A systematic review of epigenetics of endometriosis

The search strategy identified 18.639 studies, of which 18.517 did not meet the inclusion criteria and were excluded based on title or abstract review. A total of 122 studies were fully screened for inclusion. After the evaluation of exclusion criteria, 57 manuscripts remained and were included in this systematic review ( 25 – 82 ). Figure 1 illustrates the PRISMA flow diagram of study screening and selection. Among the studies included, 50 (88%) were case-control studies, and 7 (12%) were cross-sectional. This systematic review included 1.623 endometriosis patients and 1.243 control subjects. Acetylation and methylation are two key histone modifications that modulate chromatin structure and therefore alter transcriptional activity. Accordingly, we divided the included studies based on these two characteristics. Genome-wide methylation studies were used to evaluate the methylation status across an entire individual genome, whereas studies evaluating histone modification were used to analyze how gene expression can be altered without changing the nucleotide sequence. Table 1 and 2 report the list of studies included. Several studies found that gene silencing through hypermethylation of tumor suppressor genes’ promotor regions is more prevalent in endometriotic tissue. Compared to healthy endometrial stromal cells, ovarian endometriotic stromal cells had hypermethylation in the promoter regions of the tumor suppressors, ataxia telangiectasia mutated (ATM) ( 25 ) and GATA-binding factor 2 (GATA2) ( 26 ). Higher rates of hypermethylation of E-cadherin (CDH1) and Ras association domain family 1A (RASSF1A) was found in eutopic and ectopic endometrium from women with endometriosis compared from healthy controls ( 27 , 28 ), and eutopic endometrium from women with ovarian endometriosis had higher rates of grainyhead-like 2 gene (GRHL2) promoter hypermethylation compared to healthy controls ( 29 ). Xie et al. found that not only was promoter hypermethylation of AT-rich interactive domain 1A (ARID1A) more common in endometriotic tissue, but that oxidative stress further increased hypermethylation, therefore proposing a mechanism by which gene silencing may occur ( 30 ). Several studies evaluated the methylation status of tumor suppression genes in endometriosis and endometriosis-associated ovarian cancer. Both Wang et al. and Guo et al. found that hypermethylation in the promotor region of tumor suppressor runt-related transcription factor 3 gene (RUNX3) was significantly higher in endometriosis-associated ovarian cancer tissues than in endometriosis tissues (p < 0.01), suggesting that this gene silencing may be involved in endometriotic malignant transformation( 31 , 32 ) While Wang et al. found there was no significant difference in RUNX3 methylation between the ectopic and eutopic endometrial issues in the endometriosis group or between the eutopic endometrial tissues of the endometriosis group and that of the controls ( 31 ), Guo et al. did noted higher rates of RUNX3 methylation in eutopic endometrium from those with endometriosis-associated ovarian cancer( 32 ). Ren et al. showed that Ras association domain-containing protein 2 (RASSF2) was more frequently methylated in endometriosis-associated ovarian cancer as compared to ectopic endometrium ( 33 ). In a separate study, Ren at al. found that epigenetic inactivation of mismatch repair gene human mutL homolog 1 (hMLH1) via hypermethylation was significantly associated with cases of ovarian cancer when compared to cases of both endometriosis and controls (p < 0.05) ( 34 ). Martini et al., found similar results ( 35 ). Two studies evaluated oncogene methylation. The expression of syncythin-1 is typically limited to trophoblastic cells as the enzyme mediations the fusion of cytotrophoblasts into syncytiotrophoblasts. However, it also acts as a cell cycle regulator and overexpression has also been linked with cell proliferation and development of cancer. Its normal cell-specific expression is controlled by methylation of the promotor region. Zhou et al. reported that hypomethylation of the promoter region of the human endogenous retrovirus envelope gene HERV-W, the gene that encodes syncythin-1, in endometriotic tissues ( 36 ). Hypomethylation of the gene allowed for syncytin-1 gene expression d in ectopic endometrium. However, the gene remained silenced (i.e. methylated) in the eutopic endometrium, suggesting that altered methylation of this gene may contribute to the pathogenesis of endometriosis ( 36 ). In a separate study, the Twist promoter was found to be hypomethylated in some areas of