and Immunity (2001) 2, 263–268  2001 Nature Publishing Group All rights reserved 1466-4879/01 $15.00 www.nature.com/gene The T cell regulator SH2D2A contributes to the genetic susceptibility of multiple sclerosis

K-Z Dai1, HF Harbo1, EG Celius2, A Oturai3, PS Sørensen3, LP Ryder4, P Datta4, A Svejgaard4, J Hillert5, S Fredrikson5, M Sandberg-Wollheim6, M Laaksonen7, K-M Myhr8, H Nyland8, F Vartdal1 and A Spurkland1 1Institute of Immunology, The National Hospital, N-0027 Oslo, Norway; 2Department of Neurology, Ulleva˚l Hospital, Oslo, Norway; Departments of 3Neurology and 4Clinical Immunology, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark; 5Department of Neurology, Huddinge University Hospital, Karolinska Institute, Huddinge, Sweden; 6Department of Neurology, Lund University Hospital, Lund, Sweden; 7Turku Immunology Centre and Department of Virology, University of Turku, Turku, Finland; 8Department of Neurology, Haukeland Hospital, Bergen, Norway

The T cell specific adapter protein (TSAd) encoded by the SH2D2A gene is involved in the control of T cell activation. The gene is located in the 1q21 region, which has been implicated in susceptibility to experimental allergic encephalomyelitis in the mouse. We therefore evaluated whether a polymorphic GA repeat (GA13–GA33) within the region of the SH2D2A gene shows association to multiple sclerosis (MS). The frequency of the short alleles

GA13–16 was increased among 313 Norwegian MS patients compared to 277 healthy controls (0.332 vs 0.249, OR 1.5, Pc = 0.03). Transmission disequilibrium analysis in 146 Scandinavian families with at least two affected sibs showed increased transmission of GA16 to MS patients. No linkage or association of MS to four genetic markers flanking the SH2D2A gene was observed. After activation of naive CD4+ T cells, T cells homozygous for MS associated short alleles displayed lower level of TSAd ex vivo than T cells carrying at least one long allele, which were not associated to MS. Since the SH2D2A protein modulates T cell activation, this may be a mechanism for how short SH2D2A alleles confer susceptibility to develop MS. Genes and Immunity (2001) 2, 263–268.

Keywords: adapter protein; signal ; T cell; genetic susceptibility; multiple sclerosis

Introduction SH2D2A gene.9 Such length polymorphisms in the regu- latory regions of genes have been implied in the genetic Multiple sclerosis (MS) is a chronic inflammatory disease susceptibility to diseases.15–17 We therefore studied of the central nervous system. Both environmental and whether the SH2D2A gene contributes to the genetic sus- 1 genetic factors contribute to the disease. Several - ceptibility to MS by analyzing unrelated Norwegian MS wide screenings have suggested that over 20 genetic patients and controls, as well as Scandinavian families 2–6 regions could encompass susceptibility genes, but the with at least two affected siblings. Our results indicate only genetic factor which firmly has been shown to con- that short alleles of the SH2D2A gene contribute to the 7 tribute to MS is HLA. genetic susceptibility to develop MS. Interestingly, we 8 Recently, we cloned the cDNA as well as the SH2D2A also found that the MS associated short alleles confer 9 gene encoding a T cell specific adapter protein (TSAd), lower expression of TSAd in activated T cells than the which appears to be involved in inhibiting early T cell long alleles not associated with MS. signaling events.10 The SH2D2A gene is located close to the CD1 region on chromosome 1q21.9 This region con- fers susceptibility to develop chronic allergic encephalo- Results myelitis11 and viral induced encephalomyelitis12 as well 13 as chronic collagen induced arthritis in the mouse (as MS is associated with short alleles of the GA repeat 14 reviewed in Holmdahl ). A variable number of tandem in the SH2D2A promoter repeat (VNTR) is located in the promoter region of the Length polymorphisms in promoter regions have been shown to modulate the expression of the accompanying gene,15,16,18 as well as the genetic susceptibility to various Correspondence: Dr A Spurkland, Institute of Immunology, The National diseases.15–17 The SH2D2A gene is located in a genetic Hospital, N-0027 Oslo, Norway. E-mail: anne.spurklandȰlabmed.uio.no region, which has been implicated in the genetic suscepti- The project has received financial support from Statens La˚nekasse, bility to EAE in the mouse. We therefore postulated that EU-Commission (Project no. BMH4-CT97–2422), the Health and the GA13–33 repeat polymorphism in the SH2D2A pro- Rehabilitation Research Fund, Norwegian Research Council (NFR 9 project: 129081/310), Odd Fellow MS Society, Medinnova, and the moter region, which has a characteristic bimodal fre- MS Society of A˚ lesund. quency distribution (Figure 1), could confer susceptibility Received 22 March 2001; revised and accepted 1 June 2001 to develop MS. Multiple sclerosis and the SH2D2A gene K-Z Dai et al 264 Table 2 Transmission disequilbrium of SH2D2A promoter alleles in 146 Scandinavian MS sib pairs

