and Immunity (2000) 1, 330–337  2000 Macmillan Publishers Ltd All rights reserved 1466-4879/00 $15.00 www.nature.com/gene Single nucleotide polymorphisms in the coding regions of human CXC- receptors CXCR1, CXCR2 and CXCR3

H Kato, N Tsuchiya and K Tokunaga Department of Human Genetics, Graduate School of Medicine, University of Tokyo, Japan

Chemokines and their receptors have critical roles in inflammatory and immunological responses, and thus their genetic contribution to various human disorders needs investigation. In this study, systematic variation screening of the entire coding regions of CXCR1 (IL8RA), CXCR2 (IL8RB) and CXCR3 was carried out, using genomic DNA from a large number of Japanese healthy individuals and patients with rheumatic diseases. In addition to the previously reported variations in CXCR1 and in CXCR2, two non-synonymous, two synonymous substitutions and one nonsense mutation of CXCR1, one non-synonymous and two synonymous substitutions of CXCR2, two non-synonymous substitutions of CXCR3 were newly identified. The common single nucleotide polymorphisms (SNPs) at CXCR1 codon 827 and CXCR2 codon 786 were in strong linkage disequilibrium. In addition, familial analysis indicated that human CXCR3 is located on X. No significant association was observed between the variations and the tested rheumatic diseases. However, CXCR variations identified in this study will provide valuable information for the future studies in medical sciences as well as in human genetics. Genes and Immunity (2000) 1, 330–337.

Keywords: CXCR1; CXCR2; CXCR3; cSNP; haplotype; rheumatic diseases

Introduction Woude syndrome, type 2, and neoplastic diseases such as rhabdomyosarcoma and uterine leiomyomata.5,11,12 Chemokine receptors belong to the G- coupled On the other hand, CXCR3 was identified as a receptor (GPCR) family, which have seven transmem- for T--specific . It is strongly brane regions and signal through heterotrimeric G-pro- expressed in IL-2-activated T , but not in 1 teins. Among them, CXCR1 (IL8RA) and CXCR2 (IL8RB) freshly isolated peripheral blood leukocytes or related were identified as receptors for interleukin-8 (IL-8), are cell lines.13 However, its intense expression was shown both expressed in , and have 77% amino acid in T cells isolated from synovial fluid from patients with 2,3 identity with each other. CXCR1 binds only two ELR- RA, from tissues of chronically inflamed vaginal and CXC chemokines: IL-8 and granulocyte chemotactic pro- colonic mucosa, or from cerebrospinal fluid from patients tein (GCP)-2, while CXCR2 binds other ELR-CXC chemo- with .14,15 Activated Th1/Tc1 subsets 1 kines as well. CXCR1, CXCR2 and IL8RBP, the pseudo- have been shown to express CXCR3 and CCR5.14,16,17 bearing greater similarity to CXCR2, are clustered Genetic contribution of polymor- 4,5 at chromosome 2q35. Physical distances among them phisms to several human disorders has been reported, have not yet been clear. These genes are considered as such as CCR5-⌬32 allele associated with protective effect candidate genes for several human disorders. The from human immunodeficiency virus (HIV)-1 infection.18 mapped susceptibility loci for rheumatoid arthritis (RA), In autoimmune diseases, a significant association was systemic lupus erythematosus (SLE), insulin-dependent shown between CCR5-⌬32 allele and the clinical and diabetes mellitus (IDDM) and juvenile amyotrophic lat- immunological characteristics of RA, or between CCR2– 6–10 eral sclerosis include 2q35. Moreover, structural 64I allele and the susceptibility to IDDM.19,20 To assess abnormalities of this chromosomal region have been the genetic contribution of other chemokine receptors to reported in Waardenburg syndrome, type 1, van der human diseases, polymorphisms in these genes need to be screened and registered as the essential information. In our previous study, we hypothesized that genetic Correspondence: Dr Naoyuki Tsuchiya, Department of Human Genetics, Graduate School of Medicine, University of Tokyo, 7–3-1 Hongo, Bunkyo- variations of chemokine receptors CCR3 and CCR4, pref- 16,17 ku, Tokyo, Japan 113–0033. E-mail: tsuchiya-tkyȰumin.ac.jp erentially expressed in the Th2 subset of lymphocytes, Grant support: This study was supported by the Grant-in-Aid for might be associated with the susceptibility to rheumatic Scientific Research (B11470505) from the Ministry of Education, diseases.21 Since the imbalance of Th1/Th2 cells had been Science, Sports and Culture. implicated in autoimmune rheumatic diseases, it was Sequence Accession numbers: The nucleotide sequence data possible that genetic variations of CCR3 or CCR4 might reported in this paper have been submitted to the DDBJ/EMBL/GenBank nucleotide sequence databases and have cause the imbalance and might influence the suscepti- 22 been assigned the Accession number AB032728 – AB032738. bility to the diseases. However, although several new Received 18 January 2000; revised and accepted 1 March 2000 variations were detected, no association was observed cSNPs in human CXCRs H Kato et al 331 between the susceptibility to rheumatic diseases and any CXCR2 in Japanese, since they were present in 97.9% and of the variations. On the other hand, as for Th1 subset in 89.3% of healthy population, respectively. Other markers, no attempts have been reported on polymor- sequences were regarded as variants. As shown in phisms in CXCR3, while a number of reports are present Table 1, all of them were single nucleotide substitutions. for CCR5.23 In CXCR1, two non-synonymous and two synonymous Thus far, several genomic DNA or cDNA sequences of substitutions were detected. The non-synonymous substi- CXCR1 and CXCR2 have been published.2–4,24–28 How- tution, 827 G → C (S276T), was present in 19.8% of heal- ever, they have a difference at one nucleotide in CXCR1, thy population, while the two synonymous substitutions, 827 G/C (counting from ATG initiation codon) coding 741 C → T (V247V) and 915 C → T (Y305Y) were rare for amino acid 276 S/T. Similarly, in CXCR2, two variations. In CXCR2, one non-synonymous and two syn- sequences have been published possessing 786C or T, onymous substitutions were detected. 786 C → T (L262L) leading to synonymous substitution encoding amino acid was the most frequent variant observed in 52.5% of heal- 262L. In addition, restriction fragment length polymor- thy individuals. In CXCR3, two non-synonymous substi- phism (RFLP) in the region encompassing CXCR2 gene tutions were detected. 875 G → A (R292Q) was detected was reported, and a haplotype containing the RFLP and in one patient with SLE and one healthy individual, and GT dinucleotide repeat number polymorphism in the 1087 G → A (A363T) was detected in only one patient natural resistance associated protein-1 gene with scleroderma (SSc). Both patients were female, and (NRAMP1) promoter region was demonstrated to be were heterozygous for the common allele and the substi- linked with the susceptibility to RA.7,29 However, it has tution allele. Hence, CXCR3 is considered to be highly not been shown whether the CXCR2 RFLP site is pos- conserved, at least in the Japanese population. itioned within or outside of CXCR2 gene. Systematic In addition, in the process of primer design for CXCR3 screening of the entire coding sequences has not been exon 1, we confirmed that CXCR3 gene has one intron reported for CXCR1, CXCR2 and CXCR3. between exon 1 and exon 2, the size of which is approxi- In this study, we attempted to detect variations in the mately 1 kb (data not shown). One polymorphic site was entire coding regions of human CXCR1, CXCR2, CXCR3, incidentally detected within intron 1 [+235 (G/A)] (The estimated each allele frequency in Japanese, and exam- nucleotide sequence data of the 5′ portion of intron 1 ined the possible association between their variations and have been submitted DDBJ/EMBL/GenBank database; the susceptibility to rheumatic diseases. accession numbers: AB032737–AB032738).

