A large-scale analysis of codon usage bias in 4868 bacterial genomes shows association of codon adaptation index with GC content,protein functional domains and bacterial phenotypes

Multiple synonymous codons code for the same amino acid, resulting in the degeneracy of the genetic code and in the preferred used of some codons called codon bias usage (CBU). We performed a large-scale analysis of codon usage bias analysing the distribution of the codon adaptation index (CAI) and the codon relative adaptiveness index (RA) in 4868 bacterial genomes. We found that CAI values differ significantly between protein functional domains and part of the protein outside domains and show how CAI, GC content and preferred usage of polymerase III alpha subunits are related. Additionally, we give evidence of the association between CAI and bacterial phenotypes.     

Comparative analysis of the base biases at the gene terminal portions in seven eukaryote genomes

Adenine nucleotides have been found to appear preferentially in the regions after the initiation codons or before the termination codons of bacterial genes. Our previous experiments showed that AAA and AAT, the two most frequent second codons in Escherichia coli, significantly enhance translation efficiency. To determine whether such a characteristic feature of base frequencies exists in eukaryote genes, we performed a comparative analysis of the base biases at the gene terminal portions using the proteomes of seven eukaryotes. Here we show that the base appearance at the codon third positions of gene terminal regions is highly biased in eukaryote genomes, although the codon third positions are almost free from amino acid preference. The bias changes depending on its position in a gene, and is characteristic of each species. We also found that bias is most outstanding at the second codon, the codon after the initiation codon. NCN is preferred in every genome; in particular, GCG is strongly favored in human and plant genes. The presence of the bias implies that the base sequences at the second codon affect translation efficiency in eukaryotes as well as bacteria.     

Comprehensive analysis of prokaryotic mechanosensation genes: their characteristics in codon usage.

In the present study, we examined GC nucleotide composition, relative synonymous codon usage (RSCU), effective number of codons (ENC), codon adaptation index (CAI) and gene length for 308 prokaryotic mechanosensitive ion channel (MSC) genes from six evolutionary groups: Euryarchaeota, Actinobacteria, Alphaproteobacteria, Betaproteobacteria, Firmicutes, and Gammaproteobacteria. Results showed that: (1) a wide variation of overrepresentation of nucleotides exists in the MSC genes; (2) codon usage bias varies considerably among the MSC genes; (3) both nucleotide constraint and gene length play an important role in shaping codon usage of the bacterial MSC genes; and (4) synonymous codon usage of prokaryotic MSC genes is phylogenetically conserved. Knowledge of codon usage in prokaryotic MSC genes may benefit from the study of the MSC genes in eukaryotes in which few MSC genes have been identified and functionally analysed.     

Patterns of codon usage in two ciliates that reassign the genetic code: Tetrahymena thermophila and Paramecium tetraurelia

We used the recently sequenced genomes of the ciliates Tetrahymena thermophila and Paramecium tetraurelia to analyze the codon usage patterns in both organisms; we have analyzed codon usage bias, Gln codon usage, GC content and the nucleotide contexts of initiation and termination codons in Tetrahymena and Paramecium. We also studied how these trends change along the length of the genes and in a subset of highly expressed genes. Our results corroborate some of the trends previously described in Tetrahymena, but also negate some specific observations. In both genomes we found a strong bias toward codons with low GC content; however, in highly expressed genes this bias is smaller and codons ending in GC tend to be more frequent. We also found that codon bias increases along gene segments and in highly expressed genes and that the context surrounding initiation and termination codons are always AT rich. Our results also suggest differences in the efficiency of translation of the reassigned stop codons between the two species and between the reassigned codons. Finally, we discuss some of the possible causes for such translational efficiency differences.     

Nucleotides downstream of start codons show marked non-randomness in Escherichia coli but not in Bacillus subtilis

This study aimed at measuring the nucleotide non-randomness in the region downstream of start codons in bacterial genes and to see if the non-randomness differs between biased and unbiased genes, in terms of the effective number of codons (Nc) and the codon adaptation index (CAI). In Escherichia coli, there was a marked elevation in nucleotide conservation for the genes having low Nc-values compared to the genes having high Nc-values, i.e the more biased genes showed a higher level of non-randomness. Likewise, the genes displaying high CAI-values showed stronger nucleotide conservation than the genes of low CAI-values. This elevated conservation is visible up to approximately 15-17 nucleotides downstream of the start codon, after which there is little difference. This indicates that there may be distinct selectional mechanisms acting upon the first 5-6 codons within genes in E. coli. In B. subtilis, these effects are less pronounced, if present at all. Furthermore, analyses of codons used in this region were not in support of the hypothesis that the elevation in nucleotide non-randomness is a question of selection for certain optimal codons.     

