T-Cell receptor BV6S1 null alleles and HLA-DR1 haplotypes in polyarticular outcome juvenile rheumatoid arthritis

ELSEVIER

BRIEF

COMMUNICATION

T-Cell Receptor BV6Sl Null Alleles and HLA-DRl Haplotypes in Polyarticular Outcome Juvenile Rheumatoid Arthritis Alexei A. Grom, Christoph von Knorre, Kevin J. Murray, Patricia A. Donnelly, David N. Glass, and Edmund Choi ABSTRACT: JRA is a complex of disease subtypes which are normally identified by clinical features such as age of onset and extent of joint involvement both at onset and during the course of the disease. We previously identified an association between TCR BV6Sl null allele and one subgroup of early-onset pauciarticular patients positive for HLA-DQAl *0 10 I, an HLA haplotype predisposing to a potyarticular course of the disease. In this report we extend this observation by identifying an increased

prevalence of this nonfunctional or null allele in the patients with a polyarticular disease course regardless of the mode of onset. This increase was most prominent in clinical subsets that have early onset of the disease and a polyarticular outcome. In one clinical group, stratification of patients by the HLA allele DQAl*OlOl strengthened the association considerably. This implies that there is an increased genetic load defined by specific alleles of both MHC and TCR genes.

ABBREVIATIONS EOPA JRA early-onset pauciarticular JRA juvenile rheumatoid arthritis JRA MHC major histocompatibility complex

PCR TCR

polymerase chain reaction T-cell receptor

INTRODUCTION Juvenile rheumatoid arthritis (JRA) is a chronic idiopathic inflammatory disorder with a diverse clinical spectrum both at the onset and during the course of the disease

Cl]. It comprises

various

disease

subtypes

which

such as age of onset and extent of joint involvement. There are at least three major onset types: pauciarticular (4 or fewer joints), polyarticular (5 or more), and systemic onset. Pauciartitular JRA can be further subdivided into early onset are normally

identified

by clinical

features

From the Department of Molectdar Gevzetm, Biocbemistvy, and Microbiology (C. VX, EC.), ~~nl~l~~ity ofC~n~~~~ti~ CoEege of Med~r~~e,Cincinnati: md the Wi~iia~ S. Rowe Division of Rbe~~ato~agy (A.A.G., K.J. M., P. A. D., D. N.G.). Children’s Hospital Medical Center. Depautmnt of Pediatrics, University ojCincinnati, Co&e of Medicine, Cincinnati, Ohio, USA. Address reprint requests to Dr. E. Cboi, Department of Molecular Gcn&s. Biochemistry. and Microbiology. Uniwsity of Cincinniati. CO&Y of Medirine, Cincinnati, OH. 45267-0524. LISA. Received August 25. 1995: ucrepted November 8, 1995. Human immunology 45, 152-156 (1996) 0 American Society for Histocomparibiliry

(before age 6 years) and late onset. Early-onset pauciartitular (EOPA) JRA is both the most common form of JRA and the most comprehensively studied immunogenetically. EOPA-JRA patients can remain pauciarticular or can progress to a polyarticular erosive course of disease. HLA-DQA 1 *O 10 1 is positively associated with the progression to polyarticular disease in EOPA-JRA patients [2}. Furthermore, a polymorphism in one of the T-cell receptor (TCR) VP genes (TCRBV6Sl.) is associated with disease in HLA-DQAl*OlOl-positive EOPAJRA patients 133. This observation was reproduced in a population of JRA patients from the Seattle area {4}, but not in a population of Norwegian JRA patients {S]. The polymorphisms associated with disease are nonfunctional versions of the VP6.1 gene due to a tyrosine substitution of a highly conserved cysteine residue at position 92 {G, 71. The loss of this cysteine would prevent the formation 019%88S’)/‘)6/$15.00

