163

LINKAGE BETWEEN RHEUMATOID ARTHRITIS

AND DRP ALLELIC THIRD HYPERVARIABLE REGION SEQUENCES IN SOUTHERN CHINESE PERSONS

SUSCEPTIBILITY AND THE PRESENCE OF HLA-DR4

JAKOB SEGLIAS, EDMUND K. LI, MICHAEL G. COHEN, RAYMOND W. S. WONG, PAUL K. POTTER, and ALEX K. SO

Objective. To analyze HLA-DR and DQ associa- tions with rheumatoid arthritis (RA) in patients from southern China.

Methods. In 66 patients and 45 controls, restric- tion fragment length polymorphism studies were per- formed using DRB, DQA, and DQB probes, and DRB allele-specific typing of polymerase chain reaction- amplified DRB DNA.

Results. The frequency of HLA-DR4 was signif- icantly increased among RA patients (42.4% versus 17.8%). Increased frequencies of the DQA3 allele (77.8% versus 48.9%) and the DQB1*0302 allele (71.0% versus 46.3%), which are in linkage disequilib- rium with DR4, were also found. Oligonucleotide typing showed that the amino acid sequence LLEQRRAA, spanning amino acid positions 67-74 of the DRP mole-

From the Rheumatology Unit, Department of Medicine, Royal Postgraduate Medical School, London, United Kingdom, the Department of Medicine, Prince of Wales Hospital, Chinese Uni- versity of Hong Kong, the Queen Mary Hospital, Pokfulam, Hong Kong, and the Princess Alexandra Hospital, Brisbane, Australia.

Supported by the Wellcome Trust. Dr. Seglias’ work was supported by the Schweizerische Gesellschaft fur Rheumatologie and by a grant from the European League Against Rheumatism.

Jakob Seglias, MD: Rheumatology Unit, Department of Medicine, Royal Postgraduate Medical School (current address: Klinik fur Rheumatologie, Kantonsspital Aarau, Switzerland); Ed- mund K. Li, MD: Department of Medicine, Prince of Wales Hospital, Chinese University of Hong Kong; Michael G. Cohen, FRACP: Princess Alexandra Hospital; Raymond W. S. Wong, MRCP: Queen Mary Hospital; Paul K. Potter, BSc: Rheumatology Unit, Department of Medicine, Royal Postgraduate Medical School; Alex K. So, PhD, MRCP: Rheumatology Unit, Department of Medicine, Royal Postgraduate Medical School.

Address reprint requests to Alex K. So, PhD, MRCP, Rheumatology Unit, Royal Postgraduate Medical School, Du Cane Road, London W12 ONN, UK.

Submitted for publication May 10.1991; accepted in revised form October 8, 1991.

cule, was found in 19 of 49 patients and 5 of 32 controls. The main DR4 allelic subtypes found in the population were DRB1*0404 and DRB1*0405, both of which car- ried the sequence. There was no difference in subtype distribution between patients and controls.

Conclusion. Chinese RA patients have an in- creased frequency of HLA-DR4 alleles which possess the same DRB third allelic hypervariable sequence shown to be associated with susceptibility in Caucasian RA patients.

Genes located within the major histocompati- bility complex determine susceptibility to rheumatoid arthritis (RA). Studies have shown a strong associa- tion between the disease and HLA-DR antigens in different ethnic groups: DR4 in Caucasian, American black, and Japanese patients with RA, DRl in Asian Indian and Ashkenazi Jewish patients, and Dw16 in Yakima Indians (1-3). The association is not complete, and 2630% of Caucasian patients do not carry DR4.

