of closely related individuals provides a necessary stress test for the discrimination of facial identity tracking. However, the main weakness is the pipeline's reliance on controlled animal isolation and small visual markers, which raises questions about the approach's generalizability to unconstrained multi-animal environments. The authors justify the use of beads, but the dependency of facial recognition on the beads needs to be described more clearly, as it is unclear how independent facial recognition performance truly was. The overall utility of this approach therefore remains to be seen.

Reviewer #2 (Public review):

Summary:

In this study, Yang et al. develop a real-time system for automatic face detection and identification of multiple unrestrained common marmosets in a home cage setting.

Strengths:

The study aims to address an unmet need in behavioral neuroscience: the ability to non-invasively identify animals is crucial to the automated and rigorous study of neural behaviors; this is especially true for common marmosets, which are rapidly becoming a model system of choice for the study of complex social cognition. By using a YOLOv8 backbone, the study achieves human level performance, both in terms of precision and recall of the trained models.

Weaknesses:

The robustness of the system is not clear from the limited datasets presented.

Comments on revised version.

The authors have adequately addressed my comments from the previous round, and I have no further comments

Reviewer #3 (Public review):

Summary:

In the revised manuscript, the authors provide additional details and evidence regarding the robustness and utility of their method.

Strengths:

(1) The authors provide a very precise automatic identification of marmosets in their home cage, to levels comparable to animal health professional.

(2) This method is robust across lightning, camera angles etc but importantly is able to identify marmosets in naturalistic conditions, which can be of tremendous value to neuroscientists and to ecological or behavioral studies.

(3) Easy to use and implement, requiring minimal settings. Phone videos can even be used.

Weaknesses:

While the manuscript improved tremendously from the previous version, given the nature of the paper, it is still a strenuous read.

Comments on revised version.

The authors did a good job of addressing my previous concerns and I don't have more comments.

Reviewer #1 (Public review):

Summary:

Kaku and Flenniken investigate the mechanistic pathways through which specific viral infections alter the flight capabilities of honeybees. Building on their previous discovery that DWV impairs flight while SBV unexpectedly enhances it, the authors hypothesized that these behavioral shifts are driven by interactions with the insect's octopamine (OA) signaling pathway, which is responsible for the "fight-or-flight" neurohormonal stress response and energy mobilization. To test this, the authors experimentally infected adult honeybees with DWV or SBV and pharmacologically manipulated the OA pathway using either octopamine supplementation or epinastine (EP), an OA-receptor antagonist. They then evaluated the bees' flight performance (distance, duration, and speed) on custom flight mills and profiled their gene expression using qPCR and RNA sequencing.

Strengths:

A major strength of this study Is the high prevalence of preexisting background DWV and SBV infections in the honeybee cohorts, which meant there were no completely "virus-free" control groups. However, the authors successfully mitigated this limitation by rigorously quantifying viral RNA copies for every individual bee via qPCR and utilizing these viral abundances as continuous variables in powerful linear mixed-effect models.

Weaknesses:

The primary weakness lies in the methodology used for targeted pharmacological manipulations, as well as the lack of OA quantification across different treatments. Thus, their claims are not sufficiently supported by the current data.

Comments on revised version.

I appreciate the authors' efforts to address the reviewers' concerns and to revise the wording of the manuscript. The revised version is more cautious than the original, and some of the discussion has been appropriately toned down. However, I remain unconvinced that the key mechanistic conclusions are sufficiently supported by the current evidence.

(1) The specificity of epinastine remains insufficiently demonstrated.<br /> The authors argue that AmOARβ2 is the predominantly expressed octopamine receptor subtype in their RNA-seq dataset and therefore the physiological effects of epinastine are most likely mediated through this receptor. However, I do not find this argument fully convincing.

First, relatively low transcript abundance of other OA receptor subtypes does not exclude their physiological contribution. Even receptors expressed at lower levels may play important functional roles, particularly in specific neuronal populations or flight-related tissues. Therefore, the possibility that epinastine affects multiple OA receptor subtypes cannot be excluded.

Second, although epinastine is widely used as a pharmacological tool to inhibit octopamine signaling, its receptor pharmacology has not been comprehensively characterized. The study by Roeder et al. primarily employed radioligand binding assays, which provide information on receptor affinity but not on functional antagonism or subtype selectivity. Without systematic functional characterization across the insect octopamine receptor family, it remains difficult to exclude contributions from other OA receptor subtypes or potential off-target effects.

A more convincing pharmacological strategy would be to demonstrate similar results using an additional chemically distinct octopamine receptor antagonist. Concordant phenotypes obtained with two independent antagonists would substantially strengthen the conclusion and reduce concerns regarding off-target effects.

(2) The OA supplementation experiments should be interpreted more cautiously.<br /> The authors correctly acknowledge that exogenous octopamine produces only transient elevations in signaling. However, I do not find the comparison with synthetic agonists entirely appropriate.

Although synthetic agonists such as amitraz generally produce more prolonged receptor activation than endogenous octopamine, the more fundamental difference lies in their physicochemical properties. Octopamine is a highly polar endogenous amine that exhibits limited tissue penetration and is rapidly cleared through uptake and metabolic pathways. Consequently, exogenously administered OA is unlikely to efficiently reach relevant target tissues or receptor populations in a manner comparable to endogenous neurotransmitter release. In contrast, the greater lipophilicity of amitraz facilitates its distribution into target organs and enables more sustained receptor engagement following systemic administration.

More importantly, the observation that OA supplementation partially rescues flight behavior does NOT necessarily establish that altered endogenous OA signaling is the primary mechanism underlying the virus-induced phenotypes. Such rescue experiments demonstrate that pharmacological enhancement of octopaminergic signaling can modulate the phenotype, but they do NOT provide direct evidence that endogenous OA levels or OA signaling are altered by viral infection. Therefore, these experiments should be interpreted as supportive rather than mechanistic evidence.

(3) Direct quantification of octopamine remains the major missing evidence.<br /> The authors acknowledge that direct measurements of octopamine and tyramine would strengthen their conclusions but argue that technical limitations and cost prevented these analyses. While these practical considerations are understandable, they do not compensate for the absence of the critical mechanistic evidence.

Overall, I appreciate the authors' revisions and agree that the manuscript provides interesting evidence that octopaminergic signaling is associated with virus-dependent changes in honeybee flight performance. However, I do not believe that the current data are sufficient to support the stronger mechanistic claims regarding regulation of the OA pathway or the specific involvement of the AmOARβ2 receptor.

Unless direct measurements of endogenous OA (and ideally tyramine) can be provided, I recommend that the authors substantially moderate the mechanistic conclusions throughout the manuscript, including the Abstract, Results, and Discussion. The study should be presented primarily as evidence for a pharmacological association with octopaminergic signaling rather than as definitive proof of the proposed mechanistic model.

Reviewer #1 (Public review):

The authors have considered a panel of antibodies that target epitopes at the gp120/gp41 interface (8ANC195 and PGT151), the fusion peptide in the gp41 domain (VRC34), and the MPER region of gp41 (DH511.2_K3 and VRC42). They also investigate 10E8.4/iMab, which is an engineered bispecific antibody that targets the MPER and the CD4 receptor. On a technical note, they have applied a double amber codon-readthrough strategy to incorporate the non-natural TCO*A amino acid, which gets labeled through click chemistry. This approach should result in less disruption of the native Env structure as compared to the peptide insertion previously used for smFRET imaging of Env. Furthermore, previous implementations of smFRET imaging of HIV-1 Env, which focus on gp120 conformation, have yielded limited information on antibodies that target gp41. Altogether, through the cutting-edge application of smFRET imaging, the study provides novel insights into the mechanisms of action of interesting and clinically relevant antibodies.

Comments on revised version:

The authors have nicely responded to all of my concerns. I have no further issues.

Reviewer #2 (Public review):

Summary:

In this paper, Xu and co-workers unveil two distinct modes of neutralisation by gp41-targeted broadly neutralizing antibodies on HIV-1 Env. So far, it was unclear as to how the mechanism of neutralisation occurred for this subset of neutralising antibodies (that can target the fusion peptide or the membrane proximal external region of the gp41 subunit). Thanks to single-molecule FRET, the authors show that the majority of broadly neutralizing antibodies stabilize the closed Env conformation (named State 1 since the original work by Munro and colleagues PMID: 25298114). Interestingly, the bivalent 10E8.4/iMab stabilized in turn a CD4-bound open state of Env. The two modes of neutralization described for these antibodies show previously unknown allosteric mechanisms that stabilize closed and open Env conformation, stressing the importance of Env conformational dynamics and its efficiency during the process of fusion.

Strengths:

The article is well-written, and the figures fully depict the data in a convincing way. The authors have used smFRET, which is now established in the field as a good tool to assess Env dynamics.

Comments on revised version:

I am very happy with the comments, answers and the way the new manuscript is shaped after revision. I have no further questions or concerns.

Reviewer #1 (Public review):

Summary:

Fujita and colleagues investigated two selective peripheral nerve voltage-gated sodium channel inhibitors targeting either Nav1.7 or Nav1.8 on excitability of human dorsal root ganglion neurons. The authors discovered that Nav1.8 inhibition is more effective at suppressing repetitive firing of DRG neurons and this may explain the greater clinical efficacy observed for suzetrigine.

Strengths:

The study is interesting and the findings are conceptually satisfying in that they may explain one aspect of Nav1.7 vs Nav1.8 targeting success.

Weaknesses:

(1) The use of postmortem human DRG neurons provides translational relevance, but the use of these cells is also a liability given their high degree of variability. Of note are the 10 to 20-fold differences in baseline properties among cells, which dwarfs the effects of the test compounds. The experiments may suffer from under sampling.

Comments on revised version.

The revised manuscript addresses my prior concern with reasonable effort given the limitations of human postmortem DRGs.

Reviewer #3 (Public review):

Summary:

In this manuscript, Fujita/Jo/Stewart/Osorno et al., investigate the contribution of Nav1.7 in regulating the excitability and firing properties of human dorsal root ganglion (hDRG) neurons in vitro. The authors characterize the effects of a previously reported Nav1.7-selective blocker AM-2099 in recombinant human Nav1.7 channels and in cultured hDRG neurons from postmortem organ donors. The authors observed modest changes in many of the properties expected by inhibiting Nav channels, including decreased action potential upstroke rate and amplitude, while increasing the voltage and current thresholds for spike generation. However, AM-2099 did not change the maximum number of APs in response to suprathreshold stimulation, leading the authors to conclude that Nav1.7 inhibition alone has limited efficacy in reducing the firing properties of hDRG neurons at the soma, and discuss that the effects of Nav inhibition may be different at distal axons.

Strengths:

Experiments are well-designed and executed, and the results presented are convincing. The focus on voltage-gated sodium channels in native human DRG neurons is highly relevant to recent efforts to develop safer analgesic options for chronic pain in people.

Comments on revised version.

The authors have done an excellent job addressing my prior critiques.

Reviewer #2 (Public review):

The paper by Freas and Wystrach is an interesting computational study, exploring the detailed mechanisms of how simple neural circuits could explain complex behavioral patterns observed in navigating ants. The authors compare detailed, high speed video recordings of Australian desert ants (Melophorus bagoti) with predictions made by their new computational model and find convincing similarities between the model and the behavioral data, at a level of detail not previously studied. Particularly interesting are emerging properties of the model, yielding behavioral motifs it was not designed to reproduce, but which occur in natural ant behavior.

A strength of the study is that the model is based on previous models, without making major novel assumptions. It combines existing models of the insect central complex with a model of the lateral accessory lobe and adds a stochastic inhibition of forward velocity to the interaction of central complex and lateral accessory lobes. In essence, the central complex provides corrective steering signals when the goal direction and the current heading of the insect are not aligned, while the lateral accessory lobes provide an intrinsic oscillator underlying the behavioral oscillations shown by walking ants at all times. These background oscillations are modulated by the steering signals from the central complex. Depending on which phase of the intrinsic oscillations coincides with the corrective signals, and how fast the ant is moving forward during this time, a complex set of behaviors emerges.

Most prominently, scanning behaviors, which are regularly carried out by the ants, are recapitulated in great detail by the model. Additionally, other behaviors, such as full loops, emerge naturally from the model. While computational models are not to be seen as definite evidence for any biological reality, they can provide strong support for particular neural implementations. The current study is an excellent example in that it provides evidence for a serial arrangement of central complex circuits upstream of the lateral accessory lobe circuits, modulated by speed regulating input. While the latter is hypothetical, it yields a clear hypothesis that can be validated by connectomics studies and functional work in the future.

The computational model is explained in detail and information about all model parameters is provided in an accessible way. The approach is thus transparent and reproducible, leaving it to the readers to assess the assumptions made in the model and how the studied complex behaviors emerge. This also provides the possibility to combine this new model with existing models to expand the scope and to more comprehensively capture the behavioral repertoire of ants, and insects in general.

Importantly, the study shows that even complex behavioral motifs do not require dedicated neural modules, but can rather emerge from the interplay of already known circuits - highlighting the efficiency of insect brains and possibly providing the path towards embodied hardware solutions of such circuits in autonomous agents.

Reviewer #1 (Public review):

The authors sought to determine how Rif1 contributes to DNA replication timing (RT), transcriptional regulation, and embryonic development using zebrafish. They generated a maternal-zygotic rif1 knockout line and examined developmental phenotypes, genome-wide replication timing profiles, RNA-seq, and nascent transcription (SLAM-seq) during early embryogenesis.

Their major findings in this manuscript are

(1) Rif1 is not essential for zebrafish viability, unlike its partially essential role in mice.

(2) Rif1 deficiency causes defects in female sex determination, delayed epiboly, and reduced primitive erythropoiesis.

(3) Genome-wide RT is altered by Rif1, but developmental stage has a much larger influence than Rif1 itself.

(4) Rif1 is required for the proper maturation ("sharpening") of the RT program during development rather than for specific developmental RT switches.

(5) Rif1 has a much stronger effect on transcription during zygotic genome activation (ZGA) than on replication timing at these early stages.

(6) Loss of Rif1 leads to increased expression of early zygotic genes, indicating that Rif1 normally suppresses widespread transcription during ZGA.

Overall, the work proposes that Rif1 independently regulates replication timing and transcription, with these two functions becoming most prominent at different developmental stages.

The major strengths of the manuscript are as follows.

(1) the study combines multiple genome-wide approaches including whole-genome RT profiling, RNA-seq, SLAM-seq in combination with gene KO and developmental analyses.

(2) One of the strongest points is that the authors conducted the analyses at multiple developmental stages rather than a single point.

(3) The most important conclusion is that the Rif1 regulates transcription during development in a manner largely independent of its RT function, which was further strengthened by the additional data provided in the revised manuscript.

On the other hand, the weakness of the manuscript includes the followings.

(1) Limited mechanistic insight. The questions such as where Rif1 binds on the chromatin (in relation to the transcriptional promoters/ enhancers and replication origins).

(2) Which functional domains of RIf1 are involved in regulation of transcription and replication (Is PP1 recruitment required for transcription regulation?) are not addressed.

(3) Since Rif1 is known to be involved in chromatin organization/ nuclear architecture regulation, the studies addressing this (Hi-C, compartment analyses, ATAC seq etc) would provide important mechanistic information.

(4) Female sex determination phenotype is intriguing, but it remains largely descriptive, and its mechanisms are elusive at the moment.

Overall, the results support the authors' conclusions and they have successfully provided answers to the authors' original questions on developmental roles of Rif1 in RT and transcription in vertebrate.

Comments on revised version:

The authors responded to my comments in a largely satisfactory manner. They have conducted additional analyses and concluded that Rif1 regulates transcription during ZGA largely independently of its classical RT function, which is an important finding.

Although authors did not examine origin firing and replication fork rate in rif1 KO cells, which I suggested in my original review, this can be saved for their future studies.

I think the revised manuscript has been improved and provides important basic information on the functions of the conserved Rif1 protein in RT and transcriptional regulation.

I have no further recommendation for additional experiments or data analyses.

Reviewer #2 (Public review):

This study by Masser et al. analyzes global replication timing and gene expression in rif-1 null zebrafish. This work is an extension of their previous report of the normal replication timing pattern during wild-type zebrafish development. The major valuable finding here is that Rif1 is not essential for viability in zebrafish, and - counter to expectation from studies in cultured cells and other species - late replication does not strongly depend on Rif1. Instead, the data suggest that Rif1 subtly sharpens replication timing pattern during normal development rather than function generally to delay replication timing. In the absence of Rif1, the normal pattern establishment is somewhat delayed. The authors also document some changes in expression during development with more genes being repressed by Rif1 than activated at some early stages.

The study and analysis are generally rigorous, and the conclusions are supported by convincing data. Given the strong link between replication timing and cell type/development, studying timing in a whole developing organism is important. The experimental approach is technically challenging, particularly the bioinformatic analysis. The scientific advance here is largely confined to documenting the timing of Rif1-affected transcription, the unanticipated effect of the rif1 deletion on replication timing and on sex determination, though the latter is not explored. The difference in timing of the transcription phenotypes and replication phenotypes suggests they may be very distinct Rif1 roles. The overall study a useful set of findings and detailed data for future work.

Loss of Rif1 did not affect viability, but it did strongly influence sex determination, resulting in a lower population of females. This effect is the strongest organismal phenotype, but the study provides no mechanistic explanation for the loss of females from the data gathered here.

Comments on revised version:

We are generally satisfied with the revised version of this manuscript.

Reviewer #1 (Public review):

This interesting paper addresses the phenomenon of potentiation in single-cell habituation in Stentor coeruleus. This is an important "hallmark" of habituation that helps to establish single-cell learning as being similar to habituation in animals. Prior studies from Wood, as well as our own results, have shown that potentiation occurs in Stentor, but I have always remained a little bit skeptical that this effect was possibly just due to incomplete recovery after the first trial. When I first read this paper and saw the habituation curves for the first and second trials, such as in Figure 5, I thought, yes, that is definitely what is happening, and so is this really potentiation?

The authors were also clearly aware of this issue and, notably, they embraced it head-on by developing an analysis that allows potentiation effects to be detected even despite failure of the cell to fully recover after the first trial. The key is their "phase portrait" that allows the learning process to be depicted as a curve capturing how learning rates and response probability evolve over time, thus allowing the curves to be compared between trials. If my interpretation was correct that so-called potentiation was just incomplete recovery, the prediction would be that the curves for two successive trials would overlap, with the first trial curve extending beyond the second one towards higher response probabilities, which would be lost in the second trial due to failure to recover fully. But the data clearly are not consistent with that idea. I think that this result is very strong and important.

Especially nice is the approach of Figure 7C, which uses a vertical shift in the phase portrait as an indicator of potentiation. I did, however, find Figure 6 a little hard to digest at first, and I have a few suggestions about that. First, I think it would be a good idea to explicitly say which curve is the first trial and which is the second. Second, I think it would help readers if the authors could start with a cartoon that explains visually what the curves mean. For example, show a habituation curve, indicate how the slope is calculated at different parts of the curve, and then show how the slope versus response are plotted to make the phase portrait. It is all spelled out in the text, but it would help a lot of readers to see it visually, I think.

One question I have about Figure 6 is that it looks like the specific case of ITI 1 hour ISI 2 min has some kind of pathological behavior in the second trial, despite not seeing any indication of any 'weirdness' in Figure 5. I gather that this is meant to be due at least in part to the incomplete recovery seen after the first trial, but then I don't see why this would not also be an issue for ITI 1 hour ISI 3 min. I would not require the authors to explain every anomaly, but this one stands out, and I feel it could be telling us something interesting.

Reviewer #2 (Public review):

Summary:

The authors address habituation and potentiation in the single-celled organism Stentor in a large data set by systematically varying stimulus frequency and recovery duration. They analyze habituation dynamics on the level of single cells within a Bayesian inference framework to map out how the response probability of individual cells decays during training. Mapping out the progression of habituation quantified by learning rate versus decaying response probability, they observe different dynamics for different stimulus frequencies and recovery durations, which they reconcile with multiple time-scales governing the memory of prior training.

Strengths:

The authors accumulate a systematic, broad data set of Stentor habituation and potentiation, which, in combination with the Bayesian framework they developed, unfolds its power to probe underlying habituation dynamics and challenge theoretical frameworks.

Weaknesses:

The interlacing of theoretical framework, existing concepts and expectation, and experimental data in their narrative may challenge readers. The Bayesian inference of habituations is very successful in concluding that their variation with stimulus frequency and recovery duration points to multiple time scales of memory are involved. However, the authors' comprehensive analysis of potentiation may need more guidance to follow the authors' conclusions.

The combination of a dynamical systems-driven hypothesis, experimental data, and statistical analysis, as put forward in this work, is immensely powerful for uncovering the mechanisms that facilitate learning, such as habituation and potentiation, in single-celled organisms.

Reviewer #1 (Public Review):

The paper itself has a reasonable aim, to compare the inputs to the hippocampus from cortical regions across mammals. But for some reason, the conclusions that are reached are very limited. We know for example that the main laboratory rodents investigated, rats and mice, are nocturnal, live in underground tunnels, and have a very wide field of view with no fovea. In contrast, primates have a highly developed cortical system for vision and a fovea, and so have very different capabilities to rodents, as they have an ability to identify people or objects at a distance, and to remember where they have been seen. Despite this major difference in the visual cortical processing in these different mammals, somehow important points are missed in this paper about how the cortical processing is organised in these different mammals, and how this is reflected in the anatomy.

Reviewer #2 (Public Review):

Summary:

The manuscript emphasizes a phylogenetic conservation of the hippocampal region and primary sensory cortical regions in mammalian species. The authors then propose that the evident species-specific differences in behavior and memory-related functions may be due to differences in type and amount of cortico-hippocampal connectivity.

Strengths:

The authors are well-established researchers with a long history of excellent results and publications. The question (co-influence of cortical and hippocampal connections) is potentially interesting.

Weaknesses:

The treatment is very broad and macro scale, ignoring the likelihood that hippocampal-cortical connectivity and behavioral outcomes result from multiple differences at a more micro-scale. The designated "mammalian" sample is also broad. Thus, it can appear incomplete as a sample, and incompletely discussed.

Reviewer #1 (Public review):

Summary:

These authors used a binocular rivalry task with flickering stimuli in which subjects had to report the color of the target grating at the end of each trial. Target or distractor cues provided information about the orientation of the respective stimulus prior to each trial. The stated goals of this project include testing the neural mechanisms underlying strategic target and distractor processing. Behavioral enhancement was observed for target cueing, while no cost was noted for distractor cueing. These authors present evidence for reactive suppression, characterized by pronounced frontal theta activity that reduced the sensory gain (SSVEP) of the distractor. Distractor cues also increased alpha activity over parietal areas, which these authors link to attentional gating while pointing out no relationship with sensory gain.

Strengths:

This manuscript clearly reflects thoughtful analysis of the available data. Alongside a simple and effective task design, sophisticated methods provide good support for most of the claims made by these authors.

Weaknesses:

Lack of temporal precision for SSVEP effects. I would like to see how sensory gain is/isn't dynamically modulated in the moments after the initial ERP to see if there could be differences compared to the broader window used presently (1.3 to 3.1 seconds).

These authors indicate that persistence of the neural representation of cued distractor orientations into the rivalry period is evidence against a "search-and-destroy" type mechanism where distractors are enhanced to then be suppressed reactively. This claim relies on an indirect link between the maintenance of information about distractor orientation (i.e., successful orientation decoding) and the processing of sensory representations. This claim would be backed up more substantially if the SSVEP (a measure of sensory processing) could reveal temporal dynamics on a finer scale.

Reviewer #2 (Public review):

Summary:

The findings are conceptually useful - a sequential alpha-then-theta architecture for proactive gating and reactive distractor suppression would be a compelling contribution to the attention control literature - but the evidence is incomplete at best. The central dissociation rests on an inadequate proxy for perceptual dominance, the key alpha-behavior effect is small (d = 0.199) and confined to a single unprotected data quadrant, and the GLMM uses an inadequate random effects structure that inflates false-positive risk.

Strengths:

The SSVEP frequency-tagging + binocular rivalry combination is genuinely inventive for isolating sensory gain signals from the two competing stimuli simultaneously. The finding that distractor cueing enhances sensory processing of the distractor yet fails to impair behavior is a clean result that directly addresses a behavioral paradox in the attentional suppression literature. The non-phase-locked TF analysis and the use of RESS for SSVER extraction are methodologically sound.

Weaknesses:

The most consequential flaw in the paper is the operationalization of "perceptual dominance." The authors explicitly acknowledge in a footnote that trial categorization as "target-dominant" or "distractor-dominant" is based on which eye received the stimulus, not on participants' actual perceptual reports. Because participants were never asked to report which stimulus was dominant (only to reproduce the target's color), the assignment is an anatomical proxy, not a perceptual measure. This matters enormously for the paper's central claims. Specifically: (a) The entire two-mechanism dissociation (theta for target-dominant trials, alpha for distractor-dominant trials) is built on a trial-type categorization that may not reflect subjective perceptual experience on a given trial, and (b) Dominant-eye stimuli do typically win initial rivalry dominance, but dominance alternates, and in a 2-second window (the stimulus duration used), perceptual states likely fluctuate in many trials. The lack of button-press perceptual tracking (e.g., continuous dominance reports) means the authors cannot verify that their neural effects actually correspond to the perceptual states they claim. This is a major structural limitation of the design that can't be retroactively corrected, and it significantly weakens the consciousness/awareness framing of the findings.

Another significant issue is that the parietal alpha effect on behavior is confined to a very specific quadrant of the data: distractor-dominant trials where both target and distractor SSVERs are weak simultaneously. The authors present this as an elegant result - "alpha helps most under high perceptual uncertainty" - but it could equally reflect insufficient statistical power for effects in the other three SSVER-strength cells (target strong/distractor weak; target weak/distractor strong; both strong). The Cohen's d for the alpha effect on target reporting probability is only d = 0.199, which is a very small effect. With N=36 and no correction for the multiple SSVER-strength subgroupings tested, there is a real risk that this specific cell-finding is a false positive, while the null in adjacent cells reflects inadequate power rather than a genuine boundary condition.

A third major limitation is that with a design that includes 6 fixed effects and all their interactions, the random effects structure should include random slopes for at least the key predictors (cueing condition, dominance). Fitting maximal random effects models or justified reduced structures (Barr et al., 2013) is standard in within-subjects EEG research. Using only random intercepts risks inflating Type I error rates for the interaction terms that form the core of the paper's claims. The authors provide a supplementary table (Table S1) but do not describe whether model convergence was verified or alternative random effects structures were tested.

Fourth, the paper's title and central claim are that alpha and theta dynamics operate sequentially. However, the temporal ordering (preparatory alpha -> rivalry-phase theta) is primarily shown by examining each oscillation in its respective analysis window, not by a single analysis testing whether the sequence itself predicts behavior better than either mechanism alone. A path analysis or cross-lagged model linking trial-level alpha to subsequent theta, and both to behavior, would directly substantiate the "relay" framing. Without this, the sequential architecture is more of an interpretation than a demonstrated property.

Finally, the frontal theta cluster identified by permutation testing spans 3 to 16 Hz - a range that extends well into the alpha band. Calling this a "theta" effect while simultaneously discussing alpha as a separate mechanism is difficult to reconcile. At minimum, this frequency boundary issue warrants explicit discussion.

Reviewer #3 (Public review):

Summary:

Interest was especially focused on how foreknowledge of the orientation of either the target or the distractor could be used to resolve the competition between these stimuli and properly report the target color. The target or distractor was pre-cued by a solid or dashed orientation cue. They were displayed with slightly different presentation frequencies, which allowed for examining their sensory processing with steady-state visual evoked responses (SSVERs). Furthermore, orientation decoding was performed, which revealed that orientation cues selectively affected processing after stimulus onset related to the dominant but not the non-dominant eye. EEG analyses additionally focused on parietal alpha activity and frontal theta, both during the anticipatory phase and the stimulus-processing phase. Cueing the distractor vs. the target induced increased right parietal alpha power during the anticipatory phase, but this did not result in direct inhibition of distractor features. During the stimulus-processing phase, cueing the distractor resulted in increased theta activity. Finally, a generalized linear model was employed wherein trial-by-trial behavior (precision in target color report) was predicted by target and distractor SSVERs, type of pre-cued stimulus (target/distractor), preparatory parietal alpha power, stimulus processing-related frontal theta power, eye dominance, and all their interactions. Performance in the case of reduced sensory processing of the target (based on SSVER) showed more deviations when sensory processing of the distractor was high, but no such effect was observed when sensory processing of the target was high. The latter effects were modulated by eye dominance and cue. Increased theta reduced distractor sensory processing but not target sensory processing. Increased alpha was only beneficial when sensory evidence for both target and distractor was low. Results were interpreted as favoring sensory gating before stimulus onset, reflected by increased parietal alpha (i.e., pro-active control), while theta activity especially seemed relevant to suppress distractor activity (i.e., reactive control) thereby favoring target-related performance.

Strengths:

The authors convincingly show that EEG can provide crucial information about how the human brain deals with the conflict between a target and distractor in a binocular rivalry paradigm with pre-cues signaling either the target or the distractor orientation. An important aspect of the study is the focus on precision of target color report, in combination with the possibility to assess SSVERs to the target and distractor. The strength of this study may actually also be its weakness; the question is whether the presented ideas on proactive and reactive mechanisms can be generalized to paradigms that do not employ binocular rivalry. Separation of target and distractor processing by selectively presenting them to the left/right eye increases the conflict when the target is presented at the non-dominant eye, but what happens in the absence of binocular rivalry concerning the target-distractor conflict?

Weaknesses:

An important aspect of the study relates to the cue manipulation. In many studies, cues are often informative but not mandatory. Couldn't one argue that in this study task performance crucially depends on cue processing, as without the cue, it becomes difficult to tell apart the target from the distractor. It could be argued that participants are able to do this based on the slight difference in flickering frequency, but I doubt whether this is possible at all. However, if this were the case, then they might use this as an alternative cue and ignore the orientation cue. What do participants experience while performing this task? As the cue can be considered to be mandatory, the question may be raised what strategy the participants actually employed. If the target was cued, they simply may have prepared for this orienting and could ignore the distractor. However, if the distractor was cued, they could use two strategies: search for the stimulus without the cued orientation, or first detect the distractor, and then orient towards the other stimulus. The ideas and results on parietal alpha and frontal theta in combination with the other findings are certainly very interesting, but recently, it has also been argued that frontal theta may be more related to action control (e.g., see Panek et al., https://doi.org/10.1093/cercor/bhaf276) and also pro-active control (Cooper et al., 2017). So, it might be that increased theta reflects suppression of the response related to the distractor, which feeds back on its sensory processing. This raises the question whether there is possibly also some evidence on functional connectivity between frontal and posterior regions that varies depending on the precise condition. Are the results also shining a new light on the relation between attentional orienting and eye dominance (e.g., see Schintu et al., 2020)?

Joint Public Review:

Summary:

Inferring so-called "functional connectivity" between neurons or groups of neurons is important both for validating models and for inferring brain state, including in human patients. This study aims to enhance this inference process by using closed-loop perturbation-based approaches. To this end, the authors develop a framework based on linear dynamical models that minimizes the estimation error. Based on this framework, the authors provide a practical guide for applying it in realistic experiments. Modalities include non-invasive ones, such as fMRI, iEEG, and invasive ones, such as optogenetic perturbations combined with neuropixel probes or calcium imaging.

Strengths:

A main strength of this paper is the application and adaptation of an explicit error expression to system dynamics estimation from evoked neural responses, bringing a useful theoretical tool into computational neuroscience for, as far as we know, the first time. Importantly, while the analytical derivation assumes the neural dynamics is linear and the control signal is known, these assumptions do not appear to be essential: their method outperforms passive observation even when the true dynamics is nonlinear or the control input is not known perfectly. Moreover, the relative simplicity of the method makes its practical applications straightforward, as the authors illustrate in the context of brain state classification and neural control.

Besides being of practical importance, simply pointing out that passive observation can lead to large mis-estimation of functional connectivity should serve as a wakeup call to anybody engaged in this endeavor.

Weaknesses:

None.

Reviewer #1 (Public review):

Lohse et al. describe an open-source system for laser scanning photostimulation (LSPS) in head-fixed animals. Although similar systems have been developed and used by different groups, Zapit provides an open-source solution requiring few custom parts and minimal coding. This tool can clearly facilitate and speed the adoption of LSPS, particularly for the increasingly used purpose of mapping the effects of focal cortical silencing during behavior. Other potential uses include mapping optogenetically evoked movements and selectively activating genetically labeled neuronal subtypes of interest in the cortex. The design is well thought through, and the presentation is mostly clear and well written.

In general, the more modular such a system is, the better, in terms of compatibility with existing hardware and software that potential users may already have purchased - laser, galvo, and camera in particular. The system has struck a reasonable balance between allowing modularity and providing an integrated complete package, but even more flexibility would be welcome for potential users looking to cut costs, as would clearer presentation of such flexibility as already exists.

Comments and suggestions are mostly minor, as follows.

(1) Command signals:

How is the relationship between analog voltage commands and laser power determined? Is this assumed (or required) to be linear (as Figure 7F implies)? Usability and modularity would be improved by an option to measure or provide a calibration curve for systems with a nonlinear mapping between command voltage and laser power.

For the grid calibration step, how is the initial mapping from galvo voltage commands to image position determined? Presumably, some sort of initial guess or calculation based on the hardware specifications is needed for the grid calibration to be feasible. Also, how are the number of grid lines and the distance between them determined?

Why is the mapping between analog outputs and hardware (galvos, laser, masking light) fixed? This would be trivial to make configurable and allow labs with existing setups to adopt Zapit without rewiring existing hardware.

(2) Laser and optics:

In Figure 1, the authors should consider explaining the scanning principle schematically, i.e., depicting how tilting of the scan mirrors translates via the scan lens into beam displacement in the specimen plane. Perhaps Zemax can be used for accurate rendering.

Since the unexpanded beam greatly under-fills the back aperture of the lens, the z resolution is presumably terrible - which is good! That is, for the purposes of LSPS, this advantageously avoids focus-dependent effects, which might otherwise arise due to (e.g.) skull curvature. The authors should consider pointing this out, as well as providing an estimate of the z resolution.

What is the working distance?

Reviewer #2 (Public review):

Summary:

In this work, Lohse and colleagues develop a system for doing targeted photostimulation in mouse cortex. The system uses a camera image to target laser stimulation to stereotactically defined locations in mouse dorsal cortex.

Strengths:

The hardware is well designed, and the software is well documented and supported. The build guide and well-documented software package should allow for simple implementation of the technology. Without a doubt, this is a valuable community resource for the circuit neuroscience field.

Weaknesses:

No weaknesses were identified by this reviewer.

Reviewer #3 (Public review):

Zappit is an open-source implementation of arbitrary-access laser-scanning optogenetics for manipulation of neuronal activity in mice. As the method requires expertise ranging from optics, hardware control and programming, the authors make the point that this powerful strategy is underutilized in the field, and put forward a well-documented modular hardware and software platform aligned to the Allen Mouse Brain Atlas aimed at enabling the larger scientific community to use this approach (democratizing) for controlling cortical activity during behavior in mice.

The authors favor a galvanometric approach to laser targeting. The system is inexpensive, easy to build, well-documented and user friendly (Matlab based GUI and GitHub repository). The photo-stimulation laser is directed into an X-Y galvo scanner targeted to the specimen using a dichroic mirror and focused on the sample using a Plössl lens as scan lens which is also used as an objective. The scan lens/objective images the specimen onto a camera via tube lens (also a Plössl lens) in a 0.5X magnification ensuring to fit the extent of the mouse brain onto the camera sensor (USB-3 Basler acA120-40um).

The authors report short and reproducible onsite time (~ 0.5 ms) and block (mask) the stimulation source using the laser analog control (~0.5 ms). The system is reliable, aiming at up to 20 stimulation sites per sequence considered as quasi-simultaneous (10 ms). They minimize rebound by gentle ramping down of stimulation over 250 ms.

The system is fast to calibrate by mapping scanner positions to pixel space in the camera space and mapping stereotaxic coordinate onto the image of the exposed skull. The theoretical x-y PSF is 70 µm (measured ~90µm) while the authors make the point that due to scattering the photo-stimulation spot size (lateral extent) is about 1 mm in diameter. This is what they also observe in electrophysiological recordings using silicon probes. The effective radius of inactivation depends on laser power, but was about 1 mm for laser powers (1-2-4 mW) on which the authors observed significant behavioral perturbations - in several tasks: 1) a delayed response somatosensory discrimination, 2) a visual detection task assessing changes in temporal frequency of a drifting visual stimulus; and 3) a visual discrimination (International Brain Laboratory task) in which mice were tasked to report the location of visual stimuli by turning a wheel. As proof of principle, the authors used a photo-stimulation set composed of 52 bilateral sites positioned at 0.5 mm interval covering a large network of frontal, motor and somatosensory cortical areas. Indeed, photo-inhibition of frontal motor cortex sites produced robust increases in reaction time. In contrast, stimulation at other motor and somatosensory sites produced modest, but significant decreases in reaction times.

While the approach is not novel, it does serve the need of better disseminating this technique in the research community. Overall, the Zappit is well-documented and easy to build and use, and will have impact in increasing robust use of site directed photo-stimulation (exciting/inhibiting ensembles of neurons at particular ~1 mm size regions of interests across the dorsal surface of the brain). The authors also note that the axial resolution is ~1.5 mm.

Concerns & comments:

(1) While the authors argue that it offers the best utility to affordability trade-off - faster than motorized drivers and require much less power than DMDs (100X) and less expensive/easier to use compared to SLMs, in the current form, the manuscript does not clearly list the limitations of the approach. At such, in my opinion, the authors should include side by side comparisons (perhaps as a table). For example, clear statements should be included with respect to comparisons in lateral (x-y), axial (z) spatial resolution, as well as temporal sequential aspect of Zappit and other photo-stimulation techniques involving DMDs or SLMs.

(2) Is power really a limitation in terms of the laser sources? Or is this a disadvantage mainly because using less power has beneficial effects on the tissue health? It may be useful to provide metrics of comparisons along these lines between Zappit and DMD-based approaches.

(3) Arbitrary-scanning vs random scanning may be more appropriate to describe to strategy.

Reviewer #1 (Public review):

Sensory hair cells of the inner ear convert mechanical sound vibrations into electrical signals through mechano-electrical transduction (MET). While the protein components of the MET machinery have been studied extensively, much less is known about how the surrounding membrane lipid environment contributes to hair cell function. The recent discovery that TMC1 and TMC2 also function as lipid scramblases has brought renewed attention to the importance of membrane lipid asymmetry and the mechanisms that maintain it in sensory hair cells.

In this study, the authors identify the P4-ATPase ATP8B1 and its partner TMEM30B as key regulators of membrane lipid asymmetry in outer hair cells. Using complementary genetic models, HA-tagged knock-in mice, localization analyses, and functional experiments, they show that ATP8B1-TMEM30B is enriched in stereocilia and the apical membrane of outer hair cells and is required to maintain phosphatidylserine asymmetry, support hair cell survival, and preserve normal hearing. The parallels between the ATP8B1/TMEM30B loss-of-function phenotypes and TMC1 deafness-associated mutants with constitutive scrambling support a model in which ATP8B1-TMEM30B flippase activity maintains membrane lipid asymmetry and homeostasis, whereas constitutive TMC1-mediated phospholipid scrambling disrupts this balance and contributes to membrane instability.

The authors have addressed the points raised during the initial review thoroughly. The revised manuscript includes clearer methodological details, additional physiological characterization, improved presentation and quantification of several datasets, and a more balanced interpretation of the localization and mechanistic findings. These changes improve both the clarity and rigor of the study while leaving its main conclusions unchanged.

As with any study that opens a new area of investigation, important mechanistic questions remain. In particular, it will be interesting to determine how disruption of membrane lipid asymmetry ultimately impairs MET function and triggers hair cell degeneration, how flippase and scramblase activities are coordinated in vivo, and how these pathways are integrated with the broader molecular machinery underlying mechanotransduction. These questions highlight the exciting directions that this study opens for the field.

Overall, this work provides evidence that ATP8B1-TMEM30B is a critical regulator of stereocilia membrane lipid asymmetry and represents an important contribution to our understanding of membrane homeostasis in auditory hair cells. I have no further major concerns and support publication.

Reviewer #2 (Public review):

Summary:

Prior work identified TMEM30B (knockout mice) as well as ATP8B1 (human genetics and mouse model), ATP8A2 (knockout mice), and ATP811A (human genetics) as relevant for hearing. The authors also reasoned that given the recent discovery of TMC1 and TMC2's dual function as mechanotransduction channels of the inner ear and as lipid scramblases, a counterpart flippase should be in the sensory hair-cell stereocilia bundle where mechanotransduction happens. They use CRISPR/CAS to modify the endogenous mouse genes and add an HA tag at the N-terminus of the ATP8B1, ATP8A1, ATP8A2, and ATP11A proteins. Their experiments with these mice unambiguously localized ATP8B1 at the base of outer hair cell stereocilia bundles. Knockout of ATP8B1 results in loss of outer hair cells, deficient auditory function (ABR), and degeneration of outer hair cell stereocilia bundles. Similarly, hair cells from genetically modified mice with endogenous HA-tagged TMEM30B proteins show localization of this protein to outer hair cell stereocilia bundles. TMEM30B knock out mice phenocopy the ATP8B1 knock out model. Interestingly, the authors show that annexing V staining precedes hair cell loss in ATP8B1 and TMEM30B knockout mice and that proper localization of these proteins is lost in mice that lack CIB2, a protein essential for hair cell mechanotransduction.

Strengths:

(1) Use of knock-in HA-tagged proteins to unambiguously localize ATP8B1 and TMEM30B

(2) Systematic characterization of auditory function (ABR), hair cell loss, and hair-cell stereocilia bundle morphology.

(3) Advances our understanding of the role played by lipid homeostasis in auditory function.

(4) Reports on mouse models that will be helpful to further understand the mechanistic role played by ATP8B1 and TMEM30B in normal hearing and hereditary deafness.

Weaknesses:

(1) Are the HA tags causing any functional issues? Function and localization of tagged proteins can sometimes be compromised. This is checked for TMEM30B and ATP8B1, but not for ATP8A1, ATP8A2, and ATP11A.

(2) Following on the point above, is it possible that ATP8B1-HA is well localized, but localization for the other three flippases (ATP8A1-HA, ATP8A2-HA, and ATP11A-HA) is compromised by the tag? Is this potential miss-localization causing any functional phenotypes? I find surprising that there are flippases only in outer hair cells and only formed by ATP8B1. A possible explanation is that the tag is interfering with trafficking. If so, there should be a phenotype (ABRs), although this might be masked by redundancy among these flippases or caused by systemic issues (admittedly difficult to sort out).

Reviewer #1 (Public review):

Summary:

This is a study utilizing several types of analyses (computational modeling, neuronal cultures, rodent epilepsy model, and human intracranial multi-scale recordings) to address a highly relevant conceptual question: Are fast ripples (FRs) distinct pathological entities or largely emergent products of stochastic spike clustering? The results can potentially reshape current approaches to incorporating fast ripples into the epilepsy surgery evaluation.

Strengths:

The conceptualization of fast ripples as potentially arising by chance is highly novel and builds effectively on questions raised in prior studies that have never been satisfactorily resolved. Integration across biological scales and models provides a rigorous approach, now improved by addressing theoretical concerns regarding validity of the shuffling approach and state dependence. The discussion has been updated to provide a more nuanced interpretation of the study's findings.

Weaknesses:

The authors have satisfactorily and thoughtfully addressed the critiques provided in the first review. However, there remain two points that I would like authors to address:

(1) Synchronized burst firing is a key feature of an epileptic site generating interictal discharges, and one that could generate either oscillatory or stochastic FRs as documented in multiple prior publications cited in the manuscript and/or in the prior review. Paroxysmal depolarization, for example, has been very well described, and consists of strong, disorganized burst firing (resulting in summated postsynaptic potentials strong enough to generate high gamma signal) in a neuronal population coinciding with a large low-frequency deflection. I would like to see the results described in this context, and to avoid blanket dismissal of stochastic FRs without a clear oscillatory component.

(2) It would be highly useful to add a conclusion paragraph that spells out implications of the study for use of FRs as epileptic biomarkers in clinical invasive EEG recordings.

Please address the above critiques in Discussion, or elsewhere as deemed necessary by the authors.

Reviewer #2 (Public review):

Summary:

This paper asks an important question that has not been discussed much in the extensive literature on the High Frequency Oscillations (HFOs) that have been extensively studied in patients with epilepsy and experimental models of epilepsy. The question is whether the Fast Ripples (FRs), the HFOs in the 250-500 Hz frequency band, represent a pathological phenomenon or represent a physiological phenomenon that occurs in the healthy brain but happens to be more frequent in epileptic tissue. It is an important question that has not been systematically addressed until now. The authors conclude, from very extensive simulations, from extensive experimental animal studies (the systemic kianate model of epilepsy in rats), and from a modest amount of human data, that FRs occur in healthy brains as a result of the chance occurrence of bursts of action potentials, and that in epileptic tissue, their frequency of occurrence is approximately 30% higher than what is expected by chance. They conclude that FRs are not a separate phenomenon of epileptic tissue. This finding is reinforced by the recent findings of FRs in experimental models of Alzheimer's disease.

Strengths:

This is a valuable study because it asks an important and original question and because it evaluates it from several angles (simulation, tissue culture, experimental animals, and human patients). The simulations and the analyses of real data are performed very carefully and with original and solidly documented approaches, using extensive simulations and extensive data sets in the cultured cell data and in the in vivo experiments. The paper is clearly written and well-illustrated.

Comments on revised version.

The authors have appropriately addressed the questions I raised in the first review.

Reviewer #3 (Public review):

Summary:

An outstanding question in the field of high frequency oscillations (HFOs) in the context of epilepsy is how these oscillations emerge, considering that they occur at such high frequencies i.e., 250Hz well above the firing ability of single neurons. One hypothesis that has been suggested in the past is that neurons that fire in an out of phase fashion or rather at random intervals may contribute to a spectrum of HFOs ranging from 250-500Hz that observed in epilepsy. However, how possible it is that random action potentials could aggregate to the extent that they could give rise to HFOs in the so-called fast ripple (FRs) frequency range (>200 according to the authors) remains unclear. To test this hypothesis, they used computational modeling to randomly insert action potentials in a signal, and they found that this approach is sufficient to generate FRs. Some of the predictors of whether FRs could occur were neuronal count, firing rate and synchronization. Besides computational modeling, they used different model systems to test whether that would be possible to be observed in neuronal cultures, in epileptic rats (intrahippocampal kainic acid model), and human data. Neuronal cultures treated with picrotoxin did not show evidence that FRs could be generated more than chance aggregation of action potentials. They then asked whether synchronization and firing rate could play a role in the emergence of FRs. They found that changes in neural firing and synchronization, such as those occurring during differences phase of the sleep-wake cycle could affect the number of FRs occurring by chance aggregation, with more FRs seen during periods of wakefulness, a result that they replicated in human data.

The authors largely achieve their proposed aims of demonstrating that random neuronal firing can, in principle, generate FRs. Results from this study could influence current thinking around mechanisms generating FRs in epilepsy. The use of different computational approaches and model systems could offer new analytical methodologies for the study of FRs in the context of brain disease.

