Association of polymorphisms in complement component C3 gene ...

Rheumatology 2008;47;158–164 Advance Access publication 3 January 2008

doi:10.1093/rheumatology/kem321

Association of polymorphisms in complement component C3 gene with susceptibility to systemic lupus erythematosus H. Miyagawa1, M. Yamai2, D. Sakaguchi2, C. Kiyohara3, H. Tsukamoto1, Y. Kimoto1, T. Nakamura4, J.-H. Lee5, C.-Y. Tsai6, B.-L. Chiang5, T. Shimoda7, M. Harada1, T. Tahira2, K. Hayashi2 and T. Horiuchi1 Objective. Identification of the genes responsible for systemic lupus erythematosus (SLE). Methods. All the exons and putative promoter regions of 53 candidate genes (TNFRSF6/Fas, TNFSF6/FasL, Fli1, TNFSF10/TRAIL, TNFSF12/TWEAK, Bcl-2, PTEN, FADD, TRADD, CDKN1A, TNFRSF1A/TNFR1, TNFRSF4/OX40, TNFSF4/OX40L, TNFSF5/CD40L, TNFSF13B/BAFF, ICOS, CTLA4, CD28, FYN, G2A, CR2, PTPRC/CD45, CD22, CD19, Lyn, PDCD1, PTPN6, TGFB1, TGFB2, TGFB3, TGFBR1, TGFBR2, TGFBR3, CD3Z, DNASE1, APCS, MERTK, C3, C1QA, C1QB, C1QG, C2, MBL2, IGHM, IL-2, IL-4, IL-10, IFNG, TNFA, MAN2A1, TNFRSF11A/RANK, TNFRSF11B/OPG, TNFSF11/OPGL) were screened for single nucleotide polymorphisms (SNPs) and their association with SLE was assessed by case–control studies. A total of 509 cases and 964 controls of Japanese descent were enrolled. Results. A total of 316 SNPs was identified. When analysed in the Japanese population, the allele frequencies of T at rs7951 and G at rs2230201 of the C3 gene were 0.110 and 0.626, respectively, in SLE patients; significantly higher than the frequencies of 0.081 and 0.584, respectively, in controls [odds ratio (OR) ¼ 1.40, 95% confidence interval (CI) ¼ 1.05–1.86, P ¼ 0.016 and OR¼1.19, 95% CI ¼ 1.01–1.41, P ¼ 0.038, respectively]. The mean serum C3 level of carriers of the rs7951 T allele was significantly lower than that of non-carriers of the T allele in 87 SLE patients whose medical records were available (P ¼ 0.0018). Conclusion. rs7951 T allele of the C3 gene was significantly associated with SLE, and decreased serum level of C3 seems to be correlated with this allele. KEY

WORDS:

Complement C3, Systemic lupus erythematosus, Genetic analysis, Case–control study, SSCP.

(ii) identification of 316 single nucleotide polymorphisms (SNPs) in the entire coding regions and the promoter regions of these genes, encompassing a total of 194 403 bp, (iii) association study of the SNPs in a total of 509 cases and 964 controls of Japanese descent (Fig. 1). To our knowledge, this is one of the largest association studies of SLE.

Introduction Systemic lupus erythematosus (SLE) is an inflammatory disorder affecting multiple organ systems and is considered to be a prototypic systemic autoimmune disease [1]. Genetic predisposition has been implicated in the pathogenesis of SLE. Numerous candidate gene loci have been reported by genome-wide linkage studies in multiplex SLE families and by case–control studies. In addition, multiple susceptibility loci have been identified in mouse models of lupus [2–4]. The relatively low risk conferred by the susceptibility alleles indicates that the effects of the respective loci are small and therefore that the SLE is a multifactorial disease. Of the candidate loci, the human leucocyte antigen (HLA) genes have been most extensively studied in SLE patients from different ethnic groups. The association of HLA genes with SLE is undisputed, although the SLE-related HLA alleles are different between various ethnic groups [5, 6]. In order to clarify the genetic basis of SLE in more detail, with particular focus on non-major histocompatibility complex (MHC) susceptibility genes, we have performed case–control studies. The present study consisted of the following steps; (i) selection of 53 candidate genes associated with SLE on the basis of findings from mouse model studies and the likely functions of the genes,

