Epigenetic Modifications in Generalized Autoimmune Epithelitis: Sjögren’s Syndrome and Primary Biliary Cholangitis
, Wesley H. Brooks 4, George N. Dalekos 1,2 and Yves Renaudineau 3,5,*Round 1
Reviewer 1 Report
The review article entitled “Epigenetic Modifications in Generalized Autoimmune Epithelitis: Sjӧgren’s Syndrome and Primary Biliary Cholangitis” is a well written manuscript. This is the first review covering this topic on epigenetics in autoimmune epithelitis. I would recommend publishing this review in “Epigenomes” but not in current format. Over all this review looks good and is very informative, but some of the lines seems repetitive and mostly filled with lengthy sentences. This manuscript requires sincere editing to avoid lengthy sentences for easy understanding.
I have following suggestions
1) Please add a line about what exactly SjS is or PBC is and symptoms, affected organs in the 5thline of introduction, just before mentioning affected age groups. (For example: SjS is mainly associated with dryness in eyes (xerophthalmia)and dryness in mouth (xerostomia)).
2) Association of hypomethylation with higher gene expression and hypermethylation with lower/decreased gene expression are obvious, seems repetitive when mentioning for every gene separately. (lines 145-161).
3) Lines 84-87 and 130-134 are repetitive.
4) Please avoid lengthy sentences. Examples of lengthy sentences
Line 69-72
Line120-123
Line 182-185
Line 306-311
Minor:
5) Remove “to” in the first line of abstract
6) Remove tests after liver function in the line 41
7) Please check the spelling of xerophthalmia in line 55
8) Remove epigenetic in line 89
9) In line 157, Please correct CD14+ T lymphocytes??
10)Line 186 is not clear. Cell type studies??
11)In line 193, please expand DNMC’s.
12)Line 247, correct “details” (remove ‘s)
13)Line 248 is not clear. Please correct
Author Response
Revisions based on the first reviewer’s comments
1- The review article entitled “Epigenetic Modifications in Generalized Autoimmune Epithelitis: Sjӧgren’s Syndrome and Primary Biliary Cholangitis” is a well written manuscript. This is the first review covering this topic on epigenetics in autoimmune epithelitis. I would recommend publishing this review in “Epigenomes” but not in current format. Over all this review looks good and is very informative, but some of the lines seems repetitive and mostly filled with lengthy sentences. This manuscript requires sincere editing to avoid lengthy sentences for easy understanding.
Answer 1: Thanks for your kind remarks. The manuscript has been revised according to your suggestions. Long sentences have been rewritten in a more concise and comprehensible way.
2- Please add a line about what exactly SjS is or PBC is and symptoms, affected organs in the 5thline of introduction, just before mentioning affected age groups. (For example: SjS is mainly associated with dryness in eyes (xerophthalmia)and dryness in mouth (xerostomia)).
Answer 2: A few lines concerning SjS and PBC pathogenesis and symptoms were added in the introduction: “SjS is characterized by lymphocytic infiltration of lachrymal and salivary glands leading to progressive loss of secretory functions and the development of xerophthalmia and xerostomia [1]. PBC is an autoimmune cholestatic liver disease characterized by progressive inflammatory destruction of the small intrahepatic bile ducts with portal inflammation and progressive fibrosis, leading to cirrhosis and subsequent liver failure [2]. »
3- Association of hypomethylation with higher gene expression and hypermethylation with lower/decreased gene expression are obvious, seems repetitive when mentioning for every gene separately. (lines 145-161).
Answer 3 : In order to avoid unnecessary repetitions, the association of demethylation and increased gene expression was omitted for some of the described genes.
4- Lines 84-87 and 130-134 are repetitive. Please avoid lengthy sentences. Examples of lengthy sentences :line 69-72, line120-123, line 182-185, line 306-311
Answer 4: Based on your suggestion we have made some changes in the lines mentioned above so as to avoid repetition. Accordingly, in the introduction we briefly mention all epigenetic mechanisms, and in the main body of the article we have focused on the mechanisms with special emphasis on DNA methylation/demethylation.
5- Minor points: Remove “to” in the first line of abstract; Remove tests after liver function in the line 41 ; Please check the spelling of xerophthalmia in line 55 ; Remove epigenetic in line 89 ; In line 157, Please correct CD14+ T lymphocytes??; Line 186 is not clear ; Cell type studies?? In line 193, please expand DNMC’s ; Line 247, correct “details” (remove ‘s) ; Line 248 is not clear. Please correct
Answer 5: To comply with your remarks “to” was removed in the first line of the abstract; “tests” was removed (line 41); the spelling was corrected; the word “epigenetic” was removed (line 89); CD14(+) T lymphocytes was corrected; the phrase cell type studies was omitted; the abbreviation DMCs was expanded in line 189 where it is mentioned for the first time; “s” was removed (line 247); and the structure of line 248 was changed so as to make it more clear for the reader.
