Background/Purpose: CXCR4 is a recently identified susceptibility locus for juvenile idiopathic arthritis (JIA). The index single nucleotide polymorphism (SNP) used to identify this locus, rs95338, lies in an intergenic region upstream of the CXCR4 gene, suggesting that this region has regulatory functions. We sought to gain biological insight into the association between JIA and the CXCR4 locus by examining the chromatin architecture within this locus, including the presence of epigenetically-marked regulatory elements, in relevant primary cells and cell lines.

Methods: Using the SNP Annotation and Proxy Search from the Broad Institute and standard computational methods, we first defined the linkage disequilibrium (LD) block containing the rs953387 SNP. We used BedTools to query ENCODE and Roadmap Epigenomics data to identify relevant functional epigenetic marks (e.g., H3K4me1/H3K27ac, and CTCF binding motifs) within the CXCR4 haplotype. For statistical analysis, each value was compared to the mean and standard deviation of 100 randomly generated files with the same number of peaks as the ENCODE file being examined. In addition, we used quantitative rtPCR analysis to confirm the presence of a non-coding, intergenic RNA transcript that we identified on RNAseq in neutrophils and assess the effects of treatment on expression of this transcript.

Results: The LD block for the rs953387 SNP comprises a genomic region of between 54,236 (r-squared of 0.9) and 69,609 bp (r-squared of 0.8) upstream from the coding region of the CXCR4 gene. This region contains 154 JIA associated SNPs that we recently identified using whole genome sequencing, in addition to >30 variants catalogued in 1000 Genome Project. Further analysis using ENCODE and Roadmap Epigenomics data demonstrate that the LD block (r²= 0.8) includes enrichment for H3K4me1 (poised enhancer) peaks compared to randomly generated peaks from hg19 genome in primary CD4+ T-cell lines and primary CD 15+ (neutrophil) cell lines, with the LD containing 12 and 15 peaks respectively. There are 6 H3K27ac (active enhancer) peaks in the LD block from primary CD4+ cell data, which is statistically higher than the random value. CHIPSeq data from neutrophils from our own lab shows multiple H3K4me1 and H3K27ac peaks in this LD block in addition to a noncoding (intergenic) RNA. The non-coding RNA was verified using rtPCR and showed changes in expression levels when neutrophils from children with active JIA were compared to those in clinical remission on medication. Further analysis identified H3K4me1, but not H3K27ac, enrichment in CD14+ monocytes and CD20+ B cells as well as other T- lymphocyte cell lines.

Conclusion: The LD blocking containing the rs953387 SNP lies upstream of the CXCR4 gene itself and is likely a regulatory region. This region contains enhancer elements in multiple immune cell lines, especially in CD4+ primary T-cells and CD15+ neutrophils as well as an intergenic non-coding RNA. The regulatory function of this region may or may not relate to the CXCR4 gene itself, as enhancers do not always regulate the nearest gene. These findings corroborate our earlier studies demonstrating that the genetics of JIA impinges almost exclusively on gene regulatory functions.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-chromatin-landscape-around-the-juvenile-arthritis-associated-cxcr4-locus-suggests-regulatory-functions-and-genetic-roles-in-both-innate-and-adaptive-immunity/