ovarian ectopic endometrium and eutopic endometrium of ovarian endometriosis ( 37 ). Elucidating the regulation, and dysregulation, of steroid hormones is the pathogenesis of endometriosis was the focus of many studies. Maekawa et al. evaluated the methylation status of estrogen receptor 1 (ESR1), estrogen receptor 2 (ESR2), and progesterone receptor (PGR) among 16 women. Although no significant difference was observed between DNA methylation of ESR2 and PGR in eutopic endometrium and ovarian endometriomas, ESR1 methylation was found to be significantly higher in ovarian endometriomas (p < 0.050) ( 38 ). Meyer et al. also examined the methylation status of these steroid hormone receptor genes, specifically in tissue from deep intestinal endometriosis as compared to eutopic endometrium tissue from the same sources. They found no significant difference in ESR1 and ESR2 methylation, however PGR hypermethylation was observed in 17 of the 44 cases (39%) in promoter regions B ( 39 ). Methylation differences in the ESR2 gene was found by Xue et al.; a key CpG island in the promoter of the ESR2 gene was heavily methylated in endometrial stromal cells and largely unmethylated in endometriotic stromal cells, with higher levels of ESR2 gene expression in endometriosis (p < 0.0001) ( 40 ). Several studies examined specific isoforms of the progesterone receptor gene. Whereas Febri et al. found that both PGR-A and PGR-B were more intensely methylated in the endometriosis group compared to controls (p < 0.05) ( 41 ), both Rocha-Junior et al. and Wu et al., found that PGR-A was unmethylated in both the endometriosis and control group, but PGR-B was methylated in the endometriosis group ( 42 )( 43 ). In granulosa cells from women with endometriosis and infertility, Hayashi et al. found that the promoter of the androgen receptor (AR) was more heavily methylated and had associated decreased expression of AR mRNA as compared to granulosa cells in the control group ( 44 ). Genes involved with hormone regulation were also studied. Steroidogenic factor 1 (SF-1) is a nuclear receptor that regulates steroidogenic enzymes such as aromatase, the enzyme that catalyzes the final step in estrogen synthesis. Xue et al. compared SF-1 methylation in the eutopic endometrium of eight women without endometriosis to ovarian endometrioma samples from eight women with endometriosis. The SF-1 gene was significantly hypermethylated in the ovarian endometriomas (p < 0.001) and this correlated to significantly increased SF-1 mRNA levels (r = 0.980, p < 0.001) ( 45 ). In a 2008 study, Izawa et al., demonstrated that adding a demethylating agent to media containing endometriotic tissue led to a significant increase in aromatase gene mRNA and in estrogen production ( 46 ). They were later able to identify the non-promoter CpG island that became hypomethylated ( 47 ). Unlike the studies conducted by Izawa et al., Hosseini et al. found that expression of CYP19A1, the gene that encodes aromatase, in endometriosis patients was significantly lower than the control group (p = 0.04) ( 48 ). However, they were examining the cumulus cells from patients undergoing fertility treatment, whereas Izawa et al., studied stromal cells. Catechol- O -methyltransferase (COMT) inactivates estradiol metabolites that bind to ER and has been suggested to play a role in the development of endometriosis ( 49 ). Among 60 women with endometriosis, Ji et al. found significantly increased methylation levels in COMT promoter regions of ectopic tissue as compared to eutopic endometrial tissue (p < 0.001) ( 50 ). Joshi et al. specifically evaluated eutopic endometrium during the mid-secretory phase in women with endometriosis and infertility. They found that those with low endometrial integrin (which the authors used as a marker for lower pregnancy rate) had higher expression of the aryl hydrocarbon receptor (AHR) gene and that AHR promotor region tended to be hypomethylated compared to controls. ( 51 ). While AHR does have tumor suppressive functions, it is also involved with cell differentiation and proliferation as well as estrogen receptor signaling, which may explain its role in endometriosis ( 52 , 53 , 54 ). Multiple studies reported an association between aberrant methylation of the HOXA10 gene and endometriosis. HOXA10 expression in the endometrium varies across the menstrual cycle and is thought to play a role in endometrial receptivity. Previous studies have suggested that an altered cyclic expression of HOXA10 in the endometrium may explain infertility associated with endometriosis ( 55 ). Muharam et al., Kulp et al., Wu et al., and Andersson et al. found