Abstract
Objective. Endometriosis, which is a common disease affecting approximately 10% of women of reproductive age, usually causes dysmenorrhea and infertility, thus seriously harming the patients’ physical and mental health. However, there is a mean delay of 6.7 years between the onset of the symptoms and the surgical diagnosis of endometriosis. There is an increasing amount of evidence that suggests that epigenetic aberrations, including deoxyribonucleic acid (DNA) methylation, play a definite role in the pathogenesis of endometriosis. This study aimed to explore the noninvasive or minimally invasive biomarkers of this disease. Materials and Methods. Six patients with surgically confirmed ovarian endometriosis and six patients who received IUD implantation for contraception without endometriosis were recruited in the East Hospital of Tongji University in 2018. The genome methylation profiling of the eutopic and ectopic endometrium of ovarian endometriosis patients and normal endometrial specimens from healthy women was determined using a methylation microarray test. The test screened methylation-differentiated 5′–C–phosphate–G–3′ (CpG) sites and then located the target genes affected by these sites following sequence alignment. Then, an additional 22 patients and 26 healthy controls were enrolled to further verify the difference in the selected genes between endometriosis patients and healthy women. The differential DNA methylation of the selected genes was validated via direct bisulfite sequencing and analysis of their messenger ribonucleic acid (mRNA) levels using quantitative reverse transcription polymerase chain reaction (qRT-PCR). Results. Fifteen differentially methylated CpG sites were found among the patients and healthy women, and five CpG sites were mapped to the introns of the human leukocyte antigen-C (HLA-C) gene; these were highly polymorphic between different HLA-C alleles and were HLA-C07 specific. The results indicated that the HLA-C07 carrier patients exhibited significantly higher DNA methylation levels at the intron VII of HLA-C compared to the HLA-C07 carrier healthy controls. High HLA-C07 mRNA levels were also observed using qRT-PCR with HLA-C07-specific primers, which indicated that the hypermethylation of CpG in intron VII might suppress a silencer that regulates HLA-C07 expressions. Conclusion. Deoxyribonucleic acid hypermethylation in the intron VII of the HLA-C07 gene appears to regulate the expression of HLA-C07. The aberrant DNA methylation in this region was positively correlated with the occurrence of endometriosis.
1. Introduction
Endometriosis is a chronic gynecologic disease that affects approximately 10% of women of reproductive age [1]. The disease is typically characterized by the presence and growth of endometrial epithelium and stroma outside the uterus, primarily in the pelvic peritoneum and on the ovaries [2]. Endometriosis is a benign disease, however, it is aggressive, with distant metastases and a high recurrence rate. Its high prevalence and severe complications, which include infertility and pelvic pain, make this disease a major public health concern.
The etiology and pathogenesis of endometriosis require further investigation. The existing theories on the cellular origin of endometriosis include retrograde menstruation, coelomic metaplasia, embryonic cell rest theory, and stem cell theory [3, 4]. The occurrence and severity of endometriosis are influenced by hormonal levels, genetic background, and immune response, as well as certain environmental and lifestyle factors [5]. Recent evidence suggests that various epigenetic modifications play an essential role in the etiology and pathogenesis of endometriosis [6].
Deoxyribonucleic acid (DNA) methylation is a widely understood epigenetic modification, and aberrant DNA methylation represents a potential mechanism of endometriosis [7–9]. Ectopic endometrial tissues express variable levels of DNA methyltransferase enzymes, which introduce and maintain DNA methylation on the C5 position of cytosine in 5′–C–phosphate–G–3′ (CpG) dinucleotides [10]. Abnormal DNA methylation in endometriosis then affects the expression of several genes, including homeobox A10 [11–13], estrogen receptor [14], progesterone receptor [15], steroidogenic factor 1 (SF-1) [16, 17], and aromatase [18].
It has been estimated that there is a mean delay of 6.7 years between the onset of the symptoms and the surgical diagnosis of endometriosis [19, 20]. Many studies have been conducted to identify novel diagnostic markers. Although the immunologic (IL-6 [21], TGF-β [22], etc.), genetic (miR-200 family [23], exosomal miR-215-5p [24], etc.), and heterogeneous (CA125 [20], leptin [25], etc.) biomarkers may be helpful in the diagnosis of endometriosis, most of them are not highly specific and can be changed in many other pathological conditions such as inflammation and malignancy. In addition, many markers are not sensitive enough in patients with early endometriosis [26]. For example, CA125 is the most valuable marker in clinical application. CA125 ≥ 30 U/mL can highly predict female patients with pain and/or infertility, but CA125 < 30 U/mL cannot exclude endometriosis [20]. Therefore, CA125 can be used as a routine test in symptomatic patients, but is not recommended as a screening marker. Thus, there is an urgent need for novel potential biomarkers. This study involves the screening of several epigenetic candidates as biomarkers for endometriosis and evaluates the effect of DNA methylation on gene expression.
2. Materials and Methods
2.1. Study Design
2.1.1. Microarray Study
To screen and locate the genome methylation sites. Patients with ovarian endometriosis and healthy women were recruited in a small sample in the Shanghai East Hospital from 2018.7.1–2018.10.31. There were 21 eligible patients with ovarian endometriosis, and finally, 6 women were willing to participate while 6 of 48 eligible healthy women joined. The methylation-differentiated 5′–C–phosphate–G–3′ (CpG) sites screening of the eutopic and ectopic endometrium of ovarian endometriosis patients and normal endometrial specimens from healthy women was determined using a methylation microarray test. Then we located the target genes affected by these sites following sequence alignment.
2.1.2. Validation Study
