When safe is not safe together: Co-exposure to individually sub-cytotoxic compounds alters functional integrity of HepG2-derived liver 3D spheroids

This is the authors' abstract. We don't add key points for this paper.

Three-dimensional (3D) liver models provide increased physiological relevance for evaluating chemical toxicity; however, their suitability for assessing co-exposure effects remains underexplored. This study investigated the impact of co-exposure to individually sub-cytotoxic concentrations of pharmaceuticals and biopolymers in HepG2-derived 3D spheroids using multiparametric endpoints. HepG2 spheroids (∼800 cells/spheroid) were generated in ultra-low attachment 96-well plates and cultured for 5 days to reach functional steady state. Spheroids were then exposed for 48 h to rifampicin, tamoxifen, epsilon poly-L-lysine (ePL), hyaluronan (Hya), and poly-L-arginine (PAR). sub-cytotoxic concentrations were determined using 3D viability assays and applied for co-exposure studies. The following combinations were evaluated: rifampicin + tamoxifen, Hya + ePL, Hya + PAR, and ePL + PAR. Endpoints assessed included spheroid growth dynamics, viability, CYP3A4 activity, and levels of hepatic biomarkers (LDH, AST, urea and transferrin). Rifampicin-tamoxifen co-exposure increased spheroid size without affecting metabolic activity, CYP3A4, or most biomarkers, although AST was elevated, suggesting sub-cytotoxic hepatocellular stress. In contrast, Hya-ePL and ePL-PAR induced spheroid compaction and reduced metabolic activity, accompanied by increased CYP3A4 activity, indicating modulation of xenobiotic metabolism. Hya-ePL also reduced LDH and trended toward lower urea levels, suggesting altered hepatocellular turnover and urea cycle activity without overt cytotoxicity. Hya-PAR exposure did not significantly affect any endpoint. Overall, co-exposure to individually sub-cytotoxic compounds induced subtle but mechanistically relevant morphological and functional alterations in 3D liver spheroids. These findings demonstrate that mixture effects cannot be predicted from single-compound data alone and highlight the importance of mixture toxicity assessment in physiologically relevant in vitro systems.