and Immunogener~cs,

1996

SSDI 0198-S&359(95)00174-3

TCR Null Alleles and JRA

of conserved disulfide bridges found in normal Ig-like domains so that the gene product could not participate in the formation of a functional TCR. Two null alleles of this VB6.1 gene have been defined (TCRBV6S1*2P and “3P) l8.l. The goal of this study was to determine whether the described association might be observed in other clinical groups of JRA in addition to the early-onset pauciarticular patients. We typed a large group of JRA patients to determine the distribution of functional versus nonfunctional VB6.1 alleles in different clinical forms of the disease. Our results indicate that the association is strongest with the clinical subsets that have early onset of the disease and a polyarticular outcome. These include not only a group of EOPA-JRA patients that progress to a polyarticular disease course but also a group of earlyonset polyarticular patients. Both of these subsets are associated with HLA-DQAl*OlOl.

MATERIALS

AND

METHODS

Patients. Genomic

DNA samples obtained from 316 patients with different clinical forms of JRA as defined by the American College of Rheumatology criteria 111 were used in this study. The study group included 166 pauciarticular-onset patients (144 with early, less than 6 years of age, and 22 with late onset), 118 polyarticularonset patients (54 with early onset, less than 8 years of age, and 64 with late onset), and 32 patients with a systemic onset of the disease, all from the Cincinnati area. Definitive disease course was established in this group after at least 5 years of disease evolution. The control group of 190 unrelated Caucasian individuals resembled the disease population with respect to geographic and ethnic (European) origin. TCRBVGSl typing. Fifty-five patients and 130 controls were typed utilizing BV6Sl gene Bgl II polymorphism as previously described lb]. An additional 261 patients and 60 additional controls were typed utilizing a (GT), polymorphism found in the first intron of the BV6Sl gene 191. Carriers of the BV6Sl null alleles were defined as individuals with the 12.5-kb Bgl II restriction fragment length polymorphism or a (GT)ia or i3 dinucleotide repeat in the first intron of the BV6Sl gene. Oligonucleotide primers and conditions for polymerase chain reaction (PCR) amplification of genomic DNA samples for the (GT) polymorphism were reported by Charmley et al. [91. Briefly, 100 ng of genomic DNA was mixed with primers (0.3 pM final concentration), 20 pM of each dNTP, and 1 U of Taq DNA polymerase (Boehringer Mannheim, Indianapolis, IN, USA) in 50 pl total volume. One of the primers (3’ primer) was labeled using polynucleotide kinase (Boehringer Mannheim) and

153

[r-32P] ATP (Amersham, Arlington Heights, IL, USA). Primer equivalent to 5 x lo6 cpm was added to each PCR reaction. Amplification was performed for 30 cycles with 1 minute of denaturation (93”C), 1 minute of annealing (64”C), and 2 minutes of extension at 72X. Cycling was preceded by a single 3-minute denaturation (93X), and followed by a single 7-minute extension. Four microliters of the final PCR reaction product were added to an equivalent volume of formamide/dye loading buffer, heated at 90°C for 2 minutes, and loaded on a 6% denaturing (6 M urea) acrylamide gel. Gels were dried and exposed to Kodak XAR-5 film for 24 hours at -7ooc. HLA

typing. HLA-DR and -DQ antigens in our entire study population were typed serologically according to standard techniques using commercially available antisera. In two patients the absence of the HLA-DRl specificity was deduced from the available HLA-DQ splits data. HLA-DQAl splits were identified by the PCR gene amplification with sequence-specific primers. A kit containing 5’- and 3’-primers for identifying all recognized DQAl alleles was purchased from DYNAL. (Lake Success, NY, USA). Data analysis. Statistic

analysis was carried out using the chi-square test with two degrees of freedom, MantelHaenszel chi-square test, and Woolfs test for heterogeneity.

