Abstract
The question of whether transforming growth factor-α (TGF-α) might be a useful marker of preneoplastic lesions during hepatocarcinogenesis was investigated using a medium-term liver carcinogenesis bioassay (Ito test) extended to 50 weeks. Male F344 rats, 6 weeks of age, received a single i.p. injection of 200 mg/kg diethylnitrosamine (DEN), then after 2 weeks, the animals were divided into four groups. Groups 1 and 2 were maintained on tap water supplemented with 10 ppm of N-nitrosomorpholine (NMOR), a genotoxic carcinogen, for 48 or 6 weeks, respectively. Animals in Group 3 were maintained on basal diet and tap water ad libitum as controls. Group 4 was included as the vehicle control and was maintained until experimental week 50. Two-thirds partial hepatectomy was performed on all rats at the end of week 3 of the experiment. Histopathology and immunohistochemical investigations of glutathione S-transferase placental form (GST-P) and TGF-α of the liver were performed sequentially at experimental weeks 8, 12, 20, 35 and 50. The numbers of TGF-α positive lesions were significantly lower than those of GST-P positive foci at all experimental time points. In contrast, the mean size values of TGF-α positive liver lesions in the continuous and 6-week treatments of DEN/NMOR, and the DEN alone protocol, were larger than those of their GST-P positive counterparts from week 35. Bromodeoxyuridine (BrdU) labeling indices for TGF-α positive foci and hepatocellular adenomas at week 20 were clearly higher than those for TGF-α negative lesions. In conclusion, the overexpression of TGF-α can be immunohistochemically detected in large hepatocellular lesions demonstrating increased levels of DNA synthesis. Therefore TGF-α may be a marker of a relatively late, progression stage of hepatocarcinogenesis.
