Linkage and interaction of loci on 1q23 and 16q12 may contribute to susceptibility to systemic lupus erythematosus.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 46, No. 11, November 2002, pp 2928–2936 DOI 10.1002/art.10590 © 2002, American College of Rheumatology Linkage and Interaction of Loci on 1q23 and 16q12 May Contribute to Susceptibility to Systemic Lupus Erythematosus Betty P. Tsao,1 Rita M. Cantor,1 Jennifer M. Grossman,1 Sung K. Kim,1 Noel Strong,1 Chak S. Lau,2 Chung-Jen Chen,3 Nan Shen,4 Ellen M. Ginzler,5 Rose Goldstein,6 Kenneth C. Kalunian,1 Frank C. Arnett,7 Daniel J. Wallace,8 and Bevra H. Hahn1 P ⴝ 0.005 to P not significant). Evidence for linkage to 1q23 and 16q12 was stronger in 68 non-Caucasian affected sibpairs than in 77 Caucasian affected sibpairs. Exclusion mapping ruled out linkage at 14q21-23 (s [sib recurrence risk or genotypic risk ratio] ⴝ 1.8). Because the pericentromeric region of chromosome 16 has been identified by genome scans in several autoimmune diseases, we postulated that it might harbor an autoimmune modifier gene. To explore this possibility, we tested for an interaction between 16q12 and 1q23, and between 16q12 and 20p12. Haplotype sharing at 1q23 increased concomitantly with increased haplotype sharing at 16q12 (P ⴝ 0.008 by nonparametric Jonckheere-Terpstra exact statistical test). No evidence supporting an interaction between 16q12 and 20p12 was observed. Analysis of sibpairs sharing 2 alleles at 16q12 also showed increased allele sharing at 1q23 (MAS from 0.56 to 0.65). Conclusion. These data support the presence of SLE susceptibility genes at 1q23 and 16q12, particularly in non-Caucasians. The skewed distribution of haplotypes suggests that genetic interaction of these two loci may affect SLE susceptibility. Objective. Six recent genome scans of different systemic lupus erythematosus (SLE) multiplex family cohorts showed multiple putative susceptibility loci. In the present study, we examined 4 previously identified loci to replicate findings of significant linkage to 1q23 and 16q12, and to support findings of suggestive linkage to 14q21-23 and 20p12 in a cohort of 115 multiethnic nuclear families containing 145 SLE-affected sibpairs. Methods. Model-free, multipoint linkage analyses (SIBPAL2, SAGE version 4.0) and exclusion mapping (GeneHunter) were performed. Results. Linkages to 1q23 (peak at D1S2675, mean allele sharing [MAS] 0.56; P ⴝ 0.003) and to 16q12 (peaks between D16S753 and D16S757, MAS 0.57; P ⴝ 0.003) were confirmed, but linkage evidence at 20p12 was weak and inconsistent (MAS 0.52–0.56; from Supported in part by grants from the NIH (AR-43814 and AI-45916), the Southern California Chapter of the Arthritis Foundation, the Paxson Family Foundation, The RGK Foundation (Austin, TX), and the Arthritis Society (Canada). The results of sibpair linkage analyses were obtained by using program package SAGE supported by USPHS resource grant P41-RR-03655 from the Division of Research Resources. 1 Betty P. Tsao, PhD, Rita M. Cantor, PhD, Jennifer M. Grossman, MD, Sung K. Kim, BS, Noel Strong, BS, Kenneth C. Kalunian, MD, Bevra H. Hahn, MD: University of California, Los Angeles; 2Chak S. Lau, MD: University of Hong Kong, Hong Kong; 3 Chung-Jen Chen, MD: Kaohsiung Medical University, Kaohsiung, Taiwan, Republic of China; 4Nan Shen, MD: Shanghai Institute of Rheumatology, Ren Ji Hospital, Shanghai Second Medical University, Shanghai, China; 5Ellen M. Ginzler, MD: State University of New York Health Science Center, Brooklyn; 6Rose Goldstein, MD: University of Ottawa, and the Ottawa Hospital Research Institute, Ottawa, Ontario, Canada; 7Frank C. Arnett, MD: University of Texas Health Science Center, Houston; 8Daniel J. Wallace, MD: Cedars-Sinai Research Institute, Los Angeles, California. Address correspondence and reprint requests to Betty P. Tsao, PhD, Division of Rheumatology, Department of Medicine, Rehabilitation Center, Room 32–59, 1000 Veteran Avenue, UCLA School of Medicine, Los Angeles, CA 90095-1670. E-mail: BTsao@mednet.ucla.edu. Submitted for publication January 16, 2002; accepted in revised form July 15, 2002. The prevalence of systemic lupus erythematosus (SLE) in the general US population is ⬃1 in 2,000, but it varies among racial and ethnic groups (e.g., it is more prevalent in Hispanics and African Americans) (1,2). In contrast, a familial prevalence of 10–12% has been documented using surveys of several hundred SLE patients who reported having at least one first-degree relative with the disease (3,4). The prevalence of SLE is estimated to be 2.6–3.9% in the first-degree relatives of SLE probands compared with 0.3–0.4% in relatives of matched controls (5–7). The concordance rate in monozygotic twins (24–56%) is ⬃10 times the rate in dizygotic twins or in siblings (2–5%) (1,8,9). The risk of 2928 SLE SUSCEPTIBILITY GENES AT 1q23 AND 16q12 developing disease in siblings of SLE patients (s [sib recurrence risk or genotypic risk ratio]) has been estimated to be 20 times higher than that in the general population, which is similar to the risk observed in other autoimmune diseases, such as type 1 diabetes and multiple sclerosis (10). Similar to many autoimmune diseases, SLE is likely to be a multifactorial disorder in which complex interactions between multiple genetic and