Pan-cancer ion transport signature reveals functional regulators of glioblastoma aggression

Ion channels, transporters, and other ion-flux controlling proteins, collectively comprising the “ion permeome”, are common drug targets, however, their roles in cancer remain understudied. Our integrative pan-cancer transcriptome analysis shows that genes encoding the ion permeome are significantly more often highly expressed in specific subsets of cancer samples, compared to pan-transcriptome expectations. To enable target selection, we identified 410 survival-associated IP genes in 33 cancer types using a machine-learning approach. Notably, GJB2 and SCN9A show prominent expression in neoplastic cells and are associated with poor prognosis in glioblastoma, the most common and aggressive brain cancer. GJB2 or SCN9A knockdown in patient-derived glioblastoma cells induces transcriptome-wide changes involving neuron projection and proliferation pathways, impairs cell viability and tumor sphere formation in vitro, perturbs tunneling nanotube dynamics, and extends the survival of glioblastoma-bearing mice. Thus, aberrant activation of genes encoding ion transport proteins appears as a pan-cancer feature defining tumor heterogeneity, which can be exploited for mechanistic insights and therapy development.

Figures should be cited as 'Figure EV1, Figure EV2" etc. in the text and their respective legends should be included in the main text after the legends of regular figures.
-For the figures that you do NOT wish to display as Expanded View figures, they should be bundled together with their legends in a single PDF file called *Appendix*, which should start with a short Table of Content.Appendix figures should be referred to in the main text as: "Appendix Figure S1, Appendix Figure S2" etc. See detailed instructions regarding expanded view here: .
-Additional Tables/Datasets should be labelled and referred to as Table EV1, Dataset EV1, etc. Legends have to be provided in a separate tab in case of .xlsfiles.Alternatively, the legend can be supplied as a separate text file (README) and zipped together with the Table/Dataset file.10) When assembling figures, please refer to our figure preparation guideline in order to ensure proper formatting and readability in print as well as on screen: http://bit.ly/EMBOPressFigurePreparationGuidelinePlease remember: Digital image enhancement is acceptable practice, as long as it accurately represents the original data and conforms to community standards.If a figure has been subjected to significant electronic manipulation, this must be noted in the figure legend or in the 'Materials and Methods' section.The editors reserve the right to request original versions of figures and the original images that were used to assemble the figure .11) For data quantification: please specify the name of the statistical test used to generate error bars and P values, the number (n) of independent experiments (specify technical or biological replicates) underlying each data point and the test used to calculate p-values in each figure legend.The figure legends should contain a basic description of n, P and the test applied.Graphs must include a description of the bars and the error bars (s.d., s.e.m.).
Please remember: Digital image enhancement is acceptable practice, as long as it accurately represents the original data and conforms to community standards.If a figure has been subjected to significant electronic manipulation, this must be noted in the figure legend or in the 'Materials and Methods' section.The editors reserve the right to request original versions of figures and the original images that were used to assemble the figure.
Further information is available in our Guide For Authors: https://www.embopress.org/page/journal/14602075/authorguideWe realize that it is difficult to revise to a specific deadline.In the interest of protecting the conceptual advance provided by the work, we recommend a revision within 3 months (21st Sep 2023).Please discuss the revision progress ahead of this time with the editor if you require more time to complete the revisions.
I am not going into more detail, because I have reviewed this manuscript at another occasion before, with a similar assessment that was basically very positive -in the current submission, my few additional recommendations have been optimally addressed by the authors.I have no more recommendations.

Referee #2:
The manuscript entitled Pan-Cancer analysis of the ion permeome reveals functional regulators of glioblastoma aggression by Bahcheli et.provides a computational analysis of gene expression of channels the authors have classified as the ion permeome with in vitro and in vivo validation studies focused on glioblastoma.
The manuscript is interesting and the experimental evaluations are intriguing.However, the manuscript suffers major issues.The largest issue is the authors' fundamental approach of using gene expression with patient survival as the parameters for their machine learning.The field is moving away from the use of overall survival as readouts as so many confounding factors are associated with patient mortality that cannot be uncoupled.For GBM overall survival may still be a valid approach but for the many other cancers analyzed it is not...the one-size fits all approach is not appropriate across all cancers.If the authors want to continue looking at pan-cancer the better metrics would be tumour grade/stage which all that data is available via TCGA or other metrics that truly define disease state and progression on a per-cancer basis.Further to this point, the pan-cancer analysis is not really needed considering the authors spend the majority of the manuscript tailoring this to GBM.I highly recommend the authors focus on GBM and expand on their very interesting in vitro and in vivo findings.