␹2 a GA repeat Observed Expected ldf P alleles

GA13 49 55 2.10 NS GA14 1 1 0.06 NS GA16 180 159 17.10 0.00004 GA17 14 11 3.09 NS GA18 5 5 0.01 NS GA19 4 2 3.64 NS GA20 26 35 7.84 0.005 GA21 16 17 0.17 NS GA22 119 119 0.01 NS GA23 115 119 0.65 NS GA24 55 60 1.61 NS GA25 16 17 0.12 NS GA26 2 2 0.33 NS GA27 3 2 0.60 NS GA30 0 1 3.42 NS GA31 1 1 0.06 NS Figure 1 Two modal frequency distribution of SH2D2A GA repeat GA33 2 1 1.82 NS alleles among 277 healthy controls and 313 MS patients in Norway. ␹2 = = Overall 16df 39.07 P 0.001

We first examined the distribution of the SH2D2A a␹2 values presented are calculated on the basis of probability alleles among 313 Norwegian MS patients and 277 values for expected transmissions, according to the trans- Norwegian healthy controls. The frequency of short mission/disequilibrium test for uncertain-haplotype trans- mission.27 The ␹2 values are therefore not directly comparable to (GA13–16) repeats was increased among MS patients (0.332 28 = traditional TdT test. When performing TdT test with the 59 famil- vs 0.249; OR 1.5; Pc 0.03; Table 1; Figure 1). Accordingly, ies with one or two parents present, the GA16 allele was transmitted 2 the frequency of short GA13–16 homozygotes was 35 times, and not transmitted 15 times (␹ = 8.0, P Ͻ 0.005), whereas increased among the MS patients (12% vs 6%; OR 2.0; P the GA20 was transmitted four times, and not transmitted 12 times = (␹2 = 4.0, P Ͻ 0.05). 0.025; Table 1), while the frequency of the long GA17–33 homozygotes was decreased (45% vs 56%; OR 0.6; P =

0.01; Table 1). The frequency of GA13–16 homozygotes was higher among HLA-DR2+ patients than HLA-DR2− ted one patient from each of the 92 Swedish and Danish patients (14% (25/179) vs 9% (10/116)). This difference sibling pairs. The frequency of GA13–16 alleles among however did not reach statistical significance. these unrelated patients was 0.385 compared to 0.302 We then performed a transmission disequilibrium test among 158 unrelated Swedish and Danish controls (OR of the SH2D2A alleles in 146 Scandinavian MS sib pair 1.4, P = 0.06). Correspondingly, 16% of the patients were

