Review Reports - Facial Skin Microbiome: Aging-Related Changes and Exploratory Functional Associations with Host Genetic Factors, a Pilot Study

Round 1

Reviewer 1 Report (Previous Reviewer 2)

The authors have improved the manuscript according to the majority of suggestions and now the presented information is of much higher quality.

Author Response

We thank the reviewer for its/her positive comment.

Reviewer 2 Report (New Reviewer)

It is interesting and comprehensive work. Following minor comments, please check it out.

Table 1 In Smoke column, what is the meaning Ex? Please explain about the word.

Figure 1–3, 8 Please improve the figure color and increase font size or high resolution figure. Especially, these colors are difficult to understand color blindness. Please select universal colors.

Figure 4, 3D axis (PC1, PC2, PC3) more thicker and darker. Increase font size and resolution of Figure 4.

Figure 7 and 12, Text is not clear, improve high resolution or increase font size, please.

Figure 13, increase font size of asterisk.

Table 3, I suggest the author change of the gray nucleotides to black. In addition, p-value should be a period, not a comma.

Line 40,46, 70, 227, 380, 384, 407, 493, skin microbiota => SM or SM => skin microbiota. Please unify the wording.

Line 65 MEFV, line 67 IBD, please describe more like as line 68 “FUT2, which is linked to Crohn’s disease”.

Line 129, Does NGS mean 16S metagenomic sequence? Or, once spell out the NGS as next generation sequence.

Line 133, Does the DNA yield have any unit for example ng or µg?

Table 2, % column, Is the Genus 0.962451 value 96.25%, or 0.96%?

Line 162, Is the DOI information an reference article about analysis or a microbial database on the Web?

Line181, SNP fist spell out then abbreviate.

Line 215, I have no access to the supplemental material and therefore cannot determine if this paper is legitimate.

Line217 and Figure 1, In the text, classified into four phyla, but figure 1 mentioned five phyla. I suggest five phyla in body text is preferable.

What is a difference between Figure 1 and Figure 3? It looks almost the same.

Line 176, 278, 327, PICRUST2 => PICRUSt2 ?

Line 284-285, 325, Figure 6, 7, 11 and 12, please translate KEGG orthology numbers to the other words, those number are not informal for readers without Table S3 (I guess that probably some explanations are in table S3).

Figure 8-10, Please unify the wording. Figure 8, Dry skin (left), Normal (right), Figure9, Dry (left), Not dry (right), Figure 10, normal (left), Dry skin (right).

Line 413, Period before “Additionally, “ looks red, but better black.

Author Response

Table 1 In Smoke column, what is the meaning Ex? Please explain about the word.

“Ex” means “ex smoker”. We added the meaning in the text.

Figure 1–3, 8 Please, improve the figure color and increase font size or high resolution figure. Especially, these colors are difficult to understand color blindness. Please select universal colors.

According to the rightful observation of the reviewer, the appearance of every figure of the article has been improved. Moreover, the groups color has been changed according to the reviewer suggestion.

Figure 4, 3D axis (PC1, PC2, PC3) more thicker and darker. Increase font size and resolution of Figure 4.

According to the rightful observation of the reviewer, the appearance of every figure of the article has been improved.

Figure 7 and 12, Text is not clear, improve high resolution or increase font size, please.

According to the rightful suggestion of the reviewer, the appearance of every figure of the article has been improved.

Figure 13, increase font size of asterisk.

The asterisk is now bigger and then more visible.

Table 3, I suggest the author change of the gray nucleotides to black. In addition, p-value should be a period, not a comma.

We corrected the Table 3 accordingly

Line 40,46, 70, 227, 380, 384, 407, 493, skin microbiota => SM or SM => skin microbiota. Please unify the wording.

We corrected in the text accordingly

Line 65 MEFV, line 67 IBD, please describe more like as line 68 “FUT2, which is linked to Crohn’s disease”.

We corrected in the text

Line 129, Does NGS mean 16S metagenomic sequence? Or, once spell out the NGS as next generation sequence.

