The nucleotide sequence of the cloned tufA gene of Escherichia coli.

The 4 kb (8.5 % lambda units) EcoRI fragment harboring the tufA gene of Escherichia coli was cloned using plasmid pTUA1 (Shibuya et al., 1979) and its structure was analyzed. The nucleotide sequence of about 1500 base pairs, covering the C-terminal portion of elongation factor EF-G (fus gene), the intercistronic region between fus and tufA, the entire structural gene for tufA with the GUG initiation and UAA termination codons, and the 3' flanking region of tufA, was determined. Comparison of the tufA nucleotide sequence with the tufB sequence (An and Friesen, 1980) and the known amino acid sequence of EF-Tu (Arai et al., 1980) revealed that the products of genes tufA and tufB are identical except for one amino acid at the C-terminal, i.e., glycine for tufA and serine for tufB. Nucleotide differences between tufA and tufB were found at 13 positions. Among them, one in the initiation codon and the other one in the C-terminal amino acid codon had replacements at the first letter of the codons. The other eleven changes were in the third codon positions, which did not affect the amino acid coding. The pattern of codon usage in tufA and tufB is highly nonrandom, and remarkably similar to that in ribosomal protein genes, with the codons for the most abundant species of isoaccepting tRNAs being preferentially utilized (Post et al., 1979; Post and Nomura, 1980).