Editing Mechanisms in the Aminoacylation of tRNA

Abstract

There is a basic problem in protein biosynthesis in that there are insufficient differences in structure between certain amino acids to enable a sufficiently precise distinction between them by simple preferential binding of the correct substrate to the aminoacyl-tRNA synthetase. The classic example, anticipated by Pauling (1958), is the competition between valine and isoleucine for the isoleucyl-tRNA synthetase. Since the active site of the enzyme is large enough, or flexible enough, to accommodate isoleucine, it must also bind the smaller molecule of valine. The difference in binding energy between the two amino acids cannot be greater than that of the additional methylene group on the larger substrate, now known to be worth up to about 3 kcal/molecule, equivalent to a factor of 200–300. This sets the upper limit on the discrimination between valine and isoleucine by simple binding. Furthermore, this cannot be increased by invoking any of the theories of enzyme catalysis, such as “strain,” “induced fit,” and “nonproductive binding,” or by the effects of a series of sequential steps or interacting active sites (Fersht 1974Fersht 1977a). Yet, it has been found by Loftfield (1963; Loftfield and Vanderjagt 1972) that the overall error frequency for the misincorporation of valine for isoleucine in the synthesis of ovalbumin and globin is only 3 × 10⁻⁴, and this includes any errors in transcription and translation.

The increase in specificity results from the evolution of an editing or proofreading function whereby the aminoacyl-tRNA synthetase has a hydrolytic activity that specifically destroys by deacylation...