Review Reports - Integrative Multi-Omics Approach in Vascular Ehlers–Danlos Syndrome: Further Insights into the Disease Mechanisms by Proteomic Analysis of Patient Dermal Fibroblasts

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript by Chiarelli et al further expands and develops their previous findings obtained in the identical experimental system. They explore dermal fibroblasts from patients with vEDS. In the paper published in January of 2024 they used transcriptomic analysis. In the current manuscript the authors apply proteomic analysis in order to get more evidence to support the already published data.

Technically the study was performed sufficiently good. However, I still have few small comments:

First, the authors studied ONLY skin fibroblasts for any intracellular alterations that can be caused by dominant mutations within Col3a1 gene aiming at elucidating biological consequences in the vasculature, for example, in the arteries. How do the authors expand their findings obtained with skin fibroblasts to the fibroblasts in the arterial wall? Are these cell types so similar that the conclusions obtained with one cell types can by directly used for the cell from the anatomically different location? I think the authors must show either single-cell RNA data or bulk RNA data for the skin and arterial fibroblasts or thouroughly discuss this issue referring to the already published data (the best woiuld be single-cell RNA-seq data).

Second, the fibroblasts that were used for the study were not freshly isolated. But they were cultured in vitro quite extensively (at least, until confluence; and I'm wondering if, on top of this, the cells were additionally split during the culture). Considering that in the organs like skin or arteries the cells normally do not proliferate (fibroblasts, endothelial cells, inflammatory cells, smooth muscle) this in vitro procedure would certainly affect transcriptomic as well as proteomic landscape in the fibroblasts. Therefore I would rather say that the data by the authors only partially reflect the processes that are naturally occuring within the fibroblasts in the skin or in the arteries. I would suggest the authors to show somehow if the culture conditions in vitro combined with intensive proliferation DO NOT affect (or DO) transcriptome and/or proteome of the fibroblasts. And then having these data to build their respective conclusions, such as those biochemical and cellular pathways mentioned in the Results and Discussion.

Third, do the authors 100% believe that the vascular Ehlers-Danlos syndrome has purely fibroblastic nature and develops only due to intracellular events in the fibroblasts, like mutations in the Col3a1 gene cause alterations of miR-29b, and this affects expression of various mRNAs or proteins? The authors also meention that mutated variants of Col3a1 affects ECM structure, which means that biochemical events in the patient's fibroblasts may not be the primary cause of the pathobiology of the vEDS. The authors should somehow demonstrate the relevance of their approach by showing that mutated (vs. normal) Col3a1 protein do (or dop not) show any effect on the endothelial cells. This can be some type of co-culture of fibroblasts with endothelial cells and collagen (abnormal vs. normal) or finding and discussing the published data (if available) or combination of both approaches.

Fourth, mutations in Col3a1 can be different. If the authoprs believe that the primary cause of vEDS is the abundance of miR-29b, affected by Col3a1, then they should suggest a theory (or provide experimental data) that would place a scientific background to this linkage between mutated Col3a1 and miR-29b. Do various types of mutation in the gene trigger the very similar downstream event - changes in the miR-29b expression?

Author Response

Comments and Suggestions for Authors

Author reply: Dear Reviewer, thank you for your comments and feedback on our manuscript. We appreciate your acknowledgment of the technical robustness of our study.

However, I still have a few small comments:

First, the authors studied ONLY skin fibroblasts for any intracellular alterations that can be caused by dominant mutations within Col3a1 gene aiming at elucidating biological consequences in the vasculature, for example, in the arteries. How do the authors expand their findings obtained with skin fibroblasts to the fibroblasts in the arterial wall? Are these cell types so similar that the conclusions obtained with one cell type can by directly used for the cell from anatomically different location? I think the authors must show either single-cell RNA data or bulk RNA data for skin and arterial fibroblasts or thoroughly discuss this issue referring to the already published data (the best would be single-cell RNA-seq data).

