Association of tumor necrosis factor and human leukocyte antigen DRB1 alleles with Graves' ophthalmopathy

Association of Tumor Necrosis Factor and Human Leukocyte Antigen DRB1 Alleles with Graves’ Ophthalmopathy Tomasz Bednarczuk, Yuji Hiromatsu, Naoko Seki, Rafał Płoski, Tomoka Fukutani, Alina Kuryłowicz, Krystian Jażdżewski, Krzysztof Chojnowski, Kyogo Itoh, and Janusz Nauman ABSTRACT: Tumor necrosis factor (TNF)-␣ plays a central role in the development of ophthalmopathy in patients with Graves’ disease (GD). The aim of this study was to investigate the association of TNF promoter polymorphisms at positions -1031 (T-1031C), -863 (C-863A), -857 (C-857T), -308 (G-308A), and -238 (G-238A) with Graves’ ophthalmopathy (GO). We studied the distribution of TNF and human leukocyte antigen (HLA) DRB1 alleles in 228 Polish white patients with GD, 106 of whom had ophthalmopathy (NOSPECS class ⱖIII) and 248 healthy subjects. TNF -308A and HLA-DRB1*03 alleles were significantly increased in patients with GD compared with healthy subjects. Stratification analysis revealed no independent association of -308A with GD when the DRB1*03 status was considered. Subdividing GD according to eye involvement revealed that the distribution of TNF promoter haplotypes differed significantly in patients with or without ophthalmopathy. The haplotype containing the -238A allele was absent in GO. The association of G-238A with GO was independent of DRB1 alleles. These results indicate that TNF G-308A is associated with susceptibility to GD (however, this association is not independent of HLA-DRB1*03) and that TNF G-238A is associated with the development of ophthalmopathy, suggesting that G-238A or a gene in linkage disequilibrium may be disease modifying in GD. Human Immunology 65, 632– 639 (2004). © American Society for Histocompatibility and Immunogenetics, 2004. Published by Elsevier Inc. KEYWORDS: genetic polymorphisms; Graves’ disease; HLA; ophthalmopathy; TNF ABBREVIATIONS GD Graves’ disease GO Graves’ ophthalmopathy HLA human leukocyte antigen SNP TNF INTRODUCTION Graves’ ophthalmopathy (GO) is an autoimmune inflammatory disorder of the extraocular muscles and the orFrom the Department of Endocrinology, Medical Research Center in Warsaw, Poland (T.B., A.K., K.J., J.N.); Departments of Endocrinology (Y.H., T.F.) and Immunology (N.S., K.I.), Kurume University School of Medicine, Kurume, Japan; Human Molecular Genetics Laboratory of Departments of Forensic Medicine, Childhood Diabetes and Birth Defects (R.P.), and Department of Endocrinology (K.C.), Warsaw University School of Medicine, Warsaw, Poland. Present address: Department of Hypertension and Diabetology, MedicalUniversity of Gdansk, Poland (K.J.). Address reprint requests to: Dr. Tomasz Bednarczuk, Department of Endocrinology; Medical Research Center, Polish Academy of Science, Banacha 1A, 02-097 Warsaw, Poland; Phone/Fax: (48 22) 659 75 62; E-mail: bednar@amwaw.edu.pl. Received December 16, 2003; revised February 13, 2004; accepted February 19, 2004. Human Immunology 65, 632– 639 (2004) © American Society for Histocompatibility and Immunogenetics, 2004 Published by Elsevier Inc. single nucleotide polymorphism tumor necrosis factor bital fat/connective tissue that is closely associated with Graves’ hyperthyroidism. Cytokines are likely to play an important role in the initiation and propagation of the autoimmune process in the orbit [1]. The proinflammatory reactions mediated by tumor necrosis factor (TNF)-␣ include an induction of expression of adhesion molecules on endothelial cells [2]. TNF-␣ influences also the expression of a potentially important autoantigen (thyrotropin receptor) and certain immunomodulatory proteins (human leukocyte antigen [HLA] DR, ICAM-1, heat shock protein 72) on orbital fibroblasts, which are considered to be the target cell of the autoimmune attack [1, 3]. High levels of TNF-␣ were found in retrobulbar tissues samples in GO, and the enlargement of extraoc0198-8859/04/$–see front matter doi:10.1016/j.humimm.2004.02.033 TNF Gene SNPs in Graves’ Ophthalmopathy ular muscles was significantly correlated with TNF mRNA expression [4, 5]. Increased serum levels of TNF-␣ have been also reported in patients with active GO [6, 7]. Because