How Changes in Anti-SD Sequences Would Affect SD Sequences in Escherichia coli and Bacillus subtilis

Abstract

The 3' end of the small ribosomal RNAs (ssu rRNA) in bacteria is directly involved in the selection and binding of mRNA transcripts during translation initiation via well-documented interactions between a Shine-Dalgarno (SD) sequence located upstream of the initiation codon and an anti-SD (aSD) sequence at the 3' end of the ssu rRNA. Consequently, the 3' end of ssu rRNA (3'TAIL) is strongly conserved among bacterial species because a change in the region may impact the translation of many protein-coding genes. Escherichia coli and Bacillus subtilis differ in their 3' ends of ssu rRNA, being GAUCACCUCCUUA3' in E. coli and GAUCACCUCCUUUCU3' or GAUCACCUCCUUUCUA3' in B. subtilis Such differences in 3'TAIL lead to species-specific SDs (designated SD_Ec for E. coli and SD_Bs for B. subtilis) that can form strong and well-positioned SD/aSD pairing in one species but not in the other. Selection mediated by the species-specific 3'TAIL is expected to favor SD_Bs against SD_Ec in B. subtilis, but favor SD_Ec against SD_Bs in E. coli Among well-positioned SDs, SD_Ec is used more in E. coli than in B. subtilis, and SD_Bs more in B. subtilis than in E. coli Highly expressed genes and genes of high translation efficiency tend to have longer SDs than lowly expressed genes and genes with low translation efficiency in both species, but more so in B. subtilis than in E. coli Both species overuse SDs matching the bolded part of the 3'TAIL shown above. The 3'TAIL difference contributes to the host specificity of phages.

Keywords: Bacillus subtilis; Escherichia coli; Shine-Dalgarno; anti-SD-sequence; ssu rRNA; translation efficiency.

Figure 1

A model of SD sequence and aSD interactions. (A) The free 3′ end of SSU rRNA (3′TAIL) of E. coli and B. subtilis based on the predicted secondary structure of the 3′ end of the ssu rRNA of E. coli and B. subtilis from mfold 3.1, adapted from the comparative RNA web site and project (http://www.rna.icmb.utexas.edu). (B) A schematic representation of SD and aSD interaction illustrates D_toStart as a better measure for quantifying the optimal positioning of SD and aSD than the conventional distance from putative SD to start codon. SD1 or SD2, as illustrated, are equally good in positioning the start codon AUG against the anticodon of the initiation tRNA, but they differ in their distances to the start codon. D_toStart is the same for the two SDs. (C, D) D_toStart is constrained to a narrow range in E. coli (C) and B. subtilis (D); solid blue line denotes SD hits with the UCU-ending TAIL, and the dashed red line shows SD hits with the UCUA-ending TAIL. The y-axis in (C) and (D) represents the percentage of SD motif hits detected. See Materials and Methods section for details.

Figure 2

Distribution of SDs from 200 HTE genes and 200 LTE genes over SD length for E. coli (A) and B. subtilis (B). Classifying genes into HEGs and LEGs generates equivalent results, with HEGs similar to HTE genes, and LEGs similar to LTE genes. HEGs and HTE genes tend to have longer SDs than LEGs and LTE genes.

Figure 3

Distribution of E. coli and B. subtilis SDs for HEGs and LEGs. SDs that are more frequent in HEGs than LEGs match the core aSD (in bold red) of 16S rRNA. The trailing 3′ nucleotides in B. subtilis are used mainly for SD/aSD pairing in LEGs. Classifying genes into genes of HTE and LTE generates similar results.