Decreased Serum Levels of the Insulin Resistance-Related microRNA miR-320a in Patients with Polycystic Ovary Syndrome

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Authors in the article reported a significant association between lowered serum levels of the IR-associated microRNA miR-320a and the PCOS phenotype. They showed initial links between the transcriptional regulation of this microRNA and the dysregulated androgen concentrations observed in PCOS patients. These results provided a clue regarding the sub-characterization of different PCOS metabolic patient types, thereby helping to understand the metabolic changes triggered by PCOS on a molecular level and individually assign personalized treatments to PCOS patients in the clinical reproductive medicine routine.

I have read the article with interest and found that the article's methodology was rigorous and concrete, the findings were significant and sound and had good practical implications, and the discussion was also evidence-based. I was unable to find any shortcomings, so it is recommended that the article be accepted for publication.

Author Response

Comments 1: Authors in the article reported a significant association between lowered serum levels of the IR-associated microRNA miR-320a and the PCOS phenotype. They showed initial links between the transcriptional regulation of this microRNA and the dysregulated androgen concentrations observed in PCOS patients. These results provided a clue regarding the sub-characterization of different PCOS metabolic patient types, thereby helping to understand the metabolic changes triggered by PCOS on a molecular level and individually assign personalized treatments to PCOS patients in the clinical reproductive medicine routine.

I have read the article with interest and found that the article's methodology was rigorous and concrete, the findings were significant and sound and had good practical implications, and the discussion was also evidence-based. I was unable to find any shortcomings, so it is recommended that the article be accepted for publication.

Response 1:

Dear Reviewer,

Thank you for taking the time to review our manuscript “Decreased serum levels of the insulin resistance-related microRNA miR-320a in patients with polycystic ovary syndrome”. We appreciate your encouraging remarks and thank you for the positive feedback. Best regards

Thomas Greither (in behalf of the authors)

Reviewer 2 Report

Comments and Suggestions for Authors

Polycystic ovary syndrome (PCOS) is often associated with metabolic abnormalities such as obesity and insulin resistance (IR). The latter affects the serum levels of various microRNAs, but the relationship between IR-related microRNAs and PCOS is rather elusive. 

 

In this interesting study, the authors measured the serum levels of several IR-associated microRNAs in a cohort of infertility patients, including approximately 40% with PCOS. They found that serum levels of two of these microRNAs were inversely associated with dipeptidyl peptidase 4 serum concentrations, and that serum levels of another microRNA were significantly decreased in PCOS patients. In patients who achieved pregnancy after Assisted Reproductive Technology (ART) cycles, serum levels of the IR-associated microRNAs were significantly elevated. In vitro experiments demonstrated a significant upregulation of one of these microRNAs in response to testosterone stimulation. 

 

The authors conclude that there is a close relationship between metabolic disorders and dysregulation of microRNA expression in PCOS.

 

This Reviewer has a few minor concerns:

[1]. In the statistical analyses the Authors use Student's t-test in data which are not normally distributed - this is acceptable with larger samples, yet the Authors do use a non-parametric test (Mann-Whitney U-test) at times....

[2]. More to this, in the results section there are many comparisons but it is not evident if any statistical correction was applied.

Author Response

Polycystic ovary syndrome (PCOS) is often associated with metabolic abnormalities such as obesity and insulin resistance (IR). The latter affects the serum levels of various microRNAs, but the relationship between IR-related microRNAs and PCOS is rather elusive. 

In this interesting study, the authors measured the serum levels of several IR-associated microRNAs in a cohort of infertility patients, including approximately 40% with PCOS. They found that serum levels of two of these microRNAs were inversely associated with dipeptidyl peptidase 4 serum concentrations, and that serum levels of another microRNA were significantly decreased in PCOS patients. In patients who achieved pregnancy after Assisted Reproductive Technology (ART) cycles, serum levels of the IR-associated microRNAs were significantly elevated. In vitro experiments demonstrated a significant upregulation of one of these microRNAs in response to testosterone stimulation. 

The authors conclude that there is a close relationship between metabolic disorders and dysregulation of microRNA expression in PCOS.