the production of TNF-␣ has been demonstrated to be under genetic control, TNF gene may be considered as an important candidate gene contributing to the development and/or severity of GO [8]. In our previous study, we have demonstrated that TNF single nucleotide polymorphisms (SNPs) at positions -1031 (T3 C change, termed T-1031C) and -863 (C863A) were associated with ophthalmopathy in Japanese patients with Graves’ disease (GD) [9]. In the present study, we have investigated the associations of five SNPs located in the 5⬘ promoter/enhancer region at positions -1031 (T-1031C), -863 (C-863A), -857 (C-857T), -308 (G-308A), and -238 (G-238A) with GO in a Polish white population. In addition, we analyzed the frequencies of HLA-DRB1 alleles in patients with GD with and without ophthalmopathy because of the strong linkage disequilibrium within the HLA region. MATERIALS AND METHODS Subjects We studied a total of 228 randomly selected Polish patients with GD recruited from the Department of Endocrinology, Medical University of Warsaw, and 248 healthy Polish adults recruited from the Blood Transfusion Center. The diagnosis of GD was based on the presence of hyperthyroidism, diffuse goiter, detectable thyroid-stimulating hormone receptor autoantibodies (TRAK Lumitest, BRAHMS Diagnostica, Germany) and/or increased radioiodine uptake. Patients with GD were subdivided into two groups, GO or GD without GO, according to the presence of clinically evident ophthalmopathy, as previously described [10]. Group 1: GO. The GO group comprised 106 patients (81 women, 25 men) aged 14 –77 years (median, 47 years). The severity of eye changes was assessed according to the NOSPECS classification. Patients with proptosis (NOSCPECS class III), extraocular-muscle dysfunction (class IV), exposure keratitis (class V), and optic neuropathy (class VI) were considered clinically evident. Patients were categorized according to their highest ever NOSPECS class (class III, 56 patients; class IV, 41; class V, 3; and class VI, 6). The mean duration between the onset of GD and assessment for the study was 4.0 ⫾ 6.0 years. Group 2: GD without clinically evident ophthalmopathy. The GD without clinically evident ophthalmopathy group (NOSPECS class 0-II) comprised 122 patients (99 women, 23 men) aged 16 –78 years (median, 40 years). 633 The research program was approved by the local ethical committee, and all subjects provided written informed consent for genetic studies. TNF Gene Polymorphism Analysis Polymorphisms in the TNF promoter (GenBank Accession number: L11698) were identified by dot-blot hybridization with sequence-specific oligonucleotide probes, as previously described [9, 11]. HLA-DRB1 Genotyping HLA-DRB1 typing was performed by polymerase chain reaction with sequence-specific primers (PCR-SSP) with the Dynal All Set SSP DR test (Dynal Biotech, Bromborough, Wirral, UK). HLA-DRB1 alleles were identified in all patients with GD and in 125 healthy subjects, who have been previously reported [12]. Statistical Analysis The frequencies of TNF genotypes and HLA-DRB1 carriers were compared between groups by ␹2 test or Fisher’s exact probability test with a 2 ⫻ 2 contingency table. Bonferoni’s correction for multiple testing was applied. A corrected p (pc) value of ⬍0.05 was considered significant. Odds ratios (ORs) were calculated according to Woolf’s method. When one element in the 2 ⫻ 2 table was zero, the OR was calculated with the formula modified by Haldane: OR ⫽ {[(2a ⫹ 1)(2d ⫹ 1)]/[(2b ⫹ 1)(2c ⫹ 1)]} [13]. TNF haplotypes were estimated from population genotype data by PHASE version 2.02 software [14, 15]. The differences in estimated haplotype frequencies were analyzed by a case-control permutation test with 1000 iterations implemented in the PHASE software package. The Hardy-Weinberg equilibrium test was performed by Arlequin software version 2.000 (Genetics and Biometry Lab, Department of Anthropology, University of Geneva). RESULTS Association of TNF and HLA-DRB1 Alleles With GD In all studied groups, the distribution of TNF genotypes was consistent with the Hardy-Weinberg equilibrium. There were significant differences in TNF G-308A genotype distributions, with an excess of A/A and A/G genotypes in patients with GD compared with healthy controls (p ⫽ 0.0004, pc ⫽ 0.006, OR ⫽ 2.0) (Table 1). There was a tendency toward a decrease in TNF -238 A/A and A/G genotypes in GD (p ⫽ 0.018, pc ⫽ 0.27, OR ⫽ 0.4). There were no significant differences in genotype frequencies of any TNF gene polymorphism between male and female patients with GD (data not shown). 