Heart Vessels (1992) Suppl. 7:73-80
Heart andVessels © Springer-Verlag1992
HLA-Linked susceptibility and resistance to Takayasu Arteritis Rui-Ping D o n g , 1 Akinori Kimura, 1 Fujio N u m a n o , 2 Yasuharu Nishimura, 1 and Takehiko Sasazuki 1 1The Department of Genetics, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, Japan and 2The Third Department of Internal Medicine, Tokyo Medical and Dental University, Tokyo, 113 Japan
Summary. To investigate genetic factors involved in the pathogenesis of Takayasu arteritis, patients in the Japanese population were examined for HLA-A, -B, and -C alleles by serological typing and for HLA-DR, DQ, and DP alleles by DNA typing using polymerase chain reaction (PCR)/sequencespecific oligonucleotide probe (SSOP) analysis. The frequencies of HLA-Bw52, DRBI*1502, DRB5*OI02, DQAI*OI03, DQBI*0601, and DPBI*0901 alleles were significantly increased and the frequencies of HLA-Bw54, DRBI*0405, DRB4*0101, DQAI*0301, and DQBI*0401 alleles were significantly decreased. Strong linkage disequilibria among the increased alleles and among the decreased alleles were evident in the Japanese population. Therefore, the haplotype of HLA-Bw52-DRBl*1502-DRB5*OlO2-DQAl*0103DQBI*O601-DPAI*O2-DPBI*0901 may confer susceptibility to Takayasu arteritis while another haplotype of HLA-Bw54-DRBl*O405-DRB4*OlO1DQAI*O301-DQBI*040I may confer resistance to the disease. These observations clearly indicate that HLAlinked gene(s) are involved in the development of Takayasu arteritis. Key words: Takayasu arteritis - HLA Typing - Disease susceptibility gene - Sequence-specific oligonucleotide probe (SSOP) - Polymerase chain reaction (PCR)
Introduction Takayasu arteritis, a chronic inflammatory arteriopathy is prevalent in women. Autoimmunity and genetic factor(s) are suspected of playing a significant
Address correspondence to: T. Sasazuki
role in the pathogenesis of the disease. There is a relatively high incidence of the disease in Asian and South American countries [1-3], and data on multiplex families with the disease [4-7] have been reported. The HLA genes in human populations bestow a diversity of immune responses to natural antigens and genetically control susceptibility and resistance to some diseases [8, 9]. Several researchers including ourselves investigated the association between Takayasu arteritis and HLA genes at the population level and family analysis in attempts to identify the disease susceptibility and resistance genes [10--14].
Association between serologically determined HLA and Takayasu arteritis Since 1978, the association between Takayasu arteritis and HLA genes has been studied at the serological and cellular level. Isohisa et al. were the first to report that a definite positive association between patients and HLA-Bw52 in the Japanese population [10]. Castro et al. later found that HLA-Bw52 was also positively associated with Takayasu arteritis in Mexican populations [11]. These observations suggested that HLA-linked gene(s) are involved in development of the disease. In 1979, we noted that Takayasu arteritis was associated with HLA-Bw52 as well as with an HLA class II allele, HLA-DHO(Dw12) determined by the one-way mixed lymphocyte reaction, using HLA-D homozygous typing cells [12]. Moriuchi et al. in 1982 reported that HLA-DR2 and MBI(DQwl) were positively associated with the disease [13]. In 1990, Takeuchi et al. noted a statisticalty significant increase in the frequency of DQA Taq I 6.6Kb fragment associated with DQw6 (a split of DQwl) in these patients [14]. In the Japanese population, HLA-Bw52 and Dwl2(DHO) are in strong linkage disequilibria with DR2 and MB1 (DQwl) [15]. Thus, it was suggested by serological and cellular
74 that the HLA-Bw52-Dw12(DHO)-DR2DQwl(MB1) haplotype might confer susceptibility to typing
Takayasu arteritis. We used a standard complementdependent microcytotoxicity assay [16] and a panel of highly selected alloantisera to identity HLA-A, -B, and -C antigens in both patients and controls. The frequencies of HLA-A and -C specificities did not differ significantly between patients and controls. Among the HLA-B specificities tested, the significantly increased frequency of Bw52 was confirmed; this antigen was positive in 53.1% of the 64 patients and in 24.0% of the 317 controls (P < 0.00001; relative risk = 3.59). On the other hand, we found a negative association between HLA-Bw54 and the disease (3.1% vs 13.9%; P < 0.04; R.R. = 0.20). When the corrected P value (Pc) was calculated by multiplying the P value (P) by 154, number of alleles tested at the HLA class I and class II loci, the association of Bw52 with the disease was retained (Pc < 0.002) and that of Bw54 lost statistical significance. There are, however, limitations to investigating the association between HLA genes and the disease, at the serological level. Typing of HLA class II alleles, especially DP alleles, is difficult due to the limited avaiiability of well-defined serological or cellular reagents. Furthermore, serological HLA types can be subdivided into many subtypes at the DNA level, for example, serologically determined DR2 can be subdivided into DRBI*1501, "1502, "1503, "1601, and "1602 subtypes. Hence, it was difficult to know which subtype or allele is associated with the disease at the serological typing level.