families with two affected sibs. The overall transmission homozygous for GA13–16 alleles compared to 7% among of GA repeat alleles among the patients was skewed the controls (OR 2.6, P = 0.03). Thus, although statistically ␹2 = ( 16df 39.07; P 0.001). The short GA16 allele made the borderline, the strength of association of MS to GA13–16 largest contribution to the overall ␹2 statistic (Table 2). alleles as assessed by OR among the Swedish and Danish The transmission distortion was independent of the sex MS patients were comparable to that observed in the of the parent (data not shown). When the analysis was Norwegian case-control material. repeated with alleles grouped into short or long alleles, the short GA13–16 alleles were significantly more often Linkage analysis of the SH2D2A chromosomal region ␹2 = transmitted than the long alleles ( 1df 4.77; P 0.03). 1q21 In order to assess whether the observed transmission To investigate whether the observed association to distortion was mainly due to the Norwegian contribution SH2D2A alleles was due to linkage disequilibrium with to the Scandinavian sibling material, we randomly selec- another disease contributing gene in the vicinity of

Table 1 Distribution of short (GA13–16) and long (GA17–28) SH2D2A GA promoter alleles and the corresponding genotypes among Norwegian MS patients and controls

SH2D2A alleles and genotypes Norwegian OR 95% CI ␹2

MS patients Controls

= = GA13–16 2n 626 2n 554 Pc 0.332 0.249 1.5 1.2–2.0 9.4 0.03 n = 313 n = 277 P

Homozygotes GA13–16 12% 6% 2.0 1.1–3.9 5.05 0.025 Heterozygotes GA13–16/GA17–28 43% 38% 1.3 0.9–1.8 1.48 NS Homozygotes GA17–28 45% 56% 0.6 0.5–0.9 6.62 0.010

Genes and Immunity Multiple sclerosis and the SH2D2A gene K-Z Dai et al 265 SH2D2A, we performed a multi-point linkage analysis of a region extending approximately 15 cM to each side of the SH2D2A locus among the Scandinavian sib pair fam- ilies using four additional markers. Only a very weak positive LOD score was observed for the SH2D2A locus itself. None of the four other markers in the SH2D2A region on 1q21 tested in the families showed evidence of transmission distortion in MS patients (figure not shown).

The GA repeat length in the SH2D2A promoter influences the expression of the TSAd protein In order to investigate whether the SH2D2A promoter genotype influences the relative increase in TSAd expression after T-cell activation, resting CD4+ T cells enriched by negative selection from nine individuals with known SH2D2A genotypes were activated by cross-link- ing of CD3 and CD28, and the cells were harvested at different time points. The increase in TSAd expression was measured relative to the constant level of ZAP 70 expression using Western blot and digital registration of chemiluminesence signals (Figure 2a). Individuals carry- ing at least one copy of a long SH2D2A promoter allele displayed a higher induction of TSAd expression at all time points after activation compared with individuals being homozygous for short SH2D2A alleles (Figure 2b). Forty-eight hours after activation where the TSAd expression reaches its maximum,10 an eight-fold increase in TSAd expression was observed in cells having at least one long SH2D2A allele compared to a three and a half- fold increase in cells being homozygous for short SH2D2A alleles (Figure 2b).

Single nucleotide polymorphisms (SNP) are located in the promoter region of the SH2D2A gene Figure 2 TSAd expression was assessed ex vivo by Western blot In order to evaluate whether the SH2D2A promoter con- after activation of negatively selected resting CD4+ cells from nine tains other polymorphisms than the GA repeat which healthy individuals. The expression of TSAd was compared to that could influence disease susceptibility or expression of the observed for ZAP-70 before and after activation. A Jurkat TAg cell gene, we sequenced 1 kb of the promoter region line (3A3) stably tranfected with TSAd cDNA was included as a upstream of the first coding ATG in exon 1 of a total of positive control (a). In seven of the nine individuals (two SS and LL 10 individuals. Six healthy blood donors of which four and three SL genotypes) digital registration of chemiluminesence signals was performed, and the mean relative values (with standard had also been included in the ex vivo TSAd expression deviation) of TSAd expression was calculated, as shown in the dia- assay and four MS patients carrying either short GA gram. In the last two individuals (one SS and one LL genotype) alleles (GA13 or GA16) or long GA alleles (GA22 or GA23). only photographic registration of signals was available (data not Sequence data were aligned and compared to the pub- included in the figure). S stands for short alleles (GA13 or GA16), lished sequence (GenBank: AF106072). Two single nucle- while L is long alleles (GA22 or GA23)(b). otide polymorphism were identified upstream of the GA repeat (Table 3). As can be seen, the −628A and −446T alleles were only found on genotypes containing the The association of the SH2D2A promoter GA repeat