 

Dear Reviewer, 

We want to thank you for your time and insightful remarks. Revisions according to the questions raised are detailed below.

Best regards.

Thomas Greither (in behalf of the authors)

This Reviewer has a few minor concerns:

Comments 1: In the statistical analyses the Authors use Student's t-test in data which are not normally distributed - this is acceptable with larger samples, yet the Authors do use a non-parametric test (Mann-Whitney U-test) at times....

Response 1: Thank you for pointing out this important statistical issue. As the ex vivo serum level values of the individual microRNAs were not normally distributed, we used the Mann-Whitney U test to analyze them. The microRNA expression data in the in vitro analyses performed in the cell culture were normally distributed, therefore, we used Student t-test. We tried to clarify this approach in the Materials and Methods section (l. 158 - 160).

Comments 2: More to this, in the results section there are many comparisons but it is not evident if any statistical correction was applied.

Response 2: We agree with the reviewer, that this is an issue, which should be taken care of, and apologize for not pointing it out in our manuscript beforehand. As we consider our data analysis as exploratory, we did not apply corrections against multiple testing (e.g. Bonferroni correction or other). We address this approach in the limitations to give the reader the opportunity to comprehend our approach. (l. 395  397)

Reviewer 3 Report

Comments and Suggestions for Authors

 

The authors present the results of an original study on microRNAs  in a group of patients with PCOS and other feritily disorderes. They also conducted experimental studies on the expression of microRNAs following exposure to glucose and testosterone in cell cultures. Thus, they address a issue of epigenteic regulation in the pathogenesis of metaboilic diseases, which is an important area of research

The paper is written very carefully and interestingly, it is a pleasure to read. It has a high chance of being widely cited.

There are no major objections regarding the selection of microRNA molecules, other methodological issues, or statistical methods."

 

However, please request that the authors address the following comments

 1.    In a group of 136 patients with PCOS, with a significantly varied BMI range (17-47), some likely used metformin, which affects the level of microRNA, including 320 and 148a. Please consider this fact in the analysis of results or at least in the discussion.

 2.    What is  the beckround for studying the activity of DPPt in the context of microRNA concentration? This topic should be clarified in the introduction. Do incretins play any role in microRNA expression?

3.     What is the expression of microRNAs in subgroups with varying degrees of obesity (based on BMI or more precise anthropometric measurements) – in reference to lines 276-288 of the discussion.

While the authors mention this in the limitations paragraph, it would be better to provide the results of the mentioned analyses.

4.    What is the expression of microRNAs in relation to insulin sensitivity indices, even if HOMA?

Performing such analyses should be feasible in such a large group of patients.

 

Line 356 – correct typo

 

 

Author Response

The authors present the results of an original study on microRNAs  in a group of patients with PCOS and other feritily disorderes. They also conducted experimental studies on the expression of microRNAs following exposure to glucose and testosterone in cell cultures. Thus, they address a issue of epigenteic regulation in the pathogenesis of metaboilic diseases, which is an important area of research

The paper is written very carefully and interestingly, it is a pleasure to read. It has a high chance of being widely cited.

There are no major objections regarding the selection of microRNA molecules, other methodological issues, or statistical methods."

 

Dear Reviewer,

Thank you for your constructive feedback regarding our manuscript. We appreciate your interest in our study and your valuable suggestions. Following, please find a point-to-point answer on your remarks.

Best regards.

Thomas Greither (in behalf of the authors)

 

However, please request that the authors address the following comments

Comments 1:  In a group of 136 patients with PCOS, with a significantly varied BMI range (17-47), some likely used metformin, which affects the level of microRNA, including 320 and 148a. Please consider this fact in the analysis of results or at least in the discussion.

Response 1: We thank the reviewer for the valuable suggestion. We stated the following information in the results section “Information on the usage of metformin was only available for a subset of patients. While 83 of the Non-PCOS patients did not use metformin, 20 of this group used metformin. Vice versa, 68 of the PCOS patients used metformin, while 17 did not. In non-parametric tests, in the whole subset metformin usage was not significantly associated with altered serum microRNA serum levels (miR-93: p = 0.998; miR-320a: p = 0.181; miR-148a: p = 0.611; miR-216a: p = 0.985; miR-224: p = 0.213; Mann-Whitney U test). Additionally, either than analyzing the PCOS or the Non-PCOS group, there was no association between metformin usage and the mioRNA serum levels.“ (l. 224 - 231).