Strengths:

(1) The authors used a multi-level approach combining computational modeling with experimental datasets, including neuronal cultures, a rat model of temporal lobe epilepsy and human data.

(2) Identification of key parameters such as neuronal count, firing rate, synchronization and brain state in observed incidence of FRs generated through random aggregation of neural firing.

(3) Cross-species validation increases the likelihood of generalizability of the findings.

Minor weakness:

(1)The analyses conducted in human data lack direct comparison with sleep data due to no available data, but would encourage future investigations directly comparing HFOs during wakefulness and nocturnal sleep.

Comments on revised version.

The authors have addressed my comments and I have no further suggestions.

Reviewer #1 (Public review):

Summary:

This manuscript by Ghosh and colleagues investigates the transcriptional changes within the oligodendrocyte lineage that contribute to age-related declines in oligodendrocyte differentiation and myelination. Combining bulk RNA-Seq on acutely purified oligodendrocyte lineage cells with bioinformatic approaches, the authors identify groups of genes that show different patterns of dynamic regulation during differentiation (which they term "switch" genes, or "switches"). A subset of these switch genes are differentially regulated with age. The authors identify two transcription factors, Bcl11a and Foxm1 that are downregulated during differentiation, have predicted binding site enrichment at other switch genes and are downregulated in aged OPCs. Functionally testing Bcl11a, the authors show that Bcl11a knockdown inhibits the differentiation of young OPCs in culture, whereas overexpression promotes differentiation of aged OPCs. Viral expression of Bcl11a in Sox10 expressing cells accelerates the formation of Plp1+ oligodendrocytes in aged rodents following lysolecithin induced demyelination.

Strengths:

The work is clearly presented and addresses an important biological problem. The bioinformatic approaches used in the manuscript are powerful, and the identification of Bcl11a as a modulator of oligodendrocyte differentiation is a novel finding. The combined in vitro and in vivo approaches to assess the function of Bcl11a in oligodendrocyte differentiation are a substantial strength of the work.

Comment on revised version.

In the revised version the authors now provide analysis of expression of stage-specific markers for OPCs, preOls and OLs in their isolated cells. It is slightly concerning that the OPC markers show higher expression in the isolated preOLs than in the isolated OPCs, but the authors do provide some discussion on this point in the supplementary text.

Reviewer #2 (Public review):

Ageing poses a significant challenge to the regenerative capacity of oligodendrocyte precursor cells (OPCs). Myelin abnormalities accumulate with age, while the ability of OPCs to differentiate into myelinating oligodendrocytes progressively declines. This likely contributes to inefficient replacement of damaged myelin and oligodendrocytes, impaired remyelination following injury, and reduced adaptive myelination. Identifying the molecular changes associated with this decline is therefore important for understanding and potentially treating age-related deterioration of CNS white matter.

This study sought to identify transcriptional regulators involved in oligodendrocyte-lineage progression whose expression is altered in aged OPCs. The authors developed gSWITCH, a computational tool that identifies genes showing defined dynamic expression patterns across ordered biological states. By combining this analysis with comparisons of young and aged OPC transcriptomes and transcription-factor-binding-site enrichment, they identified Bcl11a as a candidate regulator. Bcl11a transcripts are abundant in young OPCs, decline during oligodendrocyte differentiation, and are markedly reduced in aged OPCs.

A major strength of the study is its combination of computational candidate identification with functional experiments. Bcl11a knockdown substantially impaired the differentiation of young OPCs without measurably affecting their proliferation. Conversely, transient Bcl11a overexpression increased the differentiation of aged OPCs in vitro. Oligodendrocyte-lineage-specific expression of Bcl11a in aged mice also increased the generation of PLP1-positive oligodendrocytes following focal demyelinating injury. Together, these complementary loss- and gain-of-function experiments support the conclusion that Bcl11a expression is functionally important for OPC differentiation and that restoring its expression can improve the differentiation competence of aged OPCs.

While the transcription-factor-binding-site enrichment analysis predicts a Bcl11a-regulated network, the current study does not establish direct binding or identify the downstream genes responsible for its effect on OPC differentiation. Similarly, the upstream mechanisms responsible for the age-associated reduction in Bcl11a expression were not investigated. Further work may help establish a more complete mechanistic framework explaining how restoration of Bcl11a expression improves OPC differentiation.

Overall, this study offers valuable insights into the age-related loss of regenerative capacity in the central nervous system and introduces a computational framework that may be broadly useful for investigating dynamic gene regulation in other biological contexts.

Comments on revised version.

The authors have addressed my previous comments, and the revised manuscript has been substantially strengthened by the inclusion of additional supporting data and an expanded discussion.

Reviewer #1 (Public review):

Summary:

This work provides a comprehensive analysis of how adult zebrafish show fear responses to conspecific alarm substances (CAS) and retain their associative memory. It shows that freezing is a more reliable measure of fear response and memory compared to evasive swimming, and that the reactivity and the type of responses depend on the zebrafish strain. It further suggests neuronal substrates of different fear responses based on c-Fos mapping.

Strengths:

The behavioral part is the most comprehensive and detailed yet in the zebrafish field, providing strong support for the authors' claim. The flow from Figure 1 to Figure 4 is very smooth. They provide extremely detailed, yet complementary and necessary, analyses of how different categories of behavior emerge over time during the CAS exposure and memory retrieval. I'm convinced that neuro researchers who study fear/stress responses will always refer to this paper to plan and interpret their future experiments.

Comments on revised version:

The authors successfully addressed my comments, including the addition of Figure S6-2, which gives us some intuition into the relationships between c-Fos levels in individual areas and the behavioral outputs.

Reviewer #2 (Public review):

In this study, Fontana et al. develop a paradigm for associative conditioning by pairing exposure to alarm substance with a novel tank. Exposure to conspecific alarm substance (CAS) in the novel tank triggers freezing and what they characterize as evasive swimming behaviour, which are subsequently seen in a re-exposure to the novel tank without the CAS present. Importantly, these states are identified via automated processes including postural tracking and a random forest classification process, which could be very useful tools for subsequent studies.

In their experiments they focus on the differences in behaviour among strains of zebrafish (both males and females), and among individual zebrafish. For males and females of different strains they find some differences, though the clearest message seems to be that the most robust measure of the behaviour in response to both the CAS and in the memory trials is the freezing behaviour, while evasive behaviour is more variable and not always seen. This may relate to their observation of significant "evasiveness" in vehicle control experiments (discussed further below).

Moving on to individual variation from within this multi-strain male/female dataset, they first examine transition matrices between states, and find this is not dramatically altered by stimulus exposure. They then use clustering to identify 4 different "classes" of zebrafish that differ in their expression (or not) of two types of behaviour: freezing and/or evasive behaviour. They show that over the three exposure epochs of the experiment this classification is somewhat stable in an individual fish, though many fish change their behaviour -- e.g. evading + freezing -> only freezing.

In the final set of experiments they move beyond behavioural analyses and perform whole-brain cFos mapping of these individual zebrafish, and perform analyses aimed at identifying correlations between individual behavioural expression and the number of cFos positive cells in different brain regions. Using partial least squares analysis they find areas associated with two types of behavioural contrasts, which differ in their weighting of different behavioural expression during the Memory trials. Covariation and network structure analysis within different classes of fish also find some differences in covariation among brain areas, providing hypotheses as to underlying network effects that may govern the expression of freezing and/or evasive behavior in the memory trial phases.

Overall, I find this to be an interesting study that employs state of the art methods of behavioural analyses and whole-brain cFos analyses. The revision has clarified the take-home message considerably: the abstract is now more careful about which behavioural groups are memory-associated, and the causal language in the conclusions has been appropriately softened. Two of my three original main concerns have been addressed. The first is not and having looked at the data again I can now be more specific about what concerns me.

Comments on revised version.

(1) My first concern related to the claim that fear memory behaviour falls into four distinct groups, and specifically to the role of evasiveness in defining them. The authors give three reasons for retaining it, but I remain unconvinced.

The first is that variable evasion in response to alarm substance is a long-standing observation (von Frisch; Suboski et al.), and that dissecting this individual variation is the purpose of the paper. I agree with the motivation, and it is a good reason to measure evasion. But it does not establish that evasion on memory day reflects fear memory, and memory day is the only day used for the clustering and neural activity mapping. The manuscript's own results point the other way: relative to pre-exposure, no strain or sex increased evasion on memory day, and relative to vehicle only female TUs did. The temporal profiles show evasion on memory day to be largely similar between vehicle and CAS-treated fish. Historical observations of variable evasion during CAS exposure do not carry over to the memory phase.

The second is that the clustering itself reveals two kinds of freezing fish - one freezing between bouts of normal swimming, the other between bouts of evasion - demonstrating that a subset of fish increase evasion. In absolute terms, this does not match the data. In Figure 4B, evading freezers are below the population mean for absolute evasion, as are freezers. The text describes evading freezers as "high in freezing and evasive behaviors," and I do not think Figure 4B supports this.

What actually separates the two freezing groups is the third measure, evasion as a percentage of active time. And this is where I have difficulty, because that measure is not an independent behavioural readout. The classifier assigns every window to normal, evasive or freezing, and active time is simply non-freezing time, so evasion-as-percent-of-active is fully determined once the other two are known.

This matters for the clustering specifically. Distance-based methods weight each input dimension equally, so a variable that carries no information beyond the other two nonetheless contributes a full third of the distance between any two fish - and it contributes it in a way that counts freezing twice, once directly and once through the denominator of the derived measure. The space is nonetheless described as three-dimensional throughout, including in the Methods and the Figure 4 legend, when there are only two independent behaviours in it.

The consequences fall hardest on exactly the animals at issue. Both freezing groups sit at 65-70% freezing, so there is very little active time to divide by, and small absolute differences in evasion - together with any noise in estimating them from a couple of minutes of non-frozen behaviour - are inflated into large differences on the rescaled measure. In terms of what the fish actually did, the two groups differ by a few percent of trial time. That is the boundary on which much of the rest of the paper rests.

I recognise that evasion as a proportion of active time is in some respects the more biologically meaningful quantity, and the authors are right that a fish freezing 70% of the time has limited opportunity to do anything else. But that is an argument for reporting it as a descriptive measure, not for entering it into the clustering alongside the two variables from which it is computed.

This impression is reinforced by Figure 4A itself. While the freezer group occupies a reasonably distinct region, the non-reactive, evader and evading freezer groups appear as a single continuous distribution with cluster boundaries drawn through it rather than around visible gaps. I appreciate that UMAP is a projection and that visual separation is not required for genuine structure, but this is the figure by which most readers will judge whether four discrete types exist, and it does not obviously support that reading - particularly given that the embedding is built from the same variables, including the rescaled measure, that most favour the separation.

I would suggest that the authors re-run the clustering using only the two directly measured behaviours, percent freezing and percent evasion of total time, and report whether four groups still emerge and, in particular, whether the evading freezer / freezer split survives.

The third is that the two groups have distinct functional networks despite equally high freezing, so the behavioural difference is real and is manifesting in the brain. This is the strongest of the three arguments, and I accept part of it: something about how a frozen fish spends its remaining active time does appear to be neurally meaningful, which is interesting in its own right. But it does not establish that these are two distinct types, nor that the difference has anything to do with the conditioning. Fish taken from either side of a cut through a continuous distribution will differ neurally if that continuum tracks brain state, so the network result is equally compatible with graded variation. More importantly, Figure 5A shows that a substantial proportion of fish are classified as evaders in the vehicle condition and at pre-exposure, before any CAS has been given. This suggests a pre-existing individual tendency toward evasive behaviour that is independent of the alarm substance, and one would expect such a tendency to persist into the memory trial. If so, the distinction the network analysis is drawing between freezers and evading freezers may simply reflect that baseline trait, and its neural correlates would be correlates of the trait rather than of fear memory. I am therefore not convinced that this distinction is related to CAS or to memory.

(2) This concern is fully resolved. I had misread the CAS preparation: it was pooled from eight donors spanning all four strains and both sexes, so every fish received identical material and the strain and sex differences cannot be attributed to donor variability. The clarification now added to the Results will prevent other readers making the same error. The addition of FDR correction to the Figure 2 comparisons also addresses my related concern about multiple testing.

(3) Somewhat resolved. The conclusion no longer states that behavioural variation is "driven by" activity in particular regions, and the added caveat that neural activity was not directly manipulated sets the right expectation for a mapping study. The scatterplots in Figure S6-2 are a useful addition and give a much better intuition for what the PLS contrasts represent. My remaining reservation is the one above: a great deal of the neural story rests on the evading freezer / freezer contrast, and I am not persuaded that this contrast marks a boundary relevant to fear memory.

Reviewer #3 (Public review):

This revised manuscript by Fontana et al. aims to study how animals respond to fearful stimuli, with a specific focus on brain regions involved in predicting animals that passively freeze or those that actively evade the threat. I continue to be enthusiastic about the study. The study addresses an important question regarding individual variation in fear-related behavior and links these behavioral phenotypes to whole-brain activity patterns in adult zebrafish. The combination of a contextual fear conditioning paradigm, strain/sex comparisons, behavioral clustering, and AZBA-based c-Fos mapping makes this a valuable contribution to the field, not just in answering the question posed by the authors, but also in formulating a framework for using adult zebrafish for whole brain analysis of complex behaviors. Overall, I find the authors have responded to my concerns:

(1) I still think that separating memory acquisition and consolidation is an interesting question, and further use of the framework will need to eventually solve that; however, I also appreciate that this may be beyond the scope of the current study, and I appreciate the authors acknowledging this in the manuscript.

(2) Regarding Figure 3, I also agree that this is difficult to present differently, and I appreciate the authors adding text to the body to clarify things. My one request is that the sentence (lines 214-215) that reads: "This increase in evasion in the vehicle group likely represents a response to the water disturbance that occurs when solution is added to the tank." Be changed to: "This increase in evasion in the vehicle group may represent a response to the water disturbance that occurs when solution is added to the tank." While it is entirely possible, there are no concrete data to support that this is "likely."

(3) I appreciate the clarification regarding the PLS-derived contrasts in Figure 6A and in the body.

Overall, this is a really interesting paper that will have a wide-ranging impact. All of my concerns have been addressed.

Reviewer #1 (Public review):

Summary:

The authors set out to evaluate whether AGES, a recently developed auxin/TIR1-based conditional GAL4 expression system, is a suitable tool for Drosophila ageing research. They characterise induction efficiency across sex, transgene insertion site, auxin dose and age, then test whether AGES can replicate a well-established pro-longevity manipulation (dominant-negative insulin receptor expression).

Strengths:

The study is thorough and methodical. The authors use appropriate genetic controls throughout, which is required to properly interpret AGES-based experiments. They identify an important issue, in that activation of the AGES machinery itself (independent of any UAS-transgene) shortens lifespan and alters protein levels, while high-dose auxin independently affects body mass, and even a moderate dose (5 mM) impairs stress resistance across all genotypes. These findings are important for researchers when interpreting their experiments. The tissue and age mapping of induction efficiency (brain, fat body, gut) is also useful, and the inclusion of driver-only positive controls at each age (Figure 2) establishes that da-GAL4 activity itself is stable across the ages tested, ruling out declining driver activity as an explanation for the reduced induction seen in older flies (though, as noted below, reduced auxin ingestion with age remains a very plausible contributing factor alongside declining AGES efficacy).

Weaknesses:

Longevity and stress assays were conducted only in females, which, combined with the finding that males show weaker and less consistent induction, means the study cannot speak to whether the metabolic and survival costs of auxin/AGES activation observed here also apply to, or differ in, males. The KCl vehicle control matches the potassium cation (K⁺) content of K-NAA across conditions; therefore, chloride (Cl⁻) concentration differs between control and auxin-fed media (both minor weaknesses).

Achievement of aims and impact:

The authors achieve their stated aim. Rather than validating AGES as unambiguously suitable for longevity work, they set out to characterise its behaviour and limitations in this context, which they do convincingly. The data support their overall conclusion that AGES can be used to conditionally induce transgene expression at advanced ages, but that its use in longevity/healthspan studies requires caution and rigorous control genotypes. This is a useful contribution with direct practical value: it will help other researchers make informed decisions about whether and how to deploy AGES in ageing-related work, and the cautionary findings regarding auxin/AGES toxicity are likely to be of broad relevance to the growing community of AGES users beyond the ageing field specifically.

Reviewer #2 (Public review):

McGilvary et al. evaluate the recently developed auxin-based gene expression system (AGES) for use in aging studies of Drosophila melanogaster. This system is based on the widely used Gal4/UAS system that enables cell-specific expression of UAS-transgenes under Gal4 activator control. AGES uses an auxin-inducible degron-tagged Gal80 repressor that should prevent Gal4-dependent activation unless flies are fed auxin, providing a useful approach for temporal control of transgene induction - something that would be highly useful for aging studies. The authors perform a comprehensive analysis of AGES-dependent transgene induction in male and female flies at different ages with multiple controls, demonstrating some moderate induction in female flies only - albeit with some substantial background induction even in the absence of auxin.

Overall, transgene induction appears to be both much lower with the AGES system compared to Gal4 driver controls and very leaky, with some tissue-specific differences in induction observed as well. Combined with their observations that auxin feeding has impacts on body mass, triacylglycerol and protein levels, and lifespan, these data raise some concerns regarding the interpretation of data obtained using the AGES system for aging or longevity studies in flies. This study provides well-needed validation for the recently developed AGES system and highlights critical caveats that will support future studies.

Most conclusions of the paper are well supported by data, but additional controls and textual edits would strengthen and clarify the findings. In addition, the abstract and conclusions of this study should more accurately reflect the limitations of transgene induction using this AGES system in adult flies.

Reviewer #3 (Public review):

Summary:

In this useful work, the authors characterize the auxin-based gene expression system (AGES) as a tool for studying ageing. They found that this system can be applied to ageing studies. In addition, they identified important drawbacks of the methods, including effects of insertion sites and sex on induction of the system, and that some auxin doses may have inadvertent effects on body mass and physiology. Overall, the study extends the AGES system for use in fly ageing studies and highlights some caveats. While the findings are solid pointers, more extensive characterization is needed to benchmark the extent of the caveats identified.

Strengths:

The study provides the first longitudinal evaluation of the AGES system's induction efficiency across the entire Drosophila lifespan. The authors also highlighted a number of caveats of the AGES system in ageing animals. These are all important points to be considered when using this system, and findings should be interpreted keeping these caveats in mind.

Weaknesses:

There were inconsistencies with auxin dosages between the figures.

(1) The authors used a higher dose of auxin (20mM) compared to the original AGES paper (McClure, 2022) in Fig 3. The auxin dose-dependent effects are not linear for TAG and protein levels, highlighting that the genotype-dependent effects may be highly variable and may yield quite different results in other studies.

(2) The results in Figure 4 showing the lack of induction in the brain are quite interesting; however, only 5mM auxin is tested. Characterizing dose-dependence for the variability of the induction in tissue types would be useful. At the very least, recapitulating prior results at 10mM should be done.

(3) Food intake and hydration status were not measured alongside body mass, TAG, and protein endpoints. The changes seen could be an effect of decreased feeding or fluid balance rather than metabolic reprogramming.

Reviewer #1 (Public review):

This study investigates the role of a specific neuronal population in the lateral septum (LS) in balancing exploratory and defensive behaviors. The authors created a mouse model (cKO) lacking Nkx2.1-lineage neurons in the LS by deleting the Prdm16 gene. They discovered that this ablation specifically eliminated Crhr2-expressing neurons, which are normally targeted by urocortin-3 (UCN-3) inputs. Behaviorally, cKO mice did not show general changes in anxiety but displayed a significantly increased exploratory drive. In a predator odor test (using TMT), cKO mice spent more time investigating the aversive stimulus compared to controls, suggesting these LS neurons normally suppress exploration during threat. Furthermore, the study found that Nkx2.1-lineage neurons in the LS are specifically activated by acute stress (body restraint), as shown by an increased number of c-Fos-positive neurons. While the loss of these neurons caused some connectivity and electrophysiological changes, the remaining Nkx2.1-lineage neurons were more excitable. Therefore, the authors demonstrate that LS Nkx2.1-lineage/Crhr2+ neurons are a distinct population crucial for calibrating behavioral responses to stress, acting to inhibit exploration in favor of defensive strategies.

This work provides new insights into the neural circuitry underlying anxiety and threat avoidance. However, some of the methods and data analyses require revision for greater clarity, and additional experiments and analyses are needed to further substantiate the conclusions.

Some of my specific questions and concerns are as follows:

(1) The authors showed a reduction in the size of LS and a specific decrease in Crhr2+ neurons in cKO mice. I would suggest examining whether the density of other types of neurons (e.g., Crhr1+ cells or other known cell types in LS) was altered in the cKO mice.

(2) For the single-cell sequencing experiment (Figure 2), it is unclear whether tissues from the 3 male and 3 female mice within each genotype were pooled together or processed individually (i.e., as 6 separate samples). This information is not clearly stated in the manuscript. Given that male and female mice exhibited behavioral differences, it would be valuable to examine sex-dependent effects in the analysis shown in Figure 2.

(3) Previous studies have shown that LS neurons exhibit distinct firing patterns, including regular spiking, bursting, complex-bursting, and phasic spiking. Since the authors recorded from both tdTomato-positive and -negative LS cells, it would be interesting to determine whether the positive cells display a unique firing pattern, thereby representing a distinct electrophysiological cell type within the LS.

(4) More detailed descriptions of the electrophysiological data analysis should be provided in the Methods section. Some LS neurons display spontaneous firing without current injection; therefore, it should be clarified how the resting membrane potential was measured in these cells. The amplitude and onset latency of the first spike are presented in the figures; however, it is unclear how the first spike was selected-whether from spiking responses to rheobase current or to a specific current pulse. I would suggest defining the first spike based on responses at a certain firing frequency. The method used to determine the spike voltage threshold should also be specified.

(5) Could the authors analyze the single-cell sequencing data to examine whether changes in ion channel expression might explain the observed alterations in spike waveforms?

Reviewer #2 (Public review):

Summary:

The manuscript "Selective loss of Nkx2.1-lineage neurons in the lateral septum alters the balance between novelty seeking and threat avoidance" is an interesting study by Miguel Turrero García and colleagues. Here, the authors report a novel mouse model allowing complete ablation of neurons pertaining to the Nkx2.1 lineage by conditionally ablating the transcriptional regulator Prdm16 from the Nkx2.1 lineage. The authors combined single-nucleus RNA sequencing, histological and electrophysiological approaches, as well as behavioral analyses to demonstrate that a large portion of LS neurons are profoundly altered by Prdm16 deletion from the Nkx2.1 lineage. This manipulation preferentially impacts Crhr2-expressing neurons, leading to electrophysiological defects. At the behavioral level, this cell population is preferentially recruited in a stressful situation, and ablation of Prdm16 from this lineage leads to enhanced exploratory behavior even in the presence of a perceived threat.

Strengths:

The strengths of this manuscript are (i) leveraging a transcriptional regulator within a specific cell lineage and restricted to an early stage of ontogeny (ii) disrupting the developmental trajectory of a discrete neuronal population identified with an elegant snRNAseq approach and (iii) without obvious compensation (iv) and its impact on behavior in adult mice, with a special emphasis on exploratory drive in the presence of an acute stressor. The manuscript is well written; the experiments are well conducted, organized, and presented in a logical framework. Statistical analyses are well described. Each experimental group includes a sufficient number of subjects, allowing robust statistical comparisons.

Overall, I very much enjoyed this manuscript and the elegant mouse model bridging developmental biology with systems neuroscience. Insights generated from this line of work could illuminate how discrete perturbations in gene expression programs at early stages of ontogeny could have a profound impact on the development, organization, and function of select neural circuits and how they may impinge on behavior at later stages of ontogeny.

Weaknesses:

Some comments and suggestions:

(1) General-

Photoinhibition of LS Crhr2-expressing neurons has no effect on anxiety-like behaviors in the absence of a stressor (Anthony et al., Cell, 2014). It would thus be interesting to reappraise the behavioral experiments performed with cKO mice in response to an acute stressor. The authors duly acknowledge this important point in the discussion section.

(2) Specific-

(a) Figure 2J: Was the increase in Crhr2 expression observed in tdTom-cells from cKO mice in the snRNAseq as well? If so, was Crhr2 expression enhanced in a specific cluster that did not belong to the Nkx2.1 lineage, or was it randomly enhanced across distributed clusters?

(b) Figure 3 and S3: Does the lack of UCN3+/ENK+ terminals reflect a downregulation of UCN3 and ENK, or does it reflect the absence of innervation? Restricting a retrograde viral vector in iLS that expresses a fluorophore to illuminate the UCN3+ cell bodies (and lack thereof) in PefAH of cKO mice could address this question.

(c) Figure 3: Does immunostaining for UCN3 in the PefAH area reveal cell bodies in cKO mice? In other words, is the loss of UCN3 terminal-specific or does it reflect a general downregulation of UCN3 in the PefAH?

(d) Figure 4: The remaining tdTomato+ neurons are more excitable in cKO mice. To what extent can alterations in the electrophysiological properties of tdTomato+ neurons lacking Prdm16 be related to their survival? Is it a general response to Prdm16 deletion that is unrelated to survival? Is it a compensation mechanism in surviving cells? Or, alternatively, is it a unique property of these specific cells that favored their survival despite Prdm16 deletion?

(e) Figure 5 and S5: Really nice figures. Great use of MoSeq with the predator odor test.

(f) Figure 6: Interesting that the decrease in cFos induction in NeuN+ cells of cKO mice is more prominently observed in LSd when tdTomato+ cells are prominently found in the LSi/LSv (Figure S2B). Could this be related to intra-septal connectivity?

(g) Figure 6: If tdTomato+ cells consist of 10-30% of neurons, and these tdTomato+ cells are preferentially found in the LSi/LSv, shouldn't we expect a decrease in c-Fos+NeuN+ density in the LSi/LSv (since there are generally fewer neurons in cKO mice)? If I am not mistaken, this could suggest that another unrelated LS population that is tdTom- displays an increase in cFos expression in cKO mice compared to WT mice. Could be interesting to see if the Crhr2+ neurons that are tdTom- are preferentially recruited in cKO mice as a compensation mechanism in LSi/LSv.

Reviewer #3 (Public review):

In the current work, Turrero Garcia et al. investigate the molecular, electrophysiological, and behavioral outcomes of conditionally knocking out cells from a unique developmentally-defined subpopulation in the lateral septum (LS). The authors focused on targeting cells from the Nkx2.1 developmental lineage that also expressed the transcriptional regulator Prdm16, which is uniquely upregulated in LS postmitotic neurons. They observed that this mutant line (Nkx2.1Cre;Prdm16fl/fl;Ai14; cKO) resulted in complete ablation of neurons from the Nkx2.1-lineage exclusively in the LS and not in the medial septum, making it an ideal model to study a lineage-defined subpopulation in the LS. They performed single-nucleus RNA-sequencing and uncovered four neuronal subtypes missing in the cKO. Furthermore, the authors validated this expression loss in the subtype expressing Crhr2 and observed a reduction of UNC-3 inputs (a neuropeptide with high affinity for Crhr2), highlighting additional disruptions in circuit connectivity. Loss of Prdm16 in Nkx2.1-lineage cells only resulted in mild electrophysiological changes in the LS. Finally, the authors performed a battery of behavioral assays to study anxiety-like behaviors and threat avoidance and observed increases in exploratory behaviors in some but not all assays. It should be noted that the cKO line also results in 30% loss of cortical interneurons, which could be contributing to the behavioral phenotype and not be exclusively due to LS loss. Overall, this manuscript takes a novel perspective by providing unique insights into how embryonic origin gives rise to mature molecular identity and distinct behaviors in adults. Therefore, it elegantly links developmental origin to mature molecular identity and function in the LS, an important question understudied in the field.

The conclusions of the work are overall supported by the data and limitations discussed, but some of the findings, in particular the histological and behavioral results, need to be extended.

Strengths:

(1) Utilizing developmental origin as a marker for mature neuronal identity and function is a valuable approach which remains under-utilized in the field and serves to provide a deeper understanding of how circuits are shaped to allow for appropriate behavioral responses.

(2) The authors perform a comprehensive analysis of the cKO mutant to determine the role of the developmentally defined Prdm16 in Nkx2.1-lineage cells at the molecular, cellular, electrophysiological, and behavioral level.

(3) The sn-RNAseq dataset in the LS of WT and cKO mice will be valuable to the neuroscience community.

(4) The authors perform an extensive array of behavioral paradigms investigating the balance between threat avoidance and exploratory behavior, performing all experiments in both male and female mice to determine whether the same developmental origin can lead to sex-specific differences.

Weaknesses:

(1) It remains unknown whether the reduction of UCN3 inputs to the LS is due to loss of the Nkx2.1-lineage in the LS itself or due to reductions in the number of UCN3 cells that provide innervation to the LS in the cKO (Figure 3). The authors speculate and include anecdotal observations that the perifornical region of the hypothalamus (PeFAH) provides inputs to the LS and could be the region driving the differences in UCN3 inputs in cKO. The authors should expand on this histological data and directly test whether the UCN3 inputs are indeed originating from PeFAH and whether the loss of Prdm16 in the Nkx2.1 lineage leads to a reduction in cell numbers in these inputs. These would disentangle the authors' claims on whether it is due to loss of Prdm16 in the Nkx2.1 lineage cells in the LS or whether it is due to loss of Nkx2.1 lineage neurons in upstream regions.

(2) The authors claim that there is an increase in exploratory drive in cKO mice, even though the dark-light test showed increases in time spent in the dark side for cKO mice in comparison to controls (Figure 5C). They discuss that this could be due to the mice being placed first in the light compartment of the chamber during the light hours, so they spend more time exploring the dark compartment of the chamber instead, which would be considered 'novel'. If this were the case, to make the results more solid, authors should place cKO mice in the dark compartment of the chamber during the dark hours and then record time spent in both chambers. If there was indeed an increase in exploratory drive in new environments, the authors should see increases in time spent in the light compartment.

(3) The result that cKO mice spend more time than controls exploring the inlet with TMT is of interest (Figure 5E, F). It needs to be highlighted that this is primarily the case in male mice and there is a trend in females. To confirm that the increases in exploration time of the inlet are not due to overall increases in general arousal, locomotion (e.g., velocity; pixels/frame) should be assessed in cKO vs controls.

(4) Statistical analysis correcting for repeated testing should be performed when running multiple t-tests in the same dataset, such as when analyzing histological results in Figure 1 and Figure 6 to increase confidence in the presented results.

(5) An important consideration that the authors address in the discussion is that the loss of Prdm16 in Nkx2.1-lineage cells is, for the most part, restricted to the LS, but other regions such as the cortex also show decreases in this population. Therefore, to strengthen the authors' conclusions that Nkx2.1-lineage neurons in the LS are indeed directly responsible for balancing threat avoidance and exploratory drive, targeted manipulation experiments, or at least additional c-Fos experiments assessing activity of Nkx2.1-derived cells in the LS will need to be performed in the future.

Reviewer #1 (Public review):

Summary:

Amadei et al investigate how excitation/inhibition balance in the prefrontal cortex plays a role in social behavior. To address this question, they developed a behavioral task where adult female mice can choose between a social reward (e.g., an adult male for sociosexual choice, or an adolescent female mouse) and a non-social reward (e.g., milk). They found that optogenetic inhibition of inhibitory neurons expressing oxytocin receptors (OXTR neurons) in the prefrontal cortex (PFC) reduces choice for sociosexual interaction compared to non-social reward and to a greater extent in sexually receptive females. They also found that this manipulation increases pyramidal neuron activity. Specifically, the authors identified a neuronal ensemble which represent the male option. Inhibition of OXTR disrupts the ability of the neuronal ensemble to represent the male option during decision-making in the behavioral task. Thus, using computational modeling, the authors proposed that OXTR neurons promote male choice by letting a male-representing pyramidal ensemble outcompete other pyramidal populations in the mPFC.

Strengths:

The study addresses an important topic in social behaviour and reward neuroscience with a focused hypothesis. The combination of behavioral testing and circuit manipulation combined with calcium imaging is a clear strength, and the work has the potential to make a solid contribution.

Weaknesses:

The main weaknesses are limited methodological clarity and details.

Reviewer #2 (Public review):

The authors aim to understand how inhibitory circuitry within the medial prefrontal cortex regulates the selection of sociosexual behaviour. Rather than studying social interaction in isolation, they develop an elegant behavioural paradigm in which female mice repeatedly choose between interacting with a male and obtaining an appetitive non-social reward. This task allows the authors to examine behavioural choice under conditions that more closely resemble natural decision-making. They combine optogenetic inhibition of oxytocin receptor-expressing interneurons, large-scale calcium imaging of pyramidal neurons, slice electrophysiology, and computational modelling to investigate how inhibition shapes cortical representations that ultimately bias behavioural choice.

The study has several notable strengths. The behavioural paradigm is novel and well-designed, allowing repeated choice measurements while controlling for general social motivation by including both male and juvenile female stimuli. The integration of multiple experimental approaches is particularly impressive. The behavioural effects of optogenetic inhibition are complemented by population imaging demonstrating elevated pyramidal activity, electrophysiological recordings confirming monosynaptic regulation of pyramidal neurons, and a computational model that provides a mechanistic interpretation of the observed circuit dynamics. The work therefore spans multiple levels of analysis, from synaptic interactions to behaviour, and the individual datasets are generally of high technical quality.

The imaging analyses identifying a putative "MALE" ensemble are particularly interesting. The observation that a relatively small subset of pyramidal neurons preferentially represents the male option before behavioural commitment provides an attractive framework for understanding how inhibition can stabilise specific behavioural representations. The temporal analysis suggesting that disruption of this representation precedes impaired behavioural choice is especially compelling, as it moves beyond simple correlations between neural activity and behaviour.

Several aspects of the mechanistic interpretation remain somewhat speculative. The central conclusion relies heavily on the computational competition model, which assumes an asymmetric competition between a relatively small male-selective ensemble and a much larger default pyramidal population. While the model successfully reproduces several experimental observations, many of its architectural assumptions are inferred rather than experimentally demonstrated. In particular, the designation of the remaining pyramidal neurons as a functional "OTHER" population representing the non-social alternative is not directly established experimentally. Alternative circuit architectures may be capable of producing similar behavioural and population-level effects, and the current data do not fully distinguish among these possibilities.

Similarly, although the identification of MALE cells is thoughtfully performed, the classification depends on an operational threshold derived from ROC analysis and correlated activity. It remains uncertain whether these neurons constitute a stable functional ensemble across sessions or merely reflect one end of a continuous representational spectrum. Longitudinal analyses examining the stability of these ensembles across days or across changes in behavioural state would strengthen the claim that they represent a dedicated neuronal population.

An additional limitation concerns the specificity of the behavioural interpretation. The reduction in male choice is interpreted primarily as impaired sociosexual decision-making. While the inclusion of juvenile female stimuli substantially improves the experimental design, it remains difficult to completely separate altered sociosexual motivation from broader changes in motivational salience, valuation, or action selection. The observed changes could reflect alterations in multiple components of the decision-making process, and this distinction deserves a somewhat more balanced discussion.

The interaction with the oestrous state is a very interesting aspect of the work and is consistent with previous studies of oxytocin-dependent sociosexual behaviour. However, this analysis is based on relatively modest numbers of animals and sessions, making it difficult to judge the robustness of these effects. The conclusions regarding hormonal modulation would therefore benefit from a more cautious interpretation.

Overall, the authors achieve their primary objective of demonstrating that oxytocin receptor-expressing interneuron-mediated inhibition contributes to the selection of sociosexual behaviour while regulating pyramidal population dynamics in the medial prefrontal cortex. The behavioural, imaging, and electrophysiological datasets provide convincing evidence that inhibition shapes cortical activity during decision-making. The computational model offers a plausible mechanistic framework linking these observations, although some aspects of this framework remain hypothetical and await further experimental testing.

The work is likely to have a significant impact on the fields of cortical circuit function, social neuroscience, and decision-making. Beyond its specific findings, the study introduces a behavioural paradigm that should prove broadly useful for investigating how competing behavioural options are represented within prefrontal circuits. The combination of behavioural neuroscience, population imaging, and computational modelling represents a valuable resource for the community and provides an important foundation for future studies examining how excitation-inhibition balance shapes flexible social behaviour.

Reviewer #3 (Public review):

Summary:

Using a combination of Miniscope imaging and optogenetic manipulation, Amadei et al. reveal how oxytocin receptor neurons in the prefrontal cortex of mice control pyramidal subpopulations and socio-sexual behavior. This work was planned and executed carefully and provides a novel and important angle to study the oxytocin system in the cortex. According to their results, oxytocin receptor neurons help discriminate between sexual and non-sexual stimuli, most likely by controlling different pyramidal subpopulations that are either most active during trials that include a sexual stimulus or that include non-sexual stimuli. I highly appreciate this article; however, I have one major concern related to the modeling part.

Strengths:

(1) Well-designed experiments.

(2) Rigorous analysis.

(3) Generates a new avenue to study socio-sexual decision making and creates a hypothesis about the connectivity of oxytocin-sensitive circuits.

Weaknesses:

(1) Major

In their last figure (Figure 4), the authors generated a computational model that, according to the authors, reveals a potential network mechanism in which oxytocin receptor (OXTR) neurons are connected to both pyramidal populations with certain connectivity rules. Although the model seems to reproduce the experimental results, some assumptions of the model seem to be poorly supported. If I understood correctly, the authors simply assumed that the strength of the connections between OXTR neurons and MALE neurons is the same as the strength of the connections between OXTR neurons and OTHER neurons. The authors neither discuss literature supporting such connectivity nor provide experimental evidence for this. I also could not find information about the magnitude of the synaptic weights to each of these populations. I guess these parameters are critical for the outcome of the simulation, and it may be worth exploring the outcome of simulating the different combinations of connectivity and synaptic weights between OXTR neurons and pyramids, as well as the degree of recurrent connectivity within the pyramidal subpopulations. Further, the authors should at least discuss in depth how inhibitory OXTR neuronal subtypes (they have different properties that could potentially be implemented in the modeling) may match their computational model best. If the current model remains the most promising, the authors should clearly discuss which experimental trajectory should be taken next to actually provide proof for its correctness (e.g. whether and how it would be possible to determine the predicted connectivity experimentally).

(2) Minor

The authors state regarding counterbalancing in Figure 1 and Figure S4C: "The social presentation order (male or female first), as well as the locations of the social and milk options (left or right relative to start arm), were fixed over sessions within a given subject, but varied over subjects (Figure S4C)". In Figure S4C, it looks as if there are fewer animals in which the male was always presented first than animals in which the female was always presented first (~ 16 vs. 20). While the difference is not very big, it may be influential. To fully exclude a sequence effect, I would suggest adding male-first animals until both groups are the same size.

Reviewer #1 (Public review):

The paper presents novel evidence that spatial representations prioritize coarse topological features (T‑junctions, holes, crosses) over precise Euclidean metrics like angle and length, using drawing-based memory tasks with adults and children. The study is interesting and well‑motivated, and the importance of topological relations is clear, but stronger and more nuanced evidence is needed before concluding that topological relations are more important than metric details, as task difficulty and the potentially distinct roles of metric and topological information in spatial representation have not yet been fully disentangled.

Introduction<br /> (1) P.5: Please explain in more detail what you mean by "What is relevant is the relative prioritization of each of these features."

Results<br /> (2) P.8: Please clarify how the "proportion of drawings with angles biased towards 90{degree sign}" was computed. Specify the criterion for counting a drawing as biased (e.g., a certain absolute deviation toward 90{degree sign} from the original angle), and explicitly state in the Results that absolute degrees of deviation were used, as described in Methods.

(3) It would help to spell out whether the findings imply that obtuse angles are typically drawn smaller (closer to 90{degree sign}) and acute angles larger (closer to 90{degree sign}). Also, would angles be more biased toward 90{degree sign} or 180{degree sign} (or 0{degree sign}) depending on the angle? (e.g., 175{degree sign} is seen more as 180{degree sign} while 95 is seen more as 90{degree sign})

(4) Figure 4B: The statement that "positive values indicate bias in the direction of 90 degrees" needs a more precise explanation. Please explain exactly how the bias metric is computed (e.g., signed difference between drawn and original angle, with the sign indicating movement toward or away from 90{degree sign}) and what the y-axis values represent. Given that the Methods refer to absolute deviations, it would be useful to reconcile where the positive/negative signs come from in this plot.

(5) Figure 4C: The description in the Results seems to use a different metric than what is plotted. Please ensure that the measure in the text matches the measure shown in the figure, and adjust labels or wording so they align clearly.

(6) P.11: Consider briefly justifying why the authors predicted that participants would also add L‑junctions, rather than only remove them.

(7) P.12: The last sentence: Weren't the overall rates of feature preservation 'higher' in the adult sample?

Methods<br /> (8) Experiment 1: Please clarify whether the angles associated with T‑ and L‑junctions were equated or differed systematically. A short description of stimulus generation (e.g., angle ranges, line lengths, junction configurations) would be helpful.

(9) It would also be helpful to specify the statistical tests used (e.g., t‑tests, ANOVAs, mixed‑effects models), including the main factors and any random effects, so readers can clearly follow your analysis pipeline.

Discussion<br /> (10) It may be important to note that task difficulty likely differs across feature types: junctions involve presence/absence or counting, whereas angle and length reproduction require finer metric precision. The authors' claim of "prioritization" and possible difficulty effects should be disentangled.

(11) Furthermore, would it be possible that people retain relative order/comparison of different angles/lengths rather than computing precise values?

(12) I agree that topological relations are extremely important. However, for above reasons, it seems like stronger/stricter evidence is needed to claim that topological relations are 'more' important than metric details. They also might serve different roles in spatial representations

(13) The Discussion would benefit from a short paragraph on where different junction types (T, L, crosses) typically appear in everyday scenes and objects (e.g., as cues to occlusion, surface intersections, 3D structure) and what functions they serve. This would help connect your experimental findings to the ecological importance of these features for natural vision and spatial cognition.

Reviewer #2 (Public review):

Summary:

This is an interesting study that uses drawings to evaluate the extent to which visual representations of letter- and graph-like figures (preferentially) include topological features, like junctions and holes.

The main claim is based on the observation that when participants are asked to draw presented figures from memory, they tend to (1) regularise angles towards 90deg and lengths towards the average length of the lines in the figure, while (2) preserving topological features like T-junctions more assiduously than non-topological features like L-junctions. A third experiment with 'serial reproductions' in which participants copy drawings made by other participants (like a visual version of the 'broken telephone' game) reproduce these patterns in exaggerated form. These findings were also reproduced in children (Experiment 4).

These findings are consistent with the idea that memory representations are low-bandwidth or noisy approximations to the original figure. I would suggest that when participants are asked to reproduce the figure, it is if they combine the noisy stored representation, with generic priors about angles and the average line length. The preferential preservation of T- over L-junctions indicates that they are somehow more salient or memorable. This is not inconsistent with the authors' preferred interpretation of an explicit representation of topological structure. However, it is also not inconsistent with the idea that in order to compress the visual signals for storage, high-information (complex) components of the source are given preferential treatment. This would be compatible with optimal use of limited resources when compressing the information. Additional comparisons and control conditions would help tease these alternatives apart.

Strengths:

+ Innovative use of drawing methods to probe internal visual representations<br /> + Experiments spanning both adults and children

Weaknesses:

- Failure to consider alternative hypotheses that are consistent with the findings

Reviewer #3 (Public review):

Kittur et al. ask whether human spatial memory is organized around topological relations (meaning coarse structural properties such as T-junctions, crosses, and holes) rather than around Euclidean properties such as angle and length. Across four experiments, adults and children studied letter-like figures and reproduced them from memory by drawing. The authors report two complementary patterns: metric features are systematically distorted, with angles pulled toward 90 degrees and line-length ratios compressed toward an average, while topologically critical features are comparatively well preserved. The central test contrasts T-junctions with L-junctions, which are visually similar but topologically distinct, since an L-junction reduces to a straight line, whereas a T-junction does not. A serial reproduction experiment amplifies both patterns across chains of participants, and a fourth experiment extends the findings to children aged five to eight.

Strengths:

The question is a good one and sits at a productive intersection of topics. It bears on debates about the representational format of cognitive maps, on proposals about the primitives of visual perception, and on a classic developmental claim from Piaget and Inhelder that has rarely been tested directly.

The drawing paradigm is well chosen and offers something that the group's earlier forced-choice work could not. Because participants produce an open-ended response, distortion of metric detail and preservation of structure can be observed within a single response, and the relationship between them can be examined directly. The serial reproduction experiment is a particularly effective use of this affordance. The choice of the T-junction versus L-junction contrast as the primary test is well-motivated, since it holds the number of junctions constant and varies only topological relevance. It is also worth noting for readers that the central claim of a representational privilege for topologically distinct features was previously established by this group using forced-choice paradigms in both adults and children. That a similar conclusion emerges from free generation is a genuine strength, since the two methods have very different sources of error.

The work is carefully executed. Sample sizes, dependent variables, and analyses were preregistered; stimuli were purpose-built for each question, including the deliberate exclusion of 90-degree angles so that no reference angle was available; drawings were double-coded; and the full set of raw drawings is being released publicly.

Weaknesses:

Drawing is treated as a transparent window onto representation, and motor limitations are not considered. Drawing is a motor act, drawing skill varies widely across individuals, and the manuscript does not discuss motor limitations at any point. As the study is designed, representational imprecision cannot be separated from difficulty of precise reproduction. The clearest way to resolve it might be asking adults to copy the figures exactly while the stimulus remains visible. If the biases persist under direct copying, then some portion of the effect is production rather than memory. Because the topological findings have already been demonstrated in keypress-only paradigms, this concern affects the metric distortion results most heavily, which are the novel contribution of the present paper.

Three distinct claims are treated as one, and the data speak mainly to the weakest of them. The paper moves between a claim about mnemonic robustness (topological features survive degradation better than metric features), one about representational architecture (topology is a base layer with metric detail superimposed on top), and a claim about priority (topology is encoded prior to metric detail). The experiments show evidence for robustness, which is a claim about what is lost first. Robustness does not entail architecture: an encoder with a single layer, whose loss happens to spare structure, produces the same pattern with no layered format and no claim about encoding order. Earlier work does address format, because false "same" judgments to topologically matched but metrically different items show that topology plays a role in what the system treats as equivalent. Preservation counting measures robustness, rather than equivalence.

Some alternative hypotheses to consider/address: (a) A capacity-limited memory that reconstructs from a prior produces the metric distortions with no commitment to topology. A literal absence of angle encoding, which the authors invoke, predicts noisy and unconstrained recall rather than recall pulled toward a particular value. The observed pattern reflects a structured prior. (b) The result that does discriminate might be confounded with local salience. A memory that adds uniform noise to all parts of a figure does not predict that T-junctions are preserved better than L-junctions; however, a three-way branch point is plausibly more locally distinctive than a corner, so a salience-weighted-but-topology-free account predicts the same ordering. (c) Motor simplification also predicts the same ordering, since omitting an L-junction converts a bend into a straight line, which is easier to draw, whereas omitting a T-junction requires dropping a stroke.