GA16 allele. polymorphism to MS could be due to a gene in linkage disequilibrium with the SH2D2A gene, because there are Discussion several immunologically important genes in this region, such as the CD1 gene cluster and the CD3Z gene. We did This study demonstrates that a polymorphism in the pro- not find evidence of linkage or association of MS to other moter region of the SH2D2A gene encoding the T cell genes in the vicinity of the SH2D2A gene. The lack of specific adapter protein TSAd may contribute to the gen- linkage to markers in the 1q21 region among MS families etic susceptibility to develop MS. The observation that is in accordance with previous genome screens.2–6 Lack the MS associated short alleles of the SH2D2A gene were of linkage in the presence of association is to be expected less expressed than long alleles not associated with MS, for genes of small effect or genes that are neither neces- suggests that this polymorphism may directly influence sary nor sufficient for the development of disease.19,20 the susceptibility to develop MS. Moreover, the finding Thus, the finding that MS was not linked to markers in that the alleles increased or decreased are clustered the vicinity of the SH2D2A gene lends support to the among short and long alleles, respectively, further streng- notion that it could be the SH2D2A gene itself and not thens the assumption that these alleles themselves con- flanking genes, which confer the disease susceptibility. tribute directly to the genetic susceptibility. The TSAd protein encoded by the SH2D2A gene

