Reviewer #1 (Public review):

In this study, Chaudun et al use optogenetics and genetically encoded neuromodulator sensors to test to what extent dopamine neuron stimulation produces striatal serotonin release, and vice versa. The study is timely given growing interest in dopamine/serotonin interactions, but important controls are missing. The authors find that striatal dopamine and serotonin afferents function largely independently, with dopamine neuron stimulation producing no striatal serotonin release and serotonin neuron stimulation (possibly) producing minimal striatal dopamine release.

In the revised manuscript draft, the authors increased sample sizes for some important experiments, removed inconclusive electrophysiology data, reorganized the figures, and added new anatomical tracing experiments. The critical fluorophore-only controls for the optostimulation experiments requested by reviewers were not done.

Unfortunately, with the exception of increased samples sizes, these revisions have done little to substantially improve the strength of the evidence behind the authors' conclusions. These conclusions, as reflected in the new title of the paper, have also slightly changed in a manner that puts them further at odds with the data presented. As a result, my opinion continues to be that the study is interesting and the experiments are useful, but important controls are missing, a more cautious interpretation is warranted, and the organization of the study is confusing. I explain my reasoning below.

(1) A major criticism of the original manuscript was that the electrophysiology experiments were only tangentially related to the question under study. The manuscript has now been reorganized in such a way that the cocaine experiments are similarly irrelevant. In the authors' own words, the central question of the study is "does activity in one [DA or 5-HT] system directly recruit transmitter release from the other, or do DA and 5-HT operate mainly through parallel circuits?". Optogenetically stimulating release of dopamine while recording serotonin activity (or vice versa) begins to answer this question, and in the original draft, cocaine-evoked dopamine and serotonin recordings were presented side by side with the opto-evoked dopamine/serotonin data. In this configuration, the drug-evoked recording data allowed readers to compare the magnitude of dopamine/serotonin release evoked by an artificial optogenetic manipulation to that evoked by a behaviorally relevant pharmacological treatment. This serves a purpose. In the revised manuscript, the cocaine experiments are presented in a standalone first figure. As the authors themselves note in their response to reviewer 2, cocaine independently evokes dopamine and serotonin accumulation is therefore poorly suited on its own to test how activating one neuromodulatory system affects the other. Thus, the cocaine data presented alone in this way neither address the paper's central question nor support the paper's conclusions.

(2 The original manuscript suggested that one way by which striatal serotonin activity could recruit striatal dopamine release would be if the same serotonin neurons project to both VTA and striatum. To test this, the authors use retrobeads to identify DS-, NAc-, and VTA- projecting dorsal raphe neurons. However, they neither identify whether the labeled neurons are serotonergic nor evaluate colocalization between the retrograde tracers as would be required to ascertain whether the same serotonin send axon collaterals to more than one region. Thus, the anatomical data presented do not answer the question of whether the same serotonin neurons innervate both targets.

(3) As noted in the original round of review, applying light stimulation to mice expressing a fluorophore-only construct without opsin is an important control in optostimulation experiments that is widely standard in the field. The omission of such a control is a weakness of the paper. The authors suggest that other results presented - such as the lack of green serotonin sensor activation when red light is applied in the midbrain in Fig 2G-H - argues against cross-talk between the red and green optical channels. This is technically correct, but ruling out cross-talk between optical channels is not the only reason such a control is necessary. The high intensity red light used for optostim travels several millimeters through tissue and can reach the retina. Although mice do not have cones to detect red light (and therefore cannot distinguish the color red from other colors), their rods are sensitive to red light. As a result, bright red intracranial optostim can be perceived by the animal as a flashing light, and this sensory stimulus can produce a neural response. The absence of such a response in the serotonin system in Fig 2G-H does not rule out the possibility that the small response in the dopamine system observed in Fig 3G-H is driven by sensory input rather than a bona fide physiological effect of serotonin neuron activation on striatal dopamine release. Crucially, sensory responses to the stimulating light would also show 2Hz entrainment to the frequency of the stimulating laser trains. In other words, without a fluorophore-only control, the entirety of the authors' results pertaining to serotonin-evoked dopamine release can be explained by artifactual effects of the optostimulation. If this opsin-negative control is absolutely impossible for the authors to perform, at the very least this major limitation should be clearly and explicitly addressed in the discussion because it pertains directly to the study's main conclusions. This brings me to my final point.

(4) Taken together, the authors' data appear to support two primary conclusions. First, optogenetic dopamine stimulation has no effect on striatal serotonin release. Second, synchronous and supra-physiological activation of the serotonin system has a very modest effect on striatal dopamine release that might be entirely explained by artifacts of the optostimulation protocol. The miniscule magnitude of the evoked dopamine response observed following maximally strong (15-20 mW) optogenetic serotonin stimulation also bears consideration here. It argues that even if the observed dopamine release effect is not an artifact of the optostimulation protocol, serotonin-evoked dopamine release is likely to be vanishingly small or absent in other conditions when serotonin neuron firing is neither supraphysiological nor synchronous. The new title strongly implies the existence of significant cross-talk, which is not well supported by the data. Claims throughout the text that call the cross-talk effect reliable or otherwise robust are similarly overstated.

Reviewer #3 (Public review):

The authors suggest that the small release of DA may be due to a release of glutamate from DRN 5-HT neurons to the VTA that stimulates weakly and in a transient fashion the VTA DA neurons, which in the end, produce a transient and small release of DA in the NAc.

Their findings give more information on the previously reported complex and partial known crosstalk between 5-HT and DA in the NAc.

Comments on revised version.

The authors have added several new experiments that address some of the prior concerns. However, there are still points of confusion. For example, the authors repeatedly argue that 5HT or DA are synaptic when in fact most of the transduction mechanisms for these neuromodulators and their transporters are extra synaptic, as indicative of volume transmission. This is not well articulated in the authors' interpretative scheme.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors investigated the relationship between the Tat system and MPIase, a glycolipid that facilitates protein integration into the bacterial cell membrane. The TAT (twin-arginine translocation) system is a unique membrane transport machinery that exports fully folded proteins containing a twin-arginine signal peptide. Using both in vivo and in vitro approaches, the authors demonstrated that a sufficient amount of MPIase is required for Tat-dependent protein translocation. Furthermore, the authors successfully reconstituted the Tat transport system by combining recombinant TatA, TatB, TatC, MPIase, and FoF1-ATP synthase.

Strengths:

The reconstituted system clearly demonstrated the requirement for each component, as substrate translocation occurred only when all components were present. Based on these findings, the authors proposed a mechanistic role for MPIase in facilitating Tat-mediated membrane translocation.

Previous studies have shown that MPIase is involved in Sec-dependent protein translocation and membrane protein integration, as well as YidC-dependent membrane insertion. The present study further demonstrated that MPIase also plays an essential role in the Tat translocation pathway.

Overall, this work highlights the central importance of MPIase in bacterial membrane protein biogenesis and provides new insights into the molecular mechanism of Tat-dependent protein transport.

Reviewer #1 (Public review):

Summary:

This study uses high-throughput bacterial cell surface display to identify LC3B interacting peptides in the human proteome. The screen is unbiased and this type of assay has not previously been used for selecting LC3B interacting peptides. The screen was done with library of 500,000 peptides and they ended up with 427 peptides that they scored as high-confidence LC3B binders. The experiments performed are solid and data are analyzed using well documented methods and statistics.

The aim for the authors was to isolate LC3B interacting peptides from the human proteome, and the screen succeeded in doing so. The selected set of peptides included several previously reported LIR motifs, but also many novel LC3B binding peptides that either contained or did not contain the canonical core LIR motif [WFY]xx[LVI].

Another aim was to identify binding determinants important for the LC3B interaction, and they made an interesting sequence logo based on selected LIR containing peptides. This study does not really extend our knowledge related to binding determinants essential for LIR motifs in LC3B binding, but it expands the number of exceptions to the canonical core LIR motif.

Strengths:

The approach used here (high throughput bacterial-surface-display) is new. The screen is unbiased and the fact that peptides are directly tested for LC3B binding may facilitate the discovery of non-canonical LIR motifs. The screen appears to be highly selective and manage to distinguish between peptides that interact with LC3B and peptides that do not interact.

Weaknesses:

It is a limitation that no proteins are analyzed in this study. Further work is therefore needed to verify that identified LIR motifs are functional in full-length proteins and in cells.

Comments on revised version.

It remains a weakness that none of the identified LIR motifs are tested in full-length proteins or cells, but all the specific comments in my initial review have been addressed.

Reviewer #2 (Public review):

Summary:

To discover peptides that interact with autophagy-related protein LC3B and profile the key binding determinants, the authors screened a library of ~500,000 36-residue peptides derived from the human proteome using bacterial cell-surface display. Analysis of the screening data revealed exceptions to the reported LIR motif and a strong preference for negatively charged residues adjacent to the LIR.<br /> These results support a refinement the LIR motif definition and expand the network of candidate LC3B interaction partners.

Strengths:

High-throughput approach

Weaknesses:

Lack of in vitro data and molecular dynamics simulations

Comments on revised version.

In the revised manuscript the authors did not provide in vitro and molecular dynamics simulations data.

Reviewer #3 (Public review):

Summary:

The LC3 family of proteins, which includes LC3B, are ubiquitin-like proteins that are covalently linked to phosphatidylethanolamine in the expanding autophagosomal membrane during autophagy. LC3 family members bind to short sequences of amino acids that reside within dynamic regions in a wide variety of proteins. These sequences, termed LC3 Interacting Regions (LIRs), were initially thought to function primarily to link LIR containing autophagy cargo receptors to LC3 family members to help facilitate their capture during autophagy. However, the functional importance of LIRs in autophagy has broadened to include more general functions in autophagy as well. While a general consensus for LIR sequences has been described as [FWY]0-X1-X2-[LVI]3 recent work has suggested that additional sequences outside of the canonical LIR sequence can bind LC3 family members and play important roles in autophagy. In this manuscript by Kosmatka et al, the authors perform a high throughput screen using bacterial surface display coupled with fluorescence associated cell sorting to identify which human sequences can bind to LC3B. They identify a variety of peptides capable of binding LC3B including peptides from proteins that have not previously been described as LC3B binding proteins. The results from the bacterial surface display were then used to guide sequence analysis, mutational analysis and structural studies to further characterize the range of LIR sequences that are capable of binding LC3B. Taken together, this work adds to the growing knowledge of how LIR sequences interact with LC3 family members and demonstrates which amino acids both inside and outside of the LIR sequence aid in binding. This work also identifies new potential LC3 binding proteins which may play unknown roles in autophagy regulation. Lastly, this work reinforces the importance of alternative LIR sequences such as the [WFY]0-X1-X2-[WFY]3 sequence which the authors have dubbed the LIR+ sequence.

Strengths:

The manuscript uses a robust approach to identify and characterize different peptide sequences that can interact with LC3B. They validate a large number of sequences using biolayer interferometry (BLI). The large number of LC3B binding sequences and their dissociation constants adds significant new information to the field that will help others understand what sequences can bind to LC3B. The authors are also very careful to accurately report on their data and not overly interpret their findings.

Weaknesses:

After the authors identify proteins from their bacterial display assay, the remainder of the manuscript is focused on characterizing the different types of sequences that are identified in addition to validating the LC3B-LIR interactions using BLI and X-ray crystallography. However, it's not entirely clear if the screen identified novel LC3B binders that interact with LC3B in cells. While I acknowledge that the focus of the manuscript is on the characterization of LIR sequences that can bind LC3B it seems like a missed opportunity to not validate a few of the novel LC3B binders in vivo.

Reviewer #1 (Public review):

Summary:

The authors present the most comprehensive phylogenomic study of free-living kinetoplastids to date, including newly generated transcriptomic data from 11 species. This dataset provides a solid foundation for future comparative evolutionary studies of kinetoplastids, a group that includes important human pathogens such as Trypanosoma brucei.

Strengths:

The study is clearly written, and the analyses are thoroughly described, well documented, and reproducible. The phylogenomic dataset comprises 444 proteins, which is larger than those used in many comparable phylogenomic studies. The main strength of the study is the substantially improved taxon sampling of neobodonids, which provides a much stronger framework for resolving evolutionary relationships among free-living kinetoplastids.

Weaknesses:

The phylogenomic analysis is based on a newly assembled dataset comprising 444 single-copy orthogroups. Using an established phylogenomic dataset, expanded to include the newly studied species, to confirm the inferred relationships would provide additional support for the robustness of the recovered topology.

Reviewer #2 (Public review):

Summary:

This is a highly competent and very throough phylogenomic study of the position of neobodonids within the Kinetoplastid phylogeny. The addition of new transcriptomic data for diverse neobodonids allows their relationships to be better understood, leading to convincing changes to neobodonid taxonomy.

Strengths:

The generation of transcriptomic data produces more phylogenetic characters than ever before from several, diverse neobodonid lineages. This represents a substantial increase in the data available for this relatively unstudied group. The methodology used to define the character set and then estimate the phylogeny is conventional, very thoroughly described and expertly executed. The phylogeny is highly robust, and we can be very confident of the result.

Weaknesses:

These changes to neobodonid taxonomy, or essentially the discovery that some 'neobodonids' actually represent distinct lineages at the base of the Metakinetoplastida, are not surprising. So undersampled are the Kinetoplastida, both in terms of molecular markers and phyletic diversity, that an investment in sampling on this scale is always likely to bring about changes. The significance of these changes is not immediately apparent. Since they do not concern the Eubodonidae, they have no implications for the trypanosomatids and the new Metakinetoplastid lineages are very distant from the origins of parasitism, such that comments on "preadaptations for disease" are not compelling.

Reviewer #3 (Public review):

Summary:

The manuscript by Zavadska et al. presents a phylogenomic study focused on the relationships among metakinetoplastids, specifically addressing the question whether Neobodonida are monophyletic. The authors collected eleven kinetoplastid strains affiliated with five out of six neobodonid clades identified in previous studies and sequenced their transcriptomes. These data along with SRAs available from public databases were processed via assembly, annotation and decontamination. The final dataset, representing all (non-trypanosomatid) kinetoplastids with genome and/or transcriptome data available as well as selected trypanosomatids and discoban outgroups, was used for protein orthology inference and identification of single-copy orthologs present in the majority of taxa, followed by phylogenetic analysis in maximum likelihood and Bayesian frameworks.

Strengths:

I find the new data and resulting phylogenetic reconstruction valuable and interesting, particularly given the paucity of genomic/transcriptomic data from free-living kinetoplastids. The study is well-designed methodologically, and the main finding (paraphyly of Neobodonida) appears indisputable, considering the provided phylogenimic evidence. Following best practices, the authors shared their primary data, scripts, and files resulting from the conducted analyses.

Weaknesses:

Several aspects of the presentation and interpretation would benefit from revision. In particular, the composition of the neobodonid lineages central to the study is insufficiently explained, and an 18S rRNA gene-based tree showing these lineages and the placement of the newly analysed strains would substantially improve the clarity of the study. The Introduction would also benefit from a more concise and better-structured account of previous phylogenetic evidence and a clearer explanation of the broader biological significance of kinetoplastids, while the Methods could be streamlined by removing redundant information and presenting some details in tables. Finally, the rationale for including the electron microscopy images in the supplement should be clarified, as these are currently not discussed in the text and represent only a single studied strain.

Reviewer #1 (Public review):

Summary:

This study addresses an important question of how recovery from conditioned fear can be characterized beyond the reduction of defensive responses. The authors combine multiple behavioral measures with a theory-guided hidden Markov model to identify recurring behavioral modes and examine how the organization of behavior changes during fear conditioning and extinction. The study provides an interesting and potentially valuable framework for examining recovery from conditioned fear as a broader change in behavioral repertoire rather than simply as a reduction in defensive behavior.

Strengths:

A major strength of the study is the attempt to characterize fear-related behavior as a dynamic organization of multiple behavioral measures rather than relying on a single conventional measure such as freezing. The integration of freezing, ultrasonic vocalizations, and food-reinforced behavior into recurring behavioral modes provides a potentially useful approach for capturing behavioral states that may be missed by analyses of individual responses. The study also presents a theoretically motivated framework for distinguishing defensive, quiescent, exploratory, and goal-directed behavioral modes, which could provide a useful conceptual basis for studying functional recovery following aversive experiences.

Weaknesses:

The main concern is whether the analyses and behavioral definitions provide sufficient support for the interpretation of these behavioral modes and, in particular, for the proposed concept of functional recovery. First, the data-driven model selection indicates substantially better predictive performance for eight states, whereas the authors retain four states primarily to conform to their theoretical framework. A stronger justification for prioritizing the four-state solution would therefore strengthen the interpretation of the results. Second, it is difficult to determine from the current behavioral measures whether the Quiescent state reflects safety and rest or behavioral inhibition associated with residual fear or anxiety. The marked reduction in exploratory and goal-directed behavior could potentially be consistent with either interpretation, making it important to consider alternative explanations for this state.

The definition of functional recovery also warrants further consideration. The authors propose the reinitiation of food-seeking behavior in the previously threatening environment as an important indicator of recovery. From an evolutionary and ethological perspective, however, an animal that has experienced a threat while foraging in a particular location may benefit from shifting its foraging to an alternative location rather than returning to the previously threatening location, even after extinction learning. Thus, the resumption of goal-directed behavior in the previously threatening environment may not necessarily represent a biologically appropriate benchmark for successful fear extinction or functional recovery, as functional recovery may involve substantial changes in motivational processes that cannot be captured by the simple return of a previously suppressed behavior. More broadly, the results may support the value of examining behavioral repertoire during extinction, but they do not yet establish that restoration of the broader repertoire, or specifically the return of goal-directed behavior in the previously threatening environment, is required for functional recovery.

Finally, the use of the no-shock control group could be more informative. The observed reduction in exploratory behavior during conditioning cannot by itself distinguish associative fear learning from habituation to novelty. A direct comparison between shock and no-shock groups would therefore help determine whether the changes in behavioral modes are specifically related to conditioning.

Overall, the study provides a potentially useful framework and an interesting conceptual perspective on recovery from conditioned fear. However, the strength of the evidence supporting the interpretation of the identified behavioral modes and the proposed definition of functional recovery is currently limited. Addressing these interpretational and analytical issues would strengthen the manuscript and help establish the extent to which this framework provides a meaningful advance in the study of fear extinction and recovery.

Reviewer #2 (Public review):

Summary:

The authors aimed to develop and validate a preclinical computational framework for assessing functional recovery after aversive learning. They argue that a reduction in defensive behavior (e.g., freezing) is insufficient to conclude recovery, and that a comprehensive assessment must also track the re-emergence of exploration and goal-directed action. To test this, they adapted a multimodal paradigm in rats, measuring freezing, 22-kHz and 50-kHz ultrasonic vocalizations (USVs), and food-reinforced lever pressing. Using a theory-guided hidden Markov model (HMM) with four states derived from tri-reference-point theory (Fear/Danger, Quiescent, Exploratory, Goal-directed), they decoded behavioral trajectories across fear conditioning, extinction, delay, and contextual relapse experiments.

Strengths:

(1) The study tackles a significant gap in animal models of fear and trauma-related disorders. It correctly identifies that a focus on a single defensive response like freezing may be misleading and that a broader behavioral assessment is necessary to understand the impact of interventions.

(2) The application of a hidden Markov model to combine multiple behavioral channels into a cohesive, theory-driven framework is a strength. This approach moves beyond simple summary statistics to capture the temporal dynamics of behavior and allows for the identification of distinct, recurring states.

(3) The authors demonstrate commendable rigor in their model selection and validation. They employ discrete-aware scoring to address the boundary issues of count data and conduct extensive stability analyses across restarts, variance floors, and studies. The association of the Fear/Danger mode occupancy with fecal boli, an independent autonomic measure, provides a convincing external validation.

(4) The study provides several important findings. For instance, it shows that during the first extinction session, although Fear/Danger occupancy declined, the dominant state was Quiescent, rather than Exploratory or Goal-directed. This directly supports the core premise that the absence of fear does not equal functional recovery.

Weaknesses:

(1) The authors transparently report that their results are not emission-family invariant. The failure of count-native models (e.g., negative binomial) to converge or produce the expected four-state solution is a major limitation. This suggests the theory-guided state structure is tied to the specific Gaussian representation chosen and may not be a robust and universal feature of the behavioral data. This limits the strength of the claim that the data naturally resolve into these four modes.

(2) The exclusive use of male rats from a single laboratory limits the generalizability of the findings to female animals and other experimental settings. The cross-study recurrence analysis is a helpful sensitivity check but does not substitute for true model transfer to an independent, out-of-sample dataset.

(3) Testing K=4 vs. K=8 raises the question of whether the four-state solution is the most appropriate level of abstraction for the questions being asked.

(4) The interpretation of the 50-kHz USV band as an Exploratory state is somewhat broad. 50-kHz calls are emitted in a variety of contexts, including social interaction and reward. While the use of the band is justified, it remains an indirect measure of exploration and should be interpreted with caution. The interpretation and link to the exploratory state should be toned down and discussed.

Reviewer #1 (Public review):

Summary:

The manuscript proposes an architecture and distance-regression-based logic for 3D cell segmentation.

The architecture combines a U-Net backbone with an xLSTM bottleneck layer and two novel components: a Zernike Phase Layer that is used for input normalization and the Scaled Grid Geocaps that link the input and bottleneck, inspired by the Capsule mechanism.

During training, the network predicts foreground vs. background and regresses a distance transform as output. During inference, these predictions are used as input to a watershed procedure to segment cells in 3D.

The method is evaluated by training on four different 3D cell/nucleus segmentation tasks and compared to StarDist3D and CellPoseSAM baselines, which are also trained on the respective datasets using the same train / val / test splits.

The authors report improved segmentation quality compared to the baselines.

Strengths:

The experiments are solid; ablation studies evaluate the main architectural components, and the software appears to be well documented (I did not have time to test it).

Weaknesses:

(1) This is not really a weakness but a general point: the main issue in microscopy segmentation is that a lot of training data is needed if training a network from scratch. This results in a very laborious data annotation process. This is what modern approaches like CellPoseSAM and microSAM address by providing strong pre-trained networks that work well out-of-the-box for many imaging settings and can then be fine-tuned with a few corrected images to achieve state-of-the-art results. Having a better 3D segmentation approach is of course still worthwhile, but does not address the main practical challenge. I would suggest clarifying this in the introduction.

(2) The authors implement a well-known strategy to segmentation in microscopy: regressing the distance transform. They acknowledge some prior work (Scherr et al., Lux and Matula), but this approach is more prevalent than acknowledged, including for 3D cell segmentation. For example, microSAM [1] implements cell instance segmentation by regressing the distance to the cell boundary and centroid, which is applied in 3D by CochleaNet [2]. See also [3] for an overview of different distance- or flow-field-regression-based cell segmentation approaches. I would suggest performing a more thorough literature review and acknowledging more prior work.

(3) The baseline comparisons are not described in any detail, are questionable for CellPoseSAM, and lack current state-of-the-art models:

3.1) The details on training and inference for the baseline methods are missing. This is particularly missing for CellPoseSAM. Here, it would be important to clarify whether the model is initialized randomly or with CellPoseSAM's pretrained weights and which of the two different 3D segmentation modes is used (see next point). It would also make sense to report both CellPoseSAM in zero-shot mode and finetuned for all cases.

3.2) CellPoseSAM supports two different 3D segmentation modes: either a "2.5D" segmentation mode, where the network is applied across all three axes, the resulting flow maps are merged, and then used for segmentation. And an "overlap-based stitching mode", where segmentation is applied only in the XY plane and slices are then merged based on overlap. It is unclear which of these modes is used. Ideally both should be reported; the "2.5D" mode often leads to worse results for anisotropic data, and we almost always prefer the overlap-based stitching in practice.

3.3) Two important state-of-the-art baselines are missing:

i) microSAM [1], which is widely used for cell segmentation based on a pretrained 2D network that is extended to 3D using a similar strategy as CellPoseSAM's overlap-based stitching mode. It would make sense to report both zero-shot and fine-tuned performance.

ii) uSeg3D [4]: this method implements a generalized 3D stitching approach that works independently of the underlying architecture and reports significant improvements.<br /> Improving and extending the baseline experiments would give a much better picture of the practical relevance of this work.

References:

[1] https://www.nature.com/articles/s41592-024-02580-4

[2] https://www.biorxiv.org/content/10.1101/2025.11.16.688700v1.abstract

[3] https://doi.org/10.6084/m9.figshare.32782521

[4] https://www.nature.com/articles/s41592-025-02887-w

Reviewer #2 (Public review):

Summary:

In their manuscript "Penumbria: Advanced 3D cell segmentation for biomedical imaging", the authors describe their 3D nuclei segmentation pipeline and compare it against state-of-the-art methods for nuclei and cell segmentation.

The introduction is clear, with state-of-the-art methods, for instance, segmentation of cells from microscopy images, and motivates the method with science that can be done with cell segmentation. Related work is appropriately mentioned with the caveat that it conflates cell and nuclei segmentation, and some related work is missing.

The Results section describes the loss, the architecture novelties, and evaluation results on nuclei segmentation, not cell segmentation. Here it becomes clear that Penumbria addresses 3D nuclei segmentation, i.e., instance segmentation of medium-sized, mostly convex-ish and well-separated objects in single-channel 3D volumes.

Penumbria combines several choices and innovations:

- loss: boundary distance regression<br /> - architecture<br /> - 3D U-VixLSTM<br /> - Zernike Layer<br /> - SGG<br /> - combine multiple scales in SGG<br /> - sampling strategy to look for bright and dark objects.

The authors show on four nuclei segmentation benchmarks from confocal images that Penumbria outperforms CellPose-SAM (in either zero-shot or fine-tuned mode) and Stardist-3D in most cases, except for some where Stardist-3D performs better. They report results of an IoU threshold sweep to measure instance segmentation performance.

Penumbria is available on GitHub under a non-commercial license. I have not tested it yet, and therefore assume that the code is in full working condition and reproduces the reported results.

Strengths:

(1) The method may outperform state-of-the-art methods for cell segmentation as shown in three of four 3D benchmarks.

(2) The code is available.

Weaknesses:

Major

(1) Similar to the established confusion across the literature, all the authors' evaluation datasets do not label cells, but cellular nuclei. This is important to understand because the methods they compare against, CellPose-SAM and Stardist-3D, demonstrate their performance on a variety of datasets, some of which label cellular membranes instead of nuclei. The cell segmentation task is often considered harder than the nuclei segmentation task, because objects consistently touch each other, the separation may have noise-induced holes, and object shapes may be more diverse than for nuclei.

(2) The addition of U-VixLSTM, SGG, multi-scale input, Zernike layers, and distance transform regression sounds great, but a comparison to architectures or conventional losses used to train any combination is missing. Five innovations/choices combined simultaneously without an attempt to attribute.

(3) There is no clear delineation of 2D and 3D approaches. E.g. CellPose-SAM is not 3D, but sequential 2D execution with optional averaging of doing this along all axes.

(4) For the 3-fold cross-validated evaluation: It sounds like they were all trained for each benchmark independently? Did Stardist-3D and CellPose SAM start from their existing 'general' weights or from scratch (not possible with SAM, I know). Not described at all.

(5) Experiments to evaluate generalization performance are missing (train on all datasets, compare across datasets, ...).

(6) The Zernike layer optimizes only coefficients {3,4,12}; the evaluation could be described as rudimentary. Table 3 is missing cross-validation. Since mouse_skull is (a lot) different and differences for the others are minor, I would be curious how significant these results are. The following text is unclear: "This reflects how we developed these modifications in turn." How so?

(7) Training and inference: input and output block size missing, GPU memory requirements undefined, discuss block size vs. batch size. 50% overlap block-inference with feathering not compared to other overlap ratios, just "works best". Test-time augmentation 8x rotation (what are those rotations? Why 8? How much does it contribute? Compare the cost for 8xs test time augmentation against skipping "background" heuristic.)

Minor

(1) What is the incoming contrast range and bit depth?

(2) Stardist-3D's weakness is not that it requires convexity (it requires star-convexity) but that it picks one hypothesis leading to low-resolution outlines. This has been overcome by Starfinity (https://linkinghub.elsevier.com/retrieve/pii/S0092867421013398; https://github.com/multiFISH/EASI-FISH).

(3) Heatmap regression is not the same as distance regression; the manuscript does not explain what it is and uses terms interchangeably. This should be clarified. The reference for heatmap is 14; for distances, there are others, e.g. doi:10.1007/978-3-030-00934-2_36.

(4) Only one evaluation example compares against CellPose-SAM zero-shot; I would add this to all of them. The mentioned place switch would be interesting to see in a plot instead of text.

(5) "We replicated this failure on multiple machines." How does the machine matter here? Did the authors do cross-validation across different computers, instead of different random seeds? This sounds expensive!

(6) Brightness-based sampling could be achieved by augmentation?

(7) In the text "We use the following bicubic upsampling factors[...]", I think the authors mean tricubic, or cubic (because they upscale only one axis). Bicubic makes no sense, but people say this in the literature; we can do better. Why integer upscaling factors instead of exact ones? Are the authors concerned about that one incomplete border pixel?

(8) "Furthermore, it is not confined to Linux-based environments such as Segmamba [32], an important consideration given that biologists often work with Windows machines."<br /> This is confusing. Is this a comment about somebody's publicly available repository? This is not relevant to evaluate the method and should be discussed separately. Does it work in WSL?

(9) Downsampling by factors >2 with bicubic interpolation sounds wrong; should the authors use averaging?

Reviewer #1 (Public review):

Summary:

The manuscript authors have used Martini 3-based CG simulations to study how S22 phosphorylation and two catalytic-pocket inhibitors, Nexturastat A and HPOB, regulate HDAC6 condensation.

In isolated IDR1 slabs, phosphorylation lowers the apparent saturation concentration without compacting single chains, and the manuscript attributes this to new contacts between the phospho-S22/D26 region and Arg/Lys-rich patches. The two ligands perturb these contacts differently. In full-length multichain systems, the C-terminal FD3 domain is the most frequent contact partner, and the ligands shift contacts from FD3-FD3 toward pocket-FD3. The manuscript proposes a two-layer mechanism by which pocket-binding ligands regulate multidomain condensates.

Strengths:

The question is well motivated by published experiments, the simulated IDR1 construct matches the experimental one, the phosphorylation contact pattern recurs across independent salt conditions, and the Discussion offers testable predictions. My main reservations are whether the full-length results are robust to the model and to sampling, and whether the evidence reaches the mechanistic and pharmacological conclusions drawn from it.

Weaknesses:

(1) The full-length conclusions rest on FD3-FD3 being the dominant domain contact. The water-interaction correction applies only to the disordered regions, so the folded domains retain standard Martini 3 behavior, which is known to overestimate protein association (e.g., Thomasen et al., J. Chem. Theory Comput. 2022). The contact maps are also normalized per bead, which favors a small domain such as FD3. Please test whether FD3-FD3 dominance survives a folded-domain correction, and show raw contact counts next to the normalized maps. Without this test, and without experimental evidence that this domain self-associates, frame the FD3 result as a model prediction.

(2) The ligand-treated IDR1 slabs are single trajectories with no statistical uncertainty on the saturation concentrations or on the contact-difference maps. The full-length systems have three replicates; contact numbers are still rising at the end of several trajectories (Figure S9), and most differences between Nexturastat A and HPOB overlap within error. The wild-type controls also reverse the ligand ordering (Figs. S12, S14), which suggests that trajectory-to-trajectory variation may drive the domain-level differences. Please add independent replicas for the ligand comparisons, show convergence over consecutive time blocks, and report uncertainties on the difference maps. Claims should then be limited to the differences that are resolved.

(3) The Discussion presents the simulations as explaining why Nexturastat A suppresses HDAC6 condensation more strongly than HPOB. However, every full-length system assembles at a high protein concentration, and total contacts and morphologies are similar across conditions, so the simulations do not show suppression. The IDR-level and domain-level mechanisms are studied in separate systems, and ligand distribution and IDR1 contacts are not analyzed in the full-length assemblies, so the coupling of the two layers named in the title is not tested. Please add an analysis that bears on condensation propensity and on this coupling, or recast "explains" as "is consistent with" and revise the title and abstract accordingly.

(4) For phosphorylation, the IDR1 fragment is strongly basic, so a simple reduction of net charge could lower the saturation concentration without a position-specific contact node. The phosphoserine charge state used is also lower than expected at neutral pH. Moving the same charge to a distal serine, and testing the fully charged phosphoserine, would distinguish these readings. For the ligands, the coarse-grained models show only weak pocket engagement, lack a non-pocket control, have unvalidated nonbonded parameters, and are simulated at millimolar concentrations. The ligand-IDR1 mechanism could therefore reflect nonspecific association. Please add a non-pocket surface control, validate the ligand parameters, and relate the simulated concentrations to those of the motivating experiments.

Reviewer #2 (Public review):

Summary:

The authors investigated the phosphorylation- and ligand-dependent regulation of HDAC6 condensates using explicit-solvent coarse-grained simulations. They found these regulations can be attributed to the reorganization of the interaction networks between multiple domains within the condensate.

Strengths:

Large-scale condensate simulations with multiple domains using explicit-solvent models are nontrivial, impressive, and provide a useful example of what can currently be studied using such simulation models.

Comparison between simulations with/without ligand and with/without phosphorylation at different salt concentrations provides a comprehensive framework for understanding conditional residue- and domain-level condensate organization.

The simulation results are mostly consistent with previous experimental observations and offer potential mechanistic understandings that extend those findings.

The proposed FD3-centered interaction network for phosphorylation- and ligand-dependent regulations is a novel hypothesis.

The study generates a few experimentally testable predictions regarding the IDR- and FD3-mediated interactions.

Weaknesses:

When going through the manuscript, especially the result section, I sometimes found it challenging to distinguish what has already been reported experimentally (mostly from Lu et al., Nature Cancer) from what is newly learned from the simulations. I do appreciate that the authors mentioned which results reproduce experimental trends; however, this is often after all simulations. It may help if each subsection first briefly states the known experimental observation and then explains the specific question addressed by or new mechanisms learned by the simulation. For instance, experiments have shown that phosphorylation enhanced and ligand suppressed phase separation.

The authors directly used Automartini to generate the ligand parameters and then checked the contact fraction between the ligand and the pocket. Such analysis only tells whether the ligand visits the pocket but does not test either specificity or binding affinity, and therefore does not support the claim of ligand pocket recognition. The simulation concentration is also in the millimolar regime, probably much higher than the experimental range, so it's not surprising to see ~10% to 20% contact fraction. Considering that no native interactions are used to restrain the ligand-pocket interactions, I suspect the ligand might also have similar contacts to the other regions of HDAC. The authors might want to be careful about their discussions.

The claim regarding the IDR dimension might be problematic. It's still an open question whether a coarse-grained model, or even current all-atom force fields, can reliably resolve the effect of a single phosphorylation on IDR dimensions. I agree that all current results, including the error estimate, support that Rg do not change that much, but the authors might still want to be careful about such claim considering such small perturbation in the sequence.

The proposed electrostatic rewiring is based on the difference between the contact maps of WT and phosphorylation variant. It looks like the difference is relatively small to the contact probabilities. There is also no error estimation or pattern validation from multiple simulations. The authors might want to quantify the uncertainties of such patch-level relative changes by either checking some independent simulations or time-block variations.

A similar comment applies to the claims on the charged-contact reorganization mechanism of two small inhibitors on LLLPS. Error estimates should be provided for these contacts when comparing different variant. In cases when errors are large (e.g. contact number with errors reported in Fig. 4K to N), statistical analysis (p-values) is needed to show the significance of the results.

Regarding the proposed "phospho-S22-centered" contact network, I had some difficulty recognizing the specific regions discussed in Figures 2 and 3. It might be helpful to mark these regions directly on the contact maps. The authors may also wish to consider whether "S22-centered" is the most accurate description, or whether the observed changes are better described as involving a broader segment spanning several residues around S22. And since phosphorylation is introduced at S22, observing changes near S22 is not unexpected and may not be sufficient to establish phospho-S22-centered network.

Within the last result section, the authors argued that FD3 plays an important role in both phosphorylation- and ligand-regulated HDAC assembly. However, only contact-based analysis instead of saturation concentration was presented. I understand the substantial challenge of obtaining converged saturation concentration for such system with multiple domains, if so, the authors need to discuss such caveats.

Though direct experimental evidence of such FD3-centered interaction network does not exist, the experimental paper did provide measurements for a couple domain-deletion constructs, including the deletion of the FD3 (BUZ) region. This might be worth discussing, considering the manuscript focuses on the role of FD3 region. In particular, deletion of the BUZ region did not appear to affect LLPS. This does not necessarily exclude a role for FD3 in phosphorylation- or ligand-dependent regulation. However, it may complicate the description of FD3 as the center of the condensate network. The authors might want to discuss how their model can be reconciled with these domain-deletion results and consider their corresponding conclusion.

Finally, the authors describe their mechanism as coupled networks. They might want to clarify the coupling they meant. If this refers to the coupling between different regions (e.g. S22 and FD3) or domains, they might want to consider more quantitive analysis, for instance by conditional/correlated contacts between domains. Changes in contact patterns at different conditions do not necessarily show the coupling between domains.

Reviewer #3 (Public review):

Summary:

In this work, the authors probe the combined role of phosphorylation of an intrinsically disordered region (IDR) and domain pocket interactions with small molecules on the assembly of a multidomain protein HDAC6 using molecular simulations. Their find that IDR phosphorylation and ligand interactions with domain pockets can together reshaped in the interaction networks; thereby, expanding mechanistic understanding of multidomain protein assembly in the context of phase separation beyond an IDR centric grammar. The work is supported by extensive simulations featuring several controls and analysis of the underlying interaction networks.

Strengths:

(1) They find an important role of phosphorylation and ligand binding in shaping multidomain protein assembly.

(2) They demonstrate that even if IDR:IDR interactions are important in isolated IDR systems, folded domain-folded domain interactions can dominate in the full-length protein. Furthermore, ligands can significantly perturb the interaction landscape, e.g., by redirecting folded domain-folded domain interactions to domain pocket ones.

(3) They also show that different ligands can yield differing impact on the interaction networks by both competing with and compensating interactions to varying degrees. This is demonstrated for two ligands (Nexturastat A and HPOB).

(4) They probe both the IDRs in isolation (including both the phosphorylated and unphosphorylated versions) and condensates formed via these IDRs. They also characterize assemblies of the full length proteins (with and without phosphorylation) in presence and absence of the ligands. The combined simulations, in general, support the key conclusions of the paper.

Weaknesses:

(1) There are areas where statistical tests can improve the quantitative presentation of the work. Specifically, the difference in the IDR ensembles for the isolated chains should be assessed for statistical significance.

(2) There appears to be considerable finite-size effects in some of the slab simulations that may affect the conclusions. In particular, the chains along the long axis appear to be interacting across the periodic boundary. This can happen if the slab long axis is not long enough/if the ratio of dilute phase to dense phase is not large enough.

(3) Some decisions in the simulation conditions require further justification/explanation. More specifically, the concentrations of small molecules used, as well as the decision to conduct simulations at 295 K, should be adequately justified.

(4) Given the size of the full-length protein, it is understandable that the full dense phase is not probed for the full-length protein. However, it would be useful to comment on how the conclusions may change in the context of the full length. For instance, there is no real interface that the small molecules need to diffuse across and so discussions on the role of partitioning in the paper should be better contextualized in light of this.

Reviewer #1 (Public review):

This manuscript describes a multi-modal study of associative learning and memory in humans that combines scalp EEG, pupillometry and behavioral analysis to explore the construct of mnemonic prediction errors (MPEs), in terms of their relationship to attention and cognitive control. Across two pooled studies, participants performed associative memory tasks in which they learned the relationship between a cue word (action verb) and subsequent picture (animate or inanimate) with a strong vs. weak (4 or 1 repetitions) encoding manipulation. At test, participants were encouraged to generate a prediction following the cue word to determine whether the subsequently presented picture was a match or mismatch. The timecourse of pupillary responses during match decisions were decomposed using temporal principal components analysis, which identified 6 distinct and overlapping processes. Some of the components (PC3/PC4) exhibited sensitivity to both the strength and mismatch conditions, as well as behavior (both RT and accuracy) and retrieval success on the subsequent trial. Furthermore, relationships were also observed between pupillary responses (specifically for PC4) and both frontal theta and posterior alpha power measures obtained from scalp EEG in Experiment 2, as well as for frontal theta and subsequent learning from mismatch stimuli (assessed using subsequent memory findings from a surprise recognition test). The authors suggest the findings indicate that MPEs elicit changes in attention, arousal and cognitive control which impact subsequent learning.

Strengths:

This manuscript has many strengths, including a clever study design, thoughtful integration of multiple neurocognitive measures, and a set of rigorous and technically sophisticated analyses, which reveal a large set of relationships among the measures and behavior. The findings demonstrating brain/physiology-behavior relationships are particularly important, in that they point to potential functional consequences of MPEs.\

Reviewer #2 (Public review):

Summary:

The authors studied cognitive control and attention in response to mnemonic prediction errors (MPEs): situations in which the external reality violates internal memory-based predictions. The behavioral task first established strong versus weak predictions, and then either confirmed or violated these predictions. The authors examined markers of cognitive control (frontal theta) and attention (posterior alpha suppression, pupil response) while strong and weak predictions were confirmed or violated. They found increased cognitive control (frontal theta) for strong MPEs, which correlated with subsequent memory. Markers of attention (alpha suppression, pupil response) also accompanied strong MPEs but did not correlate with subsequent memory. Pupil response was investigated using an interesting approach that decomposes the response into different components, finding that different components respond earlier or later and show different correlations with MPEs and their strength. The authors also investigated how EEG, reaction time, and pupil responses correlated with one another, providing further insight into the mechanism underlying the response to MPEs. Together, the study points toward multiple control and attention mechanisms involved in MPE response and memory.

Strengths:

The study has a clear behavioral paradigm with multiple measures - behavioral, EEG, and pupillometry that offer an investigation into different aspects of MPE response and memory.

The study is also very comprehensive in looking at multiple phases in processing MPEs: the prediction phase (prior to the violation), the response to MPEs, and subsequent memory of MPEs, all within one study. Specifically, the link between neural mechanisms and subsequent memory is a major advancement, as most prior studies did not include this component. Mechanisms underlying subsequent memory of MPEs are theoretically important, as a primary function of MPEs is to promote learning and memory. As the authors mention, the different neural and pupillary signals are not robustly correlated, suggesting multiple mechanisms underlying MPE detections, which is interesting, offers avenues for future research, and can facilitate a better theory of how MPEs are processed in the brain. Finally, the decomposition of pupil response into different components and their correlation with behavior (RT during match/MPE detection) is interesting.

Reviewer #1 (Public review):

Summary:

The authors aimed to re-analyse two fMRI datasets of studies investigating the effect of cognitive emotional regulation on amygdala response during the viewing of negative images, in order to assess whether such effects might have been missed in previous reports of the results of these studies. In order to accommodate potential non-typical response profiles in the amygdala, they used a modelling approach that is agnostic about the shape of the BOLD response (finite impulse response modelling). Furthermore, the authors sought to differentiate between amygdala subregions using i) two previously published probabilistic atlases of amygdala subnuclei; and ii) a data-driven analysis focusing on temporal response profiles, which identified four clusters that mostly overlapped with anatomical parcellations. The analyses confirmed their previous results insofar as they did not find any impact of emotional regulation on the amygdala response in any subregion. However, their approach revealed differences in temporal response profiles between amygdala subregions: the centromedial amygdala showed sustained activation extending into the post-stimulus rating period, whereas laterobasal and superficial subregions peaked earlier during stimulus presentation. Overall, this study helps to clarify whether cognitive emotional regulation during negative image viewing alters amygdala response, and demonstrates the value of flexible response modelling to differentiate between functional subregions of an important brain structure.

Strengths:

The strengths of this study lie in the amount of data used (a combined N of 358 participants is rather high), the flexibility of the modelling approach (FIR), the exploratory nature of the data-driven clustering based on temporal response profiles, and clarity of the text and of the findings. One weakness would be the fact that no fMRI data acquired at higher field strengths were used for the aim of differentiating responses of amygdala subunits.

In my opinion, the authors achieved their aims of re-analysing these datasets with complementary methods, and helped to discover new aspects of the BOLD response in the amygdala during emotional regulation. The results support their conclusions.

This work is likely to stimulate discussion of the response of amygdala subnuclei during emotional regulation, and might encourage researchers to resort to more flexible analysis methods by demonstrating the information gain achieved with the FIR approach.

Weaknesses:

There are 7T studies of the amygdala response that have looked at differences in BOLD response across subnuclei. While I don't know of any that have investigated emotional regulation, I think these studies should nevertheless be mentioned and discussed in the Discussion, at least to compare the differences observed across subnuclei.

Reviewer #2 (Public review):

Summary:

In this paper, Bo et al. study the spatiotemporal organization of BOLD responses in the amygdala. They apply FIR modelling to a large sample (two cohorts tested at the same site and scanner, totalling >350 individuals) who completed a negative/neutral picture viewing and cognitive reappraisal task. Averaging voxelwise BOLD responses across two distinct probabilistic atlases of the amygdala, the authors found that a corresponding centromedial region showed sustained activation to negative stimuli beyond the stimulus period. Laterobasal and superficial subregions peaked earlier, within the stimulus presentation period. In all cases, reappraisal did not change these profiles. Lateralization effects (left-side dominance) were found in the CM region at later time points, and to a lesser extent in the superficial nucleus area at earlier time points. Data-driven clustering of voxelwise time courses could partly recover findings from probabilistic atlases, but nonetheless showed partial decoupling of anatomical and functional boundaries in the amygdala.

Strengths:

This is a well-written paper with a clear objective addressed by sound methods. The authors set out to address two main questions: (1) "do amygdala subregions exhibit distinct temporal dynamics in response to negative emotional stimuli?" and (2) "does cognitive reappraisal differentially modulate these temporal profiles across subregions?" Regarding question 1, FIR modelling results indicate strong differentiation of the centromedial region of the amygdala from other subdivisions, with strong and late responses to presented negative stimuli. On question 2, cognitive reappraisal does not seem to play a major role here, although I raise a question on this in the next section. Overall, all presented conclusions are supported by study findings. I would add that lateralization effects are not mentioned in the main objectives/questions of the paper, but this analysis represents a large section of the results that may warrant mention in the introduction as well.

Weaknesses:

(1) An important concern when looking at mesiotemporal regions such as the amygdala is signal-to-noise. Yet, this challenge in analyzing data from the amygdala was not addressed here but represents a potential confound if SNR varies within the amygdala along its major subdivisions.

(2) The authors analyze BOLD evoked response time courses, but I did not find a GLM analysis showing that the task reliably engaged the amygdala, either when analyzed with a canonical HRF or with the FIR model used here. We can see from Figure 2 that the task may not engage the entire amygdala (e.g. LB/LA subregions seem to show lower amplitude evoked-responses), yet the authors focus equally on time courses of subregions engaged and not engaged in the task. As such, differences observed between subregions may reflect different underlying processes (e.g. different processes of engagement in the task, or high/low engagement).

(3) The study uses two different probabilistic atlases to parse BOLD response time courses in the amygdala. However, methodological details on the processing of the atlas data are missing from the text. For example, it is unclear if the subregions were thresholded or overlap in any way, which would affect the strength of detected effects if there is substantial overlap between areas.

(4) The authors found no effect of reappraisal on BOLD response time courses. However, if I understand correctly, not all participants showed successful reappraisal according to their behavioural ratings (lines 125-126: "Eighty-one percent of participants showed a regulation effect in the same direction as the group-level effect"). Were these participants excluded from the reappraisal analysis on evoked response time courses? This is currently unclear, but holds significance in reliably answering question 2.

Reviewer #1 (Public review):

Summary:

Wang et al. report that changes in the levels of canonical Wnt signaling in CNS ECs modulate protein expression levels of VEGFR2 and VEGFR1. An increase in Wnt/beta-catenin signaling results in a decrease in VEGFR2 and an increase in the decoy receptor VEGFR1; conversely, reduction of Wnt signaling leads to the opposite changes in VEGF receptors in brain CVOs (less clear in the retina). Next to the solid quantitative immunostaining of VEGFR2 and VEGFR1 in multiple CNS regions using genetic mosaics and loss-of-function mouse mutants, the authors employ HEK293T cells to investigate the effect of activation of Wnt/beta-catenin signaling on multiple RTK signaling systems and on the level of overexpressed VEGFR2. The principal conclusions are that Wnt/beta-catenin regulates VEGF signaling. Two distinct levels of regulation are reported: one acts in vivo on receptor protein expression levels, the other acts in 293T cells, and the precise mechanism remains to be identified.

Strengths:

Strengths of the study include the use of genetic mosaics that allow for quantification of immuno-staining intensity with an internal reference and the significance of understanding the interplay of beta-catenin-dependent and VEGF pathways.

There are also some weaknesses noted:

(1) Experiments to restore the VEGFR2/VEGFR1 ratio in vivo, or to modulate VEGF signaling in vivo, and assess to what extent Wnt/beta-catenin LOF phenotypes can be restored by modulation of VEGF signaling have not been conducted at this point. In addition, it remains to be determined to what degree an experimental recreation of the aberrant VEGFR2/VEGFR1 ratio is sufficient to recreate (partially) some of the phenotypes observed with loss of Wnt/beta-catenin signaling in ECs.

(2) It would be useful to consider and discuss other essential components of the VEGF signaling system, especially Neuropilins. Wang et al. (2012, Figure 2, PMID: 23217714) reported that CNS ECs with loss of Norrin signaling gain Sema3F binding while not displaying VEGF-B binding sites, suggesting changes in VEGF/semaphorin co-receptor biology.

(3) Martowicz et al. (2019) PMID: 31533473  also investigated the influence of Wnt/beta-catenin signaling in CNS ECs on VEGFR2 and VEGFR3 expression and reported findings regarding VEGFR2 regulation that appear to differ from the present study. This should be discussed.

(4) Norrin/beta-catenin and Wnt/beta-catenin signaling have a critical role in CNS angiogenesis, and Wnt surrogates counteract neovascularization in the oxygen-induced retinopathy model. These findings highlight context-dependent roles of Wnt/beta-catenin signaling in normal and pathological ECs. How the modulation of VEGFR2/VEGFR1 may influence developmental and pathological angiogenesis could be discussed.