Reviewer 2 Report
The review article by Arvaniti et al. provides a detailed overview on the current knowledge on epigenetic mechanisms in primary Sjögren’s syndrome (pSS) and primary biliary cholangitis (PBC) under the concept of autoimmune epithitis diseases.
General comments:
1. Overall the structure and the language of the review are appropriate. However, the manuscript may benefit from a critical linguistic revision to improve the reading flow and understandability, e.g. by reduction of the amount of very long sentences and correction minor typing and punctuation errors etc.
2. Also, the review would benefit from shortening some sections and rewriting them in a more concise and clear manner. In some parts many details of the different studies within in the field string together in a not always very coherent way; more accuracy and rigor instead of many words would be preferred.
3. In those parts of the manuscript where not mechanisms or results which are specific for pSS and PBC are discussed, but rather general information is given e.g. on mechanisms of DNA methylation per se, it is not very appropriate to mainly cite other review articles within the field of pSS epigenetics. Instead, original articles or more general reviews on the mechanisms should be referred to.
4. The impact of the analysis of different cell and tissue types could be discussed better. This question is touched upon briefly in the last part of the review. What are the advantages and short-comings of the different approaches? And which strategies are currently in use or are still lacking to handle the difficulties? A few words about context-specificity of epigenetic modifications would be valuable. On a similar note, the impact of different treatments and other factors on epigenetic patterns should be (briefly) discussed.
5. Overall a bit more critical distance and own opinion on the limitations and advantages of the studies so far performed would be highly appreciated rather than just enumerating the different studies.
Major comments:
1. Abstract and p.2: The authors state that the genetic predisposition of the diseases has been well established. However, compared to other autoimmune diseases such as SLE or RA, one may argue that the genetic basis of pSS and PBC is rather understudied. I would also suggest, to include some (brief) information on non-HLA related genetic findings in pSS in the introduction part in order to give a more complete picture about what is known.
2. p.2: In their introduction the authors state that “[epigenetic] changes are independent of the underlying DNA sequence”. They are certainly not, as the existence of highly abundant methylation quantitative trait loci (meQTLs) shows. Probably something else is meant here, and this part has to be revised to prevent misinterpretation.
3. p.2: I would not agree with the statement that DNA methylation is the major epigenetic process in eukaryotic cells; it is the one most extensively studied.
4. The paragraph regarding the question of reduction of global methylation in pSS needs to be revised in order to increase understandability and clarity for the reader. Here, it needs to be pointed out that a reduction in global methylation was only observed in cultivated salivary gland epithelial cells, but not in B or T cells and also not in whole blood and not in salivary gland biopsies. Especially regarding salivary gland biopsies versus cultured SGECs one has to be exact about the finding of the different original studies. Both, Cole et al. and Imgenberg-Kreuz et al. have analyzed salivary gland biopsies from patients with pSS compared to control samples. In these studies global hypomethylation in LSGs was not observed. “Thabet et al. (8) previously reported global hypomethylation in cultured LSG epithelial cells from SS patients. We considered whether these differences could be detected in more heterogeneous LSG tissue samples. However, no significant differences in mean genome DNA methylation were observed across all CpGs (1.01‐fold hypermethylation in SS cases; P = 0.26).” (Cole et al, 2016).
4. p.3-6: Sections 2.2. DNA methylation/hydroxymethylation in pSS and PBC and 2.3. Genome wide analysis of methylation in pSS and PBC should be restructured to enhance clarity for the reader, as the current structure/chronology and content of these parts are quite confusing. I suggest to rewrite these parts in a clear and concise language and to focus on the main results and mechanisms.
5. p.4: The manuscript only covers X chromosomal DNA methylation findings in PBC. Results from studies in pSS (e.g. Imgenberg-Kreuz et al, 2016 ARD) should be added.
6. p.5: EWAS does not allow quantification of DNA methylation across the whole genome. Techniques such as the 450k (or EPIC) array or whole-genome bisulfite sequencing allow for site specific quantification of DNA methylation across the genome. EWAS is an approach to statistically analyze these data. Again, Cole et al. did not report global hypomethylation in pSS LSG tissue, and the corresponding sections need to be corrected.
