Escape from TGF-β-induced senescence promotes aggressive hallmarks in epithelial hepatocellular carcinoma cells

Transforming growth factor-β (TGF-β) signaling and cellular senescence are key hallmarks of hepatocellular carcinoma (HCC) pathogenesis. While provoking senescence-associated growth arrest in epithelial HCC cells, elevated TGF-β activity paradoxically correlates with aggressiveness and poor prognosis in advanced tumors. Whether the transition between these dichotomous functions involves bypassing the senescence barrier during disease progression remains unknown. Exploiting the epithelial HCC cell line Huh7, we demonstrate that chronic TGF-β exposure prompts escape from Smad3-mediated senescence, leading to the development of TGF-β resistance. The resistant state is characterized by restoration of proliferative capacity and acquisition of molecular and functional traits of mesenchymal cells, coinciding with hybrid EMT, increased invasiveness, and metastasis. Mechanistically, resistant cells exhibit defective signaling of Smad molecules, as ectopic activation of the TGF-β/Smad3 axis reinstates TGF-β sensitivity. Gene expression landscape profiling reveals both shared and distinct gene signatures associated with senescent and TGF-β resistant states. Importantly, genetic ablation and molecular studies identify GRM8 (Glutamate Metabotropic Receptor 8) as a critical modulator of the resistance phenomenon, potentially by impairing subcellular spatiotemporal dynamics of Smad activity. Our findings unveil a novel phenomenon wherein epithelial HCC cells may exploit senescence evasion as a mechanism to oppose TGF-β anti-tumor responses and progress towards more aggressive HCC phenotypes. To investigate the gene expression profiles of senescent and resistant states, we established relevant cellular models. To induce cellular senescence, we treated Huh7 cells with 5 ng/mL TGF-β for 3 days. Resistant state was established by chronic TGF-β exposure of Huh7 cells for more than 50 days. We then performed gene expression profiling analysis using data obtained from RNA-seq of 3 different cellular states in three replicates.