Anti-Caspase-8 Autoantibody Response in Silicosis Patients is Associated with HLA-DRB1, DQB1 and DPB1 Alleles

Transcription

1 J Occup Health 2005; 47: Journal of Occupational Health Anti-Caspase-8 Autoantibody Response in Silicosis Patients is Associated with HLA-DRB1, DQB1 and DPB1 Alleles Ayako UEKI 1, Yumika ISOZAKI 2 and Masayasu KUSAKA 3 1 Kawasaki University of Medical Welfare, 2 Department of Hygiene, Kawasaki Medical School and 3 Department of Medicine, Kusaka Hospital, Japan Abstract: Anti-Caspase-8 Autoantibody Response in Silicosis Patients is Associated with HLA-DRB1, DQB1 and DPB1 Alleles: Ayako UEKI, et al. Kawasaki University of Medical Welfare We reported previously the autoantibodies directed to caspase-8 among patients with silicosis, systemic sclerosis (SSc) and systemic lupus erythematosus (SLE), and in healthy individuals. In this study, we analyzed the correlation between anti-caspase-8 autoantibody responses and HLA class II alleles in silicosis patients. The frequencies of HLA-DRB1*0406 were significantly higher in antibody positive patients (16.67%) than in control individuals (3.03%, p=0.0006). The lysine (K) at position 71 as in DRB1*0406 has been reported to be associated with rheumatoid arthritis (RA) and insulin dependent diabetes mellitus (IDDM). The haplotype HLA-DR4; DQB1*0302 was detected in 4 of 12 antibody positive patients. RA, IDDM, or pemphygus vulgaris link to the haplotype. The frequencies of DQB1*0401 were significantly lower in antibody positive patients (0%) than that in controls (13.33%, p=0.0390). The aspartic acid at position 57 in the DQB1 molecule as in DQB1*0401 is reported to play a role in the resistance to IDDM. The frequency of DPB1*0601 in antibody positive patients (5.88%) was significantly higher than that in controls (0.56%, p=0.0003). DPB1*0601 is reported to be a risk factor among RA patients, and glutamate at position 69 of the DPB1 molecule may be involved. Repeated and continuous screening of autoantibodies seems to be necessary among workers in contact with Si-related substances for the detection of immunological disorders in the early stage. (J Occup Health 2005; 47: 61 67) Key words: Silicosis, Autoantibodies, Caspase-8, HLA class II alleles Received July 16, 2004; Accepted Nov 13, 2004 Correspondence to: A. Ueki, Kawasaki University of Medical Welfare, 288 Matsushima, Kurashiki, Okayama , Japan ( aueki@mw.kawasaki-m.ac.jp) We reported previously the antibodies directed to caspase-8 and its intramolecular epitope spreading among the patients with silicosis, systemic sclerosis (SSc) and lupus erhtematosus (SLE), as well as in healthy individuals 1). We have studies the pathogenesis of immunological disorders, especially in silicosis patients 2, 3). Dysregulation of apoptosis through the Fas-Fas ligand pathway is considered to be relevant in autoimmune disease onset, and we reported previously the elevated serum levels of soluble Fas (sfas) in patients suffering from silicosis, SSc or SLE, and proposed that the suppression of apoptosis is involved in the pathogenesis of these diseases 4). In healthy individuals, caspase-8 is recruited into the death-inducing signal complex, a multiprotein complex formed on the Fas/CD95 receptor 5). Patients with inherited mutations in CD95 demonstrate the defects in caspase activation that causes them to suffer from autoimmune lymphoproliferative syndrome (ALPS) 6, 7). Based on this evidence, it is of great interest that the autoantibodies directed to caspase-8 were detected, whereas the biological function remains unknown. In this study, we analyzed the correlation between anticaspase-8 autoantibody response and HLA class II alleles in silicosis patients using a PCR restriction fragment length polymorphism (PCR-RFLP) method in order to establish whether specific alleles might in part explain disease susceptibility, since autoimmune responses such as the production of autoantibodies against pancreas islets in insulin dependent diabetes mellitus (IDDM ) 8), rheumatoid factor production 9) or susceptibility to rheumatoid arthritis (RA) 10), anti-scl-70 (topoisomerase I) antibody production in SSc 11), or SSA production in Sjogren s syndrome 12) have been reported to be associated with specific HLA class II alleles. We now report the association of anti-caspase-8 antibody response with HLA class II alleles.