To further verify the difference of the selected genes between patients with ovarian endometriosis and healthy women.
An additional 22 patients and 26 healthy controls were enrolled in the same hospital during 2019.2.1–2020.1.31. The differential DNA methylation of the selected genes was validated via direct bisulfite sequencing and analysis of their messenger ribonucleic acid (mRNA) levels using quantitative reverse transcription polymerase chain reaction (qRT-PCR).
We nonselectively enrolled eligible patients who were willing to participate in both two cohorts.
2.2. Participants
Diagnostic criteria were as follows: (1) patients with ovarian endometriosis were surgically and pathologically confirmed and (2) healthy women were selected those people who were about to receive IUD implantation for contraception without dysmenorrhea, infertility, or chronic pelvic pain, while no ovarian cyst detected by gynecological ultrasound, and serum CA125 was normal.
Inclusion criteria were as follows: (1) women aged 20 to 45; (2) with regular menstrual cycle within one year; (3) good mental state; and (4) willing to participate.
Exclusion criteria were as follows: (1) took oral sex hormones within 3 months; (2) with a history of the pelvic inflammatory disease; (3) women in menstrual period or pregnant; (4) with a history of the gestational trophoblastic disease; (5) women in menstrual period or pregnant; (6) suspected or previously suffered from cancer; and (7) suspected or previously suffered from other chronic diseases, such as diabetes, hypertension, hyperlipidemia, and immunological diseases.
2.3. Human Tissue Collection
Eutopic endometrium from disease-free participants and patients with endometriosis was obtained at the time of laparoscopy, and ectopic endometrium from the cyst walls of ovarian endometriomas was obtained immediately after surgery. Endometriosis was confirmed for each sample via histological examination. The general characteristics of the participants are listed in Table 1.
2.4. Deoxyribonucleic Acid Isolation and Bisulfite Treatment
The total genomic DNA from each sample was isolated using a QIAamp Fast DNA Tissue Kit (Qiagen, Valencia, California, the United States), and the DNA quality was assessed via visualization following agarose gel electrophoresis. To selectively convert unmethylated cytosine to uracil, 1 µg of the genomic DNA was subjected to bisulfite treatment using EZ DNA Methylation kits (Zymo, Orange, California, the United States).
2.5. Beadchip Arrays
The converted genomic DNA was hybridized to the Infinium Human Methylation 450 BeadChip assay and scanned using the Illumina iScan system, while the image data were processed using Genome Studio software. Methylated and unmethylated signals were used to compute beta (β) scores, which are quantitative scores of DNA methylation levels, ranging from 0 (indicating completely unmethylated) to 1 (indicating completely methylated).
2.6. Human Leukocyte Antigen-C07 Typing
Human leukocyte antigen (HLA)-C07 typing was performed using the polymerase chain reaction (PCR) method with HLA-C07-specific primers (forward, 5′-AGAAAGCAGAAGTCCTTCT-3′ and reverse, 5′-TACATCCGTCCTTCATTGTCA-3′). The PCR was carried out in a thermocycler at 94°C for 3 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, elongation at 72°C for 1 min, and, finally, extension at 72°C for 3 min. The amplified PCR products were separated in 1% agarose gel and visualized using ethidium bromide staining.
2.7. Bisulfate Sequencing Polymerase Chain Reaction
The bisulfite sequencing primers (forward, 5′-TGTGGTGGGTYGTTTAGAGTGTTATT-3′ and reverse, 5′-AATAAACRATAACACTCTAAACAACC-3′) for HLA-C07 intron VII amplified a 298-base pair product flanking eight CpG dinucleotides. The PCR conditions were 94°C for 3 min, followed by 40 cycles of denaturation at 94°C for 1 min, annealing at 52°C for 2 min, elongation at 72°C for 1 min, and then extension at 72°C for 10 min. The PCR products were cloned in a pGEM-T Easy vector (Promega, Madison, Wisconsin, the United States). Following transformation, 20 individual clones were selected at random, and 10 positive clones containing the correct insert were then sequenced using an Applied Biosystems 377 instrument. The sequence data were analyzed using Molecular Evolutionary Genetics Analysis version 5 software.
2.8. Reverse-Transcription and Quantitative Polymerase Chain Reaction
The total ribonucleic acid (RNA) was isolated using TRIzol™ (Invitrogen, the United States), according to the manufacturer’s protocol. A 1-µg volume of total RNA was used to generate complementary DNA (cDNA) with the Superscript III first-strand synthesis system (Invitrogen, Carlsbad, California, the United States). Real-time quantitative PCR was performed using a SYBR green mix on an ABI 7900 detection system (Applied Biosystems, Massachusetts, the United States). The relative gene expression was assessed using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a reference gene. The following primers were used for the HLA-C07 coding region: forward, 5′-CGGGATGGGGAGGACCAGACCC-3′ and reverse, 5′-CATAGCGGTGACCACAGCTCCA-3′, while the GAPDH primers were forward, 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse, 5′-GGCTGTTGTCATACTTCTCATGG-3′.
2.9. Data Analysis
For the microarray study, the differences in CpG methylation sites were determined with a false discovery rate of <0.05, , and Δβ > 0.04, while for the validation study, the percentage methylation of each clone obtained from every individual was treated as a single value for the statistical analysis of the bisulfite sequencing. The data were analyzed using a Student’s t-test to compare the percentage methylation of the two groups, with a statistical significance set at the level of
3. Results
3.1. Differential Screening of the 5′–C–Phosphate–G–3′ Methylation Sites in the Endometrial Tissues of Patients and Healthy Women
The methylation profiles of ectopic and eutopic endometrium from the endometriosis patients were compared with the eutopic endometrium of healthy women. The 15 common sites of comparison are shown in Figure 1. No significant differences were observed in the methylation profiles of the 15 sites with respect to the ectopic tissues, suggesting that these sites may be potential diagnostic markers of endometriosis; there were only significant differences in the methylation profiles of the patients and the healthy controls.

