Proofreading of the codon-anticodon interaction on ribosomes.

Abstract

The fidelity of protein synthesis is substantially greater than the specificity of codon-anticodon recognition that would be expected from the known energetics of base-pairing in solution. To test the suggestion that the specificity of recognition may be increased by "kinetic proofreading" associated with GTP hydrolysis [J. J. Hopfield (1974) Proc. Natl. Acad. Sci. USA 71, 4135-4139], we have studied the interaction of ternary complexes of polypeptide elongation factor Tu, aminoacyl-tRNA, and GTP with poly(U)-programed ribosomes. With most noncognate ternary complexes, including two that pair correctly with the 5' and 3' bases of UUU, rejection occurred without GTP hydrolysis, presumably by the reverse of the initial binding reaction. However, with complexes containing Leu- or Ile-tRNAs, which may pair correctly with the 3' and middle bases, GTP hydrolysis was stimulated though the aa-tRNA was not retained on the ribosome. These results demonstrate the existence of a GTP-dependent proofreading step in aminoacyl-tRNA recognition on ribosomes. They also suggest that the 5' base of the codon is more prone than the middle base to errors that can be corrected by proofreading.