Recent studies have shown that a conserved epitope spanning the third allelic hypervariable region of the DRP molecule may form the molecular basis of susceptibility to RA (4-6). This supported earlier find- ings of a common HLA-DR determinant in RA, iden- tified using serologic and cellular techniques (7,8). The sequences over this third allelic hypervariable region, which spans amino acid residues 67-74 of the DRP chain, are LLEQKRAA and LLEQRRAA in the sus- ceptibility-associated alleles DRB 1 *040 1 (Dw4) and DRB 1 *0101 (DRI)/DRB 1 *04M (Dw 14), respectively. It is unclear, however, whether these epitopes alone are sufficient to account for HLA-D-associated dis- ease susceptibility. One way to study this is to deter- mine the distribution of these 2 sequences in patients

Arthritis and Rheumatism, Vol. 35, No. 2 (February 1992)

164 SEGLIAS ET AL

Table 1. Oligonucleotide probes used in DRBl allelic typing*

Corresponding amino acid sequence and Oligonucleotide Nucleotide sequence DRB allele

Washing temperature

620 GGCCCGC'ITCTGCTCCAGGAG 67-73, DRB 1*0401 70°C 62 1 GGCCCGCCTCTGCTCCAGGAG 67-73, DRB1*0101, DRB1*0404, DRBl*0403, 68°C

622 CGGCCTAGCGCCGAGTACTGG 55-61 ,DRB 1 *O405, DRB 1*0801 72°C 623 CAGAGGCGGGCCGAGGTGGAC 7&76,DRB1*0403, DRB1*0406, DRB1*0407 14°C

DRB1*0405, DRB1*1402

* Oligonucleotides 620 and 621 were synthesized in reverse orientation to the coding sequence, and oligonucleotides 622 and 623 were in the same orientation to the coding sequence.

from different ethnic groups, which have different distributions of HLA-DR alleles, to determine if sim- ilar associations with specific sequences can be found. HLA-DR4 has been reported to be associated with RA in southern Chinese individuals (9), but the role of specific sequences has not been studied. We therefore examined the distribution of these allelic hypervari- able region sequences, using oligonucleotide hybrid- ization and restriction fragment length polymorphism (RFLP) studies in southern Chinese patients from Hong Kong.

PATIENTS AND METHODS Patients. Sixty-six southern Chinese patients with

RA were studied. All patients had RA as defined by the American College of Rheumatology (formerly, the American Rheumatism Association) criteria (10). Forty-five healthy southern Chinese subjects served as the control population.

HLA-DR typing by RFLP. Seven micrograms of genomic DNA from each subject was digested with Tuq I and electrophoresed in 0.7% agarose gels. Southern transfer onto Hybond-N membranes (Amersham, Buckinghamshire, UK) was performed using standard techniques. Filters were prehybridized and hybridized using the recommended solu- tions, with the addition of 2 x lo6 counts per minute of radiolabeled probe in the hybridization solution. The filters were washed to high stringency; the final wash was for 30 minutes at 65°C in 0 . 2 ~ SSC (lx SSC = 0.15M NaCl and 0.0194 sodium citrate), 0.1% sodium dodecyl sulfate (SDS). The DRB probe is a 530-basepair Psf I fragment of a DR4/Dw4 complementary DNA (cDNA) clone, and the DQA and DQB probes are full-length cDNA clones. Probes were labeled with "P-CTP by random oligonucleotide primer synthesis (1 1).

Allele-specific hybridization of amplied DRB genes. Genomic DNA samples were amplified by the polymerase chain reaction using DRB genomic primers (forward TTCT- TCAATGGGACGGAGCG; reverse GCCGCTGCACTGT- GAAGCTCTC), with 1 unit of Tuq polymerase (Cetus, Emeryville, CA). Amplification was performed at 94°C for 2 minutes, 55°C for 2 minutes, and 72°C for 2 minutes. Amplified DNA samples were Southern blotted onto nylon membranes after electrophoresis in 1.5% agarose gels run in TAE buffer (40 mM Tris, 1 mM EDTA, pH 7.6).

Allelic sequences were detected by oligonucleotide hybridization, using 4 oligonucleotide probes (Table I ) . One hundred nanograms of the oligonucleotide was radiolabeled using T4 DNA kinase and 50 pCi of "P-yATP (Amersham), separated from unincorporated ATP on a Sephadex G-50 spin column, and then used in hybridization experiments at a specific activity of 1 x lo6 c p d m l of hybridization buffer. Washing conditions were determined for each oligonucleo- tide, to discriminate a single base mismatch to the probe sequence. The filters were washed in 2 x SSC, 0.1% SDS at room temperature for 30 minutes, then at the appropriate temperatures (Table 1) for 60 minutes. Autoradiography was performed for 1-3 days, using Hypefilm (Amersham).