Materials and methods Patients and DNA The Japanese patients had been followed-up at the rheumatology clinic of Kyushu University Hospital and its affiliated hospitals, and fulfilled the American College of Rheumatology (ACR) 1982 and 1997 revised criteria for SLE [7, 8]. We enrolled two sets of SLE patients; case 1a (n ¼ 264; 249 females/15 males) and case 1b (n ¼ 245; 224 females/21 males). Healthy volunteers from Kyushu and Chubu area were enrolled as controls for the Japanese population. The control set (n ¼ 964) consists of three groups; control 1a (n ¼ 269; 269 females), control 1b (n ¼ 426; 306 females/ 120 males) and control 1c (n ¼ 269; 269 females). DNA was purified from peripheral blood mononuclear cells (PBMCs) by using QIAamp DNA Blood Maxi Kit (Qiagen, Hilden, Germany). The study protocol was approved by our institutional review board, and all participants provided written informed consent. Pools were constructed by stepwise dilution and combining as follows. The initial concentration of each DNA dissolved in 10mM Tris–HCl, pH 7.5 and containing 1mM ethylenediamine tetra-acetic acid [EDTA (TE)] was determined with a microplate spectrofluorometer (SPECTRAmaxÕ GEMINI XS; Molecular Devices Corporation, Sunnyvale, CA, USA) after staining with PicoGreen dsDNA Quantification Reagent (Molecular Probes, Eugene, OR, USA). The concentration was adjusted to 20 ng/
l by two-step dilution with various volumes of 0.1
TE by using a robotic system (Genesis RSP150; Tecan Trading AG, Mannedorf, Switzerland). The concentrations of all samples were confirmed to be within 10% of the expected value.

1 Department of Medicine and Biosystemic Science, Graduate School of Medical Sciences, 2Medical Institute of Bioregulation, 3Department of Preventive Medicine, Graduate School of Medical Sciences, Kyushu University, 4Kumamoto Center for Arthritis and Rheumatology, Kumamoto, Japan, 5Department of Pediatrics, National Taiwan University Hospital, 6Department of Medicine, Taipei Veterans General Hospital and National Yang-Ming University School of Medicine, Taipei, Taiwan and 7Department of Clinical Research, Fukuoka National Hospital, Fukuoka, Japan.

Submitted 1 May 2007; revised version accepted 1 November 2007. Correspondence to: T. Horiuchi, Department of Medicine and Biosystemic Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka 812-8582, Japan. E-mail: [email protected]

158 ß The Author 2008. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected]

Association of C3 gene with SLE

159

Polymerase chain reaction (PCR) and post-PCR fluorescent labelling and automated capillary electrophoresis under single-strand conformation polymorphism conditions (PLACE-SSCP) PCR was performed as described [13]. Our PLACE-SSCP analysis system is a streamlined and high-throughput procedure to detect polymorphisms by analysing DNA samples of individuals and then to quantify allele frequencies in DNA pools. The experimental procedures have been described previously [13–15].

SNP identification and allele frequency quantification The PLACE-SSCP analyses were performed for 12 individuals (11 SLE patients and a healthy control) to find polymorphic STSs that were then sequenced for SNP identification. The allele frequencies of SNPs were determined by peak-height analysis of pooled DNAs as described previously [13].

SNP genotyping and haplotype inference For SNPs of interest, we sequenced individual DNA samples. Haplotypes were re-constructed by Phase [16] and by Haploview [17].

Statistical analysis

A total of 53 genes related to apoptosis/cell proliferation (TNFRSF6/Fas, TNFSF6/FasL, Fli1, TNFSF10/TRAIL, TNFSF12/TWEAK, Bcl-2, PTEN, FADD, TRADD, CDKN1A), lymphoid signalling (TNFRSF1A/TNFR1, TNFRSF4/OX40, TNFSF4/OX40L, TNFSF5/CD40L, TNFSF13B/BAFF, ICOS, CTLA4, CD28, FYN, G2A, CR2, PTPRC/CD45, CD22, CD19, Lyn, PDCD1, PTPN6, TGFB1, TGFB2, TGFB3, TGFBR1, TGFBR2, TGFBR3, CD3Z), antigen/immune complex clearance (DNASE1, APCS, MERTK, C3, C1QA, C1QB, C1QG, C2, MBL2, IGHM), cytokines (IL-2, IL-4, IL-10, IFNG, TNFA), epitope modification (MAN2A1) and others (TNFRSF11A/RANK, TNFRSF11B/OPG, TNFSF11/ OPGL) were selected for the association study [9–12]. The genes implicated in the lupus mouse model or in knockout or transgenic mice that develop a lupus-like condition are underlined [9].

We performed Z-tests for comparison of allele frequencies obtained by PLACE-SSCP analysis of pooled DNA [18] and chi-square tests for the allele and genotype frequencies obtained by individual genotyping. One-way analysis of variance was performed to examine the relationship between the C3 SNP rs7951 and serum C3 level. To account for multiple comparisons, a Bonferroni correction was applied. Bonferroni correction concerns the question if, in the case of doing more than one test in a particular study, the -level should be adjusted downwards to consider chance capitalization. The  level is the chance taken by researchers to make a type I error. The type I error is the error of incorrectly declaring a difference, effect or relationship to be true due to chance producing the observed state of events. Customarily the -level is set at 0.05, or, in no more than 1 in 20 statistical tests the test will show ‘something’ while in fact there is nothing. In the case of more than one statistical test, the chance of finding at least one test statistically significant due to chance fluctuation in the total experiment and to incorrectly declare a difference or relationship to be true, increases. Using the Bonferroni method the -level of each individual test is corrected downwards to ensure that the overall experiment wise-risk for a number of tests remains 0.05. Even if more than one test is done the risk of finding a difference or effect incorrectly significant continues to be 0.05. The dose–response relationship between serum C3 level and the T allele of C3 SNP rs7951 was evaluated by using regression analysis. All the calculations were performed with the computer program STATA Version 8.2 (Stata Corporation, College Station, TX, USA).

