Identification of susceptibility loci for Takayasu arteritis through a large multi-ancestral genome-wide association study

Abstract

Takayasu arteritis is a rare inflammatory disease of large arteries. We performed a genetic study in Takayasu arteritis comprising 6,670 individuals (1,226 affected individuals) from five different populations. We discovered HLA risk factors and four non-HLA susceptibility loci in VPS8, SVEP1, CFL2, and chr13q21 and reinforced IL12B, PTK2B, and chr21q22 as robust susceptibility loci shared across ancestries. Functional analysis proposed plausible underlying disease mechanisms and pinpointed ETS2 as a potential causal gene for chr21q22 association. We also identified >60 candidate loci with suggestive association (p < 5 × 10^-5) and devised a genetic risk score for Takayasu arteritis. Takayasu arteritis was compared to hundreds of other traits, revealing the closest genetic relatedness to inflammatory bowel disease. Epigenetic patterns within risk loci suggest roles for monocytes and B cells in Takayasu arteritis. This work enhances understanding of the genetic basis and pathophysiology of Takayasu arteritis and provides clues for potential new therapeutic targets.

Keywords: GWAS; HLA; Takayasu arteritis; chromatin interaction; eQTL; epigenetic; genetic association; genetic risk scroe; vasculitis.

Figure 1

Manhattan plot showing the results of the meta-analysis among the five populations included in this study The −log₁₀ p value for the genetic variants analyzed are plotted against their physical chromosomal position. The red and blue lines represent the genome-wide level of significance (p value < 5 × 10⁻⁸) and the suggestive level of significance (p value < 5 × 10⁻⁵), respectively.

Figure 2

Manhattan plots showing the results of stepwise conditional analysis of the HLA region (A) Unconditioned. (B) Conditioning on rs2844678. (C) Conditioning on rs2844678 + rs12524487. (D) Conditioning on rs2844678 + rs12524487 + rs17193507. (E) Conditioning on rs2844678 + rs12524487 + rs17193507 + rs12526858. The −log₁₀ p value for the genetic variants analyzed are plotted against their physical chromosomal position. The red line represents the genome-wide level of significance (p value < 5 × 10⁻⁸).

Figure 3

Column bar plot representing the results from histone marks enrichment analysis among genetic susceptibility loci identified with a GWAS level of significance (p value < 5 × 10⁻⁸) in Takayasu arteritis in this study The y axis corresponds with the −log₁₀ corrected p value of H3K27ac, H3K4me1, and H3K4me3 enrichments in different blood cell types. These data were derived via ENCODE and Roadmap Epigenomics Consortium projects. PBMCs, peripheral blood mononuclear cells; HSCs, hematopoietic stem cells.

Figure 4

Functional annotation results point to ETS2 as a causal gene for the chr21q22 association (A) Summary of the epigenetic marks found at rs2836882 in chr21q22. (B) Circular view of the chromatic interaction between rs2836882 in chr21q22 and the ETS2 promoter obtained from Capture Hi-C data. This interaction is observed in several cell types: endothelial precursors, megakaryocytes, monocytes, M0 macrophages, M1 macrophages, M2 macrophages, neutrophils, and CD34+ cells. The red color of the line indicates a strong confidence for this interaction. (C) Violin plot representing the difference in expression levels of ETS2 depending on rs2836882 genotypes obtained from GTEx project.

Figure 5

Takayasu arteritis genetic susceptibility loci identified and the main Gene Ontology terms represented (A) Schematic illustration depicting the susceptibility loci identified in this study. Loci surpassing the genome-wide level of significance (p value < 5 × 10⁻⁸) are highlighted in red. (B) Gene Ontology (GO) enrichment analysis. Biological Process GO clusters revealed with genes annotated to genetic susceptibility loci identified in this study (p value < 5 × 10⁻⁵). Only clusters detected with a p value < 1 × 10⁻³ and that included at least three genes are shown. Each node corresponds with a GO term, and edges represent connections between GO terms. Each GO cluster is defined by a different color and the most-associated Biological Process within each cluster is highlighted.

Figure 6

Dot plot representing the cumulative genetic risk score (GRS) for Takayasu arteritis across the five major populations included in the 1000 Genomes Project Each dot illustrates the GRS value of an individual, and the black lines represent the means and ranges. Data were used from the 1000 Genomes Project phase 3 release. African, n = 661; Admixed American, n = 347; East Asian, n = 504; European, n = 503; South Asian, n = 489. There was a significant difference between the GRSs among populations (ANOVA p value < 0.0001). The multiple comparison test on mean GRS between each pair of populations revealed statistically significant differences for all pairs (p value < 0.0001) except African versus East Asian and European versus South Asian (Tukey’s adjusted p value = 0.924 and 0.457, respectively).