Author reply: Thank you for your insightful comments and for raising this important point regarding the applicability of our findings from skin fibroblasts to fibroblasts in the arterial wall. We agree that fibroblasts from different anatomical locations, such as the skin and arterial wall, may exhibit distinct transcriptional and functional profiles due to their unique microenvironments and physiological roles. While our study focused on skin fibroblasts due to their accessibility and relevance to vEDS pathology, we acknowledge the importance of addressing whether these findings can be extrapolated to arterial fibroblasts. We emphasize that our primary aim was to explore intracellular alterations due to dominant COL3A1 mutations, which are expected to broadly affect collagen synthesis and extracellular matrix (ECM) organization in fibroblasts, irrespective of their origin. For example, using primary patient fibroblast cultures as a cellular model, it was recently demonstrated that ER stress may be a key molecular mechanism contributing to the vEDS. By treating patients’ fibroblasts with the chemical chaperone 4-phenylbutyric acid, the researchers successfully reduced ER stress, restored collagen folding, and mitigated both intracellular and extracellular defects (https://www.biorxiv.org/content/10.1101/2024.06.20.599980v1). However, we acknowledge that cell-specific adaptations or responses may influence the manifestation of these alterations in arterial fibroblasts. Although we recognize that dermal fibroblasts may not perfectly replicate all vascular defects seen in vEDS, as other ECM-producing cells like vascular smooth muscle cells (VSMCs), endothelial cells, and adventitial fibroblasts may exhibit different gene expression profiles, our cellular model can still provide valuable insights. Considering the multisystemic nature of the disease that affects different cell types, and since vEDS patients exhibit significant cutaneous involvement and skin fragility like defects in vascular tissues, analyzing gene/protein expression in dermal fibroblasts may provide helpful clues regarding molecular pathways that could aid in explaining disease causes and guide future treatments. While we do not currently have access to single-cell RNA-seq or bulk RNA-seq data for arterial fibroblasts from vEDS patients, future studies integrating such data could strengthen our conclusions.

Second, the fibroblasts that were used for the study were not freshly isolated. But they were cultured in vitro quite extensively (at least, until confluence; and I'm wondering if, on top of this, the cells were additionally split during the culture). Considering that in organs like skin or arteries the cells normally do not proliferate (fibroblasts, endothelial cells, inflammatory cells, smooth muscle) this in vitro procedure would certainly affect transcriptomic as well as proteomic landscape in the fibroblasts. Therefore, I would rather say that the data by the authors only partially reflect the processes that are naturally occurring within the fibroblasts in the skin or in the arteries. I would suggest the authors to show somehow if the culture conditions in vitro combined with intensive proliferation DO NOT affect (or DO) transcriptome and/or proteome of the fibroblasts. And then having these data to build their respective conclusions, such as those biochemical and cellular pathways mentioned in the Results and Discussion.

Author reply: We appreciate the reviewer’s insightful comment regarding the potential impact of in vitro culture conditions and extensive proliferation on the transcriptomic and proteomic profiles of fibroblasts. This is indeed a thought-provoking point that we recognize as relevant to interpreting our findings. However, due to the rarity of vascular Ehlers-Danlos syndrome and the severe complications associated with this condition, particularly vascular fragility, it is currently challenging to obtain additional fibroblast strains from new patients for further studies. Additionally, the use of cells from the vascular system, which might serve as a more direct cellular model to explain vEDS-associated vascular complications, is largely inaccessible due to these same limitations. We acknowledge that our data only partially reflect the altered biological processes that naturally occur in dermal fibroblasts or vascular cells. As the reviewer has pointed out, the culture conditions used in vitro, combined with the proliferative behavior of fibroblasts in these environments, could indeed influence the transcriptomic and proteomic landscape. In tissues like skin and arteries, fibroblasts, endothelial cells, smooth muscle cells, and inflammatory cells typically exhibit low proliferation rates under physiological conditions. Culturing these cells extensively in vitro, especially through multiple passages, may alter their molecular profiles, making them less representative of their in vivo counterparts. While this is an important consideration, we currently lack the means to conduct additional experiments to determine whether or how the in vitro culture conditions specifically affect the transcriptome or proteome of fibroblasts from vEDS patients. However, we agree that such analyses would provide valuable insights and could strengthen the conclusions regarding the biochemical and cellular pathways discussed in our study. This limitation reinforces the need for future studies to assess the extent to which culture-induced proliferation influences cellular profiles and to better contextualize how these findings translate to the natural biology of fibroblasts and other cell types involved in vEDS. We remain committed to pursuing this line of inquiry in the future when access to additional patient-derived fibroblast strains becomes available. While our study provides important initial insights into the molecular alterations in vEDS fibroblasts, we recognize the inherent limitations of using extensively cultured fibroblasts as a model. We appreciate the reviewer’s suggestion and consider this a valuable direction for future research aimed at better understanding the complex biology of vEDS in its native tissue environment.