The better a feature works as a topological marker, the less variance it produces and the harder it is to test, so the method is best powered where the theoretical signal is weakest. Holes are the textbook case of a topological invariant and are reported to disappear from drawings less than one percent of the time, but they are excluded from formal analysis because they are at ceiling and have no matched comparison feature. It would be useful to see bidirectional rates for holes (both how often a hole disappears and how often participants spuriously close an open figure into a loop).

In Experiment 4, the conclusion of developmental stability rests on a nonsignificant effect of age, which is failure to detect a change rather than evidence of stability. An equivalence test or an estimate of the precision of the null is better support for developmental stability. Motor skill is confounded with age throughout. So this is an experiment that shows that the effect generalizes to childhood, but cannot adjudicate a developmental question.

In the serial reproduction experiment, chains were intermixed so that each participant contributed one drawing to each of ten chains. The final drawings are therefore linked through shared intermediate participants, and an individual with an idiosyncratic drawing style influences ten chains at once, so the reported degrees of freedom are somewhat generous. The analysis also focuses on the final drawings and sets aside the nine hundred intermediate ones, which are the data that would show where in a chain metric detail collapses and whether structure ever breaks.

To formalize the topology is to strengthen the argument: each figure is a one-dimensional complex (its underlying graph), treated intrinsically and up to homeomorphism. The homeomorphism type is what remains after suppressing all degree-2 vertices. Under this definition, every feature in this paper's taxonomy becomes one kind of object, namely a homeomorphism invariant of the graph: number of components, first Betti number, and the degree sequence of three or greater with its adjacency structure. Relatedly, the term "metric" needs to be unpacked. The 90-degree bias concerns angle, whereas the 4:2:1 result concerns length ratios, which are affine rather than strictly metric. The stronger statement available is that distortion appears at every level above topology in the transformation hierarchy while preservation occurs at the topological level, which connects directly to Chen's (2005) invariance hierarchy that is already cited.

Appraisal and impact:

The authors aimed to show that topological structure is preferentially retained in memory while metric detail is lost, and in the sense of relative preservation they succeed. The dissociation is real, replicates across two stimulus sets, amplifies under serial reproduction, and appears in young children. What the data do not establish is the stronger architectural claim that topology is a base representational layer, nor that the metric distortions specifically implicate topology rather than general properties of reconstructive memory. Separating these claims would communicate the well-supported result better. Conceptually, the work strengthens a growing case that coarse relational structure deserves a place alongside Euclidean properties in accounts of spatial representation. Practically, the public release of the full set of adult and child drawings, including excluded ones, is a resource that will support analyses well beyond those reported in this paper, and the serial reproduction design is a method that others will want to borrow.

Reviewer #1 (Public review):

Summary:

The authors developed a novel theoretical/computational procedure to count bacterial populations without introducing artificial randomness effects due to dilution. Surprisingly, this very important aspect of studies of bacterial systems has been overlooked. The proposed method provides a simple and transparent approach to eliminate the randomness of bacterial accounting procedures, allowing now to fully concentrate on the intrinsic effects of the studied systems.

Strengths:

A very simple and clear procedure is introduced and explained in full detail. This elegant approach finds an excellent compromise between mathematical rigor and computational efficiency, which is important for practical applications. The provided examples are convincing beyond a doubt, clearly indicating the potential strong impact of the proposed framework. Various complications and possible issues are also discussed and analyzed. This seems to be a very powerful novel method that should significantly advance the analysis of complex biological systems.

Weaknesses:

The only minor weakness that I found is the assumption of independence of bacterial species, which is expressed as the well-stirred approximation. One could imagine that bacterial species might cooperate, leading to non-uniform distributions that are real. How to distinguish such situations?

I believe that this method can be extended to determine if this is the case or not before the application. For example, if the bacteria species are independent of each other and one can use the binomial distributions - then the Fano factor would be proportional to the overall relative fraction of bacterial species. Maybe a simple test can be added to test it before the application of REPOP. However, I believe that this is a minor issue.

Comments on revised version.

I am satisfied with the correction proposed by the authors. The method is already quite impressive, and there is no need to complicate it at this stage.

Reviewer #2 (Public review):

I appreciate the thorough responses from the authors, which address my concerns. The expansion of Appendix B as well as the addition of text discussing how the REPOP method interacts with data collection efforts are very useful. These new sections show that relative error decreases with increasing samples, as expected, yet these error metrics, including KL divergence, describing the fit of the full distributions not just the modes, drop off fairly quickly with increasing number of samples showing that REPOP likely minimizes discrepancies between estimated and true distributions even at lower sampling efforts.

Additionally, the extension of the REPOP method to the quantification of multiple bacterial species or phenotypes shows the potential utility of the method in contexts beyond basic plate counts. Between this example and the additional information on how to implement REPOP, I believe this workflow will be attractive and accessible to the audience.

Reviewer #1 (Public review):

Summary:

The authors combine discriminative auditory fear conditioning with longitudinal in vivo calcium imaging to ask how prelimbic (PL) representations of learned and generalized threat evolve across recent and remote memory time points. Using two different CS+ frequencies and a no-shock control group, they report that PL population activity tracks graded behavioral generalization, that population similarity is highest for tones eliciting strong threat responding, and that distinct subnetworks can be identified that appear to encode tone-specific sensory features versus learned threat-related response structure.

To my knowledge, this may be the first study to comprehensively examine neural encoding of fear generalization in prelimbic cortex (PL). The manuscript is ambitious and technically interesting, and several aspects are potentially important. In particular, the suggestion that neurons showing graded, learning-related response patterns become selectively stabilized over time is intriguing. The inclusion of two CS+ training conditions and a no-shock control also strengthens the case that at least some of the reported effects are related to associative learning rather than simple sensory differences. However, in its current form, the manuscript does not yet fully support the strength of the conceptual claims. Several issues limit confidence in the interpretation, including the possibility that repeated testing itself contributes to changes across days, uncertainty about the relationship between neural activity and freezing behavior, limited quantitative documentation of longitudinal cell registration, and a number of problems in figure clarity and statistical framing. Overall, the study contains promising observations, but the claims should be narrowed, and several analyses or controls would be needed to fully support the proposed framework.

Detailed Comments

(1) A general concern is that the repeated test procedure itself may contribute to extinction. Because the animals are exposed to multiple CS frequencies across multiple test days, and each tone is presented three times per session, some of the reported changes in behavior and neural activity across days could reflect extinction or repeated nonreinforced retrieval rather than the passage of time per se. This is especially relevant given that the manuscript makes claims about recent versus remote representations and representational drift over 30 days. At a minimum, the authors should discuss this limitation explicitly and temper claims about time-dependent changes. Ideally, they would include a control group in which animals are tested only once or twice (e.g., at an early and later time point with fewer CS frequencies), or a reduced-frequency testing design that minimizes extinction while still allowing evaluation of recent versus remote memory.

(2) More generally, some of the reported learning-related neural differences may be driven by behavioral differences, particularly freezing, rather than by learning or generalization per se. For example, animals that freeze more to certain frequencies may show corresponding neural response differences simply because freezing alters PL activity. The authors should examine this possibility more directly. Analyses testing whether recorded cells encode freezing behavior, or whether tone frequency-related neural differences remain robust when comparing high- and low-freezing epochs, would help determine whether the reported effects reflect learned stimulus value rather than behavioral state differences.

(3) A central feature of the manuscript is the analysis of neural response properties over an extended period of time, up to 30 days after learning. However, aside from a brief mention in the Methods that spatial registration was used, the manuscript provides very little quantitative information about this critical aspect of the study. The paper would be strengthened by including explicit metrics describing longitudinal cell tracking, such as the number and proportion of ROIs retained across all sessions, distributions of spatial-footprint correlations or centroid distances across days, and representative examples of matched imaging fields over time. Without this information, it is difficult to assess how strongly the longitudinal claims are supported.

(4) The text states that "Figs. 1c and 1d show GCaMP6f expression in PL, representative calcium footprints, and activity traces". However, the figure as presented does not clearly show all of these elements, at least not in a way that matches the description in the Results. The correspondence between text and figure should be corrected.

(5) The labeling of Figure 2a is insufficient for interpretation. The legend states that the panel shows raster plots of sound responsiveness, but the axes and scaling are not clearly defined. It is not clear from the figure what the x-axis represents, whether the y-axis corresponds to individual neurons, where the CS period occurs, or what the activity scale at the right denotes. Also, the term 'rasters' implies that spikes were analyzed. It seems that the spike inference approach (CASCADE) was only used for later analyses. Perhaps 'heat-plot' would be more accurate here? Generally, this figure should be annotated more clearly so that the reader can understand it without referring back to the Methods.

(6) In relation to Figure 3, the analysis of population-averaged responses across tone frequencies is useful, but the manuscript would be stronger with additional statistical analyses across time and across groups. For example, if the authors want to argue that learning induces graded changes in neural responses and that these evolve across time, they should directly compare within-group responses across days and also compare matched frequencies between the conditioned groups and the no-shock controls. These analyses would help establish whether the observed differences are genuinely learning dependent and whether they change significantly over time.

(7) The inclusion of two different CS+ frequencies and a no-shock control is a strength of the study and substantially improves the interpretation that graded neural responses are related to learning and generalization rather than to simple sensory processing or passage of time. That said, I am not entirely comfortable with the use of the term "inference" throughout the manuscript. What is being measured here appears closer to sensory generalization than inference in a stronger cognitive sense. The current task does not clearly require that animals infer hidden structure or stimulus value through abstract reasoning; rather, the generalized stimulus may simply be treated as similar to the conditioned cue. The terminology should therefore be reconsidered or softened.

(8) I also found the use of the term "valence" somewhat problematic. The manuscript appears to use valence to refer to graded responding across tones with different aversive significance, but valence typically refers more broadly to distinctions between appetitive and aversive value. Here, terms such as "threat value," "aversive value," may be more precise. The authors should consider revising this language throughout.

Reviewer #2 (Public review):

Summary:

The following points are those that occurred to me across readings of the paper. They are listed in what I take to be the order of their significance. Many of the points relate to the loose use of language and invocation of concepts that are not warranted, given the study design and results obtained.

Major Comments:

(1) The concept of ensemble turnover is interesting - the way it is introduced and discussed implies some type of spontaneous change in the neural underpinnings of fear discrimination and generalization in the PL. But, of course, every trial involves an opportunity to learn about the threat CS or the generalization test stimuli, and I am troubled by the thought that stability in the neural underpinnings of fear discrimination and generalization will actually reflect the level of defensive behaviours evoked on different trial types and/or the discrepancy between those behaviours and the outcome of a given trial in the generalization test. That is, stability in the neural underpinnings may be related to an animal's certainty or uncertainty in the contingency between a stimulus and danger; or, put another way, an animal's confidence that danger will or won't occur given the presence of some stimulus. This is not uninteresting. It is, however, not considered anywhere in the paper, which is overloaded with references to inferred threat values and integration of information across different types of stimuli. The protocol is not one that requires inference about anything or integration across anything.

(2) I appreciate the link to Gu and Johansen in paragraph 3 of the Introduction, but the type of generalization under investigation here is not the same as the type of 'generalization' studied by Gu and Johansen [who used a sensory preconditioning protocol]. Nonetheless, the authors have forced the language used by Gu and Johansen into their paper, and this has created tension [at least for this reader] as the concepts introduced by Gu and Johansen [inference, integration] are simply not relevant given the generalization protocol used here. Here are a few examples of points where the tension might interfere with a reader's understanding:

a. 'We hypothesized that generalization to novel stimuli depends on stable subnetwork organization that enables comparisons between learned and inferred valence, as well as population-level features that reduce variability across related representations.'

I understand the words in the hypothesis, but can't form a representation of what is being said because of the reference to terms that stand in need of clarification [inferred valence, variability across related representations], but, ultimately, won't be clarified. This needs to be re-expressed so that the reader can appreciate what is being said.

b. 'Our results show that stable cortical subnetworks integrate the emotional "gist" of memory and inferred valence for novel cues over time, despite ongoing ensemble reorganization, and that population-level firing rate similarity across stimulus presentations determines threat generalization.'

Again, what does this mean? How is the gist of a memory integrated with inferred valence for novel cues over time? The statement simply doesn't make sense. This needs to be rewritten for clarity.

c. 'In CS⁺15 mice, positively modulated sound-responsive neurons exhibited graded tone activity reflecting the contingency learned valence as well as the inferred valence of novel tones across testing days...'.

Can this be rewritten as 'In CS⁺15 mice, positively modulated sound-responsive neurons exhibited graded activity to the tone CS and its variants that were used to assess generalization.'? The overloading of the text with references to 'contingency learned valence' and 'inferred valence' is unnecessary and makes it much harder to understand what has been shown in the results.

(3) Re the same passage of text as in 2c:

Is it the case that these neurons are simply tracking the expression of freezing to the various tones? The same question applies to the results obtained for the CS+3 mice. If this is the case, then why should the results be taken to support the banner statement that 'Sound-modulated PL population responses encode learned and inferred valence' - these analyses do not support that statement. And, as indicated, I don't believe that the language of learned and inferred valence is appropriate to such statements, given the nature of the protocol used and results obtained. It is a study looking at how populations of neurons in the PL respond during presentations of auditory stimuli that were subject to discriminative conditioning, and during tests of generalized freezing to other [intermediate] auditory stimuli.

(4) It is stated that:

'In no-shock controls, although both positive and negative responses were present, population activity was not modulated by tone frequency or valence'.

What does this mean? I can understand that population activity was not modulated by tone frequency. But what does it mean to say that it was not modulated by valence? Why should it have been when none of the tones were conditioned in this group and, hence, mice were responding to all the tones equally? And given that this is true, I don't understand the use of 'valence' here, or the subsequent statements in this paragraph that 'graded responses require associative learning' and that 'PL population responses encode graded sound-valence associations that reflect both learning and inference, closely matching behavioral generalization.' The latter statement is particularly unwarranted and, again, highlights a major issue with the paper. It could and should be rewritten as 'PL population responses reflect behavioral generalization.' There is nothing in the additional language that adds to the reader's understanding of what has been shown. The reference to 'graded sound-valence associations that reflect both learning and inference' is completely unwarranted, given the nature of this study. It is anathema to the vast literature on stimulus generalization. If the authors wished to make statements of this sort, they should have taken a different approach, perhaps using protocols like those featured in Gu and Johansen.

(5) The section titled, 'Consistently active neurons preserve valence representations as newly recruited neurons sharpen remote memory traces' ends with the following summary:

'Together, these results indicate that consistently active neurons maintain stable representations of learned and inferred sound associations across time, whereas neurons recruited after conditioning progressively acquire graded tuning at later retrieval stages. This dynamic refinement suggests that cortical memory representations become increasingly selective during systems consolidation, while a stable neuronal subpopulation preserves the core emotional content of the memory.'

Once again, the summary is not in keeping with the results obtained. The 'dynamic refinement' of representations is far more likely to reflect the repeated testing across days 1, 15, and 30 rather than anything to do with systems consolidation - at the very least, it is the simplest interpretation of the results. The impact of repeated testing is evident in the sharpening of generalization gradients over time, which is contrary to what is otherwise observed in the literature - the incredibly well -documented broadening of generalization gradients with time. Given this impact of repeated testing, surely the changes in the neuronal population that underlie performance are more likely to reflect the learning that occurs on days 1, 15, and 30, which is reflected in reduced freezing to the non-conditioned tones. If this is a reasonable take on the results, then I don't see the basis for invoking systems consolidation at all, and I don't see the basis for inferring a stable neuronal subpopulation that preserves the emotional content of the memory. Rather, non-reinforced presentations of 'never-reinforced' tones result in recruitment of additional neurons that result in suppression of freezing responses to those stimuli.

(6) In the section titled, 'Population vector similarity at stimulus onset determines degree of generalization', it is stated that:

'Because population similarity peaked shortly after stimulus onset, we quantified similarity during the first 5 s after tone onset relative to the CS⁺. In CS⁺15 mice, population similarity was highest for 15/15 and 15/11 tone pairs with no differences between them.'

Isn't this consistent with the view that the population response in the PL simply reflects the level of freezing? Freezing to the 15-15 and 15-11 tones is most likely to be similar on their first presentation prior to the effects of extinction on the 11 Hz tone; hence the results obtained. That is, these results appear to clearly indicate that neuronal responses in the PL reflect the degree of stimulus generalization, as evidenced in freezing behavior. Given all that we know about the involvement of the PL in expressing fear responses, it is not appropriate to claim that 'population vector similarity at stimulus onset *determines* the degree of generalization. The PL responses simply reflect the varying levels of performance displayed to the different types of tones. What have I missed that could be taken to support additional statements?

Later in the same section, it is stated that 'population-level similarity at stimulus onset scales with behavioral threat generalization and is maximal for tones associated with robust threat responses.' For simplicity and, therefore, clarity, this should be rewritten as 'population-level similarity at stimulus onset reflects behavioral threat generalization.'

(7) In the section titled, 'Different subnetworks encode acoustic versus learned properties of sound association', it is stated that:

'Our previous analyses show that learned and inferred associations are represented at the population level. However, these results do not resolve whether graded responses arise from pooled activity of frequency-selective neurons or from subnetworks encoding integrated learned valence across tones.'

What does it mean to say 'integrated learned valence across tones'? As it presently stands, the meaning of the phrase is unclear. It only makes sense if one supposes that generalized freezing responses to the 11 and 7 kHZ tones reflect separate associations between those tones and the aversive foot shock US. This supposition is inconsistent with the rich literature on generalization of Pavlovian conditioned fear responses. Specifically, it is inconsistent with the many theories of fear generalization, which attribute the reduction in fear as one moves away from the specific conditioned stimulus to a decrement in the ability of the test stimulus to activate the trained CS-US association. My strong impression is that the authors would do well to ground their findings in theories of stimulus/fear generalization, of which there are many. This would better serve the results obtained [and the reader's appreciation of them] - at present, the unnecessary invocation of concepts does very little to enhance the reader's appreciation or understanding of what has been found in the study.

(8) Another example of what has been a common theme in this review :

'...we hypothesized that the PL active ensemble segregates into functionally distinct subnetworks: one encoding tone-specific sensory features with dynamic characteristics, and another responding to all frequencies encoding stable core memory content and inferred emotional valence.'

What does it mean to say 'all frequencies encoding stable core memory content and inferred emotional valence'? Do the authors mean to say '...and another that tracks freezing/defensive responses regardless of whether they were elicited by the trained CS or one of the generalization test stimuli'?

(9) It is stated that - 'Graded clusters encode emotional valence but constitute only a fraction of the active population; yet valence coding at the population level remains accurate and precise. This indicates that neurons newly recruited into the population-likely frequency-selective and organized within learning-independent clusters-can be shaped by associative processes through modulation of firing activity.'

What does this mean? Are the authors trying to say that - 'Some clusters of PL neurons track freezing responses. In spite of the fact that these are only a fraction of the total active neuronal population, the population-level response of PL neurons also tracks the levels of fear to the trained tone and its variants used in the test for generalization.' If this is what one wants to say, then the final statement in the reproduced section does not follow. That is, there is no indication that 'neurons newly recruited into the population-likely frequency-selective and organized within learning-independent clusters-can be shaped by associative processes through modulation of firing activity.' As noted, the characteristics of other ensembles that become active across the repeated tests on days 1, 15, and 30 are more likely to reflect learning from non-reinforcement that occurs within and across those sessions. Perhaps this is what is meant by the phrase, 'shaped by associative processes'? If so, it should be stated explicitly instead of left to the reader to work out.

(10) The following points all relate to the Discussion and reiterate many of the points above.

a. 'A subset of neurons remains consistently active across sessions, preserving core components of the memory trace and supporting inference of emotional valence for novel sounds, while neurons recruited after conditioning progressively acquire valence selectivity at remote time points.'

'Inference of emotional valence' is unclear and unwarranted for all of the reasons provided above regarding the use of language.

b. '...Our data reconcile these views by demonstrating that cortical representations of emotional valence emerge rapidly after learning and persist within stable subnetworks, even as the broader population undergoes substantial turnover. This architecture preserves core mnemonic content while allowing flexibility in the surrounding ensemble.'

These statements assume that the PL neuronal responses reflect something more than the levels of freezing behavior to the different stimuli; what are the grounds for this assumption?

c. 'Importantly, these subnetworks encode both learned contingencies and the inferred valence of novel stimuli along a graded representational axis, suggesting that strong recurrent connectivity provides a stable scaffold for emotional memory representations.'

What is a graded representational axis, and what part of the first statement suggests that 'strong recurrent connectivity provides a stable scaffold for emotional memory representations'? If the authors' goal was to make statements about emotional memory representations vis-à-vis emotional memory content, they should have used protocols that allowed them to probe such content. The auditory fear conditioning protocol used here [followed by tests for generalization to other auditory stimuli that differ in frequency from the conditioned tone] is not one that lends itself to analysis of emotional memory representations or content.

d. 'Dynamic tone-selective responsive neurons emerge independently of learning, as they are present in both control and experimental mice, reflecting pre-existing PL sensory-driven properties (Hockley & Malmierca, 2024; Zikopoulos & Barbas, 2006).'

Maybe. They are also likely to have developed as a consequence of the repeated testing on days 1, 15, and 30, which involved intermixed exposures to the tones of different frequencies. That is, rather than 'pre-existing PL sensory-driven properties', the responses of these neurons might reflect the emergence of discrimination between the various tones across testing, and greater suppression of freezing to the non-trained tones compared to the trained tone across the various test intervals.

Reviewer #3 (Public review):

Summary:

Normandin et al. explore the coding of stimuli predicting an aversive event in the prelimbic cortex. Stimuli could either be explicitly paired, explicitly unpaired, or novel but with an inferred association with the aversive event (generalization). Long-term tracking of GCaMP-positive neurons allowed them to examine how coding evolves out to a month following training. In general, they found two types of ensemble codes. One was ensembles coding for each stimulus independently, but with enhanced responding to the one eliciting a freezing response. The other was ensembles that responded to all stimuli in proportion to their similarity to the stimulus paired with the aversive event, either increasing or decreasing their activation with the degree of freezing elicited by a stimulus. Importantly, this second set of ensembles was more stable across days, potentially providing a memory trace.

Strengths:

(1) The authors track ensembles in prelimbic cortex over long time scales, providing valuable information on the consolidation of neural codes.

(2) Neural coding of generalization is examined, which is under-examined in the field.

Weaknesses:

(1) Difficult to determine if responses treated as encoding stimulus valence are driven instead by the behavior that the stimulus elicits, freezing.

(2) The study implies that the identified ensembles are causally related to valence memory, but no experimental interventions are performed to justify this.

Reviewer #3 (Public review):

Summary:

Human and animal trypanosomiasis are fatal illnesses caused by African trypanosomes transmitted by tsetse flies during a bloodmeal. Thus, tsetse fly feeding is the key physical step in disease transmission to mammals. Tsetse fly feeding is not a new story, but it is revisited here through the application of sophisticated imaging techniques and novel biomechanical methods of analysis. The author's aim is to provide a high-resolution picture of the structures and forces involved in feeding to provide mechanistic insights into the process of feeding, from attachment, penetration, drinking and retraction of the feeding parts.

Largely the authors have achieved their aims. They (i) examine the structures and forces involved in attachment; (ii) they provide detailed multi image analysis of the proboscis providing insights into its probing ability and physical mechanism of penetration; (iii) they conduct a controlled analysis of the physical forces involved in penetration and report that they are in the low nM range, not especially strong but much higher that the mosquito bite and finally they provide a first analysis of blood uptake during feeding.

Strengths:

The study images the tsetse fly feeding structures in unprecedented detail, with resolution to the uM scale, in 3-D, and during feeding. The resulting images are dramatic and insightful (and beautiful and frightening!) that researchers interested in trypanosomes, tsetse flies or blood feeding by flies in general will want to see.

They conclude that flies attach strongly to smooth surfaces, because of interactions possible via the array of acanthae of the pulvillus pad at the ends of the tarsi. The estimated attachment forces are similar in male & female flies, in the low mM range (they look impressively strong in video 1). They provide a very striking analysis of the proboscis and labellum and associated tooth structures (Figs 4 & 5). I recall many years ago observing that tsetse flies are messy feeders, and these structures, especially the rasping teeth structures on the reverse folded labial tips explain why! This seems more like a chainsaw than a jigsaw in action, but the authors are probably correct that these structures and probing/retraction mechanism explain many features of tsetse fly feeding and their ability to feed on a wide range of hosts with very different skin types.

The impressive aspect of this paper is the range of imaging techniques, (CLSM, SEM, uCT, FIB SEM), the quality of the images which attests to the obvious care taken with sample preparation. The biomechanically analysis, especially the penetration analysis is impressive. Finally, the paper is clearly written and presented, it was a very easy read and overall, a very engaging study.

Weaknesses:

I suppose it could be said that the paper is a descriptive study; it doesn't really test a hypothesis but that is not a prerequisite for publication. Perhaps the least convincing prats are the imaging of the flexible v rigid parts of the structures, which is based on amount of resilin (flexible) and chitin-protein (stiff) based on their autofluorescence. In seems odd that the joints would be less blue (stiffer) in Fig 1i, or what the blue structures correspond to in Fig. 6B-D.

Comments on revised version.

In revised version these issues have been satisfactorily addressed

Reviewer #1 (Public Review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers.]

Summary:

This study aims to understand how cell fusion contributes to wound healing using a laser-induced injury in the notum epithelium of a developing fruit fly. The authors meticulously characterize the epithelial fusion events using a live imaging approach and report that syncytia arise by 'border breakdown' and 'cell shrinking'. The syncytial epithelial cells also appear to outcompete mononucleated cells and preferentially dissolve their tangential borders, which correlates with the accumulation of actin at the leading edge.

Strengths:

The strength of this study is the authors' live imaging approach to capture these dynamic fusion events that are a fundamental yet poorly understood biological process.

Reviewer #2 (Public Review):

Summary:

Overall, this study provides a thorough description of the formation of syncytia following wounding of the proliferation-competent diploid epithelium of the pupal notum. While this phenomenon has already been described briefly for this particular tissue by the Galko lab in Wang et al 2015, the authors provide a much more detailed description and characterisation of the process providing some novel insights (radial versus tangential border breakdown, cell shrinkage, timings, syncytia outcompeting mononucleated cells, etc.).

Strengths:

This paper provides an elegant, thorough, descriptive characterisation of syncytia-driven wound closure using state-of-the-art confocal live imaging of the pupal notum. The authors show that laser-induced wounding of this diploid, proliferation-competent epithelium results in the formation of syncytia of various sizes in the first few cell rows around the wound edge, which progressively become bigger as healing proceeds. This results in ~50% of cells becoming part of these syncytia. The cell fusion events were convincingly demonstrated by showing the disappearance of p120ctnRFP and E-Cadherin-GFP from cell-cell borders as well as cytoplasmic GFP mixing of GFP-positive cells with a GFP-negative cell.

Apart from cell-cell fusion by border breakdown that mostly happens in the first 2h following wounding, the authors also found that at later stages of wound healing cell shrinkage following cytoplasmic mixing contributed to syncytia formation.

Next, the authors provided some convincing evidence that syncytia outcompete mononuclear cells for being positioned in the first cell row around the wound.

The authors then show that radial border breakdown occurs much less frequently than tangential border breakdown. They suggest that radial border breakdown reduces the requirement for cell-cell intercalations. They also hypothesise that tangential border breakdown might allow fused cells to share resources and provide more resources to be used near the wound edge, e.g. for actomyosin cable formation. To test this, the authors generate single-cell clones that overexpress Actin-GFP. They then show convincingly how a single Actin-GFP-positive cell in the second cell row fuses with one GFP-negative cell in the first cell row. The Actin-GFP signal then spreads in the fused cell and labels some previously unlabelled actin-rich structure near the wound edge which most likely is the actomyosin cable. This provides some evidence for resource sharing by cytoplasmic mixing following fusion.

Reviewer #3 (Public Review):

In this revised manuscript, White et al. aimed to understand the wound-induced syncytia formation behavior in wound repair of Drosophila melanogaster pupal notum. For this purpose, the authors characterized two different types of adherens junctions' outcomes during syncytia formation around the wound region - border breakdown versus apical shrinking which appear to happen in different time points and for different time durations. The authors characterized cell-cell fusion events using cytoplasmic, junctional and nuclear markers. They determined that about half of the cells within 70 um radii from the wound undergo cell-cell fusion. They studied wound induction on the border between control epithelia and pnr domain suggesting that Atg1 is required for post-wound syncytia formation and wound closure. They showed that during wound closure syncytia gradually invade the wound leading edge mostly by radial fusion events. The data suggests that intercalation of cells from the leading edge slows down the wound closure process. They propose that cell fluidity of syncytial cells plays a role in wound closure speed. Finally, the authors showed that actin is concentrated to the front edge of syncytia located in the wound leading edge. The authors described some aspects of syncytia formation during wound closure using different approaches.

Reviewer #1 (Public review):

Summary:

This study asks how selection for male aggressiveness affects life-history and reproductive fitness traits in Drosophila melanogaster males.

Strengths:

Multiple comprehensive assays are used to address the question.

Weaknesses:

(1) The flies used for comparisons are inadequate. Behavioral assays compare Bully males mated top non-coevolved Cs females with Cs males mated to coevolved Cs females.

(2) Lifespan analysis is done on male progeny of Cs females mated to either genetically more distant Bully or co-evolved Cs males, the longer lifespan and performance on the former is interpreted as trade-off with aggressiveness, rather than a simple explanation of hybrid vigor.

(3) Differences in CHCs between Bully and Cs males and Cs females mated to those males are not shown to cause difference in measured behavioral outcomes.

Comments on revised version.

I appreciate authors responding to reviewer's comments. The inclusion of additional Bully lines in behavioral analysis, and Bully homozygous male progeny in lifespan analysis gives more strength to the authors' conclusions. It does not exclude other possible explanations for the observed results, but now authors note genetic drift as an alternative explanation for some of their results.

I do want to point to a potential misunderstanding of male-female co-evolution by authors. The authors state that "The Bully lines used in our work were derived from Canton-S flies and thus did co-evolve with Cs". This statement is incorrect if the process of selection and line maintenance in this study was the following:

In my understanding to create Bully lines the most aggressive males were first chosen from an ancestral Cs line and their most aggressive male progeny were mated to their sibling females, repeating the process for 37 generations. Therefore, Bully females were co-evolving with Bully males during selection process of over 37 generations, while Cs females were staying co-evolved with their own males, since they mated within the line. Moreover, after aggressive lines were created, they were kept separate from each other, and from Cs line since about the year 2010, until the experiments described in the paper were performed (which must over 10 years?). Over 10 years, a significant genetic drift can happen, that changes allele frequencies, and may results in differences in male-female co-evolved traits and in lifespan that are unrelated to selection for aggression.

Also, decapitating females does not completely prevent female influence over mating process, but just removes central brain control over it. In Drosophila, however, the main control over copulation process for males and female is not central. Therefore, you do not completely remove the effect of coevolved or non-coevolved female traits over copulatory and post-copulatory processes.

Reviewer #2 (Public review):

Summary:

The authors compare "Bully" lines, selected for male aggression, to Canton-S controls and find that Bully males have lower mating success, shorter mating durations, and remate sooner. Chemical analyses show Bully males have distinct cuticular hydrocarbon (CHC) signatures and transfer markedly less cVA to females, offering a plausible mechanistic link to weaker mate-guarding. Paradoxically, Bully males live longer and remain fertile at older ages when Cs males no longer mate, indicating a shift in the reproduction-survival trade-off in aggression-selected populations. Importantly, the work sheds light on proximate mechanisms, demonstrating that shifts in CHCs and pheromone transfer co-occur with changes in fitness traits.

Strengths:

The manuscript's strengths lie in its comprehensive and integrative approach framed within an evolutionary context. By combining behavioral assays, chemical profiling, and lifespan measurements, the authors reveal a coherent pattern linking aggression selection to life-history trade-offs. The direct quantification of cVA in the female reproductive tract after mating provides a particularly compelling mechanistic correlate, strengthening the link between behavior and chemical signaling. Findings on altered 5-T and 5-P levels further highlight how chemical communication shapes mating and mate-guarding strategies. Analytical approaches are largely rigorous, and the results provide valuable insights into the pleiotropic effects of selection on socially relevant traits.

The revision responds directly to the main concerns raised previously. The addition of a third, independently selected line (Bully C), together with the Bully × Bully data, considerably reduces the concern that the behavioral phenotypes reflect line-specific drift or founder effects rather than a correlated response to selection. The reorganized survival figure (Figure 5) is a clear improvement over the previous version, with isolated and group-housed males in separate panels and a heterozygous Bully condition added, so the longevity claim can be evaluated more directly. The isolated-male data are especially useful here, since those flies never mate, and a longevity difference under that condition argues that the effect is not simply a consequence of Bully males mating less often. The behavioral schematic, corrected symbols, and reported sample sizes also help, as does the reinterpretation of the post-mating courtship data in terms of courtship motivation rather than a refractory-period effect once no latency difference was found.

Weaknesses:

Most of the remaining weaknesses are ones I raised in the first round, and the revision has narrowed them. The links between the altered CHC profiles, the reduced cVA transfer, and the behavioral outcomes remain correlative. The causal experiments that would establish them (for example, perfuming or cVA-equalization) are acknowledged by the authors as future directions, which is reasonable, but it means the mechanistic claims should be read as candidate explanations rather than demonstrated ones. It is also worth noting that the CHC differences and the behavioral differences may both be downstream of a common selection target (for instance, genes affecting oenocyte function or CHC biosynthesis) rather than one causing the other; the Discussion would be more balanced if this alternative were stated explicitly.

My main remaining concern is with the lifespan data. The behavioral phenotypes are replicated across Bully A, B, and C, but the survival assays were done on Bully A only, so a line-specific contribution to the longevity result, including drift, cannot be excluded, even though this has been addressed for the behavioral traits. This matters because the title and abstract present the survival-reproduction trade-off as a general consequence of selection for aggression, whereas the survival evidence rests on a single line. The authors can either run the lifespan assays on a second line, or calibrate the text, title, and abstract so that the strength of the survival claim matches the single-line evidence behind it, with second-line lifespan data noted as a future step.

The Bully C line is currently underused. Its intermediate aggression, together with the absence of a significant reduction in mating duration, points to a graded rather than binary relationship between aggression intensity and mating duration. This is one of the more interesting features of the expanded dataset, and it deserves more than its present role as a justification for focusing on Bully A.

The authors have appropriately softened causal language in the title, subheadings, and much of the Discussion. A few residual passages still imply causation or directional transfer and would benefit from the same treatment.

Reviewer #1 (Public review):

Summary:

Deng and colleagues pursue the possibility that red light exposure can provide some benefits and anti-senescence effects in aged mouse models. In addition, they show how red light influence metabolism in cultured keratinocytes. The authors provide a long dissection of the potential paths involved in the changes promoted by red light exposure, identifying CytC oxidase, SIRT4, PPARa and MCD as key players.

Strengths:

The authors did a thorough exploration of the multiple potential avenues by which red light exposure influence metabolism. The in vitro and in vivo evidence nicely complement each other.

Weaknesses:

This is a challenging hypothesis that would require some additional experimental controls. The pathway dissection, while extensive, sometimes is approach in unconvincing ways and the results are not always evident to judge or interpret. Technically, the western blots and transcriptomic analyses require notable improvements.

Comments on revised version.

The revised version of the manuscript provides some improvements. However, I feel that many aspects remain poorly addressed. In the authors' favour, many of these limitations are now acknowledged in their rebuttal, as well as in the discussion section.

Reviewer #2 (Public review):

Summary:

This work identifies a previously unknown way that red light can slow ageing. The authors show that red light lowers the level of a protein called SIRT4 in skin cells. Reducing SIRT4 boosts fatty acid use and increases a type of histone modification that keeps genes active. These changes help cells clear away signs of ageing, reduce inflammation, and restore normal metabolism. The findings open the possibility of developing new treatments that target SIRT4 to reverse age‑related decline.

Strengths:

The evidence is solid because the authors use several complementary methods. They test red light in both cultured cells and naturally aged mice, and they confirm the key role of SIRT4 by silencing its gene. Measurements of metabolism, protein changes, and ageing markers all point in the same direction. However, the exact way red light lowers SIRT4 levels is not fully explained, which leaves a minor gap. Overall, the conclusions are well supported and convincing.

Weaknesses:

The paper does not evolve to use the mechanistic discoveries of the manuscript to help our community to identify the mechanism of photobiomodulation, which is not known so far.

I would like to draw your attention to a recently published paper by Herrera et al. (FEBS Letters 2025, doi:10.1002/1873-3468.70195), which shows that red light (660 nm) stimulates mitochondrial fatty acid oxidation in keratinocytes via AMPK‑dependent phosphorylation of ACC, without altering expression of electron transport chain complexes. I believe this paper is highly complementary to current study.

Herrera et al. demonstrate that red light increases basal, ATP‑linked, and maximal oxygen consumption rates in keratinocytes specifically through enhanced fatty acid oxidation (inhibited by etomoxir). This independently validates the central finding of the current manuscript ,i.e., red light boosts lipid metabolism, strengthening the robustness of this concept.

While the current manuscript focusses on the SIRT4‑MCD axis, Herrera et al. identify AMPK phosphorylation and ACC inhibition as key effectors. Authors can integrate and expand their discussion, since SIRT4 downregulation may converge on AMPK activation, or they may represent parallel, reinforcing mechanisms. This would enrich the mechanistic model and open new hypotheses.

The mechanism of photobiomodulation: Herrera et al. explicitly challenge the prevailing paradigm that red light acts solely via cytochrome c oxidase (by showing long‑lasting effects, unchanged OXPHOS protein levels, and no difference in permeabilized cells). The current finding (red light acts through SIRT4 downregulation, i.e., not direct enzymatic activation, aligns perfectly with Herrera´s critique.

Long‑term metabolic effects - Herrera et al. show that a single red light exposure elevates oxygen consumption for up to 2 days. The current study focuses on changes at 12‑24 h. Their data extend the time window and suggest that the metabolic reprogramming you describe may persist longer than currently discussed, which is clinically relevant.

Discussing Herrera et al. results would not only acknowledge independent, corroborating evidence but also allow the authors to position your SIRT4‑centric mechanism within a broader, emerging understanding of red‑light photobiomodulation.

Comments on the latest version:

The authors have made a terrific work in answering the reviewers and modifying the manuscript.

Reviewer #1 (Public review):

Summary:

The authors investigate the role of different specific dopaminergic neurons in the mushroom body of Drosophila larvae for learning and innate behavior. All the tested neurons are thought to be involved in punishment learning. The authors discover that artificial activation of single DANs in training leads to safety learning, but not punishment learning. Furthermore, activation of single DANs can lead to changes in locomotion behavior, which can affect light preference. The authors provide a deeper understanding of the functional diversity of single dopamine neurons; however, it is unclear how translatable these findings are to learning experiments with real punishment stimuli.

The authors provide a detailed behavioral analysis of locomotion in response to activation of various dopamine neurons. This analysis allows them to exclude that the locomotion defects affect memory recall behavior.

Strengths:

The authors disentangle which kind of memories are formed with the activation of different dopamine neurons - safety learning and/or punishment learning. They further investigate whether the US is required in the test for recall. They do indeed find differences, and the results will be of interest to the learning and memory community.

Interestingly, optogenetic activation of a single DAN during training leads to safety memory, but not punishment memory. Furthermore, DAN activation also affects innate locomotion, and the authors show that optogenetic activation of different DANs affects locomotion differently.

Weaknesses:

All experiments in the manuscript use optogenetic activation of DANs, thus it is not clear what kind of memories are formed. Several stimuli can be used as punishment, such as electric shock, salt, bitter, and light - it is not clear what kind of memory the authors investigate here. The findings could be discussed in the context of what DANs respond to. Furthermore, studies in adults and larvae showed that most DANs can code for both valences - etc., aversive DANs can be activated by punishment, and inhibited by reward. Thus, safety learning might be a result of a decrease in activity in DANs during odor presentation. The authors also do not discuss possible feedback loops from MBONs to DANs across compartments. Could such connections allow for safety learning in larvae?

The authors show that artificial activation with different light intensities can form different memories and that increasing the light intensity sometimes leads to no memories. Also, using different optogenetic tools reveals different results. This again raises the question of how applicable the results will be for learning with real stimuli. Is there a natural stimulus that only induces safety learning, but no punishment learning? The authors discuss these limitations.

Reviewer #2 (Public review):

Summary:

This study provides valuable context for ongoing research on the role of dopamine in memory and locomotion. DANs have been a fascinating area of study due to their complexity, and this work dissects specific DANs, exploring their roles in different memory-related behaviors while offering some explanations. The discussions provided by the authors effectively situates the study in the broader field of learning, memory, DAN circuitry and behavioral computation in insect brains. The study achieves what it sets out to and it does so unequivocally. The experiments were elegantly designed, leaving little room for doubt in the study's claims. However, the study lacks context regarding the molecular pathways underlying these results. While it strengthens current knowledge by providing robust evidence, it does little to explore the molecular mechanisms behind these effects.

Strengths:

(1) Experiment design is one of the strengths of this study. The experiments are thorough and cover the length and breadth of the core findings of the study. Although a lot of work has already been done in studying the role of dopamine in memory and locomotion, the dissection of the functions of distinct DANs in larvae has been done meticulously with well-structured experiments.

(2) This study fits quite nicely into the puzzle of memory, especially in the context of Dopamine. Previous studies in *Drosophila* adults have shown the opposing roles of DANs in locomotion depending on the context of DAN activation. This study drives that point home for larvae, providing conclusive evidence in that regard.

(3) The use of clear figures and simple language is one of the strengths of this paper. The figures are comprehensive, complete and manage to narrate the story by themselves. The flow of information is smooth. The simple and effective language used maintains scientific rigor while remaining accessible to those new to the field. A pleasant read.

Weaknesses:

(1) The authors have done a great job at structuring the figures. But some main figures would benefit from including the controls instead of placing them in supplementary.

(2) The paper would benefit from a deeper discussion regarding molecular mechanisms underlying their results. It would be interesting to see what the authors think about different Dopamine receptors and how they relate to the findings of this paper.

(3) Throughout the paper, the authors have been clear and comprehensive, but in some cases, further explanation of their choices were missing. For example, the choice to compare bending and tail velocity over other parameters within the same clusters is unclear.

Comments on revised version.

Most of the comments have been addressed.

Reviewer #3 (Public review):

Summary

Across species, dopamine release serves seemingly diverse functions, such as reinforcing memories and regulating locomotion and flight. However, whether distinct dopaminergic neurons (DANs) are allocated to each function is unclear. In this study, Toshima et al. have used the numerically simple organization of the Drosophila larval brain to answer this question. They use optogenetic activation to systematically stimulate a small set of DANs, individually and collectively, and study the effect on diverse functions such as memory formation, retrieval, and locomotion. The reproducibility of optogenetic activation is a strength of this approach. At the same time, this is a caveat, as optogenetic activation may not recapitulate natural modes of activation and may lead to outcomes not observed under natural conditions. They find that singly or collectively, DL1 DANs can induce punishment and/or safety memory formation and retrieval. DANs can even gate the expression of memory. Finally, the same DANs also modulate locomotion in the larvae. The authors speculate that dopaminergic neurons in other species may also share such overlapping functions. Their findings are nicely summarised in Figure 9.

Strengths

The study systematically activates neurons in the DL1 cluster. Individual and collective stimulation of the Dl1 DANs has been conducted to assess the induction and gating of aversive punishment memory, safety memory, and acute locomotion.

Specific adult Drosophila DANs are known to induce dual behaviors and functions. The same MP1/y1pedc DANs are recognized for gating appetitive memory expression and representing aversive teaching signals downstream of sensory stimuli such as electric shocks, bitter tastes, and heat. Neurons in the PPL1 cluster regulate adult flight and food-seeking behavior. The authors deserve credit for conducting an organized examination of dopaminergic neuronal functions in larvae, thereby making their findings more comparable and facilitating the proposal of a holistic model.

They have provided substantial evidence for their findings and have frequently presented replicated behavioral datasets. They have been transparent about the results that were difficult to explain. Additionally, they have provided an impressive body of supporting data to strengthen their main findings.

Weaknesses

As mentioned above, optogenetic activation may not recreate natural neuronal activation in response to external stimuli. This could have led to outcomes that will not occur under other natural circumstances.

Comments on revised version.

I appreciate the author's responses, and I do not have any comments or suggestions at this point.

Reviewer #1 (Public review):

Summary:

In this manuscript the applicants study two residues in the GHKL ATPase active site of Aq MutL and GyrB, and argue that the catalytic base function is shared between two conserved acidic residues that are 3 residues apart.

In the manuscript, they generated mutant versions in MutL and GyrB (both ala and the appropriate Asn/Gln version) and performed ATPase analysis. They also generated high resolution crystal structures of the GyrB NTD with AMPPnP for WT and mutants of the two acidic residues. The data show that mutation in either of these residues does not fully kill activity (with the exception of the Alanine mutation of the first of the two, that interferes with ATP (or AMPPnP) binding). When the acidic residues are mutated to Asn/Gln, the catalytic water can still be positioned, and hence these mutants are more active than the Ala mutants. In both cases the double mutation is catalytic dead.<br /> The authors then perform phylogenetic analysis and ancestral gene reconstruction and based on this they argue that HSP90 forms a different class of GHKL ATPases, and lost rather than gained this separate status.

Strengths:

The biochemical analysis seems solid.

Weaknesses:

- A major question that remains, is why the mutations have so much more detrimental effect in MutL (100-fold lower kcat/KM) than they do in GyrB (3-fold lower). Can the authors explain this? Doesn't this argue against the proposed catalytic conservation?

The authors need to discuss this issue explicitly to make it clear that conservation of the mechanism is not complete and that other interpretations are possible.

- The structure figures all have omit maps for just the AMPPnP and the water, whereas the density for the the acidic residues and their mutants are not shown.

This has been addressed.

There are some issues with figure S2B and S5.

Reviewer #2 (Public review):

Summary:

In this manuscript, Fukui et al. re-examined the ATP hydrolysis mechanism in GHKL ATPases, revealing a cooperative role of two conserved acidic residues rather than one. The authors have used a range of biochemical and structural techniques on various mutants from different members of the GHKL ATPase family to test and validate their proposed mechanism.

Through a detailed re-analysis of their previously published structure of the aqMutL NTD (ATPase domain) in complex with AMPPCP, they identified Glu29 and Glu32 as interacting with nucleophilic water for the catalysis. The authors carefully dissected the respective roles of these two acidic residues with a series of site-directed mutations. Mutations at Glu29 impaired ATPase activity without affecting protein secondary structure or ATP binding in the case of the E29Q mutant. Moreover, mutations at Glu32 did not affect secondary structure (except for E32G) but reduce ATPase activity. Activity was abolished when both residues (E29Q/E32Q) are mutated.

The authors extended their study to another GHKL ATPase, aqGyrB. Their findings further supported the cooperative function of the corresponding acidic residues in aqGyrB (Glu48 and Asp51) during ATP hydrolysis. Mutation of these residues partially impaired ATP hydrolysis without affecting protein secondary structure. ATPase activity was completely lost in the double mutant E48Q/D51M. While the E48Q mutant retained the ability to bind ATP, the E48A mutant did not. High-resolution structures of the WT and E48A, E48Q, D51A and D51N mutants of the aqGyrB NTD demonstrated that nucleophilic water positioning depended on these residues. E48 played a dominant role in water positioning and is critical for stabilising ATP lid formation and associated conformational changes, whereas D51 contributed cooperatively to catalysis.

