Author Response:

Reviewer #1:

Weaknesses:

Although the BOLD data is highly spatially specific, there is just one electrophysiological timeseries per subject. This is no doubt a bi-product of the extensive noise cancellation that is necessary to record within the scanner. The caveat therefore is that the covarying BOLD and electrophysiological changes may derive from different regions.

We recognize this is a limitation which is also not easily solved by approaches for source analysis, given the nature of the data (only 64 channels) and the usually larger imprecisions related to EEG source reconstruction. We circumvented this by choosing a task that is known from previous studies in MEG to induce changes in multiple frequency bands originating from regions the early visual cortex (Hoogenboom et al., 2006; Hoogenboom et al., 2010; Koch et al., 2009; Muthukumaraswamy and Singh, 2013). Furthermore, the EEG responses are highly similar to invasive recordings in animals from visual regions in the context tasks investigating selective attention (Fries et al., 2008). We mention this limitation now in the introduction (lines 102-111).

The analysis methods are slightly non-standard, perhaps for good reason. The main thing that stands out is the use of correlation coefficients, rather than regression coefficients, at the first level of analysis. This could potentially conflate changes in signal with changes in noise or unexplained variance.

We chose here for the correlation, since in our opinion this leads to a more interpretable measure of linear association than a regression slope. A regression slope-based analysis will yield different outcomes for the regression of y on x, than for x on y, doubling the number of analyses needed. The different results for a regression of y on x and x on y are often interpreted as implying directionality, which is not warranted and not what we would like to imply with our analysis. The asymmetry is caused by the implicit assumption that x does not contain noise in a regression of y on x. This is valid when x represents a paradigm condition vector, but not when it is a data vector. We therefore opted to use the difference in (Fisher-z-transformed) correlation as our estimate for linear association/connectivity between laminar fMRI signals.

In both a correlation as well as in a regression approach differences can be attributed to differences in true underlying coupling andin a difference in noise. This is however not different for correlation-based measures in coupling in fMRI than in for instance coupling measures like coherence and phase-locking-factor in electrophysiology. Coherence can be regarded as the frequency domain version of (squared) correlation. The fact that our measure might indeed be related to differences in noise would therefore not be resolved by opting for a regression based approach, and is not different from often used measures of coupling in electrophysiology.

Reviewer #2 (Public Review):

Introduction. The introduction provides an overoptimistic view on the current possibilities with respect to the investigation of layer-specific activation or connectivity in the living human brain. Cortical layers cannot yet be segmented, the fMRI measures only provide an indirect signal that is heavily influenced by partial voluming between cortical depths, and EEG and MEG approaches often only measure two compartments due to low spatial resolution. The introduction, however, gives the impression that layer-specific neuronal connectivity can precisely be measured in the living human brain, which is not the case. The authors should take considerably more care with respect to how they introduce the methodology with clear references to the limitations. Also, statements such as "laminar fMRI allows us to study connectivity.." should be removed. In the same vein, I would suggest to replace laminar fMRI and laminar connectivity with cortical depthdependent fMRI and connectivity to account for the above mentioned aspects.

In laminar fMRI research it is commonly accepted that what we measure are not true layers, but depth dependent fMRI between the boundaries of white/gray matter and gray matter/CSF. For the general audience we will make this distinction clearer and discuss the limitations of the technique (lines 74-80).

Concept. Whereas the authors provide a model in the introduction that specifies how different frequency bands could relate to cortical depth-dependent connectivity, they do not develop a working hypotheses based on their experimental design. One conceptual step is therefore missing in the introduction, which has to combine present knowledge on the relationship between different frequency bands and present knowledge on how attention influences frequency-specific activation in the visual system to then make statements about which analyses can be performed to test which aspect of the model.

The primary focus of our study was to investigate how oscillations across several frequency bands in the EEG relate to laminar specific activity. Recent publications on laminar fMRI have demonstrated the possibility of performing laminar level fMRI connectivity analyses, which led us to revisit our previously recorded data in order to explore whether not only laminar specific BOLD amplitude but also laminar fMRI connectivity relates to frequency specific EEG power. Since laminar fMRI, and especially connectivity derived from those measures is very novel, we started this analysis without a preconceived model or notion on how this relation would be. The results from this project should therefore be interpreted as an exploration of how these laminar fMRI derived connectivity measures relate to neural oscillations rather than directly addressing a specific cognitive process like selective attention, or prediction and/or a model of how neural oscillations play a role in these processes. Our experimental paradigm was also not designed to address such processes. We chose a paradigm that is known from previous studies using MEG and EEG to induce changes in multiple frequency bands in the early visual cortex (Hoogenboom et al., 2006; Hoogenboom et al., 2010; Koch et al., 2009; Muthukumaraswamy and Singh, 2013). Furthermore, the EEG responses are highly similar to invasive recordings in animals from visual regions in the context tasks investigating selective attention (Fries et al., 2008). The crude attention/task modulation added to the paradigm (attention On versus Off) was in the first place introduced to induce meaningful variation over subjects in a task effect across the frequency bands modulated by visual stimulation. It was not intended to investigate specific individual processes such as prediction, attention or arousal. The observed effects can therefore also not be ascribed to such specific processes, since they are co-modulated by the task. We will make this more clear in the introduction now. We make this point now explicitly in the introduction.

• Concept & Methods: With respect to both the concept and the analyses, what is missing is taking into considerations the brain areas that were investigated. Wheres in the abstract the authors only mention "within brain region connectivity" and "between brain region connectivity" also in the Methods section there is no clear relation to the anatomical areas that were investigated, being V1, V2 and V3. The authors rather classify the areas as "high level" and "low level" where V2 is sometimes classified as high-level and sometimes as low-level. The data are therefore not investigated with reference to the anatomy of the visual system. In my view, it would be beneficial if all analyses could be performed with respect specifically to V1-V2 connecitivity and V2-V3 connectivity as well as V1-V3 connectivity so that the specific anatomical interrelations are taken into account. Also, the authors should develop a conceptual framework of how layer-specific attention-driven connectivity changes should influence the visual cortex, and why.

In the results for between region connectivity we averaged over several connection pairs (V1- V2,V1-V3,V2-V3) and for within region connectivity across regions (V1-V1,V2-V2,V3-V3) before effects in connectivity were correlated with EEG power. There are several reasons why we opted for this approach: First, we wanted to maximally increase the statistical power to observe patterns of association between laminar connectivity and EEG power. Since the analyses as carried out here have not previously been performed, we had no estimate of effect size. Secondly, by averaging over region combinations we drastically limit the multiple comparisons problem, since the number of comparisons scales with the square of the number of regions connectivity is computed between. Third, by averaging over regions, we target more general effects of connectivity between and within regions that are more likely to correspond to patterns observed within other contexts and other modalities. The effect for individual region combinations would likely be more variable.

For completeness in the first submission we did include the results for every single region combination in the supplementary material (see Supporting Figures S2-S5). We have now included in the main document the results for region combinations V1-V2,V1-V3 and V2-V3 for between region connectivity, and V1-V1, V2-V2 &V3-V3 for within region connectivity, presented alongside the results for the grand average.

The results for the individual region-pairs suggest that inter- and intra-region connectivity are generally consistent with the average over individual region combinations, but also have unique features.

Similarities include: A strong negative correlation between beta power and deep-to-deep layer coupling was observed for average inter-regional connectivity. In line with this, for all three individual region pairs (V1-V2,V1-V3,V2-V3) a negative correlation is observed for deep-to-deep layer coupling. Similar patterns can be observed from alpha and beta for intra regionalconnectivity (averaged over all regions) and connectivity within V1,V2 and V3 in isolation.

Individual features include: The relation between beta and inter-regional coupling shows variation over the individual region-pairs. In particular for V2-V3 connectivity, but also for V1-V2 the relation seems to differ from the pattern observed on average. For V2-V3, deep layer V3 seems to be coupled to both deep and superficial layers in V2, a pattern that might reflect anatomical feedback projections that go from deep layer V3 to both deep and superficial layers in V2.The stronger correlation between deep V1 and more middle deep V2 is however harder to directly place, since direct anatomical connections here are largely absent here. It might therefore reflect an indirect effect.

Despite some degree of individual variation we think the overall picture is largely consistent. The strongest features present in the averaged results can clearly be observed in each of the individual region-combinations as opposed to the latter being a collection of vastly different random patterns that happen to add up to the average result (see for example the intraregional alpha results).

With respect to our classification of regions into higher and lower level cortical regions, we based on standard anatomical hierarchies like that of van Felleman & van Essen (Felleman and Van Essen, 1991). Here, V1, V2 and V3 are ordered from low to higher in the visual cortical hierarchy.

Methods. Given the missing conceptual overview over how attention-induced changes in EEG frequency bands should influence laminar connectivity in the visual system, also the methods lack a clear analyses strategy. The authors computed one correlation between power level of different frequency bands and connectivity between different brain areas without providing an explanation of which question this analysis addresses. The offered results therefore seem random to me, without a clear relationship to an investigated hypothesis.

The primary focus of our study was to investigate how oscillations across several frequency bands in the EEG relate to laminar specific activity. Recent publications on laminar fMRI have demonstrated the possibility of performing laminar level fMRI connectivity analyses (Sharoh et al., 2019; Huber et al., 2017; Huber et al., 2020), which led us to revisit our previously recorded data in order to explore whether not only laminar specific BOLD amplitude but also laminar fMRI connectivity relates to frequency specific EEG power. Since laminar fMRI and especially connectivity measures derived from it are very novel, we started this analysis without a preconceived model or notion on how this relation would be. The results from this project should therefore be interpreted as an exploration of how these laminar fMRI derived connectivity measures relate to neural oscillations rather than directly addressing a specific cognitive process like selective attention, or prediction and/or a model of how neural oscillations play a role in these processes. Our experimental paradigm was also not designed to address such processes and test hypotheses derived from these. The primary focus of the work presented here is to provide a first insight in how neural oscillations measured by electrophysiological measures relate to cortical depth resolved fMRI coupling, which is usually correlation based. We believe these results will be relevant for research focused on how neural oscillations relate to inter-and intra regional interactions (e.g. (Bastos et al., 2012)(Fries, 2015)), since depth resolved fMRI allows us to study laminar interactions within and between brain regions non-invasively in humans. For this it is important to know if and how neural oscillations relate to laminar fMRI based connectivity measures, of which our research here provides a first insight. It also provides insight into which neural processes underlie observed changes in laminar fMRI based coupling, and is therefore relevant for research using such methods in general.

Methods. The authors mention that they only analyzed the strongest two connecting vertices within a layer, which was done to improve SNR. In my view, for a connectivity analyses, this is not valid, as it can bias the effect towards superficial connectivity where the SNR and thus correlation is always higher.

We did not analyze vertex pairs within a layer. We computed vertex pairs that connect the boundary between gray matter and CSF with the boundary of gray matter and white matter based on a high resolution anatomical MRI scan. Between these vertices we sampled 21 points of functional fMRI data using nearest neighbour interpolation. Since not all parts of V1, V2 and V3 will be involved in the task, we selected the most activated vertex pairs for further analysis. This serves as a localizer to select the parts within a region where task related activation is observed. For the main analysis the top 10% activated vertex pairs were chosen based on data collapsed across all depths and all attention conditions. This selection is therefore independent of depth, task condition, and the relation with any EEG feature. For this procedure we actually excluded the top five depth bins to avoid being too biased to superficial depths since it is known that signal to noise is substantially better near the surface of the cortex in part due to larger pial veins. To investigate whether the observed results are not due to this arbitrary threshold of 10%, we repeated the analyses for top 5% and 25% activated vertex pairs, the results of which are included in the supplementary information.

Methods. The authors report 21 correlations in cortical depth, where their resolution allows to only sample perhaps 2-3 data points. The correlation analyses are therefore oversampled, which influences the statistical results. I would suggest to first run a component analyses across cortical depth, and to then correlate independent components to one another to investigate independent data points.

The correlations are not oversampled, since the correlations used for the connectivity analyses are over trials, and not over space. These analyses are not influenced by the number of laminar BOLD data points we sample. Furthermore, spatial supersampling is a very common practice in FMRI research. For instance, the default in SPM is to upsample 3 mm isotropic standard voxel (very common for initial acquisition) size to 2 mm isotropic voxel size. In laminar fMRI laminar signals are often upsampled up to several factors above the the original resolution. This is for a number of reasons, well outlined on the laminar fMRI community website, a resource maintained by L. Huber in collaboration with many layer fMRI labs (see: https://layerfmri.com/2019/02/22/how-many-layers-should-i-reconstruct/) and ~20 layers is thought to be optimal.

For our statistical test we explicitly chose a non-parametric cluster based technique to correct for multiple comparisons that takes dependencies across space into account. Laminar fMRI data are not well suited to decompose into components using techniques like PCA and ICA, since they violate assumptions of orthogonality/independence of the underlying responses in both the spatial as well as the temporal dimensions. To illustrate: in a recent laminar connectivity methods review an hierarchical, iterative ICA approach resulted in data being split up in columnar maps rather than laminar ones (Huber et al., 2020).

Methods. The authors refer to their previously published paper with respect to the methods, and do not give any speficiations on the image sequence, image resolution, and image processing in this paper. In my view, all basic methodological steps that are critical to understand the paper should be described here.

We are willing to include all relevant parts of the methodology described in our previous paper. This would involve copying large parts of the methods section, and might have to be coordinated with the publisher of the previous publication for copyright reasons. We would be pleased if the editor could advise us on this issue.

Results. The figure captions are too short and do not explain the presented data in an appropriate way. In Figure 1, details on the calculated contrasts, number of participants investigated, sampling and analyses methods should be given that allows interpreting the data. Also, it would be beneficial to explain the attention paradigm in a bit more detail in the figure caption so that panel A can be interpreted. In Figure 3, more details should be given on what data are shown, particularly for panel C where the only information given is "attention effect on laminar connectivity" with no further axes labels.

We extended the figure captions in the revised article.

Results. I do not fully understand the results as shown in Figure 3. As those form the major part of the manuscript, this needs revision. As said before, I think that the figure and results section would benefit from region-specific data analyses and presentation, but also clear axes labels are needed to allow interpretation of the data. Also, when I interpret the data correctly, correlations are done for altogether 21 different cortical depth, which would not be valid because of artificially inflating the number of correlations, as pointed out above.

We have extended our analyses and now split original Figure 3 up into current Figures 3 and 4 where we separately depict the results for intra- and inter-regional connectivity. For both intraand inter regional connectivity we have now also shown the region-specific results that underlie these results. We updated the figures and captions to make clearer what is depicted. We addressed the point raised about the 21 data points above. It is not relevant for the analysis presented here.

Reviewer #3 (Public Review):

However, a weakness of the technique as currently presented is that patterns of connectivity are only related to oscillations across subjects. It would be more powerful to examine whether the current network state (estimated by trial-by-trial power estimates) relates to laminar connectivity within subject. This would indeed speak to the nature of neuronal communication, which takes place on a moment-to-moment time scale, and which is not reflected in the current analysis. This may explain why laminar patterns of fMRI connectivity were not found to correlate with gammaband oscillatory activity. In addition, the negative effects of attention on fMRI connectivity itself are somewhat puzzling. This may related to the limitations of the task design which do not perfectly separate attention vs. arousal/expectation, as the authors readily discuss.

The reviewer suggests that a relationship between fMRI connectivity and EEG power within subjects over trials would be more indicative of a direct link between connectivity and neural communication. We agree that establishing such a link would further strengthen the link between neural oscillations and laminar connectivity. This would not be trivial however, since connectivity in (laminar) fMRI is typically expressed as a measure of linear association (e.g. correlation or regression slope) over trials or time. Even at conventional spatial resolution, single trial/time point estimates of the network state are rarely used. These single data-point measures usually indicate to what extent a single data point contributes to the measure over all data points. We did not opt for such an analysis, since such analyses in normal (e.g. resting state) fMRI studies are uncommon, introducing more complexity to a study that already includes considerable novel analytic approaches. Furthermore, research relating fMRI activation and connectivity across subjects with other variables(e.g. clinical test scores, DTI measures, personality traits) is a well established procedure. Here we followed this more common approach.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Firstly, we would like to thank the reviewers for their helpful and insightful comments.

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

In the manuscript of Ramadan et al. , authors use the ex vivo organoid approach to compare gene expression in organoids derived from adult type stem cells when these organoids are grown using different matrices. The presence of Collagen type I induces the emergence of cells with a transcriptome similar to fetal progenitors. In contrast, laminin the main component of matrigel, induces an organoid-protruded phenotype with transcriptome of stem cell type. Then, they correlate these data with expression of collagens and laminins from data publicly available. They show by qRT -PCR that laminins are more expressed in mesenchymal versus epithelial fractions postnatally. They hypothesize on this basis that the remodeling at postnatal stage is likely only dependent on the mesenchymal compartment and it involves interaction of laminins with integrity a6.

It seems that some of the presented data have already been described and could not be considered as « novel ».

For some of the statements, like this one « the basement-membrane produced by the epithelium is not sufficient to increase stem cell numbers and induce a morphological crypt formation », the conclusion is not sustained by provided experiments. To draw definitive conclusion on this particular point, authors could reproduce the experiment presented in Fig. 4d but using Cre recombinases specific for mesenchymal and epithelial compartments rather than the ubiquitous Cre line. It would be interesting to investigate if organoids grown from lamc1-/- mice can generate protruded organoids or not.

In addition, how interpret the fact that fetal organoids up is associated with « laminin interactions » in fig. 1c?

The statement that the epithelium-produced basement membrane is not sufficient to increase stem cell numbers is based on our in vitro observations. Analysis of the RNAseq data shows that the expression of several laminins is increased on collagen (see heatmap of laminin interactions below, which will be added to the manuscript). This is also the reason why ‘laminin interactions’ is highly significant in the gene set enrichment analysis (Fig. 1C). Despite this upregulation, we never observed morphological changes (or expression changes) as when laminin is added to the collagen-hydrogel. In addition, we showed that the vast majority of ECM components is produced by the mesenchyme in vivo, in line with previous literature as cited in the manuscript. The mentioned Cre lines to address the question in vivo are unfortunately not available to our collaborators with the Lamc1 k.o. mice and it would therefore take too long to perform these experiments.

However, to address this point in vitro we will grow organoids from Lamc1 fl/fl mice and induce loss of laminin in the pure epithelial cell culture. Organoids will then be analysed for morphological changes, as well as proliferation and gene expression changes.

One major point to address regards statistics. In material and methods, the paragraph describing statistical analyses is missing. Moreover, in the figures presenting qPRC data ( figs 1g 3b 3D 3g 4c and f), no statistic analysis is provided; and the number of samples for some conditions is extremely limited (n=2). In general, the term « independent experiment « should be clarified : does it correspond to one organoid line for which the experiment was repeated or one single experiment using different organoid lines?

In fig 4c , all collagen conditions are set to 1.

The avoidance of statistical inference for most of the experiments was a deliberate choice. In line with several comments (e.g. 1. Vaux, D. L. (2012) Know when your numbers are significant. Nature. 492, 180–181), we chose to show all individual data points (with exception of Fig. 3D, n=5, to ease interpretation) without statistics. In addition, for most expression data, we have data from RNAseq, single-cell RNAseq and qPCRs repeated at different hydrogel concentrations to obtain reliable results. Further, the in vivo mesenchymal qPCR expression data was validated with RNA in situ hybridization showing the mainly mesenchymal expression.

The term independent experiment was used mainly for repeated experiments with the same organoid lines (exception RNAseq data, different organoids derived from individual mice). While conducting these experiments, we realised that the variability of these experiments comes from time in culture, density of cells and even Matrigel variation. The experiment in Fig. 4c (n=4, each time with the all controls) was performed with longer intervals in between, and showed variation in the absolute levels of expression. However, relative to each control we believe the effect is clear.

As we will perform additional experiments for the revision of this paper, we will then perform statistical tests in the key experiments (e.g. Itga6 experiment) to alleviate any concerns regarding significance.

Regarding the experiment presented in fig 4c, authors should include additional control conditions : anti-a6 integrity antibody in matrigel and use of an isotype antibody.

We will conduct additional experiments regarding the Itga6. In addition to including the mentioned controls for the neutralizing antibody, we will genetically inactivate Itga6 via an inducible Crispr/Cas9. This should enable us to delete Itga6 when the cells are grown on collagen, and hence reduce the possibility of compensation in matrigel derived organoids.

Another point regards RNAscope data presented in Fig 4b, it is surprising to observe such difference in terms of expression between E19 and P0. Does this mean that birth dramatically unregulates Itga6 expression in few hours? Authors should comment this point if verified.

We do believe that birth is a timepoint where a dramatic change in the ECM and their receptors can be observed. The epithelial RNAseq data would already indicate that at 18.5 there is an increase in expression compared to E16. This upregulation of the receptor is in line with the dramatic remodelling of the ECM at birth, as is shown by the expression of the basement membrane components in Fig. 3d.

Authors should avoid the word « signaling » for laminin-integrin interactions as they do not study this aspect at all in their experiments.

The word signaling was used for the protein:receptor interaction and to distinguish it from changes to the physical characteristics of the hydrogel. But we agree with the reviewer, that we did not study laminin signaling per se and therefore will change the wording accordingly.

Regarding Col1a1, authors cannot claim that it's expression only slightly changed (fig 3d) as it is clearly upregulated between E17 and P0.

The reviewer is right, and we apologise for the misleading sentence. The contrast was meant to the basement membrane components that are very lowly expressed at E17 and then suddenly show the burst of expression at birth, whereas collagen seems to be continuously expressed with a peak at P7. We will rephrase the sentence.

Reviewer #1 (Significance (Required)):

Overall, the methodology used for the asked questions is accurate.

One potential problem for publication comes from the fact that some of the findings are already reported and hat the present data do not provide further advances.

for example, collagen and fetal-like expression profile, Ly6a sorting and replating in culture-Yui et al, 2018, Jabaji et al, 2013.

We obviously do not agree with the reviewer on this point. We build upon the work of Jabaji et al. 2013, and Wang 2017 to characterise the specific effect of collagen on the intestinal epithelium compared to a pure Matrigel culture. The emergence of Ly6a cells was nicely shown by Yui et al., however it was unclear if collagen changes the fate of all intestinal cells or only a few. We strongly feel that our data extends these findings as it associates the changes we observe in vitro to the development of the crypt morphology and intestinal stem cells.

The phenotype of Lamc1-/- mice and the observed reduced stem cell marker expression are also reported by Fields et al, 2019.

Indeed, as we cited this paper. However we predicted based on our in vitro model, that deletion of laminin would result in this specific fetal-like gene expression and hence were happy to include these findings in our manuscript.

Infine, authors do not interpret their ex vivo data in the context of fetal progenitors which grow as spheres in matrigel (containing laminin)?

Our ex-vivo (in vitro) data would suggest that adult epithelial cells express some genes that are characteristic for fetal organoids, however we do not think these cells completely revert back to a fetal stage. Regarding the comment of spheres, it is noteworthy that fetal cells from E14-16 stay as spheres in Matrigel, whereas fetal cultures from E19 initially grow as spheres and then develop into organoids within 30 days in vitro (M. Navis et al., “Mouse fetal intestinal organoids: new model to study epithelial maturation from suckling to weaning,” EMBO Rep., vol. 20, no. 2, pp. 1–12, 2019.).

In figure 5, should we interpret that there is no laminin at all in the fetal mesenchyme?

We now see how the image is a bit misleading for that stage. The levels of laminin are lower in the fetal stage as can be seen by the IF image in Fig.3f and the image will be updated. We also apologise for the lack of labeling in Figure 3f, which should be E19, P7 and adult.

Also, authors do not cite a paper reporting on the role of the epithelium ( stem cells) in regulating its own extracellular matrix composition, this process modulating the stem cell number and fate (Fernandez-Vallone et al. 2020). As this is contradictory with the claim that only mesenchyme impacts on crypt morphogenesis, authors could discuss on this point.

In the paper by Fernandez-Vallone, deletion of Lgr5 in E16.5 embryos resulted in a decrease expression of several ECM genes. Further, the authors could show that the fetal epithelium does express for example Col1a1 at this point, which decreases with maturation. However even for the example of Col1a1 it is evident in their paper that the mesenchyme expresses Col1a1 at much higher levels. Our proposed experiments with Lamc1 k.o. In organoids will show if the produced laminins of the epithelium are essential.

This manuscript could be interesting for an audience in the stem cell and developmental fields ( my field of expertise).

**Referee Cross-commenting**

Considering the pertinent and sometimes overlapping comments of the two other reviewers, the estimated time is revised to 3-6 months.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

**Summary:**

In this manuscript, Ramadan and colleagues demonstrate that depending on the extracellular matrix (ECM) composition in which mouse intestinal organoids and/or 2D intestinal epithelial cells are grown in, cellular composition of the epithelium changes. Organoids plated on 2D collagen layers show a unique cell cluster characteristic of fetal-like genes, while organoids plated with increased amount of Matrigel in 2D or in 3D exhibit a shift towards higher stem cell abundance and the absence of the fetal-like gene cluster. Specifically, the ECM component Laminin supports acquisition of stem cells identities in small intestinal epithelial cells, correlating with a transient increase in expression levels of collagen and laminin genes in vivo spanning time points of crypt formation. The authors reported the functional contribution of laminin signaling (Lamc1 KO) via integrin alpha 6 (antibody-blocking) to intestinal stem cell acquisition in vitro and in vivo.

There are a handful of comments/concerns that would need to be addressed before publication.

Major points:

1. The author claimed: "the effect of ECM components on gene expression is not due to difference in morphology (2D collogen vs. 3D Matrigel)". The conclusion of the 2D vs. 3D experiment should be toned down to that organoid morphology (2D vs. 3D) does not directly impact on the expression of fetal-like genes. Otherwise more analysis of RNAseq data with different group of genes (e.g., in different mechanosensing pathways) should be provided with Fig. S1D. Also, would it be technically feasible to perform experiments of SI in collagen (3D) in all group of experiments? Directly comparing 3D Matrigel with 3D collagen avoids the concern of the 2D vs. 3D effect.

We apologise for the too strong claim of structure of growth vs. signalling of the ECM and its effect on the transcriptome. Indeed, the main message is that a changed morphology from 3D to 2D is not responsible for the expression of fetal-like genes. The paragraph will be rephrased.

Also, would it be technically feasible to perform experiments of SI in collagen (3D) in all group of experiments? Directly comparing 3D Matrigel with 3D collagen avoids the concern of the 2D vs. 3D effect.

To address this point, we want to refer to the excellent idea of growing established organoids in collagen (3D) vs. Matrigel (3D) as suggested by this reviewer (and reviewer #3) in Minor points #5. This circumvents the need for Wnt3a addition, which affects stem cell and Paneth cell gene expression.

1. For Fig. 1f, the authors should include overlapping stainings of Lyz (or Olfm4, CD44 etc.) and Adolase B signal, or they could perform Aldolase B staining in Lgr-5-DTR-GFP and/or Lyz-RFP organoid line. From the current data provided one cannot draw clear conclusions on the crypt morphology as claimed by the authors. Additionally, when talking about crypt morphology and apical accumulation of Actin specifically in the Lyz+ cells, the authors should show a higher zoom in of the picture and either add an orthogonal slice to see apical and basal side and the specific accumulation in one of the cell types, or also co-label with apical polarity markers.

We will perform additional co-stainings to further highlight the differences in the spatial distribution of differentiated cells and undifferentiated-crypt-like cells. Further we will provide higher magnification images highlighting the apical accumulation of Actin in the crypt-like structures, which can also be seen in mature organoids.

1. Authors referred the organoid transient change to fetal-like state. To exam the similarity of ECM-induced reprogramming with the regenerative-type of reprogramming, it would be essential to compare the expression of the selected fetal-like genes (Anxa3, Ly6a/Sca1, Msln, Col4a2 et al.), as well as bulk and single-cell (if applicable) RNA-seq data.

Here, we would like to refer to the excellent study by Yui et al. ([S. Yui et al., “YAP/TAZ-Dependent Reprogramming of Colonic Epithelium Links ECM Remodeling to Tissue Regeneration,” Cell Stem Cell, vol. 22, no. 1, pp. 35-49.e7, 2018.). In this study the authors detected the same gene signature in the repairing epithelium. We can provide a GSEA for the Ly6a+ signature that was derived from this paper, if necessary.

1. For in vivo data, authors were looking at the normal development of intestine. Following the point of organoid culture recapitulates regeneration, it would be relevant to check the in vivo ECM change by staining in the process of intestinal regeneration or discuss would the fetal-like genes be involved in regeneration.

We will address this point in the discussion as it also involves the study by Yui et al.

1. For Fig2.d and e, it would be important to measure compactness vs. the emergence/probability of Ly6+ cells to see if there is correlation.

If we understand the reviewer correctly, this would address the important relationship between cell shape and cell fate/type. However, this is a topic that needs more attention than a simple correlation and would exceed the scope of this manuscript as we are not able to modulate cell shape to make any further points about its effect on the fetal gene expression program.

1. In Fig.2d, Ly6a expression is very obscure, and it would be important to show control staining for cell boundaries (eg. Phalloidin, PM) to visualize which nuclei show Ki67 staining and are high or low in Ly6a (plus quantification).

We will improve the image in Fig.2d and include the mentioned Actin staining. In addition we will perform an analysis via Flow Cytometry to quantify the level of Ly6a staining and EdU positivity.

1. In Fig. 2f-g, FACS-ed Ly6a+ and Ly6- cells embedded in Matrigel can grow into organoids with crypts. Here the imaging of Paneth cell staining is not clear, and a quantification on number of Paneth cells per crypt would be very helpful to confirm the phenotype. Also, authors should either provide data on the initial size of seeded cell clusters and report organoid growth and cell type composition in more detail when plating from Ly6a+ and Ly6- cells or report the variation in the respective populations.

This comment suggests that we may not have described the experimental settings properly. The sorted cells were embedded as single cells, not as clusters, in drops of matrigel (10k cells/25ul Matrigel). The emergence of Paneth cells together with a normal organoid architecture grown from Ly6a+ cells shows their stem cell capacity, as has been shown by Yui et al. before from the regenerating colon. In addition, organoids from both cell populations (Ly6a+ and Ly6a-) could be passaged, indicating presence of intestinal stem cells.

1. The authors could also test whether Ly6+ cells have any advantages over Ly6- cells when grown on collagen I instead of Matrigel.

We will sort Ly6a+ and Ly6- negative cells and plate them on collagen I. It will be interesting to see if the Ly6a+ cells can give rise to the other cell types when plated on collagen or if they stay Ly6+ cells. This will also answer whether Ly6a+ need the presence of Ly6a- cells in the cultures. In addition, the experiment proposed in #6 will also highlight any proliferative advantage of Ly6a cells compared to Ly6-negative cells on collagen.

1. In Fig.3f, a control of membrane protein staining should be added for the experiment. The increased Laminin signal can be caused by the global increase of protein when there are more cells, or tissues are more compact. When authors make conclusion of "Dramatic remodelling of ECM during crypt formation ", the experiment should also count cell numbers vs. Laminin (intensity). The phenotype can come from increased area of interface between epithelium and mesenchyme instead of active remodelling.

We agree with the reviewer that by itself the IF images are not enough to make such a claim. However, we would point to the qPCR data and RNA in situ, that can be more easily normalised and shows the dramatic increase in expression of all laminins at birth. To show that laminin protein is increasing is more difficult than we initially anticipated. However, in the study by De Arcangelis (A. De Arcangelis et al., “Hemidesmosome integrity protects the colon against colitis and colorectal cancer,” Gut, vol. 66, no. 10, pp. 1748–1760, 2017.) the authors use an EDTA assay to show that the epithelium detaches easily when Itga6 is deleted. Within the figure, it seems also that the epithelium detaches easily at P2, compared to P14. As EDTA is disrupting laminin polymerisation, this would further indicate increased laminin protein deposition after birth.

1. The authors claim that intestinal stem cells in vivo are controlled by Laminin signalling that goes via Integrin alpha 6. However, there is no evidence provided that supports the contribution of ITGA6 in the in vivo setting. So, the authors should either tone down on that point or show a convincing in vivo experiment (e.g., inhibit ITGA6 in vivo by inhibitor injections or by extracting the ECM of a wild-type mouse and seeding intestinal epithelial cells without vs. with ITGA6 blocking antibody which should recapitulate the phenotype in Fig. 4 c.

We apologise for this confusion. We are well aware about the limitations of our Itga6 blocking experiment in vitro and its relevance in vivo. We tried to get material of the inducible VilCreER Itga6 mouse as referenced in the discussion of the manuscript, without any luck so far. Therefore we will highlight further that any claims about the laminin:Itga6 interaction can only be made in vitro.

1. Fig. 4: For the data of ITGA6 expression and all sorts of analysis on protein expression with staining, normalization with cell numbers should be performed.

The RNAseq data that shows the upregulation of Itga6 in the epithelium at E18 is normalized within. Our RNAscope only further validates these expression changes and highlights the specific enriched expression at the bottom of the nascent crypts. We can add quantification of the RNAscope if required.

1. Two questions on mechanisms:

2. What is the mechanism from ITGA signaling to Ly6a+ cell fate?

3. And would/how Laminin induce ITGA expression? Depends on how much the authors would like to go deep with the project, could be addressed further with functional studies, or touch on the topics with discussion.

These are important questions, however we do agree that this will go to deep for the scope of this manuscript. We will address these open questions in the discussion and leave the experimental part for a follow-up study.

**Minor points:**

1. Text in Fig.S1d regarding 'in' or 'on' collagen, could be clearer by changing the terms to 2D and 3D correspondingly.

We agree and the text will be changed accordingly.

1. Fig. S1a, it is great the authors showed that similar stiffness in Matrigel and collagen I. It would be important to check the concentration of collagen I vs. stiffness (also for increasing concentrations of Laminin in Fig. 3b), since this is also the type of ECM change that might lead to the change of cell status in cancer progression or collective cell migration.

We will perform further stiffness measurements of the hydrogels and update the Fig. S1a.

1. When plating intestinal epithelial cells on collagen I, is the Ly6+ phenotype altered upon Wnt addition? This is not so clear from the RNAseq data Fig S1d., so authors should provide antibody stainings (stem cells/Paneth cells). This could give insight whether Ly6+ cells are still able to convert into stem cells/ Paneth cells by changing morphogen concentration vs. ECM composition.

We will reanalyse the RNaseq dataset further, specifically analysing the ratio of stem cell and Paneth cell gene expression. However, as mentioned before, Wnt3a specifically does reduce the expression of Paneth cell markers.

Similar to this point, also enteroendocrine cell fate is absent in collagen I condition (Fig2.ab), the authors could address this point by medium induced EE cell fate.

Due to the reduced number of secretory cells the clustering in Fig2 a/b does not separate all the different cell lineages. However, EE cells are present in the collagen cultures as characterised by expression of Chga, just reduced in their number (see Supl Fig. 2B).

1. It would be more informative to indicate the thickness of ECM layer in culture of 2D collagen I, as well as the image of the whole well, demonstrating the morphological variation in the middle and peripheral of the ECM layer.

The thickness of the collagen layer is about 1mm in a 6well plate and we do not observe any morphological differences in the cells between the periphery and center of the well.

1. After the formation of PC/SC clusters, would ECM contribute to maintenance? Putting mature organoids from Matrigel to Collagen I 3D would help to clarify.

This is an interesting experiment that we will conduct, we thank the reviewer for this suggestion. Indeed, established organoids should be able to grow in collagen I without Wnt3a addition. The paper by Sachs et al. (N. Sachs, Y. Tsukamoto, P. Kujala, P. J. Peters, and H. Clevers, “Intestinal epithelial organoids fuse to form self-organizing tubes in floating collagen gels,” Development, vol. 144, no. 6, pp. 1107–1112, 2017. et al) used extensive washing with PBS to remove Matrigel from the organoids. We will go one step further and trying to completely remove laminin specifically by EDTA incubation, as has been shown recently (J. Y. Co et al., “Controlling Epithelial Polarity: A Human Enteroid Model for Host-Pathogen Interactions,” Cell Rep., vol. 26, no. 9, pp. 2509-2520.e4, 2019.). This should then also answer whether disruption of laminin signalling is sufficient to induce fetal-gene expression without the addition of collagen I in a 3D setting.

1. Check secretome and individual culture of Mesenchyme, see if the increase of Laminin is epithelium independent.

We agree that the mesenchyme is key for laminin production, therefore these are important questions. Our prediction would be that epithelium from birth (P0) versus adult might result in different responses on the mesenchyme. However, we feel these experiments are better suited for a follow-up study.

1. In general, the authors should look at cell polarity markers to check the ECM contribution to cell polarity in different cell types.

We thank the reviewer for the suggestions and as mentioned above, we will perform additional stainings.

Reviewer #2 (Significance (Required)):

**Significance:**

The work highlights the role of ECM on stem cell niche and is of great interest to the organoid and stem cell community.

Our field of expertise is image- and seq-technology-based quantitative biology, regeneration and mechanics in organoid.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

**Summary:**

Ramadan et al present a highly informative paper detailing how the Extracellular matrix influences the development of the intestine. Specifically, the authors provide a thorough analysis of how manipulating the components of the ECM can affect organoid growth, morphology, and gene expression of the organoids. Most importantly, the authors isolate laminin as a critical component of the ECM which impacts the development of fetal-like epithelium. While the in vitro work is generally compelling and of interest to the field, the in vivo data is someone lacking in depth and novelty. Particularly, these conclusions from the abstract could be much better supported: "This laminin:ITGA6 signalling is essential for the stem cell induction and crypt formation in vitro. Importantly, deletion of laminin in the adult mouse results in a fetal-like epithelium with a marked reduction of adult intestinal stem cells." The in vivo work was largely published previously and has caveats noted below, while the in vitro association of ITGA6 signaling with crypt formation is over-interpreted based upon an antibody blocking experiment and a lack of statistical rigor. Despite these concerns, this reviewer finds the work of considerable interest in an important area of the field (epithelial/stromal interactions of the intestine).

**Major Concerns:**

The use of the Ubc-Cre Lamc1-flox mouse model is an interesting way to test the impact of loss Lamc1 on intestinal development. However, with the Ubc-Cre, does the mouse model have other deleterious effects on the mouse beyond the intestine?

Can the authors use a more localized Cre to observe specifically the impacts of Lamc1 loss in the intestine? What is the fate of these mice? Can the authors show swiss roll, low mag sections to let the reader know the extent of this phenotype? OLFM4 and Ki67 IHC should be conducted over a timecourse to show how the changes occur over time after loss of Lamc1. How long does Lamc1's protein product perdure after tamoxifen treatment? More details of this exciting, in vivo validation of the authors' in vitro studies are key to elevating the impact of this work. However, it appears that much of this mouse model was previously published, but the previous findings are not well summarized in the current manuscript.

We will describe the model in more detail and refer readers to the excellent study of our collaborators which answers most of the raised questions. It is interesting to note that although a ubiquitous Cre was used to delete Lamc1 in adult mice, a phenotype was only observed in the intestine indicating a specific role for continuous laminin production here.

Can the authors show that ITGA6 loss has functional consequences in vivo with an epithelial knockout or via an organoid knockdown? A more rigorous genetic test of this proposed function would be important for substantiating the claims made in the abstract.

As referenced in the discussion of the manuscript, there is a VilCreER Itga6 mouse described in the literature (A. De Arcangelis et al., “Hemidesmosome integrity protects the colon against colitis and colorectal cancer,” Gut, vol. 66, no. 10, pp. 1748–1760, 2017.), which mainly focus on the colon. However, the authors use an EDTA assay in the small intestine to show that the epithelium detaches easily when Itga6 is deleted (Fig. 1J). Within the figure, it seems also that the epithelium detaches easily at P2, compared to P14. As EDTA is disrupting laminin polymerisation, this would further indicate increased laminin protein deposition after birth which is dependent on Itga6.

We tried to get material of the inducible VilCreER Itga6 mouse however without any luck so far.

We will conduct additional experiments regarding the Itga6 in vitro. In addition to including additional controls for the neutralizing antibody, we will genetically inactivate Itga6 via an inducible Crispr/Cas9. This should enable us to delete Itga6 when the cells are grown on collagen, and hence reduce the possibility of compensation in matrigel derived organoids.

The authors state that the changes in gene expression are not due to differences in morphology, but rather are specific to the components of the environment. While the authors show that organoids treated with Wnt3a "in Matrigel" and "CollagenI" appear to have similar morphologies and yet still result in a different gene expression profiles, it would be of great interest to see whether that difference persists without Wnt3a when organoids are "in Matrigel" and "in CollagenI". While the reviewer understands the difficulties of culturing organoids in 3D Collagen without Wnt3a, organoids can be indeed be cultured in 3D using "floating collagenI rings" (Sachs et al 2017).

This is an interesting experiment that we will conduct, we thank the reviewer for this suggestion. Indeed, established organoids should be able to grow in collagen I without Wnt3a addition. The paper by Sachs et al. (N. Sachs, Y. Tsukamoto, P. Kujala, P. J. Peters, and H. Clevers, “Intestinal epithelial organoids fuse to form self-organizing tubes in floating collagen gels,” Development, vol. 144, no. 6, pp. 1107–1112, 2017. et al) used extensive washing with PBS to remove Matrigel from the organoids. We will go one step further and trying to completely remove laminin specifically by EDTA incubation, as has been shown recently (J. Y. Co et al., “Controlling Epithelial Polarity: A Human Enteroid Model for Host-Pathogen Interactions,” Cell Rep., vol. 26, no. 9, pp. 2509-2520.e4, 2019.). This should then also answer whether disruption of laminin signalling is sufficient to induce fetal-gene expression without the addition of collagen I in a 3D setting.

Similarly, while the authors indicate that increasing Matrigel concentrations altered the gene expression patterns in a dose-dependent manner, it is unknown whether this can fully be attributed to the Matrigel composition, or whether the layer of Matrigel is providing the capability to transition from 2D to 3D culture.

We are not entirely sure, we understood this point. The Matrigel and the collagen I are mixed before they solidify, therefore enabling a homogenous hydrogel. The different hydrogels are not layered (if that is what the reviewer is referring to).

The authors cultured organoids in different concentrations of Laminin/CollagenIV when mixed with CollagenI. Can organoids be sustained only on a matrix of CollagenIV and/or Laminin? Would this show more direct differences between CollagenI vs. Laminin cultured organoids?

Organoids cannot be grown in pure Collagen IV, but pure laminin should be feasible. We did initial experiments with 3-5mg/ml laminin in PBS and that was sufficient to allow organoid growth. We will perform additional experiments with pure laminin and show the impact on organoid growth.

With the reduction in stem-cell and Paneth cells in the Lamc1-KO mice, it would also be of interest to determine what cell types are now prominent within the heavily elongated intestinal "crypt" structures seen in the Lamc1-KO mice and whether populations are more TA-cells or enterocytes to consider differentiation status of the cells. Additionally, it would also of interest to see if the Itga6 expression is significantly altered in the absence of Lamc1.

We will test expression changes for Itga6 in the Lamc1-KO mice, in the epithelium via qPCR. Additionally we can stain tissue from these mice for Sox9, Ki67 and differentiated markers eg. CD44, AldolaseB, Villin etc .to determine whether the elongated, hyperproliferative crypts contain progenitor cells or secretory enterocytes.

**Minor concerns:**

Matrigel is a complex matrix derived from mouse tumors. In many instances in the manuscript, the authors portray it as a more simpler mix of laminin/Collagen4 (fig 3a). It should be made clearer to the reader that Matrigel is not a mix of recombinant proteins, but a more clear depiction of how Matrigel is derived will be critical for this study, given the focus on specific ECM components and how they affect intestinal epithelial growth.

We agree and will change the oversimplified view of Matrigel.

Some labels of specific conditions would be appreciated in the figures as opposed to only the figure legends (ie. Fig. 1b and 1d should be labeled with comparisons; Fig. 3f labels of fluorescence, Fig. 4b label of itga6 staining).

We apologise for this and the Figure labels will be updated.

With the light staining of Lamc1 in-situ, it is hard to appreciate the expression of laminin within the stroma of the intestine when compared to Col4. This reviewer is also curious of the biological relevance of the concentrations of Laminin/CollagenIV when culturing organoids in Fig. 4a.

Indeed, the Lamc1 due to its lower expression than Col4a1 is more difficult to see. Maybe the reviewer overlooked Suppl.Fig.4A, where the blue channel of these in situ images show more contrast. If required, we can try to optimise the hybridisation times to increase the signal a bit further.

When culturing the organoids with the mixture of collagen and laminin or collagen IV, the concentrations of the two single components were selected similar to their concentrations in Matrigel. Regarding the absolute concentrations of laminin/collagen IV in vitro versus their “concentration” in vivo is much harder to answer. In addition to the unknown concentrations in vivo, there are many more Laminin types present with specific localisation and even specific receptor interactions. For our in vitro studies we relied on the Laminin present in EHS tumours, which is Laminin alpha 1 beta 1 gamma 1. We are currently investigating decellularization protocols to purify the ECM from mouse intestinal tissue, but again this would be more suited for a follow up study.

It would be appreciated if gene expression analyses presented in figures would include p-values to provide context for differences in gene expression.

The avoidance of statistical inference for most of the experiments was a deliberate choice. In line with several comments (e.g. 1. Vaux, D. L. (2012) Know when your numbers are significant. Nature. 492, 180–181), we chose to show all individual data points (with exception of Fig. 3D, n=5, to ease interpretation) without statistical testing. For most expression data, we have data from RNAseq, single-cell RNAseq and qPCRs repeated at different hydrogel concentrations to obtain reliable results. Further, the in vivo mesenchymal qPCR expression data was validated with RNA in situ hybridization showing the mainly mesenchymal expression.

As we will perform additional experiments for the revision of this paper, we can perform statistical tests in the key experiments (e.g. Itga6 experiment) to alleviate any concerns regarding significance.

In figure 3f, the authors report "immunofluorescence of laminin". How is this measured? Can more details be given about the antibody in the text and figure legend? Laminins are a family of genes, and it's not clear what's being demonstrated in this figure panel. Developmental stages of the samples are also not clear.

We apologise for the lack of labeling in Fig.3f. The details of the antibody were hidden in the Materials and Methods of the manuscript (Slides were incubated with Laminin Polyclonal Antibody (1/200, Thermo Fisher #PA5-22901) overnight at 4C ). This pan-laminin antibody reacts with most Laminin isoforms alpha1, alpha2, beta1, gamma1. We will declare it as a pan-laminin antibody in the Figure legend to help future readers.

Reviewer #3 (Significance (Required)):

This work is in an exciting "hot" area of research to understand the role of non-epithelial cells in intestinal epithelial development and function. The audience would be those in the GI field and those studying tissue-tissue interactions.

There's some concern that the in vivo portion of the manuscript (4th figure) uses a model that was previously characterized and published by this group, and that isn't clearly disclosed. The manuscript would benefit from more disclosure and detail about the in vivo phenotype. Such changes would substantially increase the impact and novelty of the study.

We would like to point out that we cited the paper of the original study that uses the model throughout the manuscript. We will disclose in more detail that this group did the study and that the reduction in stem cell genes was already mentioned in the original publication.

Author Response:

Reviewer #2 (Public Review):

1) The authors describe their algorithm as a tool that (i) was validated "across heterogeneous populations around the world"; (ii) has an "accuracy matching or exceeding human accuracy"; (iii) "is easy to use". I take issue with these three statements. First, the authors did not test the performance of their algorithm in clinical populations with sleep disorders, despite the fact that individuals with sleep disorders represent (logically) the vast majority of sleep recordings. Crucially, such a comparison was made in the best (to my knowledge) published automated sleep staging algorithm (Stephansen et al. Nature Communications 2018, doi: 10.1038/s41467-018-07229-3). The omission of this work is very surprising. Quantifying the impact of sleep disorders on a sleep scoring algorithm is critical for its deployment in sleep clinics.

We apologize as we were not clear in describing our training and testing data set. Indeed, both the training and testing set 1 included a significant number of individuals with sleep disorders. Indeed, about 30% of the individuals had moderate to severe sleep apnea (AHI >= 15). The validation dataset (DOD, or testing set 2) also includes 55 nights from individuals with obstructive sleep apnea (average AHI = 18.5 ± 16.2). Furthermore, both the training and testing set 1 included individuals with a medical diagnosis of insomnia, depression, diabete and hypertension.

The health status and demographics data of the training and testing sets have now been clarified throughout in the manuscript to avoid any such confusion:

1) Methods: We have added an extensive description of each dataset in the training and testing sets, including data on health and sleep disorders.

2) Results: We have added a new table to report and compare demographics/health data of the training and testing set, as suggested in a later comment by the reviewer.

3) Results: Performance results of the testing set 2 are now reported separately for healthy individuals and individuals with sleep disorders.

Second, the authors wrote that their algorithm is "matching or exceeding" human accuracy but seem to present uncorrected one-to-one comparisons to support their claim. The fact that an algorithm is better than some humans do not mean it exceeds human performance.

Thanks for noting that. We have now removed all instances of “exceeding human accuracy”.

Third, although I agree that the tool seems easy to use even for individuals with limited programming skills, it still requires some. I don't think someone who is used to software with graphical interfaces and who has never used (or heard of!) python would describe the tool as easy to use. This poses an important implementation challenge.

2) An important limitation of this algorithm is that it captures only one part of the visual examination of sleep data. Indeed, especially in clinical settings, the data is not only examined to establish the hypnogram but to also identify markers of common sleep disorders (e.g. sleep apnea, leg movements, etc). Although this algorithm could significantly speed up sleep scoring, it does not allow to detect these other important markers. Currently, and in link with the previous comment, the algorithm could not replace the visual inspection of the data for clinical diagnoses.

We have now revised the manuscript such that we discussed in this possibility in “Limitations and future directions” subsection of the new Discussion:

“The algorithm is not currently able to identify markers of common sleep disorders (such as sleep apnea, leg movements) and as such may not be suited for clinical purposes. It should be noted however that our software does include several other functions to quantify phasic events during sleep (slow-waves, spindles, REMs, artefacts) as well as sleep fragmentation of the hypnogram. Rather than replacing the crucial expertise of clinicians, YASA may thus provide a helpful starting point to accelerate clinical scoring of polysomnography recordings. Furthermore, future developments of the algorithm should prioritize automated scoring of clinical disorders, in particular apnea-hypopnea events. On the latter, YASA could implement some of the algorithms that have been developed over the last few years to detect apnea-hypopnea events from the ECG or respiratory channels (e.g. Varon et al. 2015; Koley and Dey 2013).”

3) The data were curated with some recordings or portions of recordings being excluded (see p. 7). While I understand that this curation is important for the training set, I think it should not be applied to the test set. Indeed, it goes contrary to the logic of automating sleep staging. For example, cutting the beginning and end of the recording according to sleep start and end (p. 7) supposes that the start and end of sleep are already known (i.e. it has already been scored).

This truncation step has now been removed from the pipeline and all the results have been updated accordingly. In addition, we have also removed all other exclusion criteria (e.g. PSG data quality, recording duration, etc) to improve the generalization power of the algorithm, thanks to the suggestions of the reviewer.

4) Two types of EEG derivations were used (C4-M1 or C4-Fpz). Was the performance impacted by this variable? Is it fair to assume that the choice of features (spectral features or summary statistics of time series data) could explain the absence of differences but that introducing new features (i.e. phase-sensitive features) could increase the influence of the choice of the derivation?

Thanks for raising this. First, our choice of the EEG reference was determined by the datasets: the CFS, CCSHS, MrOS, CHAT and HomePAP datasets were all referenced to Fpz, while the MESA, SHHS and DOD datasets were referenced to the contralateral mastoid. The montage of each dataset has now been added to the Methods section.

Second, as rightly pointed out by the reviewer, the features implemented in the algorithm were chosen to be robust to various recording montages. This is now explicitly discussed in the “Features extraction” subsection of the Methods:

“The features included in the current algorithm were chosen to be robust to different recording montages. As such, we did not include features that are dependent on the phase of the signal, and/or that require specific events detection (e.g. slow-waves, rapid eye movements). However, the time-domain features are dependent upon the amplitude of the signal, and the algorithm may fail if the input data is not expressed in standard units (uV) or has been z-scored prior to applying the automatic sleep staging.”

5) Given that markers of sleep stages are very different in EOG, EMG and EEG time series, could the authors explain the logic behind applying the same pre-processing and extracting the same features on these three very different types of data? Could this explain why the majority of the features in the top-20 features were EEG features?

We now provide a more detailed explanation on the inclusion of EOG and EMG features in the “Features extraction” subsection of the Methods:

“These features were selected based on prior work in features-based classification algorithms for automatic sleep staging (Krakovská and Mezeiová 2011; Lajnef et al. 2015; Sun et al. 2017). For example, it was previously reported that the permutation entropy of the EOG/EMG as well as the EEG spectral powers in the traditional frequency bands are the most important features for accurate sleep staging (Lajnef et al. 2015), thus warranting their inclusion in the current algorithm. Several other features are derived from the authors’ previous works with entropy/fractal dimension metrics1. ” https://github.com/raphaelvallat/antropy

Furthermore, we have added a “Limitations and future directions” section in the Discussion in which we propose future improvements of the algorithm. One of these potential improvements is the development of EOG and EMG features that would provide a higher discrimination of the sleep stages:

“This suggests that one way to improve performance on this population could be the inclusion of more EEG channels and/or bilateral EOGs. For instance, using the negative product of bilateral EOGs may increase sensitivity to rapid eye movements in REM sleep or slow eye movements in N1 sleep (Stephansen et al. 2018; Agarwal et al. 2005). Interestingly, the Perslev 2021 algorithm does not use an EMG channel, which is consistent with our observation of a negligible benefit on accuracy when adding EMG to the model. This may also indicate that while the current set of features implemented in the algorithm performs well for EEG and EOG channels, it does not fully capture the meaningful dynamic information nested within muscle activity during sleep.”

6) Sleep scoring guidelines incorporate not only what can be observed on a given epoch of data but also what is observed in the previous epoch(s). For example, an epoch can be scored as N2 even if there is no marker of N2 but there was (i) a marker of N2 in a previous epoch, (ii) no reason to change the score since. To reproduce this, the authors employed a symmetrical smoothing approach (a combination of a triangular-weighted rolling average and asymmetrical rolling average). Why did the authors choose to incorporate data from following epochs, which is not implemented in established guidelines? How was the duration of the smoothing window chosen? Indeed, 5 minutes appear as rather long could explain the poor performance of the algorithm for fast changing portions of the data (i.e. N1 or transitions). Importantly, these transitions can be very relevant in clinical settings and to establish a diagnosis.

This is a great question. We have addressed this in the revised manuscript.

Temporal smoothing

We have also conducted a new analysis of the influence of the temporal smoothing on the performance. The results are described in Supplementary File 3a. Briefly, using a cross-validation approach, we have tested a total of 49 combinations of time lengths for the past and centered smoothing windows. Results demonstrated that the best performance is obtained when using a 2 min past rolling average in combination with a 7.5 minutes centered, triangular-weighted rolling average. Removing the centered rolling average resulted in poorer performance, suggesting that there is an added benefit of incorporating data from both before and after the current epoch. Removing both the past and centered rolling averages resulted in the worst performance (-3.6% decrease in F1-macro). Therefore, the new version of the manuscript and algorithm now uses a 2 min past and 7.5 min centered rolling averages. All the results in the manuscript have been updated accordingly. We have now edited the “Smoothing and normalization” subsection of the Methods section as follow:

“In particular, the features were first duplicated and then smoothed using two different rolling windows: 1) a 7.5 minutes centered, and triangular-weighted rolling average (i.e. 15 epochs centered around the current epoch with the following weights: [0.125, 0.25, 0.375, 0.5, 0.625, 0.75, 0.875, 1., 0.875, 0.75, 0.625, 0.5, 0.375, 0.25, 0.125]), and 2) a rolling average of the last 2 minutes prior to the current epoch. The optimal time length of these two rolling windows was found using a parameter search with cross-validation (Supplementary File 3a). [...] The final model includes the 30-sec based features in original units (no smoothing or scaling), as well as the smoothed and normalized version of these raw features.”

Reviewer #3 (Public Review):

This study presents a new sleep scoring tool that is based on a classification algorithm using machine-learning approaches in which a set of features is extracted from the EEG signal. The algorithm was trained and validated on a very large number of nocturnal sleep datasets including participants with various ethnicities, age and health status. Results show that the algorithm offers a high level of sensitivity, specificity and accuracy matching or sometimes even exceeding that of typical interscorer agreement. The conclusions are supported by the data. Importantly, a measure of the algorithm's confidence is provided for each scored epoch in order to guide users during their review of the output. The software is described as easy to use, computationally low-demanding, open source and free. This paper addresses an important need for the field of sleep research. There is indeed a lack of accurate, flexible and open source sleep scoring tools. I would like to commend the authors for their efforts in providing such a tool for the community and for their adherence to the open science framework as the data and codes related to the current manuscript are made available. I predict that this automated tool will be of use for a large number of researchers in the field. However, there are plenty of automated sleep scoring tools already available in the field (most of them are not open source and rather expensive, as noted by the authors). The current work does not provide a clear view on whether the new algorithm presented in this research performs better than algorithms already available in the field. No formal comparisons between algorithms is provided and the matter is not discussed in the paper.

Thanks so much for pointing this out. We have now added this relevant reference throughout the manuscript. To build on the reviewer’s point, the current algorithm and Stephansen’s algorithm did not use the same public data. The Stephansen 2018 algorithm was trained and validated on “10 different cohorts recorded at 12 sleep centers across 3 continents: SSC, WSC, IS-RC, JCTS, KHC1, AHC, IHC, DHC, FHC and CNC”, none of which are included in the training/testing sets of the current algorithm. Nevertheless, we certainly agree that the manuscript will benefit from a more extensive comparison against existing tools. To this end, we have made several major modifications to the manuscript. First, we have added a dedicated paragraph in the Introduction to review existing sleep staging algorithms:

“Advances in machine-learning have led efforts to classify sleep with automated systems. Indeed, recent years have seen the emergence of several automatic sleep staging algorithms. While an exhaustive review of the existing sleep staging algorithms is out of the scope of this article, we review below — in chronological order — some of the most significant algorithms of the last five years. For a more in-depth review, we refer the reader to Fiorillo et al. 2019. The Sun et al. 2017 algorithm was trained on 2,000 PSG recordings from a single sleep clinic. The overall Cohen's kappa on the testing set was 0.68 (n=1,000 PSG nights). The “Z3Score” algorithm (Patanaik et al. 2018) was trained and evaluated on ~1,700 PSG recordings from four datasets, with an overall accuracy ranging from 89.8% in healthy adults/adolescents to 72.1% in patients with Parkinson’s disease. The freely available “Stanford-stage” algorithm (Stephansen et al. 2018) was trained and evaluated on 10 clinical cohorts (~3,000 recordings). The overall accuracy was 87% against the consensus scoring of several human experts in an independent testing set. The “SeqSleepNet” algorithm (Phan et al. 2019) was trained and tested using a 20-fold cross-validation on 200 nights (overall accuracy = 87.1%). Finally, the recent U-Sleep algorithm (Perslev et al. 2021) was trained and evaluated on PSG recordings from 15,660 participants of 16 clinical studies. While the overall accuracy was not reported, the mean F1-score against the consensus scoring of five human experts was 0.79 for healthy adults and 0.76 for patients with sleep apnea.”

Second, and importantly, we now perform an in-depth comparison of YASA’s performance against the Stephansen 2018 algorithm and the Perslev 2021 algorithm using the same data for all three datasets. Specifically, we have applied the three algorithms to each night of the Dreem Open Datasets (DOD) and compared their performance in dedicated tables in the Results section (Table 2 and Table 3). This procedure is fully described in a new “Comparison against existing algorithms” subsection of the Methods. None of these algorithms included nights from the DOD in their training set, thus ensuring a fair comparison of the three algorithms. Related to point 4 of the Essential Revisions, performance of the three algorithms are reported separately for healthy individuals (DOD-Healthy, n=25) and patients with sleep apnea (DOD-Obstructive, n=50). To facilitate future validation of our algorithm, we also provide the predicted hypnograms of each night in Supplementary File 1 (healthy) and Supplementary File 2 (patients).

Overall, the comparison results show that YASA’s accuracy is not significantly different from the Stephansen 2018 algorithm for both healthy adults and patients with obstructive sleep apnea. The accuracy of the Perslev 2021 algorithm is not significantly different from YASA in healthy adults, but is higher in patients with sleep apnea. However, it should be noted that while the YASA algorithm only uses one central EEG, one EOG and one EMG, the Perslev 2021 algorithm uses all available EEGs as well as two EOGs. This suggests that adding more EEG channels and/or using the two EOGs may improve the performance of YASA in patients with sleep apnea. Though an important counterpoint is that YASA requires a far less extensive array of data (channels) to accomplish very similar levels of accuracy, which has the favorable benefit of reducing analysis computational and processing demands, improves speed of analysis (i.e. a few seconds per recording versus ~10 min for the Stephansen 2018 algorithm), and is amenable to more data recordings since many may not have sufficient EEG channels. All these points are now discussed in detail in the new “Limitations and future directions” subsection of the Discussion (see point 3 of the Essential Revisions).

There are some overstatements in the manuscript. For example, the algorithm was trained and validated on nocturnal sleep data. Sleep characteristics (eg duration and distribution of sleep stages etc.) are different, for example, during diurnal sleep (nap) and the algorithm might not perform as well on nap data. As such, the tool might not be as "universal" as stated in the title. Additionally, as human scores are used as the ground-truth for the validation step, it might be misleading to state that "this tool offers high sleep-staging accuracy matching or exceeding human accuracy". The algorithm exceeded the accuracy of some human scorers and matched the scores of the best scorer.

We have now removed the word “universal” from the title and replaced “exceeded human accuracy” with “matched human accuracy”. Furthermore, we have now added the fact that the algorithm was trained and validated only on nocturnal data in the Limitations section of the discussion, and as such, noted that there is the possibility that the algorithm may not perform at the same accuracy levels for daytime nap data.

No reflection on further improvement is offered in the paper. The algorithm performs worse on N1 stage, older individuals and patients presenting sleep disorders (sleep fragmentation) and it is unclear how this could be improved in future research. In the same vein, the current work does not present performance accuracy separately for healthy individuals and patients when it is expected that accuracy would be poorer in the patient group.

The revised manuscript now includes a dedicated section in the Discussion to propose ideas for improvements.

First, we have now added a “Limitations and Future Directions” subsection in the Discussion to present ideas for improving the algorithm, with a particular focus on fragmented nights and/or nights from patients with sleep disorders:

“Despite its numerous advantages, there are limitations to the algorithm that must be considered. These are discussed below, together with ideas for future improvements of the algorithm. First, while the accuracy of YASA against consensus scoring was not significantly different from the Stephansen 2018 and Perslev 2021 algorithms on healthy adults, it was significantly lower than the latter algorithm on patients with obstructive sleep apnea. The Perslev 2021 algorithm used all available EEGs and two (bilateral) EOGs, whereas YASA’s scoring was based on one central EEG, one EOG and one EMG. This suggests that one way to improve performance in this population could be the inclusion of more EEG channels and/or bilateral EOGs. For instance, using the negative product of bilateral EOGs may increase sensitivity to rapid eye movements in REM sleep or slow eye movements in N1 sleep (Stephansen et al. 2018; Agarwal et al. 2005). Interestingly, the Perslev 2021 algorithm does not use an EMG channel, which is consistent with our observation of a negligible benefit on accuracy when adding EMG to the model. This may also indicate that while the current set of features implemented in the algorithm performs well for EEG and EOG channels, it does not fully capture the meaningful dynamic information nested within muscle activity during sleep.”

Second, we have now conducted a random forest analysis to identify the main contributors of accuracy variability. The analysis is described in detail in the “Moderator Analyses” subsection of the Results as well as Supplementary File 3b, the revision now states:

“To better understand how these moderators influence variability in accuracy, we quantified the relative contribution of the moderators using a random forest analysis. Specifically, we included all aforementioned demographics variables in the model, together with medical diagnosis of depression, diabetes, hypertension and insomnia, and features extracted from the ground-truth sleep scoring such as the percentage of each sleep stage, the duration of the recording and the percentage of stage transitions in the hypnograms. The outcome variable of the model was the accuracy score of YASA against ground-truth sleep staging, calculated separately for each night. All the nights in the testing set 1 were included, leading to a sample size of 585 unique nights. Results are presented in Supplementary File 3b. The percentage of N1 sleep and percentage of stage transitions — both markers of sleep fragmentation — were the two top predictors of accuracy, accounting for 40% of the total relative importance. By contrast, the combined contribution of age, sex, race and medical diagnosis of insomnia, hypertension, diabete and depression accounted for roughly 10% of the total importance.”

In addition, the performance of the algorithm in the DOD testing dataset is now reported separately for healthy individuals and patients with sleep disorders.

As requested by the reviewer, we now analyze and report the performance of YASA on the DOD testing set separately for healthy individuals (DOD-healthy) and patients with obstructive sleep apnea (DOD-Obstructive), which can be found in section “Testing set 2”.

There is series of methodological choices that is not justified. For example, nights were cropped to 15 minutes before and after sleep to remove irrelevant extra periods of wakefulness or artefacts on both ends of the recording. This represents an issue for the computation of important sleep measures such as sleep efficiency and latency as the onset/offset of sleep might be missed. It is also unclear how the features were selected and a description of said features is currently missing. The custom sleep stage weights procedure is unclear. The length of the time window for the smoothing procedure seems arbitrary. Last, it is currently unclear when / how the EEG and EMG data were analyzed.

As recommended by the reviewers, the 15-min truncation step has now been removed from the pipeline. Furthermore, the Methods section has been improved to provide more details on the features. Finally, the best class-weights and smoothing windows are now found using a cross-validation analysis on the training set. For more details, we refer the reviewer to the “Justification for some methodological choices” section below.

Author Response:

Reviewer #1 (Public Review):

The manuscript provides very high quality single-cell physiology combined with population physiology to reveal distinctives roles for two anatomically dfferent LN populations in the cockroach antennal lobe. The conclusion that non-spiking LNs with graded responses show glomerular-restricted responses to odorants and spiking LNs show similar responses across glomeruli generally supported with strong and clean data, although the possibility of selective interglomerular inhibition has not been ruled out. On balance, the single-cell biophysics and physiology provides foundational information useful for well-grounded mechanistic understanding of how information is processed in insect antennal lobes, and how each LN class contributes to odor perception and behavior.

Thank you for this positive feedback.

Reviewer #2 (Public Review):

The manuscript "Task-specific roles of local interneurons for inter- and intraglomerular signaling in the insect antennal lobe" evaluates the spatial distribution of calcium signals evoked by odors in two major classes of olfactory local neurons (LNs) in the cockroach P. Americana, which are defined by their physiological and morphological properties. Spiking type I LNs have a patchy innervation pattern of a subset of glomeruli, whereas non-spiking type II LNs innervate almost all glomeruli (Type II). The authors' overall conclusion is that odors evoke calcium signals globally and relatively uniformly across glomeruli in type I spiking LNs, and LN neurites in each glomerulus are broadly tuned to odor. In contrast, the authors conclude that they observe odor-specific patterns of calcium signals in type II nonspiking LNs, and LN neurites in different glomeruli display distinct local odor tuning. Blockade of action potentials in type I LNs eliminates global calcium signaling and decorrelates glomerular tuning curves, converting their response profile to be more similar to that of type II LNs. From these conclusions, the authors infer a primary role of type I LNs in interglomerular signaling and type III LNs in intraglomerular signaling.

The question investigated by this study - to understand the computational significance of different types of LNs in olfactory circuits - is an important and significant problem. The design of the study is straightforward, but methodological and conceptual gaps raise some concerns about the authors' interpretation of their results. These can be broadly grouped into three main areas.

1) The comparison of the spatial (glomerular) pattern of odor-evoked calcium signals in type I versus type II LNs may not necessarily be a true apples-to-apples comparison. Odor-evoked calcium signals are an order of magnitude larger in type I versus type II cells, which will lead to a higher apparent correlation in type I cells. In type IIb cells, and type I cells with sodium channel blockade, odor-evoked calcium signals are much smaller, and the method of quantification of odor tuning (normalized area under the curve) is noisy. Compare, for instance, ROI 4 & 15 (Figure 4) or ROI 16 & 23 (Figure 5) which are pairs of ROIs that their quantification concludes have dramatically different odor tuning, but which visual inspection shows to be less convincing. The fact that glomerular tuning looks more correlated in type IIa cells, which have larger, more reliable responses compared to type IIb cells, also supports this concern.

We agree with the reviewer that "the comparison of the spatial (glomerular) pattern of odor-evoked calcium signals is not necessarily a true apples-to-apples comparison". Type I and type II LNs are different neuron types. Given their different physiology and morphology, this is not even close to a "true apples-to-apples comparison" - and a key point of the manuscript is to show just that.

As we have emphasized in response to Essential Revision 1, the differences in Ca2+ signals are not an experimental shortcoming but a physiologically relevant finding per se. These data, especially when combined with the electrophysiological data, contribute to a better understanding of these neurons’ physiological and computational properties.

It is physiologically determined that the Ca2+ signals during odorant stimulation in the type II LNs are smaller than in type I LNs. And yes, the signals are small because small postsynpathetic Ca2+ currents predominantly cause the signals. Regardless of the imaging method, this naturally reduces the signal-to-noise ratio, making it more challenging to detect signals. To address this issue, we used a well-defined and reproducible method for analyzing these signals. In this context, we do not agree with the very general criticism of the method. The reviewer questions whether the signals are odorant-induced or just noise (see also minor point 12). If we had recorded only noise, we would expect all tuning curves (for each odorant and glomerulus) to be the same. In this context, we disagree with the reviewer's statement that the tuning curves do not represent the Ca2+ signals in Figure 4 (ROI 4 and 15) and Figure 5 (ROI 16 and 23). This debate reflects precisely the kind of 'visual inspection bias' that our clearly defined analysis aims to avoid. On close inspection, the differences in Ca2+ signals can indeed be seen. Figure II (of this letter) shows the signals from the glomeruli in question at higher magnification. The sections of the recordings that were used for the tuning curves are marked in red.

Figure II: Ca2+ signals of selected glomeruli that were questioned by the reviewer.

2) An additional methodological issue that compounds the first concern is that calcium signals are imaged with wide-field imaging, and signals from each ROI likely reflect out of plane signals. Out of plane artifacts will be larger for larger calcium signals, which may also make it impossible to resolve any glomerular-specific signals in the type I LNs.

Thank you for allowing us to clarify this point. The reviewer comment implies that the different amplitudes of the Ca2+ signals indicate some technical-methodological deficiency (poorly chosen odor concentration). But in fact, this is a key finding of this study that is physiologically relevant and crucial for understanding the function of the neurons studied. These very differences in the Ca2+ signals are evidence of the different roles these neurons play in AL. The different signal amplitudes directly show the distinct physiology and Ca2+ sources that dominate the Ca2+ signals in type I and type II LNs. Accordingly, it is impractical to equalize the magnitude of Ca2+ signals under physiological conditions by adjusting the concentration of odor stimuli.

In the following, we address these issues in more detail: 1) Imaging Method 2) Odorant stimulation 3) Cell type-specific Ca2+ signals

1) Imaging Method:

Of course, we agree with the reviewer comment that out-of-focus and out-of-glomerulus fluorescence can potentially affect measurements, especially in widefield optical imaging in thick tissue. This issue was carefully addressed in initial experiments. In type I LNs, which innervate a subset of glomeruli, we detected fluorescence signals, which matched the spike pattern of the electrophysiological recordings 1:1, only in the innervated glomeruli. In the not innervated ROIs (glomeruli), we detected no or comparatively very little fluorescence, even in glomeruli directly adjacent to innervated glomeruli.

To illustrate this, FIGURE I (of this response letter) shows measurements from an AL in which an uniglomerular projection neuron was investigated in an a set of experiments that were not directly related to the current study. In this experiment, a train of action potential was induced by depolarizing current. The traces show the action potential induced fluorescent signals from the innervated glomerulus (glomerulus #1) and the directly adjacent glomeruli.

These results do not entirely exclude that the large Ca2+ signals from the innervated LN glomeruli may include out-of-focus and out-of-glomerulus fluorescence, but they do show that the bulk of the signal is generated from the recorded neuron in the respective glomeruli.

Figure I: Simultaneous electrophysiological and optophysiological recordings of a uniglomerular projection using the ratiometric Ca2+ indicator fura-2. The projection neuron has its arborization in glomerulus 1. The train of action potentials was induced with a depolarizing current pulse (grey bar).

2) Odorant Stimulation: It is important to note that the odorant concentration cannot be varied freely. For these experiments, the odorant concentrations have to be within a 'physiologically meaningful' range, which means: On the one hand, they have to be high enough to induce a clear response in the projection neurons (the antennal lobe output). On the other hand, however, the concentration was not allowed to be so high that the ORNs were stimulated nonspecifically. These criteria were met with the used concentrations since they induced clear and odorant-specific activity in projection neurons.

3) Cell type-specific Ca2+ signals:

The differences in Ca2+ signals are described and discussed in some detail throughout the text (e.g., page 6, lines 119-136; page 9, lines 193-198; page 10-11, lines 226-235; page 14-15, line 309-333). Briefly: In spiking type I LNs, the observed large Ca2+ signals are mediated mainly by voltage-depended Ca2+ channels activated by the Na+-driven action potential's strong depolarization. These large Ca2+ signals mask smaller signals that originate, for example, from excitatory synaptic input (i.e., evoked by ligand-activated Ca2+ conductances). Preventing the firing of action potentials can unmask the ligand-activated signals, as shown in Figure 4 (see also minor comments 8. and 10.). In nonspiking type II LNs, the action potential-generated Ca2+ signals are absent; accordingly, the Ca2+ signals are much smaller. In our model, the comparatively small Ca2+ signals in type II LNs are mediated mainly by (synaptic) ligand-gated Ca2+ conductances, possibly with contributions from voltage-gated Ca2+ channels activated by the comparatively small depolarization (compared with type I LNs).

Accordingly, our main conclusion, that spiking LNs play a primary role in interglomerular signaling, while nonspiking LNs play an essential role in intraglomeular signaling, can be DIRECTLY inferred from the differences in odorant induced Ca2+ signals alone.

a) Type I LN: The large, simultaneous, and uniform Ca2+ signals in the innervated glomeruli of an individual type I LN clearly show that they are triggered in each glomerulus by the propagated action potentials, which conclusively shows lateral interglomerular signal propagation.

b) Type II LNs: In the type II LNs, we observed relatively small Ca2+ signals in single glomeruli or a small fraction of glomeruli of a given neuron. Importantly, the time course and amplitude of the Ca2+ signals varied between different glomeruli and different odors. Considering that type II LNs in principle, can generate large voltage-activated Ca2+ currents (larger that type I LNS; page 4, lines 82-86, Husch et al. 2009a,b; Fusca and Kloppenburg 2021), these data suggest that in type II LNs electrical or Ca2+ signals spread only within the same glomerulus; and laterally only to glomeruli that are electrotonically close to the odorant stimulated glomerulus.

Taken together, this means that our conclusions regarding inter- and intraglomerular signaling can be derived from the simultaneously recorded amplitudes and the dynamics of the membrane potential and Ca2+ signals alone. This also means that although the correlation analyses support this conclusion nicely, the actual conclusion does not ultimately depend on the correlation analysis. We had (tried to) expressed this with the wording, “Quantitatively, this is reflected in the glomerulus-specific odorant responses and the diverse correlation coefficiiants across…” (page 10, lines 216-217) and “ …This is also reflected in the highly correlated tuning curves in type I LNs and low correlations between tuning curves in type II LNs”(page 13, lines 293-295).

3) Apart from the above methodological concerns, the authors' interpretation of these data as supporting inter- versus intra-glomerular signaling are not well supported. The odors used in the study are general odors that presumably excite feedforward input to many glomeruli. Since the glomerular source of excitation is not determined, it's not possible to assign the signals in type II LNs as arising locally - selective interglomerular signal propagation is entirely possible. Likewise, the study design does not allow the authors to rule out the possibility that significant intraglomerular inhibition may be mediated by type I LNs.

The reviewer addresses an important point. However, from the comment, we get the impression that he/she has not taken into account the entire data set and the DISCUSSION. In fact, this topic has already been discussed in some detail in the original version (page 12, lines 268-271; page 15-16; lines 358-374). This section even has a respective heading: "Inter- and intraglomerular signaling via nonspiking type II LNs" (page 15, line 338). We apologize if our explanations regarding this point were unclear, but we also feel that the reviewer is arguing against statements that we did not make in this way.

a) In 11 out of 18 type II LNs we found 'relatively uncorrelated' (r=0.43±0.16, N=11) glomerular tuning curves. These experiments argue strongly for a 'local excitation' with restricted signal propagation and do not provide support for interglomerular signal propagation. Thus, these results support our interpretation of intraglomerular signaling in this set of neurons.

b) In 7 out of 18 experiments, we observed 'higher correlated' glomerular tuning curves (r=0.78±0.07, N=7). We agree with the reviewer that this could be caused by various mechanisms, including simultaneous input to several glomeruli or by interglomerular signaling. Both possibilities were mentioned and discussed in the original version of the manuscript (page 12, lines 268-271; page 15-16; lines 358-374). In the Discussion, we considered the latter possibility in particular (but not exclusively) for the type IIa1 neurons that generate spikelets. Their comparatively stronger active membrane properties may be particularly suitable for selective signal transduction between glomeruli.

c) We have not ruled out that local signaling exists in type I LNs – in addition to interglomerular signaling. The highly localized Ca2+ signals in type I LNs, which we observed when Na+ -driven action potential generation was prevented, may support this interpretation. However, we would like to reiterate that the simultaneous electrophysiological and optophysiological recordings, which show highly correlated glomerular Ca2+ dynamics that match 1:1 with the simultaneously recorded action potential pattern, clearly suggest interglomerular signaling. We also want to emphasize that this interpretation is in agreement with previous models derived from electrophysiological studies(Assisi et al., 2011; Fujiwara et al., 2014; Hong and Wilson, 2015; Nagel and Wilson, 2016; Olsen and Wilson, 2008; Sachse and Galizia, 2002; Wilson, 2013).

In light of the reviewer's comment(s), we have modified the text to clarify these points (page 14, lines 317-319).

Reviewer #3 (Public Review):

To elucidate the role of the two types of LNs, the authors combined whole-cell patch clamp recordings with calcium imaging via single cell dye injection. This method enables to monitor calcium dynamics of the different axons and branches of single LNs in identified glomeruli of the antennal lobe, while the membrane potential can be recorded at the same time. The authors recorded in total from 23 spiking (type I LN) and 18 non-spiking (type II LN) neurons to a set of 9 odors and analyzed the firing pattern as well as calcium signals during odor stimulation for individual glomeruli. The recordings reveal on one side that odor-evoked calcium responses of type I LNs are odor-specific, but homogeneous across glomeruli and therefore highly correlated regarding the tuning curves. In contrast, odor-evoked responses of type II LNs show less correlated tuning patterns and rather specific odor-evoked calcium signals for each glomerulus. Moreover the authors demonstrate that both LN types exhibit distinct glomerular branching patterns, with type I innervating many, but not all glomeruli, while type II LNs branch in all glomeruli.

From these results and further experiments using pharmacological manipulation, the authors conclude that type I LNs rather play a role regarding interglomerular inhibition in form of lateral inhibition between different glomeruli, while type II LNs are involved in intraglomerular signaling by developing microcircuits in individual glomeruli.

In my opinion the methodological approach is quite challenging and all subsequent analyses have been carried out thoroughly. The obtained data are highly relevant, but provide rather an indirect proof regarding the distinct roles of the two LN types investigated. Nevertheless, the conclusions are convincing and the study generally represents a valuable and important contribution to our understanding of the neuronal mechanisms underlying odor processing in the insect antennal lobe. I think the authors should emphasize their take-home messages and resulting conclusions even stronger. They do a good job in explaining their results in their discussion, but need to improve and highlight the outcome and meaning of their individual experiments in their results section.

Thank you for this positive feedback.

References:

Assisi, C., Stopfer, M., Bazhenov, M., 2011. Using the structure of inhibitory networks to unravel mechanisms of spatiotemporal patterning. Neuron 69, 373–386. https://doi.org/10.1016/j.neuron.2010.12.019

Das, S., Trona, F., Khallaf, M.A., Schuh, E., Knaden, M., Hansson, B.S., Sachse, S., 2017. Electrical synapses mediate synergism between pheromone and food odors in Drosophila melanogaster . Proc Natl Acad Sci U S A 114, E9962–E9971. https://doi.org/10.1073/pnas.1712706114

Fujiwara, T., Kazawa, T., Haupt, S.S., Kanzaki, R., 2014. Postsynaptic odorant concentration dependent inhibition controls temporal properties of spike responses of projection neurons in the moth antennal lobe. PLOS ONE 9, e89132. https://doi.org/10.1371/journal.pone.0089132

Fusca, D., Husch, A., Baumann, A., Kloppenburg, P., 2013. Choline acetyltransferase-like immunoreactivity in a physiologically distinct subtype of olfactory nonspiking local interneurons in the cockroach (Periplaneta americana). J Comp Neurol 521, 3556–3569. https://doi.org/10.1002/cne.23371

Fuscà, D., and Kloppenburg, P. (2021). Odor processing in the cockroach antennal lobe-the network components. Cell Tissue Res.

Hong, E.J., Wilson, R.I., 2015. Simultaneous encoding of odors by channels with diverse sensitivity to inhibition. Neuron 85, 573–589. https://doi.org/10.1016/j.neuron.2014.12.040

Husch, A., Paehler, M., Fusca, D., Paeger, L., Kloppenburg, P., 2009a. Calcium current diversity in physiologically different local interneuron types of the antennal lobe. J Neurosci 29, 716–726. https://doi.org/10.1523/JNEUROSCI.3677-08.2009

Husch, A., Paehler, M., Fusca, D., Paeger, L., Kloppenburg, P., 2009b. Distinct electrophysiological properties in subtypes of nonspiking olfactory local interneurons correlate with their cell type-specific Ca2+ current profiles. J Neurophysiol 102, 2834–2845. https://doi.org/10.1152/jn.00627.2009

Nagel, K.I., Wilson, R.I., 2016. Mechanisms Underlying Population Response Dynamics in Inhibitory Interneurons of the Drosophila Antennal Lobe. J Neurosci 36, 4325–4338. https://doi.org/10.1523/JNEUROSCI.3887-15.2016

Neupert, S., Fusca, D., Kloppenburg, P., Predel, R., 2018. Analysis of single neurons by perforated patch clamp recordings and MALDI-TOF mass spectrometry. ACS Chem Neurosci 9, 2089–2096.