Primer design

Results

FIG. 1. Strategy for identifying genes responsible for SLE.

The pools were then made by combining equal volumes of each sample, and the concentration was finally adjusted to 16.6 ng/
l. We prepared three sets of DNA pools: case 1a (n ¼ 264), control 1a (n ¼ 269) and control 1b (n ¼ 426).

Selection of candidate genes

We designed screening primers [sequence-tagged site (STS) size is 100–350 bp] for single-strand conformation polymorphism (SSCP) analysis to cover all the exons and the putative transcriptional start sites (upstream region; UR), based on reference sequence data or the dbTSS database (http://dbtss.hgc.jp). For allele frequency quantification, another set of primers was designed to amplify short STS products (50–120 bp), which allowed us to calculate allele frequencies accurately for each SNP. Sequences of the primers are available upon request.

Identification of SNPs in the 53 candidate genes A total of 194 403 bp was scanned, and 316 SNPs were detected. Of these SNPs, 71 were identified in the coding regions (26 nonsynonymous, 45 synonymous), 90 in the putative promoter regions, 20 in the 50 -untranslated regions (UTRs), 55 in the 30 UTRs and their downstream regions and 80 in the intron regions (Table 1). All the SNPs identified are listed in supplementary table 1, available as supplementary data at Rheumatology Online.

H. Miyagawa et al.

160 TABLE 1. Identified polymorphisms of the 53 candidate genes

No. of identified SNPs Genea

RefSeq ID

APCS BCL2 C1QA C1QB C1QG (C1QC) C2 C3 CD19 CD22 CD28 CD3Z (CD247) CDKN1A CR2 CTLA4 DNASE1 FADD Fli-1 (FLI1) FYN G2A (GPR132) ICOS IFNG IGHM IL10 IL2 IL4 LYN MAN2A1 MBL2 MERTK PDCD1 PTEN PTPN6 PTPRC TGFB1 TGFB2 TGFB3 TGFBR1 TGFBR2 TGFBR3 TNFA (TNF) TNFRSF11A TNFRSF11B TNFRSF1A TNFRSF4 TNFRSF6 (FAS) TNFSF10 TNFSF11 TNFSF12 TNFSF13B TNFSF4 TNFSF5 (CD40LG) TNFSF6 (FASLG) TRADD Total 53 genes

NM_001639 NM_000633 NM_015991 NM_000491 NM_172369 NM_000063 NM_000064 NM_001770 NM_001771 NM_006139 NM_000734 NM_000389 NM_001877 NM_005214 NM_005223 NM_003824 NM_002017 NM_002037 NM_013345 NM_012092 NM_000619 X57086 NM_000572 NM_000586 NM_000589 NM_002350 NM_002372 NM_000242 NM_006343 NM_005018 NM_000314 NM_002831 NM_002838 NM_000660 NM_003238 NM_003239 NM_004612 NM_003242 NM_003243 NM_000594 NM_003839 NM_002546 NM_001065 NM_003327 NM_000043 NM_003810 NM_003701 NM_003809 NM_006573 NM_003326 NM_000074 NM_000639 NM_003789

a

No. of base pairs scanned

Non-synonymous

Synonymous

Upstream

50 -UTR

30 -UTR & downstream

Intron

Total

2142 7150 2806 1881 1911 4643 7668 4222 4729 4854 3316 3446 6199 1893 2287 2716 3811 3968 3738 3849 2327 2324 2633 2080 1766 4147 6180 4866 5064 2576 3806 5200 9321 2479 2857 3783 3834 3719 3262 2841 4203 2694 4214 1341 5577 3065 4631 2023 2381 4230 2778 2598 2374 194 403

0 0 1 0 1 2 0 1 0 0 0 1 2 1 1 0 0 0 0 0 0 2 0 0 0 1 2 1 3 1 0 0 1 1 0 0 0 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 26

1 0 1 1 1 2 7 1 0 1 0 0 1 0 0 0 2 0 1 0 0 2 0 1 0 0 3 2 2 1 0 0 1 0 0 1 0 1 2 0 0 0 2 1 3 1 1 1 0 0 1 0 0 45

5 1 7 0 0 1 0 2 0 2 5 5 1 0 0 0 5 0 2 2 0 2 2 1 2 1 0 7 0 2 0 2 3 1 0 1 0 2 1 1 1 2 6 0 3 4 3 0 3 1 0 0 1 90

0 0 0 0 1 0 0 0 0 1 1 0 1 0 3 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 1 0 1 0 0 1 1 0 0 0 1 0 0 0 1 0 1 1 0 0 0 20

0 6 0 0 1 0 0 1 0 4 4 1 2 0 0 0 0 0 0 3 0 0 1 0 0 1 0 9 0 1 0 0 0 0 1 0 0 0 0 0 1 0 1 0 2 5 2 3 0 5 0 1 0 55

0 0 3 1 0 0 15 0 2 1 2 1 0 0 1 0 2 5 1 2 0 0 2 0 1 0 3 1 2 1 1 5 5 0 1 2 3 3 0 0 1 1 1 2 1 2 4 0 0 1 1 0 0 80

6 7 12 2 4 5 22 5 2 9 12 8 7 1 5 1 9 6 4 7 0 6 5 2 4 3 8 20 7 6 2 8 11 2 3 4 3 8 5 1 4 4 11 3 9 12 11 4 4 8 2 1 1 316

Symbols approved by HUGO Gene Nomenclature are shown in parentheses.