Results Analysis of association between detected variations and rheumatic diseases Identification of new variations of CXCRs The detected variation sites were subsequently examined Several new nucleotide sequence variations, in addition for the possible association with rheumatic diseases using to the previously reported ones, were detected using case-control analysis. As summarized in Table 1, no sig- polymerase chain reaction (PCR)-single-strand confor- nificant association between the positivity of these vari- mation polymorphism (SSCP) method. The representa- ations and any of the tested rheumatic diseases was tive SSCP patterns are shown in Figure 1. From the observed. Among the variations, CXCR1-827G/C and observed frequency of each sequence, we considered CXCR2-786C/T were frequently observed and were con- CXCR1-827G (coding for 276S) and CXCR2-786C (262L) sidered as single nucleotide polymorphisms within the as the common standard sequences of CXCR1 and coding sequence (cSNPs). For these cSNPs, genotype fre-

Figure 1 SSCP patterns of amplified fragments of CXCR1 (a, b), CXCR2 (c, d), and CXCR3 (e). Each PCR-SSCP condition was shown in Table 5. (a) CXCR1–5: Lane 1, CXCR1-741C/T, 827G/G, Lane 2, 741C/C, 827C/C, Lane 3, 741C/C, 827G/C, Lane 4, 741C/C, 827G/G. (b) CXCR1–6: Lane 1,4,CXCR1-915C/C, 1003C/C, Lane 2, 915C/C, 1003C/T, Lane 3, 915C/T, 1003C/C. (c) CXCR2–2: Lanes 1, 2, 4, CXCR2- 238C/C, Lane 3, 238C/T. (d) CXCR2–4: Lane 1, CXCR2-768C/C, 786C/C, Lane 2, 768C/T, 786C/C, Lane 3, 768C/C, 786T/T, Lane 4, 768C/C, 786C/T, Lane 5, 768C/T, 786C/T. (e) CXCR3-exon2–5: Lane 1, CXCR3-875G/A, 1087G/G (heterozygote, female), Lane 2, 875G/G, 1087G/A (heterozygote, female), Lane 3, 875G, 1087G (hemizygote, the father), Lane 4, 875A, 1087G (hemizygote, the son [the proband]), Lane 5, 875G/A, 1087G/G (heterozygote, the mother). These SSCP patterns (Lanes 3–5) indicated that CXCR3 gene was located on chromosome X, not on autosome. 875G/G, 1087G/G (homozygote, female) showed identical pattern to Lane 3.