Genes and Immunity Multiple sclerosis and the SH2D2A gene K-Z Dai et al 266 Table 3 Single nucleotide polymorphisms in the SH2D2A pro- changes in the amino acid sequence could yield a dys- moter region functional protein. Variation in the expression level of a gene will probably not be hazardous to the same extent Nucleotide position from first ATG in SH2D2A a since this will only affect the fine-tuning of the function. exon1 In an evolutionary perspective, this type of variation could also be more convenient, since it will allow for b − − − Sample no. 628 446 341 individual variation in the response to environmental challenges without the risk of destroying the function of AH118a10 G A GA22 the gene. Thus, if polymorphisms modifying the 1 G A GA13 2 G A GA13 expression are selected, it is to be expected that the same 3 G A GA13 polymorphisms sometimes may have adverse affects, 4Rc W GA13/GA16 such as increasing the susceptibility to autoimmune dis- 5Ad Td GA16 eases like MS. 6 A T GA16 7 G A GA22 8 G A GA22/GA23 Subjects and methods 9 G A GA23 10 G A GA23 11 G A GA23 Unrelated patients and controls The 313 Norwegian unrelated MS patients and the 277 aThe evaluated region extended from position +6to−1000 of the healthy controls were living in the Oslo area. All patients SH2D2A promoter. bSample nos. 1, 2, 8 and 10 were MS patients, fulfilled the Poser criteria for clinically definite or labora- while nos. 4, 5, 7 and 9 were healthy persons. Sample nos. 3, 6 and tory supported definite MS.22 A total of 295 of the MS 11 were promoter fragments from two different healthy controls patients have previously been genomically typed for with different GA alleles subcloned into a luciferase reporter plas- HLA class II alleles.23 The Scandinavian control material mid. The published sequence from cosmid clone AH118a10 was included for comparison. cR = AorG;W= AorT.dThe accession consisted of 78 Danish and 80 Swedish healthy controls. numbers (SNP000574268 and SNP000574269) of the two single nucleotide polymorphisms (SNPs) are available at HGBASE Scandinavian sib pair families (http://hgbase.cgr.ki.se). A total of 150 Caucasian sib pair families (31 with both parents, 30 with only one parent and 89 with no parents) were included. All patients had clinically definite MS inhibits early T cell activation.10 Within 2 h after trig- according to the Poser criteria. Sib pairs from Denmark gering of the T cell receptor (TCR), the expression of (58 families) and Norway (40 families) were collected TSAd increased in T cells, and the increased expression from the whole country; sib pairs from Finland (14 is maintained for at least 24 h.8 Thus, TSAd may be families) were collected from the Turku area, whereas sib important in downregulating activated T cells. We pairs from Sweden (38 families) were recruited from the observed that the genotype of the SH2D2A promoter Lund and Stockholm areas. Where one or both parents influences TSAd expression in CD4+ T cells after acti- were missing, blood from the available parent and from vation through the TCR/CD3 complex. Since lower unaffected siblings was sampled. Genotypic data was expression of the inhibitory TSAd protein in the T cell obtained from 146 of these families, yielding the equival- after activation could increase the propensity of T cells ent of 162 nuclear families, out of which 146 included two to stay activated, this could be a mechanism whereby the or more affected offspring. Genotypes from a total of 675 associated SH2D2A alleles increase the susceptibility to individuals were included in the present study. The pro- develop MS. (Note added in proof: Recently it has been tocol was approved by the regional Ethics Committee, demonstrated that TSAd also may have a role as a tran- and informed consent was obtained from each patient scription factor in T cells (Marti et al, J Exp Med 2001; 193: and relatives of the patient. 1425–1430). The biological significance of this observation remains to be studied.) Typing of microsatelite markers in the 1q21 region The association of the GA repeat to MS and to the level Peripheral venous blood was sampled, and DNA was iso- of SH2D2A could be due to other poly- lated from leukocytes by standard procedures.24 In vitro morphisms of the SH2D2A gene in linkage disequilib- amplification by polymerase chain reaction (PCR) was rium with the GA repeat. Sequencing of 1 kb of the performed in 20 ␮l reaction volume containing 20 mm SH2D2A promoter upstream of the first coding ATG from Tris·HCl (pH 8.75), 10 mm KCl, 10 mm (NH4)SO4,2mm a total of 10 individuals, revealed two variant nucleotides MgSO4, 0.1% Triton X-100, 1 mm dNTP, 4 pmol primers, which were only present on chromosomes carrying the 0.4 U Taq polymerase and 0.003 U Pfu polymerase 25 GA16 allele. Whether these single nucleotide polymor- (added for avoiding the Taq slippery ). Thermal cycling phisms influence the expression of the SH2D2A gene is conditions were as follows: 94°C 15 sec, 58°C 30 sec, 72°C unknown. Polymorphisms located at other positions in 1 min 30 sec, 36 cycles on a PTC-200 Peltier Thermal the gene than in the promoter region, for instance in the Cycler (MJ Research, Watertown, MA, USA). 3Ј untranslated region, could also influence gene PCR products containing the SH2D2A GA repeat were expression level. Further studies to elucidate the role of obtained from genomic DNA using the following pri- polymorphisms on the regulation of the SH2D2A gene mers: primer 1 5Ј TCCCCATTCCCACTGCTC 3Ј; primer are clearly needed, and such studies are presently under 25ЈCCTGACCTTCATCCCTCC 3Ј. One of the primers way in our laboratory. was fluorescence-labelled. Four additional highly poly- There is generally a strong selection against polymor- morphic microsatellite markers (D1S252, D1S2140, phisms in coding regions of genes.21 This may be because D1S1653 and D1S196) were selected for linkage analysis