634 Bednarczuk et al. TABLE 1 Genotype frequencies of TNF promoter polymorphisms in healthy subjects and in patients with Graves’ disease (GD) with or without ophthalmopathya TNF polymorphism T-1031C C-863A C-857T G-308A G-238A Genotype C/C T/C T/T A/A C/A C/C T/T C/T C/C A/A G/A G/G A/A G/A G/G Healthy subjects (n ⫽ 248) 5 (2.0%) 76 (30.6%) 167 (67.3%) 5 (2.0%) 53 (21.4%) 190 (76.6%) 9 (3.6%) 64 (25.8%) 175 (70.6%) 4 (1.6%) 72 (29.0%) 172 (69.4%) 1 (0.4%) 22 (8.9%) 225 (90.7%) GD total (n ⫽ 228) GD without ophthalmopathy (n ⫽ 122) GD with ophthalmopathy (n ⫽ 106) 5 (2,2%) 69 (30.3%) 154 (67.5%) 2 (0.9%) 63 (27.6%) 163 (71.5%) 2 (0.9%) 61 (26.8%) 165 (72.4%) 10 (4.4%) 96 (42.1%) 122 (53.5%)b 0 (0%) 8 (3.5%) 220 (96.5%)c 4 (3.3%) 39 (32.0%) 79 (64.8%) 1 (0.8%) 33 (27.0%) 88 (72.1%) 2 (1.6%) 34 (27.9%) 86 (70.5%) 5 (4.1%) 52 (42.6%) 65 (53.3%) 0 (0%) 8 (6.6%) 114 (93.4%) 1 (0.9%) 30 (28.3%) 75 (70.8%) 1 (0.9%) 30 (28.3%) 75 (70.8%) 0 (0%) 27 (25.5%) 79 (74.5%) 5 (4.7%) 44 (41.5%) 57 (53.8%) 0 (0%) 0 (0%) 106 (100%)d,e a p values were calculated with the use of ␹2 test or Fisher’s exact probability test and corrected (pc) for the number of tests performed (n ⫽ 15). OR, odds ratio. GD versus healthy subjects: b p ⫽ 0.0004, pc ⫽ 0.006, OR ⫽ 2.0; c p ⫽ 0.018, pc ⫽ 0.27, OR ⫽ 0.4; GD with ophthalmopathy vs. GD without ophthalmopathy: dp ⫽ 0.008, pc ⫽ 0.12, OR ⫽ 0.1; GD with ophthalmopathy versus healthy subjects: ep ⫽ 0.0003, pc ⫽ 0.0045, OR ⫽ 0.05. The distribution of HLA-DRB1 carriers is listed in Table 2. A strong association for GD was seen with the DRB1*03 allele (p ⫽ 0.00001, pc ⫽ 0.0004, OR ⫽ 2.9). Because TNF G-308A is known to be in strong linkage disequilibrium with the A*01-Cw*0701-B*0801DRB1*0301-DQB1*0201 haplotype, we stratified subjects according to the DRB1*03 and -308A status (Table 3) [16]. Only the frequency of DRB1*03(⫹)/TNF308A(⫹) carriers was significantly increased in GD com- pared with healthy subjects, suggesting that any association between TNF G-308A and susceptibility to GD resulted from linkage disequilibrium with HLADRB1*03. Association of TNF and HLA-DRB1 Alleles With GO Subdividing patients with GD according to clinical eye involvement revealed that the -238A allele was absent in TABLE 2 HLA-DRB1 carrier frequencies in healthy subjects and in patients with Graves’ disease (GD) with or without ophthalmopathya HLA-DRB1 01 15 16 03 04 11 12 13 14 07 08 09 10 a Healthy subjects (n ⫽ 125) GD total (n ⫽ 228) GD without ophthalmopathy (n ⫽ 122) GD with ophthalmopathy (n ⫽ 106) 22 (17.6%) 28 (22.4%) 6 (4.8%) 26 (20.8%) 27 (21.6%) 30 (24.0%) 12 (9.6%) 38 (30.4%) 6 (4.8%) 27 (21.6%) 8 (6.4%) 2 (1.6%) 1 (0.8%) 30 (13.2%) 60 (26.3%) 20 (8.8%) 98 (43.0%)b 36 (15.8%) 62 (27.2%) 8 (3.5%)c 55 (24.1%) 12 (5.3%) 37 (16.2%) 17 (7.5%) 3 (1.3%) 3 (1.3%) 16 (13.1%) 30 (24.6%) 9 (7.4%) 52 (42.6%) 16 (13.1%) 33 (27.0%) 5 (4.1%) 35 (28.7%) 4 (3.3%) 23 (18.9%) 10 (8.2%) 1 (0.8%) 2 (1.6%) 14 (13.2%) 30 (28.3%) 11 (10.4%) 46 (43.4%) 20 (18.9%) 29 (27.4%) 3 (2.8%) 20 (18.9%)d 8 (7.5%) 14 (13.2%) 7 (6.6%) 2 (1.9%) 1 (0.9%) p values were calculated with the use of ␹2 test (with Yates’ correction where appropriate) or Fisher’s exact probability test and corrected (pc) for the number of tests performed (n ⫽ 39). OR, odds ratio. GD versus healthy subjects: b p ⫽ 0.00001, pc ⫽ 0.0004, OR ⫽ 2.9; c p ⫽ 0.033, pc ⫽ 1.0, OR ⫽ 0.3; GD with ophthalmopathy versus healthy subjects: d p ⫽ 0.044, pc ⫽ 1.0, OR ⫽ 0.5. 