Allele specific chromatin signals, 3D interactions, and motif predictions for immune and B cell related diseases

Marco Cavalli,Jan Grau,Rapolas Spalinskas,Jan Komorowski,Husen M Umer,Ola Wallerman,Claes Wadelius,Pelin Sahlén,Gang Pan,Ioana Lemnian,Nicholas Baltzer,Ivo Grosse

doi:10.1038/s41598-019-39633-0

Abstract

Several Genome Wide Association Studies (GWAS) have reported variants associated to immune diseases. However, the identified variants are rarely the drivers of the associations and the molecular mechanisms behind the genetic contributions remain poorly understood. ChIP-seq data for TFs and histone modifications provide snapshots of protein-DNA interactions allowing the identification of heterozygous SNPs showing significant allele specific signals (AS-SNPs). AS-SNPs can change a TF binding site resulting in altered gene regulation and are primary candidates to explain associations observed in GWAS and expression studies. We identified 17,293 unique AS-SNPs across 7 lymphoblastoid cell lines. In this set of cell lines we interrogated 85% of common genetic variants in the population for potential regulatory effect and we identified 237 AS-SNPs associated to immune GWAS traits and 714 to gene expression in B cells. To elucidate possible regulatory mechanisms we integrated long-range 3D interactions data to identify putative target genes and motif predictions to identify TFs whose binding may be affected by AS-SNPs yielding a collection of 173 AS-SNPs associated to gene expression and 60 to B cell related traits. We present a systems strategy to find functional gene regulatory variants, the TFs that bind differentially between alleles and novel strategies to detect the regulated genes.

Highlights

More than 15% of the variants reported today in the Genome Wide Association Studies (GWAS) catalog are associated to immune system diseases
The ChIP-seq reads from histone modifications defining promoters (H3K4me3), enhancers (H3K4me[1], H3K27ac) and domain boundaries proteins (CTCF, SA.1)[30] were aligned to the respective genomes to identify heterozygous sites with allele specific signals and the results were corrected for genome-wide testing and pruned for false positives by filtering out AS-SNPs in CNVs and different types of repeated regions
The cell specific collections of AS-SNPs obtained in this way were intersected with GWAS and eQTL SNPs identified in B cells in order to link the allele-specific behavior observed in ChIP-seq to B cell specific traits or gene expression

Summary

Introduction

More than 15% of the variants reported today in the Genome Wide Association Studies (GWAS) catalog are associated to immune system diseases. Several functions mediated by B cells, such as secretion of autoantibodies, inflammatory cytokines, presentation of autoantigens, modulation of antigen processing etc., are today consistently reported as central in the onset of several autoimmune diseases[2]. Linking genomic variation to diseases or phenotype is a complex process that involves three major steps: (i) identify the causal gene regulatory variant(s), (ii) identify the TF(s) that bind to the variants, (iii) identify the target gene(s) whose deregulation lead to the phenotype. This opens the field for functional studies of the biological mechanisms of disease

Methods

Results

Conclusion