Genes and Immunity cSNPs in human CXCRs H Kato et al 332 Table 1 Detected nucleotide substitutions and positivities among the patients with rheumatic diseases and healthy individuals (% in parentheses)

Substitutions Control RA SLE SSc BD Sjs

nucleotidea AAa position n = 242 n = 146 n = 80 n = 14 n = 12 n = 12

CXCR1 741(C→T) V247V TM6 0 (0) 0 (0) 1 (1.3) 0 (0) 0 (0) 0 (0) 827(G→C)b S276T EC3 48 (19.8) 30 (20.5) 16 (20.0) 1 (7.1) 2 (16.7) 2 (16.7) 915(C→T) Y305Y TM7 0 (0) 1 (0.7) 0 (0) 0 (0) 0 (0) 0 (0) 1003(C→T) R335C C-tail 2 (0.8) 4 (2.7) 0 (0) 0 (0) 0 (0) 0 (0)

CXCR2 238(C→T) R80C IC1 2 (0.8) 1 (0.7) 0 (0) 1 (7.1) 0 (0) 1 (8.3) 768(C→T) V256V TM6 6 (2.5) 5 (3.4) 1 (1.3) 0 (0) 0 (0) 0 (0) 786(C→T)b L262L TM6 127 (52.5) 85 (58.2) 44 (55.0) 6 (42.9) 6 (50.0) 6 (50.0)

CXCR3 875(G→A) R292Q EC3 1 (0.4) 0 (0) 1 (1.3) 0 (0) 0 (0) 0 (0) 1087(G→A) A363T C-tail 0 (0) 0 (0) 0 (0) 1 (7.1) 0 (0) 0 (0)

AA, amino acid; TM, transmembrane region (eg, TM6 stands for the sixth transmembrane region); EC, extracellular loop region; IC, intra- cellular loop region; C-tail, the intracellular carboxy-terminal tail region; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SSc, systemic sclerosis (scleroderma); BD, Behc¸et disease; Sjs, Sjo¨gren’s syndrome. aNumbering starts at the ATG translation initiation. bThese sequence variations were previously reported.2,25

quencies were compared between patients and controls. (10/575, 1.7%), which was not significantly different from Based on the similarity of the pathogenesis, patients with the positivity among the controls, excluding the possi- SLE and Sjo¨gren’s syndrome (Sjs) were grouped together bility of weak association. In affected sibs, no individuals for this analysis. Although CXCR1-827C/C was not except for the previously mentioned two sib-pairs had present in RA, significant difference in the genotype fre- 1003T allele. Incidentally, two new variations were quency was observed neither between RA and controls, detected during the screening of the additional samples: nor between SLE + Sjs and controls (Table 2). In addition, 915 C → A (Y305-stop) and 916 G → A (A306T) both pos- association with clinical features or disease severity was itioned in the seventh transmembrane region (TM7). Both not observed (data not shown). were rare mutations detected in only one sample as a het- CXCR1-1003T (335C) allele was present with slightly erozygote. higher frequency in the patients with RA (4/146, 2.7%) than that in healthy individuals (2/242, 0.8%). In Assignment of human CXCR3 gene on chromosome X addition, two of the four patients possessing 1003T allele During the analysis of CXCR3, one healthy male individ- had affected sib with RA, who also possessed 1003T ual was noted to possess only 875A (292Q) allele, and no allele. At this point, it remained possible that 1003T allele common allele (875G) (Table 1). Two conflicting chromo- is weakly associated with the susceptibility to RA, somal localizations have been reported for human especially in the patients who have family members CXCR3, 8p12-p11.2 and Xq13.30,31 Because of the low affected with RA. Thus, a large number of non-familial allele frequency of 875A, it was considered highly patients as well as affected sib-pairs with RA were unlikely that this individual was homozygous for the additionally screened for only CXCR1–6 segment, 875A allele of CXCR3 located on 8p. In order to confirm encompassing 1003C/T site. In a total of 575 unrelated that CXCR3 is located on , the genotypes patients with RA, 10 patients possessed 1003T allele of his parents were examined. As shown in Figure 1e (Lanes 3–5), his father had only the common allele, and his mother were heterozygous for the common allele and Table 2 Genotype frequencies of CXCR1-827 G/C and CXCR2-786 C/T SNPs in patients and healthy individuals (% in parentheses) 875A allele. If the proband was homozygous for 875A allele on the autosome, his father should possess 875A RA SLE + Sjsa Controls allele. Thus, our data indicated that human CXCR3 gene n = 146 n = 91 n = 242 is located on chromosome X.

CXCR1-827 G/G 116 (79.5) 73 (80.2) 194 (80.1) Allele frequencies of each IL8Rs (CXCR1 and CXCR2) G/C 30 (20.5) 17 (18.7) 43 (17.8) in Japanese C/C 0 (0) 1 (1.1) 5 (2.1) In this study, we detected two polymorphic sites in CXCR1 and three in CXCR2 in the healthy population CXCR2-786 C/C 61 (41.8) 40 (44.0) 114 (47.1) (Table 1). Since no subjects possessed two or more C/T 65 (44.5) 44 (48.4) 103 (42.6) CXCR1 variations, three CXCR1 alleles are considered to T/T 20 (13.7) 7 (7.7) 25 (10.3) be present in the healthy individuals. Allele frequencies calculated by direct counting in Japanese are shown in Significant difference was observed neither between RA and con- trols, nor between SLE, Sjo¨gren’s syndrome (Sjs) and controls (␹2 Table 3A. In the case of CXCR2, two polymorphic sites, analysis from 2 × 3 contingency tables). aPatients with SLE or Sjs. 768 C/T and 786 C/T, were present in the same PCR One patient fulfilled classification criteria for both SLE and Sjs. fragment. Only three of the four possible combinations

Genes and Immunity cSNPs in human CXCRs H Kato et al 333 Table 3 Frequencies of CXCR1 and CXCR2 alleles in 242 Japanese estimated haplotype frequencies were not significantly healthy individuals different among controls, patients with RA and patients with SLE or Sjs (Table 4B, 4C). (A) CXCR1 allele frequencies