1st screening: single locus association study using PLACE-SSCP analysis of pooled DNA The primary screening for (a) susceptibility gene(s) was performed by two-step PLACE-SSCP analysis of pooled DNAs (Fig. 1). The usefulness of DNA pooling has been reported in the association study of LRCH1 gene and osteoarthritis [19]. Two independent control groups, control 1a (n ¼ 269) and control 1b (n ¼ 426), and an SLE group, case 1a (n ¼ 264), were examined. For the first step, pooled DNA samples of case 1a and control 1a were compared for differences in all of the SNPs identified above, and a total of 279 SNPs were successfully quantified for the frequency determination. The SNPs whose minor allele frequencies were higher than 0.1 in case 1a or in control 1a were subjected to the association study in order to minimize the error. Twenty-six SNPs were associated with SLE (P < 0.05). Three of the 26 SNPs

were located in each of C3, ICOS and TNFRSF1A, two in each of CD28, IGHM, MBL2, TGFBR3 and TNFSF10, and one in each of FYN, PTPN6, TNFRSF11A, TNFRSF11B, TNFRSF4, TNFSF12 and TNFSF5. In order to confirm the association, these SNPs were further studied for allele frequencies in control 1b, another set of healthy controls for case 1a. Three SNPs (rs7951 in the C3 gene, rs3181098 in the CD28 gene and rs4149570 in the TNFRSF1A gene) showed replicated association with SLE (P < 0.05) (data not shown).

2nd screening: genotyping the SNPs of three genes by sequencing 192 individuals randomly selected from each of case 1a and control 1b To confirm the association of these three genes with SLE, 16 SNPs in the C3 gene, 3 SNPs in the CD28 gene and 9 SNPs in the

Association of C3 gene with SLE

161

TABLE 2. Association study for SNPs of C3, CD28 and TNFRSF1A with SLE Allele frequency Gene

rs

C3

rs2547438 rs2230201 rs2230204 rs2230205 rs432823 rs428453 rs366510 rs423490 rs389404 rs3745567 rs11569508 rs11569509 rs11569510 rs7951 rs2277984 rs17030 rs3181098 rs3116496 rs3181113 rs4149618 rs4149569 rs4149619 rs16932194 rs4149570 rs4149621 rs2234649 rs767455 rs12426675

CD28 TNFRSF1A

Position

Allele

Control (n ¼ 192)

Case (n ¼ 192)

6669078 6664291 6660848 6660704 6653246 6653157 6648829 6648406 6644387 6641771 6641746 6641744 6641742 6632991 6630511 6628989 204278623 204302757 204310155 6322762 6322729 6322502 6322001 6321851 6321822 6321624 6321206 6309731

C/A G/A A/G G/A G/A C/G A/C C/T T/C A/G C/T C/T C/T T/C G/A C/T A/G T/C G/T T/C G/C T/G CT/T/G G/A A/C A/G T/C

0.16/0.84 0.57/0.43 0.49/0.51 0.58/0.42 0.86/0.14 0.85/0.15 0.85/0.15 0.93/0.07 0.18/0.82 0.08/0.92 0.08/0.92 0.08/0.92 0.08/0.92 0.07/0.93 0.50/0.50 0.50/0.50 0.30/0.70 0.93/0.07 0.48/0.52 0.11/0.89 0.78/0.22 0.07/0.93 0.90/0.10 0.36/0.64 0.11/0.89 0.91/0.09 0.79/0.21 0.60/0.40

0.20/0.80 0.65/0.35 0.52/0.48 0.66/0.34 0.86/0.14 0.86/0.14 0.86/0.14 0.95/0.05 0.23/0.77 0.11/0.89 0.11/0.89 0.11/0.89 0.11/0.89 0.14/0.86 0.51/0.49 0.51/0.49 0.38/0.62 0.96/0.04 0.53/0.47 0.13/0.87 0.85/0.15 0.07/0.93 0.93/0.07 0.45/0.55 0.13/0.87 0.93/0.07 0.85/0.15 0.66/0.34

TNFRSF1A gene, including the above identified SNPs (rs7951 for C3, rs3181098 for CD28 and rs4149570 for TNFRSF1A), were selected to cover the entire genomic regions (Fig. 1). These 28 SNPs were genotyped by sequencing genomic DNA from 192 SLE patients and the same number of samples, randomly selected from the samples of case 1a and control 1b, followed by estimation of the allele frequencies in SLE and controls. As shown in Table 2, three SNPs (rs2230201, rs2230205, rs7951) in the C3 gene, one SNP in the CD28 gene (rs3181098) and three SNPs (rs4149569, rs4149570, rs767455) in the TNFRSF1A gene were shown to be significantly associated with SLE, confirming the result of PLACE-SSCP analysis using the pooled DNA. Among these SNPs, the allele frequency of T at position rs7951 was 0.07 in the controls vs 0.14 in SLE, which was the most significant difference found [odds ratio (OR) 2.17, 95% confidence interval (CI) 1.32–3.56, P ¼ 0.0017]. As this SNP was located in the C3 gene, we concentrated our effort on the analysis of association between C3 and SLE. Figure 2 shows the linkage disequilibrium (LD) of the SNPs in the C3 gene based on the genotyping data above.

