miRNA-29-CLIP uncovers new targets and functions to improve skin repair

Article activity feed

1. 

   Review Commons

   Nov 14, 2024

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

   Learn more at Review Commons

   ---

   Reply to the reviewers

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

   miRNAs are important for the control of many cellular processes, with the miR-29 family of miRNAs implicated in the regulation of cell growth in different cell types in both the epidermis and dermis of the skin. However, the roles of miRNAs in specific cell types in general, and of the miR-29 family in the skin, are currently unknown. Here, the authors use a range of cellular and molecular techniques, including miRNA cross-linking and immunoprecipitation (miRNA-CLIP) and antisense oligonucleotides (ASO), as well as RNA-Seq, qPCR, Western blotting, in situ hybridization, adhesion and ECM …

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

   Learn more at Review Commons

   ---

   Reply to the reviewers

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

   miRNAs are important for the control of many cellular processes, with the miR-29 family of miRNAs implicated in the regulation of cell growth in different cell types in both the epidermis and dermis of the skin. However, the roles of miRNAs in specific cell types in general, and of the miR-29 family in the skin, are currently unknown. Here, the authors use a range of cellular and molecular techniques, including miRNA cross-linking and immunoprecipitation (miRNA-CLIP) and antisense oligonucleotides (ASO), as well as RNA-Seq, qPCR, Western blotting, in situ hybridization, adhesion and ECM assays, ELISA and immunofluorescence, to interrogate the roles of the miR-29 family of miRNAs in controlling cell growth in epidermal keratinocytes and dermal fibroblasts, using 2D and 3D ex vivo models. The coupling of miR-CLIP with functional assays allowed the authors to identify both miRNA-mRNA complexes, and the biological pathways that these ultimately manipulate.

   The authors report the identification of unbiased, tangible miR-29/mRNA pairs, together with functional roles in cell adhesion, ECM regulation and fibroblast proliferation, that are distinct between keratinocytes and fibroblasts. miR-29 is identified as a valuable target for interventions that seek to promote healthy skin regeneration, including applications for wound healing. Many of the pathways identified here have previously been described, but the novelty of this manuscript lies in the innovative combination of miR-CLIP with functional assays, the application of these in combination to specific cell types, the identification of miR-29 as a novel master regulator of epidermal keratinocyte adhesion via a range of different pathways, and the demonstration that miR-29 inhibition in fibroblasts can influence keratinocyte adhesion via paracrine signalling.

   The experiments are well designed and reported. The interpretations are sound and appropriate for the data presented (though see the comment on potential normalisation of ECM data to cell numbers in cultures for the miR-29 mimic/inhibitor data for fibroblasts and the query about the number of direct miR-29 targets in fibroblasts that are ECM-related).

   Major Comments: I have no major concerns to raise over this manuscript. The claims and conclusions are supported by the data and no additional experiments are required (though please note the comment on normalisation mentioned above and detailed below). The methods are clearly reported and statistical reporting is adequate.

   Minor Comments: Pg3, 7th line from the bottom: "processed into three functional miRNA..." - minor edit needed here, it looks like there's a word missing somewhere. Pg3, last line on the page: "results supported..." - is there a missing 'are' here? Pg5, 15th line of the main text: "of miRNA-29-mediate repression..." - is there a missing 'd' here ('-mediated...')? There is lots on minor presentation errors like this throughout the manuscript - I won't point them out exhaustively, but the manuscript needs a good thorough proof-read, maybe from a fresh pair of eyes? - *We fully agree with the reviewer. The manuscript has been proofread and corrected throughout. * Fig. 1C: Can the figure be edited to better highlight the basal layer with lack of (nsm image) and expression of (abm image) K10? Maybe a box around that layer, rather than the current arrows only on the abm image (which are not particularly closely indicating the basal layer)? We thank the reviewer for this suggestion. The arrows on the Fig.1C point to the areas where keratin K10 filaments are reaching the basal membrane (indicated by collagen IV staining). It was difficult to box out the basal level without covering the K10 signal. We decided to explain this in the legend to clarify how the data shows this pre-mature expression of keratin K10 in the miR-29ab mimic sample. ____The basal layer of the control (nsm) sample thus remains K10-free and only shows nuclear DAPI staining. Fig. 2 legend should include definitions of abbreviations shown on the figure.* - Added* Pg8/Fig. 4A: Can the reporting of shared transcript targets of miR-29 in IFK/HFK/DF cells be better communicated? Maybe just adding the actual percentage overlap in transcriptomes for IFK/HFL and keratinocytes/fibroblasts to the main text would help .* – Actual percentages of the overlaps added in the text*. Similarly, I think a direct report somewhere (in the main text?) of total number for relevant groups shown in Fig. 4E would also be useful - e.g. there are 45 transcripts that are direct targets of miR-29 in keratinocytes and also associated with ECM, and 190 that are direct targets of miR-29 in keratinocytes and also associated with cell adhesion, but these number are difficult to come by quickly at the moment. It would be nice to be able to quickly compare these numbers for keratinocytes to their equivalents for fibroblasts__. – This is a very helpful suggestion with a good example. We incorporated the suggestion into the text and made changes to the figure to make it easier to compare pro-adhesive and miR-29-regulated functions in keratinocytes and fibroblasts.__ Fig. 4B: It's interesting that ~15% of miR-29 binding targets identified using miR-CLIP are not predicted targets based on TargetScan/microT-CDS. I'd like to see a little more information on this added to the manuscript - perhaps listing some of these or including a table of them? And perhaps some discussion of this could be added also. - Indeed, almost 170 mRNAs are in this category and are now listed in a table in Suppl. File 1. Non-canonical binding is briefly discussed in the text. Fig. 4E: I would be nice to see the Venn numbers for keratinocyte proliferation (either is a supp figure, or addition to the main text?), to help illustrate the lack of a role for miR-29 in the regulation of keratinocyte proliferation. – It is an interesting point; the cell proliferation seems to be a function of miR-29 in fibroblasts but not in keratinocytes. We did not detect cell proliferation as a significantly enriched function among keratinocyte mRNAs directly regulated by miR-29. It is consistent with the lack of change in BrdU incorporation in keratinocytes grown in 3D (Figure 2). We also never noticed any change in keratinocyte proliferation while expanding them in 2D after miR-29 transfection or inhibition. This has been further highlighted in the text. Fig. 4E: Is the reported number of direct miR-29 targets in fibroblasts that are ECM-related correct? This number is reported as 10 in the main text (pg10, 3rd paragraph), but it looks like 10 is only for direct miR-29 targets in fibroblasts that are ECM-related AND related to proliferation. Should this number be 58? The 10 that are direct miR-29 targets in fibroblasts that are ECM-related AND related to proliferation can be reported in the next sentence, where this group is specifically referred to__. – This has now been amended in the text according to the reviewer’s suggestion.__ Fig. 7 (and related main text): Did you take any steps to normalise ECM measurements to cell numbers present in cultures in the miR-29 mimic/inhibition experiments in fibroblasts? This should really be included as it would provide an answer to the speculation of whether the effects of manipulating miR-29 on ECM are due to proliferation or classical pro-fibrotic pathways - it is probably based on proliferation not pro-fibrosis because TGFb is one of the most pro-fibrotic cytokine known and it’s response is abrogated by miR-29KD. Need to check the original excel for Fig. 7D. – Yes, the concentration of the ECM was measured in ng/ml and normalized per number of cells. We calculated the concentration of oligonucleotides per cell by dividing the amount of transfected oligo per number of transfected cells counterstained with nuclear DAPI signal. We could do so because every cell showed a similar transfection rate by calculating fluorescence of Cy3 conjugated to the miR oligos. Then, we divided the ECM concentration by the number of transfected cells per well, thus normalizing the ECM deposition to the cell number. The reviewer is correct, both the increase in ECM after miRNA-29 KD and the decrease in ECM after miRNA-29 overexpression is consistent with increased and decreased cell numbers, correspondingly. As suggested, we later confirm that the increased deposition of the ECM was not a result of activated pro-fibrotic pathway (Figure 7).