Introduction:
The authors have really glossed over massive amounts of work on ion channels in breast and GI tract cancers and chose to focus on the background of GBM.There have been landmark papers that should be cited and even new papers implicating channels in metastasis, the number one cause of cancer related-death.If the authors decided to continue highlighting the pancancer approach then the introduction really needs to be better balanced and representative of pan-cancer not GBM solely.

Results:
1. (Figure 1) I would recommend focusing on GBM specifically.Would it possible to carryout these analyses on different molecular subtypes of GBM?I like this approach, if it could be more focused on GBM and different subtypes it may be more impactful.Are there cancers where there is no clear shift in ion permeome (I think in C but hard to tell with all the data piled up). 2. (Figure 2) This is not very meaningful to the reader.All of this would be better summed up in a table which is easier to read.3. (Figure 3) The differences in survival based on age group are very intriguing.What does this mean biologically?If you separate based on GBM subtype in the middle-aged group, is there a more pronounced difference in survival of middle-aged overexpressors vs none?Or is there a subtype imbalance in the different age groups?4. (Figure 3) The scRNAseq needs working.How were the subgroups determined?What genes qualified cells as differentiated vs. stem cell?Further, what's the main difference between the 2 data sets?Why does only one really have all the expected normal cells and the other not?These data need better explanations.5.I highly recommend assessing CN-alterations, mutations as well as the methylation of SCN9A and GJB2 in GBM. 6. (Figure 4) For in vitro studies, do the shRNA KD cells changes their GBM molecular phenotype or show signs of an altered molecular phenotype transcriptomically?How long after targeting the cells with shRNA do author's carry out phenotypic assessments?Do the authors know the turn-over time of these channels?In part b, there are many nodes presented of which many are not specifically labelled and hard to interpret.Have the authors considered simplifying this to perhaps a larger heatmap where it would be easier to make comparisons between the groups.7. (figure 6) The author's indicate that the low expression of these makers is more in stem cell-like states while the highest are in differentiated cells.It seems odd to me that as the in vitro data indicates that KD reduces viability so greatly, why would one proceed with a stem cell assay?Did the author's see from their KD profiling that the cells looked more stem cell-like?Did the authors think the two were going to be uncoupled, please clarify that logic in the text.Otherwise, it does not make sense to do the stem cell assay as the cells are dying.8. (figure 5) Since the KD is reducing viability, couldn't the shortened projections be related to cells dying?Those that don't die, what sort of knockdown do they have?Are the channels expressed more in the filopodia?Are there antibodies that can be used to see subcellular localization of the channels?9. (Figure 6).The data are really interesting.However, much more indepth analysis is required.This whole figure is based on survival while in the previous figure, the authors discuss nanotubes.There seems to be some missing continuity here.Based on all the experiments in Figure 5, I would expect the authors to continue down this thought process and at least just look at the tumours to see how they are different and look at the changes they observed in vitro.Simple H&E looking at the cells in the tumours and at the edges.Would the authors see less invasive tumours in the KD (e.g. cells spreading out into the parenchyma?) Do the animals with shRNAs that succumb to the tumours have knockdown or are they all escapers?Also, these models should be used to validate the findings in vitro (e.g.RAC1, RAC2, RAC3, PTCH1).

Discussion:
The parts of the discussion on GBM are good.There is room to expand on this more.The pan-cancer approach is not really highlighted.Again, I strongly suggest the author consider possibly removing this pan-cancer approach and tailoring the manuscript to GBM and expanding on their computational and experimental approaches.
Minor concerns: 1. Imprecise wording: Please do not use words like 'excessive' or 'most samples'.You have the data, please list some quantitative measures.2. Need to define 'switch-like' to the audience.3. Clearly define the number of genes analyzed in each class of channels for figure 1. 4. The wording in many of the figures is still small too read and crammed together.a. Figure 4: The networks are too small to read.Also changing color, size and shape is too much in one figure to easily follow.This needs to be reworked and labels changed so they are legible.Please include WB validation of KD if antibodies exist.b. Figure 5: in panel b the WB are sufficient, no need for the expression below.