environmental factors are involved (11,12). Numerous population-based studies have shown that SLE is associated with many genetic factors, including polymorphic alleles of certain major histocompatibility complex class II genes, Fc␥ receptors, mannosebinding ligand, interleukin-6 (IL-6), IL-10, and tumor necrosis factor ␣, as well as with deficiencies in complement components (C1q, C2, and C4) (11). Recently, 3 targeted genome scans (13–15) and 6 complete genome scans (16–21) using multiple cohorts mapped many chromosomal regions that are likely to contain SLE susceptibility genes. Using Lander and Kruglyak’s criteria for interpretation of linkage statistics (22), 6 loci (1q23 , 1q41-42 [14,18], 2q35-37 , 4p16-15.2 , 6p11-21 [16,17], and 16q12 [16,17]) reached the threshold for significant linkage to SLE (a logarithm of odds [LOD] score of 3.3 or 3.6, depending on linkage methods), and ⬎20 loci showed suggestive linkage (an LOD score of 1.9 or 2.2). Using Lander and Kruglyak’s guidelines for reporting confirmed linkage (P ⫽ 0.01 in an independent cohort) (22), confirmation of significant linkage was established for 5 loci: 1q41-42 (15,17), 1q23 (19), 2q35-37 (21), 4p16-15.2 (21), and 6p11-21 (21). Because of the large number of markers tested in whole genome scans, a replication of linkage results in an independent collection of families provides important evidence for a true susceptibility locus, and thus warrants subsequent fine-mapping studies for gene discovery. We chose to investigate 4 loci identified in such studies to extend suggestive linkage of 14q21-23 and 20p12 to SLE (16) and to seek further support for and/or confirmation of significant linkage to 1q23 and 16q12 (18). Using our independent cohort of 115 nuclear families containing 145 affected sibpairs, we confirmed linkage to SLE susceptibility at 1q23 and 16q12, observed weak evidence for linkage at 20p12, and excluded linkage at 14q21-23 of a gene conferring a genotypic risk ratio of 1.8. Because 16q12 has been mapped in genome scans of several autoimmune diseases, including Crohn’s disease (23–25), psoriasis (26), Blau syndrome (27), and rheumatoid arthritis (28,29), we postulated that it might harbor a gene modifying multiple autoimmune pheno- 2929 types. Here we report statistical evidence supporting an interaction between 1q23 and 16q12. SUBJECTS AND METHODS SLE families. This study was approved by the Human Subjects Protection Committee of the University of California, Los Angeles. Multiplex families were recruited by ascertaining ⱖ2 members with SLE, all available parents, and other siblings of nuclear families. The study cohort consisted of 115 nuclear families containing 145 SLE-affected sibpairs, in which 103 families had 2 affected sibs, 10 families had 3 affected sibs, and 2 families had 4 affected sibs. Confirmation of the classification of patients as having SLE (fulfillment of at least 4 of the 11 criteria recommended by the American College of Rheumatology [ACR] [30,31]) was performed as described previously (14). DNA preparations. DNA was isolated from blood mononuclear cells by the standard protocol. Buccal mucosa swipes were also used as the source for DNA in cases where obtaining blood samples was difficult. A swipe was obtained after 6 gentle brushings inside each cheek of a participant with a cytobrush (Medscand, Hollywood, FL). Each cytobrush with buccal mucosa cells was submerged in 1 ml of lysis buffer (0.32M sucrose, 10 mM Tris HCl [pH 7.5], 5 mM MgCl2, and 1% Triton X-100) and shaken for 30 minutes at room temperature. Following removal of the cytobrush, mucosa cells were pelleted and incubated with 200 l Chelex (Bio-Rad, Hercules, CA) solution (10% Chelex, 10 mM Tris HCl [pH 8.3], and 50 mM KCl) at 100°C for 15 minutes. Subsequently, the Chelex mixture was removed after centrifugation, and the supernatant containing DNA was used for genotyping microsatellite markers. Genotyping. Microsatellite markers at or near each potential SLE susceptibility locus were selected based on 1) their usage in the previous studies and 2) their genetic map positions (online at http://www.marshmed.org/genetics/) and their physical map positions (online at http:// genome.cse.ucsc.edu and http://www.ensembl.org). A candidate gene at 1q23, SLAM, which maps physically ⬃150-kb centromeric to the marker D1S484, was genotyped using primers flanking an intronic AT repeat. We designed the forward primer (CCTGACCAAAGCCTCTTATTT) and the reverse primer (TCTTTGAGTCATGGGCTCCT) based on the DNA sequence from clone RP11-404F10 on chromosome 1q23.1-24.1 (GenBank GenInfo Identifier 7161187). A candidate gene at 14q21-23, ESR2, contains a dinucleotide CA repeat with high heterozygosity (0.93) (32). The map position of ESR2 was obtained from the database of the National Center for Biotechnology Information (online at http:// www.ncbi.nlm.nih.gov/genemap/). Fluorescent primers were purchased from Applied Biosystems (Foster City, CA). Optimized primers of the same color were placed in the 5-l reaction mixture containing 0.1 M of each primer, 40 ng of genomic DNA, 20 mM Tris HCl (pH 8.4), 2.5 mM MgCl2, 50 mM KCl, 0.2 units of Platinum Taq DNA polymerase (Gibco BRL, Baltimore, MD), and 250 M dNTP. The polymerase chain reaction condition for microsatellite polymorphisms was 95°C for 2 