Externally Suppressive +1 “Glycine” Frameshift: Possible Quadruplet Isomers for Glycine and Proline

Abstract

WE have reported our original finding of frameshift suppression in Salmonella^1,2. The frameshift we studied initially was induced in the histidinol dehydrogenase (HDH) gene with the intercalating agent ICR-191 (ref. 3.) It is a +1 type most likely containing an extra C in an mRNA repeat of C residues². External suppressors are efficiently induced by ICR-191 (ref. 1). The suppressors restore small amounts of HDH with the normal amino-acid sequence to the mutant cell⁴. We have hypothesized a proline suppressor tRNA with a quadruplet (+G) anticodon or its functional equivalent^2,4. Prompted by our findings, Riddle and Roth showed that most frameshifts tentatively classified as +1 types by genetic criteria are externally suppressible. Almost all were induced with ICR-191 (ref. 5). Two classes of suppressible frameshift were found, each with a set of mutually exclusive suppressors⁵. Judging from the demonstrated capacity of ICR compounds to produce + 1 additions in DNA repeats of GC pairs, we have further suggested to Riddle and Roth that these two frameshift-suppressor systems represent +1 additions in RNA repeats of C residues (proline codons, glycine anticodons) and in RNA repeats of G residues (glycine codons, proline anticodons)⁴ (personal communication to J. R. Roth, Histidine Workshop, 1970); that is, the two types of +1 frameshift are genetic “isomers”, the one involving proline and the other glycine (Fig. 1). The recent demonstration by Riddle and Roth of altered proline tRNA and glycine tRNA in appropriate suppressed strains⁶ is consistent with this suggestion. Further characterization of frameshifts of the type originally investigated has implicated a proline mRNA quadruplet, CCCg, as a sufficient if not necessary condition for suppression^7,8. A requirement for neighbouring sequences, particularly chain terminating codons, cannot be completely ruled out, however⁸. I have now examined a suppressible frameshift of the second type and present evidence that it contains a +1 addition in or near a glycine codon (Fig. 2). Oddly enough, this mRNA site is followed by an extensive nucleotide sequence reminiscent of two out of three +1 “proline” sequences examined (Fig. 2)⁸. The ICR compounds seem to have a marked bias for inducing suppressible +1 frameshifts in this extensive sequence. Whether some property of this extensive sequence is crucial to suppression is not yet clear.