Comprehensive analysis of stop codon usage in bacteria and its correlation with release factor abundance

Abstract

We present a comprehensive analysis of stop codon usage in bacteria by analyzing over eight million coding sequences of 4684 bacterial sequences. Using a newly developed program called "stop codon counter," the frequencies of the three classical stop codons TAA, TAG, and TGA were analyzed, and a publicly available stop codon database was built. Our analysis shows that with increasing genomic GC content the frequency of the TAA codon decreases and that of the TGA codon increases in a reciprocal manner. Interestingly, the release factor 1-specific codon TAG maintains a more or less uniform frequency (∼20%) irrespective of the GC content. The low abundance of TAG is also valid with respect to expression level of the genes ending with different stop codons. In contrast, the highly expressed genes predominantly end with TAA, ensuring termination with either of the two release factors. Using three model bacteria with different stop codon usage (Escherichia coli, Mycobacterium smegmatis, and Bacillus subtilis), we show that the frequency of TAG and TGA codons correlates well with the relative steady state amount of mRNA and protein for release factors RF1 and RF2 during exponential growth. Furthermore, using available microarray data for gene expression, we show that in both fast growing and contrasting biofilm formation conditions, the relative level of RF1 is nicely correlated with the expression level of the genes ending with TAG.

Keywords: Bacteria; Codon Usage; Protein Synthesis; Ribosome; Stop Codon; Translation; Translation Release Factor.

FIGURE 1.

Stop codon distribution in different bacterial phyla. The frequencies of three stop codons TAA (green), TAG (black), and TGA (red) in different bacteria classified and grouped in different phyla (A–E) and within the class of Mollicutes (F) are plotted as a function of genomic GC content.

FIGURE 2.

Distribution of the three canonical stop codons in bacteria. A, the frequency of the three stop codons estimated using all ORFs in 4684 bacterial sequences (chromosomes and plasmids) plotted as a function of genomic GC content. B, distribution of stop codon frequency in highly expressed genes (comprising ribosomal protein and translation factor genes) of 220 randomly selected bacteria.

FIGURE 3.

Occurrence of downstream additional stop codons. The occurrence of additional stop codons within 15 nucleotides downstream from the primary stop codons TAA, TAG, and TGA is shown. All ORFs in E. coli (A), B. subtilis (B), and M. smegmatis (C) were analyzed. The horizontal bars represent percentages of the total genes with additional stop codons TAG (black), TGA (red), and TAA (green). Freq., frequency.

FIGURE 4.

Expression analysis and correlation of release factors and stop codon frequency. A, correlation of the stop codon frequency (TGA/TAG) with mRNA transcripts determined with qPCR (prfB/prfA) and the amount of the release factor proteins (RF2/RF1) estimated by quantitative Western blotting using antibodies against E. coli RF1 and RF2. B, Western blotting-based determination of the amount of RFs in E. coli growing exponentially (E) and in stationary (S) phase in rich (LB) and minimal (M9) media. The bars represent the amount of the release factors in pmol estimated from the blots (presented above) done in quadruplicates. C, E. coli genes ending with TAG/TGA/TAA are grouped, and the mean expression level under fast growing condition is plotted according to the linear expression value of the microarray. TAA w/o highly are the genes ending with TAA without the highly expressed genes (yellow). D, same as C but showing expression data under biofilm-forming conditions. E, classification of the genes ending with TAG/TGA/TAA according to the biological functions.