minutes, 11 cycles of 94°C for 30 seconds, and 66°C for 15 seconds (and 1°C lower for 2930 TSAO ET AL each subsequent cycle), followed by 30 cycles at 94°C for 30 seconds, 55°C for 15 seconds, and 72°C for 30 seconds, with a final extension at 72°C for 10 minutes. The amplified products were pooled and electrophoresed in 4.25% polyacrylamide sequencing gels (Gibco BRL) for 2 hours at 3 kV using an ABI 377 DNA sequencer (Applied Biosystems). The gel data were analyzed by using GeneScan version 3.1 for sizing amplified products and Genotyper version 2.5 for allele assignment. Statistical analysis. Model-free multipoint linkage analysis was performed by scanning 2-cM increments to evaluate evidence for linkage by estimating marker allele sharing in 145 SLE-affected sibpairs, using the SIBPAL2 program of SAGE Version 4.0 (33). Variance component regression analysis was conducted on the identical-by-descent (ibd) sharing estimates for each of the 414 total sibpairs (including 145 pairs with both sibs affected, 204 disease-discordant sibpairs, and 65 unaffected sibpairs) and on the mean corrected values of the sibpair trait differences (sibs were coded 0 for unaffected and 1 for affected) using the same program. Empirical P values were computed by simulation. A lack of evidence for linkage at all tested markers within 14q21-23 in the current cohort was followed by exclusion mapping using the GeneHunter software package (34). Multiple closely spaced markers in both the 1q and 16p regions showed evidence of linkage to SLE. To investigate whether there might be an interaction between genes at these loci, we first haplotyped 3 markers in each region to improve their informativeness. Haplotypes were inferred using the principle of parsimony. We then categorized each sibpair by the number of haplotypes shared ibd (0, 1, or 2) in the two regions to obtain a 3 ⫻ 3 table, ordered in 2 directions. The Jonckheere-Terpstra statistic (35) was used to evaluate whether there were differences among the distributions of allele sharing in the chromosome 1 region when stratified into those sibpairs sharing 0, 1, or 2 haplotypes ibd in the chromosome 16 region. This nonparametric statistic tests for a shift in ordered distributions (sharing 0, 1, or 2 haplotypes at chromosome 1) when stratified by ordered categories (sharing 0, 1, or 2 haplotypes at chromosome 16). The analysis was conducted using StatXact software (36), and an exact P value was generated. RESULTS A total of 115 nuclear families were included in this study (62 Caucasian, 22 Asian, 17 Hispanic, 13 African American, and 1 of mixed ethnic origin). There were 463 subjects from these families: 132 parents (12 SLE-affected individuals), 244 SLE-affected siblings, and 87 SLE-unaffected siblings. Thus, a total of 256 SLE patients were included in this cohort; their demographic characteristics and major clinical manifestations are reported in Table 1. Ethnic variations in the clinical manifestations and genetics of SLE have been well recognized. We combined all of the non-Caucasian patients into 1 group for comparisons with Caucasian patients because of the small numbers of patients from Table 1. Major demographic characteristics and clinical features of the 256 patients with systemic lupus erythematosus (SLE) in the current cohort* Caucasians (n ⫽ 137) Age at diagnosis mean ⫾ SD (range) years Duration of disease mean ⫾ SD years Sex ratio, F:M Laboratory/clinical features ANA positive Arthritis Skin involvement Hematologic disorders Thrombocytopenia Anti-dsDNA positive Renal disease Pleuritis Pericarditis Vasculitis Psychosis/seizures Secondary APS Medication history Corticosteroids Antimalarials Cytotoxic drugs Non-Caucasians (n ⫽ 119) 31 ⫾ 12 (5–62) 30 ⫾ 13 (6–73) 13 ⫾ 10 10 ⫾ 9 7.6:1 7.5:1 95 84 85 58 16 55 35 39 16 9 17 12 100 85 72† 72† 21 71† 39 30 19 13 5† 9 82 63 44 80 60 49 * Except where indicated otherwise, values are the percentage of patients with a given characteristic. ANA ⫽ antinuclear antibody; anti-dsDNA ⫽ anti–double-stranded DNA; APS ⫽ antiphospholipid syndrome. † P ⬍ 0.0002–0.0007 versus Caucasian SLE patients, using contingency table for comparisons (not corrected for nonindependence of family members or for multiple testing). each minority group. Comparisons of 137 Caucasian SLE patients from 62 families with 119 non-Caucasian patients from 53 families showed no differences in age at diagnosis, duration of disease, sex ratio, and medication history. Among 12 laboratory/clinical features (mainly ACR criteria), skin involvement (malar rash, discoid rash, or photosensitivity) and psychosis/seizures appeared more frequently in Caucasian than in nonCaucasian SLE patients (P ⬍ 0.0002 and P ⬍ 0.0004, respectively), while hematologic disorders and anti– double-stranded DNA (anti-dsDNA) positivity occurred more frequently in non-Caucasian than in Caucasian SLE patients (P ⬍ 0.0006 and P ⬍ 0.0007, respectively). The other 8 laboratory/clinical features depicted in Table 1 showed no differences between Caucasians and non-Caucasians after corrections for multiple testing. We performed model-free, multipoint linkage analyses to investigate 4 putative SLE susceptibility loci using markers identical to those previously shown to have positive linkage in reported genome scans, as well as additional