RESULTS HLA-DR frequency in RA patients. HLA-DR

typing by RFLP showed that only DR4 had a signifi- cantly increased frequency among RA patients versus controls (42.4% versus 17.8%; relative risk = 3.4) (Table 2). The frequency of HLA-DR9 was slightly increased (34.8% versus 31.1%), but this was not

Table 2. HLA-DR antigen frequencies in Chinese rheumatoid arthritis @A) patients and controls

RA patients Controls (n = 45) (n = 66)

No. % No. %

DR 1 DR2 DR3 DR4 DR5 DR6 DR7 DR8 DR9 DRlO x3* Blankt

3 26

3 28

7 6 1 6

23 3

14 2

4.5 39.4 4.5

42.4 10.6 9.1 1.5 9.1

34.8 4.5

21.2 3.0

0 17 5 8 9

10 0 3

14 2

13 1

0 37.8 11.0 17.8 20.0 22.2 0 6.7

31.1 4.4

28.9 2.2

~ ~ ~

*The X3 pattern has been shown, by sequencing, to define a haplotype containing a DRBl*1201 allele (Bidwell J: personal com- munication). t Blank indicates inability to assign HLA-DR type.

HLA-DRB DETERMINANTS IN CHINESE RA PATIENTS 165

Table 3. DQA and DQB allelic frequencies defined by Taq I RFLP*

RA patients Controls (n = 63, 62)t

No. % No. %

(n = 45, 41)t

1 2 3 4 5 6 7 8 9 1 0 DQA la lb lc 2 3

DQBS l a Ib 2a 2b 3a 3b Blank

12 30 4

12 49

23 1 2 8 44 17 17

19.0 47.6 6.3

19.0 77.8

37.1 1.6 3.2

12.9 71.0 27.4 27.4

9 24 4 I5 22

16 6 1 6

19 13 13

20 53.3 8.9

33.3 48.9

39.0 14.6 2.4

14.6 46.3 31.7 31.7

* Restriction fragment length polymorphism (RFLPMefined DQA and DQB alleles were identified according to previously established patterns (12). RA = rheumatoid arthritis. t The first n value is for DQA; the second is for DQB. $ The DQB3a pattern identifies the DQB1*0302 allele; the DQB3b pattern identifies the DQB1*0301 allele.

statistically significant. The etiologic fraction for DR4 was 0.30. The results indicated that DR4 is the only HLA-DR antigen associated with RA in southern Chinese individuals.

HLA-DQA and DQB allelic frequency. HLA- DQA and DQB alleles detected by RFLP showed patterns similar to those reported in studies of Cauca- sian subjects (12). The DQA3 allele was found in 49 RA patients (77.8%) and 22 normal subjects (48.9%). The DQB1*0302 allele, which is detected by its char- acteristic RFLP pattern (DQB3a), was found in 44 RA patients (71 .O%), compared with 19 control subjects (46.3%). Both of these differences were statistically significant (DQA3 2 = 9.72, P < 0.005; DQB1*0302 2 = 6.30, P < 0.025) (Table 3).

Allele-specific oligonucleotide probe hybridiza- tion studies of third allelic hypervariable region se- quences. Oligonucleotides 620 and 621, which are specific for third allelic hypervariable region se- quences found in HLA-DRB alleles associated with RA in Caucasians, were used to study the prevalence of these sequences in Chinese RA patients and con- trols. In only 1 subject (an RA patient), from a total of 81 subjects studied with this technique, was there hybridization with oligonucleotide 620. Samples from 19 RA patients (38.8%) and 5 control subjects (15.6%) hybridized with oligonucleotide 621 (Figure 1 and Table 4).

Figure 1. Oligonucleotide hybridization of amplified DRB DNA. Samples were hybridized with oligonucleotide 621 (upper panel), after which the same filter was stripped and rehybridized with oligonucleotide 620 (lower panel). Lanes 1-3, HLA-DFU4, DR1/4, and DR4/5, respectively, from rheumatoid arthritis (RA) patients. Lane 4, DRB1*0401 (Dw4) from a control. Lane 5 , DRB1*0404 (Dw14) from a control. Lane 6, DR114 from an RA patient. Lane 7, DW/9 from an RA patient. Lane 8, DRu4 from a control. Lane 9, DR3/10 from a control. Lane 10, Negative control for the polymer- ase chain reaction. See Table 1 for washing conditions.

