DNA polymorphsm of HLA class I1 genes in pauciarticular juvenile rheumatoid arthritis N. Morling, J . Friis, L. Fugger, J. Georgsen, C . Heilmann, F. K. Pedersen, N. Pldum, A. Svejgaard. DNA polymorphism of HLA class I1 genes in pauciarticular juvenile rheumatoid arthritis. Tissue Antigens 1991: 38: 16-23.
Abstract: We investigated the DNA restriction fragment length polymorphism (KFLP) of the major histocompatibility complex (MHC) class I1 genes: HLA-DRB, -DQA, -DQB, DPA, and -DPB in 54 patients with pauciarticular juvenile rheumatoid arthritis (PJRA) and in healthy Danes. The frequencies of DNA fragments associated with the following HLA class I1 genes were increased in PJRA when compared to normal controls: DRBI*08 (DRw8) (35.2% vs 10.3%, R R = 4.6, p < DRB3*01/02/ ,03 (DKw52) (76.3% vs 48.1%, R R 3.5, p < 10-3), DQAI*0401 (41.0% vs DQAI'0501 (55.6% vs 29.7%, R R = 3.0, 7.4%, RR = 7.9, p < p < 10--2,DQBI*0301 (DQw7) (46.2% vs 17.5%, R R = 4.0; p < DPA1+0201 (44.2% vs 7.9%, R R = 8.7, p < and DPB1*02 (DPw2) The frequency of DRBI*II was (40.7% vs 7.1%, RK = 8.5, p < not significantly increased. The frequencies of DNA fragments associated with the following HLA class I1 genes were decreased in PJRA although not statistically significantly so after 'correction' of p values: DRB1*04 (14.8% vs 40.2%, RR = 0.27; p < DRB1*07 (0'70 vs 25.9%, RR = 0.04, p < lo-'), DRB4*0101 (DRw53) (25.9% vs 53.6%, R R = 0.31, and p < lo-'), DQA1*0102 (11.6% vs 36.0%, R R = 0.25, p < Individuals who carDQA1*0201 (2.6% vs 34.2%, RR = 0.05, p < ried certain combinations of two of the risk factors seemed to have a further increased risk of developing PJRA compared to individuals carrying only one of the risk factors. The combinations of HLA class 11 genes primarily involved were: DRBI*08 and DQAI*O5OI, DRBI*08 and DQBI*0301, DRBI*08 and DPA 1*0201, DQAP0401 and DQBI*0301, DQA1*0401 and DPA1*0201, l?QAI*OSOI and DPAP0201, and DQAI*0501 and DPBI*02, although other combinations also gave odds ratios which were non-significantly increased compared to the odds ratios of the individual genetic markers. The data suggest that HLA class I1 gene products coded by genes at different loci interact in the susceptibility of 1 PJRA.
Introduction
The human major histocompatibility complex (MHC), the H L A system, encodes at least three polymorphic series of class I1 transmembraneous glycoproteins. The HLA class I1 molecules are designated DR, DQ, and DP and each consists of an a and a fl glycoprotein chain (1). A great number of diseases are associated with HLA-DR/DQ antigens (e.g. (2)), while a limited number of diseases have been reported to be associated with HLA-DP antigens: pauciarticular juvenile rheumatoid arthritis (PJRA) (3, 4), aplastic anemia (5), celiac disease (6, 7), leukemia (S), and alopecia areata (9). The mechanism by which variants of the HLA
Nials M~rling'~',Johannet Friig, Lars Fuggrr', dirgen 6eergsen', Cartten Heilmann', Fnddy Karup PeIsrsrI', Niels Bdum' and Arne Svsjgaard' 'Tissue Typing Laboratory, 'Hornbrek Hospital, 3Paediatric Department, Slate University Hospital (Rigshospitalet), 'Institute of Forensic Genetics, University of Copenhagen, Denmark
Key words: DNA polymorphism - class 1 IPJRA Received 22 February, revised, accepted for publication 29 March 1991
class I1 molecules confer susceptibility to different diseases is most probably related to the function of these molecules in the immune response. Thus, class I1 molecules are involved in the selection in the thymus of the T-cell repertoire (lo), and they bind antigenic peptide and present it to the antigen reactive CD4-positive T cells (1 1). The HLA class I1 molecules thus act like restriction elements in the presentation of antigens to T lymphocytes and thereby may confer susceptibility to various, in particular autoimmune, diseases. Pauciarticular juvenile rheumatoid arthritis is associated with HLA-B27 (e.g. IZ), DRw8/DwS, HLA-DRS/DwS, DRw6/Dw6 (e.g. 13-19), and DPw2 (3, 4). l h e combined presence of (i) DPw2