1
We would like to thank both reviewers for constructive and positive comments that helped improve our manuscript and strengthen our results.We have addressed all comments.Please find a point-by-point response document below.

Reviewer #1
This manuscript is a very strong and important contribution to the emerging field of "Cancer Neuroscience".Many findings are intriguing, and match very well to what is known.But more importantly, this manuscript provides many highly interesting roads that can be followed in the future.I am sure this is an important resource for the entire field.
I am not going into more detail, because I have reviewed this manuscript at another occasion before, with a similar assessment that was basically very positive -in the current submission, my few additional recommendations have been optimally addressed by the authors.I have no more recommendations.
We would like to thank the reviewer for the very positive comments.We appreciate the opportunity to improve the manuscript based on the reviewer's advice at a previous occasion.

Reviewer #2
The manuscript entitled Pan-Cancer analysis of the ion permeome reveals functional regulators of glioblastoma aggression by Bahcheli et al. provides a computational analysis of gene expression of channels the authors have classified as the ion permeome with in vitro and in vivo validation studies focused on glioblastoma.The manuscript is interesting and the experimental evaluations are intriguing.
We would like to thank the reviewer for the positive note and the constructive comments.Our point-by-point responses are shown below.
However, the manuscript suffers major issues.The largest issue is the authors' fundamental approach of using gene expression with patient survival as the parameters for their machine learning.The field is moving away from the use of overall survival as readouts as so many confounding factors are associated with patient mortality that cannot be uncoupled.For GBM overall survival may still be a valid approach but for the many other cancers analyzed it is not...the one-size fits all approach is not appropriate across all cancers.If the authors want to continue looking at pan-cancer the better metrics would be tumour grade/stage which all that data is available via TCGA or other metrics that truly define disease state and progression on a per-cancer basis.
Thank you for a great comment.Indeed, our models predicted different patient survival readouts based on the recommendations in previous publications of the TCGA consortium (1).Specifically, overall survival (OS) was used for some cancer types and progression-free survival (PFS) was used for others (see Methods and Table EV1).Also, we incorporated tumor grade and/or stage as covariates in our machine learning models.This is discussed in the Results and Methods sec- Further to this point, the pan-cancer analysis is not really needed considering the authors spend the majority of the manuscript tailoring this to GBM.I highly recommend the authors focus on GBM and expand on their very interesting in vitro and in vivo findings.
We would like to argue politely that our manuscript is strengthened by the pan-cancer approach as it provides a useful resource of candidate genes to the field (as also emphasized by Reviewer 1).We believe that our systematic analysis of ion permeome genes in more than 30 cancer types is highly relevant to a wide audience of basic and translational researchers who work on these in various experimental projects.Here we focused on GBM, a cancer type of dismal prognosis and strong unmet need, for which we have previous computational and experimental expertise in our group.The candidate genes we found in GBM prompted our interdisciplinary collaboration.Also, we believe that limiting our analysis to GBM and only its subtypes would be also problematic from the point of statistics, since the cohorts are relatively small (see below).
We thank the reviewer for this comment.Regarding the first part of the comment, we discuss our rationale to focus the computational prediction of IP genes in all cancer types earlier in this response letter.
Regarding the second part of the comment, we performed new analyses to include three major GBM subtypes where possible (i.e., classical, mesenchymal, and proneural GBMs).
First, we analysed the overexpression of ion permeome (IP) genes in GBM subtypes, comparing all IP genes to protein-coding genes and other major druggable genes (G-protein coupled receptors and kinases) similarly to the main analysis (see new Figure EV2a copied below).Generally, we confirmed aberrant expression of IP genes in GBM subtypes significantly exceeds other druggable gene families, consistent with our observations for GBM and the other cancer types we analysed.Of note, the GBM subtype analysis is underpowered statistically due to the small sample sizes we could include (55 classical, 49 mesenchymal, and 37 proneural GBMs were available).These sample sets are close or below our initial sample size cut-off of 50 or more samples that we used in our main Results (Figure 2).
Second, we compared patient overall survival (OS) patterns in GBM subtypes by comparing samples with high and low expression levels of GBJ2 or SCN9A, similarly to our analysis of all GBM samples (see the new Figure EV2b copied below).No significant OS differences were found, other than in the mesenchymal group for which high SCN9A expression associated with worse prognosis.Only few samples could be analysed and GBM prognosis is dismal regardless of subtypes, so the lack of strong signals is not surprising and can be perhaps attributed to a statistically underpowered analysis with overall poor prognosis apparent in the L-shaped OS curves characteristic of GBM. 2) This is not very meaningful to the reader.All of this would be better summed up in a table which is easier to read.

