and (2000) 1, 501–503  2000 Macmillan Publishers Ltd All rights reserved 1466-4879/00 $15.00 www.nature.com/gene BRIEF COMMUNICATION New single polymorphisms in the coding region of human TNFR2: association with systemic lupus erythematosus

N Tsuchiya, T Komata, M Matsushita, J Ohashi and K Tokunaga Department of Human , Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan 113-0033

We recently reported the association of the allele coding for Arg at the position 196 (196R: nucleotide [nt] 587G) of tumor necrosis factor receptor 2 (TNFR2, TNF-R75) with systemic lupus erythematosus (SLE) in Japanese. In the present study, we completed the variation screening of the entire coding region of TNFR2. Three new single nucleotide polymorphisms within the coding sequence (cSNPs), as well as several variations within the , introns and 3Ј- (3ЈUTR), were identified. Among the new SNPs, nt168G, a synonymous substitution (K56K), was in tight linkage disequilibrium with nt587G. Two other cSNPs, nt543 (C→T) (P181P) and nt694 (G→A) (E232K), were not significantly associated with SLE. Thus, among the non-synonymous cSNPs, only nt587 (T→G) (M196R) was found to be significantly associated with SLE in Japanese. Genes and Immunity (2000) 1, 501–503.

Keywords: TNFR2; polymorphism; SLE; genetics; susceptibility

In spite of extensive studies by a number of investigators, Japanese, unrelated individuals living in Tokyo area. the susceptibility genes to systemic lupus erythematosus Variation screening was performed using PCR-single (SLE) largely remain to be determined.1 Previous reports strand conformation polymorphism (SSCP)6 and PCR- suggested the role for TNF␣ in the pathogenesis of preferential homoduplex formation assay (PHFA).7 Both human2 and murine3 SLE, and the results from - methods can detect any variations, including single wide screening indicated possible linkage of nucleotide polymorphisms (SNPs). The human TNFR2 1p36, where TNFR2 gene is located, with SLE.4,5 Based gene consists of 10 , among which exons 4, 9 and on such findings, we considered TNFR2 as a candidate 5Ј-portion of 6 were already analyzed.6 Each exon for a susceptibility gene to SLE. Through the initial was amplified using flanking intronic primer sets, except screening of three exons containing the previously for exons 6 and 10, which were divided into two frag- reported variation sites, we detected a significant associ- ments due to the length of the exons (Table 1). Thus, ation of the allele coding for a non-conservative change among the introns, only 10–80 flanking each of Arg (196R) for Met (196M) at codon 196 (nucleotide exon were screened. Nucleotide sequences of detected [nt] 587 [T→G]) with SLE.6 However, it was still unclear variations were determined by direct sequencing.6 whether the codon 196 polymorphism bears primary sig- New SNPs were identified at nt168 (A→G) within exon nificance, or it represents linkage disequilibrium with 2, nt543 (C→T) within exon 5 and nt694 (G→A) within other undefined polymorphism(s) of primary signifi- exon 6. Among these SNPs, nt694 (G→A) leads to the cance. In this study, we completed the variation screen- non-synonymous substitution within the extracellular ing of the entire TNFR2 coding region in 81 Japanese domain, E232K. patients with SLE and 258 healthy individuals, and exam- We next examined whether any of these coding ined the possibility that any of the other variations may sequence SNPs (cSNPs) is associated with the suscepti- be associated with the susceptibility to SLE. bility to SLE, using the case-control association analysis. The characteristics of the patients and controls are As shown in Table 2(a), 28 of 81 patients (35%) possessed described in the previous paper.6 All of them are at least one nt168G allele, as compared with 47 of 258 controls (18%). This difference was statistically significant (P = 0.002, odds ratio [OR] = 2.37, 95% confidence interval Correspondence: Naoyuki Tsuchiya, MD, PhD, Department of Human [CI] : 1.37–4.09). No significant association was observed Genetics, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, in the other two SNPs (Table 2(b)). Bunkyo-ku, Tokyo, Japan 113-0033. E-mail: tsuchiya-tkyȰumin.ac.jp To obtain information on the relationship between This study was supported by the Grant-in-Aid for Scientific nt168G/A and the previously reported nt587T/G,6 the Research (B) (11470505) from the Ministry of Education, Science, same sets of controls and patients were genotyped for the Sports and Culture. The nucleotide sequence data reported in this paper will appear in two cSNPs. Thirty of 81 patients (37%) with SLE pos- the DDBJ/EMBL/GenBank nucleotide sequence databases with the sessed nt587G, which was significantly increased as com- accession numbers AB030949-53. pared with 53 of 258 controls (21%) (P = 0.0026, New SNPs in TNFR2 N Tsuchiya et al 502 Table 1 Primers used in this study