635 TNF Gene SNPs in Graves’ Ophthalmopathy TABLE 3 Distribution of HLA-DRB1*03 and TNF-308A carriers in healthy subjects and patients with Graves’ disease (GD) with or without ophthalmopathya OR1 GD without ophthalmopathy (n ⫽ 122) GD with ophthalmopathy (n ⫽ 106) OR2 0.5b 1.6 0.6 2.9c 56 (45.9%) 9 (7.4%) 14 (11.5%) 43 (35.2%) 52 (49.1%) 5 (4.7%) 8 (7.5%) 41 (38.7%) 1.1 0.6 0.6 1.2 Carriers HLA-DRB1*03 ⫺ ⫹ ⫺ ⫹ TNF-308A Healthy subjects (n ⫽ 125) GD total (n ⫽ 228) ⫺ ⫺ ⫹ ⫹ 79 (63.2%) 5 (4.0%) 20 (16.0%) 21 (16.8%) 108 (47.4%) 14 (6.1%) 22 (9.7%) 84 (36.8%) a Odds ratios (OR) were calculated according to Woolf’s method, comparing GD total versus healthy subjects (OR1) and GD with ophthalmopathy versus GD without ophthalmopathy (OR2). p values were calculated by ␹2 test or Fisher’s exact probability test; b p ⫽ 0.004; c p ⫽ 0.0001. patients with ophthalmopathy, and only -238G/G homozygotes were detected (GO vs. GD without ophthalmopathy: p ⫽ 0.008, pc ⫽ 0.12, OR ⫽ 0.1; GO vs. healthy subjects: p ⫽ 0.0003, pc ⫽ 0.0045, OR ⫽ 0.05) (Table 1). We have previously reported that cigarette smoking and age at onset of GD of ⬎42 years were associated with the development of GO [10]. However, T-1031C, G-868A, C-857T, and G-308A genotype frequencies remained insignificantly different between patients with GD with or without ophthalmopathy, even after stratification for cigarette smoking status, sex, and age at onset of GD (data not shown). There was also no significant association between the frequency of -1031C, -868A, -857T, and -308A alleles and the severity of eye changes in patients with GD: NOSPECS class 0-I (20%, 14%, 14%, and 25%, respectively), class II (19%, 19%, 23%, and 27%), class III (19%, 19%, 13%, and 26%), and class IV–VI (11%, 11%, 13%, and 25%). On the basis of the genotype data, we reconstructed TNF promoter haplotypes (Table 4). In all studied groups, the genetic variation in the TNF promoter could be explained by five or six haplotypes, which together accounted for ⬎99% of all haplotypes. The frequencies of individual haplotypes in healthy subjects were similar to those reported recently by Zeggini et al. [17, 18]. The distribution of TNF haplotypes differed significantly between patients with GD with or without ophthalmopathy (p ⫽ 0.002). The haplotype containing -1031C and -238A alleles and the haplotype containing only the -1031C allele were absent in patients with GO. The later haplotype was also absent in healthy controls and thus may be regarded as “specific” for Graves’ hyperthyroidism without ophthalmopathy. The distribution of HLA-DRB1 carriers did not differ significantly in patients with GD with or without ophthalmopathy (Table 2). In patients with ophthalmopathy, the frequency of HLA-DRB1*03(⫹)/TNF-308A(⫹) carriers was significantly increased compared with healthy subjects (p ⫽ 0.0002). However, the distribution of DRB1*03 and -308A carriers did not differ significantly between patients with GD with and without ophthalmopathy (Table 3). The frequency of a potentially protective HLA-DRB1*07 allele, which has been previously associated with ophthalmopathy [19 –21], tended to be lower in patients with GO compared with healthy controls (p ⫽ 0.08, OR ⫽ 0.5) (Table 2). Because TNF -238A allele has been reported to be in linkage disequilibrium with HLA-DRB1*07, we performed TABLE 4 Estimated TNF promoter haplotype frequencies in healthy subjects and in patients with Graves’ disease (GD) with or without ophthalmopathya TNF promoter haplotype ⫺1031 T T T C C C a ⫺863 ⫺857 ⫺308 ⫺238 Healthy subjects (n ⫽ 248) C C C A C C C T C C C C G G A G G G G G G G A G Sum 50.5% 16.3% 15.6% 12.0% 4.6% 0% 99% GD without ophthalmopathy (n ⫽ 122) GD with ophthalmopathy (n ⫽ 106) 39.8% 15.6% 25.4% 14.3% 3.3% 1.6% 100% 46.7% 12.7% 25.5% 15.1% 0% 0% 100% OR1 OR2 0.9 0.7 1.9 1.3 0.05 1.0 1.3 0.8 1.0 1.1 0.07 0.1 Boldface indicates the less common allele. Only haplotypes which occurred with a frequency ⬎1% in at least one studied group are shown. Odds ratios (OR) were calculated according to Woolf’s method (with Haldane’s modification where appropriate), comparing GD with ophthalmopathy versus healthy subjects (OR1) and GD with ophthalmopathy versus GD without ophthalmopathy (OR2). p values were estimated by a case-control permutation test, comparing the distribution of TNF haplotypes in GD with ophthalmopathy versus healthy subjects (p1 ⫽ 0.004) and GD with ophthalmopathy versus GD without ophthalmopathy (p2 ⫽ 0.002). 