827 1003 Allele Frequency Discussion G C 0.886 In this study, we carried out a variation screening of the C C 0.110 entire coding regions of human CXCR1, CXCR2, CXCR3, G T 0.004 and detected several new variations. GPCRs containing chemokine receptors are predominantly intronless within (B) CXCR2 allele frequencies coding regions and often they are clustered in , and therefore, they are suggested to have been 238 768 786 Allele Frequency duplicated by retroposition and by recombination events.32,33 When compared with previously published C C C 0.674 C C T 0.310 sequences of various chemokine receptor family genes, C T C 0.012 we found that the identical amino acid substitutions are T C T 0.004 present at the same or similar positions of the different receptors: CXCR1-827G/C (276S/T) and CCR3-827G- 828C/827C-828G (276S/T); CXCR3-875G/A (292R/Q) were observed; 768C–786C, 768C–786T, 768T–786C and CCR3-824G/A (275R/Q); CXCR1-741C/T (247V/V) (Figure 1d). Among all tested samples, five individuals and CXCR2-768C/T (256V/V). Considering that these (two controls and three patients) were 238 C/T: hetero- variations are not so common, it seems unlikely that the zygous for R80C substitution. Four of them were 786 variations had already existed before the gene family T/T, and the other one was 786 C/T. Based on these emerged and branched off by duplication events from genotypes, it was deduced that four alleles are present ancestral gene(s). Therefore, the presence of these vari- for CXCR2. The estimated allele frequencies in Japanese ations may indicate a mechanism of convergent evolution healthy individuals are shown in Table 3B. from the viewpoint that identical substitutions indepen- dently took place at similar positions. Such a hypothesis Linkage disequilibrium between cSNPs of CXCR1 remains speculative and should be tested by further and CXCR2 analyses, such as studies on other populations or even Because CXCR1, CXCR2 and the homologous pseudo- other species, and screening for the variations of other gene (IL8RBP) are clustered at human chromosome GPCRs. 2q35,4,5 it was considered likely that cSNPs of these genes Conflicting observations have been reported concern- are in linkage disequilibrium. To test this possibility, link- ing the chromosomal localization of human CXCR3 gene. age disequilibrium was examined using the genotypes of CXCR3 was initially reported to localize at 8p12-p11.2 by the 242 healthy population. Using the data at the most fluorescence in situ hybridization (FISH) using the gen- polymorphic CXCR1-827 and CXCR2-786 cSNPs, it was omic DNA clone GPR9 as the hybridization probe.30 demonstrated that these two cSNPs are in strong linkage However, it was later reported to be located on chromo- disequilibrium (␹2 = 104.6, P Ͻ 0.0001, Table 4A). The some X by PCR analysis using hamster/human hybrid lines,34 and subsequently localized at Xq13 by FISH.31 Table 4 Frequency of estimated haplotypes formed by CXCR1 Our data of SSCP analysis strongly supported the latter, codon 827 and CXCR2 codon 786 SNPs although the possibility that the father (shown in Figure 1e, Lane 3) has a deletion in one autosome (null Haplotype HFa LDb RLDc allele), and that the null allele was transmitted to the pro- band, cannot be completely excluded. Because human CXCR1-827 CXCR2-786 chemokine receptor (HCR) clone was identified at Xq13 as well,35 the localization of CXCR3 at Xq13 suggests the (A) Healthy individuals (n = 242) G C 0.678 0.067 0.897 presence of a novel mini-cluster region of chemokine G T 0.212 −0.067 −0.897 receptor genes at this chromosomal position. Recently, C C 0.008 −0.067 −0.897 mouse CC chemokine, secondary lymphoid tissue C T 0.102 0.067 0.897 chemokine (SLC), was shown to serve as a functional = ligand for mouse CXCR3, though human SLC does not (B) RA (n 146) have such a function.34,36 In addition, certain CC chemo- G C 0.632 0.057 0.869 G T 0.265 −0.057 −0.869 kines, especially CCR3 ligands such as eotaxin, show C C 0.009 −0.057 −0.869 higher affinity to CXCR3 than CXC chemokines other C T 0.094 0.057 0.869 than CXCR3 agonists, and act as antagonists in human.37 Thus, based on the chromosomal localization, the pres- = (C) SLE or Sjo¨gren syndrome (n 91) ence of an intron between the separated coding regions, G C 0.674 0.064 0.901 and other several unique functions, CXCR3 is considered G T 0.221 −0.064 −0.901 C C 0.007 −0.064 −0.901 to hold a unique position among the chemokine recep- C T 0.097 0.064 0.901 tors. It is interesting to examine the possible association between the susceptibility to the diseases such as asthma aHF, haplotype frequency; bLD, linkage disequilibrium parameter; and the CXCR3 variations, which might influence cRLD, relative linkage disequilibrium value. Th1/Th2 balance or antagonists’ binding to CXCR3,