The authors investigated the functional impact of mutating the corresponding residues in the human MutL homologs PMS2 and MLH1. Clinical variants consistently exhibited reduced or abolished ATPase activity, providing a potential molecular basis for Lynch syndrome, through impaired DNA mismatch repair.

Lastly, through evolutionary analysis, the authors inferred that the second acidic residue was likely present in the common ancestor of MutL, GyrB, and MORC proteins, but was lost in the case of Hsp90.

Strengths:

(1) This study contains a detailed structural and biochemical analysis of a biologically important set of GHKL ATPases. The authors identify a second acidic residue that is conserved and contributes to catalysis in a large subset of GHKL ATPases. An updated and extended mechanistic model of ATP hydrolysis by this class of enzymes is proposed, which involves cooperative and partially overlapping roles for the catalytic residue pair. This revised mechanistic model is invaluable for the interpretation of clinical variants of GHKL ATPases such as PMS2 and MLH1.

(2) The work described was performed to an excellent and rigorous technical standard. The structural and biochemical data are sound. The evidence supporting the claims is compelling.

Weaknesses:

(1) The identification in this study of a second acidic residue contributing to catalysis but not absolutely essential for catalysis is a useful finding. However, given that many structures of GHLK ATPases have been determined with different nucleotide analogs bound and that the essential role of the first acidic residue is well established, the importance and scope of the advances described here remain focused within the field of study of GHKL ATPases.

(2) The authors assessed the consequences of variants in the human MutL homologs PMS2 and MLH1, but various other human GHKL ATPases contain clinically relevant variants, some of which have stronger disease associations than the mutations examined in this study. A broader analysis of any effect of disease-linked mutations in GHKL ATPases would have strengthened this study.

(3) The effect of other aqMutL NTD E32 mutants, particularly, the E32K mutant on ATP binding remains unclear, although experimental assessment of nucleotide binding would be challenging due to the high protein concentrations required for the equilibrium dialysis assay.

Reviewer #1 (Public review):

Summary:

The authors conducted a carefully constructed experiment to test reorganization in the auditory cortex in deafness in response to task (spatial vs. temporal working memory) and modality (visual vs. somatosensory). They found a complex pattern of results, which included changes to univariate response strength in deafness that differed between the primary and association auditory cortex. HG showed a preference for the somatosensory working memory task, whereas STG/S responded more in both modalities for the temporal task. They further showed multivariate similarity of their results to models representing both task and modality in both groups, which were increased in deafness. Curiously, the task effect was for a sensorily-bound model, which shows mid-level representation, and not high-level task ("metamodal") code.

Strengths:

I appreciated the matched design, rigorous analysis, and careful interpretation of the nuanced results.

Weaknesses:

Only minor weaknesses: behavior and residual hearing can be better controlled.

Reviewer #2 (Public review):

Manini and colleagues present an interesting study on the consequences of early deafness on the organization of temporal regions chiefly engaged in audition in hearing people. Mainly relying on representational similarity analyses, they show that the auditory cortex in deaf individuals represents information about task, sensory modality, and somatosensory frequency. Critically, task and modality representations were also found in the auditory cortex of hearing individuals. There were significant differences between groups, implying that these representations are enhanced as a consequence of deafness.

Overall, I feel that the paper could gain in clarity and impact if the hypothesis space tested in the introduction and discussion was made clearer, if some new analyses were provided to support some claims, and if the authors better matched their conclusions to the observed results.

Reviewer #3 (Public review):

Summary:

This manuscript examines functional plasticity in auditory cortices of people born deaf (or early deaf). Deaf participants (N=13, all native BSL signers) and hearing controls (N=18) performed a delay-match-to-sample working memory task in visual and somatosensory modalities, attending either to frequency (temporal task) or spatial pattern (spatial task). Using fMRI and univariate and representational similarity analysis (RSA), the authors test what type of information is represented in auditory areas of hearing controls and deaf participants. Different types of tasks and stimulus features are examined, including low-level sensory features (vibration/movement frequency and spatial position of stimuli on screen/hand) as well as task features (attending to space vs. frequency) and stimulus modality (somatosensory vs. visual).

There are a number of interesting findings. In early auditory cortices (right Heschl's gyrus), only Deaf participants show above baseline responses and only in the somatosensory task. In secondary auditory/multisensory STS, only deaf participants show above-rest responses to both visual and somatosensory stimuli. In RSA analysis, d/Deaf, but not hearing participants, show sensitivity to the frequency of somatosensory vibration when finger position is held constant (SFRm model). RSA in auditory cortices finds sensitivity to sensory modality (visual vs. somatosensory) information in both hearing and deaf participants, with an enhanced effect in the d/Deaf group. A similar d/Deafness enhancement effect was observed for task (frequency vs. spatial) but only within modality, not across, suggesting a less abstract representation.

Strengths:

The paper has a number of strengths. Running tasks in multiple modalities in the same study is a technical challenge and adds valuable information, since, as it turns out, early auditory areas of deaf people are sensitive to somatosensory but not visual frequency. Moreover, it makes it possible to look for modality coding.

The RSA analysis finds sensitivity to task and modality features not detectable with univariate analyses.

Overall, the findings of differences across modalities (visual and somatosensory) in auditory cortices are very interesting. Having parallel tasks in the two modalities makes it all the more important that only somatosensory stimuli activate HG. The discussion of this finding and ideas about alternative possible paths of somatosensory and visual information to the auditory cortices in d/Deaf individuals is also interesting.

The authors conclude that there is both evidence for shared and different function across hearing and deaf groups, and this makes sense. It's refreshing that the authors acknowledge that the simplicity dichotomy of preservation vs. change present in the literature and presented in the introduction turns out not to explain the findings.

Weaknesses:

The participant sample is not very large, but this is a difficult-to-recruit population, and it is within an acceptable range since the authors have taken care to run a robust study design and collect a large amount of data from each person.

Some weakness of the paper includes incomplete presentation of the results and conclusions that do not follow from the data.

The results are presented in a way that makes it hard to track what is significant and how the auditory ROIs are different or similar to the control regions. It is also difficult to connect the written text results with the figures. In some cases, the figures look like there is no significant effect, but the results report that there is one.

It is not clear that three-way interactions were tested for (e.g., group, by task, by modality). This complicates the interpretation of significant two-way interactions, e.g., task by group. For example, in STG a task by group interaction is reported, but the plot suggests that this is driven primarily by the somatosensory modality.

The paper suggests that frequency/temporal information is coded in auditory areas of d/Deaf participants but not location/spatial information. This is stated in the Results and in the Discussion as a major point. But this is not quite true in the somatosensory case and not true in the visual case at all, as far as I can tell. In the somatosensory case, since frequency is only coded in a finger-dependent manner, location is in fact coded in this regard. This pattern differs from what is observed in the somatosensory cortex, where frequency is coded in a finger-dependent and independent manner as well as location as such. In the visual case, it seems like temporal, i.e., frequency information, is not coded at all in the auditory ROIs. Although it is also puzzling that visual frequency is not coded in the visual 'control' ROI. An incomplete presentation of results motivates some conclusions that are not warranted, e.g., coding in the auditory cortex reflects a preservation of its function - i.e., frequency but not location coding.

A second related issue is that some dimensions (e.g., visual frequency, modality-independent task) show no neural response anywhere in the brain, including in the canonical visual, somatosensory, and amodal networks. For these dimensions, the current experiment does not offer a good test. This is fine but should be clearly stated in the results and the Discussion so there is no confusion about which hypotheses are really tested. Right now, the results say things like "the control ROI shows the expected pattern", but in some cases it's more complicated and failures to observe effects constrain what can ultimately be expected in auditory areas. This is okay, but needs to be made clear in the results and discussion, and auditory results need to be interpreted in this context.

Relatedly, the paper has no whole cortex searchlight analyses, and it is not clear why. If nothing comes out in the small sample of n=13, that is okay, but at least the collapsed hearing and d/Deaf sample data should be shown for each dimension. This will give the reader a sense of what is to be expected and contextualize the results.

Some claims are made which are not supported by the data: "However, it is likely that the representation of somatosensory frequency in deaf individuals relies on the same mechanisms used to represent auditory temporal frequency in hearing individuals." Likewise, the paper goes on to say, 'the underlying computations might be the same'. True, they might be, but they also might be different. No evidence is presented to support one or the other hypothesis, so both should be stated, and it should be stated that these cannot be distinguished based on the presented data.

Conclusions-wise, the paper sometimes makes sweeping claims that go well beyond what the evidence supports and fails to provide caveats. The first paragraph of the Discussion states: "Overall, these findings suggest that crossmodal plasticity relies on representational and functional configurations that are present across individuals and modulated by sensory experience." Such a sweeping conclusion about cross-modal plasticity in general is not supported or refuted by the present data. There is no evidence that 'representations' or 'functional configurations' are the same across groups. What are functional configurations? What representations are shared? Second, it is far too general to make claims about 'cross-modal plasticity' based on one study with one population.

Reviewer #1 (Public review):

Summary:

The present report describes an investigation into the use of machine learning techniques to improve cross-racial/ethnic performance of brain models of cognitive function. The authors tested several approaches to boost prediction of NIH cognitive toolbox scores using brain imaging data (function, structure) for minoritized (Black) participants in the ABCD Study sample compared to white (majority) participants. Structural (e.g., volume) measures showed the greatest performance gap, and a balanced weighting method showed the greatest performance gain across features. The authors conclude that supervised domain adaptive methods can improve models for cognitive prediction and mitigate cross-racial/ethnic performance disparities.

Strengths:

This investigation makes some headway into issues by identifying computational methods that may help to improve some models for limited outcome variables (i.e., general cognitive performance). Addressing racial/ethnic disparities in brain imaging research has significant implications for generalizability of findings and for the practical utility of imaging findings in the wider population. A comparative approach to evaluate the improvements in a "prediction gap" across various methods could have benefits for neuroimaging beyond racial/ethnic disparities. The use of the ABCD Study, given its deep phenotyping of individuals, is also a benefit.

Weaknesses:

Despite its strengths, there are several large conceptual and related methodological issues that impact its conclusions and the overall utility of the approach. The sample selection approach limits insight into likely drivers of the performance gap (e.g., socioenvironmental disparities known to exist between groups and associated with neurodevelopment), and in so doing ignores a critical component of understanding racial brain differences, particularly in relation to cognitive functions. Further, while the relative gaps in performance of a single cognitive score across features are well described, the actual performance (and therefore relative benefit to these techniques) is unclear. Specific examples include the following.

(1) The overarching conceptual issue with the manuscript is a lack of engagement with a substantial and growing evidence base on the drivers of racial disparities in brain imaging which impact model performance. Racial/ethnic groups in the US (and other regions of the world) are not equivalent in terms of developmental environments that shape brain function and structure (see Harnett et al., 2023, Neuropsychopharmacology; Ricard et al., 2023, Nature Neuroscience; Cardenas-Iniguez & Gonzalez, 2024, Nature Neuroscience for some overview here). The socioenvironmental disparities inherent to race in the US further shape cognitive development and brain associations with cognitive performance (e.g., Marek et al., 2025, Science). The framing of the manuscript focuses almost exclusively on broad sampling issues, and in doing so treats racial/ethnic variability as if it reflects statistical abnormality rather than a critical component of understanding human brain development. This lack of contextualizing racial disparities significantly impacts the overall utility of the proposed approach and the conclusions of the manuscript.

(2) In relation to the above, another conceptual issue in this approach of using a majority to inform minority brain associations with cognitive variables is an assumption that minority brain patterns should match the majority, rather than developmental stressors inducing alternative brain-weighting to predict outcomes. This framework does not assess this possibility and may in fact obscure such an outcome, limiting our inferences into neurodevelopment.

(3) Another conceptual/methodological issue here is the use of "matched groups" for analysis. The specifics of matching are fairly vague, but given the description one would assume the w/B groups are matched on a number of behavioral/socioenvironmental variables, which is a significant issue for interpretability and applicability. As noted, w/B groups in the US (and the ABCD Study) differ substantially across variables; matching has the likely consequence of creating a highly non-generalizable sample, particularly when the minority group is restricted to N = 10.

Reviewer #2 (Public review):

In the manuscript "Supervised domain adaptation mitigates cross-ethnicity prediction errors in neuroimaging-based cognitive prediction", the authors investigated the efficacy of data adaptation techniques to reduce ethnicity-related prediction bias in neuroimaging-based cognitive prediction. They found that data adaptation algorithms, particularly balanced weighting, contributed to mitigating ethnicity-related performance disparities. Furthermore, these bias mitigations could be achieved without requiring a large set of data from the underrepresented ethnic group. This study addressed an important concern in the field of neuroimaging-based behaviour prediction, providing many intriguing results. Nevertheless, the manuscript also suffers from a lack of coherent methods design, the unorganised presentation of information, and the lack of in-depth discussion of results.

The conclusions claimed by the authors are sometimes over-generalised and not fully supported by the study outcomes. Overall, this study demonstrated strong technical designs and convincing statistical analysis for the main outcomes, although clearer presentation would be needed to convey the messages in the manuscript.

The central investigation of this study is whether domain adaptation techniques improve ethnicity-related performance disparities. However, these improvements were only measured against a very weak baseline model, where a small set of African American (AA) subjects were added to the training sample consisting purely of White American (WA) subjects. While the authors recognised that balancing the training sample could already mitigate the ethnicity-related disparities, they considered that such approaches are unfeasible in their experimental scenario, where only a small amount of AA data were available. However, as Li et al. (2022) showed, a balanced sample of around 90-150 AA subjects could already reduce the ethnicity-related bias. Even from a practical standpoint, this balanced sample approach would be a more valid baseline for domain adaptation models to compare against.

The authors made two main conclusions: that domain adaptation methods reduced ethnicity-related bias, and that balanced weighting performed the best and the most stably. Both claims were over-generalised to some extent. First, the adaptation benefit claimed in the first conclusion is not seen in the functional connectivity (FC) modality, which is the most popular modality for neuroimaging-based prediction of behaviour. This difference in adaptation benefit across modalities is an important finding that is meaningful for future studies, the omission of which also removes interesting insights that the audience could take away from this article.

Second, the judgement of prediction performance is based on the area under the improvement curve (AUIC) metric, which summarises a model's performance across different availability of labelled AA data. As a result, the analysis of prediction performance naturally favours algorithms that could perform well with a small amount of added AA data. On the one hand, this provides an easy decision point for users to pick an algorithm to use without being concerned about data availability. On the other hand, important insights could be overlooked with the oversimplified recommendation of balanced weighting. As the authors have also observed, in some cases, domain adaptation strategies do not improve ethnicity-related bias more than the non-adaptation baseline. If the message is to recommend simple, low-cost strategies to reduce ethnicity-related prediction bias, it would be misleading not to note that the simplest and lowest-cost strategy could also be non-adaptation methods sometimes.

Regardless, for the general audience, the underlying assumptions when interpreting the AUIC metric are not immediately clear, which could cause the conclusions to be misleading. Apart from aggregating over different amounts of available AA data, the statistical comparison of AUIC gain across data adaptation algorithms also did not account for the impact of brain phenotype modalities. Even though the upstream analyses have confirmed that adaptation benefits vary greatly across brain modalities, this major observation was not followed in the final analysis where conclusions were made about which algorithm performed the best. Based on visual inspection of Figure 3b, it may be suspected that PRED performed better than or comparably to balanced weighting when task contrasts based on the Destrieux atlas were used.

Finally, the findings from this study align with the common hypothesis that ethnicity-related prediction bias originates from disparities already manifested during data collection and preprocessing. As the authors have noted, the modalities with the most tendency for ethnicity-related bias are the anatomical ones, including all three volume-based modalities (cortical volume, T1 and T2 subcortical volume) in the top ten phenotypes with the largest performance gap. Most prominently, brain features in the occipital pole, frontal pole, and a range of subcortical areas were found to contribute highly to adaptation gain. Subcortical areas are often reported to show noisier measurements compared to cortical areas, whereas the poles of the brain are likely more strongly warped/distorted during alignment to a standard template. From a data quality perspective, these results support the interpretation that ethnicity-related prediction bias may stem from loss of data quality during data collection or preprocessing. In the prediction models based on anatomical brain features, data adaptation methods may have helped to address these disparities in the data, without the more resource-intensive need to improve the bias in preprocessing pipelines.

Li, J., Bzdok, D., Chen, J., ... Genon, S. (2022). Cross-ethnicity/race generalization failure of behavioral prediction from resting-state functional connectivity. Science Advances, 8(11), eabj1812.

Reviewer #3 (Public review):

The manuscript frames its work in fairness and disparities but does not show or directly test that its approach decreases differences between White and African Americans. While it is stated that the objective is not to equalize performance across groups, large parts of the paper repeatedly claim that the methods mitigate cross-ethnicity disparities and improve fairness. Improving prediction in African American participants relative to a non-adapted model is not necessarily the same as reducing the disparity between African American and White American participants. The adapted model should be evaluated in both groups, and the post-adaptation performance gap should be reported directly.

Additional prediction performance measures are needed. For example, in Li et al, different results and conclusions are made with MSE and the correlation between observed and predicted variables. In that paper particularly, aggression measures showed better correlation in African Americans but better MSE in White Americans. Such differences are important to note as they likely suggest different mechanisms.

Similarly, characteristics of the cognitive outcome need to be understood. For example, differences in MSE or MAE may reflect a difference in variance between the groups. The group with a larger variance will have a larger MSE. Correlation or other performance measures that are invariant to different mean or variance shifts can be helpful here.

While the authors note that for the paper they treat racial and ethnic backgrounds interchangeably, I do not think that is the best given the differences between them and the impact and history they have in American culture. Overall, the authors likely need to do a better job conceptualizing their results in the history of minoritized populations in the United States. It is immensely important not to treat them as biological domains without considerable qualification and to avoid language implying that observed domain differences are intrinsic properties of racial groups.

Changes in feature weight are not a proper way to identify the mechanisms of improved performance. At most, these analyses characterize how model coefficients change when target-group data are incorporated or upweighted.

The cross-validation strategy is suboptimal. First, the use of the matched splits of the ABCD data introduces data leakage. To match a validation set to the training set in such a manner requires that each split knows about the other split's characteristics. That is data leakage. Though the impact could be small. Second, African American breakdowns are not balanced across sites and scanners. Domain adaption methods may be learning a shortcut or proxy for African American like site, scanner, or something else. A likely better approach would be some sort of leave X sites out approach, where a model is trained on White Americans from a set of sites, adapted with African Americans from those sites, and applied (with and without adaptation) to the White and African Americans from the left-out sites.

Baseline models for comparisons to the domain adaptation are missing. Some simpler ones include a target-only model trained on the same 10-100 African American participants and a pooled model with a group indicator and group-by-feature interaction. Without these comparisons, it is difficult to know whether balanced weighting is learning target-specific neurobiological information or merely recalibrating the prediction distribution.

There are a few statistical issues:

(1) The repeated MAE estimates are therefore not independent observations. Paired t-tests cannot be applied across repetitions. Subject-level bootstrap or permutation procedures that repeat the complete training and testing process are needed

(2) The caption describes approximate 95% confidence intervals as {plus minus}1.96 × SD/n. Conventionally, the standard error would involve SD/sqrt(n). However, even if corrected, there would still be issues about the dependence among the overlapping resamples.

(3) Ten repetitions are likely insufficient, especially in the case of ten target participants. Results at n = 10 may be extremely sensitive to which children are selected. The authors should use substantially more repetitions and report the full distribution of results.

(4) The Friedman and Wilcoxon comparisons treat the 80 imaging phenotypes as the observational units. These phenotypes are highly dependent because they are derived from the same participants, many use overlapping images, and numerous task contrasts and structural measures are strongly correlated. This non-independence can make the comparison among adaptation methods look much more precise than it is. A hierarchical analysis by modality or a resampling strategy that preserves dependence among phenotypes would be more appropriate.

(5) The gap metric and AUIC are difficult to interpret. Gap is the absolute relative difference between target-group and source-group MAE, normalized by source-group MAE. It is sensitive to the denominator and may produce large values whenever source-group. MAE is relatively small. Reporting signed raw MAE differences and MAE ratios alongside this derived score would help. Similarly, the AUIC combines errors with an arbitrary sequence of target-sample sizes. IStatistically significant differences in AUIC do not necessarily indicate practically meaningful differences among methods.

(6) The correlation between baseline gap and adaptation gain is partly tautological. Those with the widest gaps have the most room for improvement and likely thus show the greatest improvement. While still of value, the authors may want to tone down their interpretation of the correlation and describe its limitation.

(7) Given that the sample sizes vary from approximately 4,000 to more than 11,000 depending on modality, the authors may want to consider a reduced sample matched in size across modalities. It is hard to fully know if the conclusion that connectivity is more robust given the wide-scale differences in sample size and feature dimensionality.

(8) PLS are sensitive to many factors like scaling and collinearity. Many recent papers have been written about their limitations when used for subtyping. Some of these hold for prediction too. I think showing the results are consistent with different prediction algorithms is needed. SVR and ridge regression are two common methods for regression prediction with neuroimaging data.

(9) The feature interpretation is partly circular. The method with the largest performance gain is selected, and its coefficient changes are then used to explain that gain. A method designed to give target observations greater influence will unsurprisingly change its coefficients more than naïve inclusion.

(10) The practical and ethical deployment scenario is underspecified. Supervised adaptation requires labelled cognitive outcomes from the target population and, as currently framed, may require choosing a model based on an individual's racial category. What are the implications of deploying race-specific models that need to be considered? It is not self-evident that this approach is preferable to developing a broadly representative model or directly modeling the social and technical sources of distribution shift.

(11) The paper is worded and interpreted much too strongly. The current study supports the conclusion that, within ABCD, giving a small labelled target-group sample greater influence can sometimes improve held-out target-group MAE relative to naïvely adding the same participants. It does not yet establish that the method improves fairness or identifies mechanisms of racial bias. Likewise, the abstract and conclusion overstate the results. The abstract states that all adaptation methods reduced target-group prediction error, while the Results show near-zero or negative benefits for several functional-connectivity phenotypes and instability of PRED and interpolation below 30 target participants. Similarly, "substantially reduce disparities," "improve equity," "consistently," and "practical path forward" are stronger than the analyses support. Finally, the limitations section is incomplete and omits the more consequential limitations.

(11) That only ten labelled participants are needed to change the results is troubling. This is a shockingly low number. Giving ten target observations disproportionate influence can move the fitted model, particularly when the balanced-weighting ratio is high. A measurable MAE change is therefore possible, but it may reflect a shift in intercept or slope rather than learning a stable target-group brain-cognition relationship. Further, the manuscript does not report the numerical improvement for the n = 10 condition in the text or a table. Visual inspection of Figure 4 suggests reductions of roughly 0.10-0.25 standardized MAE units for some high-gap structural phenotypes, approximately 10-20%, while low-gap connectivity phenotypes show little or no gain.

(12) The study lacks genuine external validation, which may be needed to fully convince readers that such a low number of subjects is needed to reduce biases.

Reviewer #1 (Public review):

In this manuscript, the authors explore whether GPCR signaling in astrocytes affects the production of TNF by astrocytes and, to a lesser extent, microglia. Unfortunately, the method used by the authors to acquire astrocyte-enriched cultures is known to result in meaningful rates of contamination by myeloid cells (microglia and others), oligodendrocyte-lineage cells, and neurons. Alternative methods of generating highly enriched astrocyte cultures, as well as purifying astrocytes with little to no neuronal or myeloid contamination across age and brain regions, have shown no evidence of TNF expression by astrocytes (Zhang et al., J Neurosci, 2014; Zhang et al., Neuron, 2016; Clarke et al., PNAS, 2018). In fact, the paper cited by the authors as demonstrating differences between human and rodent astrocytes found no evidence of TNF expression in immature or mature human astrocytes (Zhang et al., Neuron, 2016). The idea that the majority of the observed TNF transcriptomic signal, at least in culture, comes from myeloid or neuronal contamination also aligns with the authors' observation that myeloid-enriched cultures act identically to astrocyte-enriched cultures.

The authors also use a GFAP virus to drive GPCR signaling in astrocytes and neuronal progenitor cells in their cultures, but, given that these cultures are known to have meaningful contamination by other cell types, such signaling could be due to astrocyte → microglia/neuron signaling or other multicellular pathways that cannot be excluded. Similar concerns mean that we cannot assume the effect of DREADD activation of astrocytes in vivo (Figure 6) reflects a bulk change in TNF expression driven by astrocyte-specific changes rather than by multicellular signaling.

The most compelling evidence for their claim of astrocyte TNF expression comes from the human-induced astrocytes. However, their antibody staining is not sufficient to claim these cells are truly astrocyte-like. Antibody staining is highly prone to non-specificity, as highlighted by the fact that their ALDH1L1 antibody staining appears perfectly nuclear despite ALDH1L1 being a cytoplasmic protein.

To address both the purity concerns of the astrocyte-enriched cultures and the concerns about the astrocyte identity of the induced astrocytes, the authors should perform RNA sequencing. By profiling gene expression in these cultures at the genome-wide level, readers can truly assess the degree of contamination and thus the likelihood of the proposed mechanism (i.e., astrocyte-specific TNF production). Importantly, previous studies have suggested that very little neuronal and myeloid contamination is required to dramatically change cellular responses (Foo et al., Neuron, 2011; Liddelow et al., Nature, 2017).

Reviewer #2 (Public review):

Summary:

Abbasi et al. examine how signaling through the major G-protein pathways (Gs, Gq, and Gi) influences tumor necrosis factor expression in astrocytes and microglia. Using a combination of pharmacological receptor activation, chemogenetic manipulation, primary rodent glial cultures, human induced pluripotent stem cell-derived astrocytes, and an in vivo astrocyte-targeted Gi manipulation, the authors report a broadly consistent pattern in which Gs- and Gq-associated signaling reduces tumor necrosis factor expression, whereas Gi signaling increases it. The study's cross-species and cross-preparation design, spanning astrocytes and microglia as well as in vitro and in vivo systems, provides a potentially valuable framework for understanding how neuromodulatory pathways may regulate glial inflammatory signaling.

Strengths:

A major strength of the study is the breadth of experimental systems used, which includes primary rat glia, human induced pluripotent stem cell-derived astrocytes, and an in vivo manipulation, allowing for comparison across species and levels of biological complexity. The use of chemogenetic receptors in astrocytes provides relatively direct control over Gq and Gi signaling, and these experiments yield consistent effects on both tumor necrosis factor messenger RNA and protein, strengthening the internal validity of the astrocyte findings. The observation that similar directional effects are seen in human-derived astrocytes and in microglial cultures further supports the idea that aspects of this regulatory relationship may be conserved across glial cell types. More broadly, the study addresses an important and timely question about how neuromodulatory signaling pathways interface with glial inflammatory outputs, and it generates a coherent set of observations that could serve as a foundation for more mechanistic work.

Weaknesses:

The central claim that Gs, Gq, and Gi signaling broadly and directly constitute a general regulatory code for tumor necrosis factor expression is more expansive than the current evidence fully supports. In particular, the evidence for Gs-dependent effects is indirect, relying on beta-adrenergic receptor activation and forskolin-mediated adenylyl cyclase stimulation rather than direct manipulation of Gs itself, leaving uncertainty about pathway specificity. More generally, the use of different endogenous receptors to represent each G-protein class in microglia complicates interpretation, since individual receptors may engage additional signaling pathways beyond their canonical G-protein coupling, limiting the extent to which the results can be attributed to G-protein class alone.

The in vivo experiment also does not definitively establish the cellular source of the observed increase in tumor necrosis factor, as measurements are taken from bulk cortical tissue following astrocyte-targeted Gi activation. This leaves open the possibility that the observed changes arise indirectly from other cell types, particularly microglia, which are shown elsewhere in the study to be strongly responsive to Gi-related manipulations. In addition, the specificity of chemogenetic expression in vivo is not quantitatively demonstrated, further limiting cell-type attribution.

There are also important issues related to experimental design and statistical interpretation. Across several experiments, it is unclear whether reported sample sizes reflect independent biological replicates, technical replicates, or imaging fields, which is especially consequential for the human induced pluripotent stem cell-derived astrocyte experiments where donor-level independence is not clearly established. The in vivo design also appears to treat hemispheres as independent observations despite their paired nature, which may inflate statistical independence given the small sample size.

Finally, several conclusions would benefit from more cautious framing. The data support differential regulation of tumor necrosis factor relative to interleukin-1 rather than strict cytokine specificity, and measurements based solely on messenger RNA should not be interpreted as direct evidence of cytokine production. The comparison between glial signaling effects and neuronal excitation or inhibition also juxtaposes fundamentally different biological readouts and should not be interpreted as a direct functional opposition. Overall, while the study provides interesting and potentially important observations, the broader pathway-level and cell-type-specific conclusions are not yet fully established by the current experimental evidence.

Reviewer #1 (Public review):

Summary:

This paper leverages 7T fMRI data from the Natural Scenes Dataset to investigate whether retinotopic coding the position-selective organization of visual responses structures spontaneous resting-state interactions between the Default Network (DN) and the Dorsal Attention Network (dATN). Using individualized network parcellations and population receptive field (pRF) modeling, the authors show that DN voxels can be split into two subpopulations based on their response to visual stimulation: those with position-specific positive BOLD responses (+pRFs) and those with position-specific negative BOLD responses (-pRFs). Critically, these subpopulations relate differently to the dATN during rest: -pRFs are anticorrelated with the dATN, +pRFs are positively correlated, and non-retinotopic DN voxels show no coupling. The anticorrelation (and positive correlation) is enhanced when DN and dATN voxels share visual field preferences. An event-triggered analysis suggests that retinotopic coding shapes both "top-down" (DN-initiated) and "bottom-up" (dATN-initiated) spontaneous activity transients, supporting the claim that the retinotopic scaffold is intrinsic to the DN. These findings challenge the prevailing view of global DN-dATN antagonism and suggest retinotopic coding as an organizing principle for cross-network communication.

Strengths:

The central finding that what looks like network-level independence between DN and dATN decomposes into structured, bivalent interactions organized by voxel-level visual field preferences is a compelling demonstration that macro-scale network descriptions can hide meaningful substructure. The logic of the analysis is clean: pRF properties are estimated from retinotopic mapping data and then used to predict resting-state coupling in completely independent scanning sessions. This cross-session, cross-modality design rules out many circularity concerns.

The use of individualized multi-session hierarchical Bayesian parcellation (Kong et al.) to define DN and dATN boundaries within each subject is the right methodological choice for this question. Network boundaries in posterior cortex, where DN and dATN interdigitate most closely, vary considerably across individuals, and group-average approaches would introduce exactly the kind of misassignment that would most confound the result.

The matched-vs-random pRF analysis is well-controlled. The authors demonstrate that cortical distance between matched and randomly matched dATN pRFs does not differ, effectively ruling out spatial proximity on the cortical surface as a confound. tSNR controls further show that signal quality differences do not drive the effect.

The event-triggered analysis (Figure 3) is creative and adds genuine value. Showing that retinotopically-specific coupling persists during DN-initiated activity transients not only dATN-initiated ones is the key piece of evidence for the claim that the code is intrinsic to the DN rather than passively inherited through bottom-up visual drive.

The result is observed consistently across all individual participants, which provides strong evidence for the robustness of the qualitative pattern despite the small sample size inherent to densely sampled designs.

Comments on revised version:

I'm content with the additional analyses and alterations to the writing that the authors have performed. I'm convinced that this work will spawn a very productive thread in the literature.

Reviewer #2 (Public review):

Summary:

Using a public dataset of retinotopic mapping and resting-state data, the authors find that the default mode network has voxels that respond (positively or negatively) to visual stimulation at specific retinotopic positions, and that resting-state activity in these voxels is correlated with activity in more traditional sensory voxels with the same visual-location preference. The retinotopic specificity is bidirectional, such that high activity in default mode voxels drives activity only in voxels with matching receptive fields in sensory cortex, and vice versa. These findings are at odds with traditional views of the default mode network as having abstract (non-retinotopic) representations and competing (rather than cooperating) with external sensory representations.

Strengths:

This study continues an intriguing line of research about how default mode regions interact with sensory cortex. Demonstrating that there are structured interactions between these regions at rest, and that these interactions are in fact organized according to retinotopic location (as opposed to traditional views of representational format in the default mode network), provides a new framework for thinking about large-scale internal and external brain networks. The authors make use of a well-powered public dataset that allows for precise estimates of pRFs and individual-specific resting-state networks and develop a number of interesting analyses that characterize the relationships between DN and dATN voxels. The findings are exciting and could have a major impact on future studies in cognitive neuroimaging.

The authors mention that these findings could shed light on internal/external interactions such as "anticipatory saccades or memory-guided attention," which is true, though I would argue that constructing DN representations of external stimuli is in fact even more fundamental than these specific cases (e.g. see Barnett and Bellana, 2025, "Situation models and the default mode network"). The "highways" identified in this study could play a vital role in real-world perceptual processes that are constantly translating external input into internal mental models.

Weaknesses:

(1) The criterion used for defining voxels as retinotopic seems very liberal. The authors show that only 5% of voxels have R^2>0.14 in a null analysis and therefore define voxels with R^2>0.14 as retinotopic. Although all the networks in Fig 1C show voxel distributions that differ from the null, the number of false positives above R^2>0.14 seems problematic, especially for the DN positive pRFs (red distribution) and to a lesser extent the DN negative pRFs (blue distribution). From visual inspection of the plot, the false discovery rate (fraction of voxels labeled as retinotopic that are false positives) looks like it would be greater than 50% for the DN positive pRFs. The authors do show that the positive pRF voxels have above-chance consistency across runs and also show in a supplementary analysis (Fig S5) that applying a stricter R^2 criterion yields similar results. These help to mitigate this concern, providing evidence that there are true positive voxels in this set which are driving the effects.

(2) The claim that "voxel-level visual response profiles shape DN-dATN coupling during spontaneous resting-state activity" is well-supported for specific sub-groups of DN voxels, though it is unclear whether the overall DN-dATN correlation at rest is primarily driven by the pRF-tuned voxels investigated in this study.

(3) The event-triggered analysis is effective at testing the bidirectional relationship between DN and dATN, with high activity in either network triggering a response in the other network. However, it would be helpful to show more validation that these "events" are meaningful windows of time to study, and that 13 TRs a typical length of time that activity is elevated during one of these events.

(4) The framing of this paper relative to the authors past work, such as Steel et al. 2024 ("A retinotopic code structures the interaction between perception and memory systems") could be improved. The primary novelty here is that this paper examines resting-state data and individually defined whole-brain networks, showing that there are widespread spontaneous interactions between broad internal and external networks, but this distinction is not made explicit in the Introduction.

Reviewer #3 (Public review):

Summary:

This paper addresses an important question (relationship between DN and dATN, and the role of retinotopic coding) and uses a set of novel analyses.

Strengths:

Important question, novel analytical approaches (pRF-informed functional connectivity analysis).

Weaknesses:

Some of the analyses are not described with sufficient clarity, especially the final analysis related to Fig. 3.

Comments on revised version.

Related to my previous comment 3), the removal of the labels "bottom-up" and "top-down" in the final analysis is a big improvement. However, I still don't fully understand how the 10 most aligned pRFs and the 10 most anti-matched pRFs are selected. The methods section on this has some ambiguity: "the 10 with the smallest Euclidean distance in RF center (x,y)". Does this mean that these are the pRFs closest to fovea? If not, what is the Euclidean distance referring to? Likewise, I don't understand how the anti-matched voxels are selected. This makes the interpretation of Fig 3 difficult.

My previous comment about baseline activation was to compare the matched voxels with randomly selected voxels, instead of with anti-matched voxels. The authors responded that there was a technical difficulty with this.

Reviewer #1 (Public review):

Summary:

Forbes et al. developed an integrated approach to identify cis-regulatory elements (CREs) in the large (3.6 Gbp) genome of the crustacean Parhyale hawaiensis, addressing the challenge of pinpointing these regions among large regions of non-coding sequences. They combined ATAC-seq chromatin accessibility profiling (both bulk and single-nucleus) across embryonic and adult tissues with low-coverage genome sequencing of three congeneric species (P. aquilina, P. darvishi, P. plumicornis). Without assembling congener genomes, they mapped reads with low stringency to the P. hawaiensis reference, identifying about 55k conserved islands that overlap ATAC peaks more than expected by chance. This dual filter was used to select CRE candidates for transgenic reporter validation, yielding 6 functional elements (out of 11 tested) driving ubiquitous, neuronal, or muscle-specific expression, a major advance for non-model systems with large genomes.

Strengths:

Forbes et al. generated high-quality ATAC data across multiple scales. Using bulk ATAC-seq (from whole embryos, developing and adult legs) they identified tens of thousands of open chromatin peaks across the assembled P. hawaiensis large genome. Moreover, using single-nucleus ATAC-seq from adult legs, they could resolve differentially accessible chromatin profiles across more than 15 cell types previously identified by scRNA-seq, enabling cell-type-specific candidate selection.

Furthermore, their innovative low-coverage comparative genomics method mapped 0.46-6.4% of congener reads to P. hawaiensis without genome assembly, revealing hundreds of thousands of conserved non-coding islands, including about 55k showing conservation in all four species, far exceeding random expectation.

Using the developed approach, the authors could validate 6 (out of 11 candidates) reporter constructs, driving robust ubiquitous and tissue-specific expression, succeeding where prior promoter-only screening failed and providing immediately useful genetic tools for the Parhyale community.

Weaknesses:

The primary limitation is that functional CRE testing was performed only in P. hawaiensis. While the conservation maps provide a valuable resource for comparative analyses, functional validation in congener species was not performed, so the extent to which the identified CREs or the prioritization strategy can be functionally generalized across related species remains to be established.

The approach did not successfully identify developmental CREs among the candidates tested. None of the candidates selected using the combined ATAC-seq and conservation filtering drove reporter expression matching the expected endogenous patterns. The authors appropriately discuss possible technical and biological explanations.

Overall Assessment:

Forbes et al. fully succeed with their integrated approach to (1) generate an ATAC-seq atlas plus functional CRE discovery and (2) innovative low-coverage sequencing for conservation mapping in the large 3.6 Gbp genome of Parhyale hawaiensis. Their combination of ATAC-seq chromatin accessibility profiling (bulk and single-nucleus) across embryonic and adult tissues with low-coverage genome sequencing of three congeneric species (P. aquilina, P. darvishi, P. plumicornis), without congener genome assembly, drastically shrank the CRE search space. Using this approach, the authors could validate six out of 11 candidate transgenic reporters (ubiquitous, neuronal, and muscle-specific) where prior promoter-only screening failed.

The low-coverage mapping innovation cuts cost and labour while snATAC-seq provides cell-type resolution, making these resources valuable for building new genetic and imaging tools in Parhyale.

This compelling method also has the potential to enable labs with limited resources to identify and characterize regulatory elements in more non-model organisms, advancing our understanding of their evolution while establishing a scalable pipeline for large-genome systems.

Comments on revised version.

The authors have adequately addressed all my previous comments. I have no further specific suggestions or requests.

Reviewer #2 (Public review):

The manuscript by Forbes, Skafida, Karapidaki et al. concerns the in-silico identification of cis-regulatory elements (CREs) in large genomes using chromatin accessibility (ATAC-seq) and sequence conservation (genomic DNA sequencing) data. They exemplify this method by applying it to identify novel CREs in Parhyale hawaiensis, which they validated using reporter constructs.

The results are convincing and are well supported by the data and validations. Identified CREs are valuable for researchers interested in the regulation of the expression of genes they control.

The methodology on the whole is also valid, as suggested by the results and previous publications on various taxa. Sequence conservation, as stated by the authors, was long used as a method to identify regions of non-coding DNA with functional and evolutionary constraints. The same applies to ATAC-seq data, which has also been used as a proxy for functional regions in different animals such as sea urchins and amphioxus. The methodology proposed is likely to be successfully used by researchers working on a variety of experimental organisms.

The authors do not use existing genome assemblies and use short-read sequencing to identify conserved regions, and while it is not conceptually novel, such an approach is becoming more and more viable and useful considering the recent advances in next generation sequencing technology and the decrease in price of short-read sequencing.

The authors have addressed and discussed the limitations and weaknesses of the approach as well as explicitly indicated the advantages.

All in all, the authors provide a valid method to strengthen CRE identification via sequence conservation without the need of multiple complete close species genome assemblies, making it a compelling option for non-model organism research.

Reviewer #3 (Public review):

Summary:

Forbes et al. present a new approach for identifying cis-regulatory elements in large genomes. Using Parhyale hawaiensis, a crustacean with a large genome (~3.6 Gb, comparable in size to the human genome), the authors show that current methods for identifying cis-regulatory elements, effective in smaller genomes, are markedly inefficient in organisms with large genomes. To address this limitation, they combine bulk ATAC-seq and single-cell (sc) ATAC-seq to identify chromatin regions that are either ubiquitously accessible or specifically accessible in particular cell types. They further integrate comparative genomics across multiple Parhyale species (P. hawaiensis, P. aquilina, and P. darvishi), selected at appropriate phylogenetic distances (20-95 million years divergence), to pinpoint conserved open chromatin regions likely under functional constraint.

Using this strategy, the authors predict a set of ubiquitous and cell-type-specific cis-regulatory elements. Importantly, they validate these predictions using rigorous transgenic reporter assays, convincingly demonstrating that their approach can successfully identify functional regulatory elements where previous methods had failed.

Strengths:

The approach introduced by Forbes et al. is conceptually straightforward, efficient, and readily transferable to other organisms. The validation experiments show not only that a substantial proportion of the predicted elements are functional, but also that the method is capable of identifying both ubiquitous and cell-type-specific regulatory elements. Given that the identification of regulatory regions remains a major bottleneck in understanding the molecular mechanisms underlying processes of development and regeneration, this work has the potential to make a significant impact in developmental and regeneration biology, particularly for studies involving non-model organisms with large genomes.

An additional strength is the demonstration that only the genome of the focal species requires high-quality sequencing and assembly. In contrast, species used solely for comparative analysis can be sequenced at low coverage without assembly, substantially reducing costs and increasing the accessibility of the approach.

Weaknesses:

While the method is effective in identifying regulatory elements that are active ubiquitously or in differentiated cell types, it failed in detecting elements associated with developmentally regulated genes. This may be due to trivial reasons, such as very low level of expression of the selected genes. However, as acknowledged by the authors, it may also indicate inherent challenges in identifying regulatory elements associated with developmentally dynamic gene regulation, compared to those associated with genes expressed in differentiated cell types.

A second limitation, also acknowledged by the authors, is the absence of chromatin conformation capture data, which would help link distal regulatory elements to their target genes. This limitation may be particularly relevant for developmentally regulated genes, where long-range regulatory interactions may be critical.

Addressing these limitations will be an important direction for future work. Nonetheless, the approach as presented in this manuscript represents a key contribution that sets the stage for further methodological advances in the identification of cis-regulatory elements in large genomes.

Comments on revised version.

I am fully satisfied with the current version of the manuscript.

Reviewer #1 (Public review):

Summary:

The article is testing the relative advantages of plant lineages with differing ploidy and admixture across environmental gradients. The results show that intraspecific variation in ploidy and admixture between lineages impacts plant traits that may enable persistence and range expansion.

Strengths:

Suitable marker panel size and strong results that include attempts to analyse mixed ploidy level data which is a challenge.

Weaknesses:

The sample sizes of the common garden experiments are very low making it difficult to draw robust conclusions.

Reviewer #1 (Public review):

Summary:

The authors used single-nucleus RNA sequencing (snRNA-seq) to investigate accelerated tooth replacement following tooth plucking in cichlid fish. They analyzed four stages of regeneration using elegant and well-designed approaches to characterize cellular trajectories and interactions within the dental epithelium and mesenchyme during the accelerated replacement process. Their analyses identified cell type-specific gene expression profiles and intercellular signaling interactions associated with whole-tooth regeneration.

Strengths:

This is a highly interesting and thoughtfully executed study that provides compelling and convincing insights into the mechanisms underlying accelerated tooth regeneration.

Comments on revised version.

I noted in my initial review that "the manuscript currently lacks experimental validation of the single-nucleus RNA-seq data." In response, the authors have added a statement indicating that their cell-type annotations and pathway interpretations are supported by extensive prior experimental work in the cichlid tooth model, including histology, in situ hybridization, immunohistochemistry, and pharmacological perturbation of major developmental pathways. They have also acknowledged this limitation in the Study Limitations and Future Directions section, stating that direct experimental validation of the single-nucleus RNA-seq findings will be the focus of future studies.

The authors have carefully addressed my comments, particularly the Major Points (2), (3), and (4), as well as all of the Minor Points. I appreciate their efforts to further characterize the mesenchymal landscape surrounding the putative successional lamina and to provide additional evidence supporting the presence of a specialized stromal microenvironment associated with tooth regeneration. Overall, the revisions have substantially strengthened the manuscript.

Reviewer #2 (Public review):

Summary:

Mubeen and colleagues study the cellular basis of tooth regeneration in cichlid fish. Using an elegant tooth plunking strategy followed by single nucleus RNA-sequencing, the authors were hoping to achieve an atlas of cellular and transcriptional changes that occur within and between cells during whole tooth replacement.

Strengths:

The major strengths of the methods and results are high novelty in the approach in a vertebrate with continuous tooth replacement, the temporal analysis of analyzing at plucking and three later time points, the thorough and sophisticated analysis of the snRNA-seq data including the inferring of trajectories and signaling events, and the robust signal of transcriptional differences induced by tooth plucking.

Weaknesses:

The major weaknesses of the methods and results are no validation of any of the inferred cell types, no functional tests of whether any of the changes in signaling pathways affect the plucking-induced tooth replacement process, and perhaps no clear take-away message for biologists not necessarily interested in tooth replacement.

Conclusions:

The authors achieved their aims of identifying the changes in gene expression and cellular composition that occur during whole tooth replacement accelerated by plucking. Overall, the results support their conclusions, although some slight semantic qualifiers should probably be added (e.g. referring to "cell types" as "putative cell types").

The work should have high impact in the field of tooth and organ regeneration, and the novel methodological paradigm established here of accelerating tooth replacement three-fold by plucking has great promise for future follow up studies to further study this process. The work also could have strong impact by the computational methods used here to infer trajectories and signaling interactions. Specific pathways, genes, and cell types could be tested in other fish such as zebrafish to test function during tooth replacement.

The work is unique and interdisciplinary and also has significance by establishing that robust phenotypically plastic accelerations in regeneration rates occur upon tooth removal. There are very few studies like this one that combine genetic x environmental studies of regeneration. The result that three different species of cichlid fish that normally have very different tooth patterns all accelerate tooth replacement threefold upon tooth plucking also has significance in revealing a highly conserved plucking response.

Reviewer #3 (Public review):

Summary:

In this manuscript, Barré et al utilize the Gp1ba-Cre transgenic mouse model to build upon previous findings in a Pf4-Cre system to investigate the effects of individual and combined Shp1 and Shp2 deletion in megakaryocytes and platelets. They report decreased megakaryocyte maturation, macrothrombocytopenia, and increased blood loss primarily in association with the Shp1/Shp2 double-knockout condition. The authors further show that this phenotype appears to be driven primarily by Shp2 and implicate dysregulation of Tpo signaling and downstream Ras/MAPK pathways, including ERK1/2. They propose that Shp1 may be functioning through a distinct pathway that has yet to be identified, opening up areas for future study.

