Geographic Location Determines Differentially Methylated Gene Expressions in Autoimmune Diseases

Round 1

Reviewer 1 Report

Title: Geographic location determines differentially methylated gene expressions in autoimmune diseases

General Comments
The presented results are a huge work and effort of many scientists from different countries. I am impressed with the research planning, coordination, and results obtained. The work is written in a very clear language and the results obtained are very well summarized. When analyzing such a large group of people, the obtained data may be definitely helpful in determining diseases and the influence of various factors on these diseases. The work is graphically prepared very carefully. 
The work may be published as it is.

Author Response

1-The presented results are a huge work and effort of many scientists from different countries. I am impressed with the research planning, coordination, and results obtained. The work is written in a very clear language and the results obtained are very well summarized. When analyzing such a large group of people, the obtained data may be definitely helpful in determining diseases and the influence of various factors on these diseases. The work is graphically prepared very carefully. 
The work may be published as it is.

We thank the reviewer for his/her enthusiastic comments.

Reviewer 2 Report

In the original article of Pers et al. the authors aimed to compare DNA methylation of patients suffering from SLE, RA, pSS, and SSc to geographical location. Austria, Belgium, France, Germany, Hungary, and Switzerland are defined as north countries, while Italy, Portugal, and Spain are defined as southern countries. They found that the geographical location of patients with autoimmune diseases has an impact on DNA methylation patterns, RNA expression, and immunological profiles. This impact is explained by environmental factors, such as temperature, humidity, ozone levels or pollutants, UV exposition, and food consumption patterns. 

The study is a pre-planned substudy of the European multi-center cross-sectional PRECISESADS IMI consortium project. 

The basic objective of the study is understandable: to examine important population-level differences in the listed autoimmune diseases. 

The methodology, the used statistical analyses, the used in silico tools, as well as a large number of involved patients are all strengths of the study.

The results are clear.

However, several points need revision:

• the abstract of the manuscript is misleading as it lists the diseases studied in the PRECISESADS IMI study, however, this study only covers 4 diseases.
• the northern and southern classification of the studied countries is a bit confusing: Austria, Hungary or Switzerland cannot be said to be northern countries, but rather Central (Eastern Central) European ones.
• Why were patients from genuine northern European countries (e.g., Sweden, Norway, Finland, or Denmark) not involved?
• The geographical separation used by the authors and the environmental factors used to explain the obtained results lack the indication of the values ​​characteristic of the real environmental factors.
• What are the multi-year average temperatures, humidity, ozone levels, air pollution data in each country? This is an important aspect, since, for example, the average temperature and humidity in the Carpathian basin is much higher than in northern Italy.
• Is there a significant correlation between the DNA methylation abnormalities of the perception and the clinical parameters of the examined patient in terms of the clinical course and therapy of the diseases?

According to the criticisms, the manuscript needs major revision. 

Author Response

• The methodology, the used statistical analyses, the used in silico tools, as well as a large number of involved patients are all strengths of the study. The results are clear.

We thank the reviewer for his/her comments.

• The abstract of the manuscript is misleading as it lists the diseases studied in the PRECISESADS IMI study, however, this study only covers 4 diseases.

We agree with the reviewer on the confusion and we now clearly state in the abstract that we have focused this study on SLE, SSc, pSS and RA patients.

• The northern and southern classification of the studied countries is a bit confusing: Austria, Hungary or Switzerland cannot be said to be northern countries, but rather Central (Eastern Central) European ones.

We agree with the reviewer but we have chosen to divide the patients into two arbitrary geographic groups. It is also true that the patients from Brest and Porto are patients from Western Europe. Consequently, the significant results obtained in this manuscript are the consequences of this “arbitrary geographical separation”. We have therefore added this notion of "arbitrary separation" in the text of the manuscript.

• Why were patients from genuine northern European countries (e.g., Sweden, Norway, Finland, or Denmark) not involved?

The present study was conducted in patients with autoimmune diseases and in HCs that were included in the European multi-center cross-sectional PRECISESADS IMI consortium project, which involved patients from Austria, Belgium, France, Germany, Hungary, Italy, Portugal, Spain, and Switzerland.

• The geographical separation used by the authors and the environmental factors used to explain the obtained results lack the indication of the values ​​characteristic of the real environmental factors. What are the multi-year average temperatures, humidity, ozone levels, air pollution data in each country? This is an important aspect, since, for example, the average temperature and humidity in the Carpathian basin is much higher than in northern Italy.

These informations have been added in a supplementary table (Supplementary Table 5).

• Is there a significant correlation between the DNA methylation abnormalities of the perception and the clinical parameters of the examined patient in terms of the clinical course and therapy of the diseases?