(Figure
We thank the reviewer for this recommendation.We already include a supplementary table that shows this information comprehensively and allows the readers to explore our dataset (Table EV2).That said, we believe that our current visualisation allows the readers to look up specific IP genes while having a broader overview as well.Figure 2a-b show how various candidate IP genes compare to each other and thus we decided to keep the figure format as in our original manuscript.However, we optimised the figure to include larger gene symbols.We would be happy to continue improving the figure if necessary.In response to the comment, we now provide an additional table to present this data in another format: Table EV4 describes the data summarised in Figure 2b.3) The differences in survival based on age group are very intriguing.What does this mean biologically?If you separate based on GBM subtype in the middle-aged group, is there a more pronounced difference in survival of middle-aged overexpressors vs none?Or is there a subtype imbalance in the different age groups?

(Figure
We appreciate the reviewer's positive note and constructive comments.We addressed these in our updated manuscript.We separated GBM samples by subtypes and by age groups (Figure EV3c) and analysed the survival differences of the different age groups (Figure EV3d).No significant associations between GBM subtypes and patient age groups were found, other than the finding that the older age group (patients > 66 years) had significantly worse survival than all other age groups (Hazard Ratio (HR) 1.564, p = 0.022).The younger age group (< 56 years) had no significant differences and the middle age group (56 -66 years) had only sub-significant differences in survival (HR = 0.693, p = 0.061).The age groups were roughly equal in size (48-51 samples) and there were no significant differences in GBM subtypes based on age.We note again that these results are based on small sample sizes and therefore not fully reliable.Thus, we added these in our supplementary materials for completeness and did not emphasise these strongly in the main part of the manuscript.3) The scRNAseq needs working.How were the subgroups determined?What genes qualified cells as differentiated vs. stem cell?Further, what's the main difference between the 2 data sets?Why does only one really have all the expected normal cells and the other not?These data need better explanations.