Olsen, S.R., Bhandawat, V., Wilson, R.I., 2007. Excitatory interactions between olfactory processing channels in the Drosophila antennal lobe. Neuron 54, 89–103. https://doi.org/10.1016/j.neuron.2007.03.010

Olsen, S.R., Wilson, R.I., 2008. Lateral presynaptic inhibition mediates gain control in an olfactory circuit. Nature 452, 956–960. https://doi.org/10.1038/nature06864

Sachse, S., Galizia, C., 2002. Role of inhibition for temporal and spatial odor representation in olfactory output neurons: a calcium imaging study. J Neurophysiol. 87, 1106–17.

Shang, Y., Claridge-Chang, A., Sjulson, L., Pypaert, M., Miesenbock, G., 2007. Excitatory Local Circuits and Their Implications for Olfactory Processing in the Fly Antennal Lobe. Cell 128, 601–612.

Wilson, R.I., 2013. Early olfactory processing in Drosophila: mechanisms and principles. Annu Rev Neurosci 36, 217–241. https://doi.org/10.1146/annurev-neuro-062111-150533

Yaksi, E., Wilson, R.I., 2010. Electrical coupling between olfactory glomeruli. Neuron 67, 1034–1047. https://doi.org/10.1016/j.neuron.2010.08.041

Author Response:

Reviewer #1 (Public Review):

The paper is a tour-de-force across multiple techniques and model systems from classical forward screening in C. elegans over ChIP to targeted CRISPR mutagenesis. The data is of a very high quality and supports most of the authors' claims strongly and convincingly. Finally, the manuscript is well written and, in spite of complex experiments and genetics, interesting and easy to comprehend.

• CAMTA, as the name CaM-binding transcription activator implies, have been studied previously and across many different organisms including plants, mice and humans. It was thus presumed and in part shown that CAMTAs regulate transcription depending on CaM levels.
• The authors confirm that the gene cmd-1 (encoding CaM) is directly regulated by Camt-1 by using a combination of cell-specific RNAseq and ChIP. This allows them to identify three binding sites upstream of the cmb-1 gene that bind to Camt-1.
• Moreover, the authors show that overexpression of CaM in the nervous system fully rescues the observed behavioral phenotypes.
• Importantly, the authors make another discovery. They show that CaM can directly repress its own transcription by binding to specific residues of Camt-1. Thereby, the authors argue, Camt-1 is used to precisely and bidirectionally regulate CaM levels dependent on the cell, animal's state etc.

The reported data are interesting and, in particular, the aspect that CAMTAs likely act as activators AND repressors is a novel aspect previously not appreciated. In spite of all these strengths, a potential weakness is that it remains open whether this mechanism is primarily a house-keeping mechanism or is indeed, as the authors speculate, regulated by internal and external factors that might, through CAMTA, make cells more or less responsive to Ca2+-CaM signaling.

We are grateful for and encouraged by our reviewer’s comments. We think that our discovery that CAMTAs regulate CaM expression is thought provoking.

Reviewer #2 (Public Review):

Vuong-Brender, Flynn, and de Bono report a detailed analysis of the function of a highly conserved calcium-calmodulin-dependent transcriptional regulator in the function of the C. elegans sensory nervous system. The C. elegans homolog of this factor - CAMT-1 - emerged from a genetic screen for mutants defective in a sensory-driven aggregation behavior. The authors find that multiple chemosensory modalities are disrupted by loss of CAMT-1, and this factor has distributed functions in the nervous system, including in interneurons that receive inputs from sensory neurons. A major finding of this study is that many of the effects of CAMT-1 mutation can be linked to a critical role for CAMT-1 in regulating expression of calmodulin itself. This finding is supported by multiple lines of experimentation, including a demonstration that the effects of losing CAMT-1 can be compensated by restoring expression of calmodulin. The authors further show that what is true for CAMT-1 and calmodulin in C. elegans also applies to Drosophila, indicating that CAMT-1 is a regulator of calmodulin expression whose function has been conserved throughout evolution. This manuscript has many strengths. Key hypotheses are tested using quantitative and technically independent experimental methods. The case that CAMT-1 is a regulator of calmodulin expression is built carefully and, for the most part, the logic of the argument is made clearly and supported by compelling data. Another strength of the manuscript is its candid exposition of data that do not fit neatly into the most simple and accessible model. It is refreshing to see authors who freely admit that they haven't neatly wrapped up every question in a field. The loose ends in this study do not impact the authors' main conclusions. However, some observations seem to consume more bandwidth than warranted, and the authors should consider reorganizing the manuscript so that the loose ends do not distract from the main thread of the narrative. The paper does have a few minor weaknesses that could be addressed. These are listed below.

We thank our reviewer for their thoughtful review.

Specific comments:

1. The initial description of the isolation of camt-1 mutants seemed a bit disorganized. A description of the gene and gene product preceded descriptions of the mutants. Also, some mutants were mentioned in the text but not presented in the corresponding figure. The authors should consider minor changes to better communicate how the mutations were cloned.

We have sought to do this.

1. In Fig. 2 npr-1 baselines vary a great deal between panels A, B, and C. It is not clear why npr-1 behavior is this variable, and the authors do not mention this obvious feature of their data. Data presented in Fig. 2 indicate that heat-shock-induced expression of camt-1 restores a defect in basal locomotion, but it is unclear whether it restores O2-sensitivity - the effect of oxygen on speed of transgenics seems the same +/- heatshock (compare black traces in panels 2B and 2C). We understand the concern of the reviewer. Since the design of these experiments was different from the rest (with only one shift in O2 concentration), we repeated them with 3 O2 changes, bringing them in line with the rest of the manuscript. The results are presented in the new Figure 2. We observed a more consistent baseline speed between different conditions, however some differences still exist (for example between panel 2A and 2B). One explanation is that for heatshock experiments we keep npr-1 animals at lower temperature (20 degree Celsius, panels 2B and 2C) to minimize basal activity of the heatshock promoter, whereas in the rescue experiment in Figure 2A, and in the rest of the manuscript, animals were kept at 22 oC. Figure 2B-C of our original submission used worms raised at 15 oC for the heatshock experiment, which may explain the greater discrepancy in npr-1 speed values. Heatshock also modifies slightly the response of the npr-1 control animals to O2.

Regarding whether heat-shock-induced expression of camt-1 restores O2 responses, we found that the npr-1; camt-1; dbExhsp-16p::camt-1 heat-shocked strains aggregated much more than npr-1; camt-1 heat-shocked animals. However, the rescue is not complete. Thus expressing camt-1 using heatshock-induced expression restores some O2 sensitivity which correlates well with the partial rescue of the baseline in Figure 2C. We have noted this in the results.

1. Unlike other datasets, the responses of wild-type AFDs to CO2 do not look particularly convincing (panel 3C). There is clearly an effect of camt-1 mutation on AFD calcium, but the AFD responses seem qualitatively different from the responses of BAGs to CO2 or URXs to O2. The authors might consider moving these data to a supplementary figure and tempering their description of wild-type AFDs as CO2-sensors.

The data on AFD has been moved to Figure 3 – figure supplement 1. We should add that we agree that in the absence of an identified CO2 sensor expressed in AFD, we cannot be sure that AFD neurons are primary CO2 sensors. Although the AFD CO2-evoked responses are retained in mutants defective in synaptic transmission, they may very well still be indirectly evoked by other neurons.

1. The authors candidly present data that do not conform to a simple model for how camt-1 affects behavior. Loss of camt-1 increases calcium in sensory neurons that activate the speed-controlling interneuron RMG. However, RMG calcium is reduced in camt-1 mutants. This inversion in the effect of camt-1 mutation might be caused by a homeostatic mechanism, as the authors propose. It might be possible to test this hypothesis by testing whether reducing excitatory input into RMGs elevates resting calcium in camt-1 mutants, for example via mutations that affect sensory transduction.

In the interest of simplifying the manuscript, and given other comments, we have now removed the RMG Ca2+ imaging data. However, this is an interesting way of testing what is going.

1. In Fig. 4H RMG data are presented as fractional ratio change - all other imaging data are presented as absolute ratios of YFP and CFP fluorescence. It is not clear why these data are treated differently. It is also no clear that these data are consistent with data shown in Fig. 3F. Which dataset represents the effect of camt-1 mutation on RMG calcium? More measurements might be warranted.

As highlighted above we have removed the RMG imaging data from the paper. .

1. Nice experiments show that regulation of calmodulin in Drosophila requires a CAMT-1 homolog. The bar graphs showing unity for values normalized to themselves are a bit odd - perhaps there's a more compact way to plot these data.

We have sought to address this question in two ways. First, we have further buttressed our results by performing in situ immunofluorescence staining of dissected fly retinas with a calmodulin antibody. We see a significant decrease in calmodulin expression in fly CAMTA mutants compared to controls.

Prompted by this comment, we also realized we omitted an explanation of how we normalized the data for the qPCR graphs in the figure legend. This was done using rRNA as a control. The Yamamoto lab had previously used the same control to normalize CAMTA expression in wild type and mutant flies. We add a note saying this.

1. ChIPseq analysis of CAMT-1 is also quite nice. Is there a sequence motif for CAMT-1 binding that emerges from this study? If so, how does this motif compare to motifs from studies of CAMT-1 homologs in other species?

We used the MEME algorithm, (motif-based sequence analysis tools (https://meme-suite.org/) to seek enriched sequence motifs in our ChIPSeq data. This identified a series of enriched motifs, although none coincided with the peaks at the CMD-1 promoter. However, we did observe sequences resembling the mouse CAMTA1 binding site at the centre of each of the three CAMT-1 binding peaks upstream of cmd-1. We now say this is the discussion.

1. Figure 7 shows that CMD-1 inhibits cmd-1 expression via interaction with CAMT-1. These data are interesting, but it is not clear how this effect can be related to prior data showing that forced expression of CMD-1 can compensate for loss of CAMT-1. The authors behavioral and physiological studies suggest that in vivo CAMT-1 promotes CMD-1 expression. In Figure 7, they suggest that CAMT-1 inhibits expression of CMD-1, but there is no clear link to behavior or physiology for this repressor-function of CAMT-1. The manuscript might be more clear without these data, and the absence of these data would not affect the overall impact of the study.

We agree that the feedback control of cmd-1 gene expression by CMD-1 interacting with CAMT-1 is a part of the story that has not been fully developed. Given the feedback from our reviewers and Editors to give these findings less prominence, but not remove them entirely, we moved the data into supplementary information. We have also altered the main text and the legend of Figure 7 to explicitly say that further experiments are needed to establish if this feedback is relevant under physiological conditions.

Reviewer #3 (Public Review):

Vuong-Brender et al present a thorough study investigating how CaM-binding transcription activators (CAMTAs) in C. elegans and Drosophila are required for numerous behaviors and proper neuronal function. The study is strong in how it uses a variety of approaches to study a major underlying mechanism for CAMTA. First, they use reporters, mutant analysis, and heat-shock rescue to show how cart-1 is expressed widely in neurons and functions in adults in several behaviors. They used transcriptional profiling to show that cart-1 is required to upregulate CaM in subsets of neurons in worm. They next use ChIP-seq to zero in on where worm CAMT-1 binds regulatory regions upstream of the CaM gene cmd-1 to promote its expression. They find that overexpression of CaM compensates for behavioral and neuronal response deficits in a cart-1 mutants. Lastly, they propose that when CaM highly expressed, it may down regulate its own expression by binding CART-1.

We thank our reviewer for their critique of our work.

1. Overall, I feel that the study is excellent and most conclusions are justified by evidence. However, I do not think the title is supported by the data. It currently is listed as: CAMTA TUNES NEURAL EXCITABILITY AND BEHAVIOR BY MODULATING CALMODULIN EXPRESSION. The authors show evidence that camt-1 is required for the normal function of neurons and behavior by promoting expression of CaM. Their only evidence that camt-1 downregulates CaM is a more artificial situation where CaM is overexpressed. I don't think they provide any evidence that camt-1 is used to "tune" behavior or neuron activity up and down in a wild-type strain. Tuning implies that the molecule modulates a physiological system bidirectionally in a natural situation. I suggest using a more accurate title that better fits the experimental evidence.

We have changed the title to ‘Neuronal Calmodulin levels are controlled by CAMTA transcription factors’. We hope this more neutral title is appropriate to describe our findings.

1. They show ample evidence that cart-1 appears to promote the expression of cmd-1 in most cases. This includes showing that overexpression of cmd-1 suppresses the behavioral and imaging phenotypes of cart-1. But they didn't perform the more straight forward epistasis test with the cart-1;cmd-1 double mutant in worm or fly , presumably because there is no viable loss-of-function allele in the coding area of the cmd-1 gene. It would help the readers understand why this simpler experiment was not performed if they explain this in the paper. A good place would be near line 220, where they generate hypomorphic promoter alleles using CRISPR. If they have tried to make their own loss-of-function alleles by mutating the coding area of cmd-1, but it resulted in presumed lethality, this might be mentioned here too.

This is a good point, and one that we had overlooked. cmd-1 loss of function mutations do indeed confer lethality. We have added a sentence to say:

‘Straightforward comparison of camt-1 and cmd-1 loss of function phenotypes was not possible, since disrupting cmd-1 confers lethality (7, 8).’

1. I am most worried about the potential caveats with the calcium imaging experiments. As the authors note, it is challenging to infer absolute levels of calcium using the ratiometric sensor cameleon across different individuals and genotypes. However, the authors do not note that the YFP/CFP FRET signal from cameleon might be perturbed because it uses calmodulin to bind calcium. At the end of their study (line 244), they provide evidence that calmodulin may bind to CART-1 to suppress its own expression when calmodulin is highly expressed. This is worrisome because cameleon is probably expressed highly in some or most of these strains. The authors may want to re-examine neuronal activity for a subset of experiments with a method that is independent of a calmodulin-based sensor (if possible).

We agree that this is a potential concern. As suggested by our referee, we therefore repeated some of our Ca2+ imaging experiments using a genetically-encoded Ca2+ indicator that does not contain CaM. We opted to use TN-XL, an indicator that uses troponin C as the Ca2+ binding moiety, and which has previously been used successfully in C. elegans. We imaged CO2-evoked Ca2+ responses in BAG sensory neurons, in wild type and in camt-1 mutant animals. The data obtained using TN-XL recapitulated what we observed using YC3.60 (BAG).

1. The title of "Fig 3 - Figure supplement 1" is confusing because it suggests that they measured the levels of YC2.60 cameleon, when in fact they measured a separate GFP reporter, albeit using the same promoter. So they could clarify the figure title.

The reviewer is right – our heading was confusing. We have changed it, and now say: ‘Expression from the gcy-37 promoter is reduced when CAMT-1 is overexpressed.’

1. D. Bazopoulou, A. R. Chaudhury, A. Pantazis, N. Chronis, An automated compound screening for anti-aging effects on the function of C. elegans sensory neurons. Sci Rep 7, 9403 (2017).
2. M. S. Choi et al., Isolation of a calmodulin-binding transcription factor from rice (Oryza sativa L.). J Biol Chem 280, 40820-40831 (2005).
3. J. Han et al., The fly CAMTA transcription factor potentiates deactivation of rhodopsin, a G protein-coupled light receptor. Cell 127, 847-858 (2006).
4. N. Bouche, A. Scharlat, W. Snedden, D. Bouchez, H. Fromm, A novel family of calmodulin-binding transcription activators in multicellular organisms. J Biol Chem 277, 21851-21861 (2002).
5. T. Yang, B. W. Poovaiah, A calmodulin-binding/CGCG box DNA-binding protein family involved in multiple signaling pathways in plants. J Biol Chem 277, 45049-45058 (2002).
6. E. Kodama-Namba et al., Cross-modulation of homeostatic responses to temperature, oxygen and carbon dioxide in C. elegans. PLoS Genet 9, e1004011 (2013).
7. V. Au et al., CRISPR/Cas9 Methodology for the Generation of Knockout Deletions in Caenorhabditis elegans. G3 (Bethesda) 9, 135-144 (2019).
8. A. Karabinos et al., Functional analysis of the single calmodulin gene in the nematode Caenorhabditis elegans by RNA interference and 4-D microscopy. Eur J Cell Biol 82, 557-563 (2003).

I think this is really interesting. As a cis woman I can recognise where acting straight and feminine in society can have its benefits, rather than deviating away from the "norm". Although we find in many aspects fitting into the traditional norm can be beneficial potentially to acquiring government funds, or being more desirable for a job position because you act in a certain way. Although these are not fair, it is seen often.

However, I would argue everyone is in some way queer. When alone, or with those they are comfortable with, may present traits that are not considered as desirable by society. Whether it be based gender roles throughout the household. For an example, my partner and father of my children, doing the nightly routine with our children, playing games such as dress ups and tea parties, reading books with the children and giving them affection and more on the emotional level, all whilst I work in the office. While we believe this is equal, would society suggest these activities are "queer" for a "Male" to be doing? Or are we at a place where these activities are more accepted? (I believe it is more accepted" but based on traditional nuclear families, are these activities queer?

This is just one example but could relate to many activities that go on behind closed doors, that the public eye does not see where people would generally act out of typical "straightness, feminine or masculine roles". I would argue people are more 'queer" in those situations. I hope this makes sense.

SKIP NAVIGATION

9+

0:01 / 4:26

#ThePrettyReckless #DeathByRockAndRoll

The Pretty Reckless - And So It Went [feat. Tom Morello] (Official Lyric Video)

638,635 views

Jan 11, 2021

20K277SHARESAVE

The Pretty Reckless

1.49M subscribers

SUBSCRIBED

The Pretty Reckless - And So It Went [feat. Tom Morello] (Official Lyric Video) From the upcoming album 'Death By Rock And Roll' | Available February 12, 2021: http://found.ee/dbrr Subscribe to The Pretty Reckless on YouTube: https://found.ee/tpr_subscribeyt Photo by: Rob Fenn Pre-order/Pre-Save the album 'Death By Rock And Roll': iTunes: http://found.ee/dbrr_it Apple Music: http://found.ee/dbrr_am Spotify: http://found.ee/dbrr_sp Amazon Music: http://found.ee/dbrr_amzm Amazon: http://found.ee/dbrr_amz YouTube Music: http://found.ee/dbrr_ytm Stay connected with The Pretty Reckless Website: https://deathbyrockandroll.com/ Facebook: https://found.ee/tpr_facebook Twitter: https://found.ee/tpr_twitter Instagram: https://found.ee/tpr_instagram LYRICS And so it went the children lost their minds Begging for forgiveness was such a waste of time And the bullets start to fly And the bough's about to break When you hear them cry It's too much for me to take The world does not belong to you You are not the king I am not the fool They said the world does not belong to you It don't belong to you It belongs to me And, so it went the children lost their minds Crawling over bodies of those who gave their lives And the fists begin to throw And the fire starts to blaze Don't you think they know They're the fucking human race The world does not belong to you You are not the king, I am not the fool They said the world does not belong to you It don't belong to you It belongs to Everyone is crying out, I can hear them scream With all these eyes upon us but no one seems to see That you and me are just the same as god meant it to be But you're much too close to me. You're much too close to me So it went The children lost their minds Nowhere to run, nowhere to hide And the wind begins to howl And the wolf is at your door You have so much of everything But still you wanted more They said the world does not belong to you You are not the king, I am not the fool They said the world does not belong to you It don't belong to you It belongs to me #ThePrettyReckless #DeathByRockAndRoll

SHOW MORE

The Pretty Reckless - On Tour

OCT

3

Upcoming show - Nashville, TN

Sun 7:00 PM - The Cowan

Ticketmaster

VIEW TICKETS 

Buy The Pretty Reckless merchandise

[

The Pretty Reckless - Death By Rock And Roll Coke Bottle Clear Vinyl

$28.00

Merchbar

](https://www.youtube.com/redirect?event=product_shelf&redir_token=QUFFLUhqbWpfbjNWZ1d0eXhXV0I1a2Y2UEd0NGwtc0V3UXxBQ3Jtc0trQ0hnNm85MktDcks1MW9IYTZseGpoRHJYNkRxSWdZZUxZclZkUXpzb09HbDJNa2xkT3dFUnJSWURqX2Q5VmE0U2E4RUh3T0dQcVppeEYwQ0Y4SVNQWmZ4VFVmd0Nzb3lDRjBjYmRwMFN2UkFWdWJycw&q=https%3A%2F%2Fwww.merchbar.com%2Fhard-rock-metal%2Fthe-pretty-reckless%2Fthe-pretty-reckless-death-by-rock-and-roll-coke-bottle-clear-vinyl%3Fucc%3DUS%26utm_term%3DUCrrhJmM3H7eyxrevc4lpRIw%26utm_medium%3Dproduct_shelf%26utm_source%3Dyoutube%26utm_content%3DYT-ACRcEUr70GaLvLpWdb-jwGYTdm5aFCJNFojnyaIECNbjjwKI8jU72BoLJ12L7jH6MU5EGFwxipDrzN0NpDpElIKGxFWXGREW6018kt6-ufDHVIUy2Mh8lkTGEpCRSHUyNw-hqfR1DDvQU4U0qp-C1ZG2UBSeoKHNMVEgpLUlvKnjcPhdwz0ieiBM9eTvKJp_moCQPETZl6zHKPQYB2rBbhPL8up4yIWMlrU%253D&v=0MpJv8DW4_U) [

The Pretty Reckless - Skull-Cycle Hoodie

$60.00

Merchbar

](https://www.youtube.com/redirect?event=product_shelf&redir_token=QUFFLUhqbUQ3TG80TzJWczM2bnMzZWRCcXh6U2RhX1VjZ3xBQ3Jtc0tsVk1KOUYtTjNBLTZpVkEyU2ZIUEx1YUhEdUlIVV9vX3k3ck1fQmlJVUpfODRXdTh5dWxJbW5aLTNFclhseEFuV1Y0b211R013bURvNmMyZ01LbE4ycjBINVptalJzX2t6Wkd4STJGVlJiUGZpTDNERQ&q=https%3A%2F%2Fwww.merchbar.com%2Fhard-rock-metal%2Fthe-pretty-reckless%2Fthe-pretty-reckless-skull-cycle-hoodie%3Fv%3D1273424%26ucc%3DUS%26utm_term%3DUCrrhJmM3H7eyxrevc4lpRIw%26utm_medium%3Dproduct_shelf%26utm_source%3Dyoutube%26utm_content%3DYT-ACRcEUrDJLp45ux0a5dMJ0016ICyrRgV97AmOfXg2ju4mSw4wkzsV1E3Z6uvmNk1H0HnluO_vHPfH4EfRup20jes5zjP1yM9QZxLVaz9ClEzm3IafKMMPGNO6Genrsqu7gtpNn-Hr0GhxbM1AevJ7BPiBg4Thg998W-ekyqCf447Azl5EpVMLnOSOSKYXt2fPVyxTBlVQLzkvVV7DmukIOPaw7JAU8e1odV0E7fPpnA%253D&v=0MpJv8DW4_U) [

The Pretty Reckless - Harley Long Sleeve

$45.00

Merchbar

](https://www.youtube.com/redirect?event=product_shelf&redir_token=QUFFLUhqa1dCUGtkZEtZdzhrSFZIWkRpZVo5XzZYdzBQd3xBQ3Jtc0tua3pFZ2RpNVl6aTVMdDYydWtjYmY3U1pEZmlFc3lRWS1NOHFkWHRUd2dLSjVmZGhoZEd4N0tiV2Z3VnUzMHNxa1JyeGVpREdIcV80RzFEb2F2NjJpREtNekFwYTVCdlJmU0tLVXIwUDVPX1RvYlBucw&q=https%3A%2F%2Fwww.merchbar.com%2Fhard-rock-metal%2Fthe-pretty-reckless%2Fthe-pretty-reckless-harley-long-sleeve%3Fv%3D1273416%26ucc%3DUS%26utm_term%3DUCrrhJmM3H7eyxrevc4lpRIw%26utm_medium%3Dproduct_shelf%26utm_source%3Dyoutube%26utm_content%3DYT-ACRcEUoAlSm71DWkwiqiKv3-Tshusvr18ByCwU2l03YABaFIh0wv3kBlNSXXDSabdZgekxDaVxQdTKeg131ld61IykjLodPHlRqnABsJCpYFv57fFbcxyvMYpNMo3NzrFi_Af3mvtDkOUeQRyqdKYkE36ezRxqkG9VygCNFfgiPJOoL8UonACBVY8WJO356BiAaP9zePitV0l4c0SKmhKrARs4TofQ4MLLO_ab1_lUY%253D&v=0MpJv8DW4_U) [

The Pretty Reckless - Death By Rock And Roll Black Vinyl

$28.00

Merchbar

](https://www.youtube.com/redirect?event=product_shelf&redir_token=QUFFLUhqbUF0b09QaFlGTDBrSEtpam1oTERvdDhkblpRd3xBQ3Jtc0tuSk5Oa182dENZdFRJOEg0Nk9UY0dtRl9vR0p1R2tqMHBfWjJxNTBjT2QxNmtwcFROYVBQbDlXLUNFejdoLWprT2NWSjJVOEx1eGFWUzdTcGlNSnBIZlFoNUJfUzFjTGVLMHB4Yi1WdHFNNW5TQU9GMA&q=https%3A%2F%2Fwww.merchbar.com%2Fhard-rock-metal%2Fthe-pretty-reckless%2Fthe-pretty-reckless-death-by-rock-and-roll-black-vinyl%3Fucc%3DUS%26utm_term%3DUCrrhJmM3H7eyxrevc4lpRIw%26utm_medium%3Dproduct_shelf%26utm_source%3Dyoutube%26utm_content%3DYT-ACRcEUovlrxz0nQ4InXCWvl-mbuMilggo1lYilIUQWvAfwCWeeyWwMh95m-WEJZ9hhbK8q4ovFBWFJQVs0mh-1RNGRFIqTO9DgGoBX6tEeliATAa6ycUeaP84L51uo1mKkWsLcdbpJ0kZAduNYCM8ULbuqx_q04uSvSqybI4LLGIh4Q7_QQGQtYEXtCRGbhmzq14vEiM3zIW-fuHnqIDHHaWcVuWpug0ox1ePRZf&v=0MpJv8DW4_U) [

The Pretty Reckless LIGHT ME UP CD

$14.84

Merchbar

](https://www.youtube.com/redirect?event=product_shelf&redir_token=QUFFLUhqa21WR2lJNGhWLXhnc21CTTZSTk5jRG5uODlFd3xBQ3Jtc0trc1FKZi1iZ0M5V3R0eVFOdVQ1ZnhkQ0s1VkVEXzR4S214RVNlRUVmM0FmVHFxdjZ5ZXpDRy1TWDhEN1ZaUGIxWkpHcWM5MXZ1ak9acHdNallvVVZhWm9zeHhHa0VmY1R5ejlyQkxPOGtDaHVQZkRVZw&q=https%3A%2F%2Fwww.merchbar.com%2Fhard-rock-metal%2Fthe-pretty-reckless%2Fpretty-reckless-light-me-up-cd%3Fucc%3DUS%26utm_term%3DUCrrhJmM3H7eyxrevc4lpRIw%26utm_medium%3Dproduct_shelf%26utm_source%3Dyoutube%26utm_content%3DYT-ACRcEUo19Va_nceiNgLL_XOuiRVxi_E4tLq3xIYvnCIgV8ui9sPYX0ayGIYd1Ez-UIz2Q3RIqr2kx206Y1dHxDLjRmjJQUuEzo-EfKlU8x1CjvUZE3FoIzxiRmGWHzMF2xiOhEasZphPFAH780jN21NmOkUfakQ6t4szQ0tVWepj1AxTfxv8E8hOSyNTUSEtkzTrxjv69LjSdKt_GzRwZmKsashY2_I%253D&v=0MpJv8DW4_U)

802 Comments

SORT BY

Add a public comment...

Adam Dobrin

1 second ago

VERBATIM: "stop hurting people, stop hurting people. if you do not believe people should be hurt" ... ((ishing)) ~either disable the attackers or leave the area~

1

REPLY

Adam Dobrin

1 second ago

sometimes words have two meanings. sometimes "kill" means "excorcized the demons" @LOUDON @SALEM

REPLY

CANCELREPLY

Raziel Moreno

7 months ago

I'm so damn happy she chose to rock instead of staying as an actress.

426

REPLY

View 3 replies

sonya fuks

7 months ago

How is it possible they just keep on getting better and better with each album?!?!

341

REPLY

View 12 replies

Graziano D'Ovidio

7 months ago

I love her voice. So much.

652

REPLY

View 18 replies

Natsumi Ikari

7 months ago

Most of people know Tom Morello from RATM of Prophets of Rage, but don't forget he was also the guitarist of Audioslave, whose singer was Chris Cornell. Must be emotional for Taylor to record this song him. I'm sure Chris is proud of them both.

631

REPLY

View 28 replies

Matstrr Salazar

7 months ago

This is why TPR is one of my all time favorite bands. They release bad-ass singles and albums

322

REPLY

View reply

Mike Whicker

7 months ago

The entire death by rock n roll album is first on my list of reasons why 2021 will not stink as much as last year did!

180

REPLY

View 5 replies

Mackenzie Dinkins

5 months ago

A good theme for this year's Elimination Chamber event.

25

REPLY

View 2 replies

FDCAnselmo

7 months ago

The Anthem of Youth... Fuck me what a tune.

159

REPLY

View 8 replies

norangutan

7 months ago

I love how obviously this song draws inspiration from Audioslave's sound, and in a way it's paying homage to chris cornell especially with the "you have so much of everything still you wanted more" line - which is almost the same as a line from Audioslave's 'What you are'

165

REPLY

View 7 replies

Shiteyanyo

7 months ago

Tom Morello?! Each single they release for this album just keeps getting better and better

55

REPLY

View reply

Lau pin

7 months ago

I fell in love with every song they released

90

REPLY

View 4 replies

🕸️Haley Munster🕷️

7 months ago

I really hope this album has more Going to Hell vibes 🥺🥺🥺🖤

201

REPLY

View 22 replies

M K

7 months ago

Great tune and a good example for why I think TPR is among the best of the current rock bands. They always have catchy riffs and melodies however, especially in the bridge, you never know which turn the song is gonna take. Yet it always fits the song and never sounds out of place.

15

REPLY

Mike B

7 months ago

Underrated band

154

REPLY

View 8 replies

🕸️Haley Munster🕷️

7 months ago

THIS is the Pretty Reckless I love !! 💙

45

REPLY

View reply

pedrogoularth

7 months ago

This is SO fucking good! 💖 🇧🇷

31

REPLY

Nick Brick

7 months ago

Kicking the year off with a banger. Looking forward to the album!

19

REPLY

Jo Mill Hyde

6 months ago

The fucking power in her voice and this song

49

REPLY

View reply

HT82 Smash

6 months ago

WWE main roster stepping up their game with these theme songs. Elimination Chamber 2021 everyone

17

REPLY

View 2 replies

Hide

7 months ago

The addition of the children's choir killed and buried me. 😔🤘🏾

80

REPLY

View 5 replies

Try Google Fi

Ad

 fi.google.com

SIGN UP

AllThe Pretty RecklessPop MusicRelatedWatched

[

1:16:03NOW PLAYING

](https://www.youtube.com/watch?v=3DMM87DMF5o)

[

S.O.A.D Greatest Hits 2021 - S.O.A.D Best Songs Playlist

HARD ROCK COLLECTION

552K views6 months ago

](https://www.youtube.com/watch?v=3DMM87DMF5o)

[

1:57:34NOW PLAYING

](https://www.youtube.com/watch?v=uetFO7y8WPA&t=96s)

[

Corey Taylor - Live in London (Full Show)

Corey Taylor

12M views3 years ago

](https://www.youtube.com/watch?v=uetFO7y8WPA&t=96s)

[

109NOW PLAYING

](https://www.youtube.com/watch?v=0J2QdDbelmY&list=RDCLAK5uy_l3PeyHeqJh1dR78WjfsMJwRHJx9ofMvvc&start_radio=1&rv=0MpJv8DW4_U)

[

Modern Rock Hits

YouTube Music

](https://www.youtube.com/watch?v=0J2QdDbelmY&list=RDCLAK5uy_l3PeyHeqJh1dR78WjfsMJwRHJx9ofMvvc&start_radio=1&rv=0MpJv8DW4_U)

[

1:47:37NOW PLAYING

](https://www.youtube.com/watch?v=26nsBfLXwSQ)

[

Best of 90s Rock - 90s Rock Music Hits - Greatest 90s Rock songs

Memory Music Boxx.

19M views2 years ago

](https://www.youtube.com/watch?v=26nsBfLXwSQ)

[

49:52NOW PLAYING

](https://www.youtube.com/watch?v=zZyWxbojYH8)

[

Fall Out Boy - The Young Blood Chronicles (Uncut Longform Video)

Fall Out Boy

19M views7 years ago

](https://www.youtube.com/watch?v=zZyWxbojYH8)

[

2:54:28NOW PLAYING

](https://www.youtube.com/watch?v=KLcqekMfO9E&t=5188s)

[

Dave Matthews Band - 07/23/2021 {Full Show | 4K} Walnut Creek Amphitheatre - Raleigh, NC

Andy Mendoza

42K views3 weeks ago

](https://www.youtube.com/watch?v=KLcqekMfO9E&t=5188s)

[

1:01:46NOW PLAYING

](https://www.youtube.com/watch?v=MRQt5h-IDCg)

[

Sheryl Crow - Outlaw Music Festival - Live in Milwaukee, WI (Summerfest 2017)

Crow Archives

1M views4 years ago

](https://www.youtube.com/watch?v=MRQt5h-IDCg)

[

54:10NOW PLAYING

](https://www.youtube.com/watch?v=ofil7qB842o&t=1106s)

[

Evanescence - Acoustic Songs

Evanescence vds

2.7M views5 years ago

](https://www.youtube.com/watch?v=ofil7qB842o&t=1106s)

[

1:16:16NOW PLAYING

](https://www.youtube.com/watch?v=u7zuAlf5FBo)

[

Halestorm - Live from Download UK 2019

Halestorm

685K views1 year ago

](https://www.youtube.com/watch?v=u7zuAlf5FBo)

[

1:18:34NOW PLAYING

](https://www.youtube.com/watch?v=zYON-KnG3rM)

[

Shinedown - The Sound Of Madness (Deluxe Edition) (Full Album)

The Rock Cafe

212K views4 months ago

](https://www.youtube.com/watch?v=zYON-KnG3rM)

[

45:56NOW PLAYING

](https://www.youtube.com/watch?v=X3-Bw31fI2s)

[

Linkin Park - Hybrid Theory [Full Album] 2000

quake

5M views9 months ago

](https://www.youtube.com/watch?v=X3-Bw31fI2s)

[

46:50NOW PLAYING

](https://www.youtube.com/watch?v=GUa_AJa7uOc)

[

Matchbox Twenty - Yourself or Someone Like You (Full Album)

Gal Godonut

1.4M views1 year ago

](https://www.youtube.com/watch?v=GUa_AJa7uOc)

[

2:24:37NOW PLAYING

](https://www.youtube.com/watch?v=v5Z1LyfgAuo&t=7102s)

[

E̲v̲a̲n̲escence - The Bitter Truth (Deluxe) [Full Album]

Santiago Lee

222K views4 months ago

](https://www.youtube.com/watch?v=v5Z1LyfgAuo&t=7102s)

[

1:22:20NOW PLAYING

](https://www.youtube.com/watch?v=bf4tLUMxrDc)

[

E V A N E S C E N C E Greatest Hits Full Album - Best Songs Of E V A N E S C E N C E Playlist 2021

Parker Michael

624K views5 months ago

](https://www.youtube.com/watch?v=bf4tLUMxrDc)

[

57:12NOW PLAYING

](https://www.youtube.com/watch?v=TgJnsOy_EeM&t=452s)

[

The Pretty Reckless best lives compilation

Marianne Audouin

260K views8 months ago

](https://www.youtube.com/watch?v=TgJnsOy_EeM&t=452s)

[

1:06:00NOW PLAYING

](https://www.youtube.com/watch?v=nWjN8SA5KEU)

[

Chevelle - Stray Arrows: A Collection of Favorites (Full Album)

Ita Depeeza

345K views1 year ago

](https://www.youtube.com/watch?v=nWjN8SA5KEU)

[

12:02NOW PLAYING

](https://www.youtube.com/watch?v=76lIGyw7zL8)

[

The Pretty Reckless FULL acoustic session in Paris

Marianne Audouin

705K views3 years ago

](https://www.youtube.com/watch?v=76lIGyw7zL8)

[

53:38NOW PLAYING

](https://www.youtube.com/watch?v=OYMLjqoYhdw)

[

Guardians of the Galaxy Awesome Mix Vol 1 Vol 2

Alternitive Videos

1.8M views1 year ago

](https://www.youtube.com/watch?v=OYMLjqoYhdw)

[

2000's Rock Songs Mix 🎸 Best Rock Hits of the 2000's Playlist

Redlist - Rock Mixes

913K views1 month ago

](https://www.youtube.com/watch?v=_3A4JW9RQWM)

[

Motley Crue greatest hits full songs \m/

soulsickk\m/

2M views3 years ago

](https://www.youtube.com/watch?v=3-FQCgVZG5o)

Lynne Kelly is correct here that we need better delineations of the words we're using here.

To some of us, we're taking historical methods and expanding them into larger super sets based on our personal experiences. I've read enough of Kelly's work and her personal experiences on her website (and that of many others) that I better understand the shorthand she uses when she describes pieces.

Even in the literature throughout the middle ages and the Renaissance we see this same sort of picking and choosing of methods in descriptions of various texts. Some will choose to focus on one or two keys, which seemed to work for them, but they'd leave out the others which means that subsequent generations would miss out on the lost bits and pieces.

Having a larger superset of methods to choose from as well as encouraging further explorations is certainly desired.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Reviewer #1

This manuscript by Gulyrutlu and co-workers addresses the role of CUG expanded repeat RNA associated with DM1 in regulating the formation of higher order RNP assemblies such as stress granules and P-bodies in the cell. The authors used lens epithelial cells (hLECs) derived from a DM1 patient

We used cell lines from several patients and age-matched controls to avoid effects of individual cell-line variation. We will make sure that this is clear in the text.

or a HeLa cell inducible model of DM1 to investigate whether expression of the CUG repeat-associated protein MBNL1 and CUGBP1 affected the formation and dispersal of stress granules and P-bodies. The authors show that MBNL1 and CUGBP1 are components of SGs and PBs in hLECs and HeLa cells. In cells expressing the CUG repeat, there are minor alterations in the dispersal of stress granules as well as in the formation of P-bodies.

The alterations in the formation and dispersal of stress granules are not minor. For example, in the HeLa cell model, stress granules take more than twice as long to form in cell expressing the CUGexp repeats associated with DM1 and disperse in half the time. These data are already in the results section, but we will highlight them in a revision and have included an additional representation of the data to the figure, using graphs of ‘proportion of cells with stress granules’ against time. The changes we see are as large, or larger, then results published elsewhere (see appendix below)

MBNL1 could affect the formation and dispersal of SGs independent of the CUG repeat.

In fact, we present data in HeLa cells with MBNL1 almost completely removed by shRNA revealing that this has a much smaller effect on stress granules than does the expression of CUGexp RNA. This is an important point, as it is widely assumed that most of the cellular defects in DM1 are caused by the ‘sequestration’ of MBNL1 in the CUGexp foci. Since only . This is not what our data show. In the hexanediol experiments, both cell lines over-express MBNL1 in similar amounts. The difference between them is that one cell line expresses a DMPK1 mini-gene with a CUG expansion and the other expresses a mini-gene without the expansion. Again, our results show that the alteration to P-body responses to 1,6-hexanediol can be attributed to the presence of the CUGexp RNA, rather than altered levels of MBNL1. We will revise the results and discussion to further emphasise this point.

Finally, in HeLa cells, overexpression of MBNL1 can reduce the dispersal of P-bodies upon 1,6-hexanediol treatment.

This is not what our data show. In the hexanediol experiments, both cell lines over-express MBNL1 in similar amounts. The difference between them is that one cell line expresses a DMPK1 mini-gene with a CUG expansion and the other expresses a mini-gene without the expansion. Again, our results show that the alteration to P-body responses to 1,6-hexanediol can be attributed to the presence of the CUGexp RNA, rather than altered levels of MBNL1. We will revise the results and discussion to further emphasise this point.

Major comments:

One limitation of the work is that the perturbations seen with stress granules or P-bodies are all relatively small, and no evidence for a functional consequence on gene expression is demonstrated. Specifically, the authors observe only minor alterations in the formation or disaggregation of PBs and SGs in these DM1 models. Further, some of the effects observed are independent of the CUG repeat expression, suggesting that MBNL1 and CUGBP1 might have independent roles in modulating some properties of SG and PB formation or dispersal.

As above, the changes we see in SG formation and dispersal are not small. There are already numerous studies of the effects of DM1 on gene expression and mRNA splicing. This is not what we set out to study: we are interested in perturbations to the organisation of cellular structures associated with the expression of the CUGexp repeat RNA characteristic of DM1. We do show some data relating specifically to the proteins MBNL1 and CUGBP1 in the paper. shRNA resulting in almost complete loss of these proteins has much smaller effects that the expression of CUGexp RNA, suggesting that the major part of the effects caused by expression of the CUGexp RNA is not mediated through changes in MBNL1 or CUGBP1 levels. MBNL1 and CUGBP1 levels may well contribute to alterations in SG dynamics, but our data suggest that they are minor contributors. This is an advance in our current knowledge

1. The authors could investigate whether the CUG repeat RNA itself is localized to SGs or PBs in their models, and whether the presence of the repeat RNA is absolutely necessary for regulating the dynamics of SG or PB formation.

We have now done this. The CUG repeat RNA is not localised in stress granules or PBs to a detectable extent. This suggests that the effect we see on these structures by expression of the expanded RNA occurs despite the absence of the RNA from the structures. This is similar to the effects of the ALS-associated paraspeckle protein FUS, which can affect the integrity of nuclear LLPS structures (gems) despite not co-localising with them https://doi.org/10.1016/j.celrep.2012.08.025 We have added these data to the manuscript, as part of draft figure 8, and will add text emphasising this as an additional example of disease-causing macromolecules affecting the structure of LLPS domains in which they are not found.

1. The authors use 1,6-hexanediol to suggests that PBs and SGs in HeLa cells show behavior analogous to LLPS. However, the use of 1,6,-hexanediol to establish an assembly as a LLPS is a relatively limited analysis (despite its widespread use in the field), since this compound can affect the formation of multiple cellular substructures that are not always LLPS (for example, see Wheeler et al, 2016, eLife).

We are aware of and have cited this publication. Our comments about LLPS structures are measured, as there is still controversy about how to definitively identify them in cells. SGs and PBs have, however, previously been widely published to be formed by LLPS. The rapid exchange of SG and PB components during FRAP and the ability of SGs to both fuse and bud (seen in our supplementary movies) are also supportive of these structures behaving as LLPS structures in our models. Wheeler et showed that, in yeast, the nuclear pore complex and some cytoskeletal structures were affected by 1,6-hexanediol but membrane-bound structures such as the ER and mitochondria were not. The disruption of the nuclear pore complex is not unexpected, since phase separation is involved in cargo shuttling through the NPC (reviewed in https://doi.org/10.1016/j.devcel.2020.06.033). We will revise our discussion to make it more clear that we are not relying only on the use of 1,6-hexanediol to define SGs and PBs as LLPS structures but also on other aspects of their dynamic behaviour and on extensive prior literature.

Significance

This study would be of interest to the field if the impact of the DM! repeat RNAs on PB and SG were more substantial...

As above, the effects we see on SG formation and loss are substantial. Tissue types affected in DM1 are prone to stress, particularly the lens of the eye, so alterations to cellular response to stress associated with the presence of CUG repeats are of key importance to understanding the cellular pathology of DM1.

...and if some functional consequences were demonstrated.

In terms of function, we show altered responses to stress caused by the expression of CUGexp RNA and probably mediated through alterations in the propensity of LLPS cytoplasmic structures (SGs and PBs) to form and be resolved. Additionally, we can now show that SGs in HeLa cells expressing CUGexp RNA contain less total polyA RNA than is seen in controls, and that ‘docking’ events between SGs and PBs are compromised in cells with CUGexp RNA. These docking events are proposed to mediate transfer of RNA from SGs to PBs (reviewed in https://doi.org/10.1007/978-1-4614-5107-5_12). These new data demonstrate functional impairment of SGs and PBs associated with DM1. We have included this as an additional draft figure 8.

The lack of a strong effect on SG or PB formation in the DM1 models, along with the CUG repeat-independent effect of MBNL1 on the formation and dispersal of these complexes, argues that MBNL1/CUGBP1 may not significantly affect the formation or dispersal of SGs and PBs.

We are actually not arguing that MBNL1 and CUGBP1 are the main effectors in the changes we see to SGs and PBs, but that the CUGexp RNA is the key player, so are a little confused by this comment.

Reviewer #2:

In the current study, the authors compared the dynamics of P-bodies (PBs) and stress granules (SGs) between control and several DM1 cell lines. They found that MBNL1 and CUGBP1, two CUG repeat RNA-binding proteins that are primarily nuclear, could also co-localize with PBs in the cytoplasm and re-localize to SGs under stress. Small differences were observed in SG assembly and disassembly dynamics between control and DM1 HLECs, between HeLa cells expressing either CTG12 or CTG960, and between HeLa cells with and without shRNAs targeting CUGBP1 or MBNL1.

As detailed above, the alterations in SG assembly and disassembly in cells expressing CUGexp RNA are not small, in contrast to those in cells will lowered expression of MBNL1 and CUGBP1, which are much smaller suggesting that the changes caused by CUGexp RNA largely do not result from loss of MBNL1 (or CUGBP1). We have inserted additional graphs of ‘proportion of cells with stress granules’ against time' and will modify the text to emphasise both of these points.

Overall, the experiments were clearly described and the results properly presented. However, critical controls, as detailed below, are missing in multiple analyses. The mechanisms underlying these apparent differences are also unknown.

We do not consider that any ‘critical controls’ are missing, but can supply all of the additional analysis of our data that the reviewer requests below. We can also now provide additional mechanistic insight and will add an additional figure showing lowered amount of polyA RNA in stress granules in cells expressing CUGexp RNA and compromised docking events between stress granules and P-bodies, suggesting impaired communication between them.

Major concerns:

1. Throughout the study, the authors compared MBNL1 and CUGBP1 association with PBs and SGs without considering the potential differences in their cytoplasmic abundance between control and DM1 cell lines, which seems to be case for MBNL1 abundance in CTG960-expressing HeLa cells (Fig. 3). Provided that PBs and SGs exchange components with the cytosol at an equilibrium, if the cytoplasmic abundance of, for example, MBNL1 is decreased in DM1, one would expect the equilibrium being shifted resulting in less MBNL1 associated with PB/SG. Therefore, before measuring the association or the assembly/disassembly kinetics of PB and SG, the authors should first test whether MBNL1 and CUGBP1 abundance may be different between control and DM cell lines.

There is, in fact, no difference in the relative cytoplasmic abundance of GFP-MBNL1 between CTG12 and CTG960- expressing HeLa cells. Each has approximately a 50/50 split between nucleus and cytoplasm, with <3% of nuclear GFP-MBNL1 found in nuclear CUGexp foci when they are present. We have added a graph demonstrating this to the supplementary data. The abundance of total endogenous MBNL1 is also not altered in DM1 patient-derived lens cell lines compared to controls, as shown by semi-quantitative western blot analysis, which we have also added to the supplementary data. However, if the expression of CUGexp RNA did cause a major loss of cytoplasmic MBNL1, this change would be reflective of the situation seen in DM1 and would not invalidate our results or conclusions.

The same caveat applies to MBNL1/CUGBP1 knockdown experiments, where knocking down one may change the abundance of the other.

To carry out FRAP experiments or live cell analysis of SG formation and loss, it is necessary to over-express a tagged version of the protein being studied. For the knockdown experiments shown in figure 6, therefore, when we knocked down MBNL1, CUGBP1 was present in excess as a GFP-tagged protein and when we knocked down CUGBP1, MBNL1 was present in excess as a GFP-tagged protein. Thus, any effects of the knockdowns on expression of the endogenous proteins being analysed would be highly unlikely to influence the results.

1. Similarly, the authors did not consider the possibility that changes in SG/PB dynamics may be due to changes in the abundance/availability of essential SG/PB components such as GE1 and G3BP1.

From our immunofluorescence experiments, there was certainly no obvious reduction in GE1 or TIA1 abundance (we did not assess G3BP1). We have quantitative proteomic analysis (unpublished) from a similar pair of cell lines, expressing CUGexp RNA alongside GFP rather than GFP-MBNL1. This shows no change in GE1 or G3BP1, so we would not expect to see any here either. We can easily carry out a quantitative western blot analysis to confirm and will add this to the supplementary data

1. Most of the observed differences between control and DM cell lines were modest, leaving one wonder whether it could be simply due to cell line-to-cell line variability. Whenever possible, the authors should present results for each individual lines. For example, in Fig.2, 3 DM1 lines and 2 control lines were used. Was the difference in SG disassembly (Fig. 2B) observed in each of the 3 lines?

Some of the alterations were modest and there is cell line-to-cell line variability in the lens cell lines. This is why we pooled the data: on average, DM1 cells disperse their SGs more quickly than control cell lines do on average. This is not an unusual way to present data from patient cell lines of diverse genetic background. We have added data for stress granule loss in the individual cell lines to the supplementary data. These data show a consistent trend towards quicker dispersal of stress granules in patient cell lines. The variability between the patient lens cell lines was also the primary reason for us to develop the inducible system in HeLa cells, on a fixed genetic background, as explained in the manuscript.

Minor points:

1. Western blot in Fig. 3 shows two protein products from both endogenous and overexpressed MBNL1. Please explain.

Many of the commercially available anti-MBNL1 antibodies show this double-band in some cell lines as evidenced in numerous publications and on manufacturers’ websites (for example https://abclonal.com/catalog-antibodies/MBNL1RabbitmAb/A5149, https://www.ptglab.com/products/MBNL1-Antibody-66837-1-Ig.htm). We haven’t analysed the two bands in detail, but assume this to be a result of a post translational modification of some sort. Since GFP-MBNL1 and endogenous MBNL1 show the same thing, we do not consider it to be a major concern. We do mention the double-band as ‘characteristic’ in the figure legend for figure 3 so are not seeking to conceal anything here.

1. No data were shown to substantiate the statement that "MBNL1 localises to CUGexp foci and CUGBP1 does not" (page 6).

This has been published many times and is shown in figure 1A. However, we will add in a citation for this and have added an additional supplementary figure showing the lack of co-localisation in the foci from figure 1A more clearly together with separate data confirming that MBNL1 and CUGexp RNA do not co-localise with CUGBP1 in the nuclei of line HeLa_CTG960_GFPMBNL1.

1. The y-axis of Fig. 4D should not go beyond 1.

We will trim the axis. There are no data points above 1.0, just the indicator of statistical significance

Significance:

The nature of the current study is highly descriptive with little mechanistic insights.

Our work is not descriptive, as we observed a change in stress granules in patient cells, which we could then replicate (and enhance) in a novel inducible model of DM1 designed to abrogate the unavoidable variation in patient-derived cell lines. We also now have additional mechanistic insights (see above) and have added an additional figure (draft figure 8) detailing these.

For the subtle differences observed between control and DM1 cell lines, it remains unclear whether it may be due to cell line-to-cell line variation (see above).

We cannot completely rule out an influence of cell line-to-cell line variation in the patient-derived lens cell lines (see above), though we think this unlikely as we saw the same effect repeated and amplified in the inducible HeLa-derived cell model, which was designed to minimise this concern. Furthermore, for stress granule loss, we see a larger effect in the HeLa cell model after 72hrs of induction than after 24hrs (figure 5C). This argues strongly that the effects seen are due to the expression of CUGexp RNA and we will emphasise this point more strongly.

Some difference appear to be specific to one model but not the others (e.g., SG formation is slower in HeLa-CTG960 cells but not in DM1 HLECs).

Even for the observations that seem consistent between models, the current results yielded little novel biological insights into whether and how these subtle differences in PB/SG dynamics may relate to DM1 pathogenesis. Collectively, these weaknesses render the current study incremental at best.

The key biological insight the results provide is that the presence of the CUGexp repeat RNA results in defects in LLPS structures that are largely separable from any sequestration of MBNL1 in nuclear foci. With many researchers attributing the cellular defects in DM1 simply to the loss of MBNL1 by sequestration into nuclear foci, both this separation of altered stress response from MBNL1 levels and the involvement of altered LLPS formation (evidenced by the changes in PB behaviour on 1,6-hexanediol treatment) are novel biological insights into the cellular pathology of DM1. Additionally, our results shift the emphasis from nuclear effects to those seen in the cytoplasm.

In terms of specific DM1 pathogenesis, the eye lens is subject to constant repeated stress and is subject to continued growth throughout the life span, relying on lens epithelial cells as a stem cell pool. Epithelial cells are also vital to the homeostatic regulation of ions, growth factor and nutrient flow from the aqueous humor to the underlying fibre cells. Any alterations in the response of lens epithelial cells, in particular, to stress is highly relevant to the pathology of cataract seen in DM1. We will revise our discussion to emphasise these key points more strongly.

Reviewer #3

The manuscript entiled "Phase-separated stress granules and processing bodies are compromised in Myotonic Dystrophy Type 1" by Gulyurtlu et al., characterizes the composition and ydnamics of stress granules and P-bodies in two Myotonic Dystrophy Type 1 (DM1) cell models, human lens epithelial cells from DM1 patients and age-matched controls and HeLa_CTG12_GFPMBNL1 and HeLa_CTG960GFPMBNL1 cell lines. The manuscript is somewhat descriptive with lack of functional data and some discrepancies. For example, in the discussion section, the authors conclude that "MBNL1 appears to be absent from P-bodies in cells with CUGexp foci in their nuclei. This observation suggests that the role of MBNL1 in P-bodies may be disrupted by the presence of CUGexp RNA." Figure 4A shows that "P-bodies in the DM1 model line, HeLa_CTG960GFPMBNL1 do not contain detectable amounts of GFPMBNL1". However, Figure 4E shows similar levels of total cellular MBNL1 per PB between the control CTG12 and mutated CTG960 lines.

There is no discrepancy here. The reviewer has misinterpreted our data. PBs in the HeLa CTG960 cell line do not contain detectable amounts of GFP-MBNL1 under normal growth conditions, as shown in figure 4A. The data shown in figure 4E concern arsenite-treated cells, where some PBs in the CTG960 line do contain detectable levels of GFP-MBNL1, but significantly less than in the control CTG12 cells. We will reword these sections to make sure this is clear.

Most importantly, in Figure S3 the authors show that CUGexp foci are present in 1-2 % of the cells. The claim appears to be too strong for the data presented in the manuscript.

This is not what is shown in figure S3. The reviewer has misinterpreted our data. Figure S3 shows that in cells from line CTG960, only 1-2% of the total nuclear GFP-MBNL1 signal is found in the CUGexp foci, despite the intensity of the signal within them. Virtually all of the cells from the CTG960 cell line contain CUGexp foci (>95%). We will add a statement to this effect into the results section. We would not have continued working with a cell line in which only 1-2% of cells showed the DM1 phenotype of nuclear CUGexp RNA foci.

Although the findings are interesting and of potential impact for a better understanding of the implications of RNA-protein condensate dynamics in the pathogenesis of DM1, the work presented here is still descriptive and preliminary in my opinion. In summary, the conclusions are not so convincing and additional experiments are essential to support the authors claims.

This reviewer seems to have misinterpreted several of our data sets, including the specific points above, leading to the assertion that our conclusions are not convincing.

Several months of works will be required to consolidate data and reorganize and ameliorate the manuscript, including the way data are presented and quantified.

We already have data with which to address the majority of the queries posed, so should be able to make the adjustments relatively quickly.

Specific comments:

"On removal of stress, clearance of stress granules is mediated largely by a form of autophagy." This statement is not correct since the majority of stress granules disassemble and are not targeted to autophagy; in healthy cells only 5 % (or less) of the total SGs tend to persist in presence of autophagy or lysosome inhibitors, while the vast majority disassembles. Please rephrase carefully.

The degree and manner of dispersal of stress granules in healthy cells on removal of stress is not well understood, but is known to differ depending on the type and duration of the stress (DOI: 10.1126/science.abj2400). We do not yet know how this may be altered in DM1, however, compromised autophagy is implicated in cataract formation, which is of relevance to our study. We will re-phrase this section of the discussion carefully to reflect the complex situation.

Figure 1: RNA-protein complexes have heterogeneous composition. In HLECs, do all PBs colocalize with MBNL1 and CUGBP1 or only a fraction of them?

We do not routinely see PBs without MBNL1 or CUGBP1 in the HLECs, in contrast to the situation in the HeLa CTG960 line. We have data available in order to quantify this and will add the numbers to the text of the results.

Figure 2: Stress granules and P-Bodies are known to touch each-other, a process referred to as a "kissing event". The authors have studied the mobility of GFP-MBNL1 inside these two types of assemblies. It would be important also to quantify the "kissing" events. Is this altered in DM1 cells?

We couldn’t find reference in the literature to ‘kissing events’ between SGs and PBs, but found several references to ‘docking’ events. We have noticed such interactions between PBs and SGs in our models. We are currently quantifying this and our first experiments (one in the HeLa cell model and one comparing one of the patient-derived lens cell lines to a control) suggest that there is a change in the frequency and/or size of such interactions in the HeLa CTG960 cell line compared to the CTG12 control and in the DM1 lens cell line derived to control. If this holds true in our repeat experiments (currently in progress), this would also provide the mechanistic insight requested by reviewer 2. We have included this, together with our data showing a decrease in total polyA RNA in stress granules in HeLa cells expressing CUGexp RNA, as an additional draft figure 8.

Figure 3: In HeLa cells overexpressing CTG960_GFPMBNL1, beside the accumulation of one bright CUGexp puncta, several intranuclear GFPMBNL1 protein foci are visible. This subcellular distribution is different from the one observed in the control HeLaCTG12 GFPMBNL1. Can the author describe what these intranuclear GFPMBNL1 protein foci are?

In most cells expressing CUGexp RNA, several nuclear foci form (usually one large and several smaller) and all of them contain MBNL1 (or GFP-MBNL1 in the HeLa_CTG960_GFPMBNL1 cell line). Figure S3 shows object identification using MBNL1 in this cell line, with two clear foci detected as the reviewer points out. We have added an additional panel to supplementary figure 1 to confirm that the additional foci are also CUGexp RNA foci and will clarify in the text of the results that there is not a single focus of CUGexp RNA in each nucleus.

Is GFPMBNL1 accumulating at the level of splicing speckles? Or paraspeckles? Or other types on intranuclear condensates such as e.g. PML nuclear bodies? The different intranuclear distribution of GFPMBNL1 should be better characterized.

The sub-nuclear distribution of MBNL1 is, indeed, very complex. MBNL1 also sometimes co-localises to splicing speckles/interchromatin granule clusters as we have previously reported in lens epithelial cell lines (DOI: 10.1042/BJ20130870 ) . The details of differences in the nuclear distribution of MBNL1, beyond its accumulation in CUGexp RNA foci, in DM1 cells compared to controls is the subject of another manuscript we have in preparation but are beyond the scope of the current study.

Moreover, the % of cells expressing CTG960_GFPMBNL1 and forming intranuclear CUGexp foci is only mentioned in the discussion (Figure S3); for clarity it should be reported in the main text when describing Figure 3.

The number of cells forming nuclear CUGexp foci on expression of CTG960_GFP-MBNL1 is >95% and we will add this to the text of the results section.

"Figure S2: Quantitation of GFPMBNL1 in P-bodies in HeLa cell model of DM1." The authors report in the legend "Some, but not all, of these P-bodies contain detectable amounts of GFPMBNL1". However, the figure only shows a representative image of cells without quantification. Quantitation should be provided.

We have data available to provide this simple quantitation. Approximately 38% of PBs in arsenite-treated cells from line HeLa_CTG960_GFPMBNL1 contain detectable levels of GFPMBNL1 using a manually-assigned cut-off intensity. We will add this to the relevant figure legend (now figure S5). However, this method of analysis requires an intensity to be manually set above which GFP-MBNL1 signal is considered ‘detectable’. This is hugely subjective, and in our opinion, the automatically generated quantitative comparison of “% total cellular MBNL1 per P-body” as shown in figure 4E is a more experimentally robust way to demonstrate a small loss of MBNL1 from P-bodies in cells from line HeLa_CTG960_GFPMBNL1 treated with arsenite compared to the relevant control.

The authors report "a subtle change in stress granule architecture associated with the presence of CUGexp RNA". This statement is not supported by experimental data and should be omitted.

We will qualify this statement to make it clear that we are referring to a subtle alteration in the co-localisation between CUGBP1 and MBNL1 specifically in the SGs, as our experimental data shown in figure 4D clearly support that, showing a statistically significant increase in the Pearson’s co-efficient of cololcalisation between MBNL1 and CUGBP1 in cell containing CUGexp RNA compared to the relevant control (0.90+/-0.05 for CTG960; 0.87+/-0.07 for CTG12).

Figure 4. MBNL1 and CUGBP1 co-localise in P-bodies. What is the % of colocalization?

We’re not sure exactly what is being requested here or what biological question the reviewer is asking us to address. MBNL1 and CUGBP1 co-localise in virtually all PBs (except in the HeLa CTG960 line where MBNL1 is undetectable in PBs under normal growth conditions). Figure 4E shows that, in cells with PBs upregulated by sodium arsenite, the mean amount of total cellular MBNL1 per PB is 0.1%, so it will be similar in cells grown under normal conditions as the PB sizes are similar and they appear to be of similar brightness by immunofluorescence. Again, this would be straightforward to quantify with our existing data if this is, indeed what the reviewer is requesting, but we question the biological significance. We would be reluctant to derive a Pearson’s co-efficient for the degree of co-localisation between CUGBP1 and MBNL1 in P-bodies as the structures are too small in size for this to be meaningful within the limits of imaging capabilities. We could, however, provide this if this is a specific request.

Figure 5: "Treatment with sodium arsenite was then carried out under time-lapse microscopy, with Z-stacks of images taken every 4 minutes until stress granule formation was clearly seen (Fig.5A). This revealed a pronounced delay in formation of stress granules in cells containing CUGexp foci (HeLa CTG960 GFPMBNL1, 36 min +/- 12) compared to those without (HeLa CTG12 GFPMBNL1, 15 min +/- 2) (Fig.5B)." Data representation in Figure 5 is unclear and the pronounced delay in stress granule formation is not appreciated. Since the authors performed a live imaging taking pictures every 4 minutes, it would me more informative to plot the data and show the assembly and disassembly kinetics over time for both control and CTG960_ GFPMBNL1 cell lines (similar to what shown in e.g. Gwon et al., Science 2021, Ubiquitination of G3BP1 mediates stress granule disassembly in a context-specific manner, Figure 2G).

The bar graph in figure 5B shows that cells from the CTG960 line take more than twice as long to form SGs compared to controls and are lost in half the time, with the precise numbers given in the text. A simple bar graph seemed the clearest way to present this. However, we have plotted our existing data in a similar manner similar to that in the cited reference and added this to figure 5. These graphs clearly show that the differences we see are at least as great as in other published literature, including the reference given by the reviewer (see below).

Concerning Figure 1, the authors report no difference in the kinetic of stress granule formation in HLECs. However, they only report data after 45 and 60 min of arsenite treatment; at these time-points the assembly step is maximal. Thus, for consistency, the authors should include earlier time-points to the analysis of stress granule assembly also in HLECs, similar to what done in HeLa cells in Figure 5.

The assembly step is not ‘maximal’ in these cell lines after 45 minutes. Figure 2A clearly shows that only ~30% of cells have SGs after 45 minutes of treatment, compared with 100% of cells after 90 minutes shown in figure 2B. We have additional data at 10, 20 and 30 minutes all showing no significant differences. We had omitted them to keep the graph simple, but have now included them as a graph of ‘% of cells with stress granules against time’ in figure 2.

"Having established that MBNL1 and CUGBP1 co-localise closely in stress granules": the authors investigated the colocalization of each of these two proteins with stress granule markers but they did not verify whether MBNL1 and CUGBP1 co-localise.

In figure 1B we show that endogenous CUGBP1 and endogenous MBNL1 both co-localise with the stress granule marker TIA1 in stress granules in lens epithelial cells. It would, therefore, be highly unlikely that CUGBP1 and MBNL1 would not co-localise with each other in stress granules. We have also previously verified that GFPMBNL1 behaves identically to its endogenous counterpart (Coleman et al, 2014). Furthermore, in figure 4C and 4D, we show close co-localisation between endogenous CUGBP1 and GFPMBNL1 in stress granules in our HeLa cell model, using high-resolution AiryScan microscopy for which we provide detailed quantitation.

This aspect should be addressed experimentally since the authors also conclude that "a complex relationship between MBNL1 and CUGBP1 in stress granules" exists. Thus, the authors need to assess the colocalization of GFPMBNL1 with endogenous CUGBP1 in stress granules and the one of GFPCUGBP1 with endogenous MBNL1.

The complex relationship we propose is based on the effects of CUGBP1 or MBNL1 knockdown on the dynamic behaviours of each other by FRAP assay and not solely on their co-localisation, although we have already analysed their co-localisation in detail as above.

Figure 6: Please add antibody labeling to microscopy panels A and B.

Certainly, this was an accidental omission and has been added

Moreover, specify is the numbers refer to minutes in panel F. The data representation is also unclear - see comment above, Figure 5.

As stated in the figure legend and on the graph axes, these numbers have been normalised to the mean time taken for SG formation/loss in the control CTG12 cell line (set at 100%). The precise numbers in minutes for mean and SD are given in the text. We have added additional graphs of ‘% of cells with stress granules against time’ to this figure, with the values in minutes given to clarify the exact time-scale.

Figure 7: was 1,6-hexanediol added in presence of arsenite? Or was arsenite removed?

Arsenite was not removed (neither was Doxycycline) as we wanted to examine the effect of 1,6-hexanediol on SGs and PBs without the added complication of the effects of stress removal. We will clarify this point in the methods/results.

Aberrant persistent stress granules have been implicated in age-related (Mateju et al., 2017) and neurodegenerative diseases (Protter and Parker, 2016), such as ALS and FTD (Jain et al., 2016; Markmiller et al., 2018; Zhang et al., 2018). These are proposed to result from increased liquid-to-solid phase transitions within the stress granules (Mateju et al., 2017)." The authors should better define what are aberrant stress granules (e.g. see Ganassi et al., 2016; Turakhiya et al., 2018, PMID: 29804830).*

We will expand on this subject in the discussion

"Persistent stress granules have long been associated with degenerative conditions, notably ALS (Li et al., 2013)". I suggest updating the reference adding a more recent one.

We selected this 2013 review to emphasise that there is a long history of association of persistent stress granules with degenerative conditions. We will add in an additional, more recent review.

Significance

The work is descriptive; thus, in this form I do not consider that it is strongly advancing the field.

Having noted alterations to stress granule disassembly in lens epithelial cells from DM1 patients, we went on to develop a novel inducible model in which we replicated and enhanced these effects by expressing the large CUGexp RNA that causes DM1 as part of a DMPK mini-gene mimicking the genetic mutation seen in DM1 patients. This is not purely descriptive. Furthermore, we are now in a position to add an additional figure showing two pieces of evidence for functional defects in stress granules associated with CUGexp RNA expression 1) reduced accumulation of total PolyA RNA in stress granules indicating compromised function and 2) compromised ‘docking’ events between stress granules and P-bodies, a process proposed to be integral to the function of both structures.

Author Response:

Evaluation Summary:

This manuscript is of primary interest to readers in the field of infectious diseases especially the ones involved in COVID-19 research. The identification of immunological signatures caused by SARS-CoV-2 in HIV-infected individuals is important not only to better predict disease outcomes but also to predict vaccine efficacy and to potentially identify sources of viral variants. In here, the authors leverage a combination of clinical parameters, limited virologic information and extensive flow cytometry data to reach descriptive conclusions.

We have extensively reworked the paper.

Reviewer #1 (Public Review):

The methods appear sound. The introduction of vaccines for COVID-19 and the emergence of variants in South Africa and how they may impact PLWH is well discussed making the findings presented a good reference backdrop for future assessment. Good literature review is also presented. Specific suggestions for improving the manuscript have been identified and conveyed to the authors.

We thank the Reviewer for the support.

Reviewer #2 (Public Review):

Karima, Gazy, Cele, Zungu, Krause et al. described the impact of HIV status on the immune cell dynamics in response to SARS-CoV-2 infection. To do so, during the peak of the KwaZulu-Natal pandemic, in July 2020, they enrolled a robust observational longitudinal cohort of 124 participants all positive for SARS-CoV-2. Of the participants, a group of 55 people (44%) were HIV-infected individuals. No difference is COVID-19 high risk comorbidities of clinical manifestations were observed in people living with HIV (PLWH) versus HIV-uninfected individuals exception made for joint ache which was more present in HIV-uninfected individuals. In this study, the authors leverage and combine extensive clinical information, virologic data and immune cells quantification by flow cytometry to show changes in T cells such as post-SARS-CoV-2 infection expansion of CD8 T cells and reduced expression CXCR3 on T cells in specific post-SARS-CoV-2 infection time points. The authors also conclude that the HIV status attenuates the expansion of antibody secreting cells. The correlative analyses in this study show that low CXCR3 expression on CD8 and CD4 T cells correlates with Covid-19 disease severity, especially in PLWH. The authors did not observe differences in SARS-CoV-2 shedding time frame in the two groups excluding that HIV serostatus plays a role in the emergency of SARS-CoV-2 variants. However, the authors clarify that their PLWH group consisted of mostly ART suppressed participants whose CD4 counts were reasonably high. The study presents the following strengths and limitations

We thank the Reviewer for the comments. The cohort now includes participants with low CD4.

Strengths:

A. A robust longitudinal observational cohort of 124 study participants, 55 of whom were people living with HIV. This cohort was enrolled in KwaZulu-Natal,South Africa during the peak of the pandemic. The participants were followed for up to 5 follow up visits and around 50% of the participants have completed the study.

We thank the Reviewer for the support. The cohort has now been expanded to 236 participants.

B. A broad characterization of blood circulating cell subsets by flow cytometry able to identify and characterize T cells, B cells and innate cells.

We thank the Reviewer for the support.

Weaknesses:

The study design does not include

A. a robust group of HIV-infected individuals with low CD4 counts, as also stated by the authors

This has changed in the resubmission because we included participants from the second, beta variant dominated infection wave. For this infection wave we obtained what we think is an important result, presented in a new Figure 2:

This figure shows that in infection wave 2 (beta variant), CD4 counts for PLWH dropped to below the CD4=200 level, yet recovered after SARS-CoV-2 clearance. Therefore, the participants we added had low CD4 counts, but this was SARS-CoV-2 dependent.

B. a group of HIV-uninfected individuals and PLWH with severe COVID-19. As stated in the manuscript the majority of our participants did not progress beyond outcome 4 of the WHO ordinal scale. This is also reflected in the age average of the participants. Limiting the number of participants characterized by severe COVID-19 limits the study to an observational correlative study

Death has now been added to Table 1 under the “Disease severity” subheading. The number of participants who have died, at 13, is relatively small. We did not limit the study to non-critical cases. Our main measure of severity is supplemental oxygen.

This is stated in the Results, line 106-108:

“Our cohort design did not specifically enroll critical SARS-CoV-2 cases. The requirement for supplemental oxygen, as opposed to death, was therefore our primary measure for disease severity.”

This is justified in the Discussion, lines 219-225:

“Our cohort may not be a typical 'hospitalized cohort' as the majority of participants did not require supplemental oxygen. We therefore cannot discern effects of HIV on critical SARS-CoV-2 cases since these numbers are too small in the cohort. However, focusing on lower disease severity enabled us to capture a broader range of outcomes which predominantly ranged from asymptomatic to supplemental oxygen, the latter being our main measure of more severe disease. Understanding this part of the disease spectrum is likely important, since it may indicate underlying changes in the immune response which could potentially affect long-term quality of life and response to vaccines.”

C. a control group enrolled at the same time of the study of HIV-uninfected and infected individuals.

This was not possible given constraints imposed on bringing non-SARS-CoV-2 infected participants into a hospital during a pandemic for research purposes. However, given that the study was longitudinal, we did track participants after convalescence. This gave us an approximation of participant baseline in the absence of SARS-CoV-2, for the same participants. Results are presented in Figure 2 above.

D. results that elucidate the mechanisms and functions of immune cells subsets in the contest of COVID-19.

We do not have functional assays.

Reviewer #3 (Public Review):

Karim et al have assembled a large cohort of PLWH with acute COVID-19 and well-matched controls. The main finding is that, despite similar clinical and viral (e.g., shedding) outcomes, the immune response to COVID-19 in PLWH differs from the immune response to COVID-19 in HIV uninfected individuals. More specifically, they find that viral loads are comparable between the groups at the time of diagnosis, and that the time to viral clearance (by PCR) is also similar between the two groups. They find that PLWH have higher proportions and also higher absolute number of CD8 cells in the 2-3 weeks after initial infection.

The authors do a wonderful job of clinically characterizing the research participants. I was most impressed by the attention to detail with respect to timing of viral diagnosis as it related to symptom onset and specimen collection. I was also impressed by the number of longitudinal samples included in this study.

We thank the Reviewer for the support.

Author Response:

Reviewer #1:

The authors demonstrate deficits in perceptual tests related to fine-time perception in non-speech and speech sounds in a group of patients with stroke aphasia compared to a control group without a lesion. A subgroup of patients with deficits in spectrotemporal processing at a fine timescale have lesions mapped to the posterior STS, MTG and adjacent white matter. The area associated with deficits in spectrotemporal analysis with a fine timescale is then used as a seed for probabilistic fibre tractography based on diffusion MR. These results show connectivity of the functionally defined seed region with a number of areas including the cerebellum.

The work is carefully done and I think interesting in demonstrating the cerebellar connections of the functionally defined region associated with deficits in fine temporal analysis that might be a basis for event representation at this temporal level.

We appreciate the referee's evaluation and constructive feedback.

Reviewer #2:

Based on consideration of supportive evidence in the literature, the authors propose that a cerebellar-temporal lobe functional network plays a key role in auditory temporal processing. The precise parsing of temporal information is critical to understanding dynamic auditory processing and thus is an interesting area of study. Better understanding of how the cerebellum and temporal lobe may interact to achieve such parsing of the dynamic signal in a generative/predictive internal model is of clear interest to a broad readership. This idea is put to the test by first having individuals with lesions in the posterior portion of left temporal lobe perform speech perception and timing tasks and comparing performance with 12 healthy controls to establish the role of this region in tasks reliant on intact fine temporal processing. Typically, a lesion model will be helpful when a dissociation between structure and function can be demonstrated, and preferably this would be a double dissociation. Here, while lesions to auditory regions of the left temporal lobe are associated with impoverished performance on speech and temporal order tasks relative to a healthy control group, performance on comparably difficult auditory tasks that do not require good temporal discrimination is not tested to determine if there is such a dissociation. Given the extensive discussion of hypothesized different time sensitivities of right and left auditory cortices in the Introduction, patients with right homologous lesions might also have served as an interesting control and could have supported a double dissociation. In a second step to their study, a seed region was generated based on comparison of the lesion loci for the half of the patients who performed most poorly on the behavioral tasks to the other half, and this was used to explore anatomical tract connectivity of the seed region to the rest of the brain in the neuroimaging data from the healthy controls, with a focus on connections with the cerebellum. This approach to establishing that "temporo-cerebellar connectivity underlies timing constraints in audition" is unfortunately just not that convincing. The data are interesting, but taken alone they simply do not support such a conclusion. In the data, there is no clear functional link established or even hinted at between the temporal lobe and the cerebellum.

We appreciate the referee's evaluation and constructive feedback. We address the raised concerns point by point below. We appreciate the concerns regarding our methodological choice and our interpretation of a functional link between the temporal lobe and the cerebellum. It certainly is more reasonable to derive a functional interpretation based on disconnection measured directly in patients’ DTI. However, if unavailable, indirect measures of disconnection can also be used to establish a functional link between a lesioned region and the networks associated with it. The rationale behind this is that it reflects an indirect estimation of the effect of a lesion on structural brain networks. To make this approach clearer, we have revised the manuscript accordingly. See revised manuscript pages 6 and 12:

[...] Assessing connectivity in healthy participants based on lesion information is a relatively new method that measures structural disconnection in networks associated with given anatomical regions (Foulon et al., 2018). This allows for the indirect estimation of the lesion effect on structural brain networks. In this regard, it was shown that behavioral deficits can be explained similarly by local brain damage and indirectly measured disconnection (Salvalaggio et al., 2020). [...]

[...] We next used the respective areas as seed regions for probabilistic fiber tractography in a healthy age-matched sample to visualize the underlying common connectivity pattern (see Methods). Thus, we indirectly explored the association between posterior superior temporal disconnection and processing of sound at short timescales. [...]

We also changed the abstract and conclusion accordingly. See pages 2 and 15 of the revised manuscript.

[...] Here we tested whether temporo-cerebellar disconnection is associated with the processing of sound at short timescales. [...]

[...] The evidence we describe (i) shows that lesion-related deficits in spectrotemporal analysis occur in posterior temporal regions connected to the cerebellum [...].

Reviewer #3:

Stockert et al. investigate the cortico-cerebellar network underpinning rapid temporal auditory analysis. This study uses a well-defined group of stroke participants with mostly circumscribed lesions to the left posterior superior temporal lobe to motivate probabilistic tractography from cortical regions associated with verbal and non-verbal rapid auditory temporal analysis. Lesion-symptom mapping identifies a specific region of the posterior superior temporal sulcus and underlying white matter as statistically associated with impairment in rapid auditory temporal analysis. Tractography results demonstrate that these regions have high structural connectivity to wider regions of the left hemisphere cortical language network and ipsilateral and contralateral connectivity to postero-lateral cerebellum and dentate nucleus. It is interpreted that this cortico-cerebellar network is crucial to developing representations of fine auditory temporal structure.

The conclusions of the paper are an interpretation which is based on integrating previous neuropsychology with the current tractography results and based on well-defined models in the motor domain. Such conclusions are not unreasonable but there is no direct (associative) evidence linking this network to the cognitive function of interest.

Strengths:

The paper integrates neuropsychology and neuroimaging methodologies to build a coherent picture which is more than the sum of its parts. The stroke group has well-defined and selected lesions which enable testing of the hypotheses put forward by the authors. The behavioural measures are sensitive and suitable to identify impairments in the behaviours of interest. There has been a detailed analysis of the behavioural speech perception data in the stroke group which largely, although perhaps not entirely, conforms to the asymmetric temporal sampling hypothesis. The lesion-symptom mapping approach is suitable for the nature of the population (small group with similar lesion distributions) and has allowed neuropsychologically guided tractography in the neurotypical population. This has clearly illustrated the complexity of the structural connectivity of the posterior superior temporal sulcus and underlying regions.

Weaknesses:

The selective nature of the stroke population - relatively small, chronic lesions - has resulted in only mild impairments for a small number of participants (6/12 participants). At the group level there is no difference between the stroke and neurotypical population on speech perception measures - group statistics do not reach one tailed significance. This reduces the certainty with which the regions identified are associated with the behaviour or interest. However, the results do conform to previous neuropsychology and lesion studies and it is likely that this lack of effect is due to low statistical power.

Please refer to our response to the next point.

All the stroke participants have a similar lesion distribution, and this makes lesion-symptom mapping challenging. For example, lesion data do not give an indication of the functional integrity of perilesional regions which can be reduced, even at the chronic stage, therefore the superior temporal sulcus may not be functioning effectively, even in the proportion of the group without lesions to this area. Lesion symptom mapping is more robust with a wider distribution of lesions and the inclusion of participants with lesions remote from the area of interest. Having said that, the behavioural measures appear sensitive enough to identify mild impairments and the authors, for good reason, wished to reduce the extension of lesion into primary auditory regions. As above, given the limited sample and homogeneous lesion, the lesion symptom mapping approach is reasonable.

We agree that the small number of patients is a possible limitation to the study and add this point to the limitations section. See revised manuscript page 21.

[...] First, the study population is relatively small and lesion symptom mapping is typically applied to larger populations with wider lesion distribution. Although careful selection of circumscribed lesions has the advantage of highlighting behavioral differences without confounding other deficits (e.g., primary auditory processing), it is possible that additional regions are involved in processing of sound at short timescales. However, tractography based on healthy participants makes it possible to indirectly obtain information (i.e., structural disconnection) about brain regions contributing to the investigated function. In addition, it is likely that the small number of patients might hamper the ability to detect statistically significant differences between the behavior of controls and patients. Nevertheless, we are confident that the current results align with the fact that the posterior superior temporal cortex contributes to the processing of sound at short timescales, as indicated by previous neuropsychological evidence and lesion studies (Boemio et al., 2005; Chedru, Bastard, and Efron, 1978; Efron, 1963; Robson, Grube, Lambon Ralph, Griffiths, & Sage, 2013; Swisher & Hirsh, 1972). Further studies should however test larger populations to replicate and extend this finding. [...]

The authors suggest that the behavioural results conform to the asymmetric temporal sampling hypothesis in that only place of articulation discrimination impairments in the stroke group can be (just about) detected, whereas there were no significant stroke-neurotypical differences in other phonetic contrasts. It is not clear that the VOT differences associated with plosive voicing changes and the cues associated with place changes happen over fundamentally different time-scales and, therefore, it is important to further justify the interpretation of the data. In the future it will be helpful to have this level of analysis applied to individuals with lesions to the wider speech perception network to draw conclusions about the specificity of the impairment to these regions - for example, impairments in phoneme discrimination have been associated with frontal lobe lesions.

It appears that voicing contrasts in which shorter and longer voice onset times result in the perception of a voiced or voiceless plosive (for example [t] and [d]) are encoded in both the temporal envelope and fine structure (Rosen 1992) of the speech signal that occur in time windows of 20-500 ms and <2 ms, respectively. In words an additional cue is the closure time, which can be further used to discriminate between voiced and voiceless plosives. However, place of articulation contrasts are exclusively encoded in the temporal fine structure (i.e., very quick transitions of the frequency spectrum, formant transitions). Even though for all contrasts shorter timescale information plays a role, somewhat redundant encoding is present for voice contrasts. Ultimately, place of articulation contrasts seem to be the most difficult to discriminate. In Figure 2D it is apparent that despite highest error rates for the place of articulation contrasts, several patients also showed impaired discrimination for voicing contrast when compared to healthy controls. We do agree with the referee that it would be interesting to also extend this level of analysis to individuals with lesions in the wider speech perception network in future work.

The tractography results reveal a complex pattern of structural connectivity, including other regions associated with speech perception. The authors have a theoretical motivation to focus on the importance of the temporo-cerebellar pathway but there is no correlation evidence to link auditory temporal analysis to the integrity of this pathway in the neurotypical population. The non-verbal measures appear to be sufficiently sensitive for this type of analysis. This lack of association with behaviour makes it hard to draw conclusions about the functional role of this network.