RESULTS Distribution of TCRBVGSl fering

alleles in JRA patients with difi

onset types and course of disease. TCR

BV6Sl null allele carrier frequencies of JRA patients and controls are shown in Table 1. Patients with a polyarticular disease course regardless of onset type had an increased BV6Sl null alleles carrier frequency compared with controls: 123 out of 205 (0.60) vs 86 out of 190 (0.45) in controls (x2 = 8.0, p < 0.01). Further differences were noted when the disease onset was also considered in addition to the clinical course (Table 1). We found a significant increase in the frequency of BV6Sl null allele carriers among patients in whom early-onset pauciarticular disease had evolved to become polyarticular: 33 out of 51 (0.65) vs 86 out of 190 (0.45) in controls (X2 = 5.3,p < 0.05). In contrast, the BV6Sl null allele carrier frequency in patients with early-onset persistent pauciarticular disease was not different from controls: 45 out of 93 (0.48). The polyarticular-onset JRA group was also associated with BV6Sl null alleles, with the greatest BV6Sl null allele carrier frequency observed in the group of patients with an onset of their polyarticular JRA before 8 years of

A.A. Grom et al.

154

TABLE

1

Distribution of BVGSl null allele carriers in different clinical groups of JRA patients and controls

course

Frequency of BV6Sl null allele carriers

Pauci Poly Pauci

(n = 93) (n = 51) (n = 13)

0.48 (45) 0.65 (33)” 0.08 (1)

Pal p

(n (n (n (n (n (n

0.33 0.69 0.56 0.52

Clinical

Type of onset Early-onset Late-onset

pauciarticular pauciarticuiar

Polg Polp Poly Systemic

Early-onset polyarticular Late-onset polyarticular

Systemic onset Controis

= 9) = 54) = 64) = 27) = 5) = 130)

(3) (37)” (36) (14)

(0) 0.45 (86,

“p \ 0.05 vs controls. 6p < 0.01 “S controls.

age: 37 out of 54 (0.69) vs 86 out of 190 (0.45) in controls (x2 = 8.2, p < 0.01). In contrast, late-onset polyarticular and systemic-onsetipolyarticular course JRA patients, analyzed separately, did not show a significant increase in BV6Sl null allele frequency. The number of late-onset pauciarticular JRA patients was too small for evaluation and will not be considered any further in this report. TCR BV6Sl null de/es and HLA u~socktions. Table 2 illustrates the differences in BV6Sl null allele carrier distribution in different clinical groups in addition to demonstrating the frequencies of HLA-DR 1, previously associated with polyarticular course of JRA. The association of BV6Sl null alleles with a polyarticular course of the disease was most prominent in the groups where HLA-DRl+ haplotypes are likely to be a significant major histocompatibility complex (MHC) risk factor. HLA-DQAl*OlOl-positive haplotypes have been implicated in the HLA-DRl association in JRA. Stratification of both patients and control populations by the HLA-DQAl*OlOl haplotypes considerably strength-

TABLE

2

BV6Sl null allele and DRl haplotype frequency in different subtypes of polyarticular outcome JRA Proportion of BV6Sl null allele carriers

Type of onset

Early-onset polyarticular (n = 54)

0.65 (33) 0.69 (37)

Late-onset polyarticular (n = 64) Systemic onset (n = 27) Controls (n = 190)

0.56 (36) 0.52 (14) 0.45 (X6)

Early-onset

pauciarticular

(n = 5 1)

HLA-UK 1 frequency 0.27 0.31 0.1 9 0.07 0.14

(14) (17) (I 2) (2) (27)

the association between the BV6Sl null alleles and EOPA-JRA with a polyarticular outcome of disease (Woolfs test for heterogeneity = 3.7, p = 0.054) (Table 3). Odds ratios for HLA-DQAl*OlOl haplotypes and for the presence of BVGSl null alleles separately were 2.3 (95% confidence limits 1 .O-5.5) and 2.1 (95% confidence limits 1.1-4.4), respectively, whereas odds ratio for the combination of these two risk factors was 6.4 (95% confidence limits 1.9-22.7). In contrast, in earlyonset polyarticular JRA, HLA-DQA 1 *O 10 1 stratification did not reveal any significant heterogeneity. ened