636 Bednarczuk et al. TABLE 5 Distribution of HLA-DRB1*07 and TNF-238A carriers in healthy subjects and patients with Graves’ disease (GD) with or without ophthalmopathya TNF-238A Healthy subjects (n ⫽ 125) GD without ophthalmopathy (n ⫽ 122) GD with ophthalmopathy (n ⫽ 106) OR1 OR2 ⫺ ⫺ ⫹ ⫹ 92 (73.6%) 23 (18.4%) 6 (4.8%) 4 (3.2%) 94 (77.0%) 20 (16.4%) 5 (4.1%) 3 (2.5%) 92 (86.8%) 14 (13.2%) 0 (0%) 0 (0%) 2.4b 0.7 0.1c 0.1 2.0 0.8 0.1d 0.2 Carriers HLA-DRB1*07 ⫺ ⫹ ⫺ ⫹ a Odds ratios (OR) were calculated according to Woolf’s method (with Haldane’s modification where appropriate), comparing GD with ophthalmopathy versus healthy subjects (OR1) and GD with ophthalmopathy versus GD without ophthalmopathy (OR2). p values were calculated by ␹2 test or Fisher’s exact probability test: b p ⫽ 0.01; c p ⫽ 0.02; d p ⫽ 0.04. stratification analysis in order to examine whether the association of TNF -238A with ophthalmopathy was dependent of HLA-DRB1*07 (Table 5). Although frequencies of both DRB1*07(⫹) and -238A(⫹) carriers were decreased in ophthalmopathy compared with healthy subjects and GD without ophthalmopathy, the differences were significant only with DRB1*07(-)/238A(⫹) individuals. In addition, after removing patients with GD carrying the HLA-DRB1*07 allele, the distribution of TNF promoter haplotypes remained significantly different between patients with or without GO (p ⫽ 0.012, data not shown). These results suggest that the “protective effect” of the -238A allele was independent of DRB1*07. DISCUSSION GD is a heterogeneous autoimmune disorder affecting the thyroid, eyes, and skin with varying degrees of severity [22]. Although characteristic changes in retrobulbar tissues are detectable on orbital imaging in almost all patients with GD, clinically apparent ophthalmopathy occurs in 30%–50% of patients, with severe and potentially sight-threatening forms affecting 3%–5% of patients. The reason for this variation in the clinical presentation of eye changes is unclear. Given the pathophysiological role of TNF-␣, we hypothesized that TNF gene may be an important candidate gene contributing to the development and/or severity of GO. Therefore, we studied TNF promoter SNPs at positions -1031, -863, -857, -308, and -238 with respect to the susceptibility to GD and ophthalmopathy. In patients with GD, the frequency of TNF -308A/A and A/G genotypes were significantly increased compared with healthy controls. The association of G-308A with GD, as well as with a variety of infectious and autoimmune diseases, has been already reported [23–27]. However, the functional significance of the G-308A polymorphism remains controversial [16, 28, 29]. In addition, investigations of the role of TNF G-308A are complicated by the strong linkage disequilibrium with the HLA haplotype A*01-B*0801-DRB1*0301DQA1*0501-DQB1*0201, which is known to confer susceptibility to various autoimmune diseases [16]. In order to assess the independent role of TNF G-308A polymorphism, we performed HLA-DRB1 typing, which confirmed the known association between DRB1*03 and GD [30]. The OR conferred by DRB1*03 allele (OR ⫽ 2.9) was similar to previously reported in European white populations (2.5– 4.3). The DRB1*03 allele appeared to be more strongly associated with GD than the TNF -308A allele (OR ⫽ 1.9). In addition, the stratification analysis suggested that -308A allele did not confer susceptibility to GD independently of DRB1*03, which is in accordance with the study by Hunt and colleagues [23]. Although linkage and/or association studies implicate the HLA region in susceptibility to GD, the role of HLA in modifying the disease phenotype remains unclear [30]. Recently, TNF haplotypes have been reported to be associated with specific manifestations of ulcerative colitis, scleroderma, and asthma [31–33]. Subdividing patients with GD according to the presence of clinically evident eye diseases revealed that the distribution of TNF promoter haplotypes differed significantly between patients with or without ophthalmopathy. The -238A/1031C haplotype was absent in GO, whereas the haplotype containing only the -1031C allele was “specific” for Graves’ hyperthyroidism without eye disease. Moreover, TNF -238A/A and A/G genotypes were absent in GO. The TNF-238A allele is not known to be of functional significance but is located close to a putative repressor site [34, 35]. TNF G-238A polymorphism has been associated with susceptibility to or severity of various autoimmune diseases [17, 