Genetic constraints on protein loop evolution: out-of-frame stop codons limit Bacillus subtilis stationary-phase mutagenesis

Insertion and deletion (indel) events in protein-coding DNA are associated with loss-of-function frameshift mutations. <i>In silico</i> studies have found that protein-coding regions are littered with out-of-frame stop codons (OSCs) which may restrict OSC frameshifts from becoming non-functional polypeptides by prematurely terminating translation. However, the effect of OSCs on protein variation has not been tested <i>in vivo</i>-specifically when OSCs are found in coding regions characterized by relaxed selection like protein loops. Here, we examined the formation of indel mutations in DNA encoding the <i>Bacillus subtilis</i> loop region of the ornithine carbamoyltransferase (Otc) protein by constructing a gain-of-function genetic system in strains with or without OSCs (+OSC and -OSC, respectively). The use of a gain-of-function selection system and a DNA region coding for a protein loop in which a wide spectrum of amino acids can restore function allowed us to maximize the spectrum of indels we could detect. We found that the +OSC strain had a lower indel frequency than the -OSC strain. Furthermore, indel events in the -OSC strain were not possible in the +OSC strain because they would shift the OSC into frame and terminate translation. By analyzing all indel events in both background strains, we found that the -OSC strain had a broader indel spectrum and led to more amino acid variation in the protein loop of Otc than did the +OSC strain. This variation was not due to GC content or the presence of homopolymers. Indel formation was dependent on the transcription-coupled repair factor Mfd and increased with elevated transcription. In addition to promoting the premature translation termination of non-functional peptides, this study provides the first <i>in vivo</i> evidence that OSCs are DNA features that restrict genetic changes and constrict the evolution of new protein domains like those that evolve from protein loops.