A Comparative Genomics Analysis of Codon Reassignments Reveals a Link with Mitochondrial Proteome Size and a Mechanism of Genetic Code Change Via Suppressor tRNAs

Using a comparative genomics approach we demonstrate a negative correlation between the number of codon reassignments undergone by 222 mitochondrial genomes and the mitochondrial genome size, the number of mitochondrial ORFs, and the sizes of the large and small subunit mitochondrial rRNAs. In addition, we show that the TGA-to-tryptophan codon reassignment, which has occurred 11 times in mitochondrial genomes, is found in mitochondrial genomes smaller than those which have not undergone the reassignment. We therefore propose that mitochondrial codon reassignments occur in a wide range of phyla, particularly in Metazoa, due to a reduced “proteomic constraint” on the mitochondrial genetic code, compared to the nuclear genetic code. The reduced proteomic constraint reflects the small size of the mitochondrial-encoded proteome and allows codon reassignments to occur with less likelihood of lethality. In addition, we demonstrate a striking link between nonsense codon reassignments and the decoding properties of naturally occurring nonsense suppressor tRNAs. This suggests that natural preexisting nonsense suppression facilitated nonsense codon reassignments and constitutes a novel mechanism of genetic code change. These findings explain for the first time the identity of the stop codons and amino acids reassigned in mitochondrial and nuclear genomes. Nonsense suppressor tRNAs provided the raw material for nonsense codon reassignments, implying that the properties of the tRNA anticodon have dictated the identity of nonsense codon reassignments. Electronic Supplementary Material The online version of this article (doi:) contains supplementary material, which is available to authorized users. [Reviewing Editor: Dr. Laura Landweber]     

Selective advantages created by codon ambiguity allowed for the evolution of an alternative genetic code in Candida spp

Several species of the genus Candida decode the standard leucine CUG codon as serine. This and other deviations from the standard genetic code in both nuclear and mitochondrial genomes invalidate the notion that the genetic code is frozen and universal and prompt the questions ‘why alternative genetic codes evolved and, more importantly, how can an organism survive a genetic code change?’ To address these two questions, we have attempted to reconstruct the early stages of Candida albicans CUG reassignment in the closely related yeast Saccharomyces cerevisiae. These studies suggest that this genetic code change was driven by selection using a molecular mechanism that requires CUG ambiguity. Such codon ambiguity induced a significant decrease in fitness, indicating that CUG reassignment can only be selected if it introduces an evolutionary edge to counteract the negative impact of ambiguity. We have shown that CUG ambiguity induces the expression of a novel set of stress proteins and triggers the general stress response, which, in turn, creates a competitive edge under stress conditions. In addition, CUG ambiguity in S. cerevisiae induces the expression of a number of novel phenotypes that mimic the natural resistance to stress characteristic of C. albicans. The identification of an evolutionary advantage created by CUG ambiguity is the first experimental evidence for a genetic code change driven by selection and suggests a novel role for codon reassignment in the adaptation to new ecological niches.     

Pseudogene rescue: an adaptive mechanism of codon reassignment

Alterations to the genetic code – codon reassignments – have occurred many times in life’s history, despite the fact that genomes are coadapted to their genetic codes and therefore alterations are likely to be maladaptive. A potential mechanism for adaptive codon reassignment, which could trigger either a temporary period of codon ambiguity or a permanent genetic code change, is the reactivation of a pseudogene by a nonsense suppressor mutant transfer RNA. I examine the population genetics of each stage of this process and find that pseudogene rescue is plausible and also readily explains some features of extant variability in genetic codes.     

Searching of Code Space for an Error-Minimized Genetic Code Via Codon Capture Leads to Failure, or Requires At Least 20 Improving Codon Reassignments Via the Ambiguous Intermediate Mechanism