(Figure
We apologise for the lack of clarity.In this analysis we used data from two previous publications where the cell type annotations were reported along with the scRNA-seq data.These included 7,930 cells from IDH1/2 wildtype GBMs (Neftel et al. (2019)) and 55,284 cells from IDH1/2 wildtype and mutant GBMs (Johnson et al. (2021)).We then performed UMAP dimensionality reduction on the previously processed gene expression values (TPM) from the two studies.All the cells were labelled using the annotations provided in the original studies.As we used previously published data, the differences in the original experimental designs, data preprocessing steps, and cell type annotation strategies contribute to the differences we observed.We explained this data analysis in the Methods section under "Expression of candidate IPs in GBM at the single-cell level" on page 26 of our manuscript.We also added a sentence to the Results to discuss the differences between the two datasets (page 12). 5.I highly recommend assessing CN-alterations, mutations as well as the methylation of SCN9A and GJB2 in GBM.
We appreciate this comment.When preparing our first submission we systematically analysed the genetic and epigenetic alterations of the two candidate genes in GBM.However, as no striking findings were then apparent, we excluded these data from our manuscript and focused on gene expression patterns instead.As this comment greatly contributes to the completeness of our study, we analysed the data again and added the findings to our manuscript in Appendix Figure S4.
First, we investigated the promoter methylation of the two candidate genes in GBM and found no significant associations between promoter methylation of GJB2 and SCN9A and patient survival (Appendix Figure S4a, left).Promoter methylation associated with GJB2 expression but not with SCN9A expression (Appendix Figure S4b).
Second, we investigated genomic copy number alterations (CNAs) of GJB2 and SCN9A in GBM and found no associations between CNAs and patient survival (Appendix Figure S4a, middle).Interestingly, we found that CNAs in SCN9A co-occurred with CNAs of the prognostic GBM oncogene IDH1 due to the genomic proximity of SCN9A and IDH1 such that most CNAs in SCN9A also affected IDH1 and vice versa (19 / 20 of samples or 95%).However, these represent only a minority of the GBM cohort (20/146 or 14%).This is a potentially relevant finding for future studies as IDH1 is mostly described in the context of SNVs (i.e., the prognostic mutation hotspot IDH1-R132H) rather than CNAs, therefore we added a brief sentence about it in the manuscript.
Third, we studied somatic SNVs and indels in GJB2 and SCN9A in GBM (Appendix Figure S4a right, Appendix Figure S4c).We found no non-silent mutations in GJB2 and eight missense SNVs in SCN9A in six GBM samples.No significant survival associations were found for the SCN9A mutations.
Taken together, the aberrant expression patterns of GJB2 and SCN9A had the strongest associations with GBM patient survival while no consistent signals among the genomic or epigenomic alterations were found to explain the differential expression or OS characteristics of the two candidate genes.
6. (Figure 4) For in vitro studies, do the shRNA KD cells changes their GBM molecular phenotype or show signs of an altered molecular phenotype transcriptomically?How long after targeting the cells with shRNA do author's carry out phenotypic assessments?Do the authors know the turn-over time of these channels?
Thanks for these great comments.Regarding the first part of the comment, we witnessed transcriptome-wide changes mediated by GJB2 or SCN9A knockdown in vitro apparent in the differential expression of 4647 and 2088 genes, respectively (FDR < 0.05; FC > 1.25) (Figure 4a).
RNA-seq was performed on GBM cells at the 4-day timepoint after candidate gene knockdown.Fluorescence imaging was also carried out 4 days after lentiviral shRNA transduction, the viability assay after 7 days, and the limiting dilution assay after 14 days.We included this information in the methods section under "Cell viability, limiting dilution assay, tunneling nanotube and filopodia imaging".
Regarding the turn-over time of these proteins, we observed reduced protein expression and cellular phenotypes 4 days after shRNA treatment (Figure 5, 6, Appendix Figure S7a), which indicate that at this time point, GJB2 and SCN9A that were expressed prior to shRNA treatment had been degraded.Therefore, the average turn-over time of these proteins in GBM cells is at most around 4 days.
In part b, there are many nodes presented of which many are not specifically labelled and hard to interpret.Have the authors considered simplifying this to perhaps a larger heatmap where it would be easier to make comparisons between the groups.We respectfully argue that our "enrichment map" is a preferred visualisation to a heatmap.The major reason is that pathway descriptions are highly redundant and often provide too much information that is difficult to navigate, while the network visualisation provides fewer but more general functional themes that are easier to interpret (these are usually manually curated).In other words, there are too many nodes (i.e., enriched pathways) in the enrichment map of Figure 4b to label them all legibly, while our summaries of the sub-networks allow a higher-level overview (most individual labels are removed on purpose as is standard practice for these types of figures; see Reimand et

al. (2019) (2)).
To address this comment, we have added the full list of detected pathways and their evidence with respect to GJB2 and SCN9A to Table EV4.We also added the manual annotations from each pathway in Figure 4b to the Table EV4 so that readers may relate the pathway themes shown in the enrichment map to specific pathways of interest.We reviewed the description of this panel in the figure legend for improved clarity.6) The authors indicate that the low expression of these makers is more in stem celllike states while the highest are in differentiated cells.It seems odd to me that as the in vitro data indicates that KD reduces viability so greatly, why would one proceed with a stem cell assay?Did the author's see from their KD profiling that the cells looked more stem cell-like?Did the authors think the two were going to be uncoupled, please clarify that logic in the text.Otherwise, it does not make sense to do the stem cell assay as the cells are dying.

