Murein-metabolizing enzymes from Escherichia coli: sequence analysis and controlled overexpression of the slt gene, which encodes the soluble lytic transglycosylase

Abstract

The complete nucleotide sequence of the slt gene encoding the soluble lytic transglycosylase (Slt; EC 3.2.1.-) from Escherichia coli has been determined. The largest open reading frame identified on a 2.5-kb PvuII-SalI fragment indicates that the enzyme is translated as a preprotein of either 654 or 645 amino acids, depending on which of two potential start codons is used. The two possible translation products differ only in the lengths of their predicted signal peptides, 36 or 27 amino acids, respectively. In both cases, processing results in a soluble mature protein of 618 amino acids (Mr = 70,468). The deduced primary structure of the mature protein was confirmed by N-terminal sequencing and determination of the amino acid composition of the isolated transglycosylase. The slt gene contains a high percentage of rare codons, comparable to other low-expressed genes. A hairpin structure that could serve as a transcriptional terminator is located downstream of the slt coding region and precedes the trpR open reading frame at 99.7 min on the E. coli chromosomal map. A computer-assisted search did not reveal any significant sequence similarity to other known carbohydrate-degrading enzymes, including lysozymes. Interestingly, a stretch of 151 amino acids at the C terminus of the transglycosylase shows similarity to the N-terminal portion of the internal virion protein D from bacteriophage T7. Overexpression of the slt gene, under the control of the temperature-inducible phage lambda pR promoter, results in a 250-fold overproduction of the mature transglycosylase, whereas after deletion of the signal peptide a 100-fold overproduction of the enzyme is observed in the cytoplasm.