35–39]. Similarly to our results, TNF -238A/A and A/G genotypes have been reported to be absent in patients with severe rheumatoid arthritis and were associated with a lower rate of joint damage [35, 36, 39]. Thus, G-238A or a gene in linkage disequilibrium may have an disease-modifying effect in 637 TNF Gene SNPs in Graves’ Ophthalmopathy rheumatoid arthritis and GD. Nevertheless, the clinical relevance of the G-238A polymorphism is limited because the -238A allele is rare in white populations (⬃5%), and it represents only a small contribution to overall susceptibility to autoimmune diseases. The results of HLA-DRB1 associations with GO are contradictory [30]. In white patients with GO, the frequency of HLA-DRB1*03 allele has been reported to be increased, decreased, or unchanged, compared with GD without eye disease [21, 40 – 42]. Our data revealed a nearly identical distribution of HLA-DRB1*03 and TNF -308A alleles in patients with GD with or without ophthalmopathy, suggesting that DRB1*03 does not influence the development of eye disease in patients with GD. The HLA-DRB1*07 is a protective allele for patients with GD [23, 43]. Reduced frequencies of HLADRB1*07 have been reported especially in juvenile GD and in some studies in GO [21, 44, 45]. In our study, the frequency of DRB1*07 carriers tended to be lower in GO compared with healthy subjects (p ⫽ 0.08). Although the TNF -238A allele has been reported to be in linkage disequilibrium with HLA-DRB1*07, the stratification analyses performed suggest that the association of G-238A with GO was independent of DRB1*07 [19, 20]. In our previous study, we have reported that TNF -1031C and -863A alleles were associated ophthalmopathy in Japanese patients with GD [9]. There was no relationship between G-308A and G-238A SNP’s and susceptibility to GD or GO in the Japanese population. However, the distribution of TNF alleles differed significantly in the two populations, with -308A and -238A alleles being significantly higher in whites. Therefore, this contradictory results may reflect different genetic susceptibility to GD and GO in different ethnic groups [46]. Nevertheless, both studies suggest that TNF may play a role in the development of ophthalmopathy and further studies are required to detect “true” susceptibility or protective genes within the HLA region. In conclusion, the results our study suggest that TNF G-308A is associated with susceptibility to GD (however, this association is not independent of that seen for HLA-DRB1*03); and that TNF G-238A is associated with the development of ophthalmopathy, which is independent of HLA-DRB1*07. Although our study suggests that TNF G-238A, or a gene in linkage disequilibrium, may have an disease-modifying effect in GD, a further study of sufficient power is needed to confirm this observation. ACKNOWLEDGMENTS This work was supported by the State Committee for Scientific Research, grant 4PO5B13119. REFERENCES 1. Bahn RS: Pathophysiology of Graves’ ophthalmopathy: the cycle of disease. J Clin Endocrinol Metab 88:1939, 2003. 2. Heufelder AE, Scriba PC: Characterization of adhesion receptors on cultured microvascular endothelial cells derived from the retroorbital connective tissue of patients with Graves’ ophthalmopathy. Eur J Endocrinol 134:51, 1996. 3. Valyasevi RW, Jyonouchi SC, Dutton CM, Munsakul N, Bahn RS: Effect of tumor necrosis factor-alpha, interferongamma, and transforming growth factor-beta on adipogenesis and expression of thyrotropin receptor in human orbital preadipocyte fibroblasts. J Clin Endocrinol Metab 86:903, 2001. 4. Hiromatsu Y, Yang D, Bednarczuk T, Miyake I, Nonaka K, Inoue Y: Cytokine profiles in eye muscle tissue and orbital fat tissue from patients with thyroid-associated ophthalmopathy. J Clin Endocrinol Metab 85:1194, 2000. 5. Kumar S, Bahn RS: Relative overexpression of macrophage-derived cytokines in orbital adipose tissue from patients with Graves’ ophthalmopathy. J Clin Endocrinol Metab 88:4246, 2003. 6. Krassas GE, Pontikides N, Doukidis D, Heufelder G, Heufelder AE: Serum levels of tumor necrosis factoralpha, soluble intercellular adhesion molecule-1, soluble vascular cell adhesion molecule-1, and soluble interleukin-1 receptor antagonist in patients with thyroid eye disease undergoing treatment with somatostatin analogues. Thyroid 11:1115, 2001. 