Author reply: We appreciate the reviewer’s perspective and acknowledge the multisystemic nature of the disease, which impacts multiple cell types beyond fibroblasts. However, it is important to note that no causal relationship has been established between the loss of COL3A1 function and the dysregulation of miR-29b expression. Our previous and current findings emphasize the significance of understanding the functional impact of ncRNAs in vEDS pathogenesis. Dysregulated expression of miR-29b may impact ECM remodeling and vascular angiogenesis and contribute to the pathogenesis of abdominal aortic aneurysms and vasculopathies such as Marfan syndrome (MFS) (PMID: 26965051, PMID: 37007957, PMID: 38895538). Interestingly, an epigenetic regulation in aortic aneurysm development involving miR-29b was described in a murine model of MFS. In this experimental model, it was demonstrated that increased expression of miR-29b and decreased COLL gene expression augmented aneurysm growth, while inhibition of miR-29b and increased COLL expression reduced aneurysm formation (PMID: 28455451). Our expression and protein data align with these findings, suggesting a critical role for miR-29b in abnormal ECM remodeling and downstream processes that may affect cell survival. These pilot data provide a foundation for future studies to elucidate the biological role of miR-29b in vEDS. They also underscore the importance of characterizing ncRNAs in vEDS pathogenesis, with implications for identifying potential miRNA and lncRNA signatures associated with aberrant disease pathways. Our findings reinforce the notion that altered miRNA and lncRNA expression plays a pivotal role in vEDS etiopathology, emphasizing the broad post-transcriptional regulatory effects triggered by single-gene mutations. This highlights an additional layer of molecular complexity within this pleiotropic condition. However, our study specifically focuses on identifying a microRNA that influences the homeostasis and organization of ECM components, primarily expressed in connective tissues with elastic properties such as the dermis, arteries, and gastrointestinal tract. This underscores the intricate pathological processes at multiple levels, making it challenging to define the disease solely through fibroblast involvement. Further research is essential to unravel these interactions across affected tissues.

Unfortunately, due to time constraints for the manuscript revision, we are unable to conduct the suggested experiments. Additionally, we do not currently have access to endothelial cells to culture and evaluate the impact of collagen type III alterations. While we acknowledge the value of such analyses, they fall beyond the scope of this study and the resources available at present. We appreciate your understanding and will consider these suggestions for future research endeavors.

Fourth, mutations in Col3a1 can be different. If the authors believe that the primary cause of vEDS is the abundance of miR-29b, affected by Col3a1, then they should suggest a theory (or provide experimental data) that would place a scientific background to this linkage between mutated Col3a1 and miR-29b. Do various types of mutation in the gene trigger the very similar downstream event - changes in the miR-29b expression?

Author reply: We appreciate the reviewer's comment and would like to clarify that we do not believe the primary cause of vEDS is merely an increased expression of miR-29b or potentially other miRNAs. Rather, our findings and interpretations emphasize the complexity of the disease, which arises from multiple levels of expression and regulation that extend far beyond mutations in a single gene such as COL3A1. The pathophysiology of vEDS is highly intricate and involves dysregulation at genetic, epigenetic, and molecular levels. Mutations in COL3A1 have diverse effects depending on the nature of the mutation that may vary involving collagen synthesis, assembly, secretion, and ECM integrity. In our studies, all enrolled patients harbored COL3A1 glycine substitutions and in-frame exon skipping, which are the most common types of mutations that cause the disease and are associated with a more severe clinical presentation. These mutations disrupt ECM homeostasis, impairing vascular and tissue stability and leading to characteristic features of the syndrome, including arterial dissections, aneurysms, and gastrointestinal ruptures. However, the downstream effects of COL3A1 mutations likely involve a cascade of regulatory changes, including alterations in non-coding RNAs like miR-29b, which modulate ECM-related processes such as collagen remodeling and vascular angiogenesis. While we have observed dysregulated miR-29b expression in our study, we consider it part of a broader network of pathological changes rather than a standalone causal factor. Dysregulation of miRNAs and lncRNAs likely represents a secondary effect of the underlying genetic mutation, contributing to the pleiotropic and multisystemic nature of the disease. Moreover, the phenotypic diversity observed in vEDS may be influenced by the type and location of COL3A1 mutations, which might trigger varying downstream events, including differential miRNA expression profiles. These profiles could vary between patients, tissues, or stages of disease progression, further underscoring the complexity of vEDS pathogenesis. It is unlikely that all COL3A1 mutations result in identical downstream effects, but rather they converge on shared pathological pathways, such as ECM disorganization and impaired vascular integrity, through a multifaceted regulatory network. In this context, our study does not claim a direct causal relationship between COL3A1 mutations and increased miR-29b expression. Instead, we propose that altered miR-29b levels reflect one aspect of the disease’s broader regulatory landscape. The interplay between miRNAs, lncRNAs, and coding RNAs, along with their post-transcriptional and epigenetic modifications, likely contributes to the complex pathogenesis of vEDS. This underscores the need for further studies to dissect the regulatory networks involved in vEDS and to explore how various COL3A1 mutations interact with non-coding RNAs to influence disease mechanisms. Such research could pave the way for novel insights into the molecular complexity of vEDS and ultimately inform therapeutic approaches targeting these intricate regulatory pathways.

