Background Radiographic joint damage is highly variable among rheumatoid arthritis patients. Since the heritability of joint destruction rate was estimated to 45-58% in previous twin and the Icelandic RA-population studies, it is highly suggested that genetic variants are involved in the process of joint destruction and cause the different amount of joint damage among the patients.
Objectives We aimed to examine genetic variants and biological pathways associated with bone erosion using high-density genotype data for immune loci in Koreans.
Methods We analyzed ∼100,000 immune-loci SNPs for their associations with bone erosion in 1,142 RA patients. The subjects were genotyped by Immunochip, a high-density genotyping array for immune disease loci (n=343), or imputed for the Immunochip SNPs from our previous genomoe-wide association study (GWAS) array by SHAPEIT and IMPUTE2 (n=799). The patients were classified into two groups based on the Steinbrocker staging system: non-erosive (stages I and II) and erosive (stages III and IV) RA. The association between SNPs and bone erosion was tested for the genotyping and imputation-based data separately, by multivariate logistic regression with adjustment of age, sex, disease duration, and smoking status at symptom onset using PLINK. The association results from both datasets were combined in a fixed-effect meta-analysis by using GWAMA. Enrichment of associated variants in the known biological pathways and gene-sets was further tested from the meta-analysis results by MAGENTA.
Results By testing the associations between the 99,177 SNPs and bone erosion in 446 erosive and 387 non-erosive RA patients, rs10782763 (OR=1.64, P=2.95×10-5) and rs12095896 (OR =1.60, P=5.28×10-5) near LPHN2 and rs10263303 (OR=1.81, P=8.41×10-5) near PTPRN2 (or DNAJB6) were potential SNPs associated with bone erosion in RA, but none reached the genome-wide significance threshold. In the gene-set enrichment analysis, five candidate causal pathways were identified: the ataxia telangiectasia-mutated gene (ATM), vascular endothelial growth factor (VEGF), C-C chemokine receptor type 5 (CCR5), Src by protein-tyrosine phosphatase alpha (srcRPTP), and free radical induced apoptosis (FREE) pathways (P<0.05, BioCarta database).
Conclusions We could not find any immune-loci SNPs surpassing genome-wide significance level in our study subjects, although we identified suggestive associations at several SNPs, raising the possibilities that genetic variants for joint erosion would locate beyond the immune loci, unlike RA susceptibility genes and there is highly heterogenous genetic etiology in joint erosion. In addition, we identified five biological pathways associated with bone erosion, which might provide insight into the molecular mechanisms underlying the radiographic damage in RA.
Disclosure of Interest None declared