Genetic susceptibility to Barrett's esophagus: From GWAS to biology

Barrett’s esophagus (BE) is the only known precursor for esophageal adenocarcinoma (EAC), a highly lethal cancer with rising incidence and median survival <1 year. Substantial health-care resources are devoted to BE screening, surveillance, and treatment. Gastroesophageal reflux-induced injury of the lower esophagus and chronic inflammation are key drivers of BE development, but molecular pathways underlying risk are not well defined. Recent genome-wide association studies (GWAS) led by members of our team identified >20 novel genetic susceptibility loci for BE/EAC, providing new insights into the inherited genetic component of risk. Nevertheless, little progress has been made in bridging associations to biology. Consistent with GWAS of other complex diseases, all BE index variants map to non-coding regions, lack obvious biologic function, and are in linkage disequlibrium with many other SNPs, any of which may be causal. The vast majority of functional variants underlying GWAS signals are believed to map to and alter activity of regulatory elements including enhancers, in an allele-specific manner, and in turn modulate expression of downstream genes involved in risk. Importantly, such regulatory effects may be tissue- and condition-specific. To begin prioritizing candidate functional variants for experimental interrogation, we developed a customized informatics scoring pipeline using comprehensive in-silico annotations from multiple public resources. We selected four high-scoring BE risk loci for evaluation using luciferase reporter assays in esophageal cell lines, and found that two of four regions exhibited allele-specific enhancer activity. CRISPR-mediated deletion of the enhancer region at both loci correlated with downregulation of several candidate risk genes. Motivated by these successes, we seek to expand our integrative framework for elucidating functional consequences of BE-related genetic variation. We hypothesize that such variation is biologically expressed through alterations in transcriptional regulation and downstream gene expression. Our goal is to identify functional variants, risk enhancers, and target genes underlying BE risk, leveraging unique resources and complementary statistical/experimental approaches. In Aim 1, we will define candidate causal variants via Bayesian fine-mapping, using the largest BE GWAS world- wide, and further prioritize leading candidates via functional-potential scores. In Aim 2, we will perform new transcriptome profiling of reflux-exposed gastroesophageal junction tissues and constituent cells, and identify candidate BE risk genes and pathways via eQTL colocalization and network-based analysis. In Aim 3, we will validate candidate functional variants using luciferase reporter enhancer activity assays; identify target genes of risk enhancers via CRISPR-mediated enhancer deletion and RNA-Seq; and interrogate pathways influenced by prioritized target genes in Aims 2 & 3 via CRISPR-mediated gene knockdown/overexpression and RNA- Seq. This study will advance noncoding GWAS signals into functional biological signatures and support future efforts to develop novel preventive/interventional strategies for BE/EAC.