Device Global Header

Subscribe
or Renew

Advanced Search

Original Article

Risk Alleles for Multiple Sclerosis Identified by a Genomewide Study

  • The International Multiple Sclerosis Genetics Consortium
    • *

Abstract

Background

Multiple sclerosis has a clinically significant heritable component. We conducted a genomewide association study to identify alleles associated with the risk of multiple sclerosis.

Methods

We used DNA microarray technology to identify common DNA sequence variants in 931 family trios (consisting of an affected child and both parents) and tested them for association. For replication, we genotyped another 609 family trios, 2322 case subjects, and 789 control subjects and used genotyping data from two external control data sets. A joint analysis of data from 12,360 subjects was performed to estimate the overall significance and effect size of associations between alleles and the risk of multiple sclerosis.

Results

A transmission disequilibrium test of 334,923 single-nucleotide polymorphisms (SNPs) in 931 family trios revealed 49 SNPs having an association with multiple sclerosis (P<1×10−4); of these SNPs, 38 were selected for the second-stage analysis. A comparison between the 931 case subjects from the family trios and 2431 control subjects identified an additional nonoverlapping 32 SNPs (P<0.001). An additional 40 SNPs with less stringent P values (<0.01) were also selected, for a total of 110 SNPs for the second-stage analysis. Of these SNPs, two within the interleukin-2 receptor α gene (IL2RA) were strongly associated with multiple sclerosis (P=2.96×10−8), as were a nonsynonymous SNP in the interleukin-7 receptor α gene (IL7RA) (P=2.94×10−7) and multiple SNPs in the HLA-DRA locus (P=8.94×10−81).

Conclusions

Alleles of IL2RA and IL7RA and those in the HLA locus are identified as heritable risk factors for multiple sclerosis.

Introduction

Multiple sclerosis, the most common neurologic disease affecting young adults, is an inflammatory, presumed autoimmune disorder in which lymphocytes and macrophages infiltrate the central nervous system,1-3 sometimes together with antibodies and complement.4,5 Axonal transection with neuronal loss can be an early event in the disease.6 Systemically, there is subtle dysregulation of cellular and humoral immune responses with a loss of regulatory T-cell function.7

Studies of twins and sibling pairs suggest that genetic factors influence susceptibility to multiple sclerosis; the evidence indicates that multiple genes, each exerting only modest effects, probably play a part.3,8 Candidate-gene studies have validated associations between multiple sclerosis and polymorphic variants within the major histocompatibility complex (MHC), but no other loci with a definitive association with the disease have been found. Early efforts to screen the genome for linkage with the use of low-density maps of microsatellites were unsuccessful.9-11 On the assumption that this method lacked the statistical power to identify genetic variants with associations that are not easily detected, we used a more powerful linkage scan in which we analyzed 4506 single-nucleotide polymorphisms (SNPs) in 2692 samples from 730 multiplex families with multiple sclerosis. This analysis revealed linkage with genomewide significance in the MHC region (maximum logarithm of the odds [LOD] score, 11.66), but no other region having a significant linkage with multiple sclerosis was identified.12 These results indicate that in multiple sclerosis, linkage studies lack the statistical power to detect susceptibility loci that may reside outside the MHC region.

Association studies have greater statistical power than linkage studies to detect common genetic variants that confer a modest risk of a disease.13 Genomewide association analyses (see Glossary), which are unbiased, “hypothesis-free” scans of the genome, have identified susceptibility loci outside the MHC region in type 2 diabetes,14-17 inflammatory bowel disease,18-20 rheumatoid arthritis,21 systemic lupus erythematosus,22 and type 1 diabetes.23,24 Here we present the results of a large-scale genomewide association scan aimed at identifying alleles associated with multiple sclerosis.

Figure 1. Figure 1. Genotyping, Quality Control, and Strategies for Analysis.

Panel A shows an outline of the experiments, in which 1003 family trios were initially identified for genotyping. After quality control of DNA and SNPs, a final set of 931 family trios and a total of 334,923 SNPs survived the exclusion criteria and were included in the analysis. Panel B shows the sources of patients with multiple sclerosis and the results of transmission disequilibrium testing of the family trios, Cochran–Mantel–Haenszel testing of the case and control subjects, and SNP selection criteria. Panel C shows the results of the replication phase, in which 174 SNP assays were initially developed, with 22 of these being redundant on the basis of transmission disequilibrium testing. These assays either served as internal quality controls and validation SNPs or ultimately failed Sequenom genotyping quality checks. The remaining 152 SNPs, which included 12 that were not present on the microarray chips, were selected on the basis of their location within candidate genes (regions previously identified in other autoimmune diseases). These SNPs were genotyped in multiplexed pools, and three SNPs showing less than 97% concordance with the microarray data were dropped from the analysis. In addition, a total of seven SNPs with a call rate of less than 90% and excess mendelian errors were also dropped. Concordance for the remaining SNPs was 99.6%. As a result, 20 of the SNPs that were included in the first analysis did not meet the selection criteria for the final analysis. MAF denotes minor allele frequency, HW Hardy–Weinberg equilibrium, ME mendelian errors, WTCCC Wellcome Trust Case Control Consortium, NIMH National Institute of Mental Health, and IMSGC International Multiple Sclerosis Genetics Consortium.