   Fig. 8E: The upper and lower image need to have nsa/abc labels added to them. – This has been done, thank you for noticing! Pg12, 1st sub-heading: typo (cell-specific). -corrected.

   **Referees cross-commenting**

   All reviews appear to be fair and balanced to me. I agree that in places wording could be amended to temper the strengths of some claims, and it would also be nice to see some additional functional assays included, to complement the adhesion and ECM deposition assays that are currently presented, though I do not think this should necessarily be a requirement for publication and could be included in subsequent follow-up work from the group. I did not spot the reuse of images between Fig. 1 and 2, but clearly this should be addressed - either by replacing one set of images, or by removing the relevant panels from Fig. 1 and changing in-text reference to guide the reader to Fig. 2A. I also agree that it would be nice to see miR-29 staining of mouse dermal fibroblasts during wound healing, to complement the images already shown for keratinocytes, and to see miR-29 staining in human skin__*. – We thank Reviewer 1 for cross-checking other reviews, and we address these comments in response to Reviewers 2 and 3. *__

   Reviewer #1 (Significance (Required)):

   miR-CLIP is a powerful, recently developed technique, with enormous promise for the identification of true miRNA-mRNA pairs, that has not yet been widely adopted by the research community. As such, its application here is itself relatively novel, adding enormously to our existing knowledge of likely miR-29 targets, providing tangible information in miR-29/mRNA pairs in specific cell types in different layers of the skin, but also further adding novel functional information to this, with demonstrations of the regulation of specific relevant biological pathways through manipulation of targets identified using miR-CLIP. The methods are sound (and impressive), results are reported well and not over-interpreted. There is the potential for better characterisation of the relative importance of canonical pro-fibrotic pathways vs proliferation-related effects on ECM production, and this should not be difficult to address. This paper will be on interest to a wide readership, including those engaged in fundamental research and clinicians.

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

   Summary:

   The article entitled, "miRNA-29-CLIP uncovers new targets and functions to improve skin repair", by Thiagarajan et al. describes the characterization of the functions of miRNA-29 in keratinocytes and fibroblasts, its RNA interactors and potential mechanisms of action. Using candidate interactors and 2D cell culture and 3D skin equivalents combined with loss-of-function (inhibitor) and gain-of-function (mimic), and changes in expression analyses, the authors conclude that the major function of miRNA-29 is to regulate cell-substrate adhesion.

   Major comments:

   • While the interactors and expression changes are useful resources, the claims and the conclusions that are based on them are exaggerated. The treatments are associated with changes in expression, but no functional data support the conclusions. Additional functional experiments are required to assertively make the claims. The title is misleading when stating "to improve skin repair" and the abstract also makes some bold general claims, which are tangentially supported by the findings. For example, "protein folding" only appears in the abstract and "RNA processing" is in the abstract and figures but not referred to in the text__. – We thank the reviewer for valid criticism. While this manuscript was in preparation, we were publishing our other study showing the function of miRNA-29 in wound healing in cutaneous mouse-based model. This study demonstrated an improved re-epithelialization and wound closure in Mir29ab1 KO mice (Robinson et al, Am. J. of Pathology 2024). It was difficult not to think about the role of miR-29 in a wider context of skin repair, which was the goal of the in vivo part of the project. We could not cite the other manuscript at that time as a reference and should have toned down our claims to improved skin repair in this manuscript.__

   • The authors may want to tune their language that their data suggest the conclusions as opposed to being definitive and assertive. This should be done in the Discussion, while the Results should represent the direct conclusions__. – This has now been amended accordingly (highlighted in green).__

   • A couple of examples to the above, in the conclusion to section 1 of the Results, how was the "loss of basal adhesion" assessed? Is it by beta1-integrin localization changes? – We have not performed assays specific to activated integrins, but this is planned studies where we will address the molecular details of the miRNA-29-controlled cell-to-cell and cell-to-matrix adhesion mechanism. Also, how is "growth" defined"? proliferation is not changed and a more accurate way to describe the result is to refer to thickness__. – Indeed, our results clearly demonstrate no change in keratinocyte proliferation in response to a change in miRNA-29 levels either way. We therefore speculate that the reason for differences in 3D cultures of keratinocytes (the SEs) is pre-mature differentiation, induced by miRNA-29. While we do not have a mechanistic answer to this observation (e.g., keratin K14 is not a direct target of miRNA-29), premature expression of K10 in the basal layer may be a consequence of altered adhesion mechanisms in the basal layer. As noted earlier, we are currently investigating the mechanism of miRNA-29-regulated adhesion of mouse and human keratinocytes, but this was beyond the scope of presented study, which has identified the phenomenon at the first instance using organismal and tissue-level approach.__

   • The images in Fig 1C are reused in Fig 2A, where new examples should be shown instead. – We had erroneously inserted the same panel as in Figure 2. The correct day 6 panel is now inserted instead in Figure 1C, along with an additional control of normal human skin.

   • Fig 1C and Fig 2A are not quantified to make the claims about premature differentiation and integrin expression changes. – We struggled to find an accurate method of quantifying the fluorescent signal coming from varied cell shapes and the basal lamina of human SEs. We however see certain consistency in deposition of integrin beta 1 and alpha 6 (ITGB1and ITGA6) in our SEs. The signal for ITGB1 completely disappears in miRNA-29 treated SEs while ITGA6 goes down. Conversely, increased ITGB1 after inhibition of miR-29 coincides with a higher signal of ITGA6 (Figure 2A). ITGB1 and ITGA6 are co-expressed in basal layer of ____human skin____* and ____SEs____(____Solé-Boldo et al, Comm. Biology 2020, ____Fig. 1c____; Stabel et al, Cell Rep. 2023, Fig. 3E) and can heterodimerize to form integrin α6β1 in various tissues (____reviewed by Zhou et al. Stem Cell Res Ther. 2018____). We have changed the way we discuss the results in the text.*

   • Fig 3: It is not clear from the figure legends what statistical methods were used for which experiment or how many times the experiment was performed (not just biological replicates), especially given the variability among experiments in Fig 3C. - Adhesion assay in Fig. 3A was performed in four biological replicates with one batch of primary human keratinocytes (pooled neonatal), and in 3C, as two independent experiments (exp) with two different batches of keratinocytes (exp 1 and exp 2). Lower numbers of cells in exp 1 as compared to expt 2 are due to an unfortunate but usual variability between batches of primary cells. The variability noted by the reviewer is most likely coming from lower numbers of cells in exp 1 as compared to exp 2. We have now clarified this in the figure legend.