 

Comments 2: What is  the background for studying the activity of DPP4 in the context of microRNA concentration? This topic should be clarified in the introduction. Do incretins play any role in microRNA expression?

Response 2: DPP4 serum activity was added to the analyses, as we have previously detected an association between PCOS and an increased DPP4 serum activity (Blauschmidt et al., 2017). As the patient cohort of the previous study overlapped with the study cohort analyzed in the present manuscript, we decided to co-analyze DPP4 serum activity as important regulatory factor of glucose homeostasis and treatment target of type 2 diabetes mellitus. We mentioned this results, as well as contradictory findings, in the introduction (l.46 - 50).

Regarding the putative role on incretins on microRNA expression - where is less literature on this topic, as far as we see. It is an interesting hypothesis and worth testing it, but unfortunately is beyond the scope of our present manuscript.

 

Comments 3: What is the expression of microRNAs in subgroups with varying degrees of obesity (based on BMI or more precise anthropometric measurements) – in reference to lines 276-288 of the discussion.

While the authors mention this in the limitations paragraph, it would be better to provide the results of the mentioned analyses.

Response 3: We agree with the reviewer, that there is a strong association between IR, an enhanced BMI, the PCOS phenotype and several microRNA expressions. To tackle this issue, we performed some subanalyses regarding the association between the selected serum microRNAs and the BMI in the whole patient cohort. We did not found significant associations between a BMI>25 or BMI>30 to the respective control groups in Mann-Whitney-U tests (miR-93 - p = 0.36; miR-320a - p = 0.59; miR-148a - p = 0.55; miR-216a - p = 0.60; miR-224 - p = 0.44 / miR-93 - p = 0.56; miR-320a - p = 0.28; miR-148a - p = 0.22; miR-216a - p = 0.46; miR-224 - p = 0.86). Even when separately analyzing PCOS and Non-PCOS patients, we did not observe a significant association between the individual microRNA serum expressions and BMI>25 or BMI>30; with only a trend towards a significant association between the miR-148a serum expression and a BMI>30 (p = 0.052; Mann-Whitney U-Test). We added these results in condensed form in subchapter 3.3 (l. 221 -223).

When initially analyzing our data, we developed the idea to tailor BMI-matched PCOS and Non-PCOS subgroups. However, as the BMI was differently distributed between our PCOS and Non-PCOS cohort (see table 1), the inclusion of only matched BMIs would have resulted in smaller and - more problematic -  selected groups with a bias. This would have weakened the statistical power of our analysis. Therefore, we inserted the following sentence in the discussion section: “One limitation of this study is the heterogeneity of the study cohort. Although a relatively large number of participants could be included, stratification according to factors like the BMI was not performed. The statistical analyses showed no association between the serum microRNA levels and overweight or obesity; however, it can not be ruled out that obesity-associated confounders additionally influence the serum microRNA levels.” (l. 389 - 393).

 

Comments 4: What is the expression of microRNAs in relation to insulin sensitivity indices, even if HOMA? Performing such analyses should be feasible in such a large group of patients.

Response 4:  We agree with the reviewer, that analyses on insulin sensitivity indices would be an interesting add-on. We have data on the HOMA-IR of 162 patients in our cohort, mostly from the PCOS subgroup, as it is not clinical standard in our outclinics to perform HOMA-IR without the suspicion of an insulin resistance. In bivariate regression analyses, we found no significant association between different serum microRNAs and the HOMA-IR. Unfortunately, given the unbalanced distribution between the HOMA-IR, we consider it difficult to present these data without a proper assessment and would therefore not detail these analyses in the manuscript.  

 

Comments 5: Line 356 – correct typo

Response 5: Thank you for indicating this error. We have corrected the respective typo. (l.378)