Characterization of Polyketide Ketoreductase Gene from Midecamycin Producing Strain
麦迪霉素产生菌酮基还原酶基因的研究

This paper presents the results about the expression of the polyketide ke-toreductase gene (PKG) from midecamycin producing strain (S. mycarofaciens 1748), gene localization and nucleotide sequences analysis of the PKG. A BamH I -BamHI 4.0kb fragment isolated from genomic library of midecamycin producing strain containing the act I -homologous DNA was inserted into E. coli-Strep-tomyces shuttle vector pWHM3. A recombinant plasmid pCBR4 was obtained and introduced into a 2-hydroxyaklavinone producer 5. galilaeus ATCC31671 that was a polyketide reductase gene deficient mutant. The transformant produced aklavinone according to TLC and HPLC analyses. The BamHI-BamHI 4.0 kb fragment was inserted into pWHM3 in the reverted orientation with pCB4 and a recombinant plasmid pCBR4 was obtained. Introduction pCBR4 into 5. galilaeus ATCC31671 also resulted in the production of aklavinone. Thus we demonstrated that this gene encode a polyketide ketoreductase which results in deoxygenation of 2-hydroxyaklavinoe in C-2 position and this gene has own promoter itself. Restriction analysis of pCB4 indicated that there was no sites for EcoRI, Hind III , but there were sites for Sail, SphI, Xho I , BssH II on the cloned gene fragment. The polyketide ketoreductase gene was located on a BssH I -BamHI 1. 3kb fragment from Southern hybrization result, using act I gene as a probe, and that was confirmed by gene expression experiment in S. galilaeus ATCC31671. The nucleotide sequences analysis showed that the BssH I -BamHI 1. 3kb fragment contained an open reading frame(ORF). The protein coding region was assumed to start with an ATG start codon, end at an TAG stop codon. There was a 5 nt (GGAGG) SD sequence at the upstream of proposed start codon. A presumptive promoter consisted of 7 nt AACCGGA at the-10 region and 5 nt TTCGA at the-35 region. The deduced amino acid sequences of the polyketide ketoreductase gene from midecamycin producer consisted of 260 aa residues. Its amion acid sequences were compared with act IB gene coding protein sequences. The percentage of similarity was 77.4 and percentage identity was 66. 7.