Name Primer sequence Fragment size

promoter region TNFR2proF 5Ј-AAGCACCATCCTTGCAGGCT-3Ј 385 bp TNFR2proR 5Ј-AGATGAACAGAGGCGGGAG-3Ј exon1 TNFR2E1-F 5Ј-AGCGGAGCCTGGAGAGAAGG-3Ј 202 bp TNFR2E1-R 5Ј-ACATGCGGGGCGGCTGTC-3Ј exon2 TNFR2E2-F 5Ј-ATCAGGCATGGCAGAACCCA-3Ј 219 bp TNFR2E2-R 5Ј-TGTACACACACGCTCCTCCA-3Ј exon3 TNFR2E3-F 5Ј-ATGAGCCAGGGTCCTGGCA-3Ј 235 bp TNFR2E3-R 5Ј-AAGTTGGAGGCAGGGGTGTA-3Ј exon5 TNFR2E5-F 5Ј-CGCAGAGTGTCTGAGTGGTT-3Ј 178 bp TNFR2E5-R 5Ј-CTCCCTGCTCCTCCAGAAC-3Ј exon6 TNFR2E6-F 5Ј-CAGCCAGTGTCCACACGAT-3Ј 182 bp TNFR2E6-R 5Ј-GACAGGCAGACAGAAGGAGT-3Ј exon7 TNFR2E7-F 5Ј-TGGCCCCTGGTACATTTGA-3Ј 179 bp TNFR2E7-R 5Ј-CCTGAACAAGTGGATGAAGG-3Ј exon8 TNFR2E8-F 5Ј-CAGATGTGCCTGAGGAAGTC-3Ј 165 bp TNFR2E8-R 5Ј-ACTGCTTCCTCTGTGACAGC-3Ј exon10 TNFR2E10–1F 5Ј-GAATCTGCATCTTGGGCAGG-3Ј 278 bp TNFR2E10–1R 5Ј-GTCTCCAGCTGTGACCGAAA-3Ј TNFR2E10–2F 5Ј-ATTCCAGCCCCTCGGAGT-3Ј 283 bp TNFR2E10–2R 5Ј-TTGGCCCAGAAAGAGCCTCA-3Ј

Exons 4, 9 and 5Ј-portion of exon 6 were analyzed in the previous paper.6

OR = 2.28). In addition, it was revealed that almost all (C→T)] while the samples containing exon 2 and exon 6 individuals possessing nt168G had nt587G. We therefore SNPs were sequenced. Among the sequenced eight carried out haplotype estimation based on the EH alleles, all three alleles possessing nt168G and nt587G program.8 Relative linkage disequilibrium values were were found to possess these intronic nucleotide substi- estimated from the genotypes of the healthy Japanese tutions. In addition, nt1409 (G→A) was detected within individuals using the difference between observed and the 3Ј-untranslated region (3ЈUTR) in only one patient expected frequencies of allele combinations, standardized with SLE, but not in healthy individuals. by the maximum possible value.9 As shown in Table 3, The association of SLE with TNFR2-196R11 or another these two SNPs were in tight linkage disequilibrium; SNP within the 3ЈUTR12 was not observed in the Cauca- nt168A was positively associated with nt587T, while sian populations. It is therefore possible that this SNP nt168G was positively associated with nt587G. Such link- might possess a significant effect in the Japanese, but not age disequilibrium was also observed in the patients. in the Caucasian populations. Further studies on other Two variations, −1413 (A→C) and −1120 (G→C), were populations, especially other Asian populations, will be previously reported in the promoter region of TNFR2.10 of particular interest. However, population-based studies to estimate the allele It is interesting to note that, although located within frequencies have not been reported. In order to test the the same exon 6, SNP at nt587 is associated with SLE, possibility that the association of nt168G–nt587G allele while SNP at nt694 is not. This is because nt694A (232K) with SLE derived from linkage disequilibrium with pro- seems to be present almost exclusively in the allele carry- moter polymorphisms, we carried out PCR-SSCP analysis ing nt587T (196M) (data not shown), suggesting that of the region encompassing the two known variation sites nt694A originated from the allele carrying nt587T after (−1458 to −1073) in 81 Japanese patients with SLE and 129 the divergence of nt587G allele. Thus, it will be important healthy individuals randomly selected from 258 controls. to take the evolutionary pathway of SNPs into consider- Fifteen individuals were also examined by direct sequen- ation when any disease association is examined using a cing. A new single nucleotide substitution was detected huge number of SNPs in the future. at position −1194 (G→A) in one patient with SLE. On the In conclusion, we identified several new variations in other hand, all samples were homozygous for both of the the coding region, promoter and introns of TNFR2. previously reported −1413C and −1120C variations. Thus, Association with SLE was observed only in nt168G and at least in the Japanese population, these nucleotides are nt587G SNPs, among which the latter results in amino considered to constitute the common TNFR2 promoter acid substitution (M196R). Although sufficient experi- allele. mental data is not available concerning the biological Furthermore, single nucleotide substitutions were effects of M196R substitution, it is possible that this non- detected in intron 2 [178+13 (A→G)] and intron 5 [552−28 conservative substitution within the fourth cysteine-rich