(figure
To clarify our interpretation of the LDA assay, we performed new immunofluorescence imaging of the mitotic marker phospho-histone H3.We found that GJB2 or SCN9A knockdown led to a reduction in the total percentage of mitotic cells (Figure 6d), demonstrating that GJB2 and SCN9A may promote GBM cell proliferation.8. (figure 5) Since the KD is reducing viability, couldn't the shortened projections be related to cells dying?
Thank you for highlighting this potential confounding factor.To rule out the possibility that the shortened TNTs were the result of cells undergoing apoptosis, we performed immunofluorescence imaging of the apoptosis marker "cleaved caspase-3" with cell membrane-tagged GFP (mGFP).We observed that GJB2 knockdown reduced TNT length in cleaved caspase-3 negative, non-apoptotic cells (Appendix Figure S6a), suggesting that the shortened TNTs in GJB2 knockdown cells were not caused by cell morphological changes during apoptosis.
To further show that the shortened filopodia were not associated with apoptotic morphological changes, we repeated the mGFP time-lapse imaging with a longer imaging time.Since the morphological changes associated with apoptosis were previously demonstrated to occur within 2 hours (5,6), we performed time-lapse imaging for 2.5 hours.In cells that did not exhibit apoptotic rounding or blebbing throughout the imaging duration, GJB2 knockdown resulted in shorter filopodia extension length and lifetime (Appendix Figure S6b).These results confirm that the shortened projections in GJB2 knockdown cells were not due to morphological changes associated with apoptosis.
Those that don't die, what sort of knockdown do they have?Are the channels expressed more in the filopodia?Are there antibodies that can be used to see subcellular localization of the channels?
Thanks for a great question.To address this comment, we obtained antibodies against GJB2 and SCN9A and validated antibody specificity by shRNA KD (Appendix Figure S7a).GJB2 and SCN9A were broadly expressed on the GBM cell membrane (Appendix Figure S7b).We observed that GJB2 and SCN9A were primarily localised to the soma and not TNTs or filopodia (Appendix Figure S7b), suggesting that GJB2 may indirectly affect cellular projections, for example, through a RAC1-associated mechanism.9. (Figure 6).The data are really interesting.However, much more in-depth analysis is required.This whole figure is based on survival while in the previous figure, the authors discuss nanotubes.There seems to be some missing continuity here.Based on all the experiments in Figure 5, I would expect the authors to continue down this thought process and at least just look at the tumours to see how they are different and look at the changes they observed in vitro.Simple H&E looking at the cells in the tumours and at the edges.Would the authors see less invasive tumours in the KD (e.g. cells spreading out into the parenchyma?) Do the animals with shRNAs that succumb to the tumours have knockdown or are they all escapers?Also, these models should be used to validate the findings in vitro (e.g.RAC1, RAC2, RAC3, PTCH1).
We thank the reviewer for these insightful and helpful questions and the positive note.To complement the in vitro data in Figure 5 and establish in vivo relevance, we performed xenografts of control and GJB2 knockdown GBM cells in mice with a time-matched analysis at 12 days post implantation.We observed that RAC1 and pMLC2 (i.e., downstream effector of Rho GTPases) levels were reduced in GJB2 knockdown tumors (Figure 5e).We analysed tumor invasiveness and found that both sinuosity (i.e., a measure of tumor infiltration) and the sizes of infiltrating tumor colony were reduced in GJB2 knockdown tumors (Figure 5f).Importantly, filopodia at the invading front were shortened in GJB2 knockdown tumors as well (Figure 5g).These results validate our in vitro findings that GJB2 is involved in the regulation of RAC1 signaling and cellular projections and demonstrate that GJB2 regulates GBM invasion in vivo.
To address the question of whether the mice in the shRNA groups that did not succumb to the xenografts were due to tumor escapers of the shRNA knockdown, we harvested endpoint tumors and performed immunostaining of GJB2 or SCN9A (Figure EV5).We observed that at endpoint, GJB2 or SCN9A knockdown tumors and control tumors had comparable levels of GJB2 or SCN9A expression, respectively.These results, together with the in vitro data showing GJB2 or SCN9A knockdown reduces cell viability and proliferation, suggest that the tumors in shRNAtreated groups at endpoint are escapers.
3. Clearly define the number of genes analyzed in each class of channels for figure 1.
We added gene numbers for Figure 1.
4. The wording in many of the figures is still small too read and crammed together.
We increased the size of figures to make text more visible and adapted our figures to meet the requirements of EMBO Journal.
4a. Figure 4: The networks are too small to read.Also changing color, size and shape is too much in one figure to easily follow.This needs to be reworked and labels changed so they are legible.Please include WB validation of KD if antibodies exist.
We increased the size of the labels within the network of Figure 4d to improve legibility.We believe that having 2 shapes in addition to size and color scales is appropriate for this figure.
To validate the knockdown experiments, we obtained antibodies against GJB2 and SCN9A, and performed immunostaining 4 days after lentiviral shRNA transduction in mGFP GBM cells.To quantify GJB2 and SCN9A intensity, we used the mGFP signal to create a region of interest encapsulating the cell membrane.We then measured the fluorescence intensity of GJB2 and SCN9A within this region of interest.We observed that GJB2 or SCN9A intensity was reduced in GJB2 or SCN9A shRNA treated cells, respectively, validating knockdown at the protein level (Appendix Figure S7a).4b. Figure 5: in panel b the WB are sufficient, no need for the expression below.
We appreciate this advice.However, we note that only one of three western blot replicates are shown in the western blot of Figure 5b.In addition to these blots, the expression values below the western blots show all three biological replicates of our experiment and display significance.We believe that keeping the expression values for all three replicates is important to demonstrate the statistical significance in our analysis.We reviewed the figure legend for clarity.

