The impact of mistranslation on phenotypic variability and fitness.

Phenotypic variation is widespread in natural populations, and can significantly alter population ecology and evolution. Phenotypic variation often reflects underlying genetic variation, but also manifests via non-heritable mechanisms. For instance, translation errors result in about 10% of cellular proteins carrying altered sequences. Thus, proteome diversification arising from translation errors can potentially generate phenotypic variability, in turn increasing variability in the fate of cells or of populations. However, the link between protein diversity and phenotypic variability remains unverified. We manipulated mistranslation levels in Escherichia coli, and measured phenotypic variability between single cells (individual-level variation), as well as replicate populations (population-level variation). Monitoring growth and survival, we find that mistranslation indeed increases variation across E. coli cells, but does not consistently increase variability in growth parameters across replicate populations. Interestingly, although any deviation from the wild-type (WT) level of mistranslation reduces fitness in an optimal environment, the increased variation is associated with a survival benefit under stress. Hence, we suggest that mistranslation-induced phenotypic variation can impact growth and survival and has the potential to alter evolutionary trajectories.

The Boggarts of biology: how non-genetic changes influence the genotype.

The notion that there is a one-one mapping from genotype to phenotype was overturned a long time ago. Along with genotype and environment, 'non-genetic changes' orchestrated by altered RNA and protein molecules also guide the development of phenotype. The idea that there is a route through which changes in phenotype can lead to changes in genotype impinges on several phenomena of molecular, developmental, evolutionary and applied interest. Phenotypic changes that do not alter the underlying DNA sequence have been studied across model systems (eg: DNA and histone modifications, RNA editing, prion formation) and are known to play an important role in short-term adaptation. However, because of their transient nature and unstable inheritance, the role of such changes in long-term evolution has remained controversial. I classify and review three ways in which non-genetic changes can influence genotype and impact cellular fitness across generations, with an emphasis on the enticing idea that they may act as stepping stones for genetic adaptation. I focus on work from microbial systems and attempt to highlight recent experiments and models that bear on this idea. Overall, I review evidence which suggests that non-genetic changes can impact phenotype via their influence on the genotype, and thus play a role in evolutionary change.

Global mistranslation increases cell survival under stress in Escherichia coli.

Mistranslation is typically deleterious for cells, although specific mistranslated proteins can confer a short-term benefit in a particular environment. However, given its large overall cost, the prevalence of high global mistranslation rates remains puzzling. Altering basal mistranslation levels of Escherichia coli in several ways, we show that generalized mistranslation enhances early survival under DNA damage, by rapidly activating the SOS response. Mistranslating cells maintain larger populations after exposure to DNA damage, and thus have a higher probability of sampling critical beneficial mutations. Both basal and artificially increased mistranslation increase the number of cells that are phenotypically tolerant and genetically resistant under DNA damage; they also enhance survival at high temperature. In contrast, decreasing the normal basal mistranslation rate reduces cell survival. This wide-ranging stress resistance relies on Lon protease, which is revealed as a key effector that induces the SOS response in addition to alleviating proteotoxic stress. The new links between error-prone protein synthesis, DNA damage, and generalised stress resistance indicate surprising coordination between intracellular stress responses and suggest a novel hypothesis to explain high global mistranslation rates.

How many initiator tRNA genes does Escherichia coli need?

Multiple copies of a gene require enhanced investment on the part of the cell and, as such, call for an explanation. The observation that Escherichia coli has four copies of initiator tRNA (tRNAi) genes, encoding a special tRNA (tRNA(fMet)) required to start protein synthesis, is puzzling particularly because the cell appears to be unaffected by the removal of one copy. However, the fitness of an organism has both absolute and relative connotations. Thus, we carried out growth competition experiments between E. coli strains that differ in the number of tRNAi genes they contain. This has enabled us to uncover an unexpected link between the number of tRNAi genes and protein synthesis, nutritional status, and fitness. Wild-type strains with the canonical four tRNAi genes are favored in nutrient-rich environments, and those carrying fewer are favored in nutrient-poor environments. Auxotrophs behave as if they have a nutritionally poor internal environment. A heuristic model that links tRNAi gene copy number, genetic stress, and growth rate accounts for the findings. Our observations provide strong evidence that natural selection can work through seemingly minor quantitative variations in gene copy number and thereby impact organismal fitness.

Initiation with elongator tRNAs.

In all domains of life, initiator tRNA functions exclusively at the first step of protein synthesis while elongator tRNAs extend the polypeptide chain. Unique features of initiator tRNA enable it to preferentially bind the ribosomal P site and initiate translation. Recently, we showed that the abundance of initiator tRNA also contributes to its specialized role. This motivates the question, can a cell also use elongator tRNA to initiate translation under certain conditions? To address this, we introduced non-AUG initiation codons CCC (Pro), GAG (Glu), GGU (Gly), UCU (Ser), UGU (Cys), ACG (Thr), AAU (Asn), and AGA (Arg) into the uracil DNA glycosylase gene (ung) used as a reporter gene. Enzyme assays from log-phase cells revealed initiation from non-AUG codons when intracellular initiator tRNA levels were reduced. The activity increased significantly in stationary phase. Further increases in initiation from non-AUG codons occurred in both growth phases upon introduction of plasmid-borne genes of cognate elongator tRNAs. Since purine-rich Shine-Dalgarno sequences occur frequently on mRNAs (in places other than the canonical AUG codon initiation contexts), initiation with elongator tRNAs from the alternate contexts may generate proteome diversity under stress without compromising genomic integrity. Thus, by changing the relative amounts of initiator and elongator tRNAs within the cell, we have blurred the distinction between the two classes of tRNAs thought to be frozen through years of evolution.