2 62 J Occup Health, Vol. 47, 2005 Materials and Methods Patients After receiving informed consent, 61 silicosis patients (56 men and 5 women, average age: 67.6 ± 7.45) were analyzed for anti-caspase-8 autoantobodies using the methods of Western blotting, SELDI (desorption/ ionization) proteinchip analysis and the SPOTs system, as reported previously 1). Eighteen patients among 61 were considered positive. None of the patients showed clinical symptoms of autoimmune disease such as sclerotic skin, Raynaud s phenomenon, facial erythema or arthralgia. All patients were Asian Japanese and the severity of lung fibrosis varied from profusion rate (PR) 2 to 4. HLA typing of HLA-DRB1, DQB1 and DPB1 alleles was performed among anti-caspase-8 positive patients and other negative patients only when the informed consent was obtained (the number of samples tested was somewhat different from the allele to allele). Allele frequencies in Asian Japanese by Akaza et al. 13) were used as the control. HLA class II typing Genomic DNA was extracted from 5 ml of peripheral blood using a genomic DNA extraction kit (Takara Biomedicals, Ohtsu, Japan). HLA-DRB1, DQB1 and DPB1 typing was performed by polymerase chain reaction-restriction fragment length polymorphism (PCR- RFLP) method, using specific primers. For HLA-DRB1 typing, the DRB1 genes were amplified using the group specific 5 primers along with the common 3 primer, according to the method of Ota et al. 14) (Table 1) The PCR was carried out for 30 cycles with a DNA Thermal Cycler (Perkin-Elmer, CT, USA) using a thermostable DNA polymerase (Takara Biomedicals, Ohtsu, Japan) under the following conditions: denaturation for 1 min at 94 C, annealing for 1 min at 62 C, extension for 2 min at 72 C. After amplification, aliquots of the reaction mixtures were digested with restriction endonucleases, AvaII, PstI, FokI, Cfrl3I, HphI, Hinf I, MnlI, KpnI, HaeII, SfaNI, SacII, BsaJI, ApaI, and RsaI. Specificities were assigned to RFLP bands, and compared with HLA-D haplotypes determined by the World Health Organization Nomenclature Committee for factors of the HLA System 15). For HLA-DQB1 typing, the specific primers were used according to the method of Nomura et al. 16) (Table 1). The PCR condition was the same as in the case of DRB1 typing. Fok I, ApaI, Hae II, SfaN I, BssH II, Hph I, Bgl I, Sac I, Acy I and Hpa II were used for the digestion of genomic DNA. For HLA-DPB1 typing, the specific primers were used according to the method of Ota et al. 17) (Table 1) The PCR condition was the same as those mentioned above. For RFLP analysis, Bsp 1286 I, Fok I, Dde I, BsaJ I, BssH II, Cfr 13 I, Rsa I, EcoN I, Ava II, Fok I and Rsa I were used for the digestion of genomic DNA. Statistical analysis Comparison of allelic frequencies among anti-caspase- 8 positive versus anti-caspase-8 negative silicosis patients or the frequency in Asian Japanese was performed using the chi square analysis (without Yates correction). In cases where the sample group contained 5 or fewer members, Fisher s exact test was used. The frequencies of DRB1, DQB1 and DPB1 alleles in Asian Japanese were based on the data by Akaza et al. 13) Table 1. Primers used for HLA typing Gene Sequences (5 to 3 ) DRB1 DQB1 DPB1 for DR 1 5 primer GGTTGCTGGAAAGATGCATCT for DR 2 5 primer TTCCTGTGGCAGCCTAAGAGG for DR 3,5,6,8 5 primer ACGTTTCTTGGAGTACTCTACG for DR 4 5 primer GTTTCTTGGAGCAGGTTAAAC for DR 7 5 primer AGTTCCTGGAAAGACTCTTCT for DR 9 5 primer GAAGCAGGATAAGTTTGAGTG for DR 10 5 primer GGTTGCTGGAAAGACGCGTCC common for DRB1 3 primer CCGCTGCACTGTGAAGCTCT for DQ 1 5 primer GCATGTGCTACTTCACCAACG 3 primer CACCTGCAGATCCCGCGGTACGCCACCTC for DQ 2,3,4 5 primer GCATGTGCTACTTCACCAACG 3 primer CACCTGCAGTGCGGAGCTCCAACTGGTA common for DPB1 5 primer GTGAAGCTTTCCCCGCAGAGAATTAC common for DPB1 3 primer CACCTGCAGTCACTCACCTCGGCGCTG