Of the 15 differentially methylated CpG sites, 11 were hypermethylated in the patients, while the other four sites were hypomethylated. All the sites were then mapped to the genome sequence (Table 2). Ten sites were located in different genes or loci, but the other five CpG sites (cg03216697, cg01521131, cg09382842, cg11574174, and cg09556042) were mapped to the same gene, namely, HLA-C. The HLA-Cgene-specific CpG sites were hypermethylated in the endometriosis patients, except for cg03216697.
3.2. Human Leukocyte Antigen-C07 Typing
A potential bias in estimating DNA methylation with 450K arrays can arise from the impact of single nucleotide polymorphisms. This is of particular importance in the HLA region due to the high density of polymorphic sites found there and the sequence similarity among proximal HLA genes. Following sequence alignment (Figure 2), the CpG sites of cg03216697, cg01521131, cg09382842, and cg09556042 were found in the HLA-C07 allele cluster, but they were abolished in the other HLA-C alleles although the CpG site of cg11574174 was conserved in all the HLA-C alleles.

(a)

(b)

(c)

(d)

(e)
To rule out the effects of HLA polymorphisms on the methylation microarray test, the HLA genotypes of the 12 participants evaluated in the microarray were analyzed. Six patients were found to be HLA-C07 carriers, while there were only three such carriers in the healthy control group. This result indicates that the differential methylation level of HLA-C may be due to genetic polymorphism. Nevertheless, it was observed that of the three healthy HLA-C07 carriers, one patient had a methylation profile similar to that of the endometriosis patients, but the other two had markedly different profiles (Figure 3). It was therefore assumed that the methylation profile of HLA-C07 in endometriosis patients and healthy women is different.