   Minor comments:

   • The Introduction is focused on methodology and should include elements that pave the way to the Results. Some information that belongs in the introduction are present in the Results section. In this respect, please define the miRNA processing Dicer pathway and its components in the introduction so that the reader can follow the nomenclature (AGO2, RISC, etc.). Also, introduce human skin equivalents or organotypic culture as a model system in the Introduction.

   • Some information in the Results belongs in the Introduction, for example, the first seven lines of the Results section. - We have changed the introduction accordingly

   • The authors might want to consider including quantifications in the main figures, so they are immediately apparent to the reader, for example, Fig S1C. Also, Fig S2B is an important measure for the immediate outcome of the treatment on miRNA-29__. – We have included the quantification of the SE epidermal thickness in Fig. 1D and emphasized the KD effect of miR-29 anti-sense oligos in the text.__

   • Please change "imidiate" to "immediate", "sculp" to "scalp", "has to be releaved of miRNA-29-mediate repression" to "has to be relieved of miRNA-29-mediated repression" * - Done.*

   **Referees cross-commenting**

   I agree with my colleagues' assessments and suggestions. The miRNA-CLIP data in keratinocytes and fibroblasts are important resources. The figures and text require reconsideration to more accurately represent the data as detailed in our collective reviews

   Reviewer #2 (Significance (Required)):

   The study utilizes 2D and 3D cultures and presents an important resource for miRNA-29 interactors in keratinocytes and fibroblasts, as well as the expression changes associated with its inhibition and overexpression. However, the conclusions are exaggerated and based on expression changes. If the conclusions are rephrased, the findings would be of interest to a broad audience interested in miRNA, cell adhesion and epithelial and mesenchymal biology.

   My expertise is in skin development and maintenance, genetics and cell biology. I have limited knowledge in RNA biology.

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

   Summary:

   Thiagarajan et al. report on the functions and molecular targets of miR-29 in human primary skin cells. They first focus on the potential role of miR-29 in wound healing and in the adhesion of keratinocytes to the basement membrane using both in vivo wounding assays in the mouse and human cultures/skin equivalents. The authors report that miR-29 negatively affects adhesion in vivo and in vitro and characterise the transcriptome of fast and slow-adhering cells with or without miR-29inhibition. They proceed to identify miR-29 targets in three primary skin cell types (follicular keratinocytes, interfollicular keratinocytes and fibroblasts) by performing miRNA-clip. By comparing these targets to genes altered in keratinocytes with high adhesion capacity after miR-29 inhibition or fibroblasts after miR-29 inhibition, the authors describe a model in which miR-29 inhibits multiple adhesion-associated pathways in keratinocytes and negatively regulates proliferation and ECM deposition by dermal fibroblasts.

   Major comments:

   Overall, the paper is interesting, and the experiments performed are generally sensible for the questions being investigated. However, I thought the data was presented in a very confusing and unclear way, both in the main text and in the figures. I found the paper quite difficult to navigate, with contradictory statements between text and figures, cryptic or confounding graphs or arrangement of the figures and, in at least one instance, re-use of the same image with inconsistent labelling. The paper will thus greatly benefit from extensive tidying up and review of both text and figures to improve clarity. I highlight several points below, with many being related to this overarching issue, and I try to offer suggestions to the authors improve the quality of the manuscript.

   • The stainings in Figure 1A should be repeated in intact sections as it is difficult to understand the exact distribution of miR-29 when the whole epidermis appears to be falling apart in the section. It is possible to see the pattern the authors are describing based on the current images, but it is not convincing. – We fully agree with the reviewers that an intact section would inform the reader on the distribution of miRNA-29 inside the wound much better when the wound morphology is preserved. We have tried repeating the staining (fluorescent in situ hybridization coupled with the antibody staining). The protocol involves multiple washing steps performed at high temperature (for the FISH) and detergent (for the immunodetection step) to ensure specific miRNA probe binding and a low background for the antibody binding. As a result, we could not get a more intact section at the end unfortunately. We have however published a miRNA-29 FISH only stained mouse wounds in ____Robinson et al, Am Journal of Pathology 2024, Figure 1C and Suppl. Fig. 1B____* showing more intact sections with miRNA-29 signal against DAPI. There, one can see the same pattern of miRNA-29 expression as in Figure 1 of this manuscript, with less miRNA in the basal layer of wound keratinocytes vs more miRNA-29 in the skin peripheral to the wound.*

   * *

   The authors should comment on the fact that miR-29 signal in the inset (at the edge of the wound) appears more basal than in the wound epidermis or in the unwounded__. – We have now inserted this suggestion and discussed it where appropriate (highlighted in cyan)__

   Quantifications and statistical analysis of the intensity and distribution of miR-29 for panels A and B and K10 for panel C will need to be included to help get a better sense of the data in its entirety and strengthen the observations.* – We agree with the reviewer that such quantifications would be extremely helpful. The nature of the miRNA FISH protocol relies on signal amplification, allowing detection of mature miRNA specifically despite their short length. We could not therefore rely on conventional methods to quantify the fluorescence reliably as it can only be interpreted relatively to other areas/sections stained at the same time. We have attempted to do the miRNA FISH without amplifying the signal by attaching the FITC probe directly to the miRNA-29 probe but the signal was too weak to reliably detect and quantify miRNA-29 expression in wounds.* Importantly, Figure 1C is described as staining after 6 days of skin equivalent cultures, but the same images are used in Figure 2A, where they are described as stainings after 11 days of culture. The authors should try to harmonise the data presentation so that the same data is not presented multiple times if possible. If repeated data presentation is necessary, it should be clearly stated and justified, and the authors should be careful to correctly indicate what the images represent. *– This has been corrected. *

   • ITGB1 stainings in Figure 2 do not convincingly match the statements in the main text ("miRNA-29 mimic-transfected SE struggled to attach through the integrin beta1 (ITGB1)-mediated adhesion__*"). – This should have been phrased rather as a suggestion. We detected virtually no integrin beta 1 in miRNA-29 overexpressing cells, which strongly suggested that high levels of miRNA-29 prevent ITGB1-mediated adhesion of keratinocytes to the basal membrane. *__

   * *

   All stainings, or at least the most important ones, like ITGB1, should have quantifications and statistical analyses of their intensity and distribution to support any observations. – We thank the reviewer for this comment and fully agree it would be ideal to have quantifications of all staining. We have tried to do so but were able to reliably quantify only BrdU, ITGB1, and ITGA6. The data has now been added to results and discussion.