We clarified the meaning of NGS

Line 133, Does the DNA yield have any unit for example ng or µg?

We corrected it in the text

Table 2, % column, Is the Genus 0.962451 value 96.25%, or 0.96%?

We thank the reviewer for this remark. Those percentages are now written properly in the table

Line 162, Is the DOI information an reference article about analysis or a microbial database on the Web?

We added the reference in the text, however as the original endnote formatting is no more present in the text, it appears as a double reference 1. We will correct it with the typesetter

Line181, SNP fist spell out then abbreviate.

We corrected it in the text

Line 215, I have no access to the supplemental material and therefore cannot determine if this paper is legitimate.

Apologies to the reviewer, we have added supplementary material

Line217 and Figure 1, In the text, classified into four phyla, but figure 1 mentioned five phyla. I suggest five phyla in body text is preferable.

We corrected it in the text

What is a difference between Figure 1 and Figure 3? It looks almost the same.

In the Figure 3 the samples have been regrouped in order to highlight the similarity and differences among the groups.

Line 176, 278, 327, PICRUST2 => PICRUSt2 ?

We corrected it in the text

We agree with the reviewer. Now, every plot regarding the predicted KO functions figures the description of the function instead of the code

Figure 8-10, Please unify the wording. Figure 8, Dry skin (left), Normal (right), Figure9, Dry (left), Not dry (right), Figure 10, normal (left), Dry skin (right).

According to the reviewer suggestion, the wording in the plots are now unified.

Line 413, Period before “Additionally, “ looks red, but better black.

We corrected it in the text

Reviewer 3 Report (New Reviewer)

This is a pilot study on microbiota in women of different ages. Being a pilot study with so few participants no real conclusions can be made, especially where age is concerned. I have some major comments:

1. Methods: It is not explained how the age groups were defined. It seems randomly? A woman aged 52 may gave been in menopause for a few years, while a woman aged 53 may still not have reached full menopause. The women around this age may be very different among themselves, even when falling in the same category, and vice versa may be quite alike hormonally but fall into different categories.

Smoking, sun-exposure and hydration may have a more significant role in microbiota than absolute age. If these are taken into account you actually have diverse participants.

I suggest that you focus on those factors more than on age. Age results can be given next to these data.

2. Discussion: You do not really explain your results. Why would participants from young and old group be more alike in microbiota than the middle aged group?

Paients A5, B4 and C5, very different ages, have almost the same microbiota. Likewise B2 and C2 (41 and 54 y old and different categories). It seems that age is not the only explanatory factor here, and the paper should focus on microbiota in general, but with less significance given to age since your methods do not allow for any robust conclusions.

You mention that having three groups is a strength, but the groupings should be divided to better reflect differences between age groups.

Author Response

Methods: It is not explained how the age groups were defined. It seems randomly? A woman aged 52 may have been in menopause for a few years, while a woman aged 53 may still not have reached full menopause. The women around this age may be very different among themselves, even when falling in the same category, and vice versa may be quite alike hormonally but fall into different categories.

We thank the reviewer for this critical observation, we recruited the subjects in agreement with their age group. In addition, we carried out an analysis taking into account the menopause factor, we did not show the data as we did not obtain significant differences.

Smoking, sun-exposure and hydration may have a more significant role in microbiota than absolute age. If these are taken into account you actually have diverse participants. I suggest that you focus on those factors more than on age. Age results can be given next to these data.

Thank you for the right suggestion, but in this pilot study we focused only in host skin characteristics as hydration since we considered smoke and sun exposure as environmental factors, such as other factors that we did not considered as diet, lifestyle etc .

Discussion: You do not really explain your results. Why would participants from young and old group be more alike in microbiota than the middle aged group?

In this pilot study we decided to show this intriguing result that surely needs to be further explored in a more large cohort of subjects to give a rational explanation, being different external and intrinsic factors that could impact on the skin microbiota modulation .

