Exploring the Functional Basis of Epigenetic Aging in Relation to Body Fat Phenotypes in the Norfolk Island Cohort

Round 1

Reviewer 1 Report

In this manuscript, Cao et al. analyzed blood-based gene transcripts and intragenic CpG methylation data from 74 healthy individuals (male 50, female 24) from the Norfolk Island population and tried to link aging-related methylation sites to body fat phenotypes. This type of study has the potential to provide information of significance, but the small sample size used in this study makes the conclusions highly questionable. While I feel that the results are too preliminary to be considered for publication, I would like to leave it to the editor and other reviewer(s), who I hope have better expertise in statistics, to decide whether there would be any value in publishing such preliminary data.

Author Response

Reviewer 1:

In this manuscript, Cao et al. analyzed blood-based gene transcripts and intragenic CpG methylation data from 74 healthy individuals (male 50, female 24) from the Norfolk Island population and tried to link aging-related methylation sites to body fat phenotypes. This type of study has the potential to provide information of significance, but the small sample size used in this study makes the conclusions highly questionable. While I feel that the results are too preliminary to be considered for publication, I would like to leave it to the editor and other reviewer(s), who I hope have better expertise in statistics, to decide whether there would be any value in publishing such preliminary data.

Response: With a sample size of 74 subjects that have methylation, expression and phenotype data this study has been able to detect a set of age-related CpGs that reached genome-wide statistical significance and has linked these to gene expression and body fat phenotypes at suggestive levels of statistical significance. To further support the statistical associations, we have included additional levels of evidence including i) the requirement of adjacent CpGs associated with outcomes in the same direction, which is unlikely to occur by chance and ii) biological annotation from curated public databases and literature searches.  We have also taken care to ensure the interpretation of the findings is not overstated use terms such as "exploring". We firmly believe these findings are worthy of publication since they reveal new hypotheses with which researchers can test in more focused experiments in future.

Reviewer 2 Report

I checked your manuscript and described comments below..

Methylation sites (CpGs) related to aging are very important because they are involved in diseases. This paper does a very good job of researching in the Norfolk Island cohort.

This paper has the following problems.

1.       The author uses R for data analysis. It would be better to describe specifically what kind of library or program used. The current content does not allow us to see the reproducibility of the data.

2.       There are Supplementary Tables S1, S2, and S3 in the text, but they have not been uploaded, so I cannot confirm them.

3.       There is an error in the footnote to Table 3 on page 9.

^bknown age-related CpG (No = 0; Yes =1According to largest effect size (R2) the Collagen Type

-> known age-related CpG (No = 0; Yes =1)

According to largest effect size (R2) the Collagen Type

I don't think this paper has any major mistakes or grammatical problems.

Author Response

Reviewer 2:

The author uses R for data analysis. It would be better to describe specifically what kind of library or program used. The current content does not allow us to see the reproducibility of the data.

Response: As suggest we have now included specifics of the methods including the library and packages used. This included the missMDA package for the imputation of missing phenotype values followed by using FactoMineR package (version 1.41) to perform the Principal Component Analysis. The methylation data was processed using the default function via champ.filter through the ChAMP package. Gene expression data were taken as previously described in Benton MC et al. [31]. All association analysis was performed using the base R function.

There are Supplementary Tables S1, S2, and S3 in the text, but they have not been uploaded, so I cannot confirm them.

Response:  We have uploaded as Supplementary document.

There is an error in the footnote to Table 3 on page 9.

Response: Thanks for pointing out the error. This have now been fixed on the footnote in Table 3.

Reviewer 3 Report

The authors have investigated the relationship between DNA methylation, obesity, and aging.  They compared the age related methylation of various genes to a set of criteria normally thought to describe obesity.  While the results are interesting as the authors point out it is difficult to determine the significance of the genes identified.  As a minor point the title of the actual manuscript is not the same as the title on the review report. 

Author Response

Reviewer 3:

 

The authors have investigated the relationship between DNA methylation, obesity, and aging. They compared the age-related methylation of various genes to a set of criteria normally thought to describe obesity. While the results are interesting, as the authors point out it is difficult to determine the significance of the genes identified.

Response: As mentioned, we have taken care to frame this study as exploratory (or hypothesis generating) to allow future independent research. We have also used existing biological annotation from curated public databases and literature searches to add credibility to the key genes identified in terms of relation to body fat phenotypes.

Reviewer 4 Report

The manuscript entitled “Exploring the functional mechanisms of epigenetic aging in relation to body fat phenotypes in the Norfolk Island cohortl” by Van Cao et al. is generally well written (there are some minor typos/grammatical errors - I recommend being consistent in English style throughout) and figures were well done and easy to interpret.  While the sophistication of the statistical analyses are outside of my expertise, the methods and description of the results was easy to understand.  I only have a few comments to share that may improve the manuscript.