   Staining of basement membrane proteins at 6 days could help better visualize if indeed there are any attachment defects in the mimic-overexpressing cells – We stained 6 day section for basement proteins collagen IV and laminin 5 but could not detect any differences in attachment (data added below). Since both keratinocytes and fibroblasts contribute to the epidermal-dermal adhesion on the BM, a more sophisticated method of detecting adhesion in human skin equivalents may be needed following miRNA-29 manipulation (e.g., electron microscopy of keratinocyte-BM contacts like hemidesmosomes).

   Since the authors use transient transfections, the significance ant interpretation of the stainings performed at 11 days will be reliant on the transfection strategy employed, the rate of proliferation of the cells, and the half-life of the proteins stained.

   The transfection strategy is not clearly explained (this is a more general problem, see below) and staining for miR-29 in these sections is necessary to ensure that the treatments are still in effect after this prolonged time in culture__. – We have now clarified the transfection protocol and added the quantification of miRNA-29 levels in skin equivalents at day 6 and day 11 (Figure S2D). The overexpression and the inhibition of miRNA-29 is still evident at day 6 and day 11, probably because of the high levels of miRNA mimics and the stabilizing chemistry of miRNA-29 anti-sense oligos (MOE-PS modifications).__

   • The mimic/inhibitor transfection strategy employed by the authors throughout the paper is not clearly explained and this is a very important detail to understand the results of many of the assays they perform. The methods and Figures S2/S3 describe a 'double transfection' transfected twice on D2 and D4 strategy for the inhibitors, but it is unclear if the same approach was used for the mimics (which is important since some of the experiments where they are employed have functional assays that can last longer than a week). Additionally, the strategy used for the inhibitors described in the methods section seems different than the one described in Figure S3. In the methods, the cells are transfected at day 1 and day 3 and collected for functional assays at day 5. Figure S3 instead shows two transfections at 'day 0' and an additional one at 'day 4' with miRNA levels measured at day 0 and day 8 (this bar plot should be modified to better reflect that measurements were only taken on specific days). The legend for Figure S3 reads "keratinocytes (P3/4) were transfected twice on subsequent days" and mentions "representative images of the cells from each treatment after the third transfection". This is all extremely confusing. The authors should make sure they explain what they did clearly and univocally, for both mimics and inhibitors, and they should add a time course with miR-29 levels following transfections of mimics and inhibitors covering the span of their longest assay. – We thank the reviewer for carefully checking the flow and apologize for the confusion. The successful transfection of primary keratinocytes with miRNA mimics is more straightforward than with the anti-sense oligos as the chemistry quite differ. Mimics go in as a ‘stem loop’ RNA structures ____and require only one transfection round. Anti-sense ‘inhibitors’ oligos (ASOs) are 15-16 nt single-stranded, ____phosphorothioate (PS)-methoxyethyl (MOE)-modified ASO____* require a double-transfection. This way, ASO remain in ‘fast’ cells for days and during adhesion assay as shown here.* The additional experiment for the cell viability and proliferation was following the 2nd transfection, which is now clarified in the text and in the Suppl. Figure S3.

   • Figure 3 includes reference to morphological parameters that would be predictive of a keratinocyte ability to form a holoclone (red arrows). While the larger size and low nucleus-to-cytoplasm ratio of differentiated cells is well-established, to my knowledge there is no accepted consensus about strong predictive capacity of simple morphological parameters when it comes to holoclone formation. The consensus regarding keratinocyte clonogenicity is generally missing in the field, relying primarily on early passage, low cytoplasm/nucleus ratio, and colony boundaries. Another important characteristic is the number of passages that the cells can undergo before they growth arrest or die. We are currently performing follow up experiments to characterize the miRNA-29 KD (abc) clones and consistently observe higher growth capacity (longevity) of the miRNA-29 depleted keratinocytes. This is also consistent with the data shown in Figure 3A and S3A.

   • The inhibition of miR-29 in experiment 1 of the growth factor depletion assay seems to have failed according to Figure S2C, so the results of experiment 1 (-GF) in Figure 3 should be disregarded and the experiment repeated. *We have disregarded the failed experiment and repeated adhesion assays under -GF conditions with more controls. While the improved adhesion upon depletion of miRNA-29 was reproducible, we also found that the growth factor depletion using a specific inhibitor of epidermal growth factor receptor (EGFR) AG-1478 abrogated the fast ____adhesion effect of miRNA-29 inhibition. It possibly means that miRNA-regulated adhesion requires EGF (but not other GF) signaling; however, more experiments would be needed to uncouple the role of GF in miRNA-29 adhesion. *

   • The authors report reduced keratinocyte differentiation in the miR-29 inhibited cells. This statement is mostly supported by the cell number time course shown in Figure S3B, but this experiment is not mentioned in the main text, which instead focuses on (less reliable) morphological parameters alone. Moreover, Figure S3 only shows the morphology of cells at day 4 and does not provide any information about the cell morphology at day 6 or day 8 as suggested by the main text. Assessing differentiation based on morphology alone is prone to inaccuracy and while the cell number experiment is good support for the stated decrease in differentiation in the miR-29 inhibited cells, it should be complemented with differentiation marker staining and/or clonogenicity assays. - *We agreed with the reviewer and made the appropriate changes in the text. Figure S3 has been updated as well, and we also ran a side analysis of differentiation markers (keratin K10 and loricrin). We found that miRNA-29 does not change significantly during keratinocyte differentiation in 2D (please, see the Support Figure A below). *

