Oral Presentation The 13th International Congress of the Immunology of Diabetes Society 2013

Network analysis identifies cytokine signaling pathways involved in genetic risk common to autoimmune diseases (#38)

Stephen S Rich 1 , Wei-Min Chen 1 , Patrick Concannon 2 , Aaron R Quinlan 1 , Suna Onengut-Gumuscu 1
  1. University of Virginia, Charlottesville, VA, United States
  2. University of Florida, Gainesville, Florida, United States

Commonality in the genetic basis of autoimmune diseases was recently reported (Cotsapas et al., 2011) based upon results of genome-wide association studies (GWAS), in which SNPs in ~50% of loci contributed to risk of multiple autoimmune diseases. Many of these diseases, including type 1 diabetes (T1D), have conducted dense mapping in these risk loci to identify candidate genes and causal variants for functional studies using the ImmunoChip, a custom high density genotyping array with ~1,000 SNPs in each of 186 loci implicated in autoimmunity.  We assembled significant results from ImmunoChip analyses of T1D, IBD, celiac disease, RA, MS and autoimmune thyroid disease and determined the set of genes that exhibited significant overlap as well as unique contribution with T1D. Of 39 loci associated with T1D, 17 were unique to T1D (e.g., GLIS3, RNLS, INS) while 22 were shared with other autoimmune diseases (e.g., PTPN22, BACH2, IL2RA, PTPN2). Protein-protein interaction network analysis was conducted to determine collections of genes/proteins that cluster more often than random in order to define common mechanisms. In the collection of SNPs most associated with T1D only, there were 12 direct interactions (P = 0.07 versus random interactions) that resulted in 5 small clusters of genes. One cluster was the known HLA-INS interaction with the remaining four novel clusters of genes. In the collection of SNPs associated with multiple autoimmune diseases, there were 42 proteins contributing to the primary network with 66 direct interactions between the proteins (P<9.9 x 10-4). There were six small clusters of genes with one large cluster of 25 interacting proteins. The members in the large cluster highlight the role of cytokine signaling. These analyses suggest that the majority of T1D risk consist of genes and proteins that affect the immune system, with smaller disease-specific elements focused on the target organ.