and Immunity (2013) 14, 62–66 & 2013 Macmillan Publishers Limited All rights reserved 1466-4879/13 www.nature.com/gene

SHORT COMMUNICATION -associated single-nucleotide polymorphisms in CLEC16A correlate with reduced SOCS1 and DEXI expression in the thymus

IS Leikfoss1,2, I-L Mero1,3, MK Dahle2, BA Lie3, HF Harbo1,4, A Spurkland2 and T Berge1,2

Genome-wide association studies have revealed that the 16p13 chromosomal region, including CLEC16A, DEXI, CIITA and SOCS1, is associated with susceptibility to autoimmune diseases. As non-coding single-nucleotide polymorphisms (SNPs) may confer susceptibility to disease by affecting expression of nearby genes, we examined whether autoimmune-associated intronic CLEC16A SNPs (rs12708716, rs6498169 and rs7206912) correlate with the expression of CLEC16A itself as well as neighboring genes in whole- blood and thymic samples. Real-time quantitative PCR analyses show that SOCS1 and DEXI expression was lower in thymic samples carrying at least one of the CLEC16A risk alleles compared with non-carriers of the risk allele. Linear regression analysis revealed a significant correlation between the expression level of CLEC16A and that of SOCS1 and DEXI in thymic samples. These data indicate a possible regulatory role for multiple sclerosis-associated non-coding CLEC16A SNPs and a common control mechanism for the expression of CLEC16A, SOCS1 and DEXI.

Genes and Immunity (2013) 14, 62–66; doi:10.1038/.2012.52; published online 15 November 2012 Keywords: CLEC16A; DEXI; SOCS1; CIITA; multiple sclerosis; autoimmunity

INTRODUCTION No expression quantitative trait locus were detected for other 19,20 Polymorphisms in the C-type lectin domain family 16, member A genes within the 16p13 region in these data sets. (CLEC16A) gene on 16p13, influence susceptibility CLEC16A is widely expressed in immune cells (www.genome. for development of several autoimmune disorders, including type ucsc.edu), but its immune function is still unknown. The presence 1 diabetes,1–3 Addison’s disease,4 Crohn’s disease5 and multiple of an immunoreceptor tyrosine-based activation motif (http:// sclerosis (MS),2,6,7 among others.8–10 Fine mapping of 57 CLEC16A elm.eu.org) indicates a potential role for CLEC16A in immune single-nucleotide polymorphisms (SNPs) in a combined British and cell activation. In addition to CLEC16A itself, both CIITA and SOCS1 4,15,21,22 Norwegian MS cohort revealed that rs12708716, which lies within are attractive candidate genes for autoimmune disease, intron 19 of CLEC16A, showed the strongest association to MS, and their functions in immune cells are well established. CIITA followed by two other intronic CLEC16A SNPs, rs7206912 and encodes the major histocompatibility complex class II rs6498169 (both in intron 22).3,11 Several recent studies,2,12,13 transactivator, a transcription factor required for the activation 23 including a collaborative genome-wide association study, which of the major histocompatibility complex class II gene expression. involved 9 772 MS cases and 17 376 controls, confirmed CLEC16A SOCS1 is a suppressor of cytokine signaling important for immune 24 intron 19 (rs7200786, in linkage disequilibrium (LD) with cell homeostasis and regulation of inflammation. The DEXI gene, rs12708716; D0 1, R2 0.61 (haploview v 4.2, CEU population)) which is located between CIITA and CLEC16A (Figure 1a), ¼ ¼ 25 as an MS susceptibility locus.7 Although there are independent is a dexamethasone-induced gene with unknown function. genetic signals from other genes in the 16p13 chromosomal Whole-genome expression data have shown that expression of region, for example, CIITA, SOCS1 and the intergenic CLEC16A- the adjacent DEXI, CLEC16A and SOCS1 genes are highly correlated 17 SOCS1 region,14–17 CLEC16A has been suggested to be the most in human lymphoblastoid cell lines, but not in human 19 likely candidate gene as it contains the strongest MS-associated monocytes. SNPs.17 Common for the autoimmune-associated CLEC16A SNPs In a previous report, we investigated the correlation between is that they are located in non-coding regions of CLEC16A,1–5,7,18 expression of two CLEC16A isoforms and the CLEC16A rs12708716 11 and it is uncertain whether these SNPs could influence expression genotype in whole-blood as well as in thymic samples. In the of CLEC16A itself or its neighboring genes. We have previously present report, we have extended these analyses to study gene shown an association between the CLEC16A rs12708716 genotype expression of CLEC16A as well as selected neighboring genes, and the relative expression of two CLEC16A isoforms in samples CIITA, DEXI and SOCS1, and tested for correlation with the from thymic tissue.11 Recently, an expression quantitative trait presence of the three CLEC16A SNPs (rs12708716, rs7206912 and locus for the neighboring DEXI gene was identified within intron rs6498169) that displayed the strongest association with MS in our 19 of CLEC16A in monocytes and in lymphoblastoid cell lines. fine-mapping studies of 57 CLEC16A SNPs in a combined British