Genes and Immunity New SNPs in TNFR2 N Tsuchiya et al 503 Table 2 New single nucleotide polymorphism of TNFR2 in the variations in other genes linked with TNFR2 nt587T can- Japanese patients with SLE and controls not be excluded. Further studies should focus on the (a) Positivity and genotype frequency of TNFR2-nt168A/G (56K/K) effect of the M196R substitution as well as on the screen- in SLE and controls ing of other genes located close to TNFR2. SLE Controls ␹2 P Odds (n = 81) (n = 258) ratio Acknowledgements allele positivity The authors are indebted to Dr Tetsufumi Inoue, Dr nt168G+ 28 (35%) 47 (18%) 9.57a 0.002 2.37 Keiko Ishihara, Dr Shigeto Tohma, Dr Naoto Hirose, Dr nt168A+ 80 (99%) 253 (98%) NS Takeshi Suzuki, Dr Hiroshi Furukawa (Department of genotype frequency Allergy and Rheumatology, University of Tokyo) and Dr nt168G/G 1 ( 1%) 5 ( 2%) 11.10b 0.004 Kunio Matsuta (Matsuta Clinic) for the recruitment of the nt168A/G 27 (33%) 42 (16%) patients, to Michiko Shiota (Department of Human Gen- nt168A/A 53 (66%) 211 (82%) etics, University of Tokyo) for technical assistance, and to Dr Chika Morita and Dr Takahiko Horiuchi (The First (b) Frequencies of other TNFR2-SNPs in SLE and controls Department of Internal Medicine, Kyushu University) for helpful discussions. SLE Controls P (n = 80) (n = 91) References exon5 nt543 (C→T) (P181P) genotype frequency 1 Tsao BP. Genetic susceptibility to lupus nephritis. Lupus 1998; nt543T/T 0 (0%) 0 (0%) NS 7: 585–590. nt543C/T 4 (5%) 3 (3%) 2 Elliott MJ, Maini RN, Feldmann M et al. Repeated therapy with ␣ nt543C/C 76 (95%) 88 (97%) monoclonal to tumor necrosis factor (cA2) in patients with . Lancet 1994; 344: 1125–1127. 3 Jacob CO, Lee SK, Strassmann G. Mutational analysis of TNF- SLE Controls P ␣ Ј (n = 81) (n = 205) gene reveals a regulatory role for the 3 -untranslated region in the genetic predisposition to lupus-like . J Immunol 1996; 156: 3043–3050. exon6 nt694 (G→A) (E232K) 4 Gaffney PM, Kearns GM, Shark KB et al. A genome-wide search genotype frequency for susceptibility genes in human systemic lupus erythematosus nt694A/A 0 (0%) 0 (0%) NS nt694G/A 6 (7%) 16 (8%) sib-pair families. Proc Natl Acad Sci USA 1998; 95: 14875–14879. nt694G/G 75 (93%) 189 (92%) 5 Shai R, Quismorio FP Jr, Li L et al. Genome-wide screen for systemic lupus erythematosus susceptibility genes in multiplex families. Hum Mol Genet 1999; 8: 639–644. ad.f. = 1, bd.f. = 2 6 Komata T, Tsuchiya N, Matsushita M, Hagiwara K, Tokunaga K. Association of tumor necrosis factor receptor 2 (TNFR2) poly- morphism with susceptibility to systemic lupus erythematosus. Table 3 Linkage disequilibrium between nt168A/G (56K/K) and nt587T/G (196M/R) estimated from the genotypes of 258 healthy Tissue 1999; 53: 527–533. Japanese individuals 7 Matsushita M, Tsuchiya N, Oka T, Yamane A, Tokunaga K. New variations of human SHP-1. Immunogenetics 1999; 49: 577–579. Haplotype HF LD RLD ␹2 P 8 Terwilliger JD, Ott J. Handbook of Human Linkage Analysis. Johns Hopkins University Press: Baltimore, 1994, pp 188–193. + + 168A-587T 0.886 0.086 0.93 9 Lewontin RC. The interaction of selection and linkage. I. General 168A-587G 0.014 −0.086 −0.93 − − considerations, heterotic models. Genetics 1964; 49: 49–67. 168G-587T 0.006 0.086 0.93 10 Santee SM, Owen-Schaub LB. Human tumor necrosis factor + + Ͻ −10 168G-587G 0.095 0.086 0.93 414.8 10 receptor p75/80 (CD120b) and promoter charac- terization. J Biol Chem 1996; 271: 21151–21159. HF, haplotype frequency; LD, linkage disequilibrium parameter; 11 Al-Ansari AS, Ollier WER, Villarreal J, Ordi J, Teh L-S, Hajeer RLD, relative linkage disequilibrium value. AH. Tumor necrosis factor receptor II (TNFRII) exon 6 polymor- phism in systemic lupus erythematosus. Tissue Antigens 2000; 55: 97–99. domain in the extracellular region has some effect on the 12 Sullivan KE, Piliero LM, Goldman D, Petri MA. A TNFR2 3Ј TNF binding affinity and/or signaling pathway. On the flanking region polymorphism in systemic lupus ery- other hand, the possibility of the primary role of the thematosus. Genes Immun 2000; 1: 225–227.

Genes and Immunity