Network Analysis of Inflammatory Bowel Disease Reveals PTPN2 As New Monogenic Cause of Intestinal Inflammation

Abstract

54BACKGROUND & AIMS: Genome-wide association studies (GWAS) have uncovered multiple 55 loci associated with inflammatory bowel disease (IBD), yet delineating functional consequences 56 is complex. We used a network-based approach to uncover traits common to monogenic and 57 polygenic forms of IBD in order to reconstruct disease relevant pathways and prioritize causal 58 genes. 59 METHODS:We have used an iterative random walk with restart to explore network 60 neighborhood around the core monogenic IBD cluster and disease-module cohesion to identify 61 functionally relevant GWAS genes. Whole exome sequencing was used to screen a cohort of 62 monogenic IBD for germline mutations in top GWAS genes. One mutation was identified and 63 validated by a combination of biochemical approaches. 64 RESULTS: Monogenic IBD genes clustered siginificantly on the molecular networks and had 65 central roles in network topology. Iterative random walk from these genes allowed to rank the 66 GWAS genes, among which 14 had high disease-module cohesion and were selected as putative 67 causal genes. As a proof of concept, a germline loss of function mutation was identified in PTPN2, 68 one of the top candidates, as a novel genetic etiology of early-onset intestinal autoimmunity. The 69 mutation abolished the catalytic activity of the enzyme, resulting in haploinsufficiency and hyper-70 activation of the JAK/STAT pathway in lymphocytes. 71 CONCLUSIONS: Our network-based approach bridges the gap between large-scale network 72 medicine prediction and single-gene defects and underscores the crucial need of fine tuning the 73 JAK/STAT pathway to preserve intestinal immune homeostasis. Our data provide genetic-based 74 rationale for using drugs targeting the JAK/STAT pathway in IBD.75

Low-usage splice junctions underpin immune-mediated disease risk

“…A rare alternative splicing event likely underpins inflammatory bowel disease risk at the PTPN2 locus To demonstrate how sQTLs for lowly-used introns can dysregulate alternative splicing and predispose to IMDs, we further investigated a PTPN2 sQTL that implicates a lowly-used intron that colocalised with an inflammatory bowel disease (IBD) associated risk locus at 18p11.21. Multiple lines of evidence, including coding variants associated with monogenic IBD 38,39 and mouse knock-out models 40,41 , have suggested PTPN2 is the effector gene at 18p11.21, though this remains to be established. Moreover, It is not yet known if and how common IBD-associated SNPs affect the expression of PTPN2.…”