Genes and Immunity cSNPs in human CXCRs H Kato et al 334 because the antagonists serve as the CCR3 agonists to to the generally accepted criteria.44–48 All examined indi- Th2 function. viduals were unrelated Japanese living in Tokyo area, the Human IL8R genes (CXCR1, CXCR2, IL8RBP) are clus- central part of Japan. The area has been shown to be tered at chromosome 2q35, approximately 130 kb telo- homogeneous with respect to genetic background,49 per- meric to NRAMP1 gene and 10 ෂ 30 kb centromeric to the mitting the case-control approach employed in this villin1 gene (VIL1), which encodes a Ca2+-regulated actin- study. Informed consent was obtained from the subjects. binding protein.29,38 The mouse homologue of NRAMP1, Furthermore, for the screening of a certain polymorphic Ity/Lsh/Bcg, was found to control susceptibility to the site, additional patients with RA were recruited for the intracellular pathogens Salmonella typhimurium, Leish- case-control association analysis or familial analysis; in mania donovani, and Mycobacterium bovis BCG.39 Besides, total, the genotypes at the particular site were determined NRAMP1 was shown to upregulate mRNA of ELR-CXC in 575 unrelated RA samples and 20 affected sib-pair fam- chemokines.40 Shaw et al7 reported positive association ilies with RA (19 families with two RA sibs, one family between NRAMP1-IL8RB(CXCR2)-VIL1 haplotype and with three RA sibs). susceptibility to RA by the identity by descent analysis using affected sib pairs. Subsequently, John et al41 sug- Genomic DNA isolation gested a role for NRAMP1 polymorphism in a subset of Genomic DNA from each individual was purified from patients who do not possess HLA susceptibility alleles. peripheral blood leukocytes using a commercially avail- In the present case-control study, no association of IL8Rs able DNA extraction (QIAamp blood kit, Qiagen, variations and RA was observed in Japanese, suggesting Hilden, Germany) following the manufacturer’s protocol. that IL8Rs may not be responsible for the reported associ- ation with RA. It should be noted that none of the CXCR2 PCR-SSCP analysis variations detected in this study corresponded to the The variation screening was performed using PCR-SSCP reported CXCR2 RFLP site,29 possibly because the RFLP method. The primers used for PCR and the PCR-SSCP site was positioned out of the coding region of CXCR2, conditions are summarized in Table 5. The specific or such variation was too rare to be detected in Japanese. primer pairs were designed according to the published Although we focused on rheumatic diseases in this study, genomic DNA sequences: CXCR1 (GenBank accession it will be interesting to examine the variations for the sus- number: L19592),24 CXCR2 (M99412),28 GPR9 (U32674).30 ceptibility to the intracellular pathogens such as tubercu- Both CXCR1 and CXCR2 have no intron within their losis, because of the close relationship between polymor- coding regions of 1053 bp and 1083 bp, respectively. phisms of NRAMP1 and those of IL8Rs. GPR9 corresponds to the putative CXCR3 exon 2 and Although the frequency is rare, some of the variations consists of partial CXCR3 coding region [CXCR3 (5–368)] identified in this study seem to be potentially important of 1095 bp and the 3′ untranslated region (3′UTR).13 Since in terms of receptor functions. For example, CXCR1- the optimal size of PCR product for SSCP was deter- R335C and CXCR2-R80C may confer potentiality of mined to be 200–400 bp in the previous studies,50 specific homo- or heterodimerization through disulfide primer sets were designed to amplify six overlapping bonds,42,43 while CXCR3-A363T increases the number of segments of the entire coding region of CXCR1. CXCR2 potential phosphorylation sites in the C terminal tail. was divided into five fragments and CXCR3 exon 2 into Functional analysis as well as larger scale association five as well. Because no flanking sequence data were studies might reveal important role for such rare vari- available to design the primer set for exon 1, the sense ations in the future. In addition, the linkage disequilib- primer was designed within 5′ untranslated region rium between the haplotype of NRAMP1-IL8RB-VIL1 and (5′UTR) based on the published cDNA sequence of the nucleotide sequence variations detected in this study CXCR3 (X95876)13 and the antisense primer within intron needs to be analyzed both in Japanese and in other popu- 1 based on the results of direct sequencing (data not lations. shown). In conclusion, we detected new variations of CXCR1, PCR was carried out using commercially available CXCR2, and CXCR3 in Japanese, and some of which were reagents (GeneAmp and AmpliTaq Gold, Perkin-Elmer, considered to form specific haplotypes. Such information Norwalk, CT, USA). After preheating for 10 min at 96°C, will become important in the genomic approaches to the 35 cycles of denaturation for 30 s at 96°C, annealing for wide spectrum of immunological diseases from auto- 30 s at the temperature shown in Table 4, and extension immunity to infections, and also, in the analysis of the for 30 s at 72°C were performed, using thermal cyclers evolutionary pathway of a family of chemokine recep- (GeneAmp PCR system 9700 and 9600, Perkin–Elmer tor genes. Applied Biosystems, Foster City, CA, USA, and Thermal cycler MP, Takara, Kyoto, Japan). The amplified DNA was analyzed by SSCP method. Materials and methods SSCP analysis was performed essentially according to the method described previously.21 One microliter of solution Subjects containing the PCR product was mixed with 7 ␮l of de- A total of 242 healthy individuals and 264 patients with naturing solution (95% formamide, 20 mm EDTA, rheumatic diseases were examined. The healthy individ- 0.05% bromphenol blue, 0.05% xylene cyanol FF). The uals consisted of the researchers, laboratory workers and mixtures were thermally denatured at 96°C for 5 min and students at the institutions where this study was carried immediately cooled on ice. One microliter of the mix- out. The patient group consisted of 146 patients with RA, tures was applied to 10% polyacrylamide gel 80 patients with SLE, 14 patients with SSc, 12 patients (acrylamide : bisacrylamide = 49:1). Electrophoresis was with Behc¸et disease (BD) and 12 patients with Sjo¨gren’s carried out in 0.5 × TBE (45 mm Tris-borate [pH 8.0], syndrome (Sjs). These diseases were diagnosed according 1mm EDTA) under constant electric current of