Surely other external factors can affect skin microbiota composition (please, see answer one and two) that we will take into account in a future larger study. However, we decided to show our intriguing pilot results, taking into account the limitation due to the number of patients.

Reviewer 4 Report (New Reviewer)

This is an interesting study to explore facial skin microbiome and ageing-related changes and exploratory functional associations with host genetic factors.

However, supplementary Materials and Methods, Figures not available for review.

A few points below need to be improved.

The figures in the manuscript are not clear. The quality of figures needs to be improved.

Suggest moving Functional pathway analysis Method and Results from supplementary to the main manuscripts.

Needs to mark Figure 2A, 2B, 2C inside Figure 2 and Figure 2 legend.

Line 387 and 397. References 38 was placed twice in the same place.

Author Response

The figures in the manuscript are not clear. The quality of figures needs to be improved.

According to the rightful observation of the reviewers, the appearance of every figure of the article has been improved.

Suggest moving Functional pathway analysis Method and Results from supplementary to the main manuscripts.

The functional pathway analysis Method and results are already in the main text. We just reported tables in the supplementary material

Needs to mark Figure 2A, 2B, 2C inside Figure 2 and Figure 2 legend.

We thank the reviewer for this suggestion. The group characteristics are now further explained in the plot.

Line 387 and 397. References 38 was placed twice in the same place.

We corrected the text

Round 2

Reviewer 3 Report (New Reviewer)

The authors have not responded to my comments. The paper is only changed as to typos, quality of figures and small technical issues.

None of the points I took up have been changed so that other readers wondering on the limitations of the study.

It is not sufficient to explain the issues I mention to me alone. Changes to the text should also be made for all readers to understand those issues.

Author Response

We apologies for the misunderstanding. According to the suggestion of the reviewer, we inserted his/her the right suggestion also in the text.

In detail:

-In the methods section added details on the age group definition

--In the paragraph 3.3 we explained that in this pilot study, we focused only on host factors that could impact skin microbiota, such as hydration, neglecting external factors, such as smoking or sun exposure will be considered in a larger future study

-In the discussion section we added : Intriguingly, we also observed that subjects A5, B4 and C5, belonging to different age groups, have a similar skin microbiota composition; likewise B2 and C2. This preliminary result suggests that ageing could be not the only explanatory factor that we will consider in a fine microbiota evaluation. -Finally, we further improved the "limit" section: “However, we present the study as a preliminary research with original but pioneering data that paves the way to further insights into a greater number of subjects, being different external and intrinsic factors that could impact on the skin microbiota modulation

Round 3

Reviewer 3 Report (New Reviewer)

The authors have adequtely addressed my points.

Minor: There is an extra reference after the references that is numbered 1. Could the authors check the references for correctness?

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Thank you for sharing the interesting study. My main concern is that the number of individuals are too low to find significance according to age and skin dryness from the results.

Author Response

We thank the reviewer for its critical comment, and in agreement, we are aware of this limit; we remark it in the limitation paper section. In addition, to stress this issue more, we added in the title that our manuscript is a pilot study. Based on these original preliminary data, we are conducting a future study on a larger number of patients.

Reviewer 2 Report

The presented study investigated age related changes as well as functional and genetic associations of facial skin microbiome and the authors attempt to correlated the data with host collagen metabolism and oxidative stress. The submitted manuscript is well written and structured. The topic of the work could have significance and be of interest to the scientists in the field as the study is designed in an original manner presenting both microbial and host related factors. There are however several major points not addressed in the current manuscript and additional revisions are advised.

1. The major point to address is the authors declaration and discussion on the low sample size of the recruited cohort which gets further reduced by stratifying for age and skin type. The authors are advised to include a standard information/discussion on the limitations of the study including all study limitations (low study sample size, sampling method used to select the sample, definition of healthy, variability between individuals/microbiota…)

2. The authors should explain the presented data in more detail:

2.1. The quantity and quality of the obtained DNA

2.2. The average number of reads per sample and the reads range of the collected DNA samples and some basic information of the quality criteria used during the sequencing

2.3. Table 2 could be more informative if data on individual samples are presented. It is not clear how is this table useful for the insight into the data.