DNA (PCR/SSOP) Typing of HLA class lI genes Molecular and genetic analysis of HLA class II polymorphism has been facilitated using the polymerase chain reaction (PCR) [17-19]. PCR/sequence specific oligonucleotide probe (SSOP) analysis reveals not only that two alleles differ but how they differ, thereby allowing for a precise identification of individual alleles or polymorphic residues that are critical for disease susceptibility of resistance [9, 20] We investigated Japanese patients and a family with Takayasu arteritis for class II (DRB1, DRB3, DRB4, DRBS, DQA1, DQB1, DPA1, and DPB1) genes by DNA typing, using PCR/SSOP analysis. HLA class II loci (HLA-DR, DQ, and DP), located on the short arm of chromosome 6, encode an c~and 13 glycopeptide chain that binds foreign or self antigen peptides and presents them to T cells. One of the characteristic features of class II molecules is their extensive polymorphism in the first extracellular domain encoded by the second exon of the relevant HLA class II (DRA, DRB1, DRB3, DRB4~ DRB5, DQA1, DQB1, DPA1, and DPB1) genes. The detec-
R.-P. Dong et al.: HLA and Takayasu arteritis tion of HLA class I1 polymorphism has been developed from serological and cellular methods at the protein level to PCR/SSOP anatysis at the DNA level (HLA oligotyping). Amplification of a specific DNA sequence from genomic DNA using PCR makes feasible a precise analysis of the polymorphism. This is important for studies on disease, anthropology and transplantation. The second exon of DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, or DPB1 gene was amplified using two oligonucleotide primers synthesized toward the conserved sequences of each gene. To determine more precisely the DRB alleles from an individual, group-specific or locus-specific amplification of DRB genes was developed by selecting specific sequences as primers so that only a specific DRB gene or a group of DRB1 alleles would be amplified (Table 1). DNA was extracted from peripheral granulocytes of each subject using standard techniques [21]. Genomic DNA, as a template subjected to PCR, is first denatured by heating to 96°C, an approach that yields single DNA strands, then is cooled to 55°C or 60°C so that the primers anneal to their target sequences on the single strand, and then at 72°C so that annealed primers are extended on a base of the complementary sequences of the single strand with thermostable DNA polymerase. The denaturation, annealing and step of extension for each 1min was from 96°C to 55°C or 60°C to 72°C. Five hundred nanograms of DNA was subjected to 30 cycles of amplification with a programmable heat-block DNA thermal cycler. Following PCR, the products can be checked by electrphoresis in agarose gel. The amplified DNAs were spotted onto a nylon membrane and immobilized by alkaline denaturation in 0.4N NaOH and subsequent baking at 80°C for 1 h, making for a tight binding of DNA to the nylon membrane. Filters were prehybridized for i h at 54°C in 10 ml of hybridization buffer (50 mM Tris-HC1 (pH 8.0), 3 M tetramethylammonium chloride, 2mM EDTA, 5x Denhardt's solution, 1% SDS and 100~tg/ml heat-denatured herring sperre DNA). HLA class I! probes were end labeled with [7-32p] - ATP, using T4 polynucleotide kinase (or non-rädioactively digoxigenin-ddUTP, using terminal deoxyuncleotidyl transferase). Sequencespecific oligonucleotide probes (SSOPs) were selected to hybridize with sequences specific for one or more than one alMic variant. These SSOPs can hybridize with an allele(s) which has a sequence complementary to the SSOPs [22] (Table 2). After hybridization with labeled probes at 54°C for 1 h, the filters were washed twice at room temperature in 2x SSPE + 0.1% SDS and at 58-60°C in TMAC solution (50 mM Tris-HC1, 3M tetramethylammonium chloride, 2mM EDTA and 1% SDS) with constant gentle agitation. This procedure removed any probes that had one or more mismatches with the target sequence. The bound
R.-P. Dong et al.: HLA and Takayasu arteritis
75
Table 1. Oligounclectide primers for amplification of HLA class II genes Amplified gene
Name
Sequence 5' to 3'
DRBI generic
DRBAMP-A DRBAMP-B DRBAMP-1 DRBAMP-B DRBAMP-2 DRBAMP-B DRBAMP-4 DRBAMP-B DRBAMP-3 DRBAMP-B DRBAMP-52 DRBAMP-B DRBAMP-5 DRBAMP-B DQAAMP-A DQAAMP-B DQBAMP-A DQBAMP-B DPAAMP-A DPAAMP-B DPBAMP-A DPBAMP-B
CCCCACAGCACGTTTCTTG CCGCTGCACTGTGAAGCTCT TTCTTGTGGCAGCTTAAGTF CCGCTGCACTGTGAAGCTCT TTCCTGTGGCAGCCTAAGAGG CCGCTGCACTGTGAAGCTCT GTTTCTFGGAGCAGGTTAAAC CCGCTGCACTGTGAAGCTCT CACGTTTCTTGGAGTACTCTAC CCGCTGCACTGTGAAGCTCT CCCAGCACGTlq'CTI'GGAGCT CCGCTGCACTGTGAAGCTCT CAGCACGTTTCTFGCAGCAGGA CCGCTGCACTGTGAAGCTCT ATGGTGTAAACTTGTACCAGT TTGGTAGCAGCGGTAGAGTI'G CATGTGCTACTTCACCAACGG CTGGTAGTTGTGTCTGCACAC GCGGACCATGTGTCAACTFAT GCCTGAGTGTGGTTGGAACG GAGAGTGGCGCCTCCGCTCAT GCCGGCCCAAAGCCCTCACTC
DRB1-DR1 DRBI-DR2 DRB1-DR4 DRB1-DR3,5,6,8 DRB3 DRB5 DQA1 DQB1 DPA1 DPB1
Amplification of the DRB gene was done using DRBAMP-A and DRBAMP-B. Group specific amplification was made for the DR1 group with DRBAMP-1 and DRBAMP-B, for DR2 group with DRBAMP-2 and DRBAMP-B, for DR4 group with DRBAMP-4 and DRBAMP-B, and for DRw52 associated (DR3,5,6,w8) group with DRBAMP-3 and DRBAMP-B. Locus specific amplification of the DRB3 gene was done using DRBAMP-52 and DRBAMP-B, and that of DRB5 gene was done using DRBAMP-5 and DRBAMP-B. Amplification of DQA1, DQB1, DPA1 and DPB1 alleles was done using DQAAMP-A and DQAAMP-B, DQBAMP-A and DQBAMP-B, DPAAMP-A and DPAAMP-B, and DPBAMP-A and DPBAMP-B, respectively
probes were detected by autoradiogaphy, carried out for 10min to l h at room t e m p e r a t u r e with K o d a k X A R 5 film (or by enzymatic conversion of a colorless soluble substrate to a colored precipitate).