(5) When looking up the publicly available scRNA-seq datasets from Heng et al. (2019, PMID: 30988181) from Norrin KO mice, it appears that there is transcriptional regulation of VEGFR1/VEGFR2 consistent with the protein regulation reported here. If transcriptional regulation is important, CMV-driven VEGFR2 expression in Fig.5 supplement 2 may not model the regulatory mechanisms observed in vivo. Does CHIR99021 modulate VEGFR2/VEGFR1 transcripts and proteins in HRMVECs?

Reviewer #2 (Public review):

The manuscript by Wang, Nathans, and colleagues presents a compelling investigation into the regulatory crosstalk between Wnt/beta-catenin and VEGF signaling within central nervous system (CNS) endothelial cells. Overall, it is a very solid paper from a very respected research lab. The use of in vivo genetic mosaics to directly compare endothelial cells (ECs) with differing beta-catenin states within the same tissue is a particularly elegant approach to mapping receptor dynamics.

Major Comments:

(1) Distinguishing Receptor Abundance from Functional Signaling:

The study relies heavily on immunohistochemistry to assess total VEGFR1 and VEGFR2 levels, inferring that changes in receptor abundance directly equate to altered VEGF signaling. Because immunofluorescence does not establish plasma membrane localization, receptor phosphorylation, or downstream pathway activation, the definitive conclusion that beta-catenin "acts at the receptor level to reduce VEGF signaling" is currently overstated.

Recommendation: The authors should tone down the functional claims throughout the text (e.g., state that the receptor changes are "predicted to reduce VEGF signaling") or provide direct functional readouts of VEGFR2 activation in vivo, such as phospho-VEGFR2, phospho-PLCγ, ERK, or AKT.

(2) Limitations of the In Vitro HEK293T Model:

The HEK293T STF reporter system is utilized to argue for a "post-receptor physiological suppression" of VEGF signaling. While experimentally practical, this non-endothelial system does not definitively establish a physiological mechanism for CNS ECs. Furthermore, NFAT-luciferase is not uniquely specific to VEGFR2 and can be activated by broader calcium/calcineurin pathways. The authors also show that high Wnt1 expression reduces total VEGFR2 protein in these cells, but only at plasmid concentrations much higher than those required to suppress the VEGFR2/NFAT reporter.

Recommendation: Explicitly decouple the in vitro observations from the physiological in vivo claims. Remove the term "physiological" unless you validate post-receptor suppression in a primary endothelial system under endogenous signaling conditions.

(3) Regional Heterogeneity (Brain vs. Retina):

The retinal vascular mosaic data reveal that while VEGFR1 is reduced in BRB- ECs, VEGFR2 shows no obvious difference. The authors correctly acknowledge that Müller glial VEGFR2 expression complicates endothelial quantification, rendering Figure 4C a negative result with uncertain sensitivity. Consequently, the broad generalization that an "increase in Wnt/beta-catenin signaling produces an increase in VEGFR1 levels and a decrease in VEGFR2 levels" across all CNS ECs is not entirely accurate.

Recommendation: Embrace this biological nuance. Revise the generalized claims to reflect that beta-catenin regulates VEGFR1 consistently across the brain and retina, but VEGFR2 regulation appears context-dependent and is not clearly demonstrated in the retina. To resolve the Müller glia interference, consider orthogonal approaches to isolate the EC signal, such as EC-specific RNAscope or flow cytometry on FACS-sorted ECs.

(4) Statistical Rigor and Experimental Units:

The figure legends (e.g., Figure 1) explicitly state that data points represent the mean values from single confocal images. If multiple images were obtained from the same animal, treating them as independent units. Furthermore, the authors utilized unpaired statistics (Wilcoxon rank-sum) to compare BBB+ and BBB- ECs within the same mosaic tissue sections.

Recommendation: Reanalyze the data using the animal as the primary biological replicate if such data are available (averaging all image measurements within each animal). For mosaic comparisons within the same tissue image/animal, utilize paired statistical tests (e.g., paired t-test or paired Wilcoxon signed-rank test). Figure legends should clearly report the exact N (animals) and n (images/cells), the specific statistical tests used, and confidence intervals.

Reviewer #3 (Public review):

Summary:

The manuscript by Wang et al. describes the identification of a novel mechanism by which the canonical Wnt/beta-catenin pathway could regulate endothelial permeability in the central nervous system (CNS). Specifically, they show that in circumventricular organs (CVOs), in which the blood-brain barrier (BBB) is not established in vessels and endothelial cells are highly permeable, the previously reported low levels of Wnt/beta-catenin coincide with high levels of the vascular endothelial growth factor receptor 2 (VEGFR2, Flk-1), whereas the decoy receptor VEGFR1 (flt-1) is expressed at low levels. In order to test the hypothesis in vivo that the Wnt/beta-catenin pathway controls VEGF signaling in CNS vessels by regulating VEGF receptors, the authors analyzed mosaic cerebellar vessels of mice deficient for the Wnt receptor frizzled 4 (Fzd4), showing that BBB+ (normal condition) vessels show low VEGFR2 and high VEGFR1, whereas BBB- (Fzd4ko) vessels show high VEGFR2 and low VEGFR1. Similarly, although not as conclusively due to strong background from Müller cells, these effects were also visible in the retina of these mice. Conversely, dominant genetic activation of beta-catenin signaling in endothelial cells in vivo resulted in upregulation of VEGFR1 and downregulation of VEGFR2 in vessels of the CVOs as well as of the choroid plexus.

In order to better understand the direct relevance of Wnt/beta-catenin signaling for this regulatory theme, the authors activated the pathway in HEK293 cells, stably harboring the Super-Top-FLASH reporter construct in the presence of VEGFR2 and VEGF. Moreover, by transfecting Wnt1 into HEK293 cells, the authors analysed the effect of TGFbeta, EGF, and FGF signaling. Together, these experiments showed that Wnt/beta-catenin signaling potently suppressed VEGF signaling at low concentrations of Wnt1 transfection, whereas VEGFR2 suppression was only visible at relatively high concentrations. Also, the other pathways were suppressed, although to a lower extent.

In conclusion, the authors suggest a mechanism of VEGF signaling suppression by Wnt/beta-catenin 1) at the post-receptor level of VEGF receptors and 2) at the transcriptional level of VEGFR1 and VEGFR2, independent of VEGF signaling suppression.

Strengths:

(1) The present manuscript provides interesting additional information on top of the published work on the mechanistic function of vascular permeability control by Wnt/beta-catenin, specifically by the highlighted interaction with the VEGF pathway and the opposed regulation of the two VEGF receptors VEGFR1 and VEGFR2.

(2) They have used state-of-the-art animal models, high-resolution imaging, and appropriate in vitro analysis.

(3) This finding has potentially strong implications for future therapeutic strategies, aiming at Wnt/beta-catenin activation in various diseases of the CNS in which vascular repair is desired.

Weaknesses:

A detailed analysis of the post-receptor and transcriptional mechanisms remains largely unexplored and needs to be explored in the future.

Reviewer #1 (Public Review):

Summary:

In this paper, Li and colleagues have found mircoRNAs that affect levels of metamorphosis-regulating genes that can also affect levels of sesquiterpenoids (juvenile hormone and related compounds) and ecdysteriods, which regulate the timing and stages of insects, respectively. They first compared the transcriptomes of Drosophila at the third larval instar and at the white pre-pupa stage. They found thousands of differences in gene transcript levels between males and females, and between the two different stages. Among those genes that were differentially regulated they saw that genes involved in insect hormone biosynthesis were disproportionately represented. Many of the differentially regulated genes were involved in the insect hormone biosynthesis pathway and ascorbate and alderete metabolism. MicroRNAs were also differentially expressed during metamorphosis and were separately identified. The authors then considered genes and whether the differentially expressed microRNAs might regulate transcripts known to be involved in sesquiterpenoid production. In silico analysis of microRNAs predicted a list of 17 microRNAs that can regulate transcripts of sesquiterpenoid biosynthesis genes. The authors then used an in vitro luciferase assay to validate the binding and downregulation of 10 of the microRNAs to genes involved with sesquiterpenoid production in S2 cells.

Li and colleagues then focus on two genes they found were bound by microRNAs that have established roles in metamorphosis. The microRNAs miR-34 and miR-277 bind transcripts of two protein-coding genes that regulate metamorphosis Kr-h1, which encodes a transcription factor that is a JH-inducible transcription factor, and Allatostatin C Receptor 1, (AstC-R1), a G-protein coupled receptor that regulates the corpora allatum, the gland that produces sesquiterpenoids. Using a LAMP assay, one of the microRNAs, miR-277 was shown to bind to both AstC-R1 and Kr-h1 in in vivo whole-animal extracts. There is no mention of binding between either protein-coding transcript and the miR-34 microRNA. Temporal expression of all four transcripts shows that their abundance is anti-correlated; stages of high miR-34 or miR-277 expression correlate with low AstC-R1 or Kr-h1 expression. Homozygous deletions of both mircroRNAs result in 23% lethality, five days after adult eclosion. The authors also generated specific mutants in miR-34 or miR-277 and find differences in the expression of AstC-R1 and Kr-h1 and sex-specific differences in both sesquiterpenoids and ecdysteroids in the knock-out lines. If there were phenotypes associated with the specific knock-outs, those were not mentioned. Next, the authors examined the transcriptomes of the miR-3277 and miR-34 mutants and found several other GO-terms enriched among the differentially expressed genes. However, the sesquiterpenoid pathway and ascorbate and alderete metabolism are not listed.

Strengths:

This is an interesting manuscript that could make an important contribution to our understanding of the roles of micro RNAs at metamorphosis, and potentially of how sex-specific differences arise during metamorphosis. Strengths of the paper include the functional validation of microRNA binding, in vitro and in vivo-, as well as the characterization of sesquiterpenoid and ecdysteroid titers. The authors have also used CRISPR to generate specific knock-outs of miR-34 and miR-277. The transcriptomes will be a resource for future work to mine for differences in gene expression during metamorphosis.

Weaknesses:

(1) Spatial Expression of miR-34 and miR-277. If miR-34 and miR-277 regulate AstC-R1 and Kr-h1, then they must be expressed in the same cells. Although the authors show that the microRNAs do bind to the transcripts of AstC-R1 and Kr-h1 in S2 cells, and miR-277 binds AstC-R1 and Kr-h1 in vivo whole-animal homogenates, we do not know if the microRNAs are ever in the cells where AstC-R1 or Kr-h1 are expressed. AstC-R1 is only expressed in a few cells in the brain, so it is not at all certain that it is co-expressed with either microRNA. The creation of enhancer lines or in situ hybridization in Drosophila is straightforward and would sort this out.

(2) Phenotypes. Although a double deletion was used and specific knock-outs of both miR-34 and miR-277 were generated, the analysis of the mutants is very superficial. For the homozygous deletion of both microRNAs miR-34 and miR-277, only a decrease in survivorship was observed a full six days after adult eclosion - after the end of metamorphosis. No phenotype for either miR-34KO or miR-277-KO was given. The authors cite the work of others who have found specific phenotypes after manipulation of sesquiterpenoids or ecdysteroids, like Riddiford and Ashburner, but do not use any of these many studies to help them characterize the phenotype. If the loss of miR-34 and miR-277 affects so many pathways (including MAPK signaling, TGF-beta signaling, FoxO signaling, and Wnt signaling), as well as global titers of metamorphic hormones, then there shouldn't there be something different in the development to discuss?

(3) I think the reliance on GO term enrichment is getting in the way of biology. For instance, I would not describe Kr-h1 as a sesquiterpenoid biosynthesis pathway gene. Yet the authors say they were motivated to examine microRNA regulation of Kr-h1 because they saw differences in levels of the sesquiterpenoid biosynthesis pathway between WL3 and WPP, a period which also saw differences in expression of some microRNAs. I understand that Kr-h1 expression is regulated by JH, a sesquiterpenoid, but it is not directly involved with JH production, so relying on GO term enrichment has made the decision to focus on Kr-h1 feel arbitrary.

(4) The transcriptomes of miR-34 and miR-277 should have revealed genes encoding members of the sesquiterpenoid biosynthesis pathway as well as AstC-R1 and Kr-h1, but neither was mentioned. The functional tests of miR-34 and miR-277 were performed because they were shown to affect the levels of expression of genes in the sesquiterpenoid biosynthesis pathway. Figure 2 shows a significant decrease in AstC-R1 and Kr-h1 transcripts after the loss of miR-34 and miR-277. However, the results do not mention either (Lines 250-264). Instead, there is a list of 10 different GO terms (like arginine and proline metabolism or fatty acid degradation) that were enriched in miR-34 and miR-277 transcriptomes. If any of those ten types have any relationship to Kr-h1, AstC-R1, or metamorphosis, that has not been explained.

(5) Not enough care was taken in describing the stages. The methods describe wandering larvae (WL3) and white pre-pupa (WPP) for the transcriptomes, but in the text, different terms are used, like "larva", "pupa" and "L3 larvae instars" "early pupae" "late L3". Also, it seems like the small RNA libraries for sequencing were taken from "L3 larvae", but the stage of the L3 larvae was not mentioned. Staging is important, especially during metamorphosis, since differences in expression are expected to exist between different stages of L3, between early vs late wandering, and between WPP and early pupal stages.

Reviewer #2 (Public Review):

Summary:

This study proposes that the microRNA cluster miR-277/34 controls the generation of sexual dimorphism in Drosophila melanogaster during metamorphosis by acting on specific hormonal and developmental gene pathways.

Strengths:

Using a combination of mRNA and small RNA sequencing together with genome-wide in silico and in vitro analyses the authors identified a microRNA cluster that may be involved in metamorphosis and the generation of sexual dimorphism in Drosophila melanogaster.

Weaknesses:

Biological validation of the identified sexually dimorphic genes and a detailed understanding of how the microRNA cluster miR-277/34 might be involved in the regulation of sesquiterpenoids are needed.

Major suggestions:

(1) If AstC-R1 and Kr-h1 are targets of the miR-277/34 cluster and cause their downregulation, it is not clear why there would also be a decrease in the levels of these genes in the miR-277/34 mutants. This would suggest that the mechanism is not straightforward and that further epistatic experiments should be carried out in order to clarify this issue.

(2) The changes in the expression levels of AstC-R1 in pupae of miR-277-KO and mir-34-KO flies must be accompanied by photos of the respective larvae and pupae, as well as an analysis of the larvae-pupa transition on the mutants by gender.

(3) Biological validation of the identified sexually dimorphic genes in vivo will be necessary for the support of this work.

Reviewer #3 (Public Review):

Summary:

The authors show convincingly the complexity of gene up- and down-regulation at the outset of metamorphosis and identify substantial differences between the two sexes, even at this early time in development. The complexity of microRNA expression and the difference between the sexes are also nicely laid out. The functional significance of these differences, though, is harder to establish. The authors have focused on the roles of two families of developmental hormones, the ecdysteroids, and the juvenile hormones. The emergence of sex-specific differentiation of organs during metamorphosis is clearly downstream of the action of ecdysteroids and/or JH, but there is no evidence that the presence or lack of these hormones has any effect on the sexual identity of organ systems - i.e., that manipulations of JH or ecdysteroid result in either the masculinization or feminization of individuals or their organs. The precedence for the linkage of these hormones to sex determination is the 2002, Belgacem & Martin study, which describes the effects of JH on fly locomotion. These authors show that the number of stop/start bouts is sexually dimorphic, and removal of JH in males shifts their frequency into the female range while giving treated males exogenous JH moves it back. While this is referred to as a "feminization" of male behavior, this quantitative shift in frequency is not as compelling as would be a qualitative shift -- for example, the removal of JH causing males to show egg-laying behavior (a result that has never been seen). Also, these effects are in a fully mature system, rather than at the early metamorphic time examined in the present paper. In driving and coordination metamorphosis, JH and ecdysteroids are intimately involved in sexual differentiation, but I know of no compelling evidence that they play a role in sex determination.

While the summary of the effects or removal of specific microRNAs on the components of the biosynthetic pathway for the JHs and ecdysteroids (Figure 2E ,F) is quite compelling, I am concerned about the effects of the removal of mir-277 and mir-34 on the levels of both the JHs and 20E. My concern centers around the data from the control group (w[1118] animals in Figure 2D). These data are the first report of a marked sex difference in the titer of either JH or ecdysteroid at the start of metamorphosis in Drosophila. As expected, males show a 10-20 increase in levels of JH III, JHB3, and 20E between the L3 stage and the white puparium, but, surprisingly, the levels of these hormones in female L3 larvae are equal to or greater than that seen at pupariation! These data for females run counter to over 50 years of work on the effects of ecdysteroids in Drosophila!

As far as I can gather from the paper, the L3 data were obtained using wandering larvae. This stage lasts for about 12 hours and ends with pupariation. Larvae from this period need to be used with caution for hormone studies. Levels of both JH and ecdysteroid are low as larvae leave the food but rapidly rise to their peak levels at the white puparium stage 12 hours later. To deal with the rapidly changing hormonal landscape through this period, the researchers have used physiological markers to track this progression. Initially, it was the sage of "puffing" of the giant salivary gland chromosomes, but, for bulk collection of staged larvae, larvae are fed on food containing a blue dye, and progression is tracked by the loss of blue coloring from the gut. I could not find if the authors had any criteria for selecting larvae during the wandering period. Male and female larvae grow to different sizes. Might this difference in growth be biased when larvae were selected during their wandering phase?

The other hormone-related issue is the expression of Kr-h1 during larval stages and metamorphosis (Figure 1G). Kr-h1 is the main target of JH and Kr-h1 expression is often used as a proxy for the JH titer. The authors report that peak Kr-h1 expression occurs in the L3 (when the JH titer should be lowest!) and that it drops at wandering. This pattern is counter to that reported in the literature (e.g., FlyBase, ModEncode).

Reviewer #1 (Public review):

Summary:

In the manuscript entitled "Gene function and expression profiling in yeast spores, killifish diapause embryos, and their post-dormant offspring cells" the authors analyze dormancy at three levels: functional, transcriptomic and proteomic, using Schizosaccharomyces pombe spores, diapausing killifish embryos and human dormant cancer cells as models. The authors performed a genome-wide Bar-seq screen for gene deletions in fission yeast spores and then compared changes in mRNA and protein expression in young and old and heat-stressed cells in yeast and killifish. The main conclusions of the work include that dormancy is a dynamic not static state. The authors conclude that specific biological processes, rather than particular genes, are conserved across organisms during the dormancy, and that there is a clear disconnect between the mRNA and protein kinetics and behavior, particularly for ribosomal proteins and processes related to autophagy and mitochondrial metabolism.

In general, the manuscript is well structured and written, with carefully designed figures. The authors are also generally aware of the limitations of the work. The manuscript is interesting and valuable, but some interpretations should be moderated.

Major comments:

(1) The data presented provide in general good support for majority of conclusions presented in the manuscript. However, less support is provided for the theories that: there is a universal "core expression signature" of dormancy; the mRNA/ribosome protein uncoupling reflects a specific mechanism of preparation for exiting dormancy; and the observed changes provide evidence for an epigenetic "memory" of dormancy. These elements are currently more interpretive than evidentiary. I would recommend softening the language in these places and clearly labeling them as hypothetical.

(2) The authors propose that spores and post-dormancy cells retain a memory of time spent in dormancy or stress exposure. However, sporulation is also generally expected to reset or remove accumulated damage. Please clarify whether the observed long-term effects reflect classical "memory" of the cell, persistent regulatory state, or selection of cells with different fitness (carry-over effects). I would be careful in interpreting the effects as a form of memory associated mostly with epigenetic control - the statement seems stronger than what the data directly demonstrate.

(3) Validation of a small set of top hits by independent assays would substantially strengthen the conclusions. For example, targeted qPCR and/or biochemical assays for key genes (like ribosomal and autophagy) or pathways would help confirm the most important transcriptomic and proteomic trends. Without this, the suggestion of, for example, translational regulation remains possible but unproven.

(4) Overall, the study appears reproducible, with detailed methods, biological replicates, and appropriate statistical analyses. However, reproducibility could be improved by more clearly describing the limitations of the comparisons (especially with the published datasets) and the data-processing parameters, particularly for correlation and functional enrichment analyses. Some comparative conclusions are based on data from different studies, under different conditions and often with different depths of coverage. It is not always clear to what extent interspecies comparisons are fully methodologically similar.

Significance:

This is an ambitious work that combines a genetic screen with multi-omics analyses and cross-species comparisons. The novelty is less in known roles of autophagy, mitochondria and translation, but rather in integrating them into a comparative framework across S. pombe spores, killifish diapause embryos, and dormant human cancer cells, combined with Bar-seq screening and multi-omics. Its greatest strength is its coherent, broad concept: dormant cells are not passive, and the regulated processes and functions during dormancy are evolutionarily conserved rather than the individual set of genes. The contribution of this work lies more in demonstrating process-level conservation than in identifying a single universal gene. This significantly expands the literature on dormancy, but rather in terms of a framework of biological organization rather than a single mechanism.

The weakness of the manuscript is the tendency to draw broad, sometimes overly strong mechanistic conclusions based primarily on correlational data (sometimes with poor statistical significance) and on relative, rather than absolute, expression. The lack of biochemical validation, in my opinion, lowers the impact of the paper.

Audience: This work will be of interest especially to aging and dormancy specialists and comparative biologists. It is important to note that this is not a purely technical or model-specific work, but it has the potential to impact researchers interested in homeostasis, translation and adaptation to stress.

My expertise: My expertise is mostly in aging, sporulation, cell cycle control and ribosome biogenesis.

Reviewer #2 (Public review):

Summary:

In this manuscript, Hassan et al. examine dormancy in fission yeast spores and killifish embryos in diapause. The authors first performed a genetic screen in yeast spores to identify genetic modulators of spore longevity and resilience to heat stress, finding that there is a trade-off between stress resilience and longevity. Next, the authors performed transcriptomic and proteomic profiling of yeast spores and killifish diapause embryos, comparing active and dormant states, early and late dormancy, and with and without heat stress in each model. Comparison between the two models identified a shared set of gene expression changes in dormancy. Lastly, the authors tested whether stress in dormancy is able to reshape the transcriptome or proteome following exit from dormancy, finding that heat stress altered gene expression post-dormancy in both yeast spores and killifish diapause embryos.

This study is interesting because it provides valuable transcriptomic and proteomic datasets across a large number of conditions in both yeast spores and killifish embryos. Furthermore, these datasets reveal several interesting insights about dormancy, including shared gene expression changes in dormancy across diverse organisms, a trade-off between stress resilience and longevity in yeast spores, and memory of heat stress following exit from dormancy.

This interesting manuscript would be strengthened by additional characterization of the functional consequences of these gene expression changes.

Major comments:

(1) The finding that a set of gene expression changes in dormancy is shared between yeast spores and killifish diapause embryos is very interesting and raises the question of whether the genetic determinants of stress resilience or longevity in dormant states are also shared. For instance, do any genes identified in the genetic screen as critical for resilience to heat stress in yeast spores also show a similar phenotype in killifish diapause embryos?

(2) The authors show that gene expression changes in dormancy (old vs. young spores and late vs. early diapause) are poorly correlated with aging in stationary-phase yeast cells (Figure 4E) and killifish brain (Figure 5E). The authors suggest that this poor correlation might reflect a lack of aging during dormancy. Interestingly, the authors also show that in both yeast and killifish models, both time and stress in dormancy result in gene expression changes following exit from dormancy. Are these post-dormancy differences correlated with aging in stationary-phase yeast cells and adult killifish? Additionally, could the authors compare these changes with age-related changes in different adult killifish organs?

(3) In Figures 1C, 4F, and 5F, the authors find significant correlations between two comparisons with a shared condition (e.g. 2-week unstressed spores in Figure 4F). Using the same samples in both comparisons can inflate the correlation between the two sets of fold-changes because the two comparisons are not made independently. The authors should ensure that the correlations found in these figures remain significant even when different samples are used for the shared condition in the two comparisons.

Significance:

This study is interesting because it provides a comparison between different states of dormancy and identify shared pathways. This paper should appeal to a broad audience, including aging, stress pathways, and quiescence/dormancy.

Reviewer #3 (Public review):

Summary:

In this study, the authors use several genome-wide approaches to find commonalities and differences between distinct dormancy models, including S. pombe spores, killifish diapause embryos, and human dormant cancer cells. As G0 states are ubiquitous in nature, yet still understudied, this study provides not only a clear demonstration of an evolutionarily-conserved component in G0 regulation, but also a rich resource of genomics datasets, that will be especially of interest to S. pombe and killifish researchers.

The main conclusions of the study are that: (1) there are commonalities in dormancy programs across eukaryotic evolution; (2) the uncoupling between transcriptome and translatome is more pronounced in non-dividing cells, particularly at ribosomal protein coding genes; and (3) dormant cells retain dynamic responses to stress, changing their cellular state and maintaining this difference.

Overall, the study is clearly written, provides large data resources for future research in the dormancy field, and the conclusions are supported by the presented evidence (although point 3 would require further experimental clarification, see below).

Major comments:

A key consideration when comparing transcriptomes of cycling and dormant cells is the overall RNA quantity in the cells; however, it is not made clear in the manuscript whether this overall level was compared and/or if RNA-seq used spike-in quantification. However, this will directly affect some conclusions; for example, there is a conceptual difference between a transcript upregulated in spores and a transcript that appears upregulated in spores but is in fact 'less repressed than average' (resulting in higher FC). While the overall transcript diversity as described will be not affected, this analysis can provide important context. For the highlighted example of ribosomal protein mRNAs, this scenario could be what resulted in an apparent mRNA increase but protein decrease (which would be interpreted as both mRNA stabilization and translation inhibition); if the total mRNA is much lower, then the mRNA levels may be similar (which would then be interpreted as no change in transcription nor mRNA stability, but highlight the importance of translation inhibition).

This point was more clearly described in a previous seminal study by the authors (Marguerat et al 2012), and so should be made clearer here as well. What is the overall RNA level in cycling cells compared to quiescent cells, and compared to spores? Likewise, what is the overall reduction in total protein levels? Was the overall reduction comparable to the previous study? These overall levels should at least be quantified, which is a straight-forward experiment. The current mention of this limitation (two sentences in the last paragraph of the conclusion) is not enough.

Accordingly, the term "induced" (in Fig 2 for example) may be misleading, and could be separated into truly "induced" genes and "relatively induced" genes, when normalized to absolute levels. Likewise, Fig S2A and S2I are misleading due to this effect (normalization will tend to show a similar number of up-regulated and down-regulated genes). These panels should be clarified by plotting next to the relative quantification (current panels) a comparison taking overall RNA level into account. This will very likely strongly increase the number of downregulated genes and decrease the number of upregulated genes, and the new lists can also be subjected to GO analysis.

The same issue would be important to discuss for the killifish experiment as well.

For spore offspring stress experiments, the estimated number of cell divisions after spore germination should be indicated. It appears over 20, which would confirm the authors' conclusion that an epigenetic state is induced and persists over multiple rounds of replication and underlie improved heat resistance, indeed hinting at transcriptional memory through chromatin modification. As the phenotype is tested at a population-level, it is unclear whether every spore is equally able to impart this phenotype, and for how long the stress resistance response persists. Moreover, it may be clearer to test heat resistance in log-phase cells, as in the current experimental setup cells are grown to stationary-phase before plating; however, stationary-phase cultures are more heterogenous and more heat-resistant.

Note also that using 3-fold serial dilutions (instead of more commonly used 10-fold) means that the differences shown in Fig 3M are mild. Plating duplicates (at log-phase) and imaging at several times of growth would be necessary to have a clearer view of the result.

Significance:

Overall, the study provides a proof-of-concept that dormancy programs across vastly different eukaryotes display commonalities. This is not an entirely new concept, but this is the first demonstration of it using S. pombe and turquoise killifish and will therefore be of interest to a wide audience of researchers interested in quiescence and dormancy in general. A key finding (albeit also not entirely new) is the transcriptome/proteome difference, which appears increased in dormant cells; this hints at the existence of key post-transcriptional regulatory mechanisms in dormant and G0 cells.

A current key limitation is that only relative quantifications are discussed, which can currently give a misleading idea of overall transcriptional activity in dormant cells. Another limitation is that differentially-regulated genes, proteins and GO processes are mentioned, but with relatively limited biological interpretation, making some sections of the manuscript lengthy and descriptive.

Advance: The authors provide a vast amount of datasets for fission yeast spores (transcriptome, proteome, differential viability of a mutant collection) and killifish, which will be a very useful resource for researchers in the field. The authors identified several genes and processes that are commonly differentially regulated in dormant cells across several systems, including autophagy and ribosomal protein genes, which were previously known to be required for G0.

Reviewer #4 (Public review):

Summary:

This manuscript from Hassan et al presents a comparative study of the hallmarks of dormant cells in different systems. The authors first use a bar-coded knockout set to identify genes important for persistence of fission yeast spores in the dormant state (up to 6 months) and then use RNA seq and Mass spec to examine changes in transcript and protein abundance as fission yeast spores enter and exit dormancy. They then perform a similar transcriptomic and proteomic profile of Killifish embryos in diapause. The authors note certain functional classes of genes (e.g. autophagy-related genes or ribosomal proteins) that are similarly upregulated during dormancy in both systems and show that this applies in human dormant cancer cells as well. Importantly, this pattern applies to genes broadly in these functional classes, but not necessarily to any particular gene within the class. Finally, the authors examine the effects of stress (heat shock) on the transcriptomes and proteomes of young and aged spores as well as mitotic populations germinated from those spores.

Dormancy is relatively understudied at the cellular level and this study provides a number of potentially interesting observations, in particular the authors identify common features of dormancy-related expression changes in the different systems that may be a general hallmark of dormancy. Included in this is a fascinating disconnect between transcript and protein levels of ribosomal genes. That said, the work seems to be almost three different stories, none of which is entirely complete, and which only loosely fit together: (1) the identification of dormancy mutants in fission yeast; (2) comparative 'omics of dormancy in different systems; and (3) changes induced by stress in dormant cells.

Specific comments:

There is a dissonance between the apparent upregulation of autophagy genes during dormancy in both systems and the observation that loss of autophagy genes enhance dormancy. While the authors brush this off by noting "that genes regulated in a given cellular state often do not overlap with those functionally required for that state" (an argument that undermines their whole analysis), it raises the possibility that the process signatures they report are not, in fact, functionally linked to dormancy and could be coincidental. What would make the expression analysis more convincing would-be functional data showing that at least one of the upregulated genes in one of the systems is important for dormancy.

It is not clear how the bar-coding experiment was controlled for mutants that make defective spores (e.g. poor spore walls) rather than being required for dormancy per se. As the authors note, several of the mutants with the strongest effects (shortening lifespan of the spores) have been previously identified in screens for sporulation or spore wall defects, consistent with those genes playing a specific role in proper spore assembly rather than a general function in dormancy.

It would be helpful to make the data - or at least the lists of genes identified in the screen and the up and downregulated transcripts/proteins available as supplemental information.

There are at least two different points of diapause in Killifish. It was not entirely clear which diapause was analyzed here. This should be clarified.

The data that offspring from older or heat-shocked spores are more stress resistant (Figure 4M) is not convincing. Serial dilution assays as shown are more qualitative than quantitative and an actual quantitative measure (e.g. c.f.u.'s per cell plated) would be better and would allow statistical confidence to be calculated.

Significance:

This work provides a useful overview and comparison of gene expression changes occurring during dormancy in different systems. As this is a relatively understudied area the results are likely to be of great interest to researchers in this area. The primary weakness, as is often the case for systems studies, is that it provides a 'high level' view of changes, but it is unclear whether any of the specific changes documented are functionally relevant to dormancy.

I am a cell biologist and not qualified to critically assess the quality/appropriateness of the different statistical measures used to analyze their expression data.

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.]

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.

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):

Summary:

Desingu et al. show that JEV infection reduces SIRT2 expression. Upon JEV infection, 10-day old SIRT2 KO mice showed increased viral titer, more severe clinical outcomes, and reduced survival. Conversely SIRT2 overexpression reduced viral titer, clinical outcomes and improved survival. Transcriptional profiling show dysregulation of NF-KB and expression of inflammatory cytokines. Pharmacological NF-KB inhibition reduced viral titer. The authors conclude that SIRT2 is a regulator of JEV infection.

Strengths:

This paper is novel because sirtuins have been primarily studied for aging, metabolism, stem cells/regeneration. Their role in infection has not been explored until recently. Indeed, Barthez et al. showed that SIRT2 protects aged mice from SARS-CoV-2 infection (Barthez, Cell Reports 2025). Therefore, this is a timely and novel research topic. Mechanistically, the authors showed that SIRT2 suppresses the NF-KB pathway. Interesting, SIRT2 has also been shown recently to suppress other major inflammatory pathways, such as cGAS-STING (Barthez, Cell Reports 2025) and the NLRP3 inflammasome (He, Cell Metabolism 2020; Luo, Cell Reports 2019). Together, these findings support the emerging concept that SIRT2 is a master regulator of inflammation.

Comments on revised version.

The authors have addressed my comments.

Reviewer #2 (Public review):

The manuscript by Desingu et al., explores the role of SIRT2 in regulating Japanese Encephalitis Virus (JEV) replication and disease progression in rodent models. Using both an in vitro and an in vivo approach, the authors demonstrate that JEV infection leads to decreased SIRT2 expression, which they hypothesize is exploited by JEV for viral replication. To test this hypothesis, the authors utilize SIRT2 inhibition (via AGK2 or genetic knockout) and demonstrate that it leads to increased viral load and worsens clinical outcomes in JEV-infected mice. Conversely, SIRT2 overexpression via an AAV delivery system reduces viral replication and improves survival among infected mice. The study proposes a mechanism in which SIRT2 suppresses JEV-induced autophagy and inflammation by deacetylating NF-κB, thereby reducing Beclin-1 expression (an NF-κB-dependent gene) and autophagy, which the authors consider a pathway that JEV exploits for replication. Transcriptomic analysis further supports that SIRT2 deficiency leads to NF-κB-driven cytokine hyperactivation. Additionally, pharmacological inhibition of NF-κB using Bay 11 (an IKK inhibitor) results in reduced viral load and improved clinical pathology in WT and SIRT2 KO mice. Overall, the findings from Desingu et al. are generally supported by the data and suggest that targeting SIRT2 may serve as a promising therapeutic approach for JEV infection and potentially other RNA viruses that SIRT2 helps control.

Comments on revised version.

We believe the authors of Desingu et al. have adequately addressed the major concerns of each reviewer. Thus, we believe the current version of the manuscript is significantly improved, and the data more strongly support the authors' conclusions.

We would like to point out that the statistics in Figure 8 are difficult to interpret and should be improved to make them useful to the reader and strengthen the paper. Also, line 70 has an extra period that should be removed.

Reviewer #1 (Public review):

This work reports the finding of a dissociable human Mediator complex subunit, MED16, and its recruitment of transcription factors UBP1 and TFCP2 that form a trimeric complex regulating cellular gene transcription and inhibiting HIV promoter activity, thus promoting MED16-targeting inhibitors as HIV latency-reversing agents for anti-HIV therapy. In this revision, the authors provide new data indicating that MED16 action occurs in the holo-Mediator complex, not as a free form, for its functional association with UBP1 and TFCP2.

Summary:

Characterization of a dissociable Mediator subunit implicated in cellular pathways, particularly lung alveolar function, and HIV latency is conceptually interesting.

Strengths:

The strengths of this study are:

(1) Demonstration of MED16 dissociation from the core Mediator complex and formation of a subcomplex containing MED16, upstream-binding protein 1 (UBP1), and transcription factor cellular promoter 2 (TFCP2) by elegant biochemical fractionation and immunoblotting analysis.

(2) Defining nine N-terminal WD-40 repeats (WDRs) of MED16 as a Mediator-incorporating module and the C-terminal ⍺β-domain (157 amino acids) important for interaction with the UBP1-TFCP2 heterodimeric complex.

(3) Illustration of a weak hydrophobic interaction between MED16 and the Mediator core that could be disrupted by 1,6-hexanediol, but not by its 2,5-hexanediol isomer nor by high salt (500 mM NaCl) disruption.

(4) Classification of UBP1-upregulated cellular genes typically containing binding sites flanking the transcription start site (TSS) in contrast to UBP1-downregulated genes often containing a TSS-overlapping UBP1-binding site;

(5) Presenting evidence for Mediator complex-dissociated free MED16-repressed HIV promoter activity through functional association with UBP1 and showing bromodomain-containing protein 4 (BRD4) inhibitor JQ1 that potentially disrupts BRD4-inhibited HIV-1 transcription elongation could lead to reversal of HIV-1 latency.

Weaknesses:

(1) No clear demonstration that MED16-UBP1-TFCP2 indeed form a trimeric core subcomplex in regulating cellular gene transcription and HIV-1 promoter inhibition.

(2) No validation of transcriptomic datasets and pathways identified.

(3) Use of mostly artificial reporter gene constructs and non-HIV host cells (e.g., human 293T embryotic kidney cells, human HeLa cervical cancer cells, and mouse HT pancreatic cancer cells) for examining MED16/UBP1-regulated HIV transcription.

(4) Inconsistent use of 293T and HeLa cells in the characterization of dissociated MED16 interaction with UBP1 and TFCP2.

(5) In vitro transcription using immobilized DNA templates was not performed to a high standard thus failing to convincingly show MED16/UBP1-inhibited HIV-1 transcription preinitiation complex formation.

In this revision, the authors stipulate that the action of MED16-UBP1-TFCP2 is through the holo-Mediator complex, not as a free trimeric complex, since deletion of the N-terminal Mediator-incorporating module abolishes the functional effect of MED16 even though its interaction with UBP1-TFCP2 persists. The authors also show that a predicted UBP1-binding site is indeed present in the promoter region not overlapping with the TSS in each of the UBP1/MED16-upregulated respiratory gaseous exchange genes, such as surfactant protein A1 (SFTPA1), SFTPA2, SFTPB and SFTPC. To demonstrate UBP1/MED16-inhibited HIV transcription in a more biologically relevant system, the authors employ an HIV-1 latency model derived from CD4+ T cells. Additional experiments are also performed to address other issues raised by the reviewers.

Overall, the revised manuscript has been improved by addressing some of the reviewer's comments. Additional issues that came up while reading this revision are: a) The statement, "UBP1 (as known as LBP-1) and TFCP2 (as known as LSF) belong to a subfamily of the TFCP2/Grainyhead transcription factor family" (lines 63-64), is confusing and needs to be clarified. My understanding is that LBP-1 has four family members encoded by two different genes (see Yoon et al., MCB 14: 1776-1785, 1994): LBP-1a and LBP-1b are probably UBP1, and LBP-1c and LBP-1d are likely TFCP2/LSF. If this is correct, the authors should clearly indicate which specific isoforms are analyzed in their UBP1 and TFCP2 proteins and clarify some of the text descriptions seemingly contradictory to the earlier findings.

The use of mostly artificial reporter templates and transfection assays conducted mostly in 293T and HeLa cells remains concerning. It is also unclear why overexpression of MED16 only, but not MED23 or MED24, inhibits HIV promoter activity (Figure 5) if MED16 action is through the holo-Mediator complex and overexpression of MED16 may titrate out key cellular factors, thus diminishing their action through holo-Mediator. Results based on this squelching type of experiment relying on overexpression of a single component needs to be carefully interpreted.

Reviewer #2 (Public review):

Summary:

The article from Zheng et al. proposes that the Med16 subunit of Mediator can dissociate from the complex, associate with transcription factor UBP1, and contribute to activation or repression of specific sets of genes in human cells. The question is potentially interesting, and the possibility of a distinct role for Med16 in UBP1-dependent transcription could be of interest to the field. However, I have substantial concerns about the experimental design, reproducibility, and interpretation of the data. In particular, several of the central conclusions rely on single experiments, indirect assays, or perturbations that are expected to cause broader changes in Mediator structure and transcription. In my view, the current data therefore do not provide sufficient evidence to support several of the principal mechanistic conclusions of the manuscript.

Strengths:

The authors provide preliminary evidence that Med16 may exist in an altered or partially dissociated Mediator complex and may be associated with specific transcriptional effects. The use of biochemical assays, reporter assays, RNA-seq, and Med16 knockout/rescue experiments provides several potentially useful approaches to investigating this question. In particular, the Med16 knockout/rescue RNA-seq experiment described in the rebuttal is a potentially informative experiment. Making the underlying IP-MS abundance and peptide-count data available is also an improvement over the previous version.

However, important limitations of these experiments need to be addressed before the proposed mechanism can be established.

Comments on the revised version:

In the previous review, I raised concerns that the evidence for a specific UBP1-Med16 interaction and for a mechanistic role of this interaction was limited. The revised manuscript does not resolve these concerns. In particular, whether UBP1 interacts specifically with Med16, whether this interaction is required for activation of UBP1 target genes, and whether UBP1 represses HIV-1 transcription through Med16 remain unresolved.

A stronger case for a specific UBP1-Med16 interaction would require more direct biochemical, structural, or biophysical evidence. The current experiments do not distinguish sufficiently between a specific Med16-dependent mechanism and broader consequences of perturbing Mediator structure or transcription.

The revision also does not include the additional biological replicates needed to assess the reproducibility of several foundational experiments. Consequently, a number of the central conclusions remain based on single experiments, which substantially limits the strength of the conclusions that can be drawn.

Major concerns:

Figure 1:

All of the data in Figure 1 are from single experiments without biological replicates. Because Figure 1 provides an important foundation for the proposed mechanism, this limits confidence in the reproducibility of the observations. Western blot experiments can show substantial experiment-to-experiment variability, and such variability appears relevant to several of the observations presented here.

A key experiment is the western blot analysis of the gel-filtration experiment in Figure 1D. Rather than migrating with Mediator, the majority of the detected Med16 appears to elute at a lower molecular weight. The extract was maintained at super-physiological salt concentrations from preparation through column loading and elution, which could potentially contribute to this result. Other possibilities include experiment-to-experiment variation or detection of a cross-reacting protein by the Med16 antibody. The latter possibility warrants consideration because the bands detected in fractions 28-31 appear to migrate at a different molecular weight from the Med16 bands in fractions 16-19. Additional replicated experiments and appropriate controls would be needed to establish that the lower-molecular-weight signal represents Med16 in a distinct complex.

Figure 2:

The data in Figure 2 are also based on single experiments without biological replicates. The biochemical experiments use protein fragments that do not occur as such in the native protein. Such experiments can be useful for mapping possible protein-protein interaction surfaces, but they can also generate false-positive or false-negative interactions depending on the experimental configuration.

Some of the results raise additional questions about the proposed interaction model. For example, the observations involving TFCP2 in panel B and Med1 and Med23 in panel F appear difficult to reconcile with a simple model in which UBP1 interacts with Mediator through Med16. Panel 2F is particularly relevant because Med1 and Med23 are not detected despite the proposed model involving UBP1 interaction with Mediator. These observations should be addressed when interpreting the specificity and mechanism of the proposed UBP1-Med16 interaction.

Figure 3:

The reporter assays provide potentially useful preliminary evidence for factors that influence transcription. However, the relatively long experimental timeframes (36-48 hours after transfection), use of artificial promoters, and overexpression of proteins introduce important limitations.

The data are consistent with the authors' model, but they are also consistent with a simpler explanation in which Med16 depletion broadly affects Mediator structure and function. Previous work from the Gross, Robert, and Roeder laboratories has shown that loss of Med16 can result in dissociation of Med23, Med24, and Med25 from Mediator. Importantly, the authors' own data in the rebuttal document appear to show similar effects following Med16 loss. These observations are important because Med23, Med24, and Med25 provide major interfaces for DNA-binding transcription factors, and Med16, Med23, Med24, and Med25 together represent a substantial fraction of the molecular mass of Mediator. The functional properties and stability of the resulting Mediator complex therefore need to be considered when interpreting Med16 depletion experiments.

The authors' data also suggest broader transcriptional effects: Med16 knockout causes loss of basal transcription in panel 3E. Conversely, overexpression of Med16 or Med16 fragments can cause broad defects in transcriptional activation, including through squelching. The relatively modest effects observed despite substantial increases in protein abundance are therefore also consistent with non-specific consequences of overexpression. These alternative explanations need to be considered when interpreting the reporter assays.

Figure 4:

RNA-seq is an appropriate approach for investigating possible effects of Med16 on gene expression. However, the experimental design has important limitations because Med16 depletion is expected to affect the composition of Mediator, including loss of Med23, Med24, and Med25. The knockdown also occurs over an extended timeframe (>24 hours), making it difficult to distinguish direct effects of Med16 from secondary effects arising from altered Mediator structure or broader changes in transcription.

The RNA-seq results could therefore reflect Med16-specific functions, but they are also consistent with a model in which loss of Med16 compromises Mediator structure and function and produces downstream gene-expression changes.

The presentation of the RNA-seq data in Figure 4E is also difficult to interpret. The volcano plot shows UBP1 overexpression versus the shCTRL vector, whereas the comparison central to the authors' interpretation concerns the effects of shMed16. Showing the volcano plot for the shMed16 comparison would allow readers to assess the statistical significance and magnitude of the relevant gene-expression changes directly. Instead, the authors superimpose "x" symbols for genes that "failed to be activated by UBP1 in the shMed16 cell line." Because the underlying Med16 comparison is not displayed, the p-values and fold changes for the comparison of interest cannot be evaluated. Moreover, positioning the "x" symbols according to the coordinates of the UBP1-overexpression volcano plot does not appear to represent their corresponding positions in a volcano plot of the Med16 comparison. This makes the figure difficult to interpret and could give a misleading visual impression of the underlying data.

Finally, the RNA-seq data are not spike-in normalized. This is potentially important because Med16 depletion is expected to cause broad effects on transcription and could result in an overall reduction in transcription across the genome. Appropriate normalization is therefore important for distinguishing gene-specific effects from global changes. I also did not identify information in the manuscript indicating where the RNA-seq data have been deposited in a public repository.

Figure 5:

The reporter assays in Figure 5 are subject to similar concerns. The authors' model may explain the observations, but the data are also consistent with broader disruption of Mediator structure and function following Med16 depletion, or with more general transcriptional effects caused by overexpression of Med16 or Med16 fragments.

A separate concern is the proposed mechanism for HIV-1 repression. The authors propose that UBP1 interacts with Med16/Mediator and that overexpression of Med16 enhances Mediator recruitment, although Mediator recruitment is not directly measured. The proposed increase in Mediator recruitment would, in isolation, be expected to promote transcription rather than repression. The authors instead invoke a UBP1-dependent mechanism to explain the repression. However, a simpler explanation is that overexpression of Med16 or selected Med16 fragments perturbs normal transcription through mechanisms including squelching. The authors themselves invoke a Med16 squelching model to explain reduced transcription at the UBP1 reporter in Figure 3G-I. The possibility that the same mechanism contributes to the results in Figure 5 should therefore be addressed.

Finally, a central component of the proposed model is recruitment of Mediator by UBP1 through Med16, but neither UBP1 occupancy nor Mediator occupancy at the relevant promoters is directly measured. These experiments would provide important evidence for the proposed mechanism. There is also no legend for Figure 5G,H, and the legend for Figure 5C,D appears to be incorrect.

Figure 6:

The title of Figure 6 states that "The position of the UBP1 binding site relative to the TSS dictates whether MED16-UBP1 activates or represses transcription." The data presented do not establish this conclusion.

Only one alternative location of the UBP1 binding site is tested, and the analysis in panel D compares only the 50 most up- or down-regulated genes in UBP1-overexpression experiments. The correlations shown are relatively weak, and a rigorous statistical analysis is not presented.

The authors cite Duttke et al. (Nature 2024) in support of the concept that transcription-factor position can influence regulatory activity. That study used genome-wide mapping of native TSSs, MPRA analysis of approximately 2,000 promoter variants, and targeted follow-up experiments. By contrast, Figure 6 relies on an artificial reporter and a single change in UBP1 binding-site position. The current data therefore support the possibility that binding-site position influences the observed reporter activity, but they do not establish the broader conclusion that the position of the UBP1 binding site relative to the TSS dictates whether Med16-UBP1 activates or represses transcription.

Figure 7:

The title of Figure 7 states that Med16 "interferes with HIV-1 PIC assembly," but the data presented do not establish this mechanism.

First, the proposed mechanism is described as acting through interaction with UBP1, but UBP1 is not measured in this experiment.

Second, Med16 knockout or overexpression has relatively modest effects on the estimated levels of bound PIC components, and these changes do not consistently correspond to the transcriptional changes shown in panel C. For example, in the WT versus Med16 KO comparison, there is no apparent difference in the levels of Pol II, Med1, IIB, or IIH based on the western blot data. The reduction in Med16 and Med23 is expected following Med16 knockout and therefore does not by itself establish altered PIC assembly. In the WT versus Med16 overexpression comparison, panel C suggests a meaningful reduction in transcription, whereas the changes in PIC-associated proteins do not follow the direction predicted by the proposed model: Pol II and IIB increase with Med16 overexpression, while Med1, Med23, Med16, and IIH appear similar.

Third, Cdk8 and Med12 are not PIC components. They are subunits of the CDK8 module, which can inhibit PIC formation. Even under a simplified interpretation in which reduced Cdk8 and Med12 levels were relevant to PIC formation, the direction of these changes is opposite to that expected from the "relative mRNA level" plots in panel C. Reduced levels of these proteins in the Med16 overexpression condition would, under this interpretation, be expected to promote rather than reduce transcription.

Finally, this is a single-replicate experiment, and quantitative comparisons are made across immobilized-template pulldowns using different nuclear extracts (WT, Med16 KO, and Med16 OE). Differences in protein concentration between extracts, differences in pulldown efficiency, and experiment-to-experiment variation could all contribute to the observed differences. Replicated experiments with appropriate normalization would therefore be needed to support the proposed PIC mechanism.

Figure 8:

Panel F presents a model that, in my view, goes beyond what is supported by the experimental data. In particular, there is currently insufficient evidence to establish the proposed Med16-UBP1-TFCP2 complex as a distinct functional complex.

The model also depicts the Mediator CDK8 module bound at promoters together with Pol II. The manuscript does not provide evidence supporting this arrangement, and there is substantial prior evidence that the CDK8 module can inhibit or prevent PIC formation. This aspect of the model therefore requires either experimental support or clarification.

Finally, the model depicts a separate Med16-UBP1-TFCP2 complex. This appears inconsistent with the authors' statements in the rebuttal that they did not intend to propose such an unsupported isolated complex. If this is not the intended model, Figure 8 should be revised to avoid conveying that interpretation.

Figure S1: IP-MS analysis:

The IP-MS data raise substantial concerns regarding the specificity of the proposed UBP1-Med16 interaction.

