Conservation and evolution of the programmed ribosomal frameshift in prfB across the bacterial domain

When the ribosome reaches a stop codon, translation is terminated by a release factor. Bacteria encode two release factors, RF1 and RF2. In many bacteria, the gene encoding RF2 (prfB) contains an in-frame premature stop codon near the beginning of the open reading frame. A programmed ribosomal frameshift is therefore required to translate full-length RF2. While the molecular mechanism of the programmed ribosomal frameshift has been extensively characterized in Escherichia coli, bioinformatic analysis of the evolution and conservation of this motif has been limited to few genomes. By analyzing >12,000 bacterial genomes, we sought to thoroughly characterize the conserved frameshifting elements within the programmed frameshift motif and identify genomic features of phyla that have lost the motif altogether. We find that the programmed ribosomal frameshift in prfB was likely present in the last common ancestor of bacteria and that the motif elements are almost completely conserved, including the identity of the premature stop codon. We find that loss of the programmed frameshift motif is highly correlated with RF2-specific stop codon usage, suggesting that stop codon usage has shaped the conservation of this regulatory mechanism. In support of this model, the programmed frameshift in prfB is entirely absent in Actinomycetota, which have particularly high RF2-specific stop codon usage. Finally, we show that a model member of Actinomycetota fails to produce full-length RF2 when provided with an allele of prfB that contains the programmed frameshift motif. Altogether, our work provides a thorough characterization of RF2 regulation across the bacterial domain.IMPORTANCETranslation termination is catalyzed by one of two release factors in bacteria, RF1 or RF2. It has been known for decades that RF2 levels in Escherichia coli are regulated by a programmed ribosomal frameshift within the prfB gene that encodes RF2. We investigated the conservation and features of the prfB programmed ribosomal frameshift in >12,000 genomes across the bacterial domain. Our data suggest this autoregulatory motif was present in the common ancestor of bacteria and that organisms that lost the motif have high RF2-specific stop codon usage. We also find that overexpression of RF2 from prfB lacking the programmed frameshift motif is toxic to Bacillus subtilis, as has been observed in E. coli. The fitness cost of RF2 overexpression in distantly related species coupled with its broad conservation suggests that RF2 autoregulation imparts a strong selective advantage in organisms that do not have high RF2-specific stop codon usage.