Association between HLA class II alleles typed at DNA level and Takayasu arteritis In recent work, HLA-DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, and DPB1 alleles were directly assigned at the D N A level, so that the HLA class II alleles, including H L A - D P associated with Takayasu arteritis could be precisely identified. In particular, it was asked whether the HLA-DP locus was closely linked with the susceptible gene to Takayasu arteritis. The frequencies of HLA-DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, and DPB1 alleles in 64 patients with Takayasu arteritis were compared with those in 317 healthy, unrelated Japanese controls. Strength of the statistical association between Takayasu arteritis and genetic markers was expressed by the relative risk (R.R.) as give by Woolf [23] and the statistical significance was examined by
chi-square test with Yates' correction. Linkage disequilibrium was estimated according to the method by Mittal [24]. As shown in Tables 3, 4, 5, and 6 positive associations between HLA and Takayasu arteritis were observed with DRBl*1502 (46.9% vs 24.0%; P < 0.0003; R . R . = 2.80), DRB5*OI02 (46.9% vs 24.0%; P < 0.0003; R.R. = 2.80), DQAI*0103 (59.4% vs 39.1%; P < 0.004; R.R. = 2.27), DQBI*0601 (57.8% vs 38.8%; P < 0.006; R.R. = 2.16), and DPBI*0901 (43.8% vs 20.5%; P < 0.0001; R.R. = 3.02). On the other hand, negative associations were observed with DRBI*0405 (10.9% vs 28.0%; P < 0.006; R.R. = 0.31), DRB4*OIO1 (50.0% vs 63.7%; P < 0.05; R . R . = 0.57), DQAI*0301 (53.1% vs 68.1%; P < 0.03; R . R . = 0.53), and DQBI*0401 (10.9% vs 28.4%; P < 0.006; R . R . = 0.32). An increase of DPAI*02 (90.6% vs 84.9%) was not significant. When Pc correction was applied, only the DRBl*1502 (Pc < 0.05), DRB5*OI02 (Pc < 0.05) and DPBI*0901 (Pc < 0.02) were retained and other alleles in the patients lost statistical significance. Because this .is the first evidence for the association between an HLA-DP allele and the disease, we examined the nucleotide sequence of the second exon of DPB1 allele in a
76
R.-P. Dong et al.: HLA and Takayasu arteritis
Table 2. Sequence specific oligonucleotides probes for DPA1 and DPBI alleles Gene
SSOP
Sequence 5' to 3'
Specificitya
DPA1
DPA3101 DPA3102 DPA5001 DPA5002
AAGATGAGATGTTCTATG AAGATGAGCAGTTCTATG AGTTTGGCCAAGCCTTTT AGTTTGGCCGAGCCTTTI"
0101,0102,0103 0201 0101,0102,0103 0201
29-34 29-34 47-53 47-53
DPB1
DPB0901 DPB0902 DPB0903 DPB0904 DPB090ó DPB0905
CTFTTCCAGGGACGGCAG GTGTACCAGTTACGGCAG GTGTACCAGGGACGGCAG GTGCACCAGTTACGGCAG CCAGGGACGACAGGAATG CCAGGGACGGCAGGAATG
8-13 8-13 8-13 8-13 10-15 10-15
DPB3501 DPB3502
GGGAGGAGTTCGCGCGCT GGGAGGAGTTCGTGCGCT
DPB3503 DPB3504 DPB3505 DPB3506 DPB5501 DPB5502 DPB5503 DPB5504 DPB6901 DPB6902 DPB6903 DPB6905 DPB7601 DPB7602 DPB7603 DPB8501 DPB8502 DPB8503
GGGAGGAGCTCGTGCGCT ACAACCGGCAGGAGTACG GGGAGGAGTACGCGCGCT GGGAGGAATTCGTGCGCT GGCCTGCTGCGGAGTACT GGCCTGATGAGGAGTACT GGCCTGAGGCGGAGTACT GGCCTGATGAGGACTACT GACATCCTGGAGGAGAAGC GCTCCTCCTCCAGGATGTC GACCTCCTGGAGGAGAAGCG ACCTCCTGGAGGAGAGGC GGACAGGATGTGCAGACA GGACAGGGTATGCAGACA GGACAGGATATGCAGACA AGCTGGGCGGGCCCATGA AGCTGGTCGGGCCCATGA AGCTGGACGAGGCCGTGA
0201,0202,0401,0402,0501,0801,1601,1901 0301,0601,1101,1301 0101,1501,1801 0901,1001,1401,1701 02011 O101,02012,0202,0401,0402, 0501,0801,1501, 1601,1801,1901 0401 0201, 0301,0402,0601,0801,0901,1401,1601,1701, 1801,1901 0202, 0501 1101,1501 0101,1301 1001 0101,0401,1101,1301,1501 0201,0402, 0801,1001,1601,1801 0202, 0501,1901 0301,0601,0901,1401,1701 0101,0401,0402, 0501,1801 0201,0202, 0801,0901,1001,1301,1601,1701,1901 0301,1401 1101,1501 Other than DPB7602 or 7603 positive alleles O101,0301,0801,0901,1001,1401 1301,1901 0201,0202,0401,0402 1501,1801 Other than DPB8501 or 8502 positive alleles
A.A. b
32- 37 32- 37 32- 37 30- 36 32-37 32-37 53-59 53-59 53- 59 53- 59 64- 69 64- 69 64 - 69 64-69 73-79 73- 79 73-79 82- 88 82-88 82-88
aThe allele designations are according to the recommendations made by the WHO official Nomenclature Committee [31] b Amino acid residues corresponding to SSOP