7. p.5: Gene expression in B lymphocytes was only measured by Imgenberg-Kreuz et al, not by Miceli-Richard et al. The former showed that hypomethylation and overexpression of IFN induced genes were driven by antibody positive patients. The correlation with disease activity described by Miceli-Richard et al., relates to hypomethylation of IFN induced genes, but a correlation between gene overexpression and disease activity has not been formally shown.
8. p.5: MeQTL effects in pSS are not at all limited to the HLA region, see e.g. Imgenberg-Kreuz et al and Miceli-Richard et al.
9. p.6: The authors hypothesize that the female predominance seen in PBC is reflected by the majority of DMCs (in a very small study, though) being located on the X chromosome. This is an appealing idea. However, pSS has a similar female predominance and a whole bunch of DNA methylation studies has been performed and identified only few differentially methylated sites on the X chromosome. This discrepancy needs to be discussed.
10. p.6: It is not correct that DNA methylation in general has a repressive effect of gene expression. DNA methylation in CpG island within gene promoters may generally has a repressive effect. The importance of the location of differential methylation in relation to gene regions (e.g. promoter versus gene body) needs to be addressed. On a similar note are histone modifications not less predictive (actually, the correlation of DNA methylation status with gene expression levels is quite limited to specific genes and tissue types/circumstances) regarding their effect on gene expression, but they are more divers given the plethora of different histone modifications.
11. p.7: Table 2 is incomplete and does e.g. not include studies by Zilahi et al, Chen et al, Wang et al. and Wang-Renault et al. It would be very valuable if information about the main findings and/or targets of the differentially expressed miRNAs cloud be included in Table 2 and Table 3.
12. p.9: The majority of studies cited (ref 116-121) in the context of future directions for development of epigenetic treatments are relatively old. To add up-to-date developments (and corresponding references) would be appreciated.
13. The potential role of long interspersed nuclearelement-1 (LINE-1) in pSS pathogenesis, especially with regard to the IFN signature, should be included in the review. Mavragani et al. (Arthritis Rheum 2016 and J Autoimmun 2018) have investigated promoter methylation and gene expression of LINE-1 minor salivary gland biopsies from pSS patients and controls.
Minor:
1. p.6: Reference 54 does not refer to Selmi et al. Is ref 73 meant?
2. p.6. Is reference 77 correct? Maybe Konsta et al. is meant here instead?
3. Gene symbols need to be given in italics according to the common nomenclature.
4. p.6: “Simultaneous genomic-epigenomic analysis” may be a misleading wording; instead e.g. “integration of genomic and epigenomic data” is more accurate.
Author Response
Revisions based on the second reviewer’s comments
6- Overall the structure and the language of the review are appropriate. However, the manuscript may benefit from a critical linguistic revision to improve the reading flow and understandability, e.g. by reduction of the amount of very long sentences and correction minor typing and punctuation errors etc.
Answer 6 :We appreciate this guidance for improving the manuscript. In order to achieve better comprehensibility in our manuscript, an extensive revision has now been made. Moreover, in an attempt to increase comprehensibility, we have rewritten many of the long sentences and corrected punctuation errors (see answer to reviewer 1).
7- The review would benefit from shortening some sections and rewriting them in a more concise and clear manner. In some parts many details of the different studies within in the field string together in a not always very coherent way; more accuracy and rigor instead of many words would be preferred.
Answer 7: As our review focuses on the epigenetic studies of SjS and PBC it is necessary to mention in detail the results of many of the studies. However based on your valuable comments we have reorganized the manuscript and improved there adability by avoiding long sentences.
8- In those parts of the manuscript where not mechanisms or results which are specific for pSS and PBC are discussed, but rather general information is given e.g. on mechanisms of DNA methylation per se, it is not very appropriate to mainly cite other review articles within the field of pSS epigenetics. Instead, original articles or more general reviews on the mechanisms should be referred to.
Answer 8 :Parts of the manuscript contain general information on DNA methylation. For these parts original papers, including some of our own previous research publications, were selected first and reviews were cited when they provide additional information based on literature analysis. In other parts of this manuscript, where general information on epigenetic mechanisms is provided (e.g. miRNAs), original papers were selected.
9- The impact of the analysis of different cell and tissue types could be discussed better. This question is touched upon briefly in the last part of the review. What are the advantages and short-comings of the different approaches? And which strategies are currently in use or are still lacking to handle the difficulties? A few words about context-specificity of epigenetic modifications would be valuable. On a similar note, the impact of different treatments and other factors on epigenetic patterns should be (briefly) discussed.
