Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arthritis Rheum. Author manuscript; available in PMC 2009 July 1.
Published in final edited form as:
PMCID: PMC2570539

Lack of association of functional CTLA4 polymorphisms with juvenile idiopathic arthritis



Juvenile idiopathic arthritis (JIA) is an autoimmune disorder mediated by Th1-immune responses. Cytotoxic T-lymphocyte Antigen 4 (CTLA4), expressed on the T-cell surface, plays a negative role in regulating T-cell activation. Single nucleotide polymorphisms (SNPs) in CTLA4 have been implicated in susceptibility to several autoimmune disorders, including JIA. Our objective was to test three functional CTLA4 variants for association with JIA.


Families of 531 children with JIA were genotyped for SNPs located in the promoter region (C-318T), exon-1 (A49G), and the 3’ untranslated region (CT60) of CTLA4 by PCR amplification and digestion. Family-based association test (FBAT) was used to test CTLA4 SNPs and haplotypes for association with JIA. A second independent cohort of more than 300 children with JIA and 500 controls were genotyped for case-control analyses. Case-control analyses of the combined cohorts, as well as meta-analyses of published association studies between CTLA4 and JIA, were performed.


There were no deviations of transmission of any of the CTLA4 variants to children with JIA, or JIA subtypes, by FBAT. There were also no significant associations between CTLA4 C-318T or A49G SNPs in 650 JIA cases and 350 controls. Similarly, there were also no significant associations between CT60 variants with over 800 JIA cases and 500 controls. The meta-analysis also failed to confirm an association between JIA and CTLA4 variants.


These results suggest that C-318T, A49G or CT60 or haplotypes tagged by these CTLA4 SNPs are not associated with JIA or major JIA subtypes.

Keywords: CTLA4, JRA, genetics, autoimmune, association, juvenile idiopathic arthritis

Juvenile idiopathic arthritis (JIA) comprises a group of chronic childhood arthritides [1]. Although associations between JIA and variants in the major histocompatibility complex (MHC) have been described, non-MHC variants, such as PTPN22 also play a role in JIA susceptibility. The pathogenesis of JIA is mediated, in part, by T-cells [2]. The synovial tissue contains a high proportion of activated T-cells. Hence genes encoding proteins that regulate the interactions between T-cells and antigen-presenting cells are good candidates for genetic association studies of JIA.

CTLA4 (CD152) is a T-cell surface molecule that negatively regulates T-cell activation[3]. Three single nucleotide polymorphisms (SNPs) of CTLA4 (C-318T rs5742909, A49G rs231775, and CT60 rs3087243) are associated with many autoimmune disorders, including type-1 diabetes (T1D), rheumatoid arthritis (RA) and autoimmune thyroiditis. To date three association studies investigating CTLA4 variants and JIA have been published with mixed results [4-6]. Our objectives were to determine if these functional CTLA4 variants are associated with JIA susceptibility, and to replicate an association reported earlier between C-318T and juvenile rheumatoid arthritis (JRA) [4].

Patients and methods

Cohort 1 consisted of 531 children with JIA (143 persistent oligoarticular, 69 extended oligoarticular, 165 rheumatoid factor (RF)-negative polyarticular, 44 RF-positive polyarticular, 91 systemic, and 19 other JIA subtypes) and their families in part from a registry sponsored by the National Institute of Arthritis and Musculoskeletal and Skin Diseases and the Cincinnati Children’s Hospital Medical Center. The median onset age was 5.5 years, and 73% of the subjects were female. Cohort 2 comprised 365 children with JIA (134 persistent oligoarticular, 27 extended oligoarticular, 82 RF-negative polyarticular, 32 RF-positive polyarticular, 32 systemic, 27 enthesitis related arthritis and 31 other JIA subtypes) from the Intermountain States Database of Childhood Rheumatic Diseases. The median onset age was 5.6 years, and 66% of the subjects were female. Additionally, 551 adult controls screened for several common autoimmune diseases were ascertained from Salt Lake City, UT. All cases and controls from cohort 2 were used for CT60 genotyping, since this SNP shows the strongest association with autoimmunity [7]. Cases, family members and controls were predominantly (>90%) of Northern European ancestry. Subjects were enrolled under protocols approved by the Institutional Review Boards at the Cincinnati Children’s Hospital Medical Center and the University of Utah.