patient and her healthy relatives, and found it to be identical to the DPBI*0901 sequence [25]. No mutation was observed in the patient's D P B I alleles. Thus, the pathogenesis of Takayasu arteritis is not due to a mutant DPBI*090I allele.
Probable association between HLA haplotype and Takayasu arteritis Among the genetic markers tested, the frequencies of HLA-Bw52, DRBl*1502, DRB5*OI02, DQAI*OI03, DQBl*0601, and DPBI*0901 were significantly increased in the patients. The DQAI*OI03 and DQBl*0601 allele encodes for the alpha and beta chain of DQw6 (a split of D Q w l ) molecule, respectively. DRBI*1502-DRB5*OI02 (DR2-Dwl2) is in strong linkage disequilibrium with DQw6 (DQAl*OlO3DQBl*0601). Therefore, the increased frequencies of HLA-Bw52, DRBI*1502, DRB5*OI02, DQAI*OI03, and DQBI*O601 alleles in Takayasu arteritis were consistent with the increase in frequency of H L A -
Bw52-DR2-Dw12(DHO)-DQwl(MB1) haplotype determined by serological typing, cellular typing, and RFLP analysis. It was thought that there are no strong linkage disequilibria between DR and DP alleles, but we found specific linkage disequilibria between DR and DP alleles in the Japanese population, using the method of Mittal [24] (Table 7). Because DPAI*O2DPBI*0901 is in linkage disequilibria with both DRBI*1502 (t = 8.75) and HLA-Bw52 (t = 8.21) in the Japanese population, it is likely that DPAI*O2DPBI*0901 is also in linkage disequilibrium with the H L A - Bw52 - DRB1 *1502- DRB5 * 0102 - DQA1 * 0103DQBl*0601 haplotype. This observation strongly suggests that the HLA-Bw52-DRBl*1502-DRB5*OlO2D QAI*OIO3-DQB1 *0601-DPA1 *02-DPBl*0901 haplotype controls susceptibility to Takayasu arteritis in the Japanese population. The existence of HLA-Bw52DRB1 *1502-DRB5 *0102-DQA1 *OI03-DQB1 *0601 DPAI*O2-DPBI*0901 haplotype was also confirmed by examination of DNAs from members of the same family with Takayasu disease. Thus, the combination of alleles associated with the susceptibility to
R.-P. Dong et al.: HLA and Takayasu arteritis
77
Table 3. Association between Takayasu arteritis and DRB1 alleles Alleles~
Patients n = 64
Controls n = 317
Relative risk
~2b
7.8% 14.1 46.9 4.7 0 0 7.8 10.9 4.7 1.6 0 0 6.3 3.2 0 14.1 4.7 0 1.6 1.6 3.2 3.2 1.6 10.9 15.6 26.6 0
10.7% 13.9 24.0 0.6 0.9 4.4 3.8 28.1 5.0 0.3 2.8 3.2 7.5 3.4 0.6 12.3 4.4 0.3 3.2 4.1 3.2 0.6 0.6 6.3 17.3 30.3 0.6
0.71 1.02 2.80 7.75 0 0 2.15 0.31 0.93 5.02 0 0 0.81 0.90 0 1.17 1.06 0 0.61 0.37 0.99 5.08 2.50 1.82 0.88 0.83 0
0.23 0 14.21 3.59 0 1.82 1.19 8.30 0.04 0.10 0.83 1.02 0.01 0.06 0.10 0.15 0.06 0 0 0.38 0.29 1.24 0 1.73 0.11 0.33 0.10
010I 1501 1502 1602 0301 0401 0403 0405 0406 0407 0410 1101 1201 1202 1301 1302 1401 1402 1403 1405 1406 1407 07 0802 0803 0901 1001
pb
Heart andVessels © Springer-Verlag1992
HLA-Linked susceptibility and resistance to Takayasu Arteritis Rui-Ping D o n g , 1 Akinori Kimura, 1 Fujio N u m a n o , 2 Yasuharu Nishimura, 1 and Takehiko Sasazuki 1 1The Department of Genetics, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, Japan and 2The Third Department of Internal Medicine, Tokyo Medical and Dental University, Tokyo, 113 Japan
Summary. To investigate genetic factors involved in the pathogenesis of Takayasu arteritis, patients in the Japanese population were examined for HLA-A, -B, and -C alleles by serological typing and for HLA-DR, DQ, and DP alleles by DNA typing using polymerase chain reaction (PCR)/sequencespecific oligonucleotide probe (SSOP) analysis. The frequencies of HLA-Bw52, DRBI*1502, DRB5*OI02, DQAI*OI03, DQBI*0601, and DPBI*0901 alleles were significantly increased and the frequencies of HLA-Bw54, DRBI*0405, DRB4*0101, DQAI*0301, and DQBI*0401 alleles were significantly decreased. Strong linkage disequilibria among the increased alleles and among the decreased alleles were evident in the Japanese population. Therefore, the haplotype of HLA-Bw52-DRBl*1502-DRB5*OlO2-DQAl*0103DQBI*O601-DPAI*O2-DPBI*0901 may confer susceptibility to Takayasu arteritis while another haplotype of HLA-Bw54-DRBl*O405-DRB4*OlO1DQAI*O301-DQBI*040I may confer resistance to the disease. These observations clearly indicate that HLAlinked gene(s) are involved in the development of Takayasu arteritis. Key words: Takayasu arteritis - HLA Typing - Disease susceptibility gene - Sequence-specific oligonucleotide probe (SSOP) - Polymerase chain reaction (PCR)
