Base substitutions in the tRNA anticodon arm do not degrade the accuracy of reading frame maintenance.

Abstract

We have examined the activities of a set of 34 site-directed mutants of tRNA Su7 for their ability to shift reading frame during translation of amber codons in vivo. The set includes variants at every position in the distal three base pairs of the anticodon stem and saturates the anticodon loop, with the exception of the anticodon itself. Most anticodon-stem mutations were made pairwise to preserve the secondary structure of that region. Variants of the Hirsh (A24) coding alteration were also tested. The mutations have varied and often dramatic effects on the ability of Su7 to act in translation, which indicates that they cause distortions of the codon-anticodon complex. However, none of the tested mutations affects the intrinsic accuracy of translocation, which we show to be very high. These results suggest that translocation must be independent of the conformational detail of the codon-anticodon complex and stand in contrast to frameshifts that occur when tRNAs misread codons. We suggest that when the tRNA is properly paired to the codon, translocation proceeds normally. Thus, we conclude that selection of a cognate tRNA ensures highly accurate reading frame maintenance. As a corollary, inefficient amber suppressors are not inefficient because they frameshift. Instead, they are likely to fail because a release factor translates the amber codon.