flanking markers. At 14q21-23, 9 markers SLE SUSCEPTIBILITY GENES AT 1q23 AND 16q12 Figure 1. Multipoint systemic lupus erythematosus (SLE) exclusion map for markers at 14q21-23. The x-axis depicts relative map positions of the tested markers. The logarithm of odds (LOD) score of ⫺2.0 at each marker is depicted by the dotted line. A gene with s ⱖ1.8 was excluded in this analysis. within a 28-cM region (D14S288, D14S978, D14S276, D14S980, D14S1038, D14S290, D14S63, ESR2, and D14S258) with intramarker distances ranging from 1 cM to 6 cM were analyzed individually. None exhibited significantly increased mean allele sharing (MAS) in the affected sibpairs. Exclusion mapping using the GeneHunter program (34) indicated that this region could be excluded with a s as small as 1.8 for the entire region (see Figure 1). Because of the important role of female sex hormones in the pathogenesis of SLE (37), ESR2, encoding estrogen receptor ␤, is a positional candidate gene for SLE. The numbers of CA repeats within ESR2 tested by us have been associated with serum levels of androgen and sex steroid hormone–binding globulin in premenopausal women (38), but not with SLE susceptibility (39). Linkage to SLE at this marker at a s of 1.5 was excluded. We evaluated evidence for linkage to SLE of 4 markers mapped within an 8-cM interval on 20p12. The individual MAS estimates at markers D20S162, D20S189, D20S186, and D20S604 in 145 multiethnic affected sibpairs were 0.56, 0.52, 0.52, and 0.55, respectively. As shown in the left panel of Figure 2A, D20S162 and D20S604 had increased MAS (P ⫽ 0.005 and P ⫽ 0.03, respectively). Linkage analysis of data from all 414 sibpairs (145 affected sibpairs, 204 disease-discordant sibpairs, and 65 unaffected sibpairs) yielded P values 2931 ranging from 0.01 to not significant (NS). These results indicate weak evidence for linkage of 20p12 to SLE. However, when the families were stratified by ethnicity, a different pattern emerged. The right panel of Figure 2A depicts stronger evidence for linkage of 20p12 to SLE in Caucasian families. D20S162 and D20S604 (at a distance from each other of ⬃8 cM or ⬃2.5 Mb) had increased MAS (0.61 and 0.59, respectively; P ⫽ 0.001 and P ⫽ 0.002, respectively) in 77 Caucasian affected sibpairs, but not in 68 non-Caucasian affected sibpairs (MAS 0.52 and 0.49, respectively; P NS at both markers). Linkage analysis of the data from all 202 Caucasian sibpairs yielded P values ranging from 0.006 to NS, while a similar analysis of the data from all 212 non-Caucasian sibpairs failed to yield any evidence of linkage. At 1q23, each of the 4 markers (D1S2635, SLAM, D1S484, and D1S2675) within a 5-cM interval exhibited increased MAS in 145 multiethnic affected sibpairs (0.56, 0.55, 0.54, and 0.56, respectively), with P values ranging from 0.003 to 0.05, as shown in the left panel of Figure 2B. The variation in levels of significance may be influenced by the informativeness of each marker and the distance between the marker and the susceptibility gene. SLAM (signaling lymphocyte activation molecule) is a member of the CD2 subgroup of the immunoglobulin superfamily encoding a surface receptor involved in the activation of T cells and natural killer cells (40). When combining marker data from all sibpairs in the pedigrees, P values of 0.001–0.01 were obtained, and D1S2675 exhibited the best evidence, as shown in the left panel of Figure 2B (MAS 0.56; P ⫽ 0.003). In the full sample of 414 total sibpairs, the evidence was also strongest for D1S2675 (P ⫽ 0.002). This marker is physically close (⬃0.1–0.7 Mb) to linkage peaks reported in genome scans (18,21). Thus, our results support the presence of an SLE susceptibility gene or genes in this region. The right panel of Figure 2B indicates stronger evidence for linkage of 1q23 to SLE in non-Caucasians than in Caucasians. The 4 tested markers within the 5-cM region had increased MAS (0.57–0.59; P ⫽ 0.005– 0.02) using data from 68 non-Caucasian affected sibpairs, as well as P values ranging from 0.0005 to 0.004 for a variance component regression linkage analysis of data from a total of 212 non-Caucasian sibpairs. Similar analyses using the data from Caucasians (77 affected sibpairs and 202 total sibpairs) yielded MAS of 0.50– 0.55 (P NS). At 16p11-q12, markers D16S753, D16S3136, D16S757, and D16S415 spanning a 10-cM region each exhibited increased MAS (0.55, 0.55, 0.53, and 0.52, 2932 TSAO ET AL Figure 2. Linkage analyses of 3 chromosome regions: A, 20p12; B, 1q23; and C, 16p11-q12. Centimorgans and megabases are plotted on the x-axes. On the y-axes, logarithms of P values are plotted for excessive mean allele sharing (MAS) in 145 systemic lupus erythematosus–affected sibpairs and regression (REG) analysis in 414 sibpairs (including pairs with both sibs affected, disease-discordant sibpairs, and unaffected sibpairs) by variance component regression linkage analyses (54). Microsatellite markers are listed below the x-axes. Values between markers were inferred at 2-cM increments. Evidence for linkage of each chromosome region tested is presented in the left panels for the total cohort and in the right panels for Caucasians (C) and non-Caucasians (NC). SLE SUSCEPTIBILITY GENES AT 1q23 AND 16q12 2933 Table 2. Statistical evidence for an interaction between