Answer 9 : Based on your interesting comment concerning the analysis of different cell types, we now explain in more detail the importance of tissue selection in epigenetic studies. We mention that the selection of long-term cultured cells could reduce the heterogeneity of the results when different cell types are analyzed with the main limitation being small sample size of some cell types. Differences in survival of different cell types in culture could lead to the reduced heterogeneity. In addition, the different epigenetic therapeutic targets are discussed in more detail.
10- Overall a bit more critical distance and own opinion on the limitations and advantages of the studies so far performed would be highly appreciated rather than just enumerating the different studies.
Answer 10: In order to keep a critical point of view, a few lines regarding our opinion on the different studies mentioned were added in the text.
11- Abstract and p.2: The authors state that the genetic predisposition of the diseases has been well established. However, compared to other autoimmune diseases such as SLE or RA, one may argue that the genetic basis of pSS and PBC is rather understudied. I would also suggest, to include some (brief) information on non-HLA related genetic findings in pSS in the introduction part in order to give a more complete picture about what is known.
Answer 11: Indeed the genetic bases for other autoimmune diseases like SLE and RA is much better established compared to SjS and PBC. In order to avoid misleading conclusions the phrase was changed to “well described”. In addition, according to your suggestion, some brief information about non-HLA genetic risk loci in SjS was added in the text.
12- In their introduction the authors state that “[epigenetic] changes are independent of the underlying DNA sequence”. They are certainly not, as the existence of highly abundant methylation quantitative trait loci (meQTLs) shows. Probably something else is meant here, and this part has to be revised to prevent misinterpretation.
Answer 12: According to comment 2 the phrase “epigenetic changes are independent of the underlying DNA sequence” was removed from the text.
13- I would not agree with the statement that DNA methylation is the major epigenetic process in eukaryotic cells; it is the one most extensively studied.
Answer 13: According to your comment, we changed our statement to “DNA methylation represents the most extensively studied epigenetic process”..
14- The paragraph regarding the question of reduction of global methylation in pSS needs to be revised in order to increase understandability and clarity for the reader. Here, it needs to be pointed out that a reduction in global methylation was only observed in cultivated salivary gland epithelial cells, but not in B or T cells and also not in whole blood and not in salivary gland biopsies. Especially regarding salivary gland biopsies versus cultured SGECs one has to be exact about the finding of the different original studies. Both, Cole et al. and Imgenberg-Kreuz et al. have analyzed salivary gland biopsies from patients with pSS compared to control samples. In these studies global hypomethylation in LSGs was not observed. “Thabet et al. (8) previously reported global hypomethylation in cultured LSG epithelial cells from SS patients. We considered whether these differences could be detected in more heterogeneous LSG tissue samples. However, no significant differences in mean genome DNA methylation were observed across all CpGs (1.01‐fold hypermethylation in SS cases; P = 0.26).” (Cole et al, 2016).
Answer 14: We disagree with your interpretation regarding global DNA methylation analysis and consider that this difference is related to the technology used: anti-5mC antibodies versus the 450K array. The use of a specific anti-5mC antibody by immunohistochemistry or by ELISA supports DNA demethylation in epithelial cells and this was not only observed by our group (Thabet, 2013; Konsta, 2016) but confirmed by MJ Gonzales group (Lagos, 2018). Using a whole methylome approach (450K array), such differences are not observed, so we agree with you on this point. However Illumina’s technology used in EWAS studies is poor at testing global DNA methylation, which is not the case of the anti-5mC antibodies for example (your comment 17). In addition, Illumina’s technology tests DNA hydroxymethylation, in addition to DNA methylation, and we know that the hydroxymethylation process is increased in SjS (Lagos et al and our own data). The text was changed accordingly and it is mentioned that global hypomethylation does not affect B and T lymphocytes.
15- Sections 2.2. DNA methylation/hydroxymethylation in pSS and PBC and 2.3. Genome wide analysis of methylation in pSS and PBC should be restructured to enhance clarity for the reader, as the current structure/chronology and content of these parts are quite confusing. I suggest to rewrite these parts in a clear and concise language and to focus on the main results and mechanisms.
Answer 15: In order to make it more comprehensible for the reader we have revised the parts concerning methylation/hydroxymethylation in SjS and PBC primarily in the section about GWAS. The different studies were put in chronological order and the types of cells found to be hypomethylated are now clearly mentioned in each part.
16- The manuscript only covers X chromosomal DNA methylation findings in PBC. Results from studies in pSS (e.g. Imgenberg-Kreuz et al, 2016 ARD) should be added.
Answer 16: According to your kind remark, a section concerning the X chromosomal DNA methylation in SjS has been added.