The standard genetic code (SGC) has a fundamental error-minimizing property which has been widely attributed to the action of selection. However, a clear mechanism for how selection can give rise to error minimization (EM) is lacking. A search through a space of alternate codes (code space) via codon reassignments would be required, to select a code optimized for EM. There are two commonly discussed mechanisms of codon reassignment; the Codon Capture mechanism, which proposes a loss of the codon during reassignment, and the Ambiguous Intermediate mechanism, which proposes that the codon underwent an ambiguous phase during reassignment. When searching of code space via the Codon Capture mechanism is simulated, an optimized genetic code can rarely be achieved (0–3.2% of the time) with most searches ending in failure. When code space is searched via the Ambiguous Intermediate mechanism, under constraints derived from empirical observations of codon reassignments from extant genomes, the searches also often end in failure. When a local minimum is avoided and optimization is achieved, 20–41 sequential improving codon reassignments are required. Furthermore, the structures of the optimized codes produced by these simulations differ from the structure of the SGC. These data are challenges for the Adaptive Code hypothesis to address, which proposes that the EM property was directly selected for, and suggests that EM is simply a byproduct of the addition of amino acids to the expanding code, as described by the alternative ‘Emergence’ hypothesis.     

Evolution of mitochondrial genomes and the genetic code.

Mitochondrial genomes are clearly marked by a strong tendency towards reductive evolution. This tendency has been facilitated by the transfer of most of the essential genes for mitochondrial propogation and function to the nuclear genome. The most extreme examples of genomic simplification are seen in animal mitochondria, where there also are the greatest tendencies to codon reassignment. The reassignment of codons to amino acids different from those designated in the so called universal code is seen in part as an expression of the reduction of the number of genes used by these genomes to code for tRNA species. The driving force for the reductive evolution of mitochondrial genomes is identified with two population genetic effects which may also be operating on populations of parasites.     

How Mitochondria Redefine the Code

Annotated, complete DNA sequences are available for 213 mitochondrial genomes from 132 species. These provide an extensive sample of evolutionary adjustment of codon usage and meaning spanning the history of this organelle. Because most known coding changes are mitochondrial, such data bear on the general mechanism of codon reassignment. Coding changes have been attributed variously to loss of codons due to changes in directional mutation affecting the genome GC content (Osawa and Jukes 1988), to pressure to reduce the number of mitochondrial tRNAs to minimize the genome size (Anderson and Kurland 1991), and to the existence of transitional coding mechanisms in which translation is ambiguous (Schultz and Yarus 1994a). We find that a succession of such steps explains existing reassignments well. In particular, (1) Genomic variation in the prevalence of a codon's third-position nucleotide predicts relative mitochondrial codon usage well, though GC content does not. This is because A and T, and G and C, are uncorrelated in mitochondrial genomes. (2) Codons predicted to reach zero usage (disappear) do so more often than expected by chance, and codons that do disappear are disproportionately likely to be reassigned. However, codons predicted to disappear are not significantly more likely to be reassigned. Therefore, low codon frequencies can be related to codon reassignment, but appear to be neither necessary nor sufficient for reassignment. (3) Changes in the genetic code are not more likely to accompany smaller numbers of tRNA genes and are not more frequent in smaller genomes. Thus, mitochondrial codons are not reassigned during demonstrable selection for decreased genome size. Instead, the data suggest that both codon disappearance and codon reassignment depend on at least one other event. This mitochondrial event (leading to reassignment) occurs more frequently when a codon has disappeared, and produces only a small subset of possible reassignments. We suggest that coding ambiguity, the extension of a tRNA's decoding capacity beyond its original set of codons, is the second event. Ambiguity can act alone but often acts in concert with codon disappearance, which promotes codon reassignment. Received: 26 October 2000 / Accepted: 19 January 2001     

A unified model of codon reassignment in alternative genetic codes

Many modified genetic codes are found in specific genomes in which one or more codons have been reassigned to a different amino acid from that in the canonical code. We present a new framework for codon reassignment that incorporates two previously proposed mechanisms (codon disappearance and ambiguous intermediate) and introduces two further mechanisms (unassigned codon and compensatory change). Our theory is based on the observation that reassignment involves a gain and a loss. The loss could be the deletion or loss of function of a tRNA or release factor. The gain could be the gain of a new type of tRNA or the gain of function of an existing tRNA due to mutation or base modification. The four mechanisms are distinguished by whether the codon disappears from the genome during the reassignment and by the order of the gain and loss events. We present simulations of the gain-loss model showing that all four mechanisms can occur within the same framework as the parameters are varied. We investigate the way the frequencies of the mechanisms are influenced by selection strengths, the number of codons undergoing reassignment, directional mutation pressure, and selection for reduced genome size.     