Genes and Immunity cSNPs in human CXCRs H Kato et al 335 Table 5 Primers used for the detection of variations and PCR-SSCP conditionsa

Name Primer Fragment Annealing Electrophoresis size temperature temperaturec sequence positionb

CXCR1–5 CXCR1-F5 5′-ATGGATTCACCCTGCGTACA-3′ IC3 239 bp 55°C20°C CXCR1-R5 5′-GTTGAGGCAGCTATGGAGAA-3′ TM7 CXCR1–6 CXCR1-F6 5′-AGCGCCGCAACAACATCGG-3′ EC3 312 bp 58°C10°C CXCR1-R6 5′-CCTGTCCAGAGCCAGATCAC-3′ 3′UTR CXCR2–2 CXCR2-F2 5′-GGTCATTATCTATGCCCTGG-3′ TM1 336 bp 55°C8°C CXCR2-R2 5′-TTGACCAAGTAGCGCTTCTG-3′ IC2 CXCR2–4 CXCR2-F4 5′-CAATGTTAGCCCAGCCTGCT-3′ EC2 335 bp 59°C15°C CXCR2-R4 5′-GAATCTCGGTGGCATCCAGA-3′ TM7 CXCR3-exon2–5 CXCR3-ex2-F5 5′-TGGACATCCTCATGGACCTG-3′ EC3 319 bp 62°C15°C CXCR3-ex2-R5 5′-GAAGTCAGACTGTGGGCGAA-3′ 3′UTR aOnly the primer sequences and experimental conditions for the fragments containing variations are listed. Others are available upon request. bPrimer position: IC, intracellular loop region (eg, IC3 stands for the third intracellular loop region); TM, transmembrane region; EC, extracellular loop region; 3′UTR, 3′ untranslated region. cSSCP conditions: Electrophoresis was carried out in 10% polyacrylamide gels without glycerol at the shown temperature for 90 min duration under constant current 20 mA/gel. These conditions were applicable in only the fragments listed in this table, not always in others.

20 mA/gel, using a minigel electrophoresis apparatus a multi-locus haplotype, LD is defined as the difference with a constant temperature control system between the estimated haplotype frequency and the pro- (90 × 80 × 1 mm, AE-6410 and AE-6370; ATTO, Tokyo, duct of each allele frequency. The relative linkage dis- Japan). The optimal conditions of electrophoresis were equilibrium (RLD) value is defined as the ratio of LD to determined by preliminary experiments. To detect vari- the absolute value of possible maximum (when LD у 0) ations sensitively, each fragment was analyzed with and or possible minimum (when LD Ͻ 0) of LD.52 without 5% glycerol in the gel. Single-strand DNA frag- ments in the gel were visualized by silver staining (Daiichi Pure Chemicals, Tokyo, Japan) following the Acknowledgements manufacturer’s protocol. The authors are indebted to Drs Kunio Matsuta (Matsuta Clinic), Yasuoki Moroi, Akira Watanabe (National Ito Spa Direct sequencing Hospital), Kenji Ikebe (Nanasato Hospital), Shoji Uchida, The nucleotide sequences of the samples showing differ- Yoshiaki Kuga (Tokyo Metropolitan Bokuto Hospital), ent SSCP patterns were determined by direct sequencing. Yuichi Nishioka (Yamanashi Central Hospital), Akira PCR products were amplified, and both sense and anti- Hashimoto, Fumiaki Fukuhara, Sumako Fukuhara sense strands were directly sequenced using the same (Fukuhara Hospital) for the recruitment of the patients, primers as those for the SSCP analysis. Fluorescence- to Dr Jun Ohashi (Department of Human Genetics, Uni- based automated cycle sequencing of PCR products was versity of Tokyo) for statistical analysis, and to Michiko performed with an automated sequencer (DNA Sequen-   Shiota (Department of Human Genetics, University of cing System ABI PRISM 377 and ABI PRISM 310, Per- Tokyo) for technical assistance. kin–Elmer Applied Biosystems), using dye-terminator method according to the manufacturer’s instructions (ABI PRISM dRhodamine Terminator Cycle Sequen- References cing-Ready Reaction Kit). 1 Baggiolini M. Chemokines and leukocyte traffic. Nature 1998; 392: 565–568. Statistical analysis 2 Holmes WE, Lee J, Kuang WJ, Rice GC, Wood WI. Structure and Allele positivity was defined as the proportion of the functional expression of a human interleukin-8 receptor. Science individuals possessing the allele. Fisher’s exact prob- 1991; 253: 1278–1280. ability test was used to analyze the distribution of each 3 Murphy PM, Tiffany HL. Cloning of complementary DNA enco- variation in the patients and in healthy individuals. Odds ding a functional human interleukin-8 receptor. Science 1991; ratio (OR) was also calculated to evaluate the strength of 253: 1280–1283. associations between rheumatic diseases and variations. 4 Mollereau C, Muscatelli F, Mattei MG, Vassart G, Parmentier P Ͻ 0.05 was considered significant. Haplotype fre- M. The high-affinity receptor gene (IL8RA) maps quencies and linkage disequilibrium parameters were to the 2q33-q36 region of the human genome: cloning of a pseudogene (IL8RBP) for the low-affinity receptor. Genomics estimated from the typing results using the EH algor- 51 1993; 16: 248–251. ithm. The linkage disequilibrium (LD) value of a two- 5 Morris SW, Nelson N, Valentine MB et al. Assignment of the locus haplotype is the difference between the estimated genes encoding human interleukin-8 receptor types 1 and 2 and haplotype frequency of AiBj, f(AiBj) and the product of an interleukin-8 receptor pseudogene to chromosome 2q35. Gen- the allele frequencies of Ai and Bj,f(Ai)f(Bj). In the case of omics 1992; 14: 685–691.