EMBO Press Author Checklist USEFUL LINKS FOR COMPLETING THIS FORM
The EMBO Journal -Author Guidelines EMBO Reports -Author Guidelines Molecular Systems Biology -Author Guidelines EMBO Molecular Medicine -Author Guidelines Please note that a copy of this checklist will be published alongside your article.

Abridged guidelines for figures 1. Data
The data shown in figures should satisfy the following conditions: ➡ ➡ definitions of statistical methods and measures: -are tests one-sided or two-sided?-are there adjustments for multiple comparisons?-exact statistical test results, e.g., P values = x but not P values < x; -definition of 'center values' as median or average; -definition of error bars as s.d. or s.e.m.

Materials Newly Created Materials Information included in the manuscript?
In which section is the information available?
(Reagents and Tools New materials and reagents need to be available; do any restrictions apply?Not Applicable

Antibodies
Information included in the manuscript?
In which section is the information available?
(Reagents and Tools For antibodies provide the following information: -Commercial antibodies: RRID (if possible) or supplier name, catalogue number and or/clone number -Non-commercial: RRID or citation

DNA and RNA sequences Information included in the manuscript?
In which section is the information available?
(Reagents and Tools Short novel DNA or RNA including primers, probes: provide the sequences.

Cell materials Information included in the manuscript?
In which section is the information available?
(Reagents and Tools Have primary datasets been deposited according to the journal's guidelines (see 'Data Deposition' section) and the respective accession numbers provided in the Data Availability Section?

Methods, Data Availability Section
Were human clinical and genomic datasets deposited in a public accesscontrolled repository in accordance to ethical obligations to the patients and to the applicable consent agreement?

Yes
Methods, Data Availability Section Are computational models that are central and integral to a study available without restrictions in a machine-readable form?Were the relevant accession numbers or links provided?

Not Applicable
If publicly available data were reused, provide the respective data citations in the reference list.Yes Methods, Reference List The MDAR framework recommends adoption of discipline-specific guidelines, established and endorsed through community initiatives.Journals have their own policy about requiring specific guidelines and recommendations to complement MDAR.
tions (page 7 and page 24, respectively).For the four candidate genes in GBM, the covariateadjusted HR values are shown in the dot-and-whiskers plot in Figure3.
Table, Materials and Methods, Figures, Data Availability Section) Table, Materials and Methods, Figures, Data Availability Section) Table, Materials and Methods, Figures, Data Availability Section)