3 Ayako UEKI, et al.: Anti-Caspase-8 Autoantibody and HLA Class II 63 Table 2. Distributions of HLA class II alleles in patients with anti-caspase 8 autoantibodies detected among 61 silicosis patients No. DRB1 DQB1 DPB NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT: not tested, The sample blood could not be obtained because of the death of the patients or for some other reason. obtained from the samples of more than 1,000 Japanese volunteers. Results HLA class II alleles among silicosis patients with anticaspase-8 autoantibodies The results of HLA typing among silicosis patients with anti-caspase-8 autoantibodies (18 cases) are summarized in Table 2. Association of anti-caspase-8 autoantibodies with HLA- DRB1 The distribution of DRB1*0406 was significantly higher in anti-caspase-8 positive silicosis patients (16.67%) than that in Japanese controls (3.03%, p=0.0006). The frequency of DRB1*0406 in positive patients was slightly higher than that in negative patients (p=0.0933). DRB1*0803 was also frequently detected in anti-caspase-8 positive patients (20.83%), but without a significant difference compared with that in Japanese controls (8.29%, p=0.0555) (Table 3). From these results, a significant correlation was observed between HLA- DRB1 *0406 allele and anti-caspase-8 autoantibody response among silicosis patients. Association of anti-caspase-8 autoantibodies with HLA- DQB1 As seen in Table 4, the frequency of DQB1*0302 was slightly higher in anti-caspase-8 positive silicosis patients (21.88%) than that in Japanese controls (9.92%, p=0.0549). In contrast, the frequency of DQB1*0401 was significantly lower in caspase-8 positive silicosis patients (0%) than that in Japanese controls (13.33%, p=0.0390). The distribution of DQB1*0401 in antibody positive patients was also slightly lower than that in antibody negative patients (9.09%, p=0.0963). DQB1*0401 seems to suppress the induction of anticaspase-8 autoantibodies. These results coincide with the report that the aspartic acid at position 57 in the DQB1 molecule, as possessed in DQB1*0401, plays an important role in the resistance to IDDM. Association of anti-caspase-8 autoantibodies with HLA DPB1 DPB1*0601 was detected in 5.88% of anti-caspase-8 positive patients, and the frequency was significantly higher than that in Japanese controls (0.56%, p=0.0003), but without statistical significance compared with that in antibody negative patients (0%, p= ). The distribution of DPB1*1601 (5.88%), *1801 (5.88%) and *1901 (5.88%) in anti-caspase-8 positive patients was significantly higher than that in control individuals (0.56%, p=0.0210; 0.35%, p=0.0102; 0.56%, p=0.0210, respectively). These alleles were also frequently distributed with a significant difference in anti-caspase- 8 negative patients (4.55%, p=0.0002; 4.55%, p=0.0030;

