Codon bias in Escherichia coli may modulate translation initiation.

Abstract

Codons with low corresponding intracellular tRNA concentrations are generally used less frequently (i .e. rare) compared to synonymous codons associated with higher tRNA concentrations 11 1. The differential use of codons with defined tRNA concentrations causes a given mRNA sequence to be translated at variable rates 121. Rare codons with extremely low tRNA concentrations may stall translation until an appropriate tRNA becomes available [31. Translational pauses caused by low tRNA availability may also influence the folding of a nascent polypeptide chain, and it has been proposed that in several proteins (e.g. Rabbit a and p globins, Cytochromes C and Yeast Pyruvate Kinase 131) translational pausing may allow discrete protein domains to fold and interact. Translational pausing of this type therefore complements the kinetics of a protein folding reaction. Translational pauses could also influence the targeting of exported polypeptides. In E . coli proteins are generally exported by interactions of a N-terminal signal sequence with the translocation machinery [4,5,6]. Translational pausing could therefore favour export in two ways. With proteins exported posttranslationally, an increase in time required for translation elongation would also increase the time a nascent polypeptide chain is exposed in the cytoplasm. Chaperone proteins 17) involved specifically in protein translocation (e.g. Sec B [6]) could then bind to the protein chain, increasing both the efficiency of export and the probability of the protein adopting a loose fold suitable for translocation. Proteins exported in a co-translational fashion could also benefit from translational pauses within approximately the first 100 codons. Pauses located around this region would expose the nascent signal sequence in the cytoplasm prior to pausing and this should allow i t to act independently of the un-translated protein. The export process would then be more efficient since direct interactions with the translocation machinery could ensue and the un-translated portion of the protein would not have to be maintained in a translationally competent state. This would be of particular importance in prokaryotic species where there is no evidence for elongation arrest by a Signal Recognition Particle (SRP). We therefore developed analytical software to examine rare codon usage & distribution in a sample containing 47 exported and 46 cytoplasmic polypeptides. The necessary coding regions of Escherichia coli were selected from the Linkage map of E . ccdi K12 181, and subsequently exuacted from the Gembl database using ‘GCG Navigator’ software. To locate potential pause sites we selected 8 rarely used codons 191, with minor concentrations of tRNA [ 11. These were CTA (Leu), ATA (Ile), ACA (Thr), CCT, CCC (Pro), CGG, AGA & AGG (Arg). If translational pausing was a widely used folding strategy, it would be reasonable to assume a longer polypeptide would require more pauses during translation to achieve this. Preliminary data demonstrated that frequency of rare ctdon use appears to be independent of polypeptide chain length. This does not prevent translational pauses regulating protein folding events, but we found no evidence to support this. In order to analyse the role translational pausing may play, detailed structural data on protein domain boundaries is required. Analysis of the distribution of rare The distribution of rare codons in the first 120 codons of 47 exported and 46 cytoplasmic polypeptides.