HLA class II genes in PJRA and (ii) Dw5 and/or Dw8 is associated with a significantly higher risk of PJRA than each antigen alone, indicating interaction of DP and DR gene products (4). Since the initiation of the present work, van Kerckhove et al. (20) have reported on increased frequencies in PJRA of the HLADRBI*80I, DQAI*401, DQBI*402 genes, and Begowitch et al. (21), van Kerckhove et al. (22), and we (23) have reported increased frequency of the DPBI*0201 allele which codes for DPP-chains of the majority of the DPw2 phenotypes in Danish PJRA patients and controls. We have investigated the DNA restriction fragment length polymorphism (RFLP) of the MHC class I1 genes: HLA-DRB, -DQA, -DQB, DPA, and -DPB in Danish patients with pauciarticular juvenile rheumatoid arthritis. Material and methods Controls
Healthy, random unrelated Danish blood donors and members of the hospital staff were typed for HLA-A, -B, -C,-D, -DP antigens, and HLA class I1 genes and acted as controls. Patients
Blood samples were drawn from 54 patients with pauciarticular juvenile rheumatoid arthritis (PJRA) according to the ARA criteria (24). Pauciarticular JRA was defined as one to four affected joints during the first 6 months of the course of disease. All patients could also be classified as having juvenile chronic arthritis according to the EULAR criteria (25). All the patients had negative latex fixation and Waaler-Rose tests. None of the patients was suspected of having Reiter's disease, psoriatic arthropathy, arthritis associated with inflammatory bowel diseases, juvenile ankylosing spondylitis, or reactive arthritis due to recognized pathogens. The patients were randomly selected among the patients attending Hornbaek Hospital and the Department of Pediatrics, Rigshospitalet. All patients had previously been typed for HLAD, DR, and/or DP (4, 16, 17, 23). Serological and cellular HLA typing
HLA-ABC typing was performed with the NIH microlymphocytoxicity technique. HLA-D typing with the mixed lymphocyte culture (MLC) technique was done essentially as described by Thomsen et al. (26). HLA-DP typing with the primed lymphocyte typing (PLT) technique was done as described by 0dum et al. (27).
Typing for HLA class I1 genes
Peripheral blood was drawn in equal amounts of heparinized medium, RPMI 1640 (Gibco) and peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation. DNA was isolated from PBMC or granulocytes by solubilization of the cells in 10 mM Tris (pH 7.5), 2 mM EDTA, 10 mM NaC1, 1% SDS and 0.2 mg/ml proteinase K (Boehringer Mannheim) and incubated at 37°C for 16 h. Protein contaminants were removed from high-molecular weight DNA by phenol-chloroform-isoamyl alcohol extraction, and the DNA was recovered by ethanol precipitation. Restriction enzyme digest of chromosomal DNA was performed overnight using 5 U of enzyme per pg of DNA. The enzymes were BurnHI, BgnI, EcoRI, EcoRV, HindIII, MspI, PstI, PvuII, RsuI, and TuqI. The digested samples were loaded on agarose gels (0.6-1 .O%) and electropheresed for 25-35 h at 20-30 V in Tris acetate buffer, pH 8.0. The DNA was then transferred onto Hybond-N nitrocellulose filters (Amersham). Hybridization of the filters was for 40 h at 42°C in plastic bags with a total volume of 20 ml of 50% (vol/vol) formamide, 5 x SSPE (1 x SSPE = 180 mM NaCl/ 10 mM NaH,PO,/l mM EDTA), 0.2% Ficoll, 0.2% polyvinylpyrilidone, 0.2% bovine serum albumin, 5% (Wtlvol) dextran sulphate, and the radiolabeled probe at approximately 10 ng/ml. After hybridization, the filters were washed twice, 5 min each, in 2 x SSPE/O.5% NaDodSO, (Sodium dodecyl sulphate) at room temperature and twice, 30 min each, in 0.2xSSPE/O.S% NaDodSO, at 65°C. Autoradiography was for 1-7 days on Kodak XAR-5 film with intensifying screens. DNA probes