Experimental challenges of sense codon reassignment: An innovative approach to genetic code expansion

The addition of new and versatile chemical and biological properties to proteins pursued through incorporation of non-canonical amino acids is at present primarily achieved by stop codon suppression. However, it is critical to find new “blank” codons to increase the variety and efficiency of such insertions, thereby taking ‘sense codon recoding’ to center stage in the field of genetic code expansion. Current thought optimistically suggests the use of the pyrrolysine system coupled with re-synthesis of genomic information towards achieving sense codon reassignment. Upon review of the serious experimental challenges reported in recent studies, we propose that success in this area will depend on the re-synthesis of genomes, but also on ‘rewiring’ the mechanism of protein synthesis and of its quality control.     

Development of the genetic code: Insights from a fungal codon reassignment

The high conservation of the genetic code and its fundamental role in genome decoding suggest that its evolution is highly restricted or even frozen. However, various prokaryotic and eukaryotic genetic code alterations, several alternative tRNA-dependent amino acid biosynthesis pathways, regulation of tRNA decoding by diverse nucleoside modifications and recent in vivo incorporation of non-natural amino acids into prokaryotic and eukaryotic proteins, show that the code evolves and is surprisingly flexible. The cellular mechanisms and the proteome buffering capacity that support such evolutionary processes remain unclear. Here we explore the hypothesis that codon misreading and reassignment played fundamental roles in the development of the genetic code and we show how a fungal codon reassignment is enlightening its evolution.     

Nonstandard genetic codes and translation termination

Genetic code is not universal. Various nonstandard versions of the code are known for some mitochondrial, prokaryotic, and eukaryotic genomes. The most common deviation is stop codon reassignment; i.e., a stop codon is decoded as a sense codon rather than as a signal for translation termination. Class 1 release factors (RFs: prokaryotic RF1 and RF2 and eukaryotic eRF1) recognize the stop codons and induce hydrolysis of peptidyl-tRNA in the ribosome. The specificity of class 1 RFs changes in organisms with a nonstandard code. The rare amino acids selenocysteine and pyrrolysine utilize essentially different decoding strategies. The review considers several hypotheses of the origin of nonstandard genetic codes. A new hypothesis is advanced, assuming a change in the specificity of class 1 RFs as a starting point for stop codon reassignment.     

The non-standard genetic code of Candida spp.: an evolving genetic code or a novel mechanism for adaptation?

A number of yeasts of the genus Candida translate the standard leucine-CUG codon as serine. This unique genetic code change is the only known alteration to the universal genetic code in cytoplasmic mRNAs, of either eukaryotes or prokaryotes, which involves reassignment of a sense codon. Translation of CUG as serine in these species is mediated by a novel serine-tRNA (ser-tRNA_CAG), which uniquely has a guanosine at position 33, 5' to the anticodon, a position that is almost invariably occupied by a pyrimidine (uridine in general) in all other tRNAs. We propose that G-33 has two important functions: lowering the decoding efficiency of the ser-tRNA_CAG and preventing binding of the leucyl-tRNA synthetase. This implicates this nucleotide as a key player in the evolutionary reassignment of the CUG codon. In addition, the novel ser-tRNA_CAG has 1-methylguanosine (m¹G-37) at position 37, 3' to the anticodon, which is characteristic of leucine, but not serine tRNAs. Remarkably, m¹G-37 causes leucylation of the ser-tRNA_CAG both in vitro and in vivo , making the CUG codon an ambiguous codon: the polysemous codon. This indicates that some Candida species tolerate ambiguous decoding and suggests either that (i) the genetic code change has not yet been fully established and is evolving at different rates in different Candida species; or (ii) CUG ambiguity is advantageous and represents the final stage of the reassignment. We propose that such dual specificity indicates that reassignment of the CUG codon evolved through a mechanism that required codon ambiguity and that ambiguous decoding evolved to generate genetic diversity and allow for rapid adaptation to environmental challenges.     