To maximize genotyping efficiency, we used a staged approach (Figure 1). First, we used a DNA microarray (GeneChip Human Mapping 500K Array Set, Affymetrix) to examine most of the common genetic variants in 1003 family trios, consisting of a patient with multiple sclerosis and both parents.25 After removing SNPs and DNA samples with low genotyping rates, excessive mendelian errors, or low frequencies of minor alleles from further analysis, there remained a set of 334,923 SNPs that were genotyped in 931 family trios. These markers capture more than 2.2 million common SNPs (minor allele frequency, ≥0.05) observed in the HapMap26 CEPH (Centre d'Etude du Polymorphisme Humain) subjects, consisting of Utah residents with northern and western European ancestry (CEU, release 21), with an average pairwise coefficient of determination (r2) of 0.77 (62% with ≥0.8).

In the second (replication) stage of the investigation, 110 SNPs, most of which yielded a signal in the first phase, were genotyped in additional family trios, case subjects, and control subjects. In both stages, the analysis was supplemented by data from previously genotyped control subjects. The overall sample contained 1540 family trios, 2322 case subjects, and 5418 control subjects for a total of 12,360 samples. The combined results from this genomewide association scan identified genetic variants associated with multiple sclerosis that were not in the MHC region.

Methods

Patients and Controls

Table 1. Table 1. Demographic Characteristics of the Patients.

Table 1 lists the demographic features of the case subjects, who all received the diagnosis of multiple sclerosis on the basis of reliable clinical criteria.8,27,28 Subjects with clinically isolated syndromes or neuromyelitis optica29 were excluded from samples in the United Kingdom, whereas 4% of subjects in the United States had a clinically isolated syndrome at the time of enrollment. Healthy control subjects from Brigham and Women's Hospital in Boston and the University of California at San Francisco consisted of unrelated people who reported themselves as being non-Hispanic whites and free of chronic inflammatory disease. (For details, see the Supplementary Appendix, available with the full text of this article at www.nejm.org.)

Sample Preparation and Genetic Fingerprinting

We used methods similar to those described in a recently performed genomewide association scan.15 A minimum of 1 μg of genomic DNA (diluted in 1× TE buffer at 50 ng per microliter) from case subjects and controls was arrayed on 96-well master plates at the project's centralized DNA bank. Before scanning, DNA concentrations were determined by fluorescence measurement with molecular probes (PicoGreen, Molecular Probes). As a genetic fingerprint, a panel of 24 SNPs, including a sex-confirmation assay, was genotyped with the use of the Sequenom platform. Twenty-three of these SNPs are included on both of the chips in the Affymetrix GeneChip Human Mapping 500K arrays and served as a cross-platform sample verification.

HLA Typing

Medium-resolution typing of HLA-DRB1 and HLA-DQB1 (two to four digits) was performed on the 931 family trios in the first screening step.30 Of the 931 case subjects in these families, 531 (57.0%) carried at least one copy of HLA-DRB1*1501. Because complete HLA-DRB1 genotype data for all members of the replication sets were not available, all the consortium subjects (trio family members, case subjects, and control subjects) were genotyped for the rs3135388 (A/G) SNP to identify the DRB1*1501 allele associated with multiple sclerosis, since the presence of this allele and the rs3135388A SNP are highly correlated.31 Both HLA-DRB1*1501 and rs3135388 genotyping results were available for 2757 of 2793 subjects (98.7%) in the 931 family trios. Data from 2730 of these 2757 subjects showed complete concordance between the rs3135388A SNP and the DRB1*1501 genotype (>99% for tagging the correct number of DRB1*1501 alleles).

Methods of genomewide scanning, DNA fingerprinting concordance, quality control, SNP exclusion criteria, additional control data, statistical analysis, technical validation, and replication genotyping are described in the Supplementary Appendix.

Results

Screening Phase

Figure 2. Figure 2. Overview of the Primary Genomewide Association Scan Involving 931 Family Trios.

P values (shown as –log10 values) for results of transmission disequilibrium testing are plotted across the genome. The classic HLA-DR risk locus on chromosome 6p21 stands out with strong statistical significance (P<1×10−81).

Figure 3. Figure 3. Observed and Expected Distributions for the Results of Transmission Disequilibrium Testing from the 931 Family Trios in the Primary Genome Scan.

P values (shown as –log10 values) for all 334,923 SNPs (in gray) are plotted against the expected null distribution (red). After exclusion of the SNPs across the extended major-histocompatibility-complex region, the observed distribution (in black) matches the expected distribution closely, with an excess in the tail at P<0.001 (more associated SNPs than expected under the null hypothesis).

Figure 1 shows an outline of the experiments. Figure 2 shows results from the transmission disequilibrium testing of the 334,923 SNPs typed in the 931 family trios that were included in the screening stage. A plot of the association results for the initial genome scan is shown in Figure 3; the P values for all SNPs are plotted as a function of P values from the expected (uniform) null distribution. After exclusion of the SNPs across the extended MHC region, the observed distribution closely matches the expected (null) distribution (genomic inflation factor, 1.05) with an excess in the tail at P<0.001 (more associated SNPs observed than expected under the null hypothesis).

The initial screening phase revealed 78 SNPs outside the MHC region with P<1×10−4 (by transmission disequilibrium testing, indicating overrepresentation of risk alleles in case subjects, as compared with those in control subjects). These SNPs were selected for potential inclusion in the second-stage analysis. To maximize information from the screening phase, we compared the case subjects from the 931 family trios with control data from a random selection of 1475 subjects from the Wellcome Trust Case Control Consortium (WTCCC) and 956 subjects provided by the National Institute of Mental Health (NIMH), for a total of 2431 subjects. A comparison between data for these case subjects and for controls with the use of the Cochran–Mantel–Haenszel method identified an additional 63 SNPs with P<0.001. A further 24 SNPs with P<0.01 (in either the family trios or in case–control analyses) were selected for follow-up on the basis of proximity to previously identified loci associated with susceptibility to autoimmune diseases. Finally, nine SNPs that appeared to show highly significant associations but that were likely to represent genotyping artifacts were also included. (As expected, none of these nine SNPs showed any evidence of association in the replication phase.)