1.  Is Figure 1 necessary?  While it is not unclear, it is very simple yet takes up a lot of room. This could be redone to be smaller.  But if the journal is not concerned about space, then it is fine.

2. The legend for Figure 2 is not very clear.  "Genes were shown as the most significant in each chromosome."?  Do you mean, the genes shown in boxes are the most significant?

3.  Were no genes identified on the X or Y chromosome?  I did not find any of the supplemental data referenced in the text was provided to the reviewers to see if this was true, or if genes on the sex chromosomes were not examined.  It is interesting since there are many parameters related to sex that are highly significant.

4. What do the different colors of the bars in Figure 3 represent?  It is a bit misleading given the colors in Figure 2.  Took me a second to realize that there is no correlation. Is this necessary?  Or, since the authors make a point that over half of the genes that are methylated are in regulatory regions, maybe color those one color and the body/exon a different color?  The figure legend could be a bit more detailed, but the explanation in the text was sufficient.

5. Again, this is my lack of expertise, but Figure 4 doesn't seem well explained. Is this even necessary to include?  I didn't see any mention of the enrichment results in the discussion or why this figure is germane to the manuscript. If this remains, can this Figure or the results be discussed more thoroughly?

In general, the manuscript was well-written, which was greatly appreciated. I noted "ageing" was used in the abstract while the rest of the manuscript wrote "aging". I also saw "tumour" used (last line, paragraph 1, page 2). I didn't make a note of any other instances. Neither of these is incorrect; recommend choosing one style and being consistent throughout. 

Author Response

Reviewer 4:

1. Figure 1 necessary?  While it is not unclear, it is very simple yet takes up a lot of room. This could be redone to be smaller.  But if the journal is not concerned about space, then it is fine.

Response: The intention of figure 1 is to provide the reader with an initial summary of the study design and workflow and to contextualise how gene expression and methylation data will be analysed and linked with age and obesity-related phenotypes. We have since updated the figure to include more detail.

2. The legend for Figure 2 is not very clear.  "Genes were shown as the most significant in each chromosome."?  Do you mean, the genes shown in boxes are the most significant?

Response: We have now changed the statement "Genes were shown as the most significant in each chromosome." to “Gene names shown in boxes indicated the most significant genes  in each chromosome”.

3. Were no genes identified on the X or Y chromosome? I did not find any of the supplemental data referenced in the text was provided to the reviewers to see if this was true, or if genes on the sex chromosomes were not examined. It is interesting since there are many parameters related to sex that are highly significant.

Response: In our analysis, we chose to exclude methylation CpGs on the sex chromosomes and perform sex-specific analyses because the total sample size was already quite modest and further separating into males and females would diminish statistical power to unacceptably low levels.

 

4. What do the different colors of the bars in Figure 3 represent?  It is a bit misleading given the colors in Figure 2. Took me a second to realize that there is no correlation. Is this necessary? Or, since the authors make a point that over half of the genes that are methylated are in regulatory regions, maybe color those one color and the body/exon a different color? The figure legend could be a bit more detailed, but the explanation in the text was sufficient.

Response: The colour in the figure 3 has no correlation with the colour in the Figure 2. To clarify we have updated the figure 3 to just highlight the regulatory regions and the figure legend.

5. Again, this is my lack of expertise, but Figure 4 doesn't seem well explained. Is this even necessary to include?  I didn't see any mention of the enrichment results in the discussion or why this figure is germane to the manuscript. If this remains, can this Figure or the results be discussed more thoroughly?

Response:  We have removed figure 4 as deemed unnecessary

 

Comments on the Quality of English Language

In general, the manuscript was well-written, which was greatly appreciated. I noted "ageing" was used in the abstract while the rest of the manuscript wrote "aging". I also saw "tumour" used (last line, paragraph 1, page 2). I didn't make a note of any other instances. Neither of these is incorrect; recommend choosing one style and being consistent throughout. 

Response: The ageing has been corrected as “aging”.

I have also checked the word “tumour” throughout. These are all consistently written as “tumour”.

Round 2

Reviewer 1 Report

I am not completely satisfied with the response but agree with authors that the preliminary study has potential to provide information for future researchers to perform more focussed experiments.

Reviewer 2 Report

I checked your manuscript and described comments below.

I pointed out the following problems in the previous version.

1.       The author uses R for data analysis. It would be better to describe specifically what kind of library or program used. The current content does not allow us to see the reproducibility of the data.

2.       There are Supplementary Tables S1, S2, and S3 in the text, but they have not been uploaded, so I cannot confirm them.

3.       There is an error in the footnote to Table 3 on page 9.

^bknown age-related CpG (No = 0; Yes =1According to largest effect size (R2) the Collagen Type

-> known age-related CpG (No = 0; Yes =1)

According to largest effect size (R2) the Collagen Type

I have confirmed that points 1 and 3 have been corrected.

Regarding point 2, I don't think there is any particular problem with the supplementary file.

I don't think this paper has new various major mistakes or grammatical problems.