We thank the reviewer for this very interesting question. We have now added the physician general assessment and treatments in Table 1 and supplementary Table 1. Interestingly, patients from the north presented an increased physician global assessment score in patients and received significantly more immunosuppressive and anti-TNF drugs than those from the south. However, these treatments cannot explain that autoimmune patients from the north of Europe present a higher hypomethylated profile (see the added PCA in Figure 1).

Reviewer 3 Report

In the section on patients and methods, it is necessary to describe what clinical characteristics were used to define the group of healthy controls, although the study is derived from a previous study, it is important to describe the characteristics of these subjects, as well as the clinical criteria used to classify patients with autoimmune diseases included.

In Table 1 change the term race to population or ethnic origin.

 The differences in age described in Table 1 when comparing the subjects from south vs. from the north could mark differences in the methylation patterns observed. In the case of healthy controls and in SSc patients, the duration of the disease and the frequency of smoking were observed significant differences between southern subjects vs. From North. Were the analyzes showed in the results section adjusted for these confounding variables?

In figure 2 it is recommended to arrange the foot of the figures in a clearer way to guide the reader: Figure a) and c) would have to be a) and b) because represents the data of hypomethylated genes, and figures b) and d) would have to be c) and d) because represents the data of the hypermethylated genes, it is recommended to reorder the description of the figure captions in the order of the data presented.

The discussion section could be improved, describe more about the relationship of hypomethylation with the increase in IFN scores.

Discuss the role of UV light, in particular a strong relationship between UV light exposure and vitamin D levels has been described, and its deficiency in autoimmune diseases, particularly in patients in northern Europe where the UV radiation is lower. Inquire more about related studies and its potential role in the generation and differences of epigenetic patterns that are influenced by UV light exposure.

Discuss the role of nutrients in the generation of epigenetic marks, particularly in previous studies carried out in the population of patients with SLE, where the intake of nutrients is related with changes in the methylation patterns such as CD40L. Methyl groups such as choline, betaine and folic acid are associated to changes in the rates of hypo and hypermethylation of immune response genes involved with autoimmune diseases. The potential role of nutrients in the establishment and maintenance of epigenetic marks between populations is superficially addressed in the discussion and results, despite having data on these variables in the studied populations, even with these data they could be analyzed. Justify why they are not described in detail in results and discussion section.

Describe also what is the role of smoking in changes in methylation patterns and why it was a variable that is described in the tables (table 1), but is not addressed in depth in the results or in the discussion section, when smoking has been described as an important environmental trigger of autoimmunity, and could be related to changes in methylation rates

It is important that the authors describe in the discussion section the limitations of the study, at the methodological level, study design, establishment of causal relationships or associations, as well as the perspectives of the study. Add this section.

Author Response

• In the section on patients and methods, it is necessary to describe what clinical characteristics were used to define the group of healthy controls, although the study is derived from a previous study, it is important to describe the characteristics of these subjects, as well as the clinical criteria used to classify patients with autoimmune diseases included.

As described in the paper these criteria are detailed in ClinicalTrials.gov with numbers NCT02890121 (patients) and NCT02890147 (HCs).

Inclusion Criteria for HCs were:

• Aged 18 years or older at the time of consent
• Signed the informed consent form

Exclusion Criteria for HCs were:

• Individuals on chronic medication.
• Individuals suffering from any inflammatory, autoimmune, allergic or infectious condition and without a history of autoimmune disease, particularly thyroid disease or other diseases that may modify cellular profiles in blood.
• Pregnant women.

The diagnostic criteria used in the inclusion of patients were based: for RA on the 2010 ACR/EULAR classification criteria [1], for SLE on the 1997 update of 1982 ACR criteria [2], for SSc on ACR/EULAR 2013 classification criteria [3], for pSS on AECG pSS classification criteria [4] with at least the presence of anti-SSA and/or a positive focus on a minor salivary gland biopsy.

[1] Aletaha D, Neogi T, Silman AJ, et al. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2010;62:2569–81. doi:10.1002/art.27584

[2]     Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997;40:1725. doi:10.1002/art.1780400928

[3]     van den Hoogen F, Khanna D, Fransen J, et al. 2013 classification criteria for systemic sclerosis: an American college of rheumatology/European league against rheumatism collaborative initiative. Ann Rheum Dis 2013;72:1747–55. doi:10.1136/annrheumdis-2013-204424

[4]     Vitali C, Bombardieri S, Jonsson R, et al. Classification criteria for Sjögren’s syndrome: a revised version of the European criteria proposed by the American-European Consensus Group. Ann Rheum Dis 2002;61:554–8. doi:10.1136/ard.61.6.554

Exclusion criteria were for patients were:

• Patients unable to understand the procedures related to the protocol should not be included. The study is voluntary and patients must be able to give their informed consent.
• Pregnant women
• Neonatal lupus
• Drug-induced lupus
• Patients whose condition is so serious that they cannot take part in the study
• Severe nephrotic syndrome with proteinuria >=3,5 g/day
• Patients with stable doses of steroids >15mg/day for the last 3 months or with IV corticosteroids in the last 3 months
• Patients under immunosuppressants for the last 3 months prior to recruitment with:
  • Methotrexate ≥25mg/week
  • Azathioprine ≥2.5mg/kg/day
  • Cyclosporine A > 3mg/kg/day
  • Mycophenolate Mofetil > 2gr/day
• Treatment with cyclophosphamide (any dose or route of administration) or Belimumab in the past 6 months
• Patients with combined therapy of two or more immunosuppressants
• Patients on depletative therapy such as Rituximab in the last year
• Patients receiving experimental therapy.
• Chronic HBV or HCV infection
• Overlap syndromes

• In Table 1 change the term race to population or ethnic origin.