2.4. The methodology related to PCoA analyses should be explained in detail, either with literature recommending “per cent normalization of count data” or with the explanation if the compositional nature of the data was taken into account.

2.5. The use of hydration levels (2 categories) instead of using actual measured values could have value for certain analyses but it is not clear why the authors choose not to use actual measured values for association/correlation analyses.

3. Additionally, it would be useful for the readers if genes mentioned (discussed) are stated with their full names/functions, where this is not the case.

4. In general, the data presentation is not very reader friendly and it would be recommended to add more detail in Figure/Table captions and footnotes, especially in the supplementary document. Also, some data are missing in Figure 5 in the Families plot or some other error occurred. The readers would benefit from using gene designations rather than KO entry numbers to better appreciate the results presented.

5. 1. In my opinion, the genetic association of facial skin microbiota to host gene variation is the weakest part of the study. This is primarily due to the fact that an extremely low sample size is used. Also, the definition of healthy for the presented group is lacking and the selection of the host genes of interest is not sufficiently substantiated. The significance of selected genes for the selected participant population is not clearly presented so the information is highly speculative and potentially misleading as the authors do not present the other potential confounding factors (data from obtained questionnaires on lifestyle, sun exposure, sun protection and skincare are not presented nor associated with host gene variation).

6. The authors should avoid statements that they studied the mechanisms of skin aging as their study did not have the temporal component. This is particularly sensitive in the conclusion section where this should be re-phrased as in the current wording this is completely misleading. The same is true for the word, pioneer, which should be avoided.

Minor revisions needed:

1. Typo line 40 (MS vs SM), line 44 (hos or host)?

2. The sentence in line 76 should be rephrased to state that the skin of healthy women was studied, and which exactly functional pathways were studied.

3. Line 387, word reported should be replaced with there was or any other more appropriate statement.

Author Response

The major point to address is the author’s declaration and discussion on the low sample size of the recruited cohort which gets further reduced by stratifying for age and skin type. The authors are advised to include a standard information/discussion on the limitations of the study including all study limitations (low study sample size, sampling method used to select the sample, definition of healthy, variability between individuals/microbiota…)

According to the reviewer’s suggestion, we improved the limitation section of manuscript expanding the discussion about the critical points of this pilot study.

The authors should explain the presented data in more detail:

2.1. The quantity and quality of the obtained DNA

According to the reviewer’s suggestion, we added more info in method section: “The mean DNA yield was 33.45 ± 16.22 and the UV absorbance 260/280 ratio was greater than 1.86 (260/230 ratio over 2.11)”

2.2. The average number of reads per sample and the reads range of the collected DNA samples and some basic information of the quality criteria used during the sequencing

We thank the reviewer for this right suggestion. The number of reads before and after the FASTQ processing is now reported in the designated table (Supplementary material Table S1). Regarding the quality criteria for processing, we reported the arbitrary parameters chosen in paragraph 2.3 (e.g. trimming point and identity threshold for taxonomy) according to reads quality, as specified. The rarefaction analysis, performed to check the sample saturation after the reads discarded by DADA2, has been added to supplementary materials (Figure S1).

2.3. Table 2 could be more informative if data on individual samples are presented. It is not clear how is this table useful for the insight into the data.

We thank the reviewer for the comment because we have the possibility to better explain this point. The table purpose is to report the percentage of reads assigned to a certain taxon in the SILVA database. We think it is an interesting information because it displays both the low percentage of unassigned reads as well as the quality of 16S sequencing and processing, due to the use of a dedicated taxonomic database with a high (0.99) identity alignment threshold.