We appreciate the referee’s concerns about our interpretation of the functional link between the temporal lobe and the cerebellum regarding auditory temporal analysis. It certainly is more reasonable to derive a functional interpretation based on disconnection measured directly in patients DTI. However, if unavailable, indirect measures of disconnection can also be used to establish a functional link between a lesioned region and the networks associated with it. The rationale behind this is that it reflects an indirect estimation of the effect of a lesion on structural brain networks. To make this approach clearer, we have modified the manuscript as such. See revised manuscript pages 6 and 12:

[...] Assessing connectivity in healthy participants based on lesion information is a relatively new method that measures structural disconnection in networks associated with given anatomical regions (Foulon et al., 2018). This allows for the indirect estimation of the effect of a lesion on structural brain networks. In this regard, it has been shown that behavioral deficits are explained to a similar extent by both the local damage and indirectly measured disconnection (Salvalaggio et al., 2020). [...]

[...] We next used the respective areas as seed regions for probabilistic fiber tractography in a healthy age-matched sample to visualize the underlying common connectivity pattern (see Methods). Thus, we indirectly explored the association between posterior superior temporal disconnection and processing of sound at short timescales. [...]

We also changed the abstract and conclusion accordingly. See revised manuscript pages 2 and 15.

[...] Here we tested whether temporo-cerebellar disconnection is associated with processing of sound at short timescale. [...]

[...] The evidence we describe (i) shows that lesion-related deficits in spectrotemporal analysis occur in posterior temporal regions connected to the cerebellum [...].

Author Response:

Reviewer #1 (Public Review):

The authors interrogated an underexplored feature of CRISPR arrays to enhance multiplexed genome engineering with the CRISPR nuclease Cas12a. Multiplexing represents one of the many desirable features of CRISPR technologies, and use of highly compact CRISPR arrays from CRISPR-Cas systems allows targeting of many sites at one time. Recent work has shown though that the composition of the array can have a major impact on the performance of individual guide RNAs encoded within the array, providing ample opportunities for further improvements. In this manuscript, the authors found that the region within the repeat lost through processing, what they term the separator, can have a major impact on targeting performance. The effect was specifically tied to upstream guide sequences with high GC content. Introducing synthetic separator sequences shorter than their natural counterparts but exhibiting similarly low GC content boosted targeted activation of a reporter in human cells. Applying one synthetic separator to a seven-guide array targeting chromosomal genes led to consistent though more modest targeted activation. These findings introduce a distinct design consideration for CRISPR arrays that can further enhance the efficacy of multiplexed applications. The findings also suggest a selective pressure potentially influencing the repeat sequence in natural CRISPR arrays.

Strengths:

The portion of the repeat discarded through processing normally has been included or discarded when generating a CRISPR-Cas12a array. The authors clearly show that something in between-namely using a short version with a similarly low GC content-can enhance targeting over the truncated version. A coinciding surprising result was that the natural separator completely eliminated any measurable activation, necessitating the synthetic separator.

The manuscript provides a clear progression from identifying a feature of the upstream sequences impacting targeting to gaining insights from natural CRISPR-Cas12a systems to applying the insights to enhance array performance.

With further support, the use of synthetic separators could be widely adopted across the many applications of CRISPR-Cas12a arrays.

Weaknesses:

The terminology used to describe the different parts of the CRISPR array could better align with those in the CRISPR biology field. For one, crRNAs (abbreviated from CRISPR RNAs) should reflect the final processed form of the guide RNA, whereas guide RNAs (gRNAs) captures both pre-processed and post-processed forms. Also, "spacers" should reflect the natural spacers acquired by the CRISPR-Cas system, whereas "guides" better capture the final sequence in the gRNA used for DNA target recognition.

We thank the reviewer for this correction. We have now changed most uses of “crRNA” to “gRNA”. We decided to retain the use of the word “spacer” for the target recognition portion of the gRNA rather than changing it to “guide” as the reviewer suggests, because we think there is a risk that the reader would confuse “guide” with the non-synonymous “guide-RNA”. We have added a remark explaining our use of “spacer” (“A gRNA consists of a repeat region, which is often identical for all gRNAs in the array, and a spacer (here used synonymously with “guide region”)”)

A running argument of the work is that the separator specifically evolved to buffer adjacent crRNAs. However, this argument overlooks two key aspects of natural CRISPR arrays. First, the spacer (~30 nts) is normally much longer than the guide used in this work (20 nts), already providing the buffer described by the authors. This spacer also undergoes trimming to form the mature crRNA.

If we understand this comment correctly, the argument is that, in contrast to a ~20-nt spacer, a 30-nt spacer would provide a buffer between adjacent guides even if a separator is not present. However, even a 30-nt spacer may have high GC content and form secondary structures that would interfere with processing of the subsequent gRNA. Our hypothesis is that the separator is AT-rich and so insulates gRNAs from one another regardless of the length or GC composition of spacers. Please let us know if we have misunderstood this comment.

Second, the repeat length is normally fixed as a consequence of the mechanisms of spacer acquisition. At most, the beginning of each repeat sequence may have evolved to reduce folding interactions without changing the repeat length, although some of these repeats are predicted to fold into small hairpins.

We agree with this comment. Indeed, we propose that the separator, which is part of the repeat sequence, has evolved to reduce folding interactions. We now clarify this at the end of the Results section: “Taken together, the results from our study suggest that the CRISPR-separator has evolved as an integral part of the repeat region that likely insulates gRNAs from the disrupting effects of varying GC content in upstream spacers.”

Prior literature has highlighted the importance of a folded hairpin with an upstream pseudoknot within the repeat (Yamano Cell 2016), where disrupting this structure compromises DNA targeting by Cas12a (Liao Nat Commun 2019, Creutzburg NAR 2020). This structure is likely central to the authors' findings and needs to be incorporated into the analyses.

We thank the reviewer for this important insight. We have now performed experiments exploring the involvement of the pseudoknot in the disruptive effects of high-GC spacers.

First, we used our 2-gRNA CRISPR array design (Fig. 1D) where the second gRNA targets the GFP promoter and the first gRNA contains a non-targeting dummy spacer. We generated several versions of this array where we iteratively introduced targeted point mutations in the dummy spacer to either form a hairpin restricted to the dummy spacer, or a hairpin that would compete with the pseudoknot in the GFP-gRNA’s repeat region (new Fig. S3). We found that both of these modifications significantly reduced performance of the GFP-targeting gRNA. These results suggest that interfering with the pseudoknot indeed disrupts gRNA performance, but that also hairpins that presumably don’t interfere directly with the pseudoknot are detrimental – perhaps by sterically hindering Cas12a from accessing its cleavage site. Interestingly, the AAAT synSeparator largely rescued performance of the worst-performing of these constructs. These results are displayed in the new Fig. S3 and discussed in the related part of the Results section.

Second, we have now performed a computational analysis using RNAfold where we correlated the performance of all dummy spacers with their predicted secondary structure (Fig. 1M). The correlation between predicted RNA structure and array performance was higher when the structural prediction included both the dummy spacer and the entire GFP-targeting gRNA (R2 = 0.57) than when it included only the dummy spacer (R2 = 0.27; new figure panel S1C). This higher correlation suggests that secondary structures that involve the GFP-targeting gRNA play a more important role in our experiment than secondary structures that only involve the dummy spacer. These results are described in the Results section and in the Fig. 1 legend.

Third, we now also performed secondary structure analysis (RNAfold) of two of our worst-performing dummy spacers (50% and 70% GC), which indicated that these spacers are likely to form secondary structures that involve both the repeat and spacer of the downstream GFP-targeting gRNA (Fig. 3G-H). Interestingly, this analysis suggested that the AAAT synSeparator improves performance of these spacers by loosening up these secondary structures or creating an unstructured bulge at the Cas12a cleavage site. These results are presented in Fig. 3G-H and the accompanying portion of the Results section.

To conclude, our analyses suggest that the secondary structure in the spacer and its interference with the pseudoknot in the repeat hairpin play a role in gRNA performance, wherein the inclusion of the AAAT synSeparator can partly rescue the performance, likely by restoring the Cas12a accessibility to the gRNA cleavage site.

Many claims could better reflect the cited literature. For instance, Creutzburg et al. showed that adding secondary structures to the guide to promote folding of the repeat hairpin enhanced rather than interfered with targeting.

We thank the reviewer for this comment. Creutzburg et al. report the interesting finding that a carefully designed 3’ extension of the spacer can counteract secondary structures that disrupt the repeat. In this way, the extension rescues disruptive secondary structures that involve the repeat and any upstream sequence. Relevant to this finding, it is conceivable that the synSeparator (AAAT) exerts its beneficial effect at the 3’ end of the GFP spacer by folding back onto the GFP spacer and in this way blocking secondary structures caused by a GC-rich dummy spacer located upstream of the GFP gRNA, according to the mechanism reported by Creutzburg et al. However, we used structural prediction of the GFP-targeting gRNA with and without the AAAT synSeparator and did not find evidence that the AAAT extension would cause this spacer to fold back onto itself (data not shown). Moreover, our experimental data (Fig. 3E) demonstrate that the synSeparator exerts its main beneficial effect when located upstream of the GFP-targeting gRNA, which would not be the case if the main mechanism was the one demonstrated by Creutzburg et al. We already had a paragraph discussing the Creutzburg paper in the Discussion, but we have now added a sentence specifying the mechanism that Creutzburg et al. demonstrated: “RNA secondary structure prediction (RNAfold) did not indicate that the GFP-targeting spacer would fold back on itself when an AAAT extension is added to the 3’ end, which would have been the case for the mechanism demonstrated by Creutzburg et al. (data not shown).”

Liu et al. NAR 2019 further showed that the pre-processed repeat actually enhanced rather than reduced performance compared to the processed repeat.

The experiment referenced by the reviewer (Fig. 2 in Liu et al., Nucleic Acids Research, 2019) in fact nicely supports our findings. In Liu et al., the pre-processed repeat only shows improved performance if it is located upstream of the targeting gRNA, and the gRNA is not followed by an additional pre-processed repeat (DRf-crRNA in their Fig. 2B & C). In this situation, the pre-processed repeat (containing the natural separator) may serve to enhance gRNA processing, as would be expected based on our results. At the same time, the absence of a full-length repeat downstream of the gRNA means that after gRNA processing, there will not remain any piece of RNA attached to the 3’ end of the spacer, which might disrupt gRNA performance. In contrast, when Liu et al. added an additional pre-processed repeat downstream of their gRNA (DRf-crRNA-DRf in the same panel), this construct performed the worst of all tested variants. This is consistent with our conclusion that the full-length separator reduces performance of gRNAs if it remains attached to the 3’ end of spacers. We have added a paragraph in the Discussion about this (Line 376).

Finally, the complete loss of targeting with the unprocessed repeat appears represent an extreme example given multiple studies that showed effective targeting with this repeat (e.g. Liu NAR 2019, Zetsche Nat Biotechnol 2016).

We acknowledge that our CRISPR array containing the full, natural separator (Fig. 3B) appears to be completely non-functional in contrast to the studies mentioned by the reviewer. We think this difference may have a few possible explanations. First, this array is in fact not entirely non-functional. Re-running the same experiment with a stronger dCas12a-activator (dCas12a-VPR, full length VPR, also used in Fig. 5) shows some modest GFP activation even with the full separator (1.4% vs 20.8% GFP+ cells; see the Appendix Figure 1). But for consistency, we have used the same, slightly less effective, dCas12a-activator (dCas12a-miniVPR) for all GFP-targeting experiments. Second, both the Liu et al. and Zetsche et al. studies used CRISPR editing rather than CRISPRa. We speculate that this might explain their relatively high indel frequency: Only a single cleavage event needs to take place for an indel to occur, whereas gene activation presumably requires the dCas12a-activator to be present on the promoter for extended periods of time. Thus, any inefficiency in DNA binding caused by the separator remaining attached to the spacer might disfavor CRISPRa activity more than CRISPR-editing activity. We have added these considerations to the Discussion and referenced the suggested papers (Line 376).

Appendix Figure 1: Percentage of GFP+ cells without or with a full-length separator using dCas12a-VPR (full length) gene activation.

Relating to the above point, the vast majority of the results relied on a single guide sequence targeting GFP. While the seven-guide CRISPR array did involve other sequences, only the same GFP targeting guide yielded strong gene activation. Therefore, the generalizability of the conclusions remains unclear.

We have now performed several experiments that address the generalizability of our conclusions:

First, we now include data demonstrating that the beneficial effect of adding a synSeparator is not limited to the AAAT sequence derived from the Lachnospiraceae bacterium separator. We now include three other 4-nt, AT-rich synSeparators derived from Acidaminococcus s. (TTTT), Moraxella b. (TTTA) and Prevotella d. (ATTT) (Fig. 3I). All these synSeparators rescued the poor GFP activation caused by an upstream spacer with high GC content, though not equally effectively. The quantitative difference between the synSeparators could either be due to the intrinsic “insulation capacity” of these sequences, or the way they interact with the Lb-Cas12a protein, or to sequence-specific interactions with this particular CRISPR array. We discuss these possibilities in the Discussion (Line 437).

Second, we now include data demonstrating that nuclease-deactivated, enhanced-Cas12a from Acidaminococcus species (enAsdCas12a; Kleinstiver et al., 2019) is also sensitive to the effects of high-GC spacers (Fig. 3J). This poor performance was largely rescued by including a TTTT synSeparator derived from the natural AsCas12a separator.

Furthermore, we have now included a paragraph in the Discussion where we speculate on why the effect of adding the synSeparator was more modest for the endogenous genes than for GFP: 1) Our GFP-expressing cell line has multiple GFP insertions in its genome, and each copy has seven protospacers in its promoter. This may amplify the effect of the synSeparator. 2) The gRNAs used for endogenous activation were taken from the literature or had been pre-tested by us. These guides had thus already proven to be successful and might not be particularly disruptive (e.g., they were not selected by us for having high GC content). Therefore, researchers might experience the greatest benefit from the synSeparator with newly designed spacers that have not already proven to be effective even without the synSeparator.

Reviewer #3 (Public Review):

Magnusson et al., do an excellent job of defining how the repeated separator sequence of Wild Type Cas12a CRISPR arrays impacts the relative efficacy of downstream crRNAs in engineered delivery systems. High-GC content, particularly near the 3' end of the separator sequence appears to be critically important for the processing of a downstream crRNA. The authors demonstrated naturally occurring separators from 3 Cas12a species also display reduced GC content. The authors use this important new information to construct a synthetic small separator DNA sequence which can enhance CRISPR/Cas12a-based gene regulation in human cells. The manuscript will be a great resource for the synthetic biology field as it shows an optimization to a tool that will enable improved multi-gene transcriptional regulation.

Strengths:

• The authors do an excellent job in citing appropriate references to support the rationale behind their hypotheses.
• The experiments and results support the authors' conclusions (e.g., showing the relationship between secondary structure and GC content in the spacers).
• The controls used for the experiments were appropriate (e.g., using full-length natural separator vs single G or 1 to 4 A/T nucleotides as synthetic separators).
• The manuscript does a great job assessing several reasons why the synthetic separator might work in the discussion section, cites the relevant literature on what has been done and restates their results to argument in favor or against these reasons.
• This paper will be very useful for research groups in the genome editing and synthetic biology fields. The data presented (specially the data concerning the activation of several genes) can be used as a comparison point for other labs comparing different CRISPR-based transcriptional regulators and the spacers used for targeting.
• This paper also provides optimization to a tool that will be useful for regulating several endogenous genes at once in human cells thus helping researchers studying pathways or other functional relationships between several genes.

Opportunities for Improvement:

• The authors have performed all the experiments using LbCas12a as a model and have conclusively proven that the synSeparator enhances the performance of Cas12a based gene activation. Is this phenomenon will be same for other Cas12a proteins (such as AsCas12a)? The authors should perform some experiments to test the universality of the concept. Ideally, this would be done in HEK293T cells and one other human cell type.

We thank the reviewer for these suggestions. We have now addressed the generalizability of our findings with several new experiments. First, we now include data demonstrating that nuclease-deactivated, enhanced Cas12a from Acidaminococcus species (denAsCas12a; Kleinstiver et al., 2019) is also sensitive to the effects of high-GC spacers (Fig. 3J). This poor performance was largely rescued by including a TTTT synSeparator derived from the natural AsCas12a separator.

Second, we now include data demonstrating that the beneficial effect of adding a synSeparator is not limited to the AAAT sequence derived from the Lachnospiraceae b. separator. We now include three other 4-nt, AT-rich synSeparators derived from Acidaminococcus s. (TTTT), Moraxella b. (TTTA) and Prevotella d. (ATTT) (Fig. 3I). All these synSeparators rescued the poor GFP activation caused by an upstream spacer with high GC content, though not equally effectively. The quantitative difference between the synSeparators could either be due to the intrinsic “insulation capacity” of these sequences, or the way they interact with the Lb-Cas12a protein, or to sequence-specific interactions with this particular CRISPR array. We discuss these possibilities in the Discussion.

Third, as described above, we have now performed an in vitro Cas12a cleavage assay and present the data in a new figure (Fig. 4). We found that a CRISPR array containing a 70%-GC dummy spacer was processed less efficiently than an array containing a 30%-GC spacer, but that addition of a synSeparator could to a large extent rescue this processing defect (Fig. 4E). The fact that this result was observed even in a cell-free in vitro setting demonstrates that it is a general feature of Cas12a CRISPR arrays that is likely to work the same way in many cell types rather than being specific to HEK293T cells.

Fourth, we attempted to investigate the effect of the synSeparator in different cell types. However, either due to poor transfection efficiency or poor expression of the Cas12a activator construct, CRISPRa activity was consistently poor in these cell types, both with and without the synSeparator (e.g., we did not visually observe fluorescence from the mCherry gene fused to the dCas12a activator, which we always see in HEK293T cells). Because of the low general efficiency of CRISPRa, it was not possible to evaluate the performance of the synSeparator. Many cell types are difficult to transfect and dCas12a-VPR-mCherry is a big construct (>6 kb). To our knowledge, there have not been many reports using dCas12a-VPR in cell types other than HEK293T. While we think that it will be important to optimize CRISPRa in many cell types (e.g., by optimizing transfection conditions, Cas12a variants, promoters, expression vectors, etc.), the focus of our study has been to show the separator’s mechanism and general function; we believe that optimizing general CRISPRa for different cell types is beyond the scope of this paper. We acknowledge that this is a limitation of our study and we have added a paragraph about this in the Discussion (line 355). We nevertheless hypothesize that the negative influence of high-GC spacers and the insulating effect of synSeparators are generalizable across cell types. That is because we could observe improved array processing with the synSeparator even in the cell-free context of an in vitro expression system, as described above (Fig. 4). This suggests that the sensitivity to spacer GC content is determined only by the interaction between Cas12a and the array, rather than being dependent on a particular cellular context.

I understand that Disney may not have openly queer characters and I understand people want more representation. However, I think it is weird to compare the sisters to anything "couple" related since they are sisters and that has nothing to do with being Queer. So I don't exactly understand how someone could take that into an account and have a theory that Frozen is a queer movie.

Author Response:

Reviewer #1:

The manuscript by Takahashi et al describes the interaction between MLL fusion proteins with HBO1 and its role in leukemogenesis. Myeloid progenitor transformation assays using various MLL fusion proteins reveal that MLL fusion proteins requires the TRX2 domain of MLL for effective leukemic transformation. IP-MS identifies HBO1 as a bona fide binding partner of the MLL TRX2 domain. ChIP-seq experiments show genome-wide colocalization of HBO1 complex with MLL-ENL and the WT MLL in MLL-fusion leukemia cells and MLL WT cells, respectively. ChIP-qPCR in MLL-deficient cells suggest that recruitment of HBO1 to MLL target genes (such as MYC and CDKN2C) depends on MLL. Truncation analysis of the ELL part of the MLL-ELL fusion reveal that MLL-ELL transformation activity requires OHD domain-mediated recruitment of AF4 and EAF1. Furthermore, co-IP and ChIP experiments with various fragments show that AF4 and EAF1 form two distinct SL1/MED26-containing complexes and likely the AEP/SL1/MED26 complex is competent for transactivation. Series of transformation assays suggest that MLL-ELL transforms hematopoietic progenitors via association with AEP, but not other ELL-associated proteins. Finally, the authors also show that NUP98-HBO1 fusion transforms myeloid progenitors through interaction with MLL. Overall, this is a quite comprehensive study demonstrating that various MLL fusions and NUP98 fusions transform hematopoietic progenitors via HBO1-MLL interaction, which suggests that targeting their interaction might be s new therapeutic approach.

We appreciate the comments and inputs from the reviewers.

Reviewer #2:

In this manuscript, the authors identified an interesting interaction of MLL (a methyltransferase) with an HBO1-JADE complex (an acetyltransferase) and investigated the functional impact in leukemogenesis by fusion proteins containing MLL or HBO1. The data is clear and the connection between MLL and HBO1 is unexpected. The manuscript is also well organized and relatively easy to follow.

Comments:

1) The functional relevance of the interaction between MLL and HBO1 is still correlative. It would be important to know whether there are any results directly about the impact of the loss of the HBO1 complex on the function of MLL.

We performed a sgRNA-dropout assay which showed that HBO1 is critically required for the survival of leukemia cell lines, as depicted in Figure 2F and Figure 2-figure supplement 3.

2) It is important to show the source and specificity of the antibodies that were used for ChIP of the HBO1 complex.

The details of the antibodies are provided in Key Resource Table.

3) It might be interesting to check whether other JADE proteins and also BRD1 (another partner of HBO1) are involved.

We agree that it would be very interesting to examine the involvement of other JADE/BRPF family proteins in the future because they share the ING4/5 subunits and BRD1 plays an important role in hematopoiesis (1). This can be addressed in future studies.

4) The acronym TRX2 may be confusing as some might think that it is thioredoxin.

As advised, we have changed this to THD2 (TRX homology domain 2).

Reviewer #3:

This paper starts with a series of bone marrow transformation assays comparing MLL fusions and domain-deletion mutants thereof to define the minimal elements for robust leukemic transformation and surveying growth and attendant common fusion targets HoxA9, Meis1 in colony replanting assays. Here they discover that a region of the MLL-N portion just upstream of the well-studied CXXC domain, termed in their previous work the "TRX2 domain" is important for the transformation capacity for several different MLL-fusions (and more minimal chimeras of key modules). A small region of the MLL-N protein encompassing the TRX2 domain and the CXXC module are subjected to complex purification, it is clear from comparison to number of controls that the TRX2 domain is an important mediator of association, perhaps indirect, with the HBO complex. Drop out experiments confirm that HBO1 knockout is lethal to MLL-rearranged leukemia, nicely confirming recent work (Ay et al., MacPherson et al.).

ChIP-seq experiments in an ALL with MLL-ENL fusion, and then more extensively in a kidney cancer cell line indicate overlap with some of the HBO complex subunits and MLL, however this does not establish recruitment at these sites. ChIP-qPCR at a few MLL-fusion target genes with MLL depletion supports the recruitment hypothesis somewhat although mixed and modest effect sizes indicate that alternate pathways for HBO1 recruitment are involved, and could also be explained as reduced deposition of marks known to recruit HBO1, rather than direct recruitment. Sadly, the real potential strength of this work goes unrealized, as the recruitment of HBO1 mechanism remains tantalizingly out of reach. More experiments in this space could conclusively establish the molecular mechanism of a seemingly biomedically important recruitment paradigm, and thereby have much more impact.

As the reviewer pointed out, MLL is not the only element that recruits the HBO1 complex to the target chromatin. MLL is known to deposit H3K4me marks, and the HBO1 complex is known to recognize these marks via ING4/5 subunits. We performed a ChIP-qPCR analysis of H3K4me3 in MLL-knockout cells. At the MYC promoter, the H3K4me3 marks were substantially decreased (Figure 3F). Moreover, recruitment of HBO1 was not recovered by transient expression of an MLL mutant containing THD2, indicating that the presence of H3K4me3 marks is a prerequisite for HBO1 recruitment. In accordance with this, ING5-histone interaction is required for the stable association of MLL with the HBO1 complex (Figure 8A-C). Thus, a more appropriate molecular mechanism would be the cooperative recruitment of the HBO1 complex by ING4/5-mediated chromatin association and MLL-mediated association. Because of the multiple contacts involved in this molecular network, it is not easy to pinpoint the direct contacts as desired, but our biochemical analyses indicate that PHF16 and ING4/5 offer relatively strong binding surfaces (Figure 8A-C). The ING domain of ING5 is the most likely direct binding surface identified thus far.

At this point the paper shifts to a seemingly distinct line of inquiry, which is not closely related to the HBO1-TRX2 story to the first three figures. The new direction examines the ELL fusion partner in some detail using similar fusion protein chimeras, but a portion of Figure 4, is largely confirmatory of previously established findings about the critical regions of ELL for transformation and its AF4/EAF1 partners, adding only that portions of the MLL fusion protein are dispensable, provided that they are replaced with the PWWP of LEDGF. It is a little bit of a Frankenstein's monster experiment, and does not add much new to the field. Further experiments define potentially two distinct complexes that have already been characterized being recruited by ELL, although there is overlap here again with their previous studies, and the results are a little hard to interpret.

A portion of Figure 4 was confirmatory to previous results. We have moved this to figure supplements in the revised manuscript (Figure 4-figure supplement 1B,C). The main topic of this paper is the role of the HBO1 complex in MLL-mediated transactivation pathways. The structure/function analysis of MLL fusion proteins demonstrated that MLL-ELL is highly dependent on the HBO1-mediated function in leukemic transformation (Figures 1 and 2). Hence, it was important to clarify the mechanism of gene activation by MLL-ELL in this study to understand why HBO1 association is required for MLL-ELL-mediated transformation. Because MLL-ELL associates with AEP similarly to major MLL fusions such as MLL-AF4 and MLL-ENL, it was speculated that MLL-ELL also activates its target genes via AEP. However, ELL associates with EAF family proteins and MLL-EAF also has transforming ability (3). Thus, EAF1-mediated functions could be more important for MLL-ELL-mediated transformation rather than AEP-mediated functions. To clarify the mechanism of MLL-ELL-mediated transformation, we generated a point mutant that selectively impaired ELL-EAF interaction and demonstrated that EAF1-association is dispensable for MLL-ELL-mediated transformation (Figure 6), thereby indicating that MLL-ELL transforms via AEP-mediated functions, which demands HBO1-mediated functions. We also showed that the presence of THD2 enhances ELL-AEP association to further suggest that one of the roles of the HBO1 complex is to enhance the association of ELL with AEP (Figure 6E). These findings are not reinterpretations of our prior results and are relevant to the main topic of this paper. We believe this part adds new information to the field, and therefore we have included it in the revised manuscript.

The authors create structure-guided separation of function mutants in the ELL domain that binds both AEP and SL1, permitting them to specifically disrupt EAF1 interactions but not AF4. Further experiments solidify this interpretation, and find that this mutant shows no deficits in hematopoietic progenitor transformation or primary leukemia lethality, although there appears to be some effect upon reimplantation.

The last figure in the paper tackles the seemingly unrelated Nup98-HBO1 fusion, a rare patient mutation-they demonstrate a requirement for MLL for viability of hematopoietic progenitors transformed by this fusion, connecting back to the TRX2 interaction, and show that menin inhibitors slow growth.

Strengths:

The identification of the TRX2 region of the MLL-N protein as the major point of contact (perhaps not direct), to the HBO1 complex adds mechanistic depth to the really important recent discovery (confirmed in this work) the MLL-fusion leukemias rely on HBO1 function. This lab has published a number of technically similar types of papers defining minimal regions of MLL and distinct interacting partners by chimeric fusions, with bone marrow transformation assays, mouse model engrafting studies, IP's, ChIP etc. In my view they are very much under cited, likely because they are similarly so challenging to read.

Thank you for your pointed feedback. We will try our best to make the necessary improvements so that our papers are widely read and cited.

The mixture of Co-IP biochemistry, bone marrow transformation assays, and ChIP, to define interactions, minimal requirements for transformation, and their chromatin consequences for a host of different MLL-fusions and HBO1-fusions has the potential to define the key interfaces underlying recruitment.

Weaknesses:

The mechanistic inquiry stops short of really defining the critical MLL-HBO1 complex interface. Defining the point of contact on the HBO1 side (even which subunit) and determining whether it is direct, or bridged by some, as yet unidentified factor, as well as conclusively demonstrating that this is the mechanism of HBO1 recruitment remain the major shortcomings.

To address this criticism, we further investigated the mechanism of complex formation by MLL and the HBO1 complex. As we demonstrated in Figure 8A-C, the association appears to be mediated by multiple contacts mainly through PHF16 and ING4/5. Because this association needs an intact PHD finger of ING5, it likely occurs depending on the context where ING4/5 is bound to histone H3K4me2/3. The ING domain of ING5 was also required for the association, indicating that this portion may contains a point of direct contact. We speculate that HBO1 recruitment is mediated primarily by ING4/5-H3K4me3 interaction and MLL reinforces its chromatin association.

And the follow-on figures apart from the last one, appear disconnected from this portion of the story and distract from it.

We depicted a revised model incorporating the above-mentioned aspect in Figure 8D of the revised manuscript.

The complex nomenclature and density/organization/logic of the presentation of experiments makes this paper difficult to read. Absence of sufficient grounding in the broader literature much beyond their own lab's work further compounds the problem.

We changed some of the nomenclature and density/organization/logic of the presentation of the experiments to improve the readability.

There is a lot of overlap, particularly in parts of figure 1 and figure 4 with previously published results. So perhaps re-organizing the display of data, and the organization of presentation, putting confirmatory work in the supplementary figures, would improve accessibility.

We moved some portions of Figure 4 to figure supplement. The data for MLL-AF10 and MLL-ENL were retained in the Figure 1 as important references.

Author Response:

Reviewer #1:

1) The user manual and tutorial are well documented, although the actual code could do with more explicit documentation and comments throughout. The overall organisation of the code is also a bit messy.

We have now implemented an ongoing, automated code review via Codacy (https://app.codacy.com/gh/caseypaquola/BigBrainWarp/dashboard). The grade is published as a badge on GitHub. We improved the quality of the code to an A grade by increasing comments and fixing code style issues. Additionally, we standardised the nomenclature throughout the toolbox to improve consistency across scripts and we restructured the bigbrainwarp function.

2) My understanding is that this toolbox can take maps from BigBrain to MRI space and vice versa, but the maps that go in the direction BigBrain->MRI seem to be confined to those provided in the toolbox (essentially the density profiles). What if someone wants to do some different analysis on the BigBrain data (e.g. looking at cellular morphology) and wants that mapped onto MRI spaces? Does this tool allow for analyses that involve the raw BigBrain data? If so, then at what resolution and with what scripts? I think this tool will have much more impact if that was possible. Currently, it looks as though the 3 tutorial examples are basically the only thing that can be done (although I may be lacking imagination here).

The bigbrainwarp function allows input of raw BigBrain data in volume and surface forms. For volumetric inputs, the image must be aligned to the full BigBrain or BigBrainSym volume, but the function is agnostic to the input voxel resolution. We have also added an option for the user to specify the output voxel resolution. For example,

bigbrainwarp --in_space bigbrain --in_vol cellular_morphology_in_bigbrain.nii \ --interp linear --out_space icbm --out_res 0.5 \ --desc cellular_morphology --wd working_directory

where “cellular_morphology_in_bigbrain.nii” was generated from a BigBrain volume (see Table 2 below for all parameters). The BigBrain volume may be the 100-1000um resolution images provided on the ftp or a resampled version of these images, as long as the full field of view is maintained. For surface-based inputs, the data must contain a value for each vertex of the BigBrain/BigBrainSym mesh. We have clarified these points in the Methods, illustrated the potential transformations in an extended Figure 3 and highlighted the distinctiveness of the tutorial transformations in the Results.

3) An obvious caveat to bigbrain is that it is a single brain and we know there are sometimes substantial individual variations in e.g. areal definition. This is only slightly touched upon in the discussion. Might be worth commenting on this more. As I see it, there are multiple considerations. For example (i) Surface-to-Surface registration in the presence of morphological idiosyncracies: what parts of the brain can we "trust" and what parts are uncertain? (ii) MRI parcellations mapped onto BigBrain will vary in how accurately they may reflect the BigBrain areal boundaries: if histo boundaries do not correspond with MRI-derived ones, is that because BigBrain is slightly different or is it a genuine divergence between modalities? Of course addressing these questions is out of scope of this manuscript, but some discussion could be useful; I also think this toolbox may be useful for addressing this very concerns!

We agree that these are important questions and hope that BigBrainWarp will propel further research. Here, we consider these questions from two perspectives; the accuracy of the transformations and the potential influence of individual variation. For the former, we conducted a quantitative analysis on the accuracy of transformations used in BigBrainWarp (new Figure 2). We provide a function (evaluate_warp.sh) for BigBrainWarp users to assess accuracy of novel deformation fields and encourage detailed inspection of accuracy estimates and deformation effects for region of interest studies. For the latter, we expanded our Discussion of previous research on inter-individual variability and comment on the potential implications of unquantified inter-individual variability for the interpretation of BigBrain-MRI comparisons.

Methods (P.7-8):

“A prior study (Xiao et al., 2019) was able to further improve the accuracy of the transformation for subcortical structures and the hippocampus using a two-stage multi-contrast registration. The first stage involved nonlinear registration of BigBrainSym to a PD25 T1-T2 fusion atlas (Xiao et al., 2017, 2015), using manual segmentations of the basal ganglia, red nucleus, thalamus, amygdala, and hippocampus as additional shape priors. Notably, the PD25 T1-T2 fusion contrast is more similar to the BigBrainSym intensity contrast than a T1-weighted image. The second stage involved nonlinear registration of PD25 to ICBM2009sym and ICBM2009asym using multiple contrasts. The deformation fields were made available on Open Science Framework (https://osf.io/xkqb3/). The accuracy of the transformations was evaluated relative to overlap of region labels and alignment of anatomical fiducials (Lau et al., 2019). The two-stage procedure resulted in 0.86-0.97 Dice coefficients for region labels, improving upon direct overlap of BigBrainSym with ICBM2009sym (0.55-0.91 Dice) (Figure 2Aii, 2Aiv top). Transformed anatomical fiducials exhibited 1.77±1.25mm errors, on par with direct overlap of BigBrainSym with ICBM2009sym (1.83±1.47mm) (Figure 2Aiii, 2Aiv below). The maximum misregistration distance (BigBrainSym=6.36mm, Xiao=5.29mm) provides an approximation of the degree of uncertainty in the transformation. In line with this work, BigBrainWarp enables evaluation of novel deformation fields using anatomical fiducials and region labels (evaluate_warps.sh). The script accepts a nonlinear transformation file for registration of BigBrainSym to ICBM2009sym, or vice versa, and returns the Jacobian map, Dice coefficients for labelled regions and landmark misregistration distances for the anatomical fiducials.

The unique morphology of BigBrain also presents challenges for surface-based transformations. Idiosyncratic gyrification of certain regions of BigBrain, especially the anterior cingulate, cause misregistration (Lewis et al., 2020). Additionally, the areal midline representation of BigBrain, following inflation to a sphere, is disproportionately smaller than standard surface templates, which is related to differences in surface area, in hemisphere separation methods, and in tessellation methods. To overcome these issues, ongoing work (Lewis et al., 2020) combines a specialised BigBrain surface mesh with multimodal surface matching [MSM; (Robinson et al., 2018, 2014)] to co-register BigBrain to standard surface templates. In the first step, the BigBrain surface meshes were re-tessellated as unstructured meshes with variable vertex density (Möbius and Kobbelt, 2010) to be more compatible with FreeSurfer generated meshes. Then, coarse-to-fine MSM registration was applied in three stages. An affine rotation was applied to the BigBrain sphere, with an additional “nudge” based on an anterior cingulate landmark. Next, nonlinear/discrete alignment using sulcal depth maps (emphasising global scale, Figure 2Biii), followed by nonlinear/discrete alignment using curvature maps (emphasising finer detail, Figure 2Biii). The higher- order MSM procedure that was implemented for BigBrain maximises concordance of these features while minimising surface deformations in a physically plausible manner, accounting for size and shape distortions (Figure 2Bi) (Knutsen et al., 2010; Robinson et al., 2018). This modified MSMsulc+curv pipeline improves the accuracy of transformed cortical maps (4.38±3.25mm), compared to a standard MSMsulc approach (8.02±7.53mm) (Figure 2Bii-iii) (Lewis et al., 2020).”

Figure 2: Evaluating BigBrain-MRI transformations. A) Volume-based transformations i. Jacobian determinant of deformation field shown with a sagittal slice and stratified by lobe. Subcortical+ includes the shape priors (as described in Methods) and the + connotes hippocampus, which is allocortical. Lobe labels were defined based on assignment of CerebrA atlas labels (Manera et al., 2020) to each lobe. ii. Sagittal slices illustrate the overlap of native ICBM2009b and transformed subcortical+ labels. iii. Superior view of anatomical fiducials (Lau et al., 2019). iv. Violin plots show the DICE coefficient of regional overlap (ii) and landmark misregistration (iii) for the BigBrainSym and Xiao et al., approaches. Higher DICE coefficients shown improved registration of subcortical+ regions with Xiao et al., while distributions of landmark misregistration indicate similar performance for alignment of anatomical fiducials. B) Surface-based transformations. i. Inflated BigBrain surface projections and ridgeplots illustrate regional variation in the distortions of the mesh invoked by the modified MSMsulc+curv pipeline. ii. Eighteen anatomical landmarks shown on the inflated BigBrain surface (above) and inflated fsaverage (below). BigBrain landmarks were transformed to fsaverage using the modified MSMsulc+curv pipeline. Accuracy of the transformation was calculated on fsaverage as the geodesic distance between landmarks transformed from BigBrain and the native fsaverage landmarks. iii. Sulcal depth and curvature maps are shown on inflated BigBrain surface. Violin plots show the improved accuracy of the transformation using the modified MSMsulc+curv pipeline, compared to a standard MSMsulc approach.

Discussion (P.18):

“Cortical folding is variably associated with cytoarchitecture, however. The correspondence of morphology with cytoarchitectonic boundaries is stronger in primary sensory than association cortex (Fischl et al., 2008; Rajkowska and Goldman-Rakic, 1995a, 1995b). Incorporating more anatomical information in the alignment algorithm, such as intracortical myelin or connectivity, may benefit registration, as has been shown in neuroimaging (Orasanu et al., 2016; Robinson et al., 2018; Tardif et al., 2015). Overall, evaluating the accuracy of volume- and surface-based transformations is important for selecting the optimal procedure given a specific research question and to gauge the degree of uncertainty in a registration.”

Discussion (P.19):

“Despite all its promises, the singular nature of BigBrain currently prohibits replication and does not capture important inter-individual variation. While large-scale cytoarchitectural patterns are conserved across individuals, the position of areal boundaries relative to sulci vary, especially in association cortex (Amunts et al., 2020; Fischl et al., 2008; Zilles and Amunts, 2013) . This can affect interpretation of BigBrain-MRI comparisons. For instance, in tutorial 3, low predictive accuracy of functional communities by cytoarchitecture may be attributable to the subject- specific topographies, which are well established in functional imaging (Benkarim et al., 2020; Braga and Buckner, 2017; Gordon et al., 2017; Kong et al., 2019). Future studies should consider the influence of inter-subject variability in concert with the precision of transformations, as these two elements of uncertainty can impact our interpretations, especially at higher granularity.”

Reviewer #2:

This is a nice paper presenting a review of recent developments and research resulting from BigBrain and a tutorial guiding use of the BigBrainWarp toolbox. This toolbox supports registration to, and from, standard MRI volumetric and surface templates, together with mapping derived features between spaces. Examples include projecting histological gradients estimated from BigBrain onto fsaverage (and the ICMB2009 atlas) and projecting Yeo functional parcels onto the BigBrain atlas.

The key strength of this paper is that it supports and expands on a comprehensive tutorial and docker support available from the website. The tutorials there go into even more detail (with accompanying bash scripts) of how to run the full pipelines detailed in the paper. The docker makes the tool very easy to install but I was also able to install from source. The tutorials are diverse examples of broad possible applications; as such the combined resource has the potential to be highly impactful.

The minor weaknesses of the paper relate to its clarity and depth. Firstly, I found the motivations of the paper initially unclear from the abstract. I would recommend much more clearly stating that this is a review paper of recent research developments resulting from the BigBrain atlas, and a tutorial to accompany the bash scripts which apply the warps between spaces. The registration methodology is explained elsewhere.

In the revised Abstract (P.1), we emphasise that the manuscript involves a review of recent literature, the introduction of BigBrainWarp, and easy-to-follow tutorials to demonstrate its utility.

“Neuroimaging stands to benefit from emerging ultrahigh-resolution 3D histological atlases of the human brain; the first of which is “BigBrain”. Here, we review recent methodological advances for the integration of BigBrain with multi-modal neuroimaging and introduce a toolbox, “BigBrainWarp", that combines these developments. The aim of BigBrainWarp is to simplify workflows and support the adoption of best practices. This is accomplished with a simple wrapper function that allows users to easily map data between BigBrain and standard MRI spaces. The function automatically pulls specialised transformation procedures, based on ongoing research from a wide collaborative network of researchers. Additionally, the toolbox improves accessibility of histological information through dissemination of ready-to-use cytoarchitectural features. Finally, we demonstrate the utility of BigBrainWarp with three tutorials and discuss the potential of the toolbox to support multi-scale investigations of brain organisation.”

I also found parts of the paper difficult to follow - as a methodologist without comprehensive neuroanatomical terminology, I would recommend the review of past work to be written in a more 'lay' way. In many cases, the figure captions also seemed insufficient at first. For example it was not immediately obvious to me what is meant by 'mesiotemporal confluence' and Fig 1G is not referenced specifically in the text. In Fig 3C it is not immediately clear from the text of the caption that the cortical image is representing the correlation from the plots - specifically since functional connectivity is itself estimated through correlation.

In the updated manuscript, we have tried to remove neuroanatomical jargon and clearly define uncommon terms at the first instance in text. For example,

“Evidence has been provided that cortical organisation goes beyond a segregation into areas. For example, large- scale gradients that span areas and cytoarchitectonic heterogeneity within a cortical area have been reported (Amunts and Zilles, 2015; Goulas et al., 2018; Wang, 2020). Such progress became feasible through integration of classical techniques with computational methods, supporting more observer-independent evaluation of architectonic principles (Amunts et al., 2020; Paquola et al., 2019; Schiffer et al., 2020; Spitzer et al., 2018). This paves the way for novel investigations of the cellular landscape of the brain.”

“Using the proximal-distal axis of the hippocampus, we were able to bridge the isocortical and hippocampal surface models recapitulating the smooth confluence of cortical types in the mesiotemporal lobe, i.e. the mesiotemporal confluence (Figure 1G).”

“Here, we illustrate how we can track resting-state functional connectivity changes along the latero-medial axis of the mesiotemporal lobe, from parahippocampal isocortex towards hippocampal allocortex, hereafter referred to as the iso-to-allocortical axis.”

Additionally, we have expanded the captions for clarity. For example, Figure 3:

“C) Intrinsic functional connectivity was calculated between each voxel of the iso-to-allocortical axis and 1000 isocortical parcels. For each parcel, we calculated the product-moment correlation (r) of rsFC strength with iso-to- allocortical axis position. Thus, positive values (red) indicate that rsFC of that isocortical parcel with the mesiotemporal lobe increases along the iso-to-allocortex axis, whereas negative values (blue) indicate decrease in rsFC along the iso-to-allocortex axis.”

My minor concern is over the lack of details in relation to the registration pipelines. I understand these are either covered in previous papers or are probably destined for bespoke publications (in the case of the surface registration approach) but these details are important for readers to understand the constraints and limitations of the software. At this time, the details for the surface registration only relate to an OHBM poster and not a publication, which I was unable to find online until I went through the tutorial on the BigBrain website. In general I think a paper should have enough information on key techniques to stand alone without having to reference other publications, so, in my opinion, a high level review of these pipelines should be added here.

There isn't enough details on the registration. For the surface, what features were used to drive alignment, how was it parameterised (in particular the regularisation - strain, pairwise or areal), how was it pre-processed prior to running MSM - all these details seem to be in the excellent poster. I appreciate that work deserves a stand alone publication but some details are required here for users to understand the challenges, constraints and limitations of the alignment. Similar high level details should be given for the registration work.

We expanded descriptions of the registration strategies behind BigBrainWarp, especially so for the surface-based registration. Additionally, we created a new Figure to illustrate how the accuracy of the transformations may be evaluated.

Methods (P.7-8):

“For the initial BigBrain release (Amunts et al., 2013), full BigBrain volumes were resampled to ICBM2009sym (a symmetric MNI152 template) and MNI-ADNI (an older adult T1-weighted template) (Fonov et al., 2011). Registration of BigBrain to ICBM2009sym, known as BigBrainSym, involved a linear then a nonlinear transformation (available on ftp://bigbrain.loris.ca/BigBrainRelease.2015/). The nonlinear transformation was defined by a symmetric diffeomorphic optimiser [SyN algorithm, (Avants et al., 2008)] that maximised the cross- correlation of the BigBrain volume with inverted intensities and a population-averaged T1-weighted map in ICBM2009sym space. The Jacobian determinant of the deformation field illustrates the degree and direction of distortions on the BigBrain volume (Figure 2Ai top).

A prior study (Xiao et al., 2019) was able to further improve the accuracy of the transformation for subcortical structures and the hippocampus using a two-stage multi-contrast registration. The first stage involved nonlinear registration of BigBrainSym to a PD25 T1-T2 fusion atlas (Xiao et al., 2017, 2015), using manual segmentations of the basal ganglia, red nucleus, thalamus, amygdala, and hippocampus as additional shape priors. Notably, the PD25 T1-T2 fusion contrast is more similar to the BigBrainSym intensity contrast than a T1-weighted image. The second stage involved nonlinear registration of PD25 to ICBM2009sym and ICBM2009asym using multiple contrasts. The deformation fields were made available on Open Science Framework (https://osf.io/xkqb3/). The accuracy of the transformations was evaluated relative to overlap of region labels and alignment of anatomical fiducials (Lau et al., 2019). The two-stage procedure resulted in 0.86-0.97 Dice coefficients for region labels, improving upon direct overlap of BigBrainSym with ICBM2009sym (0.55-0.91 Dice) (Figure 2Aii, 2Aiv top). Transformed anatomical fiducials exhibited 1.77±1.25mm errors, on par with direct overlap of BigBrainSym with ICBM2009sym (1.83±1.47mm) (Figure 2Aiii, 2Aiv below). The maximum misregistration distance (BigBrainSym=6.36mm, Xiao=5.29mm) provides an approximation of the degree of uncertainty in the transformation. In line with this work, BigBrainWarp enables evaluation of novel deformation fields using anatomical fiducials and region labels (evaluate_warps.sh). The script accepts a nonlinear transformation file for registration of BigBrainSym to ICBM2009sym, or vice versa, and returns the Jacobian map, DICE coefficients for labelled regions and landmark misregistration distances for the anatomical fiducials.

The unique morphology of BigBrain also presents challenges for surface-based transformations. Idiosyncratic gyrification of certain regions of BigBrain, especially the anterior cingulate, cause misregistration (Lewis et al., 2020). Additionally, the areal midline representation of BigBrain, following inflation to a sphere, is disproportionately smaller than standard surface templates, which is related to differences in surface area, in hemisphere separation methods, and in tessellation methods. To overcome these issues, ongoing work (Lewis et al., 2020) combines a specialised BigBrain surface mesh with multimodal surface matching [MSM; (Robinson et al., 2018, 2014)] to co-register BigBrain to standard surface templates. In the first step, the BigBrain surface meshes were re-tessellated as unstructured meshes with variable vertex density (Möbius and Kobbelt, 2010) to be more compatible with FreeSurfer generated meshes. Then, coarse-to-fine MSM registration was applied in three stages. An affine rotation was applied to the BigBrain sphere, with an additional “nudge” based on an anterior cingulate landmark. Next, nonlinear/discrete alignment using sulcal depth maps (emphasising global scale, Figure 2Biii), followed by nonlinear/discrete alignment using curvature maps (emphasising finer detail, Figure 2Biii). The higher- order MSM procedure that was implemented for BigBrain maximises concordance of these features while minimising surface deformations in a physically plausible manner, accounting for size and shape distortions (Figure 2Bi) (Knutsen et al., 2010; Robinson et al., 2018). This modified MSMsulc+curv pipeline improves the accuracy of transformed cortical maps (4.38±3.25mm), compared to a standard MSMsulc approach (8.02±7.53mm) (Figure 2Bii-iii) (Lewis et al., 2020).”

(SEE FIGURE 2 in Response to Reviewer #1)

I would also recommend more guidance in terms of limitations relating to inter-subject variation. My interpretation of the results of tutorial 3, is that topographic variation of the cortex could easily be driving the greater variation of the frontal parietal networks. Either that, or the Yeo parcel has insufficient granularity; however, in that case any attempt to go to finer MRI driven parcellations - for example to the HCP parcellation, would create its own problems due to subject specific variability.

We agree that inter-individual variation may contribute to the low predictive accuracy of functional communities by cytoarchitecture. We expanded upon this possibility in the revised Discussion (P. 19) and recommend that future studies examine the uncertainty of subject-specific topographies in concert with uncertainties of transformations.

“These features depict the vast cytoarchitectural heterogeneity of the cortex and enable evaluation of homogeneity within imaging-based parcellations, for example macroscale functional communities (Yeo et al., 2011). The present analysis showed limited predictability of functional communities by cytoarchitectural profiles, even when accounting for uncertainty at the boundaries (Gordon et al., 2016). [...] Despite all its promises, the singular nature of BigBrain currently prohibits replication and does not capture important inter-individual variation. While large- scale cytoarchitectural patterns are conserved across individuals, the position of boundaries relative to sulci vary, especially in association cortex (Amunts et al., 2020; Fischl et al., 2008; Zilles and Amunts, 2013) . This can affect interpretation of BigBrain-MRI comparisons. For instance, in tutorial 3, low predictive accuracy of functional communities by cytoarchitecture may be attributable to the subject-specific topographies, which are well established in functional imaging (Benkarim et al., 2020; Braga and Buckner, 2017; Gordon et al., 2017; Kong et al., 2019). Future studies should consider the influence of inter-subject variability in concert with the precision of transformations, as these two elements of uncertainty can impact our interpretations, especially at higher granularity.”

Reviewer #3:

The authors make a point for the importance of considering high-resolution, cell-scale, histological knowledge for the analysis and interpretation of low-resolution MRI data. The manuscript describes the aims and relevance of the BigBrain project. The BigBrain is the whole brain of a single individual, sliced at 20µ and scanned at 1µ resolution. During the last years, a sustained work by the BigBrain team has led to the creation of a precise cell-scale, 3D reconstruction of this brain, together with manual and automatic segmentations of different structures. The manuscript introduces a new tool - BigBrainWarp - which consolidates several of the tools used to analyse BigBrain into a single, easy to use and well documented tool. This tool should make it easy for any researcher to use the wealth of information available in the BigBrain for the annotation of their own neuroimaging data. The authors provide three examples of utilisation of BigBrainWarp, and show the way in which this can provide additional insight for analysing and understanding neuroimaging data. The BigBrainWarp tool should have an important impact for neuroimaging research, helping bridge the multi-scale resolution gap, and providing a way for neuroimaging researchers to include cell-scale phenomena in their study of brain data. All data and code are available open source, open access.

Main concern:

One of the longstanding debates in the neuroimaging community concerns the relationship between brain geometry (in particular gyro/sulcal anatomy) and the cytoarchitectonic, connective and functional organisation of the brain. There are various examples of correspondance, but also many analyses showing its absence, particularly in associative cortex (for example, Fischl et al (2008) by some of the co-authors of the present manuscript). The manuscript emphasises the accuracy of their transformations to the different atlas spaces, which may give some readers a false impression. True: towards the end of the manuscript the authors briefly indicate the difficulty of having a single brain as source of histological data. I think, however, that the manuscript would benefit from making this point more clearly, providing the future users of BigBrainWarp with some conceptual elements and references that may help them properly apprise their results. In particular, it would be helpful to briefly describe which aspects of brain organisation where used to lead the deformation to the different templates, if they were only based on external anatomy, or if they took into account some other aspects such as myelination, thickness, …

We agree with the Reviewer that the accuracy of the transformation and the potential influence of inter-individual variability should be carefully considered in BigBrain-MRI studies. To highlight these issues in the updated manuscript, we first conducted a quantitative analysis on the accuracy of transformations used in BigBrainWarp (new Figure 2). We provide a function (evaluate_warp.sh) for users to assess accuracy of novel deformation fields and encourage detailed inspection of accuracy estimates and deformation effects for region of interest studies. Second, we expanded our discussion of previous research on inter-individual variability and comment on the potential implications of unquantified inter-individual variability for the interpretation of BigBrain-MRI comparisons.

Methods (P.7-8):

“A prior study (Xiao et al., 2019) was able to further improve the accuracy of the transformation for subcortical structures and the hippocampus using a two-stage multi-contrast registration. The first stage involved nonlinear registration of BigBrainSym to a PD25 T1-T2 fusion atlas (Xiao et al., 2017, 2015), using manual segmentations of the basal ganglia, red nucleus, thalamus, amygdala, and hippocampus as additional shape priors. Notably, the PD25 T1-T2 fusion contrast is more similar to the BigBrainSym intensity contrast than a T1-weighted image. The second stage involved nonlinear registration of PD25 to ICBM2009sym and ICBM2009asym using multiple contrasts. The deformation fields were made available on Open Science Framework (https://osf.io/xkqb3/). The accuracy of the transformations was evaluated relative to overlap of region labels and alignment of anatomical fiducials (Lau et al., 2019). The two-stage procedure resulted in 0.86-0.97 Dice coefficients for region labels, improving upon direct overlap of BigBrainSym with ICBM2009sym (0.55-0.91 Dice) (Figure 2Aii, 2Aiv top). Transformed anatomical fiducials exhibited 1.77±1.25mm errors, on par with direct overlap of BigBrainSym with ICBM2009sym (1.83±1.47mm) (Figure 2Aiii, 2Aiv below). The maximum misregistration distance (BigBrainSym=6.36mm, Xiao=5.29mm) provides an approximation of the degree of uncertainty in the transformation. In line with this work, BigBrainWarp enables evaluation of novel deformation fields using anatomical fiducials and region labels (evaluate_warps.sh). The script accepts a nonlinear transformation file for registration of BigBrainSym to ICBM2009sym, or vice versa, and returns the Jacobian map, Dice coefficients for labelled regions and landmark misregistration distances for the anatomical fiducials.

The unique morphology of BigBrain also presents challenges for surface-based transformations. Idiosyncratic gyrification of certain regions of BigBrain, especially the anterior cingulate, cause misregistration (Lewis et al., 2020). Additionally, the areal midline representation of BigBrain, following inflation to a sphere, is disproportionately smaller than standard surface templates, which is related to differences in surface area, in hemisphere separation methods, and in tessellation methods. To overcome these issues, ongoing work (Lewis et al., 2020) combines a specialised BigBrain surface mesh with multimodal surface matching [MSM; (Robinson et al., 2018, 2014)] to co-register BigBrain to standard surface templates. In the first step, the BigBrain surface meshes were re-tessellated as unstructured meshes with variable vertex density (Möbius and Kobbelt, 2010) to be more compatible with FreeSurfer generated meshes. Then, coarse-to-fine MSM registration was applied in three stages. An affine rotation was applied to the BigBrain sphere, with an additional “nudge” based on an anterior cingulate landmark. Next, nonlinear/discrete alignment using sulcal depth maps (emphasising global scale, Figure 2Biii), followed by nonlinear/discrete alignment using curvature maps (emphasising finer detail, Figure 2Biii). The higher- order MSM procedure that was implemented for BigBrain maximises concordance of these features while minimising surface deformations in a physically plausible manner, accounting for size and shape distortions (Figure 2Bi) (Knutsen et al., 2010; Robinson et al., 2018). This modified MSMsulc+curv pipeline improves the accuracy of transformed cortical maps (4.38±3.25mm), compared to a standard MSMsulc approach (8.02±7.53mm) (Figure 2Bii-iii) (Lewis et al., 2020).”

(SEE Figure 2 in response to previous reviewers)

Discussion (P.18, 19):

“Cortical folding is variably associated with cytoarchitecture, however. The correspondence of morphology with cytoarchitectonic boundaries is stronger in primary sensory than association cortex (Fischl et al., 2008; Rajkowska and Goldman-Rakic, 1995a, 1995b). Incorporating more anatomical information in the alignment algorithm, such as intracortical myelin or connectivity, may benefit registration, as has been shown in neuroimaging (Orasanu et al., 2016; Robinson et al., 2018; Tardif et al., 2015). Overall, evaluating the accuracy of volume- and surface-based transformations is important for selecting the optimal procedure given a specific research question and to gauge the degree of uncertainty in a registration.”

“Despite all its promises, the singular nature of BigBrain currently prohibits replication and does not capture important inter-individual variation. While large-scale cytoarchitectural patterns are conserved across individuals, the position of boundaries relative to sulci vary, especially in association cortex (Amunts et al., 2020; Fischl et al., 2008; Zilles and Amunts, 2013) . This can have implications on interpretation of BigBrain-MRI comparisons. For instance, in tutorial 3, low predictive accuracy of functional communities by cytoarchitecture may be attributable to the subject-specific topographies, which are well established in functional imaging (Benkarim et al., 2020; Braga and Buckner, 2017; Gordon et al., 2017; Kong et al., 2019). Future studies should consider the influence of inter- subject variability in concert with the precision of transformations, as these two elements of uncertainty can impact our interpretations, especially at higher granularity.”

Minor:

1) In the abstract and later in p9 the authors talk about "state-of-the-art" non-linear deformation matrices. This may be confusing for some readers. To me, in brain imaging a matrix is most often a 4x4 affine matrix describing a linear transformation. However, the authors seem to be describing a more complex, non-linear deformation field. Whereas building a deformation matrix (4x4 affine) is not a big challenge, I agree that more sophisticated tools should provide more sophisticated deformation fields. The authors may consider using "deformation field" instead of "deformation matrix", but I leave that to their judgment.

As suggested, we changed the text to “deformation field” where relevant.

2) In the results section, p11, the authors highlight the challenge of segmenting thalamic nuclei or different hippocampal regions, and suggest that this should be simplified by the use of the histological BigBrain data. However, the atlases currently provided in the OSF project do not include these more refined parcellation: there's one single "Thalamus" label, and one single "Hippocampus" label (not really single: left and right). This could be explicitly stated to prevent readers from having too high expectations (although I am certain that those finer parcellations should come in the very close future).

We updated the text to reflect the current state of such parcellations. While subthalamic nuclei are not yet segmented (to our knowledge), one of the present authors has segmented hippocampal subfields (https://osf.io/bqus3/) and we highlight this in the Results (P.11-12):

“Despite MRI acquisitions at high and ultra-high fields reaching submillimeter resolutions with ongoing technical advances, certain brain structures and subregions remain difficult to identify (Kulaga-Yoskovitz et al., 2015; Wisse et al., 2017; Yushkevich et al., 2015). For example, there are challenges in reliably defining the subthalamic nucleus (not yet released for BigBrain) or hippocampal Cornu Ammonis subfields [manual segmentation available on BigBrain, https://osf.io/bqus3/, (DeKraker et al., 2019)]. BigBrain-defined labels can be transformed to a standard imaging space for further investigation. Thus, this approach can support exploration of the functional architecture of histologically-defined regions of interest.”

Author Response:

Reviewer #1 (Public Review):

Overall, the authors have done a nice job covering the relevant literature, presenting a story out of complicated data, and performing many thoughtful analyses.

However, I believe the paper requires quite major revisions.

We thank the reviewer for their encouraging assessment of our manuscript. We are grateful for their valuable and especially detailed feedback that helped us to substantially improve our manuscript.

Major issues:

I do not believe the current results present a clear, comprehensible story about sleep and motor memory consolidation. As presented, sleep predicts an increase in the subsequent learning curve, but there is a negative relationship between learning curve and task proficiency change (which is, as far as I can tell, similar to "memory retention"). This makes it seem as if sleep predicts more forgetting on initial trials within the subsequent block (or worse memory retention) - is this true? Regardless of whether it is statistically true, there appears another story in these data that is being sacrificed to fit a story about sleep. To my eye, the results may first and foremost tell a circadian (rather than sleep) story. Examining the data in Figure 2A and 2B, it appears that every AM learning period has a higher learning curve (slope) than every PM period. While this could, of course, be due to having just slept, the main story gleaned from such a result is not a sleep effect on retention, which has been the emphasis on motor memory consolidation research in the last couple of decades, but on new learning. The fact that this effect appears present in the first session (juggling blocks 1-3 in adolescents and blocks 1-5 in adults) makes this seem the more likely story here, since it has less to do with "preparing one to re-learn" and more to do with just learning and when that learning is optimal. But even if it does not reach statistical significance in the first session alone, it remains a concern and, in my opinion, should be considered a focus in the manuscript unless the authors can devise a reason to definitively rule it out.

Here is how I recommend the authors proceed on this point: include all sessions from all subjects into a mixed effect model, predicting the slope of the learning curve with time of day and age group as fixed effects and subjects as random effects:

learning curve slope ~ AM/PM [AM (0) or PM (1)] + age [adolescent (0) or adult (1)] + (1|subject)

…or something similar with other regressors of interest. If this is significant for AM/PM status, they should re-try the analysis using only the first session. If this is significant, then a sleep-centric story cannot be defended here at all, in my opinion. If it is not (which could simply result from low power, but the authors could decide this), the authors should decide if they think they can rule out circadian effects and proceed accordingly. I should note that, while to many, a sleep story would be more interesting or compelling, that is not my opinion, and I would not solely opt to reject this paper if it centered a time-of-day story instead.

The authors need to work out precisely what is happening in the behavior here, and let the physiology follow that story. They should allow themselves to consider very major revisions (and drop the physiology) if that is most consistent with the data. As presented, I am very unclear of what to take away from the study.

We thank the reviewer for the opportunity to further elaborate on our behavioral results. We agree that the interpretation of the behavior in the complex gross-motor task is not straight forward, which might be partly due to less controllability compared to for example finger-tapping tasks. The reviewer is correct that, initially sleep seems to predict more forgetting on initial trials within the subsequent block given the dip in task proficiency and a resulting increase in steepness of the learning curve after the sleep retention interval. Notably, this dip in performance after sleep has also been reported for finger-tapping tasks (cf. Eichenlaub et al, 2020). The performance dip is also present in the wake first group (Figure 2) after the first interval. This observation suggests that picking up the task again after a period of time comes at a cost. Interestingly, this performance dip is no longer present after the second retention interval indicating that the better the task proficiency the easier it is to pick up juggling again. In other words, juggling has been better consolidated after additional training. Critically, our results show, that participants with higher SO-spindle coupling strength have a lower dip in performance after the retention interval, thus indicating a learning advantage.

Figure 2

(A) Number of successful three-ball cascades (mean ± standard error of the mean [SEM]) of adolescents (circles) for the sleep-first (blue) and wake-first group (green) per juggling block. Grand average learning curve (black lines) as computed in (C) are superimposed. Dashed lines indicate the timing of the respective retention intervals that separate the three performance tests. Note that adolescents improve their juggling performance across the blocks. (B) Same conventions as in (A) but for adults (diamonds). Similar to adolescents, adults improve their juggling performance across the blocks regardless of group.

We discuss the sleep effect on juggling in the discussion section (page 22 – 23, lines 502 – 514):

"How relevant is sleep for real-life gross-motor memory consolidation? We found that sleep impacts the learning curve but did not affect task proficiency in comparison to a wake retention interval (Figure 2DE). Two accounts might explain the absence of a sleep effect on task proficiency. (1) Sleep rather stabilizes than improves gross-motor memory, which is in line with previous gross-motor adaption studies (Bothe et al, 2019; Bothe et al, 2020). (2) Pre-sleep performance is critical for sleep to improve motor skills (Wilhelm et al, 2012). Participants commonly reach asymptotic pre-sleep performance levels in finger tapping tasks, which is most frequently used to probe sleep effects on motor memory. Here we found that using a complex juggling task, participants do not reach asymptotic ceiling performance levels in such a short time. Indeed, the learning progression for the sleep-first and wake-first groups followed a similar trend (Figure 2AB), suggesting that more training and not in particular sleep drove performance gains."

If indeed the authors keep the sleep aspect of this story, here are some comments regarding the physiology. The authors present several nice analyses in Figure 3. However, given the lack of behavioral difference between adolescents and adults (Fig 2D), they combine the groups when investigating behavior-physiology relationships. In some ways, then, Figure 3 has extraneous details to the point of motor learning and retention, and I believe the paper would benefit from more focus. If the authors keep their sleep story, I believe Figure 3 and 4 should be combined and some current figure panels in Figure 3 should be removed or moved to the supplementary information.

We thank the reviewers for their suggestion and we agree that the figures of our manuscript would benefit from more focus. Therefore, we combined Figure 3 and 4 from the original manuscript into a revised Figure 3 in the updated version of the manuscript. In more detail, subpanels that explain our methodological approach can now be found in Figure 3 – figure supplement 1, while the updated Figure 3 now focuses on developmental changes in oscillatory dynamics and SO-spindle coupling strength as well as their relationship to gross-motor learning.

Updated Figure 3:

(A) Left: topographical distribution of the 1/f corrected SO and spindle amplitude as extracted from the oscillatory residual (Figure 3 – figure supplement 1A, right). Note that adolescents and adults both display the expected topographical distribution of more pronounced frontal SO and centro-parietal spindles. Right: single subject data of the oscillatory residual for all subjects with sleep data color coded by age (darker colors indicate older subjects). SO and spindle frequency ranges are indicated by the dashed boxes. Importantly, subjects displayed high inter-individual variability in the sleep spindle range and a gradual spindle frequency increase by age that is critically underestimated by the group average of the oscillatory residuals (Figure 3 – figure supplement 1A, right). (B) Spindle peak locked epoch (NREM3, co-occurrence corrected) grand averages (mean ± SEM) for adolescents (red) and adults (black). Inset depicts the corresponding SO-filtered (2 Hz lowpass) signal. Grey-shaded areas indicate significant clusters. Note, we found no difference in amplitude after normalization. Significant differences are due to more precise SO-spindle coupling in adults. (C) Top: comparison of SO-spindle coupling strength between adolescents and adults. Adults displayed more precise coupling than adolescents in a centro-parietal cluster. T-scores are transformed to z-scores. Asterisks denote cluster-corrected two-sided p < 0.05. Bottom: Exemplary depiction of coupling strength (mean ± SEM) for adolescents (red) and adults (black) with single subject data points. Exemplary single electrode data (bottom) is shown for C4 instead of Cz to visualize the difference. (D) Cluster-corrected correlations between individual coupling strength and overnight task proficiency change (post – pre retention) for adolescents (red, circle) and adults (black, diamond) of the sleep-first group (left, data at C4). Asterisks indicate cluster-corrected two-sided p < 0.05. Grey-shaded area indicates 95% confidence intervals of the trend line. Participants with a more precise SO-spindle coordination show improved task proficiency after sleep. Note that the change in task proficiency was inversely related to the change in learning curve (cf. Figure 2D), indicating that a stronger improvement in task proficiency related to a flattening of the learning curve. Further note that the significant cluster formed over electrodes close to motor areas. (E) Cluster-corrected correlations between individual coupling strength and overnight learning curve change. Same conventions as in (D). Participants with more precise SO-spindle coupling over C4 showed attenuated learning curves after sleep.

and

Figure 3 - figure supplement 1

(A) Left: Z-normalized EEG power spectra (mean ± SEM) for adolescents (red) and adults (black) during NREM sleep in semi-log space. Data is displayed for the representative electrode Cz unless specified otherwise. Note the overall power difference between adolescents and adults due to a broadband shift on the y-axis. Straight black line denotes cluster-corrected significant differences. Middle: 1/f fractal component that underlies the broadband shift. Right: Oscillatory residual after subtracting the fractal component (A, middle) from the power spectrum (A, left). Both groups show clear delineated peaks in the SO (< 2 Hz) and spindle range (11 – 16 Hz) establishing the presence of the cardinal sleep oscillations in the signal. (B) Top: Spindle frequency peak development based on the oscillatory residuals. Spindle frequency is faster at all but occipital electrodes in adults than in adolescents. T-scores are transformed to z-scores. Asterisks denote cluster-corrected two-sided p < 0.05. Bottom: Exemplary depiction of the spindle frequency (mean ± SEM) for adolescents (red) and adults (black) with single subject data points at Cz. (C) SO-spindle co-occurrence rate (mean ± SEM) for adolescents (red) and adults (black) during NREM2 and NREM3 sleep. Event co-occurrence is higher in NREM3 (F(1, 51) = 1209.09, p < 0.001, partial eta² = 0.96) as well as in adults (F(1, 51) = 11.35, p = 0.001, partial eta² = 0.18). (D) Histogram of co-occurring SO-spindle events in NREM2 (blue) and NREM3 (purple) collapsed across all subjects and electrodes. Note the low co-occurring event count in NREM2 sleep. (E) Single subject (top) and group averages (bottom, mean ± SEM) for adolescents (red) and adults (black) of individually detected, for SO co-occurrence-corrected sleep spindles in NREM3. Spindles were detected based on the information of the oscillatory residual. Note the underlying SO-component (grey) in the spindle detection for single subject data and group averages indicating a spindle amplitude modulation depending on SO-phase. (F) Grand average time frequency plots (-2 to -1.5s baseline-corrected) of SO-trough-locked segments (corrected for spindle co-occurrence) in NREM3 for adolescents (left) and adults (right). Schematic SO is plotted superimposed in grey. Note the alternating power pattern in the spindle frequency range, showing that SO-phase modulates spindle activity in both age groups.

Why did the authors use Spearman rather than Pearson correlations in Figure 4? Was it to reduce the influence of the outlier subject? They should minimally clarify and justify this, since it is less conventional in this line of research. And it would be useful to know if the relationship is significant with Pearson correlations when robust regression is applied. I see the authors are using MATLAB, and the robustfit toolbox (https://www.mathworks.com/help/stats/robustfit.html) is a simple way to address this issue.

We thank the reviewers for their suggestion. We agree that when inspecting the scatter plots it looks like that the correlations could be severely influenced by two outliers in the adult group. Because this is an important matter, we recalculated all previously reported correlations without the two outliers (Figure R4, left column) and followed the reviewer’s suggestion to also compute robust regression (Figure R4, right column) and found no substantial deviation from our original results.

In more detail, increase in task proficiency resulted in flattening of the learning curve when removing outliers (Figure R4A, rhos = -0.70, p < 0.001) and when applying robust regression analysis (Figure R4B, b = -0.30, t(67) = -10.89, rho = -0.80, p < 0.001). Likewise, higher coupling strength still predicted better task proficiency (mean rho = 0.35, p = 0.029, cluster-corrected) and flatter learning curves after sleep (rho = -0.44, p = 0.047, cluster-corrected) when removing the outliers (Figure R4CE) and when calculating robust regression (Figure R4DF, task proficiency: b = 82.32, t(40) = 3.12, rho = 0.45, p = 0.003; learning curve: b = -26.84, t(40) = -2.96, rho = -0.43, p = 0.005). Furthermore, we calculated spearman rank correlations and cluster-corrected spearman rank correlations in our original manuscript, to mitigate the impact of outliers, even though Pearson correlations are more widely used in the field. Therefore, we still report spearman rank correlations for single electrodes instead of robust correlations as it is more consistent with the cluster-correlation analyses.

We now use robust trend lines instead of linear trend lines in our scatter plots. Further, we added the correlations without outliers (Figure R4ACE) to the supplements as Figure 2 – figure supplement 1D and Figure 3 – figure supplement 2 FG. These additional analyses are now reported in the results section of the revised manuscript (page 9, lines 186 – 191):

"[…] we confirmed a strong negative correlation between the change (post retention values – pre retention values) in task proficiency and the change in learning curve after the retention interval (Figure 2F; rhos = -0.71, p < 0.001), which also remained strong after outlier removal (Figure 2 – figure supplement 1D). This result indicates that participants who consolidate their juggling performance after a retention interval show slower gains in performance."

And (page 16, lines 343 – 346):

"[…] Furthermore, our results remained consistent when including coupled spindle events in NREM2 (Figure 3 – figure supplement 2E) and after outlier removal (Figure 3 – figure supplement 2FG)."

Furthermore, we now state that we specifically utilized spearman rank correlations to mitigate the impact of outliers in our analyses in the method section (page 35, lines 808 – 813)::

"For correlational analyses we utilized spearman rank correlations (rhos; Figure 2F & Figure 3DE) to mitigate the impact of possible outliers as well as cluster-corrected spearman rank correlations by transforming the correlation coefficients to t-values (p < 0.05) and clustering in the space domain (Figure 3DE). Linear trend lines were calculated using robust regression."

Figure R4

(A) Spearman rank correlation between task proficiency change and learning curve change collapsed across adolescents (red dot) and adults (black diamonds) after removing two outlier subjects in the adult age group. Grey-shaded area indicates 95% confidence intervals of the robust trend line. (B) Robust regression of task proficiency change and learning curve change of the original sample. (C) Cluster-corrected correlations (right) between individual coupling strength and overnight task proficiency change (post – pre retention) after outlier removal (left, spearman correlation at C4, uncorrected). Asterisks indicate cluster-corrected two-sided p < 0.05. (D) Robust regression of coupling strength at C4 and task proficiency of the original sample. (E) Same conventions as in (C) but for overnight learning curve change. (F) Same conventions as in (D) but for overnight learning curve change.

Additionally, with only a single night of recording data, it is impossible to disentangle possible trait-based sleep characteristics (e.g., Subject 1 has high SO-spindle coupling in general and retains motor memories well, but these are independent of each other) from a specific, state-based account (e.g., Subject 1's high SO-spindle coupling on night 1 specifically led to their improved retention or change in learning, etc., and this is unrelated to their general SO-spindle coupling or motor performance abilities). Clearly, many studies face this limitation, but this should be acknowledged.

We thank the reviewers for their important remark. We agree that it is impossible to make a sound statement about whether our reported correlations represent trait- or state-based aspects of the sleep and learning relationship with the data that we have reported in the manuscript. However, while we are lacking a proper baseline condition without any task engagement, we still recorded polysomnography for all subjects during an adaptation night. Given the expected pronounced differences in sleep architecture between the adaptation nights and learning nights (see Table R3 for an overview collapsed across both age groups), we initially refrained from entering data from the adaptation nights into our original analyses, but we now fully report the data below. Note that the differences are driven by the adaptation night, where subjects first have to adjust to sleeping with attached EEG electrodes in a sleep laboratory.

Table R3. Sleep architecture (mean ± standard deviation) for the adaptation and learning night collapsed across both age groups. Nights were compared using paired t-tests

To further clarify whether subjects with high coupling strength have a motor learning advantage (i.e. trait-effect) or a learning induced enhancement of coupling strength is indicative for improved overnight memory change (i.e. state-effect), we ran additional analyses using the data from the adaptation night. Note that the coupling strength metric was not impacted by differences in event number and our correlations with behavior were not influenced by sleep architecture (please refer to our answer of issue #7 for the results).Therefore, we considered it appropriate to also utilize data from the adaptation night.

First, we correlated SO-spindle coupling strength obtained from the adaptation night with the coupling strength in the learning night. We found that overall, coupling strength is highly correlated between the two measurements (mean rho across all channels = 0.55, Figure R5A), supporting the notion that coupling strength remains rather stable within the individual (i.e. trait), similar to what has been reported about the stable nature of sleep spindles as a “neural finger-print” (De Gennaro & Ferrara, 2003; De Gennaro et al, 2005; Purcell et al, 2017).

To investigate a possible state-effect for coupling strength and motor learning, we calculated the difference in coupling strength between the two nights (learning night – adaptation night) and correlated these values with the overnight change in task proficiency and learning curve. We identified no significant correlations with a learning induced coupling strength change; neither for task proficiency nor learning curve change (Figure R5B). Note that there was a positive correlation of coupling strength change with overnight task proficiency change at Cz (Figure R5B, left), however it did not survive cluster-corrected correlational analysis (rhos = 0.34, p = 0.15). Combined, these results favor the conclusion that our correlations between coupling strength and learning rather reflect a trait-like relationship than a state-like relationship. This is in line with the interpretation of our previous studies that SO-spindle coupling strength reflects the efficiency and integrity of the neuronal pathway between neocortex and hippocampus that is paramount for memory networks and the information transfer during sleep (Hahn et al, 2020; Helfrich et al, 2019; Helfrich et al, 2018; Winer et al, 2019). For a comprehensive review please see Helfrich et al (2021), which argued that SO-spindle coupling predicts the integrity of memory pathways and therefore correlates with various metrics of behavioral performance or structural integrity.

Figure R5

(A) Topographical plot of spearman rank correlations of coupling strength in the adaptation night and learning night across all subjects. Overall coupling strength was highly correlated between the two measurements. (B) Cluster-corrected correlation between learning induced coupling strength changes (learning night – adaptation night) and overnight change in task proficiency (left) as well as learning curve (right). We found no significant clusters, although correlations showed similar trends as our original analyses, with more learning induced changes in coupling strength resulting in better overnight task proficiency and flattened learning curves.

We have now added the additional state-trait analyses (Figure R5) to the updated manuscript as Figure 3 – figure supplement 2HI and report them in the results section (page 17, lines 361 – 375):

"Finally, we investigated whether subjects with high coupling strength have a gross-motor learning advantage (i.e. trait-effect) or a learning induced enhancement of coupling strength is indicative for improved overnight memory change (i.e. state-effect). First, we correlated SO-spindle coupling strength obtained from the adaptation night with the coupling strength in the learning night. We found that overall, coupling strength is highly correlated between the two measurements (mean rho across all channels = 0.55, Figure 3 – figure supplement 2H), supporting the notion that coupling strength remains rather stable within the individual (i.e. trait). Second, we calculated the difference in coupling strength between the learning night and the adaptation night to investigate a possible state-effect. We found no significant cluster-corrected correlations between coupling strength change and task proficiency- as well as learning curve change (Figure 3 – figure supplement 2I).

Collectively, these results indicate the regionally specific SO-spindle coupling over central EEG sensors encompassing sensorimotor areas precisely indexes learning of a challenging motor task."

We further refer to these new results in the discussion section (page 23, lines 521 – 528):

"Moreover, we found that SO-spindle coupling strength remains remarkably stable between two nights, which also explains why a learning-induced change in coupling strength did not relate to behavior (Figure 3 – figure supplement 2I). Thus, our results primarily suggest that strength of SO-spindle coupling correlates with the ability to learn (trait), but does not solely convey the recently learned information. This set of findings is in line with recent ideas that strong coupling indexes individuals with highly efficient subcortical-cortical network communication (Helfrich et al, 2021)."

Additionally, we now provide descriptive data of the adaptation and learning night (Table R3) in the Supplementary file – table 1 and explicitly mention the adaptation night in the results section, which was previously only mentioned in the method section(page 6, lines 101 – 105):.

"Polysomnography (PSG) was recorded during an adaptation night and during the respective sleep retention interval (i.e. learning night) except for the adult wake-first group (for sleep architecture descriptive parameters of the adaptation night and learning night as well as for adolescents and adults see Supplementary file – table 1 & 2)."

Reviewer #2 (Public Review):

In this study Hahn and colleagues investigate the role of Slow-oscillation spindle coupling for motor memory consolidation and the impact of brain maturation on these interactions. The authors employed a real-life gross-motor task, where adolescents and adults learned to juggle. They demonstrate that during post-learning sleep SO-spindles are stronger coupled in adults as compared to adolescents. The authors further show, that the strength of SO-spindle coupling correlates with overnight changes in the learning curve and task proficiency, indicating a role of SO-spindle coupling in motor memory consolidation.

Overall, the topic and the results of the present study are interesting and timely. The authors employed state of the art analyse carefully taking the general variability of oscillatory features into account. It also has to be acknowledged that the authors moved away from using rather artificial lab-tasks to study the consolidation of motor memories (as it is standard in the field), adding ecological validity to their findings. However, some features of their analyses need further clarification.

We thank the reviewer for their positive assessment of our manuscript. Incorporating the encouraging and helpful feedback, we believe that we substantially improved the clarity and robustness of our analyses.

1) Supporting and extending previous work of the authors (Hahn et al, 2020), SO-spindle coupling over centro-parietal areas was stronger in adults as compared to adolescents. Despite these differences in the EEG results the authors collapsed the data of adults and adolescents for their correlational analyses (Fig. 4a and 4b). Why would the authors think that this procedure is viable (also given the fact that different EEG systems were used to record the data)?

We thank the reviewers for the opportunity to clarify why we think it is viable to collapse the data of adolescents and adults for our correlational analyses. In the following we split our answers based on the two points raised by the reviewers: (1) electrophysiological differences (i.e. coupling strength) between the groups and (2) potential signal differences due to different EEG systems.

1. Electrophysiological differences

Upon inspecting the original Figure 4, it is apparent that the coupling strength of the combined sample does not form isolated clusters for each age group. In other words, while adult coupling strength is on the higher and adolescent coupling on the lower end due to the developmental increase in coupling strength we reported in the original Figure 3F, both samples overlap forming a linear trend. Second, when running the correlational analyses between coupling strength and task proficiency as well as learning curve separately for each age group, we found that they follow the same direction (Figure R3). Adolescents with higher coupling strength show better task proficiency (Figure R3A, rhos = 0.66, p = 0.005). This effect was also present when using robust regression (b = 109.97, t(15)=3.13, rho = 0.63, p = 0.007). Like adolescents, adults with higher coupling strength at C4 displayed better task proficiency after sleep (Figure R3B, rhos = 0.39, p = 0.053). This relationship was stronger when using robust regression (b = 151.36, t(23)=3.17, rho =0.56, p = 0.004). For learning curves, we found the expected negative correlation at C4 for adolescents (Figure R3C, rhos = -0.57, p = 0.020) and adults (Figure R3D, rhos = -0.44, p = 0.031). Results were comparable when using robust regression (adolescents: b = -59.58, t(15) = -2.94, rho = -0.60, p = 0.010; adults: b = -21.99, t(23 )= -1.71, rho = -0.37, p = 0.101).

Taken together, these results demonstrate that adolescents and adults show the effects and the same direction at the same electrode, thus, making it highly unlikely that our results are just by chance and that our initial correlation analyses are just driven by one group.

Additionally, we already controlled for age in our original analyses using partial correlations (also refer to our answer to issue #6). Hence, our additional analyses provide additional support that it is viable to collapse the analyses across both age groups even though they differ in coupling strength.