The authors state that they "have now reanalyzed the IP-MS data using a more precise quantification approach." However, the filtering procedure described in the Methods is difficult to evaluate and requires substantially more justification. The procedure includes multiple sequential criteria: detection in at least two of three replicates of the target bait; a minimum unique peptide count; detection in at least one additional Mediator bait; exclusion of Mediator/Pol II subunits, ribosomal proteins, histones, keratins, and common contaminants; and annotation with transcriptional or DNA-binding functions using GO terms. The rationale for this particular combination of filters and its impact on the candidate list should be explained and justified.

The resulting "bait-preferential score" is also difficult to assess. The reported mean_target and mean_other values are frequently negative, with some positive values, and the method appears to derive a positive score from the difference between these values after taking an absolute value. The biological and statistical interpretation of this score is therefore unclear. The reported scores are generally clustered around relatively narrow values, and the analysis does not appear to provide the statistical framework normally used to assess enrichment and reproducibility in IP-MS experiments.

Importantly, the candidate proteins identified by this approach do not appear to overlap with proteins identified in other published Med16 IP-MS analyses in human cell lines, including HCT116, MCF7, HEK293T, and HeLa. I was unable to identify a protein among the 37 Med16 candidates that was consistently observed in the cited resources, including BioPlex (Huttlin et al., Cell 2021) and Malovannaya et al. (Cell 2011). This lack of concordance raises questions about the specificity and physiological relevance of the proposed Med16-associated proteins and, in particular, about the evidence for a specific UBP1-Med16 interaction.

The peptide counts for UBP1 also illustrate the limitation of the current analysis. The total UBP1 peptide counts across the Cdk8, Med1, and Med16 IP-MS experiments are 12, 10, and 14, respectively, yielding bait-preferential scores of 1.07, 0.76, and 1.85. Given these relatively small differences and the absence of statistical significance measures, it is difficult to determine whether these observations provide evidence for a specific biological interaction.

The authors have now made the raw abundance values and peptide counts available as supplementary tables, which is useful. However, standard statistical measures for evaluating MS data, including FDR, p-values, and adjusted p-values, are not provided, nor is sufficient information given about how the reported scores were calculated. These additions would be important for readers to assess the robustness of the conclusions.

Concerns about the rebuttal document:

Several statements in the rebuttal appear inconsistent with the data shown in the revised manuscript or with the interpretation of the authors' own experiments. I highlight specific examples below.

(1) The proposed Med16-UBP1 complex:

I previously raised concerns that the manuscript overstated the evidence for a separate Med16 complex associated with UBP1 and lacked rigorous biochemical characterization of such a complex. The authors responded that the wording of the original manuscript may have created this interpretation and stated that this was not the mechanistic model they intended to convey. However, Figure 8 of the revised manuscript still depicts a separate Med16-UBP1-TFCP2 complex. The figure therefore appears to retain the interpretation that the authors stated they did not intend to convey. This should be clarified.

(2) Replication and experimental rigor:

I previously raised concerns about the absence of replication and the limitations of the experimental approaches used to infer roles for Med16 and UBP1. These concerns remain because no additional replicate experiments were performed for several of the key experiments. As a result, important conclusions continue to depend on single western blots, including foundational experiments in Figures 1, 2, and 7.<br /> Examples include the variable quantification in Figure 1A; the absence of detectable Med23 and Med1 in the UBP1 IP in Figure 2F despite the proposed interaction between UBP1 and Mediator; and the variable association of TFCP2 with Flag-MED16 fragments in Figure 2B, including lanes 7-9. Replication would be important for determining whether these observations are reproducible.

(3) IP-MS analysis:

The revised IP-MS analysis provides additional information compared with the previous version, but important concerns remain regarding the processing and interpretation of the data. The "bait-preferential score" approach is not sufficiently justified, and standard measures such as p-values, adjusted p-values, or FDR are not provided.

The analysis also identifies no proteins among the 37 Med16 candidates that appear to be consistently observed in other published Med16 IP-MS analyses in human cell lines. This is important when assessing the specificity of the proposed interaction.

The peptide counts for UBP1 are similarly low and do not show a clear Med16-specific enrichment: 12 peptides in the Cdk8 IP, 10 in the Med1 IP, and 14 in the Med16 IP, corresponding to scores of 1.07, 0.76, and 1.85, respectively. These observations do not by themselves provide strong evidence for a specific UBP1-Med16 interaction.<br /> In addition, the Med16 IP experiments identify approximately 800 proteins per replicate. This broad set of associated proteins highlights the importance of appropriate controls and statistical analysis when interpreting a co-IP as evidence for a specific protein-protein interaction.

(4) Interpretation of the 1,6-HD experiments:

I previously raised concerns that the IP-western data in Figure 1 did not establish a Med16-specific effect. The rebuttal states that "MED16 exhibits a uniquely pronounced dissociation by 1,6-HD that is consistent across three independent IPs using anti-CDK8, anti-MED1, and anti-MED12 antibodies." However, the quantitative data shown do not clearly support this interpretation.

In the CDK8 IP, CDK8 and MED23 appear to be lost to a similar extent as Med16. In the Med1 IP, Med12 shows a similar loss to Med16, with Med23 also showing a comparable trend. In the Med12 IP, CDK8 and Med23 show changes comparable to Med16. There are also unexplained increases in some Mediator subunits with increasing 1,6-HD concentrations. For example, Med12 and Med1 are variable across the 1,6-HD conditions, and Med1 and Med24 increase at higher concentrations. Similar trends are present in the CDK8 IP.

The Figure 1B data raise related concerns. Several Mediator subunits show substantial increases following the 2,5-HD condition compared with 1,6-HD, in some cases greater than the change observed for Med16. Thus, the available data do not clearly establish that Med16 exhibits a uniquely pronounced dissociation from Mediator.

(5) Evidence for the proposed mechanism:

I previously suggested that more rigorous approaches would be needed to distinguish a specific Med16-dependent mechanism from broader effects of Med16 perturbation. The rebuttal instead emphasizes that the authors' conclusion that the UBP1-TFCP2 heterodimer acts through Med16 is supported by multiple independent lines of evidence.

However, the different experiments do not independently establish the same mechanistic conclusion. Many rely on long experimental timeframes, overexpression, reporter assays, limited controls, or lack of biological replication. These approaches can provide useful preliminary evidence for effects on transcription but do not establish that a specific UBP1-Med16 interaction is required for UBP1 target-gene expression.

Additional approaches that directly test the proposed interaction and its requirement for transcriptional regulation would therefore be needed to establish this mechanism.

(6) Med16 rescue experiment:

The authors performed an RNA-seq comparison of WT, Med16 KO, and Med16 KO cells rescued by re-expression of WT Med16. This is a potentially informative experiment, and I appreciate that it was performed.

However, the heatmap shown in the rebuttal does not appear to demonstrate a clear rescue of UBP1 target-gene expression. The figure contains approximately 60 genes but provides no legend beyond "Gene profiling of MED16 KO and rescue effect." The authors describe the result as showing that re-expression of Med16 can largely restore expression of UBP1 target genes, but this conclusion is not readily apparent from the figure as presented. More information about the analysis, normalization, and comparisons would be needed to evaluate the claimed rescue. It would also be useful to include the relevant analysis in the revised manuscript rather than only in the rebuttal.

(7) Consequences of Med16 depletion:

A continuing concern is that the manuscript does not adequately account for the broader consequences of Med16 depletion or knockout. Loss of Med16 is known to cause dissociation of other Mediator subunits, including Med23, Med24, and Med25. The authors acknowledge this in the rebuttal and state that their new Co-IP and IP-MS experiments confirm partial dissociation of these subunits following Med16 knockout. They also acknowledge that this structural perturbation must be considered when interpreting phenotypes from Med16 depletion and that long-term knockdown or knockout approaches are susceptible to secondary effects and cellular adaptation.<br /> These are important caveats, but they are not adequately incorporated into the interpretation of the revised manuscript, and the new data demonstrating loss of Med23, Med24, and Med25 are not shown in the article.

This issue is particularly important for interpreting the transcriptional experiments. Loss of Med16 is expected to disrupt Mediator more broadly, including loss of Med23, Med24, and Med25, which provide important transcription-factor interaction surfaces. These changes could produce widespread transcriptional effects independently of a specific UBP1-Med16 mechanism. Given that the authors' own experiments now provide evidence for this structural perturbation, the possibility of indirect effects and altered Mediator function should be explicitly incorporated into the experimental interpretation.

Overall assessment:

The manuscript addresses an interesting question and presents several potentially useful observations concerning Med16, Mediator, and UBP1. However, the central mechanistic conclusions are currently not supported by sufficiently rigorous or independent evidence. The principal concerns are the extensive reliance on single-replicate experiments; the use of overexpression, artificial reporters, and long depletion timeframes; the absence of direct measurements of UBP1 and Mediator occupancy; the broad structural consequences of Med16 depletion; and the limited statistical support for the IP-MS analysis.

Taken together, the current data do not distinguish adequately between the proposed specific UBP1-Med16 mechanism and alternative explanations involving broader disruption of Mediator structure or transcription. Addressing this distinction is important because it affects the interpretation of much of the manuscript, including the RNA-seq and reporter-assay results and the proposed HIV-1 mechanism.

The authors have generated several useful datasets, including the Med16 knockout/rescue experiment and additional IP-MS information, but the manuscript would need to incorporate the associated caveats and provide stronger experimental evidence before the proposed mechanistic model can be established.

Reviewer #1 (Public review):

Summary:

This is a useful study by Angla et al that investigates the basis of observations made from previous studies where loss of Onecut (OC) transcription factors lead to changes in spinal interneuron populations that do not themselves normally express OC. The authors hypothesize that OC expression in spinal motor neurons has non-cell-autonomous effects on pre-motor interneuron (V1, V2a/b/c) population size and distribution. By knocking out OC in the motor neuron lineage (i.e., downstream of Olig2, a motor neuron progenitor marker gene), they indeed show that motor neuron-specific loss of OC expression decreases V2c interneuron number and alters the spatial distribution of V1, V2a, V2b, V2c populations. Using bulk RNA-sequencing of WT and OC conditional knockout (cKO) motor neurons, the authors identify that neurotrophic factor Ntf3 as another factor that is responsible for the non-cell-autonomous effects on the interneuron population; upon knock out of Ntf3 downstream of Olig2 they show that this leads to increased interneuron numbers and alters their spatial distribution, ultimately leading to dysregulation of spinal motor circuits and motor activity.

Strengths:

The authors use sophisticated genetic tools to precisely remove OC and Ntf3 expression in a lineage-specific manner and comprehensively assess the downstream effects across brachial, thoracic, lumbar levels of the spinal cord, as well as at two developmental timepoints, E12.5 and E14.5.

Weaknesses:

The evidence supporting their claims is incomplete. In figure 2 and 4, the control distributions (despite reporting the same populations in the same regions) look different, suggesting large sample-to-sample variance in distribution. The authors explain that this is due to several technical limitations in their study.

Furthermore, the way the Ntf3 results are framed can be misleading; it reads as though one of the main claims of this study is that OC acts through Ntf3.

Reviewer #2 (Public review):

The study by Angla et al. proposes a model in which NT-3 produced by motor neurons regulates interneuron numbers and distribution in a non-cell autonomous manner. The authors demonstrate that ablation of motor neurons (MNs) and global and conditional deletion of OC transcription factors lead to changes in interneuron distribution. They identify that NT3 is upregulated after MN-specific OC deletion in RNA-seq experiments and show that olig2-cre mediated NT3 deletion leads to increased ventral interneuron numbers, altered distribution and defects in locomotor behavior. The authors conclude that MN-derived NT3, under OC control, regulates interneuron development. While this is an intriguing hypothesis, the study does not consolidate the results from the different experimental models in a cohesive model to provide any mechanistic insights into how NT3 may accomplish this function.

(1) The authors utilize different experimental models (OC transcription factor global and conditional knockouts, DTA ablation and NT3 deletion) to show that MNs may influence interneuron numbers and positioning, but there doesn't seem to be a direct link into how each manipulation relates to the others in an informative way. For example, the relationship between OC and NT3 deletion data is not entirely clear, even though NT3 is downstream of OC transcription factors. Both deletions lead to changes in interneuron distribution but there doesn't seem to be any correlation between the two that relates to relative changes in NT3 levels. Although it is understandable that transcription factor manipulations may have additional effects, it is hard to derive any mechanistic insights into how exactly MNs influence interneuron differentiation or migration from the data presented here. Similarly, while the rationale for following up on NT3 seems to be the chick electroporation experiments, only a very minor decrease in Chx10 interneurons is shown in those experiments. Shouldn't NT3 overexpression lead to substantial decreases in IN numbers according to the authors' model?

(2) The study primarily quantifies interneuron numbers and distribution at different levels of the spinal cord and under different genetic manipulations. In different figures, the values and distributions shown for controls vary quite a lot. For example, in figure 2B vs figure 4B, the number of FoxP2+ V1 neurons at brachial levels is ~350 vs 125. Similarly, the control distributions in 2I and 4I are quite different. While this can be partially explained by technical issues, it remains challenging to determine whether the conclusions regarding the impact of each genetic manipulation on interneuron numbers and distribution are valid.

(3) It is not clear that the behavioral phenotypes seen in the olig2-cre mediated deletion of NT3 can be attributed to specific changes in interneuron development, connectivity or function. While the behavioral phenotype is in theory consistent with defects in motor circuits, there is a big gap between the embryonic changes shown here and behavior, with no in between circuit level changes in locomotor circuits shown. Thus, the behavioral experiments are not very informative as to how exactly components of motor circuits might be altered with the NT3 deletion.

Reviewer #3 (Public review):

This manuscript aims to investigate cell extrinsic mechanisms that regulate differentiation and distribution of interneuron types in the spinal cord. The authors demonstrate that the loss of motor neurons leads to changes in number and distribution of different interneuron types, specifically V0v, V1, and V2b (but not V2a). The authors then hypothesize that this phenotype may be controlled by the action of Onecut (OC) transcription factors in motor neurons. Conditional knockout of OC1 + OC2 in motor neurons using Olig2-Cre, however, does not lead to significant changes in numbers of V1, V2a, V2b interneurons, although there is a change in their spatial distribution. While the authors do not check V0v neurons in OC mutants, they do check V2c, which show a reduction in number and change in distribution. Why the same neurons are not checked across experiments is unclear. The authors then analyze existing RNA-seq data to identify factors that could be mediating effects of the OC factors in motor neurons. They identify Ntf3 as a candidate and confirm that it is upregulated in OC mutants. Conditional loss of function of Ntf3 (Olig2-Cre) leads to increases in V1, V2a, and V2b (but not V2c) interneurons and changes in distribution of all four interneuron types. However, the Ntf3 phenotypes do not necessarily suggest that it works in the same pathway as OC to control interneuron development. Finally, the authors demonstrate that these Ntf3 conditional mutants have major defects in motor function.

The main conclusions of the manuscript are that i) motor neurons non-cell autonomously control interneuron development, ii) that OC activity in Olig2 lineage cells non-cell autonomously controls aspects of interneuron development, and iii) that Ntf3 in motor neurons non-cell autonomously controls aspects of interneuron development. The mechanistic link between OC and Ntf3 activity is not clear. And the interneuron phenotype seen when motor neurons are depleted it not phenocopied by either loss of OC or Ntf3 from motor neurons. While the authors addressed some of my initial comments, the impact of this work on the field remains difficult to assess for the following reasons:

(1) The manuscript relies heavily on quantifying numbers and spatial distribution of interneuron populations. However, these do not seem to be consistent in control animals across experiments, making it difficult to interpret any changes observed in genetic manipulations. Specifically, in Fig. 2 and 4, the same markers are being used to quantify V1, V2a, V2b, and V2c interneurons in controls vs. OC (Fig. 2) or Ntf3 (Fig. 4) conditional knockouts, but the numbers of neurons and their distribution in control animals are variable between these two figures. For example, there seem to be a mean of >300 V1 neurons in E12.5 brachial sections of Fig. 2 controls, but this number is <150 in Fig. 4 controls. The cell distribution scoring is similarly variable between these controls without any explanation. The same is true for E14.5 controls used in Fig. S2 vs. Fig. S6.<br /> The authors explain in response to the 1st review that technical variability including different antibodies, different researchers performing experiments, and different mouse backgrounds may contribute to this variation, but do not include this explanation clearly in the manuscript where it would be useful for readers. It would have also helped, for example, if authors had labeled datapoints on bar plots based on the animal that they came from to understand inter-animal variability within experiments. Overall, the biological significance of the results remain difficult to interpret due to high technical variability.

(2) In the initial review, it was noted that there is no data to support the claim that OC1/2 mutants have altered interneuron development because of higher Ntf3 expression. In response to this, the authors note that the manuscript does not claim that OC exerts its non-cell autonomous effects through Ntf3 and add a new paragraph in the discussion which now addresses the fact that it is difficult to compare the phenotypes observed the different genetic manipulations presented in the work. While this paragraph addresses the incomplete mechanistic link between OC and Ntf3, it also decreases the mechanistic impact of the work.

(3) In Figure 3E-3G, the authors show that some V1, V2a, V2b interneurons express the Ntf3 receptor TrkC. It is not stated (apologies if I missed it) what age this expression is, and from the images, it looks like in addition to medially located newborn neurons being negative, laterally located Gata3 interneurons are also broadly negative for TrkC, and many Foxd3 and Chx10 interneurons are also negative. Separate images of each channel and/or scoring would help clarify this. If many relevant interneurons are indeed TrkC negative, how this affects the conclusions should be addressed in the manuscript.

(4) Behavioral phenotypes are not necessarily caused by Ntf3's role in premotor interneuron development.

Reviewer #2 (Public review):

Summary:

Vertebrate spinal cord receives inputs from many supraspinal regions. The authors used optical backfilling to trace neurons in the zebrafish larval brain sending axons to the spinal cord . With two-photon microscopy, they managed to render a comprehensive 3D map of these neurons and drew homologs with mammalian brain structures.

Strengths:

The main strength lies in the precise 3D mapping. The fact that most of previously reported neuron groups have been confirmed by their approach is a solid endorsement of their methodology.

This study provides a comprehensive alternative anatomical reference framework for studying individual groups of supraspinal neurons with projections to the zebrafish spinal cord.

Weaknesses:

The backfilling approach does not reveal the full trajectories of axons, which is available to some degree by ZExplorer Atlas.

Reviewer #1 (Public review):

Summary:

The study by Fernandes et al used multiple types of experiments to study the increased metastatic activity conferred by p95-HER2 and presented evidence that biased downstream signaling was engaged by p95-HER2. Further studies demonstrated crucial phosphorylation sites and downstream transcription targets. Overall, this study will contribute to increased understanding of HER2 biology and breast cancer metastasis.

Strengths:

Multiple types of experiments to assess the increased invasion property by p95-HER2. Solid evidence was provided that p95-HER2 elicits a different transcription program, enhancing EMT. Convincing results were achieved in knockdown studies on crucial genes such as MRTFA and TGFB1I1.

Weaknesses:

To enhance the quality of this study, several points need to be addressed.

(1) First, as the background to this study, the authors stated in the introduction that different isoforms of HER2 lead to different invasive properties. Is this conclusion based on clinical observation or animal models? This needs to be specified to help readers grasp the importance of the paper.

(2) Second, this study is mostly performed on three cell lines generated from three tumor models, WT, d16, and p95. Were multiple cell lines generated from the same tumor model (e.g., p95), and how similar are these cell lines if they are from the same tumor model?

(3) Third, Figure 2 presented data from a very interesting experiment. However, the overall conclusion of this experiment is hard to grasp from the text.

Reviewer #2 (Public review):

Summary:

This is a significant study with implications for signal transduction, breast cancer, and metastasis research. The study uses cell lines from the innovative HER2BOW model mouse to compare and contrast properties of mouse mammary tumors induced by wt HER2, dex16 HER2, and p95 HER2. The linkage between p95 Y1139 and TGFB1I1, as well as MRTFA, is very interesting. The study involves mechanistic approaches to distinguish signaling from dex16 HER2 and p95 HER2 that will be well received by the field.

Strengths:

The study describes the development of novel cell lines for mouse mammary tumors induced by different forms of HER2. In vitro analysis includes a wide variety of assays, including high-resolution fluorescent imaging and tracing of cell movement/ECM degradation, and single-cell RNA-seq analysis. Several approaches are used to confirm p95 HER2 pY1139-induced signaling through nuclear MRTFA and TGFB1I1. shRNA and CRISPR/Cas9 were used to confirm signaling through this pathway.

Weaknesses:

The authors have only described one cell line for each HER2 form, a weakness which they acknowledge. The fact that these lines mimic characteristics that were described in their previous work on tumors induced by wt HER2 vs. dex16 HER2 vs. p95 HER2 is, however, supportive of their conclusions. In addition, the specific connection between p95 and nuclear MRTFA is confirmed in HEK cells. The manuscript could be improved with some discussion on how p95 p-Y1139 is likely to signal (what binds to this site?)

Reviewer #3 (Public review):

Summary:

The study aims to explain why different HER2 isoforms produce distinct cancer-cell phenotypes, with a particular focus on the poorly understood mechanisms by which p95-HER2 promotes the invasive and metastatic behaviors of HER2-positive breast cancer. The authors propose that functional selectivity at the HER2 receptor level, particularly signaling through Y1139, preferentially engages an MRTFA-TGFB1I1 transcriptional/cytoskeletal program, thereby shifting p95-HER2-expressing cells toward individual motility, matrix degradation, invasion, and metastasis rather than simply proliferation.

Strengths:

A major strength is the multi-level experimental strategy. Rather than relying on a single invasion assay, the authors integrate phenotypic profiling, single-cell RNA-seq, imaging of MRTFA nuclear localization, protein validation, cytoskeletal analysis, matrix-degradation assays, transwell invasion, 3D spheroid/organoid assays, and orthotopic transplantation. This provides relatively coherent evidence that p95-HER2 cells occupy a distinct biological state characterized by individual motility and invasion, accompanied by a myogenic/contractile transcriptional program and increased MRTFA nuclear localization. Another strong aspect is the attempt to connect the downstream phenotype to a specific HER2 phosphorylation site. The authors identify Y1139 as a candidate determinant of the "go" phenotype and show that mutation of Y1139 to phenylalanine eliminates MRTFA translocation as well as the associated motility and invasion phenotypes

Weaknesses:

The main limitation is that the evidence for the proposed signaling mechanism is stronger for functional dependence than for direct molecular mechanism. The data establish that Y1139, MRTFA, and TGFB1I1 are required for the p95-HER2 invasive phenotype, but they do not appear to fully establish how phosphorylation at Y1139 mechanistically couples p95-HER2 to MRTFA activation. In particular, a direct biochemical link between Y1139 and the machinery controlling actin polymerization/MRTFA nuclear translocation would make the proposed pathway substantially more convincing.

A second limitation is the model system used in the study. Much of the mechanistic work relies on three particular engineered HER2 isoform-expressing cell lines, while the in vivo experiments use NOD/SCID mice. This is useful for testing tumor-cell intrinsic invasion and metastasis, but it limits conclusions about the generality of the functional regulation across different cell contexts, and also limits the understanding of how the pathway operates in the intact immune microenvironment. This is especially relevant because the manuscript itself places p95-HER2 in the broader context of immune evasion and notes that MRTFA can regulate PD-L1; however, the current models cannot adequately establish whether the proposed pathway contributes to immune-mediated metastatic progression in vivo.

Finally, the distinction between "go" and "grow" signaling is conceptually attractive, but the extent to which it represents a general property of HER2-positive human breast cancers remains less certain. The study provides strong experimental evidence for functional selectivity in the models examined, but additional validation in independent patient-derived models and clinically relevant tumors would be needed before generalizing this mechanism across HER2-positive disease.

Reviewer #1 (Public review):

Summary:

The manuscript aims to mechanistically characterize the final step of the Lol lipoprotein trafficking pathway in Gram-negative bacteria, where triacylated lipoproteins are transferred from the periplasmic chaperone protein LolA to the outer membrane receptor LolB for subsequent insertion into the inner leaflet of the outer membrane. The crystal structures of LolA R43L-LolB, LolB L114G-lipoprotein, LolB Q44A (open), and LolB Q44A (closed) are reported. A combination of structural analysis, growth assays, pull-downs, and fluorescence spectroscopy was employed to characterize the behaviour of wild-type and mutant forms of Lol proteins. The authors then used previously published molecular dynamics simulations to propose a six-step model for the transfer and insertion of lipoproteins from LolA to LolB, and finally into the membrane.

The structural work is mostly solid, but the mechanistic conclusions drawn from it rely too much on interpolation between static states and on functional assays with limited dynamic range.

Strengths:

(1) The structure-guided mutagenesis was thorough and well designed. Mutants were tested both in isolation and in combination with one another, utilizing both alanine and charge reversal mutants, to fully tease apart which residues were most important for the interaction.

(2) The dissection of the Hook interactions allowed the authors to identify the importance of the identity of the hook apex residue, as well as the critical hydrogen bond that maintained the Hook geometry.

(3) The structure of the lipoprotein-bound LolB is a valuable addition to the body of knowledge around this pathway, and the way that the complex was produced was very clever.

Weaknesses:

(1) There is no raw data shown for any of the interactions or lipoprotein transfer experiments. All of the quantitative claims about LolA association and lipoprotein transfer rely on SDS-PAGE densitometry, and not a single gel image from those experiments is provided in the main text or supplemental.

(2) Pull-down assays are a fairly poor assay to use for protein interactions when there are so many quantitative alternatives that would be amenable to this experimental system, especially when the association and transfer of lipoprotein are the central claims of the manuscript.

(3) The crystal structures are under-refined for their resolution, with large R-factor gaps and somewhat concerning statistics contained in the wwPDB reports (e.g. >20% RSRZ outliers in 2/4 structures). The data also seem to be cut at strangely high I/sigI for the two Q44A structures with no explanation.

(4) In the LolB L114G structure, the Hook residue L68 is the most poorly fit in the whole structure. The validation report indicates the entire hook is poorly resolved, but without access to the maps and models, we are unable to assess this independently.

(5) There is no rubric for what constitutes a significant result for the DAUDA fluorescence spectroscopy. What level of increase vs wild-type is considered significant? I188G is incredibly close to L144G, but one is treated as a non-result and the other as something that should be further investigated.

(6) The conformational trajectory of the proposed model is on static endpoints.

Reviewer #2 (Public review):

Summary:

The manuscript by Jepson and colleagues presents the structural and biochemical study of the outer membrane LolB in complex with LolA in its apo and lipoprotein-bound states. The Lol system (LolABCDE) is essential in the translocation of lipoproteins to the outer membrane. Several studies have looked at the inner membrane complex. This study captured the last step of the process, the translocation of lipoproteins to the outer membrane via the LolB and LolA proteins. The crystal structures of LolB in complex with LolA revealed key interactions via the hook and charged belt motifs. Mutagenesis was not able to fully dissociate the complex or prevent the transfer of lipoprotein. The complex of LolB with LolA and lipoprotein revealed the 'final' step of engaging the complex for lipoprotein insertion into the outer membrane.

Overall, this is an excellent and exciting study for the Lol field as it captures the events happening at the outer membrane.

Strengths:

The authors have determined several high-resolution structures and validated them with excellent mutagenesis experiments. The key structure is that of LolB with LolA and the lipoprotein (even partially resolved).

Weaknesses:

The only 'weakness' is that the authors have used mutants of either LolB or LolA to capture the stable states, and the discussion section would benefit from commenting if these states could be dead-end states, as validation of the mutants has been done with a WT background. In addition, can they discuss whether a2 is likely to undergo further conformational changes to open a lateral gate to the OM for the acyl chains of the lipoprotein to insert?

Reviewer #1 (Public review):

Summary

The authors set out to address a critical gap in olfactory bulb research: the lack of a comprehensive annotation of neuron types and their connectivity in the adult zebrafish. Using a large serial block-face scanning electron microscopy volume, they reconstructed 459 neurons and performed a detailed morphological and ultrastructural classification. They identified 13 major neuron subclasses, including two projection neuron types (mitral cells and ruffed cells) and 11 interneuron types. A key finding is that despite the diffuse appearance of the lateral olfactory bulb neuropil, it is organized into discrete, segregated glomeruli. The authors also performed targeted synapse annotation, revealing systematic and selective connectivity patterns between projection neurons and specific interneuron subnetworks, and discovered distinct microcircuit motifs involving reciprocal and unidirectional connectivity. This work provides a crucial anatomical framework for understanding the circuit basis of olfactory computations.

Strengths

(1) The volumetric EM reconstructions capture fine ultrastructural features (e.g., spine shapes, neurite caliber variations, and varicosity morphology) at nanometer resolution across hundreds of neurons. The descriptive detail and 3D rendering provided for each subclass are remarkably thorough.

(2) The study goes beyond simple morphological descriptions by integrating ultrastructural features, such as spine shape, neurite diameter, and synaptic arrangements, to define neuron classes. This provides a more functionally relevant taxonomy. The classification was also cross-validated by independent neuroscientists.

(3) The targeted synapse annotation translates anatomy into testable circuit hypotheses, revealing distinct connectivity motifs (reciprocal vs. unidirectional, MC-selective vs. RC-selective) that provide structural substrates for computational functions such as normalization.

Weaknesses

(1) It is important to provide an estimate of the total neuronal population of the adult zebrafish OB and justify that their sample size is sufficient to claim a "comprehensive" neuron type classification.

(2) The classification was based on morphology and ultrastructure. If feasible, we recommend an unsupervised clustering analysis using only the synaptic connectivity matrix for fully reconstructed neurons to test whether connectivity patterns recapitulate the 13 morphological subclasses. Besides, cross-validation by independent annotators is great. However, when the annotators disagreed on a neuron's subclass, how was it resolved?

(3) A summarized reference table that includes features for each subclass is highly recommended (the main text cited several tables; however, I did not find them). This would facilitate community adoption of the classification.

(4) The schematic in Figure 20 could be optimized to reflect synaptic connections among 13 subclasses with connection strength annotated.

Reviewer #2 (Public review):

Summary:

The authors have used a previously published SBEM dataset of the adult zebrafish olfactory bulb and carefully reconstructed the fine structure of a large number (459) of neurons. Through manual classification of cells into 13 different anatomical classes, they provide a carefully annotated library of cells. Classifications have been validated by comparing the labeling of 5 human annotators, which revealed a large degree of reproducibility. Synaptic identification enables assignment of connectivity, revealing a major contribution of reciprocal connections (two-way synapses) that may support key computational features during olfactory processing. In general, the work provides a clean descriptive account of the organization of the zebrafish olfactory bulb, which should help to better understand principles of olfactory processing in vertebrates.

Strengths:

The authors provide a library of neurons in the olfactory bulb of adult zebrafish that will be highly useful to the broader circuit neuroscience community. Figures of classified cell types have been convincingly arranged to appreciate the high quality of the EM volume and reconstructions. Cells are easily accessible through a web interface to facilitate the assessment of classifications. An important result is that anatomical structures in zebrafish clearly follow the glomerular arrangement known in mammals, which argues against a microglomerulus model previously proposed for the zebrafish olfactory bulb.

Weaknesses:

The paper uses Wanner et al. 2016 as a main base dataset to generate a descriptive library of cells, in a limited volume (18%) of the zebrafish olfactory bulb. Annotations of cell types and synapses are largely human-based. Using a multi-human validation strategy helps to assess labeling, but it would benefit from more rigorous statistics, in particular for the synapse annotations. In general, the paper would be easier to read with a glossary of cell types. Some features are quite intriguing, like the cilia structures in some of the somata or the ruffs, but discussion of the potential computational roles of these fine structures is missing.

Reviewer #3 (Public review):

Summary:

The authors aimed to build a comprehensive understanding of the cell classes that make up the adult zebrafish olfactory bulb (OB) and their connectivity. Using an EM volume covering the three OB layers and several glomeruli, they aim to reconstruct a sufficient sample of all cell classes in the OB. Based on 459 proofread reconstructions, they identify 13 cell classes and a small number of subclasses and offer detailed expert neuroanatomical arguments, largely qualitative, for why these classes are distinct. While they do not have a synapse classification, they further build a manual wiring diagram of connectivity based on these reconstructions, which together offer a strong baseline understanding of the organization of the OB and useful properties to build into a future quantitative assessment of fish neuroanatomy in general.

Strengths:

The paper offers a powerful example of why expert assessment remains a vital tool in neuroanatomy. While quantification is deeply important to scale data, the kind of careful multimodal human investigation of neuronal shape, connectivity, and ultrastructure forms a key basis for discovering and differentiating neuronal cell types. The authors use multi-expert classification to validate qualitative assessments as well, which is a good way to handle cross-individual uncertainty. The descriptions of each type and their relationships are generally quite thorough, and there are enough cells to offer a satisfying baseline categorization. Anyone studying zebrafish OB is likely to get a lot out of this paper.

Weaknesses:

While I am a strong believer in expert neuroanatomical intuition alongside quantitative neuroanatomical approaches to cell typing (especially given the manageable number of cells and cell types), I would have liked a section explaining why this approach was taken as opposed to a more quantitative one from the beginning. I think that this is a case study in where the expert qualitative approaches are useful, especially with fascinating details like the unusual hand-shaped spines. A lot of the strength of the paper comes from this philosophy, in my view, and it would be nice to hear it elucidated by the authors. I would also be curious if the authors came to any conclusions about what might have been possible with a more data-driven approach, although this may be beyond the scope of the work or not terribly interesting.

There were some aspects of the cell typing that I didn't totally understand. Two of the most common issues with cell typing are that (1) it is unclear if cell types are discrete and (2) unusual cells appear that don't have enough peers to classify into an unambiguous type. Did those happen here? For example, it was not obvious to me how it was decided that mitral cells fell into three subclasses: small, medium, and large, and not a continuum from small to large. Similarly, some interneuron classes had subclasses. What anatomical qualities suggested that a split was a subclass-level split vs a class-level split? Were there any cells that were impossible to classify into a group, or was everything actually tidy? How confident are the authors that they have captured the complete diversity of the OB from this sampling?

I generally appreciated the numerous detailed figures and found them generally clear and informative. However, as someone who is not deeply familiar with zebrafish olfactory bulb organization, I would have liked to have at least some global context for each cell class in terms of how typical cells relate to the layers and glomeruli. Figure 18B does a nice job of this with many cells, but it would be clearer to see this along the way with the individual classes (and subclasses) as well.

For the connectivity analysis, I commend the authors for doing as much manual synapse annotation as they did, but the lack of completeness makes it less clear how to interpret the connectivity findings. In particular, how complete is the current assessment? For example, when describing connectivity between pairs of cells, how were the pairs selected? Did they necessarily have contact between the meshes, or innervate the same glomerulus, or were they just two cells of the same type in any location? This is useful to understand how to interpret the connectivity fractions measured.

Reviewer #1 (Public review):

Summary:

The authors sequenced 888 individuals from the 1000 Genomes Project using the Oxford Nanopore long-read sequencing method to achieve highly sensitive, genome-wide detection of structural variants (SVs) at the population level. They conducted solid benchmarking of SV calling and systematically characterized the identified SVs. While short-read sequencing methods, including those used in the 1000 Genomes Project, have been widely applied, they exhibit high accuracy in detecting single nucleotide variants (SNVs) and small insertions and deletions but have limited sensitivity for SV detection. This study significantly enhances SV detection capabilities, establishing it as a valuable resource for human genetic research. Furthermore, the authors constructed an SV imputation panel using the generated data and imputed SVs in 488,130 individuals from the UK Biobank. They then conducted a proof-of-principle genome-wide association study (GWAS) analysis based on the imputed SVs and selected traits within the UK Biobank. Their findings demonstrate that incorporating SV-GWAS analysis provides additional insights beyond conventional GWAS frameworks focusing on SNVs, particularly in improving fine-mapping.

Strengths:

The authors constructed a high-sensitivity reference panel of genome-wide SVs at the population level, addressing a critical gap in the field of human genetics. This resource is expected to significantly advance research in human genetics. They demonstrated the imputation of SVs in individuals from the UK Biobank using this panel and conducted a proof-of-concept SV-based GWAS. Their findings highlight a novel and effective strategy for integrating SVs into GWAS, which will facilitate the analysis of human genetic data from the UK Biobank and other datasets. Their conclusions are supported by comprehensive analyses.

Weaknesses:

The authors have addressed many of my previous comments, and I appreciate their efforts. However, I still have two related concerns.

(1) Shortly after reviewing this manuscript last year, my laboratory obtained access to the UK Biobank (UKB) Tier 3 dataset for an unrelated project. In August 2025, I searched the UKB Research Analysis Platform (UKB-RAP) for the imputed structural variant (SV) dataset described in this manuscript but was unable to locate it. After contacting UKB, I was informed that they were developing the system for releasing the data. To the best of my knowledge, the dataset remains unavailable. A major contribution of this work is the generation of an imputed SV resource for approximately 500,000 UKB participants with extensive phenotypic information. If this resource is not accessible to the research community, even to authorized UKB users, the practical impact and utility of the study are substantially diminished.

(2) Given that the imputed SV dataset is currently unavailable, it becomes even more important for the authors to provide a detailed, ready-to-run SV imputation pipeline for UKB-RAP, even if the "data processing simply consisted of running standard bioinformatics tools with the parameters exactly as described in the manuscript". In particular, the pipeline should include practical information such as computational requirements (e.g., memory and storage), expected running time, and estimated cost. Anyone with experience using UKB-RAP will agree that reproducing large-scale analyses on the platform can be both technically complex and financially expensive. Such pipeline would greatly improve the reproducibility and accessibility of this work.

Because my initial assessment of the manuscript was generally positive, I do not wish to change my overall evaluation, summary, or assessment of its strengths. However, I would view the work even more favorably if either (i) the imputed SV dataset became publicly available to authorized UKB users, or (ii) the authors extended their SV-GWAS analyses to the full range of UKB phenotypes and released the resulting summary statistics, analogous to the Pan UKBB ("https://pan.ukbb.broadinstitute.org/") resource. Although this would require considerable additional effort and computational resources, it would substantially enhance the long-term value and impact of the study.

Finally, I would like to emphasize that these comments are not intended to create unnecessary difficulties for the authors or the editors. Rather, I believe this highlights a broader issue in the use of this kind of large public datasets: reviewers cannot independently verify key results, and readers cannot readily build upon the work if the underlying resources are inaccessible, even after obtaining authorized access to the original dataset. I hope the authors, together with the eLife editors and UK Biobank where appropriate, can help facilitate the timely release of this valuable resource.

Reviewer #1 (Public review):

Summary:

The authors sought to develop a generalisable method for clearing and labelling large, highly complex human organs and tissues while preserving their three-dimensional structure, enabling visualisation at high resolution. The resulting body of work is substantial, encompassing a diverse range of tissues and labelling approaches, and demonstrates an impressive breadth of experimental work.

Strengths:

A major strength of the method is its systematic validation across a diverse range of human tissues, including several understudied tissues, in combination with labelling using multiple antibodies and fluorescent dyes. An additional strength is that the skin was retained on the finger throughout the clearing, labelling, and imaging procedures, allowing the tissue to be analysed in an intact state. This is particularly relevant to sensory neuroscience, where preserving the cutaneous sensory structures may be important. The use of tissue from donors aged 85 years and older is also a strength, as it provides a stringent test of the method in tissues with substantial age-related pigment accumulation and autofluorescence.

Weaknesses:

A major limitation is the lack of evidence that the method can be applied to formalin-fixed, paraffin-embedded (FFPE) tissue, which represents a substantial proportion of archived human tissue and is an important potential application of tissue-clearing approaches. Furthermore, the claim that the method is cost-effective is not sufficiently supported by the data presented, as there is no comparison of the costs of the proposed approach with those of existing methods.

Appraisal:

(1) Overall, I feel that the authors have broadly achieved their stated aims, although several aspects would benefit from further qualification or supporting evidence.

(2) Although magnetic resonance imaging (MRI) was performed on four tissues, the MRI data do not appear to be fully integrated with the microscopy data, and I did not find evidence of precise spatial correlation of structural information between the two modalities.

(3) The term "photoclearing" may be useful to describe the approach, but the underlying concept does not appear to be entirely novel, given that increased optical transparency can occur as a consequence of photobleaching. Furthermore, the evidence presented is not sufficient, in my view, to establish that the observed effects are attributable to photodegradation of pigments.

(4) I question whether the broader claims regarding the scalability of existing tissue-clearing approaches are sufficiently supported. In particular, the claim that aqueous- and hydrogel-based methods have limited scalability may require further evidence, given that perfusion-based approaches can facilitate processing of large tissue volumes. Similarly, the assertion that organic solvent-based approaches are better suited to large tissue volumes does not appear to be fully supported by the data presented, unless this refers specifically to the beneficial effects of tissue shrinkage during dehydration on imaging.

(5) More generally, a direct comparison of the clearing, labelling, and imaging outcomes for the thyroid samples fixed for 48 hours versus 8 years would be valuable for assessing the extent to which fixation duration influences the performance of the method.

Discussion:

The results have potential value for medical education, comparative anatomy, and research requiring the characterisation of human anatomy across spatial scales. The method could have applications across multiple disciplines and may be of interest to researchers who have access to human tissue from sources including archival material, biopsies, surgical resections, tissue banks, autopsy specimens, and donor tissue. However, the length and complexity of the protocol may limit its practical adoption, particularly in settings where throughput is an important consideration.

Context:

The results for the inner ear and ring finger are particularly notable because they demonstrate the ability of the method to clear, label, and image anatomically complex, highly calcified, and densely structured tissues. The successful clearing and immunolabelling of tissue fixed for 15 years is also significant, given the potential challenges associated with antibody penetration and preservation of epitopes following prolonged fixation. More broadly, successful imaging of large organs, such as the heart, is technically important given the challenges that their size and geometry present for light-sheet microscopy, and demonstrates the feasibility of applying the approach to intact human organs.

Technical concern:

The manuscript should acknowledge the hazardous and potentially toxic nature of several reagents used in the protocol, together with the challenges of safely handling and disposing of the relatively large volumes of chemical waste generated. These factors may have practical implications for implementation of the method.

Conceptual concerns:

The role and added value of photogrammetry in the study are not entirely clear. It would be helpful for the authors to clarify how photogrammetry was used beyond its presentation in the figures and what additional information or utility it provides relative to the other imaging modalities employed.

Reviewer #2 (Public review):

Summary:

The authors claim to present a generally applicable tissue processing pipeline for optical tissue clearing and labelling of whole human organs. To achieve this, they combine multiple previously published tissue pretreatment steps to enhance tissue permeabilization, and hence label penetration, as well as optical transparency. Most notably, the authors include LED photobleaching, or photoclearing as they call it here, to reduce both autofluorescent background and increase tissue transparency. Extending their previously published method towards a wide variety of large human samples encompassing multiple organs and tissues, they combine their clearing protocol with several small molecule dyes for general overview stains as well as immunofluorescence labelling. For some of the processed samples, they introduce a novel multimodal imaging pipeline, which includes photogrammetry, MRI, and 3D histology/microscopy. Both the multimodal pipeline and the described tissue clearing and labelling protocols seem highly useful for both pathology as well as fundamental research, especially since the modularity of tissue-clearing protocols allows for the integration of individual steps, such as photoclearing, into already established methods.

Strengths:

The transparency of the large, archival human samples is very impressive and, to my knowledge, constitutes the best human tissue clearing to date. The positive effects of photobleaching/clearing on autofluorescence reduction as well as transparency are obvious and highly interesting for the field, as this method can easily be implemented into any tissue clearing approach. It is highly appreciated that notes are provided in both the methods and results, on which steps of the pipeline had to be adjusted for different tissue types and sizes. However, this could be presented in a more structured way, by showing a graphical representation or at least a table of all used tissues and all the steps of the pipeline, with indications for which steps were skipped/adjusted per tissue type as well as the incubation times for every step. Table 2 already does that for light exposure and decalcification times, but it would be beneficial to do this for all steps to get a better overview. This would also aid replication and implementation of their method, as other groups can just look up the closest matching tissue type in the table and go from there rather than piecing this information together with the in-text notes.

The combination of different small molecule dyes, while in itself not new, has the potential to provide 3D alternatives for conventional histological overview stains. Other groups have already shown 3D H&E and Nissl stains, and it would be interesting what other routinely used stains could be mimicked in that way. Here, the combination of Eosin Y and Fast Green seems especially interesting.

The quality of the immunofluorescence labelling is extremely good, especially given the age of some of the samples (e.g. very high-quality SMA staining in 15-year-old material). It would be interesting to know if the authors encountered antibodies that were not compatible with their archival tissue and, if so, which ones.

Weaknesses:

It is not exactly clear to me how some of the image volumes were acquired. For example, the panels 1f-h: The regions these volumes originate from are indicated in the cleared heart (1e), but were these regions blocked out or was the entire heart somehow immersed in the imaging chamber? The latter would be, to the best of my knowledge, a novel scale of light sheet imaging and should be mentioned. If the regions were somehow blocked out, this should be described. In general, a lot of the imaging volumes seemed to have been acquired mainly towards the periphery of the sample, especially the higher magnification inserts. This immediately raises the question of whether this is because imaging was not possible in the centre and why (limitation of the imaging chamber size, light penetration, label penetration, or a combination of those?). While it is understandable that it is currently not possible to image, e.g. entire kidneys in light-sheet set-ups, the limits of feasible sample sizes should be mentioned somewhere to give the reader a better appreciation of the usefulness of the method. The authors demonstrate that it is possible to clear whole organs, but is it also possible to label whole organs? Is that possible with antibodies as well, or only with small-molecule dyes? And how deep can samples be imaged in the best-case scenario? Right now, the paper uses a common scheme in the tissue clearing literature, of showing a very large cleared sample and then imaging a smaller sub-volume while implying that, in principle, it is possible to do that for the entire organ, if the objective working distance and imaging chamber would allow. However, to scale their method up, the authors would first need to show that their labels would penetrate throughout the entire sample, and this is currently not done in an unambiguous manner.

The biggest limitation in the manuscript for me is the authors' claim: "We demonstrate the voxel-by-voxel alignment between MRI and microscopy datasets, merging cellular-scale detail and protein-specific labelling with macroscopic medical imaging." While technically true, the authors only show actual overlays of microscopy and MRI data for the inner ear sample in Figure 5. No overlays between MRI and the microscopy volumes are shown for the heart and finger data, which is a missed opportunity. Similarly, no overlays between all three acquired modalities (photogrammetry, MRI, 3D microscopy) are shown for any of the samples. This is a pity, as the authors put a lot of emphasis on this novel multimodal pipeline, even though side-by-side demonstrations of MRI and 3D microscopy in a similar manner have been shown already by other groups and a truly multimodal dataset with the different modalities, e.g. as different channels in a volume, would be very exciting.

Reviewer #1 (Public review):

Summary:

The authors introduce a foraging task designed to test whether movement and harvesting durations are governed by a common objective combining reward, effort, and time or by separate objectives. Participants generate wrist torque to move a cursor to a rewarding patch and then hold it within the patch while reward accumulates at a diminishing rate. The task allows reach and harvest effort, as well as delays before or between the two phases, to be manipulated separately.

Reach effort strongly affected reach duration but had little effect on harvest duration, whereas a delay imposed after reaching altered harvest duration without affecting the completed reach. Manipulation of harvest effort did not affect either duration. The authors also found consistency across individuals within each type of duration but no reliable association between reach and harvest durations. The separate model fit the behavioral results better than the common-utility model, particularly for the post-reach delay condition. The findings therefore support partly dissociable, but coupled optimization of harvest and movement in this task, while not excluding common optimization in other contexts.

Strengths:

The task provides a controlled way to manipulate the effort and delays associated with reaching and harvesting while measuring both durations. The control experiments alter the timing of delays, allowing the experimenter to test whether their effects depend on when they occur. These manipulations produce clear behavioral dissociations, particularly the selective effect of a post-reach delay on harvest duration. The delay and effort manipulations were verified, and most participants reported noticing the changes in effort and time.

The separate model extends a previous cost-of-time framework to a foraging task involving movement and reward harvesting. Its assumptions have a clear relationship to the experimental design: applied delays and actions enter the time offset for each separate cost function. This provides an intuitive account of why a delay occurring after the reach can affect harvesting without affecting the completed movement. The separate model also provides a better fit than the common utility model in both modeled experiments.

The individual-level behavioral analyses are also useful. Reach durations are correlated across reach conditions, and harvest durations are correlated across harvest conditions, indicating stable individual differences within each measure. The lack of a corresponding relationship between reach and harvest durations raises an important question about whether the two forms of vigor necessarily reflect a common individual tendency in their task.

Weaknesses:

The block design and familiarization made the changes in effort and delay learnable, and most participants reported awareness of these manipulations. The observed dissociation may reflect adaptation to this task structure. This does not weaken the behavioral result, but it limits how directly the findings can be generalized to tasks in which changes in reward, effort, or time are less explicit.

The main modeling limitation is that both models are fitted to ten group-level median observations in each experiment, rather than to individuals' behavior. The separate model has seven fitted parameters compared with three in the common model. Although the reported AIC penalizes additional parameters, the analysis includes no participant-level fitting, resampling, or cross-validation.

Reviewer #2 (Public review):

Summary:

The manuscript by Conessa and colleagues tests whether movement time and decision time are co-regulated in humans using a novel harvesting paradigm and computational modelling of the behavioral data. The data contradict the common utility model and are better supported by an alternative model where the costs of the harvesting and movement phases are optimized separately - rather than jointly.

Strengths:

The major strength of the study is the development of the novel paradigm where the forces applied during the distinct phases of the task (reaching and harvesting) can be perfectly controlled and measured.

This elegant manipulation allows precise estimation of subjective effort costs, and later can be used to better understand how these costs shape behavior.

The novel separate-cost model is also an interesting development.

Weaknesses:

In my opinion, the manuscript suffers from too many weaknesses to solidly support the general conclusions:

(1) The manuscript oversells the significance of the results;

(2) The novel model appears to be built ad hoc based on the results, and its ingredients are poorly motivated

(3) There are important methodological shortcomings. I develop all of these points below. The paper also lacks clarity, particularly when presenting the key concepts of the novel model.

Importantly, I also recommend that the authors follow open science policy and publish their data and code on a public repository.