3rd screening: genotyping and haplotyping the C3 SNPs rs2230201, rs2230205 and rs7951 by sequencing a total of 509 cases and 695 controls of Japanese descent We then genotyped the three SNPs (rs2230201, rs2230205, rs7951) in the C3 gene by sequencing all the individuals in case 1a (n ¼ 264) and in control 1b (n ¼ 426) (Fig. 1, Table 3). The allele frequencies of these SNPs estimated by individual genotyping were in good agreement (within 2% accuracy) with those obtained by SSCP analysis of pooled DNA, consistent with previous work validating allele frequency estimation by this method [20]. As shown in Table 3, individuals carrying genotype G of rs2230201 had a significantly increased risk of developing SLE (P ¼ 0.043), as did carriers of genotype T of rs7951 (P ¼ 0.019). Allele frequency analysis showed that the rs7951 T allele was associated with an increased risk of SLE (P ¼ 0.013). To extend the association of these SNPs with SLE, we newly enrolled additional individuals for SLE (case 1b) and controls

Allelic OR [95% CI] 1.35 1.44 1.11 1.39 1.05 1.07 1.07 1.32 1.35 1.32 1.32 1.32 1.32 2.17 1.05 1.04 1.43 1.69 1.26 1.24 1.50 1.01 1.32 1.43 1.31 1.39 1.48 1.29

[0.93–1.96] [1.07–1.92] [0.84–1.47] [1.04–1.86] [0.70–1.58] [0.71–1.60] [0.71–1.62] [0.73–2.41] [0.95–1.93] [0.81–2.15] [0.81–2.15] [0.81–2.15] [0.81–2.15] [1.32–3.56] [0.79–1.40] [0.79–1.39] [1.06–1.93] [0.87–3.25] [0.95–1.68] [0.80–1.92] [1.04–2.18] [0.57–1.80] [0.79–2.22] [1.07–1.92] [0.84–2.04] [0.82–2.36] [1.02–2.16] [0.96–1.74]

P 0.11 0.015 0.47 0.027 0.81 0.76 0.74 0.36 0.097 0.26 0.26 0.26 0.26 0.0017 0.72 0.78 0.019 0.12 0.11 0.33 0.031 0.97 0.29 0.016 0.23 0.22 0.041 0.092

(control 1c), and finally genotyped a total of 509 cases (designated case 1a þ case 1b) and 695 controls (control 1b þ control 1c). All three SNPs were associated with SLE. Carriers of the rs2230201 G allele, the rs2230205 G allele and the rs7951 T allele were at a significantly increased risk of SLE (P ¼ 0.017, P ¼ 0.0499 and P ¼ 0.028, respectively). Analysis of allele frequencies revealed that the rs2230201 G allele and the rs7951 T allele were significantly associated with an increased risk of SLE (P ¼ 0.038 and P ¼ 0.016, respectively). The SNP at rs2230205 was not associated with SLE when assessed by allele frequency. Haplotypes were then reconstructed by using Haploview and Phase [16, 17]. Several haplotypes generated by the three SNPs rs2230201, rs2230205 and rs7951 were significantly associated with SLE, but did not show a more significant association than obtained from single association studies (data not shown). When estimated by Haploview based on genotyping data from 695 healthy controls, the D-values of rs2230201/rs2230205, rs2230201/rs7951 and rs2230205/rs7951 were 0.984, 0.520 and 0.535, respectively.

The C3 SNP rs7951 is associated with serum C3 level Serum levels of C3 were obtained for 87 SLE patients whose medical records were available (Fig. 3). All of them were followed at our outpatient clinic and had shown stable disease activity for at least 6 months. Serum C3 values showed a normal probability distribution. The C3 levels (mean and S.D.) of SLE patients homozygous (C/C; n ¼ 54), heterozygous (C/T; n ¼ 30) or noncarriers (T/T; n ¼ 3) for allele C at rs7951 were 88.5  24.7, 73.1  24.4 and 50.2  12.5 mg/dl, respectively. The C3 SNP rs7951 was significantly related to serum C3 level (F ¼ 6.47, P ¼ 0.0024). The mean serum C3 level of genotype C/C was higher than those of the other two genotypes (Bonferroni P ¼ 0.020 in comparison with C/T genotype and Bonferroni P ¼ 0.029 in comparison with T/T genotype), whereas no significant difference was observed between genotypes C/T and T/T. The mean serum C3 level of carriers of the rs7951 T allele (71.0  24.4 mg/dl) was significantly lower than that of non-carriers of the T allele in our 87 SLE patients (P ¼ 0.0018). The daily prednisolone doses of the