Please complete ALL of the questions below. Select "Not Applicable" only when the requested information is not relevant for your study. if
Table, Materials and Methods, Figures, Data Availability Section) Reagents and Tools Table, Materials and Methods, Figures, Data Availability Section) Provide species, strain, sex, age, genetic modification status.Provide accession number in repository OR supplier name, catalog number, clone number, OR RRID.Reagents and Tools Table, Materials and Methods, Figures, Data Availability Section) If collected and within the bounds of privacy constraints report on age, sex and gender or ethnicity for all study participants.Reagents and Tools Table, Materials and Methods, Figures, Data AvailabilitySection) , such as t-test (please specify whether paired vs. unpaired), simple χ2 tests, Wilcoxon and Mann-Whitney tests, can be unambiguously identified by name only, but more complex techniques should be described in the methods section; n<5, the individual data points from each experiment should be plotted.Any statistical test employed should be justified.Source Data should be included to report the data underlying figures according to the guidelines set out in the authorship guidelines on Data Each figure caption should contain the following information, for each panel where they are relevant: a specification of the experimental system investigated (eg cell line, species name).theassay(s)and method(s) used to carry out the reported observations and measurements.anexplicitmention of the biological and chemical entity(ies) that are being measured.anexplicitmention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner.ideally,figurepanels should include only measurements that are directly comparable to each other and obtained with the same assay.plotsinclude clearly labeled error bars for independent experiments and sample sizes.Unless justified, error bars should not be shown for technical the exact sample size (n) for each experimental group/condition, given as a number, not a range; a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).a statement of how many times the experiment shown was independently replicated in the laboratory.This checklist is adapted from Materials Design Analysis Reporting (MDAR) Checklist for Authors.MDAR establishes a minimum set of requirements in transparent reporting in the life sciences (see Statement of Task: 10.31222/osf.io/9sm4x).Please follow the journal's guidelines in preparing your the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. (((

Checklist for Life Science Articles (updated January Study protocol Information included in the manuscript? In which section is the information available?
(Reagents and Tools Table, Materials and Methods, Figures, Data Availability Section)If study protocol has been pre-registered, provide DOI in the manuscript.For clinical trials, provide the trial registration number OR cite DOI.

In which section is the information available?
(Reagents and ToolsTable, Materials and Methods, Figures, Data Availability Section)

Sample definition and in-laboratory replication Information included in the manuscript? In which section is the information available?
(Reagents and ToolsTable, Materials and Methods, Figures, Data Availability Section)In the figure legends: state number of times the experiment was replicated in laboratory.

In which section is the information available?
Include a statement confirming that informed consent was obtained from all subjects and that the experiments conformed to the principles set out in the WMA Declaration of Helsinki and the Department of Health and Human Services Belmont Report.State if relevant permits obtained, provide details of authority approving study; if none were required, explain why.
(Reagents and ToolsTable, Materials and Methods, Figures, Data Availability Section) Studies involving human participants: State details of authority granting ethics approval (IRB or equivalent committee(s), provide reference number for approval.Not Applicable Studies involving human participants: Yes Methods Studies involving human participants: For publication of patient photos, include a statement confirming that consent to publish was obtained.Not Applicable Studies involving experimental animals: State details of authority granting ethics approval (IRB or equivalent committee(s), provide reference number for approval.Include a statement of compliance with ethical regulations.Yes MethodsStudies involving specimen and field samples:

Use Research of Concern (DURC) Information included in the manuscript? In which section is the information available?
(Reagents and ToolsTable, Materials and Methods, Figures, Data Availability Section) Could your study fall under dual use research restrictions?Please check biosecurity documents and list of select agents and toxins (CDC): https://www.selectagents.gov/sat/list.htmNot Applicable If you used a select agent, is the security level of the lab appropriate and reported in the manuscript?Not Applicable If a study is subject to dual use research of concern regulations, is the name of the authority

granting approval and reference number for
the regulatory approval provided in the manuscript?

Adherence to community standards Information included in the manuscript? In which section is the information available?
(Reagents and ToolsTable, Materials and Methods, Figures, Data Availability Section) State if relevant guidelines or checklists (e.g., ICMJE, MIBBI, ARRIVE, PRISMA) have been followed or provided.Not Applicable For tumor marker prognostic studies, we recommend that you follow the REMARK reporting guidelines (see link list at top right).See author guidelines, under 'Reporting Guidelines'.Please confirm you have followed these guidelines.

and III randomized controlled trials
, please refer to the CONSORT flow diagram (see link list at top right) and submit the CONSORT checklist (see link list at top right) with your submission.See author guidelines, under 'Reporting Guidelines'.Please confirm you have submitted this list.Reagents and Tools Table, Materials and Methods, Figures, Data Availability Section) (