Oligonucleotide subtyping of DR4-bearing hap- lotypes. We further subtyped for the presence of the DRB 1 *M05 (Dw 15) allele using allele-specific oligonu- cleotide probe hybridization. Three of 5 DR4+ con- trols (60%) and 12 of 19 DR4+ RA patients (63%) possessed this allele. None of them possessed the DRB1*0402 (DwlO) or DRB1*0403 (Dw13) alleles (Ta- ble 4).

DISCUSSION Based on studies on Caucasian patients with

rheumatoid arthritis, a shared epitope on HLA-DRP chains, spanning amino acids 67-74, has been pro-

Table 4. DRB I allelic subtyping by allele-specific oligonucleotide probes*

RA patients Controls (n = 32)

No. % No. %

(n = 49)

DRBI*0401 1 2.0 0 0 DRB 1*0402 0 0 0 0 DRB 1*0403 0 0 0 0 DRB 1 *0404 7 14.3 2 6.3 DRB 1 *0405 12 24.5 3 9.4 DRB 1*0101 3 6.1 0 0 DRB 1 * 1402 2 4 1 3.1

* DRBl alleles were determined using the oligonucleotide probes listed in Table 1. See Patients and Methods for details. RA = rheumatoid arthritis.

166 SEGLIAS ET AL

posed to form the basis of the observed HLA associ- ations with RA. We have now studied HLA-DR associations with RA in southern Chinese patients, a group for which there was little previous information on HLA associations with RA. Findings in this study allow further examination of the role of specific DRP sequences in disease susceptibility.

We found that HLA-DR4 is strongly associated with RA in southern Chinese subjects, but unlike other ethnic groups, this group did not exhibit association with other DR alleles that also possess this shared epitope (as DRB1*0101, DRB1*1402). The presence of shared third allelic hypervariable region epitopes was studied by allele-specific oligonucleotide hybridiza- tion. In contrast to observations in Caucasian patients, no association with the DRB1*0401 (Dw4) allele, which possesses the LLEQKRAA sequence, was found. There was an increase in the percentage of patients who possessed the LLEQRRAA sequence (38.8%, versus 15.6% of controls). This was accounted for mainly by the DRB1*0405 (Dw15) allele, with the rest being DRB1*0404 (Dw14) alleles. Unlike in the Caucasian population, the DRB 1*0401 allele was very uncommon in the Chinese subjects, and nearly all DR4 haplotypes in Chinese RA patients possessed the sequence LLEQRRAA, spanning the third allelic hy- pervariable region of the DRP molecule.

The distribution of DR4 subtypes in the Chinese RA patients, while different from that found in Cauca- sian RA patients, was similar to that found in Chinese DR4+ controls, though the number of control DR4 haplotypes analyzed was small. A small number of both patients and controls had the DRB1*1402 (Dw16) allele, which has the same third allelic hypervariable region sequence as DRB1*0404. There was no signif- icant difference between the groups in the percentage positive for this allele.

The significantly increased frequencies of the RFLP-defined DQA and DQB alleles is due to their linkage disequilibrium with DR4. The DQA3 allele was significantly more frequent among RA patients than controls. Similarly, there was a significantly increased frequency of the RFLP-defined DQB3a pattern, which correlated with the DQB1*0302 (DQw8) allele. Both RFLP patterns are also associated with HLA-DR9. Although the frequency of DR9 was slightly increased among the patients with RA, this was not statistically significant, even after correction for the increased frequency of DR4. There was no association with the DQB3b pattern (DQB1*0301), as reported in some studies of Caucasian RA patients (13). These results

therefore suggest that HLA-DR4 is the main D-region susceptibility gene for RA, and support earlier obser- vations that susceptibility to RA is linked to the DRB subregion.

