Germalin repertoire of T-cell receptor β-chain genes in patients with insulin-dependent diabetes mellitus

Germl ine Repertoire of T-Cell Receptor ~Chain Genes in Patients with Insulin-Dependent Diabetes Mellitus E. Martinez-Naves, E. Coto, V. Guti6rrez, J. M. Urra, F. Seti6n, O. Dominguez, L E. Hood, and C. L6pez-Larrea ABSTRACT: We have investigated the genotype and allelic distribution of germline restriction fragment length pol~,morphisms of the T-cell receptor fl chain, seg- ment Cfl, and two variable segments which are in linkage disequilibrium, Vfl8 and Vfll 1, in 42 insulin-dependent diabetes mellitus (IDDM) patients and ,in 51 healthy blood donors used as controls. Recently, several works ABBREVIATIONS IDDM insulin-dependent diabetes mellitus MHC major histocompatibility complex RFLP restriction fragment length polymorphism INTRODUCTION Insulin-dependent diabetes mellitus (IDDM) is a chronic autoimmune disease characterized by a selective destruction of the pancreatic islet/8 cells, mediated by T lymphocytes [1]. The association of IDDM with certain HLA antigens, particularly with DR3 and DR4, and more recently with DQB [2] and DQA genes [3] which are in linkage dis- equilibrium with DRB locus, has been clearly estab- lished. However, the genetic control of IDDM is very complex: identical twins are 36% concordant whereas major histocompatibility complex (MHC) identical sib- lings are only about 12% concordant, as reviewed by Todd [4]. These and other data [5] indicate that predisposition F~ora the Servicio de lnmunologia, Hospital Cnvadonga. Oviedo, Spain and the Division of Biology, California Institute of Technology, Pasadena, Californfa. Address reprint requests to Dr. E. Martinez-Naves. Servicio de lm- raunologla, Hospital Covadonga, 33006 Oviedo, Spain. Received August 9, 1990; accepted October 1 l, 1990. Human Immunology 31, 77-80 (1991) © E. Mattinez-Naves et al., 1991 have reported contradictor, results showing or not show- ing an association between polymorphic alleles of the Cfl gene and diabetes type I. We found no significant differences in the allele, genc~ type, and haplotype distribution of the gene segments studied, between IDDM patients and control popula- tions. Human Immunology 31 .77-80 ¢1991J SDS sodium dodecyl sulfate SSPE sodium chloride-sodium phosphate-EDTA Tcr T-cell receptor to type I diabetes is polygenic and other non-MHC genes, in addition to environmental factors, appear to influence the susceptibility to IDDM; as suggested by animal models for type I diabetes: nonobese diabetic mice (NOD) or Bio-breeding (BB) rats, in which the disease is polygenic and environmental factors are im- portant [1]. The antigen-specific T-cell receptor (Tcr) plays a cen- tral role in immune recognition. It is composed of two chains, a and/8, linked together by a disulfide bridge and associated with CD3 complex [6]. They are formed by the rearrangement of various germline segments of variable (V), diversity (D), joining (J), and constant (C) regions. Differences in the germline configuration of T-cell receptor genes may affect the T-cell repertoire and the ability to react with islet antigens. Several reports studying associations between the Tcr/8 constant region and IDDM, using restriction frag- ment length polymorphism (RFLP) analysis, have shown contradictory results. Population studies have found positive [7-10] and negative [I 1, I2] results. In addi- 78 E. Mardnez-Naves et al. tion, family studies have found no association between Tcr a [13, 14] and Tcr/8 [14] loci and IDDM. It has been estimated that 20 haman V/8 subfamilies might exist. Two of them, V/88 and V/811, are in linkage disequilibrium and were used with the (:3 locus to de- fine informative polymorphisms [ 15, 16]. In the present work, we have studied RFLP distribu- tion of C/8, V/88, and V/811 and haplotype frequencies in IDDM patients and a healthy control population. MATERIALS AND METHODS Subjects. We studied 42 patients with type I diabetes, all of which were diagnosed before the age of 18 