The DNA probes used were: DRB: A 800 bp S a d / Hind11 fragment of the cDNA clone pII-p-4 complementary to the DRBl gene (28). DQA: A 770 bp ApuI fragment of the cDNA clone pII-a-5 complementary to the DQAI gene (29). DQB: A 1080 bp PstI fragment of the cDNA clone pII-p-1 complementary to the DQBI gene (30). DPA: A 4.8 kb Hind11 fragment of the genomic clone p412-1 complementary to the DPAl gene (31). DPB: A 640 bp PstI fragment of the cDNA clone pII-p-7 complementary to the DPBI gene (32). Assignment of RFLP types
Assignments of HLA class I1 gene associated DNA fragments with the RFLP technique was performed as previously described (34-36). Of special interest were the following correlation coefficients (r): Dw8 17
Morling et al. Table 1. Associations between HLA class II DNA-fragments and pauciarticular juvenile rheumatoid arthritis PJRA RFLP band ~~
KB
5.2 2.1 8.4 6.2 9.1 15.1 11.5 6.1 2.0 4.7 4.8 1.7 4.2 13.5 2.1
m Taql
DQB Rsal DQB sgni DPA Taql DPB Mspl
Association
N
%
DRBl'04 DRB1'07 DRB1'08 DRBl'71 DRBYO7 t 02 03 DRB4'0107 DQA 10102 DR4 1'0201 DQA 1 *0401 DQA 1'0401 0501 DR4 1'0507 DQBl'O301 DQB1*0307 DPA7'0201 DPBl'O2
54 39 54 54 38 54 43 38 39 52 54 39 54 52
14.8 0 35.2 24.1 76.3 25.9 11.6 2.6 41.0 69.2 55.6 46.2 40.7 44.2 40.7
Controls N %
Relative risk
~
~
DRB Taql DRB Rsal DRB Taql DRB Taql DRB Bghl DRB Taql DQA BamHl WA Mspl WA Rsal DQA Mspl
(10 WS)' (5.22) (8.66) (6.08) (9.16) (13.81) (10.87) (6.05) (4.74) (4.1 2) (12.07)
+
+
54
97 58 97 97 81 97 89 79 54 99 101 57 99 76 98
+:Kilobase values assigned at the 10th international histocompatibility workshop 1987 (33). ': p < 0.05; ': p < lo-'; *: Not significant after 'correction' of p values. p
Abstract: We investigated the DNA restriction fragment length polymorphism (KFLP) of the major histocompatibility complex (MHC) class I1 genes: HLA-DRB, -DQA, -DQB, DPA, and -DPB in 54 patients with pauciarticular juvenile rheumatoid arthritis (PJRA) and in healthy Danes. The frequencies of DNA fragments associated with the following HLA class I1 genes were increased in PJRA when compared to normal controls: DRBI*08 (DRw8) (35.2% vs 10.3%, R R = 4.6, p < DRB3*01/02/ ,03 (DKw52) (76.3% vs 48.1%, R R 3.5, p < 10-3), DQAI*0401 (41.0% vs DQAI'0501 (55.6% vs 29.7%, R R = 3.0, 7.4%, RR = 7.9, p < p < 10--2,DQBI*0301 (DQw7) (46.2% vs 17.5%, R R = 4.0; p < DPA1+0201 (44.2% vs 7.9%, R R = 8.7, p < and DPB1*02 (DPw2) The frequency of DRBI*II was (40.7% vs 7.1%, RK = 8.5, p < not significantly increased. The frequencies of DNA fragments associated with the following HLA class I1 genes were decreased in PJRA although not statistically significantly so after 'correction' of p values: DRB1*04 (14.8% vs 40.2%, RR = 0.27; p < DRB1*07 (0'70 vs 25.9%, RR = 0.04, p < lo-'), DRB4*0101 (DRw53) (25.9% vs 53.6%, R R = 0.31, and p < lo-'), DQA1*0102 (11.6% vs 36.0%, R R = 0.25, p < Individuals who carDQA1*0201 (2.6% vs 34.2%, RR = 0.05, p < ried certain combinations of two of the risk factors seemed to have a further increased risk of developing PJRA compared to individuals carrying only one of the risk factors. The combinations of HLA class 11 genes primarily involved were: DRBI*08 and DQAI*O5OI, DRBI*08 and DQBI*0301, DRBI*08 and DPA 1*0201, DQAP0401 and DQBI*0301, DQA1*0401 and DPA1*0201, l?QAI*OSOI and DPAP0201, and DQAI*0501 and DPBI*02, although other combinations also gave odds ratios which were non-significantly increased compared to the odds ratios of the individual genetic markers. The data suggest that HLA class I1 gene products coded by genes at different loci interact in the susceptibility of 1 PJRA.