Genes and Immunity cSNPs in human CXCRs H Kato et al 336 6 Corne´lis F, Faure S, Martinez M et al. New susceptibility locus ture, genomic organization, and expression of the human for rheumatoid arthritis suggested by a genome-wide linkage interleukin-8 receptor B gene. J Biol Chem 1994; 269: 11065–11072. study. Proc Natl Acad Sci USA 1998; 95: 10746–10750. 29 White JK, Shaw MA, Barton CH et al. Genetic and physical map- 7 Shaw MA, Clayton D, Atkinson SE et al. Linkage of rheumatoid ping of 2q35 in the region of the NRAMP and IL8R genes: arthritis to the candidate gene NRAMP1 on 2q35. J Med Genet Identification of a polymorphic repeat in exon2 of NRAMP. Gen- 1996; 33: 672–677. omics 1994; 24: 295–302. 8 Gaffney PM, Kearns GM, Shark KB et al. A genome-wide search 30 Marchese A, Heiber M, Nguyen T et al. Cloning and chromo- for susceptibility genes in human systemic lupus erythematosus somal mapping of three novel genes, GPR9, GPR10, and GPR14, sib-pair families. Proc Natl Acad Sci USA 1998; 95: 14875–14879. encoding receptors related to interleukin 8, neuropeptide Y, and 9 Copeman JB, Cucca F, Hearne CM et al. Linkage disequilibrium somatostatin receptors. Genomics 1995; 29: 335–344. mapping of a susceptibility gene (IDDM7) to 31 Loetscher M, Loetscher P, Brass N, Meese E, Moser B. Lympho- chromosome 2q31-q33. Nat Genet 1995; 9: 80–85. cyte-specific chemokine receptor CXCR3: regulation, chemokine 10 Hadano S, Nichol K, Brinkman RR et al. A yeast artificial chro- binding and gene localization. Eur J Immunol 1998; 28: 3696– mosome-based physical map of the juvenile amyotrophic lateral 3705. sclerosis (ALS2) critical region on human chromosome 2q33-q34. 32 Gentles AJ, Karlin S. Why are human G-protein-coupled recep- Genomics 1999; 55: 106–112. tors predominantly intronless? Trends Genet 1999; 15: 47–49. 11 Mitelman F, Kaneko Y, Trent J. Report of the committee on chro- 33 Brosius J. Many G-protein-coupled receptors are encoded by mosome changes in neoplasia. Cytogenet Cell Genet 1991; 58: retrogenes. Trends Genet 1999; 15: 304–305. 1053–1079. 34 Soto H, Wang W, Strieter RM et al. The CC chemokine 6Ckine 12 Ishikiriyama S. Gene for Waardenburg syndrome type I is binds the CXC chemokine receptor CXCR3. Proc Natl Acad Sci located at 2q35, not at 2q37.3. Am J Med Genet 1993; 46: 608. USA 1998; 95: 8205–8210. 13 Loetscher M, Gerber B, Loetscher P et al. Chemokine receptor 35 Fan P, Kyaw H, Su K et al. Cloning and characterization of a specific for IP10 and Mig: structure, function, and expression in novel human chemokine receptor. Biochem Biophys Res Commun activated T-lymphocytes. J Exp Med 1996; 184: 963–969. 1998; 243: 264–268. 14 Qin S, Rottman JB, Myers P et al. The chemokine receptors 36 Jenh CH, Cox MA, Kaminski H et al. Species specificity of the CXCR3 and CCR5 mark subsets of T cells associated with certain CC chemokine 6Ckine signaling through the CXC chemokine inflammatory reactions. J Clin Invest 1998; 101: 746–754. receptor CXCR3: human 6Ckine is not a ligand for the human 15 Sorensen TL, Tani M, Jensen J et al. Expression of specific or mouse CXCR3 receptors. J Immunol 1999; 162: 3765–3769. chemokines and chemokine receptors in the central nervous sys- 37 Weng Y, Siciliano SJ, Waldburger KE et al. Binding and func- tem of multiple sclerosis patients. J Clin Invest 1999; 103: 807– tional properties of recombinant and endogenous CXCR3 815. chemokine receptors. J Biol Chem 1998; 273: 18288–18291. 16 Bonecchi R, Bianchi G, Bordignon PP et al. Differential 38 Pringault E, Arpin M, Garcia A, Finidori J, Louvard D. A human expression of chemokine receptors and chemotactic responsive- villin cDNA clone to investigate the differentiation of intestinal ness of type 1 T helper cells (Th1s) and Th2s. J Exp Med 1998; and kidney cells in vivo and in culture. EMBO J 1986; 5: 3119– 187: 129–134. 3124. 17 Sallusto F, Lenig D, Mackay CR, Lanzavecchia A. Flexible pro- 39 Blackwell JM, Barton CH, White JK et al. Genetic regulation of grams of chemokine receptor expression on human polarized T leishmanial and mycobacterial infections: the Lsh/Ity/Bcg gene helper 1 and 2 lymphocytes. J Exp Med 1998; 187: 875–883. story continues. Immunol Lett 1994; 43: 99–107. 