years. Fifty-one healthy blood donors were used as a control population. All patients and controls were Caucasians and lived in the city of Oviedo (Northern Spain) area. DNA extraction and hybridization. From each subject, 30 ml of peripheral blood was collected in 5% EDTA. High-molecular-weight DNA was extracted and di- gested with the restriction endonucleses BglII and BamHI following the manufacturer's instructions (Boehringer Manheim); 8 to 10/zg of DNA was loaded per lane in 0.7% agarose gels and then transferred onto nylon membrane filters (Hybond N, Amersham). Filters were prehybridized at 420(: in 50% formamide, 5 x sodium chloride-sodium phosphate-EDTA (SSPE) (1 x SSPE = 0.18 M NaCl, 10 mM phosphate, pH 7.5, 1 mM EDTA), 5 x Denhardt's solution (Denhardt's solu- tion, 1 x = 0.02% bovine serum albumin, 0.02% Fi- coil, 0.02% polyvinylpyrrolidone), salmon sperm DNA (40/zg/ml), and 0.1% SDS and then hybridized to an oligo-32P-labeled probe [17]. The filters were washed in 2 x SSPE, 0.1% sodium dodecyl sulfate (SDS) twice at 65°C for 15 min and then in 0.2 x SSPE at 65°(: once for 15 rain. Autoradiography was done with Dupont Cronex film between two Dupont lightning plus intensifying screens at -70°(2 for 2 to 8 days. DNA probes. Filters bearing DNA digested with BglII were hybridized with a cDNA-specific probe for C/8, filters with BamHI-digested DNA were hybridized with cDNA probes for V/88 and V/811 [18]. Probes were removed from blots by washing twice in boiling 0.5% SDS for 10 to 15 rain. Data analysis. Genotype and haplotype frequencies were determined in patients and control groups and compared using the chi-square test. C[3/V~8/V/811 hap- lotypes could be assigned in individuals who were dou- bly or triply homozygous for these three loci. Haplo- typic frequencies for V#8-V/311 were estimated VB8 C8 V811 kb kb ~!~ ~ kb ,o:~ ~ 2, . • , ~ Q g FIGURE 1 Autoradiograph of hybridization pattern for C~8, V38, and '4311 locus shown by homozygotes and heterozy- gotes. *, constant bands. according to Hill [19] using data of all individuals tested. RESULTS Each of the three Tcr probe/enzyme combinations de- fines diallelic RFLPs [15]. The Tcr C/3 probe with BglII defines bands at 10 and 9 kb, designated elsewhere as 9.3 and 8.6 Kb [7, 11]. Using BamHI, V/38 probe gives two polymorphic bands of 23 and 2 kb each, and V/811 defines two polymorphic bands of 25 and 20 kb in size (see Fig. 1). No extra or missing bands were observed in the IDDM group compared with the control population; thus, deletions or duplications in gene segments studied are unlikely. The genotype frequencies of the bands are shown in Table 1. No significant differences were ob- served in the distribution of the allele and genotype frequencies between IDDM patients and the control group. Genotype frequencies for each loci were not significantly different from those predicted by the Hardy-Weinberg equilibrium. Tcr /8 haplotype frequencies. The polymorphism gener- ated by the C/3, V/38, and V/311 alleles were used to define haplotypes [15, 16]. Haplotypes could be as- signed in individuals who were doubly or triply homo- zygous for these three loci. Forty-six haplotypes in the IDDM group and 62 in controls were established (see Table 2). To compare the distribution of haplotype fre- quencies between patients and controls, data for the four rarest haplotypes (23/25/10, 2/20/10, 23/25/9, and 2/20/9) were combined. No significant differences were seen between groups. To avoid the loss of information, including only T-Cell Receptor ~8-Chain Genes in 1DDM 79 TABLE 1 Distributions of genotype frequencies defined by C/3, V~8, and V/311 eDNA probes in IDDM patients and controls Cl3/Bglll V138/BamH1 V31 l/BmnHt Kb n (Frequency) Kb n (Frequency) Kb n (Frequency) IDDM Controls 10/10 12 (0.292) 23•23 12 (0.293) 25125 8 (0.190) 10/9 19 (0.463) 23/2 19 (0.463) 25•20 19 (0.452) 9/9 10 (0.244) 2/2 10 (0.244) 20/20 15 (0.357) N 41 N 41 N 42 10/10 11 (0.216) 23123 18 (0.353) 25125 8 (0.157) 10/9 24 (0.470) 23•2 20 (0.392) 25•20 24 (0.470) 9/9 16 (0.313) 2/2 13 (0.255) 20/20 19 (0.373) N 51 N 51 N 51 haplotypes from doubly or triply homozygous individ- uals, haplotypic frequencies for V138 and Vj811 were estimated taking into account the data of all individuals tested [19]. Distribution of these frequencies (Table 3) was compared between the IDDM and control groups and no significant differences were found. DISCUSSION Several works have reported an association between RFLPs of Tcr genes and autoimmune diseases such as multiple sclerosis, myasthenia gravis [20], membranous nephropathy [21], and Graves' disease. IDDM (7-9) has been reported to be associated with a BglII RFLP for the C~ gene segment. In another study with more diabetics [10], an association between a C# RFLP and IDDM was demonstrated only in individuals who were DR3/X. In contrast, Niven et al. [11] failed to confirm an association between any C/3 gene RFLP and IDDM in Caucasians and South Indian populations, even when they divided the patients in groups DR3/X, DR3/4, or TABLE 2 Haplotype frequencies defined by the C13, V/38, and VI311 probes in IDDM patients and controls Haplowpes V~81V~I IIC/3 IDDM Controls n (Frequency) n (Frequency) 23/25/10 1 (0.022) -1 (0.016) 2125110 9 (0.195) 8 (0.129) 23120110 12 (0.260) 18 (0.290) 2/20110 3 (0.065) 1 (0.016) 2312519 0 (0.00) 1 (0.016) 2/25/9 6 (0.130) 11 (0.177) 23/20/9 13 (0.282) 17 (0.274) 2•20•9 2 (0.043) 5 (0.080) Haplowpes defined 46 62 DR4/X. Field et al. [12] have found no association be- tween Q8 gene RFLP and IDDM and Concannon et al. [14], in a family study with multiple affected sibs, have found no evidence for linkage between Tcr genes and IDDM. In the pre.~ent work a comparison of allele and ge~o- type frequencies of 42 IDDM patients and a control population showed no association between any BgllI RFLP studied for the C3 gene segment. Furthermore, no differences were found between the distributions of allele and genotype frequencies of BamHI polymor- phism for the two variable gene segments in linkage disequilibrium: V~8 and V~11. The polymorphisms generated by these probes were used to define informa- tive haplotypes [15, 16]. Recently, Beall et al. [23] found a significantly different haptotype distribution of these alleles in a group of multiple sclerosis patients when compared with a healthy comrol population. We established 46 haplotypes in the IDDM group and 62 in the control Oopulation in our study, and no significant differences were found between haplotype distributions of each group. Haplotype frequencies for V~8-VjS11 were estimated on the data from all individ- uals tested and distributions of IDDM and the control group were not significantly different. TABLE 3 Haplotype frequencies estimated for V/38-V~I I alle,es HaploWpes V~8-V~I 1 IDDM Controls 23-25 0.013 0.021 23-20 0.511 0.528 2-25 0.402 0.371 2-20 0.074 0.080 Individuals tested 41 51 80 E. Martinez-Naves et al. The contradictory results of the different studies re- garding an association between Tcr a polymorphisms and IDDM may be based on ethnic and geographical differences between the distinct populations studied. An association between germline polymorphisms of Tcr genes and autoimmune diseases is more likely to involve V than C segments. In the present work, using two probes for subfamilies Va8 and va l 1, no positive association with an allele and IDDM was found. It is possible that polymorphisms in J or D regions or poly- morphisms other than RFLPs may affect susceptibility to this disease. In conclusion, our study found no associated RFLPs of Ca, Va8, and Va 11 Tcr gene segments with IDDM. ACKNOWLEDGMENT This work was supported by agrant fr~,m the Spanish Ministry of Health FISS no. 89/0370. E. Martfnez-Naves is a fellow of FPI (Spanish Ministry of Science and Education). O. Dom~nguez is a fellow of FISS. REFERENCES 1. Castafio L, Eisenbarth G8: Type I diabetes: A chronic autoimmune disease of human, mouse and rat. 2. Todd JA, Bell JI, McDevitt HO: HLA-DQfl gene con- tributes to susceptibility and resistance to insulin-depen- dent diabetes mellitus. Nature 329:599, 1987. 