   • The authors' claim that their results "revealed the direct in vivo targetome and functions of miRNA-29 in three types of cells isolated from human skin" is not accurate. While their experiments are indeed compelling, they are performed in cultured primary cells grown for at least 3 passages, which are akin, but not the same as cells in vivo and may behave differently. – We agree and have changed this now in the text. On a similar note, while there is some evidence from mouse that miR-29 may intervene in the regulation of the wound healing response in keratinocytes in vivo (Figure 1A), no analogous in vivo data is presented for fibroblasts. The authors should consider showing miR-29 stainings of mouse dermal fibroblasts and the potential variation in its level during wound healing. - While this manuscript was in preparation, we were in the process of publishing our study showing the function of miRNA-29 in wound healing in cutaneous mouse-based model. This study shows the staining for miRNA-29 in mouse wounds during healing and includes the staining in dermal fibroblasts (____Robinson et al, Am. J. of Pathology 2024, Figure S1B____). We have isolated total RNA from mouse wounds at different points of healing and checked miRNA-29a/b levels using TaqMan assays. While we detected a change in miRNA-29 expression (Support Figure C, D), this possibly included miRNA-29 in the normal surrounding skin, inevitably present in a wound biopsy. __They should also show miR-29 staining of normal human skin to confirm that its expression pattern mimics the mouse. - __ We could not cite the other manuscript at that time, but it shows lower levels of miRNA-29 in dermal fibroblasts compared to keratinocytes in the epidermis by FISH (____Robinson et al, Am. J. of Pathology 2024, Figure S1B____). We also quantified levels of miRNA-29a/b in primary mouse keratinocytes and fibroblasts using TaqMan assays, and consistently with FISH, detected more miRNA-29 in keratinocytes (Support Figure B).__ The FISH for miRNA-29 in human skin was published earlier, also showing much lower signal of miRNA-29 in the dermis (Kurinna, S. Nuc. Acid Res. 2021, Supplementary Figure S3A). If possible, they could also 'wound' human skin explants and check what happens during re-epithelialisation to miR-29 expression and to the key targets they identified (explants may be challenging to obtain, though). These experiments could provide some more compelling (though inevitably correlative) suggestion that miR-29 could intervene in the wound healing response in vivo in humans__*. – This is a very good experiment suggested by the reviewer. The human skin explants were indeed challenging to obtain. We could only get a few sections of paraffin-embedded samples, which were suboptimal for miRNA-29 FISH. We included the data as Figure S1A. *

   Minor comments:

   • I would encourage the authors to avoid, when possible, the use of red/green colour palettes both in stainings and in graphs, as it makes the paper less accessible to colourblind individuals. – We sincerely apologise for the use of these colours in many stainings. We substituted red and green everywhere we could, but our technical capabilities did not permit changing colours on all Figures.

   • I would suggest avoiding the use of "stacked" bar plots to show data as they might lend themselves to misinterpretation. It would likely increase clarity if the bars for different conditions were plotted next to rather than on top of one another. - We replaced the stacked plots as suggested on Figures 3, 6, and Figure 8. We kept one stacked plot in Figure 6D to show variability in the nsa-treated samples for some mRNAs. The control samples on these plots were set to one (nsa) and the stacked part on top reflected the fold increase in mRNA levels after knock-down of miRNA-29 (abc).

   • The first inset in Figure 1B does not appear to match the box in the lower magnification image. – We moved the inset to the correct location.

   • The title of the section "Rescue of miRNA-29 mRNA targets improves basal adhesion of human keratinocytes" should be changed, as no rescue experiments are performed. The term is used again in the text when referring to targets upregulated (or "de-repressed") after miR-29 inhibition, but it is not accurate and should be changed__. – We followed the suggestion and highlighted changes throughout the text.__

   • The authors should specify the most important details of the adhesion assay in the Results section (for example the fact that the assay is carried out on fibronectin). – We added this to the Results.

   • The main text is imprecise when describing the RNAseq of fast/slow attaching keratinocytes, because it does not mention that the assay also includes miR-29 inhibition. - We have amended this and highlighted the changes in the text.

   • The insets in the middle of Figure 3 are not described in the figure legend and it is unclear what they are meant to be highlighting. The Authors should also double-check the accuracy of the scale bars across Figure 3A. - We described the insets in the legend and double-checked the scale bars in Figure 3A.

   • The pattern in the "abc" bars in Figure 3C makes it difficult to see the symbols *– We increased the font and adjusted the label. *

   • The area overlaps in the Venn diagram in Figure 4A should reflect the numbers. Since the diagram is comparing only three sets, accurate overlaps should improve the representation of the data. – We have re-created the Venn diagram to reflect the representation of the data on Figure 4A.

   • The colour scheme of the label borders in Figure 4E does not match the colour of set for the right-most sets in both keratinocyte and fibroblast Venn diagrams, leading to confusion. – We adjusted the colours to match the diagram in Figure 4E.

   • The figure legend for Figure 6E reads "Ingenuity Pathway Analysis (IPA) generated heat map of diseases and functions from the fast keratinocytes (abc) versus control (nsa)", but this is not what is displayed in the figure panel at all. - We apologise for the mistake; we corrected the legend.

   • The methods section for the miRNA-CLIP should include information about the number of cells used in each experiment. – The change is highlighted in the Methods.

   • The authors should carefully review the text for typos and misspellings and try to improve the readability of the manuscript__. – The manuscript has been carefully reviewed for these.__

   **Referees cross-commenting**

   I generally agree with the comments of the other reviewers: I think the paper is interesting and a valuable contribution to the field, particularly with regard to the role of miRNAs in the skin and the application of miRNA-CLIP to primary skin cells. While I did not remark on any gross overstatements, I agree that the data needs some strengthening to more adequately support some of the author's claims (I have tried to offer some realistic suggestions). There seems to be some difference of opinion regarding the data presentation, but all Reviewers thought it needed improvement in some capacity. While the way in which the paper is laid out and the results are displayed will be perceived subjectively by different readers, I believe it is in the best interest of the authors to try to reach the widest readership and thus I would maintain that the manuscript requires adjustments to increase clarity. I have tried to indicate specific sources of confusion and offer appropriate suggestions in my review.

   Reviewer #3 (Significance (Required)):

   This paper complements previous work that highlighted the role of miR-29 in desmosome formation in keratinocytes (Kurinna et al., 2014) and in skin repair in the mouse (Robinson et al., 2024), adding depth to these findings by understanding the molecular details of the key genes regulated by miR-29 in primary human skin cells. While the influence of miRNA on skin biology is well known, the details of which miRNAs and molecular mechanisms are involved are somewhat understudied. For this, I believe this paper, adequately amended, could be an interesting and useful contribution to the field and help highlight the role of miRNAs in the skin. This is also, to my knowledge, the first use of miRNA-CLIP in primary keratinocytes or fibroblasts and can provide a useful precedent for other studies looking to investigate miRNA interactomes in these cells.

   Read the original source
2. 

   Review Commons

   Nov 14, 2024

   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:

   Thiagarajan et al. report on the functions and molecular targets of miR-29 in human primary skin cells. They first focus on the potential role of miR-29 in wound healing and in the adhesion of keratinocytes to the basement membrane using both in vivo wounding assays in the mouse and human cultures/skin equivalents. The authors report that miR-29 negatively affects adhesion in vivo and in vitro and characterise the transcriptome of fast and slow-adhering cells with or without miR-29inhibition. They proceed to identify miR-29 targets in three primary skin cell types (follicular keratinocytes, interfollicular keratinocytes and …

   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:

   Thiagarajan et al. report on the functions and molecular targets of miR-29 in human primary skin cells. They first focus on the potential role of miR-29 in wound healing and in the adhesion of keratinocytes to the basement membrane using both in vivo wounding assays in the mouse and human cultures/skin equivalents. The authors report that miR-29 negatively affects adhesion in vivo and in vitro and characterise the transcriptome of fast and slow-adhering cells with or without miR-29inhibition. They proceed to identify miR-29 targets in three primary skin cell types (follicular keratinocytes, interfollicular keratinocytes and fibroblasts) by performing miRNA-clip. By comparing these targets to genes altered in keratinocytes with high adhesion capacity after miR-29 inhibition or fibroblasts after miR-29 inhibition, the authors describe a model in which miR-29 inhibits multiple adhesion-associated pathways in keratinocytes and negatively regulates proliferation and ECM deposition by dermal fibroblasts.

