DNA methylome combined with chromosome cluster-oriented analysis provides an early signature for cutaneous melanoma aggressiveness
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Predicting if a tumour has aggressive or metastatic characteristics would be of great utility in the clinic as it would help patient stratification and management. In this manuscript, Carrier and collaborators derive a signature for melanoma aggressiveness relying on methylated regions of tumour and cell line genomes. The identification of a 4-gene methylation biomarker for melanoma aggressiveness and survival is an important contribution. This manuscript is of relevance to clinicians and melanoma researchers interested in biomarker research.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Cancer Biology (eLife)
Abstract
Aberrant DNA methylation is a well-known feature of tumours and has been associated with metastatic melanoma. However, since melanoma cells are highly heterogeneous, it has been challenging to use affected genes to predict tumour aggressiveness, metastatic evolution, and patients’ outcomes. We hypothesized that common aggressive hypermethylation signatures should emerge early in tumorigenesis and should be shared in aggressive cells, independent of the physiological context under which this trait arises. We compared paired melanoma cell lines with the following properties: (i) each pair comprises one aggressive counterpart and its parental cell line and (ii) the aggressive cell lines were each obtained from different host and their environment (human, rat, and mouse), though starting from the same parent cell line. Next, we developed a multi-step genomic pipeline that combines the DNA methylome profile with a chromosome cluster-oriented analysis. A total of 229 differentially hypermethylated genes was commonly found in the aggressive cell lines. Genome localization analysis revealed hypermethylation peaks and clusters, identifying eight hypermethylated gene promoters for validation in tissues from melanoma patients. Five Cytosine-phosphate-Guanine (CpGs) identified in primary melanoma tissues were transformed into a DNA methylation score that can predict survival (log-rank test, p=0.0008). This strategy is potentially universally applicable to other diseases involving DNA methylation alterations.
Article activity feed
-
-
Author Response
Public Evaluation Summary:
Predicting if a tumour has aggressive or metastatic characteristics would be of great utility in the clinic as it would help patient stratification and management. In this manuscript, Carrier and collaborators derive a signature for melanoma aggressiveness relying on methylated regions of tumour and cell line genomes. The identification of a 4-gene methylation biomarker for melanoma aggressiveness and survival is an important contribution. This manuscript is of relevance to clinicians and melanoma researchers interested in biomarker research.
We would like to specify that the methylation of 5 CpGs was identified as potential signature of the aggressiveness and survival of melanoma in primary tumors. Three are the key and original findings of the study: 1/ the observation that robust DNA …
Author Response
Public Evaluation Summary:
Predicting if a tumour has aggressive or metastatic characteristics would be of great utility in the clinic as it would help patient stratification and management. In this manuscript, Carrier and collaborators derive a signature for melanoma aggressiveness relying on methylated regions of tumour and cell line genomes. The identification of a 4-gene methylation biomarker for melanoma aggressiveness and survival is an important contribution. This manuscript is of relevance to clinicians and melanoma researchers interested in biomarker research.
We would like to specify that the methylation of 5 CpGs was identified as potential signature of the aggressiveness and survival of melanoma in primary tumors. Three are the key and original findings of the study: 1/ the observation that robust DNA methylation traits of aggressiveness are independent of the physiological context; 2/ the methodology of combining DNA methylome analysis and chromosome cluster-based analysis that can be applied beyond melanoma; and 3/ the identification of the methylation of 5 CpGs (and not genes) that provide a predictive value of the aggressiveness of the melanoma in primary tumors.
Reviewer #1 (Public Review):
In this manuscript, Carrier and collaborators derive a methylation signature for melanoma aggressiveness from the sequential analyses on various cell lines in different organisms and test it in a set of primary and metastatic melanoma tumours. However, I think that some of the claims are a little premature as a broader sample size would need to be tested to assess the signature robustness and applicability.
Strengths
- The approach the authors take is innovative and I agree with their premise that genes that make cells be more aggressive should be detected across different organisms.
- Different organisms were evaluated.
- Figures are illustrative and the narrative is very clear.
We thank the referee for the comments
Weaknesses
- The sample size is small. In my opinion, a broader and more diverse set of samples would need to be tested if authors suggest making a diagnostic kit with the genes in their signature
We agree with the referee and we are pursuing the project with the aim to develop a assay measuring the 5 CpG simultaneously that can be easily used by dermatologists and professionals. But this is beyond the scope of the manuscript, which reports an original strategy and the novel findings listed above.
- A more comprehensive comparison with what other authors have found when doing similar studies would be needed to put in context their results.
We have clarified this point in the revised version. Only two of the five CpGs that we have identified are available in public datasets. The complete signature of the five CpGs has not been analyzed in other studies.