1. Different EEG-systems

   The reviewers also raise the question whether our analyses might be impacted by the different EEG systems we used to record our data. This is an important concern especially when considering that cross-frequency coupling analyses can be severely confounded by differences in signal properties (Aru et al, 2015). In our sample, the strongest impact factor on signal properties is most likely age, given the broadband power differences in the power spectrum we found between the groups (original Figure 3A). Importantly, we also found a similar systematic power difference in our longitudinal study using the same ambulatory EEG system for both data recordings (Hahn et al, 2020). This is in line with numerous other studies demonstrating age related EEG power changes in broadband- as well as SO and sleep spindle frequency ranges (Campbell & Feinberg, 2016; Feinberg & Campbell, 2013; Helfrich et al, 2018; Kurth et al, 2010; Muehlroth et al, 2019; Muehlroth & Werkle-Bergner, 2020; Purcell et al, 2017). Therefore, we already had to take differences in signal property into account for our cross-frequency analyses. Regardless whether the underlying cause is an age difference or different signal-to-noise ratios of different EEG systems.

To mitigate confounds in the signal, we used a data-driven and individualized approach detecting SO and sleep spindle events based on individualized frequency bands and a 75-percentile amplitude criterion relative to the underlying signal. Additionally we z-normalized all spindle events prior to the cross-frequency coupling analyses (Figure R3E). We found no amplitude differences around the spindle peak (point of SO-phase readout) between adolescents that were recorded with an ambulatory amplifier system (alphatrace) and adults that were recorded with a stationary amplifier system (neuroscan) using cluster-based random permutation testing. This was also the case for the SO-filtered (< 2 Hz) signal (Figure R3E, inset). Critically, the significant differences in amplitude from -1.4 to -0.8 s (p = 0.023, d = -0.73) and 0.4 to 1.5 s (p < 0.001, d = 1.1) are not caused by age related differences in power or different EEG-systems but instead by the increased coupling strength (i.e. higher coupling precision of spindles to SOs) in adults giving rise to a more pronounced SO-wave shape when averaging across spindle peak locked epochs.

Consequently, our analysis pipeline already controlled for possible differences in signal property introduced through different amplifier systems. Nonetheless, we also wanted to directly compare the signal-to-noise ratio of the ambulatory and stationary amplifier systems. However, we only obtained data from both amplifier systems in the adult sleep first group, because we recorded EEG during the juggling learning phase with the ambulatory system in addition to the PSG with the stationary system. First, we computed the power spectra in the 1 to 49 Hz frequency range during the juggling learning phase (ambulatory) and during quiet wakefulness (stationary) for every subject in the adult sleep first group in 10-seconds segments. Next, we computed the signal-to-noise ratio (mean/standard deviation) of the power spectra per frequency across all segments. We only found a small negative cluster from 21.9 to 22.5 Hz (p = 0.042, d = 0.53; Figure R3F), which did not pertain our frequency-bands of interest. Critically, the signal-to-noise ratio of both amplifiers converged in the upper frequency bands approaching the noise floor, therefore, strongly supporting the notion that both systems in fact provided highly comparable estimates.

In conclusion, both age groups display highly similar effects and direction when correlating coupling strength with behavior. Further, after individualization and normalization the analytical signal, we found no differences in signal properties that would confound the cross-frequency analysis. Lastly, we did not find systematic differences in signal-to-noise ratio between the different EEG-systems. Thus, we believe it is justified to collapse the data across all participants for the correlational analyses, as it combines both, the developmental aspect of enhanced coupling precision from adolescence to adulthood and the behavioral relevance for motor learning which we deem a critical research advance from our previous study.

Figure R3

(A) Cluster-corrected correlations (right) between individual coupling strength and overnight task proficiency change (post – pre retention) for adolescents of the sleep-first group (left, spearman correlation at C4, uncorrected). Asterisks indicate cluster-corrected two-sided p < 0.05. Grey-shaded area indicates 95% confidence intervals of the robust trend line. Participants with a more precise SO-spindle coordination show improved task proficiency after sleep. (B) Cluster-corrected correlation of coupling strength and overnight task proficiency change) for adults. Same conventions as in (A). Similar trend of higher coupling strength predicting better task proficiency after sleep (C) Cluster-corrected correlation of coupling strength and overnight learning curve change for adolescents. Same conventions as in (A). Higher coupling strength related to a flatter learning curve after sleep. (D) Cluster-corrected correlation of coupling strength and overnight learning curve change for adults. Same conventions as in (A). Higher coupling strength related to a flatter learning curve after sleep. (E) Spindle peak locked epoch (NREM3, co-occurrence corrected) grand averages (mean ± SEM) for adolescents (red) and adults (black). Inset depicts the corresponding SO-filtered (2 Hz lowpass) signal. Black lines indicate significant clusters. Note, we found no difference in amplitude after normalization. Significant differences are due to more precise SO-spindle coupling in adults. Spindle frequency is blurred due to individualized spindle detection. (F) Signal-to-noise ratio for the stationary EEG amplifier (green) during quiet wakefulness and for the ambulatory EEG amplifier (purple) during juggling training. Grey shaded area denotes cluster-corrected p < 0.05. Note that signal-to-noise ratio converges in the higher frequency ranges.

We have now added Figure R3E as Figure 3B to the revised version of the manuscript to demonstrate that there were no systematic differences between the two age groups in the analytical signal due to the expected age related power differences or EEG-systems. Specifically, we now state in the results section (page 13 – 14, lines 282 – 294):

"We assessed the cross frequency coupling based on z-normalized spindle epochs (Figure 3B) to alleviate potential power differences due to age (Figure 3 – figure supplement 1A) or different EEG-amplifier systems that could potentially confound our analyses (Aru et al, 2015). Importantly, we found no amplitude differences around the spindle peak (point of SO-phase readout) between adolescents and adults using cluster-based random permutation testing (Figure 3B), indicating an unbiased analytical signal. This was also the case for the SO-filtered (< 2 Hz) signal (Figure 3B, inset). Critically, the significant differences in amplitude from -1.4 to -0.8 s (p = 0.023, d = -0.73) and 0.4 to 1.5 s (p < 0.001, d = 1.1) are not caused by age related differences in power or different EEG-systems but instead by the increased coupling strength (i.e. higher coupling precision of spindles to SOs) in adults giving rise to a more pronounced SO-wave shape when averaging across spindle peak locked epochs."

Further, we added the correlational analyses that we computed separately for the age groups (Figure R3A-D) to the revised manuscript (Figure 3 – figure supplement 2CD) as they further substantiate our claims about the relationship between SO-spindle coupling and gross-motor learning.

We now refer to these analyses in the results section (page 16, lines 338 – 343):

"Critically, when computing the correlational analyses separately for adolescents and adults, we identified highly similar effects at electrode C4 for task proficiency (Figure 3 – figure supplement 2C) and learning curve (Figure 3 – figure supplement 2D) in each group. These complementary results demonstrate that coupling strength predicts gross-motor learning dynamics in both, adolescents as well as adults, and further show that this effect is not solely driven by one group."

2) The authors might want to explicitly show that the reported correlations (with regards to both learning curve and task proficiency change) are not driven by any outliers.

We thank the reviewers for their suggestion. We agree that when inspecting the scatter plots it looks like that the correlations could be severely influenced by two outliers in the adult group. Because this is an important matter, we recalculated all previously reported correlations without the two outliers (Figure R4, left column) and followed the reviewer’s suggestion to also compute robust regression (Figure R4, right column) and found no substantial deviation from our original results.

In more detail, increase in task proficiency resulted in flattening of the learning curve when removing outliers (Figure R4A, rhos = -0.70, p < 0.001) and when applying robust regression analysis (Figure R4B, b = -0.30, t(67) = -10.89, rho = -0.80, p < 0.001). Likewise, higher coupling strength still predicted better task proficiency (mean rho = 0.35, p = 0.029, cluster-corrected) and flatter learning curves after sleep (rho = -0.44, p = 0.047, cluster-corrected) when removing the outliers (Figure R4CE) and when calculating robust regression (Figure R4DF, task proficiency: b = 82.32, t(40) = 3.12, rho = 0.45, p = 0.003; learning curve: b = -26.84, t(40) = -2.96, rho = -0.43, p = 0.005). Furthermore, we calculated spearman rank correlations and cluster-corrected spearman rank correlations in our original manuscript, to mitigate the impact of outliers, even though Pearson correlations are more widely used in the field. Therefore, we still report spearman rank correlations for single electrodes instead of robust correlations as it is more consistent with the cluster-correlation analyses.

We now use robust trend lines instead of linear trend lines in our scatter plots. Further, we added the correlations without outliers (Figure R4ACE) to the supplements as Figure 2 – figure supplement 1D and Figure 3 – figure supplement 2 FG. These additional analyses are now reported in the results section of the revised manuscript (page 9, lines 186 – 191):

"[…] we confirmed a strong negative correlation between the change (post retention values – pre retention values) in task proficiency and the change in learning curve after the retention interval (Figure 2F; rhos = -0.71, p < 0.001), which also remained strong after outlier removal (Figure 2 – figure supplement 1D). This result indicates that participants who consolidate their juggling performance after a retention interval show slower gains in performance."

And (page 16, lines 343 – 346):

"[…] Furthermore, our results remained consistent when including coupled spindle events in NREM2 (Figure 3 – figure supplement 2E) and after outlier removal (Figure 3 – figure supplement 2FG)."

Furthermore, we now state that we specifically utilized spearman rank correlations to mitigate the impact of outliers in our analyses in the method section (page 35, lines 808 – 813)::

"For correlational analyses we utilized spearman rank correlations (rhos; Figure 2F & Figure 3DE) to mitigate the impact of possible outliers as well as cluster-corrected spearman rank correlations by transforming the correlation coefficients to t-values (p < 0.05) and clustering in the space domain (Figure 3DE). Linear trend lines were calculated using robust regression."

Figure R4:

(A) Spearman rank correlation between task proficiency change and learning curve change collapsed across adolescents (red dot) and adults (black diamonds) after removing two outlier subjects in the adult age group. Grey-shaded area indicates 95% confidence intervals of the robust trend line. (B) Robust regression of task proficiency change and learning curve change of the original sample. (C) Cluster-corrected correlations (right) between individual coupling strength and overnight task proficiency change (post – pre retention) after outlier removal (left, spearman correlation at C4, uncorrected). Asterisks indicate cluster-corrected two-sided p < 0.05. (D) Robust regression of coupling strength at C4 and task proficiency of the original sample. (E) Same conventions as in (C) but for overnight learning curve change. (F) Same conventions as in (D) but for overnight learning curve change.

3) The sleep data of all participants (thus from both sleep first and wake first) were used to determine the features of SO-spindle coupling in adolescents and adults. Were there any differences between groups (sleep first vs. wake first)? This might be in interesting in general but especially because only data of the sleep first group entered the subsequent correlational analyses.

We thank the reviewers for their remark. We agree that adding additional information about possible differences between the sleep first and wake first groups would allow for a more comprehensive assessment of the reported data. We did not explain our reasoning to include only the sleep first groups for the correlation analyses clearly enough in the original manuscript. Unfortunately, we can only report data for the adolescents in our sample, because we did not record polysomnography (PSG) for the adult wake first group. This is also one of the two reasons why we focused on the sleep first groups for our correlational analyses.

Adolescents in the sleep first group did not differ from adolescents in the wake first group in terms of sleep architecture (except REM (%), which did not correlate with behavior [task proficiency: rho = -0.17, p = 0.28; learning curve: -0.02, p = 0.90]) as well as SO and sleep spindle event descriptive measures (see Table R2). Importantly, we found no differences in coupling strength between the two groups (Figure R2A).

Table R2. Summary of sleep architecture and SO/spindle event descriptive measures (at electrode C4) of adolescents in the sleep first and wake first group (mean ± standard deviation). Independent t-tests were used for comparisons

The second reason why we focused our analyses on sleep first was that adolescents in the wake first group had higher task proficiency after the sleep retention interval than the sleep first group (Figure R2A; t(23) = -2.24, p = 0.034). This difference in performance is directly explained by the additional juggling test that the wake first group performed at the time point of their learning night, which should be considered as additional training. Therefore, we excluded the wake first group from our correlational analyses because sleep and wake first group are not comparable in terms of juggling training during the night when we assessed SO-spindle coupling strength.

Figure R2

(A) Comparison of SO-spindle coupling strength in the adolescent sleep first (blue) and wake first (green) group using cluster-based random permutation testing (Monte-Carlo method, cluster alpha 0.05, max size criterion, 1000 iterations, critical alpha level 0.05, two-sided). Left: exemplary depiction of coupling strength at electrode C4 (mean ± SEM). Right: z-transformed t-values plotted for all electrodes obtained from the cluster test. No significant clusters emerged. (B) Comparison of task proficiency between sleep first and wake first group after the sleep retention interval (mean ± SEM). Adolescents in the wake first group had higher task proficiency given the additional juggling performance test, which also reflects additional training.

These additional analyses (Figure R2) and the summary statistics of sleep architecture and SO/spindle event descriptives of adolescents in the sleep first and wake first group (Table R2), are now reported in the revised version of the manuscript as Figure 3 – figure supplement 2AB and Supplementary file – table 7. We now explicitly explain our rationale of why we only considered participants in the sleep first group for our correlational analyses in the results section (page 6, lines 101 – 105):

"Polysomnography (PSG) was recorded during an adaptation night and during the respective sleep retention interval (i.e. learning night) except for the adult wake-first group (for sleep architecture descriptive parameters of the adaptation night and learning night as well as for adolescents and adults see Supplementary file – table 1 & 2)"

And (page 15, lines 311 – 320):

"[…] Furthermore, given that we only recorded polysomnography for the adults in the sleep first group and that adolescents in the wake first group showed enhanced task proficiency at the time point of the sleep retention interval due to additional training (Figure 3 – figure supplement 2A), we only considered adolescents and adults of the sleep-first group to ensure a similar level of juggling experience adolescents and adults of the sleep-first group to ensure a similar level of juggling experience (for summary statistics of sleep architecture and SO and spindle events of subjects that entered the correlational analyses see Supplementary file – table 6). Notably, we found no differences in electrophysiological parameters (i.e. coupling strength, event detection) between the adolescents of the wake first and sleep first group (Figure 3 – figure supplement 2B & Supplementary file – table 7)."

4) To allow a more comprehensive assessment of the underlying data information with regards to general sleep descriptives (minutes, per cent of time spent in different sleep stages, overall sleep time etc.) as well as related to SOs, spindles and coupled events (e.g. number, density etc.) would be needed.

We agree with the reviewers that additional information about sleep architecture and SO as well as sleep spindle characteristics are needed for a more comprehensive assessment of our data. We now added summary tables for sleep architecture and SO/spindle event descriptive measures for the whole sample (Table R4) and for the sleep first groups that we used for our correlational analyses (Table R5) to the supplementary material in the updated manuscript. It is important to note, that due to the longer sleep opportunity of adolescents that we provided to accommodate the overall higher sleep need in younger participants, adolescents and adults differed in most general sleep architecture markers and SO as well as sleep spindle descriptive measures. In addition, changes in sleep architecture are prominent during the maturational phase from adolescence to adulthood, which might introduce additional variance between the two age groups.

Table R4. Summary of sleep architecture and SO/spindle event descriptive measures (at electrode C4) of adolescents and adults across the whole sample (mean ± standard deviation) in the learning night. Independent t-tests were used for comparisons

Table R5. Summary of sleep architecture and SO/spindle event descriptive measures (at electrode C4) of adolescents and adults in the sleep first group (mean ± standard deviation) in the learning night. Independent t-tests were used for comparisons

In order to ensure that our correlational analyses are not driven by these systematic differences between the two age groups, we used cluster-corrected partial correlations to control for sleep architecture markers (Figure R7) and SO/spindle descriptive measurements (Figure R8A). Critically, none of these possible confounders changed the pattern of our initial correlational analyses of coupling strength and task proficiency/learning curve. Additionally, we also controlled for differences in spindle event number by using a bootstrapped resampling approach. We randomly drew 200 spindle events in 100 iterations and subsequently recalculated the coupling strength for each subject. We found that resampled values and our original observation of coupling strength are almost perfectly correlated, indicating that differences in event number are unlikely to have an impact on coupling strength as long as there are at least 200 events (Figure R8B). Combined these analyses demonstrate that our correlations between coupling strength and behavior are not influenced by the reported differences in sleep architecture and SO/spindle descriptive measures.

Figure 7R

Summary of cluster-corrected partial correlations of coupling strength with task proficiency (left) and learning curve (right) controlling for possible confounding factors. Asterisks indicate location of the detected cluster. The pattern of initial results remained highly stable.

Figure R8

(A) Summary of cluster-corrected partial correlations of coupling strength with task proficiency (left) and learning curve (right) controlling SO/spindle descriptive measures at critical electrode C4. Asterisks indicate location of the detected cluster. The pattern of initial results remained highly stable. (B) Spearman correlation between resampled coupling strength (N = 200, 100 iterations) and original observation of coupling strength for adolescents (red circles) and adults (black diamonds), indicating that coupling strength is not influenced by spindle event number if at least 200 events are present. Grey-shaded area indicates 95% confidence intervals of the robust trend line.

We now provide general sleep descriptives (Table R4 & R5) in the revised version of the manuscript as Supplementary file – table 2 & table 6. These data are referred to in the results section (page 6, lines 101 – 105):

"Polysomnography (PSG) was recorded during an adaptation night and during the respective sleep retention interval (i.e. learning night) except for the adult wake-first group (for sleep architecture descriptive parameters of the adaptation night and learning night as well as for adolescents and adults see Supplementary file – table 1 & 2)."

And (page 15, lines 311 – 318):

"Furthermore, given that we only recorded polysomnography for the adults in the sleep first group and that adolescents in the wake first group showed enhanced task proficiency at the time point of the sleep retention interval due to additional training (Figure 3 – figure supplement 2A), we only considered adolescents and adults of the sleep-first group to ensure a similar level of juggling experience (for summary statistics of sleep architecture and SO and spindle events of subjects that entered the correlational analyses see Supplementary file – table 6)."

The additional control analyses (Figure R7 & R8) are also now added to the revised manuscript as Figure 3 – figure supplement 3 & 4 in the results section (page 16, lines 356 – 360):

"For a summary of the reported cluster-corrected partial correlations as well as analyses controlling for differences in sleep architecture see Figure 3 – figure supplement 3. Further, we also confirmed that our correlations are not influenced by individual differences in SO and spindle event parameters (Figure 3 – figure supplement 4)."

5) The authors used a partial correlations to rule out that age drove the relationship between coupling strength, learning curve and task proficiency. It seems like this analysis was done specifically for electrode C4, after having already established that coupling strength at electrode C4 correlates in general with changes in the learning curve and task proficiency. I think the claim that results were not driven by age as confounding factor would be stronger if the authors used a cluster-corrected partial correlation in the first place (just as in the main analysis).

The reviewers are correct that initially we only conducted the partial correlation for electrode C4. Following the reviewers suggestion we now additionally computed cluster-corrected partial correlations similar to our main analysis. Like in our original analyses, we found a significant positive central cluster (Figure R6A, mean rho = 0.40, p = 0.017) showing that higher coupling strength related to better task proficiency after sleep and a negative cluster-corrected correlation at C4 showing that higher coupling strength was related to flatter learning curves after sleep (Figure R6B, rho = -0.47, p = 0.049) also when controlling for age.

Figure R6

(A) Cluster-corrected partial correlation of individual coupling strength in the learning night and overnight change in task proficiency (post – pre retention) collapsed across adolescents and adults, controlling for age. Asterisks indicate cluster-corrected two-sided p < 0.05. A similar significant cluster to the original analysis (Figure 4A) emerged comprising electrodes Cz and C4. (B) Same conventions as in A. Like in the original analysis (Figure 4B) a negative correlation between coupling strength at C4 and learning curve change survived cluster-corrected partial correlations when controlling for age.

We now always report cluster-corrected partial correlations when controlling for possible confounding variables in the updated version of the manuscript (also see answer to issue #7). A summary of all computed partial correlations including Figure R6 can now be found as Figure 3 – figure supplement 3 & 4 in the revised manuscript.

Specifically we now state in the results section (page 16 – 17, lines 347 – 360):

"To rule out age as a confounding factor that could drive the relationship between coupling strength, learning curve and task proficiency in the mixed sample, we used cluster-corrected partial correlations to confirm their independence of age differences (task proficiency: mean rho = 0.40, p = 0.017; learning curve: rhos = -0.47, p = 0.049). Additionally, given that we found that juggling performance could underlie a circadian modulation we controlled for individual differences in alertness between subjects due to having just slept. We partialed out the mean PVT reaction time before the juggling performance test after sleep from the original analyses and found that our results remained stable (task proficiency: mean rho = 0.37, p = 0.025; learning curve: rhos = -0.49, p = 0.040). For a summary of the reported cluster-corrected partial correlations as well as analyses controlling for differences in sleep architecture see Figure 3 – figure supplement 3. Further, we also confirmed that our correlations are not influenced by individual differences in SO and spindle event parameters (Figure 3 – figure supplement 4)."

And in the methods section (page 35, lines 813 – 814):

"To control for possible confounding factors we computed cluster-corrected partial rank correlations (Figure 3 – figure supplement 3 and 4)."

References

Aru, J., Aru, J., Priesemann, V., Wibral, M., Lana, L., Pipa, G., Singer, W. & Vicente, R. (2015) Untangling cross-frequency coupling in neuroscience. Curr Opin Neurobiol, 31, 51-61.

Bothe, K., Hirschauer, F., Wiesinger, H. P., Edfelder, J., Gruber, G., Birklbauer, J. & Hoedlmoser, K. (2019) The impact of sleep on complex gross-motor adaptation in adolescents. Journal of Sleep Research, 28(4).

Bothe, K., Hirschauer, F., Wiesinger, H. P., Edfelder, J. M., Gruber, G., Hoedlmoser, K. & Birklbauer, J. (2020) Gross motor adaptation benefits from sleep after training. J Sleep Res, 29(5), e12961.

Campbell, I. G. & Feinberg, I. (2016) Maturational Patterns of Sigma Frequency Power Across Childhood and Adolescence: A Longitudinal Study. Sleep, 39(1), 193-201.

Dayan, E. & Cohen, L. G. (2011) Neuroplasticity subserving motor skill learning. Neuron, 72(3), 443-54. De Gennaro, L. & Ferrara, M. (2003) Sleep spindles: an overview. Sleep Med Rev, 7(5), 423-40.

De Gennaro, L., Ferrara, M., Vecchio, F., Curcio, G. & Bertini, M. (2005) An electroencephalographic fingerprint of human sleep. Neuroimage, 26(1), 114-22.

Dinges, D. F., Pack, F., Williams, K., Gillen, K. A., Powell, J. W., Ott, G. E., Aptowicz, C. & Pack, A. I. (1997) Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4-5 hours per night. Sleep, 20(4), 267-77.

Dinges, D. F. & Powell, J. W. (1985) Microcomputer Analyses of Performance on a Portable, Simple Visual Rt Task during Sustained Operations. Behavior Research Methods Instruments & Computers, 17(6), 652-655.

Eichenlaub, J. B., Biswal, S., Peled, N., Rivilis, N., Golby, A. J., Lee, J. W., Westover, M. B., Halgren, E. & Cash, S. S. (2020) Reactivation of Motor-Related Gamma Activity in Human NREM Sleep. Front Neurosci, 14, 449.

Feinberg, I. & Campbell, I. G. (2013) Longitudinal sleep EEG trajectories indicate complex patterns of adolescent brain maturation. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 304(4), R296-303.

Hahn, M., Heib, D., Schabus, M., Hoedlmoser, K. & Helfrich, R. F. (2020) Slow oscillation-spindle coupling predicts enhanced memory formation from childhood to adolescence. Elife, 9.

Helfrich, R. F., Lendner, J. D. & Knight, R. T. (2021) Aperiodic sleep networks promote memory consolidation. Trends Cogn Sci.

Helfrich, R. F., Lendner, J. D., Mander, B. A., Guillen, H., Paff, M., Mnatsakanyan, L., Vadera, S., Walker, M. P., Lin, J. J. & T., K. R. (2019) Bidirectional prefrontal-hippocampal dynamics organize information transfer during sleep in humans. Nature Communications, 10(1), 3572.

Helfrich, R. F., Mander, B. A., Jagust, W. J., Knight, R. T. & Walker, M. P. (2018) Old Brains Come Uncoupled in Sleep: Slow Wave-Spindle Synchrony, Brain Atrophy, and Forgetting. Neuron, 97(1), 221-230 e4.

Killgore, W. D. (2010) Effects of sleep deprivation on cognition. Prog Brain Res, 185, 105-29.

Kurth, S., Jenni, O. G., Riedner, B. A., Tononi, G., Carskadon, M. A. & Huber, R. (2010) Characteristics of sleep slow waves in children and adolescents. Sleep, 33(4), 475-80.

Maris, E. & Oostenveld, R. (2007) Nonparametric statistical testing of EEG- and MEG-data. J Neurosci Methods, 164(1), 177-90.

Muehlroth, B. E., Sander, M. C., Fandakova, Y., Grandy, T. H., Rasch, B., Shing, Y. L. & Werkle-Bergner, M. (2019) Precise Slow Oscillation-Spindle Coupling Promotes Memory Consolidation in Younger and Older Adults. Sci Rep, 9(1), 1940.

Muehlroth, B. E. & Werkle-Bergner, M. (2020) Understanding the interplay of sleep and aging: Methodological challenges. Psychophysiology, 57(3), e13523.

Niethard, N., Ngo, H. V. V., Ehrlich, I. & Born, J. (2018) Cortical circuit activity underlying sleep slow oscillations and spindles. Proceedings of the National Academy of Sciences of the United States of America, 115(39), E9220-E9229.

Purcell, S. M., Manoach, D. S., Demanuele, C., Cade, B. E., Mariani, S., Cox, R., Panagiotaropoulou, G., Saxena, R., Pan, J. Q., Smoller, J. W., Redline, S. & Stickgold, R. (2017) Characterizing sleep spindles in 11,630 individuals from the National Sleep Research Resource. Nature Communications, 8, 15930.

Van Dongen, H. P., Maislin, G., Mullington, J. M. & Dinges, D. F. (2003) The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26(2), 117-26.

Wilhelm, I., Metzkow-Meszaros, M., Knapp, S. & Born, J. (2012) Sleep-dependent consolidation of procedural motor memories in children and adults: the pre-sleep level of performance matters. Developmental Science, 15(4), 506-15.

Winer, J. R., Mander, B. A., Helfrich, R. F., Maass, A., Harrison, T. M., Baker, S. L., Knight, R. T., Jagust, W. J. & Walker, M. P. (2019) Sleep as a potential biomarker of tau and beta-amyloid burden in the human brain. J Neurosci.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Reviewer #1:

This paper puts together a nice set of data showing that a specific gene called Resf1 when deleted effects the ability of ESCs to self-renew and proceed to germline fates. I believe the data are sound and that they provide the evidence needed for the authors to make their conclusions.While I think the data are presented well and the manuscript is well-written, the "modest" functional results suggest this work would be more suited for a specialized journal.

We thank Reviewer 1 for their supportive comments.

• *

Reviewer #2:

1. In the presence of LIF, there is no difference between Resf1 knockout mESCs and WT mESCs except the expression of Esrrb, Nanog and Pou5f1. What about other genes? RNA-seq is needed to distinguish the two cell lines.

Fukuda et al. have shown that deletion of Resf1 leads to misregulation of ~1000 genes (adj. p-value 2) in presence of LIF. This highlights large differences between transcriptomes of Resf1 KO and WT cells that occur despite only a marginal difference in self-renewal efficiency between Resf1 KO cells and WT in the presence of LIF. It is therefore questionable whether the time and resources required to perform the requested RNA-seq would produce data that could unambiguously identify the potential causative effector difference downstream of Resf1.

As an alternative approach, we have reanalysed the Fukuda et al RNA-seq data. We find that Esrrb is significantly downregulated (in agreement with our Q-RT-PCRs), as are Klf4 and LifR (FDR 1.5). However, our meta-analysis of the Fukuda et al data did not show Pou5f1 and Nanog to be differentially expressed (FDR 1.5). This is in line with the lower level of downregulation of Pou5f1 and Nanog, compared to Esrrb in our Q-RT-PCR data. Notably, our gene expression analyses were performed in 5 biological replicates, whereas Fukuda et al. performed RNA-seq in two biological replicates. We can include the meta-analysis of the Fukuda et al data in our submission. As the change in ESC self-renewal that we see at low LIF concentrations could result from a decrease in Lifr expression, we will verify the change in expression of Lifr by Q-RT-PCR. Importantly, we will do this in a way that discriminates between expression of the transmembrane Lifr and soluble LifR, since the latter acts antagonistically (PMID: 9396734, Chambers, BJ, 1997).

1. The authors showed Resf1 is not required for Nanog function, so how does Resf1 regulate the expression of pluripotency genes? Through epigenetic modifications or signaling pathways? The authors should design experiments to explain the detailed mechanisms.

The strength of the immunoblot signal for RESF1 is low, even when Resf1 is expressed episomally. Therefore, although we could try to co-immunoprecipitate with the Resf1-v5 cell line and endogenous Nanog, the expression level of RESF1 may mean this effort is unsuccessful. Given the fact that the result will not affect the conclusions of our study, we do not think this effort is justifiable.

1. The authors showed that Resf1 interacts with Nanog, but they used forced expressed proteins. Does the endogenous Resf1 interacts with endogenous Nanog? Do they bind to some same DNA sequences?

This are important questions to answer. However, many more experiments would be required to reach firm conclusions. The reviewer is right to say that the mechanisms by which Resf1 affects pluripotency are unknown and remain to be answered in future. We therefore propose to improve the text discussing similarities in pluripotency phenotype between deletions of Trim28, SETDB1, YTHDC1 and RESF1. As deletion of RESF1 partner SETDB1 or other proteins involved in repression of retrotransposons lead to downregulation of pluripotency genes and in some cases collapse of ESCs (e.g. PMID: 19884255, Bilodeau et al. 2009; PMID: 19884257, Yuan et al. 2009), we hypothesise that the RESF1 phenotype may be explained by affecting SETDB1 chromatin binding and therefore repression of SETDB1 targets. The mild phenotype of RESF1 KO indicates that RESF1 would not be an essential component of this repressor complex but rather “a modulatory protein”.

It is also worth noting that the meta-analysis of the RNA-seq data from Fukuda et al. suggests that Resf1-null ESCs may express reduced levels of LifR mRNA, and this is something we plan to investigate.

1. In figure 5C, some Resf1 positive cells showed Nanog negative. Are these Nanog negative cells pluripotent?

Nanog-null ESCs are pluripotent (PMID: 18097409, Chambers et al., 2007). In addition, NANOG-negative cells in FCS/LIF cultures can retain pluripotency. Our purpose in this figure was therefore not to say whether NANOG-negative:RESF1-positive cells are pluripotent but to draw attention to the broader expression of RESF1 in FCS/LIF compared to NANOG. Such broader expression has also been noted for other heterogeneously expressed factors (PMID: 31582397, Pantier et al. 2019).

1. In figure 6A, the naïve mESCs are induced to EpiLCs. Is the transition efficiency of Resf1 knockout cells the same with WT mESCs? The finally obtained PGCLCs should be identified.

We show that the key TFs of EpiLC state are expressed similarly in WT and Resf1 KO cells (Supplementary figure 4) and we have data showing that WT and Resf1KO EpiLCs have a similar morphology. Together this suggests an efficient transition to an EpiLC state. Our analysis has identified expression of Blimp1/Ap2g/Prdm14 in Resf1-null cultures. Compared to wild-type cells these levels are reduced up to 3-fold. As this is from an unsorted population and the number of SSEA1/CD61-positive cells is decreased around 2x, this suggests that the PGCLC population formed by Resf1-null cells is reduced in proportion but is otherwise normal.

We will add photographs of EpiLC colonies formed by Resf1 KO and WT cells.

1. in figure 5c, the scale bar is missing.

We will add missing scale bars in the figure 5C.

Reviewer #3:

1. What was less clear was an explanation of why colonies 4 and 24 were chosen. Were there other colonies with the desired expression? Was this amount of expression repeated in replicative experiments with approximately 2 colonies only available to be selected?

Approximately 30 colonies were selected for analysis. Of these, only 2 had deletion of both Resf1 alleles. We will make this point clearer in the text.

1. Figure 1C, 5C and S2B with microscopic images should include a scale bar.

Missing scalebars in the Figure 1C will be added. Unfortunately the microscopy setup used to collect the images in Figures 5C and S2B did not allow scalebars to be added at the time of imaging and these cannot be added retrospectively. However, we do not think that inclusion of scalebars, even were it possible would affect the conclusions of our manuscript.

1. Figure 1E needs a better explanation of the significance, "less clear cut" is not adequate. Reporting statistics, or lack of significance, on the graph would help.

We will update the manuscript and the Figure 1E to include results of a statistical analysis (Wilcoxon-rank sum test) comparing formation of AP+ colonies between Resf1 KO and WT cells at different LIF concentrations. These results show that both Resf1 KO cell lines have lower median number of AP+ colonies than WT cells at LIF concentrations 0 and 1 (p.adj. *

1. It's translatability to medicine, although perhaps that is not the intention, is somewhat lacking. Is there a naturally occurring situation where LIF is absent that would require this pathway to be used? These were mouse ESC's, perhaps this study could incorporate information about relevant translation to a human condition to aid in the significance. This manuscript suggests a mechanistic evaluation by which self-renewal can occur other than the canonical pathway, which is interesting and can inform the field.

Our results suggest that RESF1 directly or indirectly supports self-renewal of ESCs. Interestingly, Human cell atlas identified RESF1 expression as a negative predictor of survival of renal cancer and was found to be expressed in testis cancer cells and other cancer tissues. Therefore, RESF1 could promote self-renewal of cancer cells similarly to ESCs. However, this is speculative and needs further studies. As this is both outside of the scope of this manuscript and our expertise, we do not think it prudent for us to pursue this line of inquiry. However, we agree that further studies could evaluate RESF1 function in human tissues, especially pluripotent cells and germ cells. As we show that RESF1 deletion leads to reduced induction of PGCLCs and previous studies showed infertility of Resf1 KO mice, investigating link between human fertility and RESF1 could have implications in reproductive medicine.

We will improve our discussion to highlight the possible significance of RESF1 function in human fertility.

Qme4bLv4wxfof9ixTMj5e2eUJLJy3U7W4kKNAoNFKH4u6q

Author Response:

Reviewer #1 Public Review:

Nakayama and colleagues report a unique screening concept utilizing conserved mechanisms between zebrafish gastrulation and cancer metastasis for identification of potential anti-metastatic drugs. They screen 1280 FDA-approved drugs using the gastrulation as a marker, and identify Pizotifen as an epiboly interrupting drug. Then they find that pharmacologic and genetic inhibition of HTR2C, a target of Pizotifen, suppresses metastatic progression in a zebrafish and mouse model through inhibition of epithelial to mesenchymal transition (EMT) via Wnt-signaling.

Their work is of interest and has the potential to appeal to a broad audience. However, additional experiments are needed to further substantiate their concept that human cancer metastasis mimic/recapitulate zebrafish gastrulation in terms of conserved mechanism, as well as to confirm the validity of their screening method regarding to the effects of global toxicity.

Major concerns:

The first major concern I have is the appropriateness to think the gastrulation as a parameter/index of cancer metastasis. While they cherry-picked some genes that they are known to be involved in both gastrulation and cancer metastasis, more broad analysis should probably be necessary to conclude so. For examples, the authors can analyze comprehensive RNA-seq data set to see if the pathways/networks are similar between gastrulation (zebrafish embryo development data set) and cancer metastasis (benign/primary tumors vs metastasis tumors in TCGA).

The conservation of embryonic EMT and tumor metastasis EMT has long been well recognized. Now we cited some of these published references (Nieto et al., 2016; Thiery et al., 2009; Yang and Weinberg, 2008). In Table 1, we compiled 50 genes based on published literature to provide further and strong evidence to support this conservation. Knockdown of these genes in Xenopus or zebrafish induced gastrulation defects; conversely, overexpression of these genes conferred metastatic potential on cancer cells and knockdown of these genes suppressed metastasis. Although this point is not really an objective of this study, we believe that the evidence for the conservation is sufficiently convincing to provide the basis for our study. Further RNA-seq comparison of zebrafish embryonic EMT and human tumor metastasis should be beyond the scope of the current study. Generally, the transcriptomic data for zebrafish embryo development at the epiboly/gastrulation stage are based on the whole embryos which include all other activities and are not specific to EMT; thus, it may not be a proper comparison with tumor metastasis data to search for more evidence.

The second concern is about the Pizotifen's effects on cancer metastasis. Since the Pizotifen suppresses gastrulation, it might have some harmful effect on the organogenesis/development of day2 embryos that they used in zebrafish transplantation model. And if so, cancer metastasis can be suppressed indirectly. The authors could examine if Pizotifen could have some side effects on day2 embryos. The drug also has some cell viability suppressive effects in vivo so as the pics in Fig.2D looks like, and it would be good if this possibility was excluded.

We had not observed any abnormality in development of Tg (kdrl;GFP) and WT zebrafish at day 2 when these fish were treated with 5µM Pizotifen. However, more than 20µM Pizotifen treatment affected approximately 10% of these fish. The affected fish show shorter tail rather than that of vehicle-treated zebrafish. In xenograft experiments, zebrafish embryos at day 2 were treated with 5µM Pizotifen. The concentration of Pizotifen did not affected development of day2 embryos.

Futhermore, we demonstrated Pizotifen did not affect primary tumor growth in a mice model of metastasis using 4T1 cells by two different experimental methods. One is that tumor measurement revealed that the sizes of the primary tumors in Pizotifen-treated mice were equal to those in the vehicle-treated mice at the time of resection on day 10 post inoculation. The other is that IF-staining showed the percentage of Ki67 positive cells in the resected primary tumors of Pizotifen-treated mice were the same as those of vehicle-treated mice (Figure 3A and B). Therefore, we conclude that Pizotifen suppress metastasis without affecting cell viability in vivo.

Finally, the mechanistic parts would need more confirmation and rescue experiments. Transplanted cells can be sorted after the treatment and the expression changes of EMT markers can be examined to see if the phenomenon happens in vivo as well. All main results can be rescued to see if the effect of Pizotifen against EMT happens through HTR2C-Wnt axis.

Figure 5C showed that 4T1 primary tumors from Pizotifen-treated mice has elevated E95 cadherin expression compared with tumors from vehicle-treated mice. Furthermore, Figure 5C also demonstrated that β-catenin accumulated in the nucleus, and phospho-GSKβ and Zeb1 expression were decreased in 4T1 primary tumors from Pizotifen-treated mice compared with vehicle-treated mice. Loss of E-cadherin plays an essential role in promoting EMT-mediated metastasis since loss of E-cadherin itself is enough to promote metastasis. In contrast, overexpression of mesenchymal markers: vimentin and N-cadherin is not sufficient to induce metastasis. Based on our data from Figure 5C and the accumulated evidences, we conclude that Pizotifen restored epithelial properties to metastatic cells through a decrease of transcriptional activity of β-catenin in vivo

Cultivating an Abundance Mindset.

During all proceeding steps and activites, remember that "Through effort and experience can the brain change." This will help to explain the motivation behind the need for discomfort if change is to be made.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

First of all, we would like to thank the each of the expert reviewers for their effort in evaluating our study. We are confident that we can positively address each of the issues and queries raised by the reviewers.

Reviewer #1 (Evidence, reproducibility and clarity (Required)): **Summary:** This study investigates the functions of the tricalbin proteins in S. cerevisiae, which are homologs extended synaptotagmins in mammals. It suggests the tricalbins modulate plasma membrane phospholipid composition and are particularly important for surviving shift to elevated temperature. The tricalbins are proposed to directly or indirectly promote phosphatidylserine transport from the ER to the plasma membrane during heat stress. They also promote the localization of the kinase Pkh1 to the plasma membrane during heat stress.

**Major comments:**

General Response: We thank the reviewer for the comments and opinions. While they are starkly different from the other two reviewers’, they have caused us to consider how we can add additional support for our conclusions and to consider alternative possibilities.

1. To determine whether the tricalbins and other tethering proteins play a role in phospholipid homeostasis, lipid distribution is measured using biosensors (FLARES) and phospholipid levels are determined by mass spec. The experiments are well done but say little about what role the tricalbins or other tethering proteins play in homeostasis. There are no measurements of lipid transport or rates of phospholipid production, degradation or modification. It is reasonable to propose the tricalbins transport lipids, since other proteins with SMP domains do, but this study does not present any evidence that they do.

Response: We thank the reviewer for stating that the quantitative microscopy and lipidomics experiments in our study are well done. However, we strongly disagree that the results do not provide new information on the roles of the tricalbins or other tethering proteins in membrane lipid homeostasis. In fact, the lipidomics results definitively show that Ist2 and Scs2/22 control phosphatidylserine (PS) levels. In contrast, loss of the tricalbins does not significantly affect PS levels. We will provide a new figure to make this even more clear to expert and non-expert readers.

While the tricalbins do not regulate PS levels, the data clearly show that the tricalbins control PS acyl chain saturation and cellular distribution. A PS reporter is reduced at the plasma membrane (PM) upon loss of the tricalbins and we have now confirmed increased localization at the endoplasmic reticulum (ER). These new results will be included in a revised manuscript. As the lipid desaturase Ole1 is localized in the ER, the lipidomics are consistent with increased ER residency of PS species. Thus, the data indicate that while the tricalbins do not regulate PS levels, they control either delivery of PS from the ER to the PM, and/or they may control the organization and stability of PS at the PM which would be a novel finding on its own.

The reviewer must be aware that there are currently no in vivo cell assays that directly (only) measure lipid transport. These experiments are subject to several factors including lipid metabolism, anterograde transport rates, bilayer organization, lipid accessibility, and retrograde transport rates. While our findings clearly show that the Tcb proteins do not control PS levels, we agree that there are alternative possibilities to explain the changes in PS distribution. In our revised manuscript, we will perform additional experiments to distinguish between these possibilities. New experiments will include mutant forms of Tcb3 bearing substitutions in the SMP domain. We will also examine whether the Tcb proteins control PS organization, availability/accessibility, and stability at the PM (also see Reviewer #3, comment 2). This latter possibility may reveal a novel concept regarding Tcb/E-Syt protein function that goes beyond their proposed conventional role as lipid transfer proteins. Based on the outcome of these experiments, we shall adjust the final cartoon model and conclusions in the discussion accordingly.

The study convincingly demonstrates that there are fewer Pkh1 puncta formed after temperature shift in cells lacking tricalbins than in wild-type cells (Fig. 6 C,D). However, there is no demonstration that this change in localization alters Pkh1 function or signaling.

Response: Regulation of Pkh1 by lipids is outside the scope of our current study that is focused on providing new understanding of Tcb protein function. However, the decrease in heat-induced Pkh1 puncta may provide insight into the PM integrity defects in cells lacking Tcb1/2/3 (as Pkh1 is required for PM integrity). To test whether Pkh1 function is compromised in the tcb1/2/3 mutant cells, we can test whether constitutively active Ypk1 (which acts downstream of Pkh1) rescues the PM integrity defects in tcb1/2/3 mutant cells.

There is no demonstration that association of tricalbins and Skh1 (Fig. 4) has any functional significance or affects phosphoinositide metabolism.

Response: We thank the reviewer for raising this issue. If the association of Tcb3 and Sfk1 has functional significance, then one would expect that loss of the proteins should phenocopy one another. Deletion of the Sfk1 cytoplasmic domain necessary for co-localization with Tcb3 should also phenocopy loss of Tcb3. This is exactly what we find. Localization of the PS probe is decreased at 42oC upon loss of Sfk1 or truncation of the Sfk1 cytoplasmic tail, similar to cells lacking Tcb3. Furthermore, we find that Tcb3 regulates sterol homeostasis at the PM (using the D4H probe), as has been recently reported for Sfk1 (Kishimoto et al, 2021). Thus, Tcb3 and Sfk1 not only co-localize, but they also share common functions in PM lipid organization. These new results will be presented in our revised manuscript.

The reviewer also inquired about potential roles for Tcb3 and Sfk1 in phosphoinositide lipid homeostasis, as Sfk1 has reportedly been implicated in heat-induced PI(4,5)P2 synthesis. However, while we find clear roles for Sfk1 and the tricalbins in PS and sterol homeostasis, we did not find a requirement for Sfk1 or the tricalbins in PI(4,5)P2 homeostasis upon heat stress conditions. These findings will be included in our revised manuscript. Importantly, our results indicate that the tricalbins and Sfk1 primarily control PS and sterol homeostasis at the PM, and may regulate phosphoinositides indirectly, and thus provide new insight into the key role of these proteins.

The study proposes the tricalbins directly or indirectly promote phosphatidylserine transport after temperature shift, but transport has not measured and other possibilities are not ruled out.

Response: While the Tcb3 SMP domain has been shown to transfer phospholipids in vitro (Qian et al., 2021), we agree that a role in PS transfer in vivo should be examined in more detail and that other possible roles in PS homeostasis should also be considered (also see responses to Reviewer #3, comment 2).

Upon heat shock, we not only observe a decrease in relative levels of the PS reporter at the PM in the tcb1/2/3 mutant cells (as shown in our original manuscript), but also a corresponding increase in relative levels of the PS probe at the ER and vacuole membrane (also see response to Reviewer #3, comment 5). This could reflect impaired delivery of PS from the ER to the PM and reduced stability of PS at the PM (i.e. increased internalization of PS into the cell).

In our original manuscript, we showed that deletion of the SMP domain (the lipid transfer domain), phenocopies deletion of the full-length Tcb3 protein in terms of PS distribution and PM integrity following heat shock. To more rigorously test whether lipid transport activity of the SMP domain is responsible for these phenotypes, we will generate amino acid substitutions within the SMP domain of Tcb3 that maintains its overall structure but impairs its ability to transport lipids (by targeting conserved key residues identified in Saheki et al., 2016). We will then assess whether SMP-mediated lipid transfer is necessary for PS homeostasis and PM integrity under heat stress.

We also agree that other possibilities should be examined. First, to rule out a defect in PS production upon heat stress, we are performing new mass spectrometry lipidomics experiments to measure levels of individual phospholipid species in the tcb1/2/3 mutant and wild type cells after heat stress.

Second, we have considered whether the Tcb proteins control phospholipid bilayer distribution (e.g. flip and flop). However, cells lacking the Tcbs are not hypersensitive to duramycin (Omnus et al. 2016) and thus phosphatidylethanolamine exposure on the extracellular leaflet is not increased. Moreover, cells lacking the Tcbs (and Scs2/22 and Ist2) are not impaired in the uptake of exogenous NBD-labelled phospholipids (and thus flip across the PM bilayer is not impaired). Possibly, there may be increased lipid ‘flop’ in the mutant cells at high temperature. We can test whether there is increased phospholipid exposure in the extracellular leaflet at high temperature, but our results thus far indicate accumulation of PS on internal membrane compartments (the ER and vacuole membrane).

Another potentially exciting possibility is that the tricalbin proteins bind and stabilize PS within the cytosolic leaflet of the PM and prevent its internalization by endocytosis or non-vesicular transfer. This mechanism would be completely independent of lipid transport to the PM and would constitute new mechanistic insight into Tcb function. We will test whether PS (and sterol) becomes more accessible (less stable or reduced sequestered pools at the PM) and internalized into the cell, upon removal of the tricalbin proteins. For example, we will monitor PS distribution in cells where endocytosis is blocked with latrunculin A.

As mentioned, there currently no cellular lipid transport assay that directly (only) measure anterograde transport. However, if the Tcb3 SMP domain mutants are impaired in PS homeostasis and PM integrity, then we can consider monitoring PS transfer in vivo. By performing the experiments outlined here, we will have thoroughly characterised the roles of the tricalbin proteins in PS homeostasis at the PM. Moreover, the new findings may even reveal novel roles that are independent of transport.

Reviewer #1 (Significance (Required)): While this study is likely to be of interest to those studying the tricalbins or phospholipid homeostasis, it is incremental and provides little conceptual advance on what is already known about the tricalbins and extended synaptotagmins. They have already been implicated in lipid homeostasis in the plasma membrane and this study provides no new mechanistic insight into how this occurs. Similarly, it has already been shown that the tricalbins play a role in maintaining cell integrity during heat stress and there is little new insight into what role the tricalbins play. Perhaps the most notable part of the study is the idea that tricalbins are necessary for phosphatidylserine transport during stress, but considerable additional work is necessary to make a strong case for this claim.

Response: We strongly disagree with the reviewer’s opinions. Indeed, Reviewer #2 found our study “novel and detailed” and Reviewer #3 found the results in our study to be “highly valuable” and “interesting”.

In contrast to the reviewer’s claims, there are certainly novel findings in our study. Foremost, this is the first study that demonstrates a role of the tricalbins in PS homeostasis. Previous studies have implicated E-Syt family members in diacylglycerol and phosphoinositide regulation. Our results indicate that the tricalbins and Sfk1 primarily control PS and sterol homeostasis at the PM, not phosphoinositides, and thus provide new insight into the key role of these proteins. Second, while a previous study by Collado et al reported a role of the tricalbins in PM integrity upon heat stress, this work did not provide mechanistic insight into this process. We performed the PM integrity assays for the Collado et al study (as co-authors). Our current study now shows that Tcb function is needed for PS homeostasis and Pkh1 recruitment at the PM upon heat stress; both factors are needed for PM integrity under these conditions. As such, our current study does provide new insight into roles of the Tcbs in PM integrity. Finally, we are exploring roles of the Tcb proteins in PS homeostasis that go beyond their proposed functions as lipid transfer proteins. We are convinced that our study will provide novel and deep mechanistic understanding of the Tcb/E-Syt protein family.

Reviewer #2 (Evidence, reproducibility and clarity (Required)): The study examines the roles of tricalbins in signalling in yeast. By using several mutations they are able to evaluate whether the interactions are caused by tethering the PM and ER or by other mechanisms. Particualrly powerful is their use of cryo-EM tomography and shot-gun mass spectrometery. They look at all the major lipids of the ER and PM and consider their interconversions. They identify large changes in lipid species with one double bond and with two, particularly for PS.

Reviewer #2 (Significance (Required)): There is little information about membrane physical properties and how it changes as a result of changes in lipid molecular species. Nevertheless, the information provided is novel and detailed. The topic of ER-PM contact sites is new and evolving and this paper advances our understanding of the yeast system considerably. It also looks at protein-protein interactions by fluorescence methods and studies the consequences of heat shock.

Response: We are pleased that the reviewer concluded that our study on the Tcb proteins “advances our understanding of the yeast system considerably” and found our use of lipidomics to be “powerful”.

This reviewer only had only one critique; there “is little information about membrane physical properties and how it changes as a result of changes in lipid molecular species”. In our revised manuscript, we will provide new data showing changes in levels of sterol (ergosterol) accessibility/availability at the PM in cells lacking the tricalbin proteins. Sterol lipids exist in distinct pools in the PM bilayer (extracellular vs. cytoplasmic leaflet, accessible vs. sequestered) that control the biophysical and mechanical properties of the PM (packing order, permeability, etc.). Moreover, PS and sterol lipids are proposed to undergo mutual associations whereby PS controls sterol accessibility (the ‘umbrella’ model) and sterol in turn stabilizes PS in the cytoplasmic leaflet of the PM. Our findings demonstrate that the primary function of the Tcb proteins is PS and sterol organization in the PM, providing new mechanistic insight into regulatory mechanisms for membrane homeostasis. We will attempt to further characterize changes in PM mechano-chemical and biophysical properties to further understand how changes in membrane lipid composition affect membrane integrity.

Reviewer #3 (Evidence, reproducibility and clarity (Required)): The Tcb/ESyt proteins play important role in contact site formation and non-vesicular lipid transport. However, their exact functions remain highly controversial. This study by Stefan and colleagues revealed a new role for the yeast Tcb proteins, especially Tcb3, in regulating plasma membrane phosphatidylserine, as well as PIP2. These results are highly valuable to people working on membrane contact sites and lipid trafficking. Overall, the results are fairly convincing and interesting. There are however some concerns and suggestions:

Response: We are pleased that the reviewer stated that our study has “revealed a new role for the yeast Tcb proteins” and that the findings are “fairly convincing and interesting”. We also thank the reviewer for providing constructive criticisms and helpful suggestions.

1. Fig. 4, for Tcb3 and Sfk1 interaction, what about Tcb1/2, which would be good controls for the specificity of the interaction.

Response: We agree that it would be useful to determine if the Tcb3-Sfk1 interaction is specific. We will perform additional BiFC experiments to address whether Tcb1 and Tcb2 associate with Sfk1. Our previous work suggested that Tcb3 forms heterodimers with Tcb1 and Tcb2 necessary for PM integrity (Collado et al., 2019). However, the functional association between Sfk1 and Tcb3 may be specific to Tcb3, and we will test this possibility. It would be interesting to identify a function specific to an individual Tcb protein.

A central question is whether Tcb3 transfers PS by itself through the SMP domain or requires other lipid transfer proteins. One possibility is that the transfer is mediated by Osh6 and Osh7. Did the expression level of Osh6/7 change between the delta tether and the ist2scs2/22 null strain? Under normal and stressful conditions?

Response: We agree that it is important to address whether Tcb3 transfers PS via its SMP domain (an issue also raised by Reviewer #1) or whether this activity is carried out by another lipid transfer protein, such as Osh6 and Osh7. We have considered several alternative possibilities, as well as designed and performed new experiments, as described below (two key experiments are in italics).

To rigorously test whether SMP domain-mediated lipid transport activity is required for PS homeostasis at the PM under heat stress, we will generate amino acid substitutions within the Tcb3 SMP domain that maintain its overall structure but impair its ability to transport lipids (targeting conserved key residues described in Saheki et al., 2016 & Bian et al., 2018). We will then assess whether SMP-mediated lipid transfer is necessary for PS homeostasis and PM integrity under heat stress.

As suggested by the reviewer (and discussed in our original manuscript), the tricalbins might serve as scaffolds for PS transfer proteins, including the Osh6 and Osh7 proteins, under stress conditions. Osh6 and Osh7 are recruited to ER-PM contacts through interactions with the Ist2 tether protein where they mediate PS delivery to the PM under non-stress conditions (D’Ambrosio et al., 2020). However, strong lines of evidence suggest that Ist2, Osh6, and Osh7 are not required for PS homeostasis at the PM under stress conditions. First, loss of Ist2 has no impact on the PS probe under heat stress conditions (this result will be included in the revised manuscript). Therefore, the Ist2-Osh6/7 interaction is not required for PS homeostasis upon heat stress. Ist2 is not required for PM integrity upon heat stress either (Omnus et al., 2016).

These results do not rule out the possibility that the tricalbins serve as scaffolds for other PS transfer proteins, such as Osh6 and Osh7, under stress conditions. In this scenario, a switch between Osh6/7 tethering proteins may occur: from Ist2 under normal growth conditions to the Tcb proteins during stress conditions. However, our findings also suggest that Osh6 and Osh7 function is impaired upon stress conditions, including heat and nutrient starvation. Notably, Osh6 and Osh7 become mis-localized from the PM under upon heat stress (we can provide this data in the revised manuscript). The mechanisms for Osh6/7 attenuation upon stress conditions is outside the scope of our current study, but our preliminary results suggest that changes in cytoplasmic pH and ion homeostasis are involved; this will be a focus of a future study. This is also in line with results from a previous study (Omnus et al., 2020) that showed Osh3 forms intracellular aggregates in response to heat stress. The activity of the Osh proteins and other lipid transfer proteins, in general, may be impaired upon stress conditions (see below).

To directly determine whether Osh6 and Osh7 are required for PS homeostasis at the PM under heat stress conditions, we will monitor the distribution of the PS probe in cells lacking the Osh6 and Osh7 (osh6 osh7 double mutant cells) upon heat stress. This is a key experiment that will directly address the reviewer’s concerns.

As suggested by the reviewer, we can also examine the expression and localization of GFP-tagged Osh6 and/or Osh7 in tcb1/2/3, scs2/22 ist2, and ‘delta tether’ mutant cells. However, we have not observed any changes in other Osh proteins, including Osh2 and Osh3, in the ‘delta tether’ mutant strain.

Finally, we have considered yet another possibility. PS transfer to the PM may be generally attenuated under heat stress conditions and a key role of the Tcb proteins may be to bind and stabilize PS at the PM. In other words, the Tcb proteins may act as a ‘buttress’ to stabilize and maintain pre-existing pools of PS at the PM under stress conditions. Consistent with this idea, the Tcb3 C2 domains are required for PS homeostasis and PM integrity upon heat stress. If the Tcb3 SMP domain mutants are not impaired in PS homeostasis or PM integrity, then this alternative mechanism may be very relevant to PM quality control and organization in response to membrane stress. To address this possibility, we will address whether PS is internalized or removed (extracted) from the PM upon heat stress in cells lacking the Tcb proteins. This may uncover a novel role of the Tcb proteins that is independent of SMP domain-mediated lipid transfer.

Figure 5A. It is not obvious that the intensity of C2 decreases in tcb null cells at 42 degree. Perhaps there is more internal staining.

Response: We thank the reviewer for pointing out this issue. We think Figures 5A and 5B together convincingly show increased cytoplasmic localization of the PS reporter in the tcb1/2/3 mutant cells upon heat stress. More importantly, we thank the reviewer for pointing out the increased localization at internal membrane compartments. We realized it would be important to identify the PS-containing membrane compartments in the tcb1/2/3 mutant cells upon heat stress. We have now confirmed that the PS probe localizes at the ER and vacuole membrane (to be included in the revised manuscript). The example in Figure 5A also shows accumulation of the PS probe at both the nuclear ER and vacuole membrane. Thus, whilst wild type cells show little PS reporter localization at the ER or vacuole membrane, loss of the tricalbin proteins leads to an increase in ER and vacuole membrane PS probe localization after heat shock. Accumulation at the ER may reflect impaired PS delivery from the ER to the PM, and possibly rerouting to the vacuole membrane. Alternatively, as discussed above, vacuole membrane localization may be due increased PS removal from the PM and delivery to the vacuole membrane in tcb1/2/3 cells upon heat stress.

It is important to determine the primary function of Tcb3 since both defects in both PIP2 and PS were observed. If the change in PIP2 is due to a lack of PS, can overexpressing Osh6/7 rescue the PIP2 defect in the tetherless mutant?

Response: We agree that is important to determine whether the primary function of the Tcb proteins is regulation of PS or PI(4,5)P2 homeostasis. Our new findings definitively indicate that their primary function is PS regulation, not PI(4,5)P2 regulation. A clear effect in the distribution of the PS reporter was observed following heat shock in the tcb1/2/3 mutant cells. In contrast, there is no difference in the localization of the PI(4,5)P2 reporter in tcb1/2/3 cells compared to wild type after heat shock (also see response to reviewer #1, comment 3). In addition, cells lacking the Tcb proteins were not impaired in heat-induced PI(4,5)P2 synthesis, as assessed by metabolic labelling and HPLC analysis. These findings will be included in our revised manuscript, as they indicate that the Tcb proteins primarily control PS at the PM, not phosphoinositides, and thus provide new insight into the main role of these proteins.

Both PS and sterol are required for the proper recruitment of type I PIP5K to the PM (Nishimura et al., 2019). Therefore, defects in PS distribution could be responsible for the PI(4,5)P2 effects observed in Figure 2. However, overexpression of Osh7 in the ‘delta tether’ mutant did not significantly rescue localization of the PI(4,5)P2 reporter (included in the revision plans). Sterol organization is also perturbed in the ‘tether’ mutant cells (Quon et al., 2018), and this may explain why Osh7 expression did not rescue. Accordingly, Osh2/3/4 (sterol transfer proteins) rescue the PI(4,5)P2 effects.

The detection of PS by LactC2 has been well-established. However, an alternative approach would be to use the 2XPH in permeabilized cells. See PMID: 33929485 for some detailed discussions on the techniques. It is not a requirement for the authors to adopt the 2XPH.

Response: We thank the reviewer for suggesting another technique to confirm the results of the LactC2 domain as a PS probe. In this study, we have primarily used a genetically encoded LactC2 probe to observe PS distribution within live, intact cells. Whilst this approach was sufficient to identify accumulation of PS on cytosolic membrane leaflets of the ER and vacuole (see above), the addition of an exogenous probe to permeabilized cells may allow the detection of PS on luminal and extracellular membrane leaflets. Therefore, we plan to repeat our heat shock experiments using permeabilized cells and a purified tagged form of the LactC2 protein. This may allow for improved imaging of intracellular PS localisation and bilayer distribution. However, these experiments are technically challenging, and fixation and permeabilization conditions have not yet been optimized for yeast cell experiments. It is not yet known whether we will be able to optimize these protocols in a reasonable amount of time while completing revisions to the manuscript.

**Minor:**

1. the discussion seems to be a bit long

Response: We will shorten the discussion and modify our final conclusions based on the results from the new experiments.

Reviewer #3 (Significance (Required)): These proteins are highly important in cell biology/contact sites. The redundancy made it difficult to pinpoint their function. Previous studies have had a number of models. The current study proposed a new function of these proteins, i.e. PS transfer, and this is very interesting and valuable. There will be a good audience for this work. I specialize in lipid storage and trafficking, lipid droplets, cholesterol and phosphatidylserine.

Response: We are pleased that this expert reviewer found our study to be “very interesting and valuable”.

Author Response:

Reviewer #1:

The paper by -Blackwell et al. develops the ideas developed in the influential paper by Dash et al. (2017) which defined a similarity matrix for CDR3s TCRdist which is based on a weighted combination of local and global similarity measurements. In this paper they use the metric to develop the idea of a meta-clonotype, a set of similar TCRs which enrich for TCRs directed at the same antigen. They demonstrate that these meta-clonotypes show greater publicity than individual clonotypes, and show evidence of HLA-restriction. The authors speculate that the metaclonotype may be a useful biomarker. They provide open-access software tools for defining meta-clonotypes in antigen enriched repertoires.

The major findings are: (1) Meta-clonotypes are more public than clonotypes, a result which seems not unexpected, given that meta-clonotypes include many different sequences; (2) Meta-clonotypes show evidence of HLA restriction, again predicted given the well-established fact that specific antigens can be recognised by sets of similar TCRs.

The concept of a metaclonotype is an interesting one which could have widespread use in analysis of TCR repertoire. However, the impact could be much greater, by sharpening the focus of the paper, and adding detail and clairty to the idea of teh clonotype. In particular, while the introduction correctly points out that prediction of SARS-Cov_2 clinical outcome, or better understanding of the role of coronavirus prior exposure in determining outcome are important unanswered questions, this paper does not address these questions.

Thank you for your careful review of the manuscript. We have submitted a major revision with greater focus on the definition of a TCR meta-clonotype. We have removed from the introduction much of the background about SARS-CoV-2 and potential implications for the pandemic. In its place we’ve added greater detail about meta-clonotypes, how they can be defined from antigen-enriched TCR data, and how they can be used to analyze bulk TCR sequencing data.

A substantial portion of the paper is devoted to analysing data obtained using the MIRA assay (Klinger et al PLoS One 10 :e0141561) to define SARS-COV-2 responses, and it is not always clear whether the objective is to evaluate the accuracy of this data set, or to test the power of the meta-clonotype approach.

Our objective with the analyses of the IMMUNEcode dataset (Nolan et al. 2020) dataset, using MIRA method from Klinger et al. 2015, is to demonstrate that TCR meta-clonotypes can be defined from antigen-enriched TCR data and that they can be used to identify and quantify antigen-specific TCRs in bulk sequenced data. Furthermore, the analysis provides evidence that meta-clonotypes have greater publicity than individual clonotypes, thus increasing sensitivity of detection for antigen-specific TCRs. Using bulk repertoires from COVID-19 patients we then demonstrated that population-level analysis can be made possible using meta-clonotypes and provided supporting evidence that the antigen specificity of the centroid TCR is retained. These analyses and their interpretation is further revised on lines 319-348. We think that in the process of evaluating meta-clonotypes, our analysis also shows that the publicly released data contains valuable information about SARS-CoV-2 TCR specificities; however, we have not systematically attempted to verify the validity of the dataset. In the current revision these objectives are made clear in the revised Introduction section.

Reviewer #2:

Summary of main aims: The main aim of this paper is to build a framework for TCR meta-clonotypes for finding similar TCRs across individuals (or different repertoires). The majority of the investigations performed in this work have the objective of showing the data properties of meta-clonotypes as well as the metaclonotypes' usefulness for the analysis of antigen-specific TCR data and disease-labeled immune repertoire data.

Major strengths: Building meta clonotypes is a possible path towards a better coverage of immune repertoire biology as well as inter-individual repertoire comparison. TCRdist3 is an efficient method for building meta-clonotypes that enables the study of the specific characteristics of meta-clonotypes. So, far clusters of similar sequences have not been investigated in depth. The author team is making a significant step forward in this direction by characterizing meta clonotypes in differentially antigen-specific-clone-enriched repertoires and by relating the results to generation probability, HLA, sex and immune status.

Major weaknesses: Although the authors show a significant amount of data, I am not sure if these data convey sufficient intuition about the characteristics and behavior of meta-clonotypes. The authors seem too focused on relating meta-clonotypes to immune status instead of focusing on the specific biological characteristics of metaclonotypes.

We agree and have shifted the focus of the manuscript away from the results of the application and towards providing greater detail about the characteristics and behaviors of meta-clonotypes; for example, we’ve removed much of the background about SARS-CoV-2 and the COVID-19 cohorts and we’ve added details about how the meta-clonotype radius can be optimized. We’ve also reframed the data analysis section to emphasize that the results demonstrate how meta-clonotypes carry important antigen-specific signals above and beyond individual clonotypes; this makes the results valuable beyond the application to SARS-CoV-2. For example, while demonstrating HLA restriction occurs in SARS-CoV-2 specific T cell responses in COVID-19 patients is not a surprising finding, it provides evidence that meta-clonotypes enable quantification of an antigen-specific and HLA-restricted T cell response from a bulk single-chain TCR repertoire. We use this example analysis to compare the strength of this signal in individual clonotypes, meta-clonotypes with radius alone and meta-clonotypes with a motif constraint. The revision of lines 98-100 and lines 461-470 provide clarity about this motivation for the analysis and interpretation of the results.

Furthermore, we have added a section to the Results that demonstrates how meta-clonotypes and tcrdist3 enable analyses that can provide biological insights about the biochemical properties that may confer antigen specificity (lines 383-418, Figure 10). Since meta-clonotypes define groups of sequences, we can use CDR3 logo plots to dissect how positions and amino acid properties in the CDR3 define the group. In the revision we demonstrate a “background-adjusted” logo plot, that is able to emphasize amino acids that define the meta-clonotype, yet are uncommon among TCRs using the same V and J genes. Visualizing the results in this way can generate hypotheses that can be experimentally validated about the amino acids that are essential for antigen recognition.

Furthermore, the authors fail to convincingly show that the background repertoires chosen for meta clonotypes are robust and to what extent meta-clonotypes are sensitive to changes in the background repertoire.

We agree that it is important to understand how the creation of meta-clonotypes, and specifically optimization of the radius, depends on the background repertoire. Therefore, we conducted sensitivity analyses varying the size (25K to 1M) and makeup (synthetic OLGA vs. cord blood vs. a blend) of the background (lines 259-294). We also empirically demonstrate the value of over-sampling background TCRs with matching V and J genes. We show that using a background of 200,000 TCRs was sufficient for reducing the bias and variability in selecting a meta-clonotype radius, compared to a reference set of 2 M background TCRs; this is important because while it is tractable to use large backgrounds for a small number of meta-clonotypes, for larger studies or analyses confined to a laptop, the smaller background set is sufficient; we also show that this can largely be attributed to the gain in efficiency that comes with using a background that includes synthetic OLGA TCRs with VJ-gene frequencies that match the TCRs included in the meta-clonotype. However, we note that “Ultimately, the best choice for the background may depend on the question being asked and the data that is available, with factors including donor HLA, age, potential antigen exposures, and other factors that may shape the repertoire.” Our goal with tcrdist3 was to make it easy for the user to customize the background to the scientific question.

The authors also do not convincingly differentiate themselves from previous approaches that have used network analysis and generation probability in order to find clusters of similar sequences (very much conceptually similar to the approach taken here).

We agree it’s important to communicate how meta-clonotypes differ from existing TCR analysis approaches. There are several important distinctions with the existing methods that use networks and generation probability, namely TCRNET and ALICE. We have highlighted these differences in the Discussion (lines 483-497), quoting:

“The meta-clonotype approach also differs from methods, such as TCRNET (Ritvo et al., 2018) and ALICE (Pogorelyy et al., 2019), that seek to identify TCRs sharing antigen-specificity within bulk repertoires. These methods were developed to identify TCR nodes in a network with an enriched number of edges compared to the expected number of edges in a background (TCRNET) or derived from a probability model (ALICE). Similarly, another recent method attempts to find antigen-associated sequences in bulk repertoires using a two-stage agglomerative clustering of a k-mer based representation of CDRs, first within and then across bulk repertoires (Yohannes et al., 2021). Our framework is designed for a different task than these algorithms. Specifically, we sought to construct definitions of TCR groupings among already antigen-associated TCRs, which would have high sensitivity and specificity for finding similar TCRs in bulk repertoires. This is an important distinction because the existing network-enrichment methods would simply find that most or all of the TCR groupings among a set of antigen-associated sequences were statistically enriched compared with their frequency in antigen-naïve repertoires. By contrast, a flexible meta-clonotype radius permits the definition of the largest possible group of antigen-associated TCRs with the constraint that the likelihood of finding a TCR within the radius in an antigen-naïve background is equally low across all meta-clonotypes.”

Finally, the authors do not provide detailed descriptions of how the comparison of meta-clonotypes across repertoires is handled as well as potential sequence redundancies across meta-clonotypes (in potentially different individuals). I believe that all of the perceived shortcomings are readily addressable in a revision.

You are quite right that many meta-clonotypes are overlapping in that a single TCR might conform to more than one meta-clonotype definition. Thus, in the application of meta-clonotypes to the COVID-19 dataset we tested each meta-clonotype individually for an association with the predicted HLA- restricting genotype. Depending on the context, if a summary across meta-clonotypes is required (e.g. finding the overall abundance of conformant TCRs in a repertoire) it may be appropriate to use meta- clonotypes to identify conformant sequences, but then tally them based on actual abundance (i.e. no double counting). In a prediction context, it may be desirable to have overlapping meta-clonotype features, and in fact many machine learning algorithms excel in this regime. With tcrdist3 we have incorporated a “join” functionality that allows for relational database-style joining of meta-clonotypes with a TCR repertoire; this makes it relatively easy to eliminate or keep redundancies, depending on the context. We have added a sentence to the Discussion pointing out that there is overlap among meta-clonotypes that needs to be considered in their application and we provide a link to an example of how to use the join functionality on https://tcrdist3.readthedocs.io/en/latest/join.html#step-by-step-example.

All in all, this manuscript is an important steps towards a better understanding of immune receptor biology. tcrdist3 is an evolution of a previously published method (tcrdist) that is here used to build meta-clonotypes. After reading the paper, it remains slightly unclear (addressable in a revision) as to how useful they are for understanding repertoire biology as well as how to use them in practice in terms of robustness and sensitivity.

Thank you for your constructive comments, and I hope we’ve addressed these issues around biological interpretability and application in the revision.

Reviewer #3:

Mayer-Blackwell et al introduce a new framework for leveraging antigen-annotated T cell receptor (TCR) sequencing data to search for similar TCRs in bulk repertoire data, which potentially recognise the same antigen peptide. They introduce the notion of meta-clonotype, a T cell receptor (TCR) feature consisting of a main TCR sequence ("centroid") and a distance radius around it (+/- a CDR3 motif), with distance measured according to their previously published TCRdist method (Dash et al, 2017). The meta-clonotypes benefit from increased publicity over exact clonotype matching, and enhance the ability to find potentially relevant TCRs in repertoires from unrelated individuals, which are usually highly diverse, predominantly private, and subject to sampling constraints. The idea of meta-clonotypes is very interesting, and will provide a very useful tool in future repertoire analyses. For example, public databases of annotated TCRs (e.g. VDJdb) can be used to derive the set of meta-clonotypes for a variety of antigens, which can in turn be searched for in bulk repertoire data to identify e.g. memory to previous antigen exposure, immune status etc.

The tool for performing the analysis, tcrdist3, is open-source, well-documented with instructions and examples, and the statistical analysis has been well-thought out. It is also useful to have the comparison to the current alternative of k-mer based TCR distance (i.e. GLIPH2), and the added flexibility for the user to define the precise distance metric to be used in the tcrdist3 tool.

The authors then apply their method to analyse TCR beta sequences from COVID-19 datasets that have been publicly released by Adaptive Biotechnologies through the immuneRACE project. They use the MIRA set, the peptide-enriched set, to identify the meta-clonotypes, and then search for these in an independent cohort of COVID-19 bulk repertoires from 694 individuals. The authors find that a large proportion of the meta-clonotypes were more abundant in patients expressing the relevant restricting HLA allele, and suggest this could potentially lead to the development of disease biomarkers. The set of sars-cov-2 related meta-clonotypes is a useful resource in itself, as researchers generating other COVID-19 TCR datasets will be able to utilize this set of meta-clonotypes to search and potentially stratify patients in their own generated data.

There are a few areas were further detail / examples would strengthen the paper's claims, in particular in the application of the tcrdist3 method to the COVID-19 data.

1) Bulk TCR data from repertoires with past antigen exposure are likely to contain varying sizes of clones due to the proliferation of responding T cells and a remaining memory population. Due to the sharp drop in size between a TCR sequencing sample and the entire repertoire, clones above a particular size relative to the sample size are highly unlikely to have been sampled by chance, and identifying significantly/meaningfully expanded clonotypes in a sample is often used to identify a potentially antigen-recognising set of TCRs. The authors demonstrate the detection of meta-clonotypes in the repertoire sets, but it is somewhat unclear how the abundance of a clonotype conforming to a particular meta-clonotype is addressed. For example, there may be rationale for treating the following cases differently: meta-clonotype A is instantiated by (i) a unique clonotype with abundance 1; (ii) a single clonotype with abundance 1000; (iii) 100 different clonotypes (i.e. a "dense neighbourhood" around this meta-clonotype). If used to develop biomarkers, perhaps some degree of granularity in how the frequency/occurrence of meta-clonotypes is calculated would be helpful here.

Thanks for this helpful suggestion. We agree that the scenarios you outlined above, which differ in the level of clonal breadth (i.e. number of unique clones), may have great immunological relevance. Though we have not specifically assessed the clinical or immunological relevance of clonal breadth vs. clonal frequency, we have noted in the revision (lines 514-519) that there are multiple ways of counting meta-clonotype conformant sequences and multiple ways of aggregating counts across meta-clonotypes, for example without double counting clones that may be conformant to multiple meta-clonotypes. We have also added a documentation page about how to tabulate abundance or breadth of conforming clones: https://tcrdist3.readthedocs.io/en/latest/join.html

2) The authors focus their analysis on detecting meta-clonotypes from MIRA sets with strong evidence of HLA-restriction. They report 59.7% of these meta-clonotypes were more abundant in patients expressing the corresponding HLA allele. This means that over 40% of meta-clonotypes with strong HLA restriction were more abundant in repertoires with other HLA types. This point could be further elucidated by comparing results with the control repertoires from the COVID-19 set, from MIRA sets with low evidence of HLA restriction, or combining the sets of low and high evidence of HLA restriction (i.e. HLA agnostic results).

We’d like to clarify that the results do not imply that 40% of meta-clonotypes were more abundant in participants lacking the restricting HLA allele; rather, these meta-clonotypes did not have a significant association with presence or absence of the HLA genotype. In the discussion we highlight several of the possible explanations for this including that meta-clonotypes were too rarely detected in the population (lines 470-478). The volcano plots in Figure 6A and 7A show that there are very few if any HLA associations of the opposite sign (i.e. meta-clonotypes more abundant in patients without the restricting HLA allele). In fact, at the chosen significance threshold (FDR <0.01), 0 of 1831 predicted HLA-associated meta-clonotype were significantly negatively associated with the predicted HLA.

3) The MIRA55 set is used as an illustrative example throughout the manuscript, which familiarises the reader with this dataset as they are reading the paper. However, the claims made by the paper about MIRA sets / strong HLA evidence MIRA sets could be strengthened by providing an indication of how measured characteristics of the MIRA55 set compare to the other sets being assessed.

This is a good point and we have tried to provide as much information about MIRA55 and the other MIRA sets to help establish that MIRA55 is a representative set. Characteristics of the other MIRA sets appear in Supporting Table S6, including:

• Input number of clonotypes (AA exact)
• Number of non-redundant, public meta-clonotypes
• Clonotypes spanned by at least one meta-clonotype
• Span (% of clonotypes conforming to a meta-clonotype definition)
• Number of public enhanced sequences that match an identified TCRβ

As well as summary statistics for other meta-clonotype properties:

• Pgen
• Radius (TCRdist units)
• TRBV-CDR length
• Number of MIRA subjects contributing at least 1 sequence

Furthermore, Table S7 provides the strength of evidence for HLA-restriction for each meta-clonotype, which is then summarized by MIRA set visually in Figure 4.

Based on these criteria, we think MIRA 55 is a reasonably representative set to focus on.

4) There is some discussion throughout the manuscript about using the sars-cov-2 meta-clonotypes to identify differing clinical outcomes such as disease severity. Perhaps the dataset does not have sufficient power to allow for such sub-analysis, but a method of using meta-clonotypes to differentiate between patients based on the occurrence of meta-clonotypes in their repertoire is not provided [e.g. the number of observed clonotypes, the density distribution around clonotypes etc.)

That is true. With this manuscript we have tried to focus on establishing the methodology, evaluating the strength of the antigen-specific signal and demonstrating its potential applications; we have tried to make these goals more explicit throughout. Specifically in the revision we note that: “Much like any biomarker study, to establish a TCR-based predictor of a particular outcome, the features must be measured among a sufficiently large cohort of individuals, with a sufficient mix of outcomes.” At this time the publicly available ImmuneRACE data lack negative controls and sufficient clinical details to allow for building a predictor of SARS-CoV-2 infection or disease severity.

Author Response:

Reviewer #1 (Public Review):

This paper aims to address the question of whether the rotational dynamics in motor cortex may be due to sensory feedback signals rather than to recurrent connections and autonomous dynamics as is typically assumed. This is indeed a question of importance in neural control of movement.

The authors employ both analyses of motor cortical data and simulation analyses where a neural network is trained to perform a motor task. For the simulations, the authors use a neural network model of a brain performing arm control tasks. Importantly, in addition to the task goals, the brain also receives delayed sensory feedback from the muscle activity and kinematics of the simulated arm. The brain is modeled either using a stack of two recurrent neural networks (RNN) or using two non-recurrent neural network layers to investigate the importance of autonomous recurrent dynamics. The authors use this framework to simulate the brain performing two tasks: 1) posture perturbation task, where the arm is perturbed by external loads and has to return to original posture, and 2) delayed center-out reach task. In both tasks, the authors apply jPCA to units of the trained network, simulated muscle activity, and simulated kinematics and investigate their rotational dynamics. They find that when using an RNN in the brain model, both the RNN layers and kinematics show rotational dynamics but the muscle activity does not. Interestingly, these conclusions for both tasks also hold when networks without recurrent connections are used instead of the RNNs. Also importantly, the rotational dynamics also exist in the sensory feedback signals about the limb state (e.g. joint position, velocity). These results suggest that recurrent dynamics are not necessary for the emergence of rotational dynamics in population activity, rather sensory feedback can also achieve the same.

The authors perform similar jPCA analyses on monkey motor cortical (MC) or somatosensory cortical activity during the same two tasks and find largely consistent results. As with simulations, neural population activity and kinematics show rotational dynamics but muscle activity, which is explored only in the posture task, does not. Importantly, population activity in both motor and somatosensory cortices shows rotational dynamics. This observation is more consistent with the view that rotational dynamics emerge due to inter-region communications and processing of sensory feedback and planning, rather than autonomous dynamics within the motor cortex.

The approach of the paper is interesting and valuable and the questions being addressed are very important to the field. To further improve the paper and the analyses, there are several major comments that should be addressed to fully support the conclusions and clarify the results:

Major:

1) In the Methods, the authors explain how they model a non-recurrent network as follows: "We also examined networks where we removed the recurrent connections from each layer by effectively setting Whh, Woo to zero for the entire simulation and optimization (NO-REC networks)". However, if this is the only modification, it still leaves recurrent elements in the network. For example, if we set W_{hh} to zero, equation 2 will be:

h_{t+1} = (1-a) h_t + a tanh(W_{sh} * s_t + b_h)

where a is a constant scalar (seems to be equal to 0.5). This is indeed still a recurrent neural network since h_{t+1} depends on ht. If their explanation in the Methods is accurate, then the current approach restricts the recurrent dynamics to be a specific linear dynamic (i.e. "h{t+1} = (1-a) ht + …") but does not fully remove them. The second layer is also similar (equation 3) and will still have recurrent linear dynamics even if W{oo} is set to 0. To be able to describe networks as non-recurrent, the first terms in equations 2 and 3 (that is (1-a)h_t and (1-a)o_t) should also be set to 0. This is critical as an important argument in the paper is that non-recurrent networks can also produce rotational dynamics, so the networks supporting that argument must be fully non-recurrent. Perhaps the authors have already done this but just didn't explain it in the Methods, in which case they should clarify the Methods. However, if the current Method description is accurate, they should rerun their NO-REC simulations by also setting the fixed linear recurrent components (that is (1-a)h_t and (1-a)*o_t) to zero as explained above to have a truly non-recurrent model.

We thank the reviewer for raising this important concern. We have re-simulated the NO-REC network while removing the dynamics related to the leaky-integration component. This removal did not impact the network’s ability to perform the tasks and yielded virtually identical neural dynamics (see Figure 8). Throughout the Results we have updated the figures for the NO-REC network to the network without the leak-integration component.

2) Assuming my comment in 1 is addressed and the results stay similar, the authors show in simulations that even without recurrent dynamics (referred to as the NO-REC case), rotational dynamics are observed in the simulated brain during both tasks (Figure 8). This result is used to suggest that the sensory feedback is what causes the rotational dynamics in the brain model in this case. However, I think to fully demonstrate the role of feedback, additional simulations are also needed where the sensory feedback is removed from the brain model. In other words, what would happen if recurrent and non-recurrent brain models are trained to perform the tasks but are not provided with the sensory feedback (only receive task goals)? One would expect the recurrent model to still be able to perform the task and autonomously produce similar rotational dynamics (as has been shown in prior work), but the non-recurrent model to fail in doing the task well and in showing rotational dynamics. I think adding such simulations without the feedback signals would really strengthen the paper and help its message.