3.3. Validation of the Differential Deoxyribonucleic Acid Methylation of Human Leukocyte Antigen-C07 by Direct Bisulfite Sequencing
To further verify the DNA methylation difference of HLA-C07 between endometriosis patients and healthy women, an additional 22 patients and 26 healthy controls were enrolled in the study, and the HLA genotypes of all 48 were typed. Eight HLA-C07 carriers in the patient group and eight in the healthy control group were selected for validation analysis, which focused on the intron VII of HLA-C07 since two adjacent differentially hypermethylated CpG sites (cg09382842 and cg11574174) were mapped to this region. The intron VII of HLA-C07 was then amplified using an HLA-C07-specific bisulfate sequencing PCR following bisulfate modification. The methylation profiles of the ectopic and eutopic endometrium of the patients were similar, but both were significantly different from the eutopic endometrium profiles of the healthy women (Figure 4). The ectopic and eutopic endometrium of the patients presented a dense methylation pattern in the intron region, while most of the CpG sites were not methylated in the eutopic endometrium of the healthy women. Of the 8 HLA-C07 endometriosis patients, 87.5% (7/8) of patients exhibited significantly higher DNA methylation levels in the intron VII of the HLA-C gene, compared to 25.0% (2/8) in healthy women.

3.4. High Expression of Human Leukocyte Antigen-C07 in Endometriosis Patients
To characterize the effects of DNA methylation at the intron VII region on HLA-C07 gene expression, the messenger ribonucleic acid (mRNA) levels of the HLA-C07 carriers were analyzed using reverse transcription qPCR. As shown in Figure 5, the expression levels of HLA-C07 in the patients were found to be slightly, but significantly, higher than the levels in the healthy carriers.

