Requirements for and kinetics of growth arrest of neoplastic cells by confluent 10T1/2 fibroblasts induced by a specific inhibitor of cyclic adenosine 3':5'-phosphodiesterase.

This study was conducted to further characterize the previously described phenomenon of growth inhibition of neoplastically transformed C3H/10T1/2 cells (T10T1/2) by nontransformed C3H/10T1/2 clone 8 mouse embryo fibroblast (10T1/2) cells in the presence of inhibitors of cyclic adenosine 3':5'-monophosphate (cAMP) phosphodiesterase. The cAMP phosphodiesterase inhibitor dl-4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone (RO20-1724) was shown to be completely nontoxic to T10T1/2 cells at 10(-4) M, yet when added to mixed cultures of T10T1/2 cells and postconfluence growth-arrested 10T1/2 cells, colony formation and [3H]thymidine incorporation into T10T1/2 cells were virtually eliminated. This effect was dose dependent and was reversible upon drug withdrawal. In 10T1/2 cells, RO20-1724 caused a dose-dependent increase in cAMP levels from about 5 to 150 pmol/10(6) cells; in T10T1/2 cells, 10(-4) M drug treatment caused a 5-fold elevation in cAMP without a clear dose dependency. Cyclic guanosine 3':5'-monophosphate levels in 10T1/2 cells fell by 50% with drug treatment but were unmeasurable in T10T1/2 cells. When intracellular cyclic AMP levels were elevated by the adenyl cyclase stimulator forskolin, growth inhibition of T10T1/2 cells was again induced in mixed cultures but was not observed when added to T10T1/2 cells alone. Addition of RO20-1724 to low concentrations of forskolin produced a greater than additive effect on growth inhibition. Growth inhibition of T10T1/2 cells by RO20-1724 was shown to (a) require contact with, or extremely close proximity to, a confluent monolayer of 10T1/2 cells, (b) be maximum when seeded upon a growth-inhibited monolayer and not an actively growing 10T1/2 culture, and (c) not be decreased by continuous agitation of the culture medium, indicating that readily diffusible inhibitory factors are not involved. A model is presented whereby transformed cells can respond to but cannot themselves generate growth-inhibitory signals produced by post-confluence growth-inhibited nontransformed cells. The existence of these cellular interactions may well explain problems in the quantitation of transformed foci encountered in the use of this cell line as an assay system for chemical and physical carcinogens.