162

H. Miyagawa et al.

FIG. 2. Linkage disequilibrium of the SNP markers in the C3 gene. Linkage disequilibrium of 16 SNPs identified in the present study (Table 2) was analysed by Haploview using genotyping data of 192 control individuals. Pairwise LD (D0 ) values are presented in each box; values of 1.0 are not shown. The standard colour scheme of the software is used, so that bright red squares illustrate very strong LD and white squares illustrate little/no LD. The recombination rate track is according to the data of HapMap. As three SNPs (rs11569508, rs11569509, rs11569510) are only 4 bp apart and were in complete linkage disequilibrium, rs11569510 is shown as a representative of these three SNPs.

individuals carrying genotypes C/C, C/T and T/T at rs7951 were 8.45  4.83, 10.05  1.44 and 8.33  5.06 mg, respectively, which were not significantly different (P ¼ 0.288). In addition, there were no statistical differences between carriers of rs7951 T allele (n ¼ 33) or non-carriers (n ¼ 54) in anti-DNA antibody titres and the incidence of renal or CNS involvement. The titres of anti-DNA antibody were 37.3  71.3 IU/l for T allele carriers and 22.2  28.3 IU/l for non-carriers, which was not statistically significant (P ¼ 0.300). The incidences of renal involvement or CNS involvement were 54.5 and 21.2% for carriers, respectively, while those for non-carriers were 63.0 and 14.8%, respectively. These phenotypes were not significantly different between the carriers and non-carriers.

Discussion The present study consisted of two major parts: identification of SNPs in the 53 candidate genes and search for susceptibility gene(s) for SLE by association study. The latter part includes a number of screening steps (Fig. 1). To our knowledge, this is one

of the largest systematic genetic analyses for SLE: (i) a total of 194 403 bp were screened for SNPs, (ii) the entire coding region and the putative promoter region of 53 genes were studied, (iii) 316 SNPs were identified and assessed for association with SLE, and (iv) a total of 509 SLE patients (case 1a þ case 1b) and 964 controls (control 1a þ control 1b þ control 1c) of Japanese descent. In addition to the advantage of scale, it should also be noted that this association study was designed to minimize errors in the process of selecting genes and subsequent genotyping; first, only the SNPs repeatedly shown to be associated with SLE (case 1a; n ¼ 264) by using two independent sets of controls (control 1a, n ¼ 269; control 1b, n ¼ 426) were considered to be significant. Estimation of allele frequency by pooling DNA was shown to be accurate by the present study (Table 3), validating the use of the DNA pooling method as the first phase of screening. Second, after selecting three genes, C3, TNFRSF1A and CD28, for further analysis, nucleotide sequencing of a total of 28 SNPs in these genes was performed in 192 individuals randomly selected from each of case 1a and control 1b, confirming the results obtained from initial screening of the pooled DNA by PLACE-SSCP

Association of C3 gene with SLE

163

TABLE 3. Association study for C3 SNP rs2230201, rs2230205 and rs7951 with our SLE patients Number (%) of

Case 1a

Control 1b

Number (%) of

P

Case 1a þ Case 1b

Control 1b þ Control 1c

P

rs2230201 Genotype A/A A/G G/G Total Allele A G Total

41 113 111 265 195 335 530

(15.5) (42.6) (41.9) (100) (36.8) (63.2) (100)

71 215 149 435 357 513 870

(16.3) 0.043 78 (49.4) 219 (34.3) 205 (100) 502 (41.0) 0.115 375 (59.0) 629 (100) 1004

(15.5) 118 (43.6) 335 (40.8) 234 (100) 687 (37.4) 571 (62.6) 803 (100) 1374

(17.2) 0.017 (48.8) (34.1) (100) (41.6) 0.038 (58.4) (100)

rs2230205 Genotype A/A A/G G/G Total Allele A G Total

38 113 115 266 189 343 532

(14.3) (42.5) (43.2) (100) (35.5) (64.5) (100)

68 206 164 438 342 534 876

(15.5) 0.128 75 (47.0) 221 (37.4) 213 (100) 509 (39.0) 0.187 371 (61.0) 647 (100) 1018

(14.7) 110 (43.4) 331 (41.8) 251 (100) 692 (36.4) 551 (63.6) 833 (100) 1384

(15.9) 0.049 (47.8) (36.3) (100) (39.8) 0.093 (60.2) (100)

rs7951 GenotypeC/C C/T T/T Total Allele C T  Total

206 53 6 265 465 65 530

(77.7) (20.0) (2.3) (100) (87.7) (12.3) (100)

372 62 5 439 806 72 878

(84.7) 0.019 405 (14.1) 94 (1.1) 9 (100) 508 (91.8) 0.013 904 (8.2) 112 (100) 1016