7. Wakelkamp IM, Gerding MN, Van Der Meer JW, Prummel MF, Wiersinga WM: Both Th1- and Th2-derived cytokines in serum are elevated in Graves’ ophthalmopathy. Clin Exp Immunol 121:453, 2000. 8. Jacob CO, Fronek Z, Lewis GD, Koo M, Hansen JA, McDevitt HO: Heritable major histocompatability complex class II–associated differences in production of tumor necrosis factor alpha: relevance to genetic predisposition to systemic lupus erythematosus. Proc Natl Acad Sci USA 87:1233, 1990. 9. Kamizono S, Hiromatsu Y, Seki N, Bednarczuk T, Matsumoto H, Kimura A, Itoh K: A polymorphism of the 5⬘ flanking region of tumour necrosis factor alpha gene is associated with thyroid-associated ophthalmopathy in Japanese. Clin Endocrinol (Oxf) 52:759, 2000. 10. Bednarczuk T, Hiromatsu Y, Fukutani T, Jazdzewski K, Miskiewicz P, Osikowska M, Nauman J: Association of cytotoxic T-lymphocyte–associated antigen-4 (CTLA-4) gene polymorphism and non-genetic factors with Graves’ ophthalmopathy in European and Japanese populations. Eur J Endocrinol 148:13, 2003. 11. Higuchi T, Seki N, Kamizono S, Yamada A, Kimura A, Kato H, Itoh K: Polymorphism of the 5⬘-flanking region 638 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Bednarczuk et al. of the human tumor necrosis factor (TNF)-alpha gene in Japanese. Tissue Antigens 51:605, 1998. Ploski R, Ronningen KS, Thorsby E: HLA class II profile of a Polish population: frequencies of DRB1, DQA1, DQB1, and DPB1 alleles and DRB1-DQA1-DQB1 haplotypes. Transplant Proc 28:3431, 1996. Haldane JBS: The estimation and significance of the logarithm of a ratio of frequencies. Ann Hum Genet 20:309, 1956. Stephens M, Smith N, Donnelly P: A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68:978, 2001. Stephens M, Donnelly P: A comparison of Bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 73:1162, 2003. Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW: Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci USA 94:3195, 1997. Zeggini E, Thomson W, Kwiatkowski D, Richardson A, Ollier W, Donn R: British Paediatric Rheumatology Study Group. Linkage and association studies of singlenucleotide polymorphism-tagged tumor necrosis factor haplotypes in juvenile oligoarthritis. Arthritis Rheum 46:3304, 2002. Ploski R, Bednarczuk T, Hiromatsu Y: Distribution of tumor necrosis factor (TNF)-alpha haplotypes in healthy Caucasians. Arthritis Rheum (in press). Bertolaccini ML, Atsumi T, Lanchbury JS, Caliz AR, Katsumata K, Vaughan RW, Kondeatis E, Khamashta MA, Koike T, Hughes GR: Plasma tumor necrosis factor alpha levels and the -238*A promoter polymorphism in patients with antiphospholipid syndrome. Thromb Haemost 85:198, 2001. Pociot F, D’Alfonso S, Compasso S, Scorza R, Richiardi PM: Functional analysis of a new polymorphism in the human TNF alpha gene promoter. Scand J Immunol 42:501, 1995. Farid NR, Balazs C: The genetics of thyroid associated ophthalmopathy. Thyroid 8:407, 1998. Weetman AP: Graves’ disease. N Engl J Med 343:1236, 2000. Hunt PJ, Marshall SE, Weetman AP, Bunce M, Bell JI, Wass JA, Welsh KI: Histocompatibility leucocyte antigens and closely linked immunomodulatory genes in autoimmune thyroid disease. Clin Endocrinol (Oxf) 55:491, 2001. Kula D, Jurecka-Tuleja B, Gubala E, Krawczyk A, Szpak S, Jarzab M: Association of polymorphism of LTalpha and TNF genes with Graves’ disease. Folia Histochem Cytobiol 39(Suppl 2):77, 2001. McGuire W, Hill AV, Allsopp CE, Greenwood BM, Kwiatkowski D: Variation in the TNFalpha promoter region associated with susceptibility to cerebral malaria. Nature 371:508, 2001. 26. Rood MJ, van Krugten MV, Zanelli E, van der Linden MW, Keijsers V, Schreuder GM, Verduyn W, Westendorp RG, de Vries RR, Breedveld FC, Verweij CL, Huizinga TW: TNF308A and HLA-DR3 alleles contribute independently to susceptibility to systemic lupus erythematosus. Arthritis Rheum 43:129, 2000. 27. Newton J, Brown MA, Milicic A, Ackerman H, Darke C, Wilson JN, Wordsworth BP, Kwiatkowski D: The effect of HLA-DR on susceptibility to rheumatoid arthritis is influenced by the associated lymphotoxin alpha-tumor necrosis factor haplotype. Arthritis Rheum 48:90, 2003. 