We apologize if the network architecture was not clear. In the case of the NO-REC network the only way they can generate the time-varying signals needed for the tasks is through sensory feedback. The network simply will not work without recurrent AND sensory feedback. For the posture task there are no additional inputs since it only receives sensory feedback. For the reaching task the task-goal input is static and the GO cue turns off on a timescale considerably shorter (~20ms) than the reach duration. Thus, the REC network would always perform better than the NO-REC network when sensory feedback was removed as the NO-REC network cannot generate any dynamics. We have now included in the Results the following statement. "Note, by removing the recurrent connections these networks can only generate time-varying outputs by exploiting the time-varying sensory inputs from the limb." (line 345-347).

We have also now included simulations to highlight how REC networks that receive sensory feedback are able to generalize better to scenarios with increased motor noise than REC networks where sensory feedback is either completely removed (reaching task) or only provided at the beginning of the trial (posture task) (Figure S8). Thus, sensory feedback makes REC networks more robust in less predictable scenarios.

We agree that this could be an interesting manipulation and have now included manipulations of the sensory feedback delays. We considered three separate delays, 0ms, 50ms and 100ms and found that there was a dependence on the rotational frequency of the top jPC plane with greater delays resulting in a general reduction in frequency (see now Supplementary Figure 10). There was less effect of delay on fit qualities to the constrained and unconstrained dynamical system. This has been added to the Results section (line 423-446).

We simulated this scenario and found the answer to be rather complex and we have added these results to the supplementary material. The network's behavioural performance in the perturbation posture task is similar to the previous networks with joint-based feedback. However, the dynamics in the output layer are not the same with a clear reduction in how well the dynamics are described as rotational (Figure S11A-B).

Oddly, rotational dynamics could still be observed in the input layer dynamics (data now shown) and the kinematic signals when they were converted to a cartesian reference frame (Figure S11D-E). Furthermore, rotational dynamics could emerge in the output layer if we used a different initialization method for the network weights. We initialized weights from a uniform distribution bound from ±1/√N, where N is the number of units. In contrast, previous studies have initialized network weights using a Gaussian distribution with standard deviation equal to g/√N where g is constant larger than 1. This alternative initialization scheme encourages strong intrinsic dynamics often needed for autonomous RNN models (Sussillo et al., 2015). We found networks initialized with this method and trained on the perturbation posture task exhibited stronger rotational dynamics with fits to the constrained and unconstrained dynamical systems of 0.5 and 0.88, respectively (Figure S11C-D). When examining the reaching task, we found similar results (Figure S11F-K). When initialized with a uniform distribution, fit quality for the constrained and unconstrained dynamical systems were 0.4 and 0.77, respectively (Figure S11F-G), which were smaller than for the joint-based feedback (Figure 7B, constrained R2=0.7, unconstrained R2=0.83). Qualitatively, the dynamics were different when the network was initialized with a Gaussian distribution (Figure S11H), however fit qualities were comparable between the two initialization methods (Figure S11 I). There was also a noticeable reduction in the fit quality for the kinematic signals particularly for the constrained dynamical system (Figure S11K, constrained R2=0.36, unconstrained R2=0.77). These findings have been added to the Results

3) A measure of how well each trained network is able to perform the task should be provided. For example, is the non-recurrent network able to perform the tasks as accurately as the recurrent models? The authors could use an appropriate measure, for example average displacement in the posture task and time-to-target in the center-out task, to objectively quantify task performance for each network. Another performance measure could be the first term of the loss in equation 5. Also, plots of example trials that show the task performance should be provided for the non-recurrent networks (for example by adding to Figure 8), similar to how they are shown for the recurrent models in Figures 2 and 6.

We have now presented and quantified the NO-REC network behavioural performance. Kinematics for the NO-REC network are shown in Figure S7A-C and E-G which are comparable to the REC network. Furthermore, quantifying the maximum displacement during the posture task yielded no obvious differences between the NO-REC and REC networks (Figure S7D). For the reaching task, the time-to-target was noticeably more variable and tended to be slower for the NO-REC network (Figure S7H). These observations have been added to the Results.

4) An important observation is that rotational dynamics also exist in the sensory signals about the limb state. This may imply that the task structure that dictates the limb state and thus the associated sensory feedback may play an important role in the rotations without the recurrent connections. While the present study will be a valuable addition regardless of what the answer is, this is an important point to address: What is the role of the task structure in producing rotational dynamics? In both the posture task and the center-out task, the task instruction instructs subjects to return to the initial movement 'state' by the end of the trial: in the posture task the simulated arm needs to return to the original posture upon disturbance, and in the center out task the arm needs to start from zero velocity and settle at the target with zero velocity. Is this structure what's causing the rotational dynamics? This is an important question both for this paper and for the field and the authors have a great simulation setup to explore it. For example, what happens if the task instructions u* instruct the arm to follow a random trajectory continuously, instead of stopping at some targets? With a simulated tracking task like this, one could eliminate obvious cases of return-to-original-state from the task. Would the network still produce rotational dynamics? Of course, I don't expect the authors to collect experimental monkey data for such new tasks, rather to just change the task instructions in their numerical simulations to explore the dependence of observed rotational dynamics on the task structure. I think this will help the message of the paper and can be very useful for the field.

We agree that a tracking task would be an interesting manipulation and have simulated this with the REC and NO-REC networks (Figure 9). Here, we trained up the network to reach from the starting position and track a target moving radially at a constant velocity for the rest of the trial (1.2seconds). Thus, the network has to move the limb at a constant velocity. We found there was a consistent reduction in how well the network’s dynamics (constrained R2=0.13, unconstrained R2=0.3) were described as rotational when compared to the previous reaching task (Figure 7, constrained R2=0.7, unconstrained R2=0.83). Also, note that this reduction in rotational dynamics remained even when we initialized the network weights using a Gaussian distribution (see Essential revision 2.3). These simulations have been added to the Results section.

5) It would be beneficial if the authors could elaborate in the discussion on intuitive explanations of why sensory feedback can produce rotational dynamics even with no internal recurrent dynamics in the brain model. To me, it seems like sensory feedback is providing a path for recurrence to exist in the overall brain-arm system, so the non-recurrent neural networks can learn to exploit that path to effectively implement some recurrent dynamics. Some intuitive explanations like this will be helpful for readers.

The main reason why rotational dynamics emerge in sensory feedback is due to the phase offset between the joint position and velocity as changes first occur in the velocity followed by position (see pendulum example Pandarinath et al., 2018a also DeWolf et al., 2016; Susilaradeya et al., 2019). This phase offset is maintained across reach directions and gives rise to the orderly rotational dynamics observed in the kinematic signals (DeWolf et al., 2016; Pandarinath et al., 2018a; Susilaradeya et al., 2019; Vyas et al., 2020). Furthermore, the tracking task disrupted this phase relationship and thus the rotational dynamics were substantively reduced in the network models. This text has been added to the Discussion (lines 519-526).

6) One main result in data from non-human primates is that there exist rotations also in the somatosensory cortex not just in motor cortex. A more thorough discussion of prior work on rotational dynamics or lack thereof across brain regions and behavioral tasks is important to add here. For example, besides the works cited by the authors, there are other works such as (Kao et al., 2015; Gao et al., 2016; Remington et al., 2018; Stavisky et al., 2019; Aoi et al., 2020; Sani et al., 2021) that discuss or show rotational dynamics in various brain regions and behavioral tasks and should be cited and discussed.

We have cited the above papers and included in the Discussion the following paragraph (lines 537-549) “Importantly, findings of rotational dynamics in cortical circuits are not trivial. Activity in the supplementary motor area does not exhibit rotational dynamics during reaching (Lara et al., 2018). The hand area of MC also does not exhibit rotational dynamics during grasping-only behaviour (Suresh et al., 2020), though it does exhibit rotational dynamics during reach-to-grasp (Abbaspourazad et al., 2021; Rouse and Schieber, 2018) which may reflect the reaching component of the behaviour. More broadly there is a growing body of work characterizing cortical neural dynamics across different behavioural tasks which have revealed rotational (Abbaspourazad et al., 2021; Aoi et al., 2020; Libby and Buschman, 2021; Remington et al., 2018; Sohn et al., 2019; Stavisky et al., 2019), helical (Russo et al., 2020), stationary (Machens et al., 2010), and ramping dynamics (Finkelstein et al., 2021; Kaufman et al., 2016; Machens et al., 2010) and these dynamics appear to support various classes of computations. Thus, finding rotational dynamics across the fronto-parietal circuit in our study is not trivial."

7) The authors state that "In contrast, rotational dynamics appear to be absent in… MC activity during grasping driven by sensory inputs (Suresh et al., 2020)." There are other papers that study dynamics during reach-grasps and still finds rotational dynamics and modes (Abbaspourazad et al., 2021; Vaidya et al., 2015) and should be cited and discussed. The recent paper on naturalistic reach-grasps (Abbaspourazad et al., 2021) also argues for the involvement of a large-scale network in these movements, which further supports the authors' interpretation that "This interpretation of motor control emphasizes that the objective of the motor system is to attain the behavioural goal and this requires feedback processed by a distributed network." A discussion of this point made in this recent paper in the intro/discussion is important. Finally, there is a recent paper that argues for the input-driven nature of motor cortex (Sauerbrei et al., 2020) and is cited/discussed by the authors but briefly and mainly in the discussion. I think given the relevance of this recent paper to the core message here, it should also be briefly discussed in the introduction to better set up the work.

We agree with the reviewer that there are discrepancies between the motor cortical dynamics reported by Suresh et al. 2020 and Abbaspourazad et al., 2021 during grasping tasks. This difference may reflect differences in task as in Suresh et al. 2020 the monkeys grasped objects whereas in Abbaspourazad et al., 2021 monkeys had to reach and grasp objects. Thus, rotations may reflect the reaching component of the behaviour. This has been elaborated on in the Discussion which now reads (lines 539-542) “The hand area of MC also does not exhibit rotational dynamics during grasping-only behaviour (Suresh et al., 2020), though it does exhibit rotational dynamics during reach-to-grasp (Abbaspourazad et al., 2021; Rouse and Schieber, 2018; Vaidya et al., 2015) which may reflect the reaching component of the behaviour.”.

We have also briefly mentioned the findings by Sauerbrei et al. 2020 in the Introduction which now reads (line 79-81) “Lastly a recent study demonstrates that motor cortical dynamics are driven by inputs coming from motor thalamus (Sauerbrei et al., 2020)."

Minor:

1) The Methods are clear and comprehensive, but just to make understanding of the simulation setup easier, it would help to have a diagram of the computation graph for the recurrent and non-recurrent networks that shows their number of units, activations/nonlinearities, RNN cell type, etc., added as supplementary figure.

We agree that this is useful and have added it to Figure 1

2) Again, to help more clearly convey the simulations, it would help to show the task goals (x*) that are inputs to the simulated brain for example trials in each task (for example added to Figures 2 and 6).

We agree that this is useful and have added it to Figure 1

3) Similar to how VAF is shown on top of all plots of jPC planes, it would be helpful to have the rotation frequency for each jPC plane noted next to it. Currently it is difficult to find the jPC frequency associated with each plot from the text.

We agree and have added it to the appropriate figures

4) I am a bit surprised by how different the null distributions are for modeling muscle activity (Figure 3F) and kinematics (Figure 3H). The null distribution is simply the R2 for a constrained or unconstrained dynamic model fit to a subsampled version of the neural activity. The only difference between the null distributions in Figure 3F and Figure 3H seems to be the downsampled dimension, which for muscle activity is 6 and for kinematics is 4 (per equation 1). Any insight will be welcome as to why down sampling the population activity to 4 (Figure 3H) results in so much worse R2 compared with down sampling it to 6 (Figure 3F)?

We thank the reviewer for raising this concern. Originally, we had applied PCA to reduce the dimensionality of the kinematic signals from 4 dimensions to 2, and the muscle signals from 6 to 4. We realize now that to be more conservative in our significance testing, we should use the full dimensionality of the kinematic and muscle signals. As such, we have changed the figures throughout to reflect this.

References:

Abbaspourazad, H., Choudhury, M., Wong, Y.T., Pesaran, B., Shanechi, M.M., 2021. Multiscale low-dimensional motor cortical state dynamics predict naturalistic reach-and-grasp behavior. Nature Communications 12, 607. https://doi.org/10.1038/s41467-020-20197-x

Aoi, M.C., Mante, V., Pillow, J.W., 2020. Prefrontal cortex exhibits multidimensional dynamic encoding during decision-making. Nature Neuroscience 1-11. https://doi.org/10.1038/s41593-020-0696-5

Gao, Y., Archer, E.W., Paninski, L., Cunningham, J.P., 2016. Linear dynamical neural population models through nonlinear embeddings, in: Lee, D.D., Sugiyama, M., Luxburg, U.V., Guyon, I., Garnett, R. (Eds.), Advances in Neural Information Processing Systems 29. Curran Associates, Inc., pp. 163-171.

Kao, J.C., Nuyujukian, P., Ryu, S.I., Churchland, M.M., Cunningham, J.P., Shenoy, K.V., 2015. Single-trial dynamics of motor cortex and their applications to brain-machine interfaces. Nature Communications 6, 7759. https://doi.org/10.1038/ncomms8759

Remington, E.D., Narain, D., Hosseini, E.A., Jazayeri, M., 2018. Flexible Sensorimotor Computations through Rapid Reconfiguration of Cortical Dynamics. Neuron 98, 1005-1019.e5. https://doi.org/10.1016/j.neuron.2018.05.020

Sani, O.G., Abbaspourazad, H., Wong, Y.T., Pesaran, B., Shanechi, M.M., 2021. Modeling behaviorally relevant neural dynamics enabled by preferential subspace identification. Nature Neuroscience 24, 140-149. https://doi.org/10.1038/s41593-020-00733-0

Stavisky, S.D., Willett, F.R., Wilson, G.H., Murphy, B.A., Rezaii, P., Avansino, D.T., Memberg, W.D., Miller, J.P., Kirsch, R.F., Hochberg, L.R., Ajiboye, A.B., Druckmann, S., Shenoy, K.V., Henderson, J.M., 2019. Neural ensemble dynamics in dorsal motor cortex during speech in people with paralysis. eLife 8, e46015. https://doi.org/10.7554/eLife.46015

Vaidya, M., Kording, K., Saleh, M., Takahashi, K., Hatsopoulos, N.G., 2015. Neural coordination during reach-to-grasp. Journal of Neurophysiology 114, 1827-1836. https://doi.org/10.1152/jn.00349.2015

Reviewer #1 (Public Review):

This paper aims to address the question of whether the rotational dynamics in motor cortex may be due to sensory feedback signals rather than to recurrent connections and autonomous dynamics as is typically assumed. This is indeed a question of importance in neural control of movement.

The authors employ both analyses of motor cortical data and simulation analyses where a neural network is trained to perform a motor task. For the simulations, the authors use a neural network model of a brain performing arm control tasks. Importantly, in addition to the task goals, the brain also receives delayed sensory feedback from the muscle activity and kinematics of the simulated arm. The brain is modeled either using a stack of two recurrent neural networks (RNN) or using two non-recurrent neural network layers to investigate the importance of autonomous recurrent dynamics. The authors use this framework to simulate the brain performing two tasks: 1) posture perturbation task, where the arm is perturbed by external loads and has to return to original posture, and 2) delayed center-out reach task. In both tasks, the authors apply jPCA to units of the trained network, simulated muscle activity, and simulated kinematics and investigate their rotational dynamics. They find that when using an RNN in the brain model, both the RNN layers and kinematics show rotational dynamics but the muscle activity does not. Interestingly, these conclusions for both tasks also hold when networks without recurrent connections are used instead of the RNNs. Also importantly, the rotational dynamics also exist in the sensory feedback signals about the limb state (e.g. joint position, velocity). These results suggest that recurrent dynamics are not necessary for the emergence of rotational dynamics in population activity, rather sensory feedback can also achieve the same.

The authors perform similar jPCA analyses on monkey motor cortical (MC) or somatosensory cortical activity during the same two tasks and find largely consistent results. As with simulations, neural population activity and kinematics show rotational dynamics but muscle activity, which is explored only in the posture task, does not. Importantly, population activity in both motor and somatosensory cortices shows rotational dynamics. This observation is more consistent with the view that rotational dynamics emerge due to inter-region communications and processing of sensory feedback and planning, rather than autonomous dynamics within the motor cortex.

The approach of the paper is interesting and valuable and the questions being addressed are very important to the field. To further improve the paper and the analyses, there are several major comments that should be addressed to fully support the conclusions and clarify the results:

Major:

1) In the Methods, the authors explain how they model a non-recurrent network as follows: "We also examined networks where we removed the recurrent connections from each layer by effectively setting Whh, Woo to zero for the entire simulation and optimization (NO-REC networks)". However, if this is the only modification, it still leaves recurrent elements in the network. For example, if we set W_{hh} to zero, equation 2 will be:

h_{t+1} = (1-a) * h_t + a * tanh(W_{sh} * s_t + b_h)

where a is a constant scalar (seems to be equal to 0.5). This is indeed still a recurrent neural network since h_{t+1} depends on h_t. If their explanation in the Methods is accurate, then the current approach restricts the recurrent dynamics to be a specific linear dynamic (i.e. "h_{t+1} = (1-a) * h_t + ...") but does not fully remove them. The second layer is also similar (equation 3) and will still have recurrent linear dynamics even if W_{oo} is set to 0. To be able to describe networks as non-recurrent, the first terms in equations 2 and 3 (that is (1-a)*h_t and (1-a)*o_t) should also be set to 0. This is critical as an important argument in the paper is that non-recurrent networks can also produce rotational dynamics, so the networks supporting that argument must be fully non-recurrent. Perhaps the authors have already done this but just didn't explain it in the Methods, in which case they should clarify the Methods. However, if the current Method description is accurate, they should rerun their NO-REC simulations by also setting the fixed linear recurrent components (that is (1-a)*h_t and (1-a)*o_t) to zero as explained above to have a truly non-recurrent model.

2) Assuming my comment in 1 is addressed and the results stay similar, the authors show in simulations that even without recurrent dynamics (referred to as the NO-REC case), rotational dynamics are observed in the simulated brain during both tasks (Figure 8). This result is used to suggest that the sensory feedback is what causes the rotational dynamics in the brain model in this case. However, I think to fully demonstrate the role of feedback, additional simulations are also needed where the sensory feedback is removed from the brain model. In other words, what would happen if recurrent and non-recurrent brain models are trained to perform the tasks but are not provided with the sensory feedback (only receive task goals)? One would expect the recurrent model to still be able to perform the task and autonomously produce similar rotational dynamics (as has been shown in prior work), but the non-recurrent model to fail in doing the task well and in showing rotational dynamics. I think adding such simulations without the feedback signals would really strengthen the paper and help its message.

3) A measure of how well each trained network is able to perform the task should be provided. For example, is the non-recurrent network able to perform the tasks as accurately as the recurrent models? The authors could use an appropriate measure, for example average displacement in the posture task and time-to-target in the center-out task, to objectively quantify task performance for each network. Another performance measure could be the first term of the loss in equation 5. Also, plots of example trials that show the task performance should be provided for the non-recurrent networks (for example by adding to Figure 8), similar to how they are shown for the recurrent models in Figures 2 and 6.

4) An important observation is that rotational dynamics also exist in the sensory signals about the limb state. This may imply that the task structure that dictates the limb state and thus the associated sensory feedback may play an important role in the rotations without the recurrent connections. While the present study will be a valuable addition regardless of what the answer is, this is an important point to address: What is the role of the task structure in producing rotational dynamics? In both the posture task and the center-out task, the task instruction instructs subjects to return to the initial movement 'state' by the end of the trial: in the posture task the simulated arm needs to return to the original posture upon disturbance, and in the center out task the arm needs to start from zero velocity and settle at the target with zero velocity. Is this structure what's causing the rotational dynamics? This is an important question both for this paper and for the field and the authors have a great simulation setup to explore it. For example, what happens if the task instructions u* instruct the arm to follow a random trajectory continuously, instead of stopping at some targets? With a simulated tracking task like this, one could eliminate obvious cases of return-to-original-state from the task. Would the network still produce rotational dynamics? Of course, I don't expect the authors to collect experimental monkey data for such new tasks, rather to just change the task instructions in their numerical simulations to explore the dependence of observed rotational dynamics on the task structure. I think this will help the message of the paper and can be very useful for the field.

5) It would be beneficial if the authors could elaborate in the discussion on intuitive explanations of why sensory feedback can produce rotational dynamics even with no internal recurrent dynamics in the brain model. To me, it seems like sensory feedback is providing a path for recurrence to exist in the overall brain-arm system, so the non-recurrent neural networks can learn to exploit that path to effectively implement some recurrent dynamics. Some intuitive explanations like this will be helpful for readers.

6) One main result in data from non-human primates is that there exist rotations also in the somatosensory cortex not just in motor cortex. A more thorough discussion of prior work on rotational dynamics or lack thereof across brain regions and behavioral tasks is important to add here. For example, besides the works cited by the authors, there are other works such as (Kao et al., 2015; Gao et al., 2016; Remington et al., 2018; Stavisky et al., 2019; Aoi et al., 2020; Sani et al., 2021) that discuss or show rotational dynamics in various brain regions and behavioral tasks and should be cited and discussed.

7) The authors state that "In contrast, rotational dynamics appear to be absent in... MC activity during grasping driven by sensory inputs (Suresh et al., 2020)." There are other papers that study dynamics during reach-grasps and still finds rotational dynamics and modes (Abbaspourazad et al., 2021; Vaidya et al., 2015) and should be cited and discussed. The recent paper on naturalistic reach-grasps (Abbaspourazad et al., 2021) also argues for the involvement of a large-scale network in these movements, which further supports the authors' interpretation that "This interpretation of motor control emphasizes that the objective of the motor system is to attain the behavioural goal and this requires feedback processed by a distributed network." A discussion of this point made in this recent paper in the intro/discussion is important. Finally, there is a recent paper that argues for the input-driven nature of motor cortex (Sauerbrei et al., 2020) and is cited/discussed by the authors but briefly and mainly in the discussion. I think given the relevance of this recent paper to the core message here, it should also be briefly discussed in the introduction to better set up the work.

Minor:

1) The Methods are clear and comprehensive, but just to make understanding of the simulation setup easier, it would help to have a diagram of the computation graph for the recurrent and non-recurrent networks that shows their number of units, activations/nonlinearities, RNN cell type, etc., added as supplementary figure.

2) Again, to help more clearly convey the simulations, it would help to show the task goals (x*) that are inputs to the simulated brain for example trials in each task (for example added to Figures 2 and 6).

3) Similar to how VAF is shown on top of all plots of jPC planes, it would be helpful to have the rotation frequency for each jPC plane noted next to it. Currently it is difficult to find the jPC frequency associated with each plot from the text.

4) I am a bit surprised by how different the null distributions are for modeling muscle activity (Figure 3F) and kinematics (Figure 3H). The null distribution is simply the R2 for a constrained or unconstrained dynamic model fit to a subsampled version of the neural activity. The only difference between the null distributions in Figure 3F and Figure 3H seems to be the downsampled dimension, which for muscle activity is 6 and for kinematics is 4 (per equation 1). Any insight will be welcome as to why down sampling the population activity to 4 (Figure 3H) results in so much worse R2 compared with down sampling it to 6 (Figure 3F)?

References:

Abbaspourazad, H., Choudhury, M., Wong, Y.T., Pesaran, B., Shanechi, M.M., 2021. Multiscale low-dimensional motor cortical state dynamics predict naturalistic reach-and-grasp behavior. Nature Communications 12, 607. https://doi.org/10.1038/s41467-020-20197-x

Aoi, M.C., Mante, V., Pillow, J.W., 2020. Prefrontal cortex exhibits multidimensional dynamic encoding during decision-making. Nature Neuroscience 1-11. https://doi.org/10.1038/s41593-020-0696-5

Gao, Y., Archer, E.W., Paninski, L., Cunningham, J.P., 2016. Linear dynamical neural population models through nonlinear embeddings, in: Lee, D.D., Sugiyama, M., Luxburg, U.V., Guyon, I., Garnett, R. (Eds.), Advances in Neural Information Processing Systems 29. Curran Associates, Inc., pp. 163-171.

Kao, J.C., Nuyujukian, P., Ryu, S.I., Churchland, M.M., Cunningham, J.P., Shenoy, K.V., 2015. Single-trial dynamics of motor cortex and their applications to brain-machine interfaces. Nature Communications 6, 7759. https://doi.org/10.1038/ncomms8759

Remington, E.D., Narain, D., Hosseini, E.A., Jazayeri, M., 2018. Flexible Sensorimotor Computations through Rapid Reconfiguration of Cortical Dynamics. Neuron 98, 1005-1019.e5. https://doi.org/10.1016/j.neuron.2018.05.020

Sani, O.G., Abbaspourazad, H., Wong, Y.T., Pesaran, B., Shanechi, M.M., 2021. Modeling behaviorally relevant neural dynamics enabled by preferential subspace identification. Nature Neuroscience 24, 140-149. https://doi.org/10.1038/s41593-020-00733-0

Stavisky, S.D., Willett, F.R., Wilson, G.H., Murphy, B.A., Rezaii, P., Avansino, D.T., Memberg, W.D., Miller, J.P., Kirsch, R.F., Hochberg, L.R., Ajiboye, A.B., Druckmann, S., Shenoy, K.V., Henderson, J.M., 2019. Neural ensemble dynamics in dorsal motor cortex during speech in people with paralysis. eLife 8, e46015. https://doi.org/10.7554/eLife.46015

Vaidya, M., Kording, K., Saleh, M., Takahashi, K., Hatsopoulos, N.G., 2015. Neural coordination during reach-to-grasp. Journal of Neurophysiology 114, 1827-1836. https://doi.org/10.1152/jn.00349.2015

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Manuscript number: RC-2021-00831

Corresponding author(s): Lu, Gan

Reviewer comments are in regular font. Our rebuttal is in bolded font. Experiments that we plan for the full revision are preceded with “FULL:”. In the revision files, the changes are highlighted in yellow.

General Statements

We thank the reviewers for their detailed feedback. There are two major concerns. First, the manuscript lacks functional analysis of the meiotic triple helices (MTHs). Second, the manuscript makes claims about the properties of synaptonemal complexes (SCs) and MTHs that are inadequately supported. In order to address the first concern, we would need extensive experiments to first identify and then perturb the genes associated with the MTH. Such experiments are beyond the scope of this manuscript and are the focus of future studies. We will address the second concern with mostly text revisions. We will also improve some of the imaging analysis with new experiments that can be done in a few months’ time.

2. Description of the planned revisions

We will acquire new cryo-ET data of pachytene-arrested cell cryolamellae using our new K3-GIF camera. These new data have higher signal-to-noise ratios and allow us to generate a higher-resolution subtomogram average of the MTHs. The achievable resolution will depend on the conformational homogeneity of the MTH segments and on the number of cryotomograms we can capture. If we are able to achieve a subnanometer-resolution reconstruction, we will narrow down the possible identities of the subunits. Even if we cannot achieve subnanometer resolutions, the new data will allow us to test if ladder-like densities were missed in our lower-resolution older data, thereby improving our understanding of the SC’s structure. We will also perform subtomogram analysis of purified ribosomes as a control to strengthen our handedness determination.

3. Description of the revisions that have already been incorporated in the transferred manuscript

The Reviewers’ original comments are reproduced in regular font. Line numbers refer to the preliminary revision.

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

Ma and coworkers report studies of budding yeast undergoing meiosis by cryo-ET. They fail to detect structures interpretable as synaptonemal complex, and instead detect feather-like bundles of what appears to be a triple helix. These structures do not appear to be related to the synaptonemal complex, as spo11 mutants that do not initiate recombination, red1 mutants that lack axial elements, and zip1 mutants that lack the central element of the SC still make these bundles. These structures are absent from cells treated with latrunculin A, which depolymerizes actin filaments, but expected structures are not visible in light microscopy of cells treated with two different F-actin-staining reagents. However, it should be noted that another study (Takagi et al, 2021, bioRxiv) did detect actin associated with these structures by immunogold labeling. The structures are also reversibly dissolved in 7% hexanediol.

This part of the paper's findings is well supported by data and is certainly of interest, although interest is somewhat limited by the unknown nature of these structures-what they contain, let alone their function, remains to be determined-in fact, it is not even determined whether or not they are made of protein. However, as an initial report of a previously unknown phenomenon, the paper is of some value.

__Thank you for raising the issue of whether MTHs are composed of protein or not. Aside from the proteins, the only other materials capable of forming large bundles of linear polymers are polysaccharides and DNA. Yeast polysaccharides are found in the cell wall, so they are unlikely to be a candidate for the MTHs. In the nucleus, DNA is abundant. While we favor that MTHs are composed of protein, we cannot rule out that the MTHs are non-chromatin DNA-protein complexes. Depending on the resolution of future subtomogram averages, we may get a better idea of the MTH’s composition.__

There is, unfortunately, a second aspect of the paper that cannot be supported. Although it is clear that synaptonemal complex is present in the cells examined (by standard cytological methods) the authors cannot detect structures consistent with SC in their cryo-ET images. Unfortunately, authors then extrapolate from their inability to detect SC to conclusions about SC, such as that it is not crystalline, and even go so far as to suggest that their failure to detect SC invalidates two models for crossover interference, and that the ladder-like structure reported for SC in many organisms using many difference approaches may be a fixation artifact. Authors show remember that the absence of evidence is not evidence for absence; the speculation described above should be removed from both the abstract and discussion.

We have removed the speculation about crossover interference and limited the scope of our discussion on SCs to budding yeast only.

The differences between traditional EM and cryo-EM are not trivial. In the introduction, we added more details to explain the differences in both the sample preparation and contrast generation:

Lines 79-87: “Meiotic nuclei have been studied for decades by traditional EM (Fawcett, 1956; Moses, 1956), but not by cryo-ET. Cryo-ET can reveal 3-D nanoscale structural details of cellular structures in a life-like state because the samples are kept unfixed, unstained, and frozen-hydrated during all stages of sample preparation and imaging (Ng and Gan, 2020). The densities seen in cryo-ET data come from electron scattering of the biological macromolecules. In comparison, the densities seen in traditional EM from electron scattering of heavy metals such as uranium, tungsten, and osmium, which have adhered to a subset of biological macromolecules that were not extracted in earlier steps.”

We have also removed the term ‘artifact’ from lines 201-202:

Original: “The ladder model is based on images of traditional EM samples, which are vulnerable to fixation and staining artifacts.”

Revised: “The ladder model is based on images of traditional EM samples.”

Reviewer #1 (Significance (Required)):

This paper reports a previously unreported structure in the nuclei of yeast cells undergoing meiosis. The composition and function of this structure remain to be determined. This considerably limits the significance of the paper.

**Referee Cross-commenting**

I agree with the concerns of the other reviewers. I also agree with reviewer 3 that to raise the significance of the paper would require much work. But I think that the raw observation is of value, albeit in a journal of record. So I would stick with my recommendation of text changes, keeping in mind that there may not be a suitable journal in LSA's portfolio.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

This work describes helical filamentous structures observed in budding yeast nuclei that were cryosectioned and imaged using cryo-electron tomography (cryoET).

The goal of this work seems to have been to conduct an ultrastructural analysis of the synaptonemal complex (SC), a meiosis-specific protein structure that holds chromosomes together during meiosis and is thought to regulate meiotic recombination. In conventional TEM images of fixed, embedded, and stained sections obtained from pachytene nuclei, SCs usually appear as long, thin, transversely striated structures. At pachytene, SCs extend along the full length of a thin (100-nm) gap between paired chromosomes (typically 1-6 µm long in yeast cells). Surprisingly, the authors did not observe SCs, perhaps because these structures do not produce much contrast in cryo-EM images. Instead, they observe abundant triple helical structures in the nucleoplasm, which they designate as "meiotic triple helices" (MTHs). The authors report that these triple helices assemble at the same time as but independently of the SC. They publised a preprint in which they indicated that these structures were somehow related to SCs, but this the revised version reports that they appear independently of SCs. They further report that treatment of cells with the F-actin depolymerizing drug Latrunculin-A (LatA) resulted in a lack of detectable triple helical structures in the nucleus, suggesting that these structures may be a form of actin or, alternatively, that they may require F-actin for their assembly.

While the work is technically mostly sound, its significance is unclear because the reported structures have no known function. Many, if not most proteins will form helical structures if their concentration rises above a threshold defined by their binding affinity for themselves (see https://doi.org/10.1186/1741-7007-11-119 and references therein), so this may simply be an example of an abundant nuclear protein that polymerizes to form helical filaments under the conditions that trigger yeast sporulation.

Thank you for raising the interesting possibility that MTHs form helices because their subunits have exceeded a critical concentration threshold. In the revised text, we discuss the possibility that the MTH is simply a protein that is highly expressed in meiotic cells and polymerize either due to exceeding a critical concentration, or having undergone a biochemical change like a post-translational modification:

Lines 408-415: “Note it is possible that the MTHs may not be directly involved in meiosis, but are instead a protein that is at a sufficient concentration or has the right biochemical modifications to form helices in pachytene because it is known that many proteins can form a helix under the right conditions {Crane, 1950; Pauling, 1953; Theriot, 2013}. These MTHs also have lateral interactions that allow them to pack with crystalline density. Their sensitivity to 1,6-hexanediol suggests that the polymerization both within and between MTHs are based on hydrophobic interactions. Further work will be needed to determine the identity of the MTH’s subunits and their potential function.”

The authors perform cryosectioning and cryoEM on yeast cells undergoing meiosis to show that the assembly and disassembly of MTHs follow a similar time course as that of the SC. The observation of these triple-helical filaments in meiotic nuclei has also been reported in another study (https://www.biorxiv.org/content/10.1101/778100v2.full.pdf+html), which proposed, based on immuno-EM labeling, that they may be actin cables. This study reports that the structures are not detected using phalloidin or Lifeact-mCherry. However, treatment with LatA did eliminate detection of MTH structures, suggesting that they may be comprised of actin.

In my view, there are a number of issues that should be addressed before publication. Many of these relate to the presentation of the findings. Detailed comments below:

1. The presentation of the work is very confusing. The authors clearly expected to observe SC structures, but did not. They conclude that MTHs are not SCs, since they do not depend on molecular components required for SC assembly. They should describe their findings in a more straightforward way rather than veering from introducing the SC to describing the MTHs.

We have restructured the manuscript to tone down the discussion about the SC. However, we have to start with the SC because it is the most iconic feature of pachytene cells and a major organizer of chromatin in meiosis. Furthermore, its presence, as indicated by Zip1-GFP signals, was key to establishing that our cells were indeed arrested in pachytene. It would have been confusing to then overlook the absence of structural features as conspicuous and prevalent as what was expected of SCs. The other sections did have room for improvement. In the sections below, we describe the changes point by point.

Similarly, the discussion section on "recombination and chromosome segregation" seems inappropriate and irrelevant, since no data are presented in this study regarding the functions of the MTHs, and there is no reason to think that they contribute to crossover interference, chromosome segregation, or other aspects of meiosis. Additionally, most of the ideas presented in this section seem very muddled. I recommend deleting this section.

We have deleted the section entitled "Recombination and chromosome segregation".

Throughout the text, we have also changed the term “meiosis-specific” to “meiosis-related” when describing MTHs. Doing so allows for the possibility that MTHs might just form as a consequence of being expressed to a high enough concentration as discussed above.

Along the same lines as comment #1: The title should be changed - absence of evidence for "ladders" is not evidence of absence. Prior work using TEM and superresolution fluorescence microscopy has clearly shown that ladder-like SC structures exist in pachytene nuclei of budding yeast and many other organisms, although they apparently cannot be visualized using the methods described here.

We have changed the title to be less forceful, yet report what we see and don’t see by cryo-ET:

“Cryo-ET detects bundled triple helices but not ladders in meiotic budding yeast”

The authors should clarify whether cryo-sectioning was performed through the full volume of pachytene nuclei.

This comment refers to our attempt at serial cryosectioning, as shown in Fig. S8. We have revised the text in lines 332-334 to reflect the estimated volume covered:

“We successfully reconstructed six sequential sections from one ndt80Δ cell (Figure S8), which represents approximately one third of a nucleus (assuming a spherical shape).”

We also changed Fig. S8’s title so that it doesn’t sound like we reconstructed an entire nucleus:

“MTH bundles are extensive throughout the cell nucleus.” → “MTH bundles are extensive.”

It is not clear from the manuscript which camera/microscope configuration was used to acquire the cryoET data that were used for sub-tomogram averaging. The authors state in the methods that Falcon II and K3-GIF was used for projection images, but it's not clear if this applies to all images. These technical details should be clarified.

We have added a new column to Table S4 that reports the camera used for all the projection imaging and tomography experiments. In the original MS, all of the subtomogram averaging was done using Falcon II data.

FULL: In the full revision, we plan to incorporate new subtomogram averages of MTHs in situ, using K3-GIF data of cryolamellae.

The analysis of the handedness of the helices seems to be questionable as the resolution of the reconstructions for 80S are also quite low. I am uncertain whether this can be used to state with confidence that the MTH are right-handed.

FULL: In the full revision, we will use purified 70S ribosomes, imaged on the same K3-GIF camera and using the same software workflow as for the new subtomogram averaging of MTHs in situ. We expect higher resolution for both ribosomes and MTHs, which will make the handedness assignment unambiguous.

The authors claim that treatment with Latrunculin-A (LatA) leads to disappearance of MTHs. However, they support this with projection images of cells treated with LatA. The projection images are of poor quality and the vitrification in these cells (as well as the DMSO treated cells) do not look appropriate. They should present data for LatA-treated cells and DMSO-treated controls obtained using the same approach and ideally imaged in parallel with untreated cells. They should also quantify the number of sections and cells imaged for all conditions.

Once we realized that the MTH bundles were visible in projections, we chose to report detections of MTH bundles by projection imaging instead of the costly tilt series. The apparent poor quality and questionable vitrification comes from the fact that the projection images show the cryosection’s crevasses and knife marks, which reside on opposite cryosection surfaces. These sectioning artifacts are computationally excluded from tomographic slices. The following line was added to the figure caption to explain this:

“These image features are not devitrification artifacts; they are absent from the tomographic slices in other figures because they can be computationally excluded.”

The quantification of the MTHs in Lat-A vs control cells are in Table 2. We have now added these numbers to both the text and the figure caption.

The similarities between the MTHs and SCs - that both are present in meiotic nuclei and sensitive to hexanediol - seem unlikely to be functioanlly relevant. Again, I think the presentation suffers from being focused on the SC which was not seen, rather than on the MTHs.

We have toned down the discussion on SCs throughout the manuscript. We have retained the motivation for using 1,6-hexanediol to probe the MTHs physico-chemical properties and the fact the previous work on SCs provided motivation for this perturbation experiment. However, we removed the comparison of their relative sensitivities to 1,6-hexanediol (see reply to point #10 below).

The absence of 100-nm-wide zones containing nucleosomes is again not evidence for lack of SCs. SCs are ribbon-like structures - they are about 100 nm in one dimension but the thickness has not been characterized reliably; even if SCs do exclude nucleosomes (which is not certain) the excluded volume might be much smaller than the authors imagine.

We did not argue for “lack of SCs”; these structures clearly exist in our cells given the fluorescent linear structures seen in Zip1-GFP expressing cells. We only say that the textbook portrayal of SCs needs revision, though we should have restricted our statement to yeast. In the literature and textbooks: whenever the SC’s central element is drawn, it is depicted without internal nucleosomes and being densely packed with SC proteins.

Does Lifeact-mCherry enter the nucleus? This information is important in interpreting the failure to detect MTHs using this probe.

While Lifeact-mCherry is small enough to passively diffuse through the nuclear pore, our data cannot rule out that this molecule is excluded from the nucleus. We added the following sentence as a caveat:

Lines 244-245 “Note that we cannot rule out that Lifeact-mCherry is excluded from the nucleus.”

The sensitivity of the MTH structures to 1,6-hexanediol treatment is potentially interesting, but it does not reveal anything about their structure or function, only that their assembly likely depends in part on hydrophobic interactions. Caution should be used in interpreting these findings.

We have toned down the discussion about the meaning of the MTH bundles’ 1,6-hexanediol sensitivity by removing this line from the original Results:

“MTH bundles are therefore sensitive to a slightly higher concentration of 1,6-hexanediol than SCs are and reform when 1,6-hexanediol is removed.”

We have also added the following line to more clearly say what sensitivity to 1,6-hexanediol means:

Lines 412-414: “Their sensitivity to 1,6-hexanediol suggests that the polymerization both within and between MTHs are based on hydrophobic interactions”

The figure legends and/or Methods sections should clarify what is represented in each figure, and how the data were acquired. In particular, cryotomographic slices of varying dimensions (6nm, 10 nm or 12 nm or 70 nm) are mentioned in the captions of several figures (2-6, and S1, S2, S3). However, is often unclear whether these represent physical or computational sections.

“Tomographic slice” refers to a rendering of a computationally extracted slice from a reconstructed tomogram. To make it clearer, we have added the term “computational” to describe the tomographic slices in each figure caption.

Page 23 has a supplemental figure but no captions. Is this the same as Fig. S8?

Yes, this is a copy of Fig. S8 that appeared due to MS Word’s jumping-figure bugs. We will manually edit the PDF in the future revision.

I do not find the model figure (Figure 7) to be helpful. Additionally, the failure to detect SCs and the presence of MTHs do not warrant a "revised model of the meiotic yeast nucleus."

We now call panel A and B the “Traditional EM” and the “Cryo-EM” models, respectively. The figure therefore reports the large nuclear bodies seen by the two methods and no longer implies correctness.

We also changed the related sentence in the Introduction:

Original: “Our work strongly suggests that current models of pachytene nuclear cytology need revision.”

Revised: “Our analysis shows that MTHs coexist with SCs, which have an unknown cryo-ET structure.”

The absence of MTHs in haploid cells induced to undergo meiosis should perhaps be studied in more detail. Even SCs are present in haploid meiotic cells, so the absence of these structures may be informative as to their function. Haploid cells should also be stained for SCs and imaged by immunofluorescence to verify that they are in meiosis.

The haploid strain that was treated with sporulation media cannot enter meiosis. Haploid cells that are capable of entering meiosis need to be disomic for chromosome III, with each copy having a different mating type at the Mat locus. We believe that the construction and studies of such strains would be more meaningful after we identify the MTH’s subunits and determine its function in diploid cells.

The yeast strain is SK1, not SK-1.

Thank you. This mistake is corrected.

What are "self-pressurized-frozen samples" (p.2)?

Self-pressurized-frozen samples are generated by an alternative to the conventional machine-based high-pressure-freezing method. We have added more details in the new lines 135-141:

“Self-pressurized freezing is a simpler and lower-cost alternative to conventional high-pressure freezing, which requires a dedicated machine that consumes large amounts of cryogen. In the self-pressurized freezing method, the sample is sealed in a metal tube and rapidly cooled in liquid ethane. The material in direct contact with the metal cools first and expands by forming crystalline ice, which exerts pressure on the material in the center of the tube (Leunissen and Yi, 2009; Yakovlev and Downing, 2011; Han et al., 2012).”

Reviewer #2 (Significance (Required)):

The observation of MTHs is novel but (as stated above) of unclear significance, given that their molecular identity and function are unknown.

This work may be of interest to the meiosis field, with the caveats described above that the functional relevance is currently unclear.

This review was co-written by referees with expertise in meiosis, chromosome organization, SC structure and function, and cryoEM.

**Referee Cross-commenting**

The reviews are strikingly concordant so I don't think much needs to be added.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

The authors report the observation of filamentous structures (further termed meiotic triple helices MTH) in the nucleoplasm during meiosis in yeast cells. Those structures are visualized using Cryo-ET. While the authors initially seem to assume an association of those structures with synaptonemal complexes, they discover that those structures rely on filamentous actin and are not affiliated with synaptonemal complexes after all.

While I think that the observation of those MTH by Cryo-ET is interesting, the overall structure of the paper and presentation of the data are not very well done. As the authors find throughout their experiments that the MTHs are not associated with synaptonemal complexes the strong focus in the first figures on synaptonemal complexes as well as the title of the paper are very mis-leading. The authors try to initially make the point that other labs have observed ladder like structures by transmission electron microscopy and want to make the claim that those observed structures might be an artifact of sample preparation, hence the title: Meiotic budding yeast assemble bundled triple helices but not ladders. However, at the end those structures seem unrelated to synaptonemal complexes.

In addition, several labs have reported the presence of nuclear actin in meiosis and mitosis and have even succeeded to show those structures by transmission electron microscopy, questioning the "artifact" argument.

Presumably, the Reviewer means intranuclear actin, as opposed to perinuclear (cytoplasmic) actin. If so, then we have only seen one paper, the one from the Takagi et at 2021 (bioRxiv) that has reported seeing structures associated with nuclear actin. Note however, that the revised Takagi bioRxiv paper is very careful in saying that the filament bundles contain actin, which is not the same as saying that the filaments are polymers of actin.

Our “artifact” argument – now removed – referred to SCs, not to nuclear actin.

In the revised manuscript, we use the term “intranuclear” to make it clear that we are referring to structures inside the nucleus. We also use the term F-actin, where appropriate, to refer to the best-studied actin polymer, which resembles a double helix. Doing so eliminates confusion about other forms of actin: G-actin, which cryo-ET cannot yet directly visualize due to its small size; and non-canonical actin polymers, for which there are no previous experimental X-ray or cryo-EM structures for comparisons.

The story line of the paper is weak and I think the authors would have been better of reporting their cryo-ET structures and making a better link to actin or determining what else they think might be a component of those structures. Immuno-EM (as actually shown in Reference 41) of actin would have been much more convincing. The authors could also use the power of Cryo-ET and the achievable higher resolution to describe those filaments in much more detail. In my opinion this would have been a much better and more exciting paper.

We disagree that “making a better link to actin” is the right approach because doing so presupposes that the structures are composed of actin, for which the present evidence is inadequate. We do agree that determining what the MTHs are (or are not) would be valuable.

FULL: Now that we have better access to a K3-GIF camera and a cryo-FIB-SEM, we will attempt higher-resolution subtomogram averaging analysis. If we are able to achieve subnanometer resolution, we will attempt to narrow down the fold of the MTH subunit. Note that this goal will require that the MTHs are conformationally homogenous and that we can image sufficient copies of the MTHs.

In summary: While I think the Cryo-ET images of those structures could be very exciting the paper unfortunately does not do a very good way in presenting this data and is at times misleading trying to proof or rather dis-proof a connection to synaptonemal complexes. Based on this I think that the paper can not be published in the current form and needs major revisions that would require a significant amount of time.

**Minor comments:**

Figure 1: The choice of timepoints is confusing and makes it hard to compare. While wild type is shown at 0,1,3,4,5,8h, the mutant is shown at 0,2,3,4,5,6,7h. It would be appropriate to select the same timpepoints for both conditions.

FULL: We will recollect fluorescence images of the mutant cells at the same time points as the wild type.

Figure 3 and 4 need a quantification of the number of observed MTHs, in particular as only selected regions of the nuclei are shown.

These images were taken from single cryosections instead of serial cryosections, which would have been too difficult to do for multiple conditions and multiple cells. Therefore, quantification would be obfuscated by the fact each cryosection samples a small fraction of the nuclear volume. We believe that reporting the number of cell cryosections that are MTH-positive (Table 2) is at present the best way to characterize their abundance and ability to polymerize. Once we are able to identify the MTH gene products, we will be able to perform GFP tagging experiments and thereby get a much better estimate of the polymer mass as a function of biochemical perturbations.

Fig 7. The data certainly does not support a "REVISED" model of the yeast nuclear organization.

We have changed the Figure title to “A cryo-ET model of the meiotic yeast nucleus.” We now also refer to panels A and B as “Traditional EM model” and “Cryo-ET model”, respectively.

Reviewer #3 (Significance (Required)):

Several publications have already shown the presence of actin in meiotic and mitotic nuclei and have even succeeded in observing those structures by transmission electron microscopy. Based on this it is not clear why the authors have not tried to put their work in context to all these observations and used their technology to obtain novel information on the structure, which might be helpful to identify which proteins compose the MTHs. Based on how the data is presented I do not think that this paper contributes anything new to the field.

Presumably, the reviewer means that F-actin has been imaged in yeast cells, because for actin to exist in nuclei in mitosis/meiosis, the organism would have to undergo closed mitosis/meiosis. Furthermore, for actin to be observable by transmission electron microscopy, it would have to be in the filamentous (F-actin) form. We could not find any publications that report transmission electron microscopy of F-actin in yeast cells. We therefore cannot relate our results to F-actin in the meiotic nucleus.

My field of expertise is meiosis and mitosis as well as imaging (light and electron microscopy).

**Referee Cross-commenting**

All reviewers seem to agree that the general observation of these structures is interesting but that there is a reduced significance as the function and identity of these structures remains unknown.

4. Description of analyses that authors prefer not to carry out

The main unanswered question is: what is the function of the MTH bundles? To address this question, we would first need to identify the gene products that are needed for MTH assembly. Next, we would need to do genetic perturbation experiments to actually determine the MTHs’ function. These experiments would constitute a complete study, which is better suited for a separate, future manuscript.

Author Response:

Reviewer #1:

Maimon-Mor et al. examined the control of reaching movement of one-handers, who were born with a partial arm, and amputees, who lost their arm in adulthood. The authors hypothesized that since one-handers started using their artificial arm earlier in life then amputees, they are expected to exhibit better motor control, as measured by point-to-point reaching accuracy. Surprisingly, they found the opposite, that the reaching accuracy of one-handers is worse than that of amputees (and control with their non-dominant hand). This deficit in motor control was reflected in an increase in motor noise rather than consistent motor biases.

Strengths:

• I found the paper in general very well and clearly written.
• The authors provide detailed analyses to examine various possible factors underlying deficits in reaching movements in one-handers and amputees, including age at which participants first used an artificial arm, current usage of the arm, performance in hand localization tasks, and statistical methods that control for potential confounding factors.
• The results that one handers, who start using the artificial arm at early age, show worse motor control than amputees, who typically start using the arm during adulthood, are surprising and interesting. Also intriguing are the results that reaching accuracy is negatively correlated with the time of limbless experience in both groups. These results suggest that there is a plasticity window that is not anchored to a certain age, but rather to some interference (perhaps) from the time without the use of artificial arm. In one-handers these two time intervals are confounded by one another, but the amputees allow to separate them. I think that the results have implications for understanding plasticity aspects of acquiring skills for using artificial limbs.

Weaknesses:

• While I found that one of the main conclusion from the paper is that the main factor that is related to increased motor noise is the time spent without the artificial arm, it felt that this was not emphasized as such. These results are not mentioned in the abstract and the correlation for amputees is not shown in a figure.

We thank the reviewer for their comment. While it is true that motor noise correlated with time of limbless experience in both groups, we were hesitant to highlight the results found in amputees, considering the small number of participants, and lack of converging evidence (e.g., contrary to the congenital group, we did not find a strong main effect). For these reasons, we have chosen to include it in the manuscript but not highlight it or base our main conclusions on it. Following the reviewer’s comment, the correlation of the amputees’ data is now visualised in Figure 3. Moreover, while the behavioural correlation might be similar in both groups, from a neural standpoint, the limbless experience of a toddler with a developing brain is qualitatively different to that of an adult, with a fully developed brain, who has lost a limb. As such, we were hesitant to link these two findings into a single framework, however in the revised manuscript we highlight this tentative link.

Discussion (4th paragraph):

“In both the congenital and acquired groups, artificial arm reaching motor noise correlated with the amount of time they spent using only their residual limb. It is therefore tempting to link these two results under a unifying interpretation; however, this requires further research, considering the neural differences between the two groups.”

Figure 3. Years of limbless experience before first artificial arm use in the acquired group. (A) Relationship between years of limbless experience and (A) artificial arm reaching errors or (B) artificial arm motor noise in the acquired group.

• The suggested mechanism of a deficit in visuomotor integration is not clear, and whether the results indeed point to this hypothesis. The results of the reaching task show that the one-handers exhibit higher motor noise and initial error direction than amputees. The results of the 2D localization task (the same as the standard reaching task but without visual feedback) show no difference in errors between the groups. First, it is not clear how the findings of the 2D localization task are in line with the results that one-handers show larger initial directional errors.

We fully take on the reviewer’s comment regarding the vague use of the term visuomotor integration. In the revised manuscript, we have opted instead for a much broader term, suggesting a deficit in visual-based corrective movements, considering we are limited in our ability to infer the specific underlying mechanism from our result. We have also made changes to the abstract based on the reviewer’s comment (see below).

With regards to discussing how the various results fit together, in the revised manuscript, these are now discussed more at length. In short, in the 2D localisation task (reaching without visual feedback), participants were not instructed to perform fast ballistic movements. Instead, participants were instructed that they could perform movements to correct for their initial aiming error (using proprioception). Together with the similar performance observed for the proprioceptive task, this strengthens our suggestion that the deficit in the congenital group is triggered by visual-driven corrections. These various considerations are now detailed as follows:

Abstract:

“Since we found no group differences when reaching without visual feedback, we suggest that the ability to perform efficient visually-based corrective movements, is highly dependent on either biological or artificial arm experience at a very young age.”

Result (section 7, 1st paragraph):

“From these results, we infer that early-life experience relates to a suboptimal ability to reduce the system’s inherent noise, and that this is possibly not related to the noise generated by the execution of the initial motor plan. Early life experience might therefore relate to better use of visual feedback in performing corrective movements. The continuous integration of visual and sensory input is at the heart of visually- driven corrective movements. Therefore, one possibility is that limited early life experience, results in suboptimal integration of information within the sensorimotor system.”

Discussion (2nd paragraph):

“When performing reaching movements without visual feedback (2D localisation task), the congenital group did not differ from the acquired or control group. This begs the question, if the congenital group has a deficit in motor planning why was it not evident in this task as well? In the 2D localisation task, unlike the main task, participants were allowed to make corrective movements. While they did not receive visual feedback, the proprioceptive and somatosensory feedback from the residual limb appears to be enough to allow them to correct for initial reaching errors and perform at the same level as the acquired and control group. Moreover, we did not find strong evidence for an impaired sense of localisation of either the residual or the artificial arm in the congenital group. As such, by elimination, our evidence suggests that the process of using visual information to perform corrective movements isn’t as efficient in the congenital group.”

Discussion (2nd paragraph):

“Lack of concurrent visual and motor experience during development might therefore cause a deficit in the ability to form the computational substrates and thus to efficiently use visual information in performing corrective movements.”

Discussion (last paragraph):

“By the process of elimination, we have nominated suboptimal visual feedback-based corrections to be the most likely cause underlying this motor deficit.”

Second, I think that these results suggest that the deficiency in one-handers is with feedback responses rather than feedforward. This may also be supported by the correlation with age: early age is correlated with less end-point motor noise, rather than initial directional error. Analyses of feedback correction might help shedding more light on the mechanism. The authors mention that the participants were asked to avoid doing corrective movement and imposed a limit of 1 sec per reach to encourage that. But it is not clear whether participants actually followed these instructions. 1 sec could be enough time to allow feedback responses, especially for small amplitude movements (e.g., <10 cm).

Please see below our response to the feedback correction analysis suggestion. Regarding corrective movements, we had the same concern as the reviewer which led us to use hand velocity data to identify first movement termination. We apologise if the experimental design and pre-processing procedures were not clear.

In short, a 1 sec trial duration was imposed on all trials to generate a sense of time- pressure and encourage participants to perform fast ballistic movements. As we were worried that participants might still perform secondary corrective movements within this 1 sec window, for each trial, we used the hand velocity profile to identify the end of the first movement. Below, we have plotted the arm velocity from a single trial to illustrate this procedure. For this trial, the timepoint indicated by the circular marker has been identified as the time of the end of the first movement (See Methods for further information). For each trial, endpoint location was defined as the location of the arm at the movement termination timepoint defined by the kinematic data and not the endpoint at the 1 sec timepoint. It is worth noting that performing the same analysis using the end- points recorded at the 1 sec timepoint did not generate different statistical results.

This has now been further clarified in the text.

Results (section 1, 1st paragraph):

“Reaching performance was evaluated by measuring the mean absolute error participants made across all targets (see Figure 1C). The absolute error refers to the distance from the cursor’s position at the end of the first reach (endpoint) to the centre of the target in each trial. The endpoint of each trial was set as the arm location at the end of the first reaching movement, identified using the trial’s kinematic data (See Methods).”

Methods (section: Data processing and analysis – main task):

“Within the 1 sec movement time constraint, in some trials, participants still performed secondary corrective movements. We therefore used the tangential arm velocities to identify the end of the first reach in each trial (i.e., movement termination).”

Reviewer #2:

This is a broad and ambitious study that is fairly unique in scope - the questions it seek to answer are difficult to answer scientifically, and yet the depth of the questions it seeks to answer and the framework in which it is founded seem out of place in a clinical journal.

And yet, as a scientist and clinician, I found myself objecting to the claims of the authors, only have them to address my objection in the very next section. The results are surprising, but compelling - the authors have done an excellent job of untangling a very complicated question, and they have tested (for our field) a large number of subjects.

The main two results of the paper, from my perspective, are as follows:

1) Persons with an amputation can form better models of new environments, such as manipulandums, than can those with congenital deficiencies. This result is interesting because a) the task did not depend on significant use of the device (they were able to use their intact musculature for the reaching-based task), and b) the results were not influenced by the devices used by the subjects (cosmetic, body-powered, or myoelectric).

2) Persons with congenital deficiency fit earlier in life had less error than those fit later in life.

Taken together, these results suggest that during early childhood the brain is better able to develop the foundation necessary to develop internal models and that if this is deprived early in childhood, it cannot be regained later in life - even if subjects have MORE experience. (E.g., those with congenital deficiencies had more experience using their prosthetic arm than those with amputation, and yet scored worse).

The questions analyzed by the researchers are excellent and the statistical methods are generally appropriate. My only minor concern is that the authors occasionally infer that two groups are the same when a large p-value is reported, whereas large p-values do not convey that the groups are the same; only that they cannot be proven to be different. The authors would need to use a technique such as ICC or analysis of similarities to prove the groups are the same.

We appreciate the reviewer’s concern about inferring the null from classical frequentist statistics. In this manuscript, we have opted to using Bayesian statistics as a measure of testing the significance of similarity across groups (See Methods: Statistical analysis) as opposed to the frequentist methods suggested by the reviewer. This approach is equivalent to the ones proposed by the reviewer and are widely used in our field. A Bayesian Factor (BF) smaller than 0.33 is regarded as sufficient evidence for supporting the null hypothesis that is, that there are no differences between the groups.

This approach is described in detail in the methods and is introduced in the first section of the results as well.

Results (1st section 2nd paragraph):

“To further explore the non-significant performance difference between amputees and controls, we used a Bayesian approach (Rouder et al., 2009), that allows for testing of similarities between groups (the null hypothesis). In this analysis, the smaller effect size of the two reported here (1.39) was inputted as the Cauchy prior width. The resulting Bayesian Factor (BF10=0.28) provided moderate support to the null hypothesis (i.e., smaller than 0.33).”

Methods (Statistical analysis section):

“In parametric analyses (ANCOVA, ANOVA, Pearson correlations), where the frequentist approach yielded a non-significant p-value, a parallel Bayesian approach was used and Bayes Factors (BF) were reported (Morey & Rouder, 2015; Rouder et al., 2009, 2012, 2016). A BF<0.33 is interpreted as support for the null-hypothesis, BF > 3 is interpreted as support for the alternative hypothesis (Dienes, 2014). In

Bayesian ANOVAs and ANCOVA’s, the inclusion Bayes Factor of an effect (BFIncl) is reported, reflecting that the data is X (BF) times more likely under the models that include the effect than under the models without this predictor. When using a Bayesian t-test, a Cauchy prior width of 1.39 was used, this was based on the effect size of the main task, when comparing artificial arm reaches of amputees and one- handers. Therefore, the null hypothesis in these cases would be there is no effect as large as the effect observed in the main task.”

Following the reviewer’s comment, we have carefully scanned through the manuscript to make sure no equivalence claims are made without the support of a significant BF. In one instance that has been the case and has been rectified.

Results (3rd section, 2nd paragraph):

“We compared artificial arm and nondominant arm biases (distance from the centre of the endpoint to the target) across groups, using intact arm biases as a covariate. The ANCOVA resulted in no significant (inconclusive) group differences (F(2,47)=2.40, p=0.1, BFIncl=0.72; see Figure 2A).”

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

Saha and colleagues investigated the functions of the long non-coding RNA (lncRNA) DRAIC in malignant glioma. They find that DRAIC expression decreases cell migration/invasion and tumorsphere/colony formation in vitro, and tumor growth in vivo using established cell lines. Mechanistically, DRAIC is known to inhibit NF-kB signaling and the authors demonstrate that DRAIC activates AMPK leading to repression of mTOR, which decreases protein synthesis and increases autophagy. This is a solid study highlighting a potentially interesting pathway of tumor growth and invasion in brain tumors.

Answer: We appreciate Reviewer 1 for the positive feedback of our study

Major Comments:

1) It is unclear whether the presented values (mean +/- SD) in the histograms refer to repeat measurements (in which case n = 1) or independent experiments (n>1). The number of replicate experiments is not stated in the methods or figure legends. This must be included.

Answer: We want to thank Reviewer 1 for pointing out this omission. We have now included this information in the Materials and methods and in figure legends section.

2) I don't think the immunoblot for p62 in Fig. 5C shows a convincing increase following DRAIC knockout, so the statement on p.8 should be revised.

Answer: We have revised the statement to say: Consistent with DRAIC decrease being associated with a decrease in autophagic flux, and despite a decrease in p62 mRNA, the level of P62 protein is increased in three of the DRAIC KO prostate cancer cells (Fig. 5C, KO1, KO2, KO4 compared to WT) and unchanged in the other two.

3) On p.8/Fig.5 the authors make a case that increased DRAIC levels increase lysosomal degradation of autophagosome core proteins LC3 II / p62 (resulting in decreased protein levels of both), while simultaneously increasing gene expression of LC3B and p62 (causing increased mRNA levels). The data for DRAIC overexpression fit this logic fairly well (even though I think more work is needed to fully support this claim), but I am finding it difficult to reconcile the DRAIC knockout data with this scenario - here, loss of DRAIC results in increased protein levels to decreased autophagy, but also decreased gene expression. To fully support this argument, rescue experiments would be needed using FoxO3a knockout/overexpression.

Answer: Note that the mRNA level is not always correlated with protein expression. This is particularly true for LC3 and p62, whose protein levels are significantly affected by the extent of fusion of autophagosomes and lysosomes and subsequent degradation in autophagolysosomes. Thus, although the mRNA of these genes is decreased in DRAIC KO cells (Fig. 5E), the proteins are increased (Fig. 5C) because of decrease of autophagic flux (and decrease of degradation in the autophagolysosomes).

The overexpression of FoxO3a in the DRAIC KO cells will not restore mRNA levels of LC3 or p62, because we show in Fig. 4H that FoxO3 phosphorylation by AMPK is suppressed by DRAIC KO. This phosphorylation is important for the induction of LC3 or p62 mRNA by FoxO3.

FoxO3 knockout or knockdown in DRAIC OE cells should decrease LC3B or p62 mRNA in Fig. 5D, but it is already known from the Literature that FoxO3a is necessary for inducing LC3B or p62 mRNA. Cell Metab. 2007 Dec;6(6):458-71. doi: 10.1016/j.cmet.2007.11.001.PMID: 18054315.

4) Similarly, the data supporting increased autophagy following DRAIC overexpression (Fig. 5F/G) are a bit weak and lack controls (is the LC3B-GFP overlapping with endogenous LC3B and autophagosomes? Was the transfection efficiency comparable? Is there fusion with lysosomes?). In the absence of stronger data, the authors should temper their claims that DRAIC increases autophagy.

Answer: LC3B of fusion protein LC3B-GFP is known to overlap with the p62 puncta (similar to endogenous LC3B). This result is in Fig. 4A of the citation that we have now added (Proc Natl Acad Sci U S A. 2016 Nov 22;113(47): E7490-E7499. doi: 10.1073/pnas.1615455113. Epub 2016 Oct 17)

To support our hypothesis that DRAIC OE induces more autophagy compared to empty vector, we used Bafilomycin A1 in Figure 5B to inhibit the autophagosome and lysosome fusion. We see the accumulation of more LC3B upon treatment with Bafilomycin A1 in the DRAIC OE cells (compared to EV containing U251 cells), consistent with the idea that autophagosome-lysosome fusion is increased by DRAIC OE.

5) No information is provided on animal numbers used in this study. How many mice were used per cohort? Were male and female mice used? Authors should follow ARRIVE guidelines in reporting animal experiments. The method for calculating tumor volume needs to be specified.

Answer: We have included the details about the animal study in methods section of our modified manuscript .

6) Student's T-test is inappropriate for comparisons of more than two groups (i.e. all experiments using DRAIC knockout cells) - for these experiments a Kruskal Wallis test or ANOVA should be used. Did the authors test for normal distribution of their data? This may affect statistical testing and should be taken into consideration.

Answer: We have now modified our statistical calculation and included in the statistical analysis section in our modified manuscript.

Minor Comments:

7) Authors mention that DRAIC expression is undetectable in immortalized astrocytes and GBM cancer stem cells (Fig. S1). What is the source of these cells and how were they cultured?

Answer: The immortalized astrocytes and GBM stem cells and their culture conditions is now described.

8) The immunoblot in Fig. 3D could be replaced with a slightly lower exposure to make the difference between WT and DRAIC KO more obvious.

Answer: We have now replaced the immunoblot with lower exposure.

9) Some immunoblots in Fig. 3 (panel E, p-S6K and S6K; panel H, actin) are not of the best quality and an effort should be made to replace them.

Answer: We have now replaced the immunoblot p-S6K as reviewer mentioned.

10) Why are different loading controls used in Fig. 3 (a-Tubulin v actin)?

Answer: We use multiple loading control to make sure that we are not underestimating or overestimating changes in the experimental protein because of unexpected changes in the loading controls.

11) Compared to other blot images in the same figure (e.g. Fig. 3E), the bands for p-mTOR and mTOR in Fig. 3F look compressed and should be shown appropriately sized.

Answer: We have modified the Figure as reviewer suggested.

12) The layout of Fig. 4 is somewhat confusing. I would suggest organizing this according to DRAIC overexpression in A172 and U373 cells versus DRAIC knockout in LNCaP cells. Each immunoblot should be clearly labelled with the corresponding cell line, and it should be clearly explained why p-FoxO3a was tested in U251 cells, rather than A172/U373 as in the rest of the figure.

Answer: We thank the reviewer for the constructive criticism. We have labeled all the cell lines in the Figure as reviewer suggested. We have now systematically alternated the prostate cancer cells (for KO) and the GBM cells (for OE), as we looked at each relevant marker. We have now included the western blot for p-FoxO3a from another glioblastoma cell line U373. Please find the modified Figure 4K for p-FoxO3a.

13) Labelling of immunoblot in Fig. 5B is confusing and should be improved.

Answer: We have modified the Fig. 5B to make the label clearer.

14) Changes in GLUT1 expression (Fig. 7A) should be validated on the protein level.

Answer: We have included the immunoblot for GLUT1 from DRAIC KO cells in Figure 7B. GLUT1 protein is increased upon DRAIC KO.

Reviewer #1 (Significance (Required)):

The authors describe a novel link between the lncRNA DRAIC and AMPK activation through inhibition of NF-kB-mediated regulation of GLUT1. This study extends their previous work on DRAIC inhibition of NF-kB in prostate cancer (Saha et al. Cancer Res 2020). There is one study describing DRAIC effects on growth and invasion in glioma cell lines (Li et al. Eur Rev Med Pharmacol Sci 2020), but the work presented by Saha and colleagues contains stronger experimental data and a more detailed and previously undescribed mechanism.

The current study presents a mechanistic advance that increases the understanding of tumor growth and protein synthesis in cancer cells. The data presented in the study are not supported by in vivo experiments (other than suppression of tumor growth by DRAIC overexpression), validation in human tissue and/or primary patient-derived human glioblastoma cells, or even substantial rescue experiments. This limits the influence of the work on the field. I'm also not sure how transferable findings from DRAIC knockout in prostate cancer cell lines are to glioma, although the results are mostly complementary to the data from glioma cell lines. This is particularly relevant to the proposed mechanism of GLUT1 regulation by NF-kB, as the bulk of experimental data in Figures 6 and 7 was generated in prostate cancer cell lines and is only poorly validated in glioma cells. The study results will be most relevant for researchers investigating cell signaling pathways and autophagy in cancer.

Answer: We like to thank reviewer for the positive comments on our study. The DRAIC KO experiments of Fig. 6 and 7 cannot be done in glioma cells, because as we show if Supp. Fig. S1, there are no glioma cells or GSC that express DRAIC to levels comparable to LnCaP. We have shown that GLUT1 mRNA decreases in the glioma cells when DRAIC is overexpressed (Supp. Fig. S4. We also show in Fig. 7G that AMP levels increase when DRAIC is overexpressed in glioma cells.__

__

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

In this manuscript, the authors describe DRAIC as a lncRNA downregulated in prostate cancer. They postulate that DRAIC expression surpasses invasion, migration and growth. Mechanistically, the authors show that DRAIC activates AMPK by suppressing NFkB target gene TOR and indirectly impacting translation and autophagy. Collectively the observation is interesting and robust. However, I have several technical requests, particularly regarding the mechanistic part of the paper.

Answer: We appreciate the positive feedback. We have addressed the reviewer’s concerns in our modified manuscript.

Major Comments:

1) The authors should rescue Ko phenotypes by over expressing DRAIC to consider potential off target effects.

Answer: DRAIC OE alone is sufficient to have exactly the opposite effect as DRAIC KO in protein translation (Fig. 3C-F), so DRAIC OE will rescue the effect of DRAIC KO. We make a similar argument for all the phenotypes, including mTOR, S6K and ULK1(S757) phosphorylation (Fig. 3G-J), AMPK and FoxO3a phosphorylation (Fig. 4B-C; J-L), autophagic flux (Fig. 5B, C) and effects on LC3B and p62 mRNAs (Fig. 5D, E). The same is true for our published phenotypes of DRAIC KO on invasion, migration and NF-kB activity (Saha, Cancer Research, 2020)

2) The blots showing TOR and ULK1 phosphorylation need to be repeated. This is an important part of the paper and I feel that these blots are hard to interpret. p-S6K typically run a bit higher in gels. there may be a technical problem.

Answer: We are not sure which specific blots the reviewer is referring to, and it is possible that the blots the other reviewers pointed to are the ones under question. We have changed those blots so that the results are clear.

3) GLUT1-related results are interesting, but the authors should provide genetic evidence that the effects are mediated by GLUT1. How do we know that glucose uptake is indeed upregulated upon knockout?

Answer: In Fig. 7 C-F we show that the effects of DRAIC KO on invasion, protein translation, AMP levels and AMPK activity are reversed by the GLUT1 inhibitor Bay-876. This is a cleaner result than using siRNA to knockdown GLUT1. siRNAs can have off-target activity and sometimes cannot decrease a protein sufficiently below the threshold necessary to see reversal of action.

Minor Comments:

4) The figures need to be updated. FOnts are all different, lots of unaligned graphs, quality of the blots are poor.

Answer: We have updated the Figures and changed fonts as reviewer mentioned.

Reviewer #2 (Significance (Required)):

The observation is interesting, but the mechanism is incompletely understood. This is a nice addition to the literature, even without the mechanism.

Answer: We want to thank the reviewer for the constructive criticism.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

**Summary:**

Shaha and colleagues present a study demonstrating the tumor suppressive role of DRAIC, a long non-coding RNA transcript, through transmission of the signal from IKK/NF-kB to the AMPK/mTOR pathway via regulation of GLUT1 expression. The inhibition of mTOR by this pathway results in the reduction of protein translation, cellular invasion and activation of autophagy. Several diseases and models as well as multiple genetic and pharmacological manipulations were used to investigate the mechanisms at play. The manuscript is well written and the experiments are well designed. The conclusions are supported by the results. The following major and minor comments should be addressed:

Answer: We appreciate the reviewer for the positive comments on our study.

Major Comments:

1) In addition to reporting the effect of DRAIC overexpression on tumor volume, the authors should present survival studies with one or more models.__

Answer: __We thought of doing the survival study in our glioblastoma model but unfortunately, the tumor growth is very rapid (exceeding the size permitted by our IACUC in 2-3 weeks). The animal ethics welfare committee did not allow us to keep the mice for a longer time to perform the survival study.

2) Since the authors study metabolic energy sensor pathways, related to glycolysis, it would be important to perform some of the key experiments in physiological level of glucose: e.g., pmTOR, pAMPK, LC3-II expression level in DRAIC overexpressing and deficient cells.

Answer: The concentration of glucose in plasma is 1G/L, while that of the RPMI medium is 2G/L. We do not think we are too far from the physiological levels of glucose.

3) In addition to RT-PCR data, GLUT1 protein levels should be investigated in the different DRAIC expressing cells.

Answer: We have incorporated the GLUT1 protein expression data from DRAIC KO cells in Figure 7B and DRAIC overexpressing cells in supplementary Figure 4G-H. The blots from the same gels were split into different panels, the loading control GAPDH remain same in Figure 4K and Supplementary Figure S4H.

4) The effect of DRAIC on GLUT1 expression is also measured in condition of glucose saturation, which does not reflect disease state. The decrease of GLUT1 in response to DRAIC overexpression and the increased GLUT1 level in DRAIC deficient cells should be investigated in physiological levels of glucose.

Answer Same as above. We are near physiological levels of glucose.

__Minor Comments:

__

5) All the data are generated with established cell lines (e.g., U87) but more clinically relevant models, such as patient-derived primary cells like the ones used in Fig. S1, could be used to replicate some of the key findings.

Answer: As we showed in Fig. S1 that DRAIC is not expressed in glioma stem cells, and so knockout experiments are not possible. We believe that the knockout experiments are the most relevant to this paper because they do not run the risk of artefacts from overexpression of an RNA far beyond physiological levels.

6) Also please provide further details about the patient-derived cells from Fig. S1.

Answer: We have mentioned the details of the cell lines in our modified manuscript.

7) The statistical analysis section states that the number of measurements is indicated however I don't see the sample size of the experiments.

Answer: We have now incorporated the number the experiments in our modified text.

Reviewer #3 (Significance (Required)): The study reports a new model of regulation of tumor via long non-coding RNA. This article adds to the growing literature The topic and content of the article is relevant and significant to the field of tumor research but the significance and impact could be enhanced with the use of more physiologically relevant models and conditions as pointed in the major comments.

Answer: We want to thank the reviewer for the positive feedback on our study.

This is something that I've been thinking about a lot lately, particularly as a white academic. Students have things to say, but if we require them to conform to "Standard American English" (I'm in the US, so I'm using SAE as an example, but this would presumably apply elsewhere) and scholarly/disciplinary academic writing expectations/form, then a teacher's fixation on "proper grammar" or "academic voice" or "scholarly form" likely DO stifle the ideas and knowledges that students may have to share ("How many points off for not using an Oxford comma?" "What if I use 'informal language' or 'slang'?") As one student explained just before they were about to graduate, they had spent most of their academic career "whitewashing" their ideas and language, often writing multiple drafts, first in their own voice, then progressively more "academic," so they were never actually able to express what they really had to say...until a two faculty said "just write it the way you want. Here are some models to consider, but you don't have to conform to these either."

Author Response:

Reviewer #1 (Public Review):

[...] The authors do a great job of listing and evaluating possible explanations, one of which is simply that the strains carry multiple mutations of small effect. All but one of the successfully mapped variants consists of missense and nonsense mutations. I think it's important to note that this represents a particular range of the effect-size distribution of mutations affecting the YFP phenotype. We know from the authors' earlier work that there are lots of mutations that can affect gene expression in cis, and so the absence of trans-acting cis-regulatory variants here is parsimoniously interpreted as due to their small effects. In general, work in other systems (particularly human genetics) has shown that even molecular traits are often hugely polygenic, affected by thousands of variants of tiny but non-zero magnitude. With a forward screen of the sort performed here, it's difficult to know how much of the phenotypic variance is due to unmapped small-effect variants, but two lines of evidence suggest it may be a lot: first, the absence of mappable causal mutations in 36/82 mutants, and second, the differences between EMS mutant strains and their matched single-site mutants. The authors commendably report and discuss these issues but to my mind they neglect them in drawing inferences and generalizations from their findings.

We thank the reviewer for these encouraging comments and also appreciate the reviewer pointing out these concerns.

With respect to the overlap of the trans-regulatory mutations we mapped and previously identified eQTL, we agree the possibility of similar mapping biases in the two BSA-seq studies contributing to the overlap of trans-regulatory mutations and eQTL warrants further exploration. We interpret the reviewer’s comment to suggest that if some regions of the genome systematically showed lower sequencing read coverage (because of poor read mappability, PCR biases or any other reason), the power to detect trans-regulatory mutations and eQTL in these regions would be decreased compared to regions of the genome with higher coverage because the G-tests used to identify significant associations with expression in both studies are based on read counts. Consequently, variation in sequencing read coverage across the genome shared in this study and the prior study identifying eQTL, both of which used BSA-seq, could lead to the enrichment of transregulatory mutations in eQTL regions. Indeed, consistent patterns of read coverage across the S. cerevisiae genome have been observed in prior work.

To determine whether trans-regulatory mutations were enriched in regions of the genome with higher sequence read coverage, we compared read coverage between regions of the genome identified as having trans-regulatory mutation or non-regulatory mutations. The identification of variants classified as non-regulatory is expected to be less dependent on the depth of sequencing read coverage because this designation does not require a statistically significant G-test. We found that the mutations identified as trans-regulatory showed 120x coverage whereas mutations identified as non-regulatory showed only 100x coverage, consistent with greater power to detect associations with expression in regions of the genome with higher sequencing read coverage. However, eliminating this difference in read coverage by excluding non-regulatory mutations with lower sequence read coverage did not eliminate the observed enrichment of trans-regulatory mutations in regions previously shown to contain eQTL. Non-regulatory mutations with higher and lower sequencing read coverage were also equally likely to be found within eQTL regions, suggesting that similar variation in sequence read coverage across the genome between the two studies is unlikely to explain the observed overlap of trans-regulatory mutations and eQTL. These analyses are now included in a new Figure 7-figure supplement 1.

With respect to better incorporating biases in what we were able to map and considerations for extending findings from this work to other systems, we have tried to better address these issues in the revised discussion.

Reviewer #2 (Public Review):

Fabien Duveau et al. tried to characterize mutations in trans-regulation effects on expression of the TDH3 by using EMS mutants with TDH3 reporter in Saccharomyces cerevisiae. This work is an extension of works of Gruber et al. (2012) and Metzger et al. (2016) with specific mutation effect on TDH3 expression. They found that these trans-regulatory mutations that have effects on expression of TDH3 reporter were enriched in coding sequences of transcription factors. They found that the trans regulatory mutations with effect are associated with natural variants of trans within S. cerevisiae. In summary, the data is well described and supports their claims. The method of study could be used for study the mechanism how regulatory network works.

[...] Although the paper does have strengths in principle, some weaknesses of the paper would cause the quality of data presented. [...]

We thank the reviewer for taking the time to evaluate this work and have the following responses to the weaknesses noted:

1) The reviewer is correct that we focused this paper on trans-regulatory mutations because cis-regulatory mutations affecting TDH3 expression were previously characterized. Furthermore, long distance enhancers with cis-acting effects on expression have not been described in S. cerevisiae and the term promoter is commonly used to encompass both the basal (core) promoter (including a TATA box for some genes) as well as other upstream activating sequences (UAS) and upstream repressing sequences (URS). In other words, the cis-acting sequences for S. cerevisiae genes are confined to a particular region much more than in multicellular eukarlyotes. In fact, our prior work with TDH3 (Metzger et al. 2015) showed that 97% of cis-acting variation affecting TDH3 expression could be explained by sequence variation in the 678 bp region we define as its promoter. Consequently, all mutations outside of this region were considered to have transregulatory effects on TDH3 expression. In the revised version, we extended the discussion to specifically compare the structure of regulatory sequences in S. cerevisiae to other eukaryotic model systems.

2) In this study, a mutation is defined as trans-regulatory if it affects TDH3 expression and is not located in the TDH3 promoter, regardless of whether or not it also affects growth rate. In fact, mutations in RAP1 and GCR1 affect growth rate (Figure 5), but are clearly trans-regulators of TDH3 with well-established binding sites in the TDH3 promoter. In other words, we do not think that mutations should be discounted as having trans-regulatory effects because they also impact growth rate.

3) (A) Prior work examining the statistics of BSA-seq has shown that G-tests are most appropriate because they take into account the independent sampling from two bulk populations inherent to bulk-segregant analysis (Magwene et al. 2011 PLOS Computational Biology). (B) We are guessing that the reviewer is asking about multiple testing corrections rather than post-hoc tests, as we used a false discovery rate correction for multiple tests in Figure 2-supplement 5A. Although we did not use a multiple test correction for the BSA-seq data, we used a conservative significance threshold of 0.001 that was expected to result in a 3.5% false positive rate. Perhaps more importantly, we functionally validated the effects of 40 of the 41 associated mutations tested. (C) We may indeed have been overly optimistic about mapping power when choosing mutants to analyze with BSA-seq given that the 36 EMS mutants for which we failed to find a significant association between a mutation and fluorescence tended to have smaller effects on PTDH3-YFP expression than the EMS mutants for which we observed one or more associated sites (Figure 3-figure supplement 3). The reviewer’s comment also made us realize that our original sentence referring to mapping power had reported the effect size for estimated RNA levels rather than fluorescence. To avoid confusion, and because our anticipated mapping power does not affect the results of the study, we deleted this statement from the revised manuscript. Regardless of our anticipated mapping power, we were ultimately able to map mutations that affected fluorescence by as little as 1.6% relative to the wildtype strain.

4) The GO enrichment analysis was performed with widely used tools on www.pantherdb.org. The statistical significance of enrichment for each GO term was computed using Fisher’s exact tests that compared 1) the proportion of genes with non-regulatory mutations and 2) the proportion of genes with trans-regulatory mutations that corresponded to the tested GO term. Because the total number of genes identified in our study with trans-regulatory mutations (42 genes) was much lower than the total number of genes with non-regulatory mutations (1043 genes), it was possible to obtain strong and statistically significant enrichment (P < 0.05 in Fisher’s exact test) even if only a small number of genes corresponded to the GO term in both categories. Although we found a large number of enriched GO terms, these GO terms were not always independent from each other. For instance, GO:0009168 (purine ribonucleoside monophosphate biosynthetic process) and GO:0009167 (purine ribonucleoside monophosphate metabolic process) refers to the same biological process and contains the same genes. For this reason, even though we reported all enriched GO terms in Supplementary File 8, we only showed GO terms that were at the tips of different branches in the GO hierarchy on Figure 6 and we grouped GO terms in four main categories that together encompassed most genes with trans-regulatory mutations.