Reviewer #2 (Public Review):
Carrier et al. sought to define the methylome associated with increased aggressiveness of melanoma, with the goal of identifying common changes in methylation and to define a methylation signature of disease progression. To do so, they analyzed 3 cell line pairs that either were established from the same patient (primary vs cutaneous metastasis) or that were a parental cell line and its derivatives generated through repeated transplantation and selection for the ability to metastasize to the lung. Among these pairs, 229 genes were identified as commonly hypermethylated. Interestingly, genomic mapping of these genes revealed that 74 of these genes localized to 9 methylation clusters, 34 of which had two CpGs and at least 40% differential methylation. Carrier et al. also performed Ingenuity pathway analysis, uncovering 116 genes among the 229 with putative cancer-associated functions. From these genes, 8 candidates were selected for validation in cell lines and patient samples. 4 out of 8 genes (MYH1, PCDHB16, PCDHB15, BCL2L10) showed differential methylation in patient samples, and their methylation status correlated with patient overall survival. Carrier et al. then devised a score based on methylation of these 4 genes, which performed better in predicting patient prognosis based on primary tumor methylation score than did the Breslow index. This methylation score could therefore be used as a biomarker of melanoma aggressiveness and this approach could be implemented in other tumor types.
Overall, the approach appears to be well designed, the results are of good quality, and generally support the claims. For some aspects of the paper, the rationale is not immediately apparent and should be better described, for instance the choice of the 8 genes selected or validation appears arbitrary and the cut-off long term vs short term survival of patients (1 year) is not justified clinically or scientifically. Providing additional information will make this study clearer for the reader.
We thank the referee for the comments. We have clarified these two points in the revised version. The 8 genes were chosen because they were distributed on chromosomes with clusters of methylation, correspond to peaks of hypermethylation and have a potential function role in cancer formation. One year was chosen because at the time of the study it was the average overall survival of diagnosticated cutaneous melanoma.
Reviewer #3 (Public Review):
The authors propose that the DNA methylation signature of tumor aggressiveness would be independent of the physiological context: starting from a human tumor, shared signatures relevant to aggressiveness should emerge independent of whether this trait was acquired in humans or whether cells have been implanted into rats or mice. In a multi-step selection process, they identified hypermethylated sites common to the most aggressive melanoma forms, analyzed the distribution of these sites in the genome, and validated these methylation peaks in cell lines and patient samples.
The weakness is related to the use of murine cells and also to The Functional annotation and pathway analysis. The list of hypermethylated genes was imported into QIAGEN's Ingenuity® Pathway Analysis (IPA®, QIAGEN Redwood City, www.qiagen.com/ingenuity). I am wondering if it would be more appropriate to use other platforms to explore the data.
No murine cells were used in the study but human derived cells. This was clarified in the text. Other platforms were used (Panther.db, Kegg pathway, DAVID, …) but the results were more complete with IPA.
The strengths are related to the main strategy that identified a DNA methylation signature of five CpG sites in four gene promoters in primary tumors that could predict the overall survival of the patients and thus has potential diagnostic application. This strategy, which overcomes heterogeneity in tumors due to the environment, can potentially be generalized to other cancers involving DNA methylation alterations
The authors combined analysis of the DNA methylome with the chromosomal location. The multistep strategy developed and used to identify differentially methylated genes predicting aggressiveness is originally identified as a common pattern or a specific signature of melanoma aggressiveness. The unique approach used in this study yielded a potential DNA methylation signature that correlates with outcomes.
The description of a novel multistep approach allowed identifying a methylation signature of five CpGs in primary melanoma tissues that has the potential to predict survival outcomes in cutaneous melanoma patients. This integrated approach can be applied not only to other cancer types but also to other diseases or biological processes such as aging and development.
We greatly appreciate the comments of the referee that underline the strengths of our study.
-
Evaluation Summary:
Predicting if a tumour has aggressive or metastatic characteristics would be of great utility in the clinic as it would help patient stratification and management. In this manuscript, Carrier and collaborators derive a signature for melanoma aggressiveness relying on methylated regions of tumour and cell line genomes. The identification of a 4-gene methylation biomarker for melanoma aggressiveness and survival is an important contribution. This manuscript is of relevance to clinicians and melanoma researchers interested in biomarker research.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
In this manuscript, Carrier and collaborators derive a methylation signature for melanoma aggressiveness from the sequential analyses on various cell lines in different organisms and test it in a set of primary and metastatic melanoma tumours. However, I think that some of the claims are a little premature as a broader sample size would need to be tested to assess the signature robustness and applicability.
Strengths
- The approach the authors take is innovative and I agree with their premise that genes that make cells be more aggressive should be detected across different organisms.
- Different organisms were evaluated.
- Figures are illustrative and the narrative is very clear.Weaknesses
- The sample size is small. In my opinion, a broader and more diverse set of samples would need to be tested if authors …Reviewer #1 (Public Review):
In this manuscript, Carrier and collaborators derive a methylation signature for melanoma aggressiveness from the sequential analyses on various cell lines in different organisms and test it in a set of primary and metastatic melanoma tumours. However, I think that some of the claims are a little premature as a broader sample size would need to be tested to assess the signature robustness and applicability.
Strengths
- The approach the authors take is innovative and I agree with their premise that genes that make cells be more aggressive should be detected across different organisms.