Replication Phase

Table 2. Table 2. Results of Replication Analysis of SNPs Chosen from the Screening Analysis.

The Sequenom genotyping platform was used to genotype these 174 SNPs in a second set of 609 family trios, 2322 case subjects, and 789 control subjects from the International Multiple Sclerosis Genetics Consortium. These data were supplemented by an independent set of 1475 control subjects from the WTCCC and 723 from the NIMH, for a total of 2987 controls. Of the 174 SNPs, 22 failed assay design or were redundant, 10 failed our replication assay quality control, and 20 failed to meet our original scan quality control; also, 12 that were not included on the Affymetrix array were added for scientific interest. Table 1 of the Supplementary Appendix shows results from the remaining 110 SNPs for this second-stage analysis alone and in combination with data from the screening phase in the form of an extension analysis. Table 2 shows results for the top 16 non-MHC SNPs (a number that was chosen arbitrarily) and the HLA-DRB1 surrogate rs3135388, showing evidence for an association with multiple sclerosis in both stages of the study.

The screening sample and part of the replication sample were family-based, but the final results could in principle be inflated owing to case–control stratification. Since we did not have an adequate set of genomic control genotyping to assess inflation, in the replication sample, we evaluated inflation in the independent case–control analysis performed during the screening to prioritize markers for follow-up (but not included in the final significance estimates). The replication cases were ascertained similarly and primarily from the same collections. Since the WTCCC and NIMH control subjects were randomly divided between the screening and replication phases, this step provided an evaluation of the potential problem. Here, the case subjects from the United Kingdom (as compared with the WTCCC control subjects) had a genomic inflation factor of 1.02, and the U.S. case subjects (as compared with the NIMH control subjects) had a genomic inflation factor of 1.09, suggesting only a modest contribution to the overall study that was confined to only a subset of the replication sample.

Combined Analysis

A combined analysis including all 1540 family trios, 2322 case subjects, and 5418 control subjects (a total of 12,360 subjects) gave final estimates of effect size. The program UNPHASED, a software application for performing genetic association analysis in nuclear families and unrelated subjects, implements maximum-likelihood inference on haplotype and genotype effects while allowing for missing data, such as uncertain phase and missing genotypes.32 Table 2 shows the results of the combined analysis.

Figure 4. Figure 4. Regional Plots for Associations in IL2RA and IL7RA.

All genotyped SNPs in the primary scan are plotted with their P values (shown as –log10 values) for transmission disequilibrium testing as a function of genomic position (with National Center for Biotechnology Information's Build 35). The SNP with the most significant association in the joint analysis is shown (blue diamond) with its initial P value in the genome scan (red diamond). Estimated recombination rates (taken from HapMap) are plotted (light blue) to reflect the local linkage disequilibrium (LD) structure around the associated SNPs and their correlated proxies (red denotes stronger linkage disequilibrium, and white denotes no linkage disequilibrium). Gene annotations were adapted from the University of California at Santa Cruz Genome Browser (http://genome.ucsc.edu/).

A number of allelic variants had a significant association with multiple sclerosis. Of these, two SNPs in intron 1 of the IL2RA gene encoding the alpha chain of the interleukin-2 receptor (also called CD25, located at chromosome 10p15) are notable: rs12722489 (P=2.96×10−8; odds ratio, 1.25; 95% confidence interval [CI], 1.16 to 1.36) and rs2104286 (P=2.16×10−7; odds ratio, 1.19; 95% CI, 1.11 to 1.26) (Figure 4). These SNPs are in strong linkage disequilibrium with each other (coefficient of determination, 0.62 from HapMap CEU26). A nonsynonymous coding SNP (rs6897932) in exon 6 of IL7RA, a gene located on chromosome 5p13 that encodes a transmembrane domain of the IL7Rα chain of the interleukin-7 receptor (CD127), also showed highly significant evidence of association with multiple sclerosis (P=2.94×10−7; odds ratio, 1.18; 95% CI, 1.11 to 1.26) (Figure 4).

There were 11 other SNPs with a final P<1×10−4. These included a SNP in the KIAA0350 region (P=3.83×10−6) at 16p13 (a locus associated with susceptibility to type 1 diabetes) and a SNP in the gene encoding the CD58 molecule (P=1.90×10−5), present within a previously identified admixture locus at 1p13.33 The HLA-DR locus was unequivocally associated with multiple sclerosis (P=8.94×10−81; odds ratio, 1.99; 95% CI, 1.84 to 2.15).

Analysis of the 925 SNPs from the MHC region (positions between 29 and 34 Mb on chromosome 6) conditional on HLA-DRB1*1501 revealed a highly significant residual association signal peaking at rs9270986 (P=1.83×10−17; odds ratio, 5.80; 95% CI, 3.53 to 9.53), which lies close to DRB1. A portion of this residual signal is probably related to allelic heterogeneity at DRB1.30

Discussion

We report on a genomewide association study of multiple sclerosis that examined a significant fraction of common variations in the human genome. Using this technique, we identified a set of SNPs located outside the MHC region that are associated with multiple sclerosis. Among the most significant associations are SNPs in genes encoding the IL2Rα and IL7Rα chains. The IL2Rα chain has been implicated in the pathogenesis of type 1 diabetes24 and Graves' disease.34 These results add to the evidence from pathological and immunologic studies that multiple sclerosis is an autoimmune inflammatory disorder.35