(1) Significance

As mentioned above, the question of the co-regulation of the vigor of different behaviors is a very interesting topic, but I believe the significance of the results is overstated, for different reasons:

a) This is a harvesting task, which means that decisions are specifically about the duration of the reward period. In other words, more time directly provides more (immediate) reward, whereas in other types of tasks more time allows gathering and processing more information, and so reaching a more informed decision, which is a more indirect way of securing better outcomes - see e.g. Drugowitsch, Moreno-Bote et al. JNeurosci 2012 for optimizing decision time in this most common setting. This major difference is not discussed in the manuscript, and it is not clear whether the results found here would hold for the most common type of decisions. This is also different from the time to initiate a response, as explored by Morvan et al. in their cited preprint.

b) The title of the study mentions "flexible" co-regulation, but it's hard to see direct evidence for flexibility in the results. This flexibility is extrapolated in the Discussion, but in order to make such a statement, it is necessary to show that the co-regulation differs between different conditions, as predicted from the model.

c) The lack of impact of the harvesting effort on harvest time is very surprising, and the null value of the sensitivity to effort (d=0) in the model is intellectually uncomfortable. As some value of effort, it SHOULD matter. I am not very satisfied with the explanation provided in the Discussion ("possible hypothesis would be that the perceived effort of harvest was outweighed by the reward value"): this trade-off should be precisely dealt with by the model.

(2) Separate-Cost Model

The separate-cost model is poorly motivated and not clearly presented, in particular the indirect coupling between the two phases.

a) Some key components are motivated by the constraint to obtain a well-behaved optimization problem ('we need to find a cost that increases with time' ) rather than based on cognitive principles. Some other key components seem to have been placed there just to fit the data better, but their plausibility is not even discussed. Notably:<br /> i- The fact that the offset o_h includes the reaching delay is key to the coupling between t_h and t_r, but is not properly motivated. Why should this offset include all preceding intervals within the trial (leading to double counting of the cost of the reaching phase, in both G_r and G_d)? In this cyclic paradigm where participants aim at maximizing cumulative reward, it is not even clear that they parse behavior into different trials starting as they need to reach for a new target.<br /> ii- Why is there a free parameter d for the cost of harvest effort? If time and motor costs are converted into the same currency, and that conversion holds for all types of phases, then I believe we should have d=1. This free parameter seems to have been introduced a posteriori to accommodate the lack of effect of harvest effort on harvest time, but is not motivated.

b) The two models differ in two fundamental aspects: whether optimization is run separately on the different phases of the trial, or jointly; and how the different costs and rewards are assembled in the to-be-optimized metric. Which aspect is more important to capture behavior is not assessed.

c) The presentation of the "Separate-cost model" is very obscure. For example, the notion of a `cost of reward` is rather odd, and consequently the formula for R(t_h) appears arbitrary. I understand that this is the difference between the maximum reward that can be collected within a trial and the actual reward that is collected, but it is not even presented as such. The effort cost of harvesting seems to be discarded in the model presentation but does appear in equation 14.

(3) Methodological Weaknesses

I believe that there are numerous methodological weaknesses that really limit the inferences that can be drawn from the experimental and modelling results.

a) Sample sizes are small, especially for control experiments, and it is not clear how they have been selected. This is especially problematic because the only manipulation that shows no significant effect on behavior (harvest duration) is only tested with the smallest sample size (n=10), severely limiting the significance of this negative result.

b) The rationale for the different manipulations in different experiments is not clear. Notably, why were modulations of the harvest effort not included in the main experiment? Because not a single participant sees a modulation of both reach and harvest efforts, we cannot evaluate whether the model can capture both effects at the same time (i.e., with the same parameter values). The number of data points used for fitting directly depends on the number of conditions performed by the participant, and is very low compared to the number of parameters to be fitted. For example, the model used very different shapes for the costs of time (Figure 8) for different sets of conditions, but this cost of time should be fixed, so it's not clear whether a single cost of time function can accommodate all manipulations.

c) Best practice in behavioral modelling is to model single-subject data, as fitting the model jointly to all subjects completely obliterates subject heterogeneity. Fitting on single-subject data would also allow investigating whether the models reproduce the consistency of harvest durations and reach durations across conditions, and most importantly the absence of correlation across subjects between reach and harvest duration. At the moment, we're left wondering what the implications of these analyses for the original question are.

d) What is the rationale for comparing reach duration in one condition to harvest duration in another condition?

e) The study does not take into account the likely impact of delays on muscular fatigue and perceived effort: the effort associated with maintaining the cursor on target during harvest might be perceived as less effortful after a longer break, partially explaining the impact of delay on harvest duration.

f) I am confused by the reference to indirect effects in the separate-cost model (L692-7, Table 2). To fit this model, was the experimental value or theoretical value of the reach duration t_d used to determine the optimal harvest value (from Equation 14)? In both cases, the reach duration DOES increase with reach effort, so we should expect an impact on the harvest duration. This is indeed seen in Figure 7A. So why does Table 2 suggest that reach effort should not impact harvest duration in this model?

g) Could the effect of pre-reach delay on reach time be alternatively explained simply by longer reaction times as reported in Figure S2, e.g., a longer period to reactivate the muscles after a longer break? In other words, as far as I understood, here reaction times are subsumed in the motor time, against the tradition of splitting them into two independent (non-overlapping) behavioral metrics.

h) The EMG-time integral in the harvest phase should be largely proportional to the harvest duration, given that the applied torque is constrained to remain on target. Is this the case? If so, then how to interpret the significant effect of reach effort on this value, given that the main analyses did not identify an effect of reach effort on harvest duration?

Reviewer #1 (Public review):

The study by Luden et al. seeks to elucidate the molecular functions of AHL15, a member of the AT-HOOK MOTIF NUCLEAR LOCALIZED (AHL) protein family, whose overexpression has been shown to extend plant longevity in Arabidopsis. To address this question, the authors conducted genome-wide ChIP-sequencing analyses to identify AHL15 binding sites. They further integrated these data with RNA-sequencing and ATAC-sequencing analyses to compare directly bound AHL15 targets with genes exhibiting altered expression and chromatin accessibility upon ectopic AHL15 overexpression.

The analyses indicate that AHL15 preferentially associates with regions near transcription start sites (TSS) and transcription end sites (TES). Notably, no clear consensus DNA-binding motif was identified, suggesting that AHL15 binding may be mediated through interactions with other regulatory factors rather than through direct sequence recognition. The authors further show that AHL15 predominantly represses its direct target genes; however, this repression appears to be largely independent of detectable changes in chromatin accessibility.

In addition to the AHL protein family, the globular H1 domain-containing high-mobility group A (GH1-HMGA) protein family also harbors AT-hook DNA-binding domains. Recent studies have shown that GH1-HMGA proteins repress FLC, a key regulator of flowering time, by interfering with gene-loop formation. The observed enrichment of AHL15 at both TSS and TES regions, therefore, raises the intriguing possibility that AHL15 may also participate in regulating gene-loop architecture. Consistent with this idea, the authors report that several direct AHL15 target genes are known to form gene loops.

Overall, the conclusions of this study are well supported by the presented data and provide new mechanistic insights into how AHL family proteins may regulate gene expression.

Reviewer #2 (Public review):

Summary:

The manuscript by Luden et al. investigates the molecular function and DNA-binding modes of AHL15, a transcription factor with pleiotropic effects on plant development. The results contribute to our understanding of AHL15 function in development specifically and transcriptional regulation in plants more broadly.

Strengths:

The authors developed a set of genetic tools for high-resolution profiling of AHL15 DNA binding and provide exploratory analyses of chromatin accessibility changes upon AHL15 overexpression. The generated data (CHiP-Seq, ATAC-Seq and RNA-Seq is a valuable resource for further studies. The data suggest that AHL15 does not operate as a pioneer TF, but is likely involved in gene looping.

Weaknesses:

The authors have extended the motif analysis to the top 1,000 shared peaks, addressing part of my previous concern. It did not erase my worries about overclaiming completely, but I think it can be considered as a terminological issue, rather than technical one. Therefore, it could be addressed through minor revision, without further analyses.

Specifically, the absence of a predominant enriched motif does not establish that AHL15 binds non-specifically or lacks sequence preferences. Low motif prevalence limits the proportion of peaks it could explain but does not exclude a genuine binding preference; a binding motif need not be rare in the genomic background (binding can be supported by other factors). Please revise the interpretation in lines 188-194, and the conclusion in lines 204-206 to state that the present analysis did not identify a strongly enriched motif accounting for a substantial proportion of AHL15-associated regions. Any equivalent claims of non-specific binding elsewhere in the manuscript should be toned down.

Additional minor points:

(1) Please provide the exact HOMER background-selection procedure, and definition of the 50-bp search windows.

(2) Figures 2B-C and the related supplementary figures, add the x axis label.

Reviewer #3 (Public review):

Summary:

This study investigated the role of AHL15 in regulation of gene expression using AHL15 overexpression lines. Their results do show that more gene are downregulated when AHL15 is upregulated and its binding is not affecting the chromatin accessibility. Further, they investigated AHL15 binds in regions depleted in histone modifications and other epigenetic signatures. Subsequently, they investigated the presence of AHL15 in the gene chromatin loops. They found overlaps with both upregulated and downregulated genes. The methods are appropriately described, but could be improved to include the analysis of self-looping gene boundaries.

Strengths:

Their study clearly showed lack of any specific sequence enrichment in the AHL15 binding sites, other than these being AT-rich, suggesting that AHL proteins do not recognize a specific DNA sequence but are recruited to their AT-rich target sites in another way. The study does suggest significant enrichment of AHL15 binding sites at TSS and TES, and AHL15 sites are depleted of any histone marks. They also identified that AHL15 binding sites overlap with self-looping gene boundaries. The authors have addressed the comments raised in the revised manuscript.

Comments on revised version.

The authors have addressed the comments raised, in the revised manuscript.

Reviewer #1 (Public review):

Summary:

The manuscript by Sustar et al. takes a methodical approach to document the types of glutamate receptor subunits that reside in Drosophila muscles, examining developmental stages spanning from larvae to adults. Prior work thoroughly documented the subunits operating in Drosophila larval body wall muscles. Most subsequent research focused on the glutamate receptor heterotetramers found in the body wall, composed of GluRIIA/C/D/E or GluRIIB/C/D/E subunits, along with auxiliary subunits like isoforms of Neto.

For the current work, the authors report that the larval muscle glutamate receptor composition is not universal for all Drosophila muscles. They examine the following muscle systems: larval body wall, adult abdomen, adult leg coxa, and adult indirect flight. They also briefly examine adult muscle structures associated with the proboscis, neck, and haltere. The authors find that the receptor subunits in the adult abdomen (mostly) match those in the larval body wall. This makes sense given that the adult abdominal muscles are derived from the larval body wall. Yet not much else matches the larval body wall. For example, all (or most) of the GluRII-type subunits are missing from the adult indirect flight muscles. Leg muscles have GluRII-type subunits, but they do not have all of them expressed prominently, and they are missing GluRIIB. Additionally, leg muscles express a glutamate-gated chloride channel, which could be a source of inhibitory glutamatergic transmission. Interestingly, when it comes to non-abdominal adult muscles, one general theme seems to be an active promoter (GAL4 driver) for the kainate-type glutamate receptor called Clumsy. The authors propose that Clumsy could be key to understanding how functional GluR complexes are assembled in adult insects.

Strengths:

(1) Documenting the types of glutamate receptors that operate in diverse insect muscle systems is important because it uncovers fundamental information.

(2) Much of the prior research focus has been on how the body wall muscle tetramers assemble and operate. It is a strength to demonstrate the other receptor solutions used by adult NMJs.

(3) The work uses GAL4 drivers and immunohistochemistry (when possible) in combination to draw conclusions.

(4) The muscle anatomical analyses are high quality. This allows the research group to reach refined conclusions.

(5) The confocal-level images of synaptic active zones and their apposed glutamate receptor clusters are high quality.

(6) The dataset adds specific detail and complementary context to bulk transcript analysis collections like FlyAtlas.

(7) The Discussion section is insightful. It is not simply a recap of the data in the paper. It is also an integration of that data with prior work, and it suggests direct experiments that the research team or others could attempt downstream.

Weaknesses:

One can draw expression-level conclusions from these data. But genetic tests (e.g., would clumsy losses of function impair leg muscles?) could help the authors and the field draw stronger conclusions about the roles of some of these glutamate receptor gene products.

Overall Assessment and Discussion:

The data in this study are high quality, and the results support the main conclusion: adult muscle glutamate receptor clusters do not recapitulate the "canonical" larval body wall clusters.

Comments on revised version.

The authors have done a serious revision of their manuscript, and they have directly addressed context/interpretation-level concerns that I had from the first round.

Reviewer #3 (Public review):

The Sustar et al. manuscript catalogs glutamate receptor composition across distinct Drosophila NMJs: larval and adult abdominal NMJs, as well as NMJs on adult leg and flight muscles. This work is important and probably overdue. The larval NMJ is the exemplar NMJ in this system, and the identity of "essential" and "alternative" subunits at this stage is assumed by many to hold across developmental stages and NMJ types. Here, the authors show that there is surprising diversification among NMJ types and that the notion of essential/alternative subunits only holds true at larval NMJs.

The study will generate interest in the Clumsy GluR subunit, which has not been well-characterized at all, but is widely expressed at adult NMJs. They also find striking extrasynaptic expression of glutamate-gated chloride channel GluRClalpha in adult leg and flight muscles, raising questions about its role. The study is interesting, logical, and well-written. The figures are clear, and the discussion was particularly thoughtful.

Comments on revised version.

This is a nicely revised manuscript that went a long way toward addressing the concerns in the first round. The response to the reviewers is clear and thoughtful. I think the question is framed very well and the authors do a much better job throughout making clear where the heterologous approaches agree and differ. I also think the Limitations section in the Discussion does a great job putting the work in context.

Reviewer #1 (Public review):

Summary:

The authors address the lack of validated tools for detection and quantification of proteins associated with amyotrophic lateral sclerosis (ALS) through an extensive screening of 303 commercially available antibodies to 33 protein targets. Their ALS-Reproducible Antibody Platform (ALS-RAP) delivers a validated antibody toolbox for ALS research, which will provide an advantageous starting point for researchers in this field. Ayoubi R. et al. showcase the characterization workflow, presenting as an example the characterization of antibodies targeting Galectin-1, encoded by the LGALS1 gene. A selection of these antibodies was also used to profile protein levels across human induced pluripotent stem cell (iPSC)-derived and primary neurological cell types, and the fundings support that ALS disease mechanism involves both neuronal and glial cells.

Strengths:

The knockout (KO)-based approach is definitively the major strength of this study, providing a high level of confidence in the data collected in human induced pluripotent stem cell (iPSC)-derived and primary neurological cell types. Important is also the focus on renewable reagents (monoclonal and recombinant antibodies). The extensive characterization of this set of antibodies will benefit any scientist interested in any of the 33 targets proteins, even in fields other than neuroscience.

The authors perform an interesting protein profiling study assessing 27 proteins, comparing RNA and protein expression data, and using two independent WB preparations of the same cell types. The conclusions that can be drawn from this first assessment might not be final, but data are compelling, because they have been collected with reliable and validated antibodies.

Another strength of this work is the data dissemination strategy, which includes the Only Good Antibodies (OGA) platform, where YCharOS data are curated and presented in an easy and intuitive manner that facilitates antibody selection by the end user for WB, IP and IF applications.

The authors mentioned the development of single-chain variable fragment (scFv) recombinant antibodies raised by the SGC against the six proteins (ANXA11, OPTN, MATR3, PFN1, UBQLN2 and VCP) that had limited renewable antibodies that are commercially available. The development was optimized to generate antibodies particularly suitable for IP, and the clone selection process was carried out using IP coupled to mass spectrometry. They provide details on the screening process for the SGC-generated scFv antibodies, plus share that scFv sequences.

Weakness:

The protein profiling study is limited to WB data, and the authors did not provide any explanation on why there was no integration with IP and IF data, not even for those targets that have validated antibodies. Also, not all the cell types have been screened by chemiluminescence-based detection and by fluorescence-based WB, and the authors do not elaborate on the reason for such choice. Technical and practical considerations for such decisions could have been briefly described.

Reviewer #2 (Public review):

Overall, this is a solid manuscript that delivers an important community resource. The execution is relatively simple, but the value is real, the work is rigorously performed, and the open dissemination through Zenodo, the F1000Research YCharOS Gateway and OGA is well executed. The effort invested in generating the knockout lines for validation experiments is a clear strength of the study.

Comments on revised version.

I thank the authors for their responses. A number of the smaller points I raised are now resolved. However, I am disappointed to see that the more substantial concerns, which relate to the cell-type expression claims running through the Results and the Discussion, remain unaddressed. My original point was that these claims are not quantitatively supported and that the identity of the iPSC-derived cell types used to make them has not been documented, and I do not think this revision changes that. Most of what I asked for was quantification of experiments the authors have already performed, or quality control on the cell types themselves, and this has been answered with scope arguments rather than with data. The other points I originally raised remain.

I give further details on my two main remaining concerns below:

(1) The cell-type expression claims across the manuscript are still supported by qualitative Western blot images only. These include statements such as "higher signal in glial populations," "enriched in neuronal lineages," "predominantly detected in microglial populations," and the "convergence on multicellular disease mechanisms" narrative in the Discussion. I understand the authors' decision to keep ALS-RAP as a tool and resource paper rather than expand the experimental work. However, if that is the framing, the biological interpretations of these expression data need to be removed. In their response, the authors themselves acknowledge that "quantification across multiple biological replicates would be required to establish definitive differences in protein abundance between neuronal and glial cell types." As long as this quantification is not provided, every claim of cell-type-specific expression should be removed from the Abstract, Results and Discussion, and the data presented as a descriptive survey.

(2) Related to the previous point, I do not think that quality control of the non-microglia iPSC lineages is beyond the scope of the study, as argued in the response. For ALS-RAP to be adopted as a resource, other laboratories need to trust that the authors' iPSC-derived neurons, astrocytes and oligodendrocytes are really neurons, astrocytes and oligodendrocytes. The fact that the differentiation protocols have been previously published does not confirm that they have been correctly implemented here, and this is precisely the kind of quality control that should be documented in the paper. In their response, the authors list the single markers already shown in Figure S3 (ISL1, TH, MBP, GFAP, CD44) as the answer to my concern, but a single positive marker per lineage does not establish purity.

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 prior to publication:

(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.

Comments on revised version.

The author's additions to the discussion are appreciated.

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.

Comments on revised version.

The authors have addressed my comments and have no further suggestions or any changes in my assessment.

Reviewer #1 (Public Review):

This work revisits a substantial part of the published literature in the field of Drosophila innate immunity from 1959 to 2011. The strategy has been to restrain the analysis to some 400 articles and then to extract a main claim, two to four major claims and up to four minor claims totaling some 2000 claims overall. The consistency of these claims with the current state-of-the-art has been evaluated and reported on a dedicated Web site known as Reprosci and also in the text as well as in the 28 Supplements that report experimental verification, direct or indirect, e.g., using novel null mutants unavailable at the time, of a selected set of claims made in several articles. Of note, this review is mostly limited to the manuscript and its associated supplements and does not integrally cover the Reprosci website.

Strengths:<br /> One major strength of this article is that it tackles the issue of reproducibility/consistency on a large-scale. Indeed, while many investigators have some serious doubts on some results found in the literature, few have the courage, or the means and time, to seriously challenge studies, especially if published by leaders in the field. The Discussion adequately states the major limitations of the Reprosci approach, which should be kept in mind by the reader to form its own opinion.

This study also allows investigators not familiar with the field to have a clearer understanding of the questions at stake and to derive a more coherent global picture that allows them to better frame their own scientific questions. Besides a thorough and up-to-date knowledge of the literature used to assess the consistency of the claims with our current knowledge, a merit of this study is the undertaking of independent experiments to address some puzzling findings and the evidence presented is often convincing, albeit one should keep in mind the inherent limitations as several parameters are difficult to control, especially in the field of infections, as underlined by the authors themselves. Importantly, some work of the lead author has also been re-evaluated (e.g., Supplements S2-S4). Thus, while utmost caution should be exerted, and often is, in challenging claims, even if the challenge eventually proves to be not grounded, it is valuable to point out potential controversial issues to the scientific community.

While this is not a point of this review, it should be acknowledged that the possibility to post comments on the ReproSci website will allow further readjustments by the community in the appreciation of the literature and also of the Reprosci assessments themselves and of its complementary additional experiments. As science is evolving continuously, this site allows the current state of knowledge to be updated as needed, not necessarily by the authors themselves.

Comments on revised version:

The major weaknesses identified in the original version of the manuscript have been adequately addressed by the authors. In a few cases, additional KO mutants have been generated and analyzed thus further strengthening the manuscript.

Reviewer #2 (Public review):

Summary:

The authors present an ambitious and large-scale reproducibility analysis of 400 articles on Drosophila immunity published before 2011. They extract major and minor claims from each article, assess their verifiability through literature comparison and, when possible, through targeted experimental re-testing, and synthesize their findings in an openly accessible online database. The goal is to provide clarity to the community regarding claims that have been contradicted, incompletely supported, or insufficiently followed up in the literature, and to foster broader community participation in evaluating historical findings. The manuscript summarizes the major insights emerging from this systematic effort.

Strengths:

(1) Novelty and community value: This work represents a rare example of a systematic, transparent, and community-facing reproducibility project in a specific research domain. The creation of a dedicated public platform for disseminating and discussing these assessments is particularly innovative.

(2) Breadth and depth: The authors analyze an impressive number of publications spanning multiple decades, and they couple literature-based assessments with new experimental data where follow-up is missing.

(3) Clarity of purpose: The manuscript carefully distinguishes between assessing evidential support for claims and judging the scientific merit of historical work. This helps frame the project as constructive rather than punitive.

(4) Metascientific relevance: The analysis identifies methodological and contextual factors that commonly underlie irreproducible claims, providing a useful guide for future study design and interpretation.

(5) Transparency: Supplementary datasets and the public website provide an exceptional degree of openness, which should facilitate community engagement and further refinement.

Comments on revised version:

In the revised manuscript the authors addressed in a satisfactory manner the weaknesses I previously highlighted in my initial review.

Reviewer #3 (Public review):

Summary:

In this ambitious study the authors set out to analyse the validity of a number of claims, both minor and major from 400 published articles within the field of Drosophila immunity that were published before 2011. The authors were able to determine initially if claims were supported by comparing to other published literature in the field and if required by experimentally testing 'unchallenged' claims that had not been followed up in subsequent published literature. Using this approach the authors identified a number of claims that had contradictory evidence using new methods or taking into account developments within the field post initial publication. They publish their findings on a publicly available website designed to enable the research community to assess published work within the field with greater clarity.

Strengths:

The work presented is rigorous and methodical, the data presentation is high quality and importantly the data presented support the conclusions. The discussion is balanced and the study is written considerately and respectfully highlighting that the aim of the study is not to assign merit to individual scientists or publications but rather to improve clarity for scientists across the field. The approach carried out by the researchers focuses on testing the validity of the claims made in the original papers rather than testing whether the original experimental methods produced reproducible results. This is an important point since there are many reasons why the original interpretation of data may have understandably led to the claims made. These potential explanations for irreproducible data or conclusions are discussed in detail by the authors for each claim investigated.

The authors have generated an accompanying website which provides a valuable tool for the Drosophila Immunity research community that can be used to fact check key claims and encourages community engagement. This will achieve one important goal of this study - to prevent time loss for scientists who base their research on claims that are irreproducible. The authors rightly point out that it is impossible (and indeed undesirable) to avoid publication of irreproducible results within a field since science is 'an exploratory process where progress is made by constant course correction'. This study is however an important piece of work that will make that course correction more efficient.

Comments on revised version:

I'm impressed with the revisions carried out by the authors. No further comments.

Reviewer #4 (Public review):

Westlake, Lemaitre and colleagues have revised their manuscript regarding the reproducibility of studies on the Drosophila immune system. As was emphasized by all reviewers at the original submission of this manuscript, this paper captures a tremendous effort and makes a valuable contribution to the community.

One strong conclusion to be drawn from the overall effort is that the scientific literature is, for the most part, very good. The vast majority of claims made in the 400 papers assembled for this project were substantiated, either with previously published follow-up work or with new experimentation presented here. Although the authors have chosen not to quantify the overall validation rate of scientific claims in the present study, that information is available on the ReproSci website and in the companion paper.

In the revised manuscript, the authors make the distinction between repeatability/direct reproducibility and what they term "conceptual replication" and "indirect reproducibility", emphasizing that this study falls in the latter category because in many cases the present authors are not attempting to reproduce or replicate the original study. Instead, they are often using new tools to reinvestigate original claims and determine the accuracy of those claims, and whether they stand up to additional data. The authors have replaced the word "irreproducible" with "irreplicable" in the revised manuscript. I continue to believe the present work can be considered "validation" of the original studies and that the present work is accurately considered a "validation project". The clarification of terminology has strengthened the revised manuscript.

As with any study, the strength and breadth of conclusions that can be drawn from the present work are a function of the strength and resolution of the experiments performed. This may vary across the supplements so readers should interpret the presented results accordingly. In my opinion, all of the supplements in the revised article are sufficiently convincing to warrant inclusion in the final manuscript.

There are several instances where the authors have declined to do additional experiments in response to review. I am comfortable with these decisions, recognizing that the extra work required would be prohibitively much. Instead, the authors have adequately revised their presentation of findings. In some instances where the authors have declined to perform additional experiments, they have provided a compelling logic for their decision in the response to reviews, which unfortunately does not appear in the main manuscript. For example, in response to a comment by Reviewer 1, the authors wrote "In response to the suggestion of scanning multiple blots for quantification, we note that no clear enhancement of ModSP autoactivation was observed across different incubation times. We also explored alternative incubation conditions, but none resulted in a detectable increase in activation. Thus, additional quantification would not change the overall conclusion." This is convincing, but no readers will see it unless it appears in the manuscript or the public response to review. At the same time, I recognize the practical limitations in available space in the main article and supplement. There are a handful of instances where reviewers have made suggestions of additional/alternative experiments that would be more rigorous, accurate, or convincing than what is currently presented but the authors have declined to take up those recommendations. In some instances, I strongly agree with the authors' rationale and in other cases it is less so. However, I did not find any instance where the absence of the suggested additional experiment substantially undermined the conclusions being put forward in this manuscript. Therefore, I believe these decisions can be left to the discretion of the authors in accordance with the eLife publication model and interpretations can be left to the discretion of the readers.

Reviewer 1 and Reviewer 4 both commented on a lack of demonstration of RNAi efficacy in the knockdown experiments and suggested that negative conclusions drawn from RNAi experiments should include controls showing that the RNAi was effective. The authors have declined to include such controls, and I recognize they may be impossible to add now without redoing the entire associated experiment. Instead, the authors point to prior use of the RNA constructs by other authors in the literature. That should be recognized as less than ideal, although it is probably substantially adequate. The authors have added statements to the Discussion that RNAi knockdowns are not equivalent to null mutations and negative results may appear because of hypomorphic effects. This is a positive addition.

The replacement of the sphinx RNAi with a deletion of Sphinx 1 and Sphinx 2 as part of a general strengthening of the section on sphinx and spheroid is an extremely strong addition to the paper. I could not expect that level of effort for each of the validations and I acknowledge and appreciate that they have made that effort here.

Reviewer 1 and Reviewer 4 both commented that some figures in the original submission were ambiguous with respect to sample size and distribution of data points, including undefined error bars and inappropriate graphical representation of experiments with low sample size. These concerns have been largely addressed in the revised manuscript, including with addition of statistical analyses in several cases. The authors note that the figures were generated by multiple individuals across several research teams so it would be a large effort unify their style. I agree that this would be an unreasonable effort at this stage, for minimal return that is primarily aesthetic. The authors have now added illustration of all data points in most figures and have defined the error bars. One response to Reviewer 1 "Data points should be shown in the figure" with respect to a figure in S16 was "Unfortunately, we could not address this point to the difficulty to reach the collaborator who provided the data." That response certainly would not be acceptable in the publication of a typical scientific paper. However, I sense a different standard is being applied to the present work and the lack of these data points in this particular figure does not decrease my confidence in the authors' overall conclusion.

Overall, this is a unique piece of research that is impressive in its scope. Scientific finding is never complete; the process is eternally iterative. Inclusion of the present study in the scientific literature is a testament to collective willingness to engage in that iteration. The authors of this paper are making an exemplar contribution to accuracy in scientific research and should be commended for their effort, dedication, and fairness in what can be a delicate topic.

Reviewer #1 (Public review):

Summary:

The authors set out on the ambitious task of establishing the reproducibility of claims from the Drosophila immunity literature. Starting out from a corpus of 400 articles from 1959 and 2011, the authors sought to determine whether their claims were confirmed or contradicted by previous or subsequent publications. Additionally, they actively sought to replicate a subset of the claims for which no previous replications were available (although this set was not necessarily representative of the whole sample, as the authors focused on suspicious and/or easily testable claims). The focus of the article is on inferential reproducibility; thus, methods don't necessarily map exactly to the original ones.

The article is a large-scale analysis of the individual replication findings, which are presented in a companion article (Westlake et al. doi.org/10.1101/2025.07.07.663442). In their retrospective analysis, the authors find that 61% of the original claims were verified by the literature, 7.5% were partially verified, and only 6.8% was challenged, with 23.8% having no replication available. This is in stark contrast with the result of their prospective replications, in which only 16% of claims were successfully reproduced.

The authors proceed to investigate correlates of replicability, with the most consistent finding being that findings stemming from higher-ranked universities were more likely to be challenged - a finding that is observed in a multivariable model and multiple sensitivity analyses. Other trends observed in the initial descriptive analysis, such as less replicability in findings from "trophy" journals and duration of engagement with the field, seem less consistent and fail to reach significance in the multivariable model, and should thus be considered tentative.

Although this is a major contribution to the field, it is important to note that the high replication rates are mostly based on a retrospective survey of the literature, while prospective replication rates in the subset of articles directly replicated by the authors were much lower. Thus, the possibility of biases relating both to the field's decisions about what to replicate (and what to publish) and to the authors' criteria in evaluating claims should be considered when comparing these results with prospective replicability estimates obtained in other fields of science.

Strengths:

The work presents a large-scale, in-depth analysis of a particular field of science that includes authors with deep domain expertise of the field. This is a rare endeavour to establish the reproducibility of a particular subfield of science, and I'd argue that we need many more of these in different areas.

The project was built on a collaborative basis (https://ReproSci.epfl.ch/), using an online database (https://ReproSci.epfl.ch/), which was used to organize the annotations and comments of the community about the claims. The website remains online and can be a valuable resource to the Drosophila immunity community. A companion article details the individual replication findings (Westlake et al. doi.org/10.1101/2025.07.07.663442).

The revised version of the preprint includes a number of sensitivity analyses as supplementary material, which help to evaluate how the main findings change according to analytical decisions.

Main concerns:

(1) Although a number sensitivity analyses have now been conducted in order to address reviewer comments, these have been relegated to the supplementary material. Their existence is acknowledged in the multivariable model section, but the actual results of the main analysis are not mentioned at all in the main text and not taken into account in the discussion. Thus, the casual reader is not made aware of how particular results are dependent on analysis choices.<br /> Although the main finding of the correlation between replicability and university ranking holds over nearly all the analyses conducted, others are not as consistent. The association of irreplicability with trophy journals, for instance, is not statistically significant in the multivariate model (suggesting that it may have been partially due to confounding of journal tier with institutional ranking). It becomes even weaker when unchallenged claims are removed from the analysis (see point 2 below). Similarly, the association of irreplicability with "exploratory" career style is already weak in the univariate analysis and becomes even more marginal in the multivariable model. The effect of the lead author having been a first author in the field, on the other hand, is stronger than that of journal impact factor or that of the continuity/exploratory style (even though it fails to reach formal significance), but seems to receive less attention in the discussion.<br /> These differences in strength of evidence could be made more clear both in the abstract and in the discussion, which seem to prioritize some trends over others when discussing results. Once more, presenting a brief description of how results are affected by the sensitivity analyses in the results would be important to make these nuances clearer to the reader.

(2) I still do not understand the authors' choice of including unchallenged claims along with verified ones in the multivariate model used to investigate predictors of replicability. In particular, the justification given for this choice in the supplementary material (i.e. the fact that a finding remaining unchallenged is not a random event and might be influenced by variables such as journal tier and institutional ranking) actually seems like an additional reason for not including unchallenged claims along with verified ones: if findings from low-tier journals or less prestigious institutions are more likely to remain unchallenged, this could lead to a spurious claim of "less irreplicability" in these journals and institutions, which could be purely due to receiving less attention. Predictably, both associations are reduced when unchallenged claims are removed (although the reduction is much more marked in the journal case) in the sensitivity analysis (whose results, again, are only mentioned in the supplementary material).<br /> Moreover, multiple other statements by the authors themselves concerning unchallenged claims seem to contradict their own decision. In multiple points in the text, they reiterate that remaining unchallenged is associated with less replicability (as suggested by the fact that replicability rates were lower in prospective replications than in the literature, even for non-suspicious claims). The decision to include unchallenged claims along with verified ones thus seems unjustified - if anything, I'd have expected the opposite approach after reading the discussion. That said, I'd argue that the best decision would be to exclude them from the analysis, as done in Figure S4, and presenting this model as the main one.

Reviewer #3 (Public review):

Summary:

The authors of this paper were trying to identify how reproducible, or not, their subfield (Drosophila immunity) was since its inception over 50 years ago. This required identifying not only the papers, but the specific claims made in the paper, assessing if these claims were followed up in the literature, and if so whether they supported or refuted the original claim. In addition to this large manually curated effort, the authors further investigated some claims that were left unchallenged in the literature by conducting replications themselves. This provided a rich corpus of the subfield that could be investigated into what characteristics influence reproducibility.

A major strength of this study is the focus on a subfield, the detailing of identifying the main, major, and minor claims - which is a very challenging manual task - and then cataloging not only their assessment of if these claims were followed up in the literature, but also what characteristics might be contributing to reproducibility, which also included more manual effort to supplement the data that they were able to extract from the published papers. While this provides a rich dataset for analysis, there is a major weakness with this approach, which is not unique to this study.

The main weakness is relying heavily on the published literature as the source of whether a claim was determined to be verified or not. Nonetheless it is understandable why the authors took this approach - it is the only way to get at a breadth of the literature. However, there are many documented issues with this stemming from every field of research - such as publication bias, selective reporting, all the way to fraud. Importantly, this limitation is highlighted throughout the paper. At the same time, it is not reasonable to expect this study to have conducted independent experimental replications for all 400 papers identified as this would have been a laborious and costly effort. To overcome this weakness, the authors leveraged their own expertise, and crowdsourcing within the Drosophila immunity community, to assess the validity of the claims. While this helps mitigate some of this weakness, it does introduce new challenges in interpretation, which is acknowledged in the limitations section of the discussion. Overall, this situates this study more as an expert assessment of the replicability of the published literature opposed to an assessment of independent experimental replications.

The authors should be applauded for the monumental effort they put into this project, which does a wonderful job of having experts within a subfield engage their community to understand the connectiveness of the literature and attempt to understand how reliable specific results are and what factors might contribute to them. This project provides a nice blueprint for others to build from as well as leverage the data generated from this subfield and thus should have an impact in the broader discussion on reproducibility and reliability of research evidence.

Reviewer #1 (Public review):

This manuscript examines EEG activity during face-to-face Diapix conversations and asks whether neural activity preceding a participant's next speaking turn predicts the latency and duration of that turn. The authors report sustained ERP and alpha/beta effects, particularly associated with upcoming response duration, together with temporal-generalization decoding beginning approximately 1-1.3 s before speech onset. They interpret these results as evidence for two stages of conversational speech planning: an early process that activates and maintains the forthcoming response, and a later process related to commitment and motor execution. The naturalistic dyadic EEG setting, relatively large number of conversational turns, trial-level analyses, and convergence across ERP, oscillatory, and multivariate approaches are important strengths.

I nevertheless have substantial concerns about the interpretation. Most importantly, the EEG analyzed during this period is recorded while participants are processing their partner's speech. Therefore, neural activity necessarily includes auditory, linguistic, semantic, pragmatic, and turn-boundary processing, in addition to any potential response-planning activity. Because properties of the incoming utterance are themselves strongly related to what the listener subsequently says, predicting subsequent response duration or latency does not by itself isolate production planning. The authors acknowledge some of these limitations, but the current title, theoretical framing, Figure 8, and conclusions remain considerably stronger than the evidence allows. Indeed, the Discussion itself acknowledges that the signals cannot be determined to reflect linguistic content, general planning demands, or motor preparation.

I therefore think the study could become valuable, but substantial additional analyses and conceptual reframing are required.

Major concerns:

(1) The central problem is dissociating speech planning from listening/comprehension

This is my main concern with the manuscript and also with the claims in the Discussion. The authors repeatedly infer that because EEG during the partner's turn predicts the duration or latency of the participant's subsequent response, this activity reflects speech planning. For example, the manuscript states that examining the period before self-speech onset "isolat[es] preparatory neural activity before articulation." I do not think this inference is justified.

During precisely this interval, the participant is also listening to and comprehending the partner. The incoming utterance is not independent of the forthcoming response; it is what determines the response. A complex statement, several questions, a request for clarification, or an information-rich description will produce different auditory/linguistic/comprehension activity and may also systematically change the length and timing of the subsequent response. Consequently, both partner speech → listening/comprehension EEG and partner speech → subsequent response properties can generate an EEG-response relationship without requiring that the measured EEG activity itself reflect production planning.

The authors control partner-turn duration, which is useful, and they show that self-duration, self-latency, and partner duration are behaviorally related. However, partner duration is only a very coarse property of the input and cannot control for what participants are actually processing. The analysis should consider, at minimum, partner word count/speech rate, acoustic properties, information density or lexical/syntactic complexity, dialogue act, question versus statement, and preferably semantic/discourse content. For example, the number of questions or propositions in the partner's turn may directly determine both comprehension demands and how much the listener subsequently needs to say.

Given that transcripts and word-level timing are already available, the authors should be able to substantially improve this analysis. One particularly useful approach would be to construct an encoding model of the EEG driven by acoustic and linguistic features of the partner's speech and then test whether the residual neural activity continues to predict subsequent response properties. At minimum, richer input-related covariates are necessary.

Without such analyses, I think the main conclusion should be reframed from "neural signatures of speech planning" to something closer to "pre-response neural activity during conversational listening predicts properties of the subsequent turn." The latter is supported by the data; the former requires stronger dissociation.

(2) It is not clear that the analyzed period is actually a "listening interval".

Relatedly, the definition of the analyzed interval needs careful reconsideration. Epochs are aligned to the participant's own speech onset and extend back 2.2 s, whereas a turn transition is accepted whenever self-onset falls between −1 and +1 s relative to partner offset.

Therefore, the interval from −2 to 0 s relative to self-onset is not necessarily an interval during which the partner is continuously speaking. For positive response latencies, part of this interval contains silence following partner offset. If the preceding partner IPU is relatively short, earlier portions of the epoch could potentially contain a different conversational state altogether. Yet the manuscript describes the whole pre-onset period as being "while they are listening to their partner."

This matters particularly for the comparison between the early and late periods. Different response-latency conditions will necessarily place the partner offset at systematically different positions within an epoch aligned to self-onset. The authors already demonstrate this problem for the anterior latency ERP, where its inflection tracks the distribution of partner offsets. This is an important observation, but I do not think the problem is restricted to that single component.

The authors should quantify, separately for TW1 and TW2, how much time is occupied by: active partner speech, silence after partner offset, overlapping speech, and potentially other conversational states.

I would strongly recommend repeating the critical analyses for periods/trials during which the partner is actually speaking, and/or explicitly modeling partner-speech presence and the temporal distance to partner offset. Analyses aligned with both partner offset and self-onset would also help separate responses to the conversational boundary from activity specifically preceding self-production.

(3) The Early-versus-Late Planning hypotheses are not sufficiently differentiated

I was confused by the theoretical contrast (e.g., Lines 53-62). The Early Planning account is described as proposing that response preparation begins once sufficient information is available, whereas the Late Planning account explicitly allows "some aspects of conceptual preparation" during listening but proposes that full articulatory planning is delayed until near the turn boundary.

Under these definitions, both hypotheses allow for early planning. They differ mainly in the level of production planning that occurs early, that is, conceptual/content preparation versus articulatory/motor preparation. This is quite different from a simple early-versus-late timing hypothesis.

This is problematic because the current EEG measures do not determine whether the early activity reflects conceptual planning, lexical formulation, articulatory planning, comprehension, or general cognitive demands. Indeed, the authors explicitly acknowledge this later. Consequently, observing EEG-behavior relationships 1 s before speech cannot adjudicate between the two hypotheses as currently formulated.

The theoretical framework should therefore distinguish explicitly among at least: conceptual/message planning → linguistic formulation → articulatory/motor preparation → initiation.

If the "Late Planning" hypothesis specifically concerns late articulatory planning, then the authors need evidence that distinguishes articulatory planning from earlier conceptual processes. Otherwise, I would describe the findings more conservatively as showing early behavior-related neural activity followed by stronger onset-proximal sensorimotor activity, rather than claiming that Early and Late Planning accounts have been reconciled.

(4) Response duration is not equivalent to the amount or content of a response that has already been planned

The strongest empirical finding concerns upcoming speaking duration. I agree that this is interesting. However, throughout the Discussion, duration is gradually interpreted as the "extent," "scale," or even the amount of content already prepared. For example, the authors argue that longer responses involve maintaining a larger set of propositions, lexical items, or planned sequences. Figure 8 goes even further and explicitly contrasts "High Content" and "Low Content," although no measure of speech content has been analyzed.

I do not think response duration alone supports this inference.

A 2-s utterance is not necessarily twice as much planned content as a 1-s utterance. Duration can vary because of speech rate, hesitations, lexical retrieval, syntactic formulation, repetitions, interactive feedback, or decisions made after speech has already begun. An eventual IPU duration is therefore not necessarily specified before articulation.

This issue is especially important because the manuscript explicitly acknowledges that the analyses cannot establish whether the signals encode lexical or semantic content. It is then inconsistent to conclude later that early activity supports "what and how much to say."

Since transcripts are available, the authors could directly test several alternatives: word count, content-word count, number of propositions, speech rate, syntactic complexity, semantic similarity between partner and self turns, and possibly dialogue-act categories. If the early signal predicts eventual linguistic content/amount rather than merely duration, the claim of maintained response specification would become considerably stronger. If not, the interpretation should remain at the level of response duration.

(5) The ROI/time-window linear mixed-model (LMM) analysis appears circular

I have a methodological concern about the relationship between the cluster analyses and subsequent LMMs. The authors first identify electrodes and time ranges showing significant differences between long versus short responses or fast versus slow responses. They then use those same data-driven clusters to define anterior/posterior ROIs and TW1/TW2 and extract EEG values from the same trials, after which the extracted EEG measures are tested as predictors of the same behavioral variables.

This creates a potentially serious double-dipping/selection bias problem. The continuous LMM outcome is not statistically independent of the median-split outcome used to select the neural features. FDR correction at the LMM stage does not correct for this feature-selection procedure.

I think the confirmatory analyses need independent feature definition, for example, predefined electrodes/time windows from previous literature, split-half analysis, leave-one-participant-out feature selection, or another cross-validated procedure in which the data used to identify the ROI/window are independent from those used to estimate and test the EEG-behavior relationship.

This is particularly important because the LMM results are used as evidence that the effect survives behavioral covariates. That conclusion should come from statistically independent tests.

(6) Several statistical and validation issues need additional attention

There are several related concerns here. First, the trial-level LMMs include only (1 | participant), despite the fact that the observations come from dyadic interactions, multiple 4-min conversations, and temporally adjacent conversational turns. Participants within the same dyad do not generate independent observations because the speech of one participant directly determines the context for the other participant. I would expect the hierarchy to include at least a dyad/conversation block, and random slopes for critical within-participant predictors should be considered. Robustness to conversational autocorrelation should also be examined.

Second, the authors describe response duration as approximately log-normal, but the model description indicates only centering/scaling rather than transformation. The authors should report residual diagnostics and consider modeling log-duration or using an appropriate non-Gaussian mixed model.

Third, the decoding uses stratified 10-fold cross-validation, but it is unclear how conversational dependence is handled. Randomly assigning turns from the same 4-min conversation to training and testing sets may permit classifiers to exploit slow neural states, discourse context, or block-specific characteristics shared by temporally adjacent turns. A more convincing analysis would use blocked cross-validation, such as leave-one-conversation/run-out. The authors should also clarify whether decoding is performed independently within participants and whether median thresholds are participant-specific or global.

These analyses are particularly important because temporal generalisation establishes stable predictive information, but not the functional identity of that information. A stable partner-speech/context signal could also produce a broad temporal-generalisation pattern.

(7) ERP baseline correction and potential speech/muscle contamination require clarification

The ERP preprocessing raises another concern. ERPs were baseline corrected using the entire −2200 to +500 ms epoch. This interval includes the first 500 ms after self-speech onset, the period most vulnerable to speech-related muscular activity. Because the main grouping variable is upcoming response duration, post-onset activity may itself systematically differ between long and short responses. Subtracting a baseline that includes this activity could introduce condition-dependent shifts in the pre-speech ERP.

I therefore think the critical ERP results should be demonstrated without using post-speech activity in baseline estimation, for example, by using an appropriate pre-speech/regression-based baseline strategy, or by analyses demonstrating that the result is insensitive to baseline choice.

There is also an intriguing inconsistency concerning EMG. The participant section states that facial hair was an exclusion criterion to ensure electrode contact for EMG recordings, whereas the Limitations section suggests that combining EEG with EMG would be useful in future work. If EMG was in fact recorded, it should be described and, ideally, analyzed to establish when peripheral speech preparation begins and whether the onset-proximal alpha/beta results remain significant after controlling for muscle activity. This would be particularly valuable for the proposed late "motor commitment" interpretation.

Reviewer #2 (Public review):

This is an excellent paper, well researched, correctly referenced and very well presented. The main addition to the existing literature, apart from supporting earlier results (see also Roberts et al. 2015), is the ability to detect the length of the upcoming conversational contribution from the neural signal. This is interesting because it shows very early and sustained planning of response during listening to the incoming turn from the interlocutor. Why is that interesting? Because this simultaneous listening and response planning is utilising the very same mental machinery - a kind of dual-tasking that humans generally find very hard, given extreme capacity constraints on short-term and working memory. This then talks to the issue of whether speech production and comprehension are more independent than current theory holds. Clearly, human language is one of the elite skills of the species, but this dual-task ability in just this domain may be a specialization evolved over deep evolutionary time. I would like to see this background sketched more clearly in the introduction - it would make the paper of more interest to the general reader.

The other main contribution is methodological, namely the ability to extract meaningful EEG out of unconstrained verbal interaction, using, e.g. machine learning to filter out movement artefacts. Earlier work was held back by scepticism that this was possible, but this paper points clearly to one way to do this. A further contribution is the suggestion that Beta modulation is associated with the duration of the turn under planning, while Alpha modulation was in addition associated with preparation for execution. This needs to be tested in further studies, but could be a valuable addition to the analytical armoury.

In the final discussion, a conceptual model is outlined in Figure 8. This looks very similar to that given in Levinson & Torreira 2015 and Levinson 2016 - at least if there is a difference, it would be good to see it clarified. Note that in that model we hypothesized that preparation of the upcoming turn was continued right up to completion and then held if necessary for ending cues from the interlocutor. Note that quite late delivery preparations have been noted both in Torreira et al. 2015 (breathing signal) and in Bögels & Levinson 2023; in the latter, there is evidence of tongue preparation sometimes occurring not tied time-wise to delivery (suggesting inhibition till incoming turn is ending). So our model already had an early and late component - the news here is that the early component can be robustly discerned in naturalistic verbal interaction.

Also, in the discussion, the Pickering & Garrod IAM model is favourably discussed, even though that model imagines the production machinery is used to predict the unfolding comprehension of the incoming turn: in that case, the production machinery would have to be doing two things at once - doing comprehension and planning one's own turn. In a way, the main lesson of the findings of the submission is precisely that this is implausible.

In general, I have no hesitation recommending publication of an important contribution to the neuroscience of human performance in ecologically more valid settings.

References:

Bögels S, Levinson SC. Ultrasound measurements of interactive turn-taking in question-answer sequences: Articulatory preparation is delayed but not tied to the response. PLoS One. 2023 Jul 5;18(7):e0276470. doi: 10.1371/journal.pone.0276470. PMID: 37405982; PMCID: PMC10321606.

Griffin, Z.M. and Bock, K. (2000) What the eyes say about speaking. Psychol. Sci. 4, 274-279.

Levinson SC. Turn-taking in Human Communication--Origins and Implications for Language Processing. Trends Cogn Sci. 2016 Jan;20(1):6-14. doi: 10.1016/j.tics.2015.10.010. Epub 2015 Dec 1. PMID: 26651245.

Levinson SC, Torreira F. Timing in turn-taking and its implications for processing models of language. Front Psychol. 2015 Jun 12;6:731. doi: 10.3389/fpsyg.2015.00731. PMID: 26124727; PMCID: PMC4464110.

Roberts SG, Torreira F, Levinson SC. The effects of processing and sequence organization on the timing of turn taking: a corpus study. Front Psychol. 2015 May 13;6:509. doi: 10.3389/fpsyg.2015.00509. PMID: 26029125; PMCID: PMC4429583.

Reviewer #1 (Public review):

In this manuscript, Yadav and coworkers investigate hydrophobicity governing membrane insertion of mitochondria-encoded polypeptides. The basis of their considerations is the question of why mitochondria maintain a co-translational insertion mechanism and what rules allow allotopic expression. The issue if mitochondrial-encoded proteins can be expressed in the cytoplasm and imported into mitochondria is a long-standing topic in the field with broad implications for the treatment of mitochondrial disorders. In the manuscript, Yadav et al. propose that aliphatic-to-threonine substitution (ATS) are a recurrent and convergent adaptation in the transmembrane helices (TMH) of the inner mitochondrial membrane (MIM) proteins, solving the conflict between hydrophobicity required for membrane insertion and the need for soluble cytosolic transit. The study combines structural comparative genomics, MD simulations, and cellular experiments.

This is a well-written and highly interesting study that provides a solid basis for considerations regarding mitochondrial protein insertion. I am sure that the study will represent an important asset for future studies on allotopic protein expression and potentially the development of genetic tools. The study provides solid evidence, from evolutionary correlation to experimental manipulation and molecular-level explanation. The study provides abundant and well-thought-out controls.