1Department of Neurology, Oslo University Hospital, Oslo, Norway; 2Institute of Basic Medical Sciences, Department of Anatomy, University of Oslo, Oslo, Norway; 3Department of Medical Genetics, University of Oslo and Oslo University Hospital, Oslo, Norway and 4Institute of Clinical Medicine, University of Oslo, Oslo, Norway. Correspondence: Dr T Berge, Institute of Basic Medical Sciences, Department of Anatomy, University of Oslo, Sognsvannsveien 9, PO Box 1105 Blindern, Oslo N-0317, Norway. E-mail: [email protected] Received 9 May 2012; revised 4 September 2012; accepted 3 October 2012; published online 15 November 2012 Multiple sclerosis-associated single-nucleotide polymorphisms IS Leikfoss et al 63

Figure 1. The CLEC16A gene is located on chromosome 16p13. (a) Schematic drawing of the chromosome 16p13 genetic region comprising CIITA, DEXI, CLEC16A and SOCS1. The CLEC16A gene covers 238 kb and the three SNPs included in the study, rs12708716 (intron 19), rs6498169 and rs7206912 (both intron 22), are depicted. The B48-kb CIITA gene is located B20 kb upstream of CLEC16A, while the B14-kb-long DEXI gene is found B2.1 kb upstream of CLEC16A. SOCS1 is a small gene, only B1.8 kb, and is positioned B72 kb downstream of CLEC16A.(b)LD plot of D0-LD of the three indicated CLEC16A SNPs is based on samples typed in.11

Figure 2. DEXI and SOCS1 expression is affected by CLEC16A genotype. The graphs show relative expression of CIITA, DEXI, CLEC16A and SOCS1 normalized to TBP in thymic tissue from 37 individuals genotyped for three CLEC16A risk SNPs. Individuals carrying the risk allele were compared with individuals homozygous for the protective allele. (a) rs12708716 (risk allele A): AA/AG: n ¼ 31, GG: n ¼ 6. (b) rs6498169 (risk allele G): GG/AG: n ¼ 21, AA: n ¼ 15. (c) rs7206912 (risk allele G): GG/CG: n ¼ 28, CC: n ¼ 9. Correlation between gene expression levels and genotypes were assessed by Mann–Whitney U-test. Significant P-values are indicated with asterisks (*P-value, 0.05–0.01; **P-value, 0.01–0.001). Horizontal lines indicate the median values within the groups. and Norwegian cohort.11 The presence of CLEC16A MS risk alleles may be overshadowed by variations in the proportions of the correlated with reduced SOCS1 and DEXI expression, and linear various cell types. regression analyses revealed that DEXI, CLEC16A and SOCS1 were Gene expression was then measured in the more homogenous co-expressed in thymic samples. thymic tissue samples, which consists primarily of T cells. For each genotype, we consistently grouped individuals homozygous for the CLEC16A risk allele together with individuals heterozygous for RESULTS AND DISCUSSION the risk allele, and compared with the individuals being Non-coding SNPs may confer susceptibility to disease develop- homozygous for the protective allele. We observed a significantly ment by affecting the expression of nearby genes in cis.26,27 lower SOCS1 expression in thymic samples carrying at least As genetic variants can mediate tissue-dependent variations in one CLEC16A risk allele (Figure 2, right diagrams). Similarly, DEXI gene expression, we analyzed the expression of CLEC16A itself and expression was lower in the thymic tissue samples from its nearby genes, CIITA, DEXI and SOCS1, both in whole-blood individuals carrying the CLEC16A rs6498169 risk allele compared (n ¼ 24) and thymic tissue samples (n ¼ 37) using quantitative real- with non-carriers (Figure 2b). A similar trend for DEXI expression time PCR. These samples were genotyped for three CLEC16A SNPs, was observed also for the other two CLEC16A SNPs; however, this that is, rs12708716 (intron 19), rs6498169 and rs7206912 (both difference did not reach statistical significance, probably owing to in intron 22), conferring the most significant CLEC16A associations the small sample size in the non-carrier group. These data clearly to MS in a combined Norwegian and British cohort.11 We found no indicate that there is correlation between the given CLEC16A significant CLEC16A genotype-dependent differences in CLEC16A, genotypes and DEXI and SOCS1 expression in thymic samples. DEXI, SOCS1 or CIITA expression in the whole-blood samples (data We cannot rule out that the discrepancy between the results in not shown). Whole-blood samples are heterogeneous in their cell whole-blood and in thymic samples might be owing to individual composition, so any cell-specific, genetically determined variation differences, as whole-blood and thymic samples are harvested