4 64 J Occup Health, Vol. 47, 2005 Table 3. Frequencies of HLA-DRB1 alleles among anti-caspase 8 positive or negative Asian Japanese silicosis patients and controls Associated HLA Frequencies in Silicosis Patients Frequencies in Class II alleles Caspase-8 (+) Caspase-8 ( ) Controls a (n=24) (n=18) (n=2432) DRB1* (0%) 1 (5.56%) 141 (5.81%) DRB1* (4.17%) 2 (11.11%) 51 (2.08%) DRB1* (4.17%) 1 (5.56%) 322 (13.26%) DRB1* (16.67%)*** 0 (0%) 74 (3.03%) DRB1* (0%) 1 (5.56%) 102 (4.18%) DRB1* (20.83%) 2 (11.11%) 202 (8.29%) DRB1* (16.67%) 5 (27.78%) 342 (14.08%) DRB1* (4.17%) 0 (0%) 63 (2.59%) DRB1* (4.17%) 2 (11.11%) 89 (3.65%) DRB1* (4.17%) 0 (0%) 43 (1.75%) DRB1* (0%) 1 (5.56%) 166 (6.83%) DRB1* (4.17%) 1 (5.56%) 82 (3.37%) DRB1* (4.17%) 0 (0%) 46 (1.91%) DRB1* (8.33%) 0 (0%) 173 (7.11%) DRB1* (8.33%) 1 (5.56%) 246 (10.13%) DRB1* (0%) 1 (5.56%) 0 (0%) other alleles 290 (11.92%) a: Japanese controls according to Akaza 13), ***p<0.001 compared with Controls Table 4. Frequencies of HLA-DQB1 alleles among anti-caspase 8 positive or negative Asian Japanese silicosis patients and controls Associated HLA Frequencies in Silicosis Patients Frequencies in Class II alleles Caspase-8 (+) Caspase-8 ( ) Controls a (n=32) (n=22) (n=2550) DQB1* (0%) 0 (0%) 11 (0.43%) DQB1* (12.50%) 1 (4.55%) 283 (11.10%) DQB1* (21.88%) 3 (13.63%) 253 ( 9.92%) DQB1* (18.75%) 5 (22.72%) 379 (14.86%) DQB1* (0%) 0 (0%) 1 (0.04%) DQB1* (0%)* 2 (9.09%) 340 (13.33%) DQB1* (9.38%) 2 (9.09%) 98 ( 3.84%) DQB1* (3.13%) 1 (4.55%) 163 (6.39%) DQB1* (0%) 0 (0%) 65 (2.55%) DQB1* (6.25%) 2 (9.09%) 107 (4.20%) DQB1* (21.88%) 3 (13.63%) 448 (17.57%) DQB1*0602= (6.25%) 2 (9.09%) 179 (7.02%) DQB1*0604= (0%) 1 (4.55%) 186 (7.29%) other alleles 37 (1.45%) a: Japanese controls according to Akaza 13), *p<0.05 compared with Controls

5 Ayako UEKI, et al.: Anti-Caspase-8 Autoantibody and HLA Class II 65 Table 5. Frequencies of HLA-DPB1 alleles among anti-caspase 8 positive or negative Asian Japanese silicosis patients and controls Associated HLA Frequencies in Silicosis Patients Frequencies in Class II alleles Caspase-8 (+) Caspase-8 ( ) Controls a (n=34) (n=22) (n=1430) DPB1* (26.47%) 6 (27.27%) 301 (21.05%) DPB1* (0%) 1 (4.55%) 50 (3.50%) DPB1* (0%) 1 (4.55%) 50 (3.50%) DPB1* (0%) 1 (4.55%) 52 (3.57%) DPB1* (14.71%) 3 (13.64%) 158 (11.05%) DPB1* (26.47%) 6 (27.27%) 547 (38.25%) DPB1* ( 5.88%)*** 0 (0%) 8 ( 0.56%) DPB1* (0%) 0 (0%) 9 (0.63%) DPB1* (5.88%) 0 (0%) 133 (9.30%) DPB1* (0%) 0 (0%) 4 (0.28%) DPB1* (2.94%) 1 (4.55%) 29 (2.03%) DPB1* (0%) 0 (0%) 25 (1.75%) DPB1* (5.88%)** 1 ( 4.55%)** 3 (0.21%) DPB1* (5.88%)** 1 (4.55%)** 5 (0.35%) DPB1* (5.88%)** 1 (4.55%)* 8 (0.56%) other alleles 48 (3.36%) a: Japanese controls according to Akaza 13), ***p<0.001, **p<0.01, *p<0.05 compared with Controls 4.55%, p= respectively) (Table 5), compared with those in control individuals but without significant differences between antibody negative patients, which suggests the association of these alleles only with the progression of silicosis instead of anti-caspase-8 autoantibody response. From these results, only the DPB1*0601 allele was considered to be associated with anti-caspase-8 autoantibody response in silicosis patients. Discussion It is noticeable in this study, that the distribution of HLA-DRB1*0406 was significantly higher in anticaspase-8 positive silicosis patients than in healthy controls. Other autoimmune diseases such as RA 18, 19) or IDDM 20) are reported to link to DR4. The predominant role of the (Q)R/KRAA sequence at position of the DRB1 chain, in particular the lysine (K) substitution at position 71 as in DRB1*0406, has been reported to be highly associated with RA susceptibility 