- Different organisms were evaluated.
- Figures are illustrative and the narrative is very clear.Weaknesses
- The sample size is small. In my opinion, a broader and more diverse set of samples would need to be tested if authors suggest making a diagnostic kit with the genes in their signature
- A more comprehensive comparison with what other authors have found when doing similar studies would be needed to put in context their results. -
Reviewer #2 (Public Review):
Carrier et al. sought to define the methylome associated with increased aggressiveness of melanoma, with the goal of identifying common changes in methylation and to define a methylation signature of disease progression. To do so, they analyzed 3 cell line pairs that either were established from the same patient (primary vs cutaneous metastasis) or that were a parental cell line and its derivatives generated through repeated transplantation and selection for the ability to metastasize to the lung. Among these pairs, 229 genes were identified as commonly hypermethylated. Interestingly, genomic mapping of these genes revealed that 74 of these genes localized to 9 methylation clusters, 34 of which had two CpGs and at least 40% differential methylation. Carrier et al. also performed Ingenuity pathway analysis, …
Reviewer #2 (Public Review):
Carrier et al. sought to define the methylome associated with increased aggressiveness of melanoma, with the goal of identifying common changes in methylation and to define a methylation signature of disease progression. To do so, they analyzed 3 cell line pairs that either were established from the same patient (primary vs cutaneous metastasis) or that were a parental cell line and its derivatives generated through repeated transplantation and selection for the ability to metastasize to the lung. Among these pairs, 229 genes were identified as commonly hypermethylated. Interestingly, genomic mapping of these genes revealed that 74 of these genes localized to 9 methylation clusters, 34 of which had two CpGs and at least 40% differential methylation. Carrier et al. also performed Ingenuity pathway analysis, uncovering 116 genes among the 229 with putative cancer-associated functions. From these genes, 8 candidates were selected for validation in cell lines and patient samples. 4 out of 8 genes (MYH1, PCDHB16, PCDHB15, BCL2L10) showed differential methylation in patient samples, and their methylation status correlated with patient overall survival. Carrier et al. then devised a score based on methylation of these 4 genes, which performed better in predicting patient prognosis based on primary tumor methylation score than did the Breslow index. This methylation score could therefore be used as a biomarker of melanoma aggressiveness and this approach could be implemented in other tumor types.
Overall, the approach appears to be well designed, the results are of good quality, and generally support the claims. For some aspects of the paper, the rationale is not immediately apparent and should be better described, for instance the choice of the 8 genes selected or validation appears arbitrary and the cut-off long term vs short term survival of patients (1 year) is not justified clinically or scientifically. Providing additional information will make this study clearer for the reader.
-
Reviewer #3 (Public Review):
The authors propose that the DNA methylation signature of tumor aggressiveness would be independent of the physiological context: starting from a human tumor, shared signatures relevant to aggressiveness should emerge independent of whether this trait was acquired in humans or whether cells have been implanted into rats or mice. In a multi-step selection process, they identified hypermethylated sites common to the most aggressive melanoma forms, analyzed the distribution of these sites in the genome, and validated these methylation peaks in cell lines and patient samples.
The weakness is related to the use of murine cells and also to The Functional annotation and pathway analysis. The list of hypermethylated genes was imported into QIAGEN's Ingenuity® Pathway Analysis (IPA®, QIAGEN Redwood City, www.qiagen.co…
Reviewer #3 (Public Review):
The authors propose that the DNA methylation signature of tumor aggressiveness would be independent of the physiological context: starting from a human tumor, shared signatures relevant to aggressiveness should emerge independent of whether this trait was acquired in humans or whether cells have been implanted into rats or mice. In a multi-step selection process, they identified hypermethylated sites common to the most aggressive melanoma forms, analyzed the distribution of these sites in the genome, and validated these methylation peaks in cell lines and patient samples.
The weakness is related to the use of murine cells and also to The Functional annotation and pathway analysis. The list of hypermethylated genes was imported into QIAGEN's Ingenuity® Pathway Analysis (IPA®, QIAGEN Redwood City, www.qiagen.com/ingenuity). I am wondering if it would be more appropriate to use other platforms to explore the data.
The strengths are related to the main strategy that identified a DNA methylation signature of five CpG sites in four gene promoters in primary tumors that could predict the overall survival of the patients and thus has potential diagnostic application. This strategy, which overcomes heterogeneity in tumors due to the environment, can potentially be generalized to other cancers involving DNA methylation alterations
The authors combined analysis of the DNA methylome with the chromosomal location. The multistep strategy developed and used to identify differentially methylated genes predicting aggressiveness is originally identified as a common pattern or a specific signature of melanoma aggressiveness. The unique approach used in this study yielded a potential DNA methylation signature that correlates with outcomes.
The description of a novel multistep approach allowed identifying a methylation signature of five CpGs in primary melanoma tissues that has the potential to predict survival outcomes in cutaneous melanoma patients. This integrated approach can be applied not only to other cancer types but also to other diseases or biological processes such as aging and development.
-