Specific points

(1) There are multiple comparisons throughout the manuscript, and not all of them appear to have been explicitly corrected. Numerous Wilcoxon tests, t-tests, and ANOVAs are reported (Figures 2, 3, 4), without a systematic correction for multiple testing or clear mention of FDR/Bonferroni corrections, except for the directional analysis of substitutions, for which the Benjamini-Hochberg procedure was applied. Given the large number of pairwise comparisons in Figures 3 A-C; and Figure 3 Supplement 1, Supplement 2, Supplement 4, and Supplement 5, with dozens of p-values per panel, it would be relevant to clarify whether a multiple-testing correction was also applied in these analyses. If not, the authors should explain the rationale for not applying one.

(2) The authors used experimental structures as one criterion but, in other cases, AlphaFold2 to compare evolutionarily distant pairs. Using non-experimental structures as a base for structural alignment among very divergent proteins is understandable but also prone to errors, especially for membrane proteins. It might wrongly identify which helices are considered "related". To address this, the authors need to clearly indicate which data are based on experimental or computational data - this is not always the case.

(3) The Kyte-Doolittle hydrophobicity scale is standard but relatively raw in comparison with scales based on experimental insertion energy (like Hessa et al., cited in the text or Bothelo et al.). It might be useful to explain why the authors preferred K-D over an experimental membrane insertion scale for the analysis. Particularly, since the Hessa et al. scale was later used in Figure 8G for the thermodynamic analysis.

(4) The phylogenetic tree of ATP6 (Figure 5A) presents low bootstrap values in some internal nodes (43, 40, 55) on which the interpretation of "independent events" of genetic transfer relies. This must be discussed, since a bootstrap of 40-50% provides not too much confidence in the deep topology of the tree, even though the point that the nuclear sequences don't form a monophyletic group with the mitochondrion seems robust (supported by distinct MTS in each clade).

(5) Definition of "core" vs "peripheric" in TMH (11 central positions) seems arbitrary; it would be worth clarifying why that cut was chosen and whether the results are robust when they vary.

(6) There is a notable gap in the literature when describing some of the findings in the Results and Discussion section:

Biophysics of threonine/serine in TMH:<br /> Senes et al., 2004, Dawson et al., 2002. The result described in Figure 8 (Threonine rotamers) is essentially a biophysical refinement of a body of work already established by the Engelman and DeGrado labs about how Ser-Thr, in comparison with other polar residues, promotes specifically the packing and stability of TMH through intra-inter helical hydrogen bonds, without compromising the membrane insertion. Chamberlain & Bowie, 2002 was included (C-H bonds), and Hessa et al. 2005 (energy of insertion), but no reference was included about why Ser/Thr are polar amino acids "tolerated" at the hydrophobic core of TMH, which is the state the authors describe.

In the argument about why TMH hydrophobicity matters for targeting (not only for membrane insertion), there is much more direct and recent mechanistic work that should be included: Botelho et al., 2011 established experimentally the hydrophobicity threshold for the "stop-transfer" pathway of TIM23 with quantitative data in an analogous way as reported in this work. Lee et al., 2020 and Yang et al., 2025 described structurally how hydrophobicity of the substrate modulates the Mgr2-Tim17 interaction to allow precursors to be embedded within the membrane.

It may be helpful to consider previous published work as an important precedent for the discussion of TMH hydrophobicity and TIM23 in allotopic expression. The final observations of incomplete mitochondrial localization despite reduced TMH hydrophobicity could also be discussed in light of previous published studies. These studies suggest that hydrophobicity is an important factor, but not the only determinant of efficient mitochondrial targeting. The role of TMH hydrophobicity in mitochondrial import had already been explored in this context. This would help clarify how the current findings build upon the existing literature.

Reviewer #2 (Public review):

Summary:

The paper by Yadav et al. provides compelling theoretical and experimental evidence for an aliphatic-to-threonine substitution (ATS) in the transmembrane helices (TMs) of membrane proteins residing in the inner mitochondrial membrane (IMM) across several species. Furthermore, using ATS, they designed variants of the human Atp6 subunit and of Atp6 subunits from other species that seem to be internalized into mitochondria.

Strengths:

In general, the results support the conclusions. The findings are particularly relevant to the field of allotopic expression and could have significant implications for the future development of gene therapies for mitochondrial disorders.

Weaknesses:

The manuscript could be substantially strengthened by the addition of further experimental evidence and by some minor revisions and adjustments to the interpretation and presentation of the results.

Major Points

(1) The protein product of atp6 is an essential component of the proton channel of ATP synthase, and more than a dozen pathogenic mutations in this gene have been associated with mitochondrial diseases. Accordingly, several studies have reported attempts to achieve allotopic expression of Atp6 in human cells. I would dedicate a pair of paragraphs in the Discussion to acknowledging the most relevant of these previous efforts. Here, in the work of Yadav et al., several Atp6 constructs were detected within mitochondria by fluorescence microscopy. However, no direct evidence was provided that the allotopically expressed Atp6 protein was incorporated into complex V. Localization within mitochondria, by itself, does not demonstrate successful allotopic expression in the functional sense, as the protein could remain unassembled, inserted in an incorrect topology, or in a process of degradation. In my view, this is the main weakness of the present study. A convincing demonstration of allotopic ATP6 expression should ideally be supported by biochemical evidence showing that the imported subunit is stably incorporated into ATP synthase and, most importantly, contributes to its function. This would require demonstrating at least partial restoration of the respiratory or OXPHOS defect.

I therefore encourage the authors to test their different constructs in cybrid cells carrying a pathogenic mutation in the mitochondrial atp6 gene and determine whether they can restore OXPHOS function or, at a minimum, respiratory capacity. Such an experiment would provide substantially stronger evidence that the imported Atp6 protein is not merely localized to mitochondria but is functionally incorporated into ATP synthase. Alternatively, Blue Native PAGE of mitochondria isolated after allotopic expression of the different constructs, followed by second-dimensional SDS-PAGE, could provide direct evidence that a tagged Atp6 subunit has been incorporated into the ATP synthase complex. Furthermore, this experiment could provide a valuable functional comparison of the three ATS constructs (T1, T2, and T3) and identify which one is most effective in restoring function. Importantly, the T3 construct, despite carrying the highest level of ATS, may not necessarily be the most effective in increasing the levels of assembled ATP synthase or in restoring ATP synthase activity.

(2) I do not agree with the interpretation of the results presented in Figure 7 - Figure Supplement 3, "Targeting tracks with hydrophobicity, not expression." I am concerned that the experimental design may be biased. The two most highly hydrophobic proteins (WT and T1), which are expected to have greater difficulty being imported into mitochondria, were selected for low-level expression. In contrast, the two constructs with extensive ATS (T2 and T3) were expressed at high levels. The hydrophobicity effect and the expression levels are confounded in this experiment, making it difficult to determine whether the observed targeting differences are attributable to one of these parameters. To disentangle these effects, this experiment should be repeated using both low and high expression levels for all four constructs, resulting in eight experimental conditions (four constructs x two expression levels).

Reviewer #3 (Public review):

Yadav et al. revisit the old conundrum of why certain genes encoding inner mitochondrial membrane (IMM) proteins are retained in the mitochondrial genome (mtIMMs; 13 genes in humans) while genes encoding other IMM proteins (nuIMMs) have been relocated to the nuclear genome during evolution. It was noted already in 1986 that nuIMMs have transmembrane helices (TMHs) of lower hydrophobicity than mtIMMs (FEBS Lett 198:1). A number of subsequent studies have shown that some mtIMMs can be converted to nuIMMs by appending a mitochondrial targeting sequence (MTS) and, in most cases, reducing the hydrophobicity of one or more of the TMHs (reviewed in Genes & Dis 7:578). Conversely, increasing the hydrophobicity of a TMH in a nuIMM protein can prevent its import into mitochondria (e.g., PNAS 99:10510). Whether the reduction in TMH hydrophobicity in nuIMMs is mainly driven by the need to bypass cotranslational targeting of the protein to the ER and trafficking along the secretory pathway, or to prevent the protein from being prematurely stuck in the IMM during mitochondrial import, is still, to some extent and open question, and experimental support for both possibilities exists (e.g., FEBS Lett 591:65).

Yadav et al. present an extensive bioinformatics study which nicely confirms the reduced hydrophobicity of nuIMM TMHs compared to mtIMM TMHs (and also compared to TMHs in proteins that traffic along the secretory pathway). The study is based on a lot more sequence data and many more kinds of comparisons than previous publications, and leaves little doubt that the reduction in hydrophobicity in nuIMM TMHs is real.

Based on their bioinformatics analysis, the authors also propose that aliphatic -> Thr substitutions (ATS) are the main driver of hydrophobicity reduction in nuIMMs, TMHs, and that Thr has been enriched also in mtIMMs. TMHs in organisms that depend only on Oxa1-type IMM insertases, possibly in order to reduce the aggregation propensity of these proteins when produced in the mitochondrial matrix. This is a novel hypothesis that has not been proposed in earlier work.

The authors additionally show that a mitochondrially encoded ATP6 protein can be converted into a nuclearly encoded, import-competent protein using ATS to reduce the hydrophobicity of its TMHs and, conversely, that import of a nuclearly encoded ATP6 protein is blocked when Thr residues in its TMHs are mutated to aliphatic residues. Similar findings have been published before; what is new is the use of aliphatic -> Thr substitutions (rather than other aliphatic -> polar substitutions).

Detailed comments:

While previous bioinformatics studies of how gene relocation from the mitochondrial to the nuclear genome impacts protein sequence have focused on TMH hydrophobicity, Yadav et al. expand this to include individual amino acid frequencies. The authors present four types of analyses:

(1) The authors identify E. coli homologs of the 13 genes retained in human mitochondrial DNA, do structure-based sequence alignments of the homologous pairs, and then compare the individual aligned TMH pairs in terms of hydrophobicity and Thr content. The authors also calculate an amino acid substitution matrix from the aligned TMH pairs. These data are presented in Figure 2 (and supplements).

Hydrophobicity is calculated based using the GRAVY algorithm and the Kyte-Doolittle scale, and by calculating the fraction of ALIVF residues. Both nuIMM and mtIMM TMHs are found to be significantly less hydrophobic than the E. coli TMHs (p ≈ 0.002). Thr is reported to be enriched in nuIMM and mtIMM TMHs relative to E. coli TMHs (as are Asn and Pro, while Val and Phe are reduced relative to E. coli TMHs, panel C). Panels F and G show an enrichment of Thr residues in nuIMMs of marginal significance (p = 0.02), and an enrichment in mtIMMs of higher significance (but the statistics are skewed by one mtIMM TMH that apparently contains almost 40% Thr residues). The substitution matrix in panel H indicates that aliphatic I, L, A, and G residues in the E. coli TMHs are often substituted by Thr and Val in the mitochondrial proteins. F does not show this substitution pattern, questioning the use of ALIVF counts in the analysis.

(2) The authors calculate hydrophobicities and amino acid frequencies for all human TMHs found in AlphaFold2 structures and grouped according to their cellular location: ER, Golgi, plasma membrane, nucleus, outer mitochondrial membrane (OMM), nuclearly encoded proteins in the inner mitochondrial membrane (nuIMM), and mitochondrially encoded proteins in the inner mitochondrial membrane (mtIMM). These data are presented in Figure 3 (and supplements).

OMM and nuIMM TMHs are significantly less hydrophobic than the secretory pathway TMHs, whereas mtIMM TMHs seem to have similar hydrophobicities as the secretory ones. nuIMM TMHs may have a slightly increased Thr content compared to secretory pathway proteins, but only one p-value is significant (Golgi vs nuIMM). mtIMM TMHs appear to have a higher Thr content than nuIMM proteins, but the statistics is skewed by a single point at a Thr fraction of 0.25. The Thr enrichment is less pronounced than for the E. coli comparison in Figure 2, Figure 3 Supplement. Fig. 8 shows that TMHs in chloroplast and mitochondrial plant proteins both have reduced hydrophobicity compared to ER-targeted proteins.

(3) Starting from the 66 mitochondrially encoded proteins in Jakobids, the authors identify homologous proteins that have become nuclearly encoded and carry an N-terminal mitochondrial targeting sequence (MTS) in organisms with smaller mitochondrial genomes. As in 1, the authors then do structure-based sequence alignments of the 63 homologous nuIMM and mtIMM TMH pairs found, and compare the individual aligned TMH pairs in terms of hydrophobicity and Thr content. These data are presented in Figure 4 (and supplements).

The reduction in hydrophobicity in nuIMM vs mtIMM TMHs upon relocation to the nucleus is robust and confirms earlier publications (panel C). The increase in Thr content in nuIMM vs mtIMM TMHs is less convincing (p = 0.012, panel G) and is opposite to what is seen in Figure 3C.

(4) By phylogenetic analysis, the authors identify three independent relocations of the ATP6 gene to the nuclear genome, do structure-based sequence alignments of nuIMM and mtIMM pairs representing these three events, and illustrate the reduced hydrophobicity of the TMHs in the nuclearly encoded proteins by hydrophobicity plots. These data are presented in Figure 5 (and supplements).

In panel C, it is not only aliphatic-to-Thr substitutions that reduce the hydrophobicity of the nuATP6 TMHs. How much is the hydrophobicity reduced if the Thr mutations are reversed (i.e., how much does the hydrophobicity reduction depend on ATS specifically and not on other kinds of substitutions)?

In a short experimental section, the authors go on to show that a mitochondrially encoded ATP6 protein can be converted into a nuclearly encoded, import-competent protein using ATS to reduce the hydrophobicity of its TMHs (but functionality is not tested) and, conversely, that import of a nuclearly encoded ATP6 protein is blocked when Thr residues in its TMHs are mutated to aliphatic residues (Figures 6, 7).

Finally, the authors carry out all-atom MD simulations to study the energetics of H-bond formation between Thr side chains and TMH backbone atoms.

In summary, the reduction in TMH hydrophobicity in nuIMMs compared to mtIMMs (and to secretory pathway membrane proteins) is robust and nicely confirms earlier studies. The purported enrichment in Thr content in both nuIMMs and mtIMMs has weaker statistical support in the data: p-values are typically > 0.01 (or may be skewed by an outlier) and appear not to have been corrected for multiple testing in all cases. As is clear from the ATP6 data in Figure 5, aliphatic -> Thr substitutions explain only a fraction of the overall hydrophobicity reduction in nuIMM TMHs, and the role of other hydrophobicity-reducing mutations in gene relocation could be better defined.

Reviewer #2 (Public review):

Summary:

The authors set out to investigate how well the onset of a self-initiated movement could be predicted at different times prior to action onset. To do so, they collected EEG and MEG data across 15 human participants who watched natural landscape images on a screen. These participants performed active self-initiated movements, or observed passive actions, to have a new image appear. By comparing the neural activity prior to active and time-matched passive actions, the authors found that even though a build-up of neural activity is visible close to 1s prior to action, action onset could only be reliably predicted around 100ms prior to action. These results confirm what was already suggested in previous literature: the commitment to action is only clear from the late stages in the visible neural ramp-up to action onset.

Strengths:

(1) The paper presents a well thought out methodology to assess the predictive value of neural activity prior to a self-initiated movement and passively observed action, while keeping all other experimental factors identical. This methodology can be applied outside the specific scope of this paper as well, in efforts to assess the correspondence of a neural signature with an observed behavior.

(2) The results are a strong confirmation of what was suggested less clearly in previous research (Trevena & Miller, 2010, Consciousness & Cognition; Schmidt et al., 2016, Neuroscience & Biobehavioral Reviews; Travers et al., 2020, NeuroImage).

Comments on revised version.

I thank the authors for addressing my concerns in the Introduction and Discussion of their paper. I don't require any further edits at this point.

Joint 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:

This manuscript couples a 32-parameter model with simulation-based inference (SBI) to identify parameter changes that can compensate for three canonical hyperexcitability perturbations (interneuron loss, recurrent-excitatory sprouting, and intrinsic depolarisation). The study demonstrates a careful implementation of SBI and offers a practical ranking of "compensatory levers" that could, in principle, guide therapeutic strategies for epilepsy and related network disorders.

Strengths:

(1) By analysing three mechanistically distinct hyper-excitable regimes within the same modelling and inference framework, the work reveals how different perturbations require different compensatory interventions.

(2) The authors adopt posterior estimation to systematically rank the efficiency of different mechanisms in balancing hyperexcitability.

(3) Code and data are available.

Reviewer #1 (Public review):

Summary:

A fundamental technique for the identification of peptide-specific CD8 T cells is the use of fluorophore-conjugated and peptide loaded MHC tetramers. Classically, refolding of specific peptides with MHC monomers can be labour intensive, and not optimal for screening large numbers of different peptides. Hence, UV-exchanged tetramers have been developed to upscale this, however, still has some associated challenges such as UV-mediated damage to peptide complexes. Here, Pothast, C.R. et al demonstrate the efficacy of using temperature exchanged tetramers for the prevalent alleles HLA-A*03:01, A*11:01, B*07:02, and C*07:02. Building upon their previous work with HLA-A*02:01, H-2Kb, and HLA-E. They first demonstrate the complex stability of tetramers with different affinity peptides at high temperature, showing complex destabilisation can be rescued with higher affinity peptides. This is followed by an optimisation of peptide exchange temperatures, tailored for each allele. The authors then demonstrate successful binding to clonal T cell lines, and then a step further with viral peptides against PBMCs from individuals with confirmed infection history. For the latter they compare to conventional tetramers and demonstrate comparable signal.

Due to the prevalence of these 4 alleles, the ease-of-handling, and short time requirements, these tetramers are likely to show high utility.

Strengths:

The manuscript is well-written and results are solid, although more detail may add clarity to some of the results, in particular figure 1 and 2. Other than the points reported below, the study uses accurate controls to demonstrate the specificity of the tetramers, and the data is convincing.

Overall, their interpretation of the results is accurate and the discussion thorough, additional comments may be included to cover potential tetramer batch variability, and differences in the stability of different alleles. Specifically, whether certain alleles require higher affinity peptides to be stable, compared to others.

Comments on revised version.

The authors addressed the weaknesses highlighted in the previous version, specifically addressing that they do not conclude that thermal peptide exchange is not superior to existing technologies, but may be more feasible, as well as including a section on the stability of the different alleles and template peptides. The authors also updated the labelling and description of figures, making them much more interpretable. Addition of the LILRB1 reporter cell line to demonstrate successful peptide exchange and stability adds extra validity to the results.

Reviewer #2 (Public review):

Summary:

This work highlights a new approach to build an array of highly stable peptide MHC-I using temperature exchange, which can be used to identify antigen specific CD8+ T cells.

Strengths:

In this work, the authors have proposed an alternative method to reload the peptide MHC-I molecule. Their temperature-exchange approach is distinct from current reloadable peptide MHC technologies involving photolabile peptide, empty MHC-I (Nat Commun 11, 1314 (2020). https://doi.org/10.1038/s41467-020-14862-4), tapasin/TAPBPR chaperone-assisted (eLife 7:e40126.), enzyme exchangeable (https://pubmed.ncbi.nlm.nih.gov/31201791/) and small alcohol (Curr Res Immunol. 2022 Aug 18;3:167-174. doi: 10.1016/j.crimmu.2022.08.002) approaches

Weaknesses:

The temperature-exchange approach identifies antigen-specific T cells at the same frequency of the photolabile peptide MHC-I molecules. The time saved using temperature-exchange protocol may not be a pull factor as the photolabile peptide MHC-I approach is not unreasonably laborious.

As with other exchange systems, performance can vary by HLA allele and complex stability, so each new allele needs empirical optimization.

In short, temperature-based pMHC exchange is a practical and fast loading strategy with good utility for multimer generation, but its main limitation is dependence on thermal stability and careful template selection.

Reviewer #1 (Public review):

This work evaluates the impact of reproductive history on growth, body weight and body composition in mammals. In mice, somatic growth is stimulated by the first pregnancy while the second pregnancy increases body weight mainly by increasing adiposity. To probe the role of pituitary growth hormone (GH), the key regulator of somatic growth in these processes, this was addressed by comparing the impact of reproduction on growth in normal ("wild type") and genetically GH-deficient females and by detailed characterization of the profile of fluctuations in circulating GH levels in both types of animals. Additional studies addressed the possible role of other endocrine pathways (ghrelin and estrogen) in the pregnancy-related growth. Surprisingly, reproduction-related growth was independent of GH, ghrelin and estrogen. To determine whether these results may apply ("translate") to human physiology, data on various parameters of somatic growth were collected from women with hereditary GH deficiency. The findings indicate that GH-independent stimulation of growth by reproductive events also occurs in women.

Use of multiple animal models, rigorous characterization of GH levels in normal and GH-deficient females, and inclusion of data derived from a unique and well-characterised population of people with hereditary isolated GH deficiency and no GH replacement therapy are important strengths of these elegant and innovative studies. The results address a broader and clinically significant issue of permanent changes in body size, composition and function that result from pregnancy and lactation. This work also provides important background for further studies aimed at the identification of the mechanism involved and the role of specific reproductive events in the regulation of growth.

Reviewer #2 (Public review):

This manuscript describes the fascinating phenomenon of growth hormone (GH)-independent growth occurring in the mother during pregnancy. This growth was most pronounced in dwarf mice that are lacking the receptor for growth hormone-releasing hormone (GHRH) and therefore showing isolated GH deficiency. However, the pregnancy-induced growth could also be observed in wild-type mice, suggesting that it is a normal part of the maternal adaptation to pregnancy. The study falls short of identifying the mechanism(s) driving this pregnancy-induced growth response, but it certainly reveals a novel insight into maternal physiology. The authors have completed a range of experiments in mice to prove that, as well as being GH independent, the pregnancy-induced growth also did not require GH signaling in the liver (i.e., not another pregnancy-specific ligand operating through the GHR to promote IGF). They also provided complementary data from a population of humans with untreated isolated GH deficiency that are broadly consistent with the hypothesis. While it is important to consider the significant species differences between rodents and humans, both in terms of growth physiology and also in terms of evolution of placental somato-mammotrophic hormones, this unique population is a valuable resource and adds credence to the study. Overall, I find this a compelling research story, but somewhat unfinished. There are some areas where additional information could improve the ability to interpret the data, and some additional concepts that could be considered in the discussion. There are also areas where additional experiments might provide important insights. However, I think that such suggestions can be considered as appropriate for future research, rather than delaying the dissemination of the findings in the current manuscript.

Reviewer #3 (Public review):

Summary:

The study describes an increase in body growth and body composition in both mice and women. In mice, the impact on growth is mainly seen during the first pregnancy, and the changes postpartum on body composition are also different during the first and second pregnancies. The study has used various knockout models in the growth hormone axis to understand these changes as well as some gene expression analysis related to GH, IGF-1 and estrogen signalling pathways.

Strengths:

(1) The inclusion of various knock-out mouse models that allow for exploration of mechanisms related to the above-mentioned changes.

(2) The investigation of gene expression of GHR, IGF-1R and ER pathways.

Reviewer #1 (Public review):

Summary:

Bussey-Sutton, Gray, Wu et al. combine RelB proximity-labeling proteomics (BioID) with a custom, BioID-informed HIV-CRISPR screen to map the chromatin/transcriptional network that links non-canonical NF-κB (RelB/p52) activation by the SMAC mimetic AZD5582 to HIV-1 latency reversal. They identify a broad basal RelB interactome (chromatin remodelers, RNA-binding proteins, transcription machinery) that is partially reshaped, rather than replaced, by AZD5582 treatment. A targeted CRISPR screen against the BioID hit list then functionally separates these interactors into pro-reactivation factors (KAT7/HBO1, NSD2, SIN3A) and restrictive factors (p300, CHD4, USP7, UHRF1, DNMT1).

Pharmacological inhibition of the two acetyltransferase hits (KAT7i and p300i) validates opposing phenotypes predicted by the screen, and a comparative TNF-α (canonical NF-κB) screen argues that at least the p300 and, more selectively, the KAT7 pharmacology are biased toward the ncNF-κB axis. The dataset is a useful resource, and the BioID, CRISPR, pharmacology pipeline is a nice template for functionally triaging proximity-proteomics hits rather than treating them as a static interactome.

Strengths:

This study takes an integrative approach in which proximity proteomics is used to define candidate RelB-associated regulators, a purpose-built targeted CRISPR library tests these candidates directly for a functional phenotype, and pharmacological inhibition cross-validates the two acetyltransferase hits with an orthogonal, catalytic-activity-specific perturbation. This funneling from unbiased proteomic discovery to targeted functional testing to mechanistic pharmacology is an efficient and logical pipeline for converting a large interactome into a smaller set of high-confidence, functionally validated regulators.

The comparison of basal versus AZD5582-stimulated RelB interactomes is conceptually useful, as it shows that RelB engages a largely pre-existing chromatin and transcriptional regulatory scaffold upon ncNF-κB activation, rather than recruiting an entirely new complex. This is a finding with potential implications for non-canonical NF-κB signaling beyond the HIV latency context specifically.

The inclusion of a parallel HIV-CRISPR screen using TNF-α to activate canonical NF-κB signaling is a valuable control, allowing the authors to distinguish regulators that are broadly required for NF-κB-dependent HIV expression from those more selectively linked to AZD5582-driven ncNF-κB signaling. This comparison strengthens the authors' claims of pathway specificity for hits such as KAT7 and p300.

Finally, pharmacological validation of the top acetyltransferase hits was performed across two independent latency models (J-Lat 10.6 and N6), which helps address, at least in part, the concern that phenotypes are specific to a single clonal integration site. Together with the deposited proteomics and screen datasets, this provides a genuinely useful resource for the field.

Weaknesses:

(1) BioID inherently reports both direct and indirect (proximal) interactions, and no individual hits are validated by an orthogonal method (e.g., co-IP, PLA). As a transcription factor, RelB would be expected to reside near transcriptional/chromatin machinery, so this enrichment alone isn't surprising, making it hard to distinguish functionally relevant RelB complexes from bystander proteins in the same active chromatin neighborhood. At least a subset of top hits should be validated independently.

(2) BioID interactome was defined in a single clonal cell line, limiting generalizability across latent cell types. The RelB BioID proteomics, which forms the entire basis for the downstream targeted CRISPR library generation, was performed in one engineered Jurkat N6 clone. Both RelA/RelB pathway dependency and the baseline expression of the chromatin regulators identified as RelB interactors (e.g., DNMT1, UHRF1, SWI/SNF subunits, NSD2) can vary considerably across cell lines, latency clones, and primary CD4+ T cells. Since the CRISPR screen, and thus all functional hits in the paper, is entirely constrained by this single-cell-line interactome, any RelB interactor that is absent or differently regulated elsewhere would never have been tested, regardless of its true biological importance. To support claims of a broadly relevant RelB/ncNF-κB regulatory network, the authors should at least partially validate the basal and AZD5582-induced interactome (e.g., by co-IP/western for key hits) in a second cell line or primary CD4+ T cells, to distinguish clone-specific interactions from generalizable ones.

(3) Is KAT7/HBO1's contribution to reactivation a direct interaction/synergy with AZD5582, or an independent parallel requirement? The data show that KAT7/HBO1 knockout and catalytic inhibition (WM-3835) both reduce AZD5582-induced reactivation, and that KAT7 was detected as AZD5582-enriched in the RelB BioID. However, it remains unclear from the current data whether KAT7/HBO1 acts as a direct downstream effector in the ncNF-κB/RelB signaling axis (i.e., is recruited by or acts together with RelB/p52 at the LTR specifically upon AZD5582 stimulation), or whether it represents a more general/parallel chromatin requirement for HIV reactivation that is simply necessary regardless of the activating stimulus (and happens to also associate with RelB). Distinguishing these possibilities matters for interpretation: the former would support a model in which AZD5582 and KAT7 act synergistically through the same pathway (i.e., acetylation is mechanistically coupled to ncNF-κB activation), whereas the latter would suggest KAT7-mediated acetylation and RelB/p52 signaling are two independent, simultaneously required mechanisms that converge on the same locus without being mechanistically linked. The comparative TNF-α data (Figure 3C), where the genetic screen shows a shared dependency but the pharmacological inhibitor does not phenocopy this for TNF-α, adds to this ambiguity rather than resolving it. Clarifying whether KAT7/HBO1 functions upstream, downstream, or in parallel with RelB/p52 (e.g., via RelB ChIP after KAT7 knockdown, or KAT7 recruitment kinetics relative to RelB/p52 at the LTR) would strengthen the mechanistic claims in this section.

(4) A major limitation of the study is that screen results were validated in only two cell line models, and neither is a primary cell system. Different latency models are known to have distinct dominant transcriptional blocks, not just different genetic backgrounds. J-Lat clones, U1, ACH2, and other transformed lines differ substantially in why the provirus is silent; some are more Tat/P-TEFb-limited, some more chromatin/Polycomb-restricted, some more dependent on transcription factor availability (NF-κB, NFAT, AP-1). A regulator that is rate-limiting in J-Lat 10.6 specifically (a single integration site, clonally selected for a particular chromatin environment around the LTR) may therefore be irrelevant, or could even act in the opposite direction, in a model where the block is elsewhere.

(5) HIV latency and LRA responsiveness are well known to be strongly integration-site and clone-dependent. Pharmacological validation of CRISPR screen hits extends to N6, but the primary screen hits themselves are not confirmed in a second genetic background. This is an important caveat that should be discussed explicitly, and ideally at least the top hits (KAT7, NSD2, SIN3A, p300, CHD4) should be individually knocked out (not just inhibited) in a second model to confirm the genetic phenotype generalizability, not just the pharmacology.

(6) More importantly, no validation was performed in primary CD4+ T cells or ex vivo resting CD4+ T cells from people with HIV, which remain the field's benchmark for confirming that a candidate regulator is relevant to authentic latency rather than to a transformed cell line. Basal chromatin state, NF-κB pathway, and expression of the chromatin regulators identified here (e.g., DNMT1, UHRF1, SWI/SNF subunits) can differ substantially between transformed lines and primary resting T cells. Given this, the current claims of a broadly applicable "chromatin regulator" network should be tempered, and at least the top hits (e.g., KAT7/HBO1, NSD2, SIN3A, p300) should ideally be tested in a primary cell model before being generalized as regulators of HIV latency and reactivation more broadly.

(7) The finding that p300 restricts AZD5582-driven reactivation is framed as consistent with Horvath & Sadowski (ref 36), but p300/CBP is also a well-established Tat/LTR coactivator in many contexts (Kiernan et al., Marzio et al). The Discussion would benefit from a slightly more developed treatment of how these findings can be reconciled rather than citing a single concordant paper.

Reviewer #2 (Public review):

Summary:

Bussey-Sutton et al. investigate the RelB interactome in the presence and absence of the SMAC mimetic AZD5582. They combined RelB BioID with CRISPR screening and highlighted multiple potential transcriptional regulators that may contribute to the maintenance or reversal of HIV latency. However, there are several technical limitations and gaps in validation in the current version of the work.

Strengths:

The overall topic is of clear interest; the RelB interactome and CRISPR screens provide interesting hits.

Weaknesses:

From a technical standpoint, RelB is expressed using a doxycycline-inducible system, yet the fusion protein (Figure S1A) appears to be expressed at high levels even in the absence of doxycycline (and does not seem to respond to doxycycline supplementation). As this is the key model system used in the paper to identify RelB interactors, this limitation should be discussed and potentially addressed.

It would also be important to include some orthogonal validation (e.g., co-immunoprecipitation) of top novel RelB interactors identified by BioID.

In Figure 2C, two non-targeting controls appear among the most highly enriched hits. This raises the question that other top hits might reflect nonspecific effects or assay noise. Consistent with this concern, the effects observed following siRNA-mediated depletion of the top CRISPR-screen candidates in Figure 2D and Figure 3 appear relatively modest.

More broadly, the absence of validation in primary-cell models is a further limitation, as the chromatin environment and transcriptional regulation of HIV latency in the two cell-line models used here are likely to differ substantially from those in primary CD4+ T cells.

Minor point:

Endogenous RelB remains present in the RelB-BioID cells and could compete with the miniTurbo-RelB fusion protein for physiological binding partners. Moreover, the effect of two sources of constitutive RelB expression (see comment above) on the physiologic RelB interactome is not obvious. Controlling for this is not trivial, but this potential confounding factor should be acknowledged (and further reinforces the need for orthogonal targeted validations of interactors in cells with endogenous RelB).

Reviewer #3 (Public review):

The study is built on the premise that the SMAC mimetic AZD5582 can reverse HIV latency via the RelB-mediated non-canonical NF-kB pathway activation; however, its molecular regulators remain undefined. The authors use an inducible BioID RelB reporter system in Jurkat N6 cells containing an integrated HIV-1 provirus and identify both the basal and the AZD5582-activated chromatin and transcriptional regulatory network of the non-canonical NF-kB pathway. To further explore the role of the identified regulators, an HIV-CRISPR latency screening approach was used, and based on enrichment or depletion of guide RNAs in the supernatant, genes that inhibited or activated virus reactivation, respectively, were identified. Genes like LRPPRC (RNA-binding protein), KAT7 (acetyltransferase), NSD2 (H3L36 methyltransferase), SIN3A (part of histone deacetylase complex), and others were identified to be essential for HIV reactivation. Nearly half of the identified factors were found to be negative regulators of reactivation, notably CBP/p300 (acetyltransferase), CHD4, USP7, UHRF1, and DNMT1. The roles of KAT7 and p300 were validated through the use of specific inhibitors. To narrow down regulators that are specific for nc NF-kB HIV reactivation vs NF-kB dependent HIV expression, a parallel CRISPR screen was performed with TNF-alpha, which reconfirmed the role of RelB, KAT7, and p300 in AZD5582-mediated HIV reactivation.

Overall, this is a solid and well-designed study that combines complementary approaches to deliver a valuable resource.

The study's strength lies in integrating two high-throughput approaches of RelB proximity labelling and targeted CSIPR screening to identify molecular regulators of the HIV reactivation pathway. A parallel CSIPR screen with TNF-alpha provides key insights into pathway-specific vs shared regulators of latency reversal.

The limitation of the study, as indicated by the authors, is that the assays are based only on HIV output. Additionally, some of the identified interactors may be indirect or context-dependent and may not represent all HIV reservoirs in vivo.

Reviewer #1 (Public review):

Summary:

Involved in various cellular pathways and processes, Ser/Thr kinase CK1δ is thought to be constitutively active and its tail autophosphorylation is suggested as a putative inhibitory mechanism of CK1δ kinase activity. Here, the authors investigate CK1δ's phosphorylation status, in relation to its location and dynamics throughout the cell cycle. Immunofluorescence and biochemistry studies showed that the subcellular distribution of CK1δ is dynamic and in equilibrium between centrosomal and nuclear pools. The authors argue that CK1δ phosphorylation protects it from degradation. Finally, the authors show differences in CK1δ phosphorylation status and location throughout the cell cycle. Combining their findings with the existing literature, they propose a model of CK1δ phosphorylation status, location and dynamics throughout the cell cycle.

Comments on latest version:

The authors have unfortunately taken a very defensive mode There are two themes that authors resort to: 1) We have published previously with some reagents - therefore these are well-established and we do not need to include controls anymore; 2) Independent biological replicates and statistics are only appropriate in some places, again because certain observations have been previously published (by the same authors) and well-established. In my opinion, this is not how rigorous research should be done. Specially with IF, it is essential in every experiment to have robust negative controls.

Reviewer #2 (Public review):

Summary:

In this study, Serrano et al. employed a combination of cell biological and molecular approaches to investigate the localization and regulation of Casein Kinase CK1 during the cell cycle using U20S cells. They show that CK1 dynamically localizes between the centrosomes and the nucleus but can be sequestered away from the centrosomes upon overexpression of its binding partner PER2. They provide evidence that CK1 WT but not a phospho-null mutant strongly accumulates in a hyperphosphorylated form upon inhibition of phosphatases (using Calyculin and Okadaic Acid) and thus conclude that CK1 tail phosphorylation protects the kinase from degradation. Using synchronized cells, they show that CK1 accumulates unphosphorylated in S-phase (APC/Cdh1 inactive) but phosphorylated at the G2-M transition. Immunostaining shows that CK1 localizes to the centrosomes during mitosis.

Strengths:

The authors reveal that the activity and abundance of dephosphorylated and phosphorylated CK1δ are regulated in a cell cycle-dependent manner. This suggests that these different pools are associated with distinct physiological functions.

Comments on latest version:

Unfortunately, problems with controls remain in the revised manuscript. At the very least, the authors should include all the necessary controls in their study instead of referring to previously published papers. I have asked for a detailed description of the kinase construct in which the phospho-Serine and Threonine of the C-terminal tail are substituted, but I did not see any more information in the revised version.

Reviewer #1 (Public review):

The study addresses the organisation of synaptic connections from medial to lateral entorhinal cortex. Classic anatomical work has suggested these connections exist but very little is known about their identity or functional impact. The manuscript argues that these projections are mediated by glutamatergic neurons, providing excitatory input from MEC to all layers of LEC, and by SST+ve interneurons sending inhibitory projections to L1 of LEC. This appears the most likely interpretation of the data. Potential concerns about confounds due to spread of virus/tracer from the injection site are addressed in the supplemental figures. My view is that the weight of evidence favours the authors' interpretation although the evidence isn't quite compelling.

Knowing the configuration of projections from MEC to LEC is important for thinking about circuit mechanisms for spatial cognition and episodic memory. This study adds to an emerging view that MEC and LEC can interact directly, indicating that the cell-type level organisation of these interactions is asymmetric and identifying an intriguing long range inhibitory pathway.

Reviewer #2 (Public review):

Summary:

The manuscript by Nilssen et al. presents a comprehensive study of the circuitry linking the medial and lateral entorhinal cortices (MEC and LEC). Using a combination of anatomical tracing, optogenetics, and in vitro electrophysiology, the authors convincingly demonstrate that the MEC sends both glutamatergic and long-range inhibitory SST+ GABAergic projections to the LEC, with distinct laminar and cell-type-specific targeting. Notably, they reveal that SST+ inhibitory projections selectively suppress the activity of layer IIa neurons, whereas excitatory inputs preferentially engage neurons in layers IIb and III, thereby differentially modulating hippocampal-projecting populations.

Strengths:

The experiments are carefully executed, the results are compelling, and the conclusions are well supported by the data. This work will be of broad interest to researchers studying memory circuits, cortical inhibition, and the organization of long-range connectivity.

Weaknesses:

Although the in vivo relevance of these connections remains to be determined, this is an important and timely contribution to our understanding of entorhinal-hippocampal interactions.

Comments on revised version.

The authors have addressed my comments satisfactorily, and I am satisfied with the changes made in the revised version.

Reviewer #1 (Public review):

The authors describe a learning algorithm that is meant to deal with the problem of inference in large, complex environments. They argue that simpler but ensemble learning systems with heterogeneous learning rates often perform better than more complex approaches. They also relate the ideas of their model to a circuit implementation in the mushroom body, an associative learning area in insects.

I found the argument about the value of sublearners remaining simple to be thoughtful and intriguing. I also appreciated the connection to biology and, although there remain details to be worked out, it seems that the proposed approach has what I'd call a "biologically plausible" instantiation.

Comments on the major claims:

The paper claims that algorithms that learn stochasticity and volatility "necessarily result in worse short-term performance than Hetlearn." The results with Autocovariance Least Squares and the model of Piray and Daw are consistent with this statement, but the authors seem to be making a more generic claim. Is there a rigorous justification of this? In particular, I found the section "Competing sublearners should stay simple" to straddle a line between opinion and fact, and I wasn't sure where it intended to be regarding that line.

Relatedly, regarding the statement just before that section: "The exponential convergence of Hetlearn weightings stands in contrast to the slow convergence of stochasticity/volatility estimates (Figures 2A and C), which result in poor initial performance of the Kalman learner." Is the claim here that the Kalman learner doesn't converge exponentially, or that it converges more slowly?

The paper would benefit from a comparison to other models of multiple learning systems with different timescales. The classic one would be the complementary learning systems ideas about hippocampus and neocortex (McClelland, McNaughten, O'Reilly 1995), and the many subsequent follow-ups. Two other relevant ones to consider are Roxin & Fusi 2013 and Lindsey & Litwin-Kumar 2024, which also consider multiple memory systems with different plasticity rates.

Comments on clarity:

(1) Page 2: There could be some more unpacking of the relationship between x and z in equations 1a, 1b. If z_t are time-varying, shouldn't theta, if it represents what is learned about "large-world factors," also depend on time?

The description of second-order conditioning on page 6 is confusing. Second-order conditioning tests whether S2 acquires a positive predicted value after pairing with S1, which had previously been paired with reward. The text makes it seem like the goal of second-order conditioning is to learn S2->S1->no reward.

(2) Page 6: "Representations decay on a slower timescale than stimulus presentation. Hence, the representation x2 after experiencing S2 → S1 holds traces of the S2 representation." I couldn't understand this. Maybe an equation would help?

(3) Page 7: The phrase "unconditioned stimulus" is introduced without being described. Is this a reward or punishment? What does it have to do with valence lower-case v, or value upper-case V? The authors seem to mix notations from different subfields, leading to a rather confusing description of the model.

Comments on relationship to biology:

(1) Page 6: It is assumed that Kenyon cells interact recurrently through plastic weights. This is an unconventional assumption for mushroom body models, and it is made without much discussion. The authors cite papers that argue for changes in the calyx during learning (although these effects are generally much weaker than the well-documented plasticity onto MBONs). However, KC-to-KC interactions are largely in the lobes, not the calyx. The authors should provide more discussion of this assumption. What happens if the recurrence is weak or absent? A relevant reference to consider is Manoim et al. Current Biology 2022, which shows that KC-to-KC interactions are inhibitory.

The model seems to imply that each MBON compartment encodes only one memory, or did I misunderstand? Is this necessary? Typically, it is thought that because of the sparse, decorrelated representations in the KCs, an individual MBON could store multiple associations without too much interference.

Reviewer #2 (Public review):

Summary:

The authors propose Hetlearn, a novel learning algorithm for unstable environments: instead of using statistically optimal methods to estimate the variability of the environment, simply have a range of "sub-learners" that learn at different rates, and rely more heavily on the sub-learners with a recent track record of success. This strategy is particularly helpful when the environment has unknown and fluctuating stochasticity (randomness around a constant mean reward) and volatility (the mean itself changes over time). For high stochasticity and low volatility, it's better to learn slowly to smooth out randomness; for low stochasticity and high volatility, it's better to learn quickly to catch onto the new trend. The authors show that existing methods work well when stochasticity and volatility are constant, or when methods can learn them over a long time, but that Hetlearn does better when stochasticity and volatility change over time or an agent has limited time or data to learn from.

They then propose that the fly mushroom body implements a form of Hetlearn, since it has different compartments that form parallel memory traces at different learning and forgetting rates. Here, the conclusions are less clear because their model doesn't appear to implement the key feature of Hetlearn (the differential weighting of sublearners based on recent success), and because their model has features that don't have an obvious biological equivalent in the real mushroom body.

Strengths:

The idea behind Hetlearn is clever and interesting, and it makes sense that a heuristic method like Hetlearn would outperform a theoretically optimal algorithm under conditions of limited data. The model is clearly defined, and they test it across a range of tasks and compare it to a range of different Kalman/particle filter learners. The application to the mushroom body is also an inspiring connection and takes advantage of the discovery of empirically described sublearners. The work will stimulate new ways of understanding how biological systems might implement heuristic tricks that may not be optimal under ideal conditions, but that succeed in "real world" conditions, and it will stimulate new thinking about the computational role of parallel memory traces in the mushroom body.

Weaknesses:

In the tests of Hetlearn vs. Kalman filters, it would be useful to more fully explore the parameter space. Specifically, in Figures 1-2, it would help define the boundaries of when Hetlearn is better or worse than Kalman filters, by systematically changing the parameters of the task (e.g. increasing/decreasing the amount of data available to the learner or the frequency of changing stochasticity/volatility), and changing the parameters of Hetlearn (the learning rates of the sublearners and the recency bias, i.e. how far back to track the recent success of the sublearners when weighting their outputs). For Figure 3, while they do check the performance when varying the recency bias, there are no conditions presented in which the particle filter does better. To define when Hetlearn is better, it is important to show when it is worse.

In the mushroom body model, there are several elements that have no apparent biological equivalent, and it's not clear whether they are incidental (in which case they could be removed to enhance biological realism) or essential to the model's performance (in which case the discrepancy should be discussed; perhaps the model can be reframed as a prediction for as-yet-undiscovered features). In addition, the model seems to lack the defining feature of Hetlearn (up/down-weighting sublearners according to their recent performance). Finally, the authors argue that Hetlearn explains experimental data better than existing models, but existing models could be easily and parsimoniously tweaked to accommodate these data, so the explanatory advantage of Hetlearn is unclear.

Reviewer #2 (Public review):

Summary:

The study by Millard et al. investigates the effects of nicotine on peak alpha frequency and pain using a highly elaborate experimental design. According to the statistical analysis, the authors found a small covariate-corrected effect for prolonged heat pain and a small effect on global peak alpha frequency in response to nicotine treatment. However, the effect on heat pain was not found for pressure pain, was not mediated by peak alpha frequency, and appears mainly based on a worsening in the placebo condition rather than a clear reduction of pain after nicotine.

Strengths:

I very much like the study design and that the authors followed their research line by aiming to provide a complete picture of the pain-related cortical impact of peak alpha frequency. This is important work, even in the absence of clear statistical or mechanistic support for the main hypothesis. I also appreciate the preregistration of the study, the well-written and balanced introduction, and the additional analyses added during revision.

Weaknesses:

The weakness of the study revolves around two main aspects:

(1) Source separation would still have been more appropriate than electrode regions or global peak alpha frequency to extract the alpha signal. The added ICA analysis is appreciated, but in my view, it does not fully solve the problem. Different alpha sources, such as mu rhythm, occipital alpha, and lateralised alpha activity, are still not clearly disentangled. The main peak alpha frequency effect remains a small global, wide-band, sensor-space effect rather than a clearly source-specific sensorimotor effect.

(2) I still think the choice of nicotine is a major weakness of the study. The rationale for using nicotine as an intervention rests partly on previous reports of nicotine effects on experimental pain, but the most relevant cited meta-analysis already suggested that these effects are small and may be affected by publication bias. It is therefore not a clean way to test whether peak alpha frequency has a causal role in pain. The heat-pain result is also difficult to interpret as analgesia. Statistically, the placebo-controlled contrast may be significant, but in practice, the nicotine condition changed very little, and the effect seems mainly driven by worsening in the placebo group. In a real intervention, one gives the drug, not the difference between drug and placebo.

Impact:

The impact of the study could be to show what did not work to answer the authors' research question. The study is valuable because it provides a careful and largely negative mechanistic test: nicotine produced at most small effects, peak alpha frequency did not mediate pain, and pressure pain was not affected. The study would have more impact with a more direct intervention model and an analysis strategy that more clearly untangles the different alpha sources.

Reviewer #1 (Public review):

In the article, the authors set out to characterize in adipocytes an enhancer, ENH5, of the gene EPAS1, a gene that was found to show strong selection in Tibetan populations. They investigate whether this enhancer contributes to adipocyte response to environmental stress. The authors show that ENH5 is active in preadipocytes and that the Tibetan high-altitude allele confers reduced activity. They then use a mouse ENH5 knockout model to show a hypoxia/thermogenesis responsive phenotype of stronger transcriptional downregulation of aerobic respiration, electron transport chain, and adipogenesis pathways. The authors interpret these findings as evidence that ENH5 conditionally regulates adipocyte energetics and thermogenic response, potentially favoring energy conservation in Tibetans exposed to the demands of high-altitude hypoxia and cold. Overall, the paper presents an interesting potential connection between EPAS1-mediated high-altitude tolerance and energy metabolism; however, more work needs to be done to establish this connection.

Major comments:

(1) The authors use mouse ENH5 enhancer knockout (ENH5-KO) as the model of the Tibetan EPAS1 locus because the high-altitude allele of human ENH5 has lower transcriptional activity than the low-altitude allele (Figure 1A) and mouse ENH5 (musENH5) has enhancer activity (Figure 1D) in mouse preadipocytes. However, it is an overstatement to claim the functional role of Tibetan ENH5 haplotype only based on these data because musENH5 is neither identical to human ENH5 nor the murine high-altitude haplotype. The title should also be revised to better reflect the function of ENH5, like "An EPAS1 enhancer mediates hypoxic and cold response in mouse adipocytes". The authors should consider in some way to actually show that the Tibetan haplotype in ENH5-KO leads to expression changes. This could be done by inserting the haplotype into preadipocytes via CRISPR (realize this is a tough one) or if they have available cells from Tibetans or some eQTL or other similar datasets. The more closely they can connect this haplotype to EPAS1 expression, the more beneficial it would be for the article. As it stands, they currently have episomal luciferase assays showing reduction of enhancer activity in mouse preadipocytes of a human allele and a complete knockout of the mouse enhancer that doesn't recapitulate the Tibetan haplotype. A bit more work is needed to connect all of these to the Tibetan adaptation. As it stands now, this is all very circumstantial.

(2) In Figure 2A, the body weight of ENH5-KO normal diet is significantly lower until four weeks in male, and until 11 weeks and 18 and 19 weeks in female than that of WT. These are slight but significant differences between ENH5-KO and WT; therefore, the authors should describe and discuss this and how it could affect their results.

(3) For the mouse work, it is not clear why the authors did not do cold-exposure or some type of hypoxia experiment for the mice themselves. This will be helpful to support their claim, and if not done, or done without significant differences in the results, the authors should add and mention this. The RNA-seq work, while substantial, again provides circumstantial support.

(4) The authors used CL316243 as a β3-AR selective agonist to mimic thermogenesis in vitro. In humans, it is not β3-AR but β2-AR that mainly drives thermogenesis (Blondin et al., Cell Metabolism, 32, 287-300. e7). Therefore, the authors should describe the limitation due to the difference in mechanisms of action of thermogenesis between humans and mice, as they use the mouse cells as a human model.

(5) The authors note in the discussion that Figure 3's CL316243 stimulation intended to simulate a thermogenic reaction to cold temperatures also generated a change in OXPHOS pathways associated with hypoxia, thus making it difficult to separate the contribution of β3-adrenergic/thermogenic effect from an indirect local hypoxia response. The authors could further interrogate this effect by measuring canonical hypoxia-responsive genes, oxygen consumption, or performing an in vivo cold challenge.

(6) Figure 4: The authors mention that reduced aerobic respiration pathways are evidence for reduced thermogenesis, but it does not directly demonstrate altered thermogenesis correlates. They do not measure heat synthesis, oxygen consumption/respiration, uncoupled respiration, UCP1 protein levels or activity, mitochondrial changes, etc. Their evidence for changed thermogenesis stems from transcriptional changes in energy consumption pathways shared with hypoxia changes. They should tone down their findings.