systemic lupus erythematosus susceptibility loci on 1q23 and 16p11-q12* No. of haplotypes shared ibd at 16q12 locus 0 1 2 Total No. of haplotypes shared ibd at 1q23 locus 0 1 2 Total 7 (29) 9 (16) 2 (7) 18 14 (58) 29 (50) 17 (57) 60 3 (13) 20 (34) 11 (37) 34 24 (100) 58 (100) 30 (100) 112 * Values are the number (%) of affected sibpairs in each category. The distribution of haplotypes shared identically by descent (ibd) in the 1q23 region (defined by markers SLAM, D1S484, and D1S2675) was stratified by the extent of ibd haplotype sharing in the 16p11-q12 region (defined by markers D16S753, D16S3136, and D16S757). The Jonckheere-Terpstra test was conducted on data for the 112 affected sibpairs from the 112 families that had parental haplotypes at both loci (P ⫽ 0.008) (35,36). respectively) in 145 multiethnic affected sibpairs. Within this chromosomal interval, the loci at 59.8 cM and 64 cM (between D16S753 and D16S3136 and between D16S3136 and D16S757) both had increased MAS of 0.57 with a P value of 0.003 in 145 multiethnic affected sibpairs, as well as P values of borderline significance (0.045 and 0.06, respectively) in 414 total sibpairs, as shown in the left panel of Figure 2C. As shown in the right panel of Figure 2C, 68 non-Caucasian affected sibpairs exhibited evidence for increased allele sharing between markers D16S753 and D16S415 (MAS 0.54– 0.61; P ⫽ 0.001–0.07), whereas 77 Caucasian affected sibpairs (MAS 0.49–0.55) did not. The locus at 64 cM in the interval between D16S3136 and D16S757 exhibited the strongest evidence for linkage within 16p11-q12 in our multiethnic cohort (MAS 0.57; P ⫽ 0.003) and in stratified non-Caucasians (MAS 0.62; P ⫽ 0.001). This region maps ⬃4 cM centromeric to D16S415, and both were at or near a putative susceptibility locus linked to SLE in a previous genome scan (16,17). Since we analyzed an independent cohort, linkage of 16p11-q12 to SLE through the linkage result between markers D16S3136 and D16S757 (P ⫽ 0.003 in 145 multiethnic affected sibpairs or P ⫽ 0.001 in 68 non-Caucasian affected sibpairs) was confirmed. Since the 16q12 region has also been mapped in other autoimmune diseases (23–29), we considered the possibility that it may contain an immunoregulatory gene that can interact with other susceptibility loci influencing the expression of multiple autoimmune phenotypes. Therefore, we tested for an interaction between 16q12 and 1q23, and between 16q12 and 20p12. As shown in Table 2, the distribution of haplotype sharing in the 1q23 region (defined by markers SLAM, D1S484, and D1S2675) shifted toward increased sharing as the degree of haplotype sharing at chromosome 16 (defined by markers D16S753, D16S3136, and D16S757) also increased (P ⫽ 0.008 by Jonckheere-Terpstra exact statistical test) (35,36). Stratification of this sample by ethnicity resulted in reduced sample sizes. While the results were weaker, a skewed distribution of haplotypes was observed both in Caucasians and in non-Caucasians. Results of a similar analysis of the joint haplotype distributions of 16q12 and 20p12 were not significant. By stratifying on families containing affected sibpairs sharing 2 parental 16q12 haplotypes, the MAS of haplotypes defined by markers SLAM, D1S484, and D1S2675 increased to 0.65 from 0.56 in the total cohort. It appears that both regions may harbor genes important for the development of SLE, and that a gene in 16q12 might facilitate the contribution of a gene in 1q23. DISCUSSION We investigated 4 SLE-linked loci identified by previous genome scans. Using an independent cohort, we confirmed linkage to SLE at 1q23 and 16q12 (Figures 2B and C), observed weak linkage at 20p12 (Figure 2A), and ruled out linkage at 14q21-23 (Figure 1). Evidence for linkage to SLE at 1q23 and 16q12 appears to be stronger in non-Caucasians (46% of our patients, including Asians, Hispanics, and African Americans) compared with Caucasians (54% of our patients), while linkage at 20p12 was stronger in Caucasians. Most interesting is our observation of statistical evidence for an interaction between SLE susceptibility genes at 1q23 and 16q12. Ethnic differences in linkage might be attributed to many factors, including genetic heterogeneity, disease heterogeneity, and random variation. Ethnic variations in clinical manifestations in SLE patients of Caucasian, Hispanic, and African American origins have been shown in several studies. Compared with Caucasian SLE patients, non-Caucasian SLE patients meet a greater number of ACR criteria, have more disease activity, are younger at diagnosis, and have more frequent major organ (renal and cardiovascular) involvement (41–43). In our cohort, non-Caucasian patients had more frequent hematologic disorders and anti-dsDNA positivity than did Caucasian patients (Table 1). However, clinical features, such as thrombocytopenia and renal disease, that have been associated with poor prognosis (44) or more severe disease (45) had similar frequencies in our 2934 Caucasian and non-Caucasian patients. Skin involvement (malar rash, discoid rash, and photosensitivity) occurred more frequently in Caucasian than in nonCaucasian lupus patients (P ⬍ 0.0002) (Table 1). Since discoid lupus is more commonly seen in African American than in Caucasian SLE patients (43), we excluded discoid rash and found a