The "Clever Hans Phenomenon" revisited.

In the first decade of the 20th century, a horse named Hans drew worldwide attention in Berlin as the first and most famous "speaking" and thinking animal. Hans solved calculations by tapping numbers or letters with his hoof in order to answer questions. Later on, it turned out that the horse was able to give the correct answer by reading the microscopic signals in the face of the questioning person. This observation caused a revolution and as a consequence, experimenters avoided strictly any face-to-face contact in studies about cognitive abilities of animals-a fundamental lesson that is still not applied rigorously.

Unconventional initiator tRNAs sustain Escherichia coli.

Of all tRNAs, initiator tRNA is unique in its ability to start protein synthesis by directly binding the ribosomal P-site. This ability is believed to derive from the almost universal presence of three consecutive G-C base (3G-C) pairs in the anticodon stem of initiator tRNA. Consistent with the hypothesis, a plasmid-borne initiator tRNA with one, two, or all 3G-C pairs mutated displays negligible initiation activity when tested in a WT Escherichia coli cell. Given this, the occurrence of unconventional initiator tRNAs lacking the 3G-C pairs, as in some species of Mycoplasma and Rhizobium, is puzzling. We resolve the puzzle by showing that the poor activity of unconventional initiator tRNAs in E. coli is because of competition from a large pool of the endogenous WT initiator tRNA (possessing the 3G-C pairs). We show that E. coli can be sustained on an initiator tRNA lacking the first and third G-C pairs; thereby reducing the 3G-C rule to a mere middle G-C requirement. Two general inferences following from our findings, that the activity of a mutant gene product may depend on its abundance in the cell relative to that of the WT, and that promiscuous initiation with elongator tRNAs has the potential to enhance phenotypic diversity without affecting genomic integrity, have been discussed.

Functional significance of an evolutionarily conserved alanine (GCA) resume codon in tmRNA in Escherichia coli.

Occasionally, ribosomes stall on mRNAs prior to the completion of the polypeptide chain. In Escherichia coli and other eubacteria, tmRNA-mediated trans-translation is a major mechanism that recycles the stalled ribosomes. The tmRNA possesses a tRNA-like domain and a short mRNA region encoding a short peptide (ANDENYALAA in E. coli) followed by a termination codon. The first amino acid (Ala) of this peptide encoded by the resume codon (GCN) is highly conserved in tmRNAs in different species. However, reasons for the high evolutionary conservation of the resume codon identity have remained unclear. In this study, we show that changing the E. coli tmRNA resume codon to other efficiently translatable codons retains efficient functioning of the tmRNA. However, when the resume codon was replaced with the low-usage codons, its function was adversely affected. Interestingly, expression of tRNAs decoding the low-usage codon from plasmid-borne gene copies restored efficient utilization of tmRNA. We discuss why in E. coli, the GCA (Ala) is one of the best codons and why all codons in the short mRNA of the tmRNA are decoded by the abundant tRNAs.

Analysis of the fusA2 locus encoding EFG2 in Mycobacterium smegmatis.

The translation elongation factor G (EFG) is encoded by the fusA gene. Several bacteria possess a second fusA-like locus, fusA2 which encodes EFG2. A comparison of EFG and EFG2 from various bacteria reveals that EFG2 preserves domain organization and maintains significant sequence homology with EFG, suggesting that EFG2 may function as an elongation factor. However, with the single exception of a recent study on Thermus thermophilus EFG2, this class of EFG-like factors has not been investigated. Here, we have characterized EFG2 (MSMEG_6535) from Mycobacterium smegmatis. Expression of EFG2 was detected in stationary phase cultures of M. smegmatis (Msm). Our in vitro studies show that while MsmEFG2 binds guanine nucleotides, it lacks the ribosome-dependent GTPase activity characteristic of EFGs. Furthermore, unlike MsmEFG (MSMEG_1400), MsmEFG2 failed to rescue an E. coli strain harboring a temperature-sensitive allele of EFG, for its growth at the non-permissive temperature. Subsequent experiments showed that the fusA2 gene could be disrupted in M. smegmatis mc(2)155 with Kan(R) marker. The M. smegmatis fusA2::kan strain was viable and showed growth kinetics similar to that of the parent strain (wild-type for fusA2). However, in the growth competition assays, the disruption of fusA2 was found to confer a fitness disadvantage to M. smegmatis, raising the possibility that EFG2 is of some physiological relevance to mycobacteria.