   Major comments:

   Overall, the paper is interesting, and the experiments performed are generally sensible for the questions being investigated. However, I thought the data was presented in a very confusing and unclear way, both in the main text and in the figures. I found the paper quite difficult to navigate, with contradictory statements between text and figures, cryptic or confounding graphs or arrangement of the figures and, in at least one instance, re-use of the same image with inconsistent labelling. The paper will thus greatly benefit from extensive tidying up and review of both text and figures to improve clarity. I highlight several points below, with many being related to this overarching issue, and I try to offer suggestions to the authors improve the quality of the manuscript.

   • The stainings in Figure 1A should be repeated in intact sections as it is difficult to understand the exact distribution of miR-29 when the whole epidermis appears to be falling apart in the section. It is possible to see the pattern the authors are describing based on the current images, but it is not convincing. The authors should comment on the fact that miR-29 signal in the inset (at the edge of the wound) appears more basal than in the wound epidermis or in the unwounded. Quantifications and statistical analysis of the intensity and distribution of miR-29 for panels A and B and K10 for panel C will need to be included to help get a better sense of the data in its entirety and strengthen the observations. Importantly, Figure 1C is described as stainings after 6 days of skin equivalent cultures, but the same images are used in Figure 2A, where they are described as stainings after 11 days of culture. The authors should try to harmonise the data presentation so that the same data is not presented multiple times if possible. If repeated data presentation is necessary, it should be clearly stated and justified, and the authors should be careful to correctly indicate what the images represent.
   • ITGB1 stainings in Figure 2 do not convincingly match the statements in the main text ("miRNA-29 mimic-transfected SE struggled to attach through the integrin beta1 (ITGB1)-mediated adhesion"). All stainings, or at least the most important ones, like ITGB1, should have quantifications and statistical analyses of their intensity and distribution to support any observations. Staining of basement membrane proteins at 6 days could help better visualise if indeed there are any attachment defects in the mimic-overexpressing cells. Since the authors use transient transfections, the significance ant interpretation of the stainings performed at 11 days will be reliant on the transfection strategy employed, the rate of proliferation of the cells, and the half-life of the proteins stained. The transfection strategy is not clearly explained (this is a more general problem, see below) and staining for miR-29 in these sections is necessary to ensure that the treatments are still in effect after this prolonged time in culture.
   • The mimic/inhibitor transfection strategy employed by the authors throughout the paper is not clearly explained and this is a very important detail to understand the results of many of the assays they perform. The methods and Figures S2/S3 describe a 'double transfection' strategy for the inhibitors, but it is unclear if the same approach was used for the mimics (which is important since some of the experiments where they are employed have functional assays that can last longer than a week). Additionally, the strategy used for the inhibitors described in the methods section seems different than the one described in Figure S3. In the methods, the cells are transfected at day 1 and day 3 and collected for functional assays at day 5. Figure S3 instead shows two transfections at 'day 0' and an additional one at 'day 4' with miRNA levels measured at day 0 and day 8 (this bar plot should be modified to better reflect that measurements were only taken on specific days). The legend for Figure S3 reads "keratinocytes (P3/4) were transfected twice on subsequent days" and mentions "representative images of the cells from each treatment after the third transfection". This is all extremely confusing. The authors should make sure they explain what they did clearly and univocally, for both mimics and inhibitors, and they should add a time course with miR-29 levels following transfections of mimics and inhibitors covering the span of their longest assay.
   • Figure 3 includes reference to morphological parameters that would be predictive of a keratinocyte ability to form a holoclone (red arrows). While the larger size and low nucleus-to-cytoplasm ratio of differentiated cells is well-established, to my knowledge there is no accepted consensus about strong predictive capacity of simple morphological parameters when it comes to holoclone formation.
   • The inhibition of miR-29 in experiment 1 of the growth factor depletion assay seems to have failed according to Figure S2C, so the results of experiment 1 (-GF) in Figure 3 should be disregarded and the experiment repeated.
   • The authors report reduced keratinocyte differentiation in the miR-29 inhibited cells. This statement is mostly supported by the cell number time course shown in Figure S3B, but this experiment is not mentioned in the main text, which instead focuses on (less reliable) morphological parameters alone. Moreover, Figure S3 only shows the morphology of cells at day 4 and does not provide any information about the cell morphology at day 6 or day 8 as suggested by the main text. Assessing differentiation based on morphology alone is prone to inaccuracy and while the cell number experiment is good support for the stated decrease in differentiation in the miR-29 inhibited cells, it should be complemented with differentiation marker staining and/or clonogenicity assays.
   • The authors' claim that their results "revealed the direct in vivo targetome and functions of miRNA-29 in three types of cells isolated from human skin" is not accurate. While their experiments are indeed compelling, they are performed in cultured primary cells grown for at least 3 passages, which are akin, but not the same as cells in vivo and may behave differently. On a similar note, while there is some evidence from mouse that miR-29 may intervene in the regulation of the wound healing response in keratinocytes in vivo (Figure 1A), no analogous in vivo data is presented for fibroblasts. The authors should consider showing miR-29 stainings of mouse dermal fibroblasts and the potential variation in its level during wound healing. They should also show miR-29 stainings of normal human skin to confirm that its expression pattern mimics the mouse. If possible, they could also 'wound' human skin explants and check what happens during re-epithelielisation to miR-29 expression and to the key targets they identified (explants may be challenging to obtain, though). These experiments could provide some more compelling (though inevitably correlative) suggestion that miR-29 could intervene in the wound healing response in vivo in humans.

   Minor comments:

   • I would encourage the authors to avoid, when possible, the use of red/green colour palettes both in stainings and in graphs, as it makes the paper less accessible to colourblind individuals.
   • I would suggest avoiding the use of "stacked" bar plots to show data as they might lend themselves to misinterpretation. It would likely increase clarity if the bars for different conditions were plotted next to rather than on top of one another.
   • The first inset in Figure 1B does not appear to match the box in the lower magnification image.
   • The title of the section "Rescue of miRNA-29 mRNA targets improves basal adhesion of human keratinocytes" should be changed, as no rescue experiments are performed. The term is used again in the text when referring to targets upregulated (or "de-repressed") after miR-29 inhibition, but it is not accurate and should be changed.
   • The authors should specify the most important details of the adhesion assay in the Results section (for example the fact that the assay is carried out on fibronectin).
   • The main text is imprecise when describing the RNAseq of fast/slow attaching keratinocytes, because it does not mention that the assay also includes miR-29 inhibition.
   • The insets in the middle of Figure 3 are not described in the figure legend and it is unclear what they are meant to be highlighting. The Authors should also double-check the accuracy of the scale bars across Figure 3A.
   • The pattern in the "abc" bars in Figure 3C makes it difficult to see the symbols.
   • The area overlaps in the Venn diagram in Figure 4A should reflect the numbers. Since the diagram is comparing only three sets, accurate overlaps should improve the representation of the data.
   • The colour scheme of the label borders in Figure 4E does not match the colour of set for the right-most sets in both keratinocyte and fibroblast Venn diagrams, leading to confusion.
   • The figure legend for Figure 6E reads "Ingenuity Pathway Analysis (IPA) generated heat map of diseases and functions from the fast keratinocytes (abc) versus control (nsa)", but this is not what is displayed in the figure panel at all.
   • The methods section for the miRNA-CLIP should include information about the number of cells used in each experiment.
   • The authors should carefully review the text for typos and misspellings and try to improve the readability of the manuscript.

