Associations between patients with endometriosis and HLA class II; the analysis of HLA‐DQB1 and HLA‐DPB1 genotypes

Human Reproduction Vol.18, No.5 pp. 985±989, 2003 DOI: 10.1093/humrep/deg192 Associations between patients with endometriosis and HLA class II; the analysis of HLA-DQB1 and HLA-DPB1 genotypes Keisuke Ishii1, Koichi Takakuwa, Katsunori Kashima, Masaki Tamura and Kenichi Tanaka Department of Obstetrics & Gynecology, Niigata University School of Medicine, 1±757, Asahimachi-dori, Niigata, 951±8510, Japan 1 To whom correspondence should be addressed. E-mail: BACKGROUND: Although the aetiology of endometriosis remains unclear, many immunological abnormalities involving changes in cell-mediated and humoral immunity may be associated with endometriosis. Several disorders are thought to be associated with particular HLA antigen types. This study examines the possible association between HLA-DQ and DP. METHODS: A total of 83 patients diagnosed with endometriosis following laparoscopic examination were typed for the HLA-DQB1 and DPB1 alleles using PCR±restriction fragment length polymorphism (RFLP). The HLA DQB1 and DPB1 allele frequencies in these patients and in 222 controls were compared. RESULTS: The prevalence of HLA-DQB1*0301 in the patient group was 16.3% (27/166 alleles), compared with 8.3% in the overall control group (37/444 alleles) and 7.7% in the females of the control group (18/234 alleles). Thus, the prevalence of the HLA-DQB1*0301 allele was signi®cantly greater in patients with endometriosis, compared with the general controls [OR 2.13, 95% CI 1.25±3.64, P = 0.004 (c2 analysis), Corrected P-values; Pc = 0.049] and with the general female controls [OR 2.33, 95% CI 1.24±4.39, P = 0.008 (c2 analysis), Pc; NS]. There was no signi®cant association in the frequencies of DPB1 alleles between the patients and controls. CONCLUSIONS: The HLA systems may be involved in the aetiology of endometriosis, although further study is needed. Key words: endometriosis/HLA-DPB1/HLA-DQB1/PCR±RFLP Introduction Endometriosis is a gynaecological disorder of unknown aetiology and poorly understood histogenesis. Many immunological abnormalities have been associated with endometriosis; abnormalities in cell-mediated immunity, in particular, have been reported (Giudice et al., 1998; Senturk and Arici, 1999). In rhesus monkeys with endometriosis, decreased T-cell activity was observed, and in infertile women with endometriosis, T-cell-mediated cytotoxity to autologous endometrial cells is reduced in comparison with infertile women without endometriosis (Dmowski et al., 1981). Supposing that the suppression of cellular immunity in the peritoneal space causes endometriosis, it is possible that particular types of HLA may signal these cells to decrease their cytotoxic property. Several disorders, including Vogt±Koyanagi±Harada's disease, insulin-dependent diabetes mellitus (IDDM), systemic lupus erythematosus (SLE), and pre-eclampsia, as well as recurrent miscarriages, are thought to be associated with particular HLA types (Yao et al., 1993; Shindo et al., 1994; Tisch and McDevitt, 1996; Christiansen et al., 1998; Takakuwa et al., 1999a,b), particularly HLA-DR, which is thought to be an immune response-related gene. The association between ã European Society of Human Reproduction and Embryology endometriosis and HLA antigens has not been proven (Moen et al., 1984; Simpson et al., 1984; Maxwell et al., 1989), and none of the previous studies using serological analysis (microcytotoxicity tests) have found a statistically signi®cant association between endometriosis and HLA allotype frequency. In our previous study, however, we noted that the prevalence of the HLA-DRB1*1403 allele was signi®cantly greater in patients with endometriosis than in the general controls (Ishii et al., 2002). In the present study, we applied the PCR±restriction fragment length polymorphism (RFLP) method to the genotype analysis of HLA-DQB1 alleles and DPB1 alleles in patients with endometriosis. Materials and methods Patients and controls A total of 83 patients diagnosed with endometriosis following laparoscopic examination were typed for the HLA-DQB1 and DPB1 antigens. All of the patients had advanced stage disease, as classi®ed by the Revised American Society for Reproductive Medicine Classi®cation (American Society for Reproductive Medicine, 1996). A total of 38 patients were classi®ed at stage IV of the disease, and 45 patients were at stage III. The patients' mean age (6SD) was 34.4 (63.7) years old. Patients with indicated autoimmune abnormalities 985 K.Ishii et al. (three patients positive for anti-phospholipids antibodies and four positive for anti-nuclear antibodies) were excluded from this study. A total of 44 patients suffered from infertility. The remaining patients in this cohort complained of dysmenorrhoea and/or pelvic pain. A control population of 222 healthy individualsÐ105 males and 117 femalesÐfrom the Niigata prefecture of Japan were studied to determine the distribution of HLA-DQ and HLA-DP antigens in the general population. They were selected as controls randomly and their mean age was 33.0 (65.0) years. All individuals in this study were Japanese, and all gave their informed consent before participating in the study. Analysis of HLA-DQB1 and DPB1 genotypes Analysis of HLA-DQB1 and DPB1 genotypes was performed using PCR±RFLP analysis (Nomura et al., 1991; Ota et al., 1991). Genomic DNA isolated from peripheral lymphocytes was ampli®ed by the PCR procedure. A 241-bp fragment from the second exon of the HLA-DQB1 gene was ampli®ed by using DQ1-group speci®c primers, and a 237-bp fragment was ampli®ed by using DQ2, DQ3 and DQ4 group-speci®c primers. A 299-bp fragment from the second exon of the HLA-DPB1 gene was ampli®ed with DPB1-speci®c PCR primers. Following the ampli®cation, the PCR products were digested with appropriate restriction endonucleases (5 units) for 3 h, electrophoresed through a 12% polyacrylamide gel (Minigel apparatus AE-6450; Atto Corporation, Tokyo, Japan) and visualized by staining with ethidium bromide. HLA-DQB1 and HLA-DPB1 genotypes were determined by comparing the restriction fragment patterns of RFLP obtained in tested individuals with those of ampli®ed genes. The number of DQB1 alleles that could be differentiated by this method was 12, and the number of DPB1 alleles was 19. We were able to type all of the individuals who participated in this study. Institutional review board approval was obtained for this study. Statistical analyses The HLA DQB1 and DPB1 allele frequencies in patients with endometriosis and in the general controls were compared using c2 analysis with Yates' correction. Fisher's exact probability test was used for small expected frequencies that were <5. Corrected P-values (Pc) were obtained by multiplying by the number of alleles tested for each locus (Svejgaard et al., 1974). The odds ratio (OR) was calculated with a 95% con®dence interval (CI). Results The determination of the HLA-DQB1 alleles is mentioned as an example. The products of PCR a (...truncated)