& 2013 Macmillan Publishers Limited Genes and Immunity (2013) 62 – 66 Multiple sclerosis-associated single-nucleotide polymorphisms IS Leikfoss et al 64 from different donor groups of different age and health state. Lack thymocyte development and suggests a potential mechanism for of power could also explain the lack of correlation in the whole- development of autoimmune disease. blood samples. Permutation analyses using the R-software (R core Correlation between CLEC16A expression and that of SOCS1 and team, Vienna, Austria, 2012) indicates that we have the power to DEXI has previously been observed in lymphoblastoid cell lines, detect a correlation in SOCS1 expression in samples sorted for but not in monocytes.17,19 As B30% of genes show discordant the rs7206912 genotype (lower right graph in Figure 2). However, tissue-dependent regulation,32 we investigated whether there for the other tests giving significant correlation between gene were any correlations between the levels of the four selected expression and CLEC16A genotype in the thymic tissue, we lack genes in whole-blood and in thymic tissue samples. A linear power for the whole-blood samples (data not shown). regression analysis was performed for each pair-wise combination Expression of total CLEC16A and CIITA was not affected by the of CIITA, DEXI, CLEC16A and SOCS1gene expression levels. No CLEC16A genotype in thymic tissue (Figure 2), which is correlations were found in the whole-blood samples (data not in accordance with previous reports from us and others.11,17,19 shown), whereas in the thymic tissue samples the expression As we applied expression assays covering all the CIITA variants and levels of CLEC16A and DEXI (R2 ¼ 0.27, Po0.001) as well as the two long CLEC16A isoforms, we cannot exclude the possibility CLEC16A and SOCS1 (R2 ¼ 0.63, Po0.0001) were found to be that disease-associated CLEC16A SNPs could affect the expression correlated. In addition, a significant correlation between DEXI and of particular isoforms of these genes. Of note, in a previous SOCS1 expression levels (R2 ¼ 0.21, P ¼ 0.0041) was observed paper,11 CLEC16A expression was measured in the same thymic (Table 1). DEXI is reported to be upregulated in an adenocarci- tissue samples. In this previous study, the grouping of the samples noma alveolar epithelial cell line upon dexamethasone treat- was different; individuals homozygous for the risk allele were ment.25 This strong immunosuppressive drug also regulates SOCS1 compared with non-risk homozygotes carriers grouped together expression in macrophages,33,34 suggesting a common regulation with heterozygotes. When we compared CIITA, DEXI, CLEC16A of SOCS1 and DEXI. (both isoforms) and SOCS1 gene expression in these groups, no In conclusion, our data support the existence of a regulatory significant correlation with genotype was found. However, there element in the CLEC16A region of chromosome 16p13,17 which was a trend toward higher DEXI and SOCS1 expression in the also influences the expression of DEX1 and SOCS1. These findings non-risk carrier group, which did not reach statistical significance imply that DEX1 and SOCS1 may represent possible autoimmune (data not shown). candidate genes. The fact that several SNPs in the CLEC16A gene A recent study reported that DEXI expression is reduced in region have shown association to other autoimmune diseases monocytes carrying the CLEC16A rs12708716 risk allele.19 In our in addition to MS1–10 indicate common mechanisms between thymic tissue samples, we also observed reduced DEXI expression autoimmune diseases and the importance of further investigation in CLEC16A risk carriers, but for the CLEC16A rs6498169 genotype. of this genetic region. As rs12708716 and rs7206912 are in strong D0-LD with rs6498169 (Figure 1b), a correlation to rs12708716 and rs7206912 could also be expected in our samples. Indeed, DEXI expression was lower in MATERIALS AND METHODS samples carrying the rs12708716 and rs7206912 risk alleles, Subjects and genotyping although not significantly reduced. Interestingly, chromosome 19 Thymic tissue samples from Norwegian children under the age of conformation-capture experiments performed by Davison et al. 13 undergoing corrective cardiac surgery (n ¼ 37) were collected in RNA proposed that a loop can be formed between the DEXI promoter later (Qiagen, Hilden, Germany) and peripheral whole-blood samples from region and a 20-kb region of intron 19 of CLEC16A (containing healthy Norwegian individuals from the bone marrow donor registry rs12708716), suggesting a potential mechanism for how intronic (n ¼ 24) were collected on Tempus blood RNA tubes (Applied Biosystems, sequences within CLEC16A might affect expression of neighboring Foster City, CA, USA).11,35 The study was approved by the local ethical genes. Our findings together with the previously reported DEXI authorities, and informed consent was given either by the parents of the expression quantitative trait locus data for disease-associated children or the patients/donors. SNP genotyping of the whole-blood and thymic tissue samples was performed in a previous study.11 One of CLEC16A SNPs and the finding of the close physical proximity of 19,20 the thymic tissue samples was not successfully genotyped for the CLEC16A intron 19 and the DEXI promoter indicate that rs6498169 SNP. CLEC16A intron 19 harbors a regulatory region for DEXI in immune cells and highlights DEXI as an autoimmune candidate gene. Although the function of DEXI is unknown, it has been shown Gene expression analysis that absence of Socs1 leads to aberrant thymocyte development RNA from whole blood was isolated using the Tempus spin RNA kit and systemic inflammation in mice.28 More specifically, Socs1 (Applied Biosystems) and RNA from thymic tissues was isolated with TRIzol 11,35 has shown to influence the Foxp3 þ CD4 þ T-regulatory cell reagent (Invitrogen, Carlsbad, CA, USA) as described. RNA differentiation in the thymus.29 Interestingly, loss of Socs1 is also concentration, quality and integrity were verified by Nanodrop DN-1000 associated with increased differentiation of Th17 cells, a pro- Spectrophotometer (Thermo Fisher Scientific Inc., Madison, WI, USA) and þ Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). inflammatory CD4 T-cell subset found in higher amounts in the Reverse transcriptase (RT) PCR was performed with 50 ng RNA using the peripheral blood of patients with active MS compared with Reverse Transcriptase Core kit (RT-RTCK) with random nonamer primers 30,31 healthy controls. The observed reduced SOCS1 expression in and the EuroScript reverse transcriptase (all from Eurogentec, Liege thymic samples harboring the CLEC16A risk alleles might affect Science Park, Seraing, Belgium). A standard curve was generated for each