3. Iman K, d'Auroi L, Gobert M, Morin L, Lepage V, Deschamps I, Park MS, Degos L, Galibert F, Hors J: A combination of HLA-DQB Asp 57-negative and HLA DQA Arg 52 confers susceptibility to insulin dependent diabetes mellitus. J Clin Invest 85:1315, 1990. 4. Todd JA: Genetic control ofautoimmunity in type 1 dia- betes. Immunol Today 11:122, 1990. 5. Suarez BK, Eerdewgh P: Type 1 (insulin-dependent) dia- betes mellitus. Is there strong evidence tora non-HLA linked geae? Diabemlogia 20:524, 1981. 6. Wilson RK, Lai E, Concannon P, Barth RK, Hood LE: Structure, organization and polymorphism of murine and human T-cell receptor (x and/3 chain gene families. Im. munol Rev 111:149, 1988. 7. Ito M, Tanimoto M, Kamura M, Yoneda M, Morishima Y, Takatsuki K, Itatsu T, Saito M: Association of HLA- DR phenotypes and T-lymphocyte receptor B-chain re- gion RFLP with IDDM in Japanese. Diabetes 37:1633, 1988. 8. Hoover MC, Angelini G, Ball E, Stastny P, Marks J, Rosentock J, Raskin P, Ferrara GB, Tosi R, Capra DJ: HLA-DQ and T-cell receptor genes in insulin-dependent diabetes mellitus. Cold Spring Harbor Syrup Quant Biol 5:803, 1986. 9. Millward BA, Welsh KI, Leslie RDG, Pyke DA, Demain AG: T-cell receptor B-chain polymorphisms are associ- ated with insulin-dependent diabetes. Clin Exp Immunol 70:152, 1987. I0. Hoover ML, Black KE, Ball E, Alford M, Marks J, Statsny P, Capra DJ: Polymorphism of the human T-cell receptor c~ and B-chain genes and their relationship to insulin-dependent diabetes mellitus. In Dupont B (ed): lmmunobiology of HLA, vol 2. Immunogenetics and Histocompatibility. New York, Springer-Verlag, 1989, p 411. 11. Niven MJ, Caffrey C, Moore RH, Sachs JA, Mohan V, Festenstein H, Hoover ML, Hitman GA: T-cell receptor B subunit gene polymorphism and autoimmune disease. Hum lmmuno127:360, 1990. 12. Field LL, Bech-Hansen NT, Tobias R, McArthur RG: No association between T-cell receptor B-chain RFLP and insulin-dependent diabetes mellitus. Am J Hum Genet 43:A214, 1988. 13. Sheehy MJ, Meske LM, Emler CA, Rowe JR, Neme de Gimenez MH, Ingle CA, Chan A, Trucco M, Mak TW: Allelic T-cell receptor (x complexes have little or no influ- ence on susceptibility to type I diabetes. Hum Immunol 26:261, 1989. 14. Concannon P, Wright JA, Wright LG, Sy!vester DR, Spielman R8: T-cell receptor genes and insulin-depen- dent diabetes mellitus (IDDM): No evidence for linkage from affected sib pairs. Am J Hum Genet 47:45, 1990. 15. Concannon P, Gatti RA, Hood LE: Human T-cell recep- tor VB gene polymorphism. J Exp Med 165:1130, 1987. 16. Charmley P, Chao A, Concannon P, Hood L, Gatti RA: Haplotyping the human T-cell receptor B-chain gene complex by use of restriction fragment length polymor- phisms. Proc Natl Acad Sci USA 87:4823, 1990. 17. Feinberg AP, Wogelstein B: A technique for radiolabel- ling DNA restriction endonuclease fragments m high specific activity. Annal Biochem 137:266, 1984. 18. Concannon P, Pickering LA, Kung P, Hood LE: Diversity and structure of human T-cell receptor/3-chain variable region genes. Proc Natl Acad Sci USA 83:6598, 1986. 19. Hill WG: Estimation of linkage disequilibrium in ran- domly mating populations. Heredity 33:229, 1974. 20. Oksenberg JR, Sherrit M, Begovich AB, Erlich HA, Bernard CC, Cavalli-Sforza H,, 8teinman L: T-cell recep- tor Va and Ca alleles associated with multiple sclerosis and myasthenia gravis. Proc Natl Acad Sci USA 86:988, 1989. 21. Demaine AG, Vaughan RW, Taube DH, Welsh led: As- sociation of membranous nephropathy with T-cell recep- tor constant/3 .chain and immunoglobulin heavy chain switch region polymorphisms. Immunogenetics, 27:19, 1988. 22. Demaine AG, Welsh KI, Hawe DG, Willianms DG, Faird NR: The T-cell receptor//-chain genes are associ- ated with Graves' disease. Clin 8ci 72:62, 1987. 23. Beall SS, Concannon P, Charmley P, McFarland HI, Gatti RA, Hood LE, McFarlin, Biddison WE: The germline repertoire of T-cell receptor B-chain genes in patients with chronic progressive multiple sclerosis. J Neuroimmunol 21:59, 1989.