Abstract

Objectives Cholangiocarcinoma (CCA) is a heterogeneous malignancy with high mortality and dismal prognosis, and an urgent clinical need for new therapies. Knowledge of the CCA epigenome is largely limited to aberrant DNA methylation. Dysregulation of enhancer activities has been identified to affect carcinogenesis and leveraged for new therapies but is uninvestigated in CCA. Our aim is to identify potential therapeutic targets in different subtypes of CCA through enhancer profiling. Design Integrative multiomics enhancer activity profiling of diverse CCA was performed. A panel of diverse CCA cell lines, patient-derived and cell line-derived xenografts were used to study identified enriched pathways and vulnerabilities. NanoString, multiplex immunohistochemistry staining and single-cell spatial transcriptomics were used to explore the immunogenicity of diverse CCA. Results We identified three distinct groups, associated with different etiologies and unique pathways. Drug inhibitors of identified pathways reduced tumour growth in in vitro and in vivo models. The first group (ESTRO), with mostly fluke-positive CCAs, displayed activation in estrogen signalling and were sensitive to MTOR inhibitors. Another group (OXPHO), with mostly BAP1 and IDH -mutant CCAs, displayed activated oxidative phosphorylation pathways, and were sensitive to oxidative phosphorylation inhibitors. Immune-related pathways were activated in the final group (IMMUN), made up of an immunogenic CCA subtype and CCA with aristolochic acid (AA) mutational signatures. Intratumour differences in AA mutation load were correlated to intratumour variation of different immune cell populations. Conclusion Our study elucidates the mechanisms underlying enhancer dysregulation and deepens understanding of different tumourigenesis processes in distinct CCA subtypes, with potential significant therapeutics and clinical benefits. Data are available in a public, open access repository. Sequencing datasets can be found at the European Genome-phenome Archive (EGA, Accession number: EGAS00001007309). Data used in this manuscript also includes previously published studies from public repositories: EGA (EGA00001000950) and Gene Expression Omnibus (GEO, GSE89749 and GSE89803). Codes used are available on GitHub ([https://github.com/chernycherny/CCA\_enhancers\_2023][1]). [1]: https://github.com/chernycherny/CCA_enhancers_2023