Subjects were genotyped for C-318T and A49G, as follows. Polymerase chain reaction (PCR) amplification of genomic DNA was performed using these primers: AGGGCTCAGAAAGTTAGCAGCCT and AATACAGAGCCAGCCAAGCCAGAT. Cycling conditions were: 94°C for 15 minutes, followed by 40 cycles: 94°C for 30 seconds, annealing at 62°C for 1 minute and extension at 72°C for 1 minute, followed by a final extension at 72°C for 10 minutes. After PCR amplification, the product was digested separately with Mse1 and Tse1. CT60 was genotyped as previously described, using the primers AGTGCTTGATTGCGTGG and TGCTGAGACTATACATTGGTTAAG, and HpyCH4IV for digestion [8].


The program Haploview was used to test for Hardy-Weinberg Equilibrium (HWE) and to infer haplotypes [9]. SNPs, haplotypes and genotypes were tested for association with JIA in cohort 1 by a transmission-disequilibrium test, implemented in the family-based association test (FBAT) program [10]. Briefly, FBAT calculates a transmission score for each informative family. FBAT determines the expected scores under the null hypothesis of no association or linkage by conditioning on offspring trait values and parental genotypes, assuming Mendelian transmission of alleles to offspring. The difference between the observed and expected score is summed over all families to generate a test statistic, which has a chi-square distribution. FBAT allows for incorporation of different genetic models. Both SNPs and haplotypes can be analyzed.

Case-control analyses were performed to investigate if alleles, genotypes or haplotypes of CTLA4 were associated with JIA. Haplotype logistic regression was also performed to test CTLA4 haplotypes for association with JIA. Published association studies of CTLA4 and JIA were identified using a PUBMED search[4-6], and allele frequency data was obtained from the individual studies. Odds ratios (OR) and 95% confidence intervals (CI) were calculated for each SNP. The pooled OR was calculated under a fixed effects model, weighting the studies by sample-size. Mantel-Haenszel Chi-square tests were performed for the individual studies and the studies combined. The Breslow-Day test was performed for heterogeneity among the studies, with significance set at p < 0.05. Analyses were performed using SAS 9.1.


Allele and haplotype frequencies of the three polymorphisms

The minor allele frequencies (MAF) of the C-318T, A49G and CT60 SNP among the unaffected parents were 0.09, 0.39 and 0.43 respectively. MAF of these three SNPs in controls were 0.1, 0.39 and 0.44 respectively. These frequencies are similar to the observed MAF of 0.09, 0.36 and 0.48 respectively in 652 “White” controls in the study by Ueda et al[7]. None of the SNPs deviated from HWE. Three-locus haplotype frequencies in parents were as follows: CAA = 0.41, CGG = 0.38, CAG = 0.11, TAG = 0.08 and CGA = 0.02. The three SNPs were in high linkage disequilibrium with each other (D’ 0.8 to 1.0). Controls had similar haplotype frequencies.

Tests of association between CTLA4 SNPs and JIA

No deviation of transmission of any of the CTLA4 SNPs to JIA probands was observed under an additive model using FBAT (Table 1). Analyses under other models, (dominant, recessive, or genotypic) also did not reveal any associations in JIA probands. Analyses performed after stratifying by JIA subtypes revealed an association between C-318T and RF-positive polyarticular JIA (p < 0.04). For the RF-positive poly JIA families, under a genotypic model, there was a borderline association with the A49G-G/G genotype (12 informative families; observed score (S) = 8, expected score (E(S)) = 5; p < 0.07) and for the CT60-G/G genotype (12 families, S = 8, E(S) = 4.8; p < 0.05). Other JIA subtypes revealed no associations. There were no transmission deviations of CTLA4 haplotypes to probands with JIA or any JIA subtype, including the haplotype (G-G) containing the susceptibility alleles at both A49G and CT60.

Table 1
Family based association tests under an additive genetic model for C-318T, A49G and CT60 variants in JIA families.

Case-control analysis using cohort 2 revealed no associations between CTLA4 variants and JIA. Case-control analyses were repeated after combining the JIA cases from both cohorts. No associations were found between C-318T or A49G variants and JIA or JIA subtypes, using 650 cases and 345 controls (Table 2). Similarly the CT60 variants showed no associations in 818 cases and 518 controls. Logistic regression analyses did not demonstrate an association between CTLA4 haplotypes and JIA.

Table 2
Results of case-control analysis of C-318T, A49G and CT60 variants and JIA

Meta-analyses of C-318T and A49G SNPs and JIA

Cases and controls were mostly of European descent, from Germany [4, 5], Northern Ireland [6] and the United States (present study). Allele frequency information was available on 1915 individuals (933 cases, 982 controls) for the analysis of C-318T (Table 3). The pooled OR was not significant (1.01, 95% CI 0.80-1.27). There was significant heterogeneity between the studies (p <0.01), mainly due to the study by Miterski et al, which found an association in the opposite direction compared to the remaining studies [4]. Excluding that study resulted in a pooled OR of 0.73 (0.55-0.97, p <0.03) for the presence of the T allele. Allele frequency information was available on 2457 individuals (1005 cases, 1452 controls) for the A49G analysis. There was no heterogeneity between studies (p >0.20). The pooled OR was not significant 0.96(0.84-1.09). The allele frequencies of the CT60 SNP were not published by Suppiah, et al. [6], precluding a meta-analysis of SNP CT60.