more frequent occurrence of malar rash and photosensitivity in Caucasian than in non-Caucasian patients (P ⬍ 0.00001). However, the frequency of discoid lupus did not differ significantly between Caucasian and non-Caucasian SLE patients in this cohort. In summary, ethnic variations in clinical manifestations between Caucasian and non-Caucasian SLE patients may contribute to differences in evidence for linkage of 1q23, 16p11-q12, and 20p12. The approach of whole genome scans has been widely used to map susceptibility genes in human complex diseases. Moser et al reported an LOD score of 3.45 at FCGR2A (mapped at 171.3 Mb) using a maximized parametric linkage analysis of 94 multiethnic pedigrees multiplex for SLE (18), establishing significant linkage (22) of 1q23 to SLE. Investigators in this group subsequently extended their sample to 126 pedigrees multiplex for SLE (32% African Americans and 61% European Americans) and reported linkage at 1q22-24 (peak at D1S1679, 170.5 Mb, with an LOD score of 2.75; P ⫽ 0.009) (21). Shai et al also observed linkage to this region in an independent sample of 80 multiplex, multiethnic pedigrees (46% Caucasians and 54% Hispanics), reporting a P value of 0.005 at D1S484 (mapped at 161.6 Mb), at which both ethnic groups contributed equally to evidence for linkage (19). Our results as shown in the left panel of Figure 2B (peak at D1S2675, 170.6 Mb; P ⫽ 0.003), confirm linkage of 1q23 to SLE in a third independent SLE multiplex cohort. Our evidence for linkage to 1q23 is contributed primarily by non-Caucasians (Figure 2B, right panel), which is consistent with previous reports that African Americans show the strongest evidence for linkage of this region to SLE compared with European Americans (18,21). Based on the current available results, linkage of 1q23 to SLE does not appear to be limited to a particular ethnic group, although results are consistently stronger in non-Caucasian (or enriched) cohorts. The distances from D1S2675 to FCGR2A, D1S1679, and D1S484 have been estimated to be 0.7 Mb, 0.1 Mb, and 9 Mb, respectively (based on the human genome working draft). This region contains FCGR2A and FCGR3A, which have been associated with SLE in multiple populations (46–48). Interestingly, this region is syntenic to murine SLE susceptibility loci—the NZW- TSAO ET AL derived Sle1b and the NZB-derived Nba2 and Lbw7 (49–51). It is unclear at present whether a common susceptibility gene (or genes) is shared by murine and human SLE. The locus 16q12 also meets Lander and Kruglyak’s criterion for significant linkage in genome scans (22). In a study of a cohort of 105 SLE-affected sibpair families, the markers D16S3136 and D16S415 at 16q12 (mapped at 62.1 cM and 67.6 cM, respectively) had respective LOD scores of 3.2 and 3.6, whereas the respective LOD scores were 2.9 and 3.8 in the enlarged cohort of 187 sibpair families (16,17). Families in both cohorts are ⬃80% Caucasian. Support for linkage at D16S3136 was reported by Shai et al in an independent cohort, with a P value of 0.017 (19). Our results confirm linkage to this region as shown in Figure 2C, in which the MAS at 64 cM was significantly increased in 145 multiethnic affected sibpairs (0.57; P ⫽ 0.003) and in 68 non-Caucasian affected sibpairs (0.62; P ⫽ 0.001). This linked locus is ⬃2 cM telomeric from D16S3136 and 4 cM centromeric to D16S415, which is within the chance variation in location estimates for replication of significant linkage results of complex traits (52). We interpret our results as replicating the significant linkage of 16q12 with SLE susceptibility reported by Gaffney et al (16,17). Of note, and not previously reported, nonCaucasian affected sibpairs exhibited stronger linkage to 16q12 than did Caucasians (Figure 2C, right panel). Linkage of 20p12 to SLE, reported by Gaffney et al (16,17), peaked at marker D20S186 (at 32 cM), with an LOD score of 2.62 in those investigators’ first cohort of 105 SLE-affected sibpair families and an LOD score of 1.77 in their combined cohort of 187 families. Shai et al reported linkage to SLE at D20S115 (at 21 cM) in their cohort, with a P value of 0.012. In our multipoint linkage analysis of 145 multiethnic affected sibpairs (Figure 2A, left panel), D20S162 (at 25 cM) showed weak evidence for linkage, as did D20S604 (at 33 cM). Evidence for linkage of 20p12 to SLE appears stronger in Caucasians than in non-Caucasians (Figure 2A, right panel), in contrast to linkage to SLE of 1q23 and 16q12 (Figures 2B and C, right panels). Gaffney et al showed linkage to 14q21-23 in their first cohort, with LOD scores ranging between 1.7 and 2.8 (markers D14S288, D14S276, D14S63, and D14S258) within a 28-cM interval, but their LOD scores were not significant in their second cohort or in the combined cohort (16,17). The findings by Shai et al supported linkage of this region to SLE (markers D14S63 and D14S258 had P values of 0.04 and 0.02, respectively) (19). Our findings (Figure 1) indicate that this region is SLE SUSCEPTIBILITY GENES AT 1q23 AND 16q12 unlikely to harbor a major SLE susceptibility gene and exclude linkage to SLE at ESR2 (s ⫽ 1.5). Genetic analyses simultaneously considering multiple loci may provide insights that can facilitate identification of genes contributing to complex diseases. The observation that 