Strengths:

Overall, the experiments combine in vitro, in vivo, and ex vivo approaches and appear to have been carefully designed and carried out, with multiple technical and biological replicates where relevant. The authors make a compelling argument for using the Gp1ba-Cre as opposed to the Pf4-Cre system and demonstrate both the dose- and stage-dependent effects of Shp1 and Shp2 on megakaryopoiesis and thrombopoiesis. They find that Shp1 and Shp2 are required in late-stage megakaryocyte maturation and that even low levels of expression compared to baseline are likely sufficient to yield generally normal megakaryocytes. Their findings also lead to specific future directions, such as the mechanism by which Shp1 regulates megakaryopoiesis and thrombopoiesis that is distinct from Tpo-mediated signaling. Figure 8 is particularly effective in summarizing the different models and pathways presented.

Weaknesses:

The effects of Shp1 and Shp2 knockouts are described as "synergistic," but it is not always clear that the effects are synergistic vs. additive, especially as the specific mechanism by which Shp1 functions in megakaryocyte development has yet to be identified. On a more minor point, although a significant part of the introduction focuses on the role of Mpl signaling in human disease, there is ultimately limited reference to Mpl (although there is of course a strong focus on Tpo) and the potential clinical implications of the findings presented here.

Reviewer #1 (Public review):

Summary:

This study provides valuable evidence that hilar mossy cells play important roles in maintaining the structural organization of the dentate gyrus and regulating the maturation of adult-born granule cells. The evidence for the structural reorganization and for the accelerated dendritic maturation of adult-born granule cells is convincing: it rests on converging anatomical, viral tract-tracing, retroviral birth-dating, and electrophysiological measurements, with appropriate controls for viral spread, off-target CA3 expression, and axonal degeneration. Support for the study's broader interpretive claim - that the dentate circuit functionally compensates for mossy cell loss - is incomplete. That claim rests on two null results obtained under baseline conditions (home-cage cFos and PTZ seizure metrics) in small cohorts, without behavioral assessment and without a stimulus-driven activity readout, and the manuscript does not engage with published work showing that mossy cells regulate neural stem cell activation and are required for stimulus-evoked neurogenic and behavioral responses.

Strengths:

(1) The study is technically rigorous and employs multiple complementary approaches, including selective genetic manipulations, viral tracing, immunohistochemistry, retroviral labeling of adult-born neurons, electrophysiology, and anatomical analyses. The comparison between complete mossy cell ablation and chronic synaptic silencing is particularly powerful, allowing the authors to examine the significant role of mossy cells in structural and functional organization in the dentate gyrus.

(2) One of the most notable findings is the identification of a previously unrecognized collapse of the inner molecular layer following extensive mossy cell ablation. This observation substantially expands current understanding of dentate gyrus structural plasticity. The demonstration that adult-born granule cells undergo accelerated dendritic maturation after both mossy cell loss and silencing also provides important insight into how mossy cells regulate adult neurogenesis.

Weaknesses:

(1) The functional significance of the observed structural remodeling remains incompletely addressed. Mossy cells have been strongly implicated in pattern separation, spatial information, and emotional behavior, yet no behavioral analyses were conducted. Consequently, it remains unclear whether the dramatic anatomical changes observed following mossy cell ablation translate into meaningful behavioral alterations.

(2) The conclusion that the dentate gyrus exhibits remarkable homeostatic compensation is reasonable but remains indirect. Although cFos expression and PTZ-induced seizure susceptibility are unchanged despite altered E:I balance, the mechanisms responsible for maintaining network stability are not investigated. Additional analyses of inhibitory circuit remodeling or compensatory synaptic adaptations would strengthen this conclusion.

Reviewer #2 (Public review):

Summary:

The authors examine how hilar mossy cells (MCs) influence adult-born dentate granule cell (abDGC) maturation and dentate gyrus (DG) structural integrity. Using both MC ablation and chronic functional silencing, they find that lacking MC inputs accelerates early abDGC maturation without altering mature cellular or intrinsic properties. MC silencing specifically decreased inner molecular layer (IML) spine density, whereas MC ablation led to IML collapse and an increased E/I ratio. However, neither intervention altered overall network excitability (measured via c-Fos and seizure induction) or seizure thresholds. These results advance our understanding of DG circuit plasticity during neurodegeneration.

Strengths:

(1) The side-by-side comparison of ablation vs. silencing provides a clear distinction between structural synapse loss and functional inactivation.

(2) The multi-level analysis spanning structural anatomy, single-cell physiology, and network-level assays yields a rich, comprehensive dataset.

Weaknesses:

(1) Measuring composite E/I ratios without parsing isolated EPSCs and IPSCs limits direct evaluation of MC-driven excitatory inputs. Furthermore, electrical stimulation in the IML likely recruits local interneuron axons directly alongside MC fibers, complicating the attribution of these responses solely to feed-forward MC circuits.

(2) The dramatic structural reorganization and IML collapse observed following MC ablation make it difficult to attribute changes in the E/I ratio purely to functional synaptic remodeling rather than physical circuit distortion.

(3) Layer boundary shifts following MC ablation complicate the interpretation of site-specific spine density (Figure 4); without accounting for IML collapse, classifying spine loss purely by traditional layer boundaries rather than proximal vs. distal dendrites may obscure local structural changes.

(4) The convulsive dosing protocol used for the seizure threshold test lacks the sensitivity required to reveal subtle changes in excitability.

Reviewer #1 (Public review):

Summary:

This manuscript addresses how internally generated evaluative signals can arise during self-guided learning in the absence of external reward. Using zebra finch song learning as a model system, the authors propose that tutor-song memorization and vocal performance evaluation are not separate processes, but instead emerge from a shared local circuit that learns to predictively cancel tutor-song-related auditory input. The comparison across several candidate circuit architectures, the quantitative comparison to experimental calcium imaging data, and the decomposition of the learned recurrent connectivity into modes shaping the error landscape are all strong aspects of the work. The final demonstration that the learned error signal can guide a downstream reinforcement learning agent also provides a useful proof of principle.

Strengths:

The idea that tutor-song memorization and performance evaluation can emerge from a shared predictive-cancellation circuit is interesting, and the combination of circuit modeling, comparison to experimental data, and error-landscape analysis is compelling.

Weaknesses:

(1) A central conclusion of the manuscript is that the E→I→E model best matches experimental data. This establishes model fit, but it does not yet explain why E→I and I→E plasticity are important for tutor-song cancellation and error-signal formation. Does E→I plasticity primarily teach the inhibitory population to represent tutor-song-related excitatory activity? Does I→E plasticity then implement the negative image required to cancel expected excitatory responses? Does the closed E/I loop primarily control gain, shift the minimum of the error landscape, or both? A useful analysis would be to compare models in which only E→I synapses are plastic, only I→E synapses are plastic, both are plastic, or neither is plastic.

(2) The analysis in Fig. 5 does not yet explain how the identified modes arise from the specific E→I/I→E plasticity mechanism. For example, are the landscape modes mainly produced by E/I gain-control dynamics? Are the memory modes related to an inhibitory negative image of the tutor song? Are these modes localized to particular blocks of the recurrent connectivity, such as E→I or I→E weights, or are they distributed across the full network?

(3) The manuscript emphasizes the emergence of sparse population error codes. However, in Fig. 6, the downstream actor-critic model uses the population mean excitatory activity as a scalar negative reward. This compresses the high-dimensional sparse population response into a single scalar. If the downstream system only uses the mean response, why is a sparse high-dimensional error code functionally important, beyond matching the observed response distribution? Conversely, if the sparse population pattern contains richer information about the direction or structure of vocal errors, how might downstream reinforcement pathways read out this information?

The manuscript should clarify whether sparsity is proposed to have a functional role in motor learning, or whether it is primarily a biological feature of the evaluative circuit. This point is particularly important because the broader framing of the paper concerns internal evaluative signals, whereas the final reinforcement learning demonstration uses a scalar reward.

(4) The actor-critic model in Fig. 6 is useful because it demonstrates that the learned error signal contains enough information to guide motor learning. However, the reinforcement learning module is attached downstream of the auditory circuit and is highly simplified. Therefore, it remains somewhat ambiguous whether Fig. 6 should be interpreted as a circuit model of song learning or as a demonstration of sufficiency. The latter interpretation seems appropriate and valuable, but the manuscript should state this more explicitly. The central contribution appears to be the bootstrapping of an internal evaluative signal, rather than a complete model of sensorimotor song learning. Clarifying this distinction would prevent overinterpretation of the actor-critic results.

Reviewer #2 (Public review):

Summary:

The paper proposes a network model that explains how birdsong learning can be guided by reinforcement signals.

Strengths:

It is well known that self-generated motor actions typically suppress their associated sensory input (for example, in the mammalian auditory cortex; see Eliades & Wang, 2003). This study presents a mechanism that effectively reverses this process. The theory posits that, initially, the motor signal generated in HVC, although not yet sufficient to produce an accurate song, nevertheless sends an efference copy to auditory areas, where it acts to cancel external auditory input from the tutor. This establishes a "scaffold," such that only an accurate replica of the tutor song can successfully suppress the corresponding auditory activity.

During learning, poorly generated plastic songs produce residual auditory activity that cannot be fully suppressed. This remaining activity then serves as an error signal that guides the refinement of motor output. The idea is elegant and is supported by experimental evidence.

Weaknesses:

The authors compare several possible sites of synaptic plasticity within the auditory network and conclude that the E-to-I-to-E model provides the best fit to the existing data. In this model, the auditory network consists of recurrent excitatory (E) and inhibitory (I) neurons, and Hebbian plasticity at E-to-I and I-to-E synapses is required to establish the cancellation pattern necessary to reproduce the tutor song.

However, the manuscript's presentation of the underlying plasticity mechanisms is somewhat puzzling. The authors repeatedly emphasize anti-Hebbian learning, even though their most successful model fundamentally relies on Hebbian plasticity. Although the resulting functional relationship may be described as anti-Hebbian, the biological learning mechanism implemented in the model is Hebbian. The repeated emphasis on anti-Hebbian learning therefore distracts from the central message and may confuse readers about the actual mechanism responsible for learning.

This emphasis may reflect an effort to distinguish the present work from previous anti-Hebbian models, but I suggest restructuring the manuscript. The authors should first present the optimal E-to-I-to-E model in detail, clearly explaining its mechanism and biological interpretation. Subsequent sections could then compare this model with the less successful alternative architectures. Such a reorganization would substantially improve the clarity and overall structure of the manuscript.

Finally, the abstract presents self-guided reinforcement learning as a novel concept, although this general idea has been described in previous work (e.g., Fiete et al., 2007). The abstract should therefore be revised to more precisely identify the specific novelty and contribution of the present study, rather than attributing novelty to the broader concept of self-guided reinforcement learning.

Reviewer #1 (Public review):

Summary:

Marschall et al. develop a theoretical framework for analyzing multi-task dynamics in nonlinear recurrent neural networks (RNNs). In the RNN model, recurrent connectivity is a linear superposition of multiple non-overlapping low-rank components, each corresponding to a separate "task". Each task is an autonomous dynamical system that does not incorporate external inputs. The "multi-task computation" setting examines whether multiple tasks (dynamical systems) can operate concurrently or how a network can switch between tasks.

Within this framework, the authors show that when connectivity consists of two low-rank components implementing two tasks (a limit cycle and a bistable attractor), the network exhibits winner-takes-all dynamics, resulting in only one task being active and the other one suppressed. A task consistently dominates this competition when the magnitude of its low-rank connectivity ("task strength") exceeds that of the other task. When one task is dominant and many other tasks with weaker low-rank connectivity are present, increasing the number of tasks destabilizes the dominant-task dynamics, leading to chaotic fluctuations in network activity. Supported by the dynamical mean-field theory analysis, the authors show that as the dominant-task strength increases, the network transitions through three dynamical regimes: chaotic spontaneous activity, chaotic task-selected dynamics, and non-chaotic task-selected dynamics. The theoretical analysis additionally predicts how latent task-related dynamics manifest in single-neuron activity and how the dimensionality of population activity changes across the three dynamical regimes.

The results are interesting, the analyses and simulations are rigorous, and the text is clear and easy to follow. Overall, this study is a significant and timely contribution to the literature on low-rank RNNs, an influential model class for low-dimensional neural dynamics in computational neuroscience.

Main comments:

(1) In this modeling framework, only one dominant task can be selected while all other tasks are suppressed. In contrast, several previous studies constructed RNNs (either through gradient-descent optimization or reservoir computing) that simultaneously generate outputs for multiple tasks across the corresponding task-specific readouts. Of course, what counts as a task is arbitrary, and one could view the dynamics of a reservoir network as implementing a single high-dimensional "task" with multiple readouts. Nevertheless, it would be helpful to explicitly clarify the distinction and similarities between the current modeling framework and networks that simultaneously solve multiple tasks.

(2) The role of external inputs in task selection appears to be underdeveloped. It is only briefly examined in Fig. S4, with the conclusion that external inputs aligned with the task subspace cannot enable selection of the desired task. However, previous multi-task RNN models (e.g., optimized through gradient descent) are clearly able to switch across many tasks using external inputs. In these models, external inputs modulate RNN activity along specific directions, shaped through gradient descent, to select the relevant task for each input. In contrast, this study only considers inputs aligned with the m-direction (left loading vector in the low-rank connectivity for a task). Such input cannot enhance the corresponding task's activity through recurrent amplification (Fig. S4). Yet, low-rank RNN theory predicts that inputs aligned with the n-direction (task's right loading vector) are selectively amplified by the recurrent dynamics. Why inputs aligned with the n-direction were not examined? More broadly, focusing only on inputs aligned with either m- or n-direction appears too narrow, as trained multi-task RNNs indicate that task selection through external inputs is possible, but may require input directions different from m and n.

(3) It is unclear what the reason is for the task interference: overlaps between task loading vectors for different tasks or other nonlinear effects? This issue is especially prominent in the analysis of task capacity (Fig. 2D and Fig. 4A). The text states that the loading vectors are independent across tasks, i.e. there is zero expected overlap between loading vectors of different tasks (line 126). For a network of N neurons, a rank-R task requires 2R loading vectors. Under the assumption of independence, the maximum number of tasks is P = N/(2R). Then α=P/N can be at most 1/(2R). In the simulations, it is chosen R=2, such that alpha can be at most 1/4 for the loading vectors to remain independent. However, the range of alpha reaches up to 1 in Fig. 2D and Fig. 4A, suggesting that loading vectors are no longer independent across tasks for larger alpha. Is the linear dependence between loading vectors (i.e. overlap across tasks) the reason for the task-1 component norm to drop sharply around alpha=1/4 in Fig. 2D? More broadly, is it possible to isolate the contribution of overlap in task loading vectors versus other nonlinear effects?

(4) In the current version of the paper, it is difficult to understand how the analysis based on the dynamical mean-field theory in Methods explains the key observations from the numerical simulations. For example, the mechanism underlying the transition from the spontaneous state to chaotic task-selected state, and to the non-chaotic task-selected state remain opaque. Further, it would be helpful to specify which section in Methods is being referred to at each mention throughout the main text. Finally, Fig. 7 and Eqs. (11-13) clearly state the input sources that drive single unit activity, non-dominant task latent states and dominant task latent state. The decomposition is potentially very informative, but its implications are not discussed in sufficient detail. It would be helpful to provide an intuitive explanation of how contributions of these input sources evolve as connectivity strength of one task increases, and which of them eventually leads to the loss of stability of the previous network state.

(5) The text states that the results can be easily extended to include task-specific inputs and outputs (lines 117-119). However, such an extension does not seem to be straightforward and requires additional explanations. The dynamical mean-field theory analyses here are stationary and describe the steady-state of network dynamics. In contrast, common input-output tasks typically involve transient dynamics, in which time-dependent external inputs keep changing the RNN flow field, and steady-state is never reached. Under these transient conditions, it is unclear whether the same conclusions apply. For example, if a network is at a low-activity baseline when a task-input begins to drive activity in the corresponding task subspace, it is unclear whether the activity in other task-subspaces would grow sufficiently fast to cause interference, or whether such interference would not be observed. Thus, a more detail analyses are necessary to support the extension of the results to input-driven transient tasks beyond autonomous dynamical systems.

(6) When task strength is the same for all tasks, what determines which task will win the competition? Is it frozen noise in connectivity such that one task always wins, or do initial conditions determine which task wins, based on which task's activity grows faster?

(7) Does the theory require the activity of all neurons to operate in the saturating part of nonlinearity? For example, the text states "increased activity reduces the gain factor <Φ'(t)>" (line 169). This statement is only true when Φ'<0. For sigmoid nonlinearity used in the paper, Φ'>0 when firing rate is small. If a substantial fraction of neurons in the network is near the rest state, would the theory still apply? Similarly, this statement does not hold for ReLU nonlinearity, and it is unclear how the theory applies to ReLU networks in Fig. S3. The paper states that the results are not specific to the choice of nonlinearity (lines 176-178). However, the dynamics being studied (bistability and limit cycles) both operate on the saturating part of the nonlinearity. Could the authors clarify the assumptions on the nonlinearity for the theory to apply?

(8) Is it possible to interpret the results in Fig. 4? What does this dependence on the overlap matrix mean? Is there an intuition for this particular dependence, or is it just an observation without general interpretation?

(9) The results in Fig. 5E appear underdeveloped and somewhat arbitrary. It is unclear how the dependence of dimensionality on the recording time would change as a function of time spent in a task. If this time is long, then the curve grows slowly and total dimensionality is high. If this time is very short, then the network may not have sufficient time for all task variables to grow sufficiently large to contribute significantly to the total variance. Thus, the grows may be faster and the total variance may saturate at a lower value. Hence, it is unclear whether there will be always a qualitative difference from the spontaneous activity curve. Furthermore, since only one of two curves is measured, what quantitative criteria should be used to determine whether it is consistent with task switching or spontaneous state?

(10) On line 368: "For sufficiently large number of tasks, the dimensionality associated with sequential task selection can greatly exceed that of the spontaneous state (Fig. 5E inset)" - it seems that Fig. 5E inset shows the opposite that the dimensionality of spontaneous state can saturate at a very high value for large N, exceeding the dimension of task-switching network in the main plot. Although it is hard to say, since the inset has many lines with only two labels and no ticks on axis, so it is unclear what exactly does it show.

(11) Related to discussion on line 368-370: In a task-switching state, would the switching between tasks also be reflected in behavior? In addition, the time-correlation functions would not be stationary in task-switching state, i.e. they would change over time, whereas they will be stationary in the spontaneous state. Thus, could the two mechanisms be dissociated in experiments using behavior or metrics beyond dimensionality?

(12) On line 385, could the authors provide more details on how they envision the two potential mechanisms-synaptic plasticity and targeted neuromodulation-to reinforce a task-specific low-rank connectivity pattern? If neuromodulation changes the gain of individual neurons, this modulation corresponds to scaling of connectivity by a diagonal matrix, not strengthening of a specific low-rank component. Short-term facilitation or depression also modulates synaptic strength depending on the activity of the presynaptic neuron, thus also scaling connectivity by a diagonal matrix. It is unclear how these two mechanisms could produce a specific low-rank modulation.

Reviewer #2 (Public review):

Summary

The authors ask what recurrent connectivity supports many distinct task-related manifolds when the associated dynamics interfere, how a circuit engages one task while suppressing others, and what produces high-dimensional activity. Extending previous theoretical studies on low-dimensional dynamics in large networks, they use a solvable model whose weight matrix is a weighted sum of many low-rank, task-specific components and develop a dynamical mean-field theory that relates connectivity, dynamics, and measurable population signatures of multi-tasking.

Strengths

(1) The question is timely. Low-rank networks are a leading model for low-dimensional latent dynamics, and the composition of dynamical systems has been proposed as a mechanism allowing for rapid, flexible learning; the paper connects these two ideas under a single theoretical framework.

(2) The proposal that sequential transitions between low-dimensional, low-rank dynamics can account for the _apparent growth of dimensionality with recording time_ is novel and is the paper's most valuable conceptual contribution.

(3) The mathematical analysis is rigorous, and the spontaneous-state theory is convincingly validated against simulation.

(4) The model produces concrete, falsifiable predictions - heterogeneous, syllable-dependent single-neuron tuning, low within-state dimensionality despite single-neuron variability, and distinct dimensionality-versus-recording-time signatures for the spontaneous versus task-switching accounts.

Weaknesses - whether the claims are supported by the data

(1) Chaos is named but not demonstrated._ The large-P and intermediate task-selected regimes are labeled "chaotic," but the manuscript does not establish chaos. In a homogeneous network, it is known that once the fixed point loses stability, the surviving solution is chaotic (Sompolinsky, Crisanti & Sommers 1988); that guarantee does not transfer here. The DMFT noise term is not computed analytically, and the single-neuron correlation functions (Fig. 5) show disorder, not a demonstrated decay of the fluctuation autocorrelation to zero, nor a positive largest Lyapunov exponent. The concern is sharpened by the possibility of _transient_ chaos: orthogonal to a dominant limit cycle, fluctuations may be locally unstable only at certain amplitudes or phases, so the global attractor could remain a stable cycle visited with chaotic excursions. As it stands, the claim of chaos in the intermediate regime is unsupported; it may well hold for some range of the selected-task strength, but this is neither shown numerically nor proven.

(2) "Analytical theory" overstates what is solved in closed form._ For the task-selected state, the kernels are non-stationary: The DMFT is entrained to the dominant task's dynamics, with an O(1) time-dependent quantity inside the nonlinearity. To my knowledge, there is no closed-form DMFT solution under these conditions. The Methods section supports this, explaining that the general scheme is solved by iterative numerical self-consistency (and described there as prohibitively expensive), and tractability is recovered only in a special block-Haar ensemble with Gaussian currents. This is entirely reasonable, but the main text presents it as an analytical theory; the reliance on numerical solutions of the self-consistency equations should be stated plainly.

(3) The spontaneous-state transition is the classical critical-gain transition, only reparametrized._ The onset of the no-task-dominant state is governed by $g_{eff}^2 = \alpha R\langle D^2\rangle$. It appears to depend on the number of tasks only because per-task strength D is held fixed as tasks accumulate; under a normalization that holds g_eff fixed, the transition reduces to a critical-gain point independent of P, as in extensive random networks. Relatedly, the result that chaos "arises solely from learning many tasks" is, mechanistically, random-network chaos: the random task components raise the weight variance and play the role of effective disorder. This is a legitimate and appealing reframing, but it is not a new transition, and the manuscript should make the relationship to the standard criterion explicit.

(4) Significance of the selection mechanism._ That boosting a task's gain selects it is intuitive, and the authors note the extreme (one $D^\mu$ dominating) is trivial. The non-trivial and genuinely useful contribution is quantitative - that only a small, O(1/P) modulation near criticality is required. This deserves to be foregrounded rather than left to the Discussion.

Reviewer #1 (Public review):

Summary:

The authors present MiPS, a platform combining DMD-based patterned illumination, automated microscopy, retrained DeLTA segmentation, and mother-machine microfluidics to selectively inhibit or eliminate cells based on dynamic phenotypes. The system enables targeted UV or red-light illumination in real time using segmentation-informed projection masks, allowing selective enrichment directly within mother-machine devices. The manuscript demonstrates proof-of-concept enrichment of mCherry cells from mixed GFP/mCherry populations, characterizes off-target effects, and performs computational simulations of iterative enrichment rounds. Overall, the engineering and systems integration are impressive, and the platform has strong potential for applications in directed evolution, biosensor optimization, and dynamic phenotype-based selection workflows.

Overall, I believe the work is suitable for publication after minor revisions and clarification of several aspects of the manuscript. In particular, the paper would benefit from additional context in the Introduction and Methods sections, clearer positioning relative to existing platforms, improved figure readability/captions, and a more careful revision of the English throughout the manuscript.

Major comments:

(1) The manuscript should better position MiPS relative to recent microscopy-based and DMD-enabled selection/control systems, particularly Lugagne et al., Nature Communications (2024), DOI: 10.1038/s41467-024-46361-1. That work also combines mother-machine microfluidics, DeLTA-based real-time image analysis, and DMD projection. The key distinction here appears to be physical selection/enrichment through targeted killing rather than optogenetic control, and this difference should be stated more explicitly.

(2) The manuscript currently compares MiPS mostly to FACS/MACS. However, the more relevant comparison may be recent image-based and microfluidic photoselection systems. A dedicated comparison table discussing throughput, temporal phenotyping, iterative selection, dynamic phenotype tracking, and enrichment capabilities would strengthen the paper.

(3) The enrichment experiment in Figure 4 represents a relatively simple classification problem (GFP vs mCherry). Since the proposed applications involve subtle continuous phenotypes, it would considerably strengthen the manuscript to include at least one experiment selecting for high vs. low expressors within a single fluorescent reporter population.

(4) The strongest enrichment result (~170-fold enrichment in Figure 5) is entirely simulation-based. Since the manuscript already states that ~45 min is sufficient between rounds for growth evaluation, a real 2-3-round enrichment experiment seems feasible and would substantially strengthen the platform's practical relevance. This experiment appears realistic within a relatively short time investment.

(5) The bimodal distributions in Figure 2 suggest that a fraction of cells may be stress-resistant rather than simply surviving randomly. It would be useful to discuss whether repeated rounds could progressively enrich UV-resistant subpopulations.

(6) The manuscript repeatedly uses the term "killed," although the data shown in Figures 2 and 4 mostly demonstrate strong growth arrest/inhibition. Please clarify how the cutoff of division rate <0.4 h⁻¹ was selected and whether an independent viability assay was performed.

(7) The off-target analysis in Figure 3 is one of the strongest parts of the paper and should probably be emphasized more. The conclusion that the dominant effects are global rather than local is interesting, but additional discussion about optical scattering, ROS diffusion, or device-wide coupling effects would strengthen the interpretation.

(8) UV exposure is inherently mutagenic in E. coli, and untargeted cells still receive a substantial fraction of the UV dose at high targeting fractions. Please discuss whether the MB/red-light modality may be preferable in applications where preserving genotype integrity is important.

(9) The manuscript discusses that methylene blue (MB) improves the on:off target ratio, but MB also appears to reduce baseline growth by ~40% even without red-light exposure. This is potentially important for iterative selection workflows. Please discuss whether this effect is reversible after washout and how rapidly cells recover.

(10) The manuscript states that the retrained DeLTA model used ~3,000 annotated fluorescence images, but no train/validation/test split or segmentation performance metrics are reported. Since segmentation directly impacts phenotype classification and projection targeting, these details are important for reproducibility.

(11) The manuscript would benefit from a stronger Methods description regarding DMD calibration, alignment procedures, projection accuracy validation, and computational timing requirements for the real-time analysis pipeline.

Significance:

General assessment: This is a creative and technically impressive study that combines mother-machine microfluidics, automated microscopy, real-time image analysis, and DMD-based photoselection into a unified platform for dynamic, phenotype-based enrichment. The strongest aspects of the work are the systems integration, the quantitative characterization of off-target effects, and the conceptual demonstration that dynamic microscopy-derived phenotypes can be linked to physical enrichment workflows.

The main limitations are that the biological validation remains largely proof-of-concept and the most compelling enrichment results are currently simulation-based rather than experimentally demonstrated across multiple rounds. In addition, the manuscript would benefit from stronger positioning relative to recent image-based and DMD-enabled microfluidic control systems.

Advance: The study extends the field of single-cell microfluidics and image-based selection by introducing a platform that links longitudinal microscopy measurements directly to physical enrichment decisions within mother-machine devices. To my knowledge, the combination of iterative feedback-driven selection, DMD-based targeted elimination, and dynamic phenotype tracking in this context is novel.

The closest related systems appear to be recent DMD-enabled mother-machine platforms for real-time optogenetic control, particularly those reported by Lugagne et al. (Nature Communications 2024, DOI: 10.1038/s41467-024-46361-1). However, MiPS introduces a distinct conceptual advance by using patterned illumination for selective enrichment/elimination rather than gene-expression modulation alone.

The advance is primarily technical and conceptual, with potential downstream applications in directed evolution, synthetic biology, biosensor engineering, and dynamic phenotype screening workflows that are difficult or impossible to implement using FACS alone.

Audience: The work will likely be of strongest interest to researchers working in synthetic biology, microfluidics, single-cell analysis, systems biology, bioengineering, and automated microscopy. It may also be of broader interest to communities developing dynamic phenotype screening technologies, closed-loop biological control systems, and next-generation directed evolution platforms.

The audience is likely specialized but multidisciplinary, spanning both engineering-oriented and biology-oriented researchers. The methods and conceptual framework may also influence future development of automated selection systems beyond the specific mother-machine context.

Expertise - My expertise includes: Microfluidics, Synthetic biology, Single-cell systems, Automated microscopy, Real-time image analysis, Bioengineering platforms, Dynamic phenotype characterization.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors reported Microscopic PhotoSelection (MiPS), a closed-loop automated robotic platform designed to link time-resolved imaging with physical sample recovery in mother machine microfluidic devices. By pairing a standard mother machine layout with a custom DMD optical path, an LED array, and an optimized DeLTA deep-learning model, the system tracks dynamic single-cell phenotypes and isolates specific cells via automated, targeted phototoxicity, i.e. selection by elimination. This is a novel technical development that addresses a clear limitation of snapshot sorting methods like FACS or MACS when screening for time-resolved, lineage-dependent traits. However, several methodological limitations and presentation errors must be addressed before publication.

Major Comments:

(1) Definition of 'Optimal' Dose (Figure 2D): The authors identify 8.0 W*cm-2 UV light for 300s as the optimal condition. However, this data point lies at the absolute boundary of the tested parameter space. In classical dose-response characterization, an optimum is defined by a local peak or a plateau followed by a decline in performance (typically due to rising off-target toxicity or scatter). Because the performance curve has not rolled over, this represents a boundary condition rather than a demonstrated mathematical optimum. The authors should either extend the parameter sweep to locate the true peak or soften their language to reflect that this is simply the highest performing condition tested.

(2) UV Exposure Time Gap: The exposure time sweep skips directly from 60s to 300s. While the closely spaced early timepoints are appropriate for capturing initial cell-death kinetics, the large gap to 300s leaves a significant engineering blind spot. Figure 3D demonstrates that off-target scattering damage scales linearly with cumulative light energy. If complete target cell arrest can be achieved at an intermediate exposure (e.g., 120s, 180s or 240s), operating the system at 300s unnecessarily subjects neighboring "surviving" cells to secondary global UV stress via device-wide scattering. An intermediate temporal sweep is recommended to optimize the selection window and properly balance target lethality with background library viability.

(3) Baseline Chemical Toxicity of Methylene Blue (MB): The photosensitizer workflow shows a clear improvement in contrast at lower power densities and exposure times. However, lines 151-153 note that the addition of 2 uM MB alone, even without light activation, stunts the baseline bacterial growth rate by ~40%. This is a major biological confounder. For applications like directed evolution or dynamic physiological screening, introducing a chemical stressor that nearly halves fitness imposes an unintended selective pressure. This baseline stress may activate pathways that mask or alter the phenotypes of interest. The authors must expand their discussion on how this baseline toxicity impacts multi-round iterative selections, and should ideally evaluate lower concentrations (e.g., 0.5uM or 1uM) or alternative photosensitizers to identify a more viable operational window.

(4) Negative Selection Framework and Search Space Scale: The MiPS platform relies entirely on negative selection by destroying unwanted variants. While effective for the demonstrated 1:1 binary proof-of-concept mixture, negative selection scales poorly when screening for rare variants within large libraries. For instance, isolating a single high performer from a library of 105 cells requires the system to successfully target and kill 99,999 individual cells; any statistical leak or failure in killing efficiency directly leads to heavy contamination of the recovered sample. The Discussion section requires a quantitative evaluation of these search space constraints, outlining how they limit the system's utility compared to positive selection mechanisms (such as optical tweezers or droplet sorters) when scaling to rare mutations (<1 in 104).

Significance:

This study presents a significant methodological advance in single-cell analysis and microfluidics by integrating long-term live-cell imaging, automated image analysis, and phenotype-guided cell recovery into a closed-loop platform. Existing approaches such as FACS and MACS are largely limited to endpoint or snapshot measurements, whereas MiPS enables selection based on dynamic and lineage-dependent cellular behaviors, thereby addressing an important gap in current single-cell screening technologies.

A key strength is the effective integration of mother machine microfluidics, custom optics, and deep-learning-based tracking into an automated and functional system. While the individual components are established, their combination into a phenotype-driven selection platform is innovative and expands the utility of live-cell microscopy from passive observation to active cell selection. The advance is therefore primarily methodological and technological, with potential to enable future conceptual discoveries in cellular heterogeneity and lineage dynamics.

However, limitations remain regarding scalability, robustness, selection accuracy, and generalizability across biological systems. Additional benchmarking and validation would strengthen the work further.

Overall, the study will be of interest to researchers in microfluidics, single-cell biology, microbial systems biology, bioengineering, quantitative imaging, and synthetic biology.

My expertise is in microfluidics, cell sorting and disease mechanobiology.

Reviewer #3 (Public review):

Summary:

The work describes an optofluidic automation setup to optically inhibit and enrich selected bacterial populations in confined microchannels through negative selection using light stimulation. The work is well described and the manuscript is well constructed.

Major comment:

The authors reported that methylene blue with 2uM incubation has superior performance than UV light. But it's also noted on line 152 there is an inhibition effect from the chemical affecting ~40% of the growth rate.

It will be noteworthy what is the growth curve or at least the MIC of methylene blue used on the MG1655 E. coli by the authors.

Significance:

The optics part of the work is well described, however the materials and methods details of the biological and microfluidic part can be extended.

Overall, the system demonstrated the practical use of combining microfluidics for enrichment of microbial population as a novel alternative method, despite that the efficiency is currently subpar to conventional methods.

But combining further with deep learning phenotype or growth rate monitoring, the technology represents a new path for phenotypic selection which is also novel that conventional methods cannot offer. The work will benefit readers in applied science seeking for new target enrichment based on optofluidics.

Reviewer #1 (Public review):

Summary:

The study by McKim et al (eLife-RP-RA-2024-102684) 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 its 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 shortcomings of individual datasets. The study also acknowledges the limitations that inherent to studying hormonal signaling and provide interpretations within the context of these limitations.

Reviewer #2 (Public review):

Summary:

The authors provide a comprehensive description of the neurosecretory network in the adult Drosophila brain. They assigned and verified the types of neurosecretory cells (NSCs) found in three publicly available drosophila brain connectomes. They then describe the organization of synaptic inputs and outputs for across NSC types. They show that NSCs are regulated by multiple sensory modalities, including enteric neurons. The authors then focus on a concise pathway from corazonin-expressing NSCs to a set of descending neurons, DNg27 and demonstrate that this pathway has the capacity to regulate egg-laying in female flies. Leveraging existing transcriptomic data, they also describe the hormone and receptor expressions in the NSCs and show putative paracrine signaling between NSCs. Taken together, this study provides a framework for future functional experiments, which may demonstrate whether and how NSCs, and the circuits to which they belong, shape physiological function and behavior.

Strengths:

This study uses three Drosophila brain connectomes to assign cell types to ten classes of 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 adult 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 and generate predictions regarding sensory inputs and downstream physiological and behavioral functions. They test one newly identified NSC-premotor pathway, from corazonin-expressing NSCs to the descending neuron DNg27, with loss-of-function experiments and demonstrate that this pathway has the capacity to regulate female egg-laying.

The authors illustrate expression patterns of neuropeptides and receptors across NSC cell types from existing transcriptomic data and present a putative paracrine signaling network among NSCs. The authors also catalog hormone receptor expression across tissues.

Taken together, this study provides a comprehensive account of the neurosecretory system of the adult fly.

Weaknesses:

In Figure 6 authors use a linear dynamical modeling approach (described in Bates et al. 2026) to quantify the influence of different sensory source neuron types on the different NSC classes. The authors should discuss the two main assumptions baked into this approach: 1) all path segments (connections) from sources to targets are given the same sign and therefore result in activation, despite likely biological variation in their synaptic valences. 2) Each connection is given the same time constant for the response kinetics. Therefore, the model assumes uniform intrinsic "biophysical" properties.

Although the actual intrinsic properties (e.g. complements of voltage-gated ion channels) of the intermediate and target neurons are unknown, they are likely heterogenous. Such heterogeneity would have consequences on the steady-state responses. Thus, the response magnitudes measured in this model are unlikely to provide an accurate representation of feedforward "influences" in this circuit.

Although the intrinsic properties of all nodes in these paths will remain unknown in the absence of electrophysiological recordings, one could still consider the signs of connections using neurotransmitter predictions in the connectome (Eckstein et al. 2024). It would then be useful to compare the relative influences calculated with the Bates et al. approach to 1) simple weight propagation methods which are agnostic to time (as in Hoeller et al. 2026; doi: https://doi.org/10.64898/2025.12.22.696097) and 2) this Bates et al. approach and weight propagation methods that conserve the signs of the connections.

In Figure 8 and associated supplements, the authors probe the function of CRZ-expressing NSCs > DNg27 pathways in female and male flies. Although the authors test the effects of silencing both CRZ-expressing cells and DNg27 on feeding, egg-laying, and flight behaviors in females. They recapitulate a previous finding that CRZ-expressing cells regulate feeding behavior and then identify potential regulatory roles for this pathway in egg-laying. However, the authors did not test this full palette of behaviors in males. The authors do not test feeding or flight behaviors in males. They do, however, confirm previously reported activation phenotypes (copulation-like behaviors), via optogenetic activation of CRZ-expressing cells in males. These experiments would be more ethological if executed in freely walking male flies, rather than males that were glued, on their backs. It is unclear why the authors did not also test for activation or loss-of-function phenotypes for DNg27 in males. Taken together: the authors show compelling loss-of-function phenotypes for feeding and egg-laying for the CRZ-expressing NSC > DNg27 pathway in females, but evaluation in males remains incomplete.

Reviewer #1 (Public review):

Summary:

This study extends the authors' prior work on transgenic nhomie/homie boundary pairing (Fujioka et al. 2016 PLoS Genetics), which showed that these elements - corresponding to the eve TAD's left and right boundaries - can pair with endogenous copies over large genomic distances (142 kb here), bridging a linked reporter gene to endogenous eve enhancers for long-range gene activation (shown again here in Figure 1). Physical pairing was previously confirmed (Chen et al. 2018 Nat Genet) and further resolved by Micro-C (Bing et al. 2024 eLife), supporting the hypothesized "stem-loop" or "circle loop" topologies used to explain homie/nhomie directional pairing (shown here for nhomie in Figures 2-3). The authors recently showed that a Su(Hw) binding site is required for homie-mediated reporter gene activation by eve enhancers (Fujioka et al. 2025 Genetics); here, they extend this finding to nhomie (Figures 4-6), further showing that Su(Hw) motifs are required for Micro-C-detectable looping between transgenic and endogenous eve boundaries (Figures 7-9). Finally, they show that while cis-pairing over 142 kb is highly specific to homie/nhomie elements, transvection between homologous transgene insertions is more permissive (functioning with the Su(Hw)-bound gypsy insulator) but still shows some specificity (failing with the CTCF-bound Fab8 boundary, Figures 10-11).

Strengths:

The question of how pairs of loci can specifically physically pair over relatively long genomic distances is an interesting fundamental question. The study's strengths are the clarity and meticulous interpretation of the results, and the authors' conclusions are compelling.

Weaknesses:

A major weakness is that some figures reproduce previously published findings; in some cases it is unclear whether the same fly lines were used, and in others, the lines differ only slightly from those used previously (e.g., a shorter version of the Homie transgene than the one used previously). Most conclusions in this manuscript have already been published elsewhere. As a result, the paper does not report a genuine new discovery, and only incrementally advances our understanding of boundary pairing.

Reviewer #2 (Public review):

The results in Ke et al., build on 15 years of work focused on dissecting the pairing properties of the Drosophila Homie insulator. Here, the authors use similar methods to those shown in Fujioka et al., 2016, Ke et al., 2024, and Fujioka et al., 2025, but with a focus on nHomie pairing and the role of Su(Hw) in both Homie and nHomie long-range interactions. The main question the authors hope to address is what the mechanisms are behind the physical interactions involved in boundary:boundary pairing. They attempt to answer this question through mutating the Su(Hw) binding sites located within the nHomie and Homie transgenic sequences and observing how pairing is altered.

The work presented is thorough and thought out; however, some of the conclusions that the authors focus on are not what makes the work interesting and could be reprioritized. For example, the authors spend several paragraphs in the discussion (lines 531-595) addressing how the data presented does not support an argument for cohesion-mediated loop extrusion. While the interactions shown throughout the manuscript do not support cohesion-mediated loop extrusion occurring at the Homie locus, the authors have already made this point in both Bing et al., 2024 and Ke et al., 2024 and thus do not need to expound on this point.

Instead, the authors have a more compelling story in their specificity vs promiscuity arguments. Homie is a unique insulator in Drosophila and even when located 142kb away will still find its unique pairing partners (itself and nHomie). The authors have shown this several times prior, yet here they show that some level of this long-distance homing interaction is dependent upon the Su(Hw) binding site. Additionally, the authors show in this study that addition of gypsy sequence, in a less demanding assay, is sufficient for transvection pairing with Homie. This transvection result is a novel finding, as gypsy was previously shown to be insufficient for long-distance pairing with Homie based on the authors' prior studies. It is likely different architectural proteins that bind within the Homie sequence and allow it to pair specifically with itself, regardless of assay type, and these elements are likely absent from the gypsy sequence, leading to pairing that is more situational (see point 8 in recommendations).

Finally, to no fault of the authors, the art of visualizing complex 3D pairing configurations is difficult. Unfortunately, that can at times mask the ultimate points that the authors are trying to make about pairing early in the manuscript.

Overall, the work mainly supports the authors' claims, and the findings are a useful addition to the insulator and Drosophila 3D genome organization field.

Reviewer #3 (Public review):

Summary:

This manuscript investigates the function of Su(Hw) binding sites found in two boundaries/insulators, homie and nhomie, in TAD formation that encompasses the eve gene. They tested the hypothesis that Su(Hw) binds homie and nhomie, thereby forming a stem-loop TAD. The authors used transgene reporters with various mutations, and the results support the hypothesis strongly.

Strengths:

They combine reporter assays (GFP and LacZ expression) with MicroC contact profiling to robustly support their conclusion. Overall, they propose how Su(Hw) mediates physical interaction between boundary elements (homie and nhomie).

Weaknesses:

The writing is quite dense and not easily accessible to outside readers.

Reviewer #1 (Public review):

Summary:

Mast cells have previously been reported to play an important role in bacterial immune defence and act protectively in sepsis. However, many of these findings were based on studies using Kit mutant mice. In this study, the authors conducted a detailed investigation using mast cell-deficient Cpa3 Cre-Master mice. As a result, the authors found that the Cpa3 Cre-Master mice exhibited responses similar to wild-type mice in terms of bacterial immune defense. This suggests that the observed phenotype is not due to mast cell-dependent bacterial immune defense, but rather is associated with dysbiosis of the gut microbiota.

Strengths:

Mast cells have long been reported to play an important role in the protective response against sepsis, and their function in infection defense has been demonstrated. However, Kit mutant mice have been reported to exhibit impaired peristalsis, and several mast cell-specific genetically modified mouse lines have since been developed and examined in detail. This study presents an important finding by logically demonstrating that the exacerbation of sepsis in Kit mice is due to alterations in the gut microbiota, and that the phenotype previously thought to be mast cell-dependent was, in fact, not.

In addition, the experiments were carefully designed using mice with matched genetic backgrounds. These findings underscore the importance of microbiota composition in interpreting immune phenotypes and highlight the need for co-housing controls in mutant mouse studies.

A major strength of this work is the robustness of the CLP data, generated over eight years by three independent researchers across two institutions with large sample sizes, lending strong support to the conclusions.

Weaknesses:

The study assesses only a limited subset of gut bacterial species, leaving the extent to which E. coli expansion contributes to the observed phenotype unclear. Moreover, in the cohousing experiments, there is no evidence provided to confirm successful microbiota normalization between groups. A more detailed analysis of the microbial composition would be necessary to strengthen the reliability of the findings.

It is also important to note that Cpa3-deficient mice exhibit not only mast cell depletion but also defects in basophils and T cells. These additional immunological alterations may counterbalance one another, potentially masking phenotypic changes and complicating interpretation.

Furthermore, it remains to be determined whether the altered gut microbiota observed in KitW/Wv mice is a consequence of impaired intestinal motility, whether a similar phenotype is observed in KitW-sh/W-sh mice, and whether comparable results occur in SCF-deficient models. Addressing these questions would provide greater clarity on the contribution of mast cells versus secondary factors in the observed phenotypes.

Given that KitW/Wv mice exhibit impaired peristalsis, is the observed increase in E. coli a consequence of this dysfunction?

Previous studies with BMMC reconstitution experiments have indicated that mast cells are a source of TNF-how does this align with the current findings?

Comments on revised version.

The authors have made substantial efforts to address the concerns raised in the previous review, and the revised manuscript has been substantially strengthened by the additional microbiome analyses. In particular, the new data provide a more comprehensive characterization of the intestinal microbiota in both Kit mutant and mast cell-deficient mice and strengthen the interpretation that the microbiota alterations observed in Kit mutant mice are associated with Kit deficiency rather than mast cell deficiency per se. Although Enterobacteriaceae were significantly increased based on the unadjusted Welch's t-test, this difference did not remain statistically significant after correction for multiple comparisons. The authors appropriately acknowledge this limitation in the revised manuscript. Overall, I consider the major concerns regarding the microbiome analysis to have been adequately addressed.

Reviewer #2 (Public review):

Summary:

The authors showed that the high susceptibility to CLP sepsis of Kit-mutant mice is not due to mast cell deficiency, but to dysbiosis.

Recommendations:

(1) The authors showed that E. coli increases in the cecum of Kit-mutant mice, which causes high CLP susceptibility. However, they did not provide any evidence E. coli is responsible for the high susceptibility. In the Figure 3 experiments, the authors administered the same number of cecal bacteria and did not show the number of E. coli after the administration. The authors should provide evidence showing that depletion of E. coli decreases susceptibility.

(2) The author should provide direct evidence of dysbiosis by, for example, shotgun sequencing of cecal and fecal contents.

(3) In case the authors find dysbiosis, they should analyze the mechanisms by which Kit mutation causes dysbiosis.

Comments on revised version.

The revised manuscript focuses on refuting the notion that mast cells play important roles in sepsis. The reviewer agrees with this claim.

Reviewer #3 (Public review):

In this work, Bryant, et al. investigate genetic interactions between non-essential members of the outer membrane protein biogenesis pathway and other genes in the genome using a transposon-directed insertion sequencing (TraDIS) approach in E. coli K-12. The authors identify interactions with other components of the envelope including LPS, peptidoglycan, and enterobacterial common antigen biogenesis, and they tie these interactions to specific members of the outer membrane biogenesis pathway. Although many of these interactions are known and have been previously investigated in the field, the study provides several synthetic phenotypes that could be useful for further investigations.

