Objective. To identify new genes associated with susceptibility to rheumatoid arthritis (RA), using a 2-stage genome-wide association study.Methods. Following a liability-based study design, we analyzed 317,503 single-nucleotide polymorphisms (SNPs) in 400 patients with RA and 400 control subjects. We selected a group of candidate SNPs for replication in an independent group of 410 patients with RA and 394 control subjects. Using data from the 3 previous genome-wide association studies in RA, we also looked for genomic regions showing evidence of common association signals. Finally, we analyzed the presence of genome-wide epistasis using the binary test implemented in the PLINK program.Results. We identified several genomic regions showing evidence of genome-wide association (P < 1 ؋ ؊5). In the replication analysis, we identified KLF12 SNP rs1324913 as the most strongly associated SNP (P ‫؍‬ 0.01). In our study, we observed that this SNP showed higher significance than PTPN22 SNP rs2476601, in both the genome-wide association studies and the replication analyses. Furthermore, the integration of our data with those from previous genome-wide association studies showed that KLF12 and PTPRT are the unique loci that are commonly associated in 3 different studies (P ‫؍‬ 0.004 and P ‫؍‬ 0.002 for KLF12 in the Wellcome Trust Case Control Consortium study and the Brigham and Women's Rheumatoid Arthritis Sequential Study genome-wide association study, respectively). The genome-wide epistasis analysis identified several SNP pairs close to significance after multiple test correction.Conclusion. The present genome-wide association study identified KLF12 as a new susceptibility gene for RA. The joint analysis of our results and those from previous genome-wide association studies showed genomic regions with a higher probability of being genuine susceptibility loci for RA.Rheumatoid arthritis (RA) is one of the most prevalent autoimmune diseases in the world (1). In RA, chronic inflammation of the synovial joints leads to progressive articular damage, which can result in major functional disability (2). The etiology of RA is unknown, but several family aggregation and twin studies (3,4) clearly demonstrate a heritable component of the disease. Part of this genetic component of susceptibility has been consistently associated with the HLA class II locus variation. The remaining 50-75% of the genetic component includes several other genomic regions that are more difficult to identify due to their lower penetrance or more complex models of action (5,6).

No abstract

No abstract

Objective Paraoxonase 1 (PON‐1) is a high‐density lipoprotein (HDL)–associated antioxidant enzyme that plays an important role in HDL‐mediated cardioprotection. Although genetic polymorphisms are known to modulate PON-1 activity, its involvement in cardiovascular disease (CVD) in rheumatoid arthritis (RA) is controversial, suggesting that other factors may modulate its function. Since anti‐HDL antibodies have been found to be related to an impaired lipid profile and occurrence of CVD in RA, this study was undertaken to examine the associations between PON-1 activity, anti‐HDL antibodies, and CVD according to PON1 genetic variants in patients with RA. Methods Serum PON-1 activity, using paraoxon as substrate, and IgG anti‐HDL antibodies were quantified in 212 RA patients and 110 healthy controls. The PON1 rs662 genotype (Q>R) was determined with TaqMan probes. An additional group of 13 biologics‐naive patients with RA was prospectively followed up for 3 months. Results PON‐1 activity was decreased in RA patients compared to healthy controls (P = 0.005), and an effect of the rs662 genotype was noted in both groups, with Q/Q homozygotes exhibiting the lowest PON‐1 activity. The distribution of rs662 genotypes did not differ between RA patients and healthy controls (P = 0.215). In patients carrying the Q/Q genotype, anti‐HDL antibodies were associated with impaired PON‐1 activity (P = 0.010), and levels of anti‐HDL antibodies were associated with decreased HDL levels (r = −0.680, P < 0.001) and higher prevalence of cardiovascular events, as determined in univariate and multivariate models. Furthermore, change in anti‐HDL antibody levels upon tumor necrosis factor blockade was an independent predictor of improved PON‐1 activity (β = −0.369, 95% confidence interval −0.669, −0.069; P = 0.024). Conclusion PON‐1 activity is impaired in RA in association with the rs662 genotype and anti‐HDL antibodies, the latter being recognized as a pivotal player in the link between rs662 and CVD in patients with RA.

IntroductionAn increased expression of interferon (IFN)-responding genes (IRGs), the so-called IFN signature, has been reported in rheumatoid arthritis (RA). However, some controversy exists concerning its clinical relevance. The main aim of this study is to evaluate whether quantitative and qualitative differences in the activation of the IFN pathway may account for these findings.MethodsThe expression of IFN-induced protein 44 (IFI44), IFN-induced protein 44 like (IFI44L), IFN alpha inducible protein 6, and MX dynamin-like GTPase 1 (MX1) was determined in peripheral blood in 98 RA patients (IFI6) and 28 controls. RA patients were classified into groups according to their clinical stage and treatments received: very early RA (VERA), biological disease-modifying antirheumatic drug (bDMARD) naive, and bDMARD. An additional group of 13 RA patients candidates for tumor necrosis factor alpha (TNFα) blockade was also recruited. The associations among IRGs were evaluated by network and principal component analyses.ResultsThe expression of all IRGs was increased in RA to different levels. The IFN score was increased in all RA groups (VERA, bDMARD-naïve, and bDMARD), but important differences in their degree of activation and in the relationships among IRGs were observed. The IFN score correlated with the accumulated disease activity score 28-joints, and it was found to be a predictor of clinical outcome in VERA. No differences in the IFN score were observed between the bDMARD-naive and bDMARD groups, but opposite associations with the clinical parameters were noted. Interestingly, the correlations among IRGs delineate different pictures between these two groups. The IFN score at baseline predicted poor clinical outcome upon TNFα blockade. Although no absolute changes in the IFN score were found, TNFα-blockade shifted the associations among IRGs.ConclusionA certain heterogeneity within the IFN signature can be recognized in RA, depending on the clinical stage. The structure of the IFN signature may be a potential explanation for the controversy in this field and must be considered to decipher its clinical relevancein RA.