In agreement with the reviewer’s comment, we further specified the methodology regarding the beta diversity analysis. The compositional nature of this microbial dataset has been already taken in account by using the Hellinger transformation (addressed in the paper as “relative abundances adjusted with square root transformation”) which is reported to be appropriated for this kind of data (e.g. DOI: 10.1016/B978-0-444-53868-0.50007-1). In order to account for the drawbacks of different index, we checked the reported the result with both euclidean distance, which in this case corresponds to the Hellinger asymmetric distance, in hierarchical clustering and Bray-Curtis similarity in PCoA ordination. Finally, following the reviewer’s suggestions, we added the previous mentioned reference to the paper.

We thank the reviewer for this adequate comment. Due to the possibility of having false positives in correlation analyses, linked to low sample numbers, we preferred to perform only correlations whose results can be further checked by a logical and biologic critical interpretation. Accordingly, we chosen to transform the continuous variable “hydration level” in a factor with two conventional categories, losing part of the information in the variable but also achieving the possibility to test the “associations” by various further analyses.

2.6 Additionally, it would be useful for the readers if genes mentioned (discussed) are stated with their full names/functions, where this is not the case.

As rightly suggested, we stated the full names/functions of mentioned genes

In general, the data presentation is not very reader friendly and it would be recommended to add more detail in Figure/Table captions and footnotes, especially in the supplementary document. Also, some data are missing in Figure 5 in the Families plot or some other error occurred. The readers would benefit from using gene designations rather than KO entry numbers to better appreciate the results presented.

We thank the reviewer for the comment. We are sorry but we cannot find errors in the families’ section of the figure 5: the abundances in group A and B are almost biological or technical “zeros”, so the only points displayed are the “outlier” of the corresponding group, according to the box and whisker plot. However, in agreement with reviewer suggestion, we have chosen to not further highlight this significant yet noisy result in the discussion paragraph. Regarding the gene designations, we have added a table with brief descriptions of functions in supplementary materials. However, we though to not translate the KO entry number in gene designations, being those functional orthologous.

In my opinion, the genetic association of facial skin microbiota to host gene variation is the weakest part of the study. This is primarily due to the fact that an extremely low sample size is used. Also, the definition of healthy for the presented group is lacking and the selection of the host genes of interest is not sufficiently substantiated. The significance of selected genes for the selected participant population is not clearly presented so the information is highly speculative and potentially misleading as the authors do not present the other potential confounding factors (data from obtained questionnaires on lifestyle, sun exposure, sun protection and skincare are not presented nor associated with host gene variation).

The selection of the interesting host geneshas been based on DisGenet Skin Wrinkling dataset (CUI: C0037301). Our choice is derived by the consistent link between the composition of this dataset and the primary aim of the paper. We did not further checked other dataset to avoid an involuntary “cherry picking” approach to this analysis. As regard the other confounding factors, we agree with the reviewer, however we cannot add other independent variables to the fitted model due to the low samples’ number of this pilot study. The pioneering nature of our data has now been remarked in the manuscript .

The authors should avoid statements that they studied the mechanisms of skin aging as their study did not have the temporal component. This is particularly sensitive in the conclusion section where this should be re-phrased as in the current wording this is completely misleading. The same is true for the word, pioneer, which should be avoided.

In accordance with the reviewer’s suggestion, we rephrased the conclusion section

Minor revisions needed:

Typo line 40 (MS vs SM), line 44 (hos or host)?

We corrected it in the text

The sentence in line 76 should be rephrased to state that the skin of healthy women was studied, and which exactly functional pathways were studied.

We rephrased the sentence

Line 387, word reported should be replaced with there was or any other more appropriate statement.

We corrected it in the text

Round 2

Reviewer 2 Report

The authors did improve the manuscript and have addressed the major points. The impact of the presented work is not expected to be high due to the low sample size of the study. The added reference should be formatted properly.

Author Response

We thank the reviewer for his/her comment. We agree with him/her regarding the low sample size.

We inserted the missing reference as a comment, if the article is accepted, we will make arrangements with the tipesetter who has the program for formatting the references.

Best regards