Genes and Immunity Multiple sclerosis and the SH2D2A gene K-Z Dai et al 267 of the chromosome 1q21 region. The markers cover 30 10.0, Genetics Computer Group (GCG), Madison, WI, cM (155–186 cM). Sequences of fluorescent labelled pri- USA), and variant positions were directly identified. mers and PCR amplification protocols of the microsatel- lite loci markers were obtained from the Genome Data- base (http://www.gdb.org/). Statistical analysis PCR products were analyzed on an ABI automatic The frequencies of short and long SH2D2A alleles in  sequencing machine (ABI Prism XL 377; Perkin–Elmer, unrelated patients and controls were compared by the ␹2 Foster City, CA, USA). Data was collected and analyzed test, using the Public Domain Software for Epidemiology  with the Gene Scan 2.10 program, and was visualized and Disease Surveillance EPI Info Version 5.01b (Center  with the Genotyper 2.0 program. for Disease Control, Epidemiology Program Office, Atlanta, GA, USA). The level of significance was set to Ex vivo assay of TSAd expression 0.05. Odds ratio (OR) was calculated according to Wolf + Resting CD4 enriched T cells from nine healthy persons, with Haldane’s continuity correction.26 The P values were carrying either two long alleles (either of GA22 or GA23), corrected where appropriate with the number of alleles two short alleles (either of GA13 or GA16) or one long observed minus 1 (ie 15) which corresponds to the num- allele and one short allele were assessed for expression ber of possible divisions of the alleles into two groups of TSAd after activation through the TCR/CD3 complex. based on their length. Transmission disequilibrium Peripheral blood mononuclear cells (PBMC) were pre- analysis in the Scandinavian sib pair material was done pared freshly by Ficoll gradient centrifugation. Mono- using the Transmit program (version 2.5), which per- cytes, B cells, activated T cells and other MHC II bearing  forms transmission-disequilibrium test for uncertain- cells were depleted using Dynabeads HLA class II 27 ␤ haplotype transmission. This method allows the (Dynal, Oslo, Norway) specific for the MHC II chain inclusion of families with one or both parents missing, (bead:cell-ratio was 6:1). CD8+ T cells were subsequently  such that all available information from the material can depleted using Dynabeads HLA class I (Dynal) specific be extracted. We also analyzed the 59 families with one for CD8 (bead:cell-ratio was 9:2). + or two parents with the original TdT test (http:// The quality of the CD4 cell enrichment was monitored spielman07.med.upenn.edu/TDT.htm).28 by flow cytometry on a FACSort flow cytometer (Becton Linkage analysis was done using an affected sib pair Dickinson, San Jose, CA, USA) by staining PBMC and + method. Under the null hypothesis of no linkage, 25% of CD4 enriched cells with monoclonal antibodies specific sib pairs are expected to share two alleles, 50% one allele for CD4 (clone SK3; Becton Dickinson), CD8 (clone SK1; and 25% no alleles. If evidence for linkage is found, a Becton Dickinson), CD16 (clone 3G8; PharMingen, San skew in favor of sib pairs sharing alleles will occur. In Diego, CA, USA) or TCR␥␦ (clone 11F2; Becton Dickinson). The CD4+ enriched cells were stimulated by families where parental genotyping data are missing the cross-linking CD3 (clone OKT3, mouse IgG2a, ATCC, probability of identity by decent (IBD) sharing can be Manassas, VA, USA) and CD28 (clone CD28.2, Mouse estimated from the underlying population allele fre- ␬ 10 quencies, which in our analysis were calculated from the IgG1, , PharMingen) as previously described. At cer- 29 tain time points after stimulation, cells were harvested family data using the SPLINK program (version 1.04). and cell pellets were stored at −70°C for subsequent The information extracted from the set of families can be assessment of TSAd and ZAP-70 expression by Western increased by performing a multipoint analysis which blot as previously described.10 All samples from one indi- allows by descent sharing of alleles from homozygous vidual were analyzed on the same gel. A standardized parents to be estimated by combining knowledge of the amount of lysate from a Jurkat Tag cell line (3A3) stably genetic map with the sharing observed at flanking mark- transfected with TSAd cDNA10 was included as a posi- ers. In our analysis we used the non-parametric multi- tive control on each blot. A Chemi Doc 2000 (Bio-Rad, point sib pair program MAPMAKER/SIBS (version 30 Milan, Italy) was used to register and analyze the light 1.09). signal of gels from seven of the nine individuals. Relative expression of TSAd was measured against the constitut- ive expression of ZAP-70. Results from different blots were compared after normalization against the relative Acknowledgements expression of TSAd observed in the positive control. We thank Dr Stephen Sawcer (University of Cambridge Sequencing of the SH2D2A promoter Neurology Unit, Addenbrooke’s Hospital, Cambridge, The promoter region of four healthy controls and four UK) for kind assistance with the linkage and trans- mission analysis. Tor Lea (Institute of Immunology, MS patients homozygous for short (either of GA13 or National Hospital, Oslo, Norway) provided the antibody GA16) or long (either of GA22 or GA23) alleles were directly sequenced from genomic DNA. Sequencing tem- for T cells as a generous gift, and he has also given valu- plates were PCR products covering the previous pub- able advice. Vibeke Sundvold (Institute of Immunology, lished SH2D2A promoter sequence9 were amplified with National Hospital, Oslo, Norway) is acknowledged for suitable primers (GenBank: AF106072). Sequencing was technical advice on the transfection procedure. Jan performed using the ABI Prism BigDye Terminator Brinchmann (Institute of Immunology, National Hospi- Cycle Sequencing Ready Reaction kit (Perkin–Elmer). The tal, Oslo, Norway) is acknowledged for advice on the + sequencing reactions were analyzed on an ABI automatic procedure for isolation of CD4 T cells and the FACS anti- sequencing machine (ABI Prism XL 377; Perkin–Elmer). bodies. Ingebjørg Knutsen (Institute of Immunology, The sequences were aligned using the pileup program of National Hospital, Oslo, Norway) is acknowledged for the GCG program package (Wisconsin Package Version her excellent technical assistance.