The term has been changed.

• The differences in age described in Table 1 when comparing the subjects from south vs. from the north could mark differences in the methylation patterns observed. In the case of healthy controls and in SSc patients, the duration of the disease and the frequency of smoking were observed significant differences between southern subjects vs. From North. Were the analyzes showed in the results section adjusted for these confounding variables?

 A new Table 1 with the characteristics of all patients has been added. No difference between southern and northern subjects was observed regarding frequency of smoking regardless of the disease. Furthermore, while we agree that age was different in HCs (47 for HCs from the south and 43 for HCs from the north, only six DMPs, each with low ΔBeta (< 0.065), were found in HCs from the north group compared to HCs from the south group. Consequently, HCs were merged to be compared with autoimmune patients from both groups. No difference between southern and northern subjects was observed regarding age or gender regardless of the disease. However, it was interesting to observe that PGA was significantly different in patients from the north vs south with an increase in RA and a decrease in SLE, pSS and SSc patients from the south compared to the north.

• In figure 2 it is recommended to arrange the foot of the figures in a clearer way to guide the reader: Figure a) and c) would have to be a) and b) because represents the data of hypomethylated genes, and figures b) and d) would have to be c) and d) because represents the data of the hypermethylated genes, it is recommended to reorder the description of the figure captions in the order of the data presented.

The figure 2 and its legend have been modified accordingly.

• The discussion section could be improved, describe more about the relationship of hypomethylation with the increase in IFN scores.

We thank the reviewer for his/her comment. We have now improved the section concerning the IFN signature in the discussion.

• Discuss the role of UV light, in particular a strong relationship between UV light exposure and vitamin D levels has been described, and its deficiency in autoimmune diseases, particularly in patients in northern Europe where the UV radiation is lower. Inquire more about related studies and its potential role in the generation and differences of epigenetic patterns that are influenced by UV light exposure.

  We thank the reviewer for his/her comment. We have now added a section concerning the impact of vitamin D deficiency in autoimmune diseases in the discussion and a new table (Supplementary Table 5) reports now all the environmental characteristics of our cohort (UV, humidity, ozone, rainfall, pollutants, …).

• Discuss the role of nutrients in the generation of epigenetic marks, particularly in previous studies carried out in the population of patients with SLE, where the intake of nutrients is related with changes in the methylation patterns such as CD40L. Methyl groups such as choline, betaine and folic acid are associated to changes in the rates of hypo and hypermethylation of immune response genes involved with autoimmune diseases. The potential role of nutrients in the establishment and maintenance of epigenetic marks between populations is superficially addressed in the discussion and results, despite having data on these variables in the studied populations, even with these data they could be analyzed. Justify why they are not described in detail in results and discussion section.

We thank the reviewer for his/her comment. We have now added in the discussion a section concerning the impact of nutrients in the generation of epigenetic marks in autoimmune diseases.

• Describe also what is the role of smoking in changes in methylation patterns and why it was a variable that is described in the tables (table 1), but is not addressed in depth in the results or in the discussion section, when smoking has been described as an important environmental trigger of autoimmunity, and could be related to changes in methylation rates.

No difference between southern and northern subjects was observed regarding frequency of smoking regardless of the disease. Consequently, even if we agree with the reviewer that smoking has been described as an important environmental trigger of autoimmunity, this parameter cannot explain why, as compared to patients from the south group, 1465 DMPs were found for the patients from the north group, with 1018 hypomethylated sites corresponding to 624 genes and 447 hypermethylated sites corresponding to 376 genes.

• It is important that the authors describe in the discussion section the limitations of the study, at the methodological level, study design, establishment of causal relationships or associations, as well as the perspectives of the study. Add this section.

 We thank the reviewer for his/her comment. We have now added a section concerning the limitations and the perspectives of our study in the discussion.

Round 2

Reviewer 2 Report

The revised version of the manuscript is now acceptable for pubilcation. 

Reviewer 3 Report

The authors made the adjustments in the document and in the supplementary material referring to the observations made in the first round of the review. It is an interesting work with perspectives to be continued in subsequent studies to evaluate gene-environment interactions and their impact on changes in methylation patterns in autoimmune diseases.