a significant difference in HOXA10 methylation status among patients with endometriosis when compared to controls ( 56 – 59 ). Specifically, Kulp et al. and Wu et al. reported significant hypomethylation of the HOXA10 gene in eutopic endometrium among women with endometriosis, as compared to women without endometriosis (p < 0.010 and p = 0.030, respectively) ( 57 , 58 ). Andersson et al. and Ji et al. reported decreased HOXA10 methylation levels in ectopic endometrium as compared to eutopic endometrium among women with endometriosis (p < 0.001 and p = 0.001, respectively) ( 50 , 59 ). Esfandiari et al. found that most HOX clusters (A-D) and HOX cofactors showed methylation alterations in ectopic/eutopic endometrial tissues and endometriosis organoids compared with normal endometrium ( 60 ). However, Signorile et al. found no significant difference between patients with endometriosis and healthy controls in the methylation of the HOXA10 regulatory elements ( 61 ). Given the fluctuations in HOXA10 expression throughout the menstrual cycle, Fambrini et al. specifically examined HOXA10 promoter methylation levels during the mid-luteal phase. They reported higher rates of promoter hypermethylation in women with ovarian endometriomas than in healthy controls ( 62 ). Two studies evaluated the methylation status of the NF-IL6 site in the promotor region of prostaglandin-endoperoxide synthase 2 or cyclooxygenase-2 (COX-2). Zidan et al. reported significant hypomethylation in eutopic and ectopic endometrial samples from 60 women with endometriosis as compared to 30 controls (p = 0.002 and p = 0.000, respectively). ( 63 ). Wang et al. found lower levels of promotor methylation in the eutopic endometrium from women with endometriosis compared to women without ( 64 ). In both studies, higher expression of COX-2 mRNA was noted. ( 63 , 64 ). Zhao et al. found hypomethylation of the interleukin-12B (IL-12B) promoter in both ectopic and eutopic endometrial samples from women with ovarian endometriosis when compared to the endometrium from controls (p < 0.001 and p = 0.041)) This resulted in increased IL-12B expression in the tissue from those with endometriosis ( 65 ). IL-12B encodes the protein IL-12p40, which negatively regulates IL-12, a key immunomodulator that inhibits the development of endometriotic lesions ( 66 ). Matrix metalloproteinases (MMPs) are critical to tissue remodeling and extracellular matrix remodeling, key processes for endometriosis progression. Two papers evaluated methylation of the MMP-2 gene. Zahrah et al. found that MMP-2 mRNA expression was increased in peritoneal endometriosis tissue as compared to endometrial tissue from health controls. While the MMP-2 gene was hypermethylated, there was no statistically significant difference between peritoneal endometriosis tissue and endometrial tissue ( 67 ). However, Tang et al. did find that four CpG sites of MMP-2 were hypomethylated in in ectopic proliferative-phase endometriotic lesions as compared to eutopic endometrium from unaffected controls. They also found that four CpG sites in tissue inhibitor of metalloproteinase 3 (TIMP3) were hypermethylated in the cases compared to controls ( 68 ). Zhao et al. demonstrated that hypomethylation of the glutathione S-transferase M1 gene (GSTM1) and increased GSTM1 mRNA and protein levels were associated with endometriosis ( 69 ). GSTM1 is a detoxification enzyme and regulates apoptosis pathways; altered expression in the endometrium has been suggested to play a role in endometrial progression to endometriosis. Izawa et al. detected hypomethylated CpGs within the GATA6 gene which correlated with an upregulation of GATA6 mRNA expression by 50-fold in endometriotic cells when compared to endometrial cells ( 70 ). This finding is echoed by Dyson et al. ( 26 ). Among nine women with peritoneal endometriosis, Deraya et al. reported significant hypermethylation of RAC1 in peritoneal samples when compared to normal endometrium from 20 women without endometriosis (p = 0.044). ( 71 ). Kawano et al. reported lower levels of acetylated histone H3 and H4 in endometriotic cyst stromal cells in comparison to normal endometrial stromal cells (P < 0.025) ( 72 ). Xiaomeng et al. detected significant reduction of global histone H4 acetylation level in endometriosis patients. However, they found no difference in H3 acetylation status between endometriosis patients and controls ( 73 ). Interestingly, a similar study by Monteiro et al. reported global hypoacetylation at H3 but not H4 in endometriotic lesions compared to eutopic endometrium from controls. Endometriotic lesions were found to contain significantly