Errors During Elongation Can Cause Translational Frameshifting

Abstract

In general, the translational machinery more efficiently eliminates frameshift errors than any other sort. It is actually difficult to estimate the frequency of these errors because the system used to measure them probably introduces specific sites which stimulate frameshifting much more efficiently than at random sites in mRNAs, as discussed below. We can only guess what the rate of frameshifting is at any randomly chosen codon, but it is certainly much lower than the value 5 τ 10⁻⁵ quoted in chapter 2. What this means is that the process of translational elongation is fundamentally structured so that the ribosome may not alter its reading frame. This perhaps should have been expected. In the theoretical discussions of the coding problem Crick et al 8 proposed that genes were read into proteins by successive reading of three-base, nonoverlapping codons on an unpunctuated mRNA—specifically stating that “(t)here are not special ‘commas’ to show how to select the right triplets.” The concept of an unpunctuated code explained the behavior of frameshift mutations. Once the frame were disrupted by the insertion or deletion of a base, translation continued in that frame reading a series of codons not present in the normal gene until encountering a nonsense codon in that frame and terminating translation.26 If, as Crick et al contended, the identity of the protein produced depended entirely on the selection of the site of initiation, with slavish, mechanical reading of successive codons following, then the translational machinery must include on a very fundamental level a mechanism for ensuring this repetitive reading of adjacent triplets. It follows that the occurrence of these errors might be expected to be very low since the structure of the ribosome evolved to preclude them.