Genes and Immunity Multiple sclerosis and the SH2D2A gene K-Z Dai et al 268 References phism in the heme oxygenase-1 gene promoter is associated with susceptibility to emphysema. Am J Hum Genet 2000; 66: 187–195. 1 Tienari PJ. Multiple sclerosis: multiple etiologies, multiple 17 Lesch KP, Bengel D, Heils A et al. Association of anxiety-related genes? Ann Med 1994; 26: 259–269. traits with a polymorphism in the serotonin transporter gene 2 Chataway J, Feakes R, Coraddu F et al. The genetics of multiple regulatory region. Science 1996; 274: 1527–1531. sclerosis: principles, background and updated results of the 18 Shimajiri S, Arima N, Tanimoto A et al. Shortened microsatellite United Kingdom systematic genome screen. Brain 1998; 121: d(CA)21 sequence down-regulates promoter activity of matrix 1869–1887. metalloproteinase 9 gene. FEBS Lett 1999; 455:70–74. 3 Ebers GC, Kukay K, Bulman DE et al. A full genome search in 19 Risch N, Merikangas K. The future of genetic studies of complex multiple sclerosis. Nat Genet 1996; 13: 472–476. human diseases. Science 1996; 273: 1516–1517. et al 4 Haines JL, Ter-Minassian M, Bazyk A . A complete genomic 20 Greenberg DA. Linkage analysis of “necessary” disease loci ver- screen for multiple sclerosis underscores a role for the major sus “susceptibility” loci. Am J Hum Genet 1993; 52: 135–143. histocompatability complex. The Multiple Sclerosis Genetics 21 Cargill M, Altshuler D, Ireland J et al. Characterization of single- Group. Nat Genet 1996; 13: 469–471. nucleotide polymorphisms in coding regions of human genes. 5 Sawcer S, Jones HB, Feakes R et al. A genome screen in multiple Nat Genet 1999; 22: 231–238. sclerosis reveals susceptibility loci on chromosome 6p21 and 22 Poser CM, Paty DW, Scheinberg L et al. New diagnostic criteria 17q22. Nat Genet 1996; 13: 464–468. for multiple sclerosis: guidelines for research protocols. Ann 6 Kuokkanen S, Gschwend M, Rioux JD et al. Genomewide scan Neurol 1983; 13: 227–231. of multiple sclerosis in Finnish multiplex families. Am J Hum 23 Spurkland A, Celius EG, Knutsen I, Beiske A, Thorsby E, Vart- Genet 1997; 61: 1379–1387. dal F. The HLA-DQ (alpha 1*0102, beta 1*0602) heterodimer 7 Dyment DA, Sadovnick AD, Ebers GC. Genetics of multiple may confer susceptibility to multiple sclerosis in the absence of sclerosis. Hum Mol Genet 1997; 6: 1693–1698. the HLA-DR(alpha 1*01, beta 1*1501) heterodimer. Tissue Anti- 8 Spurkland A, Brinchmann JE, Markussen G et al. Molecular gens 1997; 50:15–22. cloning of a T cell-specific adapter protein (TSAd) containing an 24 Coligan JE, Kruisbeek AM, Margulies DH, Shevach EM, Strober Src (SH) 2 domain and putative SH3 and phospho- W. Isolation of whole mononuclear cells from perpheral