Mechanism of the paradoxical, inhibitory effect of phenobarbital on hepatocarcinogenesis initiated in infant B6C3F1 mice with diethylnitrosamine.

Phenobarbital (PB), a classical rodent hepatopromoter, remarkably enhances hepatocarcinogenesis initiated by diethylnitrosamine (DEN) in adult B6C3F1 mice. However, it is also known to strongly inhibit liver tumor development in the B6C3F1 mice initiated with DEN in their infancy. The present study aimed to elucidate the unknown biological mechanisms for this paradoxical, inhibitory effect of PB on B6C3F1 mouse hepatocarcinogenesis. Male 12-day-old infant B6C3F1 mice were injected i.p. with DEN and, at 6 weeks of age, divided into PB-treated (PB+ group) and untreated (PB- group) animals. At 24 weeks, PB treatment was ceased for half of the PB+ animals (PB+/- group) and started for half of the PB- animals (PB-/+ group). Finally, all mice were sacrificed at 36 weeks and examined for the development of liver tumors. The mean multiplicity of gross tumors in the PB+ group was only one-fifteenth of that for the PB- group. PB-/+ animals developed fewer than half of the tumors found in PB- mice, indicating that the PB effect depends solely on the treatment duration, rather than the animal age. The effect was proven to be reversible, because the mean tumor multiplicity for the PB+/- group was seven times larger than that for the PB+ group. Stereological analysis revealed the mean volume of hepatocellular proliferative lesions in the PB- animals to be 7.7- and 4.1-fold the values for the PB+ and PB-/+ groups, respectively. The mean proliferating cell nuclear antigen-labeling indices for hepatocellular adenomas in PB+ and PB-/+ animals were also one-third of that for tumors in PB- animals, whereas no significant differences were observed with regard to the mean apoptotic index. In conclusion, the inhibitory effect of PB seemed to be primarily caused by the suppression of tumor cell proliferation. Irrespective of the group, most lesions observed were basophilic hepatocellular adenomas or foci, positive for Bcl-2 oncoprotein. They were thus distinct from the eosinophilic Bcl-2- lesions that predominate with PB promotion after the initiation of adult B6C3F1 mice. This age-dependent nature of initiation, together with the differential responses of Bcl-2+ and Bcl-2- lesions, may be responsible for the apparently contradictory outcomes of PB treatment in infant and adult B6C3F1 mice.