5) We agree with the reviewer that trans-regulatory mutations can affect either the function of a gene product (including the ability of a transcription factor to bind to DNA) as well as the abundance of that gene product, but we do not think this is a weakness of the study. In fact, we think one of the strengths of the study is that we have empirical data testing the relative frequency of these two types of possible changes, finding that mutations in coding regions (presumably more likely to affect the function of the gene product than its expression) are the primary source of changes in TDH3 expression greater than 1%.

6) The goal of the study was to characterize the effects of individual trans-regulatory mutations, thus we did not look at the combined effects of mutations in proteins that might work in a complex. We do, however, mention transcription factors working in a complex: "Transcription factors encoded by the TYE7 and GCR2 genes found to harbor trans-regulatory mutations affecting expression of PTDH3-YFP are known to regulate the expression of glycolytic genes (including TDH3) by forming a complex with transcription factors encoded by the RAP1 and GCR1 genes” (line 461). We think that looking at the combined effects of mutations that all impact the same complex of regulatory proteins is an interesting direction for future work.

Finally, we’d like to point out that the reviewer’s statement in their opening summary about mutations being enriched in the coding sequence of transcription factors is not quite correct: the mutants we mapped were enriched in coding sequences, and we found more mutations in transcription factors previously shown to regulate (directly or indirectly) expression of TDH3 than expected by chance, but trans-regulatory mutations were not significantly enriched in genes encoding transcription factors relative to non-regulatory mutations (as described in the manuscript).

Reviewer #3 (Public Review):

[...] The mutagenesis approach in yeast the authors used is very powerful, but it naturally has drawbacks. The regulatory landscape in yeast is arguably simpler compared to e.g. metazoa or plants, in that the cis-regulatory regions are predominantly closely linked to target genes, the genes in majority do not have introns and post-transcriptional regulation of mRNA through e.g. splicing is rare. These features distinguish the systems, as in animals and plants introns are a very prominent source of regulatory elements (close to half of all enhancers are intronic in many animals), and alternative splicing of e.g. transcription factors are known to play major roles in transcriptional regulation. Further, chromatin is a very important layer in metazoan and plant gene regulation. To benefit the general readership, it would be informative to further elaborate on the significance of the findings for researchers studying other organisms. In addition, it would help to clarify what aspects of the differences in the regulatory landscape the authors think are important to distinguish.

We thank the reviewer for their kind words and recognition of the novelty of this work. We have modified the introduction to try to clarify the relationship of this work to eQTL studies, which we hope addresses the reviewer’s first concern. Specifically, we’ve tried to clarify that the complex, polygenic nature of trans-regulatory variation segregating within a species is well established by prior eQTL studies. We also sought to clarify that our work (which maps single mutations from mutagenized strains rather than natural variation) provides complementary insight into the distribution of regulatory mutations within the genome and within a gene’s regulatory network. Revisions have also been made to try to clarify that the single mutations we mapped were from EMS-induced mutants containing only ~24 mutations per genome, which is more than 1000-fold less than the number of single nucleotide polymorphisms between two strains of S. cerevisiae. That is, this study was designed to identify single trans-regulatory mutations rather than to characterize the genetic architecture of naturally occurring trans-regulatory variation. Although we intentionally focused on characterizing properties of single mutations here, we agree with the reviewer that testing for epistatic interactions among trans-regulatory mutations will be an interesting avenue for future work, and have added this point to the revised discussion. We have also added text to the discussion describing some similarities and differences in gene regulation as among eukaryotes that should be considered when trying to generalize from this work.

Reviewer #1 (Public Review):

This paper aims to characterize mutations that act in trans to affect a single gene's expression in yeast. Trans-acting mutations potentially play an important role in variation and disease within species and in phenotypic evolution. The authors have previously described the mutational architecture and natural variation in the gene's cis-regulatory activity, creating a powerful experimental model for the causes of phenotypic variation. Trans-acting variation is much more challenging, because the mutational target space is the whole genome. The authors use a forward-genetic screen and bulked segregant analysis to identify 52 point mutations that affect their focal transgene's activity, and they identify an additional 17 by directly searching within a handful of candidate genes. The paper includes elegant validation using genome engineering to confirm the mapped variants are causal.

With this collection of trans-acting mutations in hand, the authors can compare their characteristics to a set of mutations that do not have detectable effects on the transgene. Overall, they conclude that trans-acting mutations are enriched for genes that are known to sit upstream of the focal gene in transcriptional cascades, but the majority of mutations are in other kinds of genes, outside the network of transcription factors.

This work is a valuable contribution to the authors' important experimental assault on the genetics of regulatory variation, a useful complement to their previous work on cis-regulatory mutations and polymorphisms. They provide evidence that experimentally defined regulatory networks have predictive value for the location of trans-acting mutations, and they reinforce the result (well established and widely accepted, but important to show in this kind of rigorous way) that trans-acting variation is distributed across a wide range of cellular and molecular functions. There are also some useful fine-grained results, such as the absence of mutations in a known regulator, RAP1, probably due to pleiotropic constraints, and an excess of mutations in iron homeostasis.

Because the dataset of trans-acting mutations is relatively modest in size (necessarily- it's a heroic effort to identify this many), many of the enrichments are also modest. In particular, the finding that mutations are enriched in eQTL regions holds for only two of three previous eQTL studies, and involves a slight elevation over the baseline that 66% of the genome is in eQTL regions. Because both the eQTL and the mutations were discovered by bulked-segregant analysis, biases in mappability will affect both similarly, and so I do not find the enrichment for overlapping hits to be completely persuasive.

This work is important in substantial measure because of its contribution to the larger yeast TDH3 model trait project, which is a landmark research program for understanding phenotypic variation and evolution. On its own, the results in this manuscript would be difficult to generalize to regulatory variation more broadly. There are narrow reasons for this (yeast has a distinctive compact CDS-dense genome; the focal transcript is YFP and so has no endogenous post-transcriptional regulation; only one class of mutations assayed), but the bigger reason is that the researchers are only able to discover mutations with effects above a particular size. Even among the 82 mutant strains they start with, some 36 strains have altered YFP levels but no successfully mapped causal variants. The authors do a great job of listing and evaluating possible explanations, one of which is simply that the strains carry multiple mutations of small effect. All but one of the successfully mapped variants consists of missense and nonsense mutations. I think it's important to note that this represents a particular range of the effect-size distribution of mutations affecting the YFP phenotype. We know from the authors' earlier work that there are lots of mutations that can affect gene expression in cis, and so the absence of trans-acting cis-regulatory variants here is parsimoniously interpreted as due to their small effects. In general, work in other systems (particularly human genetics) has shown that even molecular traits are often hugely polygenic, affected by thousands of variants of tiny but non-zero magnitude. With a forward screen of the sort performed here, it's difficult to know how much of the phenotypic variance is due to unmapped small-effect variants, but two lines of evidence suggest it may be a lot: first, the absence of mappable causal mutations in 36/82 mutants, and second, the differences between EMS mutant strains and their matched single-site mutants. The authors commendably report and discuss these issues but to my mind they neglect them in drawing inferences and generalizations from their findings.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

Summary

Most laboratory research using T. b. brucei has made use of two strains, the monomorphic Lister 427 strain and the pleomorphic EATRO1125 strain. These strains have been passaged in vitro for decades in separate laboratories, thus selecting for populations that differ from both the original isolate and between laboratories.

In this study, Mulindwa et al. describe the isolation of two new T. b. brucei strains from cattle in Uganda, MAK65 and MAK98, which show differences in virulence and propensity for stumpy form differentiation in rodents. To assess the effect of culture adaptation on trypanosomes, the researchers compared the gene copy number of the MAK65 and MAK98 isolates before and after culture adaptation. The study found that isolates cultured for as little as a week already demonstrated a broader gene copy number distribution than those that were not culture adapted. Broader gene copy number distributions, compared to the non-culture-adapted isolates, were also observed for a number of routinely-passaged Lister 427 and EATRO1125 cultures. The researchers observed reproducible increases in copy number for certain genes, such as those encoding histones, HSP70 and PFR proteins. The study postulated that changes in gene copy number observed across the genome increased bias towards rapid proliferation and stress tolerance upon culture adaptation.

Major comments

I have some concerns about the method that was used for the copy-number calculations. Firstly, I can imagine that a non-uniform distribution of the reads across the genome for any of the datasets could influence the results. Was this checked? Secondly, in the methods section lines 394/395 it is said that the modal RPKM for each dataset was 'adjusted slightly' to get a symmetrical distribution. Upon checking the modal RPKMs and the adjusted values used for the calculations, the adjusted values appear to have been adjusted to different extents between the different datasets to fit the authors assumptions. Do the authors believe that this could perhaps account for some of the subtle gene copy number changes observed (as is discussed in lines 287-290)? Next, why were the reads aligned 20 times? I think in general the method needs to be explained far more clearly so that the audience can understand what you did.

Minor comments

Additional experiments:

Would it be possible to generate a phylogenetic tree comparing these new isolates with the Lister 427, TREU927 and EATRO1125 isolates in circulation to get an idea of how these strains may have evolved? I this could add to the strength of the manuscript.

Title: Since no other unicellular eukaryotes are mentioned throughout the text, I think a more appropriate title for this study might be 'Adaptation of Trypanosoma brucei brucei bloodstream forms to in vitro culture results in gene copy-number changes.'

Line 59: There have been more recent studies than the referenced paper (Cross et al., 2014) that quote far higher numbers of alternative VSG genes or pseudogenes in the Lister 427 strain. In Müller et al (2018, doi: 10.1038/s41586-018-0619-8) over 2500 VSGs are quoted and in Cosentino et al (2021, doi: 10.1101/2021.04.13.439624) 2872 VSGs are identified.

Line 109/110: The dates of isolation written here, Feb 1st and July 30th 2016, are different to what is written under the Date of Isolation column in Supplementary text 1, table 1- 25/5/2016 and 30/7/2017. Do these two dates represent different things?

Line 233: Should Figure 5 A, B, C and E (rather than A, B, D and E) be referenced here to show all the initial, not cultured results?

Line 270/271: PIP39 does not promote differentiation to stumpy forms, but instead contributes towards the efficient differentiation of stumpies to procyclic forms (Szoor et al, 2010, doi: 10.1101/gad.570310).

Discussion:

It should be discussed in the text that changes in gene copy number have also been observed upon Leishmania culture adaptation (see, for e.g., Gerald Späth's work). This will strengthen the authors' conclusions. Furthermore, similar observations of aneuploidy and triploidy in some T. brucei Lister 427 strains have recently been reported in Cosentino et al (2021, doi: 10.1101/2021.04.13.439624). This could also be referenced somewhere in the text.

Line 364: I think the Nijuru et al (2005, doi: 10.1007/s00436-004-1267-5) paper would be the correct reference here since it describes the use of the ITS-1 PCR. Reference 13 is actually referring to another paper, I cannot find Nijuru et al in the references list.

Line 379: PAD staining should be referenced to be more informative and allow reproducibility.

Figure 1, Line 580: How many cells were counted for determining the % PAD positivity?

Figure 1, Line 581: Scale bar should be included in D.

Supplement Text 1:

I found Supplement Text 1 a little confusing for two reasons. Firstly, I was never sure if the tables or references being referenced were from the main text or from the supplement text 1, perhaps this could be made a little clearer to aid the reader. Secondly, the names of the isolates switched between, for e.g., MAK 65 and Tb065. For simplicity it could help to try and stick to one naming system.

It might be worth adding a sentence about why Tb236B was not followed up. Was this because it could not easily be distinguished from MAK65 by microsatellite analysis?

S3 Figure: Legend describes red bars but there are none in the figure.

Table 2: In the legend for Sheet 2, the cut-off for the increase is missing.

Significance

Though the T. b. brucei Lister 427 and EATRO1125 strains are used most commonly for laboratory-based research, they have been extensively passaged in vitro without characterisation of the changes that have occurred between them and the original isolate, or indeed, between laboratories.

Mulindwa et al. demonstrated that changes to gene copy number occur rapidly upon adaptation to culture of field isolates, and that different cultures of the same isolate can furthermore have different ploidies. This is an important advance and raises awareness a) that trypanosomes undergo changes upon laboratory adaptation, b) of the nature of some of these changes and c) that the changes can occur rapidly. Changes to gene copy number have also been shown to effect Leishmania donovani upon culture adaptation (Prieto Barja et al, 2017, doi: 10.1038/s41559-017-0361-x).

This study will be of interest to the trypanosome community in general, but particularly those who work on biology that we already know is impacted by prolonged in vitro passage-differentiation, virulence and antigenic variation.

Reviewer expertise: BSF differentiation, antigenic variation

Author Response:

Reviewer #1 (Public Review):

[...] Authors' rigorous experimental design (based on bacterial genetics and structural biology), solid biochemical assays (including photo-crosslinking, cysteine crosslinking, and Western blotting), and carefully drawn interpretation and conclusions are impressive. Finally, authors delineate the mechanisms of BepA activation and LptD biogenesis, which are supported by the current and previous studies by the authors and other research groups.

Thank you for the nice summarization and the positive evaluation of our study.

While this is overall a wonderful piece of work, this manuscript would be further improved by clarifying the following points:

1. Authors examined how mutations (Pro and Cys scanning) on the edge-strand of BepA affected degradation and maturation of LptD.

It was assumed that these mutations impact the structure of BepA only locally. However, a mutational effect can be propagated in an unexpected way affecting the structural integrity of other regions. Although authors tested that A106P retains proteolytic activity as shown by self-cleavage, a similar test (for example, in vitro experiments using a structureless substrate) may need to be extended to other mutations to support the conclusions.

Thank you for the suggestion. Unfortunately, we have not succeeded in reproducibly detecting the proteolytic activity of BepA with purified BepA even when an unstructured substrate (-casein) is used and the assay was conducted at an elevated temperature, possibly because the protases activity of isolated BepA is tightly repressed by the mechanism that included His-246-mediated regulation as described in our paper (please see Introduction). Although BepA mutants with a mutation of His-246 or a deletion of H9 loop (these mutations release the His-246-mediated repression) significantly degrade -casein, a combination of these mutations with the edge-strand mutations should make the interpretation of the results complicated. We thus think that the suggested experiments cannot be conducted soon.

Instead, we described the following points in the revised manuscript. Although we mentioned the self-cleavage activity of only the A106P mutant in the original manuscript, our results showed that the other edge-strand Pro mutants (other than F107P) exhibited significant self-cleavage activities as well (Figure 1-figure supplement 2B). In addition, the Pro mutants other than the A106P mutant degraded mis- or un-folded BamA at a detectable level (Figure 1-figure supplement 2A). Furthermore, all the Pro mutants accumulated at a level comparable to that of wild-type BepA. These observations together indicate that most of the Pro mutations specifically affected the edge-strand structure, but not drastically altered the active site or the protein's overall structures. We described the above points in the revised text (line 174 in p7 to line 181 in p8).

1. In the result (Line 159), authors report chaperone-like activity of BepA. Here, the term "chaperone-like" is rather obscure regarding whether this activity facilitates LptD maturation without proteolysis (i.e., via holdase activity), or involves proteolysis as a part of quality control mechanisms. In another experiment, authors show that the chaperone-like activity may not necessarily involve proteolysis. It would be good to describe a possible molecular principle of how the edge-stand binding to the substrate can lead to chaperone activity.

We suppose that the interaction of BepA (via the edge-strand) with an assembly intermediate of LptD on the BAM complex stabilizes the partially unfolded assembly intermediate of LptD on the BAM complex to help the association of LptE with LptD. This was explained in Discussion (lines 388–392 in p16) and the legend to Figure 5.

Reviewer #2 (Public Review):

The authors found that a conserved β-strand (edge-strand β2 of BepA) directly contacts with the N-terminal half of the β-barrel-forming domain of an immature LptD; the C-terminal region of the β-barrel-forming domain of the BepA-bound LptD intermediate interacts with a "seam" strand of BamA in the BAM complex. By combining crosslinking and mutational studies, they showed that the edge-strand of BepA may have both the proteolytic and the chaperone-like functions. Based on the authors' previous studies of BepA, they proposed a model that the edge-strand and His switch of BepA regulate BepA in LptD assembly and degradation.

Thank you for the nice summarization of our study.

Reviewer #3 (Public Review):

[...] By performing an impressive systematic cross linking analysis, combined with previous known findings, the authors are able to dissect the general architecture of how BepA interacts with beta-barrel substrates as they are being assembled by the Bam complex. The experiments presented are nicely executed and the data are of high quality. I am convinced that the edge strand of BepA interacts with LptD, likely as it is assembling on the Bam complex. It is also clear that this interaction is functionally important because mutations in this region that disrupt the BepA-LptD interaction interfere with LptD maturation and degradation. This suggests that the substrate binding to the protease domain of BepA is important for both its chaperone and proteolytic activity. The work is well executed and will be of value to others interested in the regulation of membrane protein folding and, more broadly, in the biogenesis of the bacterial cell envelope.

Thank you for the nice summarization and the positive evaluation of our study.

While the authors conclusively establish a link between this region of BepA and its function, the data do not explain the underlying mechanism of how BepA discriminates between substrates targeted for integration into the membrane and those targeted for destruction. The model proposed at the end incorporates the presence of the edge strand of BepA, but its role in the process remains vague. As mentioned in the discussion, the mechanisms that control the switch from chaperone to protease function in BepA is likely governed by the loops that gate access to the catalytic residues proximal to the edge strand. It is possible that the edge strand may just be reporting on substrate binding to the protease domain active site. While this may be important for substrate recognition, it does not mean that the edge strand-substrate interaction plays a deterministic role in subsequent protein triage during LptD assembly.

Our data demonstrated that the edge-strand of BepA directly binds a substrate. As pointed out by the reviewer, the involvement of the edge-strand in substrate binding has been known for other proteases. However, it was not known whether the substrate binding at the edge-strand contributes to the chaperone-like function; it was possible that the binding sites of a substrate on BepA during its proteolysis and its maturation are totally different as the chaperone-like activity of BepA is independent of its protease activity (it was conceivable, for example, that substrate binding during it maturation occurs on the surface of the C-terminal TPR domain that has been shown to interact with LptD）. Our results showed that the defective binding of a substrate (LptD) at the edge-strand impairs not only its proteolysis but also its normal maturation (assembly). Because the edge-strand-bound substrate would be directly presented to the proteolytic active site for its degradation, this binding step should be important for the determination of the fates of the bound substrate. Our results strongly suggest that the substrate binding by the edge-strand is a crucial common step required for the subsequent protein triage during the LptD assembly.

I think this is a super key consideration, and I think we all have to acknowledge the possibility that our findings for this project are only a part of the story, and may not be representative of the entire campus community. I'm really looking forward to seeing the final project and learning about your interpretation of the results.

Author Response:

Reviewer #1:

Insulin-secreting beta-cells are electrically excitable, and action potential firing in these cells leads to an increase in the cytoplasmic calcium concentration that in turn stimulates insulin release. Beta-cells are electrically coupled to their neighbours and electrical activity and calcium waves are synchronised across the pancreatic islets. How these oscillations are initiated are not known. In this study, the authors identify a subset of 'first responders' beta-cells that are the first to respond to glucose and that initiate a propagating Ca2+ wave across the islet. These cells may be particularly responsive because of their intrinsic electrophysiological properties. Somewhat unexpectedly, the electrical coupling of first responder cells appears weaker than that in the other islet cells but this paradox is well explained by the authors. Finally, the authors provide evidence of a hierarchy of beta-cells within the islets and that if the first responder cells are destroyed, other islet cells are ready to take over.

The strengths of the paper are the advanced calcium imaging, the photoablation experiments and the longitudinal measurements (up to 48h).

Whilst I find the evidence for the existence of first responders and hierarchy convincing, the link between the first responders in isolated individual islets and first phase insulin secretion seen in vivo (which becomes impaired in type-2 diabetes) seems somewhat overstated. It is is difficult to see how first responders in an islet can synchronise secretion from 1000s (rodents) to millions of islets (man) and it might be wise to down-tone this particular aspect.

We thank the reviewer for highlighting this point. We acknowledge that we did not measure insulin from individual islets post first responder cell ablation, where we observed diminished first phase Ca2+. We do note that studies have linked the first phase Ca2+ response to first phase insulin release [Henquin et al, Diabetes (2006) and Head et al, Diabetes (2012)], albeit with additional amplification signals for higher glucose elevations. Thus a diminished first phase Ca2+ would imply a diminished first phase insulin (although given the amplifying signals the converse would not necessarily be the case).

Nevertheless there are also important caveats to our experiment. Within islets we ablated a single first responder cell. In small islets this ablation diminished Ca2+ in the plane that we imaged. In larger islets this ablation did not, pointing to the presence of multiple first responder cells. Furthermore we only observed the plane of the islet containing the ablated first responder. It is possible elsewhere in the islet that [Ca2+] was not significantly disrupted. Thus even within a small islet it is possible for redundancy, where multiple first responder cells are present and that together drive first phase [Ca2+] across the islet. Loss of a single first responder cell only disrupts Ca2+ locally. That we see a relationship between the timing of the [Ca2+] response and distance from the first responder would support this notion. Results from the islet model also support this notion, where >10% of cells were required to be ablate to significantly disrupt first-phase Ca2+.

While we already discuss the issue of redundancy in large islets and in 3D, we now briefly mention the importance of measuring insulin release.

Reviewer #2:

Kravets et al. further explored the functional heterogeneity in insulin-secreting beta cells in isolated mouse islets. They used slow cytosolic calcium [Ca2+] oscillations with a cycle period of 2 to several minutes in both phases of glucose-dependent beta cell activity that got triggered by a switch from unphysiologically low (2 mM) to unphysiologically high (11 mM) glucose concentration. Based on the presented evidence, they described a distinct population of beta cells responsible for driving the first phase [Ca2+] elevation and characterised it to be different from some other previously described functional subpopulations.

Strengths:

The study uses advanced experimental approaches to address a specific role a subpopulation of beta cells plays during the first phase of an islet response to 11 mM glucose or strong secretagogues like glibenclamide. It finds elements of a broadscale complex network on the events of the slow time scale [Ca2+] oscillations. For this, they appropriately discuss the presence of most connected cells (network hubs) also in slower [Ca2+] oscillations.

Weakness:

The critical weakness of the paper is the evaluation of linear regressions that should support the impact of relative proximity (Fig. 1E), of the response consistency (Fig. 2C), and of increased excitability of the first responder cells (Fig. 3B). None of the datasets provided in the submission satisfies the criterion of normality of the distribution of regression residuals. In addition, the interpretation that the majority of first responder cells retain their early response time could as well be interpreted that the majority does not.

We thank the reviewers for their input, as it really opened multiple opportunities for us to improve our analysis and strengthen our arguments of the existence and consistency of the first responder cells. We present more detailed analysis for these respective figures below and describe how these are included in the manuscript.

As it is described below, we performed additional in-depth analysis and statistical evaluation of the data presented in figures 1E, 2C, and 3B. We now report that two of the datasets (Fig.1 E, Fig.2 C) satisfy the criterion of normality of the distribution of regression residuals. The third dataset (Fig.3 B) does not satisfy this criterion, and we update our interpretation of this data in the text.

Figure 1E Statistics, Scatter: We now show the slope and p-value indicating deviation of the slope from 0, and r^2 values in Fig.1 E. While the scatter is large (r^2=0.1549 in Fig.1E) for cells located at all distances from the first responder cell, we found that scatter substantially diminishes when we consider cells located closer to the first responder (r^2=0.3219 in Fig.S1 F): the response time for cells at distances up to 60 μm from the first responder cells now is shown in Fig.S1 F. The choice of 60 μm comes from it being the maximum first-to-last responder distance in our data set (see red box in Fig.1D).

Additionally, we noticed that within larger islets there may be multiple domains with their own first responder in the center (now in Fig.S1 E) and below. Linear distance/time dependence is preserved withing each domain.

Figure 1E Normality of residuals: We appreciate reviewer’s suggestion and now see that the original “distance vs time” dependence in Fig.1 E did not meet normality of residuals test. When plotted as distance (μm)/response time (percentile), the cumulative distribution still did not meet the Shapiro-Wilk test for normality of residuals (see QQ plot “All distances” below). However, for cells located in the 60 μm proximity of the first responder, the residuals pass the Shapiro- Wilk normality test. The QQ-plots for “up to 60 μm distances” are included in Fig.S1 G.

Figure 2C Statistic and Scatter: After consulting a biostatistician (Dr. Laura Pyle), we realized that since the Response time during initial vs repeated glucose elevation was measured in the same islet, these were repeated measurements on the same statistical units (i.e. a longitudinal study). Therefore, it required a mixed model analysis, as opposed to simple linear regression which we used initially. We now have applied linear mixed effects model (LMEM) to LN- transformed (original data + 0.0001). The 0.0001 value was added to avoid issues of LN(0).

We now show LMEM-derived slope and p-value indicating deviation of the slope from 0 in Fig.2 C. Further, we performed sorting of the data presented in Fig.2 C by distance to each of the first responders (now added to Fig.2D). An example of the sorted vs non-sorted time of response in the large islet with multiple first responders is added to the Source Data – Figure 1. We found a substantial improvement of the scatter in the distance- sorted data, compared to the non-sorted, which indicates that consistency of the glucose response of a cell correlates with it’s proximity to the first responder. We also discuss this in the first sub-section of the Discussion.

Figure 2C Normality of residuals: The residuals pass Shapiro-Wilk normality test for LMEM of the LN-transformed data. We added very small number (0.0001) to all 0 values in our data set, presented in Fig.2C, D, and Fig.S4 A, to perform natural-log transformation. Details on the LMEM and it’s output are added to the Source data – Statistical analysis file.

Figure 3B Statistic and Scatter: We now show LMEM-derived slope and p-value, indicating deviation of the slope from 0, values in Fig.3 B (below). The LMEM-derived slope has p-value of 0.1925, indicating that the slope is not significantly different from 0. This result changes our original interpretation, and we now edit the associated results and discussion.

Figure 3B Normality of residuals: This data set does not pass Shapiro-Wilk test.

A major issue of the work is also that it is unnecessarily complicated. In the Results section, the authors introduce a number of beta cell subpopulations: first responder cell, last responder cell, wave origin cell, wave end cell, hub-like phase 1, hub-like phase 2, and random cells, which are all defined in exclusively relative terms, regarding the time within which the cells responded, phase lags of their oscillations, or mutual distances within the islet. These cell types also partially overlap.

To address this comment, we added Table 1 to describe the properties of these different populations.

Their choice to use the diameter percentile as a metrics for distances between the cells is not well substantiated since they do not demonstrate in what way would the islet size variability influence the conclusion. All presented islets are of rather a comparable size within the diffusion limits.

We replaced normalized distances in Fig.1 D with absolute distance from first responder in μm.

The functional hierarchy of cells defining the first response should be reflected in the consistency of their relative response time. The authors claim that the spatial organisation is consistent over a time of up to 24 hours. In the first place, it is not clear why would this prolonged consistency be of an advantage in comparison to the absence of such consistency. The linear regression analysis between the initial and repeated relative activation times does suggest a significant correlation, but the distribution of regression residuals of the provided data is again not normal and non-conclusive, despite the low p-value. 50% of the cells defined a first responder in the initial stimulation were part of that subpopulation also during the second stimulation, which is rather random.

We began to describe our analysis of the response time to initial and repeated glucose stimulation earlier in this reply. Further evidence of the distance-dependence of the consistency of the response time is now presented in Fig.S4 A: a response time consistency for cells at 60 μm, 50μm, and 40 μm proximity to the first responder. The closer a cell is located to the first responder, the higher is the consistency of its response time (the lower the scatter), below.

If we analyze this data with a linear regression model, where the r^2 allows us to quantitatively demonstrate decrease of the scatter, we observe r^2 of 0.3013, 0.3228, 0.3674 respectively for cells at 60 μm, 50μm, and 40 μm proximity to the first responder (below). This data is not included in the manuscript because residuals do not pass Shapiro-Wilk Normality test for this model (while they do for the LMEM).

One of the most surprising features of this study is the total lack of fast [Ca2+] oscillations, which are in mouse islets, stimulated with 11 mM glucose typically several seconds long and should be easily detected with the measurement speed used.

Our data used in this manuscript contains Ca2+ dynamics from islets with a) slow oscillations only, b) fast oscillations superimposed on the slow oscillations, c) no obvious oscillations (likely continual spiking). Representative curves are below. Because we focused our study on the slow oscillations, we used dynamics of type (a) in our figures, which formed an impression that no fast oscillations were present. In our analysis of dynamics of type (b) we used Fourier transformation to separate slow oscillations from the fast (described in Methods). Dynamics of type (c) were excluded from the analysis of the oscillatory phase, and instead only used for the first-phase analysis. We indicate this exclusion in the methods.

And lastly, we should also not perpetuate imprecise information about the disease if we know better. The first sentence of the Introduction section, stating that "Diabetes is a disease characterised by high blood glucose, …" is not precise. Diabetes only describes polyuria. Regarding the role of high glucose, a quote from a textbook by K. Frayn, R Evans: Human metabolism - a regulatory perspective, 4rd. 2019 „The changes in glucose metabolism are usually regarded as the "hallmark" of diabetes mellitus, and treatment is always monitored by the level of glucose in the blood. However, it has been said that if it were as easy to measure fatty acids in the blood as it is to measure glucose, we would think of diabetes mellitus mainly as a disorder of fat metabolism."

We acknowledge that Diabetes alone refers to polyurea, and instead state Diabetes Mellitus to be more precise to the disease we refer to. We stated “Diabetes is a disease characterized by high blood glucose, ... “ as this is in line with internationally accepted diagnoses and classification criteria, such as position statements from the American Diabetes Association [‘Diagnosis and Classification of Diabetes Mellitus” AMERICAN DIABETES ASSOCIATION, DIABETES CARE, 36, (2013)]. We certainly acknowledge the glucose-centric approach to characterizing and diagnosing Diabetes Mellitus is largely born of the ease of which glucose can be measured. Thus if blood lipids could be easily measured we may be characterizing diabetes as a disease of hyperlipidemia (depending how lipidemia links with complications of diabetes).

Reviewer #3 (Public Review): 

Pancreatic beta cells in each islet of Langerhans act cooperatively to produce a coordinated response to blood glucose, which takes the form of roughly synchronous electrical oscillations that result in coherent pulses of insulin secretion. However, it has become more appreciated recently that the cells are heterogeneous and that there may subgroups of cells within the islet that contribute in different ways or control different aspects of the collective behavior. 

This paper addresses a subset of cells, termed first responders, that are the earliest to transition into activity when glucose is stepped up from a sub-stimulatory level. It is shown that the first responders determine the first transient phase of electrical activity, and implicitly secretion, that precedes the start of steady-state oscillations of the second phase. The first phase is of interest for the pathogenesis of diabetes because it is (or claimed to be) one of the first indications of the disease. The clinical data on this are actually ambiguous, but nonetheless the first phase is an important aspect of how secretion from islets is organized. This is clear from the existence of a subset of readily releasable insulin vesicles, but the electrical activity correlates that synergize with vesicle availability are less well understood. As such, the paper is an important contribution to both islet biology and potentially diabetology. 

The approach is to use a genetically encoded calcium indicator, GCamps, and confocal imaging to identify cells that show the earliest rise in calcium and then to verify that this property remains consistent when the glucose stimulus is repeated about an hour later. Further testing over 48 h shows the first responder slowly fading. This thus appears to be a matter of continuous variation of cell properties within the islet and moreover one that is persistent but decaying over long-enough periods of time, rather than a discrete sub-type of beta cell. This is confirmed by simulations using an islet model in which first responders emerge from random variation of properties. The text is a bit ambiguous about whether we should think of the glass as half full or half empty and could be clarified in this regard. 

The properties that matter are the density of KATP channels and gap junctional coupling which are both lower. This is also confirmed by simulations. The intriguing suggestion is made that there may be a reciprocal negative feedback relationship between these quantities that regulates their variation over time. This would be rather different from a persistent genetic difference and would be a good subject for future investigation. 

Interestingly and perhaps surprisingly, increased sensitivity to glucose is not a feature of the first responders. In contrast, other putative "leader" or "hub" cells that have been identified for the second phase of electrical activity are proposed to have increased glucose sensitivity. This and other features lead to the conclusion that the two types of leader cells are probably distinct. The much-discussed topic of hub cells provides interesting context and relevance to the paper, but its results stand by themselves and can be judged independent of hubs.

Author Response:

Reviewer #1 (Public Review):

Sokolsky et al. propose a new statistical model class for descriptive modeling of stimulus encoding in the spiking activity of neural populations. The main goals are to provide a model family that (G1) captures key activity statistics, such as spike count (noise) correlations, and their stimulus dependence, in potentially large neural populations, (G2) is relatively easy to fit, and (G3) when used as a forward encoder model for Bayesian decoders leads to efficient and accurate decoding. There are also three additional goals or claims: (C1) that this descriptive model family can serve to quantitatively test computational theories of probabilistic population coding against data, (C2) that the model can offer interpretable representations of information-limiting noise correlations, (C3) that the model can be extended to the case of temporal coding with dynamic stimuli and history dependence.

The starting point of their model is a finite mixture of independent Poisson distributions, which is then generalized and extended in two ways. Due to the "mixture", the model can account for correlations between neurons (see G1). As any mixture model, the model can be viewed in the language of latent variables, which (in this case) are discrete categorical variables corresponding to different mixture components. The two extensions of the model are based on realizing that the joint distribution (of the observed spike counts and the latent variables) is in the exponential family (EF), which opens the door to powerful classical results to be applied (e.g. towards G2-G3), and allows for the two extensions by: (E1) generalizing Poisson distributions in mixture components to Conway-Maxwell-Poisson distributions, and (E2) introducing stimulus dependence by allowing the natural parameters of the EF to depend on stimulus conditions. They call the resulting model a Conditional Poisson Mixture or CPM (although the "Poisson" in CPM really means Conway-Maxwell-Poisson). E1 is key for capturing under-dispersion, i.e. Fano Factors below 1. For the case of discrete set of stimulus conditions, they propose minimal, maximal versions of E2; depending on which natural parameters are stimulus dependent. In the case of a continuum of stimuli (they only consider 1D continuum of stimulus orientations, e.g. in V1 encoding) they also consider a model-based parametric version of the minimal E2 which gives rise to Von Mises orientation tuning curves.

Strengths:

-Proposing a new descriptive encoding model of spike responses that can account for sub-poissonian and correlated noise structure, and yet can be tractably fit and accurately decoded.

-Their experiments with simulated and real (macaque V1) data presented in Figs. 2-5 and Tables 1-2 provide good evidence that the model supports G1-3.

-Working out a concrete Expectation Maximization algorithm that allows efficient fits of the model to data.

-Exploiting the EP framework to provide a closed form expression for the model's Fisher Information for the minimal model class, a measure that plays a key role in theoretical studies of probabilistic population coding.

As such, the papers makes a valuable contribution to the arsenal of descriptive models used to describe stimulus encoding in neural population, including the structure and stimulus dependence of their higher-order statistics.

Thank you very much for your thorough, exact, and positive evaluation of our manuscript!

Weaknesses:

1) I found the title and abstract too vague, and not informative enough as to the concrete contributions of this paper. These parts should more concretely and clearly describe the proposed/developed model family and the particular contributions listed above.

We found your summary of the paper and model to be highly accurate, and we rewrote the abstract to summarize the key strengths as you’ve listed them. We found it difficult to develop a more exact title which wasn’t overlong, so we left it as is.

2) I was not convinced about claims C1 and C2 (which also contribute to the vagueness of abstract), but I think even without establishing these claims the more solid contributions of the paper are valuable. And while I can see how the model can be extended towards C3, there are no results pertaining to this in the current paper, nor even a concrete discussion of how the model may be extended in this direction.

2.1) Regarding C1, the claim is supposed to follow from the fact that the model's joint distribution is in the exponential family (EF), and that they have reasonably shown G1-G3 (in particular, that it captures noise correlations and its Bayesian inversion provides an accurate decoder). While I agree with the latter part, what puzzles me is that in the probabilistic population coding (PPC) theoretical models that claim can be quantitatively tested using their descriptive model are, as far as I remember/understand, the encoder itself is in EF. By contrast here the encoder is a mixture of EF's and as such is not itself in EF. Perhaps this distinction is not key to the claim - but if so, this has to be clearly explained, and more generally the exact connection between the descriptive encoder model here and the models used in the PPC literature should be better elaborated.

This claim was indeed poorly explained in our manuscript, and not self-evident. There is a deeper connection between our conditional models and PPCs, which we now make explicit in a new section of the manuscript (Constrained conditional mixtures support linear probabilistic population coding, line 364), which includes an equation (Equation 4) that shows their exact relationship.

2.2) Regarding C2, I do not see how their results in Fig 5 (and corresponding section) provide any evidence for this claim. As a theoretical neuroscientist, I take "interpretable" to mean with a mechanistic or computational (theoretical) interpretation. But, if anything, I think the example studied in Fig 5 provides a great example of the general point: that even when successful descriptive models accurately capture the statistics of data, they may nevertheless not reveal (or even hide or mis-identify) the mechanisms underlying the data. In this example's ground-truth model, the stimulus (orientation) is first corrupted by input noise and then an independent population of neurons with homogeneous tuning curves (and orientation-independent average population rate) responds to this corrupted version of the stimulus. That is a very simple AND mechanistic interpretation (which of course is not manifest to someonw only observing the raw stimulus and spiking data). The fit CPM, on the other hand, does not reveal the continuous input noise mechanism (and homogeneous population response) directly, but instead captures the resulting noise correlation structure by inferring a large (~20) number of mixture components, in each of which population response prefers a certain orientation. For a given stimulus orientation, the fluctuations between (3-4 relevant) mixture components then approximate the effect of input noise. This captures the generated data well, but misses the true mechanism and its simpler interpretation. Let me be clear that I don't take this as a fault of their descriptive model. This is a general phenomenon, despite which their descriptive model, like any expressive and tractible descriptive model, still can be a powerful tool for neural data analysis. I'm just not convinced about the claim.

This is a very fair point, and we’ve reformulated a few passages to emphasize that the model is primarily descriptive, at least in our applications in the paper (see new section title at like 393, the first corresponding paragraph).

2.3) Regarding C3, I think the authors can at least add a discussion of how the model can be extended in this direction (and as I'm sure they are aware, this can be done by generalizing the Von Mises version of the model, whereby the model I believe can be more generally thought of as a finite mixture of GLMs).

In Appendix 4 we detail the relationship between CPMs and GLMs. We also note here that, at least as far as we understand, CPMs are formally distinct from finite mixtures of GLMs — the easiest way to see this distinction is to note that the index probabilities of a CPM depend on the stimulus, whereas the equivalent index probabilities in a finite mixture of GLMs would not. We have also explained this in Appendix 4.

Reviewer #2 (Public Review):

Sokoloski, Aschner, and Coen-Cagli present a modeling approach for the joint activity of groups of neurons using a family of exponential models. The Conway-Maxwell (CoM) Poisson models extend the "standard" Poisson models, by incorporating dependencies between neurons.

They show the CoM models and their ability to capture mixture of Poisson distributions. Applied to V1 data from awake and anesthetized monkeys, they show it captures the Fano Factor values better than simple Poisson models, compare spike count variability and co-variability. Log-likelihood ratios in Table 1 show on-par or better performance of different variant of the CoM models, and the optimal number of parameters to use for maximizing the likelihood [balancing accuracy and overfitting] and are useful for decoding. Finally, they show how the latent variables of the model can help interpret the structure of population codes using simple simulated Poisson models over 200 neurons.

In summary, this new family of models offer a more accurate approach to the modeling and study of large populations, and so reflects the limited value of simple Poisson based models. Under some conditions it gives has higher likelihood than Poisson models and uses fewer parameters than ANN model.

However, the approach, presentation, and conclusions fall short on several issues that prevents a clear evaluation of the accuracy or benefits of this family of models. Key of them is the missing comparison to other statistical models.

1) Critically, the model is not evaluated against other commonly used models of the joint spiking patterns of large populations of neurons. For example: GLMs (e.g. Pillow et al Nature 2008), latent Gaussian models (e.g. Macke et al Neural Comp 2009), Restricted Boltzmann Machines (e.g. Gardella et al PNAS 2018), Ising models for large groups of neurons (e.g. Tkacik etal PNAS 2015, Meshulam et al Neuron 2017), and extensions to higher order terms (Tkacik et al J Stat Mech 2013), coarse grained versions (Meshulam et al Phys Rev Lett 2019), or Random Projections models (Maoz et al biorxiv 2018).

. Most of these models have been used to model comparable or even larger populations than the ones studied here, often with very high accuracy, measured by different statistics of the populations and detailed spiking patterns (see more below). Much of the benefit or usefulness of the new family of models hinges on its performance compared to these other models.

We agree very much with this point, and have done our best to address it by thoroughly comparing our model with a factor analysis encoding model in Appendices 1 and 2, and summarizing these results at appropriate points in the manuscript (lines 196–199 and 325–328). In particular, we visualized and compared the performance of factor analysis with our mixture models, and found that (i) factor analysis is better at capturing the first and second order statistics of the data, but (ii) when evaluated on held-out data, the performance gap more-or-less vanishes. Moreover, we found that an encoding model based on FA performs poorly as a Bayesian decoder, and we provided preliminary evidence that this is because our mixture models can capture higher-order statistics that FA cannot. We believe that these results have been very valuable to conveying the strengths and weaknesses of the mixture model approach.

We have also extended the introduction to explain the differences between other model families suggested by the reviewer and our approach, to explain how the different assumptions about the form of data make it difficult to compare them quantitatively (see lines 42–63). To wit, GLMs and latent Gaussian models are both models that critically depend on modelling spike trains, and not spike counts. On the other hand, Restricted Boltzmann machines, Ising models, and random projection models all assume binary, rather than counting spiking data. As such, any comparison would depend on coming up with methods for either (i) reshaping our datasets and comparing spike- train/binary spike-count likelihoods to trial-to-trial likelihoods, or (ii) extending our conditional mixture approach to temporal/binary data, both of which are beyond the scope of our paper. We instead used factor analysis because it has been applied widely to modelling trial-to-trial spike counts, and thus avoid further transformations that might reduce the validity of our comparisons.

2) As some of these models are exponential models, their relations to the family of the models suggested by the authors is relevant also in terms of the learned latent variables. Moreover, the number of parameters that are needed for these different models should be compared to the CoM and its variants.

In our comparisons with factor analysis we also compared number of latent states/dimensions required to achieve maximum performance. Overall FA was consistently the most efficient, at least when evaluated on the ability to capture second-order statistics, although our mixture models also performed quite well with modest numbers of parameters.

3) The analysis focuses on simple statistics of neural activity, like Fano Factors (Fig. 2) and visual comparisons rather than clear quantitative ones. More direct assessments of performance in terms of other spiking statistics for single neurons and small groups (e.g., correlations of different orders ) and direct comparison to individual spiking patterns (which would be practical for groups of up to 20 neurons) would be valuable

In the Appendix 2 we evaluated the ability of our mixtures to capture the empirical skewness and kurtosis of recorded neurons, and found that the CoM-based mixture performs quite well (r2 for the CoM-Based mixture was between 0.6 and 0.9). Because FA cannot capture these higher-order moments, we speculate that modelling these higher-order moments is critical for maximizing decoding performance. This adds another perspective on the strengths of our approach, and we appreciate the suggestion.

Reviewer #3 (Public Review):

The authors use multivariate mixtures of Poisson or Conway-Maxwell-Poisson distributions to model neural population activity. They derive an EM algorithm, a formula for Fisher information, and a Bayesian decoder for such models, and show it is competitive with other methods such as ANNs. The paper is clear and didactically written, and I learned a lot from reading it. Other than a few typos the math and analyses appear to be correct.

Thank you for the positive evaluation!

Nevertheless there are some ways the study could be further improved.

Most important, code for performing these analyses needs to be publicly released. The EM algorithm is complicated, involving a gradient optimization on each iteration - it is very unlikely people will rewrite this themselves, so unless the authors release well-packaged and well-documented code, their impact will be limited.

We very much agree, and we have done this. We provide a link to our gitlab page, where all relevant code can be downloaded, and installation instructions are provided (we indicate this in the manuscript at lines 799–803).

Second, it would be nice to extend the model to continuous latent factors. It seems likely that one or two latent factors could do the work of many mixture components, as well as increasing the interpretability of the models.

We certainly agree that in some cases continuous latent variables could be much more parsi- monious. However, to the best of our knowledge most of the expressions that we rely on would no longer be closed-form, and so the machinery of the model would require suitable approximations. Nevertheless, it’s an interesting possibility that we now address in the Discussion (lines 482–491).

Third, it would be interesting to see the models applied to more diverse types of population data (for example hippocampal place field recordings).

We certainly agree with the importance of applying our model to other datasets, and indeed the purpose of our manuscript is to offer a method that can be applied broadly, and our goal in making the code available publicly is to facilitate that. However, we have decided to maintain the focus of this manuscript on the method itself, and limit the application to one kind of data (V1), for which we also now provide more extensive analysis and quantification of the response statistics (Figure 2 C-D, Figure 3 G-H, Appendix 2), a study of the sample sizes required to fit the model (Appendix 3), and model-comparison (Appendix 1–2). Overall we feel that the paper is already quite long and dense even when limited to a single kind of data. We believe applications to multiple kinds of data would perhaps be better suited for a different study, focusing on the comparisons between them. In that regard, we are certainly open to future collaborations on large-scale recordings from various stimulus-driven brain areas.

Fourth, how does a user choose how many mixture components to add?

To clarify this, we’ve added a section in the methods (Strategies for choosing the CM form and latent structure), and in particular the number of mixture components.

Author Response:

Reviewer #2 (Public Review):

[...] 1) A weakness of the paper is the disruption of the complex during cryoEM grid preparation resulting in about half of the observed particles missing the membrane arm and likely also contributing to the disorder and biased orientation seen in the intact complexes. This leads to poor density in the membrane arm for all of the intact complex I structures presented and large variations in the local resolution of the membrane arm focused refinement.

Purified E. coli complex I has always been known to be labile in particularly at the junction of peripheral and membrane arms (https://pubmed.ncbi.nlm.nih.gov/12637579/).

Air-water interface likely plays a role in disrupting the complex in addition to other possible causes. Indeed, the dissociated arms, preferred particle orientation, and low protein concentration (~0.1 mg/ml) used to produce grids with high particle density all indicate that reconstituted complex I does interact with air-water interface. While disruption and denaturation of protein complexes on air-water interface has been well documented, (https://pubmed.ncbi.nlm.nih.gov/3043536/, https://pubmed.ncbi.nlm.nih.gov/30932812/ ), we are not aware of examples where air-water interfaces caused higher mobility of a complex or induced a stable conformation, different from the one in bulk solution. Therefore, we think that air-ware interface is neither the cause of the observed high arms mobility nor of their relative rotation.

Preferential orientation was observed in the cryo-EM studies of most complex I homologs (Gutiérrez-Fernández et al., 2020; Parey et al., 2019; Zhu et al., 2016) as well as of other proteins, suggesting that adsorption of complex I on air water interface is a common phenomenon. In this case it is not clear why relative movement of the arms observed in all the structurally characterized complex I homologs is not due to the air-water interface, but in the case of E. coli complex it is.

To provide additional support to our interpretation of the structural data we purified complex I in detergent LMNG, showed that it catalyzes redox reactions and solved its structure to resolution of 6.7 Å (Figure 6 and corresponding figure supplements). Because cryo-EM grids had to be prepared at a protein concentration of 2-3 mg/ml and the particles displayed nearly homogeneous distribution of orientations, we conclude that the interaction with the air-water interface was reduced. Still, the complex assumes a very similar, or even somewhat more uncoupled conformation and the relative mobility of the arms remained comparable to that in the nanodisc-reconstituted complex reconstructions. These data allow us to rule out the air-water interface and reconstitution of the protein into lipid nanodiscs as the possible causes of the high mobility and the unusual relative position of the arms.

The corresponding modifications were added to the manuscript on lines 372-382:

“To better understand the reasons for the observed uncoupled conformation and the missing density for HTMH1, we purified E. coli complex I in detergent LMNG, showed that it can catalyze redox reactions (Figure 6 - figure supplement 1) and solved its structure to resolution of 6.7 Å (Figure 6 - figure supplement 2). The detergent-solubilized complex also displays high relative mobility of the arms (Figure 6 - figure supplement 3) and has uncoupled conformation (Figure 6). Its peripheral arm is rotated even further away from the expected coupled state position than in the nanodisc-reconstituted structures. Both the cryo-EM sample preparation conditions and more homogeneous distribution of particle orientations indicate that interaction of the complex with air-water interface was significantly reduced when compared with the complex in nanodiscs. This allows us to conclude that neither air-water interface nor reconstitution into nanodiscs cause the uncoupled conformations.”

It is not very clear what referee means by “poor”, when referring to the focused density of the membrane arm. The density corresponds well to the reported resolution of 3.7 Å. Indeed, it is in a stark contrast with the quality of the density obtained for the peripheral arm at 2.1 Å resolution. Given high mobility of the membrane arm it had to be refined essentially independently of the peripheral arm which remains still challenging for a ~200 kDa membrane protein without water-soluble domains in lipid nanodiscs. The density is heterogeneous as clearly stated at the beginning of the section “Structure of membrane arm” from line 264:

“The model of complete membrane arm, including the previously missing subunit NuoH (Efremov and Sazanov, 2011), was built into the density map with local resolution better than 3.5 Å at the arm center and approximately 4.0 Å at its periphery (Figure 1A, Figure 1 - figure supplement 4).”

Finally, for most complex I homologs the resolution was gradually improved over several years, as reflected in multiple publications of essentially the same structures. In contrast, no high-resolution structure information was available for the intact E. coli complex I until now. Therefore, it would be unreasonable to expect the complete structure to be solved at resolution of 2 Å at once.

The resolution of the membrane domain in reconstructions of complete complex I is indeed lower due to high flexibility of the complex and the fact that refinement naturally focuses on more stable peripheral arm that does not have heterogeneous nanodisc around and that contains Fe-S clusters enhancing particle alignment power. Still, these conformations clearly resolve the interface between subunits albeit at lower resolution.

This fact was also clearly stated at the beginning of results section lines 102-106:

“Three conformations of the entire complex were reconstructed to average resolutions between 3.3 and 3.7 Å (Figure 1 - figure supplement 4) resolving the interface between the arms; however, due to high-residual mobility of the arms, the antiporter-like subunits were resolved at below 8 Å (Figure 1 - figure supplement 4).”

2) A weakness of the paper is the disorder of important functional regions of the complex, namely the NuoH TMH1, whose disorder is unique to these nanodisc E. coli structures, and the NuoA TMH1-TMH2 loop. As the NuoH TMH1 forms part of the entry to the quinone tunnel of the complex, its absence in the structure leads to concerns regarding the function of the nanodisc preparation. Its absence it curious as this suggests flexibility of the helix, as pointed out by the authors, but the authors also state that there is not enough room in the nanodisc to accommodate this helix (given the visible density for the lipid and membrane scaffold protein). These observations suggest denaturation or unfolding in this region of the complex as opposed to simple flexibility.

According to the usual definition of complex I activity our preparation in nanodiscs is active. We complemented our data with additional measurements and included NADH:DQ assays (see next point) that also indicate that our preparation is active. Additional 3D reconstruction of E. coli complex I that we obtained for protein solubilized in LMNG does resolve HTMH1 and its environment appears to be more similar to other detergent-solubilized structures of complex I homologues. At the same time, the helices around HTMH1 appear to be more tightly packed and more curved than in the nanodiscs which may reflect suppressed dynamics and distorted protein conformation. Most importantly, the overall conformation of the complex remains nearly the same and still corresponds to what we call the uncoupled conformation. That of course does not allow us to say where HTMH1 is positioned within the nanodisc, but it does enable us to conclude that the local changes in the vicinity of HTMH1 do not influence the global conformation of the complex.

The additional structure is not described on lines 383-388:

“The HTMH1 helix is resolved in the detergent-solubilized complex (Figure 6A). Its density is weaker than that of the surrounding helices and it is strongly bent (Figure 6B). Simultaneously, HAH1 takes the conformation resembling other complex I homologs while ATMH1 bends towards the arm core. The arrangement of helices in detergent-solubilized reconstruction appears to be more compact and more bent than in the lipid environment which may restrain the otherwise more flexible HTMH1.”

In the revised discussion the environment of HTMH1 is described more clearly on lines 426-433:

“The absence of HTMH1 density in nanodiscs, but not in detergent, is another unique feature of E. coli complex I. HTMH1 is exposed to the lipid environment and the width of the nanodisc next to HTMH1 is similar to other regions around the membrane arm (Movie 1). Moreover, homology modelled HHTM1 fits the empty space without steric clashes suggesting that HHTM1 is dynamic rather than displaced or unfolded. By comparing the detergent-solubilized and reconstituted complexes we can conclude that position and dynamics of this helix is neither the cause of the uncoupled conformation nor of the high relative mobility of the arms.”

Disorder of ATMH1-TMH2 loop is not unique to E. coli complex I but also observed in some conformations of ovine complex I PDB 6zkd, 6zke, 6zkf.

3) Unfortunately, the NADH:Q1 functional data do not fully address these concerns at Q1 is far more soluble that the native Q8 substrate of the complex. Although the Q1 activity is sensitive to the inhibitor Piericidin A, which clearly demonstrates that the Q1 reduction is occurring in the native quinone binding site as Piericidin A binds specifically at that site, this does not preclude the possibility of Q1 accessing this binding site via a different path. In fact, the structures indicate that given the flexibility in the connection between peripheral and membrane arms of the complex, the quinone binding site is likely open to the cytoplasm. This leads the authors themselves to conclude that the structures presented are likely disrupted/uncoupled states in which the energy converting mechanism of the complex is not likely possible.

To address the raised concern, we have measured the activity of complex I in nanodiscs with less soluble decylubiquinone (DQ) as well as its inhibition. Small amounts of LMNG was used to increase the DQ solubility. Our results have confirmed that E. coli complex I in nanodiscs is active and the NADH:DQ activity is sensitive to piericidin A (see the modified Figure 1-figure supplement 2). We have also remeasured the Q1 activity and its inhibition which showed lower values than previously, due to a flaw in the activity measurements reported in the original submission (qualitatively, the results remained unchanged). Moreover, we have observed a similar activity results with somewhat higher values for E. coli complex I in LMNG (Figure 6-figure supplement 1). These data demonstrate that in the reconstituted complex I quinone analogues can enter the Q-site through the membrane. It is worth noting that due to extremely low solubility of longer quinones, including native ones, they are not used for activity measurements in purified preparations.

Regarding the complex I conformation, we do think our reconstruction represents uncoupled state which is not able to pump protons (as states in the title). We have improved the clarity of this point throughout the manuscript including the discussion lines starting from the line 412.

“The high mobility of the interfacial regions and the relative rotation of the arms disrupts conserved interfacial interactions and exposes Q-cavity to the solvent (Figure 5A). This differentiates E. coli complex I from its structurally characterized homologs in which the Q-cavity is sealed from the solvent. Thus, we interpret the observed conformation as an uncoupled state.”

And from line 469: “We also observed the relative rotation of the membrane and peripheral arms disrupting the conserved interface and trapping the complex in an uncoupled conformation. Whether this conformation is biologically relevant or is a result of protein purification is to be clarified by further research.”

4) A weakness of the paper is the building of atomic models into regions of the map which do not contain sufficient detail to warrant atomic models. This is particularly the case for the intact models of complex I as well as the membrane arm focused maps and results in low map-model correlations (0.58-0.71). The models were clearly highly restrained during refinement, resulting in good geometry, as is necessary for low resolution regions. But being able to restrain the geometry is not sufficient for placing atoms into regions where the density is weak or absent. If additional information was used in building/constraining the model, such as the X-ray structure, the regions of the model that are biased towards the X-ray structure model needs to be made clearer. Also, in several places in the membrane arm map residues bulge out of the density (side chain and main chain) leading to possible frame shifts with respect to the match between subsequent residues in the model and the map (see NuoM Ile168 for example).

A large part of the membrane domain has been solved using X-ray crystallography to resolution of 3.0 Å which was used as a starting model for model building, therefore we don’t think there are register shifts in our model. We used standard setting for model refinement in phenix_refine. Our building and refinement procedure has been described in fine details in the original submission, see from line 674:

“For the membrane domain, the previously obtained E. coli model (PDB ID: 3RKO) was real-space-refined in PHENIX. The missing NuoH subunit was homology-modelled using the T. thermophilus structure (PDB ID: 4HEA) in Coot 0.9. The final model was obtained after several rounds of manual rebuilding and real-space refinement using standard parameters with Ramachandran restrains, secondary-structure restrains applied to the NuoL TMH9-13, without ADP restrains, and with the optimized nonbonded_weight parameter. To generate the model of the complete complex I, the separate peripheral and membrane arm structures were combined and the missing parts at the interface (Table 2) were built manually. As the density of NuoL and NuoM was very poor in all the resolved full conformations, these subunits were subjected to rigid-body refinement in PHENIX, whereas the others were subjected to real-space refinement with minimization_global, local_grid_search, morphing, and ADP refinement. Ramachandran, ADP, and secondary-structure restrains were used. After manual rebuilding in Coot, real-space refinement of the full complex was performed with standard parameters and restrains.”

To improve clarity, we added a following sentence to the Results section from line 116:

“Using the resulting maps, atomic models of the peripheral and membrane arms have been built. The entire E. coli complex I was modelled by fitting models of the arms and extending additionally resolved loops and termini. Due to limited resolution, the antiporter-like subunits were refined as rigid bodies.”

The model has been improved and side chains with absent density were truncated to C position.

The density for focused refinement density of the membrane fragment is relatively week, but of sufficient quality to allow building side chains for most of the map. It even visualizes lipid densities (not described in the manuscript). Such weaker densities are common for small membrane proteins. While fully usable for model building, they naturally result in lower model map FSC and consequently, in lower real-space correlation. In addition, real-space correlation is lower when the map is heterogeneous, and it strongly depends on the way the heterogeneous map has been filtered. Therefore, lower cross correlations do not necessarily mean that the model fit is poor. In our case they reflect weaker signal to noise of the density. Model-map FSCs (Figure 1 figure supplement 4) are more informative than a single number and show that model-map cross correlations remain above 0.5 for the complete resolution range for all models.

5) A weakness of the paper is that several specific claims are made about the positions of side chains but, when investigated, the density for those side chains is poorly resolved. An example of this is NuoH Lys274, which is in a low-resolution region of the map and although is fit as well as possible must be considered low confidence given the local resolution (nearby residues Phe277 and Phe282 have almost no side chain density for example).

At lower resolution, a presence of residues density strongly depends on their mobility. Well-ordered residues may have well-defined densities while others, even in the proximity, may have a poor density. In the case of Lys274, there is a clear density for the side chain, its position makes chemical sense, and it is hydrogen-bonded to the backbone oxygen of Gly258. In fact, if examined closely, this is also the only meaningful position for Lys274 side chain. At the same time, the conformations of Phe277 and Phe282 are not restrained by interactions with other residues in their vicinity which is likely why their densities are weaker.

6) A weakness of the paper is that the conformational changes seen between the membrane and peripheral arm of the complex in the different 3D classes are difficult to interpret. It is unclear if they are mechanistically significant or, perhaps more likely given the amount of broken complex observed, due to partial disruption of the complex before it completely breaks apart.

As we discussed above, the observed multiple conformations are not due to the complex disruption. It is not very clear what the reviewer means by ‘difficult to interpret’. Many conformations of the peripheral and membrane arms observed for the complex I homologues are likely not mechanistically meaningful per see, but rather reflect overall flexibility of such a large complex. Here, our goal was to describe our structural data as accurately as possible which resulted in several resolved conformations.

We do think they all represent the uncoupled complex I, in this respect they do not have different mechanistic meanings. However, they do permit us to understand how the arms move relative to each other and what degree of freedom exists between them.

7) A strength of the paper is the interesting and original mechanistic proposal put forward by the authors. But a weakness is that it is unclear how this proposal stems from the structural data presented. Also, the arguments presented are difficult to follow in their current form and warrant a more detailed discussion with the requisite thermodynamic treatment. This may warrant a more complete discussion in an appendix or unless the authors can more convincingly show how the data presented in the paper suggests their proposed mechanism perhaps a separate review article. Furthermore, the proposed mechanism, as presented would make a simple prediction that in the absence of NuoM and NuoL (or equivalent subunits in other species) complex I would not pump any net protons. Experiments that are relevant to this prediction have been done in E. coli (NuoL deletion) and Y. lipolytica (nb8m deletion that results in loss of both NuoM and NuoL subunits). See https://pubmed.ncbi.nlm.nih.gov/21417432/ and https://pubmed.ncbi.nlm.nih.gov/21886480/. In both cases the complex is still able to pump protons. The behavior of the NuoL deletion in E. coli is reconcilable with their proposed mechanism as NuoM is still present, however, the case of the nb8m deletion in Y. lipolytica is more difficult to reconcile with their proposed mechanism. The authors would need to address these experiments in order to include their proposed mechanism.

The description of the mechanism has been modified. It is very briefly outlined in the main text along with the Figure 7 and more detailed description, including thermodynamic considerations, is moved to the supplementary text. We have also explained more clearly how the model stems from the experimental data on line 435:

“The absence of a continuous proton-translocation pathway between the Q-site and subunit NuoN, as well as high flexibility of the peripheral arm interface are not consistent with the recently proposed coupling mechanisms relying on specific movements of the interfacial loops (Cabrera-Orefice et al., 2018; Kampjut and Sazanov, 2020). This led us to ask whether a coupling mechanism consistent with known complex I properties, but without the movements of interfacial loops is conceivable.”

Furthermore, we state that at this point this is a hypothetical mechanism.

Supplementary data describing mechanism in more details now also includes the discussion of both papers mentioned by the reviewer from line 1368.

“Experiments with engineering E. coli complex I lacking subunit NuoL and Y. lipolytica complex I lacking homologs of subunits NuoM and NuoL (Dröse et al., 2011; Steimle et al., 2011)(Dröse et al., 2011; Steimle et al., 2011) (correspond to n=2 and 1, respectively) both suggested that the engineered complexes were active and for both constructs stoichiometry was estimated as 2H+/2e-. While NuoL deletion experiments support our model, the NuoL/M deletion clearly contradicts it. Both experiments should be interpreted cautiously, however. Results of NuoL deletion for E. coli complex I were not reproducible (Verkhovskaya and Bloch, 2012). In the case of Y. lipolytica, the homologs of NuoL/M dissociated from the complex along with another 11 subunits upon deletion of supernumerary subunit NB8M located at the tip of NuoL (Zickermann et al., 2015). Since the proton-translocating modules were not deleted per se, the presence of contaminating amounts of assembled complex I in the preparations that generated observed proton pumping cannot be completely excluded. It is important to note that mutation of the conserved ionizable residues on the interface between NuoN and NuoM, i.e. ME144 (Torres-Bacete et al., 2007) or its counter ion NK395 (Amarneh and Vik, 2003), result in a completely inactive complex I suggesting that dissociation of subunits NuoL/M also should render complex I inactive (Verkhovskaya and Bloch, 2012).”

The main problem with these experiments that that they have never been reproduced by other laboratories and are not completely consistent with the mutagenesis data. Deletion of subunits may also result in distinct pumping behavior of the remaining subcomplex. For example, it was shown for the bovine complex I that it can translocate Na+ ions in the deactive state (https://pubmed.ncbi.nlm.nih.gov/22854968/).

Appraisal of whether the authors achieved their aims, and whether the results support their conclusions:

8) Overall, despite the many strengths of this paper detailed above it is unclear whether the authors achieved their goal of a structure of functional E. coli respiratory complex I reconstituted in lipid nano-discs. It appears that under the current grid preparation conditions that the complex is under excessive stress resulting in partial denaturation and partial-to-complete dissociation. Given the clear biophysical data presented on the intactness of the complex in solution, this disruption likely occurs during grid preparation and further optimization of grid conditions may resolve this issue. With the current maps more work needs to be done to improve the map-to-model correlation and to clearly indicate the regions in the models where this correlation is low.

Additional reconstruction of complex I solubilized in LMNG help us to exclude the interaction of the complex with water-air interface and its reconstitution into lipid nanodiscs as the causes of the relative subunit rotation and high flexibility between the arms. At this moment, whether the structure represents an artifact of purification or is a biologically-relevant state remains an open question. However, answering it goes beyond the current study and will require additional research. This is now explained in the discussion section.

Author Response:

Reviewer #1 (Public Review):

[...] The major limitation of the manuscript lies in the framing and interpretation of the results, and therefore the evaluation of novelty. Authors claim for an important and unique role of beliefs-of-other-pain in altruistic behavior and empathy for pain. The problem is that these experiments mainly show that behaviors sometimes associated with empathy-for-pain can be cognitively modulated by changing prior beliefs. To support the notion that effects are indeed relating to pain processing generally or empathy for pain specifically, a similar manipulation, done for instance on beliefs about the happiness of others, before recording behavioural estimation of other people's happiness, should have been performed. If such a belief-about-something-else-than-pain would have led to similar results, in terms of behavioural outcome and in terms of TPJ and MFG recapitulating the pattern of behavioral responses, we would know that the results reflect changes of beliefs more generally. Only if the results are specific to a pain-empathy task, would there be evidence to associate the results to pain specifically. But even then, it would remain unclear whether the effects truly relate to empathy for pain, or whether they may reflect other routes of processing pain.

We thank Reviewer #1's for these comments/suggestions regarding the specificity of belief effects on brain activity involved in empathy for pain. Our paper reported 6 behavioral/EEG/fMRI experiments that tested effects of beliefs of others’ pain on empathy and monetary donation (an empathy-related altruistic behavior). We showed not only behavioral but also neuroimaging results that consistently support the hypothesis of the functional role of beliefs of others' pain in modulations of empathy (based on both subjective and objective measures as clarified in the revision) and altruistic behavior. We agree with Reviewer 1# that it is important to address whether the belief effect is specific to neural underpinnings of empathy for pain or is general for neural responses to various facial expressions such as happy, as suggested by Reviewer #1. To address this issue, we conducted an additional EEG experiment (which can be done in a limited time in the current situation), as suggested by Reviewer #1. This new EEG experiment tested (1) whether beliefs of authenticity of others’ happiness influence brain responses to perceived happy expressions; (2) whether beliefs of happiness modulate neural responses to happy expressions in the P2 time window as that characterized effects of beliefs of pain on ERPs.

Our behavioral results in this experiment (as Supplementary Experiment 1 reported in the revision) showed that the participants reported less feelings of happiness when viewing actors who simulate others' smiling compared to when viewing awardees who smile due to winning awards (see the figure below). Our ERP results in Supplementary Experiment 1 further showed that lack of beliefs of authenticity of others’ happiness (e.g., actors simulate others' happy expressions vs. awardees smile and show happy expressions due to winning an award) reduced the amplitudes of a long-latency positive component (i.e., P570) over the frontal region in response to happy expressions. These findings suggest that (1) there are possibly general belief effects on subjective feelings and brain activities in response to facial expressions; (2) beliefs of others' pain or happiness affect neural responses to facial expressions in different time windows after face onset; (3) modulations of the P2 amplitude by beliefs of pain may not be generalized to belief effects on neural responses to any emotional states of others. We reported the results of this new ERP experiment in the revision as Supplementary Experiment 1 and also discussed the issue of specificity of modulations of empathic neural responses by beliefs of others' pain in the revised Discussion (page 49-50).

*Supplementary Experiment Figure 1. EEG results of Supplementary Experiment 1. (a) Mean rating scores of happy intensity related to happy and neutral expressions of faces with awardee or actor/actress identities. (b) ERPs to faces with awardee or actor/actress identities at the frontal electrodes. The voltage topography shows the scalp distribution of the P570 amplitude with the maximum over the central/parietal region. (c) Mean differential P570 amplitudes to happy versus neutral expressions of faces with awardee or actor/actress identities. The voltage topographies illustrate the scalp distribution of the P570 difference waves to happy (vs. neutral) expressions of faces with awardee or actor/actress identities, respectively. Shown are group means (large dots), standard deviation (bars), measures of each individual participant (small dots), and distribution (violin shape) in (a) and (c).*

In the revised Introduction we cited additional literatures to explain the concept of empathy, behavioral and neuroimaging measures of empathy, and how, similar to previous research, we studied empathy for others' pain using subjective (self reports) and objective (brain responses) estimation of empathy (page 6-7). In particular, we mentioned that subjective estimation of empathy for pain depends on collection of self-reports of others' pain and ones' own painful feelings when viewing others' suffering. Objective estimation of empathy for pain relies on recording of brain activities (using fMRI, EEG, etc.) that differentially respond to painful or non-painful stimuli applied to others. fMRI studies revealed greater activations in the ACC, AI, and sensorimotor cortices in response to painful or non-painful stimuli applied to others. EEG studies showed that event-related potentials (ERPs) in response to perceived painful stimulations applied to others' body parts elicited neural responses that differentiated between painful and neutral stimuli over the frontal region as early as 140 ms after stimulus onset (Fan and Han, 2008; see Coll, 2018 for review). Moreover, the mean ERP amplitudes at 140–180 ms predicted subjective reports of others' pain and ones' own unpleasantness. Particularly related to the current study, previous research showed that pain compared to neutral expressions increased the amplitude of the frontal P2 component at 128–188 ms after stimulus onset (Sheng and Han, 2012; Sheng et al., 2013; 2016; Han et al., 2016; Li and Han, 2019) and the P2 amplitudes in response to others' pain expressions positively predicted subjective feelings of own unpleasantness induced by others' pain and self-report of one's own empathy traits (e.g., Sheng and Han, 2012). These brain imaging findings indicate that brain responses to others' pain can (1) differentiate others' painful or non-painful emotional states to support understanding of others' pain and (2) predict subjective feelings of others' pain and one's own unpleasantness induced by others' pain to support sharing of others' painful feelings. These findings provide effective subjective and objective measures of empathy that were used in the current study to investigate neural mechanisms underlying modulation of empathy and altruism by beliefs of others’ pain.

In addition, we took Reviewer #1’s suggestion for VPS analyses which examined specifically how neural activities in the empathy-related regions identified in the previous research (Krishnan et al., 2016, eLife) were modulated by beliefs of others’ pain. The results (page 40) provide further evidence for our hypothesis. We also reported new results of RSA analyses(page 39) that activities in the brain regions supporting affective sharing (e.g., insula), sensorimotor resonance (e.g., post-central gyrus), and emotion regulation (e.g., lateral frontal cortex) provide intermediate mechanisms underlying modulations of subjective feelings of others' pain intensity due to lack of BOP. We believe that, putting all these results together, our paper provides consistent evidence that empathy and altruistic behavior are modulated by BOP.

Reviewer #2 (Public Review):

[...] 1. In laying out their hypotheses, the authors write, "The current work tested the hypothesis that BOP provides a fundamental cognitive basis of empathy and altruistic behavior by modulating brain activity in response to others' pain. Specifically, we tested predictions that weakening BOP inhibits altruistic behavior by decreasing empathy and its underlying brain activity whereas enhancing BOP may produce opposite effects on empathy and altruistic behavior." While I'm a little dubious regarding the enhancement effects (see below), a supporting assumption here seems to be that at baseline, we expect that painful expressions reflect real pain experience. To that end, it might be helpful to ground some of the introduction in what we know about the perception of painful expressions (e.g., how rapidly/automatically is pain detected, do we preferentially attend to pain vs. other emotions, etc.).

Thanks for this suggestion! We included additional details about previous findings related to processes of painful expressions in the revised Introduction (page 7-8). Specifically, we introduced fMRI and ERP studies of pain expressions that revealed structures and temporal procedure of neural responses to others' pain (vs. neutral) expressions. Moreover, neural responses to others' pain (vs. neutral) expressions were associated with self-report of others' feelings, indicating functional roles of pain-expression induced brain activities in empathy for pain.

1. For me, the key takeaway from this manuscript was that our assessment of and response to painful expressions is contextually-sensitive - specifically, to information reflecting whether or not targets are actually in pain. As the authors state it, "Our behavioral and neuroimaging results revealed critical functional roles of BOP in modulations of the perception-emotion-behavior reactivity by showing how BOP predicted and affected empathy/empathic brain activity and monetary donations. Our findings provide evidence that BOP constitutes a fundamental cognitive basis for empathy and altruistic behavior in humans." In other words, pain might be an incredibly socially salient signal, but it's still easily overridden from the top down provided relevant contextual information - you won't empathize with something that isn't there. While I think this hypothesis is well-supported by the data, it's also backed by a pretty healthy literature on contextual influences on pain judgments (including in clinical contexts) that I think the authors might want to consider referencing (here are just a few that come to mind: Craig et al., 2010; Twigg et al., 2015; Nicolardi et al., 2020; Martel et al., 2008; Riva et al., 2015; Hampton et al., 2018; Prkachin & Rocha, 2010; Cui et al., 2016).

Thanks for this great suggestion! Accordingly, we included an additional paragraph in the revised Discussion regarding how social contexts influence empathy and cited the studies mentioned here (page 46-47).

1. I had a few questions regarding the stimuli the authors used across these experiments. First, just to confirm, these targets were posing (e.g., not experiencing) pain, correct? Second, the authors refer to counterbalancing assignment of these stimuli to condition within the various experiments. Was target gender balanced across groups in this counterbalancing scheme? (e.g., in Experiment 1, if 8 targets were revealed to be actors/actresses in Round 2, were 4 female and 4 male?) Third, were these stimuli selected at random from a larger set, or based on specific criteria (e.g., normed ratings of intensity, believability, specificity of expression, etc.?) If so, it would be helpful to provide these details for each experiment.

We'd be happy to clarify these questions. First, photos of faces with pain or neutral expressions were adopted from the previous work (Sheng and Han, 2012). Photos were taken from models who were posing but not experience pain. These photos were taken and selected based on explicit criteria of painful expressions (i.e., brow lowering, orbit tightening, and raising of the upper lip; Prkachin, 1992). In addition, the models' facial expressions were validated in independent samples of participants (see Sheng and Han, 2012). Second, target gender was also balanced across groups in this counterbalancing scheme. We also analyzed empathy rating score and monetary donations related to male and female target faces and did not find any significant gender effect (see our response to Point 5 below). Third, because the face stimuli were adopted from the previous work and the models' facial expressions were validated in independent samples of participants regarding specificity of expression, pain intensity, etc (Sheng and Han, 2012), we did not repeat these validation in our participants. Most importantly, we counterbalanced the stimuli in different conditions so that the stimuli in different conditions (e.g., patient vs. actor/actress conditions) were the same across the participants in each experiment. The design like this excluded any potential confound arising from the stimuli themselves.

1. The nature of the charitable donation (particularly in Experiment 1) could be clarified. I couldn't tell if the same charity was being referenced in Rounds 1 and 2, and if there were multiple charities in Round 2 (one for the patients and one for the actors).

Thanks for this comment! Yes, indeed, in both Rounds 1 and 2, the participants were informed that the amount of one of their decisions would be selected randomly and donated to one of the patients through the same charity organization (we clarified these in the revised Method section, page 55-56). We made clear in the revision that after we finished all the experiments of this study, the total amount of the participants' donations were subject to a charity organization to help patients who suffer from the same disease after the study.

1. I'm also having a hard time understanding the authors' prediction that targets revealed to truly be patients in the 2nd round will be associated with enhanced BOP/altruism/etc. (as they state it: "By contrast, reconfirming patient identities enhanced the coupling between perceived pain expressions of faces and the painful emotional states of face owners and thus increased BOP.") They aren't in any additional pain than they were before, and at the outset of the task, there was no reason to believe that they weren't suffering from this painful condition - therefore I don't see why a second mention of their pain status should increase empathy/giving/etc. It seems likely that this is a contrast effect driven by the actor/actress targets. See the Recommendations for the Authors for specific suggestions regarding potential control experiments. (I'll note that the enhancement effect in Experiment 2 seems more sensible - here, the participant learns that treatment was ineffective, which may be painful in and of itself.)

Thanks for comments on this important point! Indeed, our results showed that reassuring patient identities in Experiment 1 or by noting the failure of medical treatment related to target faces in Experiment 2 increased rating scores of others' pain and own unpleasantness and prompted more monetary donations to target faces. The increased empathy rating scores and monetary donations might be due to that repeatedly confirming patient identity or knowing the failure of medical treatment increased the belief of authenticity of targets' pain and thus enhanced empathy. However, repeatedly confirming patient identity or knowing the failure of medical treatment might activate other emotional responses to target faces such as pity or helplessness, which might also influence altruistic decisions. We agree with Reviewer #2 that, although our subjective estimation of empathy in Exp. 1 and 2 suggested enhanced empathy in the 2nd_round test, there are alternative interpretations of the results and these should be clarified in future work. We clarified these points in the revised Discussion (page 41-42).

1. I noted that in the Methods for Experiment 3, the authors stated "We recruited only male participants to exclude potential effects of gender difference in empathic neural responses." This approach continues through the rest of the studies. This raises a few questions. Are there gender differences in the first two studies (which recruited both male and female participants)? Moreover, are the authors not concerned about target gender effects? (Since, as far as I can tell, all studies use both male and female targets, which would mean that in Experiments 3 and on, half the targets are same-gender as the participants and the other half are other-gender.) Other work suggests that there are indeed effects of target gender on the recognition of painful expressions (Riva et al., 2011).

Thanks for raising this interesting question! Therefore, we reanalyzed data in Exp. 1 by including participants' gender or face gender as an independent variable. The three-way ANOVAs of pain intensity scores and amounts of monetary donations with Face Gender (female vs. male targets) × Test Phase (1st vs. 2nd_round) × Belief Change (patient-identity change vs. patient-identity repetition) did not show any significant three-way interaction (F(1,59) = 0.432 and 0.436, p = 0.514 and 0.512, ηp2 = 0.007 and 0.007, 90% CI = (0, 0.079) and (0, 0.079), indicating that face gender do not influence the results (see the figure below). Similarly, the three-way ANOVAs with Participant Gender (female vs. male participants) × Test Phase × Belief Change did not show any significant three-way interaction (F(1,58) = 0.121 and 1.586, p = 0.729 and 0.213, ηp2 = 0.002 and 0.027, 90% CI = (0, 0.055) and (0, 0.124), indicating no reliable difference in empathy and donation between men and women. It seems that the measures of empathy and altruistic behavior in our study were not sensitive to gender of empathy targets and participants' sexes.

Figure legend: (a) Scores of pain intensity and amount of monetary donations are reported separately for male and female target faces. (b) Scores of pain intensity and amount of monetary donations are reported separately for male and female participants.

1. I was a little unclear on the motivation for Experiment 4. The authors state "If BOP rather than other processes was necessary for the modulation of empathic neural responses in Experiment 3, the same manipulation procedure to assign different face identities that do not change BOP should change the P2 amplitudes in response to pain expressions." What "other processes" are they referring to? As far as I could tell, the upshot of this study was just to demonstrate that differences in empathy for pain were not a mere consequence of assignment to social groups (e.g., the groups must have some relevance for pain experience). While the data are clear and as predicted, I'm not sure this was an alternate hypothesis that I would have suggested or that needs disconfirming.

Thanks for this comment! We feel sorry for not being able to make clear the research question in Exp. 4. In the revised Results section (page 27-28) we clarified that the learning and EEG recording procedures in Experiment 3 consisted of multiple processes, including learning, memory, identity recognition, assignment to social groups, etc. The results of Experiment 3 left an open question of whether these processes, even without BOP changes induced through these processes, would be sufficient to result in modulation of the P2 amplitude in response to pain (vs. neutral) expressions of faces with different identities. In Experiment 4 we addressed this issue using the same learning and identity recognition procedures as those in Experiment 3 except that the participants in Experiment 4 had to learn and recognize identities of faces of two baseball teams and that there is no prior difference in BOP associated with faces of beliefs of the two baseball teams. If the processes involved in the learn and reorganization procedures rather than the difference in BOP were sufficient for modulation of the P2 amplitude in response to pain (vs. neutral) expressions of faces, we would expect similar P2 modulations in Experiments 4 and 3. Otherwise, the difference in BOP produced during the learning procedure was necessary for the modulation of empathic neural responses, we would not expect modulations of the P2 amplitude in response to pain (vs. neutral) expressions in Experiment 4. We believe that the goal and rationale of Exp. 4 are clear now.

Reviewer #2 (Public Review):

The authors performed six experiments examining the influence of beliefs regarding pain experience on behavioral and neural indices of empathy for pain and altruistic behavior. They demonstrate that manipulations that to reduce beliefs that individuals making painful expressions are actually in pain (e.g., revealing them to be actors, indicating that their treatment has been successful, etc.) attenuates subjective judgments of pain intensity, real monetary donations to these targets, and P2 amplitudes, and further, that regions involved in perspective-taking and emotion regulation are sensitive to representations of pain beliefs. While I think that the authors have done an admirable job in laying out the evidence for their argument across six well-devised experiments, I do think that the manuscript has some room for improvement. In particular, I hope that the authors can offer stronger grounding in the background literature and clarify some task and stimulus details.

1. In laying out their hypotheses, the authors write, "The current work tested the hypothesis that BOP provides a fundamental cognitive basis of empathy and altruistic behavior by modulating brain activity in response to others' pain. Specifically, we tested predictions that weakening BOP inhibits altruistic behavior by decreasing empathy and its underlying brain activity whereas enhancing BOP may produce opposite effects on empathy and altruistic behavior." While I'm a little dubious regarding the enhancement effects (see below), a supporting assumption here seems to be that at baseline, we expect that painful expressions reflect real pain experience. To that end, it might be helpful to ground some of the introduction in what we know about the perception of painful expressions (e.g., how rapidly/automatically is pain detected, do we preferentially attend to pain vs. other emotions, etc.).

2. For me, the key takeaway from this manuscript was that our assessment of and response to painful expressions is contextually-sensitive - specifically, to information reflecting whether or not targets are actually in pain. As the authors state it, "Our behavioral and neuroimaging results revealed critical functional roles of BOP in modulations of the perception-emotion-behavior reactivity by showing how BOP predicted and affected empathy/empathic brain activity and monetary donations. Our findings provide evidence that BOP constitutes a fundamental cognitive basis for empathy and altruistic behavior in humans." In other words, pain might be an incredibly socially salient signal, but it's still easily overridden from the top down provided relevant contextual information - you won't empathize with something that isn't there. While I think this hypothesis is well-supported by the data, it's also backed by a pretty healthy literature on contextual influences on pain judgments (including in clinical contexts) that I think the authors might want to consider referencing (here are just a few that come to mind: Craig et al., 2010; Twigg et al., 2015; Nicolardi et al., 2020; Martel et al., 2008; Riva et al., 2015; Hampton et al., 2018; Prkachin & Rocha, 2010; Cui et al., 2016).