Introduction Takayasu arteritis, a chronic inflammatory arteriopathy is prevalent in women. Autoimmunity and genetic factor(s) are suspected of playing a significant
Address correspondence to: T. Sasazuki
role in the pathogenesis of the disease. There is a relatively high incidence of the disease in Asian and South American countries [1-3], and data on multiplex families with the disease [4-7] have been reported. The HLA genes in human populations bestow a diversity of immune responses to natural antigens and genetically control susceptibility and resistance to some diseases [8, 9]. Several researchers including ourselves investigated the association between Takayasu arteritis and HLA genes at the population level and family analysis in attempts to identify the disease susceptibility and resistance genes [10--14].
Association between serologically determined HLA and Takayasu arteritis Since 1978, the association between Takayasu arteritis and HLA genes has been studied at the serological and cellular level. Isohisa et al. were the first to report that a definite positive association between patients and HLA-Bw52 in the Japanese population [10]. Castro et al. later found that HLA-Bw52 was also positively associated with Takayasu arteritis in Mexican populations [11]. These observations suggested that HLA-linked gene(s) are involved in development of the disease. In 1979, we noted that Takayasu arteritis was associated with HLA-Bw52 as well as with an HLA class II allele, HLA-DHO(Dw12) determined by the one-way mixed lymphocyte reaction, using HLA-D homozygous typing cells [12]. Moriuchi et al. in 1982 reported that HLA-DR2 and MBI(DQwl) were positively associated with the disease [13]. In 1990, Takeuchi et al. noted a statisticalty significant increase in the frequency of DQA Taq I 6.6Kb fragment associated with DQw6 (a split of DQwl) in these patients [14]. In the Japanese population, HLA-Bw52 and Dwl2(DHO) are in strong linkage disequilibria with DR2 and MB1 (DQwl) [15]. Thus, it was suggested by serological and cellular
74 that the HLA-Bw52-Dw12(DHO)-DR2DQwl(MB1) haplotype might confer susceptibility to typing
Takayasu arteritis. We used a standard complementdependent microcytotoxicity assay [16] and a panel of highly selected alloantisera to identity HLA-A, -B, and -C antigens in both patients and controls. The frequencies of HLA-A and -C specificities did not differ significantly between patients and controls. Among the HLA-B specificities tested, the significantly increased frequency of Bw52 was confirmed; this antigen was positive in 53.1% of the 64 patients and in 24.0% of the 317 controls (P < 0.00001; relative risk = 3.59). On the other hand, we found a negative association between HLA-Bw54 and the disease (3.1% vs 13.9%; P < 0.04; R.R. = 0.20). When the corrected P value (Pc) was calculated by multiplying the P value (P) by 154, number of alleles tested at the HLA class I and class II loci, the association of Bw52 with the disease was retained (Pc < 0.002) and that of Bw54 lost statistical significance. There are, however, limitations to investigating the association between HLA genes and the disease, at the serological level. Typing of HLA class II alleles, especially DP alleles, is difficult due to the limited avaiiability of well-defined serological or cellular reagents. Furthermore, serological HLA types can be subdivided into many subtypes at the DNA level, for example, serologically determined DR2 can be subdivided into DRBI*1501, "1502, "1503, "1601, and "1602 subtypes. Hence, it was difficult to know which subtype or allele is associated with the disease at the serological typing level.