DISCUSSION We previously reported an increase in frequency of the nonfunctional alleles of TCR BVGSl gene in HLADQAl*OlOl’ EOPA-JRA patients 131. The present study performed on a larger group of JRA patients from Cincinnati shows that this clinical group represents those early-onset pauciarticular patients whose disease had progressed to become polyarticular. In addition, the increase in BVGSL null allele frequency has been also found in a group of polyarticular-onset JRA patients. The polyarticular-onset group is heterogeneous both clinically and immunogenetically, with two main haplotypes commonly predisposing to a polyarticular course: HLA-DR4’ and HLA-DRl’ {lo}. However, their relative contributions in different clinical groups vary. HLADRl is more important in patients with the onset of polyarricular JRA before the age of 8 years (K. Murray, unpublished observations). In contrast, in older patients with late-onset polyarticular disease (i.e., after 8 years) or systemic-onsetipolyarticular course, HLA-DR4 haplotypes appear to be more important, with no significant role for HLA-DRl. In addition, HLA-DRl is important in EOPA-JRA patients whose arthritis transforms into polyarticular disease. Thus, the association between a polyarticular course and TCR BV6Sl null alleles is especially pronounced in those groups where HLA-DR 1 + (or HLA-DQAl*OlOl’) haplotypes are felt to be the most important MHC risk factor. Published reports of TCR genetic contributions to certain autoimmune disease susceptibility imply that polymorphic variations may have functional significance. Allelic polymorphisms in the germ-line TCR genes provide additional diversity to the human TCR repertoire. It has been shown that even modest allelic variations in TCR VP coding regions sequences can have a significant impact on the expression of VP genes in the peripheral repertoire 1111. Polymorphism that leads to the formation of null or nonfunctional alleles can decrease the number of T cells bearing TCRs of a certain VP family {7} and thus may influence the course of an inflammatory disease.

TCR Null Alleles and JRA

TABLE

3

155

Proportion of BV6Sl null allele carriers in subtypes polyarticular outcome JRA patients stratified by HLA-DQA*OlOl DQ Al*0101

Type of onset

frequency

Early-onset pauciarticular (n = 49) Early-onset polyarticular (n = 5 1) Late-onset polyarticular (n = 63) Controls (n = 104)

0.33 (16) 0.35 (18)

n Woolfs test for heterogeneity

0.19 (12) 0.17 (18)

DQ Al*0101 + null allele carriers 0.75 0.72 0.58 0.28

(12/16)” (13/l@ (7112) (5/l@

of

DQ Al*0101 null allele carriers 0.58 0.70 0.55 0.49

(19133) (23/33) (28/51) (42186)

= 3.7,p = O.OS.

As previously discussed [3], one hypothesis raised to explain RA pathogenesis implies that the disease may indeed develop through “a hole” in TCR repertoire. In this scenario, the immune system may exhibit a selective impairment in its ability to recognize specific antigenic peptides which are ordinarily involved in the elimination of a pathogen. Such failure may permit persistence of the antigen and the development of chronic inflammation rather than resolution of the process. This hypothesis could explain an association between the presence of a nonfunctional BV6Sl allele and worse articular prognosis. An alternative possibility to a direct role of null TCR variable region genes is that closely linked genes may be involved. Although there is little linkage disequilibrium in this part of the genome, some extended haplotypes may still be defined. One of the described cassettes of BV genes includes the BV6Sl gene [ 12,131. In this instance, the linked BVlSl, BVSS5, and BV6S7 genes are also polymorphic and the true disease association may lie with one of these genes. Two MHC loci, HLA-DRBl and HLA-DQAl, have been implicated in the HLA-DRl association in JRA, with the HLA-DRBl*OlOl or HLA-DQAl*OlOl being the specific alleles most closely associated with disease 12, 101. Due to the strong linkage disequilibrium between these two genes, it has not been possible to distinguish their effects. In the present study the HLADQAl*OlOl allele has been preferred for the analysis. Stratification by HLA-DQAl*OlOl strengthens the association between TCRBVGSl null allele in the early pauciarticular-onset/polyarticular course JRA. However, this trend was not observed in the early-onset polyarticular group, where the null allele frequency was nearly identical, irrespective of the presence or absence of the HLA-DQAl*OlOl allele. Whether this indicates different genetic effects of the TCR and HLA genes in each of these clinical subgroups is unclear at this moment. Nevertheless, this may indicate an increased genetic load due to specific MHC and TCR VP alleles.

