Polylysine and CVIM sequences of K-RasB dictate specificity of prenylation and confer resistance to benzodiazepine peptidomimetic in vitro.

BZA-5B, a benzodiazepine peptidomimetic, inhibits CAAX farnesyltransferase (FTase) and blocks attachment of farnesyl groups to oncogenic and wild-type H-Ras in animal cells. This compound slows the growth of cells transformed with oncogenic H-Ras at concentrations that do not affect the growth of nontransformed cells. This finding suggested that nontransformed cells may produce a form of Ras whose prenylation is resistant to BZA-5B. In the current studies, we found that FTase had a 50-fold higher affinity for K-RasB than for H-Ras in vitro. Farnesylation of K-RasB was inhibited by BZA-2B, the active form of BZA-5B, but only at concentrations that were 8-fold higher than those that inhibited farnesylation of H-Ras. K-RasB, but not H-Ras, was also a substrate for CAAX geranylgeranyltransferase-1 (GGTase-1), and its affinity for the enzyme was equal to that of Rap1B, an authentic leucine-terminated substrate for GGTase-1. Inhibition of the geranylgeranylation of K-RasB occurred only at high concentrations of BZA-2B. All of these properties of K-RasB were traced to the combined effects of its COOH-terminal CVIM sequence and the adjacent polylysine sequence, neither of which is present in H-Ras. These studies provide a potential explanation for the resistance of nontransformed cells to growth inhibition by BZA-5B. Inasmuch as the majority of Ras-related human cancers contain oncogenic versions of K-RasB rather than H-Ras, the current data suggest that in vitro studies of FTase inhibitors with potential anti-cancer activity should use authentic K-RasB as a substrate.