The strengths of the paper include the unbiased, TraDIS approach, and follow up on the interactions observed. The interactions with genes of unknown function also are of interest as they may suggest experiments to find the functions of these genes. Although the paper could better address the relation of its findings to existing literature, the work in the paper is well controlled and the findings will be of interest to the field.

Reviewer #1 (Public review):

Summary:

This study aimed to determine whether bacterial translation inhibitors affect mitochondria through the same mechanisms. Using mitoribosome profiling, the authors found that most antibiotics, except telithromycin, act similarly in both systems. These insights could help in the development of antibiotics with reduced mitochondrial toxicity.

They also identified potential novel mitochondrial translation events, proposing new initiation sites for MT-ND1 and MT-ND5. These insights not only challenge existing annotations but also open new avenues for research on mitochondrial function.

Strengths:

Ribosome profiling is a state-of-the-art method for monitoring the translatome at very high resolution. Using mitoribosome profiling, the authors convincingly demonstrate that most of the analyzed antibiotics act in the same way on both bacterial and mitochondrial ribosomes, except for telithromycin. Additionally, the authors report possible alternative translation events, raising new questions about the mechanisms behind mitochondrial initiation and start codon recognition in mammals.

Weaknesses:

All the weaknesses I previously highlighted were adequately addressed.

Reviewer #3 (Public review):

Summary:

Recently, the off-target activity of antibiotics on human mitoribosome has been paid more attention in the mitochondrial field. Hafner et al applied mitoribosome profiling to study the effect of antibiotics on protein translation in mitochondria as there are similarities between bacterial ribosome and mitoribosome. The authors conclude that some antibiotics act on mitochondrial translation initiation by the same mechanism as in bacteria. On the other hand, the authors showed that chloramphenicol, linezolid and telithromycin trap mitochondrial translation in a context-dependent manner. More interesting, during deep analysis of 5' end of ORF, the authors reported the alternative start codon for ND1 and ND5 proteins instead of previously known one. This is a novel finding in the field and it also provide another application of the technique to further study on mitochondrial translation.

Strengths:

This is the first study which applied mitoribosome profiling method to analyze multiple antibiotics treatment cells. The mitoribosome profiling method had been optimized carefully and has been suggested to be a novel method to study translation events in mitochondria. The manuscript is constructive and well-written.

Comments on revisions:

The authors added a discussion to the revised manuscript, and also carefully investigate structural data from others. I have no more comment. Congratulations to the team for a good manuscript!

Reviewer #1 (Public review):

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

Summary:

D. Fuller et al. set out to study the molecular partners that cooperate with ATG2A, a lipid transfer protein essential for phagophore elongation, during the process of autophagy. Through a series of experiments combining microscopy and biochemistry, the authors identify ARFGAP1 and Rab1A as components of early autophagic membranes, which accumulate at the periphery of aberrant pre-autophagosomal structures induced by loss of ATG2. While ARFGAP1 has no apparent function in autophagy, the authors show that RAB1A is implicated in autophagy, although the precise mechanisms are not explored in the manuscript.

Strengths:

The work presented by Fuller et al. provides new insights into the composition of early autophagic membranes. The authors provide a series of MS experiments identifying proteins in close proximity to ATG2A, which is a valuable dataset for the field. Furthermore, they show for the first time the interaction between ATG2A and RAB1A both in fed and starved conditions, which extends the characterisation of the pre-autophagosomal structures observed in ATG2 DKO cells.

Reviewer #2 (Public review):

The mechanisms governing autophagic membrane expansion remain incompletely understood. ATG2 is known to function as a lipid transfer protein critical for this process; however, how ATG2 is coordinated with the broader autophagic machinery and endomembrane systems has remained elusive. In this study, the authors employ an elegant proximity labeling approach and identify two ER-Golgi intermediate compartment (ERGIC)-localized proteins-Rab1 and ARFGAP1-as novel regulators of ATG2 during autophagic membrane expansion.

Their findings support a model in which autophagosome formation occurs within a specialized subdomain of the ER that is enriched in both ER exit sites (ERES) and ERGIC, providing valuable mechanistic insight. The overall study is well executed and offers an important contribution to our understanding of autophagy. I support its publication in eLife and offer the following minor comments for clarification and improvement.

Reviewer #3 (Public review):

The manuscript by Fuller et al describes a crosstalk between ARTG2A with components of the early secretory pathway, namely RAB1A and ARFGAP1. They show that ATG2A is recruited to membranes positive for RAB1A, which they also show to interact with ATG2A. In agreement with earlier findings by other groups, silencing RAB1A negatively affects autophagy. While ARFGAP1 was also found on ATG2A positive membranes, silencing ARFGAP1 had no impact autophagy. Notably, these ARFGAP1 positive membranes are not Golgi membranes.

The findings are interesting and the data are in general of good quality.

Comments on the previous version:

The revisions carried out by the authors are fine. The new data on ArfGAP1 and about the indirectness of the ATG2A and Rab1A interaction improve both clarity and strength of the manuscript. I have no further comments.

Reviewer #1 (Public review):

Summary:

King and colleagues generated a mouse with a point mutation in IL21R and investigate the influence on IL-21-mediated T and B cell activation and differentiation. They find that mutant mice show a reduced T and B cell response with CD4 T cell differentiation into T follicular helper cells being primarily affected.

Strengths:

The authors combine in vitro and in vivo analysis, including bone-marrow chimeric mice.

Weaknesses:

The effect of the IL21R EINS mutant does not specifically affect STAT1, as clearly shown in Figure 1 H, I. Particularly at lower doses of IL21 which may be more relevant in vivo, the effects are very similar. A second key weakness is the very small Tfh response, a not very clear PD-1 and CXCR5 staining to identify Tfh and a lack of a steady-state (prior to immunisation) comparison of Tfh numbers in the different mouse strains. The latter makes it impossible to know what fraction of the response is antigen-specific.

Comments on revised version.

Thank you for responding to some of my suggestions/comments.

Unfortunately, these responses do not address the concerns raised by the other reviewer and me (see 'weaknesses'). The argument that statistical analysis shows that there is no difference in p-STAT3/5 does not make any sense - it simply reflects the absence of sufficient statistical power.

No further experiments have been performed, and the authors didn't appropriately adjust the conclusions to reflect the limitations of the study.

To provide an example, the authors decided to keep the title 'An IL-21R hypomorph circumvents functional redundancy to define STAT1 signalling in germinal center responses' although both reviewers clearly highlighted that no conclusion about STAT1 can be made and the authors responded that 'We agree that further experiments are needed to definitively show that the effect is attributable to reduced STAT1 activation alone. Rescue experiments will be a focus of future experiments.'

These rescue experiments (and additional analysis of the Tfh response) are essential to allow firms conclusions to be made.

In its current form, the manuscript is unfortunately misleading.

Reviewer #2 (Public review):

Summary:

In the manuscript, "An IL-21R hypomorph circumvents functional redundancy to define STAT1 signaling in germinal center responses," Cecile King and colleagues identify a cytoplasmic site of the IL-21 receptor that differentially regulates STAT1 and STAT3 activation upon IL-21 stimulation. They further examine the immunological consequences of this site-specific alteration on Tfh differentiation and Tfh-dependent humoral immunity, raising important questions about how gene-knockout models may obscure nuanced functional roles of signaling molecules.

Strengths:

The study convincingly highlights a non-redundant role for STAT1 downstream of IL-21-IL-21R signaling in the Tfh differentiation pathway. This conclusion is supported by in vitro analyses of STAT1 and STAT3 activation in CD4 T cells stimulated with IL-21 or IL-6; by in vivo assessments of Tfh and germinal center B cell responses in WT and IL21R-EINS mutant mice, including bone-marrow chimera systems; and by investigating the expression of Tfh-related molecules in WT versus IL21R-EINS CD4 T cells.

Weaknesses:

Although the experiments were carefully executed with appropriate controls, a key question remains unresolved: whether the Tfh differentiation defect in IL21R-EINS mice is directly attributable to reduced STAT1 activation. Rescue experiments that restore STAT1 signaling in IL21R-EINS TCR-transgenic CD4 T cells would provide strong evidence linking the mutation to impaired STAT1 activation and, consequently, defective Tfh differentiation. Without such evidence, it remains formally possible that additional, uncharacterized mutations introduced during ENU mutagenesis contribute to the phenotypes observed, particularly given the discrepancies between IL21R knockout and IL21R-EINS mutant mice.

Comments on revised version.

The revised manuscript failed to address the key question, whether the Tfh differentiation defect in IL21R-EINS mice results from the reduced STAT1 activation in CD4 T cells.

Reviewer #1 (Public review):

Summary:

This manuscript is an excellent follow-up to your 2022 study, in which Sox17 expression was localized to the rete testis and shown to be required for proper formation of the Sertoli cell valve (transition region). By using Nr5a1-Cre to drive conditional deletion of Sox17 specifically in rete testis cells, you demonstrate that testis weights remain normal at 2 weeks of age but become significantly reduced by 8 weeks in Sox17-cKO males. At the later time point, the seminiferous epithelium is severely disrupted, with apparent arrest of spermiogenesis: the epididymal lumen is essentially devoid of sperm, and most tubules lack elongated spermatids.

Strengths:

Clearly shows the role of Sox17 in Sertoli cells being important to the SV function. The SV (transition region) between the rete testis and seminiferous tubules remains an understudied domain of testicular biology. The present work, together with your prior study, highlights intriguing mechanisms operating in this specialized niche.

Weaknesses:

The available data do not fully explain either the developmental assembly of the Sertoli valve or the precise consequences of its functional disruption. These studies are nonetheless valuable precisely because they raise more questions than they answer; the conceptual implications are thought-provoking.

Reviewer #2 (Public review):

This manuscript investigates the role of SOX17 in the formation and function of the Sertoli valve (SV) at the interface between seminiferous tubules and the rete testis (RT). Building on previous work showing that rete testis-specific deletion of Sox17 disrupts SV formation, leading to defective spermiogenesis and male infertility, the authors explore how SOX17 overexpression in Sertoli cells regulate SV of rodent testes.

Using transgenic mouse models with ectopic Sox17 expression in Sertoli cells, the study demonstrates that SOX17 is not only required but can also modulate SV formation. Ectopic expression in Sertoli cells induces expansion of the SV structure and partially rescues SV defects and spermatogenesis in RT-specific Sox17 conditional knockout animals. The data support a model in which SOX17 acts through paracrine signaling to regulate SV formation, although the precise mechanisms remain to be clarified.

Overall, this is a well-executed study with novel and significant findings. The ability to experimentally manipulate SV size is particularly compelling and provides a valuable framework to study fluid dynamics and epithelial interactions in the testis. This work will be of broad interest to the reproductive biology and developmental biology communities.

Reviewer #1 (Public review):

Summary:

T cells that recognize lipids-CD1c are frequent in circulation; however, their role in infection is unclear. This study aims to understand how Mtb infection can shape the responses of CD1c-specific T cells. CD1c is expressed in MTB granuloma, but in lower amount than in nearby inflamed tissue. Mtb infection downregulates the expression of CD1c on monocyte derived DCs. Single cell RNA sequencing revealed the cytotoxic program inherent to the lipids-CD1c-specific T cells. Using an in vitro APC system where CD1c expression remains intact upon Mtb infection, the authors suggest that these T cells react better to Mtb-infected than uninfected Cd1c-expressing APC and reduce Mtb burden in infected cells. Therefore, Cd1c downregulation could be an immune evasion strategy used by Mtb.

Strengths:

This study asks an important question. The single cell transcription analysis suggests the inherent cytotoxic program of lipid-CD1c cells and provides insights into their phenotypic and potential functional profiles. Function experiments suggest that these autoreactive T cells can react to Mtb infection, adding to the paradigm of infection control by these non-conventional T cell population. In summary, this study reports potential role of lipid-CD1c specific T cells in Mtb infection., and the evidence overall supports the conclusions.

Weaknesses:

(1) The study engineered THP-1 cells that lacked endogenous beta2 microglobulin (beta2m) and stably expressed chimeric CD1c-beta2m (CD1c-THP-1 cells). T cells isolated from healthy donors showed stimulation when mixed with CD1c-THP-1, suggesting the presence of CD1c reactive T cells in healthy individuals (Fig. 1). It appears that antibody staining profiles for different markers are not from the same experiment (Fig. 1). It is better to show all the representative antibody staining from a single experiment. And a quantification of MFI values from at least three independent experiments should be shown with statistics. Further, in each of the representative antibody profile, unstained/isotype control should be incorporated. The characterization of beta2m KO cells - using DNA sequencing and western blotting - should be shown in supplementary.

(2) Next, using immunohistochemistry, the study suggests expression of CD1c in lung biopsy from TB patients, and that CD1c expression is reduced on primary MoDCs upon Mtb infection (Fig. 2).

(3) To assess CD1c-reactive T cell cytotoxicity in Mtb infection, the authors generated CD1c-T cell lines (from T cells from healthy individuals) using two different approaches. In one, CD1c-specific T cells were expanded upon mixing with CD1c-THP-1 cells, were sorted using CD1c-endo tetramer, and expanded. In another approach, CD1c-specifc T cells were sorted (using CD1c-endo dextramer) directly from blood-derived T cells and were expanded. The line obtained using one of the approaches to assess cytotoxicity towards CD1c-THP-1 cells, uninfected or infected with Mtb. It is seen that CD1c-T cells exert cytotoxicity towards infected, and not towards uninfected, CD1c-THP-1 cells (Figs. 3,4).<br /> The authors should clearly indicate the T line, obtained from which one of the two approaches, was used for the cytotoxicity assay, and what was the result with the line obtained from other approach. Fig. S4 shows the sorting strategy and final profile of the T line obtained after CD1c-endo tetramer-based sorting and expansion. This figure is mistakenly referred as supplementary fig. 3 in the caption of Fig. 3; should be corrected. Further, to complete the Fig. S4, the intermediate step of sorting scheme, which shows negative and positive fraction as defined by streptamer binding, should be included. And Y-axis of left-most lower-most panel (no tetramer) should be labelled. The Profile of the T line shows three distinct population: tetramer-low, tetramer-intermediate, and tetramer-high. The authors should comment on this, particularly on the tetramerlow population that can be CD1c-negative cells that have non-specific weak binding to the tetramer.<br /> The profile of the T line obtained from the other approach should also be shown, and the experiments that used those should be clearly indicated.

(4) To assess whether the elevated response of CD1c-T cells is mediated by the T cell receptor (TCR) that they harbour, the authors performed single T cell sequencing for TCR. They could identify 11 αβ pairs, two of them (named as EM1 and EM2) constituted 10 out of the 11 αβ pairs. It is surprising that such a shallow sampling (despite of the filtering mentioned in Fig. S9) such a high degree of clonal expansion and yielded two context-relevant TCR sequences. These TCRs were indeed DC1c specific, as indicated by their ability to upregulate CD69 upon engagement with CD1c (Fig. 5C). However, they respond very weakly to CD1c-THP-1 cells - the difference (THP-1 KO versus CD1c-THP-1 cells) is small for EM1, and the upregulation of CD69 per se is very weak (within the noise range) for EM2 (Fig. 5D). Nonetheless, the difference shown is statistically significant for both EM1 and EM2.

(5) The study further indicate that the CD1c-specific T cells are enriched for the marker associated with their cytotoxic functions, and they perform slightly better in controlling Mtb growth in culture.

The authors are suggested to carefully go through the manuscript a couple of times so that supplementary figures are referred appropriately and accurately.

Reviewer #2 (Public review):

Summary:

The study by Milton et al titled "Human CD1c-autoreactive T cells recognise Mycobacterium tuberculosis-infected antigen-presenting cells and display cytotoxic effector programmes" characterises CD1c-restricted autoreactive T cells and their potential role in controlling Mtb infection. The authors develop a well-controlled system to assay for the functioning/activation of autoreactive T cells. They report the presence of CD1c-restricted autoreactive T cells in the circulating blood of healthy donors. They show that these T cells respond to CD1c and get activated even in the absence of any exogenous antigen. They next show that CD1c, along with CD1a and b, are typically downregulated on APCs during Mtb infection. These autoreactive T cells are cytotoxic, indicating they respond to Mtb treatment and/or to changes in the T cell ratio. The autoreactive T cells could effectively lyse Mtb-infected or PAMP-stimulated CD1c+APCs. Next, using TCR sequencing, they show that T cell responses were mediated by specific TCR clones with common sequence features. They show that these autoreactive T cells could curtail Mtb growth as measured by luminescence. Finally, using scRNAseq, they selectively identify the CD1c-reactive T cell pool and detect enrichment of typical effector memory CD4 and CD8 cells expressing cytolytic markers such as Granzyme, granulolysin, etc. The lung biopsy staining, along with the other data presented here, suggests that while CD1c-restricted T cells could have potential anti-bacterial roles, Mtb downregulation effectively shuts down this mechanism for TB control.

Strengths:

The study is designed well and has developed many exciting tools to generate specific information.

Weaknesses:

The revised manuscript addresses many concerns, but one section remains weak. The efficiency of these CD1c-restricted T cells in controlling TB remains very limited. The only result that addresses the bacterial control through this mechanism is Fig. 6C, which shows a very modest impact. Even THP1-KO cells show a decline in CFU when cultured with autoreactive CD1c-autoreactive T cells, and the further dip in THP1 CD1c cells is very minimal.

Another issue left unaddressed is the cytolytic response on Mtb-infected cells. How efficient are lytic responses in controlling Mtb infection? Usually, bacteria can emerge from lysed cells and divide extracellularly. How would one show this mechanism in vivo?

Reviewer #3 (Public review):

The authors have addressed most concerns from the initial review, significantly enhancing the manuscript. The expanded characterisation of the engineered THP1-CD1c system provides strong evidence that the observed T-cell responses are unlikely to result from residual conventional MHC recognition. The specificity of the CD1c-autoreactive T-cell lines is further confirmed by multimer staining, their inactivity against THP1-KO cells, and TCR-transfer experiments.

The revised data support the main conclusion that human CD1c-autoreactive T-cells recognise Mtb-infected cells that express Cd1c and exhibit cytotoxic effector functions. The difference between TCR-dependent recognition, shown by EM1 and EM2, and cytotoxicity, demonstrated by the original T-cell lines, is now more clearly presented.

Several limitations remain. The specific CD1c-associated lipid signalling molecule that enhances recognition of Mtb-infected cells has yet to be identified. Additionally, the bacterial luminescence assay lacks validation against CFU counts and cannot differentiate between intracellular and extracellular bacteria. The single-cell RNA sequencing was conducted with only two donors, and the lung immunohistochemistry remains qualitative without comparison to healthy or non-TB inflammatory tissues. These constraints limit detailed mechanistic insights and broader applicability, but they are appropriately reflected in the manuscript.

Overall, this is an important study that advances understanding of human CD1c-autoreactive T-cells in the context of mycobacterial infection. The evidence supporting the principal conclusions is solid, provided that altered CD1c-associated lipid presentation remains as a mechanistic hypothesis and the reduction in Mtb luminescence is not taken as direct evidence of selective intracellular bacterial killing.

Reviewer #1 (Public review):

Summary:

This study asks whether synapses formed by the same broad neuronal class (excitatory pyramidal neurons, PN) adapt their presynaptic organization in a cortex-specific manner, comparing prefrontal cortex (PFC) with primary somatosensory cortex (S1). The authors combine sophisticated electrophysiology (paired recordings and extracellular minimal stimulation), pharmacological perturbations of presynaptic Ca²⁺-secretion coupling, bouton Ca²⁺ imaging, and mechanistic modeling. Across two prominent excitatory connections (Layer 5 (L5) PN-L5PN and L2/3-L5PN), they provide convergent evidence that mature PFC synapses operate with looser Ca²⁺ channel-release sensor coupling than their S1 counterparts.

Overall, the study provides an appealing mechanistic link between synaptic nano/micro-architecture and cortical-area specialization. The idea that PFC synapses retain a more "plasticity-favoring" presynaptic state, while primary sensory cortex emphasizes reliability and timing precision, is potentially impactful for how we think about circuit computation and plasticity across cortical hierarchies.

Strengths:

A major strength is the multi-pronged experimental strategy. The paper first establishes robust, area-dependent differences in synaptic efficacy, reliability, timing, and short-term plasticity (facilitation prevailing in PFC versus depression in S1), using both paired recordings and minimal extracellular stimulation paradigms. The coupling interpretation is then directly supported by differential sensitivity to EGTA (and appropriate positive-control effects of fast chelators). Finally, volume averaged calcium signals are reported to be similar across areas, arguing against trivial explanations based on gross differences in calcium influx, and the modeling provides a quantitative framework for interpreting the observed chelator effects.

Weaknesses:

Limitations are minor and concern interpretation/clarity rather than core results. Some key inferences rely on indirect readouts (chelator sensitivity, fluctuation analysis-derived parameters, bouton-averaged calcium signals), each of which carries assumptions and potential confounds that should be discussed more explicitly. In particular, the, repatching paradigm for the paired-recording EGTA experiment, though very impressive, and the limited number of extracellular calcium conditions used for fluctuation analysis (three concentrations) can influence quantitative estimates and the confidence intervals around them.

Reviewer #2 (Public review):

Schwarze et al. investigated whether synaptic efficacy is brain-region specific. To this end, they compared synaptic connections established by layer 5 (L5) neocortical pyramidal cells and between L5 and L2/3 pyramidal cells. In order to identify the mechanism of this brain region specificity, the authors employed several experimental approaches, including paired electrophysiological recordings, extracellular stimulation, low- and high-affinity intracellular calcium chelators (EGTA and BAPTA), multiple probability fluctuation analysis (MPFA), and intracellular measurements of calcium transients as well as computational modelling. The findings of the present study indicate that synaptic connections in the primary somatosensory cortex (S1) are significantly stronger and more reliable than those in the prefrontal cortex (PFC).

The study is timely and the topic is of significant interest to the neuroscience community. Despite the extensive research that has been carried out on the neuroanatomy and receptor distribution of different brain regions, comparatively little attention has been paid to differences in synaptic physiology. The authors' approach is characterised by its elegance and comprehensive nature, and the conclusions drawn are compelling.

Comments on revised manuscript:

I have no further issues with the present version of the manuscript. All my concerns and/or recommendations were satisfactorily addressed.

Reviewer #3 (Public review):

Summary:

In this manuscript, Max Schwarze and colleagues examined the coupling distance between presynaptic Ca²⁺ channels and the vesicular release sensor at neocortical synapses in mouse. They propose that Ca²⁺ channel-release sensor coupling differs across cortical areas, with relatively loose (microdomain) coupling in prefrontal cortex (PFC) and tighter (nanodomain) coupling in primary somatosensory cortex (S1) for comparable pyramidal-neuron synapse types. To test this, they combine paired recordings and minimal stimulation with chelator manipulations (EGTA/BAPTA), mean-variance/MPFA-style analyses, presynaptic Ca²⁺ imaging, and computational modeling. They conclude that presynaptic coupling organization is area-specific in the mature cortex and contributes to regional differences in synaptic timing, reliability, and short-term plasticity.

Strengths:

This study tackles an important question and is strengthened by a cohesive body of evidence assembled from multiple complementary approaches. A major asset is the inclusion of high-value datasets, particularly the paired recordings between L5 pyramidal neurons and the systematic assessment of EGTA sensitivity, which provide a solid functional foundation for the authors' central claims. The work is further distinguished by its genuinely multimodal design: combining electrophysiology with presynaptic calcium imaging (and integrating these observations with quantitative analyses and modeling) offers a more mechanistic view of neurotransmitter release than any single method could provide. Overall, the direct, within-framework comparison of presynaptic release-control mechanisms across cortical areas for comparable synapse types is compelling and gives the conclusions a level of robustness and interpretability that is often difficult to achieve in studies of cortical synaptic diversity.

Weaknesses:

The principal limitation is incomplete cellular and synaptic specificity in parts of the study. The L2/3-L5PN experiments rely on minimal extracellular stimulation and therefore do not unambiguously identify the presynaptic neuron, its subtype or the number of recruited axons. Similarly, calcium imaging was performed at boutons on L5PN axon collaterals without identifying their postsynaptic targets. The imaging measurements could therefore combine boutons contacting pyramidal neurons and interneurons, potentially obscuring target-dependent differences in presynaptic calcium regulation. Recent connectomic studies demonstrate that local L5 pyramidal-cell axons can distribute substantial fractions of their output to inhibitory neurons, although the exact proportions depend strongly on pyramidal-cell subtype and distance along the axon.

The quantitative coupling-distance estimate is also model-dependent. The approximately 50-nm estimate for PFC synapses follows from a particular ring-like VGCC geometry and release-sensor model. The simulations demonstrate that this configuration is compatible with the data, but they do not uniquely identify the underlying molecular architecture.

Overall, the experiments support the narrower conclusion that the examined PFC and S1 synapses differ in functional Ca²⁺-channel-release-sensor coupling. The associated differences in synaptic timing, efficacy and plasticity are compelling, although coupling distance is not isolated causally from other regional differences in release-site number and quantal properties. The proposal that loose coupling is a general correlate of higher-order cortical function remains an interesting but currently speculative interpretation. Further comparisons across additional cortical regions and genetically or projection-defined synapse types will be particularly helpful in establishing the broader generality of this concept

Comments on revised version.

The authors have addressed most of my comments in the revised manuscript. I have only one remaining, relatively minor suggestion concerning point 5. I appreciate that the authors now acknowledge the possibility that the imaged boutons may contact different postsynaptic targets. However, the argument that interneuron-targeting boutons are likely to make only a minor contribution, based on the overall proportions of excitatory neurons or inhibitory synapses in the cortex, may not fully resolve this concern. Excitatory pyramidal neurons can distribute their outputs non-randomly across excitatory and inhibitory targets, and this distribution may depend on pyramidal-cell subtype and axonal distance. For example, a recent MICrONS/Allen Institute connectomic analysis of L5 extratelencephalic neurons in mouse visual cortex found that approximately two-thirds of their proximal synaptic outputs contacted inhibitory neurons. The proportion was close to 80% near the soma and decreased progressively with distance along the axon.

These findings concern a specific L5 pyramidal-cell subtype in visual cortex and therefore cannot be transferred directly to the PFC and S1 preparations examined here. Nevertheless, they illustrate that the postsynaptic target distribution of L5 pyramidal-neuron boutons cannot necessarily be inferred from the overall abundance of excitatory and inhibitory neurons or synapses.

Reviewer #2 (Public review):

Summary:

The study aimed to assess the associations between meteorological drivers and influenza is important although not new. The authors used 6 years of surveillance data and deep learning models, combining distributed lag non-linear models (DLNM) with Bayesian-optimized LSTM neural networks for predictive modeling. The key interest in this area is to explore the subtropical locations, where influenza is less common and circulates year-round. The authors further claimed that such an association could be able to provide an early warning in the community.

Strengths:

Study design based on a prospective cohort to analyse the data for retrospective outcomes.

[Editors' note: The Reviewing Editor has assessed the revised article without further input from the original reviewers. The authors have made considerable efforts to address the methodological and reporting concerns raised by the initial peer review, resulting in a substantially revised manuscript.]

Reviewer #1 (Public review):

Summary:

This work explores the relationship between sexual conflict and species diversification in a clade of small water striders. They use multiple phylogenetic methods to establish a potential phylogenetic tree and explore incomplete lineage sorting and introgression. They then compare their calibrated species tree with phenotypic measures of many species to try to infer the relationship between species diversity and sexual conflict. They show evidence for both ILS and introgression within the broader clade. They also find evidence that male leg diversity is associated with speciation, potentially due to sexual conflict, and that male genital grasping structures as well as female anti-grasping traits have less evidence for an association with speciation. This paper seems like a generally rigorous and interesting addition to the literature on sexual conflict and speciation, and I believe the conclusions are well supported with only a few minor concerns with methodology.

Strengths:

This paper verifies results using multiple methodologies to ensure robustness to changes in software use, and presents convincing evidence for the hypotheses tested.

Weaknesses:

(1) Lines 299-304: For the categorization of sexual conflict traits, were categories chosen by a blinded participant or by a researcher who knew the species they were observing? If any of these traits are subtle, this could add bias to the categorisation of traits.

(2) Lines 325-327: It is unclear to me if you control for phylogenetic relationships in this model. Diversification rate could be lineage-specific regardless of sexual conflict, so it seems like potentially including phylogenetic relationships in a model would control for that. And similarly, lines 331-333, I may be mistaken, but it sounds like you are using lm() to model a binary outcome (presence/absence), but lm() doesn't do logistic regression as far as I know, so it may be better to model this as a logistic regression (controlled for phylogeny) using glm().

(3) There are a few areas where the reporting of inference or statistics could be improved, and throughout the manuscript there are often mentions of 'significant' without any measure of uncertainty such as confidence intervals (example on lines 449-451). In lines 385-390, because the intervals are so wide, I would suggest saying these clades diverged between X-mya and Y-mya, rather than giving an actual estimate. It seems to me like giving a specific date is a bit overconfident when the authors have such wide intervals. On line 436, I think the authors should report confidence intervals for their 5mya estimate. In lines 455-456, what is this correlation and what are the authors' uncertainties around the estimate?

(4) Lines 549-545: When the two possible explanations for this result are reported, it seems like the authors are saying that because they can't think of how to test this possibility, the other possibility is more likely. But I don't think that is an argument against the first possibility.

(5) Lines 591-602: This paragraph confuses me. I thought that much of the introduction and Figure 1 seemed to be putting forth that the terminal segment complexity was a conflict structure that we were interested in. However, here it is stated that it does not strongly influence mating success, so is it necessarily a conflict trait? Is it demonstrated that the leg structures influence mating success?

Reviewer #2 (Public review):

Summary:

This study uses comparative phylogenetic methods to examine the evolution of male and female antagonistic traits in a group of small water striders. Water striders have long been a model system for studies into the sexual conflict that arises through anisogamy, the differential investment in gametes by males and females. Here, the authors aimed to reveal the evolutionary rates and trajectories of male grasping and female anti-grasping traits across species of the minute water-strider subgenus Pseudovelia. This was done by combining multiple genomic techniques to generate phylogenies to test trait evolution, quantify rates of evolution, and identify instances of incomplete lineage sorting (a result of rapid diversification) and introgression (the result of interbreeding between genetically different populations/species).

Strengths:

The strengths of this study lie in its comparative macroevolutionary framework, in particular the generation of multiple phylogenetic hypotheses using different methods (mitochondrial genes, USCOs, and SNPs), and contrasting these to glean insights into evolutionary patterns across species.

Weaknesses:

The main weakness of the study is the lack of underlying experimental evidence to explicitly show the grasping and anti-grasping functions of the various male and female traits, relying instead on studies of similar structures in more distantly related taxa. Without explicitly showing the functional mechanisms and reproductive costs of these traits, the resulting interpretations are wholly speculative. However, I would argue that such macroevolutionary studies are still very useful, and provide the groundwork for future studies untangling the relative roles of sexual conflict, cryptic female choice, sperm competition and reproductive interference in trait evolution and ultimately in speciation.

Reviewer #2 (Public review):

The authors initial goal was to demonstrate loss of PG during the slow sporulation process of Myxococcus xanthus, with examination of the PG degradation products in order to implicate possible enzymes involved. Upon finding a predominance of LTG products, they examined sporulation in strains lacking each of the 14 candidate LTGs encoded in the genome, leading to the identification of two sporulation-linked LTGs. An extensive characterization of the roles played by these LTGs. One LTG is responsible for the slow sporulation PG degradation, while another is required for the rapid sporulation process. Interestingly, the "slow" LTG seems to provide an important regulatory brake on the rapid enzyme. Single molecule fluorescent tracking of these enzymes was used to develop a model for their interaction with PG that mimics their observed activity. The rate of PG synthesis activity was also shown to impact the rate of PG degradation, suggesting potential interplay between the synthetic and degradative enzymes.

Strengths:

The genetic analysis to identify sporulation-linked LTGs and their effects on growth sporulation, and spore properties was well done and productive. The fluorescence microscopy to track LTG mobility, presumably tied to activity, produced a convincing argument about the mechanism of regulation of one LTG by another. The authors have responded well to most points of the previous review.

Weaknesses:

While the impact of LTGs on sporulation was clearly demonstrated, the PG analysis that resulted in the study of LTGs raised some important unanswered questions. The analyses suggest that the PG is degraded to quite small fragments, which would normally be lost during the purification of PG. The conclusions concerning the PG degradation during sporulation needs to be clarified, as described below. The authors suggest a "new mechanism of sporulation" when they have actually simply identified an important factor (PG degradation by LTGs) within a complex "process of sporulation". This needs to be reflected also in title of the paper.

Reviewer #1 (Public review):

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

Meiotic recombination at chromosome ends can be deleterious, and its initiation-the programmed formation of DSBs-has long been known to be suppressed. However, the underlying mechanisms of this suppression remained unclear. A bottleneck has been the repetitive sequences embedded within chromosome ends, which make them challenging to analyze using genomic approaches. The authors addressed this issue by developing a new computational pipeline that reliably maps ChIP-seq reads and other genomic data, enabling exploration of previously inaccessible yet biologically important regions of the genome.

In budding yeast, chromosome ends (~20 kb) show depletion of axis proteins (Red1 and Hop1) important for recruiting DSB-forming proteins. Using their newly developed pipeline, the authors reanalyzed previously published datasets and data generated in this study, revealing here-to-fore-unseen details at chromosome ends. While axis proteins are depleted at chromosome ends, the meiotic cohesin component Rec8 is not. Y' elements play a crucial role in this suppression. The suppression does not depend on the physical chromosome ends but on cis-acting elements. Dot1 suppresses Red1 recruitment at chromosome ends but promotes it in interior regions. Sir complex renders subtelomeric chromatin inaccessible to the DSB-forming machinery.

The high-quality data and extensive analyses provide important insights into the mechanisms that suppress meiotic DSB formation at chromosome ends.

Comments on latest version:

I have checked the authors' responses and the revised analyses. I think they have adequately addressed my main concerns, particularly regarding the quantitative analyses of the chromosome fusion and SK1/S288c comparisons. I have no further comments and am content for you to proceed.

Reviewer #2 (Public review):

Summary:

In this manuscript, Raghavan and his colleagues sought to identify cis-acting elements and/or protein factors that limit meiotic crossover at chromosome ends. This limitation is important for avoiding chromosome rearrangements and preventing chromosome mis-segregation.

By comparing protein axis recruitment in SK1 and S288C background, which differ in their number and distribution of Y' elements, the authors show that Y' element have a limited impact on axis protein enrichment. Genetic analyses coupled with ChIP experiments revealed that the differential binding of the Red1 protein in subtelomeric regions requires the methyltransferase Dot1. Interestingly, the lack of Red1 depletion in subtelomeric regions in this mutant does not impact DSB formation. Another surprising finding is that deleting DOT1 has no effect on Red1 loading in the absence of the silencing factor Sir3. Unlike Dot1, Sir3 directly impacts DSB formation, probably by limiting promoter access to Spo11. As now clearly stated in the abstract and the discussion, this explains only a small part of the low levels of DSBs forming in subtelomeric regions and the main mechanisms suppressing crossover close to the ends of chromosomes remain to be deciphered.

Strengths:

This work provides intriguing observations, such as the impact of Dot1 and Sir3 on Red1 loading and the uncoupling of Red1 loading and DSB induction in subtelomeric regions.

The separation of axis protein deposition and DSB induction observed in the absence of Dot1 is interesting because it rules out the possibility that the binding pattern of these proteins is sufficient to explain the low level of DSB in subtelomeric regions.

The demonstration that Sir3 suppresses the induction of DSBs by limiting the openness of promoters in subtelomeric regions is convincing.

Weaknesses:

Sir3's impact on DSB induction is compelling, yet it only accounts for a small proportion of DSB depletion in subtelomeric regions. Thus, the main mechanisms suppressing crossover close to the ends of chromosomes remain to be deciphered. [Update: these limitations have been added to the text.]

Reviewer #1 (Public review):

Summary:

The Voltage-Dependent Anion Channel 1 (VDAC1) is the most abundant β-barrel protein in the outer mitochondrial membrane and the main conduit for metabolite and ion exchange between the cytosol and mitochondria. Its oligomerization has been proposed to control mitochondrion-mediated apoptosis, making it a prime target for therapeutic intervention in diseases associated with excessive cell death, such as neurodegenerative disorders and autoimmunity. VBIT-4 is a small molecule developed to inhibit VDAC oligomerization and has shown therapeutic potential in various preclinical models. Despite its widespread use, the mechanism of action of VBIT-4 has not yet been fully elucidated. In this paper, Ravishankar et al. combine a suite of biophysical approaches with computer simulations to demonstrate that VBIT-4 forms water-permeable defects in membrane bilayers without any detectable effects on VDAC1 channel properties or oligomerization. Furthermore, cytotoxicity assays revealed identical VBIT-4 IC50 values in wild-type and VDAC1-KO cells, indicating that its activity does not depend on VDAC1. Collectively, these findings cast significant doubt on the widely held assumption that VBIT-4 is a specific inhibitor of VDAC1 oligomerization. Instead, it appears that VBIT-4 functions as a membrane-active compound.

Strengths:

This is a carefully conducted and well-written study that highlights potential side effects of VBIT-4, a compound that has been used to study the role of VDAC1 in a range of physiological and pathological conditions. The work is of interest to a broad readership by showcasing the importance of a systematic assessment of drug-membrane interactions to identify potential off-target membrane-driven effects of small molecules that may be mistakenly attributed to the inhibition of specific proteins. Its strength lies in the variety of complementary approaches the authors used to rigorously challenge the effect of VBIT-4 on VDAC1 organization and function. Overall, the experimental data are compelling and of high quality.

Weaknesses:

The authors used high-speed atomic force microscopy (HS-AFM) to study the impact of VBIT-4 on VDAC1 oligomerization in real time at nanoscale resolution. Toward this end, they adsorbed POPC:POPE:cholesterol membranes reconstituted with or without VDAC1 on mica. This revealed that addition of VBIT-4 produced small perforations in the bilayer that were independent of VDAC1. In the absence of VBIT-4, VDAC1 showed the characteristic honeycomb topography that the authors described in a previous study (Ref. 17). To quantitatively assess whether VBIT-4 affects VDAC1 organization, they analyzed protein compaction within clusters using inter-protein distance measurements. This analysis revealed no significant difference in VDAC1 organization between control conditions, 1 uM and 10 uM VBIT-4, supporting a model in which VBIT-4 primarily perturbs the lipid matrix rather than VDAC1 assemblies. This conclusion is based on the assumption that VDAC channels retain some lateral mobility in bilayers adsorbed onto mica, for which the manuscript does not provide direct evidence. In their rebuttal, the authors cite previous studies indicating that VDAC channels and beta-barrel proteins with larger extracellular domains exhibit measurable lateral diffusion in supported lipid bilayers formed on mica. Based on this, they conclude that their methodology does not constitute a limiting factor for lateral diffusion. It would be appropriate to cover this point and cite the corresponding references in the manuscript.

Reviewer #2 (Public review):

Summary

This manuscript re-evaluates the mechanism of action of VBIT-4, a compound widely used as a putative inhibitor of VDAC1 oligomerization. The authors test whether VBIT-4 acts directly on VDAC1 assemblies or instead perturbs lipid membranes more generally. Using high-speed atomic force microscopy, electrophysiology, liposome leakage assays, Laurdan fluorescence, microscale thermophoresis, coarse-grained molecular dynamics simulations, and cell-based assays in wild-type and VDAC1-knockout HeLa cells, they show that VBIT-4 partitions into lipid bilayers, induces membrane defects and leakage, and causes VDAC1-independent cytotoxicity at concentrations commonly used in the literature to infer VDAC1-specific effects.

Strengths

The main strength of the study is the convergence of multiple independent approaches on the same central conclusion. Atomic force microscopy directly visualizes VBIT-4-induced defects in lipid regions while VDAC1 assemblies remain apparently intact. Electrophysiology separates VDAC1 channel behavior from background membrane conductance and shows that VBIT-4 does not measurably alter VDAC1 conductance or voltage gating, while increasing nonspecific membrane permeability. Lipid-only membranes, lipid nanodiscs lacking VDAC1, and VDAC1-knockout cells provide important controls supporting a VDAC1-independent mechanism.

The wild-type versus VDAC1-knockout cytotoxicity comparison is a particularly strong test of VDAC1 independence at concentrations above 10 µM. The manuscript also usefully emphasizes that VBIT-4 is poorly soluble, aggregation-prone, pH-dependent, membrane-partitioning, and storage-sensitive. These properties are important for interpreting variability across previous studies using this compound.

The manuscript is careful in defining the scope of its conclusions. It distinguishes AFM- and simulation-based measurements of VDAC1 cluster organization from cross-linking-defined proximity, which is important because these are related but non-equivalent readouts of VDAC1 organization. It also explicitly discusses how VBIT-4 solubility, aggregation, protonation, membrane partitioning, and storage sensitivity complicate comparisons based on nominal compound concentration. These points help readers interpret both the current data and the broader literature using VBIT-4.

Limitations

The cellular data strongly support VDAC1-independent cytotoxicity above 10 µM, but the lower-dose mitochondrial functional phenotypes, including effects on respiration, mitochondrial calcium, and mitochondrial membrane potential, were not directly compared between wild-type and VDAC1-knockout backgrounds. The manuscript appropriately avoids overinterpreting these mitochondrial effects as directly VDAC1-independent, but readers should note that VDAC1 independence is more firmly established for cytotoxicity than for the lower-dose mitochondrial phenotypes.

The coarse-grained simulations provide useful mechanistic support for membrane partitioning, aggregation, and defect formation. However, the partitioning validation relies on the neutral VBIT-4 species and comparison with empirical partition-coefficient predictors rather than a matched all-atom octanol-water transfer calculation using the same atomistic model. This is a reasonable modeling choice, but it does not eliminate the likely importance of atomistic-level details for accurately describing pore formation. This is especially relevant for a compound with pH-dependent protonation, aggregation, and interfacial membrane localization. The simulation-derived partitioning and pore-formation results should therefore be interpreted as strong qualitative and mechanistic support rather than as a definitive quantitative description of VBIT-4 behavior across all protonation states, concentrations, and membrane environments.

Overall assessment

Overall, this is an important and timely study that provides a strong reassessment of VBIT-4 as a tool compound. The evidence that VBIT-4 perturbs lipid membranes independently of VDAC1 is compelling and should be useful for researchers interpreting past and future studies that use VBIT-4 as a probe of VDAC1 function.

Reviewer #1 (Public review):

Summary:

In this study, Sidwell and Rothenberg report a genetically rigorous study whose central finding - reduction of Bcl11b at the positive selection stage reroutes CD8 T cells to the T cell virtual memory (TVM) fate - is well supported by the convergence of elegant mouse models (WT, Bcl11bΔEnh, Bcl11b+/-, Bcl11bR3S). In wildtype mice, Bcl11b acts as a transcriptional repressor that limits the reprogramming of CD8 T cells to the TVM cell type. Reducing Bcl11b dosage partially relieves this repression, resulting in increased development toward the TVM fate. Their analysis reveals several interesting aspects: 1)<br /> Strengths:

Factors that drive CD8 TVM fate are important and relatively poorly understood. This study makes the unexpected and interesting observation that Bcl11b gene dosage impacts TVM during T cell selection in the thymus. The conclusions are strongly supported by multiple independent lines of evidence.

Weaknesses:

The direct genomic targets impacted by Bcl11b heterozygosity were not identified.

Reviewer #2 (Public review):

This manuscript by Sidwell and Rothenberg demonstrates that commitment of CD8 T cells to the virtual memory TVM cell lineage is fine-tuned in a dose-dependent manner by the transcription factor Bcl11b during intrathymic positive selection. Using multiple mouse models, the authors show that a subtle, less than two-fold reduction in Bcl11b expression or disruption of its corepressor-recruitment domain biases developing CD8 single-positive thymocytes toward a TVM cell fate without requiring peripheral activation, lymphopenia, or external cytokine signaling. Mechanistically, this modest decrease in Bcl11b does not alter global chromatin accessibility but instead enhances downstream T-cell receptor (TCR) signal responsiveness, effectively mimicking a high-affinity selection response to divert late-cycling CD8SP thymocytes into the TVM pathway. These data suggest that Bcl11b essentially serves to attenuate the interpretation of TCR (and cytokine) mediated signals to prevent the excessive differentiation characterised by virtual memory T cells and the CD44int naïve T cells. This is distinct from alternative pathways of Tvm development that are driven predominantly by exposure to cytokines, namely IL-4, in the thymus, and serves to reinforce our understanding that Tvm cells are an alternate lineage of T cells that arise during development, in part as a consequence of strong TCR signalling. There are some issues arising, not least of which is why the attenuated Bcl11b expression is insufficient to drive negative selection rather than Tvm formation.

This paper was an absolute pleasure to read given its engaging narrative style. However, in some parts it was a bit long-winded and took a while to get to the destination. Some effort should go into making the narrative more concise, while retaining the thoroughly clear explanation and interpretation of the data.

Reviewer #3 (Public review):

Summary:

The authors explore the impact of a modest (<2-fold) reduction in the expression of Bcl11b on the differentiation of CD8+ T cells, with a special focus on the generation of memory-like cells (sometimes called "virtual" memory cells - or TVM). The manuscript covers a lot of ground, but highlights are using diverse models to show that reduced Bcl11b expression during thymic development (but not in naïve CD8+ T cells that have accessed the periphery) leads to enhanced generation of cells with phenotypic, transcriptional, epigenetic and functional characteristics of TVM; that this is a cell-intrinsic effect, but not apparently driven by enhanced responsiveness to cytokines (which promote TVM in some models); decreased Bcl11b improves T cell sensitivity at the mature (and likely in immature thymocytes). Myriad approaches and controls are used, providing a very thoroughly explored model.

This is a tour-de-force in applying the geneticists tool-box for investigating how a tantalizingly modest decrease in Bcl11b expression impacts the generation of TVM-like cells during thymic development. It is unreasonable to request additional data, but clarification of some key conclusions is needed.

Reviewer #1 (Public review):

Summary:

This work builds a theory to implement planning trajectories towards a goal in a known environment, inspired by analyses of prefrontal neural recordings. Unlike standard neural architectures for this task, such as value-based learning and successor representations, their proposed theory is able to adapt to novel goal locations within-trial. The key to the theory is that future times and locations are represented by disjoint groups of neurons. The recurrent connectivity between groups of neurons selective to specific future time and locations reflects the learned knowledge of the task. Finally, the authors show that standard networks trained on the task approximate their proposed theory.

Strengths:

The structure of the work is clear, and the article is very well written, which is particularly noticeable given the consequential amount of results presented. The authors are able to link their theory to experimental findings in neural recordings. The reverse-engineering of trained recurrent neural networks is very thorough, by analyzing both dynamics and connectivity. The assumptions and predictions of their model are clearly stated.

Weaknesses:

I believe the article shows no major weaknesses. There are important points that are beyond the scope of this study, that may limit its impact. For instance, while very little previous literature has linked planning to RNNs, their proposed theory of "space-time attractors" in RNNs is about input-driven stable attractors. The authors clarify this in the text, highlighting differences between classic attractor networks and the mechanism studied here. Related to this, an assumption of this mechanistic theory is that rewards are flexibly rerouted to the corresponding neural populations after every action, which seems difficult to implement biologically. Both aspects are discussed in the current manuscript.