(7) In the Discussion, the authors should interpret and discuss their data carefully. For example, in the GSEA analysis, the authors identified enriched gene sets in ENH5-KO. Therefore, the authors should discuss which genes might contribute to each pathway, since some genes show large logFC changes. In the current discussion, there is little mention of their results, and it mostly focuses on prospects.

Reviewer #2 (Public review):

Summary:

This study extends previous work on the adaptive EPAS1 locus by examining the pleiotropic activity of the ENH5 enhancer in adipocytes and its potential role in metabolic and thermogenic responses. The authors progress from demonstrating enhancer activity and evolutionary conservation to in vivo phenotyping and environmentally dependent transcriptional responses in primary adipocytes. The work provides an interesting example of how pleiotropic regulatory effects can contribute to the complexity of adaptation, with a single adaptive regulatory locus influencing multiple biological processes in an environmentally dependent manner.

Strengths:

The study is well executed and is clearly presented, with a logical experimental progression. A particular strength is the genotype-by-treatment interaction analysis demonstrating that ENH5 loss alters metabolic and adipocyte-associated transcriptional programs following both hypoxia and beta-3-adrenergic stimulation. The convergence of these responses is particularly interesting in the context of regulatory pleiotropy and suggests that selection at the EPAS1 locus may have consequences extending beyond the canonical hypoxia response. The inclusion of negative findings, including the absence of an overt baseline metabolic phenotype and the lack of an additive response to combined stimulation, also provides a balanced presentation of the results.

Weaknesses:

The principal conclusions are generally supported by the data, and the weaknesses are relatively minor and primarily relate to the scope of interpretation. The study assesses transcriptional programs associated with thermogenic signaling using CL316243 rather than directly measuring physiological thermogenesis. Nonetheless, CL316243 is a rational and well-established approach for experimentally inducting beta-3-adrenergic thermogenic signaling. In addition, the murine ENH5 knockout is a useful model of reduced enhancer activity that phenocopies the Tibetan ENH5 haplotype, but it is not genetically equivalent to the naturally occurring Tibetan ENH5 haplotype. The authors generally recognize these limitations, and they do not substantially detract from the central findings.

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors investigated factors required for neural progenitors to exit cell cycle before the adult stage. They first show that Kr is expressed in pupal stage MBNBs, and depletion of Kr from pupal stage NBs leads to retention of MBNBs into the adult stage. Then they demonstrate that these retained NBs maintain the expression of Imp, and co-depletion of Imp abolishes the extended neurogenesis. Further, they show that co-depletion of kr-h1 significantly reduces the retained MBNBs caused by loss of kr, suggesting antagonistic genetic interactions between these two. In addition, they demonstrate that over-expressing Kr-h1 leads to the striking phenotype of tumor-like neuroblast overgrowth in adult brains. In the revised manuscript, they provide evidence suggesting that Kr does not regulate Kr-h1 expression, but rather Kr and Kr-h1 may regulate the late temporal gene E93 expression in parallel.

Strengths:

(1) The authors leveraged well-controlled powerful genetic tools (including temporal control of RNAi knock down using the Gal80ts system), and provided strong evidence that Kr is required for the Imp to Syp transition and to promote the end of neurogenesis through E93 in mushroom body neuroblasts. Similarly, the experimental result of co-depleting Kr-h1 and Kr, and the striking phenotype upon Kr-h1 mis-expression, support the antagonistic roles played by Kr-h1 and Kr in this process.

(2) The sample sizes, quantification methods and p-values are well documented for all experiments. In most parts, the data presented strongly support their conclusions.

(3) Identification of two transcription factors with opposite roles in controlling cell cycle exit, and their possible interactions with the Imp/Syp axis, is highly significant for the study on how the proliferation of neural progenitors is regulated and limited before the adult stage.

Weaknesses:

(1) In the revised manuscript, Kr immunostaining was done in KrIf-1 mutant brains using tyramide signal amplification (TSA) to enhance the weak endogenous Kr signal, and they conclude that Kr is ectopically detected in MBNBs and their progeny. However, a wildtype control with TSA amplified Kr immunostaining was not provided, so it is not known whether the elevated signal comes from TSA amplification or the mutant condition. If Kr is indeed mis-expressed in MBNBs in the KrIf-1 mutant, the authors should provide a hypothesis to explain why both loss of Kr and mis-expression of Kr in MBNBs lead to the same NB retention phenotype.

Reviewer #2 (Public review):

In this paper, the authors study the role of Krüppel in regulating the termination and survival of mushroom body neuroblasts. They first confirm that adult wild-type brains have no proliferation and report that Kr mutants and Kr RNAi specifically in neuroblasts prolong MBNB lifespan, enabling continued neurogenesis in the adult brain. They further show that Kr acts during pupal stages to regulate the elimination to the mushroom body neuroblasts, as its depletion leads to neuroblast retention, revealing a previously unrecognised postembryonic function of Kr distinct from its established role in embryonic neurogenesis. Mechanistically, this is achieved by the regulation of the Imp/Syp transition and E93 expression. Finally, they show that Kr acts antagonistically to Kr-h1, which is expressed predominantly in larval stages. Together, these results identify Kr as a mushroom body neuroblast regulator coordinating intrinsic and extrinsic signal to regulate neuroblast termination.

Strengths:

The main strength of this paper is that it identified a novel regulator of Imp expression in the mushroom body neuroblasts. Imp is a conserved RNA binding protein that has been shown to regulate neural stem cell proliferation and survival in different animals.

Moreover, the authors provide a series of evidence that this regulator, Kruppel, acts by integrating extrinsic signals (ecdysone) with intrinsic timers (Imp-Syp transition) to regulate neuroblast termination.

Weaknesses:

No significant weaknesses were identified.

Reviewer #3 (Public review):

Summary:

Drosophila neuroblasts (NBs) serve as a well-established model for studying neural stem cell biology. The intrinsic genetic programs that control their mitotic potential throughout development have been described in remarkable detail, highlighting a series of sequentially expressed transcription factors and RNA-binding proteins that together constitute the temporal patterning system.

However, the mechanisms that limit the number of NB divisions remain largely unknown in a specific subset of NBs known as mushroom body neuroblasts (MB NBs). Unlike other NBs, which terminate proliferation before or shortly after the onset of metamorphosis, MB NBs continue dividing until the end of metamorphosis, ceasing only just before adulthood.<br /> In this study, the authors identify the transcription factor Krüppel (Kr), a member of the conserved Krüppel-like family, as temporally regulated in MB NBs. They demonstrate that Kr knockdown during pupal stages maintains expression of the RNA-binding protein Imp and results in prolonged MB NB proliferation into adulthood. Their data suggest that Kr contributes to the timely silencing of Imp during metamorphosis. The authors further identify Kr-h1, a related transcription factor, as a potential antagonist. While Kr-h1 appears dispensable for the timely termination of MB NBs under normal conditions, its overexpression leads to their continued proliferation and tumor-like expansion in adults.

This work provides the first evidence for a transcription factor-driven temporal regulation mechanism in MB NBs, offering new insight into the control of neural stem cell self-renewal. Given the evolutionary conservation of Krüppel-like factors, this study may have broader implications for the neural stem cell field.

Strengths:

(1)The study possibly identifies a new series of temporal transcription factors that are specific for mushroom body neuroblasts.

(2) The mechanism could be conserved in vertebrates.

Comment on revised version.

The authors have clarified the expression status of Kr in the KrIf-1 mutant, as well as the regulatory interactions among Kr, Kr-h1, and E93. My other concerns have also been satisfactorily addressed. I congratulate the authors on this excellent work and on discovering a novel mechanism that regulates temporal progression and the termination of division in mushroom body neuroblasts.

Reviewer #2 (Public review):

Summary:

This paper shows that imiquimod, a compound often used to induce a psoriasis-like skin inflammation in mice, has a TLR7-independent effects that induce the unfolded protein response and amplify cytokine expression in dendritic cells. Although these effects of imiquimod have been described in the literature before, this study provides more detailed evidence and different contexts to this observation. These findings add to existing literature that imiquimod has a pleotropic mechanism of action involving changes in mitochondrial functions and cellular stress responses. Specifically, the authors show that imiquimod can induce calcium signaling in immune cells and potentiate two branches of the unfolded protein response in a TLR7-independent and MyD88-independent manner. They also show that some of these effects might be partially mediated by direct binding of imiquimod to Gelsolin. These findings expand our understanding of imiquimod-mediated inflammation and are useful for the field of experimental skin immunology and mouse models of psoriasis. However, molecular and cellular mechanisms connecting Gelsolin to the unfolded protein response and skin inflammation presented in this paper requires further investigation in the context of TLR-mediated inflammation.

Strengths:

(1) TLR7-independent effects of imiquimod to the expression of genes and proteins involved into the unfolded protein response are well demonstrated.

(2) Gelsolin is identified as a new imiquimod-binding protein in mouse cells.

Weaknesses:

(1) Effects of imiquimod on mitochondrial Ca signaling are not clear form the presented data.

(2) The mechanism of action connecting imiquimod to Gelsolin on the unfolded protein response and cytokine production remains not fully explained.

(3) It remains unclear if Gelsolin contributes to regulating TLR7 (or other types of TLR-mediated) inflammation in vivo.

Reviewer #1 (Public review):

Summary:

Garcia-Alcala, Kratz and Cluzel investigate to what extent our understanding of bacterial physiology in bulk experiments can be applied to single-cell observations. They find that intrinsic noise may be powerful enough to even inverse the trends found in the bulk. The authors hypothesize that asymmetric distribution of ribosomes to daughter cells during the cell division plays the dominant role in the intrinsic noise and is able to generate the observed phenomenon. They do not show it directly, but the data and its agreement with the model suffice to support this claim.

Strengths:

The experimental part is convincing: the positive correlation between the elongation rate and promoter activity of unnecessary protein is clear, as well as the negative correlation between the mean values while changing the promoter strength. This was demonstrated in both rich and poor media. The causality between the growth rate and the promoter activity was shown using the negative lag time of the cross-correlation function. A simple, reasonable model accounts well for the data. This paper demonstrates an interesting phenomenon and provides a plausible theory for it, advancing our understanding of bacterial physiology on the single-cell level.

Weaknesses:

(1) Mean-reversion timescales were assumed to be longer than the simulation time and much longer than the cell cycle time. It is not clear whether the results robust in case mean-reversion timescales become of the order of cell cycle or smaller. If not, is there an argument for such practically infinite reversion timescales?

(2) It is not easy to understand the simulation part unless one reads Ref. [14]. Is k(t) assumed to follow Eq. (1) from ref. [14]? Is this crucial that the ribosome noise appears only at the division? The ribosome noise strength \sigma_R=0.06 - is it lower or higher than the naively expected binomial division?<br /> Also, more intuitive explanation of the Simpson paradox would help the reader.

(3) It would be useful for the reader to see the raw data and not only the filtered one to appreciate the measurement noise level.

(4) Negative lag time of the cross-correlation function is visible, but consider adding statistical test for it.

(5) Can you make similar cross-correlation plots using the model? Can you infer using it whether the data agrees better with the assumption that ribosomes noise appear only at division or continuous fluctuations during the cell cycle?

Comments on revised version:

The authors addressed the five comments listed above.

Reviewer #2 (Public review):

Summary:

The manuscript by Garcia-Alcala et al. reports an interesting paradox: the cost of gene expression slows the population-average growth rate, whereas at the single-cell level, expression levels from these genes positively correlate with the growth rate. The effect is observed in the expression of flagellar genes and a gene under a synthetic promoter in E. coli. The findings are explained by the inheritance of growth factors, including ribosomes, during asymmetric division.

Strengths:

(1) The manuscript adds strength to an emerging body of literature showing that the population-level bacterial growth laws do not match correlations based on single-cell data. The evidence presented here is more striking than in previous works (such as Pavlou et al., Nat. Commun. 2025), as the trends in population-level data and single-cell data are reversed.

(2) A relatively simple model correctly explains the trends in the data.

The findings raise an interesting question of whether similar effects occur in other bacterial species and, more broadly, in other organisms.

Comment on latest version:

The additional analysis has strengthened the conclusions.

Reviewer #1 (Public review):

[Editors' note: this version has been assessed by the Reviewing Editor without further input from the original reviewers. We appreciate the care with which the authors have addressed the remaining concerns. The additional analyses and the revisions to the framing and interpretation have strengthened the manuscript and clarified several of the issues raised during the previous round of review. In particular, the added analyses in Figure 4 and the more precise treatment of repeated retrieval, learned threat value, and the relationship between neural activity and freezing have improved the paper substantially.]

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.

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.

Reviewer #2 (Public review):

Summary:

In this study, the authors developed a human forebrain organoid model that incorporates both iPSC-derived microglia and CD8⁺ T cells, allowing them to recreate and investigate multicellular aspects of AD pathology in a human-relevant system.

Their findings show that microglia help clear amyloid-β deposits, but they also promote inflammatory responses. Activated microglia recruit CD8⁺ T cells by releasing the chemokines CCL4, CCL5, and CXCL10, which signal through the receptors CCR1/CCR5 and CXCR3. Pharmacological inhibition of CCR5 or CXCR3 prevents T-cell recruitment and alters autophagy pathways in a microglia-dependent manner.

Strengths:

The manuscript has several strengths, including its innovative multicellular organoid model, the combination of complementary experimental approaches, and the identification of potentially relevant immune signaling pathways.

Comment on the revised version.

Overall, the revised manuscript is considerably improved, and the major conceptual concerns raised in the initial review have been adequately addressed.

Reviewer #1 (Public review):

Ma et al. use human-chimpanzee tetraploid cells, across different cell types, to identify the genetic causes and then transcriptomic consequences of divergence in DNA methylation. They conclude that the evolution of DNA methylation is driven primarily by cis-regulatory changes, and that the evolution of CpG sites contributes to cis-regulation while transcription factor expression underlies some trans changes. They then argue that divergence in DNA methylation is associated with changes in gene expression and may contribute to human phenotypes.

The tetraploid model is able to provide compelling evidence that most regulatory evolution occurs due to cis-regulatory changes, and that sites that have diverged in DNA methylation level cluster together into DMRs which share similar divergence patterns and genetic architecture. This stands in intriguing contrast to trans-mechanisms which explain most variation within modern human populations. The authors proceed to show that trans-effects can be explained, in part, by nearby TF binding motifs and many cis-regulatory changes are explained by CpG-disrupting variants. While these mechanisms clearly contribution to divergence in DNA methylation, the degree to which they explain cis- and trans-regulatory divergence would be valuable.

Next, the authors seek to show that differences in DNA methylation are functionally relevant. Consistent with previous results, they show that differences in DNA methylation are (weakly) associated with changes in gene expression. They hypothesize that genes with concordant regulatory should exhibit great methylation-expression coupling than other genes and show that cis-expression/cis-methylation pairs are more strongly correlated than trans/trans pairs. I think that looking at cis/trans or trans/cis changes would also be useful. Another limitation is that this analysis is limited to promoter regions. It is not clear how many divergent DMRs that includes and how many of those genes have differences in expression. The key question is whether differences in DNA methylation are functionally important, and the answer provided by these analyses is "sometimes".

Finally, the authors make a case for lineage-specific selection on DNA methylation that is connected to human traits. Using a sign-test, they identify sets of genes which have likely experienced selection in humans and may underlie human phenotypes.

In conclusion, I think this study provides a valuable resource for differences in DNA methylation between humans and chimpanzees across tissues and provides important insight into the relative abundance of cis and trans regulatory divergence. Additional research is necessary to investigate the underlying regulatory mechanisms and more care needs to be taken in exploring the functional consequences.

Reviewer #3 (Public review):

Summary:

Ma et al. use human-chimpanzee tetraploid cells to examine species differences in DNA methylation. They identify differentially methylated regions under cis or trans regulation. Cis-DMRs are enriched near SNVs that disrupt or create CpGs, providing a plausible mechanism for cis changes in methylation. They also seek to identify transcription factors that might affect methylation in trans, as well as gene sets with evidence for consistent changes in methylation and expression between humans and chimpanzees, suggesting that they may underlie lineage-specific traits.

Strengths:

The authors have generated a new dataset across multiple different cell types examining differences in DNA methylation between humans and chimpanzees using human diploid cells, chimpanzee diploid cells, and human-chimpanzee tetraploid cells. Using this dataset, they identify that cis-DMRs are enriched near SNVs that disrupt or create CpGs compared to trans-DMRs and identify transcription factors as candidate trans-acting factors. Both identified SNVs and transcription factors are good candidates for future experimentation. The authors also find that cis-DMRs are more highly correlated with cis-expressed genes than trans-DMRs with trans-expressed genes, providing evidence that methylation and expression are linked genome-wide. Further, they apply a clever sign test approach to identify gene pathways with evidence for lineage-specific selection.

Comments on revised version.

They authors have addressed my concerns from their initial manuscript. In particular, the authors have noted the limitations of their study as appropriate, including the use of permissive FDR cut-offs.

Reviewer #1 (Public review):

Summary

This computational modelling study investigates the functional role of metaplasticity in a calcium-dependent synaptic plasticity rule, using a biophysically detailed model of a striatal projection neuron. The learning rule is deliberately simple: local calcium signals derived from different sources regulate LTP and LTD through separate plasticity thresholds, while dopamine provides a reward-related signal that determines the direction of synaptic updates. The author uses linear and nonlinear feature-binding problems (FBP and NFBP) as learning tasks to illustrate the mechanics of competing LTP and LTD processes in this rule. The central aim is thus to use these classification tasks to understand what metaplasticity in the two calcium thresholds contributes to learning.

The study identifies complementary roles for metaplasticity in the LTP and LTD thresholds, enabling synapses exposed to competing plasticity processes to converge toward a stable outcome. In reversal-learning tasks, relaxing the conditions for metaplasticity allows previously stabilized synaptic states to become flexible again and permits relearning. Finally, the manuscript investigates whether two separate calcium thresholds are themselves necessary. While a modified rule with a single adaptive calcium threshold can solve the same tasks comparably well, two thresholds allow separate control over synaptic strengthening and weakening.

Strengths

A major strength of the study is the simplicity of the learning rule, which makes it possible to dissect the roles of individual components while retaining a connection to experimentally motivated mechanisms of cortico-striatal plasticity. The use of shared features in the FBP and NFBP is useful for exposing the interaction between competing LTP and LTD processes and for showing how adaptation of one plasticity threshold can indirectly permit expression of the opposite form of plasticity.

The manuscript also provides a much more systematic analysis of the rule than the original version. In particular, the expanded parameter analyses explore both successful and failing regimes for the learning and metaplasticity rates, while the scan over the maximum synaptic weight provides a clear mechanistic motivation for the additional upper LTP threshold.

Importantly, the single-threshold simulations directly address whether two calcium thresholds are computationally necessary. These additional simulations clarify that the principal advantage of two thresholds is not that they uniquely enable nonlinear learning, but that they permit independent regulation of strengthening and weakening. This considerably strengthens the conceptual conclusions of the study.

Weaknesses

The main limitations concern the scope and biological generalizability of the results. The simulations in this paper are based on a particular detailed SPN model and depend on several specific biological and modelling assumptions. Most notably, solving the nonlinear task requires appropriately clustered excitatory inputs, but the learning rule itself does not provide a mechanism for generating these clusters. The manuscript now discusses this limitation explicitly and appropriately identifies structural plasticity as an important direction for future work.

A second important assumption is the upper calcium threshold that prevents LTP once sufficiently strong supralinear responses are reached. This mechanism plays an important functional role in the model but has not been experimentally established for cortico-striatal synapses. The manuscript appropriately identifies this as a major assumption and discusses potential molecular mechanisms that could implement such an effect.

Overall assessment

The revised manuscript is substantially stronger. The conclusions are carefully matched to what is demonstrated by the simulations, and several important alternative explanations and model configurations are tested directly. In particular, the revised framing provides a more specific mechanistic account of what metaplasticity in two separate thresholds contributes within the proposed rule. The work therefore provides a valuable contribution to computational studies of synaptic plasticity and metaplasticity, with convincing evidence supporting its principal conclusions.

Reviewer #2 (Public review):

Summary:

The manuscript proposes interesting synaptic plasticity rules grounded in experimental data. Its main features are: (1) plasticity depends on local calcium concentration driven by presynaptic activity and is independent of somatic action potentials, (2) the rules incorporate metaplasticity, and (3) they demonstrate how a single neuron could address the feature-binding problem at the dendritic level.<br /> The work extends a previous study (https://doi.org/10.7554/eLife.97274.2), to which the author also contributed.

The author models two calcium thresholds (LTP/LTD) from two different calcium sources (NMDA/VGCC) and these thresholds are flexible (metaplasticity rule, similar to BCM) which is claimed to be necessary for successful learning of both FBP and NFBP (linear and nonlinear feature binding problem with 1 or 2 patterns). The role of each threshold seems to be is opposite and complementary. One extra condition has been added : an upper threshold for LTP. This threshold serves to stop synaptic strengthening once synapses are strong enough to evoke a plateau. With that, synapses are not strengthened to the maximal value, avoiding strong supralinear integration for irrelevant patterns.

Strengths:

The current model implements not only local synaptic plasticity but also metaplasticity and solves the FBP at the dendrite level. Another strong aspect of the model is that metaplasticity in the LTD threshold protects strengthened synapses from weakening. In this way, as the author mentioned, metaplasticity is able to protect learned patterns from being forgotten or weakened and prevent irrelevant patterns from being stored. This is a nice modelling example of metaplasticity being helpful in preventing the catastrophic interference or forgetting (as has been explicitly discussed in a recent article https://doi.org/10.1016/j.tins.2022.06.002). The author might want to briefly mention or emphasize this aspect of the model, which might be interesting also for the AI community.

Comments on revised version:

The author improved the paper significantly. He conducted new simulations on relearning. He addressed the problems and limitations, added new simulation results and reasoned better in the discussion and the introduction. Also, the figures are improved.

Reviewer #1 (Public review):

Summary:

In this study, the authors elegantly combined latent variable models (i.e., HMM, GPFA and dynamical system models) with a calcium imaging observation model (i.e., latent Poisson spiking and autoregressive calcium dynamics (AR)).

Strengths:

Integrating a calcium observation model into existing latent variable models improves significantly the inference of latent neural states compared to existing approaches such as spike deconvolution or Gaussian assumptions.

The authors also provide an open-source access to their method for direct application to calcium imaging data analysis.

Weaknesses:

As acknowledged by the authors, their method is dependent on the quality of calcium traces extraction from fluorescence videos. It should be noted that this limitation applies to alternative strategies.

While the contribution of this study should prove useful for researchers using calcium imaging, the novelty is limited, as it consists of an integration of the calcium imaging model from Ganmor et al. 2016 with existing LVM frameworks.

Comments on revised version.

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

Reviewer #2 (Public review):

Summary:

This compelling study proposes a framework to implement latent variable models using population level calcium imaging data. The study incorporates autoregressive dynamics and latent Poisson spiking to improve inference of latent states across different model classes including HMMs, Gaussian Process Factor Analysis and nonlinear dynamical systems models. This approach allows for a more seamless integration of existing methods typically used with spiking data to apply on calcium imaging data. The authors test the model on piriform cortex recordings as well as a biophysical simulator to validate their methods. This approach promises to have wide usability for neuroscientists using large population level calcium imaging.

Strengths:

The strength of this study is the flexibility in the choice of models and relatively easy adaptation to user-specific use cases.

Weaknesses:

The weakness of the study lies in its limited validation of biological calcium imaging data. Calcium dynamics in a task-specific context in a sensory brain region might be very different from slower dynamics in a region of integration.

Reviewer #3 (Public review):

Summary:

S. Keeley & collaborators propose a computational approach to infer time-varying latent variables directly from calcium traces (e.g., obtained with 2p imaging) without the need for deconvolving the traces into spike trains in a preliminary, independent step. Their approach rests on 1 of 3 families of latent models: GPFA, HMM and dynamical systems - which they augment with an observation model that maps latent variables to fluorescence traces. They validate their approach on simulated data as well as a single real dataset, showing that the approach improves latent variable inference and model fitting, compared to more traditional approaches (although not directly compared with the 2-step one; see below). They provide a GitHub repository with code to fit their models (which I have not tested).

Strengths:

The approach is sound and well-motivated. The authors are specialists of latent variable models. The manuscript is succinct, well-written and the figures are clear. I particularly liked the diversity of latent models considered, in particular latent models with continuous (GPFA) vs. discrete (HMM) dynamics, which are useful for characterizing different types of neural computations. The validation on both simulated and real data is convincing.

Weaknesses:

The main weakness point that I see is that the approach is tested only on a single real dataset (odor response dataset). The other model fits are obtained from simulated data. While the results are convincing, it would be useful to see the approach tested on other datasets, for instance datasets with different brain areas, different behavioral conditions, or different calcium indicators. This would help assess the generality of the approach and its robustness to different experimental conditions.

Reviewer #1 (Public review):

In this work, Frey and colleagues have carried out a very large study of α-synuclein polymorphism as a function of aggregation conditions and sample preparation. They provide valuable insight into the many critical factors affecting α-synuclein polymorphism, thereby illuminating the need for detailed reporting in the literature as well as both rigorous and detail-oriented protocols when working with this protein. Indeed, their observations are in line with the difficulties of reproducing structural outcomes across different laboratories and experiments. The authors must be complemented on their openness about the difficulties experienced and the thoroughness of their work. Efforts like these are going to be crucial to achieve an understanding of the unparalleled structural plasticity of α-synuclein amyloid fibrils. It is particularly notable that the authors have managed to optimize protocols to form single-polymorph aggregation reactions with high reproducibility.

In this work, the authors focus on the influence of α-synuclein purity in aggregation reactions. They find that using reverse-phase HPLC to purify the protein significantly alters the aggregation behaviour. It is interesting, and rather uncommon in the field, to use reverse-phase HPLC as a final purification step, rather than SEC, which is commonly used in many laboratories. It would be useful to compare this new protocol even more directly and extensively with the commonly used SEC protocol. When mentioning their previously published work, it should be mentioned explicitly how the protein was purified in these previous studies.

On the point of protein purification, the authors lyophilize their protein prior to storage. In their work, they also find that pre-aggregation oligomer formation alters the aggregation pathway of α-synuclein. While they demonstrate that this can be solved by appropriate filtration, it should be discussed why the lyphilizaiton step was not reconsidered/omitted given that this process is known to facilitate oligomer formation. Do the authors have experience with aggregation studies using α-synuclein that has not undergone lyophilisation and are able to comment on the influence of this step in the protocol?

The authors make note of several degradation products affecting their aggregation reactions, which is why they employ a much more thorough purification protocol. However, they also point out that some of the degradation products found at the end of their reaction could form during the reaction itself. Unfortunately, they never investigate this further. In particular, it would be very useful to know if different aggregation conditions (pH, salt, agitation) lead to different and characteristic degradation patterns. HPLS/mass spec of the supernatant at the end of each aggregation reaction would have been a very insightful thing to do.

With respect to degradation products, in this work a NΔ4-variant is produced to mimic a disease-relevant degradation product and indeed it is found to alter the structural outcome even at low relative concentrations (5%). Have the authors investigated the minimal fraction of the NΔ4 variant necessary to still influence the structural outcome of the predominant WT protein? This type of analysis could have significant relevance to disease-related analysis, where several variants (truncations and PTM variants) are present in trace amounts.

While on this topic, the authors note that in several of their type 5 fibrils, they find unresolved peptide fragments in their cryo-EM structures. Can the authors speculate if these fragments are indeed peptide degradation products or residues wrapping around the fibril originating from the fibril-incorporated protein?

In this work, the authors have performed an extensive study of α-synuclein polymorphism. However, despite generating what is likely the largest single data set of fibril structures, they perform very little quantitative analysis of their data. It would be interesting to analyse the relative abundance of fibril polymorphs produced in each reaction. Perhaps from the particle-picking data it could be estimated the relative abundance of each polymorph as well as non-resolved fibrils to generate a more nuanced view of polymorphism beyond overall classifications of the resolved structures. Indeed, from such data it could also be studied if certain protofilaments are more prone to pair in asymmetric fibril structures over others or if fibril asymmetry can be attributed to random pairing of protofilaments in accord with their abundance. This latter point is particularly interesting for the type 1 fibrils. Perhaps, the propensity of α-synuclein to form specific symmetries could also be estimated.

The authors note that pH is a strong factor in determining polymorph selection. This does indeed appear to be the case, but other parameters do not appear to show any clear trend. Have the authors investigated the influence of aggregation parameters (agitation, duration) on structural outcomes systematically or quantitatively, such as with principal component analysis? Indeed, they also find that some fibril types that otherwise are not compatible at the same pH appear to co-exist when the shaking parameter is modified. Are the authors then confident in the claim that pH is a deterministic parameter?

It is evident from this and other work that amyloid aggregation is highly sensitive to kinetic effects. It is therefore curious that the effect of protein concentration and reaction time has not been systematically investigated. The authors have some data studying dilution series (Figure 6A) and different reaction times (reaction 56 & 57). Could the authors comment on the effects of these two parameters, and might there be more information touching upon this that could be highlighted in this work?

The authors state that this work likely underreports fibril polymorphisms in samples due to population size or data quality challenges. Could the authors, based on their extensive experience, try to quantify this statement?

Additionally, the authors point out that the current framework for classifying α-synuclein fibril polymorphism is not sufficient to describe the real complexity of this protein system. However, they do not seem to address some of the recent literature aiming to solve such issues (see Scheres 2026, Connor et al. 2025, Milchberg et al. 2025 & Price et al. 2025).

Was any biophysical/biochemical analysis performed of the many structures produced here, such as CD spectroscopy, Proteinase K digestion, dye binding or FTIR, which could act as low-resolution structure identifiers and might help to retrospectively explain some findings in the older literature? Such data would be very useful for the vast majority of researchers, who do not have access to cryo-EM.

Reviewer #2 (Public review):

Summary:

This manuscript describes insights gained during efforts to reproduce disease-relevant alpha-Synuclein (aSyn) fibrils using recombinant protein in vitro. It follows up on a similar article from this team published in 2024. Although the authors have not been able to produce fibrils with the structure of ex vivo fibrils isolated from patients, they share insights gained into which factors influence the formation of specific fibril polymorphs.

Strengths:

This is quite an unusual manuscript because it goes into minute detail about sample preparation that are usual just mentioned in the Materials and Methods sections of other manuscripts (if at all). This makes it very valuable for the scientific community working on exactly the problem of reproducing disease-relevant aSyn fibrils in vitro (which will be a major breakthrough in the field). The authors present an impressive array of cryo-EM fibril structures, some of which have not been described before.

Weaknesses:

A major concern with the manuscript is that its story and messaging are a bit murky. The authors describe a few new polymorphs, show that some polymorphs (type 1) have small variations, show that sample purity and fragmentation will influence polymorph formation, and present a helical-symmetry mystery. This all reads like a loose collection of findings without any major takeaway. Looking at Table 1, it still seems that the authors do not have a good control over any of these polymorphs. Are they able to make any of these polymorphs reliably? I think the impact of this work could be strengthened if it ended with a reliable protocol for the production of any of the polymorphs described.

A second major concern is the quality of the aggregation kinetics and their interpretation. Are these kinetics just done once per concentration? The figure caption talks about 'three independent samples' but it is unclear if this refers to the three different concentrations or NΔ4 percentages used, or true repetitions. Looking at the curves themselves, it seems that only one of the conditions was actually done in triplicate, which should be the minimum to draw conclusions. Further, it would have been helpful to characterize the kinetic data quantitatively. Finally, because there are only kinetic data for a fraction of the conditions tested, it is not clear what they add to the overall manuscript. My recommendation is to either remove the kinetics from the manuscript or substantially expand this section.

Reviewer #1 (Public review):

Summary:

The authors have assembled the genome of four local chicken breeds from China and analysed their gene content. They come to the conclusion that thousands of genes present in the current reference genome of chicken have become pseudogenized during chicken evolution and domestication. They argue that their study provides strong support for the importance of the "less-is-more" hypothesis for adaptive evolution.

Strengths:

The paper provides medium-quality genome assemblies for four individuals representing four local populations of chickens and analyses their gene content.

Weaknesses:

They have not excluded the possibility that the high rate of putative pseudogenes reflects the presence of errors in gene models, in particular in GC-rich microchromosomes that are challenging to assemble correctly. The paper contains no genotype-phenotype analysis, which means that the adaptive significance of a high rate of pseudogenization, if it exists, is unknown.

Reviewer #2 (Public review):

Summary:

The authors set out to investigate the evolutionary role of gene presence-absence variation and pseudogenization in chicken evolution and domestication. By comparing draft genome assemblies of four indigenous Chinese chicken breeds against the red junglefowl reference genome (GRCg6a) and four PacBio HiFi assemblies, the study proposes that the common ancestor possessed nearly 22,000 genes, and that each domestic lineage independently lost thousands of genes (identifying ~8,000 dispensable genes). The authors conclude that massive loss of function and pseudogenization represent major drivers of chicken evolution under the "less-is-more" hypothesis.

While the concept that gene loss can drive phenotypic diversification during domestication is compelling, the results do not convincingly support the central conclusions. The scale of reported gene loss and the specific patterns of pseudogenization appear to be potentially driven by well-known genome assembly gaps, annotation errors, and sequencing dropouts rather than genuine evolutionary events, and the authors do not provide enough convincing evidence that this is not the case.

Strengths:

(1) The study addresses an important and timely evolutionary question regarding the role of gene loss and loss-of-function variation in animal domestication.

(2) The inclusion of multiple indigenous Chinese chicken breeds alongside high-accuracy PacBio HiFi assemblies provides a valuable comparative genomic dataset.

(3) The authors attempt to evaluate pseudogene transcription using RNA-seq data and check for transcript isoforms that bypass candidate loss-of-function mutations.

Weaknesses:

(1) Unvalidated gene and pseudogene annotations: Long-read and consensus genome assemblies are known to suffer from residual indel errors that create artificial frameshifts and premature stop codons. The frequency of predicted pseudogenes in this study (~3.5%-4.5%) aligns closely with expected baseline annotation error rates. Although the authors state in their Methods that mutations were validated using short reads, this validation is never quantitatively demonstrated or shown in the results. The fact that most proposed pseudogenes are actively transcribed and lack a paralog strongly suggests that many are intact, functional genes affected by sequencing or annotation artifacts.

(2) Assembly gaps and GC-bias mistaken for gene loss: The claim that ancestral chickens possessed ~22,000 protein-coding genes and lost thousands of genes in only ~10,000-50,000 years is inconsistent with the evolutionary conservation of avian genomes. The missing genes are enriched for high GC content and preferentially located on microchromosomes. Avian microchromosomes and dot chromosomes are notoriously GC-rich, repeat-dense, and prone to severe assembly gaps in non-telomere-to-telomere assemblies. The reported gene absences reflect assembly fragmentation and coverage dropouts rather than evolutionary deletions.

(3) Lack of synteny validation: Genuine gene absence requires demonstrating conserved collinear synteny of flanking orthologous genes with an unambiguous sequence deletion at the locus. Relying on sequence alignment or short-read mapping failures across fragmented scaffolds substantially inflates false-positive gene loss calls.

(4) Positional bias of pseudogenization mutations: The observed concentration of pseudogenization mutations in the terminal 10% of coding sequences (the "bathtub" distribution) is characteristic of alignment boundary artifacts and non-canonical translation start/stop annotations, rather than positive selection to disrupt gene ends. Mutations in the terminal 3' region often produce functional proteins with slightly altered C-termini rather than complete loss of function.

(5) Inconsistent terminology: The manuscript alternates between identifying pseudogenes as unitary (lacking a functional paralog in the same genome) and evaluating sequence identity against "parental genes," creating substantial confusion regarding whether loci are duplicated paralogs or orthologous reference genes.

Reviewer #3 (Public review):

Summary:

The authors reanalyze genome assemblies of four indigenous chicken breeds from Yunnan Province together with the red jungle fowl reference (GRCg6a), and search for genes with disrupted protein-coding sequences. They catalog candidate pseudogenes and missing genes to estimate that 7,993 of the ancestral genes are dispensable. They characterize the positional distribution of pseudogenization mutations along coding sequences, their fixation in breeds, their estimated ages, and their pathway enrichments. Most results are replicated in four independent PacBio HiFi-based chicken assemblies. From the biased position of pseudogenization mutations toward CDS ends, their frequent fixation, and the phylogenetic signal in gene-loss patterns, the authors conclude that large-scale loss of function is a major driver of chicken evolution and domestication, consistent with the "less-is-more" hypothesis.

Strengths:

The catalog itself is a substantial resource: the comparison spans multiple closely related genomes, the main patterns are checked in a second, independently assembled set of HiFi genomes, and population resequencing data are used to ask whether the pseudogenization mutations are fixed rather than segregating. The finding that candidate loss-of-function genes are under relaxed purifying selection is well supported. The question of what gene loss contributes to domestication is worth asking, and this is a useful dataset for asking it.

Weaknesses:

The finding that genes carrying disruptive mutations are under relaxed selection is not particularly surprising, and the more interesting claim, that the observed patterns reflect positive selection for gene loss, is less certain in my opinion.

(1) Relaxed purifying selection versus positive selection. The bias of pseudogenization mutations toward the two ends of coding sequences is interpreted as positive or artificial selection, along with elevated dN/dS. But the alternative, that disruptive mutations at gene ends are simply better tolerated, is equally consistent with the results. Several mechanisms would produce this pattern under relaxed constraint without positive selection per se: alternative downstream start sites that rescue 5-prime disruptions; the small fraction of protein truncated by 3-prime disruptions; and the enrichment of disordered regions at protein termini.

(2) The functional status of the pseudogenes is assumed, not demonstrated. Genome-scale work cannot be expected to validate individual genes, but the language of the paper should reflect the candidate status of these calls. Nearly all predicted pseudogenes (~95%) were reported as transcribed in multiple tissues. It is possible for a nonfunctional coding sequence to retain intact regulatory sequences, but the observation deserves more attention in the paper, particularly because transcripts carrying premature termination codons can be the targets of nonsense-mediated decay, which is not discussed. These remain candidate pseudogenes defined by the presence of a putatively large-effect mutation (e.g. premature stop or frameshift).

(3) What "missing" means. For a gene to be scored as completely absent, it could be genuinely deleted, or its allele could be diverged enough that annotation and orthology/mapping no longer detect it. These are different phenomena. Related, since most pseudogenization mutations are reported as fixed or nearly fixed in their populations, the history of alleles matters: it is not clear how the authors established which state is derived, and whether the reference sequence assumed to be functional is in fact the "functional" version.

(4) Limited biological insight into domestication. The main biological interpretation rests on hierarchical clustering of dispensable genes followed by ontology enrichment within clusters (Figure 7a), with narrative connections to breed phenotypes. Only 19.6% of dispensable genes have Gene Ontology assignments, and the phenotype links are speculative. The section on the subspecies origin of the GRCg6a reference is only loosely connected to the loss-of-function story, and the population-genetic analysis supporting it is thin as described.

Reviewer #1 (Public review):

Summary:

The CATCHR complexes are a family of five multisubunit complexes that act in vesicle tethering in several key transport steps. COG and GARP act at the Golgi, EARP acts on early endosomes, EXOCYST acts in transport to the cell surface, and DSL1 acts on the endoplasmic reticulum (ER). The authors have used in vivo proximity biotinylation and mass spectrometry to look for new neighbours of three of these complexes, COG, GARP, and EARP (despite the title, only two of the three are on the Golgi). The authors then follow up two of the hits, CCDC186 and WWOX, with more directed experiments.

Strengths:

The strength of the paper is that proximity biotinylation is of a high standard. To avoid overexpression, the authors express TurboID-tagged CATCHR subunits in cell lines from which the genes have been deleted. This allows them to confirm that the tagged proteins are functional and correctly located. Mass spectrometry is used to identify the proteins biotinylated in each cell line with four replicates, and the data are clearly presented in figures and supplementary tables. The authors make a good choice of proteins to follow up, as both CCDC186 and WWOX appear potentially interesting.

Weaknesses:

Overall, although the paper is based on a high-quality initial set of data, it seems somewhat incomplete and preliminary. There is undoubted value in presenting a through if descriptive set of in vivo proximity labelling data for a comprehensive overview of proteins or complexes. However, in this case the authors have only addressed three of the five CATCHR complexes, and so it is not a complete overview. It was also somewhat unclear why they examined VPS52 and VPS53, as these are present in both GARP and EARP, which adds some ambiguity, even if they did at least provide useful confirmation of some of the hits with EARP. Of course, a proximity biotinylation analysis does not need to cover all members of a family if it generates substantial biological insight.

However, the investigation of CCDC186 and WWOX does not provide significant insight into either function or mechanism. CCDC186 has already been identified in C. elegans as a protein involved in dense core vesicle biogenesis (CCCP-1, as the authors acknowledge), and work in C. elegans and mammalian cells has already linked it to EARP function. Less has been published on WWOX, but all that is found here are some small changes in Golgi appearance and glycosylation when it is reduced by RNAi. Thus, the paper falls between two stools: it is neither a comprehensive application of proximity biotinylation to the CATCHR family, nor is it the application of proximity biotinylation to reveal new insight into membrane traffic. I feel that for a broad-interest journal such, it should be one or the other of these.

Reviewer #2 (Public review):

In this manuscript, Aragon-Ramirez et al. present the first systematic proximity-interaction map of the three Golgi/endosomal CATCHR tethering complexes - COG, GARP, and EARP. They generated hTERT-RPE1 knockout lines rescued with C-terminally TurboID-tagged subunits (COG4, COG6, VPS50, VPS52, VPS53, VPS54) expressed from the COG4 promoter at near-endogenous levels, verified that each construct rescues its KO phenotype, and confirmed expected localization by immunofluorescence. Fifteen-minute biotin pulses followed by streptavidin capture and label-free DIA mass spectrometry, benchmarked against GFP-TurboID, yielded compartment-resolved neighbor lists. The central claim of this work is that each complex sits within a distinct "trafficking module" of coiled-coil tethers (CCTs), SNAREs, SM proteins, Rab GTPases, coats, and homeostasis regulators. COG neighbors nearly all Golgi golgins plus the STX5-SCFD1 fusion machinery, with COG4 and COG6 lobes showing overlapping but non-identical hierarchies - this is offered as in vivo support for a two-lobe model. GARP associates with TGN golgins GOLGA1/GOLGA4, STX16 and partners, TBC1D23, and CLINT1. EARP associates with GRIPAP1, RELCH, RAB11FIP5, and the VPS33B-VIPAS39 (CHEVI) SM complex; a VPS53 MUN-domain mutant loses VPS33B proximity while retaining VPS50 labeling. Two hits are followed up in this study. CCDC186, a poorly characterized coiled-coil protein, localizes to the TGN and, when ectopically anchored to mitochondria, captures ~60 nm vesicles - presented as direct evidence of tethering activity. WWOX, a tumor suppressor with no prior trafficking role, colocalizes with COG8 in the medial Golgi; siRNA knockdown reduces Golgi area, increases HPA and GNL lectin binding (O- and N-glycosylation defects), and displaces COPB2 from the Golgi. Based on these findings, the authors conclude that CATCHRs are not isolated tethers but organizing hubs that assemble compartment-specific tethering-and-fusion modules. Overall, this is a solid piece of work that will be of interest to cell biologists.

(1) The CCDC186 mitochondrial assay demonstrates sufficiency in a non-native context but includes no loss-of-function work, and the cargo-specificity result carrying most of the interpretation is "data not shown." At a minimum, the authors should show KO/KD and rescue data.

(2) The WWOX section rests on a single siRNA with no second oligo, no rescue, knockdown shown only by RT-PCR, and localization based entirely on overexpressed myc-tagged protein.

(3) Proximity of labeling does not distinguish direct from indirect interactions. When referring to "interactions", the authors either need to show recombinant protein binding data, or soften the tone to acknowledge potential indirect interactions.

(4) The VPS53 MUN-domain mutant design rests on a "manuscript in preparation," and the mutant's localization isn't shown (VPS50 labeling establishes complex incorporation, not correct targeting). The HEK293T WWOX replication is "data not shown." The authors should either show the data or delete these claims.

(5) References 73 and 78 are duplicates.

(6) "NZR" is used in the introduction. Should this be NRZ?

Reviewer #3 (Public review):

Summary:

This manuscript presents a comprehensive proximity proteomics analysis of Golgi- and endosome-associated CATCHR complexes, including COG, GARP, and EARP, using near-endogenously expressed TurboID-tagged subunits. The study aims to define how these tethering complexes organize distinct trafficking modules involving coiled-coil tethers, Rab-associated proteins, SNAREs, SM proteins, and other trafficking regulators. In addition to validating known associations, the authors identify CCDC186 as a candidate vesicle tether and WWOX as a potential regulator of Golgi homeostasis and glycosylation. The work provides a broad resource for understanding the spatial organization of CATCHR-associated trafficking networks. However, several conclusions require further experimental and statistical support, particularly those interpreting proximity-labeling data as physical interactions or evidence of discrete functional complexes.

Strengths:

The manuscript addresses an important question in membrane trafficking and provides a systematic comparison of the proximity interactomes of COG, GARP, and EARP complexes. The use of near-endogenously expressed TurboID-tagged subunits is a strength, as it may reduce artifacts associated with protein overexpression. The dataset is comprehensive and has the potential to serve as a valuable resource for investigators studying Golgi and endosomal trafficking. The comparative analysis identifies both known trafficking factors and potentially novel regulators, including CCDC186 and WWOX. The functional follow-up experiments add biological relevance to the proteomic findings and extend the study beyond descriptive mapping. Overall, the scope is well defined, and the manuscript proposes an interesting model in which CATCHR complexes act as organizing hubs for vesicle tethering and fusion.

Weaknesses:

A major limitation is that TurboID proximity labeling is repeatedly interpreted as evidence of physical interaction or assembly into discrete complexes, although the method primarily reports spatial proximity. The soluble GFP-TurboID control may not adequately account for enrichment caused by membrane confinement and high local protein concentration in Golgi/endosomal microdomains. Some spatial enrichment analyses do not reach statistical significance, weakening claims of compartment-specific labeling. Several low-fold-change SNARE hits are emphasized despite limited enrichment. Additional quantitative imaging is needed, including colocalization analysis of streptavidin labeling with Golgi/endosomal markers and localization of GFP-TurboID. The WWOX knockdown phenotype requires validation by rescue, independent siRNA, or CRISPR-based approaches. Several supporting data, quantifications, figure corrections, and nomenclature revisions are also needed.

Reviewer #1 (Public review):

Gary, Soret et al., present an interesting and ambitious study investigating the cellular and signaling mechanisms by which GDNF induces enteric neurogenesis in a disease model of HSCR. The manuscript combines scRNA-Seq, pharmacological inhibition, and several lineage-tracing strategies. The lineage tracing experiments in particular are technically impressive and clearly represent a substantial amount of work. The figures are generally beautiful and high quality, and the biological question is an important one.

However, while I found the study interesting, I also think that several of the major conclusions are substantially stronger than what the current data support. In particular, the scRNA-Seq dataset that the study is founded on is difficult to interpret because of the RFP-based enrichment strategy, the lack of WT reference controls, and the very non-stringent FACS gating. Several mechanistic claims are also based primarily on expression patterns or sequencing results.

Major concerns:

(1) The G4-RFP based enrichment strategy raises some concerns. The authors note that 88-90% of SOX10+ glial cells (where did this number come from, it doesn't seem to match the range in the bar graph?) and only 70% of GDNF-induced neurons. They then sort in "yield" mode using a non-stringent gating strategy, but do not provide the flow graphs in the supplements for interpretation. Consistent with this, the initial dataset contains multiple non-ENS populations, including lymphoid, myeloid cells etc, suggesting that their strategy, in addition to missing cell populations, additionally includes potential negative cells based on background fluorescence.

(2) It's not clear to me why the authors did not include a wild-type control for comparison. Due to the RFP-enrichment strategy, it is additionally difficult to compare to integrated with scRNA-seq datasets. The comparison seems important because the central question is not only whether GDNF induces neurons in the mutant colon, but whether the induced cells and associated populations resemble those found in a normal ENS.

(3) The proportional changes in Figure 2 are hard to interpret. The authors write that GDNF treatment leads to an enrichment in enteric neurons, "at the expense of" SCPs. This seems to be a strange conclusion or writing, unless the authors are suggesting that certain cell populations are additionally depleted by GDNF treatment. I think the authors should provide stronger support, or dial down on conclusions based on cluster proportions, given the sorting strategy.

(4) Some of the mechanistic conclusions are far too strong for the evidence. For instance, the claim that GDNF signals through GFRa1/2 based on only expression UMAPs is not justified. The authors should at least confirm some of the major findings via smFISH (HCR or RNA-Scope), and dial down the mechanistic claims. The pharmacological data is more supportive but still indirect. This section would be stronger if the authors validated NCAM1/FAK activation in the specific progenitor populations proposed to respond to GDNF.

(5) The inhibitors are administered during the same P4-8 window as GDNF. Given that the authors report cellular-level responses as early as 6 hours post GDNF treatment, it is important to know whether the inhibitors were already active at the time of GDNF administration. Why were the inhibitors not administered prior to the start time of GDNF treatment?

(6) The authors draw inferences and conclusions based on IHC images, but do not provide any details on the analysis procedures. For example, in Figure 3, the authors interpret changes in NCAM and FAK expression level. However, it does not appear that there are notable differences in intensity between the two timepoints. There are however, differences in signal coverage. Without knowing what was actually measured and quantified, it is not possible to interpret the results. Please provide the necessary information for all experiments in the methods section.

(7) The final results section introduces one of the more surprising claims in the paper, that a non-neural crest-derived progenitor contributes to regeneration, but the section ends quite quickly and abruptly after introducing the observation. The authors should provide additional validation or interpretations of the data.

Reviewer #2 (Public review):

Summary:

Previous work from this group demonstrated regeneration of the ENS following exogenous GDNF treatment within the aganglionic portion of the bowel in several Hirschsprung (HSCR) murine models as well as colon from HSCR patients (Soret, et. al., Gastro. 2020). Focusing on the Holstein (HolTg/Tg) HSCR mouse model, the authors build upon their prior work utilizing lineage tracing, immunohistochemistry, scRNA sequencing, and pharmacological methods to further delineate the mechanisms and cell types contributing to ENS regeneration in this context. This work adds to the knowledge base regarding natural ENS development as well as ENS generation outside of natural ENS development. Furthermore, this work provides important clinical implications for potential curatives treatments for HSCR disease.

