Takayasu arteritis is a rare inflammatory disease affecting large arteries, most frequently the aorta and its major branches. Clinical manifestations result from vascular complications such as stenoses, occlusions, and aneurysms and may also include non-specific symptoms such as fatigue, fever, and weight loss. Disease onset typically occurs in the second or third decade of life and the disease is more common in women. Takayasu arteritis has been reported worldwide, but disease prevalence varies across populations and is the highest in East Asia. Although the etiology of Takayasu arteritis is incompletely understood, there is strong evidence for a role of genetic factors in the disease pathophysiology.
The first suggestion of a genetic component to Takayasu arteritis arose in the 1960’s with the publication of a case report of familial aggregation of the disease. Later in the 1970’s, the human leukocyte antigen (HLA) was the first genetic region identified as a susceptibility factor. The association with the HLA-B∗52 allele remains the most robust genetic signal in Takayasu arteritis. In the last decade, four large-scale genetic analyses have been conducted, but only nine non-HLA genetic loci have been revealed with a genome-wide level of significance. Awareness of the importance of including diverse and under-represented ancestry populations in genetic studies is growing and is positioned high on the priorities of genomic research.
The total study population comprised 1,226 individuals diagnosed with Takayasu arteritis from five populations (Turkish, Northern European descendant from North America and the United Kingdom, Han Chinese, South Asian, and Italian) and 5,444 ancestry-matched unaffected individuals. Samples were removed if they had a genotyping call rate < 95%. In addition, single nucleotide polymorphisms (SNPs) with a genotyping call rate < 98%, minor allele frequency (MAF) < 1%, and those deviated from Hardy-Weinberg equilibrium (HWE) in affected and control individuals (p value < 1 × 10−3) were filtered out.
Sex chromosomes were not analyzed. In addition, duplicated or multi-allelic SNPs as well as those A/T-C/G SNPs with MAF > 0.4 were removed to avoid errors in the imputation process. Conducted logistic regression by using the 10 first principal components as covariates for each population independently. The genomic inflation factor (λ) was calculated for each cohort, and quantile-quantile (Q-Q) plots were generated with and without the HLA region. Next, the results of the logistic regression were meta-analyzed by means of the inverse variance method.
Takayasu arteritis’ genetic relationship with other immune-mediated diseases - Projected Takayasu arteritis into a previously constructed lower-dimensional representation of immune-mediated disease (IMD) genetics, an “IMD basis,” by using the project_sparse function from the cupcake R package. To represent all the ancestral groups included in this study, we used the summary statistic of the meta-analysis for these analyses. Projection requires knowledge of the linkage disequilibrium (LD) between variants, and we used an LD matrix estimated from our GWAS samples to allow for the different ancestries included.
We report a cross-ancestry meta-analysis comprised of individuals from five different populations: Turkish, Northern European descendant, Han Chinese, South Asian, and Italian. From the original 6,670 individuals (1,226 affected individuals), a total of 6,221 individuals (1,091 affected individuals) and around 4 million variants were maintained after stringent QC filtering (a summary of sample/variant QC is shown in Table S3). First, we performed logistic regression analyses in each population independently (Figure S2) followed by a meta-analysis including all cohorts. Results of genomic control analysis showed no evidence of population stratification for any cohort (Figure S3 and Table S3). The results of the meta-analysis are illustrated in Figure 1. As expected, the most robust association signal was observed within the HLA region.
Functional annotations and epigenetic enrichment analysis of genome-wide associated SNPs
In line with most of the genetic susceptibility loci detected in IMDs, Takayasu arteritis-associated genetic variants reside in non-coding regions, suggesting that they might influence the disease by disrupting regulatory elements. To assign biological significance to those variants, we carried out a comprehensive functional annotation by using in silico approaches.
Takayasu arteritis genetic relationship with multiple IMDs
Relating different diseases through their GWAS summary statistics has been used to reveal etiological relationships between diseases. Compared Takayasu arteritis to other IMDs through projecting it into a previously constructed lower-dimensional representation of IMD genetics, an “IMD basis. This representation uses 13 components to summarize axes of risk shared between different combinations of diseases. We found Takayasu arteritis differed significantly (FDR < 1%) from controls on eight of the 13 components, five of which were also associated with Crohn disease and/or ulcerative colitis. In the original report of the IMD basis, relationships between different IMD GWASs and self-reported disease traits in European UK Biobank subjects were inferred from hierarchical clustering of their projections, and we reclustered the same set of traits here together with Takayasu arteritis to examine with which IMDs it shared genetic risk components. We found that Takayasu arteritis clustered with Crohn disease, ulcerative colitis, and ankylosing spondylitis.
Performed a large multi-ancestral GWAS in Takayasu arteritis. We revealed risk factors within the HLA region and four non-HLA genetic associations with a GWAS level of significance, extended the genetic association of previously reported risk loci to other populations, and uncovered over 60 additional candidate genetic susceptibility loci for the disease. Using functional and epigenetic enrichment characterization, we identified pathways, biological processes, and target genes affected by these genetic risk loci. Our results also suggest a close genetic relationship between Takayasu arteritis and inflammatory bowel diseases. Finally, the results of this work highlight immune cell types that are crucial in mediating genetic risk in Takayasu arteritis.