Table 3
Results of meta-analysis of CTLA4 C-318T and A49G polymorphisms and JIA


CTLA4, expressed on activated T-cells negatively regulates T-cell activation [3]. Functional CTLA4 variants could contribute to the pathogenesis of disorders characterized by abnormal T-cell responses, including JIA. CTLA4 variants are associated with T1D, autoimmune thyroiditis, and RA [7, 11-13]. Variants tested in our study have functional consequences. For instance, the T allele at C-318T is associated with up-regulation of CTLA4 transcription [14]. The A49G SNP, which changes threonine to alanine at position 17 in the leader sequence of CTLA4, affects the CTLA4-driven down-regulation of T-cell activation [15]. The strongest signal of association in the CTLA4 locus is with the CT60 SNP [7]. In a comprehensive study of CTLA4 variants, Ueda et al. demonstrated that the G allele at CT60 was associated with susceptibility to autoimmunity. The ratio of soluble CTLA4 to full length CTLA4 in CD4 T-cells was 50% lower in individuals with the susceptible genotype (G/G) compared with individuals who had the A/A genotype (p < 0.002). Thus, CTLA4 variants investigated in our study have functional consequences and known disease associations.

To date three studies have investigated associations between CTLA4 variants and juvenile arthritis[4-6]. These studies were primarily focused on adult RA or asthma patients and included relatively small numbers of children with JIA. Only limited phenotypic information about JIA/JRA was available in these studies. Miterski et al. reported an association of C318T with JRA in a study of 200 subjects from Germany [4]. An association with A49G was not observed. In another study, Suppiah et al. tested 72 children with JIA, 289 adults with RA, and 475 controls from Northern Ireland for CTLA4 variants[6]. In contrast to other studies on RA, Suppiah et al found an increased frequency of the A allele of the A49G SNP in the RA group and in the combined arthritis group, although an association was not found with the JIA subgroup alone. That study did not find any associations with the CT60 SNP, although they found differences in frequencies of A49G-CT60 haplotypes between arthritic cases and controls. Finally, Schubert et al. studied C-318T and A49G SNPs in a cohort of 321 children with asthma, 86 children with JIA, and 270 controls from Germany[5]. That study found no associations between JIA and either the C-318T or A49G variants.

We did not find an association between C–318T, A49G, or CT60 variants and JIA in a large cohort. We also did not replicate the association described between C-318T and JRA by Miterski et al. Lack of replication of genetic studies occurs due to myriad reasons, including type-1 statistical errors or hidden population stratification in the original study, inadequate power in replication studies, and population differences. The discrepancies in the reported results for C-318T in JIA could be due to true population differences or co-occurrence of other autoimmune disorders in the cohort studied by Miterski et al [4]. The mean age of disease onset (10.3 years) was higher in the cohort studied by Miterski et al. than in our study, raising the possibility that there were more individuals with RF-positive JIA in their cohort.

Inadequate power can cause failure to replicate, but our combined cohort is among the largest used for JIA genetic studies. Our study had 80% power to detect OR of 1.8, 1.5 and 1.4 for C-318T, A49G and CT60 respectively. Furthermore, our meta-analysis failed to find an association between either variant and JIA. The meta-analysis had 80% power to detect an OR of 1.53 and 1.27 for C-318T and A49G respectively, but an association of lower magnitude cannot be ruled out, especially for C-318T. We did observe an uncorrected P < 0.05 for C-318T when we combined our cohort and the cohort reported by Schubert et al. [5], but this result suggests that the T allele is protective, in contrast to the finding by Miterski et al that it increases susceptibility. Although this finding might reflect a statistical fluctuation, it would be interesting if this observation can be replicated in other JIA cohorts because the T allele at C-318T has been suggested to protect against autoimmune diseases by increasing CTLA4 transcriptional activity [14].