The conclusions in the paper are overall well supported by the data with the majority of technical and model limitations openly acknowledged. Strengths include the use of scRNA sequencing, several genetic mouse models with ample, well-planned time points within experiments. However, additional analysis, clarity or more detail around experimental protocols, and broader discussion of prior studies would strengthen the manuscript by improving interpretation of the findings and placing them more clearly within the context of existing work in the field.

First, utilizing the G4-RFP transgene for ENS cell selection and subsequent scRNA sequencing of RFP+ cells, the authors nicely demonstrate the presence of known cell clusters (based on markers and analysis from prior sequencing studies) including Schwann cell precursors (SCPs), enteric glial cells (EGCs) and neuron subtypes in the aganglionic colon of HolTg/Tg GDNF-treated and untreated mice. Further use of this dataset led to the discovery of Ncam1 expression, a GDNF receptor, within all ENS cell clusters, suggesting importance of this protein in the GDNF-induced ENS regeneration phenotype. Interestingly, Ret was limited to some neuron types. On the protein level, temporal and GDNF-treatment changes in NCAM1 expression and downstream phospho-FAK(Y397) were nicely demonstrated via Western blot. Mechanistically, the importance of NCAM1 in ENS regeneration was demonstrated given pharmacologic blockade of downstream FAK signaling via PF-562271 resulted in significantly fewer GDNF-induced neurons. The paper would be strengthened here, however, if the following areas were addressed:

1) Although the GDNF-induced regeneration appears likely through an NCAM1 signaling mechanism, inhibition of FAK phosphorylation by PF-562271 could affect more than NCAM1 signaling. Blocking NCAM signaling through another method (such as through NCAM1 targeted monoclonal antibodies or small molecules) would support the authors' interpretations further.

2) A point of ambiguity is the exact cells included in the final scRNAseq analysis. The authors noted they flowed and gated broadly, including RFP negative cells, in the initial set of cells selected for sequencing (given RFP expression is not visualized within all neurons and glia in GDNF-treated HolTg/Tg mice.) Pan-neuronal (Elavl4, Tubb3) and pan-glial (Sox10) marker expression were utilized to select cells for final analysis (unsupervised analysis, pseudotime trajectories, etc.) Here, it is unclear if RFP expression was examined in these cells, either by looking at unmapped reads or if included in the initial alignment. If this is discernable within the authors' data set, this provides an opportunity for an interesting additional analysis to determine if progenitors or neuronal or glia populations differ in the RFP+ and RFP- groups.

3) The authors presumably sequenced both the submucosal and myenteric plexus ENS components within their scRNAseq analysis. Given the relatively sparse numbers of ENS cells in the submucosal plexus compared to myenteric plexus in the distal colon, this is unlikely to change their analysis or conclusions. However, given their tissue analysis focuses on the myenteric plexus, this difference should be acknowledged.

The authors go on to further to utilize their scRNA sequencing data set with pseudotime trajectory analysis to determine that SCPs appear to go through an EGC-like state prior to GDNF-induced neurogenesis. This was followed nicely by use of transgenic mice (Dhh-Cre) and immunohistochemistry within their HolTg/Tg model to capture the SCP population and demonstrate in vivo this progression. Additionally, they were able to demonstrate that temporally, SCPs generated neurons much sooner (within 6 hours of GDNF-treatment) compared to neural crest derived EGCs (labeled by Slc18a2-Cre and GFAP-CreERT2 lines).

Fascinating here as well is that the authors unearthed that most generated neurons following GDNF-treatment appeared to derive from direct cell transdifferentiation (suggested by EdU incorporation investigation at various time points) and that these neurons did not seem to originate from SCPs or EGCs (i.e. a non-neural crest origin.) Acknowledging that their mouse genetic tools and sequencing could have missed a NCC-derived population, the authors go on to label neural crest derivates utilizing the Wnt1-Cre2 transgene in their wildtype and HolTg/Tg mouse model and find that while Wnt1 labels the vast majority of neurons in wildtype mice, nearly 25% of neurons in P20 HolTg/Tg mice appear non-NCC derived. Finally, once again utilizing various genetic mouse lines for lineage tracing and immunohistochemistry, the authors demonstrate NCC-derived neural progenitors (SCPs, EGCs) appear to preferentially give rise to cholinergic neurons where non-NCC derived neurons tend to be more nitrergic.

4) The high magnification of the majority of images provides the reader with extremely compelling evidence in regard to visible overlap or lack thereof across various reporter lines, antibody markers, and EdU labeling. Unfortunately, this can somewhat create a trade off with the area or number of cells examined which appears somewhat less than the number of cells and/or area examined by many prior studies in the field. The near equal averages across the three mice in each group in many of their experiments makes this less of a concern, but the paper may be strengthened by noting exactly how the 3-11 images per mouse were selected. Were these randomly selected across the tissue, moving from mesenteric to anti-mesenteric? Or caudally to distally?

5) In line with this, the authors report a near 100% of Wnt1-Cre2 driven reporter expression in wildtype mice but a large portion (~25%) of neurons in HolTg/Tg model appear non-NCC derived (i.e. not Wnt1-Cre2 labeled.) The original Wnt1Cre line (Danielian, et.al., Curr Biol. 1998) has been reported to incompletely label all neural crest derivatives/ENS (Hari, et. al., Development. 2012.; Deal, et. al., Dev Bio. 2021). To my knowledge, the same type of analysis has not been carried out with the same rigor in the ENS with the Wnt1-Cre2 line. It could be -- depending on tissue sampling, area covered via imaging, and if there is any patchy or unique patterning to non-NCC derived neuro regeneration -- any incomplete labeling of neural crest by Wnt1-Cre2 could be missed. Of course, alternative interpretations are that the non-labeled cells are truly non-NCC derived. Utilizing other lineage tracing models for other non-NCC derived lineages may help address this issue, but would require many mouse crosses and additional experiments, and thus likely outside the scope of this manuscript. However, this caveat should be addressed.

6) A very important and fascinating find by the authors is that the NCC versus non-NCC derivatives appear to preferentially give rise to cholinergic or nitrergic neurons respectively. This will have important implications in treatment options for ENS regeneration not only within HSCR disease, but other ENS disorders as well. A point of discussion the authors missed out on is several previous reports regarding skewing of ENS neurons (typically toward higher nitrergic numbers) in ganglionic and/or hypoganglionic segments of HSCR bowel in murine models and patients (Zaitoun, et.al., Neurogastroenterol. Motil. 2013.; Musser, et.al, CMGH. 2015.; Cheng, et.al., J Pediatr Surg, 2017.; Sukhada, et. al., Front. Cell Dev. Biol. 2022.) Could ganglionic portions of the bowel in HSCR disease contain more of the non-Wnt1 labeled population? Why do proportions in the aganglionic region generate correct proportions following GDNF-induced treatment compared to ganglionic bowel? This does not need to be experimentally addressed in the manuscript, but including in the discussion impresses upon the audience the importance of the authors findings and that these factors need to be further experimentally delineated and considered with any therapeutic interventions.

Reviewer #3 (Public review):

Summary:

In this work, the authors aim to build on their previous work to further define the mechanism by which GDNF can induce enteric nervous system migration and differentiation in the Holstein mouse model (Hol Tg/Tg) of Hirschsprung's disease (HSCR). They show that GDNF can cause neuronal differentiation from either Schwann cell precursors (SCP) that migrate from the periphery and/or enteric glia (or neural crest precursors located in the intestine). Surprisingly, the rescue effect of GDNF is mediated through signaling with NCAM1 rather than the canonical receptor RET, with blockade of NCAM1 significantly reducing the number of GDNF-treatment induced neurons in the distal colon of Holstein mice. The authors then perform scRNA-seq data of the colonic ENS from Holstein mice +/- GDNF, and pseudotime analysis of this data indicates the SCP downregulate SCP genes and upregulate classical enteric glial genes on their transition to becoming enteric neurons. Following this, the authors employ Dhh-Cre mice (to label SCP), Slc18a2Cre and GFAPCreERT2 (to label enteric glia) and finally Wnt1Cre2 mice (to label neural crest derived cells) to assess the kinetics and end identity of enteric neurons driven by GDNF treatment. Using these animal models, the authors propose a model by which transdifferentiation of SCP gives rise to enteric neurons first, followed by transdifferentiation of enteric glia; the resulting enteric neurons labelled by Dhh, Slc18a2 or GFAP Cres are all biased to a cholinergic lineage, raising a questions about the source of nitrergic neurons. Finally, the authors postulate that at least some nitrergic neurons come from a non-neural crest lineage in GDNF treated Hol Tg/Tg mice as they are not labelled by a Wnt1Cre2 model, while all ENS cells are labelled in three wildtype Wnt1Cre2Tg/+;R26YFP/+ mice.

Strengths and Weaknesses:

This work convincingly shows that NCAM1 is required for GDNF-induced rescue of the Holstein mouse model of HSCR and, along with the NCAM1 staining in human HSCR tissue, highlights the potential of this treatment even in HSCR patients with loss of function RET mutations. The major weakness of this study is in the interpretation of the mouse models; many of these models are not fully penetrant/show mosaicism in other studies (like the DhhCre and Wnt1Cre) or may not only label the specific subpopulation they present (in this case, the GfapCreERT2 and Slc18a2Cre; these genes have been shown in Schwann cells/SCP in other studies, and the expression pattern of these two genes outside of the gut at the relevant developmental timepoints are not shown). While these are unavoidable limitations with our current mouse models, the authors need to clearly acknowledge these limitations in the interpretation of their data (with the DhhCre and Wnt1Cre) or more definitively show the models are only labelling enteric glia (in the case of the Gfap and Slc18a2Cre). In addition, in some figures the authors should add untreated Hol Tg/Tg and ideally WT mice so readers can contextualise the effect of GDNF on Hol Tg/Tg mice. Finally, it appears the statistics throughout the paper are performed on using each ROI taken from n=3-5 animals - this is inappropriate, as images from the same animal are not independent samples. The statistical unit in question for every analysis performed should be 1 animal, thus the statistics should be done on the n=3-5 animals, not the n=3-11 images from 3-5 animals (which leads to an n of 6-55).

Specific comments:

In general I find the author's conclusions are relatively well supported, albeit a little overstated for the evidence presented in the later figures and with the need to correct the statistical analysis. For Figures 1-5, I have only minor comments, as follows.

In Figure 1-2, the Gata4-RFP reporter is used to sort cells for scRNA-seq, and the authors get a large number of non-ENS cells so utilise Elavl4 and Sox10 expression to pull out neurons and glia to analyse. Do other putative neuronal/glial genes pull out the same cells, and can the authors find evidence of non-neural crest derived glia/neurons in this dataset? This would ensure there isn't a bias in this first analysis and strengthen the later claims respectively.

In Figure 3, the interpretation of increased NCAM1 and pFAK would be aided by adding the MFI for the untreated Hol Tg/Tg and ideally WT mice at p10 to assess if this is a normal developmental increase or induced by GDNF treatment - I would ask the authors add at least untreated Hol Tg/Tg at p10 as authors already present this data as images in figure 3D. Similarly, it would be nice to have neuronal counts of untreated Hol Tg/Tg and WT mice presented in Figure 4c to assess if NCAM1 inhibition completely or only partially prevents the effect of GDNF.

In the text accompanying Figure 2 and 5, the authors use a variety of markers of assign SCP or enteric glia identity to the clusters, but some of the enteric glia genes can be expressed by Schwann cells/SCP in other contexts - it is worth adding references to the genes chosen to define SCP and enteric glia so readers can understand why those were chosen and the strength of evidence for each marker. For me, with the exception of Dhh I do not find any of these genes definitive as, at least on the SCP/Schwann cell side, they have been shown to express Gfap, Slc18a2, Sox2 and single cell analysis indicates they also express Cpe. The ability for SCP/Schwann cells to express both Gfap and Slc18a2 is important for the interpretation of the upcoming experiments.

In Figure 6, the authors state in line 213 that "quantitative analysis as a function of time revealed that GFAP+ cells in the distal colon are not all derived from Dhh-expressing SCPs on the first day of treatment"; however, the DhhCre is restricted to specific SCP subsets (Xie, Meng, et al. "Schwann cell precursors contribute to skeletal formation during embryonic development in mice and zebrafish." Proceedings of the National Academy of Sciences 116.30 (2019): 15068-15073) and thus may not be labelling all SCP coming into the intestine. As far as I know there are no models that would allow labelling of all SCP without labelling enteric glia so would require the generation of a new models or techniques that is out of scope for this work - however, the authors should acknowledge the possibility of Dhh negative SCP contributing to the ENS.

Figure 7 and 8 focuses on the contribution of Dhh+ cells, Slc18a2+ cells and cells expressing GFAP between p3-p8 to the ENS of GDNF treated Hol Tg/Tg mice. The authors first show that by p20 the number of neurons per mm2 is equivalent to those at p60, indicating p20 as an appropriate timepoint to assess the relative contribution of each pool of glia to the resulting neurons. The authors show that neurons derived from a DhhCre are detectable first, then Slc18a2, followed finally by GFAP. This data supports the authors claims that neurons first arise from Dhh+ SCP, but as I raised earlier, both Slc18a2 and GFAP can be expressed by schwann cells later in development - the authors in fact show that GFAP is in fact expressed by Dhh+ cells in figure 6, and showed with pseudotime analysis that SCP undergo a transition to Gfap+ Slc18a2+ glial cells before becoming neurons. The data presented here negates that analysis, or at least suggests some Dhh+ cells skip this step, otherwise the kinetics of neuronal differentiation would align across the 3 lines used. Could the authors address this discrepancy? For example, one addition could be to show that SCP/schwann cells on extrinsic gut innervating nerve fibres do not express Gfap or Slc18a2 at the timepoints used. In addition, their scRNA-seq analysis suggests that Gfap & Slc18a2 are expressed in the same cells, so one would expect similar kinetics in both populations - what is the overlap of GFAP staining with Slc18a2-YFP expression at the earlier stages (say p6 and p10?). Alternatively, the inducible nature of the GFAPCreERT2 means some bona fide GFAP expressing cells may be missed if the tamoxifen dose is not saturating, thus underestimating the contribution of this population to the total neuron count. As such, could the authors please show co-staining with GFAP in some induced GFAPCreERT2 samples to assess the effectiveness of this Cre. As with other figures, it would also be nice to add Holstein negative reporters at p20 as a control to assess the level of contribution each one of these cell lineages has in a normally developing animal compared to GDNF treated Hol Tg/Tg mice.

In Figure 9, the authors show that neurons derived from labelled cells in the DhhCre-YFP or Slc18a2Cre-YFP skew strongly cholinergic (81%) compared to nitrergic (13%), but only 52% of the total neurons in GDNF treated Hol Tg/Tg mice are cholinergic, while 42% are NOS1+. The authors say that a similar trend is seen in Slc18a2Cre-YFP mice with a 'slightly higher percentage' reflecting Slc18a2 expression in neurons. However, this 'slightly higher' percentage looks to be about 40%, which is the amount expected in GDNF treated Hol Tg/Tg. I think the authors underplay the level of nitrergic neurons that are labelled in the Slc18a2Cre mice and should consider the possibility that Slc18a2+ glia can give rise to the nitrergic neurons seen in GDNF treated Holstein mice. This data also contrasts with prior reports that indicate SCP tend to become NOS1 expressing in a Ret or Ednrb loss of function HSCR model (Uesaka, Toshihiro, et al. "Enhanced enteric neurogenesis by Schwann cell precursors in mouse models of Hirschsprung disease." Glia 69.11 (2021): 2575-2590), a discrepancy worth discussing.

In Figure 10, the authors assess the contribution of neural crest to the ENS of WT and Hol Tg/Tg mice using a Wnt1Cre2 line, and conclude there is non-neural crest contribution to the ENS of GDNF treated Hol Tg/Tg mice. This is an interesting idea and supported by a few studies the authors mention in the discussion; however, the labelling of neural crest derived cells in Wnt1Cre2 may not be fully penetrant. While this line has a well established leaky phenotype, others have reported it can have patchy recombination in the neural crest too (potentially in a background dependent manner, according to the line information on Jackson laboratories). For example, recent works shows significant mosaicism in the embryonic phase (Gandhi, S., Du, E. J., Pangilinan, E. S., & Harland, R. M. (2024). The Wnt1-Cre2 transgene causes aberrant recombination in non-neural crest cell types. bioRxiv, 2024-11.) This may not be the case in the authors home institution in a FVB background, as they show all HuCD+ and Sox+ cells overlap with the Wnt1Cre2 reporter, but the Hol Tg/Tg mice have aberrant neural crest development - this could interfere Wnt1 expression thus the effectiveness of the Wnt1Cre2, thus leading to non-labelled, but actually neural crest derived cells. It is also possible that those neurons are truly from a non-neural crest background, and the neural crest labelling tools are all flawed in one way or another - as such, I think the authors just need to discuss this alternative possibility in the discussion.

Finally, Figure 11 shows NCAM1 expression in human tissue resections from HSCR patients. This data bolsters the idea that GDNF treatment could be an effective treatment for HSCR. The methods indicate that 3 tissue samples from HSCR patients were obtained, but staining is only shown for one - could all 3 be included, along with secondary only controls? This is important as human tissue resections are often processed many hours after removal, so are prone to higher background. It would also be interesting to include tissue from non-HSCR patients (the controls used most often in other studies are resections from anorectal malformations which are performed at approximately the same age as HSCR resections) to see if NCAM1 expression is influenced by the absence of an ENS in human patients, but I understand these samples are not always possible to obtain. If not possible, the authors could consider analysing publicly available RNA sequencing datasets to assess if RET is solely expressed in mature neuronal populations while NCAM1 is expressed by glial and neuronal populations like in their mice data.

Overall, the authors clearly show that GDNF can drive differentiation of neurons from different glial populations through NCAM1 not RET, and supports continued effort to translate this work to the clinic as a potential therapy for HSCR patients. I am not convinced that the relative contributions of each glial subpopulation can be extrapolated from the mouse models used - this is a problem across the field, and will only be rectified with the development of new models, which is out of scope for this work. This work complements recent work showing NCAM1-GDNF signaling promotes neurogenesis in glia in Ednrb-/- mice (Mueller, Jessica L., et al. "Intramuscular enteric glia persist in Hirschsprung disease and undergo neurogenesis in response to GDNF-NCAM1 signaling." Scientific Reports 15.1 (2025): 33200), and will add to our understanding of neural development.

Reviewer #4 (Public review):

Summary: The manuscript by Gary et al works to reveal processes that generate enteric neurons in the distal colon of Holstein mutant mice that experience aganglionosis and are an established model of Hirschsprung disease (HSCR) following GDNF enema infusion. The primary goal of the study is to determine the how rectal infusion (enema) of GDNF drives neurogenesis in these mutants. The studies focus on determining the molecules that mediate this GDNF effect and identifying the cellular origins of the newly formed neurons. Given that prior work from this group and others has indicated formation of postnatal enteric neurons is feasible, understanding this process is an important step for the field. The data presented partially support the claims of the study. The analysis described do not support unequivocable conclusion that "transdifferentiation" is occurring. To instill confidence in the stated conclusions and make the text appropriately clear, clarifications of the methods and additional information is needed.

Strengths:

(1) The authors use in vivo experimentation with multiple mouse lines including Dhh-cre, Slc18A2-cre, and GFAP-creERT2 to trace the production of enteric neurons in the context of GDNF enema in the established Holstein mouse model of Hirschsprung disease over a time line of postnatal days 4 (P4) to P10.

(2) The authors generate novel molecular profiles of enteric neurons induced in the distal colons of Holstein HSCR mice after GDNF enema induction by flow sorting for fluorescently labeled and adjacent populations with novel gating parameters followed by single cell RNA-sequencing.

(3) The authors recognize broad expression of NCAM1 amongst the induced neurons in scRNA-seq data and apply experimental approaches to further investigate the role NCAM1 in generation of GDNF-induced ENS neurons.

(4) The authors apply a widely utilized experimental approach, EdU-labeling, to assess proliferation in the immediate timeframe of GDNF-inducation in an effort to discern the timing of when the newly produced neurons exit the cell cycle.

Weaknesses:

(1) The article lacks sufficient information justifying use of the Holstein HSCR mouse model for the study. Authors need be more thorough when introducing the readers to the various HSCR mouse models they consider analyzing and provide an expanded justification for use of the Holstein model in the introduction so the audience can appreciate the rationale behind the analysis presented.

(2) The text minimally describes the G4-RFP reporter that is a critical element of the analysis and upon which the scRNA-seq profiling hinges. Given the data presented in this submission it's not clear how much of the ENS this G4-RFP line labels.

(3) The authors state that "NCAM1 is already present in SOX10+ cells before GDNF treatment begins at P4, both within and outside extrinsic nerve fibers (Fig.3b,c)." However, the data shown in Figure 3 panels B and D do not allow one to conclude co-localization of Sox10 with NCAM1 or FAK. Because ENS cells are so closely positioned with one another the signal the authors present could be due to adjacent cells or processes of cells above and below the plane of Sox10+ nuclei.

(4) The authors use the inhibitor PF-562271 in an attempt to specifically inhibit phospho-FAK[Y397] shown in Figure 4. This particular inhibitor has the known side effect of causing apoptosis in phospho-FAK[Y397]+ cells as shown by Hu et al 2017 Cancer Sci (DOI: 10.1111/cas.13256). Because this compound causes apoptosis in phospho-FAK[Y397]+ cells, the data presented do not prove that the lack of neurons produced in this condition is due to inhibition of phospho-FAK[Y397]+ cells making the neurons versus those phospho-FAK[Y397]+ cells simply dying very early in the process.

(5) In the text describing the results of Figure 1A compared to Figure 1B the authors have missed an opportunity to elaborate on the spatial distribution of what appear to be small ganglia in the GDNF treated colon of Holstein mutant mice. The schematic shown is rather simple and it's unclear from the text where these ganglia are distributed circumferentially around the gut wall.

(6) The immunohistochemical labeling for Phox2b shown in Figure 6 is odd. Phox2b is expressed in neuronal progenitors, enteric glia, and ALL enteric neurons as shown in multiple publications. The images shown in Figure 6 offer an outlined region that appears to be a ganglion; however, within that encircled area fewer than half of the cells are labeling with Phox2b by this study.

(7) The authors conclude that "transdifferentiation" is the origin of the ENS neurons that appear in GDNF-treated Holstein mice. However, the data show simply that most of the newly produced neurons have not recently gone through cell division based on lack of EdU incorporation. Given the data presented, other mechanisms may be occurring and should be considered as possibilities.

(8) The authors utilized a Wnt1-cre2 transgenic line that has known issues with expression in the male germline and ectopic expression in cells that are dependent upon the reporter line utilized, like the Rosa26-YFP of this study. The methods lack information on whether crosses were performed in such a manner as to avoid issues with male germline activation of this reporter and the potential for ectopic expression cannot be excluded based on the information provided in the study.

(9) The study lacks data on how GDNF enemas affect enteric neuron density and ganglia distribution in wildtype animals. If the signaling mechanism that produces new neurons in the Holstein model is also operating in wildtype animals, this could be crucial information for investigators interested in neuronal replacement to treat ENS damage resulting from environmental damage, disease, or age.

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:

From a big picture viewpoint, this work aims to provide a method to fit parameters of reduced models for neural dynamics so that the resulting tuned model has a bifurcation diagram that matches that of a more complex, computationally expensive model. The matching of bifurcation diagrams ensures that the model dynamics agree on a region of parameter space, rather than just at specially tuned values, and that the models share properties such as qualitative features of their phase response curves, as the authors demonstrate. A notable point is the inclusion of extracellular potassium concentration dynamics into the reduced model - here, the quadratic integrate-and-fire model; this is straightforward but nonetheless useful for studying certain phenomena.

Strengths:

The paper demonstrates the method specifically on the fitting of the quadratic integrate-and-fire model, with potassium concentration dynamics included, to the Wang-Buzsaki model extended to include the potassium component. The method works very well overall in this instance. The resulting model is thoroughly compared with the original, in terms of bifurcation diagrams, production of various activity patterns, phase response curves, and associated phase-locking and synchronization properties.

Weaknesses:

It is important to note that the proposed method requires that a target bifurcation diagram be known. In practical terms, this means that the method may be well suited to fitting a reduced model to another, more complicated model, but is not likely to be useful for fitting the model to data.

Reviewer #2 (Public review):

Summary:

The authors derive an integrate-and-fire model to describe the dynamics of a more complex Wang-Buzsaki model and compare the two models. A detailed discussion of bifurcation schemes in both models is convincing and allows us to evaluate the simpler model.

Strengths:

The idea is interesting, and the mathematical approach appears to be convincing. In addition, differences between the simple and original models are also discussed.

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:

In this study, the authors examine what happens when two facultative endosymbionts, Rickettsiella viridis and Regiella insecticola, are introduced into a novel aphid host, the Russian wheat aphid (Diuraphis noxia). They ask whether these introduced symbionts affect aphid performance, plant damage, alate production, dispersal, plant defense responses, and symbiont dynamics. The main result is that the two symbionts have contrasting effects: Rickettsiella tends to increase plant damage and reduce dispersal-related traits, whereas Regiella tends to reduce plant damage and aphid population growth, with less evidence for an effect on dispersal.

Strengths:

The manuscript presents successful establishment of stable transinfected populations of an agriculturally important aphid species, which is a substantial technical achievement in itself. I also appreciated that the authors examined the system across several experimental contexts, including different host plants, mixed cages at two temperatures, whole-plant assays, and a mesocosm dispersal experiment, rather than relying on a single laboratory setup. Taken together, these experiments provide a useful and reasonably convincing demonstration that novel symbiont associations can generate contrasting phenotypes in this system.

Reviewer #2 (Public review):

Summary:

The authors generated two novel aphid-symbiont associations and examined the impact of these new symbiotic associations on plant-insect-symbiont interactions. The authors notably provide detailed phenotypic assessments of the insect hosts and host plants. They show that one introduced symbiont, Rickettsiella, increases aphid-induced damage to host plants, while the other, Regiella, ameliorates aphid damage. The authors suggest that such novel insect-symbiont pairings may be used as tools to mitigate crop damage in the future.

Strengths:

Although a few experiments seem to have limited sample sizes and limited statistical power, these are often complemented with highly replicated smaller-scale experiments. The combination of larger mesocosm and population-level experiments along with assessments of individual insects generally provides a comprehensive depiction of the effects of these symbionts on their hosts. The opposing impacts of Regiella and Rickettsiella infection on the aphid host plant are of broad interest. It is also surprising that the host plants did not exhibit strong differences in canonical defensive signalling, despite these differences.

Weaknesses:

It is a little surprising that mesocosm-dispersal experiments were not also conducted using Regiella-infected lines. At several points throughout the manuscript, the idea of using symbiont transfections to reduce plant harm is raised. I can understand that these experiments are likely time-, space-, and resource-intensive, but that seems like these would have been relevant experiments, especially in the context of controlling damage to plants.

Reviewer #3 (Public review):

Summary:

The authors were investigating the impact of introducing novel facultative bacterial endosymbionts into the pest aphid, Diuraphis noxia, to explore the possibility of using facultative symbionts as a crop protection tool. They successfully established the vertical transmission of both endosymbionts and performed a series of aphid performance and dispersal experiments together with measurement of aphid feeding on host plant health, growth, and metabolism. While most of the experiments revealed no effect of the endosymbionts, some significant treatment effects were found, showing that Rickettsiella reduced aphid dispersal, and Regiella reduced aphid population growth and feeding damage.

Strengths:

The team worked with two novel facultative symbionts (Rickettsiella viridis and Regiella insecticola) that they were able to successfully establish in D. noxia. The data were collected and analyzed using solid, well-described methodology.

Weaknesses:

While interpretation of the data is reasonable, the few experiments which revealed significant treatment effects rest on relatively small sample sizes.

Measuring symbiont density is difficult. The authors use quantitative PCR to measure the "density" of endosymbionts relative to a host gene. This is a standard approach in the field; however, recent work has shown that endosymbionts like the aphid primary endosymbiont, Buchnera, are variably polyploid [1]; further the aphid cells that house the symbionts are also highly polyploid and variable in their ploidy [2]. It is important to understand that what is being measured when using qPCR is DNA copy number and not quantification of the number of symbiont cells. Alternative approaches to measuring symbiont density include flow cytometry [3], and SymbiQuant [4], a machine vision tool that can quantitatively characterize symbiont populations from DAPI-stained confocal images. These alternate approaches also have their limitations. Currently, there is no perfect approach to measuring symbiont density, which remains an important measure in experiments such as these. Put simply, it is important for a reader to be aware of the limitations of each approach and interpret data accordingly.

[1] Komaki, K., and H. Ishikawa. 2000. Genomic copy number of intracellular bacterial symbionts of aphids varies in response to developmental stage and morph of their host. Insect Biochemistry and Molecular Biology 30:253-258.

[2] Nozaki, T., and S. Shigenobu. 2022. Ploidy dynamics in aphid host cells harboring bacterial symbionts. Scientific Reports 12:9111.

[3] Simonet, P., G. Duport, K. Gaget, M. Weiss-Gayet, S. Colella, G. Febvay, H. Charles, J. Viñuelas, A. Heddi, and F. Calevro. 2016. Direct flow cytometry measurements reveal a fine-tuning of symbiotic cell dynamics according to the host developmental needs in aphid symbiosis. Scientific Reports 6:19967.

[4] James, E. B., X. Pan, O. Schwartz, and A. C. C. Wilson. 2022. SymbiQuant: A machine learning object detection tool for polyploid independent estimates of endosymbiont population size. Frontiers in Microbiology 13:816608.

Impact and Significance:

Food security and production, and pest control are major challenges facing the human population. This work contributes knowledge that will benefit the development of alternate pest control strategies in agriculture.

Reviewer #1 (Public review):

[Editors' note: This revised version has been assessed by the Reviewing Editor without further input from the original reviewers. The authors have satisfactorily addressed the substantive comments raised in the previous round of review. The qualifications and limitations are now more clearly acknowledged in the revised manuscript.]

Summary:

This work compiles a comprehensive atlas of ncORFs across mammalian tissues and cell types, derived from reanalysis of ~400 public ribosome profiling datasets. The authors then evaluate cross-species conservation and functional signatures, proposing that evolutionarily ancient ncORFs tend to have higher translation potential, stronger expression, and closer relationships with canonical coding sequences.

Strengths:

In general, the study provides a large-scale and timely resource of annotated ncORFs, which could be broadly useful for the community. The authors collected ~400 public ribosome profiling datasets for annotations of ncORFs, which, to my best knowledge, is the largest collection of data for such purpose. The catalog could facilitate future investigations into ncORF biology and broaden understanding of the coding potential of the "non-coding" genome.

Comments on previous version:

The authors have made efforts to address most of the previous concerns, and several points have been clarified or improved in the revision. However, in a number of cases, the responses rely more on acknowledgment and reframing rather than substantive analytical strengthening. Overall, the manuscript is improved, particularly in terms of clarity, transparency, and positioning of claims. I support its publication and look forward to seeing how the field engages with and discusses these claims.

Reviewer #2 (Public review):

Summary:

Chang et al. attempted to analyze a large number of ribo-seq datasets through a standardized pipeline, identifying novel non-canonical ORFs and elucidating their evolutionary and expression characteristics.

Strengths:

(1) The datasets analyzed by the authors are sufficiently comprehensive, and the use of standardized pipelines ensures excellent analytical consistency.

(2) Their analyses of ORF evolution and co-expression further deepen our understanding of these ORFs.

Weaknesses:

(1) The authors primarily conducted analyses through bioinformatics, lacking sufficient wet-lab experimental evidence.

Reviewer #1 (Public review):

Summary:

The manuscript entitled "Autonomic reflex plasticity associates with time-dependent SUDEP susceptibility in a murine model with hyperactive stress circuits" by Dr. Saunders and colleagues combined a traditional mouse model of SUDEP, ventral intrahippocampal kainite (vIHKA) injection, with a genetic model of chronic hyperactivity of central corticotropin-releasing hormone (CRH) neurons (Kcc2/Crh) that further increases the risk of SUDEP in the weeks following seizure.

Strengths:

Their results show during spontaneous seizures Kcc2/Crh mice had more pronounced reflex-like ictal bradycardias compared to WT controls that notably occurred prior (~10 sec) to seizure termination and had greater autonomic disturbances compared to WT controls, including a pronounced serotonin-mediated Bezold Jarisch reflex. These results show chronic hyperactivity of central corticotropin-releasing hormone (CRH) neurons (Kcc2/Crh) increased autonomic disturbances and risk of SUDEP in a kainic acid model of epilepsy.

Weaknesses:

This study could be improved with a more thorough assessment of heart rate, blood pressure and breathing during and following the seizures, and in particular the fatal event. It is unclear if the bradycardias were spontaneous, or a result of preceding central or obstructive apneas, oxygen desaturations, hypercapnia, arrhythmias, or other possible triggers.

Considerable prior work in the literature suggests SUDEP could be mediated, in some patients, by a burst of parasympathetic activity to the heart. Were the heart rate changes in these animals during seizures inhibited or blocked by atropine, or atenolol? The injection of the 5HT agonist phenylbiguanide into the right jugular is not a selective approach for activating the Bezold Jarisch Reflex (BJR) which is caused by increased activity of intracardiac sensory neurons (generally activated with ischemia or a combination of low preload with high contractility). The results should be interpreted more cautiously, as a response to systemic administration of phenylbiguanide only.

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors set out to evaluate the role of hypothalamic pituitary axis hyperactivity on cardiac and autonomic changes during epileptogenesis and following seizures in a mouse model of temporal lobe epilepsy. Epilepsy is very common. It can frequently result in death from sudden unexpected death in epilepsy, or SUDEP. SUDEP is thought to be at least in part due to seizure related cardiac and autonomic instability. Increased stress states are well known to be comorbid with epilepsy. This comorbidity is thought to increase the risk of SUDEP. Here the authors hypothesized that a mouse model of heightened stress in which there is hyperactivity of the CRH neurons in the hypothalamus would demonstrate exaggerated cardiac and autonomic effects of seizures and epilepsy.

Strengths:

For the chronic stress model, they employed the Kcc2/Crh mice that have a genetic deletion of the potassium chloride cotransporter in CRH neurons. They treated these mice and their wild type littermates with intra hippocampal kainic acid or saline, as epileptic and sham-treated animals respectively. The assessed cardiac activity, blood pressure, baroreflex, and the Bezold-Jerisch reflex during epileptogenesis. This in general is an interesting study. They make some interesting and potentially important observations regarding heart rate and blood pressure in seizures and epilepsy.

Weaknesses:

While the revised manuscript is much improved, there are still some concerns that should be addressed.

(1) The low-pressure baroreceptor responses they show in Figure 4 are somewhat confusing. Should they not be seeing a reflex increase in heart rate when blood pressure is lowered with sodium nitroprusside? It would be helpful if they could describe whether the control responses were as expected or not, and if not, why? This makes it difficult to assess changes seen in the different genotypes and conditions.

(2) It does not seem appropriate to label the assessments associated with Figure 5 as the Bezold Jarisch Reflex. This reflex involves bradycardia, hypotension, vasoconstriction, and hypopnea. They seem to be only looking at the cardiac component, which is likely mediated though peripheral 5HT3 receptors. Did they measure blood pressure and breathing? Can they include these? If they can only comment on HR, then the discussion should reflect this.

(3) In Figure 1B, it would be helpful to show some short (e.g., 0.5 sec) snippets of ECG traces that exemplify the changes in HR (spikes/second).

(4) The day 21 examples given in Figure 1B, do not seem to be representative of the data depicted in Figure 1C.

(5) From the top panel examples in Figure 2A it looks like there might be greater EEG suppression following seizures in the Kcc2/CRH mice. Was this consistent? It might be worth looking into.

(6) Can the authors include scale bars for the top panels in Figure 2A?

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:

This work asks the question of how different organelles and structures in the apicomplexan parasite Toxoplasma gondii are recycled and/or segregated to the daughter cells during cell replication. In particular, they consider an unusual cell structure called the residual body that links replicating cells during the intracellular infection stage of this parasite. The residual body has historically been considered a 'dumping ground' for unnecessary relics of the mother cell during division, but this notion is increasingly being revised. Indeed, cell replication in Toxoplasma is often misinterpreted as cell division (cytokinesis), but in fact, the cell replicates its organelles and structures to multiple 10s of copies in seemingly distinctly formed daughter cells, but cytokinesis is delayed for many such cycles and typically only occurs simultaneously with parasite egress from its host cell. The residual body is, in fact, the connection between these pre-cytokinetic replicated daughters, and effectively, this is still a single cell at this stage. The authors have previously shown that an actin network extends through the residual body between these daughter cells, and ER and mitochondria common to all cells are also linked through this structure. This study examining the fates of organelles during cell replication is timely for continuing our understanding of how this fascinating component of the cell participates in these processes. The authors use Halo-tags as their principal tool to track discrete populations of proteins, labelling their organelle locations, and this provides beautiful insight into these processes.

Strengths:

Using dyes conjugated to Halo tags this work elegantly tracks the fates of proteins synthesised by an original 'mother' cell over several replication cycles of pre-cytokinetic 'daughters'. Using this tool, they show that some organelles are made intact just once and that some of these can be subsequently sorted to the daughters (micronemes and rhoptries) while others are dismantled (IMC) and the daughters must make their own. A third set of organelles (largely synthesis, sorting and metabolic compartments) are divided and inherited, and new daughter-synthesised proteins are added to the preexisting maternal proteins in these structures. A role for actin and myosin is clearly demonstrated for micronemes and rhoptries, and this correlates with their relatively late inheritance into the developing daughters. Overall, this work gives clarity to the behaviours of several cell structures during replication and paves the way to better understanding the mechanisms that drive the differences between structures and the universality of these processes in other apicomplexan parasites. In particular, this study shows that the residue body is a region of the cell syncytium that organelles can be actively transported from. Therefore, it is a space that can actively contribute to the segregation of the late segregating micronemes and rhoptries.

Reviewer #2 (Public review):

Summary:

Toxoplasma gondii is an obligate intracellular parasite and the causative agent of toxoplasmosis. Parasite invasion of host cells, intracellular replication, and subsequent egress, which results in destruction of the infected cell, are central to pathogenicity. This manuscript focuses on understanding how maternal resources, specifically cellular organelles, are shared between daughter parasites during cell division. Many organelles are present as a single copy, making their division and inheritance essential for successful replication. In T. gondii, our understanding of how organelles are divided during cell division remains limited, and this study helps address this important knowledge gap.

Strengths:

The major strength of this study is the use of a Halo-based pulse-chase assay to characterize patterns of organelle inheritance and to monitor protein synthesis, turnover, and movement. This approach will be of considerable interest to the field. Using this method, the authors identify three major modes of organelle inheritance:

(1) Organelles present in multiple copies (such as micronemes and rhoptries) are partitioned between daughter parasites, with additional contributions from newly formed vesicles. Newly synthesized and pre-existing material remain as distinct populations within the cell.

(2) Single-copy organelles, such as the Golgi and apicoplast, are expanded through the incorporation of newly synthesized material before division.

(3) Cytoskeletal structures are synthesized de novo during each round of cell division.

These findings provide a more refined understanding of organelle inheritance and demonstrate that secretory organelles are not generated entirely de novo during each round of division, as was previously thought.

The paper places particular emphasis on the fate of maternal micronemes and rhoptries during division. The data show that (1) during division in wild-type cells, maternal micronemes and rhoptries are detectable in the residual body (RB); however, the majority of these organelles are localized within the parasite body, either at the apical or basal ends of the daughter parasites (Fig. 6). (2) In the absence of the myosin motor MyoF, micronemes and rhoptries accumulate in the residual body and are not properly trafficked to the daughter cells. Upon restoration of MyoF protein levels, these organelles redistribute to the daughter cells, although in an uneven manner.

Weaknesses:

The second half of the paper focuses on a more detailed characterization of microneme and rhoptry recycling. While the authors propose that the RB acts as a central hub for recycling both organelles, the current data do not fully support this conclusion.<br /> The model that microneme and rhoptry recycling is RB-dependent relies largely on the MyoF depletion phenotype and the limited detection of maternal organelles in the RB of wild-type parasites. Alternative models remain plausible, including direct trafficking to daughter cells, with RB accumulation upon MyoF depletion reflecting impaired trafficking rather than an obligatory RB-dependent recycling pathway, as now discussed by the authors.

Reviewer #3 (Public review):

Summary:

Knoerzer-Suckow et al. explore the mechanisms of organelle inheritance during endodyogeny in Toxoplasma gondii using an innovative dual-labeling approach to track the distribution of maternal organelles into daughter parasites. They can clearly distinguish between maternal and daughter-derived organelles using their dual-labeling Halo Tag approach. They reveal that different organelles are trafficked to daughter parasites in three broad patterns they have binned into groups. Their findings reveal a role for MyoF in the inheritance of micronemes and rhoptries, and notably, they observe that the inner membrane complex (IMC) is not recycled. Instead, the IMC undergoes a pronounced relocalization to the posterior of the maternal cell, where it is likely targeted for degradation.

Strength:

The data surrounding their MyoF knockdown experiments, IMC degradation, and trafficking of MIC2 after auxin washout are convincing. These data add to the knowledge of how organelle inheritance occurs in T. gondii, increasing the field's understanding of endodyogeny.

Weakness:

The inability to achieve higher temporal resolution due to phototoxicity precluded tracking of single micronemes, thus it remains possible that some micronemes follow a path similar to rhoptries and enter daughter cells before development of the residual body while others are recycled via the residual body.

Reviewer #1 (Public review):

Summary:

This study demonstrates that nutrient resorption efficiency (NuRE) in Phragmites australis is genetically canalized rather than plastic to salt stress. Using 110 genotypes in a common garden, the authors show that intraspecific variation in NuRE is explained by phylogeographic lineage, ecotype, and latitude, not by effective salinity. Element specific regulatory strategies further reveal how N, P, and K resorption are differentially controlled. At the population level, this is an important study that fundamentally advances our understanding of plant functional trait evolution and its implications for ecosystem nutrient dynamics under global change.

Strengths:

This study is the first to demonstrate genetic determination of a key nutrient conservation trait under effective salt stress in a widespread macrophyte, directly testing the 'plastic acclimation versus inherent conservatism' paradigm in a non-nutrient stress context. The experimental design is rigorous: each genotype was paired across control and salt treatments, and multilevel stress effectiveness (metabolomics, biomass, Na accumulation) was confirmed before evaluating NuRE. The large sample size of a macrophyte and dual classification (phylogeography + ecotype) allow robust disentangling of genetic versus plastic sources of variation.

The analysis comprehensively tests three resorption control hypotheses using appropriate SMA regression, revealing element specific and condition dependent patterns. The latitudinal gradient and variation partitioning provide strong evidence that genetic origin and geographic context outweigh short term plasticity, with important implications for predicting ecosystem nutrient cycling under global change. This study provides a clear empirical demonstration that a key nutrient conservation trait can remain homeostatic under non nutrient stress, and that intraspecific variation is primarily a product of population differentiation rather than short term plasticity.

Weaknesses:

First, the salinity treatment spanned only one growing season. The conclusion of genetic canalization therefore specifically refers to the absence of plasticity to an acute salt shock. Whether long term, multigenerational chronic salinity could act as a selective agent or induce transgenerational plasticity remains an open and interesting question for further research. Likewise, the physiological mechanisms underlying the observed lack of plastic increase in NuRE (for example, phloem loading or senescence gene expression) are not directly resolved, leaving some inference about trade-offs versus true unresponsiveness. These points do not weaken the study's main conclusion. Instead, they suggest productive future directions, such as longer-term field manipulations and targeted molecular investigations.

Second, the test of nutrient limitation control relies on resorbed N:P and N:K ratios as proxies, an established but indirect approach. Direct nutrient addition experiments would provide stronger causal evidence. Also, the metabolomic analysis is used primarily to validate stress effectiveness; deeper integration of specific metabolites with NuRE variation across genotypes could have offered mechanistic insights but was not pursued. Additionally, the potential collinearity between ecotype and phylogeographic lineage among Chinese populations is not quantitatively addressed. None of these considerations undermine the main finding, which is supported by a robust experimental design and widely accepted analytical approaches.

Comments on revised version.

The author carefully revised the parts that might cause confusion.

Reviewer #2 (Public review):

Summary:

The manuscript shows that nutrient resorption efficiency in Phragmites australis is largely genetically canalized rather than plastic under effective salt stress, with variation mainly associated with phylogeographic lineage, ecotype, and latitude.

Strengths:

The common-garden experiment with 110 genotypes, paired control and salt treatments, multi-level stress validation, and element-specific analyses provides compelling support for the central conclusions.

Weaknesses:

The generality is limited by the focus on a single species and a single growing season, and some mechanisms are inferred indirectly.

Reviewer #1 (Public review):

Summary:

In this study, the authors trained mice to perform a memory-guided navigation task, in which they must navigate to a previously cued arm after a delay period. They optogenetically inhibited the dorsal hippocampus or mPFC (targeting PL) in different task epochs. They found that for both regions, inactivating at the beginning of the navigation epoch impaired performance and induced mice to revert to habitual side biases. Inactivating during other epochs, including a delay period before the navigation phase, had little or no impact on behavior. The relationship between trial duration and behavioral performance was differentially impacted by hippocampal and mPFC inactivation, suggesting that the nature of the deficits was somewhat different.

Strengths:

The effects of perturbations are robust across animals and generally convincing. The lack of effect at some task epochs serves as a nice internal control. The finding that hippocampus and mPFC inactivation produced subtly different effects is interesting.

Weaknesses:

The simplicity of the behavior makes it difficult to resolve how exactly the hippocampus and mPFC contribute to working memory. Also, the language does not always reflect the trends in the data: the authors claim that optogenetic perturbations cause mice to repeat previous choices, but the data show that perturbations increase the likelihood of choosing a preferred side (which is left for most mice). A side bias is not the same as choice repetition. This has implications for interpreting the nature of the behavioral effects.

Reviewer #2 (Public review):

Summary:

The study uses transient optogenetic silencing of the dorsal hippocampus or prefrontal cortex in mice using a delayed response task in a T-maze, with a temporal delay of 1s after a visual cue, followed by a central stem run period before choice execution. Silencing of either the dorsal hippocampus or the dorsal prefrontal cortex is executed for varying time periods during the stem/ central arm running epoch or the 1s temporal delay epoch by targeting either PV+ or Dlx interneurons in a block-wise or random-trial design. The main result reported is that silencing of either region during long periods of stem running impaired choice behavior, and silencing during the temporal delay period did not have an effect.

Strengths:

The major strength of the study is using the optogenetic silencing strategy to target different temporal periods of the task.

Weaknesses:

(1) A major weakness of the study is the lack of a balanced design with equal time periods of silencing during the stem running period and temporal delay period in many of the animals, which precludes any conclusion about distinct functional roles of the regions during these two phases of the task. The central question of this study is not new, with many previous studies investigating distinct and overlapping roles of hippocampus and prefrontal cortex in spatial working memory tasks and memory-guided navigation, using inactivation of one or both regions, crossed inactivation approaches, as well as targeting direct and indirect connections between the regions (PMIDs: 20074655, 9030646, 17045348, 30179661, 27511010, 10491611, 26017312, etc.), in addition to several physiology studies. A key extension for the current study would have been to show a distinction between roles in the temporal delay period after the cue and the stem-running working memory period. However, the inactivation period during stem running shows effects only for long inactivation periods, 2s initial periods and later ~1.6s periods (run after 0.8s, >2/3 total running periods), whereas the temporal delay period inactivation is 1s for the large majority of animals, which is a clear mismatch in inactivation periods, obviating this conclusion of distinction.

(2) It is not clarified why such short delay periods were used compared to long periods of ~10s in T-maze spatial alternation tasks with delay, and whether the temporal delay period of 1 sec is strictly distinct from the spatiotemporal delay period during stem running in terms of short-term memory function. The choice of time windows for inactivation needs to be better justified, which currently appears to be rather random (2s initial running period, run after 0.8s, run after 1.6s; for an average reported running period of ~2.4-2.5s). The ideal design clearly would have been to use a 2s temporal delay period so that the inactivation time in this epoch matched the initial 2s run period. Only a subset of 4 animals were run with a longer temporal delay period, and too with only with hippocampal inactivation (Figure 4d). The main conclusion of distinction between temporal delay and stem running delay periods is therefore not adequately tested for the prefrontal cortex, and the statistics in terms of number of animals for this important control for hippocampal inactivation are also not comparable to the main experiment.

(3) The mixture of PV-Cre animals (5 animals), Dlx targeting (2 animals), and one WT animal is also suggestive of a fragmented approach, and inactivation efficacy cannot be assumed to be similar for different animals. Importantly, there is no physiological evidence for confirmation of suppression in the optogenetic experiments, even in exemplar animals.

Reviewer #3 (Public review):

The authors sought to determine when the dorsal hippocampus and mPFC are causally required during a delayed spatial working memory task. Using temporally precise optogenetic silencing in mice performing a delayed cue-guided T-maze task, they tested the effects of transient perturbations during distinct behavioral epochs. Contrary to the common view that these regions are primarily required during the delay period to maintain working memory representations, they report that silencing during the delay had little effect on performance, whereas perturbation during the early phase of central-arm traversal consistently impaired performance. The authors conclude that hippocampal and prefrontal contributions to memory-guided behavior are dynamically engaged during active navigation rather than passive maintenance of information.

The study has several strengths. The behavioral paradigm is carefully designed to dissociate cue, delay, and movement epochs, allowing temporally specific causal manipulations. The systematic comparison of multiple task epochs represents a major strength and provides compelling evidence that the behavioral effects of perturbation are epoch-dependent. The authors also include several important controls, including stimulation during multiple task phases, a longer-delay condition to dissociate task epoch from elapsed time since cue presentation, and analyses of movement trajectories and perseverative behavior that provide additional insight into the nature of the behavioral deficits. Together, these experiments convincingly demonstrate that transient dorsal hippocampal and mPFC perturbations have markedly different behavioral consequences depending on when they occur within a trial.

The evidence is generally solid and supports the primary finding that perturbations during early navigation produce larger impairments than perturbations during the delay period. However, some aspects of the broader interpretation are less well supported. Most notably, the study lacks a non-memory control task, such as a visually guided version of the maze, making it difficult to determine whether the observed deficits specifically reflect disruption of memory-guided behavior or more general impairments in action selection, behavioral flexibility, or movement planning. The observed increase in perseverative responding and delayed commitment to a turn are consistent with either interpretation. In addition, several experimental conditions rely on relatively small numbers of animals, limiting confidence in some negative findings, particularly for the longer-delay and later-run manipulations. Finally, while the Discussion proposes that hippocampal-prefrontal circuits become engaged during the transformation of stored information into action, this mechanistic interpretation remains speculative because no neural recordings accompany the causal manipulations.