Table 1. Linear regression analysis of pair-wise comparison of CIITA, DEXI, CLEC16A and SOCS1 expression in thymic tissue samples

CIITA DEXI CLEC16A

DEXI R2 ¼ 0.08775, P ¼ 0.0750 CLEC16A R2 ¼ 0.08910, P ¼ 0.0719 R2 ¼ 0.2696, Po0.0010* SOCS1 R2 ¼ 0.01693, P ¼ 0.8089 R2 ¼ 0.2119, P ¼ 0.0041* R2 ¼ 0.6328, Po0.0001* R2 represents the correlation coefficient and P is uncorrected P-values; *, significant correlations.

Genes and Immunity (2013) 62 – 66 & 2013 Macmillan Publishers Limited Multiple sclerosis-associated single-nucleotide polymorphisms IS Leikfoss et al 65 RNA collection as a 1:2-fold dilution series (100–1.5625 ng) using a mixture 6 Rubio JP, Stankovich J, Field J, Tubridy N, Marriott M, Chapman C et al. Replication of RNA for the specific collection. RT was performed using the GeneAmp of KIAA0350, IL2RA, RPL5 and CD58 as multiple sclerosis susceptibility genes in PCR system 9700 thermo cycler (Applied Biosystems) for a one-step PCR Australians. Genes Immun 2008; 9: 624–630. reaction (25 1C for 10 min, 48 1C for 30 min and 95 1C or 5 min). Quantitative 7 Sawcer S, Hellenthal G, Pirinen M, Spencer CC, Patsopoulos NA, Moutsianas L et al. real-time PCR analyses were performed using the real-time PCR standard Genetic risk and a primary role for cell-mediated immune mechanisms in multiple curve method and the 7900HT Fast Real-time PCR System (Applied sclerosis. Nature 2011; 476: 214–219. Biosystems) with the following PCR conditions: 50 1C for 2 min, 95 1C for 8 Ferreira RC, Pan-Hammarstrom Q, Graham RR, Gateva V, Fontan G, Lee AT et al. 10 min followed by 40 cycles of 95 1C for 15 s and 60 1C for 1 min. The Association of IFIH1 and other autoimmunity risk alleles with selective IgA samples were run in triplicates on a MicroAmp Optical 384 well reaction deficiency. Nat Genet 2010; 42: 777–780. plate (Applied Biosystems) and analyzed by sequence detection systems 9 Skinningsrud B, Lie BA, Husebye ES, Kvien TK, Forre O, Flato B et al. A CLEC16A (SDS) v. 2.3 (Applied Biosystems). Primers and probes against TBP variant confers risk for juvenile idiopathic arthritis and anti-cyclic citrullinated (4326322E), total CLEC16A (HS00389799_m1), DEXI (HS00360234_m1), peptide antibody negative rheumatoid arthritis. Ann Rheum Dis 2010; 69: total CIITA (HS00172094_m1) or SOCS1 (HS00705164_s1) (all from 1471–1474. Applied Biosystems) were added to each reaction, including 6.6-ng 10 Wellcome Trust Case Control Consortium. Genome-wide association study of complementary DNA template, 5 ml TaqMan Gene Expression Mastermix 14,000 cases of seven common diseases and 3000 shared controls. Nature 2007; (Applied Biosystems), 0.5 mlof20 Primer Probe mix (Applied Biosystems) 447: 661–678. and 4 ml RNase-free water (Qiagen). A negative control without 11 Mero IL, Ban M, Lorentzen AR, Smestad C, Celius EG, Saether H et al. Exploring the complementary DNA and a no RT control was included. Relative CLEC16A gene reveals a MS-associated variant with correlation to the expression of CLEC16A, DEXI, CIITA and SOCS1 was calculated as the ratio relative expression of CLEC16A isoforms in thymus. Genes Immun 2011; 12: between the target gene and the TBP reference gene, which was selected 191–198. as reference gene owing to low variance in CT between the different 12 Hafler DA, Compston A, Sawcer S, Lander ES, Daly MJ, De Jager PL et al. Risk alleles thymic samples (data not shown). PCR specificity was verified by a single for multiple sclerosis identified by a genomewide study. N Engl J Med 2007; 357: band after gel electrophoresis. 