   Referees cross-commenting

   I generally agree with the comments of the other reviewers: I think the paper is interesting and a valuable contribution to the field, particularly with regard to the role of miRNAs in the skin and the application of miRNA-CLIP to primary skin cells.

   While I did not remark on any gross overstatements, I agree that the data needs some strengthening to more adequately support some of the author's claims (I have tried to offer some realistic suggestions). There seems to be some difference of opinion regarding the data presentation, but all Reviewers thought it needed improvement in some capacity. While the way in which the paper is laid out and the results are displayed will be perceived subjectively by different readers, I believe it is in the best interest of the authors to try to reach the widest readership and thus I would maintain that the manuscript requires adjustments to increase clarity. I have tried to indicate specific sources of confusion and offer appropriate suggestions in my review.

   Significance

   This paper complements previous work that highlighted the role of miR-29 in desmosome formation in keratinocytes (Kurinna et al., 2014) and in skin repair in the mouse (Robinson et al., 2024), adding depth to these findings by understanding the molecular details of the key genes regulated by miR-29 in primary human skin cells. While the influence of miRNA on skin biology is well known, the details of which miRNAs and molecular mechanisms are involved are somewhat understudied. For this, I believe this paper, adequately amended, could be an interesting and useful contribution to the field and help highlight the role of miRNAs in the skin. This is also, to my knowledge, the first use of miRNA-CLIP in primary keratinocytes or fibroblasts and can provide a useful precedent for other studies looking to investigate miRNA interactomes in these cells.

   Read the original source
3. 

   Review Commons

   Nov 14, 2024

   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 #2

   Evidence, reproducibility and clarity

   Summary:

   The article entitled, "miRNA-29-CLIP uncovers new targets and functions to improve skin repair", by Thiagarajan et al. describes the characterization of the functions of miRNA-29 in keratinocytes and fibroblasts, its RNA interactors and potential mechanisms of action. Using candidate interactors and 2D cell culture and 3D skin equivalents combined with loss-of-function (inhibitor) and gain-of-function (mimic), and changes in expression analyses, the authors conclude that the major function of miRNA-29 is to regulate cell-substrate adhesion.

   Major comments:

   • While the interactors and expression changes are useful resources, …

   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 #2

   Evidence, reproducibility and clarity

   Summary:

   The article entitled, "miRNA-29-CLIP uncovers new targets and functions to improve skin repair", by Thiagarajan et al. describes the characterization of the functions of miRNA-29 in keratinocytes and fibroblasts, its RNA interactors and potential mechanisms of action. Using candidate interactors and 2D cell culture and 3D skin equivalents combined with loss-of-function (inhibitor) and gain-of-function (mimic), and changes in expression analyses, the authors conclude that the major function of miRNA-29 is to regulate cell-substrate adhesion.

   Major comments:

   • While the interactors and expression changes are useful resources, the claims and the conclusions that are based on them are exaggerated. The treatments are associated with changes in expression, but no functional data support the conclusions. Additional functional experiments are required to assertively make the claims. The title is misleading when stating "to improve skin repair" and the abstract also makes some bold general claims, which are tangentially supported by the findings. For example, "protein folding" only appears in the abstract and "RNA processing" is in the abstract and figures but not referred to in the text.
   • The authors may want to tune their language that their data suggest the conclusions as opposed to being definitive and assertive. This should be done in the Discussion, while the Results should represent the direct conclusions.
   • A couple of examples to the above, in the conclusion to section 1 of the Results, how was the "loss of basal adhesion" assessed? Is it by beta1-integrin localization changes? Also, how is "growth" defined"? proliferation is not changed and a more accurate way to describe the result is to refer to thickness.
   • The images in Fig 1C are reused in Fig 2A, where new examples should be shown instead.
   • Fig 1C and Fig 2A are not quantified to make the claims about premature differentiation and integrin expression changes.
   • Fig 3: It is not clear from the figure legends what statistical methods were used for which experiment or how many times the experiment was performed (not just biological replicates), especially given the variability among experiments in Fig 3C.

   Minor comments:

   • The Introduction is focused on methodology and should include elements that pave the way to the Results. Some information that belongs in the introduction are present in the Results section. In this respect, please define the miRNA processing Dicer pathway and its components in the introduction so that the reader can follow the nomenclature (AGO2, RISC, etc.). Also, introduce human skin equivalents or organotypic culture as a model system in the Introduction.
   • Some information in the Results belongs in the Introduction, for example, the first seven lines of the Results section.
   • The authors might want to consider including quantifications in the main figures, so they are immediately apparent to the reader, for example, Fig S1C. Also, Fig S2B is an important measure for the immediate outcome of the treatment on miRNA-29.
   • Please change "imidiate" to "immediate", "sculp" to "scalp", "has to be releaved of miRNA-29-mediate repression" to "has to be relieved of miRNA-29-mediated repression"

   Referees cross-commenting

   I agree with my colleagues' assessments and suggestions. The miRNA-CLIP data in keratinocytes and fibroblasts are important resources. The figures and text require reconsideration to more accurately represent the data as detailed in our collective reviews

   Significance

   The study utilizes 2D and 3D cultures and presents an important resource for miRNA-29 interactors in keratinocytes and fibroblasts, as well as the expression changes associated with its inhibition and overexpression. However, the conclusions are exaggerated and based on expression changes. If the conclusions are rephrased, the findings would be of interest to a broad audience interested in miRNA, cell adhesion and epithelial and mesenchymal biology.

   My expertise is in skin development and maintenance, genetics and cell biology. I have limited knowledge in RNA biology.

   Read the original source
4. 