Reviewer #2 (Public review):

This well-written manuscript proposes to use attractors in space and time (STA) as a mechanistic explanation for planning in the prefrontal cortex. The main conceptual hypothesis is that planning is implemented as attractor dynamics in a representation that encodes states at each time step jointly. Depending on inputs the network relaxes to a trajectory that already contains future states that will be visited at each time step, rather than computing a scalar value at each point in time and space like other classical approaches from RL. The authors compare this approach to implementations such as TD learning and successor representation, and further show that trained recurrent neural networks on specific tasks involving planning develop structured subspaces resembling the ones postulated in STA.

The idea of treating attracting trajectories unfolding in time as the computational substrate for planning is very interesting and potentially important. The explicit construction of a state x time representational space and its implementation via recurrent dynamics are appealing and convincing in the idealized tasks considered. I found the ms to be refreshingly explicit regarding several of the assumptions and limitations of the models, for example the fact that certain advantages can be viewed as properties of the state space itself and not necessarily of a fundamentally new planning mechanism.

I thank the authors for their reply and their thorough rebuttal. It answered most of my previous questions and greatly enhanced the understanding of the paper.

I have just two remaining concerns:

(1) The ms shows attractor dynamics in the trained RNN during planning, but it is less clear how these relate to the execution phase. It would be helpful to clarify whether the network state during execution is expected to effectively be close to a FP or at a FP for each input, or whether the RNN implements transient dynamics shaped by the underlying attractor landscape.

(2) Regarding the previously raised point of calling their result a "Mechanistic theory of planning", I did not mean to suggest that a theory cannot be mechanistic, or that "mechanistic theory" is not a valid term, especially in the context of this paper (although I believe this topic would deserve an entire separate discussion in the neuroscience field).

My point was about whether STA should primarily be interpreted as a mechanistic theory of planning, or as a candidate neural mechanism for implementing the planning as inference theory. I am aware that mechanistic theory and mechanistic models are often used interchangeably in neuroscience, and I certainly do not claim that my interpretation is the only valid one. My opinion is that the manuscript presents a convincing and interesting candidate neural mechanism for planning, which can be strongly related to planning as inference. The reason why I am not fully convinced about the framing as a mechanistic theory of planning is mainly that the adjacency-based connectivity isn't emerging or derived, but is instead introduced based on practical and empirical considerations. It's not a major issue, but I would personally frame it as a mechanistic account or model of planning (and/or planning-as-inference), rather than a theory, mechanistic or not.

Reviewer #1 (Public review):

Summary:

The immune system represents a source for melanocyte-extrinsic determinants of melanoma. Multiple immune cell types, including natural killer cells and CD8+ cytotoxic T lymphocytes, destroy cancer cells directly, and this anti-tumor activity is widely thought to eliminate many nascent tumors before they become clinically detectable. Regulatory T (Treg) cells, which are defined by expression of the transcription factor Foxp3, function as critical suppressors of lymphocyte activation in both homeostatic and disease contexts. Intratumoral Treg cell accumulation has been associated with disease progression in the clinic, and Treg cell depletion inhibits melanoma outgrowth in transplantable models of the disease.

However, the role of Treg cells during the early, premalignant stage of melanocyte expansion has not been examined. To study the interplay between incipient melanoma and cutaneous inflammation, the authors subjected an autochthonous murine model of melanoma [LSL-BrafV600E;Ptenfl/fl;Tyr::CreERT2 (BPT)mice] to three distinct inflammatory immune perturbations. Each of these perturbations accelerated premalignant melanocyte outgrowth, which was unexpected given that both Treg cell depletion and DNFB treatment markedly enhanced conventional T (Tconv) cell infiltration into the skin. Detailed analysis of each inflammatory response revealed a shared cellular and molecular signature comprising myeloid infiltration, characteristic cytokines and tissue remodeling factors, and vascular permeability. Altogether, the results support the hypothesis that oncogenic mutations in melanocytes along with altered immune response and inflammation synergistically drive melanomagenesis.

Strengths:

(1) The use of the three distinct inflammatory immune perturbations, such as transient Treg cell depletion, acute UV- B irradiation, and 2,4-dinitrofluorobenzene (DNFB)-induced contact hypersensitivity.

(2) The re-examination of the role of Treg cells in transplantable models of tumor growth by Subcutaneous (s.c.) implantation of syngeneic B16F10 melanoma cells as widely used to study anti-tumor immunity and to interrogate the effects of Treg cells on tumor suppression.

(3) The use of the LSL-TdTomato mice for measuring the TdTomato fluorescence in each immune cell type to assess uptake of melanocyte antigen.

(4) scRNA-seq data document a broad and rapid myeloid inflammatory response in Treg cell-deficient skin.

Weaknesses:

(1) The expression of inflammatory mediators Il1b, Il6, and TNFa, and angiogenic mediators Hif1a and Ang2 in all of the three models of immune perturbation has been verified at the transcript level by qRT-PCR, and it remains to be determined whether it correlates with the same at the protein level.

(2) scRNA-seq data to profile the diversity of immune cells (CD45+) in the ear skin of the DNFB-treated contact hypersensitivity model are currently missing.

(3) The authors indicate that at least 2 prior studies directly implicated inflammatory macrophages in the melanocyte proliferation response. However, no attempts were made for the identification of the UVB-driven factors underlying myeloid recruitment by which these cells activate melanocytes.

(4) It is very surprising to observe that altered immune responses, such as enhanced Tconv cell priming and activation in Treg cell-deficient skin, failed to antagonize mutant melanocyte outgrowth in the BPT model. While UVB-induced skin inflammation differs somewhat from the response to Treg cell depletion, both feature the infiltration of tissue remodeling macrophages and vascular instability. The findings that contact hypersensitivity can promote the expansion of non-malignant BRAF (V600E)Pten- Het melanocytes have interesting implications for benign hyperpigmentation conditions, such as post-inflammatory hyperpigmentation, Riehl's melanosis, and melasma.

Reviewer #2 (Public review):

Summary:

Tran and colleagues investigate how inflammation alters the earliest stages of melanoma tumorigenesis in mice carrying LSL-BrafV600E, Ptenfl/fl, and Tyr-CreERT2 alleles. They compare transient regulatory T cell depletion, acute UVB irradiation, and DNFB-induced contact hypersensitivity. Each perturbation increases ear pigmentation and Tyrp1 expression after oncogene induction. The inflammatory settings also share recruitment of monocytes and macrophages, expression of inflammatory and tissue-remodeling programs, and increased vascular permeability. Dexamethasone attenuates the DNFB-associated phenotype. A secondary finding of particular interest is that regulatory T cell depletion accelerates the premalignant BPT phenotype but inhibits B16F10 tumor growth, suggesting that regulatory T cells can have different effects during tumor initiation and established transplantable disease.

The study addresses an important question that is difficult to approach using transplantable tumor models. The data convincingly show that each perturbation produces substantial inflammation in the skin and that vascular leakage accompanies the response. At present, though, the central biological endpoint is not sufficiently separated from melanogenesis. Darkening of the ear and increased Tyrp1 RNA can reflect more pigment or altered differentiation within the existing oncogene-carrying melanocytes rather than an increase in their number, particularly given that pigment content is itself variable in transformed melanocytes, which range from heavily pigmented to nearly amelanotic. This issue is especially important in the UVB and DNFB experiments, where inflammatory signals can alter pigmentation directly.

Strengths:

The autochthonous BPT model is a major strength. It preserves the native relationship between melanocytes and the surrounding stromal and immune compartments during lesion initiation. Including three distinct inflammatory perturbations makes the recurring association with melanocyte-associated readouts more persuasive than any single model would be. The paired-ear DNFB design is efficient and controls for inter-animal variability. The combination of flow cytometry, single-cell RNA sequencing, intravital imaging, and Evans Blue assays provides useful complementary evidence that the inflammatory interventions remodel the local tissue environment. The B16F10 experiments help establish that the unexpected effect of regulatory T cell depletion is specific to the early autochthonous setting rather than a general failure of the depletion model. The BT-Het experiment is also thoughtful in asking whether inflammation can enhance the phenotype of oncogene-carrying melanocytes in a nevus-stage context that does not proceed to full malignant progression after oncogene induction alone.

Weaknesses:

The strongest caveat concerns the central claim. The outgrowth readouts are ear darkening and bulk Tyrp1 expression, but both may report pigment or differentiation state rather than the number of oncogene-carrying melanocytes. Pigment content is not a reliable proxy for cell number here, since the same population can darken or lighten without any change in cell number. No direct count or lineage-reporter measurement is provided for the regulatory T cell, UVB, or DNFB comparisons. Until that gap is filled, the data support increased pigmentation of oncogene-carrying melanocytes more firmly than the premalignant expansion named in the title, and this concern is most pronounced in the UVB and DNFB settings, where inflammation can change pigmentation on its own.

Secondly, the proposed shared mechanism is largely associative. Dexamethasone appropriately shows that inflammation as a whole is required for the DNFB phenotype, but as a broad anti-inflammatory it cannot isolate any single component. The manuscript singles out blood vessel remodeling as particularly important, and that specific attribution exceeds what a non-selective drug can show, especially as no individual pathway is selectively blocked in a tumor-initiation experiment and Il6 is reduced only modestly. The authors acknowledge that the precise chain of causation is unresolved, so the vascular claim should be softened to match or tested directly.

Also, several of the mechanistic conclusions rest on thin or single cohorts and on single-cell data whose replication is not fully reported, making them less convincing than the inflammatory phenotypes themselves. The systemic regulatory T cell model shows the consequences of body-wide depletion rather than a skin-specific regulatory T cell function, and the inferred monocyte-to-macrophage trajectory reflects transcriptional similarity rather than a demonstrated lineage path. The interpretation of dendritic-cell TdTomato uptake as evidence of antigen presentation or T cell priming is not supported by a direct measure of reactivity.

Finally, the nevus-stage framing should be corrected. The manuscript frames the BT-Het experiment as testing non-oncogenic conditions, but those melanocytes carry BrafV600E, so it is better read as inflammation-enhanced behavior of oncogene-carrying melanocytes at the nevus stage.

Reviewer #3 (Public review):

Summary:

Tran et. al. investigate how inflammation in the skin influences the early stages of melanomagenesis. They use an autochthonous, tamoxifen-inducible mouse melanoma model (LSL-BrafV600E;Ptenfl/fl;Tyr::CreERT2, "BPT") to examine three inflammatory perturbations: transient depletion of regulatory T cells, acute ultraviolet-B irradiation, and contact hypersensitivity induced by 2,4-dinitrofluorobenzene (DNFB). They report that each perturbation promotes the recruitment of immune cells, especially inflammatory monocytes and macrophages, increased expression of inflammatory and tissue-remodeling factors, and enhanced vascular permeability, which ultimately increases the outgrowth of premalignant melanocytes measured by local pigmentation and expression of the melanocyte-associated gene Tyrp1. In the DNFB model, the authors showed that treatment with dexamethasone reduces the effects of contact hypersensitivity on pigmentation, inflammatory gene expression, and vascular leakage, potentially providing a translational angle.

Strengths:

An interesting observation is that transient Treg depletion promotes premalignant melanocyte outgrowth in the autochthonous BPT model while inhibiting the growth of transplantable B16 F10 tumors. This contrast is consistent with a role for Tregs in limiting inflammatory disruption of the skin during early tumorigenesis and highlights the value of autochthonous models. These findings may also have broader implications for understanding the stage- and context-dependent functions of Tregs in cancer.

Another strength of this manuscript is the comparison of three mechanistically distinct inflammatory perturbations. Treg depletion, UVB irradiation, and DNFB-induced contact hypersensitivity engage different inflammatory pathways but converge on myeloid-cell recruitment, inflammatory gene expression, and increased vascular permeability. This convergence strengthens the conclusion that an acute inflammatory microenvironment is associated with enhanced melanocyte outgrowth during the early premalignant phase.

Weaknesses:

The paper convincingly establishes a correlation between the inflammatory signature and melanocyte outgrowth across three distinct perturbations. However, the mechanistic claim that myeloid cells and/or vascular remodeling drive melanocyte expansion rests primarily on the dexamethasone experiments in the DNFB model. Because dexamethasone broadly affects immune, stromal, endothelial, and melanocytic compartments, these experiments do not establish that inflammatory monocytes/macrophages or vascular destabilization are specifically required for the melanocyte response.

A related limitation is that the proposed monocytic origin of the inflammatory macrophage population following perturbation remains inferred. Although the scRNA-seq data and pseudotime analysis in Figure 4 - Supplement 2 are consistent with a trajectory from monocytes to macrophages, they do not exclude local reprogramming of resident macrophages into an inflammatory state. This alternative is particularly relevant because resident macrophage populations have been implicated in vascular remodeling and tumor outgrowth (PMIDs: 36493773 and 40216154).

Finally, the assessment of melanocyte outgrowth is largely through increased pigmentation and whole-ear Tyrp1 expression. Although these measurements may reflect increased melanocyte abundance, they may also be influenced by melanogenic activity or increased Tyrp1 expression per cell. More direct evidence of melanocyte proliferation, such as Ki67 or EdU/BrdU staining specifically within TdTomato-positive melanocytes, would support the use of "expansion" and "proliferation" throughout the manuscript. Histopathological characterization of lesion architecture, atypia, proliferation, and invasion would also help establish the premalignant nature of the lesions.

Reviewer #1 (Public review):

Summary:

This study described the genomic features of K. pneumoniae in tissue samples from children in Zambia who had died from pneumonia, where K. pneumoniae was believed to be the causative organism. This was a substudy of a previous study which described more broadly the aetiologies of community-acquired pneumonia in this population, using minimally invasive tissue sampling.

In this study, the authors used culture-independent molecular tools to characterise the K. pneumoniae genomes from post-mortem tissue samples from 7 children who had died of pneumonia. They described a diversity of strain lineages in the 7 children, with different capsular types and virulence profiles. Notably, all strains had a wide array of antimicrobial resistance genes. This is probably not surprising given the propensity of this organism to acquire AMR genes, but it is concerning in isolates from community-acquired infections.

The study also highlights the value of using advanced culture-independent molecular techniques, and the scope of data that can be obtained. In settings where culture is not always available, storage and subsequent remote analysis of these samples is an option to better understand the microbiology (although expense is still a barrier, and culture-based microbiology should not be completely neglected).

Strengths:

Ascribing aetiologies in respiratory infections is not (yet) an exact science, and there is likely to always be some doubt about whether an organism (whether identified by culture or nucleic acid detection) is a true pathogen. However, the authors have used a robust combination of histology, microbiology, radiology, clinical assessment and verbal autopsy, and this is probably as good as we can get it at present.

The molecular techniques employed and the analysis were robust and technically sound. There are some areas where there is inconsistency between the results from two samples from the same patient, and this is likely due to the lower number of reads - important information for future similar studies. It highlights the potential limitations of this technique.

The data obtained (albeit from a small sample set) are consistent with the other data from Africa, which provides support for the value and reliability of the technique itself.

Weaknesses:

The major weakness is the small sample size. The 7 patients analysed in this study are a subset of the children in the larger study who had been identified as having died from K. pneumoniae respiratory infection. So while the findings of this study highlight the potential role of this organism as a respiratory pathogen and provide some insight into the distribution of lineages in the community, the results don't really allow for major changes to clinical or diagnostic practice as yet. The data highlight a potentially under-recognised problem and would be useful to inform future studies.

A minor weakness is more related to the journal layout, where methods are presented last. The results are not as easy to follow without reviewing the methods - in particular the description of histopathological findings and results of PCR on the biopsies. Until one realises that 6 biopsies had been taken from each of the deceased children, and a subset of these biopsies used for the study, the results seemed confusing.

Reviewer #2 (Public review):

Summary:

This manuscript applies a culture-independent hybridization-capture metagenomic sequencing approach to characterize Klebsiella pneumoniae detected in post-mortem lung tissue from fatal pediatric pneumonia cases in Lusaka, Zambia. The study addresses an important challenge in retrospective genomic investigations where cultured isolates are unavailable and demonstrates the potential of targeted sequencing to recover clinically relevant genomic information directly from archived tissue specimens. The authors report sequence types, capsular loci, antimicrobial resistance determinants, virulence-associated genes, and evidence of closely related isolates in two cases. The work is valuable as a proof-of-concept application of targeted sequencing in challenging post-mortem specimens and provides useful descriptive genomic data from a setting where such information remains limited. However, several epidemiological and public health interpretations extend beyond what can be supported by the available data. The study includes only seven successfully sequenced children from a single setting and was not designed to determine the source of acquisition, transmission pathways, or population-level distributions of antimicrobial resistance or capsular types. The manuscript would therefore be strengthened by more consistently framing the findings as a descriptive genomic investigation of K. pneumoniae detected in children who died outside hospital settings, rather than as evidence of community-acquired infection or broader epidemiological shifts.

Strengths:

The principal strength of the manuscript is its methodological contribution. The authors demonstrate that hybridization-capture metagenomic sequencing can recover informative genomic data from post-mortem lung tissue in cases where conventional culture-based sequencing is not available. This is an important technical advance for retrospective studies, minimally invasive tissue sampling platforms, and settings where sample degradation, prior antibiotic exposure, or lack of routine culture limits genomic surveillance.

The study also addresses an important public health problem. K. pneumoniae is a major cause of severe infection and antimicrobial resistance globally, yet its role in fatal pediatric pneumonia outside hospital settings remains difficult to define. The generation of sequence type, capsular locus, antimicrobial resistance, and virulence-associated gene data from post-mortem specimens is therefore useful and may inform future study designs. The identification of closely related isolates in two infants is also potentially important and raises hypotheses about shared sources or transmission that could be explored in larger studies.

Another strength is that the authors appropriately acknowledge several technical challenges, including low numbers of K. pneumoniae-assigned reads in some specimens and unresolved or discordant capsular locus calls. These issues are important for readers considering the utility of this approach in low-input or mixed-specimen contexts.

Weaknesses:

The main weakness is that the epidemiological framing is stronger than the data allow. The manuscript repeatedly refers to community-acquired K. pneumoniae pneumonia and broader community epidemiology. However, the available data do not establish community acquisition, community transmission, or an epidemiological shift from nosocomial to community disease. Several children appear to have had prior healthcare contact or other potential healthcare-associated exposures, and the study design cannot determine where acquisition occurred. The findings would be more accurately framed as K. pneumoniae detected in post-mortem lung tissue from children who died outside hospital settings.

Causal attribution also requires more careful wording. Detection of K. pneumoniae in post-mortem lung tissue, together with histopathology and DeCoDe findings, provides important supportive evidence that the organism may have been in the causal chain leading to death. However, this does not necessarily establish that K. pneumoniae was the sole or direct cause of fatal pneumonia, particularly where multiple pathogens were detected.

The small sample size and case selection strategy limit the generalizability of the findings. Only seven children were successfully sequenced, and specimens appear to have been selected partly based on molecular signal. This is technically understandable, but it may introduce selection bias by enriching for cases with higher bacterial burden, better DNA preservation, or other specimen characteristics. As a result, the observed lineage diversity, resistance gene profiles, virulence-associated loci, and capsular locus distribution should not be interpreted as representative of community-acquired infections or broader population epidemiology.

The validation of the hybridization-capture approach also requires strengthening. Comparing outputs from different genomic analysis tools applied to the same sequencing data may assess bioinformatic concordance, but it does not independently validate the method. Ideally, the approach should be benchmarked against clinical K. pneumoniae isolates or matched specimens with conventional whole-genome sequencing data. Without this, it is difficult to assess the accuracy of sequence type, capsular locus, antimicrobial resistance determinant, virulence locus, and plasmid marker recovery, especially in low-read or mixed-specimen contexts.

Species-level attribution of antimicrobial resistance, virulence-associated genes, and plasmid replicons is another important limitation. In a culture-independent metagenomic study, these features cannot automatically be assigned to the identified K. pneumoniae lineage because many such elements are shared across Enterobacterales and may originate from co-detected organisms. This affects interpretation of antimicrobial resistance, hypervirulence, and MDR-hypervirulence convergence.

Overall, the authors achieved their methodological aim of demonstrating that targeted sequencing can recover useful genomic information from challenging post-mortem specimens. However, the epidemiological, transmission, antimicrobial resistance, and vaccine-related conclusions should be tempered.

Reviewer #1 (Public review):

Summary:

Overall, this is an interesting and well-written manuscript on a fascinating question in a "charismatic" model system.

Strengths:

(1) The Introduction is concise, though it might be helpful to the non-specialist reader to learn a bit more about what is known about the social control of somatic growth across diverse species (including humans), which would help to make this work more generally interesting.

(2) The experiment is well designed.

(3) The data collected are comprehensive.

(4) The complementary analysis of both feeding and aggression/submission data with and without known social roles is a neat idea and compelling!

Weaknesses:

The authors have addressed my concerns quite well. They now discuss the HPA/stress axis in some detail and also examined the extent to which growth, food intake, agonistic behavior, and/or gene expression patterns are coordinated across P1 vs P2 pairs. While still not ideal, they now provide a reasonable rationale for using whole bodies for the transcriptome analysis. Finally, the Discussion has been streamlined and is easy to follow.

Reviewer #2 (Public review):

In this manuscript, the authors test growth, behavior, and gene expression in pairs of clownfish as they establish social dominance hierarchies, examining patterns of gene expression in these pairs after dominance has been established. The authors show solid evidence that emerging dominant clownfish show increased growth, aggression, and food consumption compared to their submissive or solitary counterparts, eventually adopting distinct gene expression profiles.

Major Comments:

(1) The Introduction is comprehensive, but it could be condensed. Likewise, the discussion could be condensed. There is considerable redundancy between the methods, the results, and the legend in Figure 1. The authors should consolidate and remove the redundancy.

(2) For Figure 3, the authors are showing PC2 and PC3; why is PC1 not shown? There is so much overlap between the three groups in PC2 vs PC3; it seems unlikely that researchers could conclusively identify any individual as belonging to a group based on the expression profile. The ovals shown do not capture all the points within each of the groups, and particularly the grey S oval seems misaligned with the datapoints shown.

(3) The authors indicate that the 15 replicates exhibiting the greatest size difference between P1 and P2 were selected for gene profiling. Does this mean that each of the P1 and P2 were pairs with each other? Have the authors tried examining the gene expression patterns in a paired manner? E.g., for the pairs that showed the greatest size differences, do they also show the greatest differences in gene expression? Do the P1s show the most extreme differences from P2s that also show the most extreme P2 differences? Perhaps lines on Figure 3A connecting datapoints from the P1 and P2 pairs would be informative.

(4) For the specific target pathways that are up- and downregulated in the different backgrounds, I recommend that the authors include boxplots (or heatmaps) showing the actual expression values for these targets. Figure 6 shows a heatmap for appetite-related genes, and it would be great to see a similar graph for the metabolism and glycolysis genes; it would also be informative to see similar graphs for hormonal and sexual maturation pathways as well.

(5) Particularly given that there is a relatively small number of genes enriched in the different rank conditions, I did not understand the need to do the WGCNA module analysis. I thought that an analysis of GO terms across the dataset would have been more meaningful than the GO term analysis shown in Figure 4, which considers only genes assigned to the "brown WGCNA module". This should be simplified or clarified.

(6) The authors say that they have identified coordinated changes in behaviors and the "underlying gene expression, leading to the emergence" of social roles. This is a little bit misleading, since the gene expression analysis occurred well after the behavioral and phenotypic differences emerged. Presumably, the hormonal and genetic shifts that actually caused the behavioral and phenotypic difference occurred during the weeks during which the experiment was underway, and earlier capture of the transcriptome would presumably reveal different patterns, and ones that would be considered more causative. The authors acknowledge this in 434-435, but it could be emphasized further.

(7) The authors have measured a number of differences between the different dominance classes of fish. All these differences were measured relative to the other classes, but in my view, the Solitary group was the closest to a baseline control. So, I'm not sure that it is fair to say that "P2 and S individuals showed consistent downregulation of these genes and pathways" (line 401). I encourage the authors to emphasize the differences in gene expression from the "perspective" of the P1 individuals compared to the baseline of P2 and S individuals. Line 474 says that "P2 fish showed significant upregulation" of a number of pathways. It should be very clear what that is compared to (compared to P1, presumably?)

(8) Along the same lines, the authors say in line 514 that subordinates and solitaries strategically downregulate their growth. I'm not convinced that this is the case: I would consider this growth trajectory to be the default and the baseline. I would interpret that under certain social conditions, a P1 dominant pattern of growth, behavior, and gene expression is allowed to emerge.

Comments on revised version:

The manuscript has been carefully revised. The authors have also responded adequately to all of my previous comments.

Reviewer #3 (Public review):

Summary:

The authors tested the hypothesis that interactions among size- and age-matched rivals will lead to the emergence of social roles, accompanied by divergence in four aspects of individual phenotypes: growth, feeding behavior, fighting behaviors, and gene expression in clownfish.

Strengths:

The data of growth, feeding rate, and fighting behaviors support the authors claim.

Weaknesses:

The results obtained solely from the whole-body transcriptome are limited in supporting the authors' research question. However, the revised manuscript explicitly states this as a limitation.

Reviewer #1 (Public review):

Summary:

This article purports to show that ML-SA8, a synthetic activator of the lysosomal TRPML1 channel, results in AMPK activation and glucose uptake in hepatocytes, and that this action has therapeutic potential for metabolic disease. The final figure shows that glucose levels are improved in db/db mice, although it is not entirely clear whether this is due to an effect on the liver, on other tissues, or on glucose production or uptake. The earlier figures try to make the case that SA8 causes activation and GLUT4 translocation and glucose uptake in liver cells; however, these data are not convincing. GLUT4 is expressed at such low levels in liver that it is likely not physiologically important. The authors use a fluorescent glucose analog to measure glucose uptake, and this molecule has been shown to enter cells largely by fluid phase endocytosis. Overall, this reviewer finds the premise misguided and the data unconvincing.

Strengths and Weaknesses:

The initial figures show phosphorylation of AMPK on Thr172, but no downstream effects are shown. Usually, to convincingly show that AMPK activity is increased, it would be appropriate to immunoblot phospho-ACC or some other substrate. This is minor.

Lines 135-148: GLUT4 is not expressed at levels that are significant for physiology in liver cells, and its function in liver is not particularly relevant. The authors cite references 38-40 to support that it may be expressed at low levels in liver, but no knockout studies have been done to show that this expression is physiologically important.

Figure 1e is not convincing. No controls are included to show the specificity of the antibody for immunofluorescent staining. No intracellular GLUT4 is visible in the unstimulated samples.

In Figure 1f, again, the data are not convincing. The bands seem too sharp for GLUT4, which has 12 membrane-spanning domains as well as an N-linked glycosylation, so that it usually runs as a smear.

Figure 1h. Data are not convincing. 2-NBDG is not a valid approach to measure glucose uptake. 2-NBDG enters cells largely via fluid phase endocytosis, and its accumulation is independent of known GLUT inhibitors such as cytochalasin B (Yazdani et al., MBoC 2022; PMID: 35921166; see also PMID: 42287154). The idea that such a bulky derivative of glucose could enter the transporter channel is not compatible with known structural data.

Supplementary Figure 5 uses 2-NBDG glucose uptake again. This reviewer is not convinced that the data reflect transporter-mediated glucose uptake, as suggested by the authors. As well, although palmitate treatment of cells can cause an insulin-resistant-like phenotype in some cell types, this is not characterized in the present work. Finally, as noted, one would not expect hepatocytes to exhibit insulin-responsive glucose transport. Glycogen synthesis is the main insulin-regulated step that might be affected.

The data in Figures 2b,c,f,g,k,l are not convincing. Again, 2-NBDG is used.

For the glucose consumption measurements in other panels of Figure 2, the methods section states that cells were cultured in 10 mM glucose. What volume was used? It is difficult to believe that a monolayer of cells would consume very much of the glucose that is present in the culture medium. Data are shown as a percent of controls, and look reasonable, but it would be helpful to include absolute as well as relative units.

In Figure 2, in experiments using the TRPML1 KO cells, no panel is shown to demonstrate knockout. The authors cite a previous paper for the construction of these cells, but the control immunoblot should still be shown here.

In Figure 3, controls are missing in the BAPTA experiment in Figure 3a (only SA8-treated cells were treated with BAPTA and with EGTA). Again, it would be helpful to have p-ACC or some other readout of AMPK activity, and not just AMPK phosphorylation. 2NBDG is again used in this figure.

Line 212-213 the text states "considering our finding that TRPML1-mediated Ca2+ release is essential for AMPK activation." This has not been shown. The work uses chelators and does not necessarily indicate a role for TRPML1. The drug may be specific, as suggested by the authors, but the way this phrase is worded is too strong. As well, AMPK was shown to be phosphorylated, but full activation towards its various substrates has not been shown.

Figure 4cd suggests that GLUT4 expression is increased by 2 or 3-fold in the liver of DB+SA8-treated mice, compared to controls. This may be the case, but its abundance is still likely ~1000-fold less in liver compared to skeletal muscle or adipose tissue. This reviewer is still not convinced that this is physiologically relevant. The images in Supplementary Figure 8 suggest a larger increase, but it remains uncertain whether the staining really represents GLUT4.

Data showing that blood glucose and HbA1c are reduced in SA8-treated mice are reasonable, and GTTs and ITTs are shown. Unfortunately, there are no insulin concentrations, and it remains uncertain whether glucose production is reduced or uptake is increased (or if both effects are present).

In the discussion, the authors again state that GLUT4 is present in the liver and that it regulates hepatic glucose homeostasis, and they cite reference 63. This review article does not argue that GLUT4 acts in the liver to regulate hepatic glucose homeostasis, but that its actions in muscle and fat have secondary effects on the liver.

Reviewer #2 (Public review):

Summary:

The manuscript contains interesting studies suggesting that pharmacological activation of TRPML1 could be useful to treat T2D by increasing glucose uptake via activation of AMPK. Preclinical studies suggest the inhibitor improved blood glucose in Db/Db mice. Ex vivo studies in cell lines examine both pharmacologic and genetic manipulations, both to activate and to inactivate TRPML1, and the results consistently suggest that TRPML1 activates AMPK and increases glucose uptake.

Strengths:

The manuscript is well written, and the studies are carefully performed.

Weaknesses:

All mechanistic studies were performed in transformed cell lines; conclusions would be stronger if performed in primary cells. The in vivo studies were only performed in male mice. Performing metabolic studies in both sexes is standard practice now. Whether the findings would extend to females was not tested and remains uncertain. Some controls are missing, such as plasma membrane loading controls for fractionation studies. The GLUT4 staining was performed after fixation and permeabilization, yet control cells appear to be devoid of intracellular (and all) staining, a confusing result that doesn't reflect the expected biology.

Reviewer #3 (Public review):

Summary:

Zhu et al. present a proof-of-concept for targeting the lysosomal calcium channel MCOLN1/TRPML1endolysosomal ion channels to restore type 2 diabetes mellitus (T2DM). Using synthetic TRPML1 agonists (ML-SA8) and genetic manipulation, the authors demonstrate that TRPML1 stimulation triggers localized lysosomal calcium release. This calcium efflux sequentially activates CaMKKβ and phosphorylates AMPK at Thr172 in various cell models, including palmitic acid-induced insulin-resistant HepG2 cells. This signaling pathway promotes GLUT4 translocation to the plasma membrane and increases intracellular glucose uptake. When administered daily to diabetic db/db mice over six weeks, ML-SA8 lowers fasting and random blood glucose, improves oral glucose and insulin tolerance tests, reduces hepatic steatosis, and lowers serum ALT and AST levels.

Strengths:

Based on the TFEB-independent pathway activated by TRPML1 and the experimental approaches described by Medina's group (PMID: 31822666), the authors use a combination of pharmacological and genetic tools to dissect such an intracellular signaling pathway. Additionally, the animal experiments show consistent phenotypic improvements across independent metabolic parameters. The ability of ML-SA8 to restore glycogen deposition and clear hepatic lipid accumulation in db/db mice without causing weight loss or overt toxicity provides a strong rationale for exploring lysosomal targets in metabolic disease.

Weaknesses:

(1) The authors focus almost exclusively on hepatic GLUT4 to explain the observed glucose disposal. However, other glucose transporter isoforms such as GLUT2 dominate basal glucose transport. While the authors show increased AMPK phosphorylation in skeletal muscle and adipose tissue, they do not measure GLUT4 translocation or glucose uptake in these primary disposal organs. As a result, attributing systemic glycemic recovery primarily to hepatic GLUT4 translocation overlooks the major physiological roles of peripheral tissues.

(2) In both HepG2 cells and mouse liver tissues, ML-SA8 treatment increases total GLUT4 protein expression in addition to plasma membrane localization. Because total protein pools expand, the enrichment of GLUT4 in plasma membrane fractions cannot be cleanly attributed to acute vesicular translocation alone. The manuscript does not explain the timescale or mechanism behind this rapid total protein upregulation, leaving a mechanistic gap between acute ion channel gating and protein expression.

(3) While the in vitro specificity of ML-SA8 is well-controlled, the systemic animal experiments lack a specific rescue or knockout control. Small-molecule agonists administered intraperitoneally over six weeks can exert off-target effects. Without demonstrating that co-administering the TRPML1 inhibitor ML-SI5 blunts the therapeutic effect in vivo, or showing that ML-SA8 lacks efficacy in TRPML1-null mice, the definitive link between in vivo glycemic recovery and TRPML1 activation remains incomplete.

Reviewer #1 (Public review):

Summary:

This work by Beaudet and colleagues aims at exploring the effect of phosphorylation on the formation of tau envelopes and consequently on axonal transport both in vitro on reconstituted microtubules and in human excitatory neurons derived from IPSCs.

The authors found that a relatively widely used construct in which 14 serine or threonine residues often hyperphosphorylated in Alzheimer's disease are mutated to alanines (phosphodeficient) increases the density of tau envelopes compared to wildtype tau whereas a phosphomimetic (same residues mutated to glutamic acid) reduces envelopes density both in vitro and in human excitatory neurons derived from IPSCs.

By analysing the trafficking of different kinesins (KIF1a and KIF5C), they observed different effects of tau phosphorylation status on the movement of these two motors.

They then analyse transport of lysosomes by employing live imaging of lysotracker in human excitatory neurons derived from IPSCs transfected with wildtype, phosphodeficient or phosphomimetic tau observing that phosphodeficient tau seems to reduce transport of lysosomes while phosphomimetic increases transport compared to wildtype tau.

Strengths:

(1) The work aims to study a novel and underexplored topic in the tau field, tau envelopes, and investigate their relevance to Alzheimer's disease pathology.

(2) Experiments are well conducted and of high quality.

Weaknesses:

Relying only on in vitro reconstituted microtubules and human neurons derived from IPSCs leaves some doubts about the relevance of these results for Alzheimer's disease considering the embryonic state of IPSCs-derived neurons, but the authors clearly discuss this point.

Reviewer #2 (Public review):

This manuscript examines how disease-associated hyperphosphorylation disrupts tau's role as a cooperative microtubule-binding regulator of intracellular transport. Using in vitro reconstitution assays and live-cell imaging in iPSC-derived neurons, the authors employ phosphomutant tau constructs (E14 to mimic hyperphosphorylation, AP to prevent phosphorylation) at 14 disease-associated residues to isolate phosphorylation effects independent of expression system-dependent PTM heterogeneity. The results show that hyperphosphorylated tau fails to form cooperative envelope-like structures on microtubules, instead binding diffusely and dissociating rapidly. In contrast, wild-type and phospho-resistant tau form cohesive envelopes that regulate motor protein access. At the single-molecule level, hyperphosphorylation reduces KIF5C inhibition while maintaining or enhancing KIF1A inhibition through altered processivity and detachment rates. In live neurons, hyperphosphorylated tau phenocopies tau knockout conditions, weakening tau-mediated inhibition of lysosome transport and increasing processive motility. The authors quantify tau binding using Gaussian mixture model-based image analysis and measure tau kinetics via FRAP, demonstrating that hyperphosphorylation-induced loss of cooperative binding correlates with dysregulated organelle transport. These findings establish a mechanism by which phosphorylation-driven disruption of tau's gatekeeper function on microtubules compromises axonal transport prior to aggregation in tauopathies.

Comments on revised version.

The authors did a good job responding to my comments and I support publication of the revised manuscript.

Reviewer #1 (Public review):

Summary:

This work presents a flexible spike-sorting framework that allows users to run, swap, and benchmark individual modules commonly used in spike sorting. The paper argues that "opening the black box" is essential for understanding which components drive performance differences and for making progress toward more accurate and transparent spike sorting.

Using this modular benchmarking pipeline, the work identifies electrode drift as a primary bottleneck for accurate sorting, and introduces an end-to-end sorter ("Lupin") that combines the best-performing modules and is reported to be on par or maybe even outperform existing spike-sorting packages on their benchmark. While the modules forming Lupin were chosen based on a single benchmark, end-to-end evaluation of different sorters now use multiple benchmarks, including different available datasets.

Overall, this is a strong tool/resource contribution with clear potential to accelerate spike-sorting development and enable more rigorous comparisons. While not the main point of the paper and therefore less important, remaining claims regarding Lupin outperforming other sorters are not well supported. Lupin does not necessarily outperform all other sorters in real data if you consider 'finding most good units' more important than decreasing false positives - something that quality metrics post-sorting can take care off.

Strengths:

This work has high community value and practical utility. The effort to make benchmarking and spike sorting modules accessible and standardized is substantial, and likely to be broadly useful.

Treating spike sorting as a set of interchangeable modules is a useful approach to some extent, and it enables targeted improvements rather than 'new sorters' popping up which are difficult to fully understand.

Implementing this resource within SpikeInterface, an already widely used tool, will facility uptake and community contributions.<br /> Overall, I am positive about this manuscript as a resource paper. The core framework is compelling and timely.

Weaknesses:

(1) I appreciate the use of automatic curation tools to define the quality of units obtained from different spike sorters. However, a discussion on what is considered a good trade-off is missing. For example, one could argue that as long as you apply these curation tools post-sorting, finding more good-quality units is more important than keeping the number of false positives low, as these can be filtered out at a later stage by these curation methods. The manuscript currently seems to argue the balance between high number of good and low number of false positives is more important. This needs to be discussed and rationalized more explicitly, rather than using terms like 'clearly outperforms' and 'good trade-off'.

(2) Although I agree that overfitting to specific data is a general problem with spike sorting (as you can also change parameters of individual sorters), and perhaps this is even required for the best results, I still miss an explicit discussion of this.

(3) While the end-to-end evaluation is a very good addition, defining the best strategy for each module (which in turn led to choices what to use for Lupin) is still based on one benchmark only. This remains a weakness.

Cmments on revised version:

Previously identified weaknesses in limited support for Lupin being superior to other spike sorters, especially using only a single benchmark, have been largely addressed by 1) making superiority of Lupin less of a point in the manuscript and 2) adding more simulations and real datasets. Also, clarification on serial versus iterative spike sorting has been added to the discussion.

Reviewer #2 (Public review):

Summary

Spike sorting, that is, assigning events detected in extracellular electrophysiology data to firing of individual neurons, is an inherently difficult computational problem involving multiple steps. The difficulty arises from low signal to noise, instability in signal due to relative motion of the tissue and recording sites, and large volumes of data. Experimental ground truth data - where the correct assignment of spikes in known - is not available in large enough quantities to test algorithms. This paper describes a tool for creating fully synthetic ground truth data and benchmarking the individual steps of spike sorting to dissect the impact of signal to noise, firing rate, and motion correction on each step. This information is used to construct an optimized algorithm for sorting these ground truth data. One result of particular interest is the dominant role of motion correction in degrading accuracy. Another important technical result is that motion correction via interpolation of the voltages traces yields similar accuracy to interpolation of the spike templates.

Strengths

The paper shows that useful insight can be gained through analyzing process step by step. While this analysis has also been done in papers presenting spike sorters (for example, Pachitariu (2024)) the tools presented here allow users and developers to do similar studies for their own work. This toolset will be useful to many labs, especially those working in less studied brain areas or model systems, cases where the tuning of standard spike sorting tools is not a good match to the data.

Weaknesses/Limitations:

The model ground truth data used in testing spike sorting and its components does not need to be a perfect match to experimental data to provide useful benchmarking. However, as with all measurements of spike sorting accuracy, extrapolation to experimental data can be complicated. Therefore, the insights gained concerning optimization of the individual steps should be interpreted as "correct for that model data. The comparison of the paper's new sorter to standard sorters on experimental recordings suggests that the benchmarking data is reasonable. Nevertheless, users of these tools will need to assess how well the simulated data matches their recordings.

Reviewer #3 (Public review):

Summary:

In this manuscript, the authors describe two additions to an existing toolbox (SpikeInterface, Buccino et al., 2020, eLife). The first addition is an empirical simulator for extracellular recordings, in which spikes from predefined templates are added up with Gaussian noise. The second addition involves granting user-level access to intermediate processing steps along spike sorting algorithms. The authors demonstrate the toolbox by evaluating functions (e.g., event detection) or sets of functions (e.g., feature extraction + clustering) on their simulated data and suggest that a specific combination of function implementations provides performance improvement relative to kilosort4 (Pachitariu et al., 2024, Nature Methods).

The validity of the work is poor. In particular, the simulator is unrealistic and the ground truth dataset is too short. Several spike sorters are used as straw men, and most sorters compared have never been described in peer-reviewed literature or in sufficient detail. The purely feedforward architecture of the modules is very limiting and irrelevant for modern sorters. Finally, the reporting of results is sporadic and does not follow scientific reporting standards.

General comments:

(1) Abstract, lines 14-16: "We then leverage these results to create a modular component-based spike sorter that can outperform Kilosort 4 on dense and large simulated recordings and produce similar quantitative results on real data." However, the artificial data are not "large" - they are very short. And "similar quantitative results" cannot be assessed on real data because those data do not have any ground truth. Because this is a revision and the authors have already received similar feedback from this Reviewer, I am not sure what to recommend.

(2) In a previous comment, I indicated that the simulator itself is overly simplistic, and indicated that as far as I am concerned, the authors must improve it in one of two ways: (1) use a set of biophysical equations, with multi-compartmental modeling of currents and return currents; (2) use noised data from extracellular recordings; or (3) some combination thereof. The authors explained in their answer - but not in the MS - some of the shortcomings of biophysical simulators, but chose to do neither. My comment therefore remains unaddressed in the MS.

(3) In a previous comment, I indicated that the duration of 10 minutes is too short. The authors chose to extend the duration to 30 minutes, which is insufficient for units that have low firing rates. Units that fire e.g., 0.1 spikes/s would have fewer than 200 spikes. If this MS is to be taken seriously, the simulation should be done on durations that (1) are similar to the potential applications - which may be many hours or even days; (2) allow identification of low-firing neurons. Therefore, about 3 hours is the bare minimum. Therefore, at present, my comment remains unaddressed.

(4) In a previous comment, I indicated that some sorters have never been described in peer reviewed papers and therefore, they must either be removed from the present comparisons or be described in full. The authors chose to persist in relying on un-reviewed online documentation. Thus, my comment remains unaddressed, and the comparisons that involve those sorters (e.g., TDC, TDC2, SpyKING Circus 2) are simply invalid.

(5) In a previous comment, I indicated that some sorters (SpyKING Circus 2 and TDC2) are effectively straw men and suggested to reorganize the MS to demonstrate its main goal. Specifically, I suggested to reorganize the manuscript so that after every module is evaluated separately based on a limited ground truth dataset, a single "best" sorter would be constructed, and then tested extensively (and compared to the de facto state of the art). Such reorganization would both demonstrate the utility of a modular approach and clarify the general usefulness of the outcome. The authors agreed that these are straw men and gave a historical account of why these were included. However, they did not explain these considerations in the MS itself, and chose not to reorganize the MS. Therefore, my comment remains unaddressed.

(6) In a previous comment, I commented about the presentation, description, and interpretation of the results, and indicated that the choice to report point estimates makes any conclusions based on those results invalid. The authors did not make any changes to the reporting. Therefore, none of the results reported in the MS can be taken as an outcome of scientific inquiry. In other words, the MS does not deliver any solid findings - neither scientific nor methodological.

Reviewer #1 (Public review):

Summary:

The authors utilize both human iPSC-derived RPE and human fetal RPE cultures to interrogate the effect of various types of commonly used cell culture media on several key biological- and disease-relevant RPE properties. These include a comparison of RPE morphology, polarity, transepithelial electrical potential, lipid metabolism, autophagy, and targeted metabolomic profiles across 6 different media compositions.

Strengths:

This manuscript is very well written, and data are presented in a well-organized manner. The authors address media composition as a fundamental variable that will influence the interpretation of assays performed in RPE cell cultures, particularly metabolic studies. Figure 6 provides a useful summary of the study's findings across commonly used media types, and the manuscript's discussion offers insight into which media may be best suited to address specific experimental questions. Overall, this manuscript will not only serve as an important resource for vision scientists utilizing RPE culture models, but it also serves to remind the broader cell biology community of the importance of considering the potential (confounding) experimental effect(s) of various culture media and to consider tailoring the selection of culture media types to the specific experimental question.

Weaknesses:

(1) While the authors report that iPSCs were obtained from several healthy patients and that at least two clones were generated from each individual, it is not clear to this reviewer whether the experiments with each culture media type were performed on the same set of iPSC-RPE in each case. The authors mentioned iPSC differentiation variability as a limitation, but it would be helpful to understand (and quantify) the experimental variability that may exist with the same culture media using iPSCs from different patients and/or separate iPSC clones from the same patient.

(2) Since a major purpose of the manuscript is to highlight how cell culture conditions influence RPE biology and metabolism, it would be helpful to also report whether Mycoplasma testing was performed and confirmed to be negative across all cell lines.

(3) The effect of culture media on mean RPE area and hexagonality was compared in this study. Interestingly, Figure 1E demonstrates higher mean RPE cell area but also substantial variability in cell area for media 2 (MEM-alpha and B27) and media 4 (HPLM and B27). It would be helpful to include a discussion of the potential biological implications of variable cell area across these 2 media types.

(4) The authors speculate that FBS-containing media may encourage a more mesenchymal or de-differentiated state. This could be experimentally determined by interrogating mesenchymal markers (alpha-SMA, fibronectin, etc) by immunoblot and/or immunofluorescence microscopy, similar to how the RPE markers were evaluated in Figure 1.

(5) It would be helpful for the discussion section to include a comparison of key differences (where they exist) between iPSC-RPE and fetal RPE across culture media types.

Reviewer #2 (Public review):

Summary:

In this study, Lim et al. provide a comprehensive analysis of the metabolic and physiologic effects of different media compositions on iPSC-RPE. This analysis includes commonly used iPSC-RPE media bases (MEMα, DMEM-HG/F12 basal media) as well as human plasma-like medium (HPLM) in attempts to establish a more physiologically relevant culture environment.

Strengths:

The analyses in this study provide a very thorough survey of metabolic function as well as an RPE-relevant physiologic characterization. This will be a great resource for optimizing assay conditions for disease-based studies using iPSC-RPE.

Weaknesses:

In the Seahorse studies provided in Figure 3. basal readings for OCR are abnormally low compared to Oligomycin treatment and background, suggesting difficulties with the assay. Findings should be taken with caution.