lower levels of H3K9ac and H4K16ac in comparison to eutopic endometrium, indicating hypoacetylation ( 74 ). Hosseini et al. echoed these results by reporting significant hypoacetylation of histone 3 in a promoter region of the CYP19A1 gene ( 48 ). Four studies reported on the methylation patterns of histones. In 2014, Monteiro et al. reported hypermethylation of H3K4, H3K9, and H3K27 in endometriotic lesions compared to endometrium from controls (P=0.0001, 0.0015, 0.0001, respectively) ( 74 ). Xiaomeng et al. reported significant reduction in methylation of H3K4 and H3K9 in ectopic endometrial tissue in comparison to control tissue (p < 0.001) ( 72 ). A study by Colon-Caraballo et al. in 2015 found no significant differences in H3K27me3 levels between endometriotic lesions and eutopic endometrium from patients and controls by immunoassay. However, by immunohistochemistry analysis, they observed higher percentages of H3K27me3-positive nuclei in endometriotic lesions and secretory endometrium compared to eutopic endometrium from patients ( 75 ). Hosseini et al. studied the methylation levels of H3K9 in different promoter regions of the CYP19A1 gene and found it to be significantly elevated in the PI.4 promoter (P=0.02) ( 48 ). Three studies reported variations in the levels of enzymes related to histone modifications. Colon-Diaz et al. reported that basal gene expression levels of both histone deacetylase 1 and 2 (HDAC1 and HDAC2) are higher in endometriotic vs. endometrial stromal cells ( 76 ). The study by Xiaomeng et al. demonstrated increased HDAC2 mRNA levels in the eutopic endometrium of endometriosis patients (P < 0.001). It was also reported significantly decreased levels of HDAC1 mRNA in the ectopic endometrium (P = 0.006) of endometriosis patients ( 73 ). In the same study, Xiaomeng et al. also reported increased levels of the histone acetyltransferase PCAF and decreased levels of several histone methyltransferases in the ectopic endometrium of endometriosis patients vs. controls ( 73 ). A more recent study by Colon-Caraballo et al. examined the levels of the histone methyltransferase EZH2α and found increased levels in the glands of the endometriotic lesions compared to both proliferative endometrium from patients (P < 0.0001) and controls (P < 0.05) and to secretory endometrium from controls (P < 0.05) ( 77 ). A 2013 study by Kawano et al. found that increased acetylation of histones H3 and H4 in the promoter region of the CCAAT/enhancer-binding protein α (C/EBPα) led to the inhibition of cell proliferation and induction of apoptosis. Knockdown of C/EBPα then led to stimulation of cell proliferation and resistance to apoptosis in normal eutopic endometrial stromal cells via the downregulation of several genes, including proliferator-activated receptor-γ (PPARγ), period homolog 2 (PER2), p53, apoptosis-inducing factor, mitochondrion-associated 1 (AIFM1), Bax, caspase-8, caspase-10, p16(INK4a), p21(Waf1/Cip1), cyclin-dependent kinase (cdk) 2, and cdk4 ( 78 ). A similar study by Kai et al. found that death receptor 6 (DR6) is epigenetically suppressed by hyperacetylation of histone H4 in endometriotic cyst stromal cells ( 79 ). In the 2018 study by Colon-Caraballo et al., increased expression of the histone methyltransferase EZH2α was correlated with enrichment of H3K27me3 in the promoter regions of candidate genes ESR1, CDH1, and PGR ( 77 ). The study by Monteiro et al. found hypoacetylation of H3 and H4 in several candidate genes known to be downregulated in endometriosis including HOXA10, ESR1, CDH1, and p21 . Conversely, hyperacetylation of H3 and H4 was correlated with high expression of the stereoidogenic factor 1 (SF1) in endometriotic lesions ( 74 ). Barjaste et al. created DNA methylation heatmaps for ectopic and eutopic endometrial tissue from nine patients with endometriosis and compared it with endometrial biopsies from six healthy fertile women ( 80 ). The study revealed that DNA methylation has the greatest difference when comparing ectopic to control samples in comparison to eutopic and control tissue ( 80 ). Naqvi et al. found that in seven women with endometriosis, 120 genes expressed significant methylation. Ten of those genes were examined further due to their association with endometriosis. MGMT, DUSP22, CDCA2, ID2, and RBBP7 were found to be significantly hypermethylated and BMPR1B, TNFRSF1B, ZNF681, IGSF21, and TP73 were significantly hypomethylated in women with endometriosis as compared to controls ( 81 ). Among 16 women with ovarian endometriomas, Mihara et al. evaluated transcriptome and epigenome data to analyze upstream regulators involved in the