The use of initiated cells as a test system for the detection of inhibitors of gap junctional intercellular communication.

The effects of five non-mutagenic carcinogens--Aroclor 1260, benzoyl peroxide (BP), phenobarbital (PB), 12-O-tetradecanoyl-phorbol-13-acetate (TPA) and 1,1'-(2,2,2-trichloroethylidene)bis[4-chlorobenzene] (DDT)--on gap junctional intercellular communication (GJIC) were tested in a cell line consisting of initiated cells (3PC). Four agents suspected of tumor promotion activity--o-anisidine, clofibrate, L-ethionone and d-limonene--were also tested for their effects on GJIC. Finally sodium fluoride (NaF), whose carcinogenic property is still unclear, was tested for its effects on GJIC in the 3PC cell line. Four of the five selected tumor promoters (Aroclor 1260, BP, DDT and TPA) decreased GJIC between these initiated epidermal cells. The four non-mutagenic carcinogens with tumor-promoting activity in vivo (o-anisidine, clofibrate, L-ethionine and d-limonene) all inhibited GJIC, whereas NaF had no effect. Seven compounds (o-anisidine, Aroclor 1260, BP, DDT, L-ethionine, d-limonene and TPA) had a dose-dependent as well as time-dependent inhibitory effect on GJIC. Under the experimental conditions used, clofibrate showed only a dose-related inhibition of GJIC. PB showed no inhibitory effect on GJIC in the 3PC cell line. In order to determine the role of biotransformation in the tumor-promoting activity of PB, its effect on GJIC was also examined in the presence of an Aroclor 1254-induced rat liver homogenate (S9 mix) and in the hepatoma cell line HepG2. In the presence of rat liver homogenate PB decreased GJIC in the 3PC cell line, whereas in the HepG2 cells PB showed a time- and dose-dependent inhibitory effect. To study the potential differences in susceptibility of cells representing different stages in the process of tumor formation, the effect of the selected tumor promoters on GJIC was also investigated in primary mouse keratinocytes and in a mouse skin carcinoma-derived cell line (CA3/7). Primary keratinocytes were sometimes more (BP and clofibrate) and sometimes less sensitive (ethionine and limonene) for inhibitory effects on GJIC compared to the effects in the cell line 3PC. Except for TPA and anisidin, GJIC between the CA3/7 cells was less affected by the selected agents compared to the 3PC cell line. These results show that, during the process of tumor formation the susceptibility of cells to inhibition of GJIC by tumor promoters is variable. Overall the CA3/7 cells are less sensitive compared to 3PC cells. The susceptibility of primary keratinocytes is variable compared to 3PC cells, depending on the agent used. These results also show that GJIC is a valid parameter for testing the tumor-promoting activity of compounds. Finally, this study demonstrates that mouse keratinocyte cell lines could serve as an in vitro model for the detection of non-mutagenic carcinogens with diverse target organs in vivo. For this use the cell line consisting of initiated cells (3PC) is more sensitive than the carcinoma-derived cell line CA3/7.