Our results show that in southern Chinese RA patients, the HLA association is with DR4, and these patients possess a shared epitope, as has been re- ported for Caucasian RA patients. However, only 39% of our patients have this epitope, compared with nearly 80% of Caucasian patients (ref. 4 and Seglias J et al: unpublished observations). The third allelic hypervariable region sequence found in Chinese pa- tients is LLEQRRAA. It is found in DRB1*0404 and DRB 1 *O405 alleles predominantly, and together they constitute the vast majority (95%) of DR4 alleles found in patients. The lack of association with other DRB alleles that possess this sequence (i.e., DRB1*0101 and DRB1*1402) is likely to be due to the low frequen- cies of these alleles in the population. Analyzed indi- vidually, the DRB 1 *0404 and DRBl *a05 alleles were both more frequent in patients than controls, with DRB1*0405 being more prominent (DRB1*0404 14.3% versus 6.3%, DRB1*0405 24.5% versus 9.4%), but the differences between patients and controls did not reach statistical significance. These findings do not resolve the question of whether genetic susceptibility is due to an individual allele bearing the allelic hyper- variable region sequence, but taken in conjunction with similar results from studies in other ethnic groups, they suggest that a common third allelic hy- pervariable region sequence, found usually on DR4 alleles, underlies susceptibility to RA.

REFERENCES 1. Stastny P: Joint report: rheumatoid arthritis, Histocom-

patibility Testing 1980. Edited by PI Terasaki. Los Angeles, UCLA Tissue Typing Laboratory, 1980

2. Woodrow JC, Nichol FE, Zaphiropoulos G: DR anti- gens and rheumatoid arthritis: a study of two popula- tions. Br Med J 283:1287-1288, 1981

3. Willkens RF, Nepom GT, Marks CR, Nettles JW, Nepom BS: Association of HLA-Dwl6 with rheumatoid arthritis in Yakima Indians: further evidence for the “shared epitope” hypothesis. Arthritis Rheum 34:43- 47, 1991

4. Wordsworth BP, Lanchbury JSS, Sakkas LI, Welsh KI, Panayi GS, Bell JI: HLA-DR4 subtype frequencies in rheumatoid arthritis indicate that DRBl is the major susceptibility locus within the HLA class I1 region. Proc Natl Acad Sci USA 86:10049-10053, 1989

HLA-DRB DETERMINANTS IN CHINESE RA PATIENTS 167

5. Watanabe Y, Tokunaga K, Matsuki K, Takeuchi F, Matsuta K, Maeda H, Omoto K, Juji T: Putative amino acid sequence of HLA-DRp chain contribution to rheu- matoid arthritis susceptibility. J Exp Med 1690:2263- 2268, 1989

6. Nepom GT, Byers P, Seyfried CE, Healey LA, Wilske KR, Stage D, Nepom BS: HLA genes associated with rheumatoid arthritis: identification of susceptibility al- leles using specific oligonucleotide probes. Arthritis Rheum 32:15-21, 1989

7. Duquesnoy RJ, Marrari M, Hackbarth S, Zeevi A: Serological and cellular definition of a new HLA-DR associated determinant, MC1, and its association with rheumatoid arthritis. Hum Immunol 10: 165-167, 1984

8. Goronzy J , Weyand CM, Fathman CG: Shared T cell recognition sites on human histocompatibility leucocyte antigen class I1 molecules of patients with seropositive rheumatoid arthritis. J Clin Invest 77:1042-1049, 1986

9. Hawkins BR: HLA and autoimmune disease, Human

Leucocyte Antigens in Chinese. Edited by BR Hawkins. Hong Kong, Hong Kong University Press, 1987

10. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS, Medsger TA Jr, Mitchell DM, Neus- tadt DH, Pinals RS, Schaller JG, Sharp JT, Wilder RL, Hunder GG: The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31:315-324, 1988

11. Feinberg AP, Vogelstein B: A technique of radiolabel- ling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 137:26&267, 1984

12. Bidwell JL, Bidwell EA, Bradley BA: HLA class I1 genes: typing by DNA analysis. Baillieres Clin Haema- to1 3:355-384, 1990

13. Singal DP, Reid B, Kassam YB, D’Souza M, Bensen WG, Adachi JD: HLA-DQ beta-chain polymorphism in HLA-DR4 haplotypes associated with rheumatoid ar- thritis. Lancet 11: 11 18-1 120, 1987