3. I had a few questions regarding the stimuli the authors used across these experiments. First, just to confirm, these targets were posing (e.g., not experiencing) pain, correct? Second, the authors refer to counterbalancing assignment of these stimuli to condition within the various experiments. Was target gender balanced across groups in this counterbalancing scheme? (e.g., in Experiment 1, if 8 targets were revealed to be actors/actresses in Round 2, were 4 female and 4 male?) Third, were these stimuli selected at random from a larger set, or based on specific criteria (e.g., normed ratings of intensity, believability, specificity of expression, etc.?) If so, it would be helpful to provide these details for each experiment.

4. The nature of the charitable donation (particularly in Experiment 1) could be clarified. I couldn't tell if the same charity was being referenced in Rounds 1 and 2, and if there were multiple charities in Round 2 (one for the patients and one for the actors).

5. I'm also having a hard time understanding the authors' prediction that targets revealed to truly be patients in the 2nd round will be associated with enhanced BOP/altruism/etc. (as they state it: "By contrast, reconfirming patient identities enhanced the coupling between perceived pain expressions of faces and the painful emotional states of face owners and thus increased BOP.") They aren't in any additional pain than they were before, and at the outset of the task, there was no reason to believe that they weren't suffering from this painful condition - therefore I don't see why a second mention of their pain status should *increase* empathy/giving/etc. It seems likely that this is a contrast effect driven by the actor/actress targets. See the Recommendations for the Authors for specific suggestions regarding potential control experiments. (I'll note that the enhancement effect in Experiment 2 seems more sensible - here, the participant learns that treatment was ineffective, which may be painful in and of itself.)

6. I noted that in the Methods for Experiment 3, the authors stated "We recruited only male participants to exclude potential effects of gender difference in empathic neural responses." This approach continues through the rest of the studies. This raises a few questions. Are there gender differences in the first two studies (which recruited both male and female participants)? Moreover, are the authors not concerned about *target* gender effects? (Since, as far as I can tell, all studies use both male and female targets, which would mean that in Experiments 3 and on, half the targets are same-gender as the participants and the other half are other-gender.) Other work suggests that there are indeed effects of target gender on the recognition of painful expressions (Riva et al., 2011).

7. I was a little unclear on the motivation for Experiment 4. The authors state "If BOP rather than other processes was necessary for the modulation of empathic neural responses in Experiment 3, the same manipulation procedure to assign different face identities that do not change BOP should change the P2 amplitudes in response to pain expressions." What "other processes" are they referring to? As far as I could tell, the upshot of this study was just to demonstrate that differences in empathy for pain were not a mere consequence of assignment to social groups (e.g., the groups must have some relevance for pain experience). While the data are clear and as predicted, I'm not sure this was an alternate hypothesis that I would have suggested or that needs disconfirming.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Reviewer 1

This paper proposes a noise-aware approach SCRaPL for modelling the associations of single cell multi-omic data. For gene expression, it uses Poisson-lognormal model. For DNAm data, it uses Binomial noise model which explicitly takes into account the average within the region. The Bayesian hierarchical framework employed by SCRaPL could achieve higher sensitivity and better robustness in identifying correlations, and also offer a template for the application of more complex analysis techniques to multi-omics data. The symbols of this paper are a little bit confusing, and I suggest authors to carefully check them.

We thank the reviewer for his/ her appreciation, and apologise for the confusion arising from the dense notation, which we will thoroughly revise.

1. The symbols used in this paper are messy. For example, "1" and "2" are subscripts in Eq.(2) but become superscripts in Figure 5. Besides, there are many symbols not explained such as mj, Hj, Ψ0, etc. Also, I don't know if x{j,i}^{(1)} , x{j,i}^{(2)} in Figure 5 are same with x{ij1} and x{ij2} in Eq.(3). There are many places mismatch, authors should check carefully.

1. Why the equations in Fig.5 are totally different with Section 4.2? For example, pj ∼Beta(αj ,βj ) in Fig.5 but ρj ∼ Beta−1,1 in Eq.(8).

We apologise for the notational confusion, this will be fully revised.

The paper involves a lot of hyper-parameters which doesn't demonstrate their selection. For example, c1, c2, d1, d2.

This is a good point. We will include a sensitivity analysis on the hyperparameters, justifying the choices on both simulated and real data.

In section4.8, I am confused about $ρ_j$ the experiment 2, 5, 8, 11. Why $ρ_j$ both represents ZI rate and correlation?

We apologise for the notational oversight, which will be rectified.

In Section 4.5, it is difficult to understand the sentence "for me threshold u". Besides, what is $r$ represent in Section 4.5?

We apologise for the confusing sentence. $r$ is the Pearson correlation coefficient, as explained at the start of 4.5

Why there is "(6a)Agreement between SCRaPL and Pearson" in Fig. 4?

This simply means that the panel shows a methylation/ expression scatterplot for a gene where estimation by SCRaPL and Pearson return both a significant association. We will expand the caption to explain further.

For Fig.1, I cannot see the text in the rectangle.

Apologies, we will improve the readability of the figures

I would like to see the efficiency analysis for SCRaPL.

As part of part of re-implementation in a more accessible programming language, we have preformed preliminary efficiency analysis for MCMC , demonstrating linear scalability. Results will appear in the revised manuscript.

Reviewer 2

The authors present a Bayesian model to determine noise-corrected correlation coefficients for gene expression (RNA) and DNA-methylation data at single-cell resolution. The authors present a series of simulation data and an example of matched multi-omics data, and compare their results with Pearson correlation. Noise modelling allows the model to determine gene-methylation correlation patterns more accurately. While the authors demonstrate a neat application on accurate quantification of correlation coefficients, I see a limited use of the model for the broader single-cell community. The authors may therefore improve their manuscript on several aspects.

We thank the reviewer for the encouraging words, and thank him/ her for the critical observations, which we have taken at heart, considerably broadening the scope of our paper to make it more attractive to a larger community.

- Abstract: please specify the omics layers that you are analyzing (RNA + DNA methylation) in the abstract

We acknowledge that, while SCRaPL is potentially general, in the first submission we focused only on RNA and DNA methylation. We have now decided to expand our analyses to include 10X data of simultaneous chromatin accessibility (ATAC-seq) and RNA.

- What is the benefit of using a Bayesian model formulation in this setting?

The benefit is twofold: a principled treatment of noise, and a quantification of the resulting uncertainty which allows for a meaningful way to compute Bayesian significance levels. We will expand the discussion of the relative merits of a Bayesian vs frequentist approach.

- Does it also apply to unmatched data?

In principle, given measurements with the same number of cells in all modalities, it is possible to apply SCRaPL. However, unless there is a natural pairing between different cells, the scaling of this approach will be quadratic in the number of cells, hence potentially expensive (although largely parallelizable). We will discuss this now, particularly in the light of applying SCRaPL in conjunction with other suites such as Seurat.

• Would SCRaPL allow for differential correlation testing?

At the moment, SCRaPL does not allow for differential correlation testing. Of course, one may run SCRaPL separately on two groups of cells and compare the resulting estimates, which would be informative. Nevertheless, extending SCRaPL to perform differential correlation testing (e.g. using Bayesian model selection) would be a non-trivial effort. We will add a comment on this issue to the discussion section.

• Figure 1: The graphical description of the model is rudimentary. I believe that the model description could profit from a graphical model representation of SCRaPL (as presented in figure 5).

We will redraw Fig. 1 and incorporate the graphical model from Fig 5.

- Simulated data: all experiments seem to have rather low cell numbers (max. 200) and genes (max. 300). Given that 10X Genomics is the most widely-used sequencing platform with approx. 10,000 cells and 3,000 (highly variable) genes per experiment, and given that the authors show a use-case with 9480 genes in 487 cells, it seems appropriate to extend the simulations and runtime estimates of the presented model to several thousands of cells and genes, respectively.

Thank you for this comment. The original simulation settings were designed with scMT data in mind, where indeed only a few hundred cells can be assayed at most. Partly because of this feedback, and also because of the request of implementing SCRaPL in a different language, we are working on a more scalable Tensorflow implementation which will be able to handle thousands of cells and genes in a matter of tens of minutes . The new simulated data will therefore extend into this regime with larger data sets.

- Figure 4: Please revise the figure legend as I did not understand the plotted results based on the description.

We will do so.

- Results section 2.5: Please formulate your whole argument about epigenetic regulators. I do not think that "For further information please refer to supplementary figure XYZ." Is an appropriate closing statement for a paragraph, nor does it motivate the reader to look at the supplementary figures (I did look at them and I do not see how they support the point made in the paragraph). Please elaborate and consider a "take home message" for the paragraph such that the reader is able to understand the benefit of SCRaPL without revisiting the original data publication.

Thank you for this pointer, we will take it on board in the full revision.

- Conclusion: The authors mention that SCRaPL would further offer a "template for the application of more complex analysis techniques (such as clustering, dimensionality reduction and network inference)". If that was the case, the authors should consider a comparison to other tools, which offer exactly that (e.g. Seurat's CCA or non-negative matrix factorization in LIGER). Further, the authors should set their work into context with tools like bindSC.

Thank you for the suggestion. As far as we can tell, all of these methods are thought for unmatched data, rather than multi-omics assays performed in the same cells. Having said that, it is in principle possible to “preprocess” data with SCRaPL and then feed to Seurat or other tools the latent means computed by SCRaPL. We will include an example of how this may be done in the revision.

- Implementation: Matlab is used in about 6% of the single-cell RNAseq tools (according to scrna-tools.org). To reach a larger scientific community, do the authors plan to provide an R or Python implementation of their model?

We are now implementing SCRaPL in Python using Tensorflow probability, hoping to achieve substantial speedups (see response to previous point).

Additional minor points about formatting by Reviewer 2 will all be addressed.

Reviewer 3

Maniatis et al propose a sound strategy to analyse single-cell multi-comic data sets. A key advance is to use bespoke error models for each of the omics data. These are integrated into a multivariate gaussian model. This method is a novel and, in my opinion, a valuable addition to the analyses of the growing multi-omics single-cell data sets.

We thank this reviewer for his/ her appreciation of our work.

- Authors make a convincing argument of the importance of principle methods and in particular to use noise models that tailored to the data at hand. To further support this, can authors elaborate on how results would be different from using commonly applied methods ? Eg those embedded in the Seurat, OSCA, and scanpy 'suites'? Authors compare to Pearson correlation-based methods but is not clear if that is the true state-of-the-art on those methods

As far as we know, volcano plots of p-value versus Pearson correlation are the most commonly employed approaches to assess correlations amongst different molecular modalities in single-cell multi-omics (see e.g. Argelaguet et al, Nature 2020). Seurat and other methods normally do not deal with single-cell multi-omics (i.e., multiple omics measured in the same cell), rather with multiple single-cell omics (different molecular modalities assayed in different cells). Nevertheless, it is possible to pre-apply SCRaPL to non-matched data and then use another suite; as an illustration, we will perform an analysis on scMT data using SCRaPL followed by Seurat.

- In the case study on mouse embryonic stem cells, authors excluded the chromatic accessibilty. Why not using it to more clearly show the value of the method?

We did use SCRaPL also on chromatin accessibility, however the signal was weaker and we did not include it in the manuscript, we will now present these results as supplementary material.

- It would also be great if authors would use a different single-cell multi-comic data sets, using other dat modalities, e.g. CITE-Seq data. If this not possible, at least they should elaborate on which omics SCRAPL can handle, what would be the noise models for different data types, etc.

We have started analysing a joint scATAC-scRNA- seq data set generated using the new 10X commercial platform, and will add the results of this analysis to the revised manuscript. We will also expand the description of the suitability for different data types.

*- As the authors acknowledge, computational burden is high, which presumably limits scalability. Are authors able to further explore this (scalability on Insilico data)? Or how complex is adopting the variational inference method suggested? I appreciate that the variational inference implementation might be out of the scope of this paper, though.

• It is a pity that the method is in Matlab. Nearly no-one in single-cell omics use Matlab. Our own lab is largely invested in this topic and we do not even have Matlab licenses. I strongly encourage authors to implement their method in e.g. R or python, ideally compatible with the broadly used 'suites' (Seurat, OSCA, and scanpy,...).*

We are addressed these two comments jointly by re-implementing SCRaPL in Tensorflow probability (Python based), which allow us to leverage powerful libraries for variational inference. We hope that this will lead to a substantial increase of scalability, providing the possibility of running on thousands of cells / genes in under one hour (results will appear in ).

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

First of all, we would like to thank the editor and all reviewers for the effort to evaluate our paper in this difficult era of COVID-19.

Reviewer #1

(Significance): Overall, this manuscript is very clear and easy to follow. The manuscript could be improved by making the following changes:

We thank the reviewer for the favorable comment and will revise the manuscript according to the suggestions.

Reviewer #2

(Evidence, reproducibility and clarity): The use of genetics is particularly impressive but the lack of major discoveries dampens the enthusiasm. Additional efforts to mechanistically define wave initiation and wave propagation would significantly improve the impact of the manuscript. Moreover, some of the conclusions are not fully supported by the data and require further experimentation and/or analysis.

We admit that marked redundancy of function among the EGFR ligands and their essential roles in cell growth prevent us from obtaining very clear results. Considering the importance of EGFR ligands in biology, we believe, our observation will give invaluable suggestions to whom wishes to clarify the roles played by EGFR-family protein in other biological contexts.

(Significance): While it is known that ADAM17 is critical to process EGFR ligands, the specific or redundant roles of different ligands remains an open question. The authors find that all ADAM17 ligands contribute to ERK signaling waves but may have specific contributions to other phenotypes. This work would be of interest to the signaling dynamics, epithelial and developmental biology communities.

We thank the reviewer for the favorable comment.

Reviewer #3

(Evidence, reproducibility and clarity): Overall, this study is carried out with a high degree of rigor and technical excellence, with clear reporting of experimental details and replication. The writing and figures are very clear, and there are no obvious technical problems. However, there are some areas in which the strength and clarity of the conclusions could be strengthened by relatively simple experiments.

We thank the reviewer for the favorable comment. We have already performed some of the experiments suggested by the reviewer. As the reviewer might have anticipated, co-culture with the wild type MDCK cells helps mutant cells to survive. We believe we could propose a clearer model in the revised paper.

(Significance): This study definitively establishes the role of 4 EGFR ligands in the generation of ERK activity waves in MDCK cells. While other studies, including some from the senior author's lab, have strongly indicated that EGFR autocrine signaling is important for these waves, this study goes further in comparing the roles of these ligands using knockouts to unambiguously establish the autocrine factors involved. Others who use this common experimental system (MDCK) to study epithelial dynamics will find this study of great interest. A wider audience of those who work on EGFR-mediated signaling will also find the data quite fascinating as an example of the complex relationship between ERK activation and its downstream effects. The technical excellence of the paper will make it a must-read for those in these fields. However, there are some factors that limit the scope of the significance. MDCK cells are an important experimental model system but differ in substantial ways from other epithelial cells, particularly in the expression of EGFR ligands. Because different ligands such as amphiregulin dominate in other systems (as noted by the authors, and PMID 27405981), the ability to extrapolate from these findings to other cell types is somewhat limited. Also, the paper avoids addressing the major question of how ERK waves relate to collective migration rate. From the data presented it is clear that this relationship is complex; for example, bath application of the ligands restores a high migration rate but not ERK waves. Given this lack of a clear relationship it is an understandable decision to leave this question for future work; however this does limit the conclusions that can be drawn from the study.

We completely agree with the reviewer’s view. It is uncertain to what extent the observation with MDCK cells can be generalized to other cell types. We also admit that the conclusion is not very simple because EGFR signaling is required for various cellular functions including cell survival and migration. Even though the gene editing becomes so easy, it is still labor consuming work to knock out many genes in a single cell line with extensive characterization. We believe the data shown in our work will provide a basis for the understanding of EGFR ligands.

Reviewer #1

For Fig 1F, 3 individual experiments should be conducted to confirm results.

We will follow the reviewer’s suggestion and repeat the experiment.

For Fig 1G, could the authors please show the original western blot data in full rather than just the densitometry graphs?

We did not show just for the sake of brevity. We are happy to will include the images as a supplementary data.

The authors should explain the origin/phenotype of MDCK cells for those who are not familiar with the cell line.

We will modify the text according to the reviewer’s suggestion.

The authors should give a future outlook/direction for future experimentation to further confirm redundancy in EGF ligands in the propagation of ERK activation waves.

We will discuss on the redundancy in other cell types based on available NGS data.

Some mention of the use of biosensors in the abstract and introduction is recommended as this is a major part of the experimental work.

We will refer to the biosensors in the abstract and introduction.

Reviewer #2

There are conflicts with some of the conclusions made about ligands. dEGFR cells have basal ERK activity as high as WT which argues against EGF being responsible for basal ERK activity. Further, basal ERK activity was not rescued by restoration of EGF in the 4KO-EGF cells. The authors should address this discrepancy.

We agree that some new questions have arisen from our observations. The discrepancy of the phenotypes between dEGFR cells and dEGF cells is an example. We are currently establishing dEGF cell lines, in which different genomic sequences of the EGF gene were targeted. We have already started to develop these cell lines and will obtain them within a month. The result will provide some clues to answer the questions. However, even if we could not solve the question, we believe, it is worth reporting observations that could not be easily understood, because such questions are often leading to another discovery.

Besides the ones genetically disrupted in this work, other EGFR ligands seem to play functional roles given that dEGFR cells less migration and fewer ERK waves than 4KO cells. The authors could test if other ligands are upregulated in 4KO cells to compensate. On a similar note determining whether ADAM17 deficient cells are more similar to 4KO cells or dEGFR cells could provide some insight.

According to the reviewer’s suggestion, we will conduct qPCR of growth factors in mutant cell lines to see the expression levels of seven EGFR ligands might have changed significantly. At the same time, as the reviewer suggested, we will establish ADAM17 knockout cell lines and compare the phenotype with those of cell lines deficient from EGFR ligand genes.

• The authors propose that Nrg1 is responsible for ERK waves in QKO, 4KO, dEGFR, and 4KO-EGF cells but are limited in testing this due to Nrg1 being essential in 4KO cells. First, Nrg1 should have been deleted in TKO cells to confirm that it is only essential in the absence of the four EGFR ligands. Additionally, Nrg1 could be knocked out in 4KO-EGF cells to demonstrate the claim that EGF-induced ADAM17 cleavage of Nrg1 is responsible for ERK waves.*

We do not think the deletion of Nrg1 in the TKO cells will abolish the ERK activation waves because EREG in TKO cells could transmit the waves. To overcome the problem of cell growth, we will try to obtain 5KO cells by Cre-induced deletion of NRG1 in 5KO-loxP-NRG1 cells, wherein EGF is supplied exogenously. We already had preliminary data suggesting that co-culture with wild type MDCK cells helps 5KO cells grow.

The authors state that ERK activation waves are important for collective migration and seek to understand the roles of each EGFR ligand, but despite measuring migration and properties of ERK activity, there is very little analysis or commentary on the relationship between the two. The ability of HB-EGF to restore migration without ERK waves suggests that waves are not required per se. It is interesting to note that with restoration of ligands, migration is higher than WT but ERK activity is lower.

We refrained from spending much space about the essential role of ERK activation waves in collective cell migration, because several papers have already described this issue.

Probably, we should have spent more space to emphasize that the collective cell migration is comprised of at least two different phenomena. The migration of leader cells and the follower cells. The ERK activation waves are essential for the follower cells but not the leader cells. In 4KO cells, both the leader cell and follower cell migrations are impaired. We showed that GFs expression restore the leader cell migration, but not the follower cells. We will revise the text to include this issue.

It is suggested that the total amount of EGFR ligands may be the primary determinant of migration, but deletion of TGFα alone causes a significant decrease in migration comparable to the DKO cells. TGFα has the lowest expression of the four ligands studied but is the only ligand to have a significant impact on migration in the single knockout context, which disagrees with that conclusion.

Each EGFR ligand has different affinity to EGFR, which makes it difficult to link the mRNA levels directly to the effect of each EGFR ligand. We will modify the discussion to include this argument.

Other:

Fig. S3B needs clarification that the WT (black) and 4KO (green) did not receive a stimulus.

We will follow the reviewer’s advice.

Reviewer #3:

The experiments in Fig. 5 are undertaken with the purpose of assessing whether NRG acts as an additional ligand that mediates the residual ERK waves in 4KO/QKO cells. However, this question is never addressed in the NRG/4KO cells. While it might be challenging due to the proliferative defect, it seems important to attempt this experiment in some way; measuring the ERK waves for these cells would establish whether all of the critical autocrine factors have been identified. Can the proliferation be rescued by application of high amounts of growth factors?

This question is similar to a question raised by reviewer #2.

To overcome the problem of cell growth, we will try to obtain 5KO cells by Cre-induced deletion of NRG1 in 5KO-loxP-NRG1 cells, wherein EGF is supplied exogenously.

The bath exposure to EGFR ligands shown in Fig. S3A is an important experiment, but it is surprising that ERK signaling is not maintained under these conditions. Is this due to depletion of the added ligands, perhaps locally? Or is the intermittent nature of paracrine signaling needed to maintain ERK activity? These possibilities could be distinguished by checking whether the added EGF or the other ligands are depleted after several hours, or by restimulating with a new bolus of ligand after several hours.

We thank the reviewer for this invaluable suggestion. We will conduct the experiments suggested by the reviewer.

The connection between ERK activity and migration is somewhat confusing. It would be helpful to show the dose sensitivity of migration to a MEK or ERK inhibitor. Are other pathways downstream of EGFR such as PI3K involved in the autocrine-mediated migration? This could also be established with the appropriate inhibitors.

We should have spent more space to emphasize that the collective cell migration is comprised of at least two different phenomena. The migration of leader cells and the follower cells. The ERK activation waves are essential for the follower cells but not the leader cells. In 4KO cells, both the leader cell and follower cell migrations are impaired. We showed that GFs expression restore the leader cell migration, but not the follower cells. We will emphasize this issue in the revised manuscript.

Reviewer #1:

Line 47 in Abstract should read "Aiming for" not "Aiming at".

We have corrected the mistake as suggested.

Some in the field call fluorescence lifetime microscopy "FLIM", you could adopt the same wording in your manuscript to attract more readers.

We have included FLIM according to the reviewer’s suggestion.

Reviewer #1 :

Figure 1D, the images should be presented using the same scale for both the EKAREV and EKARrEV constructs so that they can be directly compared.

Because the basal FRET/CFP ratio is significantly different between EKAREV-NLS and EKARrEV-NLS, the changes during mitosis become unclear if we applied the same scale. This figure is prepared to show the reactivity to Cdk1 during mitosis; therefore, we believe the current scale is better for presentation.

The names QKO and 4KO are a bit confusing. Could the authors please change the naming of the knockout cells so that readers understand that QKO and 4KO are two separate cell types? Perhaps instead of 4KO use FKO for "full knockout" or something similar. The 5KO line might also need to be named something else if you change to FKO.

We have discussed this issue with the co-authors, but could not reach a better idea. Instead of changing the names, we will include a detailed explanation for each cell line.

Reviewer #2:

The interpretation of the RA-SOS coculture experiments is confusing. Based on the author's reasoning, I would expect ADAM17 shedding in the RA-SOS cells to trigger signaling at the interface of both WT and 4KO cells but the 4KO should be unable to propagate the wave farther away from the interface. This does not seem to be the case. Do RA-SOS ADAM17KO cells still trigger waves of ERK signaling in the WT cells? Do ADAM17KO cells behave as the 4KO cells in this coculture system?

Probably, the reviewer misunderstood the method. The GF-less 4KO cells were co-cultured with wild type cells harboring the RA-SOS system. We will describe more in detail to avoid misunderstanding.

Finally, the growth curve in Fig. 5B indicates that 5KO-loxP-NRG1-CreERT2 cells are viable for about two days after Cre induction. The authors could perform a confinement release assay of these cells 1-1.5 days after Cre induction to look for further reduction of ERK waves and migration to demonstrate the role of Nrg1.

This experiment may not be necessary. It is clear that NRG1 is required for the survival of 4KO cells. The reason why cells are still alive 1 to 2 days after 4-OHT application is simply because NRG1 protein is remaining. The interpretation of the results would be difficult during the phase of NRG1 reduction.

In Fig. 1G, the normalization of all WT pERK samples to 1 artificially lowers the variance to zero when performing the T-test.

For the comparison of immunoblotting data derived from independent experiments, the signals must be normalized to the control. We believe the use of pERK/ERK of the wild type cells as the control is reasonable for this experiment.

Author Response:

Reviewer #1 (Public Review):

I think that it is important for the authors to consider that for most (if not all) SARS-CoV-2 variants, increased transmissibility of the virus has not been directly demonstrated. While it is clear that numerous variants have emerged and will continue to emerge, the rapid upsurge of cases with a variant may be related to many factors (e.g. host susceptibility due to immunity or genetic factors, virus seeding events, predominant replication in particular age cohorts, ...) that cannot simply be captured as "transmissibility of the virus". Even for B.1.1.7 and D614G mutants, the direct evidence of increased transmissibility in humans is extremely limited if available at all. Most studies erroneously simply take the increasing occurrence of particular lineages or mutations in sequence databases as a measure of increased "transmissibility", which should be avoided, also in the present manuscript. Increased transmissibility can only be derived from field studies where transmission is measured directly.

We thank the reviewer for pointing out that this is a controversial area. We have adjusted the text throughout to accommodate the fact that the published evidence of increased transmissibility/infectivity is not definitive.

On several occasions in the manuscript (e.g. page 3, page 4 L58-59, page 9 in submitted version), the authors seem to suggest that changes that lead to increased "transmission" or binding affinity and changes that lead to immune escape are mutually exclusive. But the opposite might be true. Viruses may escape from antibody-mediated immunity by amino acid substitutions in linear or structural antibody-binding epitopes. However, viruses may also escape from antibody-mediated immunity through altered protein density on virion surfaces (e.g. less Spike) and/or altered affinity, making it harder for antibody to inhibit virus attachment. As an example, increased affinity may facilitate virus replication with less dense Spike protein, allowing more effective antibody escape. Lower affinity but more dense coverage of Spike may reduce accessibility of critical virus parts by antibodies. Several viruses are known to escape from antibody-mediated neutralization through changes in affinity/avidity.

We agree with this point and have modified the text to avoid implying that increased transmissibility and antibody escape are mutually exclusive.

In relation to the previous point, it is important that authors mention some limitations of the present work in the discussion. SARS-CoV-2 virion attachment to cells is not just a matter of spike protein binding and certainly not of a monomeric RBD. Escape from antibodies and effects on affinity are heavily influenced by the entire (trimeric) spike protein, including its N-terminal domains. Such components are not taken into account in the present experimental designs, and this should be discussed, as e.g. the NTD can be important in attachment and antibody-mediated neutralization.

We thank the reviewer for this suggestion. We have added an appropriate caveat to the Discussion.

The authors suggest that the pandemic virus as it spread across the globe initially did not have "optimized" affinity. However, in the first months of the pandemic, there was relatively limited variation in spike protein sequences. The major variants emerged only later and mostly in areas where population immunity was building up. Again, this begs the question whether natural selection is occurring as a consequence of receptor affinity or immune escape?

We thank the reviewer for making this point. However, we do not think it is that surprising that it took a few months for the first Spike variants to be detected, for the following reasons. Firstly, the number of infections would have been relatively low early in the pandemic and SARS-CoV-2 replicates with a comparatively low error rate for an RNA virus. Secondly, the introduction of strict non-pharmacological measures (social-distancing etc), which would have increased the selective pressure on the virus, was somewhat delayed. Thirdly, it would take some time for any variant that emerged by chance to expand sufficiently to be detected by sequencing. While there is evidence suggestive of broader immunity in populations were the Beta and Gamma variants emerged, which we cite, we are not aware of evidence of widespread immunity in populations where the D614G, S477N and Alpha variants first emerged.

Reviewer #2 (Public Review):

Barton and colleagues investigated the effect of common SARS-CoV-2 RBD mutations and two ACE2 mutations on the RBD/ACE2 interaction. They concluded that the N501Y, E484K and S477N increased receptor binding while the K417N/T had the opposite effect. Double and triple mutants were also included. The ACE2 mutations (that are rare in the human population) also increased binding to most RBD mutants. The study is well-performed and written clearly.

The primary conclusions of the manuscript were supported by the results. However, the interpretation was too speculative. In the abstract (lines 14-17), the authors suggest that the 501 and 477 mutations enhance transmission solely based on data on the RBD-ACE2 interaction. It is unknown whether increased affinity to ACE2 is beneficial for transmission. In addition, increased RBD affinity to ACE2 does not mean that the whole spike or virus particle also binds stronger to ACE2. Lastly, increasing ACE2 affinity does not necessarily increase binding to cells (for example S1A binding to sugars or spike abundance can also influence this).

We agree that it would be inappropriate to assume, based on our affinity/kinetic studies alone, that 501 and 477 enhance transmission. That is why the relevant sentence in the abstract starts with the phrase, “Taken together with other studies”. We summarises the evidence from these other studies in the Discussion. We acknowledge that we have not examined the effects of the mutations on binding of the whole Spike protein to ACE2 or viruses to cells, and have added a suitable caveats to the Discussion.

The overall impact on the field will be limited as there is substantial overlap with already published studies. The observation that the N501Y and E484K increase receptor binding while the K417N/T mutations decrease binding was already made prior by Laffeber et al (2021; J Mol Biol). Laffeber et al also investigated double and triple mutants and came to similar conclusions. Liu et al (2021) confirmed that the N501Y increases binding whereas the K417N/T have opposing effects (Liu et al., 2021 mAbs). The observation that the Y501N increases ACE2 affinity has been made by several groups (e.g. Liu et al 2021 Cell research; Starr et al 2020 Cell).

We thank the reviewer for highlighting these addition studies, two of which are very recent. We have now cited these studies.

Starr et all 2020 was a high throughput study in which the affinity measurements were semi-quantitative, and no kinetic analysis was performed. Liu et al (2021) and Laffeber et al (2021) were performed at 25 C and without rigorous controls for mass-transport and protein aggregation. Liu et al (2021) did not report kinetic measurements. Their results are broadly consistent with ours but their affinity and kinetic measurments are ~ 10 fold different. While we accept that some of the measurements of the effects of mutations have been made before, our measurements of affinity and especially kinetics are performed more rigorously than in previous studies and, for the first time, at a physiological temperature (37 C). Thus, the affinity and kinetic data that we have obtained for single and combinations variants are more definitive. As noted in our Discussion there is a wide variation in reported binding affinities and kinetics in previously published studies. We think the comprehensive data that we report here, the same robust method to measure binding properties of all these variants, adds significant value.

Reviewer #3 (Public Review):

[...] 1) The ACE2 receptor exists naturally as a dimeric form and the RBD is a component of the SARS-CoV-2 spike trimer. The assay format here was monomeric RBD binding against monomeric ACE2 throughout this study. While the measurements are indeed carefully executed and under more physiological conditions than many other reported studies, the authors should discuss potential avidity effects, the consequences of mutations on the accessibility of the RBD in VOC versus wildtype, and impact of other domains such as the NTD, in the context of their monomeric ACE2 measurements with isolated RBD here.

We thank the reviewer for raising this issue. We have added a section to the Discussion addressing these important points.

2) As shown in Figure S2, RBD WT, K417N, K417T, KN/EK, KT/EK, and S477N, the ~30kDa monomeric proteins were flanked by additional ~60kDa bands (which correspond to the smaller peaks to the left of the main peaks) some of which bleed through to the main fraction to different extents, whereas RBDs SA, UK1, UK2, BR, and E484K, do not seem to have as much or any of these extra species. Can the authors comment on whether these contaminants are RBD-dimers as observed before (Dai et al. 2020)? If yes, would such dimers affect the affinity and kinetics?

We thank the reviewer for pointing out these larger ~60 kDa bands in some RBD preps. We think that it is unlikely that these are RBD dimers as these are reducing gels. The strictly monophasic kinetics of all RBD preps, also argues against this being an RBD dimer. We have confirmed by densitometry that the larger band comprises less that 5% of the protein in all the preparations. This will have only a minor effect on estimated of RBD concentration. We have added this information to the Figure S2 legend.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

**Summary:**

Provide a short summary of the findings and key conclusions (including methodology and model system(s) where appropriate). Please place your comments about significance in section 2.

In this paper the authors used a targeted approach to identify rare mutations in a cohort of glioma patients. Using this approach they identified a recurrent mutation in the TOP2A gene encoding for Topoisomerase 2A, and suggest that this mutation creates a more effective protein, binding DNA strongly and maybe more enzymatically active. RNAseq analysis of TOP2A WT and TOP2A mut tumor samples suggest different transcription patterns and points to possible splicing defects. The most recurrent variant (E9448Q) is described in depth and some experimental information shows this variant might be a gain-of-function mutation.

**Major comments:**

• Are the key conclusions convincing? The validation of both the methodology and the presence of never described TOP2A variations in HGG is done quite successfully. Interesting evidence about relevance of the most frequent mutation is provided. However, besides having computational and biochemistry assays performed, lack of details about in vitro experiment statistics (no p-values are provided in figures 4 and 5, neither sample size, repetitions) weakens the conclusions claimed by the authors about the properties of the mutated topoisomerase. Ad. In the revised version we provided more details about in vitro experiments, including statistics when is applicable, sample size and a number of repetitions. In the fig. 4 we show the results of two repetitions (so we can’t calculate statistics) but I would like to stress that we tested independently two fragments of the protein and the results were similar, so our conclusion was justified. However, we do agree with the reviewer that a statistical analysis of those biochemical tests is required. We already started to produce a new batch of recombinant proteins and we will add repetitions to reinforce our claims. We will provide statistical analysis details once all experiments are performed. __

• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? Claims about E948Q variant function should be revised. Data is not presented in a convincing way, plus there is ambiguous language used from the results ("We conclude that the E9448Q TOP2A protein is functional, and MIGHT have a higher activity than the WT protein") to the rest of the paper where they strongly support the claims about the TOP2A activity. Ad. We will provide more data on biochemical features of the TOP2A variant to confirm the impact of the E948Q substitution on enzyme activities, which would allow more strong conclusions. This will present our results in more convincing way. A language of the manuscript has been critically revised and modified (see a version with tracked changes).

• Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation. In line with the presented data in the paper, additional experiments that show catalytic changes of the E9448Q variation must be added. It is shown that there are differences in the DNA binding capacity by EMSA compared to the WT form, however, the DNA supercoil relaxation activities is not that different, at least the way the results are presented. The authors suggest that TOP2A mutation is a driver mutation but no validation in vitro of this claim is shown. Can this mutation alone or in combination with e.g. tumor suppressors transform normal cells to cancer cells? Do cell lines expressing this mutation (compared to parental TOP2A wt expressing cells) display increased transcription? Increased invasion? Ad. In the revised version we moderated our conclusions and we do not state that the mutated TOP2A is an oncogenic driver. We suggest this mutation (and possibly other TOP2A mutations, as we analyzed the impact of other variants on the TOP2A protein function) contribute to gliomagenesis. This conclusion is based not only on the changes in biochemical properties, but also on the observation of the impact of the mutation of transcription and patient survival. We expanded the analysis of TOP2A mutations and expression levels on TCGA datasets and those new results support our conclusions about a pathogenic nature of TOP2A overexpression and mutations (the supplementary fig.4). We believe in such situation, there is no rationale to make a classical oncogenic driver experiment.

Due to a rarity of the TOP2A mutations it is impossible to find a patient derived cell line with such defect. We attempted to overexpress TOP2A in glioma cells but apparently there is some autoregulation preventing overexpression of this protein is cells with endogenous TOP2A expression. Therefore, we can’t verify if cell lines expressing this variant (compared to parental TOP2A wt expressing cells) have increased transcription. Moreover, such experiments are costly and require more time investment for substantial experiments

I would like to stress that modeling some events in cell cultures is difficult and we found in GBMs the link between the mutated TOP2A and increased transcription along with decrease of splicing factors expression.

We have attempted to make CRISPR/Cas9 mediated knock-in in glioma cells but without success. This is a difficult and time consuming procedure. Although in principle, we agree on the rationale for such experiment, we think that the current data are consistent and convincing. If reviewers find it necessary we may attempt to create glioma cell lines with TOP2A knock-out and overexpression of the mutated TOP2A gene and study it functionally, but it would require more time.

• Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments. If the authors can complement the already presented in vitro experiments with additional ones supporting their hypothesis, this should be feasible. The authors can use patient derived glioma cells or glioma cell lines manipulated to express either the parental TOP2A wt enzyme or the identified mutated form. __Ad. Due to a rarity of the TOP2A mutations it is impossible to find a patient derived cell line with such defect. Our findings partly relied on frozen historical samples, so it is not possible to develop patient-derived cell lines. As mentioned above, we can create a TOP2A knock-out cell line and overexpress a wild type or mutated version but there is no certainty that TOP2A deficient cells would survive (this is an essential enzyme) and such manipulation would be feasible.__

• Are the data and the methods presented in such a way that they can be reproduced? Yes, the authors provide a quite detailed explanation of the methods implemented to reach each one of the results they are presenting.

• Are the experiments adequately replicated and statistical analysis adequate? No, there is no information about the statistical analysis or number of replicates in any of the in vitro experiments performed. This information should be added to the manuscript.

Ad. In the revised version the requested information was added where was possible and additional repetitions for biochemical experiments are currently in progress.

**Minor comments:**

• Specific experimental issues that are easily addressable.
• Are prior studies referenced appropriately? Yes, authors clearly address the state of the art regarding previous NGS methodologies and let us know the advantages and novelty of their approach.

• Are the text and figures clear and accurate? There are some discrepancies between the strength of the language used in different sections of the paper to refer the conclusions they can infer from the results they are showing. While they are all valid, authors should revise it. Ad. The text of the manuscript has been unified and revised.

• Do you have suggestions that would help the authors improve the presentation of their data and conclusions? First of all, describe the statistical analysis used in every figure, include number of biological and technical replicates. I would also suggest to change the title or the scope of the discussion, there is too much focus on the TOP2A in the introduction, neglecting all the technical NGS work that actually lead to several new variants being described. This may be confusing when it collides with a conclusion that is heavily focused on the first half describing potential implications of at least another 3 proteins where genetic alterations were described. Given the fact there is not much experimental work that shows TOP2A mutations relevance in HGG or strong enough evidence of the variant's function I would suggest to change a bit the scope of the title. Ad. The description of the results and discussion have been revised to include additional data/discussion on technicalities and other finding not related to TOP2A. We performed additional computational analyses of TOP2A expression/mutations in the TCGA datasets. We believe that the planned experiments on genetically modified cell lines would provide additional support for our claims. We think that in the revised version a balance between landscape/NGS content and TOP2A content is well balanced.

Reviewer #1 (Significance (Required)):

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field. The authors describe a methodology that proved to be sensitive and specific enough in order to allow them to detect rare genetic alterations in patient glioma samples. This information could be valuable to describe new driver mutations or infer in genetic pathway alterations that could be potential therapeutic targets. As the authors state at the beginning of the paper, given the poor therapeutical approaches existing for HGG currently, information of this kind could still be highly useful and provide a better outcome to a specific cohort of patients.

On a personal note, I think there is too much speculation about how TOP2A mutations could be interesting from a biological point of view (authors referred to evidence about implications of this mutation in other forms of cancer) but since no experimental validation is provided in glioma cells, it is difficult to conclude that this enzyme gain-of-function mutation could have a relevant role in HGG and thus make these variants a potential therapeutic target. There are no experiments conducted in glioma cells that express TOP2A variants, it would be interesting to see if it has an effect in the migratory/invasive phenotype like described in other cancer types or like it is suggested by analysis of the genetic pathways activated in the HGG patients samples harboring TOP2A mutation. In addition, there is no evidence of the TOP2A mutations possible role as a driver mutation, which is an interesting aspect that could be further explored from both a computational and an experimental approach.

Ad. As mentioned above, there is no glioma cells that express TOP2A variants and we are not convinced that such experiment will be feasible taking into account an essential role of TOP2A. We will attempt to perform experiments with CRISP/Cas9 knock-in cell lines and functional validation, but until now we did not accomplish knock-in in glioma cells. We will try to knock-out the endogenous TOP2A using CRISPR and express a TOP2A WT or E948Q variant from plasmids encoding these proteins, but we can’t predict if TOP2A KO cell would survive. If we manage to produce such cells, then we will investigate proliferation, migration and invasion of cells expressing TOP2A WT or mutated variant.

We do agree with the reviewer that our previous conclusions were too strong, and in the revised version we moderated our claims. We do not say that the mutated TOP2A is an oncogenic driver. We suggest this mutation (and possibly other TOP2A mutations, as we analyzed the impact of other variants on the TOP2A protein structure) contribute to gliomagenesis.

__Data on the Fig. 1A suggests that TOP2A has a mutational hotspot in the position E948Q in our dataset. In the revised version of the manuscript we have extended RNA-seq analysis of our datasets and TCGA PanCancer datasets to search for TOP2A mutations/ overexpression. We found that another computational prediction using CADD algorithm strongly confirms that TOP2A E948Q is in the top 1% of most deleterious variants in the human genome (CADD score >20). This results was added to Supplementary Table 2.__

• Place the work in the context of the existing literature (provide references, where appropriate). The quality of the paper is high and in line with other studies in the literature that perform genome and transcriptome analysis of tumor samples. It is only the experimental validation that is lacking data supporting the "in silico" findings. Ad. We would like to point that we provided the results of experimental, biochemical validation (2 assays) showing that the variant TOP2A proteins have different properties. The associations of transcriptional dysregulation in variant TOP2A bearing gliomas was not a in silico prediction but the result of the analysis of real tumor samples.

As stated above, we are ready to perform further biological validation if the editors find it necessary.

• State what audience might be interested in and influenced by the reported findings. Computational biologists are the right audience to target this paper. If additional experimental work further validating their initial bioinformatic findings is added to the manuscript then probably a wider population could be targeted.

Ad. As stated above, we are working now on providing more replicates of biochemical assays and we are ready to perform further biological validation if the editors find it necessary. I would like to stress that genome editing by knock-in is not always possible/feasible, and these type of experiments is time and money consuming.

• Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate. Brain tumors, immunotherapy, cancer stem cells, tumor microenvironment, tumor heterogeneity. I do not have sufficient expertise to evaluate the bioinformatic analysis and software/programs used to analyze the NGS data.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

By exon targeted resequencing of 664 genes frequently mutated in cancer, authors identify novel mutations associated to Glioma in a cohort of 182 Polish and Canadian samples. Most of these novel mutations have been identified as potential rare germline mutations, somatic mosaicism or loss-of-heterozygosity variants. Among them, authors focus on mutations associated to the TOP2A gene, which encodes one of the two Type II topoisomerases paralogs present in humans. By a limited number of in vitro experiments, authors conclude that TOP2A recurrent variant E948Q, displays increased binding to DNA and topoisomerase activity. Therefore, authors suggest that the TOP2A E948Q variant is a gain-of-function mutation.

**Major comments:**

• Authors show an interesting plethora of new exon mutations associated with High Grade Glioma. Nevertheless, the characterization of TOP2A E948Q variant, which is the main focus of the study, although very interesting and potentially clinically relevant, remains incomplete. Association of the TOP2A E948Q glioma variant with a gain-of-function mutation would require to improve the statistical power of the presented experiments (increase number of replicates). With the existing experimental evidence, the increased DNA binding and activity of the TOP2A E948Q variant should be considered as preliminary, especially in the case of 431-1193 aa fragment. I would consider mandatory to increase experimental replicates and to analyse statistical significance in the case of DNA binding experiments and DNA relaxation assays with the TOP2A 431-1193 aa fragment. A more detailed biochemical characterization should be performed. A titration of different amounts of protein should be included in these experiments, and at least two batches of purified proteins should be analysed. Decatenation assays should also be performed to characterize the activity of the mutant protein in more detail. Recapitulation of DNA binding and activity results with other TOP2A variants obtained in this study will significantly reinforce authors claims too. This improved biochemical characterization should not take longer than two months.

Ad. We would like to stress that while two replicates are presented, we were testing two forms of TOP2A proteins and the results were similar, confirming our conclusions. But we agree that additional replicates would strengthen our claims. Therefore, we are in the process of producing another batch of recombinant proteins to increase a number of replicates and calculate statistics for the biochemical assays (binding and relaxation assay). We will perform titration of different amounts of the protein using two batches of purified proteins.

The occurrence of other TOP2A variants is low (identified in only a single patient sample), therefore we will perform experimental validation only for E948Q. However, we performed additional computational analysis for other TOP2A variants showing the influence of the substitution on DNA binding by docking the DNA fragment into TOP2A binding pocket (Supplementary table 4).

• To increase the significance of the results, I would encourage authors to include experiments showing the functional impact of this TOP2A mutation in cells. The connection with transcriptomic alterations is merely correlative, and would be greatly strengthened by functional experiments in cellular models. To draw definitive conclusions regarding the changes in transcription, I would encourage authors to complement the results with experiments that point to the physiological impact of TOP2A variants within the cell. Overexpression of WT and E948Q variants in a cell model and transcriptomic analysis would be desirable, but validation in these experimental models of some of the target genes identified as deregulated in patients could suffice. These experiments could be accomplished in no more than 3-4 months.

Ad. We agree that the connection of the TOP2A mutation with transcriptomic alterations is correlative, and would be greatly strengthened by functional experiments in cellular models. If we develop a TOP2A E948Q knock-in cell line or TOP2A KO cell line with E948Q over-expression, we are planning to evaluate transcriptomic changes on selected genes by qPCR or whole transcriptome by RNAseq. We estimate that developing a stable CRISPR/Cas9 cell line may take up to 6 months.

We provided additional results showing that the connection of the TOP2A mutation with transcriptomic alterations may be due to different expression of splicing factors (Supplementary Fig. 6).

• Some of the methods are not presented with sufficient detail. Regarding the DNA and RNA sequencing experiments, I consider necessary to specify the DNA fragmentation method, reference for the indexed adapters and ligation and amplification procedures (ligase reference, number of PCR cycles, etc). It would be helpful to clarify or reference which are the "special oligonucleotide probes" that are mentioned. Finally, a reference for the "special beads" and final amplification number of cycles is needed. The sequence of primers used for TOP2A cloning and mutagenesis should be included. The reference for the "site mutagenesis kit" used is missing. When studying the survival rate of glioma patients depending of TOP2A expression levels, it should be clarified what is considered HIGH or LOW expression (i.e: which percentiles are used).

Ad. We expanded the description of methodological aspects of DNA and RNA sequencing experiments. This description was revised and more details are provided in the revised version. Regarding cloning and mutagenesis, we added a table with primer sequences (Supplementary Table 5). We did not use any kit for cloning and mutagenesis. Standard methods and primers with modified nucleotides were used.

__We have included information about the partitioned groups in the survival analyses in the figure 2 caption. “D - Kaplan-Meier overall survival curve for patients with high (> TOP2A mRNA median expression x 1.25) or low (- There is a major concern about how the experiments are replicated and about the statistical analysis, which is inexistent in some cases. Indeed, Figures 4 and 5 do not present any statistical analysis, it is therefore hard to draw any conclusion. In Figure 4b, the results for the 890-996 aa fragment looks qualitatively clear, but this is not the case for the 431-1193 aa fragment. More replicates and statistical analysis are mandatory, together with a protein titration. The replicates should be performed with at least two independent batches of protein purifications. The individual values of each experiment should be included in the graph to provide a better understanding of experimental variability. All this also applies to Figure 5.

Ad. We will increase a number of replicates for the binding and relaxation assay. We will perform a titration of different amounts of protein in these experiments using two batches of purified proteins.

**Minor comments:**

• The effect on transcription of co-occurrence of TOP2A mutations with other mutations could also be analysed with the already available data. Also, a more detailed analysis of genome-wide transcription could also be used to at least partially address the proposed hypotheses of increased transcriptional rate or splicing aberrations.

Ad. We don’t have enough samples with the TOP2A mutation to analyze the effect on transcription of co-occurrence of TOP2A with other mutations.

We addressed the hypothesis of increased transcriptional rate or splicing aberrations by performed additional analyses of RNA-seq data to confirm splicing aberrations. Indeed we found splicing machinery genes down-regulated in the E948Q TOP2A glioma samples (Supplementary Fig.6).

• There is no reference for the following argument "As the identified germline variants were exceptionally rare in the general population ... it is likely that these variants are pathogenic". I also find low number of references to support the suggested high frequency of altered genes in gliomas compare to other cancer types. I miss specific works relating TOP2 activity with transcriptional regulation.

Ad. The appropriate references are provided to back-up these statements.

• At several points in the text there are quantitative and comparative statements that should be backed up by the actual numbers (e.g. "The results of the targeted sequencing indicate a high frequency of altered genes", "The most altered gene was TP53, followed by IDH1...", "Other genes that were found to be frequently altered included KDM6B...", "These partial results combined with a low frequency of this variant in the Polish population suggest a somatic mutation"). The same thing applies to the co-occurrence of mutations, in which the percentage of co-occurrence and significance is not indicated. This lack of detail in the description is also observed in the description of the transcriptomic alterations in which no detail is provided regarding how many of the 105 analyzed samples correspond to low or high gliomas.

Ad. We apologize that the frequencies of mutated genes were not specified. This information is included in the main text of the revised version. We now provide a gnomAD frequency for all variants of interest, confirming the low frequency in the population (AF__ __

Regarding the total number of samples in the transcriptomic analysis, we provided an updated supplementary table covering also samples that were used for transcriptomic analyses (Supplementary Table 1).

• For TOP2A mutation analysis, sometimes is not clear when the analysis is done with the 9 mutated samples and when with the 4 recurrent TOP2A E948Q variants. For example, in figure 2b and 2c analysis are done with 9 samples while the figure 2e is based on the 4 E948Q variants. At least this is what I have deduced from the main text, it should be clarified in the figure legend).

Ad. This information has been included in the captions of Figure 2B, 2C and 2E and now we specify how many samples were used in each analysis.

• Fig1. In figure 1b it would be interesting to color-code patients by glioma grade. This would also apply to Figure S1a, S1c, 2a, S3 and S4. In figure 1D it would be very informative to distinguish mutations that passed the quality control or not with different colors.

Ad. Following reviewer’s suggestions, we have added this information, and oncoplot figures derived from the germline analysis have a distinct color for each glioma grade. In the figure 1D, all of the presented mutations have passed a quality control in terms of quality of sequencing. One additional criterion that was used for all genomic results (except some of the TOP2A variants) was a criterion of 20% variant penetration (20% of reads in the position had to come from the alternative allele). We corrected the description in the Supplementary Table to “passed 20% penetration criterion”. The rationale behind this criterion for TOP2A variants was a fact that for one of the E948Q samples it was ~13% and we didn’t want to lose this sample from the analysis due to rarity of the mutation.

• Fig2. In figure 2b and 2c the statistical significance of differences between TOP2A and the rest of genotypes should be included. Looking at Figures 2d and 2e it looks surprising how similar is the overall survival of HIGH TOP2A mRNA expression (500 days, fig 2d) with the overall survival of the TOP2A WT samples (400 days, fig 2e). Here a I would include a graph that summarizes the TOP2A mRNA expression levels of each group in fig 2d and 2e.

Ad. We agree that median overall survival is similar comparing patients with high TOP2A mRNA expression to TOP2A WT patients in our cohort. It is worth noting, however, that both datasets were produced using different library protocols, and the methodology is different, so it can’t be expected the levels to be equal. We think that adding two more graphs, as suggested, would add another layer of information to this section of the analysis. We have included two boxplots depicting TOP2A mRNA RPKMs, and it is clear now that the medians of High TOP2 mRNA and TOP2A mutant (E948Q) are more closely related, despite the fact that we only have a few patients with the mutation.

• Fig3. It would be interesting to include the same simulation for the rest of TOP2A mutations as supplementary figure.

Ad. We agree that the other TOP2A SNPs could potentially affect DNA binding. We focused on the recurrent mutation and did not analyze those occurring in a single patient. In the revised version we included predictions whether these variants could affect TOP2A DNA binding. For WT TOP2A and variants, we calculated the Gibbs free energy (ΔG). This information can be found in Supplementary Table 4. We have extended description in the Results section: “The TOP2A E948Q substitution may affect protein-DNA interactions”

• Fig4 and Fig5. Include statistical analysis and dots representing individual replicates.

Ad. For Fig 4 we have two replicates for two protein fragments, so we can’t present statistics now. As mentioned above we are preparing a new batch of proteins and will make more repetitions of EMSA and relaxation assays. For Fig 5. we have 3 replicates but despite a trend there is no statistical significance. We intent to make more replicates and a separate protein preparation. After including additional repetitions we will present the results as dots representing individual replicates.

• Fig6. In Figure 6d I would increase the size differences in the dots representing the gene counts, as it is not easily perceived with current parameters.

Ad. The dot size in Fig 6d did not reflect the true meaning. To make it easier to understand, we changed a plot type to a barplot, which now represents the number of differentially expressed genes involved in each pathway.

• FigS2. In figure S2B, it would be informative to establish which dots are significatively above or below the diagonal.

Ad. The purpose of this figure was to show which oncogenic signaling pathways from TCGA cohorts were affected in our cohort. The pathway's size is a variable that is used to normalize the calculation (shown in abscissa axis in S2B). RTK-RAS and NOTCH pathways contain hundreds of genes, whereas other pathways, such as the NRF2 oncogenic pathway, contains only a few. On the other hand, we counted how many genes in each pathway in our cohort were mutated (shown in ordinate axis, S2B). We used logarithms in both axes for visualization purposes, but this has no effect on the enrichment of these pathways, which is shown in the color-coded legend.

• FigS3. How were the samples shown selected from the total?

Ad. In this plot we show only somatic variants that were found in at least two different patients. We apologize that this information was missing, and we have added it to the figure's caption.

• FigS4. I would include a line with the TOP2A mutation to have an idea of how these mutations are distributed between groups.

Ad. Based on the feedback of the reviewer, this figure has been modified and improved. A new row has been added to the figure, displaying TOP2A mutations alongside other highly frequent mutations in other genes.

Reviewer #2 (Significance (Required)):

In this work authors have identified new mutations associated to gliomas by targeted exome sequencing using an important cohort of 182 samples. Among these new mutations epigenetic enzymes and modifiers are found. These results potentially increase the repertoire of putative molecular targets for future cancer therapies. Authors focus in mutations associated to TOP2A gene, that provides stronger DNA binding and DNA relaxation capacity in vitro. Although further characterization is needed, tumours harbouring this kind of mutations could show higher level of sensitivity to TOP2 drugs, providing potentially interesting clinical implications. Although the link between TOP2A expression and cancer prognosis is well established, the relevance of specific mutations in still largely unexplored.

On one hand this work brings novelties in the field of Glioma providing a series of putative new players in the development of this type of cancer. Audience interested in basic or clinical aspects of these tumours would be a good target for this work. On the other hand, this putative gain-of-function mutation of TOP2A represent an interesting aspect for the DNA topology and topoisomerases field. Although, as stated above a more detailed biochemical and functional characterization would be required to draw the attention of this audience-

Scientifically, I have experience in the DNA topology and topoisomerases field, 3D genome organization and gene regulation. I have no experience in Gliomas or any other clinical aspect of cancer, so it is difficult for me to properly establish the potential impact of the newly discovered mutations. Technically I have no capacity to critically evaluate the aspects related to the targeted exome sequencing and the suitability of the analysis performed for mutation identification.

**Referee Cross-commenting**

I fully agree with the comments of the other reviewer, which are perfectly aligned with my own regarding the preliminary nature of the conclusions about the biochemical and functional characterization of the TOP2A mutations.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

Summary:

Provide a short summary of the findings and key conclusions (including methodology and model system(s) where appropriate). Please place your comments about significance in section 2. In this paper the authors used a targeted approach to identify rare mutations in a cohort of glioma patients. Using this approach they identified a recurrent mutation in the TOP2A gene encoding for Topoisomerase 2A, and suggest that this mutation creates a more effective protein, binding DNA strongly and maybe more enzymatically active. RNAseq analysis of TOP2Awt and TOP2Amut tumor samples suggest different transcription patterns and points to possible splicing defects. The most recurrent variant (E9448Q) is described in depth and some experimental information shows this variant might be a gain-of-function mutation.

Major comments:

• Are the key conclusions convincing? The validation of both the methodology and the presence of never described TOP2A variations in HGG is done quite successfully. Interesting evidence about relevance of the most frequent mutation is provided. However, besides having computational and biochemistry assays performed, lack of details about in vitro experiment statistics (no p-values are provided in figures 4 and 5, neither sample size, repetitions) weakens the conclusions claimed by the authors about the properties of the mutated topoisomerase.

• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? Claims about E948Q variant function should be revised. Data is not presented in a convincing way, plus there is ambiguous language used from the results ("We conclude that the E9448Q TOP2A protein is functional, and MIGHT have a higher activity than the WT protein") to the rest of the paper where they strongly support the claims about the TOP2A activity.

• Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation. In line with the presented data in the paper, additional experiments that show catalytic changes of the E9448Q variation must be added. It is shown that there are differences in the DNA binding capacity by EMSA compared to the WT form, however, the DNA supercoil relaxation activities is not that different, at least the way the results are presented. The authors suggest that TOP2A mutation is a driver mutation but no validation in vitro of this claim is shown. Can this mutation alone or in combination with e.g. tumor suppressors transform normal cells to cancer cells? Do cell lines expressing this mutation (compared to parental TOP2A wt expressing cells) display increased transcription? Increased invasion?

• Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments. If the authors can complement the already presented in vitro experiments with additional ones supporting their hypothesis, this should be feasible. The authors can use patient derived glioma cells or glioma cell lines manipulated to express either the parental TOP2A wt enzyme or the identified mutated form.

• Are the data and the methods presented in such a way that they can be reproduced? Yes, the authors provide a quite detailed explanation of the methods implemented to reach each one of the results they are presenting.

• Are the experiments adequately replicated and statistical analysis adequate? No, there is no information about the statistical analysis or number of replicates in any of the in vitro experiments performed. This information should be added to the manuscript.

Minor comments:

• Specific experimental issues that are easily addressable.

• Are prior studies referenced appropriately? Yes, authors clearly address the state of the art regarding previous NGS methodologies and let us know the advantages and novelty of their approach.

• Are the text and figures clear and accurate? There are some discrepancies between the strength of the language used in different sections of the paper to refer the conclusions they can infer from the results they are showing. While they are all valid, authors should revise it.

• Do you have suggestions that would help the authors improve the presentation of their data and conclusions? First of all, describe the statistical analysis used in every figure, include number of biological and technical replicates. I would also suggest to change the title or the scope of the discussion, there is too much focus on the TOP2A in the introduction, neglecting all the technical NGS work that actually lead to several new variants being described. This may be confusing when it collides with a conclusion that is heavily focused on the first half describing potential implications of at least another 3 proteins where genetic alterations were described. Given the fact there is not much experimental work that shows TOP2A mutations relevance in HGG or strong enough evidence of the variant's function I would suggest to change a bit the scope of the title.

Significance

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field. The authors describe a methodology that proved to be sensitive and specific enough in order to allow them to detect rare genetic alterations in patient glioma samples. This information could be valuable to describe new driver mutations or infer in genetic pathway alterations that could be potential therapeutic targets. As the authors state at the beginning of the paper, given the poor therapeutical approaches existing for HGG currently, information of this kind could still be highly useful and provide a better outcome to a specific cohort of patients.

On a personal note, I think there is too much speculation about how TOP2A mutations could be interesting from a biological point of view (authors referred to evidence about implications of this mutation in other forms of cancer) but since no experimental validation is provided in glioma cells, it is difficult to conclude that this enzyme gain-of-function mutation could have a relevant role in HGG and thus make these variants a potential therapeutic target. There are no experiments conducted in glioma cells that express TOP2A variants, it would be interesting to see if it has an effect in the migratory/invasive phenotype like described in other cancer types or like it is suggested by analysis of the genetic pathways activated in the HGG patients samples harboring TOP2A mutation. In addition, there is no evidence of the TOP2A mutations possible role as a driver mutation, which is an interesting aspect that could be further explored from both a computational and an experimental approach.

• Place the work in the context of the existing literature (provide references, where appropriate). The quality of the paper is high and in line with other studies in the literature that perform genome and transcriptome analysis of tumor samples. It is only the experimental validation that is lacking data supporting the "in silico" findings.

• State what audience might be interested in and influenced by the reported findings. Computational biologists are the right audience to target this paper. If additional experimental work further validating their initial bioinformatic findings is added to the manuscript then probably a wider population could be targeted.

• Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate. Brain tumors, immunotherapy, cancer stem cells, tumor microenvironment, tumor heterogeneity. I do not have sufficient expertise to evaluate the bioinformatic analysis and software/programs used to analyze the NGS data.

SciScore for 10.1101/2021.07.21.21260691: (What is this?)

Please note, not all rigor criteria are appropriate for all manuscripts.

Table 1: Rigor

Table 2: Resources

Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

However, this study also had several limitations. First, because we conducted a cross-sectional study, causality could not be determined. However, since it is theoretically unlikely that treatment interruption experienced by an individual will increase the COVID-19 infection rate in a region, we think it is likely that high regional infection rates cause treatment interruption. Second, we did not identify workers’ reasons for discontinuing treatment in this study. As discussed above, there are various possible causes of treatment interruption, which may vary by region. Third, we did not inquire about the diseases being treated. Treatment interruption may vary depending on the presence or absence of symptoms and the potential disadvantages of discontinuing treatment for a particular disease.

Results from TrialIdentifier: No clinical trial numbers were referenced.

Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

Results from JetFighter: We did not find any issues relating to colormaps.

• Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
• Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
• No protocol registration statement was detected.

Results from scite Reference Check: We found no unreliable references.

About SciScore

SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

We are grateful for the constructive and highly supportive reviews provided by our Reviewers. We especially appreciate the efforts they have made to provide suggestions on how to make our revised manuscript even more robust. We have incorporated many of these suggestions into the revised manuscript that will post to Biorxiv and will be submitted to an affiliate journal. We have provided point-by-point responses to each Reviewer below each item (starting with Response: …), along with any changes made in response to that comment/suggestion (starting with In our revised manuscript, …).

Finally, we agree with all Reviewers that this work should be of broad interest to the molecular biology, cell biology, and parasitology communities. Our discovery that Plasmodium and two related genera have taken the unorthodox approach of duplicating their NOT1 protein, and that Plasmodium has dedicated it for its unique transmission strategy, is a fascinating adaptation of the use of this core eukaryotic complex. We believe that those that focus on diverse aspects of RNA biology, including RNA preservation/decay, the maternal to zygotic transition, translational repression, and beyond will find this work to be of interest and relevant to their own research questions.

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

The manuscript „The Plasmodium NOT1-G paralogue acts as an essential nexus for sexual stage maturation and parasite transmission" investigates the two forms of NOT1 in rodent malaria parasites. The authors found out that the original NOT1 is crucial for gametocyte induction as well as transmission to the mosquito, they therefore renamed it NOT1-G. The paralogous proteins, on the other hand, appears to be crucial for intraerythrocytic growth, since it cannot be knocked out. The authors then investigated NOT1-G in more detail, using standard phenotyping assays. They found a slightly increased gametocytemia and a minor effect on transmission to the mosquito.

Response: In our submitted manuscript, we do focus on PyNOT1-G because of the exciting role it has for both sexes of gametocytes, which results in a complete defect in transmission to mosquitoes. Our investigations of what domains of PyNOT1-G focused on the most likely suspect: the putative tristetraprolin-binding domain (TTPbd). It was through deletion of this domain that we observed only a minor defect in the prevalence of infection of mosquitoes, indicating that the portion of PyNOT1-G that is required for transmission lies elsewhere (in part or in total). It is also important to correct Reviewer 1’s statement regarding the other (perhaps canonical) PyNOT1. To our surprise, PyNOT1 could be deleted, but resulted in a parasite that has an extreme fitness cost and a very slow growth phenotype. This is in stark contrast to other eukaryotes, where NOT1 is essential.

Reviewer #1 (Significance (Required)):

If the authors are able to provide convincing data that NOT1-G is indeed important for gametocyte induction and transmission to the mosquito, then the report would be of high significance for the malaria and molecular cell biology fields.

Response**: We have in fact shown this and more in the originally submitted manuscript, and thus we are grateful that Reviewer 1 considers this work to be of high significance in a broad readership (molecular and cell biology, parasitology). In our revised manuscript, we have added text throughout to make these results even more apparent and clear for the reader.

My expertise: molecular cell biology of gametocytes, translational regulation, parasite transmission

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

**Summary**

The manuscript by Hart et al. builds upon a fascinating finding presented in a previous manuscript by the same authors, in which they show that CCR4 seems to be able to associate with two members of the NOT1 family. In this work, the authors first re-annotate the two NOT1 paralogs in Plasmodium yoelii and then perform an in depth characterization of the role of NOT1-G during gametocytogenesis and early mosquito development. Using gene knockout and different genetic crosses, the authors show that NOT1-G is essential for male gametocyte development and leads to an arrest of development in zygotes arising from female gametocytes. Using RNA-seq the authors show that NOT1-G leads to lower transcript abundances, leading to the hypothesis that NOT1-G might be involved in preserving mRNAs in a larger RNA-binding complex. Lastly, the authors characterize a NOT1-G defining TPP domain and find that it is not essential for either male/female phenotype observed for the whole gene KO.

Response**: We appreciate the concise and accurate summary of these findings.

**Major comments:**

• Are the key conclusions convincing?

  The phenotypic characterization of NOT1-G during gametocytogenesis / early mosquito development is nicely presented and the experiments are well performed. Because a duplication of NOT1 with possibly opposing roles of the paralogs is a very unique feature with broad implication on RNA metabolism, it would have been great to see two select experiments on the molecular level adding evidence that 1) NOT1/NOT1-G are mutually exclusive in a complex with CCR4/CAF1 and 2) NOT1-G acts post-transcriptionally in an antagonistic way to NOT1 (i.e. as a mRNA 'stabilizer' as proposed by the authors).

Response**: We agree that inclusion of those two aspects would make for a more complete story about these two NOT1 paralogues.

First, we also think that it is highly likely that NOT1 and NOT1-G are mutually exclusive, as in other eukaryotes NOT1 acts as a scaffold protein upon which effector proteins bind and bridging interactions are made. In our original manuscript, we did not include a mention of our previous attempts to address this question through colocalization and proteomic approaches, as they were largely unsuccessful. Specifically, we generated rabbit polyclonal antisera to PyNOT1-G’s tristetraprolin-binding domain but it did not pass our rigorous quality control (e.g. too much staining persisted in pynot1-g- parasites). Using both asexual and sexual blood stage parasites, we also attempted immunoprecipitation (with and without chemical crosslinking) and proximal labeling approaches via BioID and TurboID but all approaches did not produce rigorous results and thus we did not report them in our original manuscript. However, this question of whether the two NOT1 paralogues were mutually exclusive in complexes was also taken up by the Bozdech Laboratory in their 2020 preprint (Liu et al.) where they were able to capture the P. falciparum NOT1-G and NOT1 proteins (called Not1.1 and Not1.2 in that work). While their proteomic evidence showed that they could capture these bait proteins and that the NOT1 paralogues were not in the same complex, these results should be taken with a grain of salt: all mass spectrometry-based proteomic approaches are limited in that an absence of evidence does not mean that the protein is not present/interacting. Moreover, these efforts only identified a few other proteins that were already known to interact with the CAF1/CCR4/NOT complex, but even so, they did not use statistically rigorous methods in an attempt to quantify these results. In our revised our manuscript, we have included additional text to describe our unsuccessful efforts to do these capture proteomics experiments, and we have expanded our discussion of the Liu et al findings that provide some evidence in support of a mutually exclusive complex.

Second, we also hypothesize that PyNOT1-G acts post-transcriptionally to affect mRNA abundance and translation. However, it is important to emphasize that NOT1 proteins typically act as scaffolds, with the recruited effector proteins acting to hasten the degradation and/or to preserve associated transcripts. We believe that studying these effector proteins is the next important effort to undertake. In fact, we hypothesized that these antagonistic effector proteins would be analogous to TTP and ELAV/HuR-family proteins as are found in other eukaryotes, and that the critical interaction with PyNOT1-G would be via its putative TTP-binding domain. It was for that reason that we interrogated the TTP-binding domain itself, and were surprised that its deletion did not phenocopy the complete gene deletion. Ongoing work will be focused on identifying these antagonistic effector proteins that likely are expressed in a stage-enriched manner, and to define how they interact with PyNOT1-G in order to direct specific mRNAs to their fates. Additionally, it would be very important and exciting to directly test if PyNOT1 and PyNOT1-G are functionally opposed. However, this would be exceptionally challenging to study from a technical standpoint. While we were able to delete the pynot1 gene after many repeated attempts, these parasites are very sickly and grow very slowly. Because of this, we believe that assessing direct versus indirect effects of PyNOT1 in these cells would not be feasible or robust. Given this, comparing functions between PyNOT1 and PyNOT1-G could not be done in a conclusive manner.** In our revised manuscript, we have expanded our descriptions of the mechanisms by which we believe PyNOT1-G and its complex affects mRNA fates. In particular, we have expanded our Discussion section to incorporate the results that indicate that the TTP-binding domain is not required for the essential functions of PyNOT1-G.

• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

  The authors describe the role of NOT1-G as 'preserving' mRNA. The lower abundance of many transcripts in the NOT1-G knockout suggest this, but experimental proof is not provided (see suggestions below). Maybe rephrase to 'putatively preserved/stabilized' or 'has a potentially stabilizing function'. The same is true for the mutually exclusive association of the two paralogs with CCR4/CAF1. The authors refer to a protein co-IP of CCR4 showing that CCR4 can interact with both NOT1 and NOT1-G, but a reciprocal experiment is lacking.

Response**: In our first publication on the deadenylase members of this complex, we also saw a similar effect on specific mRNAs when pyccr4-1 was deleted: the abundance of specific mRNAs went up in pyccr4-1- parasites. In that work and here in this manuscript, we have carefully decided to apply the word “preserved” to the fate of these mRNAs as it describes in a general way what is happening. In order to robustly state that mRNAs are stabilized by PyNOT1-G (directly or indirectly) would require additional experiments designed to test this (more description on this is provided on a response below). Second, as described above, we agree that doing a reciprocal IP for mass spectrometry-based proteomics would be ideal, we attempted four different approaches to do this to no avail. However, the composite proteomics data that is already available in the literature and via the Liu et al. preprint from the Bozdech Lab all indicate that these interactions occur, and perhaps that NOT1 and NOT1-G are mutually exclusive as expected. In our revised manuscript, we have provided further explanation in the Discussion for our use of the descriptor “preserve” instead of “stabilize”, and as noted above, and we have expanded our Discussion to more comprehensively define the interaction network depicted in Figure 7.

In both cases, the conclusions of the authors are very likely (e.g. downregulation of many genes as seen by RNA-seq), but the final experimental evidence is not provided and a network such as in Figure 7 is not fully supported. If the authors would like to maintain these statements, then they should be rephrased and made clear or the additional experimental evidence suggested below is necessary.

Response**: We hold that the published proteomic datasets do support such a network, with further support offered from the preliminary proteomic evidence from the Liu et al preprint. Therefore, we have not modified our manuscript beyond the additional text now provided in the Discussion as noted above.

• Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

  The essential claim that NOT1-G is important for gametocytogenesis and early mosquito development is well presented and fully supported by the experiments. As for the role of NOT1-G in 'preserving' mRNA, an mRNA half-life experiment would be necessary (or the text should be adjusted as mentioned above). In a short-term in vitro culture, pynot1-g- and WT parasites could be treated with ActD and abundances of select transcripts are measured by RT-qPCR.

Response**: We appreciate that Reviewer 2 considers the rigor of our experiments to be high. Regarding the use of the term “preserve” vs “stabilize”, we agree that to shift from our more general descriptor (preserve) to one that has specific connotations (stabilize) would require additional experimentation. To correctly and most robustly make the claim of stabilization would require work on par with that done by Painter et al. (PMID: 29985403) that uses a thiol-containing nucleotide (4-TU) along with a yeast-derived fusion enzyme (yFCU) to convert it for use by Plasmodium. Previously we have shown that an associated deadenylase (PyCCR4-1) also acted to preserve mRNAs, and moreover that deletion of its gene resulted in no discernable effect upon the poly(A) tail or 3’ UTR of an mRNA that is bound by this complex (p28).

While understanding mRNA stability is an exciting area of study, this 4-TU labeling experiment alone warranted a standalone, high impact publication for Painter et al. As this has not been adapted for any rodent-infectious Plasmodium species to date, and as adaptation of this labeling approach took several years for Dr. Painter while in the Llinas Laboratory (personal communication), we believe this work is beyond the scope of this study. Moreover, the additional information that it would provide to understand NOT1-g functions (preserve vs stabilize) would be incremental beyond the major storyline presented in this manuscript. In our revised manuscript, we have added text to ensure that our choice of “preserve” is well defined and explained.

To support the idea that NOT-1 and NOT1-G associate in a mutually exclusive way or to just show that they act in distinct complexes despite their similar expression patterns, an IFA with a double stained NOT1/NOT-1G cell line could be performed. Alternatively, the authors could perform a protein co-IP using the already existing NOT1/NOT1-G-GFP cell line and show that the proteins don't interact with each other or even have certain distinct interaction partners.

Response**: We agree, and these studies were attempted but were unsuccessful (described in our responses above). In our revised manuscript, we have included this information as noted above.

• Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

  All necessary cell lines for a NOT1/NOT1-G co-IP and the ActD experiment are already present. The authors already present a ring to schizont in vitro culture (for ActD) and also have substantial experience in protein co-IP and proteomics.

  I am not sure about the cost for a proteomics experiment at the author's institute and I don't want to make a guess on time investment given the still on-going COVID situation.

Response**: We agree that these experiments would be interesting, and would be costly to do at a transcriptome-wide scale and would require substantial time to conduct. We believe that the 4-TU approach noted above is the most rigorous, but is well beyond the scope of this study as it has not yet been adapted to rodent-infectious malaria parasites. As noted above, we have attempted four different proteomics approaches to provide reciprocal evidence for the complex composition which were unsuccessful. In our revised manuscript, we have added text to ensure that our choice of “preserve” is well defined and explained, and have noted the unsuccessful reciprocal proteomics approaches.

• Are the data and the methods presented in such a way that they can be reproduced?

  The MM section is well structured and presented and the supplemental material includes all data.

Response**: Thank you. We want to ensure that our work is clearly described and can be reproduced with the information reported.

• Are the experiments adequately replicated and statistical analysis adequate?

  There is hardly any test of significance presented in the main text of the manuscript (e.g. Figure 3B and 4A). Please show the individual data points for these graphs and make sure the n= and the statistical test is described in the figure legend. If you use the term significant in the text, then just add the p-value behind it. This is also true for the RNA-seq data: Genes are sorted by fold-changes, leaving it unclear if these changes are significant. These data are however presented in Table S1 and could be incorporated in the main text.

Response**: We agree. In our revised manuscript, we have incorporated additional details about the statistical tests used, p-values for noteworthy comparisons, and have included more panels for our comparative RNA-seq datasets (heatmap, PCA, MA plots). We have also made adjustments to our plots to make individual data points more readily observed, especially when error bars may block them (e.g. Figure 3B). And as in the original submission, all of the pertinent values, including fold changes, statistics and more are provided in our comprehensive supplementary files. We have structured the Supplementary Tables to flow from one tab to the next with the filtering/threshold applied noted both in the tab name and in the README tab that is found first among the tabs.

**Minor comments:**

• Specific experimental issues that are easily addressable.

  One idea that is also not discussed but could be added is for example that NOT1-G itself doesn't even have a stabilizing effect itself, but act as a decoy for other components of the CCR4/Caf1 complex, keeping them from associating with NOT1. In the NOT1-G knockout, the decrease in RNA abundance might then be just a result of an 'overactivity' of CCR4/Caf1/NOT1.

Response**: This hypothesis proposed by Reviewer 2, that PyNOT1-G is acting as a decoy or a binding partner sponge, is certainly feasible. For this scenario to be effective, PyNOT1-G would need to be in excess of PyNOT1 and/or would need to be able to bind to the critical effector protein(s) better than does PyNOT1. However, our microscopy data, along with the transcriptomic data presented here and previously published proteomic data would indicate that these two gene products are in approximately balanced proportions and are similarly localized. This does not exclude the possibility that PyNOT1-G could act as a sponge for relevant binding partners. In our revised manuscript, we have raised this possibility as an alternate explanation for the phenotype in the Discussion section.

• Are prior studies referenced appropriately?

  Throughout the manuscript, the authors should make clear what results come from which organism. Just as an example, the genome wide KO screens were performed in P. berghei and P. falciparum, CCR4/CAF1 experiments were performed in P. yoelii, whereas the original DDX6 work was done in P. berghei.

Response**: We agree. In our revised manuscript, we have added additional text to further clarify what data comes from which Plasmodium species.

• Are the text and figures clear and accurate?

  The Introduction is a bit long and partially turns into a minireview of eukaryotic RNA degradation. In the main text on page 13, the authors introduce a model for proteins involved in translational repression. This in not fully accurate, since for many of the proteins in this network, an effect on translation has actually not been shown. This includes NOT1-G characterized in the present work that most likely has an effect on mRNA stability, but for which a role in regulating translation is not presented.

Response**: We believe the length and content of this Introduction is appropriate to provide the context that some readers outside of the parasitology field will need to appreciate these findings. Regarding designations for these proteins as being related to translational repression, we think that the ample proteomic evidence tying them to translationally repressive complexes warrants this. In our revised manuscript, we have made it more clear that these proteins themselves have not been directly implicated in translational repression.

• Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

  Overall the RNA-seq is underrepresented and Figure 5 could easily be expanded by adding several panels that would help the future reader getting a better idea of the data:

1. Summary graphs such as PCA/MDS plots of the different replicates and MA-plots (all of which can be easily generated in DESeq2)
2. Heatmaps comparing the expression patterns of pynot1-g-, pbdozi-, pbcith-, pyalba4- highlighting some key gametocyte genes mentioned in the text

3. Alternatively to 2., a simple Venn Diagram would already be very informative

   An informative representation might also be to sort the differentially expressed genes as predominant male and/or female. The P. berghei data by Yeoh et al (PMID: 28923023) could be a starting point.

Response**: We agree. In our revised manuscript, we have expanded Figure 5 to include additional plots that speak the rigor of these datasets. Specifically, we have added a comprehensive heatmap and PCA plots, as well as MA plots as recommended. We have chosen not to include a Venn diagram for the overlap of affected mRNAs across these transgenic parasite lines, as we hold that this information is best provided in the text (high level observations) and the Supplement (details).

Reviewer #2 (Significance (Required)):

**Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.**

Technically this manuscript builds on standard methods of the field that are well executed. There is no direct clinical advancement, although one might argue that a unique adaptation of the parasite could always be a novel therapeutic target. Conceptually this is great advancement for the parasitology field as it is, providing additional evidence for the importance of post-transcriptional regulation for parasite transmission. With the two experiments suggested above and the additional evidence gained from it, this manuscript could also gain great interest to readers outside the field by clearly showing how alternative ways to regulate RNA stability evolved.

Response**: We are grateful for your careful review of our work and for the recommendations that you provided. We have incorporated many of them into the revised manuscript to make it even more rigorous and comprehensive. We also appreciate hearing that this work would be of great interest to a broader community. We feel that this is already the case, as the duplication of NOT1 and the dedication of one paralogue to an essential function is exciting and novel among eukaryotes.

**Place the work in the context of the existing literature (provide references, where appropriate)**

The work builds on the early reports of the particular RNA metabolism in gametocytes performed in the groups of Andy Waters. Since then, the authors themselves have published a great set of manuscripts extending our knowledge of the proteins involved in gametocytogenesis and nicely place the current work into this framework.

Response**: We appreciate this positive feedback. This is a fascinating topic to study.

**State what audience might be interested in and influenced by the reported findings.**

The manuscript as it stands is particularly interesting for the parasitology and potentially the evolutionary biology field. For a broader readership for example in the RNA field, the possibly antagonistic roles and mutually exclusive association with CAF1/CCR4 are likely most interesting.

Response**: We agree that this should be interesting to readers beyond our own field, as the duplication and specialization of NOT1, and the finding that the “canonical” PyNOT1 can be deleted, are both of general interest to how eukaryotes have adapted and deployed a highly conserved and essential RNA metabolic complex.

**Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.**

**Expertise:**

RNA biology, Plasmodium falciparum, Bioinformatics

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

In this manuscript, the authors investigate the requirement of two possible Not1 paralogs for the development of asexual blood stages and for the sexual transmission stages of Plasmodium yoelii. While Not1 is critical for asexual blood stages, its putative paralog, Not1G is important for the development of sexual transmission stages. In the absence of Not1G, male gametes are not formed while female gametes are formed and can be fertilised by wt male gametes. However, the resulting zygote cannot develop further into ookinete. The in vitro genetic cross assay to show this is elegant! A transcriptomic analysis further indicates that the transcriptomes of Not1G deficient parasites are significantly different from their WT counterpart.

Response**: We are thrilled that you found our evidence and approaches to be rigorous and compelling. Thank you.

**Major comments:**

The discussion section is very nice and the authors describe well what is speculative and should be further confirmed by additional experiments. However, I did find this was not the case in the results section where the authors are proposing conclusions that are not supported by the results. I think the reading of this manuscript would be much more enjoyable if the authors only describe the results shown and move all the discussions to the dedicated section. Below are some examples. The data presented in this manuscript is not showing a nexus, this is a suggestion based on the results of other articles, the word should thus be removed from the title (and kept for a future review!). The last two sentences of the localisation section should be moved to the discussion because they do refer to results not shown in this manuscript. The last sentence of the second paragraph of the zygote development section should also be moved to the discussion. For the transcriptomic analysis there is also no formal comparison with transcriptomes of other previously analysed mutants: the results of the comparisons should either be shown or not discussed in the result section. Finally, the discussions mentioning interactors of the complex should be removed from the result section and moved to the discussion unless the results are formally analysed.

Response**: We again thank you for the complement. In our original manuscript, we opted to provide some limited interpretations and context within the Results section in order to help guide readers along our train-of-thought and line-of-experimentation. While a more traditional split of keeping essentially all discussion and interpretation for the Discussion is a tried-and-true approach, we prefer this more narrative method and have opted to keep these short sections in the Results section.

I would strongly suggest the author the better present and describe their transcriptomic results. There is only one volcano plot indicating the overall defect in mixed gametocytes in the main figure. Apart from this, the results are only described in the main text or in supplementary tables. It is therefore difficult to understand the subtilities of the analysis. For example, the authors frequently mention dysregulated genes, but without specifying whether it is up or down-regulated in the mutant. To address this issue, I would suggest the authors to better describe their results in the figures. They could show the GO term enrichment analysis they mention and show how they assign GO term or transcripts to male and female parasites. It would also be nice to discuss some of the results a bit more in details. For example, it is not surprising to see a reduction in transcripts that are under the control of AP2-O in retort-arrested ookinetes as the parasite do not reach this stage. It is thus highly speculative to specifically link this observation with ALBA4 without further detailed analysis. On the other hand, it is more surprising to see a decrease in ap2g transcripts, while the authors observe an increased gametocytaemia. Could the authors comment this observation? It may also be nice to better present the comparison between gametocytes and schizonts to possibly speculate on the early requirement of Not1G in committed schizonts.