ACKNOWLEDGMENT

This research was supported in part by National Institutes of Health Grants ROl AI30622 (EC), AR39979 (DNG), and P20-AR-42632(DNG); the Schmidlapp Foundation; the Arthritis Rheumatism and Aging Information Systems ([ARAMIS] AM21393); Arthritis Foundation, Ohio River Valley Chapter; and Children’s Hospital Research Foundation of Cincinnati. Dr. Grom is a recipient of an Arthritis Foundation Fellowship and Dr. Murray was supported by a Saw Medical Fellowship from the University of Western Australia. C. von Knorre was a Fulbright scholar. The authors would like to thank Dr. Ed Giannini for invaluable assistance in statistic analysis.

REFERENCES 1. Cassidy JT, Levinson JE, Bass JC, Baum J, Brewer EJ Jr, Fink CW, Hanson V, et al: A study of classification criteria for diagnosis of juvenile rheumatoid arthritis. Arthritis Rheum 29:274, 1986. 2. Van Kerckhove C, Luyrink L, Taylor J, Melin-Aldana H, Balakrishnan K, Maksymowych WP, Elma MS, Love11 DJ, Choi E, Glass DN: HLA-DQAl*OlOl haplotypes and disease outcome in early onset pauciarticular juvenile rheumatoid arthritis. J Rheumatol 1816, 1991. WP, Gabriel CA, Luyrink L, Melin3. Maksymowych Aldana H, Elma M, Giannini E, Love11 DJ, Van Kerckhove C, Leiden J, Choi E, Glass DN: Polymorphism in a T-cell receptor variable gene is associated with susceptibility to a juvenile rheumatoid arthritis subset. Immunogenetics 35:257, 1992. 4. Charmley P, Nepom BS, Concannon P: HLA and T cell receptor P-chain polymorphisms identify a distinct subset of patients with pauciarticular-onset juvenile rheumatoid arthritis. Arthritis Rheum 37:695, 1994. 5. Ploski R, Hansen T, Forre 0: Lack of association T-cell receptor TCRBV61*2 allele in HLA-DQAl*OlOlpositive Norwegian juvenile chronic arthritis patients. Immunogenetics 38444, 1993. WP, Gabriel CA, Luyrink L, Van Kerck6. Maksymowych hove C, Leiden J, Choi E, Glass DN: Polymorphic markers related to a single Tcrb-V6 gene segment. Immunogenetics 33:281, 1991.

156

7. Luyrink L, Gabriel CA, Thompson S, Grom AA, Maksymowych WP, Choi E, Glass DN: Reduced expression of a human VB6.1 T-cell receptor allele. Proc Nat1 Acad Sci USA 90:4369, 1993. 8. Barron KS, Robinson MA: The human T-cell receptor variable gene segment TCRBVGSl has two null alleles. Hum Immunol 40:17, 1994. 9. Charmley P, Concannon P: Polymorphism and phylogeny of dinucleotide repeats in human T-cell receptor Vb6 genes. Immunogenetics 38:92, 1993. 10. Ploski R, Vinje 0, Ronningen KS, Spurkland A, Sorskaar D, Vartdal F, Forre 0: HLA class II alleles and heteroge-

A.A. Grom et al.

neity of juvenile rheumatoid arthritis: DRBl*OlOl may define a novel subset of the disease. Arthritis Rheum 36: 465, 1993. 11. Vissinga CS, Charmley P, Concannon P: Influence of coding region polymorphism on the peripheral expression of a human TCR V p gene. J Immunol 152:1222, 1994. 12. Li Y, Szabo P, Posnett D: The genomic structure of human VP6 T cell antigen receptor genes. J Exp Med 174: 1537, 1991. 13. Wei S, Charmley P, Robinson MA, Concannon P: The extent of the human germline T-cell receptor V beta gene segment repertoire. Immunogenetics 40:27, 1994.