16q12 has been linked to multiple autoimmune diseases raises the possibility of an autoimmune modifier gene in this region (23–29). Evidence for epistasis has been shown between IBD1 and 1p for susceptibility to inflammatory bowel disease (53) and between 4p16-15.2 and 5p15 for SLE susceptibility (21). As shown in Table 2, linkage at 1q23 is maximized in pedigrees showing haplotype sharing at 16q12, and this finding will be used to enhance our SLE gene-mapping efforts. ACKNOWLEDGMENTS We thank all participating patients and their family members, many physicians for referring patients and verifying their diagnoses, and Louise Ozaki for her technical help in this study. REFERENCES 1. Hochberg MC. The epidemiology of systemic lupus erythematosus. In: Wallace DJ, Hahn BH, editors. Dubois’ lupus erythematosus. Baltimore: Williams & Wilkins; 1997. p. 49–65. 2. Kotzin BL. Systemic lupus erythematosus. Cell 1996;85:303–6. 3. Buckman KJ, Moore SK, Ebbin AJ, Cox MB, Dubois EL. Familial systemic lupus erythematosus. Arch Intern Med 1978;138:1674–6. 4. Pistiner M, Wallace DJ, Nessim S, Metzger AL, Klinenberg JR. Lupus erythematosus in the 1980s: a survey of 570 patients. Semin Arthritis Rheum 1991;21:55–64. 5. Beckmann RP, Mizzen LA, Welch WJ. Interaction of hsp70 with newly synthesized proteins: implications for protein folding and assembly. Science 1990;248:850–4. 6. Hochberg MC. The application of genetic epidemiology to systemic lupus erythematosus. J Rheumatol 1987;14:867–9. 7. Lawrence JS, Martins CL, Drake GL. A family survey of lupus erythematosus. 1. Heritability. J Rheumatol 1987;14:913–21. 8. Block SR, Winfield JB, Lockshin MD, D’Angelo WA, Christian CL. Studies of twins with SLE: a review of the literature and presentation of 12 additional sets. Am J Med 1979;59:533–52. 9. Deapen D, Escalante A, Weinrib L, Horwitz D, Bachman B, Roy-Burman P, et al. A revised estimate of twin concordance in systemic lupus erythematosus. Arthritis Rheum 1992;35:311–8. 10. Vyse TJ, Kotzin BL. Genetic basis of systemic lupus erythematosus. Curr Opin Immunol 1996;8:843–51. 11. Tsao BP. The genetics of human lupus. In: Wallace DJ, Hahn BH, editors. Dubois’ lupus erythematosus. Philadelphia: Lippincott Williams & Wilkins; 2002. p. 97–120. 12. Cooper GS, Dooley MA, Treadwell EL, St. Clair EW, Parks CG, Gilkeson GS. Hormonal, environmental, and infectious risk factors for developing systemic lupus erythematosus. Arthritis Rheum 1998;41:1714–24. 13. Tsao BP, Cantor RM, Kalunian KC, Chen CJ, Badsha H, Singh R, et al. Evidence for linkage of a candidate chromosome 1 region to human systemic lupus erythematosus. J Clin Invest 1997;99: 725–31. 2935 14. Tsao BP, Cantor RM, Grossman JM, Shen N, Teophilov NT, Wallace DJ, et al. PARP alleles within the linked chromosomal region are associated with systemic lupus erythematosus. J Clin Invest 1999;103:1135–40. 15. Moser KL, Gray-McGuire C, Kelly J, Asundi N, Yu H, Bruner GR, et al. Confirmation of genetic linkage between human systemic lupus erythematosus and chromosome 1q41. Arthritis Rheum 1999;42:1902–7. 16. Gaffney PM, Kearns GM, Shark KB, Ortmann WA, Selby SA, Malmgren ML, et al. A genome-wide search for susceptibility genes in human systemic lupus erythematosus sib-pair families. Proc Natl Acad Sci U S A 1998;95:14875–9. 17. Gaffney PM, Ortmann WA, Selby SA, Shark KB, Ockenden TC, Rohlf KE, et al. Genome screening in human systemic lupus erythematosus: results from a second Minnesota cohort and combined analyses of 187 sib-pair families. Am J Hum Genet 2000;66:547–56. 18. Moser KL, Neas BR, Salmon JE, Yu H, Gray-McGuire C, Asundi N, et al. Genome scan of human systemic lupus erythematosus: evidence for linkage on chromosome 1q in African-American pedigrees. Proc Natl Acad Sci U S A 1998;95:14869–74. 19. Shai R, Quismorio FP Jr, Li L, Kwon OJ, Morrison J, Wallace DJ, et al. Genome-wide screen for systemic lupus erythematosus susceptibility genes in multiplex families. Hum Mol Genet 1999; 8:639–44. 20. Lindqvist AK, Steinsson K, Johanneson B, Kristjansdottir H, Arnasson A, Grondal G, et al. A susceptibility locus for human systemic lupus erythematosus (hSLE1) on chromosome 2q. J Autoimmun 2000;14:169–78. 21. Gray-McGuire C, Moser KL, Gaffney PM, Kelly J, Yu H, Olson JM, et al. Genome scan of human systemic lupus erythematosus by regression modeling: evidence of linkage and epistasis at 4p1615.2. Am J Hum Genet 2000;67:1460–9. 22. Lander ES, Kruglyak L. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet 1995;11:241–7. 23. Curran ME, Lau KF, Hampe J, Schreiber S, Bridger S, Macpherson AJ, et al. Genetic analysis of inflammatory bowel disease in a large European cohort supports linkage to chromosomes 12 and 16. Gastroenterology 1998;115:1066–71. 24. Hampe J, Schreiber S, Shaw SH, Lau KF, Bridger S, Macpherson AJ, et al. A genomewide analysis provides evidence for novel linkages in inflammatory bowel disease in a large European cohort. Am J Hum Genet 1999;64:808–16. 25. Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC, Beaugerie L, et al. Mapping of a susceptibility locus for Crohn’s disease on chromosome 16. Nature 1996;379:821–3. 26. Nair RP, Henseler T, Jenisch S, Stuart P, Bichakjian CK, Lenk W, et al. Evidence for two psoriasis susceptibility loci (HLA and 17q) and two novel candidate regions (16q and 20p) by genome-wide scan. Hum Mol Genet 1997;6:1349–56. 