Reviewer 2 Report

Comments and Suggestions for Authors

In the present study, Chiarelli et al. focused on the molecular mechanisms that play a key role in this disease by performing a comprehensive proteomic analysis of dermal fibroblasts from vEDS patients and healthy individuals. Moreover, they investigated the expression level and role of microRNAs (e.g. miR-29b-3p) in fibroblasts. The authors could show that inhibition of miR-29b-3p in patient fibroblasts increased the protein levels of extracellular matrix components, i.e. the fibrillar collagens type V and I. The authors concluded that microRNAs targeting could be a promising treatment strategy to restore the structural integrity of the extracellular matrix in vEDS patients.

General Comments

This is an interesting paper with a clear structure and presentation, which describes the molecular mechanism underlying the vascular Ehlers-Danlos Syndrome (vEDS). The experiments are straightforward and well performed.

Minor comments

The authors could somewhere mention that collagens type V and I are fibrillar collagens.

Line 29: It should read: Notably, the inhibition…..

L 430: It should read:………. three miR-29b-3p putative target……

Author Response

Comments and Suggestions for Authors

The vascular Ehlers-Danlos Syndrome (vEDS) is a rare genetic connective tissue disorder associated with the risk of dissection and rupture of certain organs. vEDS is caused by COL3A1 mutations that code for collagen type 3, a major component of the extracellular matrix of the vascular system and hollow organs. In the present study, Chiarelli et al. focused on the molecular mechanisms that play a key role in this disease by performing a comprehensive proteomic analysis of dermal fibroblasts from vEDS patients and healthy individuals. Moreover, they investigated the expression level and role of microRNAs (e.g. miR-29b-3p) in fibroblasts. The authors could show that inhibition of miR-29b-3p in patient fibroblasts increased the protein levels of extracellular matrix components, i.e. the fibrillar collagens type V and I. The authors concluded that microRNAs targeting could be a promising treatment strategy to restore the structural integrity of the extracellular matrix in vEDS patients.

General Comments

Author reply: We thank the reviewer for their thoughtful comment and for highlighting key aspects of our findings. We sincerely thank the reviewer for their positive and encouraging feedback on our manuscript. We are pleased to hear that the structure, presentation, and experimental design were clear and well-received. Your acknowledgment of the quality and relevance of our work is greatly appreciated, and we are encouraged by your comments to continue exploring the molecular mechanisms underlying vascular Ehlers-Danlos Syndrome.

Minor comments

The authors could somewhere mention that collagens type V and I are fibrillar collagens.

Line 29: It should read: Notably, the inhibition…..

L 430: It should read:………. three miR-29b-3p putative target…

Author reply: We thank the reviewer; the revised manuscript was modified accordingly.

Reviewer 3 Report

Comments and Suggestions for Authors

Total 9 patients and 9 controls were studied. But only 3 patients were studied for inhibition of the microRNA. The authors may like to give a reason on how these 3 patients were selected. And were them representative of the changes in the rest 6 patients.

Some questions raised while reading the manuscript,

1. line 170. Differentially expressed proteins (DEPs) with a log2 difference ≥ ±1 and a -log10 q-value <0.05 were considered significantly enriched.

What is the meaning of a -log10 q-value <0.05 ?

Is there a typo?

2. There is no mention of the normalisation procedure for the proteomic data. Please supplement some information or algorithm used.

3. Also which algotihm is used to generate values in the Diff. protein analysis worksheet from the original matrix or raw data in table S2.

4. What is the proposed mechanism of activation of the miRNA miR-29b-3p in from patients with vascular Ehlers-Danlos syndrome. May be the authors already mentioned in their previous paper. However, it is good to mention in the paper.

Author Response

Comments and Suggestions for Authors

The authors used fibroblasts from patients with vascular Ehlers-Danlos syndrome to study transcriptome, miRNA expression and proteomics changes in this disease. In addition, inhibition of miR-29b-3p was performed in-vitro to see if some phenotype of the cultured fibroblasts could be reversed or rescued. Total 9 patients and 9 controls were studied. But only 3 patients were studied for inhibition of the microRNA. The authors may like to give a reason on how these 3 patients were selected. And were them representative of the changes in the rest 6 patients.