28. de Jong BA, Westendorp RG, Bakker AM, Huizinga TW: Polymorphisms in or near tumour necrosis factor (TNF) gene do not determine levels of endotoxin-induced TNF production. Genes Immun 3:25, 2002. 29. Knight JC, Keating BJ, Rockett KA, Kwiatkowski DP: In vivo characterization of regulatory polymorphisms by allele-specific quantification of RNA polymerase loading. Nat Genet 33:469, 2003. 30. Vaidya B, Kendall-Taylor P, Pearce SH: The genetics of autoimmune thyroid disease. J Clin Endocrinol Metab 87:5385, 2002. 31. Ahmad T, Armuzzi A, Neville M, Bunce M, Ling KL, Welsh KI, Marshall SE, Jewell DP: The contribution of human leucocyte antigen complex genes to disease phenotype in ulcerative colitis. Tissue Antigens 62:527, 2003. 32. Sato H, Lagan AL, Alexopoulou C, Vassilakis D, Ahmad T, Pantelidis P, Veeraraghavan S, Renzoni E, Denton C, Black C, Wells AU, du Bois RM, Welsh KI: The TNF863A allele strongly associates with anticentromere antibody positivity in scleroderma. Arthritis Rheum 50:558, 2004. 33. Shin HD, Park BL, Kim LH, Jung JH, Wang HJ, Kim YJ, Park HS, Hong SJ, Choi BW, Kim DJ, Park CS: Association of tumor necrosis factor polymorphisms with asthma and serum total IgE. Hum Mol Genet 13:397, 2004. 34. Fong CW, Siddiqui AH, Mark DF: Characterization of protein complexes formed on the repressor elements of the human tumor necrosis factor alpha gene. J Interferon Cytokine Res 15:887, 1995. 35. Kaijzel EL, van Krugten MV, Brinkman BM, Huizinga TW, van der Straaten T, Hazes JM, Ziegler-Heitbrock HW, Nedospasov SA, Breedveld FC, Verweij CL: Functional analysis of a human tumor necrosis factor alpha (TNFalpha) promoter polymorphism related to joint damage in rheumatoid arthritis. Mol Med 4:724, 1998. 36. Brinkman BM, Huizinga TW, Kurban SS, van der Velde EA, Schreuder GM, Hazes JM, Breedveld FC, Verweij CL: Tumour necrosis factor alpha gene polymorphisms in rheumatoid arthritis: association with susceptibility to, or severity of, disease? Br J Rheumatol 36:516, 1997. 37. Huizinga TW, Westendorp RG, Bollen EL, Keijsers V, Brinkman BM, Langermans JA, Breedveld FC, Verweij CL, van de Gaer L, Dams L, Crusius JB, Garcia-Gonzalez TNF Gene SNPs in Graves’ Ophthalmopathy 38. 39. 40. 41. A, van Oosten BW, Polman CH, Pena AS: TNF alpha promoter polymorphisms, production and susceptibility to multiple sclerosis in different groups of patients. J Neuroimmunol 72:149, 1997. Zuniga J, Vargas-Alarcon G, Hernandez-Pacheco G, Portal-Celhay C, Yamamoto-Furusho JK, Granados J: Tumor necrosis factor-alpha promoter polymorphisms in Mexican patients with systemic lupus erythematosus (SLE). Genes Immun 2:363, 2001. Fabris M, Di PE, D’Elia A, Damante G, Sinigaglia L, Ferraccioli G: Tumor necrosis factor-alpha gene polymorphism in severe and mild-moderate rheumatoid arthritis. J Rheumatol 29:29, 2002. Buzzetti R, Nistico L, Signore A, Cascino I: CTLA-4 and HLA gene susceptibility to thyroid-associated orbitopathy. Lancet 354:1824, 1999. Villanueva R, Inzerillo AM, Tomer Y, Barbesino G, Meltzer M, Concepcion ES, Greenberg DA, MacLaren N, Sun ZS, Zhang DM, Tucci S, Davies TF: Limited genetic susceptibility to severe Graves’ ophthalmopathy: no role for CTLA-4 but evidence for an environmental etiology. Thyroid 10:791, 2000. 639 42. Allahabadia A, Heward JM, Nithiyananthan R, Gibson SM, Reuser TT, Dodson PM, Franklyn JA, Gough SC: MHC class II region, CTLA4 gene, and ophthalmopathy in patients with Graves’ disease. Lancet 358:984, 2001. 43. Marga M, Denisova A, Sochnev A, Pirags V, Farid NR: Two HLA DRB 1 alleles confer independent genetic susceptibility to Graves disease: relevance of cross-population studies. Am J Med Genet 102:188, 2001. 44. Lavard L, Madsen HO, Perrild H, Jacobsen BB, Svejgaard A: HLA class II associations in juvenile Graves’ disease: indication of a strong protective role of the DRB1*0701,DQA1*0201 haplotype. Tissue Antigens 50:639, 1997. 45. Badenhoop K, Schwarz G, Schleusener H, Weetman AP, Recks S, Peters H, Bottazzo GF, Usadel KH: Tumor necrosis factor beta gene polymorphisms in Graves’ disease. J Clin Endocrinol Metab 74:287, 1992. 46. Villanueva R, Tomer Y, Greenberg DA, Mao C, Concepcion ES, Tucci S, Estilo G, Davies TF: Autoimmune thyroid disease susceptibility loci in a large Chinese family. Clin Endocrinol (Oxf) 56:45, 2002.