pathogenesis of endometriosis and identified thirty-four genes as possible upstream regulators. Of those, GLI3, HOXC8, CEBPD, and NR3C2 were significantly upregulated by greater than 2-fold when compared to controls, and HOXA10, MAPK8, ETS2, GATA2, ESR1, HOXA9, TFAP2C, and PRDMI were significantly downregulated by less than 0.5-fold when compared to controls. Mihara et al. reported increased HOXC8 expression in ovarian endometrioma ( 82 ). These ovarian endometrium stromal cells overexpressing HOXC8 also demonstrated activation of the TGF-β signaling ( 82 ). Zelenko et al. identified 14 genes that were upregulated in endometriosis and five were downregulated. Specifically, THRA expression was decreased in women with endometriosis whereas HDAC1, HDAC2, and DDX5 were upregulated in endometriosis in the proliferative and early secretory phases ( 83 ). Borghese et al. found hundreds of differently methylated regions in the DNA of patients with endometriosis ( 84 ). Hypomethylated genes included E2F3, RXRA, while hypermethylated genes included RXRB, HOXD10, HOXD11, TFDP3, PCGF3, and ZNF426 ( 84 ). Comparatively, Saare et al. identified 28 differentially methylated regions between endometriosis and control patients, 16 of which were associated to known genes (PI3, SLC43A3, MGAT5B, MUC4, HIVEP3, FGG, CLCF1, CANT1, LTK, AHRR, AKR1B1, APEH, CST11, ELOVL4, HBE1 and NEGR1) ( 85 ). Yotova et al. found differently methylated genes in encoding transporters (SLC22A23), signaling components (BDNF, DAPK1, ROR1, and WNT5A) and transcription factors (GATA family, HAND2, HOXA cluster, NR5A1, OSR2, TBX3) ( 86 ). In a large survey conducted by Ramioglu et al., 27.493 significantly differentially methylated probes were found, corresponding to 8.133 independent genes in endometriotic disease tissue compared to endometrium samples ( 87 ). These results represented enrichment in the WNT signaling pathway, angiogenesis, Alzheimer disease presenilin pathway, cadherin signaling, integrin signaling, gonadotropin-releasing-hormone-receptor pathway, and CCKR signaling ( 84 ).

Studies were identified by searching publications in PubMed, EBSCOhost, Cochrane Library, Embase, Scopus, and Web of Science Core Collection from inception to January 15th, 2023. The review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines ( 23 ) ( Figure 1 ). A comprehensive list of terms related to genetic mutations in endometriosis was designed in collaboration with a data informationist. Medical Subject Headings (MeSH) terms and text related to endometriosis and genetic mutations were included. A sample search strategy used includes: endometriosis, epigenomics, gene expression profiling, genome-wide studies, histone modification etc. The complete list of search strategies is included in supplemental materials Appendix Table 1 . We considered English-language publications which assessed genetic or epigenetic analysis of patients with endometriosis or adenomyosis. To review the most recent data, we conducted the search from 1995 onwards, which resulted in the inclusion of studies from the last 20 years. Studies were restricted to human subject research (both clinical and preclinical) and study types included in-vitro studies, case-control studies, cohort studies, randomized control trials (RCTs). Review articles, case series, reports and letters, animal studies, transcriptome, proteome, and metabolome studies were excluded. After the initial search strategy, three authors (LC, JZ, BB) independently reviewed all titles, and abstracts according to the eligibility criteria in two rounds. Full-text articles were selected for studies not excluded after the title and abstract review to ensure complete inclusion criteria was met. Any discrepancies were resolved by the senior author. A total of 57 studies pertaining epigenetic mechanisms and endometriosis in humans and human tissues were ultimately included. Studies without an English language full-text, scientific abstracts and clinical trial protocols were also excluded. Studies were then classified as preclinical study or clinical study. Figure 1 presents the PRISMA flow diagram of study screening and selection. Study quality and risk of bias was assessed by three reviewers (LC, JZ, MS) and any discrepancies were resolved by the senior author. Risk of bias was assessed in the domains of the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome and selection on the reported result. The quality of non-observational case control and cohort studies was determined to be good, fair or poor based on the Newcastle-Ottawa scale by the assessment of selection, comparability, and outcome sections ( 24 ).