17- EWAS does not allow quantification of DNA methylation across the whole genome. Techniques such as the 450k (or EPIC) array or whole-genome bisulfite sequencing allow for site specific quantification of DNA methylation across the genome. EWAS is an approach to statistically analyze these data. Again, Coleet al.did not report global hypomethylation in pSS LSG tissue, and the corresponding sections need to be corrected.
Answer 17: We totally agree with you and, in order to avoid misleading conclusions, the statement that EWAS allows quantification of methylation across the whole genome has been removed.
18: Gene expression in B lymphocytes was only measured by Imgenberg-Kreuzet al, not by Miceli-Richardet al. The former showed that hypomethylation and overexpression of IFN induced genes were driven by antibody positive patients. The correlation with disease activity described by Miceli-Richardet al., relates to hypomethylation of IFN induced genes, but a correlation between gene overexpression and disease activity has not been formally shown.
Answer 18: We apologize for that, and in compliance with your remarks, the cited articles were changed in the text so as to make clear that one study has measured gene expression in B lymphocytes and that it is the hypomethylation of IFN regulated genes that is associated with disease activity.
19: MeQTL effects in pSS are not at all limited to the HLA region, see e.g. Imgenberg-Kreuz et al and Miceli-Richard et al.8)
Answer 19: The statement that genetic risk variants are associated only with HLA regions was changed and the link between SNPs and meQTL was introduced.
20- The authors hypothesize that the female predominance seen in PBC is reflected by the majority of DMCs (in a very small study, though) being located on the X chromosome. This is an appealing idea. However, pSS has a similar female predominance and a whole bunch of DNA methylation studies has been performed and identified only few differentially methylated sites on the X chromosome. This discrepancy needs to be discussed.
Answer 20: Based on you interesting comment a few lines were added so as to discuss the discrepancy between DMCs of the X chromosome in SjS and PBC.
21. p.6: It is not correct that DNA methylation in general has a repressive effect of gene expression. DNA methylation in CpGisland within gene promoters may generally has a repressive effect. The importance of the location of differential methylation in relation to gene regions (e.g. promoter versus gene body) needs to be addressed. On a similar note are histone modifications not less predictive (actually, the correlation of DNA methylation status with gene expression levels is quite limited to specific genes and tissue types/circumstances) regarding their effect on gene expression, but they are more divers given the plethora of different histone modifications.
Answer 21: We are really thankful for this comment. Accordingly, the statement that methylation leads to gene silencing was rewritten so as to make clear that the methylation of gene promoters leads to gene silencing. In addition the statement that histone modifications have a less predictable effect on gene expression was omitted.
22 :Table 2 is incomplete and does e.g. not include studies by Zilahi et al, Chen et al, Wang et al. and Wang-Renault et al. It would be very valuable if information about the main findings and/or targets of the differentially expressed miRNAs cloud be included in Table 2 and Table 3.
Answer 22: Table 2 has been corrected with inclusion of the studies of Zilahi et al, Chen et al, Wang et al and Wang-Renault et al. In addition a new column with the main findings of each study was added to tables 2 and 3. The list of references was changed accordingly with the studies added.
23: The majority of studies cited (ref 116-121) in the context of future directions for development of epigenetic treatments are relatively old. To add up-to-date developments (and corresponding references) would be appreciated.
Answer 23: The references 117-121 (primary version) were changed with more up-to date ones ref 125-132.
24: The potential role of long interspersed nuclearelement-1 (LINE-1) in pSS pathogenesis, especially with regard to the IFN signature, should be included in the review. Mavragani et al. (Arthritis Rheum 2016 and J Autoimmun 2018) have investigated promoter methylation and gene expression of LINE-1 minor salivary gland biopsies from pSS patients and controls.
Answer 24: The recent studies of Mavragani et al, regarding the role of LINE-1 expression in SjS are now included in the text.
25: Minor comments
Answer 25: Answer: The reference 54 was replaced by the reference 75; The reference 77 was also replaced by the reference 71; Gene symbols were put in italics; The phrase “simultaneous genomic-epigenomic analysis” was replaced by the phrase “integration of genomic and epigenomic data”;
Round 2
Reviewer 1 Report
I am satisfied with the revision. Authors answered most of my concerns.
Author Response
Thanks for your positive comments
Reviewer 2 Report
The revised version of manuscript is clearly improved. I would recomment to accept the manuscript for publication after careful proof reading regarding spelling mistakes and other minor errors, e.g. page 4, line 167 ("In addtion, a more [...]"); p 5 line 210 ("Imgenberg-Kreuz et al,"); p 6 line 241 "EWAS" and "GWAS" have been switched, etc.
Author Response
We appreciate your comments and have make change accordingly.
