CTLs- and lymphokine-induced apoptosis of infected hepatocytes during the course of chronic viral hepatitis is thought to be important for both disease termination and prevention of hepatocellular transformation. We therefore studied apoptosis induced by Fas (APO-1 or CD95)-a widely expressed cell surface receptor whose ligand is involved in lymphocyte cytotoxicity-in a set of human hepatoma cell lines. As normal hepatocytes, all of the human hepatoma cell lines tested do express detectable amounts of Fas on their surface. Nevertheless, only PLC/PRF/5 cells undergo apoptosis following treatment with anti-Fas. Systematic cloning and sequence analysis of the Fas cDNA did not show mutations in the Fas gene in any of the cells lines tested. However, due to alternative splicing, 5 to 10% of the Fas cDNAs are deleted of 63 internal nucleotides corresponding to the transmembrane domain, thus encoding for a soluble and secreted form of Fas (Fas delta TM), potentially able to neutralize anti-Fas or Fas-Ligand. Although we could not demonstrate a direct correlation between resistance of different hepatoma cell lines to Fas mediated death and endogenous expression of this transcript, we show that PLC/PRF 5 stable transfectants overexpressing Fas delta TM are less sensitive to anti-Fas than control cells. In three different cell lines, resistance to anti-Fas was overcome by treatment with the protein synthesis inhibitor cycloheximide. Although this could suggest the existence of short-lived repressors of the Fas-activated apoptotic signalling pathway(s), we show that translational inhibition is not required for the synergistic effect of cycloheximide to take place, and that resistant hepatoma cells can be sensitized to anti-Fas by subinhibitory concentrations of this protein synthesis inhibitor. Since cycloheximide is able to activate intracellular signalling independently on its effects on protein synthesis, we suggest that it might provide a costimulatory signal that cooperates with Fas in the induction of cell death and that, at least in the cells we tested, resistance to Fas is not an active process involving gene transcription and translation but only the consequence of an inadequate apoptotic stimulation.