Although the results discussed in this review were wide-ranging, several key similarities emerged. Several studies repeatedly found that hypermethylation in the promoter region of tumor suppressor genes and hypomethylation in the promoter region of oncogenes was associated with endometriosis. Studies of steroid-hormone related genes tended to demonstrate methylation patterns that promoted pro-estrogenic states. Histone modifications were commonly found in endometriotic lesions. Additionally, many studies found that both histone modification and methylation of genetic sequencing occurred concurrently, suggesting multifactorial epigenetic changes and dysregulation in the pathogenesis of endometriosis. Despite extensive research on DNA methylation and endometriosis, there remains some disagreement among the results. Although multiple studies associated HOXA10 hypomethylation with endometriosis, Fambrini et al. reported hypermethylation of the HOXA10 gene. In addition, there are discrepancies on the methylation status of MMP-2. Among women with endometriosis, Zahrah et al. found MMP-2 to be hypermethylated and Tang et al. observed hypomethylation. Similarly, among studies on histone modification, there was little consensus on the acetylation status of H3 and H4, as well as the methylation patterns of H3K4 and H3K9. These differences may be attributable to variability in study methods and overall quality. Histone modifications act in concert with DNA methylation to regulate gene expression. In general, acetylation and methylation of histones lead to activation and suppression of transcriptional activity, respectively. Several studies in this review examined differential acetylation and methylation patterns in histones, especially H3 and H4, as well as variations in levels of histone-modifying enzymes, such as histone acetyltransferases and histone methyltransferases. However, robust evidence remains lacking regarding the extent of involvement of these histone modifiers in the pathophysiology of endometriosis. It is important to consider these histone modifications in the context of candidate genes, such as PPARγ, HOX10, ESR1 , levels of which were all shown to be affected by aberrant histone modifications. The differential expression of these candidate genes has direct implications on cell cycle growth, cell cycle arrest, and apoptosis, all of which play key roles in the pathogenesis of endometriosis. In addition to shedding light on disease etiology, these findings offer insight into the role of histones as potential biomarkers and treatment targets for endometriosis. Several studies in this review discussed the use of histone deacetylase inhibitors (HDACi), such as valproic acid, Trichostatin A, and apicidin, in leading to the inhibition of proliferation of endometriotic lesions. Further research evaluating the efficacy of these therapies will offer valuable insight into clinical treatment of this disease. This review offers a comprehensive and up-to-date collection of currently available evidence regarding the role of epigenetics in the pathophysiology of endometriosis. However, it does have its limitations. Due to the large number of studies included in this review, there is a lack of standardization regarding characteristics of diseased and healthy tissue used in the studies. Methodology varied significantly across studies and even across studies examining the same genes. For example, tissue source varied considerably, with some studies comparing ectopic endometrial tissue to eutopic tissue from the same sample, while others compared ectopic endometrial tissue to eutopic tissue from women without endometriosis. Ovarian endometriomas were a tissue source for cases, but the pathophysiology of ovarian endometrioma development is likely different than that of deep infiltrating endometriosis. Furthermore, the cells used to perform epigenetic testing varied between studies. While the majority of studies utilized stromal cells, others used granulosa cells or cumulus cells—cell types with considerably different functions and genetic expression. Studies included seldom reported clinical outcomes associated with endometriosis such as endometriosis associated pain, dysmenorrhea, infertility etc. Control groups varied considerably as well, with some women having undergone surgery for fibroids or for cervical dysplasia, which may affect the results. Several studies only focused on women who and endometriosis and infertility, limiting comparison to studies conducted in women without infertility. Many genes studied are hormonally mediated, such as HOXA10, but the phase of the menstrual cycle varied considerably between studies, with many studies not controlling for this confounder at all. Some studies controlled for age, hereditary genetic burden, prior hormonal therapy, concurrent conditions, while others were less stringent. Given these methodological limitations, it is challenging, if not impossible to make directional assessments of the data or to summarize findings in a clinically meaningful way for many genes. Greater methodological standardization is needed.