The evidence that certain alleles of the genes encoding IL2Rα and IL7Rα are associated with multiple sclerosis supports the idea that polymorphisms within genes related to the regulation of the immune response are important factors in multiple sclerosis.36 In particular, regulatory T cells expressing CD4 and CD2537 show a loss of function in a number of autoimmune disorders.7,38-41 Moreover, the dominant effect of disrupting the function of the interleukin-2 gene in mice is an autoimmune disease characterized by dysfunction of CD4+CD25high regulatory T cells.42,43 This evidence suggests a link between a susceptibility variant in IL2RA and the pathogenic events that result in multiple sclerosis. It is important to acknowledge, however, that CD25 — the protein encoded by IL2RA — is not a specific marker of regulatory T cells. A SNP in the IL2RA gene that has recently been implicated in multiple sclerosis44 (P=0.04) appears to be incorrectly identified as monomorphic on the HapMap26 and may represent a chance observation that is unrelated to multiple sclerosis. The IL7Rα chain, a component of the receptor for interleukin-7, has also been implicated in multiple sclerosis by measurements of messenger RNA expression and by candidate-gene approaches.45-47 (Of the 12,360 case subjects and control subjects that we analyzed in this study, 6717 were also analyzed by Gregory et al.47) Interleukin-7 is important for homeostasis of the memory T-cell pool48 and may also be important for the generation of autoreactive T cells in patients with multiple sclerosis.49 Moreover, the interleukin-7 receptor is critical for the development of gamma and delta T cells,50 which are among the earliest T cells observed in the inflammatory lesions of patients with multiple sclerosis.51

Although we chose the majority of SNPs for replication on the basis of the P values identified in the screening phase, it is notable that the SNPs at IL2RA and IL7RA, which ultimately had the most significant associations with multiple sclerosis, originally gave modest P values of 0.0013 and 0.0058, respectively, in the transmission disequilibrium testing. Moreover, a recent study showed an association between the same SNP in the IL7RA gene and the risk of multiple sclerosis (P=2.9×10−7).47 A combined analysis of these two studies (with the use of the nonoverlapping union of the data sets) gives a P value of 1.92×10−10 (odds ratio, 1.20; 95% CI, 1.14 to 1.27) for the total of 2027 family trios, 2842 case subjects, and 6717 control subjects. Although this P value has not been adjusted for multiple hypothesis testing, it is clear that the same allelic variant in the interleukin-7 receptor has been identified in several studies. Of note, this SNP introduces a coding change (T244I) that alters the ratio of soluble to membrane-bound interleukin-7 receptor47 and has also shown a strong association with type 1 diabetes.52 Whether the allelic variants found in this study have a primary role in initiating multiple sclerosis or influence susceptibility to multiple sclerosis is unknown.

Given the strong heritability of some autoimmune diseases, we speculate that there are common and unique allelic variants that contribute to the particular autoimmune disease phenotype. Besides the association between MHC variants and autoimmune diseases, PTPN22 encoding lymphoid protein tyrosine phosphatase, a suppressor of T-cell activation and development,53 has emerged as an example of a gene harboring a susceptibility variant in many autoimmune diseases. The 620Trp variant (rs2476601) of PTPN22 is associated with type 1 diabetes,54,55 rheumatoid arthritis,21,56 Graves' disease,55,57 and systemic lupus erythematosus56,58 but does not contribute to susceptibility to multiple sclerosis.56,59,60 In contrast, allelic variation at IL2RA occurs in type 1 diabetes,24 Graves' disease,34 and multiple sclerosis but current results in rheumatoid arthritis do not show this association (Gregersen PK, Klareskog L: personal communication). Fine-mapping studies in large DNA collections for these diseases will shed light on the possibility of allelic heterogeneity at this locus.

Because of their modest risk ratios, each of the alleles of IL2RA and IL7RA that were identified in this genomewide scan explains a small proportion (less than 0.2%) of the variance in the risk of the development of multiple sclerosis. For each locus, our initial screen (931 family trios) had a power of only about 6% to detect these loci at P<1×10−4 and a power of less than 50% to reach P<0.01. It is highly likely that other loci with similar low risk ratios exist. Nevertheless, associations of the magnitude we found are undetectable in linkage studies; each locus confers a sibling-recurrence risk ratio of less than 1.01 and would require the scanning of hundreds of thousands of sibling pairs before a meaningful effect on regional LOD scores would be expected.

The effect sizes of the allelic variants we identified in this scan are similar to those associated with polygenic autoimmune disorders and other complex traits.61 These variants are not rare mutations of the type that occur in diseases caused by a defect in a single gene, such as muscular dystrophy or sickle cell anemia. Rather, they are polymorphic variants that also occur in normal populations. However, each is more common in patients with multiple sclerosis than in control subjects, and each has a small effect on the risk of the disease. In considering the complex genetic architecture of multiple sclerosis, we recognize that our approach has little if any statistical power to detect rare variants that could contribute to susceptibility — even those conferring a relatively large genetic risk. We anticipate that multiple sclerosis will show some degree of genetic heterogeneity and that with increasing sample sizes and better statistical power, alternative genetic mechanisms will be revealed for certain subgroups of patients with the disease. However, for most patients, we expect that the variants identified in our study and those that may emerge in follow-up studies could account for a substantial part of the heritability of multiple sclerosis in the general population.