DNA (PCR/SSOP) Typing of HLA class lI genes Molecular and genetic analysis of HLA class II polymorphism has been facilitated using the polymerase chain reaction (PCR) [17-19]. PCR/sequence specific oligonucleotide probe (SSOP) analysis reveals not only that two alleles differ but how they differ, thereby allowing for a precise identification of individual alleles or polymorphic residues that are critical for disease susceptibility of resistance [9, 20] We investigated Japanese patients and a family with Takayasu arteritis for class II (DRB1, DRB3, DRB4, DRBS, DQA1, DQB1, DPA1, and DPB1) genes by DNA typing, using PCR/SSOP analysis. HLA class II loci (HLA-DR, DQ, and DP), located on the short arm of chromosome 6, encode an c~and 13 glycopeptide chain that binds foreign or self antigen peptides and presents them to T cells. One of the characteristic features of class II molecules is their extensive polymorphism in the first extracellular domain encoded by the second exon of the relevant HLA class II (DRA, DRB1, DRB3, DRB4~ DRB5, DQA1, DQB1, DPA1, and DPB1) genes. The detec-
R.-P. Dong et al.: HLA and Takayasu arteritis tion of HLA class I1 polymorphism has been developed from serological and cellular methods at the protein level to PCR/SSOP anatysis at the DNA level (HLA oligotyping). Amplification of a specific DNA sequence from genomic DNA using PCR makes feasible a precise analysis of the polymorphism. This is important for studies on disease, anthropology and transplantation. The second exon of DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, or DPB1 gene was amplified using two oligonucleotide primers synthesized toward the conserved sequences of each gene. To determine more precisely the DRB alleles from an individual, group-specific or locus-specific amplification of DRB genes was developed by selecting specific sequences as primers so that only a specific DRB gene or a group of DRB1 alleles would be amplified (Table 1). DNA was extracted from peripheral granulocytes of each subject using standard techniques [21]. Genomic DNA, as a template subjected to PCR, is first denatured by heating to 96°C, an approach that yields single DNA strands, then is cooled to 55°C or 60°C so that the primers anneal to their target sequences on the single strand, and then at 72°C so that annealed primers are extended on a base of the complementary sequences of the single strand with thermostable DNA polymerase. The denaturation, annealing and step of extension for each 1min was from 96°C to 55°C or 60°C to 72°C. Five hundred nanograms of DNA was subjected to 30 cycles of amplification with a programmable heat-block DNA thermal cycler. Following PCR, the products can be checked by electrphoresis in agarose gel. The amplified DNAs were spotted onto a nylon membrane and immobilized by alkaline denaturation in 0.4N NaOH and subsequent baking at 80°C for 1 h, making for a tight binding of DNA to the nylon membrane. Filters were prehybridized for i h at 54°C in 10 ml of hybridization buffer (50 mM Tris-HC1 (pH 8.0), 3 M tetramethylammonium chloride, 2mM EDTA, 5x Denhardt's solution, 1% SDS and 100~tg/ml heat-denatured herring sperre DNA). HLA class I! probes were end labeled with [7-32p] - ATP, using T4 polynucleotide kinase (or non-rädioactively digoxigenin-ddUTP, using terminal deoxyuncleotidyl transferase). Sequencespecific oligonucleotide probes (SSOPs) were selected to hybridize with sequences specific for one or more than one alMic variant. These SSOPs can hybridize with an allele(s) which has a sequence complementary to the SSOPs [22] (Table 2). After hybridization with labeled probes at 54°C for 1 h, the filters were washed twice at room temperature in 2x SSPE + 0.1% SDS and at 58-60°C in TMAC solution (50 mM Tris-HC1, 3M tetramethylammonium chloride, 2mM EDTA and 1% SDS) with constant gentle agitation. This procedure removed any probes that had one or more mismatches with the target sequence. The bound
R.-P. Dong et al.: HLA and Takayasu arteritis
75
Table 1. Oligounclectide primers for amplification of HLA class II genes Amplified gene
Name
Sequence 5' to 3'
DRBI generic
DRBAMP-A DRBAMP-B DRBAMP-1 DRBAMP-B DRBAMP-2 DRBAMP-B DRBAMP-4 DRBAMP-B DRBAMP-3 DRBAMP-B DRBAMP-52 DRBAMP-B DRBAMP-5 DRBAMP-B DQAAMP-A DQAAMP-B DQBAMP-A DQBAMP-B DPAAMP-A DPAAMP-B DPBAMP-A DPBAMP-B
CCCCACAGCACGTTTCTTG CCGCTGCACTGTGAAGCTCT TTCTTGTGGCAGCTTAAGTF CCGCTGCACTGTGAAGCTCT TTCCTGTGGCAGCCTAAGAGG CCGCTGCACTGTGAAGCTCT GTTTCTFGGAGCAGGTTAAAC CCGCTGCACTGTGAAGCTCT CACGTTTCTTGGAGTACTCTAC CCGCTGCACTGTGAAGCTCT CCCAGCACGTlq'CTI'GGAGCT CCGCTGCACTGTGAAGCTCT CAGCACGTTTCTFGCAGCAGGA CCGCTGCACTGTGAAGCTCT ATGGTGTAAACTTGTACCAGT TTGGTAGCAGCGGTAGAGTI'G CATGTGCTACTTCACCAACGG CTGGTAGTTGTGTCTGCACAC GCGGACCATGTGTCAACTFAT GCCTGAGTGTGGTTGGAACG GAGAGTGGCGCCTCCGCTCAT GCCGGCCCAAAGCCCTCACTC
DRB1-DR1 DRBI-DR2 DRB1-DR4 DRB1-DR3,5,6,8 DRB3 DRB5 DQA1 DQB1 DPA1 DPB1
Amplification of the DRB gene was done using DRBAMP-A and DRBAMP-B. Group specific amplification was made for the DR1 group with DRBAMP-1 and DRBAMP-B, for DR2 group with DRBAMP-2 and DRBAMP-B, for DR4 group with DRBAMP-4 and DRBAMP-B, and for DRw52 associated (DR3,5,6,w8) group with DRBAMP-3 and DRBAMP-B. Locus specific amplification of the DRB3 gene was done using DRBAMP-52 and DRBAMP-B, and that of DRB5 gene was done using DRBAMP-5 and DRBAMP-B. Amplification of DQA1, DQB1, DPA1 and DPB1 alleles was done using DQAAMP-A and DQAAMP-B, DQBAMP-A and DQBAMP-B, DPAAMP-A and DPAAMP-B, and DPBAMP-A and DPBAMP-B, respectively
probes were detected by autoradiogaphy, carried out for 10min to l h at room t e m p e r a t u r e with K o d a k X A R 5 film (or by enzymatic conversion of a colorless soluble substrate to a colored precipitate).