9, 10). It seems that the antigen presentation of caspase-8 through HLA- DRB1 takes the same process as in the case of RA or IDDM. However, these amino acid sequences are not always indispensable for the progression of autoimmunity, but other HLA-DR alleles are reported as a risk factor for other autoimmune diseases such as the assosiation of DRB1*03 with primary Sjogren s syndrome 11, 12). The most abundant DQB1 sequences in healthy individuals have an aspartic acid at position 57 which forms a salt bridge across the end of the peptide-binding cleft of the DQ molecule, and such a conformation in the DQ molecule is associated with resistance to IDDM 21). In this study DQB1*0401 with the aspartic acid at position 57 in the DQB1 molecule was not detected among anticaspase-8 positive silicosis patients, which also suggests the association of the aspartic acid at position 57 with resistance to anti-caspase-8 autoimmune response. Diabetic patients (studied among Caucasian populations) mostly possess valine, serine or alanine at position 57 in the DQ B1 molecule instead of aspartic acid, and therefore the DQ molecule lacks the salt bridge 21). In anti-caspase- 8 positive silicosis patients the relative high frequency of the DQB1*0302 allele, which possess alanine at position 57 in the DQB1 molecule, compared to that in the control (p=0.0549) might be acceptable. Interestingly, the other group of autoantibody responders, however, such as anti-topoisomerase I (antitopo I) positive SLE, SSc and silicosis patients demonstrated a completely different association with HLA-DQB1 alleles as we and other authors reported previously 3). Anti-topo I responders among SLE, SSc or silicosis patients significantly associate with DQB1*0301, *0601 or *0402, the alleles of which possess aspartic acid at position 57 without resistance to autoantibody response. This evidence constitutes a paradox. It is speculated that not only one but several positions of the

6 66 J Occup Health, Vol. 47, 2005 DQB1 molecule play the role independently in antigen presentation according to the nature of the antigen, and that aspartic acid at position 57 in the DQB1 molecule is one of them. In this study, we observed a significantly increased frequency of the DPB1*0601 allele among anti-caspase- 8 positive silicosis patients. Rihs et al. 22) reported the enhanced frequency of DPB1*0601 or *1701 alleles among anti-scl-70 positive systemic sclerosis patients, and mentioned that glutamate at position 69 of the DPB1 molecule might be involved in susceptibility to antibody expression. Seidl et al. 10) described HLA-DPB1*0401 and *0601 alleles as risk factors among RA patients. This suggests that antigen presentation of the caspase-8 molecule takes also the similar process by HLA-DPB1 as in the case of RA. It is of interest that the autoantibody response against caspase-8 in silicosis patients seems to take the similar process as in the cases of IDDM or RA, but not as in anti-topo I autoantibody response, although both autoantibodies could be detected among silicosis patients. The correlation between autoimmune responses and exposure to silica or silicate has long been studied 23, 24). We reported the findings of polyclonal activation of human T cells by silicate in vitro 2), elevated soluble Fas (sfas) levels in silicosis patients 4), and dominantly expressed sfas mrna in silicosis patients 25). We speculate that the repeated activation of helper T