   Review Commons

   Nov 14, 2024

   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

   miRNAs are important for the control of many cellular processes, with the miR-29 family of miRNAs implicated in the regulation of cell growth in different cell types in both the epidermis and dermis of the skin. However, the roles of miRNAs in specific cell types in general, and of the miR-29 family in the skin in particular, are currently unknown. Here, the authors use a range of cellular and molecular techniques, including miRNA cross-linking and immunoprecipitation (miRNA-CLIP) and antisense oligonucleotides (ASO), as well as RNASeq, qPCR, Western blotting, in situ hybridization, adhesion and ECM assays, ELISA and …

   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

   miRNAs are important for the control of many cellular processes, with the miR-29 family of miRNAs implicated in the regulation of cell growth in different cell types in both the epidermis and dermis of the skin. However, the roles of miRNAs in specific cell types in general, and of the miR-29 family in the skin in particular, are currently unknown. Here, the authors use a range of cellular and molecular techniques, including miRNA cross-linking and immunoprecipitation (miRNA-CLIP) and antisense oligonucleotides (ASO), as well as RNASeq, qPCR, Western blotting, in situ hybridization, adhesion and ECM assays, ELISA and immunofluorescence, to interrogate the roles of the miR-29 family of miRNAs in controlling cell growth in epidermal keratinocytes and dermal fibroblasts, using 2D and 3D ex vivo models. The coupling of miR-CLIP with functional assays allowed the authors to identify both miRNA-mRNA complexes, and the biological pathways that these ultimately manipulate.

   The authors report the identification of unbiased, tangible miR-29/mRNA pairs, together with functional roles in cell adhesion, ECM regulation and fibroblast proliferation, that are distinct between keratinocytes and fibroblasts. miR-29 is identified as a valuable target for interventions that seek to promote healthy skin regeneration, including applications for wound healing. Many of the pathways identified here have previously been described, but the novelty of this manuscript lies in the innovative combination of miR-CLIP with functional assays, the application of these in combination to specific cell types, the identification of miR-29 as a novel master regulator of epidermal keratinocyte adhesion via a range of different pathways, and the demonstration that miR-29 inhibition in fibroblasts can influence keratinocyte adhesion via paracrine signalling.

   The experiments are well designed and reported. The interpretations are sound and appropriate for the data presented (though see the comment on potential normalisation of ECM data to cell numbers in cultures for the miR-29 mimic/inhibitor data for fibroblasts and the query about the number of direct miR-29 targets in fibroblasts that are ECM-related).

   Major Comments:

   I have no major concerns to raise over this manuscript. The claims and conclusions are supported by the data and no additional experiments are required (though please note the comment on normalisation mentioned above and detailed below). The methods are clearly reported and statistical reporting is adequate.

   Minor Comments:

   Pg3, 7th line from the bottom: "processed into three functional miRNA..." - minor edit needed here, it looks like there's a word missing somewhere.

   Pg3, last line on the page: "results supported..." - is there a missing 'are' here?

   Pg5, 15th line of the main text: "of miRNA-29-mediate repression..." - is there a missing 'd' here ('-mediated...')? There is lots on minor presentation errors like this throughout the manuscript - I won't point them out exhaustively, but the manuscript needs a good thorough proof-read, maybe from a fresh pair of eyes?

   Fig. 1C: Can the figure be edited to better highlight the basal layer with lack of (nsm image) and expression of (abm image) K10? Maybe a box around that layer, rather than the current arrows only on the abm image (which are not particularly closely indicating the basal layer)?

   Fig. 2 legend should include definitions of abbreviations shown on the figure.

   Pg8/Fig. 4A: Can the reporting of shared transcript targets of miR-29 in IFK/HFK/DF cells be better communicated? Maybe just adding the actual percentage overlap in transcriptomes for IFK/HFL and keratinocytes/fibroblasts to the main text would help. Similarly, I think a direct report somewhere (in the main text?) of total number for relevant groups shown in Fig. 4E would also be useful - e.g. there are 45 transcripts that are direct targets of miR-29 in keratinocytes and also associated with ECM, and 190 that are direct targets of miR-29 in keratinocytes and also associated with cell adhesion, but these number are difficult to come by quickly at the moment. It would be nice to be able to quickly compare these numbers for keratinocytes to their equivalents for fibroblasts.

   Fig. 4B: It's interesting that ~15% of miR-29 binding targets identified using miR-CLIP are not predicted targets based on TargetScan/microT-CDS. I'd like to see a little more information on this added to the manuscript - perhaps listing some of these or including a table of them? And perhaps some discussion of this could be added also.

   Fig. 4E: I would be nice to see the Venn numbers for keratinocyte proliferation (either is a supp figure, or addition to the main text?), to help illustrate the lack of a role for miR-29 in the regulation of keratinocyte proliferation.

   Fig. 4E: Is the reported number of direct miR-29 targets in fibroblasts that are ECM-related correct? This number is reported as 10 in the main text (pg10, 3rd paragraph), but it looks like 10 is only for direct miR-29 targets in fibroblasts that are ECM-related AND related to proliferation. Should this number be 58? The 10 that are direct miR-29 targets in fibroblasts that are ECM-related AND related to proliferation can be reported in the next sentence, where this group is specifically referred to.

   Fig. 7 (and related main text): Did you take any steps to normalise ECM measurements to cell numbers present in cultures in the miR-29 mimic/inhibition experiments in fibroblasts? This should really be included as it would provide an answer to the speculation of whether the effects of manipulating miR-29 on ECM are due to proliferation or classical pro-fibrotic pathways.

   Fig. 8E: The upper and lower image need to have nsa/abc labels added to them.

   Pg12, 1st sub-heading: typo (cell-specifcic).

   Referees cross-commenting

   All reviews appear to be fair and balanced to me. I agree that in places wording could be amended to temper the strengths of some claims, and it would also be nice to see some additional functional assays included, to complement the adhesion and ECM deposition assays that are currently presented, though I do not think this should necessarily be a requirement for publication and could be included in subsequent follow-up work from the group. I did not spot the reuse of images between Fig. 1 and 2, but clearly this should be addressed - either by replacing one set of images, or by removing the relevant panels from Fig. 1 and changing in-text reference to guide the reader to Fig. 2A. I also agree that it would be nice to see miR-29 staining of mouse dermal fibroblasts during wound healing, to complement the images already shown for keratinocytes, and to see miR-29 staining in human skin.

   Significance

   miR-CLIP is a powerful, recently developed technique, with enormous promise for the identification of true miRNA-mRNA pairs, that has not yet been widely adopted by the research community. As such, its application here is itself relatively novel, adding enormously to our existing knowledge of likely miR-29 targets, providing tangible information in miR-29/mRNA pairs in specific cell types in different layers of the skin, but also further adding novel functional information to this, with demonstrations of the regulation of specific relevant biological pathways through manipulation of targets identified using miR-CLIP. The methods are sound (and impressive), results are reported well and not over-interpreted. There is the potential for better characterisation of the relative importance of canonical pro-fibrotic pathways vs proliferation-related effects on ECM production, and this should not be difficult to address. This paper will be on interest to a wide readership, including those engaged in fundamental research and clinicians.

   Read the original source
5. 

   Version published to 10.1101/2024.05.02.591622 on bioRxiv

   May 2, 2024