Non-standard genetic codes and translation termination

Genetic code is not universal. Various non-standard versions of the code were found in mitochondrial, prokaryotic and eukaryotic genomes. Stop codons are used to signal the ribosome stop translation of the coding sequence and prone to reassignment to sense codons. Class-1 termination factors recognize stop codons and promote hydrolysis of the peptidyl-tRNA in ribosome (RF1, RF2 in prokaryotes and eRF1 in eukaryotes). The class-1 factor termination specificity is changed in non-standart codes organisms. Pyrrolysine and selenocysteine use dissimilar decoding strategies. The various non-standart code origin hypotheses are described. It was proposed that specificity alteration of the class-1 release factor was a starting point for stop codon reassignment.     

Origin of an Alternative Genetic Code in the Extremely Small and GC–Rich Genome of a Bacterial Symbiont

The genetic code relates nucleotide sequence to amino acid sequence and is shared across all organisms, with the rare exceptions of lineages in which one or a few codons have acquired novel assignments. Recoding of UGA from stop to tryptophan has evolved independently in certain reduced bacterial genomes, including those of the mycoplasmas and some mitochondria. Small genomes typically exhibit low guanine plus cytosine (GC) content, and this bias in base composition has been proposed to drive UGA Stop to Tryptophan (Stop→Trp) recoding. Using a combination of genome sequencing and high-throughput proteomics, we show that an α-Proteobacterial symbiont of cicadas has the unprecedented combination of an extremely small genome (144 kb), a GC–biased base composition (58.4%), and a coding reassignment of UGA Stop→Trp. Although it is not clear why this tiny genome lacks the low GC content typical of other small bacterial genomes, these observations support a role of genome reduction rather than base composition as a driver of codon reassignment.     

Origin of an Alternative Genetic Code in the Extremely Small and GC–Rich Genome of a Bacterial Symbiont

The genetic code relates nucleotide sequence to amino acid sequence and is shared across all organisms, with the rare exceptions of lineages in which one or a few codons have acquired novel assignments. Recoding of UGA from stop to tryptophan has evolved independently in certain reduced bacterial genomes, including those of the mycoplasmas and some mitochondria. Small genomes typically exhibit low guanine plus cytosine (GC) content, and this bias in base composition has been proposed to drive UGA Stop to Tryptophan (Stop→Trp) recoding. Using a combination of genome sequencing and high-throughput proteomics, we show that an α-Proteobacterial symbiont of cicadas has the unprecedented combination of an extremely small genome (144 kb), a GC–biased base composition (58.4%), and a coding reassignment of UGA Stop→Trp. Although it is not clear why this tiny genome lacks the low GC content typical of other small bacterial genomes, these observations support a role of genome reduction rather than base composition as a driver of codon reassignment.     

Natural reassignment of CUU and CUA sense codons to alanine in Ashbya mitochondria

The discovery of diverse codon reassignment events has demonstrated that the canonical genetic code is not universal. Studying coding reassignment at the molecular level is critical for understanding genetic code evolution, and provides clues to genetic code manipulation in synthetic biology. Here we report a novel reassignment event in the mitochondria of Ashbya (Eremothecium) gossypii, a filamentous-growing plant pathogen related to yeast (Saccharomycetaceae). Bioinformatics studies of conserved positions in mitochondrial DNA-encoded proteins suggest that CUU and CUA codons correspond to alanine in A. gossypii, instead of leucine in the standard code or threonine in yeast mitochondria. Reassignment of CUA to Ala was confirmed at the protein level by mass spectrometry. We further demonstrate that a predicted is transcribed and accurately processed in vivo, and is responsible for Ala reassignment. Enzymatic studies reveal that is efficiently recognized by A. gossypii mitochondrial alanyl-tRNA synthetase (AgAlaRS). AlaRS typically recognizes the G3:U70 base pair of tRNA^Ala; a G3A change in Ashbya abolishes its recognition by AgAlaRS. Conversely, an A3G mutation in Saccharomyces cerevisiae confers tRNA recognition by AgAlaRS. Our work highlights the dynamic feature of natural genetic codes in mitochondria, and the relative simplicity by which tRNA identity may be switched.     