Association between HLA class II alleles typed at DNA level and Takayasu arteritis In recent work, HLA-DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, and DPB1 alleles were directly assigned at the D N A level, so that the HLA class II alleles, including H L A - D P associated with Takayasu arteritis could be precisely identified. In particular, it was asked whether the HLA-DP locus was closely linked with the susceptible gene to Takayasu arteritis. The frequencies of HLA-DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, and DPB1 alleles in 64 patients with Takayasu arteritis were compared with those in 317 healthy, unrelated Japanese controls. Strength of the statistical association between Takayasu arteritis and genetic markers was expressed by the relative risk (R.R.) as give by Woolf [23] and the statistical significance was examined by
chi-square test with Yates' correction. Linkage disequilibrium was estimated according to the method by Mittal [24]. As shown in Tables 3, 4, 5, and 6 positive associations between HLA and Takayasu arteritis were observed with DRBl*1502 (46.9% vs 24.0%; P < 0.0003; R . R . = 2.80), DRB5*OI02 (46.9% vs 24.0%; P < 0.0003; R.R. = 2.80), DQAI*0103 (59.4% vs 39.1%; P < 0.004; R.R. = 2.27), DQBI*0601 (57.8% vs 38.8%; P < 0.006; R.R. = 2.16), and DPBI*0901 (43.8% vs 20.5%; P < 0.0001; R.R. = 3.02). On the other hand, negative associations were observed with DRBI*0405 (10.9% vs 28.0%; P < 0.006; R.R. = 0.31), DRB4*OIO1 (50.0% vs 63.7%; P < 0.05; R . R . = 0.57), DQAI*0301 (53.1% vs 68.1%; P < 0.03; R . R . = 0.53), and DQBI*0401 (10.9% vs 28.4%; P < 0.006; R . R . = 0.32). An increase of DPAI*02 (90.6% vs 84.9%) was not significant. When Pc correction was applied, only the DRBl*1502 (Pc < 0.05), DRB5*OI02 (Pc < 0.05) and DPBI*0901 (Pc < 0.02) were retained and other alleles in the patients lost statistical significance. Because this .is the first evidence for the association between an HLA-DP allele and the disease, we examined the nucleotide sequence of the second exon of DPB1 allele in a
76
R.-P. Dong et al.: HLA and Takayasu arteritis
Table 2. Sequence specific oligonucleotides probes for DPA1 and DPBI alleles Gene
SSOP
Sequence 5' to 3'
Specificitya
DPA1
DPA3101 DPA3102 DPA5001 DPA5002
AAGATGAGATGTTCTATG AAGATGAGCAGTTCTATG AGTTTGGCCAAGCCTTTT AGTTTGGCCGAGCCTTTI"
0101,0102,0103 0201 0101,0102,0103 0201
29-34 29-34 47-53 47-53
DPB1
DPB0901 DPB0902 DPB0903 DPB0904 DPB090ó DPB0905
CTFTTCCAGGGACGGCAG GTGTACCAGTTACGGCAG GTGTACCAGGGACGGCAG GTGCACCAGTTACGGCAG CCAGGGACGACAGGAATG CCAGGGACGGCAGGAATG
8-13 8-13 8-13 8-13 10-15 10-15
DPB3501 DPB3502
GGGAGGAGTTCGCGCGCT GGGAGGAGTTCGTGCGCT
DPB3503 DPB3504 DPB3505 DPB3506 DPB5501 DPB5502 DPB5503 DPB5504 DPB6901 DPB6902 DPB6903 DPB6905 DPB7601 DPB7602 DPB7603 DPB8501 DPB8502 DPB8503
GGGAGGAGCTCGTGCGCT ACAACCGGCAGGAGTACG GGGAGGAGTACGCGCGCT GGGAGGAATTCGTGCGCT GGCCTGCTGCGGAGTACT GGCCTGATGAGGAGTACT GGCCTGAGGCGGAGTACT GGCCTGATGAGGACTACT GACATCCTGGAGGAGAAGC GCTCCTCCTCCAGGATGTC GACCTCCTGGAGGAGAAGCG ACCTCCTGGAGGAGAGGC GGACAGGATGTGCAGACA GGACAGGGTATGCAGACA GGACAGGATATGCAGACA AGCTGGGCGGGCCCATGA AGCTGGTCGGGCCCATGA AGCTGGACGAGGCCGTGA