Reviewer #3 (Public review):

Summary:

The authors systematically compare six culture-media formulations using induced pluripotent stem cell-derived retinal pigment epithelium and fetal retinal pigment epithelium. They examine cell morphology, marker expression, barrier function, polarized secretion, lipid accumulation, ultrastructure, mitochondrial respiration, glycolytic function, and intracellular and extracellular metabolites. The results demonstrate that culture-medium composition and the choice of serum or B27 supplementation substantially influence retinal pigment epithelium phenotype and metabolism. Rather than identifying a single optimal medium, the study provides a comparative framework to guide medium selection according to the biological question being investigated.

Strengths:

The head-to-head comparison of six media under otherwise similar culture conditions addresses an important source of variability in retinal pigment epithelium research. The study uses a broad range of complementary approaches, including imaging, transepithelial resistance, electron microscopy, extracellular flux analysis, and targeted metabolomics. The inclusion of both induced pluripotent stem cell-derived and fetal retinal pigment epithelium increases the potential relevance of the findings across different cell sources. The matched comparisons of serum and B27 supplementation within MEMα and human plasma-like medium are particularly informative because they help distinguish supplement-associated effects from those caused by the basal medium. Overall, the dataset has the potential to serve as a valuable resource for selecting culture conditions and interpreting findings across retinal pigment epithelium studies.

Weaknesses:

The most important limitation is that the experimental unit and degree of biological replication are not clearly defined. It is unclear whether individual observations represent independent donors, clones, differentiated lines, culture preparations, wells, images, or sections. This makes it difficult to determine the independence, robustness, and generalizability of several comparisons.

The metabolic analyses also require additional methodological clarification. For intracellular metabolomics, the culture format, cellular biomass, extraction volume, pooling strategy, and normalization method are not reported sufficiently. Normalization of extracellular measurements to unspent medium accounts for differences in starting metabolite abundance but not for differences in cell number or biomass. Similarly, normalization of intracellular signals to medium 1 does not correct for differences in the amount of cellular material extracted.

For the Seahorse experiments, the main figures present unnormalized values even though the media produce differences in cell number, size, and protein content. These raw measurements represent total metabolic activity per well and may not reflect activity per cell. It is also unclear how normalization was performed because the Methods describe cell-count and protein measurements from two wells, whereas the stress tests included five to six wells per condition. In addition, measurements obtained after transfer into a common assay medium reflect metabolic adaptations retained from the preceding culture conditions rather than real-time metabolism within the original media.

Other limitations include insufficient information about the biological replication underlying the sub-RPE deposit analysis and the inability to fully interpret the effects of X-VIVO 10 because its composition is proprietary. Finally, public availability of the underlying metabolomics data would be important for a study intended to serve as a community resource.

Reviewer #1 (Public review):

Summary:

Thach et al. report on the structure and function of trimethylamine N-oxide demethylase (TDM). They identify a novel complex assembly composed of multiple TDM monomers and obtain high-resolution structural information for the catalytic site, including an analysis of its metal composition, which leads them to propose a mechanism for the catalytic reaction.

In addition, the authors describe a novel substrate channel within the TDM complex that connects the N-terminal Zn²⁺-dependent TMAO demethylation domain with the C-terminal tetrahydrofolate (THF)-binding domain. This continuous intramolecular tunnel appears highly optimized for shuttling formaldehyde (HCHO), based on its negative electrostatic properties and restricted width. The authors propose the hypothesis that this channel facilitates the safe transfer of HCHO, enabling its efficient conversion to methylenetetrahydrofolate (MTHF) at the C-terminal domain as a microbial detoxification strategy. Experimental data that shows an involvement of TDM in the reaction of HCHO with THF is less convincing.

Strengths:

The authors provide convincing high-resolution cryo-EM structural evidence (up to 2 Å) revealing an intriguing complex composed of two full monomers and two half-domains. They further present evidence for the metal ion bound at the active site and articulate a hypothesis for the catalytic cycle. Substantial effort is devoted to optimizing and characterizing enzyme activity, including detailed kinetic analyses across a range of pH values, temperatures, and substrate concentrations. Furthermore, the authors validate their structural insights through functional analysis of active-site point mutants.

In addition, the authors identify a continuous channel for formaldehyde (HCHO) passage within the structure and support this interpretation through molecular dynamics simulations. These analyses suggest an exciting mechanism of specific, dynamic, and gated channeling of HCHO. This finding is particularly appealing, as it implies the existence of a unique, completely enclosed conduit that may be of broad interest, including potential applications in bioengineering.

Weaknesses:

Although the idea of an enclosed channel for HCHO is compelling, the experimental evidence supporting enzymatic assistance in the reaction of HCHO with THF is less convincing. The linear regression analysis shown in Figure 1C demonstrates a THF concentration-dependent decrease in HCHO; however, it is well established that HCHO and THF can react spontaneously in a non-enzymatic manner, raising the possibility that the observed effect does not require enzymatic involvement. I appreciate the authors' clarification that the data in Figure 1 were not intended to demonstrate enzymatic channeling or catalytic involvement in the HCHO-THF reaction, and that the assay does not distinguish between changes in HCHO production and downstream consumption. The authors' revised statement "Overall, these findings suggest that TDM-mediated TMAO demethylation generates HCHO, which can subsequently react with THF, potentially linking TMAO breakdown to one-carbon metabolism." is appropriately cautious and leaves open the mechanism underlying this process, which is consistent with the evidence presented.

Overall, the authors were successful in advancing our structural and functional understanding of the TDM complex. They suggest an interesting oligomeric complex composition which in my opinion should be investigated with additional biophysical techniques.

Additionally, they provide an intriguing hypothesis for a new type of substrate channeling. However, additional kinetic experiments focusing on HCHO and THF turnover by enzymatic proximity effects are required to strengthen this potentially fundamental finding. If this channeling mechanism can be supported by stronger experimental evidence, it would substantially advance our understanding and knowledge of biologic conduits and enable future efforts in the design of artificial cascade catalysis systems with high conversion rate and efficiency, as well as detoxification pathways.

Comments on revised version.

It is unfortunate that no additional experimental evidence supporting the proposed channeling mechanism could be provided. In the absence of such evidence, I remain somewhat hesitant to regard the evidence as "solid" rather than "incomplete," particularly given that the authors themselves acknowledge that the current experiments do not establish enzymatic channeling. However, the rest of the manuscript is stronger despite this limitation.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers and the authors have satisfactorily answered the previous reviewer's comments.]

Summary:

They use cultures of insulinoma MIN6 cells that form spheroids in a micro-patterned PEG-hydrogel to measure Ca2+ oscillations in multiple cells simultaneously.

Strengths:

They demonstrate that insulinoma spheroids are formed in multi-well plates and that Ca2+ imaging can be performed on them.

Reviewer #2 (Public review):

Summary:

The study by Robben et al., show 3D beta-cell spheroid platform, a valuable tool allowing high-throughput monitoring of cytoplasmic Ca concentrations and insulin secretion, with Ca signals comparable to those recorded in primary islets. The authors demonstrate a solid method to culturing MIN6 cells in a 3D culture system, recording Ca signals in a high-throughput format and characterizing these Ca signals using pharmacological tools, including TRPM3 channel and K-ATP channel modulators. This highlights the utility of the 3D beta-cell spheroid for screening new ion channel modulators in beta-cells of the pancreas.

Strengths:

- The study shows that the MIN-6-based 3D beta-cell model is better to study Ca-signaling and insulin secretion compared to 2D culture of single MIN-6 cells.<br /> - The method allows imaging of Ca signaling in many spheroids in parallel followed by collecting medium to measure insulin release and correlate both effects.<br /> - The authors demonstrate that this system is suitable for screening new pharmacological modulators and used as an agonist of the ATP-sensitive potassium channel (diazoxide) and the agonist and antagonist of the TRPM3 channel.

Reviewer #3 (Public review):

Summary:

The primary objective of this study is to develop high-throughput screening assays utilizing homogeneous 3D cell cultures that more accurately replicate the intricate architecture and cellular communication found in tissues. The authors have chosen pancreatic islet β-cells as a model system to evaluate agents that modulate insulin release, which is particularly relevant given the increasing prevalence of diabetes mellitus-a significant global health concern. Moreover, the incorporation of human-based 3D spheroids, organoids, or organ-on-chip technologies into drug discovery protocols is essential for enhancing clinical translation, as candidate compounds identified using animal models have often demonstrated limited success in clinical settings.

Strengths:

This study was thoughtfully planned and skillfully carried out. The use of micropatterned hydrogels to observe 19 spheroids at once is an ingenious aspect, which has been effectively validated with Ca microfluorography. Overall, I found this investigation to be exceptionally well-executed and free from notable flaws, as the results clearly back up the conclusions. Additionally, the developed method achieved the proposed aims, providing a high-throughput format with 3D cultures. I believe this study deserves publication.

Reviewer #1 (Public review):

Short overview:

This work demonstrates that exposure to the odor diacetyl in C. elegans first induces the expression of genes that act in the DHAP (dihydroxyacetone phosphate) - glycerol shunt, followed by induction of the flavin-containing monooxygenase fmo-2. This diacetyl exposure also increases the survival of starved animals. However, the mechanism through which diacetyl is sensed by the animal does not involve any of the known receptors of diacetyl, and it is unknown whether any sensory neurons are involved in these responses.

Summary:

This manuscript shows that diacetyl exposure induces metabolic remodeling in C. elegans, where genes in the DHAP (dihydroxyacetone phosphate) - glycerol shunt pathway are activated. This leads to a secondary activation of the flavin-containing monooxygenase fmo-2, which is a longevity-promoting gene upon dietary restriction. The activation of fmo-2 is consistent with the authors' observations that diacetyl exposure increases the survival of starved animals, but not of fed animals. The authors also show that diacetyl exposure improves the animals' resistance to hyperosmotic stress, which is consistent with an increase in glycerol and glycerolipids in these animals. Together, the authors show that a volatile odor or odors can induce gene expression changes that promote metabolic changes and survival under certain environmental conditions.

Strengths:

Through transcriptomics, metabolomics and lipidomics of wild type, with or without diacetyl, and phenotypic analyses of mutant or RNAi-treated animals, the authors delineate a mechanism through which an odor or odors remodel metabolism and increase survival. They show that genes of the DHAP-glycerol shunt are activated during diacetyl exposure, and that these in turn activate fmo-2, which is known to promote longevity under different stressors, like food deprivation. The authors also show that an upstream transcription factor, the co-activator MDT-15, is required for the diacetyl-mediated initial activation of the DHAP-glycerol shunt, and thereby of fmo-2. The involvement of MDT-15 and, to a lesser extent, one of its partner nuclear hormone receptors (NHRs), NHR-49, might also explain the lipidomic changes that accompany diacetyl exposure.

Weaknesses:

The authors tested the involvement of the ODR-10 receptor, which is the known receptor for attractive concentrations of diacetyl, the same concentrations used in this study. Surprisingly, ODR-10 is not required for any of the DHAP-glycerol shunt or fmo-2 induction. In addition, they found that loss of sensory cilia (in the background mutation of the NHR daf-12) enhances diacetyl-mediated fmo-2 induction, which suggests that wild-type sensory cilia inhibit fmo-2 and likely synergizes with the DAF-12 receptor. Considering that chemosensory receptors, like ODR-10, are normally localized to the ciliary endings, their observations suggest a different mechanism through which the animals process this odor. There are at least two possibilities.

One, diacetyl might affect the animals by permeating their cuticles. However, the authors did not test if any neurons are involved in their phenotypes. The AWA neuron, where ODR-10 is expressed, is required to sense attractive concentrations of diacetyl. Thus, what happens when the AWA neurons are ablated or lost?

Two, diacetyl is both light- and temperature-sensitive and can break down into two other odors, acetoin and 2, 3-butanediol. Acetoin is sensed by the chemoattractive AWA neurons (Siddiqui et al, eLife 2024, 101936.1), but not by ODR-10 (Zhang et al., PNAS 1997, vol 94, pp 12162-12167). Thus, this odor would have a different receptor. Although C. elegans chemosensory receptors have been found within the cilia, there is a formal possibility that some of the sensory receptors will also be found on other parts of some sensory neurons, like the dendrites, as in Drosophila (Joseph and Carlson, Trends Genet 2016, vol 31, pp 683-695). Regarding the other diacetyl derivative, 2, 3-butanediol, it has not been shown to elicit any chemosensory responses in C. elegans (Siddiqui et al. eLife 2024, vol 13, RP101936), but 2, 3-butanediol has been linked to the microbiome of fmo knockout mice (Said et al., Metabolomics 2025, vol 21, 170). Thus, it is possible that it is the breakdown products of diacetyl that elicit the responses the authors see.

Finally, the authors' data contradict the observations made by Park et al (Aging Cell 2021, vol 20, e13300). Park et al previously showed that diacetyl exposure reduces the survival of food-deprived animals, although this phenotype is also independent of ODR-10 or of the SRI-14 receptor for aversive concentrations of diacetyl.

Reviewer #2 (Public review):

Short overview:

This study presents potentially important findings showing that DHAP-glycerol shunt involved in energy balance is regulated by food availability in a widely used C. elegans model. The genetic evidence supporting this conclusion is solid and is based on an extensive set of experiments; however, key metabolic measurements and comprehensive metabolic profiling are not provided, limiting the strength of the conclusions about the underlying metabolic and redox changes.

Comments:

Giorda and colleagues report interesting findings demonstrating that the DHAP-Gro3P shuttle is modulated by food availability in C. elegans. Although the authors provide multiple interesting observations in worms, supported by an extensive number of experiments, the metabolic aspect of the study requires additional development. It appears that targeted lipidomics and metabolomics analyses were performed, but the corresponding datasets are largely absent from the manuscript. Only a very limited subset of lipid species is presented in Fig. 2D. What about triglycerides? It would be highly informative to include comprehensive lipidomic profiles covering major lipid classes. A similar concern applies to the metabolomics data. Where are the measurements of Gro3P, DHAP, and glycerol? The authors state that their LC-MS method was unsuccessful and that glycerol levels were ultimately measured using a commercial kit. Given that glycerol production and excretion appear to be major output across many of the experiments presented, this approach is not entirely satisfactory. Reliable GC-MS based methods are available for the quantification of all major components of this pathway, including Gro3P, DHAP, and glycerol (derivatization helps to preserve these species, especially glycerol).

Furthermore, comprehensive LC-MS/GC-MS-based metabolic profiling should be included. Metabolites reported and organized by pathway (e.g., glycolysis, TCA cycle, pentose phosphate pathway) would provide a broader understanding of the metabolic consequences of DHAP-glycerol shunt activation.

Finally, because the DHAP-glycerol shunt is closely linked to cellular redox homeostasis, it would be important to determine how its activation affects intracellular pyridine nucleotide pools, and measurements of NAD+, NADH, NADPH, NADP+ would substantially strengthen the mechanistic conclusions and provide direct evidence for alterations in cellular redox state.

Reviewer #3 (Public review):

Short overview:

This study asks whether the perception of a volatile and attractive cue, diacetyl, leads to changes in metabolic nutrient-responsive programs in C. elegans. Using multi-omic and genetic evidence, they connect the transcriptional response to diacetyl to early activation of the DHAP-glycerol shunt, suggesting worms may activate this metabolic pathway in preparation for food intake. This work also identifies transcription factors involved, and while it does not test whether this response is diacetyl-specific, could identify a conserved pathway of food intake preparation.

Summary:

This study asks whether C. elegans can use a volatile food cue alone, in the absence of ingestion, to anticipate nutrient availability. Using the attractive odorant diacetyl, the authors show that fasting worms rapidly induce the DHAP-Glycerol shunt, a metabolic pathway normally associated with glucotoxicity and hyperosmotic stress, and that this response depends on the transcription factor MDT-15. This drives measurable metabolic rewiring (glycerol and phosphatidylglycerol accumulation) and confers protection against subsequent hyperosmotic stress. Prolonged diacetyl exposure further triggers a second, HLH-30-dependent wave of fmo-2 expression linked to depleted NTP levels, resulting in enhanced heat tolerance and food-seeking behavior upon refeeding. Together, the authors build a testable model for a pathway linking sensory cue detection to gene expression, metabolism, and physiological changes.

Strengths:

The central finding that smell alone without ingestion activates a metabolic-stress adaptation pathway is highly interesting and supported by a convergence of methods including RNA-seq, transcriptional reporters, metabolomics, and functional behavior assays. The transcriptomic time course distinguishes two temporally separable gene expression waves (the shunt at 30 min and fmo-2 at 90 min), and epistasis experiments comparing food status and osmotic stress (Fig. 1i-j) convincingly argue diacetyl acts as a food-predictive cue rather than mimicking osmotic stress. A particular strength is the test of the relationship between the two pathways identified in the study. RNAi knockdown of DHAP-Glycerol shunt enzymes block fmo-2 induction, and pretreatment with salt then switching to food deprivation shows that it is the shunt's activation state that drives fmo-2 expression. This is further reinforced by depletion of energetic mechanism (NTP/ATP). Figure 5 extends this upstream to MDT-15 validated through gene expression, metabolomics, and functional readouts. Throughout the study, transcriptional and metabolic findings are paired with functional outcomes such as hyperosmotic protection, heat stress resistance, or survival, strengthening the paper's model. Together, the authors largely achieve their aim of establishing that a food-related olfactory cue is sufficient to trigger an anticipatory metabolic and transcriptional program. The core claim that diacetyl sensing activates the shunt and subsequent fmo-2 induction and physiological protection is well supported by convergent genetic and biochemical experiments.

Weaknesses:

The authors show that diacetyl-induced fmo-2 induction does not require canonical olfactory sensing since neither diacetyl receptor mutants (odr-10, sri-14) nor a cilia-deficient strain (daf-19; daf-12) blocked the response. However, the pathway characterized here is defined almost entirely through diacetyl, and it remains unclear whether the anticipatory response reflects general food-predictive olfaction or a diacetyl-specific effect. This distinction is important because the authors' central hypothesis is whether volatile food cues in general can be used to anticipate nutrient availability. Testing a limited number of additional attractive odorants, ideally sensed through distinct chemosensory receptors, for a limited number of phenotypes, would establish whether this pathway is generalized to food-predictive smells or only diacetyl. This would increase the impact of the work. Identifying the mechanism(s) for diacetyl perception would as well, but is much more challenging and less likely to be feasible in this work.

The RNA-seq and reporter data disagree on when fmo-2 induction begins, and thus do not fully support that there are two temporally distinct waves. RNA-seq (sampled at 5, 15, 30, and 90 min) shows fmo-2 is not a significantly differentially expressed gene at 30 min and only reaches significance at 90 min, while the fmo-2 transcriptional reporter (Fig.S3a) shows detectable induction as early as 30 minutes. Other readouts in the paper use the reporter for fmo-2 but qPCR for shunt genes, further muddling the timing since mRNA should precede reporter visualization. Since reporter signal generally lags transcript level detection, this discrepancy could be further clarified through a time-course qPCR (like what was tested for the shunt genes) between 30 and 90 minutes. This would help establish whether the two waves are separated by time or whether this reflects a difference in detection method.

Minor weaknesses:

In Figure 1i, gpdh-1 induction is compared between control and 200 mM NaCl (3 hr exposure), with food present or absent. While this directly tests food status, a 3-hour exposure approaches the ~5-6-hour window previously shown to be sufficient for salt-food associative learning to form (worms move toward the high-salt side of a gradient plate if pretreated with high salt and food). This raises the possibility that, in the food-present condition, part of the measured gpdh-1 response could reflect an emerging learned association between salt and food forming during the assay itself, rather than purely reflecting an interaction between nutrient status and osmotic stress signaling. Testing a shorter exposure window (e.g., under 1 hour, as used for the diacetyl exposure in panel j) would help clarify this.

Thrashing is used throughout the paper as the primary readout for hyperosmotic stress resistance, including the central result that diacetyl protects against subsequent hyperosmotic stress (Fig. 2g). However, the authors don't explain why this was chosen as the main stress-resistance readout.

nhr-49 knockdown reduces gpdh-1 induction in diacetyl-mediated hyperosmotic protection but has no effect on development or survival on sustained hyperosmotic stress, raising the question of how the role of nhr-49 may be distinct from mdt-15 in this context.

Reviewer #1 (Public review):

Summary:

This work investigated the associations between abstraction and metacognition in the context of reward-guided learning and transdiagnostic symptom dimensions in a sample of N = 249 participants. Participants completed a reward-guided learning task and confidence judgements. Transdiagnostic dimensions used to examine associations with abstraction and metacognition were based on a large existing dataset. Findings showed that in the examined sample, a Compulsive Hypersensitivity dimension was negatively associated with abstraction and metacognitive sensitivity, while a Social Withdrawal dimension was positively associated with metacognitive sensitivity. These data add to the existing literature on associations between metacognition and transdiagnostic symptom dimensions and extend previous work on the association between abstraction and these dimensions.

Strengths:

(1) The study addresses an interesting research question and uses a transdiagnostic approach. While part of the research question is a replication (metacognition), the additional inclusion of an abstraction parameter is highly valuable.

(2) Methodologically, the study is strong. Specifically, the implementation of an experimental task to estimate parameters, the hierarchical Bayesian modelling, the parameter recovery analyses, the bootstrap regression (including corrections for multiple comparisons) and the control of relevant covariates and response tendencies are quite impressive.

(3) The Open Science approach is laudable in general. The study was preregistered and provides open data and open code. Deviations from the preregistration are transparently reported (e.g., bootstrap regression, exclusion criteria). In this vein, the high number of robustness analyses provided in the supplements is very much appreciated.

(4) More generally, the extensiveness of the supplements is particularly valuable.

(5) Finally, it is very useful that the discussion takes into account potential alternative explanations of the findings.

Weaknesses:

(1) High number of exclusions:

As the authors mention (also in their limitations section), more than half of the participants had to be excluded (only 249 out of 512 participants remained), which is substantial. Specifically, 203 participants were excluded because they failed attention checks, 58 failed comprehension questions on the confidence scale, 25 had a reading time of the instruction page below 5 seconds, and 68 showed a performance that was too poor (the sum is probably higher than 263 because these overlap). This high number of excluded cases resulted in a substantially smaller analysed sample than planned (corresponding to approximately 77% power instead of 90%) and potentially limited both statistical power and generalisability. It might even be the case that highly impulsive individuals were excluded systematically. That is, the exclusion criteria may themselves be associated with psychiatric symptoms, so the analysed sample may no longer (fully) represent the target population. It would be interesting to see whether the results also hold when including the excluded participants (because excluded and included participants differ in attentional and impulsivity-related symptoms). A sensitivity analysis would be beneficial.

(2) Deviations from preregistration:

While the preregistration of the study is positive in general, some aspects raise doubts here. First, there have been quite a few deviations from the preregistration. For instance, the primary outcome variable for abstraction has been changed (initially: proportion of blocks in which the Abstract RL model had a better fit than the Feature RL model; instead: mean posterior responsibility of the Abstract RL model across blocks), and this appears to have influenced the findings (e.g., Supplementary Figures: 8 vs. 9). Generally, these deviations introduce additional researcher degrees of freedom and therefore warrant a more detailed justification (or replication in future studies). Second, it appears that the preregistration was uploaded only to an OSF folder and not formally registered with a timestamp. However, while an update to the document has been made according to the metadata, the content still seems to be the same as the original one (created on Jun 11, 2025).

(3) Combined measurement of choices and metacognition:

Choice behaviour and metacognition were not measured independently (i.e., they were measured using a single slider). This challenges the interpretation of results concerning metacognitive sensitivity, as the authors note themselves in the limitations section (even if the effect of starting position was small). It should be clarified whether responses exactly at the centre of the slider were possible or not (and what range the slider had).

(4) Generalisability of findings:

It is questionable whether transdiagnostic factors derived from a Japanese population can be transferred to the sample in this work, especially because it seems to be composed of a heterogeneous international sample comprising participants from multiple countries (South Africa, United States, United Kingdom, Poland, others). While it is appreciated that the authors validate the transdiagnostic factors through a new exploratory factor analysis and by correlating item loadings across samples, the correlations are only moderate and point at least to some degree of variation. In addition, for factor 2 (social withdrawal?), the correlation seems to be mainly existent due to two clusters of items, challenging the validation approach in itself. Also, apart from the correlations themselves, the absolute values of the item loadings appear to vary largely. Considering this, the conclusion that "transdiagnostic symptom dimensions appear broadly consistent across samples (even across countries)" (p. 9) definitely goes too far.

(5) Parameter recovery analysis: The parameter recovery analysis is appreciated; however, the recovery of the learning rate in particular seems to be rather low (r=.40). This raises concerns regarding the validity of the parameter estimates.

(6) Effect sizes: The reported regression coefficients appear relatively small in magnitude (i.e., betas of the associations between metacognition/abstraction and transdiagnostic dimensions appear to be in the range around .02-.04). If these are standardised estimates, the corresponding effect sizes are relatively small. If not, I recommend reporting standardised effect sizes in addition.

Overall, the study provides relevant replications and new insights into the association between reward-guided learning behaviour (abstraction, metacognition) and transdiagnostic symptom dimensions. It should be noted that effect sizes appear to be rather low, however. The analytic approach is generally strong. At the same time, the high number of exclusions, the limitations regarding the validation of the transferred transdiagnostic factor structure, and the limited recovery of the learning rate parameter clearly raise important questions regarding the robustness and generalizability of findings. Given this, any interpretations regarding potential therapeutic implications (e.g., p. 10) should be made with caution.

Reviewer #2 (Public review):

Summary:

In this study, Oka and colleagues recruited an online sample to complete a previously validated abstraction task (Cortese et al., 2021) alongside confidence ratings and a large psychiatric questionnaire battery, which included a variety of methods to screen out inattentive or otherwise biased responders. Questionnaire item scores were combined with factor weights from a large dataset to estimate transdiagnostic factor scores. A computational model was then fit in a hierarchical manner to the abstraction task data, with individuals' fit to an "Abstract RL" model used as a metric of individual-level abstraction ability, and metacognitive bias and sensitivity were estimated from the confidence ratings. Associations between these task-derived measures and both dimensional and symptom-level measures of psychopathology were then estimated using multiple regression. The key findings were that, while metacognitive sensitivity and abstraction ability were associated with symptom-level scores, the associations with transdiagnostic dimensions - higher compulsivity associated with lower abstraction ability and metacognitive sensitivity; higher social withdrawal was associated with higher metacognitive sensitivity - were interpreted by the authors as more coherent.

Strengths:

(1) Robust screening for inattentive responders through catch questions (Zorowitz et al., 2023), as well as incorporating recent recommendations regarding response bias (Sarna et al., 2026).

(2) Assessed the cross-cultural generalisability of the imported factor weights by comparing item loadings from a large external sample against a de novo exploratory factor analysis in their sample.

(3) Directly compares a theory-driven model-defined abstraction metric to metacognition in relation to dimensional and symptom-level measures of psychopathology.

(4) Pre-registered analyses, with deviations from pre-registration clearly stated.

Weaknesses:

(1) The abstraction metric (mean posterior responsibility of the Abstract RL model) differed from the pre-registered metric and has not been validated here for reliability (e.g., split-half across the blocks or similar).

(2) Model recovery is not shown, so it's not clear whether the Abstract RL and Feature RL models are fully dissociable in this task design.

(3) Metacognitive measures are behaviourally defined (AUROC2 for sensitivity and mean confidence for bias), but models do not correct for task accuracy, which may be related to both.

(4) Dimensional and symptom regressions differ: the dimensions are entered in one model, but the symptom measures are entered into separate regressions and the marginal effects corrected for multiple comparisons. If I've understood this correctly, this means that the dimensional coefficients are partial associations adjusting for the other two factors, whereas the symptom-level coefficients are marginal and FDR-corrected, making it difficult to directly compare them.

Additional questions and context:

(1) Could split-half reliability (e.g. odd vs even blocks) be reported for the abstraction measure? Relatedly, a model recovery/confusion analysis for the two abstraction models, and/or posterior predictive checks showing that the two models generate behaviour resembling that of participants would help establish that the responsibility metric is able to dissociate the different abstraction strategies.

(2) Supplementary Table 2 shows the results for the pre-registered discrete proportion metric - here, there is limited evidence (p=0.220) of an association between abstraction and compulsivity, so saying they are "almost consistent" is perhaps a little overstated. Though the argument for using the alternative continuous metric is justified in the text, it's not quite clear whether the difference is due to the inference method or the abstraction metric itself - the bootstrapped analysis of the pre-registered metric is not reported, nor is the analysis without bootstrapping of the continuous metric (I think this may have been what Supplementary Table 1 was meant to report, but currently it's identical to Supplementary Table 5). In addition, it might be helpful if the correlation between the two metrics were presented graphically.

(3) In the Methods and Supplement, the authors mention that they had pre-registered running a sensitivity analysis including excluded participants. This might be interesting given the high exclusion rate, and given that most exclusions were not based on task behaviour (chance-level choosing). If there is concern about shifting group-level parameter distributions, then this could be explicitly included in the model by including an offset on group-level parameters (i.e., interaction term) on excluded participants, which would allow them to systematically differ in model parameters. Alternatively, one could at least estimate the abstraction and metacognition metrics in the excluded sample (perhaps restricted to those excluded on questionnaire-based criteria rather than task performance) to see whether they do indeed differ.

(4) How do factor scores relate to task accuracy - do those with higher compulsivity perform worse, and is this plausibly related to less abstraction?

(5) How do the factors extracted here compare to those in other studies, such as those from Gillan et al. (2016, eLife)? In particular, it'd be interesting to know what questionnaires/symptoms in the "Compulsive hypersensitivity" factor in the present study overlap with the Compulsive behaviour/intrusive thoughts factor from that earlier work, as the latter has been strongly associated with metacognitive measures - higher metacognitive efficiency for anxious/depression, lower metacognitive efficiency for compulsive behaviour - in previous work (Rouault et al., 2018). That three-factor structure also included a factor they labelled "Social withdrawal" - is it similar to the one presented here, or is the one here (including distress) more like their anxious depressive factor?

(6) In the Discussion, the authors state "Our findings are also consistent with previous converging evidence linking compulsive tendencies to less efficient computation and a preference for familiar over goal-directed action". That the Feature RL model might fairly be called less efficient is reasonable, but I'm not sure how the reduction of features in the Abstract RL model relates to goal-directed action (they're both model-free RL algorithms).

(7) The Discussion also mentions "models that integrate abstract and metacognitive representations" - was there a reason these could not be applied in the present study?

Reviewer #3 (Public review):

Summary:

This is an interesting study investigating the relationship between transdiagnostic symptom profiles and computational parameters of abstraction and metacognition. The authors found that a transdiagnostic dimension they term compulsive hypersensitivity was negatively related to abstraction ability and metacognitive sensitivity, and a dimension termed social withdrawal was positively related to metacognitive sensitivity. Overall, the question of whether and how higher-level cognitive processes relate to transdiagnostic psychiatric factors is interesting and addresses a relevant gap in the literature.

Strengths:

This is an overall well-designed study, and the authors have clearly given considerable thought to data quality and validation during the study design. This is evident, for instance, in the use of multiple approaches to assess inattentive responding and acquiescence tendencies. In addition, the use of advanced modelling approaches for the task data represents a strength and allows the authors to capture individual differences in task behaviour in a more nuanced and mechanistic manner in comparison to what would have been possible with model-agnostic analyses.

Weaknesses:

Nevertheless, I would like to raise several concerns, detailed below.

(1) My most pressing concern regards the exclusion rate of roughly half the sample. Of the 512 participants who were tested, only 249 (48.6%) were included in the final analyses, meaning that more than half of the recruited participants were excluded. Although the authors state that these exclusions were based on preregistered criteria, this high exclusion rate still warrants further investigation. Pre-registering exclusion criteria does not eliminate the potential for selection bias or establish that the resulting sample is representative of the recruited sample. In fact, the authors even report that the excluded participants differed significantly on several psychiatric scores. I believe that a detailed account of the exclusion, including how included and excluded participants differed on relevant demographic or study variables, would be beneficial.

Additionally, the authors state that a sensitivity analysis including participants excluded from the primary analysis was preregistered but was not conducted because the excluded and included participants differed significantly. This does not seem a sufficient justification for omitting a preregistered sensitivity analysis; in fact, systematic differences between included and excluded participants warrant this kind of sensitivity analysis. I agree with the authors that this may lead to changes in task parameter estimates. However, I believe this change would be an informative result of the sensitivity analyses rather than a methodological problem to avoid.

Finally, I would like the authors to clarify some inconsistencies in the preregistered exclusion criteria. In the pre-registration, they first state that all participants who fail any infrequency question will be excluded. Later in the pre-registration, they state that participants with two or more failed questions will be excluded and that a sensitivity analysis will be done, in which participants with only one failed question are retained. Neither of these criteria matches what is reported in the manuscript ("Attention check + infrequency item mistake more than 2").

(2) The current sample size of n = 249 participants is not sufficient for a factor analysis with 176 questionnaire items, and I believe that the solution of using factor weights from a previous larger sample is generally sensible. I also commend the authors for wanting to assess the validity of this approach. However, both the methods and results reported for the comparison between the factor analysis based on the current, smaller sample and the large, previous sample are insufficient and warrant substantially more detail. Firstly, it is unclear to me whether the three-factor solution in the factor analysis with the current, smaller sample was empirically grounded or whether three factors were extracted to match the three-factor solution from the larger sample. For both factor analyses, I would recommend reporting the factor extraction methods, the criterion used to determine the number of factors, and whether alternative factor solutions were considered. Secondly, it is unclear what exactly was correlated across the two factor analyses (e.g., factor scores, item loadings, factor weights?). Finally, I do not believe that the result of significant correlation is sufficient to conclude that the dimensions replicate between samples, particularly given they indicate only moderate correspondence (r = 0.46, for instance, corresponds to only 21% shared variance), thereby providing very limited evidence for replication of the factor structure.

These concerns are particularly pressing given that the authors state the consistency of transdiagnostic symptom dimensions across samples and countries as a primary result in the discussion.

(3) It is also unclear to me whether any exclusion was applied based on the first part of the Deary-Liewald reaction time task. The supplement implies that it was ("This task result was used to [...] exclude participants who exhibit problematic behaviours"), but I could not find a corresponding report in the manuscript.

(4) The manuscript refers to an attentional check questionnaire used to identify and exclude inattentive participants, but does not provide a reference for this measure or describe the questions included. Given the substantial overall exclusion rate, it is particularly important that all exclusion procedures and criteria are described in sufficient detail to allow readers to assess their appropriateness and reproducibility. The authors should therefore provide the relevant reference and/or report the specific questions and criteria used to determine inattention.

Relatedly, when describing the control-neutral items in the Supplementary Materials, the authors state that further details are provided in the Supplementary Materials. However, I cannot find another, separate Supplementary Material that may contain this information

Reviewer #1 (Public review):

Howell et al investigate the functional capacities of CD16dim CD56dim NK cells, including their activity against HIV-1-infected T cells. The authors provide an extensive characterization of the functional activity of different NK cell subsets derived from peripheral blood. CD16 is an important receptor expressed on NK cells, and previous studies have demonstrated that CD16 expression changes depending on the activation status of NK cells - one important regulator of CD16 expression is proteolytic shedding/cleavage of CD16 on activated NK cells by the metalloprotease ADAM17. In vitro activation of NK cells, for example in response to K562 cells or other target cells, results in a rapid downregulation of the expression of CD16 on the surface of NK cells, unless an ADAM17 inhibitor is added. This is an important point to consider in the interpretation of the presented results. Overall, the manuscript includes many data in nine figures plus supplemental figures, and would benefit from some focusing of the results.

(1) Figure 1<br /> The observation that CD16dim NK cells responded more strongly by degranulation to K562 cells and HIV-1-infected cells could be due to the shedding of CD16 following activation. In other words, more strongly activated NK cells express higher levels of CD107 but also shed CD16, resulting in higher CD107 expression in CD16low NK cells. The authors should investigate this, for example by performing the degranulation assays shown in Figure 1 in the presence and absence of an ADAM17 inhibitor.

(2) Figure 2<br /> The authors sorted CD16dim and bright NK cells for these experiments and observed higher lysis of HIV-1-infected CD4+ T cells. Important controls should be included in these experiments - how strong was the lysis of HIV-1-uninfected CD4+ T cells by these different NK cell subsets? It also appears that the results shown were derived using NK cells from one donor, and "representative of two independent sort experiments performed with separate donors, each yielding similar results". Why are the authors now showing the respective data? One or two experiments appear too few to come to these conclusions. To support the broad conclusions drawn by the reviewers, the experiments should be performed in a larger number of individuals.

(3) Figure 3<br /> It appears that experiments were performed again using bulk NK cell populations, and superior degranulation and killing frequencies by CD16dim NK cells might reflect different levels of activation again, as described above for Figure 1. The same applies to Figure 4 - lower degranulation events in CD16bright NK cells are consistent with lower activation of these cells, resulting in less CD16 downregulation. Also, it is not clear to the reviewer why CD107a expression and killing frequencies decrease with higher effector-to-target ratios (Figure 3).

(4) Pages 19-25<br /> It would be helpful if the authors could provide some conclusions regarding their findings - it is very difficult for the reader to follow the many reported frequencies and p-values. What does this actually mean? Overall, the results appear to follow prior observations that licensed (KIR3DL+) NK cells respond more strongly than unlicensed (KIR3DL1neg) NK cells. The consistent observation within these different subanalyses that CD16dim NK cells degranulate more than CD16bright NK cells is probably the result of activation-induced CD16 downregulation in these assays, as mentioned above. Providing two-way ANOVA analysis results for these very many observations would furthermore require, in the opinion of the reviewer, adjustments for multiple comparisons.

(5) Figures 5 and 6<br /> These figures demonstrate that NK cell-mediated activation by HIV-1-infected cells depends on NKG2D ligands and can be inhibited by blocking this interaction - this is consistent with data presented by the Barker group and others previously, and does not provide new information.

(6) Figure 7<br /> The authors extended their functional analyses of NK cells to ADCC function. It is very well established that CD16 is downregulated in the context of ADCC following activation of NK cells. Consistent with this, higher degranulation is observed by CD16dim NK cells.

(7) The data using ADAM17 inhibition in the final figures of the manuscript<br /> These data are of interest, but should be presented in a more structured way. First of all, does the addition of ADAM17 inhibitors change the overall proportion of CD16bright and dim NK cells following activation, independent of whether these cells degranulate or not? Overall, the proportion of CD16dim NK cells that degranulate appears to be reduced in the presence of the ADAM inhibitor, which is consistent with reduced CD16 shedding and maintenance of CD16 expression on activated NK cells - and this is supported by the increase in CD107a-positive NK cells that express CD16 (Figure 8a). Overall, the differences between CD16bright and dim NK cells in their level of activation appear to disappear in the presence of an ADAM17 inhibitor, based on the data shown in Figure 8b, suggesting that CD16 downregulation is occurring in response to activation of NK cells as a consequence of CD16 shedding, and can be inhibited by an ADAM17 inhibitor.

Taken together, many of the data presented in the manuscript are consistent with the very well-established downregulation of CD16 expression on activated NK cells, suggesting that the observed association between reduced CD16 expression on CD56dim NK cells and enhanced effector functions is a consequence of higher activation of these NK cells.

Reviewer #2 (Public review):

Summary:

This study investigates the cytotoxic activity of human NK-cell subsets against autologous HIV-1-infected CD4 T cells and identifies CD56dimCD16dim NK cells as the dominant effector population. The authors propose that this subset possesses superior cytotoxic activity compared with CD56dimCD16bright NK cells and could therefore represent an attractive target for HIV cure strategies. While the study addresses an important and clinically relevant question, several of its major conclusions rely on assumptions that are not adequately supported by the experimental design. In particular, CD16 is treated as a stable phenotypic marker throughout most of the study despite its well-established and rapid downregulation following NK cell activation.

Strengths:

(1) The study addresses an important and clinically relevant question regarding which NK cell subset is responsible for the elimination of autologous HIV-1-infected cells. To the best of my knowledge, this is the first study directly comparing the anti-HIV functional activities of CD56dimCD16dim vs CD56dimCD16bright NK cells.

(2) The experiments performed with purified NK cell subset (Figure 2) provide some evidence that CD56dimCD16dim NK cells possess enhanced cytotoxic activity relative to CD56dimCD16bright NK cells. This experimental approach is considerably more convincing than the analyses performed on mixed NK cell populations and should be expanded throughout the study.

Weaknesses:

(1) The central conclusion is weakened by the use of CD16 as a stable phenotypic marker. CD16 is well established to be rapidly downregulated following NK-cell activation and target cell (K562 or infected cells) engagement through ADAM17-mediated shedding. NK cell shedding regulates NK cell effector functions by promoting target cell detachment, boosting serial killing capacity, and preventing overstimulation. Therefore, NK cells displaying a CD56dimCD16dim phenotype after co-culture cannot be assumed to represent a pre-existing subset with intrinsically superior cytotoxic activity, but may instead correspond to activated CD56dimCD16bright NK cells that have downregulated CD16 during the assay. Because the vast majority of the functional experiments classified NK cell subsets based on post-assay CD16 expression, it is difficult to distinguish intrinsic functional differences between NK cell subsets from activation-induced phenotypic conversion. This limitation affects the interpretation of most of the study's principal findings.

(2) The "killing frequency" analysis presented in Figure 3 is based on a mathematical estimate rather than a direct experimental measurement. Since total target cell killing is measured in mixed NK cell populations, it cannot be attributed to individual NK cell subsets. This experiment must be repeated using purified NK cell subsets.

(3) The serial degranulation assay presented in Figure 4 does not directly measure serial target cell killing and therefore does not support the conclusion that CD56dimCD16dim NK cells possess superior serial killing capacity. Furthermore, the increased serial degranulation observed in the CD16dim population could simply reflect activation-induced CD16 downregulation rather than an intrinsic property of this subset. This experiment should therefore be repeated using purified NK cell subsets.

(4) The finding that CD56dimCD16dim NK cells exhibit greater ADCC activity is somewhat counterintuitive given the central role of CD16 in mediating ADCC. Moreover, these experiments are likely confounded by activation-induced CD16 downregulation, which is expected to be even more pronounced during ADCC. Thus, the apparent superiority of the CD56dimCD16dim subset may simply reflect the conversion of activated CD56dimCD16bright NK cells into the CD16dim gate rather than intrinsically greater ADCC activity. To directly compare the intrinsic ADCC capacity of each subset, these experiments should be repeated using purified NK cell populations prior to target-cell stimulation.

Reviewer #1 (Public review):

This work identifies two lectin receptor-like proteins as cell wall -plasma membrane anchors that contribute to persistent attachment sites maintained during plasmolysis. The authors propose that these attachment sites are important for plant resistance to hyperosmotic stress. Although the existence of persistent attachment sites between the cell wall and the plasma membrane, particularly evident in plasmolysed cells, was recognised long ago, the molecular components tethering the two cellular components remain largely unknown. Therefore, the findings presented by Arico et al. address an important question in plant cell biology.

Through a screening of potential anchors, they found that the overexpression of fluorescently tagged LecRK-I.9, LecTM, AGP18 and AT14A in Nicotiana benthamiana increased the density of Hechtian strands in plasmolysed cotyledon epidermis cells. They show that for LecRK-I.9* (* indicates kinase-dead version) and LecTM, this effect depends on the Lectin domain. Focusing on LecRK-I.9*, the overexpressed Lectin domain localized to cell walls, accumulating in certain foci. The authors interpret this as a possible preference for certain cell wall composition. I find these results convincing and the methodology robust.

The second part of the manuscript, however, relies on interpretations that, in my opinion, are not fully supported by the presented evidence. The authors move to Arabidopsis thaliana and show that overexpression of both LecRK-I.9* and LecRK-I.9*ΔLec fluorescent reporters also localizes to the plasma membrane and Hechtian strands in plasmolysed cells, although the density of Hechtian strands is not quantified. Thus, it is unclear whether LecRK-I.9* promotes Lectin domain-dependent strong cell wall-plasma membrane attachment sites in Arabidopsis.

The authors focus on the formation of big signal clusters at the plasma membrane in response to hyperosmotic treatment. They studied the dynamics of the clusters upon treatment application and observed increased mobility of LecRK-I.9* compared to LecRK-I.9*ΔLec and a plasma membrane marker. They then analysed the abundance and size (not the mobility) of these clusters on different plasma membranes facing cell walls that have or are predicted to have different mechanical and chemical properties, identifying differences between LecRK-I.9* and LecRK-I.9*ΔLec. Although the reduced mobility of LecRK-I.9 relative to LecRK-I.9ΔLec is consistent with an interaction between the lectin domain and the cell wall, it does not by itself demonstrate that the observed clusters correspond to CW attachment sites. Moreover, the relation between the clusters and Hechtian strands (bona fide cell wall-plasma membrane attachments) is not explored. Likewise, the differential clustering observed on different cell faces is intriguing but could have different interpretations.

Finally, the physiological relevance of the proposed anchoring mechanism is supported by osmotic stress assays, but the scoring method relies on manual classification of resistant seedlings and could benefit from a more objective quantitative readout.

In summary, although I find all these results valuable, I find that the methodology is not completely adequate and that several aspects of the data interpretation require additional support or clarification before the central conclusions can be fully justified.

Reviewer #2 (Public review):

Summary:

The manuscript submitted by Arico and co-workers describes the impact of two lectin-domain-containing proteins (LecRK-I.9* and LecTM) on the formation and persistence of Hechtian strands. Based on a survey of selected candidates, overexpression of these two proteins resulted in an increase in Hechtian strand formation. Removal of the lectin domains and expression of this variant did not alter HS formation compared to the WT. In addition, the existence of the lectin domain reduced protein mobility, probably due to interactions with the wall. Last but not least, overexpression of LecRK-I.9 increased the resistance of plants towards water loss conditions.

Strengths:

The study seems well conducted, but may require some small additions. While the results themselves seem not surprising, I think that this is a valuable demonstration of the cell wall binding ability of lectin proteins and its physiological and microscopical consequences.

Weaknesses:

At this stage, some of the study would benefit from some additional quantification.

Reviewer #3 (Public review):

Summary:

The cell wall and plasma membrane are tightly associated in plant cells, but even after plasmolysis, sites of strong contact remain between the plasma membrane and cell wall. Several hypotheses have been introduced about the molecular makeup of these sites of PM-CW adhesion (e.g., Rui et al 2026 Cell; Qin et al 2026 Current Biol; Pérez-Sancho et al 2025 Cell). Here, the authors implicate two transmembrane lectin proteins in PM-CW adhesion via overexpression in Nicotiana benthamiana and via Arabidopsis knockout phenotypes for one of these candidates, the lectin receptor kinase LecRK-1.9. They further show that the PM-CW adhesion function of LecRK-1.9 requires the extracellular domain, suggesting that this lectin-like domain may interact with the cell wall. Interestingly, LecRK-1.9 has also been implicated in extracellular ATP binding in the context of biotic and abiotic stress responses (e.g., Choi et al 2014 Science).

Strengths:

Overall, the work is carefully conducted with high-quality imaging and quantitative image analysis. The results present an interesting candidate for future studies of plasma membrane-to-cell-wall attachment.

Weaknesses:

There are two major caveats to this work. First, all work was conducted with the kinase-dead version of LecRK-1.9, which eliminates a significant biological function of this protein, as evidenced by the major differences in expression pattern of wild-type vs kinase-dead LecRK-1.9. Second, controls are essential to document the expression levels of different protein variants and controls, since LecRK-1.9 expression is correlated with Hechtian strand density.