Targeting the Cholangiocarcinoma Epigenome

Lay summary

Cholangiocarcinoma (CCA) is a lethal disease, with gene mutations that are now well known. These include key mutations that can be treated with specific drugs. However, treatments targeting these genetic defects have had limited success.

The sequence of DNA in each cell is the primary code that is used to make proteins. In normal cells, access to the DNA code is carefully controlled by the folding and shape of the DNA. DNA can be open to allow access of proteins that turn on genes in a cell. Or DNA can be closed to turn off genes. This epigenome pattern of open and closed DNA is carefully controlled in normal cells, involving master protein switches that set the patterns of open and closed DNA. It's now clear that cancer cells reset these switches in their favor, prompting the search for drugs that target the epigenome switches of a cancer cell and reset them to normal.

The theory is that many important cancer genes will be affected at once. Studies suggest that a cancer-promoting epigenome is a key feature of cholangiocarcinoma. In this study in the Developmental Research Program in the Mayo Clinic Hepatobiliary SPORE, we are focusing on the epigenome of cholangiocarcinoma. The goal of our study is to describe the epigenome and its switches in cholangiocarcinoma and test if they can be reset to normal.

Head and shoulders photograph of Michael T. Barrett, Ph.D.


The genomic landscape of cholangiocarcinoma is defined by somatic lesions in a diverse set of driver genes targeting a variety of signaling pathways. This intertumor heterogeneity presents a major challenge to the development of effective therapeutic strategies for this lethal cancer. Comprehensive cholangiocarcinoma genomic studies have identified a series of targetable lesions, including fibroblast growth factor receptor 2 (FGFR2) fusions and mutations and copy number aberrations, targeting RAS/MAPK and ERBB signaling. These provide improved classification of CCA subtypes, with the potential of delivering more personalized care for patients.

However recent clinical trials report only modest responses to targeted therapies even in selected patients carrying these mutations. As part of an ongoing study with members of the Gastrointestinal Cancer Program of the Mayo Clinic Comprehensive Cancer Center, we have been flow sorting and profiling CCA biopsies and patient-derived xenografts (PDXs) and functionally characterizing prevalent mutations.

Our analyses of these samples have highlighted the intertumor heterogeneity of CCA genomes. These include high-level amplification of a variety of oncogenes (for example, ERBB2, CCND1, MDM2 and MYC) and homozygous deletions targeting diverse tumor suppressor genes (for example, SAV1, VGLL4, ASCC3 and FAT1). Interestingly, superimposed on this heterogeneous genomic landscape are a large number of mutations and copy number aberrations converging on genes that regulate the epigenome and maintain chromatin organization and drive oncogenic gene expression. These include homozygous deletions targeting ARID family proteins and histone demethylases (for example, KDM6A), and pathogenic IDH1 mutations in biopsies and preclinical models. Notably, prior work has suggested that BAP1, PBRM1 and ARID1A deficiency all result in sensitivity to EZH2 inhibition across cancer types. Similarly, loss of KDM6A activity can lead to increased sensitivity to bromodomain and extra terminal (BET) inhibitors. In addition, missense mutations in isocitrate dehydrogenase 1 and 2 (IDH1/IDH2) genes, which occur in a significant number of CCAs, are associated with ARID1A hypermethylation and may also contribute to dysregulated epigenomes in CCAs.

The presence of IDH mutations and loss of function of ARIDs and related BAF subunits supports our primary hypothesis that targeting the aberrant CCA epigenomes can be exploited to improve responses to current and emerging therapies.

To test our hypothesis, we propose to characterize the genetic and epigenetic landscape of organoid cultures representing clinically relevant cholangiocarcinoma subtypes. Our preliminary results show that these models are stable in culture and retain the somatic lesions and phenotypic features of the primary tumors. These organoids are being used to explore the therapeutic potential of inhibition of master chromatin regulators, including EZH2 and BET proteins in CCAs.