(79.7) 588 (84.6) 0.028 (18.5) 101 (14.5) (1.8) 6 (0.9) (100) 695 (100) (89.0) 1277 (91.9) 0.016 (11.0) 113 (8.1) (100) 1390 (100)

Allele frequencies estimated by pooling DNA were as follows (case 1a/control 1b):  (0.362/ 0.428) for rs2230201 allele A,   (0.367/0.392) for rs2230205 allele A, and    (0.121/0.080) for rs7951 allele T.

analysis. Third, three SNPs of the C3 gene were genotyped by sequencing a total of 509 cases and 695 controls, including all the members of case 1a and control 1b as well as newly enrolled cases (case 1b) and controls (control 1c). Together, this series of studies, which included identification of SNPs, a number of association studies and analysis of clinical data for C3 concentration in serum, allowed us to conclude that the C3 gene was associated with SLE in our cases and controls. The allele frequency of HapMap JPT samples (0.141) seems close to that of our SLE samples. However, the sample size of HapMap data is small (39 individuals). A rough approximation to the 95% CI for their minor allele frequency is calculated as (0.064–0.218) under binomial random sampling. Allele frequency in our control samples (0.08) are within this range and assumed to be more reliable because of the larger sample size. We also would like to stress that success rate of genotyping rs7951 in our control samples (99.9%) is greater than that in HapMap (86.7% for rs7951 in JPT samples) and genotype frequencies conforms to Hardy–Weinberg equilibrium (P ¼ 0.598). Complete deficiency of C3 is associated with SLE-like disease [21–23], suggesting that the T allele of rs7951 that leads to lower serum C3 levels, is likely to be a risk factor for SLE. Binding of complement C3 activation fragment iC3b to its receptor (CR3) on antigen-presenting cells is essential for the induction of tolerance [24]. In addition, complete deficiencies of components of the classical complement pathway (C1q, C1r, C1s, C4 and C2) are the strongest risk factors for SLE [25–27], although such deficiencies are very rare. The pathogenetic mechanism in such cases is considered to be defective immune complex clearance and/or defective B-cell tolerance [28]. Partial deficiencies of complement components are also associated with an increased risk of developing SLE. C4A deficiency has been reported to be associated with SLE, regardless of ethnicity [29–31]. Moreover, the uptake of apoptotic cells by monocyte-derived macrophages from SLE patients is reduced when cultured in autologous

FIG. 3. Serum C3 levels of individuals carrying C3 rs7951 C/C, C/T or T/T. Serum C3 levels were obtained for 87 SLE patients whose medical records were available. The C3 levels (mean and S.D.) of SLE patients homozygous (C/C; n ¼ 54), heterozygous (C/T; n ¼ 30) or non-carriers (T/T; n ¼ 3) for allele C at rs7951 were 88.5  24.7, 73.1  24.4 and 50.2  12.5 mg/dl, respectively.

serum but is restored by addition of normal human serum [32]. This effect is dependent on disease activity and is associated with the levels of complement components C1q, C4 and C3, suggesting that phagocytic defects in SLE may be induced by even a partial reduction in complement level. These lines of evidence suggest that, as for complete C3 deficiency, decreased levels of C3 may be associated with SLE via a mechanism involving impaired clearance of immune complexes. Indeed, SLE patients carrying one copy of the rs7951 T allele had significantly lower levels of serum C3 than did SLE patients with no rs7951 T allele, as shown in Fig. 3 (P ¼ 0.020). SLE patients carrying two copies of the rs7951 T allele have even lower levels of serum C3 (P ¼ 0.026). Many confounding factors may participate for C3 concentrations; however, the phenotypes such as antiDNA antibody titre, prednisolone dosage and renal or CNS involvement, were not significantly different between the carriers and the non-carriers of rs7951 T allele. The three SNPs rs2230201, rs2230205 and rs7951 are synonymous and are located in exons 9, 14 and 35, respectively. It is likely that this particular region or a nearby region contains a defect responsible for the pathogenesis of SLE. In conclusion, rs7951 T allele of the C3 gene was significantly associated with SLE, and decreased serum level of C3 seems to be correlated with this allele. Deficiencies in the complement system may be actively involved in the pathogenesis of SLE.

Rheumatology key messages  A total of 53 candidate genes was studied for the association with SLE by case–control studies (509 cases and 964 controls).  Complement C3 gene was associated with SLE.

Acknowledgements Funding: This work was supported in part by a Grant-in-Aid for Scientific Research on Priority Areas ‘Applied Genomics’ from the Ministry of Education, Culture, Sports, Science and Technology of Japan and in part by the Taiwan National Science Council (NSC94-2314-B075-109) and the Taipei Veterans General Hospital (VGH94-248), Taiwan.

164

H. Miyagawa et al.

Disclosure statement: The authors have declared no conflicts of interest.

Supplementary data Supplementary data are available at Rheumatology Online.

