Contribution of normal and error-prone ribosomes to translational error formation in vivo

Introduction of tRNA missense suppressors, and/or a protease deficiency into Escherichia coli strains has no significant effect on misreading of non-sense codons. An increased cellular level of faulty proteins therefore does not seem to have much secondary effect on translational accuracy. A genetic test system with two UGA non-sense mutations in the same fused lacIlacZ gene does not demonstrate any enrichment of error-prone ribosomes after read-through of the first non-sense codon in such strains. In contrast, the addition of sublethal amounts of streptomycin to a wild type strain appears to enrich error-prone ribosomes at the second non-sense codon, indicating the existence of a subpopulation of streptomycin-binding ribosomes. Ribosomes in a ribosomal ambiguity mutant strain (rpsD) with or without tRNA missense suppressors appear to be functionally homogeneous with respect to error production, as judged by read-through of the double UGA codons. The results that the major contribution to translational error formation in vivo originates from normal ribosomes and not from error-prone defective particles. An increased translational error in a bacterium results in very little, if any, increased functional heterogeneity of the ribosomal population with respect to error production. This suggests that an autocatalytic formation of translational errors is unlikely to occur in a growing bacterium.