Author reply: We appreciate the reviewer’s insightful comment regarding the selection of the three patient fibroblast strains used for the miR-29b-3p inhibition experiments. To clarify, all patients were previously molecularly characterized for structural mutations in COL3A1, which are the most observed mutations in vEDS patients and are representative of the broader patient cohort included in this study. Additionally, all patient-derived fibroblast strains underwent immunofluorescence analysis, which confirmed the typical disorganization of ECM components and alterations in integrin expression. Specifically, we observed disruptions in the fibrillar collagen components, including type I, III, and V collagens (COLLI, COLLIII, COLLV), elastin (ELN), fibrillins (FBNs), fibronectin (FN), glycosaminoglycan chains (GAGs), and core proteoglycans such as perlecan, versican, and decorin. Integrin expression analysis further confirmed the absence of α2β1 and α5β1 integrins, alongside the presence of αvβ3 integrin. These analyses indicate that the selected fibroblast strains are representative of the general patient population in terms of their ECM defects and integrin expression, supporting the relevance of our findings. The three fibroblast strains selected for the miR-29b-3p inhibition experiments were chosen to ensure that they reflect the typical mutation profile seen in the majority of vEDS patients. Therefore, the biological effects observed upon miR-29b-3p inhibition in these three strains should be considered representative of the findings in the remaining patient fibroblasts. We hope this explanation addresses the reviewer’s concern and clarifies that the selected fibroblast strains are indeed representative of the larger cohort, ensuring the validity of the results for the broader population of vEDS patients included in this study.

Some questions raised while reading the manuscript,

line 170. Differentially expressed proteins (DEPs) with a log2 difference ≥ ±1 and a -log10 q-value <0.05 were considered significantly enriched.

What is the meaning of a -log10 q-value <0.05?

Is there a typo?

Author reply: We apologize for the typo, which was amended. Thank you for pointing that out, and we appreciate your understanding.

There is no mention of the normalisation procedure for the proteomic data. Please supplement some information or algorithm used. 3. Also, which algorithm is used to generate values in the Diff. protein analysis worksheet from the original matrix or raw data in table S2.

Author reply: Thank you for your observation regarding the normalization procedure and the algorithm used for the proteomic data. In the original document we did not explicitly detail the raw data processing and normalization steps. We have now added this information to the revised manuscript. Specifically, we indicated that raw mass spectrometry data were uploaded in Spectronaut (18.3.23), ran in library-free search to determine protein identity and quantity. The MS2 level was used for the quantification, whilst a cross-run normalization was applied. Protein assignment was performed by correlating spectra with the UniProt Homo sapiens database (UP000005640). Searches were conducted with tryptic specifications and default settings for mass tolerances for both MS and MS/MS spectra. Carbamidomethylation was set as a fixed modification, while oxidation and N-terminal acetylation were defined as variable modifications. The minimum and maximum peptide masses were set respectively to 7 and 470 Da, and the false discovery rate (FDR) for proteins and peptide-spectrum matches to 1%. Protein quantities were exported and additional elaborated (Perseus 1.6.2.3). Specifically, a first level of data filtering was applied to exclude contaminant proteins/peptides, reverse entries as well as the identification only by a modified peptide. The signal/noise values were normalized using log2 transformation, while the protein abundances were grouped according to experimental conditions. Missing values have been replaced by random numbers that are drawn from a normal distribution. Differences between two groups were inspected using an unpaired Student’s t-test and features changes with q-value < 0.05 and log2 fold change≥ ±1 were considered significant.

What is the proposed mechanism of activation of the miRNA miR-29b-3p in patients with vascular Ehlers-Danlos syndrome. May be the authors already mentioned in their previous paper. However, it is good to mention in the paper.

Author reply: We thank the reviewer for raising this important point. Currently, we do not know of a specific pathological mechanism linking the activation of miR-29b-3p in vascular Ehlers-Danlos syndrome. However, insights from other vascular disorders, such as Marfan syndrome, suggest that microRNA dysregulation, including the upregulation of miR-29b, may contribute to vascular complications through the regulation of extracellular matrix remodeling and angiogenesis. As mentioned in our response to Reviewer 1, it is important to clarify that we do not propose a direct causal relationship between COL3A1 loss of function and the dysregulation of miR-29b-3p. This underlines the complexity of vEDS, where the involvement of epigenetic modifications, such as altered miRNA expression, extends beyond the mutations in a single causative gene. Further research efforts are needed to elucidate these pathobiological aspects and understand how such epigenetic factors may contribute to the disease process.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors adequately responded to all my four concerns.