blood tyrosine binding sites. J Biol Chem 1998; 273: 4539–4546. and cord blood. In: Coligan JE, Kruisbeek AM, Margulies DH, 9 Dai KZ, Vergnaud G, Ando A, Inoko H, Spurkland A. The Shevach EM, Strober W, eds. Current Protocols in Immunology. SH2D2A gene encoding the T cell specific adapter protein Current Protocols (a joint venture between Greene Publishing (TSAd) is localized centromeric to the CD1 gene cluster on chro- Associates, Inc. and John Wiley & Sons, Inc.): New York, 1992, mosome 1. Immunogenetics 2000; 51: 179–185. pp 7.1.1–7.1.7. 10 Sundvold V, Torgersen KM, Post NH et al. Cutting edge: T cell- 25 Lundberg KS, Shoemaker DD, Adams MW, Short JM, Sorge JA, specific adapter protein inhibits T cell activation by modulating Mathur EJ. High-fidelity amplification using a thermostable lck activity. J Immunol 2000; 165: 2927–2931. DNA polymerase isolated from Pyrococcus furiosus. Gene 1991; 11 Sundvall M, Jirholt J, Yang HT et al. Identification of murine loci 108:1–6. associated with susceptibility to chronic experimental auto- 26 Tiwari JL, Terasaki PI. The data and the statistical methods. In: immune encephalomyelitis. Nat Genet 1995; 10: 313–317. Tiwari JL, Terasaki PI, (eds). HLA and Disease Associations. 12 Teuscher C, Rhein DM, Livingstone KD et al. Evidence that Springer Verlag: New York, 1985, pp 18–27. Tmevd2 and eae3 may represent either a common locus or 27 Clayton D. A Generalization of the transmission/disequilibrium members of a gene complex controlling susceptibility to test for uncertain-haplotype transmission. Am J Hum Genet 1999; immunologically mediated demyelination in mice. J Immunol 65: 1170–1177. 1997; 159: 4930–4934. 28 Spielman RS, McGinnis RE, Ewens WJ. Transmission test for 13 Jirholt J, Cook A, Emahazion T et al. Genetic linkage analysis of linkage disequilibrium: the insulin gene region and insulin- collagen-induced arthritis in the mouse. Eur J Immunol 1998; 28: dependent diabetes mellitus (IDDM). Am J Hum Genet 1993; 52: 3321–3328. 506–516. 14 Holmdahl R. Genetics of susceptibility to chronic experimental 29 Holmans P, Clayton D. Efficiency of typing unaffected relatives encephalomyelitis and arthritis. Curr Opin Immunol 1998; 10: in an affected-sib-pair linkage study with single-locus and mul- 710–717. tiple tightly linked markers. Am J Hum Genet 1995; 57: 1221– 15 Kennedy GC, German MS, Rutter WJ. The minisatellite in the 1232. diabetes susceptibility locus IDDM2 regulates insulin transcrip- 30 Kruglyak L, Lander ES. Complete multipoint sib-pair analysis tion. Nat Genet 1995; 9: 293–298. of qualitative and quantitative traits. Am J Hum Genet 1995; 57: 16 Yamada N, Yamaya M, Okinaga S et al. Microsatellite polymor- 439–454.

Genes and Immunity