Elucidating the Enhancer Code in 9p21 Locus Underlying Age-Related Disease Risks
MetadataShow full item record
The chromosomal region 9p21.3 is one of the first identified and most replicated risk loci for coronary artery disease (CAD) by genome-wide association studies (GWAS). This ~150 kb long non-coding region was later found to be associated with multiple disease phenotypes, most of which are age-related, such as cancers, glaucoma, and type II diabetes (T2D). Despite the strong evidence of disease associations from population studies, the mechanisms underlying the 9p21.3 locus remained largely elusive, similar to many other non-coding GWAS loci. Recent epigenetic annotation studies suggested an enrichment of enhancers in 9p21.3 and, at the genome-wide level, a general enrichment of GWAS variants around enhancer regions. I thus hypothesized that functional disease variants in the 9p21.3 locus cause alterations to enhancer functions and transcriptional regulatory networks, thus contributing to the risk of age-related diseases. In Chapter 2, I performed a comprehensive functional study of enhancers in the 9p21.3 locus. Using a human embryonic stem cell (hESC) differentiation panel, I confirmed CDKN2B expression was reduced by CAD causal variants specifically in vascular cells. A large fragment deletion study using CRISPR identified a causal enhancer region inside the CAD GWAS hotspot, together with other cell type specific enhancer elements. Together with the results from Chapter 3, I showed that these enhancers formed a synergistic interaction network that was significantly dampened by the risk genotype. In order to characterize the enhancer interaction network in 9p21.3, I designed a novel chromatin conformation capture method with substantially improved resolution. As described in Chapter 3, this new method, named Tri-4C, was able to map distal contacts to a hundred basepair resolution, leading to a much improved accuracy in the capture of enhancer loop interactions than existing methods. I demonstrated that Tri-4C was capable of revealing quantitative enhancer loop alterations under dynamic gene controls, as well as differentiating allele-specific loop interactions in heterozygote loci. In Chapter 4, I investigated the function of variants associated the diseases other than CAD. Using an in-silico prediction approach, I identified rs6475604, a glaucoma-associated enhancer variant showing reduced binding of the transcription factor YY1. Dampened YY1 activity in this locus may cause upregulation of its known repressive target, p16, leading to cell cycle arrest and premature senescence. Finally, in Chapter 5, to study whether 9p21.3 associates with extreme longevity, I performed a sequencing study of the locus in an Ashkenazi longevity cohort. I discovered an interesting mild depletion of combined age-related disease variants of this locus in centenarians. This finding supports the idea that reaching longevity is associated with evasion from risks of multiple, if not all, common age-related diseases. Taken together, these works established how causal variants in 9p21.3 were involved in the regulatory process of their target genes. The study resolved a key aspect of enhancer mechanisms underlying the disease risk associations of the 9p21.3 locus, which has remained largely a mystery in the past decade since its discovery. The effort to elucidate the enhancer regulatory networks in this locus used state-of-the-art model systems and techniques, which provide insights to the general question of how to address challenges in the functional interpretation of non-coding GWAS associations.