Identifying Causal Variants in Juvenile Arthritis Using a Massively Parallel Reporter Assay

Abstract This project aims to take an important step toward a goal that the field of pediatric rheumatology has been working towards for more than 30 years: identifying the actual causal variants that confer genetic risk in juvenile idiopathic arthritis (JIA). As is typical for complex traits, we have found that most of the genetic risk for JIA resides within non-coding regions of the genome. These regions are enriched for H3K4me1/H3K27ac histone marks, epigenetic signatures of enhancers. Thus, genetic risk for JIA involves changes in the regulation of genes, rather than amino acid substitutions that alter the function(s) of proteins. Rather than viewing JIA as primarily an autoimmune disease, we hypothesize that JIA emerges because leukocytes suffer genetically and epigenetically-mediated perturbations that blunt their capacity to regulate and coordinate transcription across the genome. This loss of coordinate regulation leads to inappropriate expression of inflammatory mediators in the absence of the normal external signals typically required to initiate or sustain an inflammatory response. A significant impediment to our testing that hypothesis is that the actual causal variants for JIA are unknown. There are more than 38 known risk loci for JIA, and these loci contain more than 18,000 variants. Our recent completion of whole genome sequencing on 48 patients with JIA has further added to the number of candidate loci and variants. Given these challenges, and the fact that the causal variants lie within regulatory (rather than protein-coding) regions, the field is in urgent need of a high throughput method to query tens of thousands of variants, determine their effects on transcription, and identify the cells in which these variants exert their effects. In this pilot study, we will introduce a massively parallel reporter assay (MPRA), developed in the laboratory of our collaborator, Dr. Pardis Sabeti, as a way of identifying potential causal variants through their capacity to alter gene expression in myeloid and lymphoid cell lines. This assay was specifically designed to query genetic variants within non-coding regions of the genome. We will test variants that demonstrate strong linkage disequilibrium with index SNPs within all known JIA associated risk loci as identified from previous studies (n= 7,312), examining their effects on gene expression. WE will then use luciferase assays to corroborate the results from the MPRA. At the end of this 2-year pilot study, we expect to have significantly narrowed the list of potential causal variants in JIA and will be prepared to undertake the studies required to elucidate a causal link between specific alleles and alterations in specific transcriptional networks and cellular functions. !