With the identification of a larger set of genetic variants, a systems biology approach will be needed to characterize common pathways amenable to therapeutic intervention. As for our identification of a variant of IL2RA as a susceptibility element in multiple sclerosis, it is intriguing that clinical efficacy has been observed in phase 2 studies assessing a monoclonal antibody targeting the IL2Rα chain.62,63

Funding and Disclosures

Supported by grants (AP-3758-A16 and RG-2899), a Collaborative Research Award (CA-1001-A14), and a postdoctoral fellowship (FG-1718-A1, to Dr. McCauley) from the National Multiple Sclerosis Society; grants (NS049477, NS032830, and NS26799) from the National Institute of Neurological Disorders and Stroke; grants (AI067152 and P01-AI039671) from the National Institute of Allergy and Infectious Diseases; a grant (076113) from the Wellcome Trust; a grant (U54-RR020278-1) from the National Center for Research Resources; the Penates Foundation; the Nancy Davis Center Without Walls; and a Jacob Javits Merit Award (NS2427, to Dr. Hafler) from the National Institute of Neurological Disorders and Stroke.

No potential conflict of interest relevant to this article was reported.

The writing group (David A. Hafler, M.D., Alastair Compston, F.Med.Sci., Ph.D., Stephen Sawcer, M.B., Ch.B., Ph.D., Eric S. Lander, Ph.D., Mark J. Daly, Ph.D., Philip L. De Jager, M.D., Ph.D., Paul I.W. de Bakker, Ph.D., Stacey B. Gabriel, Ph.D., Daniel B. Mirel, Ph.D., Adrian J. Ivinson, Ph.D., Margaret A. Pericak-Vance, Ph.D., Simon G. Gregory, Ph.D., John D. Rioux, Ph.D., Jacob L. McCauley, Ph.D., Jonathan L. Haines, Ph.D., Lisa F. Barcellos, Ph.D., Bruce Cree, M.D., Ph.D., Jorge R. Oksenberg, Ph.D., and Stephen L. Hauser, M.D.) assumes responsibility for the overall content and integrity of the article.

This article (10.1056/NEJMoa073493) was published at www.nejm.org on July 29, 2007.

We thank the Wellcome Trust Case Control Consortium and the investigators who contributed to the generation of the data (listed at www.wtccc.org.uk); the National Institutes of Mental Health for generously allowing the use of their genotype data; the patients and their families for participating in this study; Susan Pobywjlo (Brigham and Women's Hospital) for organizing patient collections; Kathryn Irenze (Broad Institute) for technical assistance; Brendan Blumenstiel, Matt DeFelice, Melissa Parkin, and the Affymetrix production team; Marcia Nizzari, George Grant, Pei Lin, and the Broad Institute Genetic Analysis Platform Informatics team; Stacey Donnelly (Broad Institute) and Andrew J.P. Lowe (Harvard Center for Neurodegeneration and Repair) for administrative support; Sandra West (Duke University and University of Miami) and Robin Lincoln (University of California, San Francisco) for assistance with sample management; Justin Giles, David Sexton, and Yuki Bradford (Vanderbilt University) and James Jaworski (University of Miami) for assistance with analysis; and a small number of key private donors whose early vision, partnership, and contributions ultimately made this project possible.

Author Affiliations

The affiliations of the writing group and other members of the International Multiple Sclerosis Genetics Consortium are listed in the Appendix.

Appendix

The writing group's affiliations are as follows: the Division of Molecular Immunology, Center for Neurologic Diseases, Department of Neurology, Brigham and Women's Hospital, and Harvard Medical School, Boston (D.A.H., P.L.D.J.); Broad Institute of Harvard University and Massachusetts Institute of Technology, Cambridge, MA (D.A.H., E.S.L., M.J.D., P.L.D.J., P.I.W.B., S.B.G., D.B.M., J.D.R.); Department of Clinical Neurosciences, Addenbrooke's Hospital, University of Cambridge School of Clinical Medicine, Cambridge, United Kingdom (A.C., S.S.); Massachusetts General Hospital, Harvard Medical School, Boston (M.J.D., P.I.W.B.); Harvard Partners Center for Genetics and Genomics, Boston (P.L.D.J., P.I.W.B.); Harvard Center for Neurodegeneration and Repair, Harvard Medical School, Boston (A.J.I.); Duke University Medical Center, Durham, NC (M.A.P.-V., S.G.G.); University of Miami School of Medicine, Miami (M.A.P.-V.); Université de Montréal, Montreal Heart Institute, Montreal (J.D.R.); Center for Human Genetics Research, Vanderbilt University Medical Center, Nashville (J.L.M., J.L.H.); University of California at Berkeley, Berkeley (L.F.B.); and University of California at San Francisco, San Francisco (L.F.B., B.C., J.R.O., S.L.H.).

The following groups participated in this study: Clinical and Sample Collection Groups (in order of the number of samples collected): University of Cambridge School of Clinical Medicine, Cambridge, United Kingdom — S. Sawcer (project coleader), M. Ban, A. Compston; University of California at San Francisco, San Francisco — J.R. Oksenberg (project coleader), B. Cree, S.L. Hauser; Brigham and Women's Hospital, Boston — P.L. De Jager (project coleader), H.L. Weiner, D.A. Hafler. Project Management and Genotyping Centers: Harvard Center for Neurodegeneration and Repair, Boston — A.J. Ivinson (project leader); Brigham and Women's Hospital, Boston — D.A. Hafler; Broad Institute of Harvard University and Massachusetts Institute of Technology, Cambridge, MA — S.B. Gabriel, D.B. Mirel; Duke University Medical Center, Durham, NC — S.G. Gregory, M.A. Pericak-Vance. Analysis Group: Massachusetts General Hospital, Boston — M.J. Daly (project coleader), P.I.W. de Bakker; Brigham and Women's Hospital, Boston — P.L. De Jager, L.M. Maier; University of California at Berkeley, Berkeley — L.F. Barcellos, J.R. Oksenberg; University of Cambridge School of Clinical Medicine, Cambridge, United Kingdom — S. Sawcer; University of Miami School of Medicine, Miami — M.A. Pericak-Vance; and Vanderbilt University Medical Center, Nashville — J.L. McCauley, J.L. Haines (project leader).

