|Home | About | Journals | Submit | Contact Us | Français|
An association between major histocompatibility complex (MHC) genes, particularly those within the class II HLA region, and rheumatoid arthritis (RA) is well established, and accounts for an estimated 30% of the genetic component in RA. The MHC class II transactivator gene (MHC2TA) on chromosome 16p13 has recently emerged as the most important transcription factor regulating genes required for class II MHC-restricted antigen presentation. Previous studies of a promoter region polymorphism (–168A/G, rs3087456) in the MHC2TA gene and RA have yielded conflicting results.
To assess the association of the MHC2TA –168A/G polymorphism (rs3087456) and risk for RA by meta-analysis.
Meta-analysis was performed for 6861 patients with RA and 9270 controls from 10 case–control studies. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for each study. Summary ORs and 95% CIs were calculated for random effects models.
No effect was observed for the G risk allele (OR 1.02, 95% CI 0.93 to 1.12, p=0.70) or the GG risk genotype (OR 1.14, 95% CI 0.95 to 1.36, p=0.16).
Our results indicate that the MHC2TA –168A/G polymorphism (rs3087456) is not associated with RA yet underscore the importance of including shared epitope allele carrier status, secondary phenotypes and more complete characterisation of MHC2TA variation in future studies.
Rheumatoid arthritis (RA) is the most common systemic autoimmune disease with a worldwide prevalence approaching 1%.1 This chronic inflammatory disease can cause substantial disability from erosive and deforming processes in joints, and is associated with increased mortality.2 The prevalence of RA, in conjunction with the impact on the health status of affected individuals, results in tremendous associated costs to patients, their families and society. While the aetiology of RA is unknown, a significant genetic contribution to RA development is established and accounts for an estimated 60% of disease risk.3 A strong association between major histocompatibility complex (MHC) genes, particular those in the class II HLA region, and RA has been well established, and accounts for an estimated 30–50% of the genetic component in RA, at least in northern European Caucasians.3 4 Thus, non-MHC genes also contribute to disease risk.
The MHC class II transactivator gene (MHC2TA or CIITA) on chromosome 16p13 is an attractive candidate for genetic studies of RA because it regulates expression of proteins presenting and processing antigens, including MHC class II molecules.5 MHC2TA encodes the CIITA protein, which chaperones assembly of several transcription factors at MHC promoters.6 Swanberg et al 7 initially demonstrated that the G allele of the MHC2TA –168A/G promoter polymorphism (rs3087456) increased risk for RA, and that the GG genotype influenced expression of the CIITA protein and MHC class II molecules. Since then studies of MHC2TA and RA have focused on this single nucleotide polymorphism (SNP) but have yielded conflicting results; see tables 1 and and22 for a brief summary. Though previous studies describe this SNP as an A→G change, the dbSNP database indicates that the G allele is ancestral; thus this SNP is actually a G→A change.8 A meta-analysis was performed to test whether the G risk allele (by comparing AA against GG+GA carriers) or the GG risk genotype (by comparing AA+AG against GG carriers) increased risk for RA.
A literature search of the PubMed database was conducted using the search terms “rheumatoid arthritis” and “MHC2TA.” Nine publications were identified before 30 May 2007.7 9-15 All languages, geographic areas and publication types were included. Seven publications were original research articles and two were letters to the editor.12 14 One letter presented results for three studies.14 Study selection criteria were: (1) patients with RA fulfilled 1987 American College of Rheumatology classification criteria for RA (excluding juvenile RA);16 (2) controls were derived from a population in the same geographic area and with the same ethnic background as patients with RA; (3) authors provided original genotype frequencies; and (4) RA patient and control groups did not overlap between studies. Two authors (PB, LB) applied these criteria independently. Ten studies met our selection criteria. The most recent study was not included because it did not provide genotype frequencies.17
One author (PB) extracted the following data from each study: authors, publication year, journal, publication type, place of study, study design, genotyping method, genotype frequencies, number of cases and controls and sample description (recruitment method, age, race and ethnicity). Cases were described by RA classification criteria, age at onset and RA secondary phenotypes. Controls were described as population-based or matched, and by matching characteristics. All 10 studies had a case–control study design. For the two studies presenting genotype frequencies for a set of matched controls and a larger set of population-based controls, population-based controls were used because genotype frequencies were similar between the two sets of matched and population controls, and a larger sample size provided more power.7 15
For each individual study, the odds ratios (OR) and 95% confidence intervals (95% CI) for the G risk allele and the GG risk genotype in patients with RA against controls were calculated. Statistical analyses were performed in R (http://www.r-project.org/, accessed October 2006). ORs were calculated using unconditional maximum likelihood estimation. 95% CIs were calculated using normal approximation. Two-tailed p-values were calculated using the χ2 test of independence. The summary ORs and 95% CIs for the G risk allele and the GG risk genotype in patients with RA against controls were calculated for random effects models because this model assumes effects are randomly distributed.18 Between-study heterogeneity was assessed with the Cochran’s Q test statistic. Forest plots were generated in R to display the OR and 95% CI for each study (shown by a black square on a horizontal line, respectively) and the summary OR and 95% CI (shown by a dotted vertical line extending from a black diamond, respectively).
Genotype frequencies and study results for the 10 studies used in the meta-analysis are summarised in tables 1 and and2.2. Studies were weighted by size to contribute to the overall combined result. There was evidence of between-study heterogeneity for the G risk allele (p=0.04) but not for the GG risk genotype (p=0.10). There was no evidence of association for the MHC2TA –168A/G polymorphism (rs3087456) with RA risk when comparing GG+AG against AA carriers (OR=1.02, 95% CI=0.93 to 1.12, p=0.70) or when comparing GG against AG+AA carriers (OR=1.14, 95% CI=0.95 to 1.36, p=0.16). The forest plots in fig 1 show a schematic representation of the data that are presented in the bottom line of table 2. Results did not differ when the meta-analysis was limited to studies of European populations (data not shown).
MHC class II gene expression appears to be regulated almost exclusively by MHC2TA, and differential expression of class II genes has shown some evidence for association with both RA susceptibility and progressive disease in a recent study with direct implications for MHC2TA.19 There is substantial evidence for genetic contribution of the MHC to RA. The MHC spans ~4.5 mega base pairs on chromosome 6p21.3 and encodes >180 expressed genes; 40% are related to immune activation and response.20 21 The MHC class II gene HLA-DRB1 demonstrates the strongest association with RA, highlighting antigen presentation and subsequent T cell activation as a potential pathway in RA pathogenesis.4 All RA-associated HLA-DRB1 alleles encode a shared epitope (SE) not present on non-RA associated alleles.22
Whole-genome microsatellite linkage screens in families with multiple patients with RA have indicated linkage on chromosome 16p, with a stronger effect in families where patients carry two copies of SE alleles.23 Although this linkage peak may be due to another as of yet unidentified candidate gene on chromosome 16p, thus far only two studies of MHC2TA variation and RA have accounted for SE allele carrier status in the data analysis. Because these data were missing from the majority of studies included here, we were not able to incorporate SE carrier status into our analysis. Eyre et al10 did not find any effect of the MHC2TA –168A/G polymorphism (rs3087456) on RA susceptibility regardless of SE carrier status. Martinez et al13 did not find any effect of the MHC2TA –168A/G polymorphism (rs3087456) on RA susceptibility but reported two MHC2TA haplotypes (comprised of the –168A/G polymorphism (rs3087456) and rs4774) were associated with RA, with a stronger effect revealed in analyses of SE-positive patients.
Additionally, common secondary phenotypes (rheumatoid factor (RF) positivity, anti-cyclic citrullinated peptide positivity and erosive joint disease) associated with severe disease outcomes have evidence of specific genetic associations, which may reflect differences in underlying disease mechanisms based partly on the influence of different genetic factors.24 Thus far only two studies of MHC2TA variation and RA have examined secondary phenotypes; therefore, we were not able to incorporate secondary phenotypes into our analysis. Harrison et al11 reported slight trends when comparing the G risk allele in RF-negative patients with controls (OR 0.8, 95% CI 0.6 to 1.0, p=0.08) and 559 RF-positive patients to 159 RF-negative patients (OR 1.3, 95% CI 1.0 to 1.7, p=0.08). Yazdani-Biuki et al15 reported no evidence for association between the MHC2TA –168A/G polymorphism (rs3087456) and RF positivity or anti-cyclic citrullinated peptide positivity.
Although our results indicate that the MHC2TA –168A/G polymorphism (rs3087456) does not increase risk for RA, additional functional variants may exist in MHC2TA. Three studies have investigated other MHC2TA polymorphisms.10 13 Eyre et al10 identified five frequency validated SNPs from public databases mapping to the 5′ region of the gene (rs7501204, rs6498114, rs6416647, rs7404672, rs6498116) and performed single-locus and haplotype analyses in 813 patients with RA and 532 controls with negative findings. Swanberg et al7 evaluated haplotypes comprised of the MHC2TA –168A/G polymorphism (rs3087456) and two additional SNPs (rs4774 and rs34654419 (rs2229320plus27 bp) in exon 11) in 1288 RA cases and 709 controls; single-locus association results were not significant for these SNPs, and association with the MHC2TA –168A/G polymorphism (rs3087456) was not improved by haplotype analysis. These results are in contrast with evidence mentioned above for association with the MHC2TA –168A/G (rs3087456)/rs4774 haplotypes in 350 patients with RA and 509 controls.13 Our analysis of haplotype block structure utilising publicly available data suggest that at least 21 haplotype tagging SNPs are needed to fully capture common MHC2TA variation.
In this study, a meta-analysis of 6861 patients and 9270 controls was performed to assess association between the MHC2TA –168A/G promoter region polymorphism (rs3087456) and RA risk. Given that the large sample size of our study provided sufficient power to detect modest effects, our negative results indicate that the MHC2TA –168A/G polymorphism (rs3087456) is not causally associated with RA. Nevertheless, undiscovered functional variants may exist in the MHC2TA. In conclusion, a complex role for MHC2TA in RA, in conjunction with SE carrier status, remains plausible. Future studies that incorporate SE carrier status, secondary phenotypes and a more comprehensive screening of MHC2TA variation with haplotype tagging SNPs will help elucidate what role, if any, MHC2TA plays in RA.
This work was supported by the NIH/NIAID grant R01 AI065841.
Competing interests: None.