This review offers a comprehensive and updated collection of the current evidence regarding the role of epigenetics in the pathophysiology of endometriosis. Several studies reported a significant difference in the level of methylation of specific genes when comparing endometrial biopsies to normal tissue, which might suggest that DNA methylation has an important role in the modulation of the genotype in endometriotic tissue. Acetylation and methylation were the two key histone modifications that were accountable for the differential expression of the genes studied. These alterations in gene expression in endometriotic tissue can have direct implications on cell cycle growth, cell cycle arrest, and apoptosis and, therefore, might play a key role in the pathogenesis of endometriosis. This review offers insight that histone modifications need further research to evaluate their role as potential biomarkers and treatment targets for endometriosis. Although several key similarities regarding the epigenetics of endometriosis were reported, there were some disagreements among the results, which might be attributable to the heterogeneity between the studies. Further research with a more robust standardization is needed to validate the epigenetic changes in endometriosis.

Endometriosis, a disease caused by the presence of endometrial-like tissue at extra-uterine sites, such as the ovary or the pelvic peritoneum, affects 6% to 10% of reproductive-age women ( 1 , 2 ). Symptoms of endometriosis vary from subfertility to incapacitating pain and disability, which negatively impact the quality of life, education, and employment ( 3 – 5 ). Endometriosis is the third leading cause of gynecologic hospitalization in the United States ( 6 , 7 ). The direct and indirect healthcare costs for treating endometriosis are estimated to be higher than twelve thousand United States dollars per patient per year ( 8 ). The etiology of endometriosis is complex and multifactorial. The most commonly proposed theories for the pathogenesis of endometriosis are retrograde menstruation, coelomic metaplasia, vascular and lymphatic dissemination, induction, and embryonic cell rests ( 9 ). Since there is no unique theory that can explain all disease phenotypes, it is proposed that genetic, immunologic, hormonal, and environmental factors contribute to disease susceptibility ( 9 , 10 ). The role of epigenetics has emerged as a promising contributor to the pathogenesis of endometriosis ( 11 , 12 ). Epigenetics are inheritable alterations to genetic expression that, without affecting the genome itself ( 13 ), modulate the expression of a genotype through DNA methylation, histone modifications, and non-coding RNA. The physical structure of the DNA molecule can be transformed by alterations in DNA and histone methylation, phosphorylation, ubiquitylation, and acetylation. Non-coding RNA, such as microRNA and small interfering RNA (siRNA), impact the expression of key epigenetic regulators such as DNA methyltransferase, recruit chromatin modifiers, and attract RNA polymerase to specific genes ( 14 , 15 ). As a result, epigenetic modifications play a pivotal role in cellular differentiation and may silence or activate key genes at the cellular or organism level. Given the expansive impact of epigenetic modifications, it is unsurprising that the epigenome has been implicated in cancer development and behavior ( 16 , 17 ), rheumatologic and cardiovascular diseases ( 18 , 19 ), and gynecologic conditions ( 20 , 21 ). Unlike mutations in DNA, epigenetic changes are reversible, making them a promising target for disease therapy ( 22 ). The study of epigenome is important for elucidating the underpinnings of endometriosis and finding targets for disease modification and treatment. The specific aim of this systematic review is to evaluate and report the key epigenetic changes associated with endometriosis, with a focus on DNA methylation and chromatin alterations.