Supplementary Material

Glossary

References (63)

  1. 1. Prineas JW, Wright RG. Macrophages, lymphocytes, and plasma cells in the perivascular compartment in chronic multiple sclerosis. Lab Invest 1978;38:409-421

  2. 2. Lucchinetti CF, Bruck W, Rodriguez M, Lassmann H. Distinct patterns of multiple sclerosis pathology indicate heterogeneity on pathogenesis. Brain Pathol 1996;6:259-274

  3. 3. Hafler DA. Multiple sclerosis. J Clin Invest 2004;113:788-794

  4. 4. Genain CP, Cannella B, Hauser SL, Raine CS. Identification of autoantibodies associated with myelin damage in multiple sclerosis. Nat Med 1999;5:170-175

  5. 5. O'Connor KC, Appel H, Bregoli L, et al. Antibodies from inflamed central nervous system tissue recognize myelin oligodendrocyte glycoprotein. J Immunol 2005;175:1974-1982

  6. 6. Trapp BD, Peterson J, Ransohoff RM, Rudick R, Mork S, Bo L. Axonal transection in the lesions of multiple sclerosis. N Engl J Med 1998;338:278-285

  7. 7. Viglietta V, Baecher-Allan C, Weiner HL, Hafler DA. Loss of functional suppression by CD4+CD25+ regulatory T cells in patients with multiple sclerosis. J Exp Med 2004;199:971-979

  8. 8. The genetics of multiple sclerosis. In: Compston A, Confavreux C, Lassmann H, et al. McAlpine's multiple sclerosis. 4th ed. Philadelphia: Churchill Livingstone/Elsevier, 2005:113-81.

  9. 9. Sawcer S, Jones HB, Feakes R, et al. A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22. Nat Genet 1996;13:464-468

  10. 10. Haines JL, Ter-Minassian M, Bazyk A, et al. A complete genomic screen for multiple sclerosis underscores a role for the major histocompatibility complex. Nat Genet 1996;13:469-471

  11. 11. Ebers GC, Kukay K, Bulman DE, et al. A full genome search in multiple sclerosis. Nat Genet 1996;13:472-476

  12. 12. Sawcer S, Ban M, Maranian M, et al. A high-density screen for linkage in multiple sclerosis. Am J Hum Genet 2005;77:454-467

  13. 13. Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science 1996;273:1516-1517

  14. 14. Sladek R, Rocheleau G, Rung J, et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature 2007;445:881-885

  15. 15. Saxena R, Voight BF, Lyssenko V, et al. Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science 2007;316:1331-1336

  16. 16. Scott LJ, Mohlke KL, Bonnycastle LL, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science 2007;316:1341-1345

  17. 17. Zeggini E, Weedon MN, Lindgren CM, et al. Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes. Science 2007;316:1336-1341

  18. 18. Duerr RH, Taylor KD, Brant SR, et al. A genome-wide association study identifies IL23R as an inflammatory bowel disease gene. Science 2006;314:1461-1463

  19. 19. Hampe J, Franke A, Rosenstiel P, et al. A genome-wide association scan of nonsynonymous SNPs identifies a susceptibility variant for Crohn disease in ATG16L1. Nat Genet 2007;39:207-211

  20. 20. Rioux JD, Xavier RJ, Taylor KD, et al. Genome-wide association study identifies new susceptibility loci for Crohn disease and implicates autophagy in disease pathogenesis. Nat Genet 2007;39:596-604

  21. 21. Begovich AB, Carlton VE, Honigberg LA, et al. A missense single-nucleotide polymorphism in a gene encoding a protein tyrosine phosphatase (PTPN22) is associated with rheumatoid arthritis. Am J Hum Genet 2004;75:330-337

  22. 22. Graham RR, Kyogoku C, Sigurdsson S, et al. Three functional variants of IFN regulatory factor 5 (IRF5) define risk and protective haplotypes for human lupus. Proc Natl Acad Sci U S A 2007;104:6758-6763

  23. 23. Smyth DJ, Cooper JD, Bailey R, et al. A genome-wide association study of nonsynonymous SNPs identifies a type 1 diabetes locus in the interferon-induced helicase (IFIH1) region. Nat Genet 2006;38:617-619

  24. 24. Vella A, Cooper JD, Lowe CE, et al. Localization of a type 1 diabetes locus in the IL2RA/CD25 region by use of tag single-nucleotide polymorphisms. Am J Hum Genet 2005;76:773-779

  25. 25. Pe'er I, de Bakker PI, Maller J, Yelensky R, Altshuler D, Daly MJ. Evaluating and improving power in whole-genome association studies using fixed marker sets. Nat Genet 2006;38:663-667

  26. 26. International HapMap Consortium. A haplotype map of the human genome. Nature 2005;437:1299-1320

  27. 27. McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121-127

  28. 28. Polman CH, Reingold SC, Edan G, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria.” Ann Neurol 2005;58:840-846

  29. 29. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Weinshenker BG. Revised diagnostic criteria for neuromyelitis optica. Neurology 2006;66:1485-1489