Introduction
The human major histocompatibility complex (MHC), the H L A system, encodes at least three polymorphic series of class I1 transmembraneous glycoproteins. The HLA class I1 molecules are designated DR, DQ, and DP and each consists of an a and a fl glycoprotein chain (1). A great number of diseases are associated with HLA-DR/DQ antigens (e.g. (2)), while a limited number of diseases have been reported to be associated with HLA-DP antigens: pauciarticular juvenile rheumatoid arthritis (PJRA) (3, 4), aplastic anemia (5), celiac disease (6, 7), leukemia (S), and alopecia areata (9). The mechanism by which variants of the HLA
Nials M~rling'~',Johannet Friig, Lars Fuggrr', dirgen 6eergsen', Cartten Heilmann', Fnddy Karup PeIsrsrI', Niels Bdum' and Arne Svsjgaard' 'Tissue Typing Laboratory, 'Hornbrek Hospital, 3Paediatric Department, Slate University Hospital (Rigshospitalet), 'Institute of Forensic Genetics, University of Copenhagen, Denmark
Key words: DNA polymorphism - class 1 IPJRA Received 22 February, revised, accepted for publication 29 March 1991
class I1 molecules confer susceptibility to different diseases is most probably related to the function of these molecules in the immune response. Thus, class I1 molecules are involved in the selection in the thymus of the T-cell repertoire (lo), and they bind antigenic peptide and present it to the antigen reactive CD4-positive T cells (1 1). The HLA class I1 molecules thus act like restriction elements in the presentation of antigens to T lymphocytes and thereby may confer susceptibility to various, in particular autoimmune, diseases. Pauciarticular juvenile rheumatoid arthritis is associated with HLA-B27 (e.g. IZ), DRw8/DwS, HLA-DRS/DwS, DRw6/Dw6 (e.g. 13-19), and DPw2 (3, 4). l h e combined presence of (i) DPw2
HLA class II genes in PJRA and (ii) Dw5 and/or Dw8 is associated with a significantly higher risk of PJRA than each antigen alone, indicating interaction of DP and DR gene products (4). Since the initiation of the present work, van Kerckhove et al. (20) have reported on increased frequencies in PJRA of the HLADRBI*80I, DQAI*401, DQBI*402 genes, and Begowitch et al. (21), van Kerckhove et al. (22), and we (23) have reported increased frequency of the DPBI*0201 allele which codes for DPP-chains of the majority of the DPw2 phenotypes in Danish PJRA patients and controls. We have investigated the DNA restriction fragment length polymorphism (RFLP) of the MHC class I1 genes: HLA-DRB, -DQA, -DQB, DPA, and -DPB in Danish patients with pauciarticular juvenile rheumatoid arthritis. Material and methods Controls
Healthy, random unrelated Danish blood donors and members of the hospital staff were typed for HLA-A, -B, -C,-D, -DP antigens, and HLA class I1 genes and acted as controls. Patients
Blood samples were drawn from 54 patients with pauciarticular juvenile rheumatoid arthritis (PJRA) according to the ARA criteria (24). Pauciarticular JRA was defined as one to four affected joints during the first 6 months of the course of disease. All patients could also be classified as having juvenile chronic arthritis according to the EULAR criteria (25). All the patients had negative latex fixation and Waaler-Rose tests. None of the patients was suspected of having Reiter's disease, psoriatic arthropathy, arthritis associated with inflammatory bowel diseases, juvenile ankylosing spondylitis, or reactive arthritis due to recognized pathogens. The patients were randomly selected among the patients attending Hornbaek Hospital and the Department of Pediatrics, Rigshospitalet. All patients had previously been typed for HLAD, DR, and/or DP (4, 16, 17, 23). Serological and cellular HLA typing
HLA-ABC typing was performed with the NIH microlymphocytoxicity technique. HLA-D typing with the mixed lymphocyte culture (MLC) technique was done essentially as described by Thomsen et al. (26). HLA-DP typing with the primed lymphocyte typing (PLT) technique was done as described by 0dum et al. (27).