cells induces the elevated levels of sfas in silicosis patients by unknown mechanisms, and that such a suppressed condition of activation-induced cell death in T cells, including self-reactive clones, could play a role in inducing autoantibodies among workers after exposure to silica dust. Caspase-8 is recruited into the deathinducing signal complex, a multiprotein complex formed on the cytoplasmic protein of the Fas/AOP-1/CD95 receptor 26). Patients with inherited mutations in CD95 demonstrate the resultant defects in caspase activation that cause them to suffer from autoimmune lymphoproliferative syndrome (ALPS) 27, 28). The pathological roles of anti-caspase-8 autoantibodies in the occurrence of immune disorders are unclear, but it is speculated that these autoantibodies could enter the cells through the cell membrane as soon as the cellular permeability increased with an apoptotic signal and interact with autoantigens such as caspase-8, but, details remain to be clarified. It seems that the repeated and continuous screening of autoantibodies is necessary for the workers who are in contact with Si-related substances such as silica, silicone or asbestos fibers for the detection of immunological disorders in the early stage. References 1) A Ueki, Y Isozaki, A Tomokuni, T Hatayama, H Ueki, M Kusaka, M Shiwa, H Arikuni, T Takeshita and K Morimoto: Intramolecular epitope spreading among anti-caspase-8 autoantibodies in patients with silicosis, systemic sclerosis and systemic lupus erythematosus, as well as in healthy individuals. Clin Exp Immunol 129, (2002) 2) A Ueki, M Yamaguchi, H Ueki, Y Watanabe, G Ohsawa, K Kinugawa, Y Kawakami and F Hyodoh: Polyclonal human T-cell activation by silicate in vitro. Immunology 82, (1994) 3) A Ueki, Y Isozaki, A Tomokuni, H Ueki, M Kusaka, S Tanaka, T Otsuki, H Sakaguchi and F Hyodoh: Different distribution of HLA Class II alleles in antitopoisomerase I autoantibody responders between silicosis and systemic sclerosis patients, with a common distinct amino acid sequence in the HLA-DQB1 domain. Immunobiology 204, (2001) 4) A Tomokuni, T Aikoh, T Matsuki, Y Isozaki, T Otsuki, S Kita, H Ueki, M Kusaka, T Kishimoto and A Ueki: Elevated soluble fas/apo-1 (CD95) levels in silicosis patients without clinical symptoms of autoimmune diseases or malignant tumors. Clin Exp Immunol 110, (1997) 5) JP Medema, C Scaffidi, FC Kischkel, A Shevchenko, M Mann, PH Krammer and ME Peter: FLICE is activated by association with the CD95 death-inducing signaling complex (DISC). EMBO J 81, (1997) 6) GH Fisher, FJ Rosenberg, SE Straus, S Gerwinski, T martin, C Berljung, A Gruneiro, SJ Chou, EH Man and S Pappatheodorou: Dominant interfering Fas gene mutations impair apoptosis in a human autoimmune lymphoproliferative syndrome. Cell 81, ) F Rieuz-Iaucat, F Le Deist, C Hivroz, IA Roberts, KM Debatin, A Fisher and JP de Villartay: Mutation s in Fas associated with human lymphoproliferative syndrome and autoimmunity. Science 268, ) KM Gillespie, EAM Gale and PJ Bingley: High familial risk and genetic susceptibility in early onset childhood diabetes. Diabetes 51, (2002) 9) A Perdriger, P Guggenbuhl, G Chales, P Le Dantec, J Yaouanq, B Genetet and Y Pawlotsky: The role of HLA-DR-DR and HLA-DR-DP interactions in genetic susceptibility to rheumatoid arthritis. Human Immunol 46, (1996) 10) C Seidl, U Koch, G Bruennler, T Buhleier, R Frank, B Moller, E Markert and G Koller-Wagner: HLA-DR/ DQ/DP interactions in rheumatoid arthritis. Eur J Immunogen 24, (1997) 11) JD Reveille and FC Amett: The immunogenetics of Sjogren s syndrome. Rheum Dis Clin North Am 18, (1992) 12) HI Kang, HM Fei, I Saito I, S Sawada, SL Chen, D Yi, E Chan, C Peebles, TL Bugawan and HA Erlich: Comparison of HLA class II genes in Caucasoid,