Suppression of the negative effect of minor arginine codons on gene expression; preferential usage of minor codons within the first 25 codons of the Escherichia coli genes.

AGA and AGG codons for arginine are the least used codons in Escherichia coli, which are encoded by a rare tRNA, the product of the dnaY gene. We examined the positions of arginine residues encoded by AGA/AGG codons in 678 E. coli proteins. It was found that AGA/AGG codons appear much more frequently within the first 25 codons. This tendency becomes more significant in those proteins containing only one AGA or AGG codon. Other minor codons such as CUA, UCA, AGU, ACA, GGA, CCC and AUA are also found to be preferentially used within the first 25 codons. The effects of the AGG codon on gene expression were examined by inserting one to five AGG codons after the 10th codon from the initiation codon of the lacZ gene. The production of beta-galactosidase decreased as more AGG codons were inserted. With five AGG codons, the production of beta-galactosidase (Gal-AGG5) completely ceased after a mid-log phase of cell growth. After 22 hr induction of the lacZ gene, the overall production of Gal-AGG5 was 11% of the control production (no insertion of arginine codons). When five CGU codons, the major arginine codon were inserted instead of AGG, the production of beta-galactosidase (Gal-CGU5) continued even after stationary phase and the overall production was 66% of the control. The negative effect of the AGG codons on the Gal-AGG5 production was found to be dependent upon the distance between the site of the AGG codons and the initiation codon. As the distance was increased by inserting extra sequences between the two codons, the production of Gal-AGG5 increased almost linearly up to 8 fold. From these results, we propose that the position of the minor codons in an mRNA plays an important role in the regulation of gene expression possibly by modulating the stability of the initiation complex for protein synthesis.     

Efficiency of translation initiation by non-AUG codons in Saccharomyces cerevisiae.

The quantitative levels of initiation of protein synthesis at codons other than AUG were determined with a CYC7-lacZ fused gene in the yeast Saccharomyces cerevisiae. AUG was the only codon which efficiently initiated translation, although some non-AUG codons allowed initiation at very low efficiency, below 1% of the normal level. Since translation initiates at codons other than AUG in at least two wild-type genes from eucaryotes, other factors presumably play a role in enhancing the activity of non-AUG codons.     

Compositional variation in bacterial genes and proteins with potential expression level

Usage of guanine and cytosine at three codon sites in eubacterial genes vary distinctly with potential expressivity, as predicted by Codon Adaptation Index (CAI). In bacteria with moderate/high GC-content, G(3) follows a biphasic relationship, while C(3) increases with CAI. In AT-rich bacteria, correlation of CAI is negative with G(3), but non-specific with C(3). Correlations of CAI with residues encoded by G-starting codons are positive, while with those by C-starting codons are usually negative/random. Average Size/Complexity Score and aromaticity of gene-products decrease with CAI, confirming general validity of cost-minimization principle in free-living eubacteria. Alcoholicity of bacterial gene-products usually decreases with expressivity.     

Synonymous codon usage in chloroplast genome of Coffea arabica

Synonymous codon usage of 53 protein coding genes in chloroplast genome of Coffea arabica was analyzed for the first time to find out the possible factors contributing codon bias. All preferred synonymous codons were found to use A/T ending codons as chloroplast genomes are rich in AT. No difference in preference for preferred codons was observed in any of the two strands, viz., leading and lagging strands. Complex correlations between total base compositions (A, T, G, C, GC) and silent base contents (A₃, T₃, G₃, C₃, GC₃) revealed that compositional constraints played crucial role in shaping the codon usage pattern of C. arabica chloroplast genome. ENC Vs GC₃ plot grouped majority of the analyzed genes on or just below the left side of the expected GC₃ curve indicating the influence of base compositional constraints in regulating codon usage. But some of the genes lie distantly below the continuous curve confirmed the influence of some other factors on the codon usage across those genes. Influence of compositional constraints was further confirmed by correspondence analysis as axis 1 and 3 had significant correlations with silent base contents. Correlation of ENC with axis 1, 4 and CAI with 1, 2 prognosticated the minor influence of selection in nature but exact separation of highly and lowly expressed genes could not be seen. From the present study, we concluded that mutational pressure combined with weak selection influenced the pattern of synonymous codon usage across the genes in the chloroplast genomes of C. arabica.     