References 1 Kotzin BL. Systemic lupus erythematosus. Cell 1996;85:303–6. 2 Tsao BP. The genetics of human systemic lupus erythematosus. Trends Immunol 2003;24:595–602. 3 Wandstrat A, Wakeland E. The genetics of complex autoimmune diseases: non-MHC susceptibility genes. Nat Immunol 2001;2:802–9. 4 Tao K, Fujii M, Tsukumo SI et al. Genetic variation of Toll-like receptor 9 predispose to systemic lupus erythematosus in Japanese population. Ann Rheum Dis 2007;66:905–9. 5 Grumet FC, Coukell A, Bodmer JG, Bodmer WF, McDevitt HO. Histocompatibility (HL-A) antigens associated with systemic lupus erythematosus. A possible genetic predisposition to disease. N Engl J Med 1971;285:193–6. 6 Nath SK, Kilpatrick J, Harley JB. Genetics of human systemic lupus erythematosus: the emerging picture. Curr Opin Immunol 2004;16:794–800. 7 Tan EM, Cohen AS, Fries JF et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1272–7. 8 Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997;40:1725. 9 Wakeland EK, Liu K, Graham RR, Behrens TW. Delineating the genetic basis of systemic lupus erythematosus. Immunity 2001;15:397–408. 10 Marrack P, Kappler J, Kotzin BL. Autoimmune disease: why and where it occurs. Nat Med 2001;7:899–905. 11 Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell 2001;104:487–501. 12 Yasutomo K, Horiuchi T, Kagami S et al. Mutation in DNASE I in people with systemic lupus erythematosus. Nat Genet 2001;28:313–4. 13 Sasaki T, Tahira T, Suzuki A et al. Precise estimation of allele frequencies of singlenucleotide polymorphisms by a quantitative SSCP analysis of pooled DNA. Am J Hum Genet 2001;68:214–8. 14 Baba S, Kukita Y, Higasa K, Tahira T, Hayashi K. Single-stranded conformational polymorphism analysis using automated capillary array electrophoresis apparatuses. Biotechniques 2003;34:746–50. 15 Higasa K, Kukita Y, Baba S, Hayashi K. Software for machine-independent quantitative interpretation of SSCP in capillary array electrophoresis (QUISCA). Biotechniques 2002;33:1342–8.

16 Stephens M, Donnelly P. A comparison of Bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 2003;73:1162–9. 17 Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–5. 18 Sham P, Bader JS, Craig I, O’Donovan M, Owen M. DNA Pooling: a tool for largescale association studies. Nat Rev Genet 2002;3:862–71. 19 Spector TD, Reneland RH, Mah S et al. Association between a variation in LRCH1 and knee osteoarthritis: a genome-wide single-nucleotide polymorphism association study using DNA pooling. Arthritis Rheum 2006;54:524–32. 20 Tahira T, Okazaki Y, Miura K et al. QSNPlite, a software system for quantitative analysis of SNPs based on capillary array SSCP analysis. Electrophoresis 2006;27:3869–78. 21 Sano Y, Nishimukai H, Kitamura H et al. Hereditary deficiency of the third component of complement in two sisters with systemic lupus erythematosus-like symptoms. Arthritis Rheum 1981;24:1255–60. 22 Pickering MC, Walport MJ. Links between complement abnormalities and systemic lupus erythematosus. Rheumatology 2000;39:133–41. 23 Tsukamoto H, Horiuchi T, Kokuba H et al. Molecular analysis of a novel hereditary C3 deficiency with systemic lupus erythematosus. Biochem Biophys Res Commun 2005;330:298–304. 24 Sohn J-H, Bora PS, Suk H-J, Molina H, Kaplan HJ, Bora NS. Tolerance is dependent on complement C3 fragment iC3b binding to antigen-presenting cells. Nat Med 2003;9:206–12. 25 Bowness P, Davies KA, Norsworthy PJ et al. Hereditary C1q deficiency and systemic lupus erythematosus. Q J Med 1994;87:455–64. 26 Ross SC, Densen P. Complement deficiency states and infection: epidemiology, pathogenesis and consequences of neisserial and other infections in an immune deficiency. Medicine 1984;63:243–73. 27 Sullivan KE. Complement deficiency and autoimmunity. Curr Opin Pediatr 1998;10:600–6. 28 Carroll MC. A protective role for innate immunity in systemic lupus erythematosus. Nat Rev Immunol 2004;4:825–31. 29 Davies EJ, Steers G, Ollier WE et al. Relative contributions of HLA-DQA and complement C4A loci in determining susceptibility to systemic lupus erythematosus. Br J Rheumatol 1995;34:221–5. 30 Doherty DG, Ireland R, Demaine AG et al. Major histocompatibility complex genes and susceptibility to systemic lupus erythematosus in southern chinese. Arthritis Rheum 1992;35:641–6. 31 Dunckley H, Gatenby PA, Hawkins B, Naito S, Serjeantson SW. Deficiency of C4A is a genetic determinant of systemic lupus erythematosus in three ethnic groups. J Immunogenet 1987;14:209–18. 32 Bijl M, Reefman E, Horst G, Limburg PC, Kallenberg CG. Reduced uptake of apoptotic cells by macrophages in systemic lupus erythematosus: correlates with decreased serum levels of complement. Ann Rheum Dis 2006;65:57–63.