Integrated miRNA-mRNA analysis revealed distinct molecular signatures and regulatory networks in cancer stem-like cell models of hepatocellular carcinoma

Purpose: Tumour micro-niche and cancer therapeutics have been implicated in cancer stem cells (CSCs) enrichment, contributing to cancer heterogeneity, aggressiveness, and resistance, eventually leading to metastasis and tumour relapse. We aim to clarify how these cellular pressures differentially influence the enrichment of CSC-like cells and elucidate the molecular regulatory networks underlying distinct CSCs. Method: Three HepG2-derived CSC-like models were established to recapitulate CSC induction from different pressures including i) tumoursphere (3D-HepG2), ii) TGF-β1-induced epithelial-mesenchymal transition (EMT; TβT-HepG2) and iii) cisplatin-resistant (CisR-HepG2) models. These models were then subjected to an integrated miRNA and mRNA profiling and molecular regulatory network analysis. Results: : All models exhibited enrichment of aldehyde dehydrogenase positive cells with differential upregulations of CD133 and distinct EpCAM species upregulated between different models. These models shared common propensity for increased stemness, quiescence, invasion, migration, and chemoresistance, while proliferation and EMT features were differential. Profiling analysis revealed the differentially expressed miRNA and mRNA of the 3D-HepG2 clustered closely to the parental, whereas distinct profiles were displayed by the TβT-HepG2 and CisR-HepG2. Seven miRNAs (miR-19b-3p miR-23a-3p, miR-29a-3p, miR-92a-3p, 125a-5p, miR-181a-5p, miR-483-3p) and 24 mRNAs - among which include CCNB1, PI3KR5, MAP2K6, MAPK10 and MAPK11 , were identified as key drivers underlying the molecular regulatory networks of these CSC-like models. Notably, the integrated analysis revealed the FoxO/PI3K-Akt/MAPK signalling and regulation G2/M transition and mitotic spindle checkpoints of cell cycle were identified as common network through the regulation of the key drivers. Conclusion: Altogether, the molecular regulatory networks from the three CSC-like models provide insights into the mechanisms regulating cancer plasticity and heterogeneity that may potentially be harnessed for the development of improved therapeutic approaches in targeting CSC for cancer treatment.