Dissecting the heterogeneity of tumor-initiating cells in hepatobiliary cancers: molecular characterization and targeting of stemness features

Primary liver cancers (PLCs) are among the most rapidly evolving malignant tumors worldwide. An underlying chronic inflammatory liver disease precedes liver cancer development for several decades and creates a pro-oncogenic microenvironment that renders the liver susceptible for malignant transformation. Depending on the primary cellular targets, a diverse range of different morphological and molecular phenotypes is observed in PLCs that frequently impairs therapeutic progress. In this context, the cancer stem cell (CSC) hypothesis provides a plausible explanation for the resulting tumor heterogeneity and implies a rationale for therapeutic strategies beyond the currently used anti-proliferative agents. Thus, eradication of tumors that follow the hierarchic model would require novel therapeutic strategies that specifically target the CSCs. Therefore, a detailed understanding of key molecular features that drive CSC characteristics is imperative to improve the dismal outcome of PLC patients. In the here presented project, we will focus on different key aspects of CSC-biology in PLCs, including intra-tumoral CSC-heterogeneity, prospective isolation of CSCs as well as therapeutic targeting of stemness features. First, we will develop and evaluate a novel robust and reproducible in vitro model of PLC that closely resembles individual phenotypic and molecular characteristics of individual patients. We will then prospectively isolate purified CSCs from human PLCs and apply next-generation whole transcriptome sequencing to define common molecular features and prognostic implications of PLC-CSCs. Subsequently, we will evaluate the therapeutic potential of targeting the identified signaling pathways for CSC-directed therapeutic approaches by using small molecules as well as RNAi. After successful evaluation of the most promising candidates for molecular targeting we will test the applicability of the obtained results for other solid tumors as well as liver metastasis from different primary cancers. Finally, we will directly test the importance of the tumor microenvironment for the heterogeneity of CSCs in PLC by comparing the molecular profiles of CSCs isolated from different intra-tumoral, peri-tumoral and non-tumorous regions. By the end of these investigations we aim at providing a detailed description of the molecular features of PLC-CSCs and delineate the potential of targeting stemness features to complement the existing therapeutic strategies in PLC.

Hepatocellular carcinoma (HCC) and cholangiocarcinoma (CCA) are two major types of primary liver cancers (PLC). PLCs rank among the deadliest cancers worldwide with poor clinical outcome. Absence of clinical symptoms and effective biomarkers for early detection are main causes of late diagnosis and detrimental prognosis. Therefore, identification of potentially druggable targets and subsequent development of next-generation targeted therapies, as well as translation of these findings into clinical practice remain crucial. To further govern translational advancements, development of primary culture models that closely recapitulate phenotypic and molecular diversities as well as the cell of origin, i.e. putative cancer stem cell (CSC), of PLC is urgently needed in order to support a translation from laboratory studies to the bedside and improve patient outcome. Major obstacle still represents a lack of models that truly and accurately represent morphological and molecular heterogeneity of original tumors. As an integral part of the here presented project, we report successful establishment and detailed characterization novel patient-derived primary liver cancer cell lines (PPCL) from HCC and intrahepatic CCA. The model was subsequently utilized to address and modulate important aspects of CSC biology in PLC. Morphologic and phenotypic characteristics of newly established cell lines were compared to original tumors and confirmed a high level of similarity. Tumor grading and genomic stability were further assessed to determine background setting for successful PPCL establishment. Our results provide evidence that less differentiated tumors with higher genomic instability possess a higher likelihood of successful PPCL establishment. Time course analyses of selected CSCs markers as well as transcriptomic and genomic changes were performed using FACS as well as nextgeneration sequencing (NGS) to characterize and generate molecular profiles of newly derived PPCL and investigate how accurately they recapitulate original cancer tissue. Analyses confirmed that PPCL retain similar profile as corresponding primary tumors during long-term culturing. Key oncogenic alterations were identified by targeted NGS (e.g. TP53, KRAS, CTNNB1, MET) and cell lines carrying potentially actionable mutations were treated with corresponding specific inhibitors. Overall, our integrative analyses demonstrate that PPCLs represent refined model for discovery of relevant molecular subgroups and exploration of precision medicine approaches for the treatment of primary liver cancer and modulation of putative CSCs.