Comparison of codon usage and tRNAs in mitochondrial genomes of Candida species

To gain insight into the nature of the mitochondrial genomes (mtDNA) of different Candida species, the synonymous codon usage bias of mitochondrial protein coding genes and the tRNAs in C. albicans, C. parapsilosis, C. stellata, C. glabrata and the closely related yeast Saccharomyces cerevisiae were analyzed. Common features of the mtDNA in Candida species are a strong A+T pressure on protein coding genes, and insufficient mitochondrial tRNA species are encoded to perform protein synthesis. The wobble site of the anticodon is always U for the NNR (NNA and NNG) codon families, which are dominated by A-ending codons, and always G for the NNY (NNC and NNU) codon families, which is dominated by U-ending codons, and always U for the NNN (NNA, NNU, NNC and NNG) codon families, which are dominated by A-ending codons and U-ending codons. Patterns of synonymous codon usage of Candida species can be classified into three groups: (1) optimal codon-anticodon usage, Glu, Lys, Leu (translated by anti-codon UAA), Gln, Arg (translated by anti-codon UCU) and Trp are containing NNR codons. NNA, whose corresponding tRNA is encoded in the mtDNA, is used preferentially. (2) Non-optimal codon-anticodon usage, Cys, Asp, Phe, His, Asn, Ser (translated by anti-codon GCU) and Tyr are containing NNY codons. The NNU codon, whose corresponding tRNA is not encoded in the mtDNA, is used preferentially. (3) Combined codon-anticodon usage, Ala, Gly, Leu (translated by anti-codon UAG), Pro, Ser (translated by anti-codon UGA), Thr and Val are containing NNN codons. NNA (tRNA encoded in the mtDNA) and NNU (tRNA not encoded in the mtDNA) are used preferentially. In conclusion, we propose that in Candida species, codons containing A or U at third position are used preferentially, regardless of whether corresponding tRNAs are encoded in the mtDNA. These results might be useful in understanding the common features of the mtDNA in Candida species and patterns of synonymous codon usage.     

Analysis of synonymous codon usage in <Emphasis Type="Italic">Zea mays</Emphasis>

It is important and meaningful to understand the codon usage pattern and the factors that shape codon usage of maize. In this study, trends in synonymous codon usage in maize have been firstly examined through the multivariate statistical analysis on 7402 cDNA sequences. The results showed that the genes positions on the primary axis were strongly negatively correlated with GC3s, GC content of individual gene and gene expression level assessed by the codon adaptation index (CAI) values, which indicated that nucleotide composition and gene expression level were the main factors in shaping the codon usage of maize, and the variation in codon usage among genes may be due to mutational bias at the DNA level and natural selection acting at the level of mRNA translation. At the same time, CDS length and the hydrophobicity of each protein were, respectively, significantly correlated with the genes locations on the primary axis, GC3s and CAI values. We infer that genes length and the hydrophobicity of the encoded protein may play minor role in shaping codon usage bias. Additional 28 codons ending with a G or C base have been defined as “optimal codons”, which may provide useful information for maize gene-transformation and gene prediction.     

Codon usage and protein length-dependent feedback from translation elongation regulates translation initiation and elongation speed

Essential cellular functions require efficient production of many large proteins but synthesis of large proteins encounters many obstacles in cells. Translational control is mostly known to be regulated at the initiation step. Whether translation elongation process can feedback to regulate initiation efficiency is unclear. Codon usage bias, a universal feature of all genomes, plays an important role in determining gene expression levels. Here, we discovered that there is a conserved but codon usage-dependent genome-wide negative correlation between protein abundance and CDS length. The codon usage effects on protein expression and ribosome flux on mRNAs are influenced by CDS length; optimal codon usage preferentially promotes production of large proteins. Translation of mRNAs with long CDS and non-optimal codon usage preferentially induces phosphorylation of initiation factor eIF2α, which inhibits translation initiation efficiency. Deletion of the eIF2α kinase CPC-3 (GCN2 homolog) in Neurospora preferentially up-regulates large proteins encoded by non-optimal codons. Surprisingly, CPC-3 also inhibits translation elongation rate in a codon usage and CDS length-dependent manner, resulting in slow elongation rates for long CDS mRNAs. Together, these results revealed a codon usage and CDS length-dependent feedback mechanism from translation elongation to regulate both translation initiation and elongation kinetics.     