Typing for HLA class I1 genes
Peripheral blood was drawn in equal amounts of heparinized medium, RPMI 1640 (Gibco) and peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation. DNA was isolated from PBMC or granulocytes by solubilization of the cells in 10 mM Tris (pH 7.5), 2 mM EDTA, 10 mM NaC1, 1% SDS and 0.2 mg/ml proteinase K (Boehringer Mannheim) and incubated at 37°C for 16 h. Protein contaminants were removed from high-molecular weight DNA by phenol-chloroform-isoamyl alcohol extraction, and the DNA was recovered by ethanol precipitation. Restriction enzyme digest of chromosomal DNA was performed overnight using 5 U of enzyme per pg of DNA. The enzymes were BurnHI, BgnI, EcoRI, EcoRV, HindIII, MspI, PstI, PvuII, RsuI, and TuqI. The digested samples were loaded on agarose gels (0.6-1 .O%) and electropheresed for 25-35 h at 20-30 V in Tris acetate buffer, pH 8.0. The DNA was then transferred onto Hybond-N nitrocellulose filters (Amersham). Hybridization of the filters was for 40 h at 42°C in plastic bags with a total volume of 20 ml of 50% (vol/vol) formamide, 5 x SSPE (1 x SSPE = 180 mM NaCl/ 10 mM NaH,PO,/l mM EDTA), 0.2% Ficoll, 0.2% polyvinylpyrilidone, 0.2% bovine serum albumin, 5% (Wtlvol) dextran sulphate, and the radiolabeled probe at approximately 10 ng/ml. After hybridization, the filters were washed twice, 5 min each, in 2 x SSPE/O.5% NaDodSO, (Sodium dodecyl sulphate) at room temperature and twice, 30 min each, in 0.2xSSPE/O.S% NaDodSO, at 65°C. Autoradiography was for 1-7 days on Kodak XAR-5 film with intensifying screens. DNA probes
The DNA probes used were: DRB: A 800 bp S a d / Hind11 fragment of the cDNA clone pII-p-4 complementary to the DRBl gene (28). DQA: A 770 bp ApuI fragment of the cDNA clone pII-a-5 complementary to the DQAI gene (29). DQB: A 1080 bp PstI fragment of the cDNA clone pII-p-1 complementary to the DQBI gene (30). DPA: A 4.8 kb Hind11 fragment of the genomic clone p412-1 complementary to the DPAl gene (31). DPB: A 640 bp PstI fragment of the cDNA clone pII-p-7 complementary to the DPBI gene (32). Assignment of RFLP types
Assignments of HLA class I1 gene associated DNA fragments with the RFLP technique was performed as previously described (34-36). Of special interest were the following correlation coefficients (r): Dw8 17
Morling et al. Table 1. Associations between HLA class II DNA-fragments and pauciarticular juvenile rheumatoid arthritis PJRA RFLP band ~~
KB
5.2 2.1 8.4 6.2 9.1 15.1 11.5 6.1 2.0 4.7 4.8 1.7 4.2 13.5 2.1
m Taql
DQB Rsal DQB sgni DPA Taql DPB Mspl
Association
N
%
DRBl'04 DRB1'07 DRB1'08 DRBl'71 DRBYO7 t 02 03 DRB4'0107 DQA 10102 DR4 1'0201 DQA 1 *0401 DQA 1'0401 0501 DR4 1'0507 DQBl'O301 DQB1*0307 DPA7'0201 DPBl'O2
54 39 54 54 38 54 43 38 39 52 54 39 54 52
14.8 0 35.2 24.1 76.3 25.9 11.6 2.6 41.0 69.2 55.6 46.2 40.7 44.2 40.7
Controls N %
Relative risk
~
~
DRB Taql DRB Rsal DRB Taql DRB Taql DRB Bghl DRB Taql DQA BamHl WA Mspl WA Rsal DQA Mspl
(10 WS)' (5.22) (8.66) (6.08) (9.16) (13.81) (10.87) (6.05) (4.74) (4.1 2) (12.07)
+
+
54
97 58 97 97 81 97 89 79 54 99 101 57 99 76 98
+:Kilobase values assigned at the 10th international histocompatibility workshop 1987 (33). ': p < 0.05; ': p < lo-'; *: Not significant after 'correction' of p values. p