851–862. 13 Nischwitz S, Cepok S, Kroner A, Wolf C, Knop M, Muller-Sarnowski F et al. More Statistical analysis CLEC16A gene variants associated with multiple sclerosis. Acta Neurol Scand 2011; 123: 400–406. Correlation between gene expression levels and genotypes were assessed 14 Bronson PG, Caillier S, Ramsay PP, McCauley JL, Zuvich RL, De Jager PL et al. CIITA by two-sided Mann–Whitney U-test using GraphPad Prism 5 (GraphPad variation in the presence of HLA-DRB1*1501 increases risk for multiple sclerosis. Software, Inc., San Diego, CA, USA). For permutation analyses, a subset of Hum Mol Genet 2010; 19: 2331–2340. 24 data points (equal to the number of data points in the whole-blood 15 Vandenbroeck K, Alvarez J, Swaminathan B, Alloza I, Matesanz F, Urcelay E et al. samples) from the thymic samples showing significant correlation between A cytokine gene screen uncovers SOCS1 as genetic risk factor for multiple genotype and gene expression were randomly picked 1000 times, keeping sclerosis. Genes Immun 2012; 13: 21–28. the ratio in number between the groups equal to the ratio in the thymic 16 Zuvich RL, McCauley JL, Oksenberg JR, Sawcer SJ, De Jager PL, Aubin C et al. samples. A two-sided Mann–Whitney U-test was performed on those data Genetic variation in the IL7RA/IL7 pathway increases multiple sclerosis sets (R software, R core team). The median P-value of 1000 tests 0.05 was o susceptibility. Hum Genet 2010; 127: 525–535. considered significant. Linear regression analysis of gene expression 17 Zuvich RL, Bush WS, McCauley JL, Beecham AH, De Jager PL, Ivinson AJ et al. levels was performed using SPSS (IBM SPSS Statistics 19, Chicago, IL, USA). Interrogating the complex role of chromosome 16p13.13 in multiple sclerosis CLEC16A expression was correlated with the expression of the CIITA, DEXI susceptibility: independent genetic signals in the CIITA-CLEC16A-SOCS1 gene and SOCS1 predictor genes, and pair-wise correlations between CIITA, DEXI complex. Hum Mol Genet 2011; 20: 3517–3524. and SOCS1 were also analyzed. Correlation coefficients (R2) with P-values 18 Hoppenbrouwers IA, Aulchenko YS, Janssens AC, Ramagopalan SV, Broer L, Kayser below 0.05 were considered significant. M et al. Replication of CD58 and CLEC16A as genome-wide significant risk genes for multiple sclerosis. J Hum Genet 2009; 54: 676–680. 19 Davison LJ, Wallace C, Cooper JD, Cope NF, Wilson NK, Smyth DJ et al. Long-range CONFLICT OF INTEREST DNA looping and gene expression analyses identify DEXI as an autoimmune Inger-Lise Mero has received an unrestricted grant for running expenses in this disease candidate gene. Hum Mol Genet 2012; 21: 322–333. project from Biogen Idec Norway AS. 20 Nica AC, Montgomery SB, Dimas AS, Stranger BE, Beazley C, Barroso I et al. Candidate causal regulatory effects by integration of expression QTLs with complex trait genetic associations. PLoS Genet 2010; 6: e1000895. ACKNOWLEDGEMENTS 21 Dubois PC, Trynka G, Franke L, Hunt KA, Romanos J, Curtotti A et al. Multiple common variants for celiac disease influencing immune gene expression. Nat This study was funded by grants from The South-Eastern Norway Regional Health Genet 2010; 42: 295–302. Authority, Ullevål University Hospital Scientific Advisory Council (VIRUUS), Biogen Idec 22 Swanberg M, Lidman O, Padyukov L, Eriksson P, Akesson E, Jagodic M et al. Norway AS, The Norwegian Diabetes Association and the Odd Fellow society. MHC2TA is associated with differential MHC molecule expression and suscept- We would like to thank Harald Lindeberg and the Norwegian Bone Marrow Registry ibility to rheumatoid arthritis, multiple sclerosis and myocardial infarction. for providing samples; Marte Viken, Siri Flåm, Hege D Sollid and Hanne Sæther for Nat Genet 2005; 37: 486–494. collecting and preparing the RNA samples; and Christian Page for help with statistical 23 Steimle V, Otten LA, Zufferey M, Mach B. Complementation cloning of an MHC analyses. class II transactivator mutated in hereditary MHC class II deficiency (or bare lymphocyte syndrome). Cell 1993; 75: 135–146. 24 Fenner JE, Starr R, Cornish AL, Zhang JG, Metcalf D, Schreiber RD et al. Suppressor REFERENCES of cytokine signaling 1 regulates the immune response to infection by a unique 1 Hakonarson H, Grant SF, Bradfield JP, Marchand L, Kim CE, Glessner JT et al. inhibition of type I interferon activity. Nat Immunol 2006; 7: 33–39. A genome-wide association study identifies KIAA0350 as a gene. 25 Edgar AJ, Birks EJ, Yacoub MH, Polak JM. Cloning of dexamethasone-induced Nature 2007; 448: 591–594. transcript: a novel glucocorticoid-induced gene that is upregulated in 2 The International Multiple Sclerosis Genetics Consortium. The expanding genetic emphysema. Am J Respir Cell Mol Biol 2001; 25: 119–124. overlap between multiple sclerosis and type I diabetes. Genes Immun 2009; 10: 26 Cookson W, Liang L, Abecasis G, Moffatt M, Lathrop M. Mapping complex 11–14. disease traits with global gene expression. Nat Rev Genet 2009; 10: 3 Todd JA, Walker NM, Cooper JD, Smyth DJ, Downes K, Plagnol V et al. Robust 184–194. associations of four new chromosome regions from genome-wide analyses of 27 Nicolae DL, Gamazon E, Zhang W, Duan S, Dolan ME, Cox NJ. Trait-associated type 1 diabetes. Nat Genet 2007; 39: 857–864. SNPs are more likely to be eQTLs: annotation to enhance discovery from GWAS. 4 Skinningsrud B, Husebye ES, Pearce SH, McDonald DO, Brandal K, Wolff AB et al. PLoS Genet 2010; 6: e1000888. Polymorphisms in CLEC16A and CIITA at 16p13 are associated with primary 28 Catlett IM, Hedrick SM. Suppressor of cytokine signaling 1 is required for the adrenal insufficiency. J Clin Endocrinol Metab 2008; 93: 3310–3317. differentiation of CD4 þ T cells. Nat Immunol 2005; 6: 715–721. 5 Marquez A, Varade J, Robledo G, Martinez A, Mendoza JL, Taxonera C et al. 29 Zhan Y, Davey GM, Graham KL, Kiu H, Dudek NL, Kay TW et al. SOCS1 negatively Specific association of a CLEC16A/KIAA0350 polymorphism with NOD2/CARD15(-) regulates the production of Foxp3 þ CD4 þ T cells in the thymus. Immunol Cell Crohn’s disease patients. Eur J Hum Genet 2009; 17: 1304–1308. Biol 2009; 87: 473–480.