4. Discussion
While the pathogenesis of endometriosis is not clear, an increasing number of studies have confirmed that it may be an epigenetic disease. Deoxyribonucleic acid methylation, which affects the occurrence and development of diseases by regulating gene transcription, is the most common modification mode in epigenetics The acquisition of abnormal methylation patterns in patients with endometriosis through noninvasive methods may help in exploring the pathogenesis of the disease and searching for new diagnostic markers and new treatments.
Using a methylation microarray, this study found 15 CpG sites that were differentially methylated in patients with endometriosis and healthy women. Five CpG sites mapped to HLA-C, an HLA class I gene, were confirmed as HLA-C07 specific. The frequency of intron VII methylation for HLA-C07 in the endometrial tissues of the endometriosis patients was significantly higher than that of the healthy controls, as detected via bisulfite sequencing. There was no significant difference in HLA-C07 methylation between the eutopic and ectopic endometrium of the patients. This finding suggests that HLA-C07 intron VII hypermethylation may be an epigenetic marker for endometriosis. This special methylation pattern is likely to be involved in the onset of endometriosis, but this requires further investigation.
Genetic variation in HLA genes is of important clinical significance since it is associated with susceptibility to autoimmune and infectious diseases and plays a major role in both transplantation and cancer. The expression of HLA class I has been found to be higher in ectopic endometrial tissues when detected via immunohistochemistry [27], and HLA class I expression appears to be induced by interferon and tumor necrosis factor in normal endometrium [28, 29]. However, the levels of these cytokines in the endometrium are similar in women with or without endometriosis [30]. In recent studies, the methylation status of HLA promoters was investigated in terms of esophageal squamous cell carcinomas [31], gastric cancer [32], and psoriasis [33], and the HLA promoter was observed to be hypomethylated or nonmethylated in the healthy controls. While the hypermethylation of promoters leads to HLA class I antigen downregulation in cancer, the aberrant promoter methylation cannot explain the high expression of HLA in endometriosis and HLA-C in psoriasis [33]. The current study found that the hypermethylation of intron VII led to the high expression of HLA-C07 in women with endometriosis.
Deoxyribonucleic acid methylation at the promoter region is a widely accepted epigenetic mechanism affecting gene expression. However, the binding sites for transcription factors might be a great distance from the transcriptional start sites [34], and those located in introns or those that are distal to genes might interact with promoters via looping chromatin [35, 36]. Xue et al. reported that the hypermethylation of CpG islands spanning exon II to intron III activated SF-1 gene mRNA expression in endometriosis [17]. The present study observed a similar mechanism with HLA-C07, and it was hypothesized that intron VII of HLA-C07 might enclose a silencer that is suppressed when hypermethylated, leading to increased HLA-C07 expression.
Depending on the amino acid found at position 80, HLA-C alleles can be divided into two groups, namely, C1Asn80 and C2Lys80 [37], and HLA-C07 is grouped in C1, which is the ligand of the killer cell immunoglobulin-like receptor (KIR) of natural killer (NK) cells. The interaction of C1 and KIR2DL2 or KIR2DL2 then inhibits NK cell activity [38, 39].
Sampson’s hypothesis is the most widely accepted theory for explaining the origins of endometriosis [4]. According to this hypothesis, impaired NK activity in women with endometriosis is thought to promote the implantation and progression of endometrial tissue. The higher HLA-C07 mRNA expression observed in the current study partially explains the resistance of endometrium to NK cells. We hypothesized that the intron VII of HLA-C07 might contain a silencer that downregulates the expression of HLA-C07 (a ligand for the inhibitory KIR signal on the surface of NK cells). Due to the low expression of HLA-C07 in the healthy group, NK cells can normally clear the endometrial tissue that is countercurrent to the pelvic cavity with menstrual blood, reducing the risk of endometriosis. However, the hypermethylation in an intron of HLA-C07 could relieve the inhibition of HLA-C07 expressions, and then by acting on KIR, inhibit the activity of NK cells and promote the occurrence of endometriosis. Of course, this hypothesis needs further study in the future.
Due to the lack of sensitive and specific biomarkers in clinical practice, many patients are already at the advanced stage of the disease when they are diagnosed with endometriosis. This research indicated that multiple abnormal DNA methylation sites in patients with endometriosis were HLA-C07 specific, while the HLA-C07 carrier patients exhibited significantly higher DNA methylation levels at the intron VII of HLA-C compared with the HLA-C07 carrier healthy controls. This may present a new method for the early and noninvasive clinical diagnosis of endometriosis.
The hypermethylation of CpG in the intron VII of HLA-C07 was associated with high HLA-C07 mRNA levels, which could be attributed to the suppression of a silencer that regulates HLA-C07 expressions. This finding contributes to the current understanding of the pathogenesis of endometriosis.
It is worth mentioning that in the microarray study, we nonselectively enrolled eligible patients who were willing to participate. To our surprise, all 6 patients in the microarray study were coincidentally HLA-C07 carriers, not due to any subjective selection behavior by the investigators. Then it seems that the allele frequency of HLA-C07 in endometriosis patients (6/6, 100%) was higher than in healthy controls (3/6, 50%). However, the allele frequencies of HLA-C07 in endometriosis patients (36.4%) and healthy controls (30.1%) were similar in the validation study.
This study involved a number of limitations. First, the number of HLA-C07 carriers evaluated was relatively small and additional studies are needed to support the conclusions. Second, since HLA-C genes are highly variable, the methylation statuses of other alleles also need to be investigated. Finally, the test requires the acquisition of organization and loses the advantage of noninvasiveness. The finding is still a long way from being a diagnostic marker for endometriosis.
5. Conclusions
In conclusion, DNA hypermethylation in the intron VII of the HLA-C07 gene appears to regulate the expression of HLA-C07. The aberrant DNA methylation in this region was positively correlated with the occurrence of endometriosis. This finding may be of more significance to advance the research on the immunopathological mechanism of endometriosis.
Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Authors’ Contributions
The authors Wenrong Zhao and Lei Lei contributed equally to this manuscript.
Acknowledgments
This study was supported by the Foundation of Shanghai Municipal Health Commission (No. 202040132).