Search

We have introduced a variety of amino acid substitutions into carboxyl-terminal CA1A2X sequence (C = cysteine; A = aliphatic; X = any amino acid) of the oncogenic [Val12]Ki-Ras4B protein to identify the amino acids that permit Ras processing (isoprenylation, proteolysis, and carboxyl methylation), membrane association, and transformation in cultured mammalian cells. While all substitutions were tolerated at the A1 position, substitutions at A2 and X reduced transforming activity. The A2 residue was important for both isoprenylation and AAX proteolysis, whereas the X residue dictated the extent and specificity of isoprenoid modification only. Differences were observed between Ras processing in living cells and farnesylation efficiency in a cell-free system. Finally, one farnesylated mutant did not undergo either proteolysis or carboxyl methylation but still displayed efficient membrane association (approximately 50%) and transforming activity, indicating that farnesylation alone can support Ras transforming activity. Since both farnesylation and carboxyl methylation are critical for yeast a-factor biological activity, the three CAAX-signaled modifications may have different contributions to the function of different CAAX-containing proteins.

p60src, the transforming protein kinase of Rous sarcoma virus, contains the 14-carbon saturated fatty acid, myristic acid, linked through an amide bond to the alpha-amino group of its NH2-terminal glycine residue. Myristic acid is known to be attached to four other eukaryotic proteins. In each case the fatty acid is also linked through an amide bond to an NH2-terminal glycine. We have used oligonucleotide-directed mutagenesis to examine the amino acid specificity of the enzyme that myristoylates the NH2 terminus of these proteins. Replacement of the NH2-terminal glycine in p60src with either alanine or glutamic acid prevented myristoylation completely. This indicates that the myristoylating enzyme may have an absolute specificity for glycine. Strikingly, neither nonmyristoylated mutant src protein induced morphological transformation of infected cells, even though wild-type levels of phosphorylation of cellular proteins on tyrosine were observed in these cells. Since conversion of the NH2-terminal residue from glycine to alanine should have little effect on the conformation of p60src, the inability of this mutant p60src protein to induce morphological transformation suggests that the myristoyl moiety is essential for the transforming activity of the protein.

Association of oncogenic ras proteins with cellular membranes appears to be a crucial step in transformation, ras is synthesized as a cytosolic precursor, which is processed to a mature form that localizes to the plasma membrane. This processing involves, in part, a conserved sequence, Cys-Ali-Ali-Xaa (in which Ali is an amino acid with an aliphatic side chain and Xaa is any amino acid), at the COOH terminus of ras proteins. Yeast a-factor mating hormone precursor also possesses a COOH-terminal Cys-Ali-Ali-Xaa sequence. However, while the COOH-terminal cysteine has been implicated as a site of palmitoylation of ras proteins, in mature a-type mating factor this residue is modified by an isoprenoid, a farnesyl moiety. We asked whether the Cys-Ali-Ali-Xaa sequence signaled different modifications for the yeast peptides (farnesylation) than for ras proteins (palmitoylation) or whether ras proteins were similar to the mating factors and contained a previously undiscovered isoprenoid. We report here that the processing of ras proteins involves addition of a farnesyl moiety, apparently at the COOH-terminal cysteine analogous to the cysteine modified in the yeast peptides, and that farnesylation may be important for membrane association and transforming activity of ras proteins.