Background/Purpose: GWAS have identified multiple genetic regions that confer risk for juvenile idiopathic arthritis (JIA).However, identifying the single nucleotide polymorphisms (SNPs) that drive risk has been impeded by the fact that the SNPs that identify risk loci are in linkage disequilibrium (LD) with hundreds of other SNPs. Since the causal SNPs remain unknown, it is difficult to identify target genes or use genetic information to inform treatment. GWAS are also unable to identify the cells whose function is impacted by the disease-driving variants. We used genotyping and functional data in primary human CD4+ T cells to nominate disease-driving SNPs and identify their likely target genes.

Methods: We identified JIA risk haplotypes using Immunochip data from Hinks et al (Nature Get 2013) and McIntosh et al (Arthritis Rheum 2017), setting r² at 0.80.We used existing genotyping data from 3,939 children with JIA and 14,412 healthy controls to identify SNPs that were present within open chromatin in multiple immune cell types and more common in children with JIA than the controls (p< 0.05) .We next sought to identify those SNPs most likely to occur within regulatory regions in both resting and activated human primary CD4+ T cells by overlaying the chosen SNPs (n=846) with bi-directional transcription initiation characteristic of non-coding regulatory regions detected using dREG to analyze precision run-on sequencing (PROseq) data. Finally, we used MicroC to detect gene promoters interacting with the regulatory regions harboring the candidate causal SNPs.

Results: We identified 138 SNPs situated within dREG peaks in resting CD4+ T cells, 119 in PMA-stimulated CD4+ T cells (2 hr stimulation), and 150 that occurred in dREG peaks in CD3-CD28-IL2-stimulated CD4+ T cells (7 days’ stimulation). We identified n=74 SNPs that appeared in dREG peaks under all 3 conditions, suggesting that their biological effects are exerted on both resting and activated cells, while those unique to activation phases (20 in resting cells, 20 in PMA-activated cells, and 39 in CD3-CD28-IL2-activated cells) may exert their effects only under specific biological conditions. MicroC studies identified n=41 genes that interacted with enhancers harboring the candidate causal SNPs in resting CD4+ T cells, n=39 in CD4+T cells activated with ionomycin and PMA (30 min) and n=38 in CD4+ T cells activated with CD3/CD28/IL2 (7 days).There was considerable overlap in these gene sets, as shown in the figure, while other gene-enhancer interactions were cell-state-specific.

Conclusion: We demonstrate that SNPs on JIA risk haplotypes that are seen more frequently in patients than controls can be further filtered by their presence within functional regulatory elements (detected by PROseq) in primary human CD4+ T cells. Some of these regulatory elements are unique to the activation state of the cells, supporting the idea that some SNPs may exert their biological effects only in specific biological contexts. Finally, we demonstrate that newer methods for querying 3D chromatin structures (e.g., MicroC) can crisply identify the genes influenced by regulatory elements harboring these SNPs, and, thus, the immunological processes impacted by disease-driving genetic variants.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/using-genotyping-and-chromatin-data-in-cd4-t-cells-to-nominate-causal-variants-on-jia-risk-haplotypes-and-to-identify-their-target-genes/