& 2013 Macmillan Publishers Limited Genes and Immunity (2013) 62 – 66 Multiple sclerosis-associated single-nucleotide polymorphisms IS Leikfoss et al 66 30 Durelli L, Conti L, Clerico M, Boselli D, Contessa G, Ripellino P et al. T-helper 17 33 Bhattacharyya S, Zhao Y, Kay TW, Muglia LJ. Glucocorticoids target suppressor of cells expand in multiple sclerosis and are inhibited by interferon-beta. Ann Neurol cytokine signaling 1 (SOCS1) and type 1 interferons to regulate Toll-like receptor- 2009; 65: 499–509. induced STAT1 activation. Proc Natl Acad Sci USA 2011; 108: 9554–9559. 31 Takahashi R, Nishimoto S, Muto G, Sekiya T, Tamiya T, Kimura A et al. SOCS1 is 34 Rhen T, Cidlowski JA. Antiinflammatory action of glucocorticoids--new mechan- essential for regulatory T cell functions by preventing loss of Foxp3 expression as isms for old drugs. N Engl J Med 2005; 353: 1711–1723. well as IFN-{gamma} and IL-17A production. J Exp Med 2011; 208: 2055–2067. 35 Viken MK, Olsson M, Flam ST, Forre O, Kvien TK, Thorsby E et al. The PTPN22 32 Fu J, Wolfs MG, Deelen P, Westra HJ, Fehrmann RS, Te Meerman GJ et al. Unra- promoter polymorphism 1123G4C association cannot be distinguished from veling the regulatory mechanisms underlying tissue-dependent genetic variation the 1858C4T association in a Norwegian rheumatoid arthritis material. Tissue of gene expression. PLoS Genet 2012; 8: e1002431. Antigens 2007; 70: 190–197.

Genes and Immunity (2013) 62 – 66 & 2013 Macmillan Publishers Limited