18 Samson M, Libert F, Doranz BJ et al. Resistance to HIV-1 infec- 40 Roach TI, Chatterjee D, Blackwell JM. Induction of early- tion in caucasian individuals bearing mutant alleles of the CCR- response genes KC and JE by mycobacterial lipoarabinoman- 5 chemokine receptor gene. Nature 1996; 382: 722–725. nans: regulation of KC expression in murine by 19 Garred P, Madsen HO, Petersen J et al. CC chemokine receptor Lsh/Ity/Bcg (candidate Nramp). Infect Immun 1994; 62: 1176– 5 polymorphism in rheumatoid arthritis. J Rheumatol 1998; 25: 1184. 1462–1465. 41 John S, Marlow A, Hajeer A, Ollier W, Silman A, Worthington 20 Szalai C, Csaszar A, Czinner A et al. Chemokine receptor CCR2 and CCR5 polymorphisms in children with insulin-dependent J. Linkage and association studies of the natural resistance asso- diabetes mellitus. Pediatr Res 1999; 46: 82–84. ciated macrophage protein 1 (NRAMP1) locus in rheumatoid 21 Kato H, Tsuchiya N, Izumi S et al. New variations of human arthritis. J Rheumatol 1997; 24: 452–457. CC-chemokine receptors CCR3 and CCR4. Genes Immun 1999; 42 Bockaert J, Pin JP. Molecular tinkering of -coupled 1: 97–104. receptors: an evolutionary success. EMBO J 1999; 18: 1723–1729. 22 Charlton B, Lafferty KJ. The Th1/Th2 balance in autoimmunity. 43 Mellado M, Rodriguez-Frade JM, Vila-Coro AJ, de Ana AM, Curr Opin Immunol 1995; 7: 793–798. Martinez AC. Chemokine control of HIV-1 infection. Nature 23 Mummidi S, Ahuja SS, Gonzalez E et al. Genealogy of the CCR5 1999; 400: 723–724. locus and chemokine system gene variants associated with alt- 44 Arnett FC, Edworthy SM, Bloch DA et al. The American Rheu- ered rates of HIV-1 disease progression. Nat Med 1998; 4: 786– matism Association 1987 revised criteria for the classification of 793. rheumatoid arthritis. Arthritis Rheum 1988; 31: 315–324. 24 Sprenger H, Lloyd AR, Meyer RG, Johnston JA, Kelvin DJ. Gen- 45 Tan EM, Cohen AS, Fries JF et al. The 1982 revised criteria for omic structure, characterization, and identification of the pro- the classification of systemic lupus erythematosus. Arthritis moter of the human IL-8 receptor A gene. J Immunol 1994; 153: Rheum 1982; 25: 1271–1277. 2524–2532. 46 Subcommittee for scleroderma criteria of the American Rheu- 25 Ahuja SK, Shetty A, Tiffany HL, Murphy PM. Comparison of matism Association Diagnostic and Therapeutic Criteria Com- the genomic organization and promoter function for human mittee: Preliminary criteria for the classification of systemic interleukin-8 receptor A and B. J Biol Chem 1994; 269: 26381– sclerosis (scleroderma). Arthritis Rheum 1980; 23: 581–590. 26389. 47 International Study Group for Behc¸et’s Disease. Criteria for 26 Ahuja SK, Ozcelik T, Milatovitch A, Francke U, Murphy PM. diagnosis of Behc¸et’s disease. Lancet 1990; 335: 1078–1080. Molecular evolution of the human interleukin-8 receptor gene 48 Vitali C, Bombardieri S, Moutsopoulos HM et al. Preliminary cluster. Nat Genet 1992; 2: 31–36. criteria for the classification of Sjo¨gren’s syndrome. Results of a 27 Lee J, Horuk R, Rice GC, Bennett GL, Camerato T, Wood WI. prospective concerted action supported by the European Com- Characterization of two high affinity human interleukin-8 recep- munity. Arthritis Rheum 1993; 36: 340–347. tors. J Biol Chem 1992; 267: 16283–16287. 49 Tokunaga K, Imanishi T, Takahashi K, Juji T. On the origin and 28 Sprenger H, Lloyd AR, Lautens LL, Bonner TI, Kelvin DJ. Struc- dispersal of East Asian populations as viewed from HLA haplo-

Genes and Immunity cSNPs in human CXCRs H Kato et al 337 types. In: Akazawa T, Szathmary EJ (ed). Prehistoric Mongoloid 51 Terwilliger JD, Ott J. Handbook of human linkage analysis. Johns Dispersals. Oxford University Press: Oxford, 1996, pp 187–197. Hopkins University Press: Baltimore, 1999, pp 188–193. 50 Bannai M, Tokunaga K, Lin L et al. Discrimination of human 52 Lewontin RC. The interaction of selection and linkage. I. General HLA-DRB1 alleles by PCR-SSCP (single-strand conformation considerations, heterotic models. Genetics 1964; 49: 49–67. polymorphism) method. Eur J Immunogenet 1994; 21: 1–9.

Genes and Immunity