  30. 30. Yeo TW, De Jager PL, Gregory SG, et al. A second major histocompatibility complex susceptibility locus for multiple sclerosis. Ann Neurol 2007;61:228-236

  31. 31. de Bakker PI, McVean G, Sabeti PC, et al. A high-resolution HLA and SNP haplotype map for disease association studies in the extended human MHC. Nat Genet 2006;38:1166-1172

  32. 32. Dudbridge F. Pedigree disequilibrium tests for multilocus haplotypes. Genet Epidemiol 2003;25:115-121

  33. 33. Reich D, Patterson N, De Jager PL, et al. A whole-genome admixture scan finds a candidate locus for multiple sclerosis susceptibility. Nat Genet 2005;37:1113-1118

  34. 34. Brand OJ, Lowe CE, Heward JM, et al. Association of the interleukin-2 receptor alpha (IL-2Ralpha)/CD25 gene region with Graves' disease using a multilocus test and tag SNPs. Clin Endocrinol (Oxf) 2007;66:508-512

  35. 35. Ascherio A, Munger KL. Environmental risk factors for multiple sclerosis. I. The role of infection. Ann Neurol 2007;61:288-299

  36. 36. Baecher-Allan C, Hafler DA. Human regulatory T cells and their role in autoimmune disease. Immunol Rev 2006;212:203-216

  37. 37. Baecher-Allan C, Brown JA, Freeman GJ, Hafler DA. CD4+CD25high regulatory cells in human peripheral blood. J Immunol 2001;167:1245-1253

  38. 38. Haas J, Hug A, Viehover A, et al. Reduced suppressive effect of CD4+CD25high regulatory T cells on the T cell immune response against myelin oligodendrocyte glycoprotein in patients with multiple sclerosis. Eur J Immunol 2005;35:3343-3352

  39. 39. de Kleer IM, Wedderburn LR, Taams LS, et al. CD4+CD25bright regulatory T cells actively regulate inflammation in the joints of patients with the remitting form of juvenile idiopathic arthritis. J Immunol 2004;172:6435-6443

  40. 40. Ehrenstein MR, Evans JG, Singh A, et al. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNFalpha therapy. J Exp Med 2004;200:277-285

  41. 41. Lindley S, Dayan CM, Bishop A, Roep BO, Peakman M, Tree TI. Defective suppressor function in CD4(+)CD25(+) T-cells from patients with type 1 diabetes. Diabetes 2005;54:92-99

  42. 42. Yu A, Malek TR. Selective availability of IL-2 is a major determinant controlling the production of CD4+CD25+Foxp3+ T regulatory cells. J Immunol 2006;177:5115-5121

  43. 43. Bayer AL, Yu A, Malek TR. Function of the IL-2R for thymic and peripheral CD4+CD25+ Foxp3+ T regulatory cells. J Immunol 2007;178:4062-4071

  44. 44. Matesanz F, Caro-Maldonado A, Fedetz M, et al. IL2RA/CD25 polymorphisms contribute to multiple sclerosis susceptibility. J Neurol 2007;254:682-684

  45. 45. Teutsch SM, Booth DR, Bennetts BH, Heard RN, Stewart GJ. Identification of 11 novel and common single nucleotide polymorphisms in the interleukin-7 receptor-alpha gene and their associations with multiple sclerosis. Eur J Hum Genet 2003;11:509-515

  46. 46. Zhang Z, Duvefelt K, Svensson F, et al. Two genes encoding immune-regulatory molecules (LAG3 and IL7R) confer susceptibility to multiple sclerosis. Genes Immun 2005;6:145-152

  47. 47. Gregory SG, Schmidt S, Seth P, et al. Allelic and functional association of the interleukin 7 receptor gene α chain (IL7Rα) with multiple sclerosis. Nat Genet (in press).

  48. 48. Seddon B, Tomlinson P, Zamoyska R. Interleukin 7 and T cell receptor signals regulate homeostasis of CD4 memory cells. Nat Immunol 2003;4:680-686

  49. 49. Bielekova B, Muraro PA, Golestaneh L, Pascal J, McFarland HF, Martin R. Preferential expansion of autoreactive T lymphocytes from the memory T-cell pool by IL-7. J Neuroimmunol 1999;100:115-123

  50. 50. He YW, Malek TR. Interleukin-7 receptor alpha is essential for the development of gamma delta + T cells, but not natural killer cells. J Exp Med 1996;184:289-293

  51. 51. Wucherpfennig KW, Newcombe J, Li H, Keddy C, Cuzner ML, Hafler DA. Gamma delta T-cell receptor repertoire in acute multiple sclerosis lesions. Proc Natl Acad Sci U S A 1992;89:4588-4592

  52. 52. Todd JA, Walker NM, Cooper JD, et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 2007;39:857-864

  53. 53. Gregersen PK. Gaining insight into PTPN22 and autoimmunity. Nat Genet 2005;37:1300-1302

  54. 54. Bottini N, Musumeci L, Alonso A, et al. A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes. Nat Genet 2004;36:337-338

  55. 55. Smyth D, Cooper JD, Collins JE, et al. Replication of an association between the lymphoid tyrosine phosphatase locus (LYP/PTPN22) with type 1 diabetes, and evidence for its role as a general autoimmunity locus. Diabetes 2004;53:3020-3023

  56. 56. Criswell LA, Gregersen PK. Current understanding of the genetic aetiology of rheumatoid arthritis and likely future developments. Rheumatology (Oxford) 2005;44:Suppl 4:iv9-iv13

