Non-coding mutations at enhancer clusters contribute to pancreatic ductal adenocarcinoma

Non-coding mutations (NCMs) that perturb the function of cis-regulatory elements (CRE, enhancers) contribute to cancer. Due to the vast search space, mutation abundance and indirect activity of non-coding sequences, it is challenging to identify which somatic NCMs are contributing to tumour development and progression. Here, we focus our investigation on the somatic NCMs that are enriched at enhancers from 659 pancreatic ductal adenocarcinoma (PDAC) tumours. We identify cis-regulatory NCMs within PDAC-specific enhancers derived from high and low-grade PDAC cell lines and patient derived organoids using two independent computational approaches. Five such CREs enriched for PDAC associated NCMs are also frequently mutated in other common solid tumours. Functional validation using STARR-seq reporter assays enables the prioritisation of 43 NCMs (7.3%) from a pool of 587 NCMs with 6,082 oligos, that significantly alter reporter enhancer activity compared to wild-type sequences. CRISPRi perturbation of an enhancer cluster harbouring NCMs over long non-coding RNA gene MIR100HG, which hosts a microRNA cluster (mir100-let7a-2-125b-1), leads to the downregulation of MIR100HG accompanied by a significant reduction in the TGF-{beta} pathway (known to induce MIR100HG) and other PDAC critical pathways, including KRAS, p53, MTOR and TNF signalling. Collectively, we have reported here cis-regulatory NCMs in PDAC proximal to many cancer-relevant genes, and our integrated approach paves way to explore CRE-associated NCMs in other human cancer genomes.