Introduction: Genome-wide association studies suggest that there is dysregulation of distinct biological circuits involved in the maintenance of intestinal homeostasis in patients with inflammatory bowel diseases (IBD). Broadly, we seek to understand the relationship between patient genotype and disease-specific, tissue-dependent gene expression to elucidate critical pathways at play in health and disease, identify novel, clinically-applicable biomarkers, guide therapeutic discovery, and move towards personalized care for patients with IBD. Methods: Using the most-recent GWAS data for IBD, identifying 163 risk loci, a custom NanoString probeset of 683 IBD-associated genes and 15 housekeeping genes was designed; 25% of these genes were not assayed by conventional microarrays. Patients and controls were recruited at five IBD centers nationwide as part of the Sinai-Helmsley Alliance for Research Excellence (SHARE) consortium. To date, we have collected 352 specimens from 172 unique patients (Crohn's disease, n=142, ulcerative colitis, n=28, indeterminate colitis, n=2). Total mRNA was isolated from 2-3 mucosal biopsies obtained at colonoscopy or taken from surgical specimens and used for downstream processing. A custom bioanalytic pipeline was designed for quality control and normalization to permit differential expression, data-driven clustering, and genotype-based expression quantitative trait loci (eQTL) analysis. Results: To serve as a basis for analysis in patients, we defined the expression patterns of these selected genes in healthy controls in multiple anatomic locations, including terminal ileum, ascending and descending colon, observing distinct expression patterns between ileum and colon and high correlation between controls. In the second phase of this project, we compared the expression patterns between diseased tissue versus healthy controls, inflamed versus uninflamed tissues from the same patient, and ulcerative colitis versus Crohn's colitis. Preliminary analysis demonstrates a strong differential gene expression pattern with 120 upregulated genes and 8 down-regulated genes in the inflamed tissue as compared to the uninflamed tissue. Pathway analysis suggests differences in activity between inflamed and uninflamed tissues in pathways related to cell adhesion, immune system signaling, and biosynthetic processing. With targeted examination for eQTLs for four disease-related SNPs (NOD2, ATG16L1, IRGM, MST1), we have identified genotype-dependent, fold-change gene expression differences between paired inflamed and uninflamed specimens. Conclusions: Selected profiling of IBD-related genes in disease-relevant tissues by NanoString technology and in patients with known genotypes can robustly detect novel patterns of gene epistasis, identify eQTLs, and highlight dominant pathways perturbed in disease.
Read full abstract