  57. 57. Velaga MR, Wilson V, Jennings CE, et al. The codon 620 tryptophan allele of the lymphoid tyrosine phosphatase (LYP) gene is a major determinant of Graves' disease. J Clin Endocrinol Metab 2004;89:5862-5865

  58. 58. Kyogoku C, Tsuchiya N, Wu H, Tsao BP, Tokunaga K. Association of Fcgamma receptor IIA, but not IIB and IIIA, polymorphisms with systemic lupus erythematosus: a family-based association study in Caucasians. Arthritis Rheum 2004;50:671-673

  59. 59. Begovich AB, Caillier SJ, Alexander HC, et al. The R620W polymorphism of the protein tyrosine phosphatase PTPN22 is not associated with multiple sclerosis. Am J Hum Genet 2005;76:184-187

  60. 60. De Jager PL, Sawcer S, Waliszewska A, et al. Evaluating the role of the 620W allele of protein tyrosine phosphatase PTPN22 in Crohn's disease and multiple sclerosis. Eur J Hum Genet 2006;14:317-321

  61. 61. Wang WY, Barratt BJ, Clayton DG, Todd JA. Genome-wide association studies: theoretical and practical concerns. Nat Rev Genet 2005;6:109-118

  62. 62. Rose JW, Watt HE, White AT, Carlson NG. Treatment of multiple sclerosis with an anti-interleukin-2 receptor monoclonal antibody. Ann Neurol 2004;56:864-867

  63. 63. Bielekova B, Richert N, Howard T, et al. Humanized anti-CD25 (daclizumab) inhibits disease activity in multiple sclerosis patients failing to respond to interferon beta. Proc Natl Acad Sci U S A 2004;101:8705-8708[Erratum, Proc Natl Acad Sci U S A 2004;101:17565.]

Citing Articles (931)

    Letters

    Figures/Media

    1. Figure 1. Genotyping, Quality Control, and Strategies for Analysis.
      Figure 1. Genotyping, Quality Control, and Strategies for Analysis.

      Panel A shows an outline of the experiments, in which 1003 family trios were initially identified for genotyping. After quality control of DNA and SNPs, a final set of 931 family trios and a total of 334,923 SNPs survived the exclusion criteria and were included in the analysis. Panel B shows the sources of patients with multiple sclerosis and the results of transmission disequilibrium testing of the family trios, Cochran–Mantel–Haenszel testing of the case and control subjects, and SNP selection criteria. Panel C shows the results of the replication phase, in which 174 SNP assays were initially developed, with 22 of these being redundant on the basis of transmission disequilibrium testing. These assays either served as internal quality controls and validation SNPs or ultimately failed Sequenom genotyping quality checks. The remaining 152 SNPs, which included 12 that were not present on the microarray chips, were selected on the basis of their location within candidate genes (regions previously identified in other autoimmune diseases). These SNPs were genotyped in multiplexed pools, and three SNPs showing less than 97% concordance with the microarray data were dropped from the analysis. In addition, a total of seven SNPs with a call rate of less than 90% and excess mendelian errors were also dropped. Concordance for the remaining SNPs was 99.6%. As a result, 20 of the SNPs that were included in the first analysis did not meet the selection criteria for the final analysis. MAF denotes minor allele frequency, HW Hardy–Weinberg equilibrium, ME mendelian errors, WTCCC Wellcome Trust Case Control Consortium, NIMH National Institute of Mental Health, and IMSGC International Multiple Sclerosis Genetics Consortium.

    2. Table 1. Demographic Characteristics of the Patients.
      Table 1. Demographic Characteristics of the Patients.
    3. Figure 2. Overview of the Primary Genomewide Association Scan Involving 931 Family Trios.
      Figure 2. Overview of the Primary Genomewide Association Scan Involving 931 Family Trios.

      P values (shown as –log10 values) for results of transmission disequilibrium testing are plotted across the genome. The classic HLA-DR risk locus on chromosome 6p21 stands out with strong statistical significance (P<1×10−81).

    4. Figure 3. Observed and Expected Distributions for the Results of Transmission Disequilibrium Testing from the 931 Family Trios in the Primary Genome Scan.
      Figure 3. Observed and Expected Distributions for the Results of Transmission Disequilibrium Testing from the 931 Family Trios in the Primary Genome Scan.

      P values (shown as –log10 values) for all 334,923 SNPs (in gray) are plotted against the expected null distribution (red). After exclusion of the SNPs across the extended major-histocompatibility-complex region, the observed distribution (in black) matches the expected distribution closely, with an excess in the tail at P<0.001 (more associated SNPs than expected under the null hypothesis).

    5. Table 2. Results of Replication Analysis of SNPs Chosen from the Screening Analysis.
      Table 2. Results of Replication Analysis of SNPs Chosen from the Screening Analysis.
    6. Figure 4. Regional Plots for Associations in IL2RA and IL7RA.
      Figure 4. Regional Plots for Associations in IL2RA and IL7RA.

      All genotyped SNPs in the primary scan are plotted with their P values (shown as –log10 values) for transmission disequilibrium testing as a function of genomic position (with National Center for Biotechnology Information's Build 35). The SNP with the most significant association in the joint analysis is shown (blue diamond) with its initial P value in the genome scan (red diamond). Estimated recombination rates (taken from HapMap) are plotted (light blue) to reflect the local linkage disequilibrium (LD) structure around the associated SNPs and their correlated proxies (red denotes stronger linkage disequilibrium, and white denotes no linkage disequilibrium). Gene annotations were adapted from the University of California at Santa Cruz Genome Browser (http://genome.ucsc.edu/).

      Weekly table of contents

      Sign up for the free NEJM
      Weekly Table of Contents email