Response**: We (and Reviewer 2) agree. In our revised manuscript, we have expanded Figure 5 to include additional plots that speak the rigor of these datasets. Specifically, we have added a heatmap, and PCA and MA plots as recommended. We have chosen not to include a Venn diagrams for the overlap of affected mRNAs across these transgenic parasite lines for the reasons stated above in our response to Reviewer 2. Similarly, we have opted to keep the specifics of the GO Term analyses in the Supplement as we believe these should always be taken with a grain of salt (especially high level GO Terms, as many choose to report). Finally, we have expanded our discussion on our observation that pyapiap2-g transcript levels are lower in the pynot1-g- line, despite seeing a slight increase in gametocytemia.

The conclusion regarding the similar localisation of Not1 and Not1G with other members of the CAF1/CCR4/NOT complex is not really convincing for two reasons. First, there is not colocalization shown and, second, the distribution is not very peculiar so it is difficult to draw any conclusion with this level of resolution. The presence of alpha-tubulin in the nucleus of male gametocytes is also very surprising as it is rather nucleus-excluded in both P. falciparum and P. berghei, could the authors comment this peculiar localisation?

Response**: We agree and disagree here. First, we agree that no colocalization data is presented here to place NOT1-G within the limit of resolution of fluorescence microscopy. What we can (and do) state is that these proteins are all localized to cytosolic puncta, which matches what is observed for essentially all other studied eukaryotes. In further support of this, our published, quantitative proteomic data indicates that the bioinformatically predictable members of the CAF1/CCR4/NOT complex do associate as anticipated. In the same vein, the micrographs presented were not captured by confocal microscopy, and thus the apparent localization of alpha tubulin “in” the nucleus is most likely attributed to being above and/or below the nucleus. Taken together, we do feel that the combined evidence is convincing. As we have already made all of these points in the original manuscript, we have not adjusted the revised manuscript further.

One of my major frustration when reading this manuscript was that the authors are not trying to discriminate between an early role of Not1G during gametocytogenesis or later in gametogenesis. The fact that the transcriptomes of gametocytes and schizonts seem to show similarities suggests that the phenotype observed during both male gametogenesis or ookinete development are probably linked to early knock-on defects during gametocytogenesis. Could the authors test whether male gametocytes replicate DNA or female activate translation? These are of course non-essential experiments as the authors are careful with their conclusions and mention possible defects during both gametocytogenesis or gametogenesis. Addressing this question may however add significant insights into the requirement for Not1G.

Response**: We are sorry for the frustration. We wrote the manuscript so as to state what we feel we could robustly say, and where we are drawn to speculate, we made that speculation clear. As Reviewer 3 notes, we have not attempted to discriminate between functions that PyNOT1-G may be playing in different stages or substages of development because we do not believe the experiments allow that discrimination. While we could investigate finer and finer aspects of possible defects in both male and female gametocyte development, the most impactful take home messages remain the same. We continue to address questions related to translational repression and its release, and anticipate that PyNOT1-G will play a substantial and essential role in this. As Reviewer 3 noted, we have already discussed these possibilities in the original manuscript, and thus have not added anything further about this in our revised manuscript.

**Minor comments:**

Please use page and line numbering for your next submissions! Please describe what "bioinformatics" was used. I would show the nice localisation in oocyst and sporozoite in the main section. The conclusions drawn from the genetic cross seem to come from a single biological replicate, if this is the case please indicate it clearly.

Response**: We apologize for these oversights. In our revised manuscript, we have provided page and line numbering, have expanded on what bioinformatic processes were done in the manuscript, and have made it more clear that the genetic crosses come from multiple biological replicates (biological triplicate for the transmission-based genetic cross, biological duplicate for the in vitro culture genetic cross). However, we have opted to retain the oocyst and sporozoite IFA data in the Supplement, as the rest of the story is focused on blood stage and early mosquito stage.

Reviewer #3 (Significance (Required)):

This manuscript highlights the requirement of a Not1 paralog in the transmission stages of a Plasmodium parasite. More specifically it describes a new player in the control of RNA biology during this process where our knowledge is scarce. It will be a valuable manuscript for molecular parasitologists interested in transmission or RNA biology.

Response**: We agree and are grateful that our colleagues find this study to be a valuable addition in our efforts to understand how malaria parasites have adapted classic eukaryotic mechanisms to suit their purposes.

Our expertise is largely in molecular and cellular parasitology.

I think this is a framework that could use more emphasis. It’s one I am cueing into more after Caviola et al. (2021) included it in their review, “the psychology of (in)effective giving.”

People have a strong aversion to prioritizing some lives over others (see Tetlock et al., 2003, "Thinking the unthinkable: sacred values and taboo cognitions"). With limited resources, we of course do this all the time. But CBA makes it uncomfortably explicit. To prioritize some recipients as a result of CBA means to deprioritize others, which feels unfair. This is one explanation for why people prefer “distributed helping” when there are multiple possible recipients, even at the expense of helping more, since then at least no one is fully deprioritized (Caviola et al. 2020a, obstacle 5; Sharps & Schroder, 2019, “The Preference for Distributed Helping”). This could also be an explanation for Berman et al., 2018 finding that people prefer to prioritize investments rather than charities, since deprioritizing an investment isn’t nearly as aversive.

A moral aversion to (de)prioritization may also explain social judgments of people who donate effectively seeming “cold” (section 7.1). This is evidenced by the differences in instinctive judgments of “coldness” based on what is deprioritized. For example, deprioritizing investing in textbooks because it isn’t an effective intervention feels much different than deprioritizing investment in childhood cancer treatment because one could help more kids dying of malaria. People would likely make harsher judgments about someone doing the latter even though the reasoning is the same – it’s what is deprioritized that is different.

There might also be something else at play related to ‘CBA’ discomfort: choosing whom to help makes it clear to individuals that they can’t help everyone. It reminds people of all the suffering in the world that they can’t alleviate, whereas just choosing a neat charity only introduces one cause of suffering and then gives the donor the satisfaction that they have done something to alleviate it. I can imagine that CBAs role in revealing the reality of triage (1) makes people less inclined to engage in CBA and (2) less likely to donate a lot in accordance with CBA because there is less warm glow/ that one cause just isn’t as sexy anymore. (2) is related to the idea of Pseudoineffecay developed in (Slovic, 2007; Västfjäll et al., 2015). People are less inclined to help when they learn about others they can’t help.

A key idea that I think is relevant here is the “affect heuristic,” the importance of instinctive emotional cues of “goodness” or “badness” informing decisions (LINK). Deprioritization of emotional cause → instinctively violates moral value → aversion → less likely to engage with, worse social . Similarly, reminder of all the suffering in the world → feeling of sadness + helplessness → avoidant behavior. These oversimplified decision pathways can be overruled by rational, deliberate processing (see Tetlock, 2003 for discussion specific to sacred values) , but charitable giving is largely a system 1/emotional arena.

I feel this is one of the most important parts of documentation and observation. I feel we should give children the space to think about what they make to give them a sense of ownership and accomplishment. If we constantly share our ideas, the child may not see theirs as valid.

I think that awareness is key for being able to teach within a context we as adults may be familiar with. In this particular experience there was almost a system of checks and balances to make sure the students hypotheses and ideas stayed at the forefront of their research.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

We thank all four reviewers for their positive and constructive comments! We have carefully considered these comments and provided a point-by-point response below.

Reviewer #1 (Evidence, reproducibility and clarity):

This paper explores an interesting problem of SHP1/SHP2 preferences of inhibitory immunoreceptors. The author are quick to point out that many of their individual data points confirm published results at some level, but the power of the paper is in the parallel analysis of both PD1, which is strongly biased towards SHP2 and BTLA, which is biased towards SHP1. This gives them the opportunity to test the predictions of descriptive experiment by making simple mutated receptors with swapped ITIM or ITSM domains.

The work is very well done and generally the authors are quite careful and precise about the language used to describe results, in general.

The results are quite striking in that the find plenty of evidence for transient interaction of SHP1 with PD1 based on the biophysical measurements, but don't detect the interactions in pull down or in "in cell" microcluster recruitment experiments. In describing the pull-downs they discuss the issue of dissociation during washing potentially missing interactions that are taking place. I would prefer that the pull down is fine evidence for binding, but lack of pull down is not evidence for lack of binding. They should double check that this language is consistent. Also, unless something has changed in the microcluster binding experiments, this in situ recruitment of SHP2 to PD1 is only observed or a 2-3 minutes and then can't be detected, the situation for SHP2 becoming the same as it is for SHP1. If the kinetics are different in the cleaner systems that have now developed they should show this in a primary figure as this would be then different when what is reported previously.

We agree with the reviewer that pull down is evidence for binding. Indeed, in most, if not all of our assays, our results with pull down were consistent with those in the microcluster imaging. As suggested by the reviewer, we will check through the manuscript and ensure the language is accurate and consistent. In our recent study (Xu et al., JCB, 2020, PMID: 32437509), we conducted a side-by-side comparison of SHP2 and SHP1 recruitment kinetics to PD-1 in a similar system as the current study. Both microcluster imaging and co-IP assays showed that PD-1:SHP2 association lasted at least 10 minutes, whereas PD-1:SHP1 recruitment was nearly undetectable. The duration of PD- 1:SHP2 association was in good agreement with Takashi Saito’s finding in CD4+ mouse T cells (Yokosuka et al., JEM, 2012, PMID: 22641383). Regardless the somewhat different kinetics in different studies, SHP2 recruitment was transient, as pointed out by the reviewer. We believe that some other effectors contribute to PD-1 inhibitory signaling. In supportive of this notion, we recently found that PD-1 remains partially inhibitory in CD8+ T cells deficient in both SHP1 and SHP2 (Xu et al., JCB, 2020).

The gap in this study is lack of any functional analysis. The Jurkat model could be quite useful as they have a relatively clean system for asking if the transient binding of SHP1 to PD1 has any functional impact, which they have not yet followed through on. Does PD-1 recruited SHP2 have any impact on function after the 5 minutes? Furthermore, the authors need to keep in mind that mice deficient in SHP2 respond to anti-PD1 checkpoint therapies (Rota, G., Niogret, C., Dang, A. T., Barros, C. R., Fonta, N. P., Alfei, F., Morgado, L., Zehn, D., Birchmeier, W., Vivier, E., & Guarda, G. (2018). Shp-2 Is Dispensable for Establishing T Cell Exhaustion and for PD-1 Signaling In Vivo. Cell Rep, 23(1), 39-49. https://doi.org/10.1016/j.celrep.2018.03.026). This is an important issue to discuss in light the the very interesting binding analysis the authors have performed. But I think the functional analysis can be part of a future paper.

In our recent publication (Xu et al. JCB, 2020, PMID: 32437509), we found that deletion of SHP1 from Jurkat cells had little, if any effect on PD-1 mediated suppression of IL-2 production. As the reviewer alluded to, we did observe SHP2 dissociation from PD-1 after 10 minutes, so the question of whether and how PD-1:SHP2 complex influence T cell function in a longer term is a great one. We currently are pursuing a hypothesis that there is a SHP2-independent mechanism of PD-1 inhibitory function, and indeed, in our recent study (Xu et al. JCB, 2020, PMID: 32437509), we found that PD-1 retains its partial inhibitory function in SHP1/SHP2 double knockout murine primary T cells. These results are consistent with the in vivo data by Rota et al. cited by the reviewer. We will also briefly discuss this point in a revised manuscript.

I would suggest that the title be modified slightly from "SHP1/SHP2 discrimination" to "differential SHP1/SHP2 interaction" and leave discussion of discrimination until they have the functional data integrated over times that are relevant to T cell transcriptional regulation (1-2 hrs). The functional analysis can be in another paper, but it would be interesting to have a paragraph in the discussion raising the outstanding issues beyond stable binding detected by the pull-down and microcluster recruitment experiments- what are the implications for function. Could the transient interactions in the noise of the steady state and equilibrium measurements be functional?

We thank the reviewer for the suggestion, even though reviewer #3 felt that our current title is appropriate. We will be happy to change the title at the editors’ discretion.

I would summarise that the work is outstanding as biochemistry and biophysics and it should be published nearly as is. I'm suggesting minor revisions in that the changes are just to text, but I think this is important and somewhat nuanced aspect of the paper that will make it even more helpful to readers.

We appreciate the positive and insightful comments!

Reviewer #1 (Significance):

The authors generate a detailed descriptive data set about the component interaction of SHP1 and SHP2 SH2 domains with PD1 and BTLA intracellular domains. They then test hypotheses generated from the descriptive data set to better define the nature of the interactions and why PD1 recruits primarily SHP2, while BTLA mainly recruits SHP1. PD1 is a major driver or the cancer immunotherapy revolution and SHP2 is the major candidate for a signalling effector of PD1. This paper can become the reference paper for the specificity and engineering of this interaction, which will make it highly significant in a very active and still expanding field.

Referee Cross-commenting

I still feel that "discrimination" has a functional/activity connotation that is not addressed at all in this paper, but can be addressed. I'm happy to have the suggestion stand and let the authors decide. They need to live with it once its published. Another suggestion- the citations on regulation are mostly old. A good recent paper is Pádua, R. A. P., Sun, Y., Marko, I., Pitsawong, W., Stiller, J. B., Otten, R., & Kern, D. (2018). Mechanism of activating mutations and allosteric drug inhibition of the phosphatase SHP2. Nature Communications, 9(1),

1. https://doi.org/10.1038/s41467-018-06814-w .

We believe that some of the functional questions raised by this reviewer, including the SHP1 and SHP2 contribution in PD-1 signaling, was addressed in our recent publication (Xu et al., JCB, 2020). Using SHP1 KO and SHP2 KO T cells, we showed that PD-1 inhibitory function is contributed by SHP2, but very little if any by SHP1. Thus in the current study, we focus on the mechanism behind the striking SHP2 preference by PD-1. We thank this reviewer for suggesting this excellent reference. We will cite this reference in the revised manuscript.

Reviewer #2 (Evidence, reproducibility and clarity):

In this study, Xu and co-workers investigate the biophysical nature of the interaction between the structurally-related non-transmembrane PTPs Shp1 and Shp2 with the ITIM/ITSM-containing inhibitory receptors PD-1 and BTLA using cell-based, biochemical, biophysical and domain swapping assays. The primary aim being to better understand how these receptors discriminate between binding Shp1 and/or Shp2, and the orientation of Shp1 and Shp2 engagement. These are major unresolved questions in the field that the authors go some way to addressing in a methodical, rigorous, clear and concise manner. Findings are convincing, correlate well with previous findings and internally, and are complemented with excellent schematics, making it easy to comprehend.

Major comments

The authors focus primarily on binding affinities to explain differential binding of Shp1 and Shp2 by PD-1 and BTLA ITIMs and ITSMs, but this is only part of the story. Avidity, compartmentalization, stoichiometry of kinases, and relative abundance of Shp1 and Shp2 are also important aspects of the discriminatory mechanism that are not addressed. Competition assays would go some way to addressing the latter point and should be at least be considered and discussed.

We agree that various parameters mentioned by this reviewers, such as compartmentalization and relative expression levels would be a concern for purely cell-free assays such as SPR, however, we feel that our cell-based assays already integrate these parameters. This is also precisely the reason why we chose to examine the recruitments of Shp1/2 in a cellular context instead of a purely cell-free system.

Regarding the competition, we have confirmed our key results in both WT and SHP2 KO background, with or without the potential competition from endogenous SHP2, suggesting that competition might not be a dominant mechanism for the recruitment specificity we observed.

Similarly, authors do not address how distortion of the pY binding pocket of Shp1 and Shp2 nSH2 domains in the auto-inhibited conformation is released, allowing the domain to engage with phopho-ITIM/ITSM. Again, this should be at least discussed. Current binding studies do not address this issue.

We feel that the overall recruitment to the PD-1 microclusters as we observed in cells already integrate this auto-inhibition mechanism of Shp1 and Shp2, because we used full length proteins. We do agree with the reviewer that future studies are warranted to address the contributions of each mechanism, including auto-inhibition, concentration, competition, etc., to the overall recruitment. This might require careful and extensive biophysical analyses coupled with mathematical modeling.

Minor comments:

Phosphorylation should be indicated in schematic representations in Figures 3, 6 b, c.

We thank the reviewer for this advice, we will indicate phosphorylation in the revised figure 3.

Cellular and physiological significance should be further discussed, as well as broader implications of findings to other ITIM/ITSM-containing receptors in other lineages.

We will further discuss this as suggested.

Reviewer #2 (Significance)

Findings from this study advance our knowledge of how inhibitory checkpoint regulatory receptors discriminate between Shp1 and Shp2, which has important implications for understanding how the unique biochemical, cellular and physiological functions of these receptors and phosphatases are dictated. Indeed, findings lay the foundation for a universal mechanism, that may apply to all ITIM/ITSM receptors in other cell lineages, and perhaps novel ways of targeting these interactions therapeutically.

Compare to existing published knowledge

Although largely correlative with previous studies, findings from this study start to fill major gaps in our knowledge of these biochemical processes, in a highly rigorous, concise and clear manner. Findings from previous studies were more 'piecemeal', whereas this study consolidates and advances important nuances of these interactions. Moreover, it lays the foundation for further structural, physiological and therapeutic studies.

Audience

The immune receptor signaling community and beyond, including any lineage in which ITIM/ITSM-containing receptors play a major role in regulating cellular responses.

Your expertise

ITIM/ITSM-containing receptors, kinase-phosphatase molecular switches, cellular reactivity to extracellular matrix proteins

Referee Cross-commenting

Generally agree with reviewer's comments. Constructive overall and fair. Although I was thinking additional competition experiments, I do not think necessary. Over the top for this study. Hence, 1 month should suffice to revise accordingly.

We thank this reviewer for the excellent comments and understanding!

Reviewer #3 (Evidence, reproducibility and clarity):

Summary:

Inhibitory immune receptors containing ITIMs function through recruiting the phosphatases SHP-1 and SHP-2. SHP-1 and SHP-2 are remarkably similar yet have different roles in vivo. How can ITIM-containing immune receptors specifically recruit SHP-1 or SHP-2? In this paper, Xu et al ask how SHP-1 vs SHP-2 specificity is achieved. They use very thorough biochemical assays to measure the affinity of SHP-1 and SHP-2 for various ITIM/ITSMs and finally pin point some key amino acids that switch an ITIM/ITSM from SHP-2 to SHP-1 specificity. The in vitro biochemical assays are augmented by in cell assays that support their conclusions. Overall, this paper is an incredibly elegant and straight forward paper addressing how SHP-1/SHP-2 specificity is achieved.

Major Comments: none

Minor Comments:

• Could the western blots in Figure 1 be quantified as the western blots in other figures?

We will quantify the western blots in Figure 1 as suggested in the revised manuscript.

• The data that the y+1 reside is essential for SHP-1/2 specificity is very convincing. We are curious if the other residues of the ITIM/ITSM also contribute to this specificity, albeit less potently. The PD-1 G224A mutant is still less potent than the PD-1 BTLA ITIM swap, suggesting that while the y+1 position is most important, the other residues contribute some specificity. The authors also included data on a PD-1 variant with the BTLA ITIM A224G mutation (8f), which is slightly better at recruiting SHP-1 than the PD-1 ITIM. It may be worth mentioning this data in the text of the paper as well as displaying it in the figure.

The reviewer raised an excellent point, yes, our data does suggest that other pY-flanking residues within the ITIM also contribute to SHP1 binding. However, the pY+1 residue replacement produced the strongest effect as the reviewer noted. In the revised manuscript, we will acknowledge the potential contributions of other residues.

• A brief introduction to ITIM vs ITSM in the introduction of the paper may be helpful background for readers. For example, ITIM receptors are reasonably well known but how ITSM functionally differs is probably less well known.

We will rewrite the introduction about ITIM and ITSM for better clarity.

• Although not the major focus of the paper, broadening out this SHP-1/2 specificity to other immune receptors in the discussion is fascinating. (a) The authors find that a Valine, Leucine, or Isoleucine in place of the Alanine in y+1 is very close to equivalent, yet the A is highly conserved. The authors speculate that there may be an advantage to sub-maximal SHP-1 affinity because it is more easy to regulate. I think this is reasonable speculation but a little unsatisfying given the very small observed difference in SHP-1 binding. If the authors have additional thoughts, I would be interested to hear them. (b) The authors note that PD-1 is the only ITIM with a glycine in the Y+1 position. Are there other receptors that function primarily through SHP-2, and how might they achieve this specificity?

Response to a: Even though valine, leucine or isoleucine did not produce a striking enhancement in Shp1 recruitment over alanine, the differences were statistically significant. In fact, when we performed these point mutations at a BTLA ITIM background, valine, leucine or isoleucine markedly enhanced the SHP1 recruitment (see unpublished data below). We speculate that other pY-flanking residues in BTLA, as this reviewer alluded to above, creates an environment that amplifies the differences. The strong sensitivity on pY+1 residue, as observed in BTLA, might be true for other SHP1-recruiting receptors too. If they were to have leucine or isoleucine at the pY+1 position of ITIM, they may recruit too much SHP1 that presumably decreases the fitness/growth of the cells. We propose to show this unpublished data as a supplemental figure in the revised manuscript. We will also discuss the potential contributions of other pY-flanking residues as this reviewer suggested.

{{images cannot be rendered at this time in reply letters}}

Response to b: Among the several receptors that we tested, PD-1 is the only receptor that exhibited no recruitment of SHP1. The lack of SHP1 recruitment is also true for murine PD-1, which has a glutamate residue (charged) at Y+1 position. In addition, earlier work reported that PECAM1 also selectively recruits SHP2, but not SHP1. We have noted that PECAM1 contain a threonine (polar) at the pY+1 position of their ITIMs. Thus, their inability to recruit SHP1 is consistent with our model that a nonpolar residue at Y+1 position is required for strong SHP1 recruitment. We will discuss these points in the revised manuscript.

• Figure 9 b Val not Vla, Figure 3a - a legend for the color code may be nice (ie, 20-1000 nM) Thanks for catching this, we will fix the error in Figure 9b and provide the color code in Figure 3a in the revised manuscript.

Reviewer #3 (Significance):

Significance:

SHP-1 and SHP-2 play a critical role in regulating immune system function. In addition, the receptors recruiting these phosphatases (like PD-1) are important immunotherapy targets. Previously, the question of SHP-1/SHP-2 specificity has been primarily described for ITIM bearing receptors individually. Other studies have predicted consensus sequences for the tSH2 domains of SHP-1 or SHP-2, but not addressed the defining molecular characteristics of these consensus sites or how these could be combined on ITIM receptors to generate selectivity between these related phosphatases. This paper represents a significant step forward because it provides a unifying mechanism explaining how ITIM-bearing immune receptors specifically recruit SHP-1 or SHP-2. I expect this paper will be broadly interesting to biochemists, immunologists and cancer biologists.

Referee Cross-commenting

I generally think the other reviewers comments are reasonable and insightful. Together, they suggest no new experiments are necessary. As for the proposed title change, I prefer the authors title and find it to be justified given their data.

Reviewer #4 (Evidence, reproducibility and clarity):

In this manuscript, Xu and college performed an elaborate study to investigate the molecular basis of Shp1 and Shp2 discrimination by immune checkpoints PD-1 and BTLA. The paper is original, clear, and well written. I only have a few minor comments:

1. Please label the molecular weights to all the western blots/IPs results.

We will label the molecular weights to all the blots in the revised manuscript.

1. Please add scale bars to all the microscopy pictures.

We will add scale bars to all the microcopy images in the revised manuscript.

1. For the SPR data, please add the fitting curves.

We thank the reviewer for the suggestion. However, we did not use the fitting curve to calculate the Kd, we plotted the maximum response as a function of concentration to determine the Kd. This is another well accepted method for Kd calculation. In fact, some of the SPR curves fit poorly with the existing algorithm. Thus, showing the fitting curve might distract the readers.

Reviewer #4 (Significance):

The strength of this paper relies on the details they dissected by using a series of mutagenesis screening experiments, which should be interesting to cell biologists and cancer immunologists.

Referee Cross-commenting

I think the other reviewer's comments are insightful and constructive, the suggested experiments are necessary and will improve the paper.

We thank this reviewer for the positive comments!

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

Summary:

Inhibitory immune receptors containing ITIMs function through recruiting the phosphatases SHP-1 and SHP-2. SHP-1 and SHP-2 are remarkably similar yet have different roles in vivo. How can ITIM-containing immune receptors specifically recruit SHP-1 or SHP-2? In this paper, Xu et al ask how SHP-1 vs SHP-2 specificity is achieved. They use very thorough biochemical assays to measure the affinity of SHP-1 and SHP-2 for various ITIM/ITSMs and finally pin point some key amino acids that switch an ITIM/ITSM from SHP-2 to SHP-1 specificity. The in vitro biochemical assays are augmented by in cell assays that support their conclusions. Overall, this paper is an incredibly elegant and straight forward paper addressing how SHP-1/SHP-2 specificity is achieved.

Major Comments:

none

Minor Comments:

• Could the western blots in Figure 1 be quantified as the western blots in other figures?
• The data that the y+1 reside is essential for SHP-1/2 specificity is very convincing. We are curious if the other residues of the ITIM/ITSM also contribute to this specificity, albeit less potently. The PD-1 G224A mutant is still less potent than the PD-1 BTLA ITIM swap, suggesting that while the y+1 position is most important, the other residues contribute some specificity. The authors also included data on a PD-1 variant with the BTLA ITIM A224G mutation (8f), which is slightly better at recruiting SHP-1 than the PD-1 ITIM. It may be worth mentioning this data in the text of the paper as well as displaying it in the figure.
• A brief introduction to ITIM vs ITSM in the introduction of the paper may be helpful background for readers. For example, ITIM receptors are reasonably well known but how ITSM functionally differs is probably less well known.
• Although not the major focus of the paper, broadening out this SHP-1/2 specificity to other immune receptors in the discussion is fascinating. (a) The authors find that a Valine, Leucine, or Isoleucine in place of the Alanine in y+1 is very close to equivalent, yet the A is highly conserved. The authors speculate that there may be an advantage to sub-maximal SHP-1 affinity because it is more easy to regulate. I think this is reasonable speculation but a little unsatisfying given the very small observed difference in SHP-1 binding. If the authors have additional thoughts, I would be interested to hear them. (b) The authors note that PD-1 is the only ITIM with a glycine in the Y+1 position. Are there other receptors that function primarily through SHP-2, and how might they achieve this specificity?
• Figure 9 b Val not Vla, Figure 3a - a legend for the color code may be nice (ie, 20-1000 nM)

Significance

SHP-1 and SHP-2 play a critical role in regulating immune system function. In addition, the receptors recruiting these phosphatases (like PD-1) are important immunotherapy targets. Previously, the question of SHP-1/SHP-2 specificity has been primarily described for ITIM bearing receptors individually. Other studies have predicted consensus sequences for the tSH2 domains of SHP-1 or SHP-2, but not addressed the defining molecular characteristics of these consensus sites or how these could be combined on ITIM receptors to generate selectivity between these related phosphatases. This paper represents a significant step forward because it provides a unifying mechanism explaining how ITIM-bearing immune receptors specifically recruit SHP-1 or SHP-2. I expect this paper will be broadly interesting to biochemists, immunologists and cancer biologists.

Referee Cross-commenting

I generally think the other reviewers comments are reasonable and insightful. Together, they suggest no new experiments are necessary. As for the proposed title change, I prefer the authors title and find it to be justified given their data.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Editor comments:

Thank you for sending your manuscript entitled "In situ imaging of bacterial membrane projections and associated protein complexes using electron cryo-tomography" to Review Commons. We have now completed the peer review of the manuscript. Please find the full set of reports below.

We thank the editors of Review Commons and all the reviewers for their insightful comments which helped us to improve our manuscript. We have now modified our manuscript based on the Reviewers’ comments and would like to ask you to consider our revised manuscript for publication.

Reviewer #1:

This manuscript by the Jensen lab surveys a plethora of bacterial outer-membrane projections captured over the years by in situ cryo-tomography under near-native conditions. The authors classify the different visualized structures, highlighting both similarities and differences among them. They further describe molecular complexes that are associated with these projections. The manuscript highlights the abundance of such understudied structures in nature, indicating the need to deepen our exploration into their biological functions and mechanisms of action.

We thank the reviewer for her/his insightful comments that allowed us to improve our manuscript.

The authors should state in the Abstract and Introduction that only diderm bacteria and outer- membrane extensions are included in the study.

Done. We have modified the title, the abstract and the introduction to explicitly highlight this point.

In the Introduction or Discussion the authors should mention the limits of the in situ cryo-tomography, such as the difficulty to observe regions in between neigbouring bacterial cells, and into the thick bacterial cell body.

Done. We have added the following to our revised manuscript:

“Currently, only electron cryo-tomography (cryo-ET) allows visualization of structures in a near-native state inside intact (frozen-hydrated) cells with macromolecular (~5 nm) resolution. However, this capability is limited to thin samples (few hundred nanometers thick, like individual bacterial cells of many species) while thicker samples like the central part of eukaryotic cells, thick bacterial cells, or clusters of bacterial cells are not amenable for direct cryo-ET imaging. Such thick samples can be rendered suitable for cryo-ET experiments by thinning them first using different methods including focused ion beam milling and cryosectioning [30]. Cryo-ET has already been invaluable in revealing the structures of several membrane extensions, including Shewanella oneidensis nanowires [6], Helicobacter pylori tubes [15], Delftia acidovorans nanopods [25], Vibrio vulnificus OMV chains [16], and more recently cell-cell bridges in the archaeon Haloferax volcanii [31].” (Lines 108-118)

Please provide a legend to Table S1 explaining the numbers (organelles?), how many cells were viewed? I think that at least part of it should be included in the main text. Also, there are examples of vesicles emanating from H. pylori. This information is missing from Table S1.

Done. We added a column to the table indicating the number of cells available for each species. We also added the information about the vesicles in H. pylori to the table. This table is now incorporated into the main text of the manuscript as Table 1.

Please provide an ordered list including all the strains (and IDs of the specific isolates) used in this study and their genotypes.

Done. We added Table S1 to the revised manuscript that contains this information. This table also includes relevant references to all the published papers where these strains were previously used.

The authors describe in detail the H. pylori tubes that seem to be flagellum-core independent. However, the authors found previously (ref 15) that during infection, these structures are dependent on CagA T4SS, and they visualized T4SS sub-complexes in proximity to the point of tube emanation. This should be described and discussed in the text. Also, please indicate if the "host-independent" tubes are similarly dependent on T4SS.

Done. We added the following to the revised manuscript:

“The scaffolded uniform tubes of H. pylori that we observed were formed in samples not incubated with eukaryotic cells, indicating that they can also form in their absence. However, the tubes we found had closed ends and no clear lateral ports, while some of the previously-reported tubes (formed in the presence of eukaryotic host cells) had open ends and prominent ports [15]. It is possible that such features are formed only when H. pylori are in the vicinity of host cells. Moreover, while it was previously hypothesized that the formation of membrane tubes in H. pylori (when they are in the vicinity of eukaryotic cells) is dependent on the cag T4SS [15], we could not identify any clear correlation between the emanation of membrane tubes and cag T4SS particles in our samples where H. pylori was not incubated with host cells. We also show that the tubes of H. pylori are CORE-independent, indicating that they are different from the CORE-dependent nanotubes described in other species.” (Lines 303-313)

Is there any difference in the frequency or length of the tubes in the mutants presented in Figure 4? The flgS mutant in the image exhibits a very short filament; is that typical?

We did not see any significant statistical difference in the number or lengths of the tubes in these different mutants. We added Table S2 to the revised manuscript which details the number of cells we visualized for each mutant and the number of the tubes seen there. In all these mutants the lengths of the tubes ranged between few tens to hundreds of nanometers. In addition, we added Fig. S2 to show more examples of these tubes in each of these mutants.

Minor points:

-Please check full bacterial names that are sometimes missing (e.g., lines 110-112).

Done.

-There is no reference to panel 2G. Please check the references to all panels.

Done. Please see lines 154 and 183 in the main text.

-Lines 181-184: There is no figure related to the formation of teardrop-like extensions from C. pinensis. Please review the text accordingly.

Done. Corrected.

-Line 235, not clear to what "as these" refers to.

Done. We modified the text as the following:

“As these MEs/MVs from S. oneidensis were purified” (Lines 246-247)

-Line 241, not clear what "a secretin-like complex" is, and no reference is provided.

Done. We modified the text as the following:

“In the third category, we observed a secretin-like complex in many tubes and vesicles of F. johnsoniae. Secretins are proteins that form a pore in the outer membrane and are associated with many secretion systems like type IV pili and type II secretion systems (T2SS) [39–41]” (Lines 252-254)

Reviewer #1 (Significance)

As described in this manuscript, even in model bacteria these structures are generated (e.g., Caulobacter forms the hardly studied nanopod extensions). The manuscript also provides visual categories of these structures, defining "extension types" that are likely to be used by the scientific community for years to come, similar to the initial pili classification during the 1960s-70s. It is a "descriptive study," in the positive sense of the term, as it significantly contributes to the field of bacteriology.

We thank the reviewer for her/his kind words and enthusiasm about our work. It is an honor to have our work compared to the seminal pili classification work done in the 1960s-70s by pioneers in the field of bacteriology.

Reviewer #2:

The manuscript "In situ imaging of bacterial membrane projections and associated protein complexes using electron cryo-tomography" by Kaplan et al., identifies and catalogues membrane extensions (MEs) and membrane vesicles (MVs) from 13 different species using cryo-electron tomography. Furthermore, they identify and discuss several protein complexes observed in these membrane projections.

The manuscript is beautifully written, interesting, and genuinely got this reviewer excited about the biology. I applaud the authors on their manuscript and have only minor comments and a few thoughts that the authors may wish to think on and discuss.

We thank the reviewer for her/his kind words and insightful comments that allowed us to improve our manuscript.

Some schematics throughout the introduction would be useful to readers new to the field/ outside the field who are not used to these different membrane structure features.

We thank the Reviewer for this suggestion. First, we made an extra figure with schematics showing the cell body and membrane tubes but that was rather redundant with Figure 8. For this reason, we added explicit labels to figure 1 highlighting the cell body and the tubes in these examples to help the reader following that figure and the subsequent ones. However, if the Reviewer has an explicit suggestion/view about the schematics then we would be very happy to do that.

The size of scale bars should be indicated on the figure panels themselves rather than in the figure legend to assist the reader.

Done.

In reference to lines 193-196 - what was the extracellular environment like in these micrographs? Were other cells present? Could it be the extracellular environment/surrounding cells that stimulate pearling? Have the authors considered this? Please discuss if relevant/insightful.

This is a good point. The cells were usually plunge-frozen in their standard growth media (except in H. pylori where the cells were resuspended in PBS and subsequently plunge-frozen). Yes, there are other cells present in the sample, however, usually, only one cell is present in the field of view of the tomogram as areas with multiple cells have thick ice and therefore not amenable for cryo-ET imaging. We added the following to the revised manuscript:

“As usually only one (or part of a) cell is present in the cryo-tomogram, we can’t exclude that differences in the extracellular environments, like the presence of a cluster of cells in the vicinity of the individual cells with pearling tubes, might play a role in this observation” (Lines 198-201).

"Randomly-located complexes" in this reviewers opinion should actually be described "seemingly randomly-located complexes" given there may be an organization present that is beyond the resolution limit of this study.

The is a good point. Indeed, we can’t exclude that these complexes have a preferred localization in specific lipid patches that we can’t detect in our cryo-tomograms. We added the following statement to the revised manuscript:

“These complexes, which were also found in the OM of intact cells, did not exhibit a preferred localization or regular arrangement within the tube at least within the fields of view provided by our cryo- tomograms (Fig. 5a & b).” (lines 227-230).

In reference to lines 287-292 - is it possible this has to do with lipid composition? Have the authors considered this? Please discuss if relevant/insightful.

Done. We added the following to the revised manuscript:

“In addition, differences in the lipid compositions among the various species investigated here might also play a role in the formation of these different forms of projections” (Lines 299-301).

Reviewer #2 (Significance ):

These results advance the field by shedding new light on bacterial membrane extension morphologies. The authors use a cryo-ET to catalogues membrane extensions and membrane vesicles which has not been done before.

This paper is likely to be of interest to structural biologists, biophysicist, membrane protein biologists, virologists and microbiologists.

This reviewer is a single-particle cryo-EM structural biologist with interest in membrane proteins._

We thank the reviewer for her/his enthusiasm about our work described here.

trails

Reviewer #3 (Public Review):

The biochemical and genetic characterization of BRCA2 has been an ongoing challenge in the DNA repair field as the protein is large, prone to degradation, and expressed at low levels in most cell types. While certain features of BRCA2 have been described previously including its ability to bind and load RAD51 onto resected DNA substrates, much remains to be discovered. In this study, the authors combine genetic studies in mouse ES cells with biochemical analysis to examine the spatial dynamics and molecular architecture of BRCA2. Notably, they utilize an innovative approach coupling endogenous tagging of mouse BRCA2 with a HALO tag to monitor BRCA2 movement within live cells by single particle tracking.

I applaud the authors for achieving a highly technical approach to epitope tagging both endogenous BRCA2 alleles in mouse ES cells and combining this strategy with a HALO tag providing additional utility for a variety of cell biological experiments. By analyzing the endogenous alleles, the authors' system provides physiological levels of protein expression as transcription will be driven by the endogenous promoter thus preserving stoichiometric protein interactions within the cell and avoiding artifacts caused by overexpression.

The authors determine the influence of the DNA binding domain (DBD) and c-terminal binding (CTD) on the dynamic activities of BRCA2. They begin by exposing cells containing 3 different deletion mutants ΔDBD, ΔCTD, and the double mutant ΔDBDΔCTD to four different types of DNA damage (IR, PARPi, MMC, and cisplatin). Notably, ΔDBD displays significant impairment in survival in response to all 4 types of DNA damage. The ΔCTD, in contrast, demonstrates less sensitivity to IR and Olaparib, however, complements as well as WT BRCA2 in response to crosslinking agents MMC and cisplatin. My only criticism in this aspect of the work is that it would have been informative to include a truncated BRCA2 (mimic of a patient pathogenic mutation) or null allele to compare to the survival of the ΔDBD and ΔCTD mutants. I realize that these alleles may be inviable but the authors should clearly state if that was indeed the case.

The authors then go on to demonstrate that the ΔDBD and ΔCTD mutants are recruited to sites of IR damage in a similar manner to WT BRCA2 based on number and intensity of foci. I think it would be informative if the authors provided statistical significance for the graphs depicting the quantitation of foci number and intensity as there do appear to be differences between the mutants and the WT protein. There appears to be a delay in the kinetics of recruitment, especially at the 2 hr timepoint, for the mutants compared to WT BRCA2, which could indicate a defect in the recognition of the DNA damage. Only at the 2 hr timepoint following IR are there less RAD51 foci, and of a lesser intensity, in the three deletion mutants compared to WT BRCA2. Another possibility is the results could be interpreted as a defect in RAD51 loading and/or stabilization of the nucleoprotein filament. While immunofluorescence imaging of DNA repair foci have become common practice to measure protein recruitment to damage, it is impossible to know exactly what is happening in these foci with any granularity.

Next, the authors measure BRCA2 movement in the mouse ES cells taking advantage of the HALO tag to track single particles. While technically and visually alluring, it is difficult to extract mechanistic insight from the results. DNA damage induces changes in diffusion leading to BRCA2 molecules with restricted mobility; the authors demonstrated this phenomenon in a prior publication. The deletion mutants appear to have little effect upon BRCA2 mobility.

Finally, the authors utilize scanning force microscopy to analyze binding of the purified human BRCA2 proteins to RAD51 and ssDNA. In the absence of RAD51/ssDNA binding, there is a notable shift in the deletion mutants from oligomeric forms to monomeric compared to full length WT BRCA2. Upon binding to RAD51, there is a dramatic change from multimeric to monomeric forms for the WT BRCA2 (~7% to 74%) with a slight suppression of these changes shown for the deletion mutants. While WT BRCA2 forms extended molecular assemblies upon binding ssDNA, not surprisingly, deletion of the DBD or CTD fail to demonstrate any significant changes in physical architecture. In both situations, the mutant proteins respond to RAD51 and ssDNA in a dampened manner likely due to altered or loss of binding. While the architectural effects of RAD51 and ssDNA binding to BRCA2 are measurable by SFM, it is difficult to reconcile these changes in shape and oligomerization to defects in response to DNA damage and at which specific steps in homologous recombination these physical forms would impact.

Strengths:

1. Generation of mouse ES cells with both endogenous alleles of BRCA2 containing the deletion mutations in addition to a HALO tag is an incredible technical breakthrough and will be a highly valuable reagent for genetic and cell biological studies of mouse BRCA2.<br> 2. The deletion mutants ablating either the DBD or the CTD, or both, is a great genetic approach to understanding the role of these key domains in BRCA2. The response of these mutants (versus WT BRCA2 as a benchmark) to various DNA damage (IR, PARPI, MMC, cisplatin) provides interesting information delineating the roles of these two important domains in BRCA2. For example, the ΔCTD mutant is significantly sensitive to IR and Olaparib, yet complements as well as WT BRCA2 in response to the crosslinking agents MMC and cisplatin.<br> 3. The BRCA2 protein is notoriously difficult to purify and yet the authors succeeded in purifying 4 different forms of the protein for biophysical analysis. While it is difficult to interpret the various forms of BRCA2 by SFM, there are clear differences in the architecture between WT and the three c-terminal mutants. These differences are highlighted upon binding to RAD51 or ssDNA.

Weaknesses:

1. While the separation-of-function result for the CTD deletion in response to crosslinking agents MMC and cisplatin is a novel and compelling result, it would have been informative to compare the survival results and gene targeting assay using a BRCA2 null or mimic of patient mutation (truncating mutation) to see how these 3 mutants stack up against a completely non-functioning BRCA2 allele. Likely, the BRCA2 null alleles are inviable but perhaps a conditional system or truncating allele similar to a patient germline mutation would give a window into response compared to the DBD and CTD deletion mutants.<br> 2. It's not clear in the manuscript what new information we are learning about the mechanisms of BRCA2 in the single particle tracking (SPT) data. The differences in mobility between the mutants and WT BRCA2 seem minimal, but more importantly, it is not immediately clear how these data help us understand the normal cellular functions of BRCA2. No doubt, the technology and innovation to track single particle proteins in the nuclei of cells is impressive, but the authors should clearly explain how we can gain mechanistic insight from the SPT data that is presented in this manuscript.

General Comments:

It is unclear how missing the c-terminal domain (CTD) or the DNA binding domain (DBD) of BRCA2 can be interpreted as having "roles beyond delivering strand exchange protein RAD51" unless a complete biochemical workup of the deletion mutants was performed to detect any alterations in DNA binding, stimulation of RAD51 dependent strand exchange, etc... While interesting and certainly an impressive technical feat, foci imaging and single particle tracking do not provide much information on mechanism (i.e. whether BRCA2 is binding DNA and loading/nucleating RAD51).

The interpretations in the discussion are not overstated, however, I somewhat disagree with the notion that the data, as presented, clarifies the role of BRCA2 beyond its canonical functions of RAD51 loading and nucleation on resected DNA substrates. I would have liked if the authors discussed the idea that it is surprising that mouse ES cells can tolerate complete loss of the DBD, CTD, and loss of both together. Questions that should be addressed in include some of the following: Are proliferation rates compromised compared to WT cells? Are they experiencing replication stress in the absence of any exogenous damage? Further, is there something unique about mouse ES cells that may differentiate BRCA2 behavior that would be expected in somatic human cells?

It is interesting to note that many years ago Ashworth and Taniguchi published back-to-back papers in Nature (2008) describing BRCA2 reversion alleles from in vitro screens of BRCA2 mutant cells selected in cisplatin or PARPi such that some of these reversions resulted in huge deletions of the entire DBD of BRCA2, and yet, they promoted resistance to PARPi. In this context, I would much appreciate if the authors commented on their findings that their constructed DBD deletion is not resistant to PARPi and if they offered some speculation as to why the reversions in those previous studies were.

Author Response:

Reviewer #3 (Public Review):

[...] I have only minor concerns regarding sources of error, particularly with respect to interpretation of the small effects the authors observe in many of their FRET experiments.

• Figure 2D shows rather small changes in ΔF/F-15 mV between fluorescent protein labels inserted at different positions in the ASIC sequence, particularly for the YFP constructs. As this metric is determined from the top and bottom asymptotes for the Boltzmann fits shown in Figure 2C, it would be useful to have some estimate as to the error associated with the fits at extreme values. Perhaps the authors could provide fits to their data (as in Figure 2C), including confidence intervals, or some similar estimate as to the size of the expected error compared to the effect size in Figure 2D.

Thank you for this point. We did use Boltzman’s fits to get the asymptotes for each cell and calculate a ΔF/F. However, we could also use a ‘fit free’ approach of simply taking the difference between fluorescence values measured at -180 mV and that at +120 mV, divided by that at -15 mV to normalize for each cell. This approach completely avoids any error associated with fitting the data or imposing any model at all. Using this approach results in slightly different ΔF/F values but the pattern of statistical significance is identical. This new analysis is included in Figure 2 figure supplement 4. It has also been corrected for multiple comparisons.

• Along those same lines, the authors use an interesting (and potentially generalizable) approach to reducing background from intracellular proteins in their experiments: co-transfecting their channels with empty plasmid DNA. What percentage of the remaining fluorescence signal is the result of intracellular background? How would that affect the data in Figure 2 and 3? Is the ΔF/Fnorm curve for YFP labeled positions in Figure 2-figure supplement 4 so flat because of contaminating background fluorescence?

This is a great question. We originally hoped that the CFP and YFP quenching data from different positions could be used to triangulate both a distance from the membrane and a value for background fluorescence assuming that CFP and YFP would yield similar background fluorescences. An analogous approach was used in Zachariassen et al. Proc Natl Acad Sci, 2016 where an equal background was assumed between conformational states within a recording. In the end, the YFP quenching appeared to have a greater background than CFP. We speculate that this may be because the YFP variant we used matures faster than the CFP (mVenus, 17.6 min verses mTurquiose2, 33.5 min; FPbase.org) and hence the YFP matures faster than the ‘new’ channels get to the plasma membrane. However, at present we are uncertain how much of the background fluorescence signal to confidently attribute to this intracellular FP issue.

• In Figure 3D, the FRET efficiency between CFP-cA1-cA1 and N YFP at a 1:15 ratio of the two plasmids is higher than the FRET efficiency between CFP and YFP in the same subunit, even though the authors conclude that fluorescent proteins on the same subunit show considerably more FRET than fluorescent proteins on neighboring subunits. Could this indicate that the N-termini of adjacent subunits are closer together than the N- and C-termini of a single subunit? If, on the other hand, this effect were entirely the result of crowding in the membrane why is FRET efficiency substantially lower when CFP-cA1-cA1 is co-expressed with C4 YFP? Wouldn't this construct produce a similar crowding effect?

We strongly suspect the N termini of adjacent subunits are closer to each other than N and C of single subunit simply because the N FPs would all be at the same ‘height’ or same depth with respect to the plasma membrane. Thus the measured FRET in this case primarily reflects distances in the x-y plane. This contrasts with the N and C FPs on the same or different subunits where both x-y distances and axial distances come into play.

• On page 23, the authors state that they detected no pH-dependent changes in FRET between their GFP tag on the N-terminus of ASIC1 and an RFP tag on the channel's C-terminus. However, Figure 4 shows a small, but significant change in fluorescence between pH 8 and pH 7.

We have corrected for multiple comparisons within a figure. As a result, this effect is no longer statistically significant (adjusted p value is 0.063).

• The interpretation of distances between various tagged position on ASIC and the plasma membrane in Figure 2 is based on using two different colored tags with two different distance dependences. However, the interpretation of the data from Figure 5 provided on page 25 is less clear. For example, the reduction in fluorescence from the N-terminal tag is interpreted as the tag moving closer to the plasma membrane. Without similar data from a YFP tag to verify, it seems equally likely that the reduction in fluorescence (at steady state) could result from a movement away from the plasma membrane.

This is a very good point. We tried to perform DPA quenching of YFP-containing constructs at pH 6.0, but the acidification resulted in proton-quenching of the YFP fluorescence (Figure 4). We didn’t feel confident in measuring DPA quenching with the concomitant loss of YFP fluorescence due to acidification. Therefore, we relied on the pH 8.0 CFP and YFP data as a starting point (Figure 2). Given the C1 insertion gives the greatest extent of CFP quenching, it is reasonable to place it around the top of the curve. The N position could then be on the left or right side of the hump or peak in the CFP distance curve. The N quenching is comparable to the C2 insertion quenching (Figure 2D, left) yet the N FP is ~ 16 amino acids from the pore-forming membrane helices while the C2 insertions is ~ 40 amino acids away. For reference, the C1 is ~ 24 amino acids. Thus we are reasonably confident the N insertion is on the left side of the hump or peak. A reduction in ΔF/F would indicate movement closer to the plasma membrane. While technically possible that the N position could move further away from the membrane, this would have to be a >25 Å movement. Given there are only 16 amino acids between the CFP and the beginning of TM1 of the channel, we do not think such a dramatic movement outward could occur.

interesting narrative shaping here.

scientifically, yes our understanding of our autonomic nervous system is super important to the healing the things that are wild and disturbing above — but jumping to the next part is where I get lost.

Agree a lot of our pain is from: "is the belief that we do not belong in this world." which I think we'll have to work on to fix the climate thing.

Author Response

Reviewer #1 (Public Review):

[...] My main technical concern lies in the choice of decomposition filter for SEP and alpha oscillations, and the conclusions the authors draw from that. Specifically, a CCA spatial filter is optimized here for the N20 component, which is then identically applied to isolate for alpha sources, with the logic being that this procedure extracts the alpha oscillation from the same sources (e.g., L359). I have no issues (or expertise) with using the CCA filter for the SEP, but if my understanding of the authors' intent is correct, then I don't agree with the logic that using the same filter isolate for alpha as well. The prestimulus alpha oscillation can have arbitrary source configurations that are different from the SEP sources, which may hypothetically have a different association with the behavioral responses when it's optimally isolated. In other words, just because one uses the same spatial filter, it does not imply that one is isolating alpha from the same source as the SEP, but rather simply projecting down to the same subspace - looking at a shadow on the same wall, if you will. To show that they are from the same sources, alpha should be isolated independently of the SEP (using CCA, ICA, or other methods), and compared against the SEP topology. If the topology is similar, then it would strengthen the authors' current claims, but ideally the same analyses (e.g., using the 1st and 5th quintile of alpha amplitude to partition the responses) is repeated using alpha derived from this procedure. Also, have the authors considered using individualized alpha filters given that alpha frequency vary across individuals? Why or why not?

Indeed, applying the same spatial filter to EEG signals with different spatial arrangements of the sources can lead to the extraction of neuronal activity which does not originate from the very same sources. We had chosen our approach, as it is well known that the generators of the early SEP components and the generators of the prominent somatosensory alpha rhythm co-reside at similar sites in the primary somatosensory cortex (e.g., Haegens et al., 2015). Therefore, we considered our approach appropriate to specifically focus on neural activity from the somatosensory region both in the frequency band of the SEP as well as of the alpha rhythm. Yet, we agree with the reviewer that it should be acknowledged that we may have missed or mixed-up effects of alpha activity from other sources by using this procedure (which might have led to different conclusions otherwise). In order to account for this, we repeated our analyses with an SEP-independent reconstruction of the oscillatory effects in source space (“whole brain analysis”). For this, we first reconstructed the sources of alpha activity using eLORETA and head models based on participant-specific MRI scans, and estimated the respective effects independently for all sources across the cortex using both linear-mixed effects models (LME) as well as a binning approach for the Signal Detection Theory (SDT) parameters sensitivity d’ and criterion c (consistent with the previous analyses in our manuscript). In the LME analyses, both the effects of pre-stimulus alpha activity on N20 amplitudes as well as on perceived stimulus intensity were strongest in the right primary somatosensory cortex – in accordance with the sources of the originally extracted tangential CCA component of the SEP (see Supplementary Figure 1 for Peer Review). Also, using the binning approach to examine the relation or pre-stimulus alpha activity with SDT parameter criterion c, the effects were most pronounced around the right somatosensory regions (Supplementary Figure 2 for Peer Review), yet these effects did not survive statistical correction for multiple comparisons (FDR-correction with p<.01). However, when performing the same binning analysis for our region of interest (ROI), the hand area in BA 3b of the right somatosensory cortex, a significant effect or pre-stimulus alpha on criterion c was indeed confirmed, t(31)=-2.951, p=.006, CI95%=[-.173, -.032]. Furthermore, in line with our previous CCA results, for sensitivity d’, neither the whole brain analysis nor the ROI analysis showed effects of pre-stimulus alpha amplitude, t(31)=0.633, p=.531, CI95%=[-.083, .157]. Taken together, the findings we report in our original manuscript for pre-stimulus alpha activity obtained with the spatial CCA filter can thus be replicated with a SEP-uninformed source reconstruction, both using LMEs for a “whole-brain analysis” as well as SDT analyses in a ROI-based approach. We therefore conclude that the relationships between pre-stimulus alpha activity, N20 potential of the SEP, and perceived stimulus intensity can indeed be attributed to neural activity from the same (or at least very similar) sources in the primary somatosensory cortex.

Addressing the question on filtering alpha activity in individualized frequency bands, we considered this option, too. However, the rather short length of our pre-stimulus window (-200 to -10 ms) constitutes a natural limit for the frequency resolution in the alpha range and slightly different filter ranges (adjusted with regards to the individual alpha peak frequency) are thus unlikely to lead to large differences in the estimation of pre-stimulus alpha amplitudes. Therefore, we refrained from using individualized frequency bands here and focused on the more generic approach using one common alpha band (8-13 Hz) for all participants, which should also facilitate direct comparisons with previous studies on pre-stimulus oscillatory effects.

In the same vein, both alpha and N20 amplitude relate to perceptual judgement, and to each other. I believe this is nicely accounted for in the multivariate analysis using the SEM, but the analysis that partitions the behavioral responses using the 20% and 80% are done separately, which means that different behavioral trials are used to compute the effect of N20 and alpha on sensitivity and criterion. While this is not necessarily an issue given that there IS a multivariate analysis, I would like to know how many of those trials overlap between the two analyses.

This is an interesting point indeed. We included both the binning analyses and the multivariate analyses in our manuscript as we believe they offer complimentary views on the data, and also allow a direct comparison to previous studies in the field (e.g., Iemi et al., 2017). In fact, the trial overlap between the extreme bins of the alpha and N20 data were rather small.

Since the expected trial overlap is 20% when partitioning the data into quintiles randomly, the effect-driven increments and reductions in trial overlap in our data appear to be rather small. However, they showed the expected directions: Larger alpha amplitudes were associated with more negative N20 amplitudes (and vice versa). Presumably, these small differences in trial overlap reflect the rather small effect sizes we also observed in the multivariate analyses. We have added this information to our revised manuscript in the following way to give the reader a better picture of the underlying data for the binning analyses (page 9, lines 137 ff.): “(Please note that this procedure resulted in a different trial selection as compared to the SDT analysis of pre-stimulus alpha activity. Please refer to Fig. 2—figure supplement 2 for further details on the trial overlap.)”

At multiple points, the authors comment that the covariation of N20 and alpha amplitude in the same direction is counterintuitive (e.g., L123-125), and it wasn't clear to me why that should be the case until much later on in the paper. My naive expectation (perhaps again being unfamiliar with the field) is that alpha amplitude SHOULD be positively correlated with SEP amplitude, due to the brain being in a general state of higher variability. It was explained later in the manuscript that lower alpha amplitude and higher SEP amplitude are associated with excitability, and hence should have the opposite directions. This could be explicitly stated earlier in the introduction, as well as the expected relationship between alpha amplitude and behavior.

Thank you for pointing out this unclarity. We have now made this rationale more explicit already at an early point in the introduction (page 3, lines 26 ff.): “According to the baseline sensory excitability model (BSEM; Samaha et al., 2020), higher alpha activity preceding a stimulus indicates a generally lower excitability level of the neural system, resulting in smaller stimulus-evoked responses, which are in turn associated with a lower detection rate of near-threshold stimuli but no changes in the discriminability of sensory stimuli (since neural noise and signal are assumed to be affected likewise).”

Furthermore, I have a concern with the interpretation here that's rooted in the same issue as the assumption that they are from the same sources: the authors' physiological interpretation makes sense if alpha and N20 originated from the same sources, but that is not necessarily the case. In fact, the population driving the alpha oscillation could hypothetically have a modulatory effect on the (separate) population that eventually encodes the sensory representation of the stimulus, in which case the explanation the authors provide would not be wrong per se, just not applicable. A comment on this would be appreciated in the revision.

Our extensive additional analyses suggest that the sources of behaviorally relevant alpha and N20 activity were located at very similar cortical sites. Nevertheless, this is not a proof that exactly the same neuronal populations were involved (for example, alpha and N20 effects could originate from different cortical layers). Therefore, we have added this potential limitation to our revised manuscript in the following way (page 19, lines 379 ff.): “Furthermore, with the present data, we cannot unambiguously conclude that the observed relation between pre-stimulus alpha activity and initial SEP indeed involved the very same neuronal populations – which may represent a limitation of the hypothesized mechanism. However, all approaches to localize these effects pointed to very similar cortical regions as discussed in the following section.”

In addition, given how closely related the investigation of these two quantities are in this specific study, I think it would be relevant to discuss the perspective that SEPs are potentially oscillation phase resets. Even though the SEP is extracted using an entirely different filter range, it could nevertheless be possible that when averaged over many trials, small alpha residues (or other low freq components) do have a contribution in the SEP. If the authors are motivated enough, a simulation study could be done to check this, but is not necessary from my point of view if there is an adequate discussion on this point.

Indeed, the phase reset mechanism may be a possible alternative explanation for relations between oscillations and later parts of the ERP. However, the N20 potential reflects the very first excitation of the cortex in response to a somatosensory stimulus and should therefore represent a textbook example of an additive response (EPSPs are added to ongoing background activity). Moreover, the N20 response should be over long before a possible phase reset in lower frequencies (such as alpha frequencies) would start to play a role (Hanslmayr et al., 2007; Sauseng et al., 2007). Nevertheless, we ran additional control analyses (including a simulation study) in order to exclude that some odd combination of phase-locking and filter residues led to the present findings: Please see Essential Revision #4 for details and how we included these considerations in our revised manuscript.

Reviewer #2 (Public Review):

[...] The main weaknesses of the manuscript becomes most apparent with respect to the stated impact that "The widespread belief that a larger brain response corresponds to a stronger percept of a stimulus may need to be revisited.". I am not really sure if there are many cognitive neuroscientists, that would actually subscribe to such a simplistic relationship between evoked responses and perception and that temporal differentiation (early vs late responses) and the biasing influence of prestimulus activity patterns are becoming increasingly recognized. So rather than actually changing a dominant paradigm, this work is an (excellent) contribution to a paradigm shift that is already taking place.

Thank you for this feedback. We agree that the paradigm shift away from simplistic assumptions about the relationship between variability of neural responses and perception is already taking place and that this is already being appreciated by many scientists in the field. Also, we agree that the present study contributes more evidence to this emerging notion rather than changing the whole field. However, we do think that particularly the observation of opposite amplitude modulations of initial somatosensory evoked responses associated with presented stimulus intensity on the one hand and pre-stimulus excitability state on the other, provides a novel perspective for our understanding of how fundamental features of sensory stimuli are processed at initial cortical levels. Following your suggestions to tone down claims about the controversiality as well as to avoid over-generalization, we have therefore adjusted the impact statement of this manuscript to: “Larger evoked responses during initial cortical processing may reflect states of lower excitability.”

Furthermore, we have adjusted similar statements throughout the manuscript accordingly.

Also it should be considered that with regards to the analysis approach using CCA, the claims are mainly restricted to BA3b: i.e. while I also think that this is a strength of the current study, one should refrain from overinterpreting the results in a very generalized manner. The authors do include some "thalamus" and "late" evoked response patterns as well, however that presentation of the results is somewhat changed now as compared to the N20 (e.g. using LMEs rather than comparison of extremes; not using SEMs). The readablity of results and especially the comparison of effects would profit from a more coherent approach.

We agree that our findings indeed have the specific focus on the N20 component and thus on its generators in BA3b. We did not intend to suggest that the effects we observed for this initial cortical response can be readily generalized to other (later) ERP components, too. However, we do believe (and hypothesize) that similar mechanisms may be in place for corresponding initial cortical responses in other sensory modalities, too – yet it is clear that we cannot test this generalization with the current study. To avoid misunderstandings of these interpretations and their limitations, we have further specified these aspects in the Discussion.

Regarding our analyses of the later SEP (i.e., N140 component) and thalamus-related activity (i.e., P15 component), we initially decided to use linear-mixed effects models as they are mathematically equivalent to the way the sub-equations of the structural equation model were constructed (Table 2 in the manuscript). Nevertheless, we have now additionally run binning analyses to make a direct comparison also with Signal Detection Theory (SDT) parameters possible: For the N140 component, there was a significant effect on criterion c, t(31)=-3.010, p=.005, but no effect on sensitivity d’, t(31)=0.246, p=.807. For the P15 component, no effects emerged either for criterion c or sensitivity d’, t(12)=1.201, p=.253, and t(12)=-0.201, p=.844, respectively. These findings correspond well to the previous LME analyses and may indeed further facilitate the comparison with the findings for the N20 potential and pre-stimulus alpha activity. Therefore, we have added these complimentary analyses to our manuscript in the following way:

Results: “In addition, the SDT analysis based on binning of the P15 amplitudes into quintiles neither suggested a relation with criterion c nor with sensitivity d’, t(12)=1.201, p=.253, and t(12)=-0.201, p=.844, respectively.” (page 14, lines 241 ff.)

“These findings were in line with a separate SDT analysis: N140 amplitudes were associated with an effect on criterion c, t(31)=-3.010, p=.005, but no effect on sensitivity d’ emerged, t(31)=0.246, p=.807.” (page 15, lines 263 ff.)

Discussion: “Crucially, our data are at the same time consistent with previous studies on somatosensory processing at later stages, where larger EEG potentials are typically associated with a stronger percept of a given stimulus (e.g., Al et al., 2020; Schröder et al., 2021; Schubert et al., 2006), as both our SDT and LME analyses of the N140 component showed.” (page 19, lines 367 ff.)

“Yet, neither our SDT analyses nor the LME models of the thalamus-related P15 component supported this notion.” (page 21, lines 414 ff.)

Methods (page 32, lines 681 ff.): “The effects of the EEG measures pre-stimulus alpha amplitude, N20 peak amplitude, P15 mean amplitude, and N140 mean amplitude on the SDT measures sensitivity d’ and criterion c were examined using a binning approach: […]”

I have some concerns whether the relationship between large alpha power and more negative N20s could be driven by more trivial factors rather than the model explanations the authors develop in the discussion. Concretely the question whether phase locking of large alpha power along with >30 Hz high pass filtering could produce a similar finding as shown e.g. in Figure 2c. This is an important issue, as prestimulus alpha influences the N20 amplitudes as well as the perceptual reports.

Indeed, potential phase-locking of alpha oscillations to stimulus onset and filter-related effects are important issues that could potentially offer an alternative explanation for the observed relationship between amplitudes of pre-stimulus alpha activity and the N20 potential of the SEP. Although such pre-stimulus alpha locking is rather unlikely in a paradigm with jittered stimulus onsets (in our case uniformly distributed between -50 ms and +50 ms; corresponding to a whole alpha cycle), we have run the following control analyses to fully exclude this possibility:

First, we analyzed whether pre-stimulus alpha phase values were distributed uniformly and whether these phase distributions differed between high and low alpha amplitudes as well as between high and low N20 amplitudes. The phase of pre-stimulus alpha activity was obtained from a Fast-Fourier transform in the pre-stimulus time window from -200 to -10 ms, applied to unfiltered, but otherwise identically pre-processed data as in the original manuscript (i.e., applying the spatial filter of the tangential CCA component). For the FFT, we used zero padding (extending the pre-stimulus data segments to 2048 data points each) in order to obtain an interpolated frequency resolution of around 3 Hz. The phase was extracted at the frequency 9.766 Hz (i.e., the closest available frequency to 10 Hz). As visible from Supplementary Figure 3 for Peer Review, pre-stimulus alpha phases were distributed uniformly across all five quintiles of both alpha and N20 amplitudes. This observation was confirmed by the Rayleigh test (testing for deviations from a uniform distribution; Berens, 2009): Neither in the concatenated phase data of all participants, z=1.130, p=.323, nor in single-participant analyses within every alpha amplitude or N20 amplitude bin, we found evidence for a non-uniform distribution of alpha phase, all p>.367 (after Bonferroni correction for multiple testing). Thus, there was no phase-locking of pre-stimulus alpha activity that could serve as a trivial alternative explanation of the relationship between pre-stimulus alpha amplitude and N20 amplitude.

Second, in order to examine whether the combination of our temporal filters (30 to 200 Hz band-pass for the SEP, and 8 to 13 Hz band-pass for alpha activity) could have led to the present findings, we additionally re-ran our analysis pipeline with simulated data: We mixed exemplary SEP responses with constant amplitudes (unfiltered; derived from within-participant averages), with simulated alpha band activity with randomized amplitude fluctuations, and pink noise, reflecting neural background activity as is typical for the human EEG. The SEP onsets were chosen according to our original experimental paradigm with inter-stimulus intervals of 1513 ms and a jitter of ±50 ms. Next, we filtered these mixed signals between 30 and 200 Hz in order to extract the single-trial SEPs, and estimated the pre-stimulus alpha amplitudes between -200 and -10 ms in the same way as was done in the original manuscript (i.e., by filtering the mixed signal between 8 and 13 Hz). This procedure was repeated for 32 generated data streams, containing 1000 SEPs each (corresponding to our empirical dataset of 32 participants). The resulting average SEPs did neither show a visually detectable difference between the five alpha amplitude quintiles nor indicated a random-slope linear-mixed-effects model any relation between pre-stimulus alpha amplitude and N20 amplitude on a single-trial level, βfixed=-.0005, t(255.16)=-.094, p=.925. Therefore, our findings cannot be explained by filter artifacts or residual activity leaking from the alpha frequency band to the frequency band of the N20 potential.

Third, we re-analyzed our empirical EEG data in time-frequency space to obtain a more detailed view of the effects of pre-stimulus alpha activity on N20 amplitudes. For this, we decomposed our pre-processed but unfiltered data with wavelet transformation (complex Morlet wavelets) and calculated linear-mixed effects models on the relation between signal amplitudes in the time-frequency domain and single-trial N20 amplitudes as obtained from our original analyses. As shown in Supplementary Figure 5 for Peer Review, the time-frequency representations of the effects on N20 amplitudes indeed indicated a specific role of the alpha band, with its effects (i.e., already 200 ms before stimulus and in the upper alpha frequency range) separated from the time- and frequency range of the N20 potential of the SEP (i.e., from ~20 ms after stimulus onwards and above ~20 Hz). In addition, we ran the same analysis for the behavioral effect (i.e., perceived stimulus intensity). Also here, pre-stimulus effects were predominantly visible in the alpha band. Of note, there were also strong effects in the beta band. These may be interesting to study further in future studies – in particular, whether they reflect independent physiological processes or rather harmonics of the alpha band. Furthermore, these time-frequency representations suggest that the studied pre-stimulus effects might have been even more pronounced if we had analyzed the data in pre-stimulus time windows from -300 to -10 ms. However, in order to avoid inflating effect sizes by post-hoc data digging (“p-hacking”), we prefer to keep the original, a priori chosen time window for the main analyses of the manuscript. Yet, these onsets of pre-stimulus effects at around -300 ms may be of interest for future work. Taken together, these time-frequency analyses further support the notion that the observed relation between pre-stimulus alpha activity and N20 amplitudes is not due to technical issues (such as filter leakage and phase-locking) but rather reflects genuine neurophysiological effects of alpha oscillations on SEPs.

We have added the time-frequency analysis, as well as the SEP simulation analysis as figure supplements to Figure 2 in our revised manuscript (page 8) since we believe that these control analyses comprehensively show that the observed effects were (a) specific to the alpha band and (b) not due to any data processing-related artifacts.

It is important to emphasize that the model develop is a post-hoc one, i.e. the authors do not develop already in the discussion various alternative scenario results based on different model predictions. Therefore there is no strong evidence in support of the specific one advanced in the discussion.

Thank you for raising this issue. Indeed, we cannot prove with the current findings that our proposed physiological model of the relation between alpha oscillations and the SEP is the correct model (or that it is at least the best one out of a selection of possible alternative models). To do so, future studies would be needed that can actually directly measure and/or manipulate differences in membrane potentials and trans-membrane currents. Rather, we aimed with the present study to associate a physiological meaning with the concept of excitability changes in the human EEG – offering a hypothesis that may be worthwhile to be studied (and either confirmed or rejected) in future studies. We have tried to make this motivation more explicit in the Discussion section (page 20, lines 384 ff.): “Also, we would like to emphasize that the presented mechanism reflects a hypothesized model, which shall be further supported or falsified with more targeted studies, for example, directly quantifying membrane potentials and trans-membrane currents in relation to different excitability states in somatosensation.”

Author Response

Reviewer #1 (Public Review):

[...] The manuscript is excellently written and discusses the simulation results clearly and succinctly. The resolution of the simulations is very impressive and yields unprecedented insight into the effect of merozoite shape on alignment dynamics, which has important implications for how effectively the parasite can survive and multiply. The conclusions reached by the authors are certainly justified by the simulation data. In particular, the authors are careful not to draw conclusions beyond the limits of their study, and acknowledge other factors which may influence the merozoite shape, such as internal structural constraints and the energy of invasion following successful alignment.

We thank the reviewer for a thorough reading of our manuscript and the very positive judgement.

Regarding weaknesses of the manuscript, some of the explanations of the trends observed in the simulation data could be expanded slightly, to help gain a deeper understanding of the competition between adhesion and RBC deformability underlying the alignment dynamics. These are described in more detail below.

1. Line 114 and lines 120-129: The discussion here of the trends observed in Figure 1 (including why the LE shape has a larger energy compared to the OB shape despite having a smaller adhesion area) is somewhat vague and should be developed further. For example, currently there is only a video showing the egg-like shape and a second video comparing the LE shape to a spherical shape - it would be helpful to have a further video comparing the LE and OB shapes and the different RBC deformations they cause. Moreover, the explanation of the energy/mobility of each shape in terms of curvatures (e.g. the OB shape having "lower curvature at its flat side") could be made more precise. I would expect that the adhesion area depends on how close the principal curvatures of the merozoite surface are to being equal and opposite to the natural curvatures of the RBC, since this determines the bending energy associated with wrapping the merozoite and forming short bonds. This would explain why the spherical shape is most mobile (its principal curvatures are constant so there is no region where at least one is relatively small), and why alignment is most likely to occur in the dimple of the RBC where the membrane is naturally concave-outward. For a given adhesion area, the deformation energy should depend on the difference in principal curvatures in the contact region, with a larger difference causing more bending of the RBC membrane. This difference is larger for the LE shape, since one principal curvature remains large at each point on the surface, compared to the OB shape whose principal curvatures are both small on the 'flat side' where contact is most likely to occur.

We have expanded the discussion of these results to make it clearer. Furthermore, a new video was generated to visually see differences between different shapes.

1. Lines 175-176: Given that the ratio A_m/A_s (adhesion area to total surface area) plays a key role in the probability of alignment, the authors should be more quantitative at this point. How does the ratio A_m/A_s (as measured directly, or indirectly e.g. by the area under the probability distributions inside the alignment region in figures 3a,b) scale with the system parameters, such as the adhesion strength and the off-rate k_off? Can it be estimated from an energy balance between RBC bending/stretching and the average adhesion energy?

A change in A_m as a function of adhesion strength can be estimated analytically for a sphere, as was done in Hillringhaus et al. Biophys. J. 117:1202, 2019. For small deformations, there is essentially a competition of bending and adhesion energies, while for strong adhesion, stretching-elasticity contribution becomes important. We have included this theoretical result into the manuscript and discuss its implications.

1. Line 197-198 and Figure 4c: Why is the deformation energy associated with the OB shape much lower than all other shapes for values of k_off/k_on^{long} smaller than 2?

For k_off/k_on^{long} < 2, the magnitude of local curvature has a pronounced effect. For the OB shape, a large adhesion area is formed over the area with very low curvature, and close to the rim where the curvature is large, the adhesion strength may not be strong enough to induce membrane wrapping and deformation. For other shapes, the adhesion strength is large enough to lead to partial wrapping of the parasite by the membrane over moderate curvatures. As a result, the integrated deformation energy is significantly lower for the OB shape than for the other shapes in this regime of adhesion strengths. We have added this clarification to the manuscript.

1. Alignment requires that the distance between the merozoite apex and RBC membrane is very small, and the alignment criteria necessitate examining small changes in the apex angle \theta from \pi. Can the authors comment on how sensitive are the results to the numerical discretisation used?

The discretization length does affect the tightness of the alignment criteria. In our simulations, the average discretization length of the RBC membrane is about l0=0.2 m. The half circumference length of a parasite (corresponding to angle ) is R, which is equal to about 12 l0 for R=0.75 m, such that our angle resolution with respect to the parasite size is 0.1. Therefore, we use 0.2 for the alignment criteria, which is large enough to avoid strong discretization effects. Simulations with a finer discretization are possible, but they become very expensive computationally.

Reviewer #2 (Public Review):

[...] A major strength of the results is that it investigates an unstudied problem in malarial pathogenesis. The results pertaining to adhesion strength may be informative for preventing the organism from invading red blood cells. A primary weakness is that there is too little detail provided in the methods for this reviewer to adequate assess the computational method. Secondly, the results are somewhat inconclusive. While the egg-shape performs better than certain other shapes, there is no clear final understanding why this shape is preferred over the spherical or short ellipsoidal shapes. However, this possibly provides some clues as to why a certain malarial species does actively adopt a spherical shape during red blood cell binding and invasion.

We thank the reviewer for a positive judgment of our manuscript. We have significantly expanded the methods section, so it should contain now all necessary simulation details. We agree with the reviewer that the conclusions about shape advantages/disadvantages are equivocal to some extent, but this is exactly what our simulation data show. However, from our data it is clear that the two shapes (i.e. egg-like and sphere) stand out, and they also correspond to real examples of merozoite shapes. As the reviewer points out, we do discuss some clues for the importance of parasite shape in the alignment process.

Overall, the authors achieved their aims by quantitatively assessing the affect of parasite shape and adhesion strength on cell alignment, which is a proxy for invasion. The discussion at the end of the manuscript provides an accurate evaluation of the results that puts them into the context of invasion. While to some extent the results presented here are inconclusive, I do think that this paper achieves an important goal for its field. This is an understudied area pertinent to a major disease. This manuscript has the potential to bring questions of the biophysics of malarial invasion out to the broader community, specifically introducing these questions to biophysicists as well as microbiologists. Furthermore, the results naturally lead to new questions. If the spherical and egg shapes do not confer a strong advantage, then these specific shapes must also play a role in other processes. The authors do suggest some possibilities in the Discussion. That their remain interesting questions is a great spur for future work.

Thank you for emphasizing the importance of multidisciplinarity. We also hope that our work will ignite interest in different communities, as only a multidisciplinary effort can bring us much closer to understanding of parasite alignment and invasion, which clearly include a combination of different mechanical and biochemical processes.

Author Response

Reviewer #1 (Public Review):

[...] Their studies were complemented by transcriptomics and metabolomics and these results support the general conclusions that pollen contains diverse carbon sources which could be used in complementary ways by the different species, which have diverse metabolic capabilities encoded in their genomes.

Reply: We thank the reviewer for the positive assessment of our manuscript.

One of the points that was not completely explored in the paper is what happens in the simplified diet both in vitro and in the Bee gut. They propose in the discussion that in the presence of few and simple carbon sources (sugars) there is competition for nutrients and competitive exclusion is driving loss of some species. But this is not fully addressed in the paper.

Reply: All four species can colonize the gut individually and grow on their own in axenic cultures when providing the simple sugars or the pollen as the only carbon source. When cultured together, all four strains are stably maintained in the presence of pollen. However, three of the four strains steadily decrease in abundance in the simple sugars. These findings are, in our opinion, consistent with the consumer-resource model (more resources = more species that can coexist) and the competitive exclusion principle which predicts that if two or more strains compete for the same nutrients they will not be able to coexist. We have added a corresponding section on line 423-425.

The system they use (with 4 closely related bacterial species) is a simplified system. Therefore, it is not clear if the same general findings will hold in more complex systems. But the results supporting that nutrient complexity (in diet) and metabolic diversity (from the microbial side) are key factors to enable co-existence and persistence of complex microbiota communities are strong and likely generalizable. Although, it is possible that with other communities and other hosts other factors will also come into play. Nonetheless, the current study is important because it sets a good example for how these questions can be addressed to study more complex systems.

Reply: It is true that bacterial coexistence does not necessarily need to be dependent on the nutrient complexity and that in other communities the host, the structure of the environment, or cross-feeding activities may play a more important role. We have discussed this point in the revised manuscript starting on line 423 and on line 427.

Overall, the study described here is complete, and rigorous, except for a few points that still need to be addressed and clarified. Namely, it would be interesting to understand what drives exclusion of some members of the community in the simplified diet.

Reply: See our reply above.

Importantly, the current study opens the door for new studies (including in vitro studies) on the identification of network interactions that are important for Microbe-Microbe interactions that enable co-existence in other systems. Additionally, this study also highlights the importance of identifying the relevant nutritional (and metabolic) conditions for addressing those questions given the importance of the metabolic context in shaping microbe-microbe interactions.

Reply: Thank you. We agree.

Reviewer #2 (Public Review):

[...] Strengths: The use of community profiling, transcriptomics, and metabolomics adds depth, as does the comparison of defined culture conditions to the host environment. The main conclusions drawn by the authors is that the presence of pollen is necessary for gut species to coexist, and that the different species, although closely related, respond in distinct ways to nutrients in pollen and consume different profiles of nutrients from pollen.

Reply: We thank the reviewer for the positive feedback and the many valuable comments which helped us to further strengthen our manuscript.

Weaknesses: The main weakness I see with this work is the choice of in vitro comparison conditions. The strains are cultured either on pollen or sugar water, whereas in vivo bees are fed a diet of pollen and sugar water, or only sugar water. A direct comparison is possible between the strains grown on sugar water in vitro or in vivo, but I think that in several places, the authors may have to reconsider or modify their interpretations comparing in vitro culture on pollen/pollen extract with the in vivo growth of the community on pollen and sugar water. Because there is sugar in the bee diet, differences in assembly dynamics, transcription, or metabolite consumption between pollen-containing culture conditions and the bee gut might stem from the dietary intake of sugar, or from an aspect of the host environment.

Reply: We agree with the reviewer that the nutrient conditions that were used in vitro and in vivo are not identical and may have impacted the relative abundance of some of the community members, the transcriptional profiles, or the metabolite changes. Nevertheless, we believe that our experimental design is valid to test the main hypothesis of our study, i.e. a complex, pollen-based diet facilitates coexistence, while simple sugars lead to the dominance of a single strain independent of the environment (culture tube versus host). An important point to consider here is that bees will pre-digest the consumed pollen, and partially absorb dietary nutrients such as amino acids, glucose, and fructose, before they reach the bacteria in the hindgut. Consequently, the in vivo and in vitro conditions will never be the same even if we would have used the identical nutrients in our treatments. Also, pollen by itself contains glucose, fructose, and sucrose. So, although we have not added glucose to the in vitro pollen condition, this simple sugar was present in the corresponding condition. We have added a corresponding section in the discussion on line 402-422. This said, while we cannot recapitulate the exact same nutritional conditions in vitro, we still think that our main conclusions hold which is that we can recapitulate the pollen-dependent coexistence found in vivo.

Reviewer #3 (Public Review):

[...] Overall, the paper is strong and the arguments and conclusions put forth are well supported by the data. I only have a few suggestions:

Reply: We thank the reviewer for the positive evaluation of our manuscript.

1) The study focuses on one strain each of the 4 Firm-5 species; however, there is diversity within each species. This is only briefly mentioned in the paper at the very end, and I think the authors should address this a bit more directly. In particular, they have previously generated a large amount of genomic data from some of these other strains, so it is likely possible to infer or speculate, based on this data, whether they expect different strains within each species to utilize similar nutrients. Also, I'm wondering if the authors can comment on how their findings could extend to the related bumble bee gut microbiome. Such a discussion would help enhance the applicability and importance of this study.

Reply: We agree that the large amount of strain-level diversity within a given species is an important point to consider. However, we would like to not expand this point much further as it would require a relatively complex genomic analysis. Also, considering that many of the strain-specific transcriptional changes are in genes shared with the other species, I am not sure how much such an analysis would reveal. Anyways, we plan to compare the coexistence between strains from the same versus another lineages in a follow-up study.

As for the bumble bees, we currently do not know how many strains or species of Lactobacillus Firm5 can coexist in bumble bees. Therefore, we feel that a discussion extending to bumble bees would be too speculative. However, we included a sentence in the discussion which states that since pollen facilitates coexistence, it follows that dietary differences are likely to influence the diversity of Lactobacillus Firm5 and give the example of the Asian honey bee, which seems to only harbor one species of this phylotype. See line 479-488.

2) It is interesting that different species ended up dominating in the in vivo vs. in vitro simple sugar-based communities. What do the authors think may be behind this difference?

Reply: This is indeed an interesting point. We have not used the same sugars in vivo (sucrose) and in vitro (glucose). Moreover, the nutritional and physicochemical conditions in the hindgut are likely different from those found in a culture tube. We have mentioned that these are potential reasons for the observed differences in the relative abundance of different community members between in vivo and in vitro conditions on line 402-422 of the manuscript.

3) Since the observed coexistence of these gut microbes is largely due to nutritional niche partitioning, it would be helpful if the authors can comment on the natural variation of key pollen derived metabolites, and if/how we could expect ecological variation in the bee microbiome due to plant pollen availability based on biogeography and seasonality.

Reply: We agree and have included a corresponding sentence in the discussion on line 479. See also our reply to point 1.

4) The supplementary information is nicely documented and accessible, but I think it would be even more useful if genome-wide data for the RNA-seq results, not just for select genes, are made available. Furthermore, I suggest including descriptive titles and labels within the supplementary Excel files, as there are many separate sheets and it is not always clear what each one shows.

Reply: This has been included in the revised manuscript.

Identity, Mexican;

Time in the US, Immigration Status, Being secretive, Hiding/lying, In the shadows, Living undocumented; Reflections, The United States, US government and immigration; Feelings, Frustration; Time in the US, Jobs/employment/work, Documents, Driver's license, Social security card/ID

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers