Identification of a two-loci epistatic interaction associated with susceptibility to rheumatoid arthritis through reverse engineering and multifactor dimensionality reduction

Altered synovial fibroblast (SF) transcriptional activity is a key factor in the disease progression of rheumatoid arthritis (RA). To determine the transcriptional regulatory network associated with SF response to an RA proinflammatory stimulus we applied a CARRIE reverse engineering approach to microarray gene expression data from SFs treated with RA synovial fluid. The association of the inferred gene network with RA susceptibility was further analyzed by a case-control study of promoter single-nucleotide polymorphisms, and the presence of epistatic interactions was determined using the multifactor dimensionality reduction methodology. Our findings suggest that a specific NF-κB transcriptional regulatory network of 13 genes is associated with SF response to RA proinflammatory stimulus and identify a significant epistatic association of two of its genes, IL6 and IL4I1, with RA susceptibility. © 2007 Elsevier Inc. All rights reserved.