Analysis of codon usage pattern in the radioresistant bacterium Deinococcus radiodurans

The main factors shaping codon usage bias in the Deinococcus radiodurans genome were reported. Correspondence analysis (COA) was carried out to analyze synonymous codon usage bias. The results showed that the main trend was strongly correlated with gene expression level assessed by the "Codon Adaptation Index" (CAI) values, a result that was confirmed by the distribution of genes along the first axis. The results of correlation analysis, variance analysis and neutrality plot indicated that gene nucleotide composition was clearly contributed to codon bias. CDS length was also key factor in dictating codon usage variation. A general tendency of more biased codon usage of genes with longer CDS length to higher expression level was found. Further, the hydrophobicity of each protein also played a role in shaping codon usage in this organism, which could be confirmed by the significant correlation between the positions of genes placed on the first axis and the hydrophobicity values (r=-0.100, P<0.01). In summary, gene expression level played a crucial role, nucleotide mutational bias, CDS length and the hydrophobicity of each protein just in a minor way in shaping the codon usage pattern of D. radiodurans. Notably, 19 codons firstly defined as "optimal codons" may provide useful clues for molecular genetic engineering and evolutionary studying.     

Revisiting the codon adaptation index from a whole-genome perspective: analyzing the relationship between gene expression and codon occurrence in yeast using a variety of models

Highly expressed genes in many bacteria and small eukaryotes often have a strong compositional bias, in terms of codon usage. Two widely used numerical indices, the codon adaptation index (CAI) and the codon usage, use this bias to predict the expression level of genes. When these indices were first introduced, they were based on fairly simple assumptions about which genes are most highly expressed: the CAI was originally based on the codon composition of a set of only 24 highly expressed genes, and the codon usage on assumptions about which functional classes of genes are highly expressed in fast-growing bacteria. Given the recent advent of genome-wide expression data, we should be able to improve on these assumptions. Here, we measure, in yeast, the degree to which consideration of the current genome-wide expression data sets improves the performance of both numerical indices. Indeed, we find that by changing the parameterization of each model its correlation with actual expression levels can be somewhat improved, although both indices are fairly insensitive to the exact way they are parameterized. This insensitivity indicates a consistent codon bias amongst highly expressed genes. We also attempt direct linear regression of codon composition against genome-wide expression levels (and protein abundance data). This has some similarity with the CAI formalism and yields an alternative model for the prediction of expression levels based on the coding sequences of genes. More information is available at http://bioinfo.mbb.yale.edu/expression/codons.     

Unexpected correlations between gene expression and codon usage bias from microarray data for the whole Escherichia coli K-12 genome

Escherichia coli has long been regarded as a model organism in the study of codon usage bias (CUB). However, most studies in this organism regarding this topic have been computational or, when experimental, restricted to small datasets; particularly poor attention has been given to genes with low CUB. In this work, correspondence analysis on codon usage is used to classify E.coli genes into three groups, and the relationship between them and expression levels from microarray experiments is studied. These groups are: group 1, highly biased genes; group 2, moderately biased genes; and group 3, AT-rich genes with low CUB. It is shown that, surprisingly, there is a negative correlation between codon bias and expression levels for group 3 genes, i.e. genes with extremely low codon adaptation index (CAI) values are highly expressed, while group 2 show the lowest average expression levels and group 1 show the usual expected positive correlation between CAI and expression. This trend is maintained over all functional gene groups, seeming to contradict the E.coli-yeast paradigm on CUB. It is argued that these findings are still compatible with the mutation-selection balance hypothesis of codon usage and that E.coli genes form a dynamic system shaped by these factors.     

