Cloning the enterotoxin gene from Clostridium perfringens type A
A C. perfringens type A genomic library was constructed in E. coli by banking overlapping 6-10 kbp Hind III fragments of chromosomal DNA from the enterotoxin (CPE) positive strain NCTC 8239 into the pUC derived vector pHG165. The library was screened by colony hybridization with a degenerate 26 bp oligonucleotide probe, derived from the amino acid sequence CPE9_17A. complex mixture of plasmid DNA was isolated from the only hybridization positive clone. A second round of screening picked out a single plasmid, with an apparently altered copy number, pLWl, that carried the CPE gene, cpe, on a 6.8 kbp insert. A sequence deduced primer strategy for direct plasmid sequencing was initiated using a primer deduced in a similar manner to the 26 bp probe, obviating the need for prior mapping and subcloning of the insert. The amino acid sequence for the conceptual gene product of the single open reading frame differed only slightly from the known CPE sequence but lacked the C terminal residues. The biased cpe codon usage reflected the low %G+C content of the DNA. The %G+C content was even lower in the upstream region and possessed properties characteristic of bent DNA. The region 5' to the ATG translational start codon contained a Shine-Dalgarno sequence and several sequences with significant homology to the putative transcriptional control regions for the tetanus toxin gene. The N-terminal coding region contained a direct repeat of an upstream sequence that shared considerable homologies with the crossover point in site 1 of the Tn3 res region. Southern blot analyses of chromosomal and plasmid DNAs from several isolates indicated that the majority of strains were cpe-. The chromosomal location and architecture of cpe appeared identical in all cpe+ strains. A second copy, pLW2, of the 5' end of cpe, on a 4.5 kbp Pst I/Eco RI restriction fragment, was cloned during one of many unsuccessful attempts to clone the 3' end. A separate re-cloning experiment isolated several different clones that contained the 0.6 kbp Hind III located = 2.5 kbp 5' to the ATG codon of both cloned copies of cpe but none of them carried the CPE gene. The fragment was used as a DNA probe to show that it was present in high copy number in some strains of C. perfringens but completely absent from others. An hypothesis describing the possible involvement of a mobile genetic element in C. perfringens enterotoxin production offers explanations for the cloning of a complex mixture of plasmids, the apparent alteration in plasmid copy number, the identification of putative DNA crossover points, the failure to clone the 3' end of cpe and the isolation of a novel DNA fragment.