CTLA4 variants are associated with adult RA. However, a meta-analysis found that the association between the CT60-G allele and RA was relatively modest, with an OR of ~1.1 [13]. When RA patients were stratified, only patients positive for anti-citrullinated cyclic peptide (CCP) antibodies demonstrated a significant association with the CT60 variant, while CCP-negative patients did not show an association [13]. Only a small proportion of children with JIA have RF or anti-CCP antibodies, and this could explain the lack of association of CTLA4 variants with JIA. Although our RF-positive polyarticular JIA cohort was small, borderline associations were observed between this subtype and both the C-318T and A49G polymorphisms. Furthermore, FBAT revealed an association between the CT60-G/G genotype and RF-positive polyarticular JIA. These observations suggest that while RF-positive polyarticular JIA might share genetic associations with adult RA, the other JIA subtypes are likely distinct phenotypes and consequently have different genetic associations. It is conceivable that genetic factors associated with RF-negative RA are associated with JIA.


The authors wish to thank Matthew Honeggar for technical assistance.

Supported by, The National Institute of Arthritis and Musculoskeletal and Skin Diseases (AR50177, AR42218, AR47363, AR42272), The National Center for Research Resources (RR00064), The National Institute of Allergy, Immunology, and Infectious Disease (U54 AI065357), The Arthritis Foundation, and The Val A Browning Charitable Foundation, The Clinical Genetics Research Program, The Primary Children’s Medical Center Foundation, and the Children’s Health Research Center, Salt Lake City, UT.


1. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol. 2004;31:390–2. [PubMed]
2. Grom AA, Hirsch R. T-cell and T-cell receptor abnormalities in the immunopathogenesis of juvenile rheumatoid arthritis. Curr Opin Rheumatol. 2000;12:420–4. [PubMed]
3. Sharpe AH, Freeman GJ. The B7-CD28 superfamily. Nature Rev Immunol. 2002;2:116–26. [PubMed]
4. Miterski B, Drynda S, Boschow G, Klein W, Oppermann J, Kekow J, et al. Complex genetic predisposition in adult and juvenile rheumatoid arthritis. BMC genetics. 2004;5:2. [PMC free article] [PubMed]
5. Schubert K, von Bonnsdorf H, Burke M, Ahlert I, Braun S, Berner R, et al. A comprehensive candidate gene study on bronchial asthma and juvenile idiopathic arthritis. Dis Markers. 2006;22:127–32. [PMC free article] [PubMed]
6. Suppiah V, O’Doherty C, Heggarty S, Patterson CC, Rooney M, Vandenbroeck K. The CTLA4+49A/G and CT60 polymorphisms and chronic inflammatory arthropathies in Northern Ireland. Exp Mol Pathol. 2006;80:141–6. [PubMed]
7. Ueda H, Howson JM, Esposito L, Heward J, Snook H, Chamberlain G, et al. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature. 2003;423:506–11. [PubMed]
8. Suppiah V, Alloza I, Heggarty S, Goris A, Dubois B, Carton H, et al. The CTLA4+49 A/G*G-CT60*G haplotype is associated with susceptibility to multiple sclerosis in Flanders. J Neuroimmunol. 2005;164:148–53. [PubMed]
9. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2004 [PubMed]
10. Horvath S, Xu X, Laird NM. The family based association test method: strategies for studying general genotype--phenotype associations. Eur J Hum Genet. 2001;9:301–6. [PubMed]
11. Kavvoura FK, Ioannidis JP. CTLA-4 gene polymorphisms and susceptibility to type 1 diabetes mellitus: a HuGE Review and meta-analysis. Am J Epidemiol. 2005;162:3–16. [PubMed]
12. Marron MP, Raffel LJ, Garchon HJ, Jacob CO, Serrano-Rios M, Martinez Larrad MT, et al. Insulin-dependent diabetes mellitus (IDDM) is associated with CTLA4 polymorphisms in multiple ethnic groups. Hum Mol Genet. 1997;6:1275–82. [PubMed]
13. Plenge RM, Padyukov L, Remmers EF, Purcell S, Lee AT, Karlson EW, et al. Replication of putative candidate-gene associations with rheumatoid arthritis in >4,000 samples from North America and Sweden: association of susceptibility with PTPN22, CTLA4, and PADI4. Am J Hum Genet. 2005;77:1044–60. [PubMed]
14. Wang XB, Zhao X, Giscombe R, Lefvert AK. A CTLA-4 gene polymorphism at position -318 in the promoter region affects the expression of protein. Genes Immun. 2002;3:233–4. [PubMed]
15. Maurer M, Loserth S, Kolb-Maurer A, Ponath A, Wiese S, Kruse N, et al. A polymorphism in the human cytotoxic T-lymphocyte antigen 4 (CTLA4) gene (exon 1+49) alters T-cell activation. Immunogenetics. 2002;54:1–8. [PubMed]