Codon usages in different gene classes of the Escherichia coli genome

A new measure for assessing codon bias of one group of genes with respect to a second group of genes is introduced. In this formulation, codon bias correlations for Escherichia coli genes are evaluated for level of expression, for contrasts along genes, for genes in different 200 kb (or longer) contigs around the genome, for effects of gene size, for variation over different function classes, for codon bias in relation to possible lateral transfer and for dicodon bias for some gene classes. Among the function classes, codon biases of ribosomal proteins are the most deviant from the codon frequencies of the average E. coli gene. Other classes of ‘highly expressed genes’ (e.g. amino acyl tRNA synthetases, chaperonins, modification genes essential to translation activities) show less extreme codon biases. Consistently for genes with experimentally determined expression rates in the exponential growth phase, those of highest molar abundances are more deviant from the average gene codon frequencies and are more similar in codon frequencies to the average ribosomal protein gene. Independent of gene size, the codon biases in the 5′ third of genes deviate by more than a factor of two from those in the middle and 3′ thirds. In this context, there appear to be conflicting selection pressures imposed by the constraints of ribosomal binding, or more generally the early phase of protein synthesis (about the first 50 codons) may be more biased than the complete nascent polypeptide. In partitioning the E. coli genome into 10 equal lengths, pronounced differences in codon site 3 G+C frequencies accumulate. Genes near to oriC have 5% greater codon site 3 G+C frequencies than do genes from the ter region. This difference also is observed between small (100–300 codons) and large (>800 codons) genes. This result contrasts with that for eukaryotic genomes (including human, Caenorhabditis elegans and yeast) where long genes tend to have site 3 more AT rich than short genes. Many of the above results are special for E. coli genes and do not apply to genes of most bacterial genomes. A gene is defined as alien (possibly horizontally transferred) if its codon bias relative to the average gene exceeds a high threshold and the codon bias relative to ribosomal proteins is also appropriately high. These are identified, including four clusters (operons). The bulk of these genes have no known function.     

Local absence of secondary structure permits translation of mRNAs that lack ribosome-binding sites

The initiation of translation is a fundamental and highly regulated process in gene expression. Translation initiation in prokaryotic systems usually requires interaction between the ribosome and an mRNA sequence upstream of the initiation codon, the so-called ribosome-binding site (Shine-Dalgarno sequence). However, a large number of genes do not possess Shine-Dalgarno sequences, and it is unknown how start codon recognition occurs in these mRNAs. We have performed genome-wide searches in various groups of prokaryotes in order to identify sequence elements and/or RNA secondary structural motifs that could mediate translation initiation in mRNAs lacking Shine-Dalgarno sequences. We find that mRNAs without a Shine-Dalgarno sequence are generally less structured in their translation initiation region and show a minimum of mRNA folding at the start codon. Using reporter gene constructs in bacteria, we also provide experimental support for local RNA unfoldedness determining start codon recognition in Shine-Dalgarno--independent translation. Consistent with this, we show that AUG start codons reside in single-stranded regions, whereas internal AUG codons are usually in structured regions of the mRNA. Taken together, our bioinformatics analyses and experimental data suggest that local absence of RNA secondary structure is necessary and sufficient to initiate Shine-Dalgarno--independent translation. Thus, our results provide a plausible mechanism for how the correct translation initiation site is recognized in the absence of a ribosome-binding site.     

The effect of codon usage on the oligonucleotide composition of the E. coli genome and identification of over- and underrepresented sequences by Markov chain analysis.

As shown in the accompanying paper (5), the oligonucleotide composition of the E. coli genome is highly asymmetric for sequences up to 6 bp in length when ranked from highest to lowest abundance. We show here that this largely reflects codon usage because heavily used codons were found in the highly abundant oligomers whereas rarely used codons, with some exceptions, occurred in sequences in low abundance. Furthermore, linear regression analysis revealed a strong correlation between the frequencies of each trinucleotide and its usage as a codon. Dinucleotides are also not randomly distributed across each codon position and the dinucleotide composition of genes that are transcribed but not translated (rRNA and tRNA genes) was highly related to that seen in genes encoding polypeptides. However, 45 tetra-, 8 penta-, and 6 hexanucleotides were significantly over- or underabundant by Markov chain analysis and could not be accounted for by codon usage. Of these underrepresented sequences, many were palindromes, including the Dam methylation site.