0201,0202,0401,0402,0501,0801,1601,1901 0301,0601,1101,1301 0101,1501,1801 0901,1001,1401,1701 02011 O101,02012,0202,0401,0402, 0501,0801,1501, 1601,1801,1901 0401 0201, 0301,0402,0601,0801,0901,1401,1601,1701, 1801,1901 0202, 0501 1101,1501 0101,1301 1001 0101,0401,1101,1301,1501 0201,0402, 0801,1001,1601,1801 0202, 0501,1901 0301,0601,0901,1401,1701 0101,0401,0402, 0501,1801 0201,0202, 0801,0901,1001,1301,1601,1701,1901 0301,1401 1101,1501 Other than DPB7602 or 7603 positive alleles O101,0301,0801,0901,1001,1401 1301,1901 0201,0202,0401,0402 1501,1801 Other than DPB8501 or 8502 positive alleles
A.A. b
32- 37 32- 37 32- 37 30- 36 32-37 32-37 53-59 53-59 53- 59 53- 59 64- 69 64- 69 64 - 69 64-69 73-79 73- 79 73-79 82- 88 82-88 82-88
aThe allele designations are according to the recommendations made by the WHO official Nomenclature Committee [31] b Amino acid residues corresponding to SSOP
patient and her healthy relatives, and found it to be identical to the DPBI*0901 sequence [25]. No mutation was observed in the patient's D P B I alleles. Thus, the pathogenesis of Takayasu arteritis is not due to a mutant DPBI*090I allele.
Probable association between HLA haplotype and Takayasu arteritis Among the genetic markers tested, the frequencies of HLA-Bw52, DRBl*1502, DRB5*OI02, DQAI*OI03, DQBl*0601, and DPBI*0901 were significantly increased in the patients. The DQAI*OI03 and DQBl*0601 allele encodes for the alpha and beta chain of DQw6 (a split of D Q w l ) molecule, respectively. DRBI*1502-DRB5*OI02 (DR2-Dwl2) is in strong linkage disequilibrium with DQw6 (DQAl*OlO3DQBl*0601). Therefore, the increased frequencies of HLA-Bw52, DRBI*1502, DRB5*OI02, DQAI*OI03, and DQBI*O601 alleles in Takayasu arteritis were consistent with the increase in frequency of H L A -
Bw52-DR2-Dw12(DHO)-DQwl(MB1) haplotype determined by serological typing, cellular typing, and RFLP analysis. It was thought that there are no strong linkage disequilibria between DR and DP alleles, but we found specific linkage disequilibria between DR and DP alleles in the Japanese population, using the method of Mittal [24] (Table 7). Because DPAI*O2DPBI*0901 is in linkage disequilibria with both DRBI*1502 (t = 8.75) and HLA-Bw52 (t = 8.21) in the Japanese population, it is likely that DPAI*O2DPBI*0901 is also in linkage disequilibrium with the H L A - Bw52 - DRB1 *1502- DRB5 * 0102 - DQA1 * 0103DQBl*0601 haplotype. This observation strongly suggests that the HLA-Bw52-DRBl*1502-DRB5*OlO2D QAI*OIO3-DQB1 *0601-DPA1 *02-DPBl*0901 haplotype controls susceptibility to Takayasu arteritis in the Japanese population. The existence of HLA-Bw52DRB1 *1502-DRB5 *0102-DQA1 *OI03-DQB1 *0601 DPAI*O2-DPBI*0901 haplotype was also confirmed by examination of DNAs from members of the same family with Takayasu disease. Thus, the combination of alleles associated with the susceptibility to
R.-P. Dong et al.: HLA and Takayasu arteritis
77
Table 3. Association between Takayasu arteritis and DRB1 alleles Alleles~
Patients n = 64
Controls n = 317
Relative risk
~2b
7.8% 14.1 46.9 4.7 0 0 7.8 10.9 4.7 1.6 0 0 6.3 3.2 0 14.1 4.7 0 1.6 1.6 3.2 3.2 1.6 10.9 15.6 26.6 0
10.7% 13.9 24.0 0.6 0.9 4.4 3.8 28.1 5.0 0.3 2.8 3.2 7.5 3.4 0.6 12.3 4.4 0.3 3.2 4.1 3.2 0.6 0.6 6.3 17.3 30.3 0.6
0.71 1.02 2.80 7.75 0 0 2.15 0.31 0.93 5.02 0 0 0.81 0.90 0 1.17 1.06 0 0.61 0.37 0.99 5.08 2.50 1.82 0.88 0.83 0
0.23 0 14.21 3.59 0 1.82 1.19 8.30 0.04 0.10 0.83 1.02 0.01 0.06 0.10 0.15 0.06 0 0 0.38 0.29 1.24 0 1.73 0.11 0.33 0.10
010I 1501 1502 1602 0301 0401 0403 0405 0406 0407 0410 1101 1201 1202 1301 1302 1401 1402 1403 1405 1406 1407 07 0802 0803 0901 1001
pb