7 Ayako UEKI, et al.: Anti-Caspase-8 Autoantibody and HLA Class II 67 Chinese, and Japanese patients with primary Sjogren s syndrome. J Immunol 150, (1993) 13) Akaza T. Frequency of the HLA allele in Japanese. In: Nakajima I, Takahashi T, Yoshikai Y, eds. Simple Immunology. Tokyo: Nankado, 2001: 211 (in Japanese). 14) M Ota, T Seki and H Fukushima: HLA-DRB1 genotyping by modified PCR-RFLP method combined with group-specific primers. Tissue Antigens 39, (1992) 15) Cohen D, Simons M, Lalouel M. Nomenclature for HLA-RFLP. In: Dupont B, ed. Immunobiology of HLA. vol 1. Histocompatibility Testing New York: Springer-Verlag, 1989: ) N Nomura, M Ota, K Tsuji and H Inoko: HLA-DQB1 genotyping by mdified PCR-RFLP method combined with allele-specific primer. Tissue Antigens 38, (1991) 17) M Ota, T Seki, N Nomura, K Sugimura, N Mizuki, H Fukushima, K Tsuji and H Inoko: Modified PCR-RFLP method for HLA-DPB1 and DQA1 genotyping. Tissue Antigens 38, (1991) 18) MA Gonzalez-Gay, AH Hajeer and C Garcia-Porrua: Patients chosen for treatment with cyclosporine bacause of severe rheumatoid arthritis are more likely to carry HLA-DRB1 shared epitope alleles, and have earlier disease onset. J Rheumatol 29, (2002) 19) DL Mattey, PT Dawes and S Clarke: Relationship among the HLA-DRB1 shared epirope, smoking, and rheumatoid factor production in rheumatoid arthritis. Arthritis & Rheum 47, (2002) 20) J Setterfield, J Theron and RW Vaughan: Mucous membrane pemphigoid: HLADQB1:0301 is associated with all clinical sites of involvement and may be linked to antibasement membrane IgG production. Brit J Derm 145, (2002) 21) Mathis D, Goodnow CC, Ravetch J. Autoimmunity and Transplantation. In: Janeway CA, Travers P, Walport M, Shlomchik MJ, eds. Immunobiology 5th. New York: Garland Publishing, 2001: ) HP Rihs, K Conrad and J Mehlhorn: Molecular analysis of HLA-DPB1 alleles in idiopathic systemic sclerosis patients and uranium miners with systemic sclerosis. Int Arch Allergy Immunol 109, (1996) 23) UF Haustein, V Ziegler, K Herrmann, J Mehlhorn and C Schmidt: Silica-induced scleroderma. J Am Acad dermatol 22, (1990) 24) J Snachez-Roman, I Wichmann, J Salaberri, JM Varela and A Nunez-Roldan: Multiple clinical and biological autoimmune manifestations in 50 workers after occupational exposure to silica. Ann Rheum Dis 52, (1993) 25) T Otsuki, H Sakaguchi, A Tomokuni, T Aikoh, T Matsuki, Y Kawakami, M Kusaka, H Ueki, S Kita and A Ueki: Soluble Fas mrna is dominantly expressed in cases with silicosis. Immunology 94, (1998) 26) JP Medema, C Scaffidi, FC Kischkel, A Shevchenko, M Mann, PH Krammer and ME Peter: FLICE is activated by association with the CD95 death-inducing signaling complex(disc). EMBO J 16, (1997) 27) GH Fisher, FJ Rosenberg, SE Straus, S Gerwinski, T Martin, C Berglung, A Gruneiro, SJ Chou, EH Man and S Pappatheodorou: Dominant interfiring Fas gene Mutations impair apoptosis in a human autoimmune lymphoproliferative syndrome. Cell 81, ) F Rieux-Laucat, F Le Deist, C Hivroz, IA Roberts, KM Debatin, A Fisher and JP de Villartay: Mutations in Fas associated with human lymphoproliferative syndrome and autoimmunity. Science 268,