Consequences of stop codon reassignment on protein evolution in ciliates with alternative genetic codes

Tetrahymena thermophila and Paramecium tetraurelia are ciliates that reassign TAA and TAG from stop codons to glutamine codons. Because of the lack of full genome sequences, few studies have concentrated on analyzing the effects of codon reassignment in protein evolution. We used the recently sequenced genome of these species to analyze the patterns of amino acid substitution in ciliates that reassign the code. We show that, as expected, the codon reassignment has a large impact on amino acid substitutions in closely related proteins; however, contrary to expectations, these effects also hold for very diverged proteins. Previous studies have used amino acid substitution data to calculate the minimization of the genetic code; our results show that because of the lasting influence of the code in the patterns of substitution, such studies are tautological. These different substitution patterns might affect alignment of ciliate proteins, as alignment programs use scoring matrices based on substitution patterns of organisms that use the standard code. We also show that glutamine is used more frequently in ciliates than in other species, as often as expected based on the presence of the 2 new reassigned codons, indicating that the frequencies of amino acids in proteomes is mostly determined by neutral processes based on their number of codons.     

The molecular mechanism of evolution of changes in the genetic code

Alterations to the standard genetic code have been found in both prokaryotes and eukaryotes. This finding demolished the central dogma of molecular biology, postulated by Crick in 1968, of an immutable and universal genetic code and raised the question of how organisms survive genetic code alterations? Recent studies suggest that genetic code alterations are driven by selection using a mechanism that requires translational ambiguity. In C. albicans, the leucine CUG codon is decoded as serine through structural alterations of the translational machinery, in particular, of a Ser-tRNACAG which has dual identity and novel decoding properties. Here, we review the molecular mechanism of CUG reassignment focusing on the structural change of the translational machinery and on the impact that such alteration had on the evolution of the Candida albicans genome.     

Codon reassignment in Candida species: An evolutionary conundrum

A number of Candida species translate the standard leucine CUG codon as serine rather than as leucine. Such codon reassignment in nuclear-encoded mRNAs is unusual and raises a number of important questions about the origin of the genetic code and its continuing evolution. In particular we must establish how a codon can come to be reassigned without extinction of the species and what, if any, selective pressure drives such potentially catastrophic changes. Recent studies on the structure and identity of the novel CUG-decoding tRNA^Ser from several different Candida species have begun to shed light on possible evolutionary mechanisms which could have facilitated such changes to the genetic code. These findings are reviewed here and a possible molecular mechanism proposed for how the standard leucine CUG codon could have become reassigned as a serine codon.     

Codon reassignment in the Escherichia coli genetic code

Most organisms, from Escherichia coli to humans, use the ‘universal’ genetic code, which have been unchanged or ‘frozen’ for billions of years. It has been argued that codon reassignment causes mistranslation of genetic information, and must be lethal. In this study, we successfully reassigned the UAG triplet from a stop to a sense codon in the E. coli genome, by eliminating the UAG-recognizing release factor, an essential cellular component, from the bacterium. Only a few genetic modifications of E. coli were needed to circumvent the lethality of codon reassignment; erasing all UAG triplets from the genome was unnecessary. Thus, UAG was assigned unambiguously to a natural or non-natural amino acid, according to the specificity of the UAG-decoding tRNA. The result reveals the unexpected flexibility of the genetic code.     

Mitochondrial Genetic Codes Evolve to Match Amino Acid Requirements of Proteins

Mitochondria often use genetic codes different from the standard genetic code. Now that many mitochondrial genomes have been sequenced, these variant codes provide the first opportunity to examine empirically the processes that produce new genetic codes. The key question is: Are codon reassignments the sole result of mutation and genetic drift? Or are they the result of natural selection? Here we present an analysis of 24 phylogenetically independent codon reassignments in mitochondria. Although the mutation-drift hypothesis can explain reassignments from stop to an amino acid, we found that it cannot explain reassignments from one amino acid to another. In particular—and contrary to the predictions of the mutation-drift hypothesis—the codon involved in such a reassignment was not rare in the ancestral genome. Instead, such reassignments appear to take place while the codon is in use at an appreciable frequency. Moreover, the comparison of inferred amino acid usage in the ancestral genome with the neutral expectation shows that the amino acid gaining the codon was selectively favored over the amino acid losing the codon. These results are consistent with a simple model of weak selection on the amino acid composition of proteins in which codon reassignments are selected because they compensate for multiple slightly deleterious mutations throughout the mitochondrial genome. We propose that the selection pressure is for reduced protein synthesis cost: most reassignments give amino acids that are less expensive to synthesize. Taken together, our results strongly suggest that mitochondrial genetic codes evolve to match the amino acid requirements of proteins.