Overall, the authors achieve their primary aim of demonstrating that the behavioral consequences of dorsal hippocampal and mPFC silencing depend strongly on task epoch. The data convincingly support the conclusion that these structures are more vulnerable to perturbation during early navigation than during the brief delay period used in this task. The broader conclusion that these findings redefine when hippocampal-prefrontal circuits support working memory should be interpreted more cautiously, as alternative explanations involving action selection or behavioral state remain plausible in the absence of additional control tasks.

The findings are potentially important because they challenge the common assumption that hippocampal and prefrontal contributions to delayed-response tasks are centered on delay-period maintenance. Instead, the work supports the idea that these circuits may be recruited when remembered information is translated into goal-directed behavior. This framework is broadly consistent with recent distributed models of working memory and provides an interesting perspective that may help reconcile previous studies reporting effects during different task phases. The behavioral paradigm and temporally precise perturbation approach should also be useful for future studies aimed at dissecting the dynamic contributions of hippocampal-prefrontal circuits during memory-guided behavior.

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:

This manuscript uses sci-L3-Strand-seq to map sister chromatid exchange events following CRISPR/Cas9-induced DNA damage. Because exchanges between identical sister chromatids are largely invisible to conventional sequencing, the study addresses an important blind spot in the assessment of genome editing outcomes. The authors compare single-locus Cas9 cleavage, simultaneous targeting of 237 repetitive genomic sites, and Cas9 nickase variants. They further use reciprocal daughter-cell pair analysis to ask whether Cas9-associated SCEs are copy-neutral or linked to larger structural alterations. Overall, this is a valuable study that introduces an important additional layer to the analysis of CRISPR/Cas9 repair outcomes. The central finding that Cas9-induced DSBs can trigger frequent local SCE is well supported and likely to be of broad interest.

Strengths:

The major strength of the manuscript is the application of a strand-resolved, single-cell method to a question that is difficult to address with standard genome sequencing. The evidence that a single Cas9-induced DSB can trigger strong local SCE is compelling in concept and supported by multiple guide RNAs targeting distinct loci. The reported on-target SCE frequencies, reaching up to 41%, suggest that inter-sister exchange is a substantial and underappreciated outcome of Cas9 cleavage.

Of particular interest is the comparison between single-site and multi-site targeting. The finding that 237 programmed Cas9 targets produce only mild bulk enrichment of on-target SCE but stronger enrichment in a subpopulation of cells with elevated SCE burden is interesting and may have wider biological implications, particularly if the findings extend beyond Cas9-induced SCE to spontaneous SCEs.

The reciprocal daughter-cell pair analysis is another notable feature of the study. The observation that some Cas9-associated SCEs are accompanied by structural alterations could challenge the assumption that SCE after a programmed break reflects error-free homologous recombination.

Reviewer #2 (Public review):

Summary:

In this short paper, a clever single-cell Strand-seq method was used to study the number and location of sister chromatid exchange events (SCEs) in cells after CRISPR/Cas9-induced DNA double-strand breaks (DSBs). Unique as well as multiple genomic loci were targeted. Cas9-induced cuts at unique genomic locations led to statistical enrichment of SCEs at the target site, whereas Cas9 targeted at repetitive targets revealed only mild enrichment of on-target SCEs unless analysis was restricted to a subset of cells with >8 SCEs per cell. Interestingly, reciprocal daughter-cell pair analysis revealed large-scale structural alterations on some chromosomes. Whereas disruption of DNA repair genes, including LIG3, LIG4, XRCC1, and XRCC4, did not measurably alter SCE frequency per cell within 24 hrs, consistent with delayed functional loss following editing and selection against essential genes. Together, these findings demonstrate that Cas9-induced DSBs are potent local triggers of SCE at unique loci and can be associated with structural alterations, highlighting the influence of lesion type and genomic context on recombination outcomes during genome editing.

Strengths:

The data in this paper represent a very rich resource of how parental DNA template strands are distributed in paired daughter cells after various treatments. Abnormalities observed in only one of such paired daughter cells provide a novel and exciting approach to study mechanisms of DNA instability and DNA repair at a genome-wide level in general and following Cas9-induced DSB in particular.

Weaknesses:

The effect of Cas9-induced DSBs in the cells that are used will depend on the cell cycle stage of the cells that are targeted, as well as the number of times cuts are made. The latter could happen before, during, and after DNA repair reactions on one or both alleles in a diploid cell. As a result, it is very difficult to extrapolate the mechanisms of DNA instability and DNA repair from the observed genomic rearrangements. Novel approaches are needed to limit the number and timing of Cas9-induced breaks to overcome some of these limitations.

Reviewer #3 (Public review):

Chovanec and Yin used their newly developed sci-L3-Strand-seq powerful method to characterize SCE after Cas9 cleavage in a human cell line, using either a single target site or an element repeated 237 times in the genome. SCE are often neglected in DNA repair analyses since they are « genetically silent ». Interestingly, the authors found enrichment of SCE at unique Cas9 sites, but only a modest enrichment of SCE when Cas9 targets 237 sites in the genome. The genetic control of SCE formation at Cas9 sites is not deliberately addressed in this paper. However, the authors found that targeted SCE seem to be enriched in a subpopulation of cells, particularly « permissive » for SCE, but the determinants of such a population are unknown. Finally, the power of the sci-L3-Strand-seq allowed the authors to characterize a specific type of SCE based on the analysis of reciprocal daughter-cell pairs' genomes that is associated with a specific type of chromosomal rearrangement compatible with the ones observed in HR defective BRCA1/2 deficient cells.

Reviewer #1 (Public review):

In this manuscript, the authors present ClustIRR, a computational tool that analyzes multiple TCR repertoires together, instead of one at a time. It builds one shared similarity graph across all the repertoires, then finds communities (CJs) on that graph that can be compared across samples. Building one shared graph across repertoires, rather than clustering each repertoire on its own, is a real improvement over existing tools. However, there are a few concerns that need to be addressed.

(1) The Introduction motivates ClustIRR by contrasting it with beta-binomial regression, Fisher's exact test, and scCODA/tascCODA (lines 62-77), but none of these are actually run on the same data in the Results. Could the authors include a direct comparison, e.g., applying a standard beta-binomial or Fisher's exact test to the Dataset 1 CJ occupancy matrix, to show the Bayesian model gives a lower false-positive rate or better-calibrated intervals than the alternatives it's positioned against?

(2) Prior predictive and posterior predictive checks are both reported as "(data not shown)" (lines 644, 728). Since the manuscript's central claim is rigorous, uncertainty-aware inference, it would help to include these diagnostic plots, along with Rhat and ESS values, in the supplement rather than stating they were checked.

(3) With 7,505 CJs tested simultaneously for differential occupancy in Dataset 1 alone (Figure 1 legend), what is the expected false discovery rate under the non-overlapping-HDI criterion used throughout? A short discussion of multiple-comparisons correction, or an argument for why it isn't needed under this framework, would strengthen the statistical claims.

(4) Line 115 states the Dataset 1 joint graph produced 10,301 CJs, of which 3,038 were singletons, leaving 7,263 non-singleton CJs. The Figure 1 legend reports 7,505 CJs used for the β modeling. Could the authors clarify how these two numbers relate - whether some singletons were included in the model, or a filtering step was applied that isn't described in Methods?

(5) The Methods section states that archival pretreatment tumor tissue was available for four patients (Pt4, Pt32, Pt36, Pt38; line 531), but the T+/T- DCJ analysis in Fig. 3B-C is only shown for Pt4. Was this analysis attempted in the other three patients? Extending it, even partially, would substantially strengthen the claim that contracting DCJs are enriched for tumor-infiltrating TCRs, which is currently based on a single patient.

(6) Dataset 1 was generated by deliberately stimulating T cells with EBV or MART1 antigen, so recovering EBV/MART1-annotated CJs from VDJdb is closer to a positive control than a blinded validation. Do the authors have, or could they obtain, any independent confirmation (e.g., tetramer data or an orthogonal cohort) for the CJs with large β that lack VDJdb annotation (orange dots, Figures 1B-C)?

Reviewer #2 (Public review):

Summary:

The study confronts a major obstacle in repertoire analysis. Given that individual TCR sequences are diverse and sparsely detected across repertoires, identifying sample-specific enrichment of TCRs based on sample-to-sample comparison of exact clonotype sequences can be intractable. This paper attempts to build on insights that sequence-similar TCRs can share antigen recognition, such that aggregating similar sequences derived from multi-sample joint-graph communities (i.e. clusters of tightly connected nodes) could reduce sparsity and boost signal.

Strengths:

The study is a well-motivated effort to address a need in the field. The paper takes a unique approach. The core method is modeling community occupancy with a hierarchical Dirichlet-Multinomial model that attempts to account for the high level of overdispersion present in repertoire sampling, a technique that has been previously applied to compositional microbiome data.

The authors apply this framework to both single-cell paired-chain and bulk single-chain TCR data. They develop a set of examples from public and synthetic data, with the most promising real-world application shown in reanalysis of longitudinal data during treatment of cancer patients with checkpoint inhibitors.

The manuscript is well structured and cogent. The authors are to be commended for contributing a well-documented, open-source R/Bioconductor package and for providing the underlying analysis datasets in a well-organized repository. In the joint graph construction step, the authors opt to use an existing implementation of the BLAST algorithm on CDR3 sequences, which is a slightly odd choice since it ignores potential contributions of other V-gene germline-encoded CDRs, but the authors also envision that their statistical package could be extended to include community graphs developed with other established TCR clustering tools. This will allow others to potentially explore the utility of Bayesian hierarchical Dirichlet-Multinomial models for differential occupancy analysis of immune receptors under varied clustering criteria.

The methods described here were demonstrated on relatively small datasets from 2-5 samples, and future work is likely needed to extend the joint graph differential occupancy concept to larger datasets. The authors are transparent about this and some of the other limitations in their current tool, most notably the computational cost of graph construction based on an all-versus-all sequence alignment to construct a joint sequence graph and the challenge of Bayesian parameter estimation as the number of subgraph entities scales with input data size. Since efficient approximate methods exist to find edges between similar text strings, the underlying idea of applying uncertainty-aware statistical inference to subgraph communities is promising, and the paper advances its primary goal.

Weaknesses:

The paper proposes the utility of the joint-graph community occupancy framework through three examples. I discuss potential weaknesses apparent in each separate example in turn.

(1) Weaknesses in Example 1

A broad weakness of the first results section ("Detecting EBV- and MART1-antigen reactive T cell communities from single cell datasets") is its reliance on a single vendor-generated dataset generated by the company ParseBio with no published experimental protocols and limited, if any, prior peer review. For reasons I will explore in greater detail below, the EBV-sample data may be particularly prone to chimeric pairings that confound the authors' primary analysis goals, and, at the very least, may not reflect physiologically realistic conditions for identifying antigen-reactive TCRs in other contexts.

Let us first consider Dataset 1 in more detail. The authors compare 2 antigen-stimulated repertoires with 3 unstimulated controls. To improve on single clonotype-level comparisons, the authors propose comparing the cell count aggregated across cells within joint graph communities constructed from paired CDR3 sequences across all the samples. Thus, one of the most relevant questions one hopes the authors answer in this section is whether the resolved graph communities are made up of many distinct clonotypes (i.e., are they polyclonal), allowing the method to function as intended by aggregating across multiple clones with putative shared antigen-reactivity.

Supplementary Figure 1B shows the size of all the communities with callouts for the putative EBV-expanded communities. The authors listed communities strongly enriched in the EBV-stimulated sample as e1, e2, e3, and e5. Each contains {greater than or equal to}100 clonotypes, and the authors note they contain at least one clonotype with a CDR3 sequence matching an EBV-annotated clone in VDJdb - a database of TCRs with some experimental evidence of epitope-reactivity. Community "e2" is notable for its remarkable size, including 1,339 unique clonotypes. At first glance, this seems promising for a method attempting to boost signal through community detection. However, it is worth re-investigating the individual clone sizes and sequences within this extraordinary community.

In the EBV-associated community "e2", a look at the data provided by the authors on the paper's GitHub repository suggests a single clonotype (clonotype_9; TRAV12-3 CATQGSNDYKLSF / TRBV9 CASSTGQVATNEKLFF) comprises 26,917 cells. As such, it makes up 29% of the sample with a total of 90,588 cells. If one clone supplies most of a graph community's cell counts, the community-level posterior estimate of β (Figure 1B) effectively tracks a single-clone estimate, and the premise of borrowing statistical power across a polyclonal expansion in this example is hard to assess.

There is also considerable evidence to believe that the apparent mega-polyclonality of cluster e2 may be partially an illusion, stemming from an artifact of this hyper-expanded clone's massive size and the experimental method used to assign TCRα-TCRβ chain pairings. In fact, the same α-chain CDR3 (CATQGSNDYKLSF) appears in ~1,219 clonotypes paired to distinct β chains, generating much of e2's remarkable 1,339-clonotype count. I believe two features warrant caution here. First, a single clonotype making up ~29% of total T cells in the sample is highly unexpected in ex vivo repertoires, suggesting intense non-physiological expansion conditions unlikely to generalize to other settings. That is, one would almost never expect to see a signal this strong.

Second, one TCR-α chain paired to ~1,220 distinct and diverse β chains in one sample is also biologically unexpected given what we know of the best-characterized epitope-specific responses for EBV, including to the well-known HLA-A*02 EBV BMLF-1 epitope, which recruits a tetramer-stained repertoire with conserved CDR3 motifs in both chains and at least some constraint on favored V-gene/α-β pairing (See Extended Data Figure 5 in Dash et al., Nature 2017). This raises the concerning possibility of barcode collision and mispairing against a hyperexpanded clone in the ParseBio split-pool method, unlikely to be robust to a clone occupying a third of the sample. Most of the ~1,219 β chains paired to the dominant α have a cell count of 1, further raising the concern of artifactual pairing versus genuine convergence. The second largest community "e1" also seems to suffer from the same issue, with a TCRβ sequence from one super clone making up 7% of the sample potentially being artifactually over-paired to >500 rare single-cell-count TCR α chains.

Taken together, these observations suggest that the authors' first positive control example passes but probably for the wrong reason since at least some of the "antigen-specific communities" are strongly anchored by a single hyper-clone. This could be remedied by repeating the same type of analysis on an ex vivo single-cell repertoire following more modest stimulation or natural infection (e.g., yellow-fever vaccine, influenza, or SARS-CoV-2 single-cell TCR datasets). I would advise future work using an alternative data source with better-documented experimental protocols, given the concerns above.

(2) Weaknesses in Example 2

Example 2 explores the application of Bayesian methods for identifying sample-enriched joint graph communities found in longitudinal data from many participants at two time points and longitudinal data from 1 person (Pt4) at 5 time points. A potential weakness of Case Study 2 is that it yields limited additional biological insight compared to what was previously shown by the authors of the underlying input data. Previously, Formenti et al. 2018 showed that the number of expanded clones after treatment strongly reflects responder status in this cohort, greater in patients with CR/PR versus SD or PD (See Figure 2b of the study). This 2018 primary analysis showed that tracking individual clonotypes was sufficient to reveal biological insight without the need for the computational demands of constructing a massive sequence similarity graph, finding communities on joint graphs, or Bayesian statistical inference. Thus, the impact of Case Study 2 in proving the unique utility of ClustIRR is somewhat diminished.

It is not clear how this study's result is more "robust" than the original analysis. Perhaps the authors could further clarify what is learned from the uncertainty-aware approach that could not be learned from exact clone tracking in time series.

Thus, example 2 shows that a complex method recapitulated the finding of a much simpler method for analyzing longitudinal TCR data where a strong signal of expansion was already present at the single-clonotype level. Since the Dirichlet method is sensitive to absolute counts, the large expanding clone in each community at 22 days may alone have carried most of the signal, which is not fully explored.

In this section, the authors make an interesting observation that CDR3β detected in both PBMC and patient-matched tumor samples were enriched in contracting communities (7/14) versus expanded communities (1/41). The authors do not indicate whether a similar enrichment of tumor-infiltrating lymphocytes (TILs) matched Day 0-22 contracting communities in the other 4 patients with tumor-matched samples, which, if consistent, would strengthen their finding.

(3) Weaknesses in Example 3

The final example explores "convergent repertoire differences between species." This is intriguing in principle but less informative due to the use of synthetic data with somewhat predictable properties that some may reasonably consider baked in by the data-generating process. That is, some of the results might be guaranteed by the way the data is constructed using the OLGA/IGoR generative model. For instance, the authors observe a positive correlation between CJ community size and Pgen of constituent clonotypes, stating: "This indicates that CDR3 sequences with high Pgen are statistically more likely to be generated, leading to convergence of similar sequences into public CJs." I may be mistaken, but this conclusion is almost guaranteed by the way the data-generating OLGA model outputs more similar high-Pgen sequences and fewer lower-Pgen sequences.

An interesting finding in this section is shown in Figure 4B, where community-level aggregation allowed for discrete clustering of human samples away from mouse samples that was not possible by comparing cosine similarity of a sparse clonotype occurrence matrix, a result that would be higher impact if it could be shown to separate real repertoire samples from humans with differential serology, vaccination status, or HLA backgrounds.

(4) Weaknesses in General

More generally, one aspect of the method that seems under-emphasized is the fact that many of the nodes in a multi-sample joint sequence similarity graph may have no edges. These zero-degree nodes would probably frequently occur in only one sample but be absent in other samples. It is not strongly emphasized in the paper how the model would infer whether such a singleton found in only one sample in the graph could be reliably inferred to be sample-specific enriched (see, for example, the large single node in Figure 3D, the orange node labeled "GQYF" in the far-right position of the lowest row in panel D). Presumably the number of cell counts represented in this single-sequence node is so great at sampled timepoint Day 22 as to yield a statistically strong signal in the multinomial model; however, the authors may wish to comment on how, for such singleton sequences, the power to detect sample-specific enrichment differs from prior single-clonotype-based methods.

With any large effort to find statistically significant features from a large candidate set examined all at once, a reader might be concerned with the potential for false discovery. Throughout, the authors seem to assign statistical significance when the 95% high-density interval (HDI) of the posterior estimate excludes zero or when the 95% HDIs of two features do not overlap. There is little discussion of how this implicitly handles multiplicity adjustment via shrinkage, which the authors could address more directly and explain more clearly to a broad audience, including many non-statisticians, who will read this paper.

Reviewer #3 (Public review):

Summary:

Analysis of immune receptor repertoires (IRR) needs to take into account the underlying diversity of the repertoires analysed, and the limitations inherent to the technologies used to measure IRRs: limited sampling depth relative to total number of cells and clonotypes, and experimental noise. In this work, Kitanovski and colleagues present ClustIRR. ClustIRR proposes to improve the analysis of immune receptor repertoires, specifically TCRs in the presented applications, by performing two steps: (1) consistent and comparable sequence clustering across repertoires, to account for sparsity of sampling, and (2) estimation of sequence clusters of interest using a Bayesian approach. They showcase ClustIRR performance in 3 scenarios: detection of antigen-specific T cells in a peptide-stimulation, detection of T cells responding to immunotherapy in the context of lung cancer, and analysis of mouse and human T cell repertoires.

Strengths:

(1) The cluster occupancy calculation presents an important conceptual framework which would be of interest and useful to the TCR repertoire field as it smoothly integrates information over a set of experimental conditions or time points. The application to longitudinal TCR sequencing datasets is particularly interesting, and could easily be extended to BCR sequencing datasets. Moreover, the calculation can be performed with any user-defined grouping of TCR clones, which allows for usage of other existing methods as the user wishes.

(2) The results of the human and mouse repertoires provide a very insightful argument for the use of metaclones as opposed to single clones for analysis of repertoires compared to single clones.

Weaknesses:

(1) While ClustIRR provides an interesting framework to analyse TCR sequencing datasets, it is not clear whether ClustIRR can perform more informative sequence clustering than state-of-the-art methods. A comparison of obtained clusters with existing methods would provide a useful benchmark. Moreover, computation time scales quite fast with the number of sequences included. This is a major limitation, as the authors state that a time of 2.5 hours is required for clustering of ~100,000 sequences, a number of clones that can easily be reached when analysing multiple samples together.

(2) The authors claim in the abstract that ClustIRR is integrated with gene expression data. However, in the results presented, the gene expression and TCR sequencing data are analysed separately, and the results are simply correlated. No real integration in the analysis exists for these two data types. The claim should be removed from the abstract. Moreover, the differential gene expression section, while it presents interesting results, lacks clarity and transparency. Presentation of the data in more transparent ways (such as showing violin plots or clustering on the UMAP) would increase the strength of the claims.

(3) The score calculation does not seem to be normalized to take into account the underlying diversities of CDR3a and CDR3b. While still a useful metric for sequence clustering, I worry about the impact of the lack of normalisation on the conclusion that the "alpha chain drives functional convergence through germline bias". The observation that clustering is mostly driven by the J gene is known and expected (https://pmc.ncbi.nlm.nih.gov/articles/PMC5553937/), as the J gene has lower diversity and greater overlap with the definition of CDR3. Because CDR3a is lower diversity generally from CDR3b, it will likely dominate the sequence similarity signal. Thus, it will appear that Ja drives the signal. The authors do try to address this by looking at clusters driven by CDR3b similarity. However, a low number of CDR3b-driven clusters is consistent with the similarity definition. I wonder if instead the appropriate control for this analysis would be to run the same analysis disregarding CDR3a sequence altogether, and quantify whether similar species-specific DCJ are identified when only CDR3b similarity is used? Absence or reduction of species-specific clustering would confirm that the effect is driven by the CDR3a sequence.

Reviewer #1 (Public review):

This meta-analysis addresses long-standing questions about the reliability and interpretation of choice probability in macaque visual areas, and provides some important findings (e.g., the cross-study consistency of the CP-sensitivity relationship, V1 distinctiveness). However, the evidence for several claims is incomplete: the analysis does not consider the statistical dependence of data points from the same studies and monkeys, and both the bistable-stimulus effect and the stimulus-duration effect rely on interpretive assumptions.

Strengths:

The paper's transparency about its own limitations is a genuine strength. Several sections of the paper and the supplement report null results (task exposure, lapse rate, eccentricity) rather than omitting them. This kind of self-scrutiny is uncommon in meta-analyses and substantially increases confidence in the parts of the analysis that do hold up.

Weaknesses:

(1) No mixed/hierarchical statistical models for nested data. The paper considers 150 data points from 59 studies and treats them as independent samples, though many share monkeys and brain areas. This reduces the confidence in the reported p-values. A standard way of dealing with this would be to use mixed-effect models with random intercepts rather than OLS.

(2) Evidence for one of the main findings in the abstract ("First, CPs were higher in tasks involving bistable percepts, reinforcing the link between CP magnitude and subjective perception.") is weak. This effect relies entirely on studies using bistable rotating cylinder stimuli performed in a single lab (lines 702 - 711). I would suggest making this more explicit in the abstract / discussion and in Figure 8b,c.

(3) The interpretation of main drivers of CP is unclear. The discussion summarizes the 4 main drivers of CP as "four systematic drivers of this variability: neuronal sensi758tivity, brain area, stimulus duration, and task type." The independent contribution of task type is however, questionable. In line 623, it is stated that the difference between coarse and fine discrimination can be entirely explained by the difference in sensitivity (explained possibly by differences in optimizing the stimuli). Again, detection tasks (line 658) show a trend for higher CP because most studies used tailored stimuli from single-recording experiments. Bistable task: see point 2. Thus, all "task effects" can be attributed to confounds, and the claim of the "four drivers of CP" should be revised.

(4) The paper could be improved by a Discussion that synthesizes the results in a concise manner. Now it seems more like a re-iteration of the results. Overall, I appreciate that the paper is thorough and discusses many of the caveats. However, those are somewhat buried in the long subsections, and I fear that the quick reader may walk away with a stronger impression of "four robust independent drivers" than the text, read carefully, actually supports.

(5) Datapoints are not weighted according to their standard error (common practice in meta-analysis is inverse-variance weighting). The concern is that underpowered studies with high variance (e.g., due to a low number of recorded neurons) have the same impact as well-powered studies, and this may change some of the estimates. For example, Supplementary Figure 4 shows that mean CP values decrease with statistical power of the study, consistent with the concern. If SEMs are not available, could the authors show the robustness of the results by weighting by sample size as a partial check?

(6) Interpretation of feed-forward vs. feedback origin of CP [Disclosure: I am an author of Wimmer et al. 2015.]. This paper presents a mechanistic network model of area MT and a decision area that decomposes CP into two components with distinct time courses, and, directly relevant to Section 2.6, shows how a combination of an early feedforward and a late feedback component can produce a roughly time-invariant (flat) CP. This is a specific, quantitative instance of the "sustained plateau" pattern the authors themselves note is inconsistent across studies (lines 480-486) but don't develop further. Engaging with this model in Section 2.1/2.6 would let the authors contrast their duration-effect interpretation against an explicit dynamical model rather than the generic feedforward/feedback dichotomy in Figure 7a.

(7) Reaction-time experiments. I am worried that differences in CP in RT vs. fixed duration tasks (Supplementary Figure 12) could have an influence on the main regression analysis (because RT experiments are mostly from detection tasks, and because RT experiments presumably include less of a post-decision period). Could this factor be included in the main analysis?

Reviewer #2 (Public review):

Summary:

This manuscript by Pletenev et al. provides a meta-analysis of 59 published studies on decision-related activity in the macaque visual cortex. The work does not contain original research material, but by conducting extensive analyses and compilations of previously published data, it provides several new insights not already conveyed in recent reviews on this topic.

Strengths:

The work is scholarly and helps organize and integrate a broad set of findings. A difficulty in such an undertaking is making sure that the original studies are accurately characterized and tabulated. The lead authors should be commended for the rigorous approach they've taken; the inclusion of many of the authors of the original studies as co-authors provides additional reassurance that trends visible across studies reflect an accurate quantification of what each study has shown.

Weaknesses:

My sole scientific concern is on the 'duration' section (lines 487 to 536). Specifically, the authors relate the dependence of choice probability (CP) on time to two competing (but not mutually exclusive) views of how CPs arise-the 'feedforward' vs 'feedback' views. The basis for the predictions in this section was not clear to me. In particular, it was not obvious that the predictions fully considered all the relevant factors. For instance, did the feedforward predictions consider how the response covariance depends on duration and how this would affect CP values (Equation 1)? (See for example Figure 4 of Kang and Maunsell, 2012, JNP.)

I would suggest either explaining the basis of the predictions much more carefully or moving the predictions to a supplementary section where they can be presented in more detail (i.e., more convincingly). Alternatively, since the section is inconclusive in the end (there is not strong evidence in favor of FF or FB), the authors could mention the theoretical predictions in passing only, i.e., much more briefly, just to make the reader aware that the different theories can provide predictions about the duration dependence.

Reviewer #3 (Public review):

Summary:

This study presents a comprehensive meta-analysis of choice probability (CP), a classic metric of the relationship between single-neuron responses and an animal's perceptual judgment. The authors compiled data from 59 macaque electrophysiology studies and identified several factors that consistently influence CP magnitude, including neuronal sensitivity, brain area, stimulus duration, and task type. These results provide evidence that helps settle several long-standing debates about how CP should be interpreted.

Strengths:

This work is a rare example of meta-analysis in macaque electrophysiology, focusing on an important and long-debated metric, choice probability (CP), in the study of sensory and decision-making mechanisms. CP has been measured across many studies, but its interpretation remains contentious because it depends on numerous task and recording factors in addition to sensory and decision-making mechanisms. Individual macaque studies also typically include few subjects, and CP effects are generally small, which prevents any single study from drawing strong conclusions about general patterns. The authors identified an ideal use case for meta-analysis and combined fragmented findings from individual primate studies into a coherent picture of which factors matter most for CP and which do not. This work can also serve as a guide for future meta-analyses of macaque electrophysiology data.

Weaknesses:

While the survey and discussion of CP's interpretation are comprehensive, the paper would benefit from a clearer conclusion on why measuring CP remains important and what future directions could make CP more useful for revealing sensory and decision-making mechanisms.

Reviewer #1 (Public review):

Summary:

The manuscript presents several reinforcement learning (RL) approaches to control activity in small, ring-shaped neural networks. The paradigm is to find temporally and spatially patterned stimuli applied to axons that maximize the length of activity propagation along the network. Several RL versions were compared. Some RL-designed stimulation patterns worked better than others.

Strengths:

The work is technically and statistically very sound, and important controls were done that are missing in some earlier work along this line of research. The statistics and technical solutions seem solid, though I am no specialist in RL. The figures are well done and informative, with matching quality of the captions. The work presented is of value mainly for someone who wants to build a good closed-loop stimulation system to experiment with neuronal networks in-vitro.

Weaknesses:

The manuscript appears undecided about whether it wants to be about RL control, about a technical implementation of long-term stimulation in vitro, or about the properties of neuronal networks and interaction with them. The introduction is well written, comprehensive and insightful, focusing on biological aspects. Methods are then extensively about RL algorithms, without explaining why several were used or why these in particular, but with specialist language hard to understand for neuroscientists. The results then quantitatively compare the performance of the RL but do not really explain what this teaches us about neuroscience or what we learn about the networks beyond that they can be stimulated for longer propagation patterns. Extensive supplementary material almost advertises the hardware built by the team. Some figures suggest, though I'm not 100% certain about this, that different algorithms find different optimal stimulation patterns in the same network - which I find puzzling. What then does this tell us about the stimulation patterns and the variability of the responses? There is some speculative mechanistic reasoning, but no data to support this further. The conclusions hardly address the initial motivation of the manuscript. To me, it is not clear what can be learned that was not, in one way or another, presented previously, with as well as without RL.

Reviewer #2 (Public review):

The authors developed a reinforcement learning framework to identify stimulation patterns that produce a predefined activity pattern in neuronal networks cultured on a microelectrode array. The main contribution of this study lies in the development and validation of the experimental framework, rather than in providing new insight into neuronal response properties or the biological mechanisms underlying network activity.

A major strength of the work is that the framework is implemented using open-source software. Electrical stimulation through microelectrode arrays is already widely used, and the proposed approach is therefore likely to be of interest to researchers seeking to automate the exploration and optimization of stimulation parameters. The authors also provide the software, hardware configuration, and experimental data, which substantially improves transparency and should facilitate reproduction and further development of the system by other laboratories.

The experiments provide a useful proof of concept showing that the framework can search for stimulation patterns associated with the predefined task in the tested neuronal network. The evidence is solid for demonstrating the feasibility of the approach within this specific experimental configuration. In particular, the closed-loop integration of stimulation, recording, evaluation of the neuronal response, and subsequent selection of stimulation patterns is clearly implemented and experimentally tested.

An important limitation is that the framework was evaluated using a relatively small and highly structured network. Four stimulation electrodes were positioned around a single network, and stimulation was delivered to microchannels in which axons were concentrated. This configuration is well suited to the initial demonstration, but it remains uncertain whether the same approach will perform similarly in conventional monolayer dissociated cultures, larger networks, or systems with different numbers and spatial arrangements of electrodes. Additional validation across a broader range of network structures and experimental configurations would therefore be needed to establish the general applicability of the framework.

Overall, this study presents a valuable and reproducible methodological framework for the closed-loop optimization of electrical stimulation in cultured neuronal networks. Its likely impact lies primarily in providing an accessible experimental and computational platform that can be adapted for studies requiring systematic exploration of stimulation patterns, although the extent to which the current findings generalize beyond the tested network configuration remains to be determined.

Reviewer #3 (Public review):

Summary:

This study integrates living neuronal circuits with a real-time reinforcement-learning framework to enable closed-loop optimization of directional firing sequences at single-spike resolution. By using the spatiotemporal structure of neuronal firing as feedback, the system adaptively searches for stimulation patterns that increase a predefined propagation reward. The work represents an exciting step toward systematic and adaptive control of living neuronal networks.

Strengths:

The study presents a cutting-edge closed-loop platform that combines neuronal cultures, high-temporal-resolution electrophysiology, and reinforcement learning with millisecond-scale latency. The ability to optimize directional activity propagation at single-spike resolution is particularly novel. More broadly, the work provides a compelling framework for interacting with biological neural networks in an adaptive rather than purely predefined manner.

Weaknesses:

Several aspects of the analysis and interpretation require further clarification. In particular, the definition and computation of directional propagation and reward are not always clear. The generalizability of the optimized stimulation patterns across cultures also remains unclear.

Reviewer #1 (Public review):

Summary:

The authors investigated somatosensory processing along the afferent somatosensory pathway (cuneate nucleus, thalamus, S1) in a group of spinal cord injury patients and a group of controls. They propose that reduced motor function in SCI patients would reduce bottom-up activity; thus, recorded activity in SCI patients would reflect top-down modulation of overt or attempted movements.

Strengths:

(1) Strong methods.

(2) Experimental and control groups.

(3) Strong writing.

(4) Results well presented.

(5) Appropriate statistics.

Weaknesses:

Some results (or lack of) cast doubt about the ability of the used technique (3T fMRI) to detect the desired effects (bottom-up vs top-down activity).

Reviewer #2 (Public review):

Summary:

This study addresses a question that has been essentially inaccessible in humans: whether the early somatosensory relay nuclei are engaged by anything other than peripheral drive. Using functional and quantitative MRI in individuals with chronic cervical spinal cord injury, the authors show that the cuneate nucleus and VPL are robustly engaged during overt or attempted hand movement, and that this engagement persists in a participant with complete hand paralysis and no detectable muscle activity. They further report structural degeneration of the cuneate nucleus that is unrelated to the preserved task-evoked activity, and interpret the residual activity as reflecting top-down corticocuneate signalling.

Strengths:

The work is well conceived and clearly written, and the demonstration that the cuneate nucleus and VPL are robustly engaged during (attempted) hand movement after chronic cervical spinal cord injury is, to my knowledge, novel at this level of anatomical resolution. The finding is convincing. The dissociation between preserved task-evoked activity and marked structural degeneration of the cuneate nucleus is a highly interesting result with clear implications for rehabilitation. The manuscript is straightforward to follow, and the imaging protocol is carefully executed.

Weaknesses:

My main reservation concerns the inferential step from "not peripheral" to "corticocuneate". The data establish the former convincingly; the latter may not.

(1) Attribution of the observed activity to corticocuneate projections.

The central claim rests on an argument by elimination: because bottom-up drive is excluded in PT01, the residual activity must be top-down and, by extension, corticocuneate. Two distinct gaps should be addressed. First, "top-down" is not equivalent to "direct corticocuneate". Descending influence could reach the cuneate nucleus through multiple indirect pathways. Second, the activity observed at the three levels (cuneate, VPL, S1) need not be serially propagated, since layer 6 corticothalamic projections, for example, could drive VPL independently of any cuneate contribution. I would ask the authors to either provide evidence bearing on the routing, or to consistently use a route-neutral term (e.g. "descending" or "top-down") and reserve "corticocuneate" for the discussion of candidate mechanisms.

(2) Afferent input arising above the lesion level.

The EMG control in PT01 was restricted to hand and forearm muscles. Musculature innervated above C4 (cervical paraspinals, trapezius, and to a variable extent the shoulder girdle) remained available to this participant, and attempted hand movement is frequently accompanied by increased proximal co-contraction, postural stabilization, and altered respiratory effort. Afferent to the upper cervical cord is known to project to the ipsilateral cuneate nucleus, and its activity would produce lateralized, ipsilaterally dominant cuneate input - that is, precisely the pattern reported. This alternative is not excluded by the present control and should be addressed directly, ideally with proximal EMG in PT01 (and, if possible, in the other participants), or at minimum with an explicit discussion. Relatedly, the authors recorded respiratory and cardiac signals: please report whether respiratory volume or heart rate differed between movement and rest blocks, and between groups, since the dorsal medulla lies adjacent to cardiorespiratory nuclei.

(3) Functional significance of the preserved top-down signal.

The discussion establishes that top-down input persists but says relatively little about why it should. If the principal role of descending input to the cuneate nucleus is the gating of incoming afferent traffic, then in the absence of afferents there is nothing left to gate, and one might have expected the signal to be lost. Its persistence is the most interesting aspect of the finding and deserves fuller discussion. Candidate accounts the authors may wish to consider include: an efference copy or predictive signal delivered to a comparator that no longer receives its input, in the framework the authors already invoke (references 27, 28); engagement of the non-lemniscal outputs of the dorsal column nuclei (e.g., cuneocerebellar, cuneo-olivary projections); attempted movement engages motor imagery and attention, in which case the relevant question becomes what distinguishes these from movement-related gating. A related interpretational point: in behaving primates, movement-related modulation of cuneate transmission is bidirectional and includes prominent suppression (refs. 7/12). Note also that BOLD increases are compatible with increased inhibition, so they do not indicate facilitated throughput.

Reviewer #1 (Public review):

Summary:

This manuscript uses a valuable dataset of simultaneous LC single-unit recordings and pupillometry in awake monkeys to examine one aspect of the relationship between LC activity and pupil diameter: whether baseline LC activity predicts evoked pupil and whether baseline pupil predicts evoked LC. These relationships are largely absent in the present dataset, and the authors conclude that this type of prediction is not reliable.

Strengths:

This is a valuable dataset of simultaneous LC single-unit recordings and pupillometry in awake monkeys, the within-modal and within-epoch analyses are sound. The new results further caution the use of pupil diameter to infer LC activity.

Weaknesses:

There is an obvious rationale for asking whether baseline LC relates to baseline pupil or evoked LC relates to evoked pupil. It is also obvious to ask about the relationship between baseline and evoked LC activity, and between baseline and pupil responses. However, the rationale for the cross-modal and cross-epoch analysis is not clear. Why should one expect baseline LC to predict evoked pupil, or baseline pupil to predict evoked LC? What is the biological importance of such predictions? These analyses simultaneously change the modality and the temporal domain.

Furthermore, some recent studies, which are not referenced in this manuscript, showed variability in the relationship between pupil and LC within the same epoch, and that tonic and phasic pupil responses could be regulated by different inputs to LC. Given that the pupil-LC coupling is readily imperfect and potentially controlled by different cellular and circuit mechanisms, it is not surprising that their cross-epoch relationship is even more variable.

One main analysis that correlated baseline pupil with evoked LC showed statistically different results between the two monkeys, raising the question of to what extent a general claim for the cross-modal cross-epoch analysis can be made.

Reviewer #2 (Public review):

The manuscript by Thompson and Gold reported that, although baseline and evoked LC activity were positively correlated with baseline pupil size and evoked pupil dilation, respectively, there were no reliable cross-epoch relationships between LC activity and pupil size - that is, between baseline LC activity and evoked pupil dilation or between baseline pupil size and evoked LC activity. A major strength of the study is its large dataset, comprising recordings from more than 100 LC single units collected across 83 recording sessions in two monkeys. The authors further strengthened their conclusions by performing several control analyses to rule out potential confounds, including whether the findings were influenced by (1) the use of residuals in the main analyses, (2) the possibility that the auditory stimulus failed to evoke the full dynamic range of LC activity and pupil responses, and (3) nonlinear cross-epoch relationships between LC activity and pupil size.

These findings are important because they remind researchers to exercise caution when assuming that non-illuminance-mediated fluctuations in pupil size can reliably serve as a proxy for LC activity under all experimental conditions

With that being said, I have two suggestions that I believe would further strengthen the manuscript.

(1) Previous studies have suggested that the LC is organized into functionally distinct subpopulations (e.g., PMID 28920933 and PMID 40770025). It is therefore possible that a subset of the LC neurons recorded in this study does not participate in controlling pupil size. I encourage the authors to discuss this possibility in the Discussion. In addition, this possibility could be tested by repeating the cross-epoch relationship analyses using only LC units that exhibit a strong correlation with pupil size (the darker points in the last column of the top two rows of Figure 1?).

(2) The manuscript would benefit from a clearer visualization of the analyses addressing the possibility that the auditory stimulus failed to evoke the full dynamic range of LC activity and pupil responses. A supplemental figure illustrating the distributions or ranges of LC activity and pupil responses, together with the corresponding control analyses, would help readers better understand this important point.

Reviewer #1 (Public review):

Summary:

The authors asked whether neurofeedback during competing continuous speech can help to modulate the attention-related N1-component in the temporal response function (TRF), which is an event-related-response-like estimate of the phase-locked EEG activity following the envelope. The research question is relevant because it asks to what degree the strength of attention can be controlled beyond the binary decision to attend or ignore something, and whether this control is beneficial for the behavioral outcome.

Strengths:

(1) Sample size of 56 participants.

(2) Control group with sham feedback.

(3) Novelty: Under-explored field of neurofeedback in selective speech tracking.

(4) Pragmatic and reasonable methodological decisions.

(5) Transparent results not hiding the fact that effect sizes are small.

Weaknesses:

Besides some need for clarification, I could only find one methodological weakness, which the authors discuss anyway:

(1) Overall, speech tracking-based neurofeedback may lead to more robust results, because the N1-extraction does not have to be handcrafted and all components would be taken into account. As the authors state, the P2-component has been related to effort, and this may provide more "room to play" for voluntary modulation.

The following "weaknesses" are related to the impact of the results:

(2) Non-translating effects to post-training trials, neither neurally nor behaviorally.

(3) Neurofeedback-related Modulation of N1

Reviewer #2 (Public review):

Summary

This manuscript investigates whether neurofeedback based on the N1 component of the temporal response function can be used to modulate neural responses during selective attention to continuous speech. Participants listened to two competing audiobooks and were instructed to attend to one of them. In the neurofeedback group, trial-by-trial N1 responses were converted into visual feedback, whereas the sham-feedback group received replayed feedback from other participants. The authors found a significant interaction between group and block for the N1 response to target speech over a small fronto-central cluster, with larger N1 responses during feedback blocks in the genuine neurofeedback group but not in the sham group. No neurofeedback effect was found for the distractor response. The authors also reported exploratory post-training effects and an association between changes in N1 and speech-comprehension performance at right fronto-central electrodes.

Overall, the study is conceptually interesting and novel. The online, trial-by-trial estimation of neural responses from continuous speech is an attractive development for auditory neurofeedback, and the inclusion of a randomised sham-feedback group is an important strength. However, the manuscript provides stronger evidence for modulation of a neural response during feedback than for learning or training of selective attention. Some aspects of the analysis and interpretation also require further consideration/clarification, particularly the use of group-specific N1 time windows, the spatial confound between target and distractor streams, the absence of artefact correction in the signal used for feedback, and the relatively weak behavioural evidence.

Strengths

The main strength of this study is its novel use of neurofeedback during continuous competing speech. Rather than providing feedback based on a general measure of brain activity, the authors targeted a specific neural response associated with selective auditory attention. The online implementation is technically impressive, allowing neural responses to be estimated from 22-second speech segments and converted rapidly into feedback. The inclusion of a sham-feedback group is another important strength, as it helps distinguish effects of genuine neurofeedback from nonspecific effects such as task engagement or motivation. The relatively large sample for a neurofeedback study and the use of natural continuous speech also increase the robustness and ecological relevance of the work.

Weaknesses

(1) The effect was present during the feedback blocks but did not increase across training blocks, and the post-training effect was not found at the same electrodes used for feedback. The evidence therefore supports online modulation more strongly than learning or lasting self-regulation, and claims about successful training or persistent learning should be interpreted cautiously.

(2) The offline analysis used different N1 time windows for the neurofeedback and sham groups. This introduces a potential bias in the group comparison because the dependent measure was defined differently between groups. Confirmation of the main result using a common, independently defined N1 window would strengthen the evidence.

(3) The target speech was always presented from the front and the distractor from behind. Differences between target and distractor responses therefore cannot be attributed entirely to attention because spatial location is also different. This limits the interpretation of target-versus-distractor differences and may also contribute to the weaker reliability of the distractor response.

(4) Feedback blocks always contained two speakers, whereas half of the baseline trials contained only one speaker. Since the presence of a distractor altered the neural response, it is important that the neurofeedback comparison is based on acoustically matched multi-speaker baseline trials. If this were not the case, differences between baseline and feedback could partly reflect differences in the acoustic condition rather than neurofeedback.

(5) No artefact correction was applied to the signal used for online feedback. Because feedback was derived from fronto-central electrodes, eye or muscle activity could potentially contribute to the measured signal. An offline demonstration that the main neural effect remains after appropriate artefact control would strengthen the interpretation that the effect reflects neural modulation rather than systematic changes in non-neural activity.

(6) The behavioural evidence is weaker than the neural evidence. There was no significant overall improvement in speech comprehension in the neurofeedback group. The reported behavioural effect is instead based mainly on associations between changes in the neural response and changes in comprehension, and some of these effects were weak before the whole-scalp analysis. Therefore, these findings are better interpreted as exploratory associations rather than evidence that neural modulation directly caused improved comprehension.

(7) Adding the distractor did not significantly reduce comprehension performance. The absence of a significant distractor effect on comprehension suggests that the listening condition may not have produced a strong behavioural cocktail-party difficulty in this sample. The large spatial separation between speakers and the nature of the behavioural task may have reduced sensitivity to distraction, which could limit the strength of the conclusions regarding improvement of speech understanding in challenging listening conditions.

(8) The study is described as double-blind, but the manuscript provides limited detail on how blinding was maintained, particularly when the experimenter manually checked the N1 estimate. In addition, participants' belief in or perceived control over the feedback was not formally assessed. These factors make it difficult to determine how effectively expectancy or motivation-related effects were controlled.

Overall assessment:

This study provides a valuable methodological and conceptual advance by showing that online, trial-by-trial neurofeedback can modulate the neural response to attended speech during competing speech. The evidence is solid for an immediate neural effect during feedback, and it was supported by comparison with a sham-feedback group. However, it remains incomplete for broader claims about learned self-regulation, persistent effects, distractor suppression, and improved speech understanding.

Reviewer #1 (Public review):

This work provides a valuable toolkit for endogenous isolation of projection neuron subtypes. With further validation, it could present a solid method for low-input ribosome affinity purification using a ribosomal RNA (rRNA) antibody. The experimental evidence for the distinct ribosomal complexes is limited to this method and indirect support from complementary analyses of pre-existing data. However, with additional experimental data to support the specificity of ribosomal complex pulldown and confirmation of the putative ribosomal complex proteins of interest, the study would provide compelling evidence for translation regulation of neuronal development through compositional ribosome heterogeneity. This work would be of interest to neuroscientists, developmental biologists, and those studying translational networks underlying gene regulation.

Strengths

(1) This in vivo labeling of specific projection neurons and ribosomal rRNA affinity purification method accommodates a low input of <100K somata per replicate, which is useful for the study of neuronal subtypes with limited input. In principle, this set of techniques could work across different cell types with limited input depending on the molecule used for cell type labeling.

(2) The authors are also able to isolate endogenous neurons with minimal perturbation up to the point of collection, preserving the native state for the neuron in vivo as long as possible prior to processing.

(3) This study identified over a dozen potential non-ribosomal proteins associated with SCPN ribosomal complexes, as well as a ribosomal protein enriched in CPN.

Limitations

(1) In this study, the authors address the advantages of their ribosomal complex isolation method in SCPN and CPN against RPL22-HA affinity purification. While this does show more pull down of the ribosomal RNA by the Y10B rRNA antibody, the authors claim this method identifies cell-type specific ribosomal complex proteins without demonstrating a positive control for the method's specificity. There are very limited experiments to truly delineate how "specific" this method is working and whether there could be contamination from other complexes bound by the antibody. I see this as the major limitation that should be addressed. To boost their claims of capturing cell-type specific ribosomal complexes, the authors could consider applying their rRNA affinity purification pipeline to compare cell-types with well-characterized ribosome-associated proteins, like mouse embryonic stem cells and HELA cells. The reviewer can completely appreciate the elegance in the neural characterization here, but it seems there needs to be a solid foothold on the specificity of the method, perhaps facilitated by cell types that can be more readily scaled up and tested.

(2) The authors followed up on their differentially enriched ribosomal complex proteins by analyzing ribosome association of these proteins in external datasets. While this analysis supports the ribosome-association of these proteins, there is limited experimental validation of physical association with the ribosome, much less any functional characterization. The reciprocal pulldown of PRKCE is promising; however, I would recommend orthogonal validation of several putative ribosomal complex proteins to increase confidence. Specifically, the authors could use sucrose gradient fractionation of SCPN and CPN, followed by western blot to identify the putative interaction with the 80S monosome or polysomes. This would also provide evidence towards the pulldown capturing association with mature ribosome species, which is currently unclear. This experiment would provide substantial evidence for the direct association of these non-ribosomal proteins with subtype-specific ribosomal complexes.

(3) The authors state interest in learning more about the differences underlying translational regulation of projection neuron development. This method only captures neuronal somata, which will only capture ribosomes in the main cell body. There are also ribosomes regulating local translation in the axons, which may also play a critical role in axonal circuit establishment and activity. These ribosomal complex interactions may also be rather transient and difficult to capture at only one developmental stage. Therefore, this method is currently limited to a single developmental snapshot of ribosomal complexes at P3 within the main cell body. It would be exciting to see extended utility of this method to sample neurites and additional developmental stages to gain further resolution on the developmental translation regulation of these projection neurons.

Likely impact of the work on the field, and the utility of the methods and data to the community

The authors introduce a unique pipeline of techniques to identify cell-type specific ribosomal complex compositions. With more validation, there is certainly potential for those studying neuronal translation to leverage this method in limited primary cells as an alternative to existing methods that do not rely on ribosomal protein tagging, such as ARC-MS (Bartsch et al., 2023), RAPIDASH (Susanto and Hung et al., 2024), and RAPPL (Nature Communications, 2025).

Comments on revised version.

We thank the authors for their thorough response to our comments. The revised manuscript satisfactorily addresses most reviewer comments through clarification and expanded discussion, although we believe some important limitations remain. In particular, we continue to view experimental validation of the identified ribosome-associated proteins as an important component of introducing this methodology to the field, rather than work that falls beyond the scope of the study. The authors have acknowledged that these experiments are future directions, and it is clear they plan to pursue additional validation outside of the current manuscript. Given their emphasis that the primary contribution is the development of a methodological framework, we believe the work may be appropriately framed as a Tool and Resource article. Such positioning would better align reader expectations with the manuscript's strengths as a technical advance while recognizing that further validation and functional characterization will be needed in future studies. Despite these limitations, I believe the manuscript makes a valuable methodological contribution.