27. Tromp G, Kuivaniemi H, Raphael S, Ala-Kokko L, Christiano A, Considine E, et al. Genetic linkage of familial granulomatous inflammatory arthritis, skin rash, and uveitis to chromosome 16. Am J Hum Genet 1996;59:1097–107. 28. Cornelis F, Faure S, Martinez M, Prud’homme JF, Fritz P, Dib C, et al. New susceptibility locus for rheumatoid arthritis suggested by a genome-wide linkage study. Proc Natl Acad Sci U S A 1998;95: 10746–50. 29. Jawaheer D, Seldin MF, Amos CI, Chen WV, Shigeta R, Monteiro J, et al. A genomewide screen in multiplex rheumatoid arthritis families suggests genetic overlap with other autoimmune diseases. Am J Hum Genet 2001;68:927–36. 30. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997;40:1725. 31. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield 2936 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. NF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1271–7. Tsukamoto K, Inoue S, Hosoi T, Orimo H, Emi M. Isolation and radiation hybrid mapping of dinucleotide repeat polymorphism at the human estrogen receptor beta locus. J Hum Genet 1998;43: 73–4. Department of epidemiology and biostatistics, Case Western Reserve University. Statistical analysis for genetic epidemiology (SAGE). Cleveland: Case Western Reserve University; 2000. Kruglyak L, Lander ES. Parametric and nonparametric linkage analysis: a unified multipoint approach. Am J Hum Genet 1996; 58:1347–63. Hodges JL, Lehmann EL. Nonparametric statistical methods. New York: John Wiley & Sons; 1963. Cytel Software Corporation. StatXact 3 for Windows. Cambridge (MA): Cytel Corporation; 2000. Walker SE. The importance of sex hormones in systemic lupus erythematosus. In: Wallace DJ, Hahn BH, editors. Dubois’ lupus erythematosus. Philadelphia: Lippincott Williams & Wilkins; 2002. p. 307–18. Westberg L, Baghaei F, Rosmond R, Hellstrand M, Landen M, Jansson M, et al. Polymorphisms of the androgen receptor gene and the estrogen receptor beta gene are associated with androgen levels in women. J Clin Endocrinol Metab 2001;86:2562–8. Chang M, Lubahn DB, Hoffman RW. DNA microsatellite markers for estrogen receptor-beta are not associated with systemic lupus erythematosus. J Rheumatol 2001;28:924–5. Tangye SG, Phillips JH, Lanier LL. The CD2-subset of the Ig superfamily of cell surface molecules: receptor-ligand pairs expressed by NK cells and other immune cells. Semin Immunol 2000;12:149–57. Alarcon GS, Friedman AW, Straaton KV, Moulds JM, Lisse J, Bastian JM, et al, LUMINA Study Group. Lupus in Minority populations: Nature vs. Nurture. Systemic lupus erythematosus in three ethnic groups III. A comparison of characteristics early in the natural history of the LUMINA cohort. Lupus 1999;8:197–209. Alarcon GS, McGwin G Jr, Petri M, Reveille JD, RamseyGoldman R, Kimberly RP. Baseline characteristics of a multiethnic lupus cohort: PROFILE. Lupus 2002;11:95–101. Cooper GS, Parks CG, Treadwell EL, St. Clair EW, Gilkeson GS, Cohen PL, et al. Differences by race, sex and age in the clinical and immunologic features of recently diagnosed systemic lupus ery- TSAO ET AL 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. thematosus patients in the southeastern United States. Lupus 2002;11:161–7. Reveille JD, Bartolucci A, Alarcón GS. Prognosis in systemic lupus erythematosus: negative impact of increasing age at onset, black race, and thrombocytopenia, as well as causes of death. Arthritis Rheum 1990;33:37–48. Gladman DD, Goldsmith CH, Urowitz MB, Bacon P, Fortin P, Ginzler E, et al. The Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) Damage Index for systemic lupus erythematosus international comparison. J Rheumatol 2000;27:373–6. Salmon JE, Millard S, Schachter LA, Arnett FC, Ginzler EM, Gourley MF, et al. Fc gamma RIIA alleles are heritable risk factors for lupus nephritis in African Americans. J Clin Invest 1996;97:1348–54. Wu J, Edberg JC, Redecha PB, Bansal V, Guyre PM, Coleman K, et al. A novel polymorphism of FcgRIIIa (CD16) alters receptor function and predisposes to autoimmune disease. J Clin Invest 1997;100:1059–70. Tsao BP. Lupus susceptibility genes on human chromosome 1. Int Rev Immunol 2000;19:319–34. Morel L, Blenman KR, Croker BP, Wakeland EK. The major murine systemic lupus erythematosus susceptibility locus, Sle1, is a cluster of functionally related genes. Proc Natl Acad Sci U S A 2001;98:1787–92. Vyse TJ, Rozzo SJ, Drake CG, Izui S, Kotzin BL. Control of multiple autoantibodies linked with a lupus nephritis susceptibility locus in New Zealand Black mice. J Immunol 1997;158:5566–74. Kono DH, Burlingame RW, Owens DG, Kuramochi A, Baldera RS, Balomenos D, et al. Lupus susceptibility loci in New Zealand mice. Proc Natl Acad Sci U S A 1994;91:10168–72. Roberts SB, MacLean CJ, Neale MC, Eaves LJ, Kendler KS. Replication of linkage studies of complex traits: an examination of variation in location estimates. Am J Hum Genet 1999;65:876–84. Cho JH, Nicolae DL, Gold LH, Fields CT, LaBuda MC, Rohal PM, et al. Identification of novel susceptibility loci for inflammatory bowel disease on chromosomes 1p, 3q, and 4q: evidence for epistasis between 1p and IBD1. Proc Natl Acad Sci U S A 1998;95:7502–7. Elston RC, Buxbaum S, Jacobs KB, Olson JM. Haseman and Elston revisited. Genet Epidemiol 2000;19:1–17.