Midecamycin A3

Josamycin and midecamycin are consisted of three groups of components with different ultraviolet maximum absorption wavelengths (λmax), which are 231 nm, 280 nm and 205 nm. The quantitative analysis of all these components is challengeable due to the absence of the respective reference substances. To address this problem, universal and reliable methods were developed using high performance liquid chromatography coupled with charged aerosol detector (HPLC-CAD) for the quantitative analysis of components in josamycin and midecamycin. The chromatographic conditions and CAD parameters setting were optimized. Subsequently, the components were identified using HPLC coupled with ion trap/time-of-flight mass spectrometry (IT/TOF MS). The developed methods were validated by assessing linearity, limit of quantitation (LOQ), accuracy, precision and robustness. Good separations were achieved for all components and the adjustment of the filter valve and power function value efficiently improved sensitivity. The developed methods were more comprehensive than current HPLC-UV method. The experimental results demonstrated good linearity with coefficients of determination (R2) greater than 0.999 in the range of 0.002-0.30 mg mL-1. The limits of detection (LOD) were ranging from 1.8 to 2.0 μg·mL-1. The intra-day and inter-day RSD values were less than 2.0 % (n = 6) and 5.6 % (n = 9) respectively. The recoveries were 95.0 %-124.0 % at the spiked concentration levels of 0.05 %, 0.50 %, 0.10 % and 2.5 % with relative standard deviations (RSDs, n = 3) lower than 2.0 %. Finally, the developed methods were successfully applied to the quantitative analysis of minor components and used main components (leucomycin A3 and midecamycin A1) as alternative reference substance of minor components. The overall results demonstrated that the HPLC-CAD was a good alternative for the quantitative analysis of multi-components in 16-membered macrolides.

Development of host microorganisms for heterologous expression of polyketide synthases (PKS) that possess the intrinsic capacity to overproduce polyketides with a broad spectrum of precursors supports the current demand for new tools to create novel chemical structures by combinatorial engineering of modular and other classes of PKS. Streptomyces fradiae is an ideal host for development of generic polyketide-overproducing strains because it contains three of the most common precursors--malonyl-CoA, methylmalonyl-CoA and ethylmalonyl-CoA--used by modular PKS, and is a host that is amenable to genetic manipulation. We have expanded the utility of an overproducing S. fradiae strain for engineered biosynthesis of polyketides by engineering a biosynthetic pathway for methoxymalonyl-ACP, a fourth precursor used by many 16-membered macrolide PKS. This was achieved by introducing a set of five genes, fkbG-K from Streptomyces hygroscopicus, putatively encoding the methoxymalonyl-ACP biosynthetic pathway, into the S. fradiae chromosome. Heterologous expression of the midecamycin PKS genes in this strain resulted in 1 g/l production of a midecamycin analog. These results confirm the ability to engineer unusual precursor pathways to support high levels of polyketide production, and validate the use of S. fradiae for overproduction of 16-membered macrolides derived from heterologous PKS that require a broad range of precursors.

A 3.2-kb DNA fragment of the carbomycin biosynthetic region including the 3-O-acyltransferase gene (acyA) from Streptomyces thermotolerans was sequenced, and four ORFs were found in the fragment. The second ORF, designated ORF-A, was transcribed in the opposite direction to the other three ORFs. The first ORF was identified as carA, a gene for carbomycin resistance. The amino acid sequence of ORF-A was homologous to proteins of the cytochrome P-450 family. Streptomyces lividans transformed with pCB20, in which ORF-A was subcloned, epoxidized carbomycin B at its C-12, 13 positions, thus producing carbomycin A. The third ORF, the amino acid sequence of which showed a homology to macrolide antibiotics O-acyltransferases was identified as acyA. The last ORF (ORF-B), which starts just 3 bp downstream from the TGA termination codon of acyA, was thought to be a carbomycin 4-O-methyltransferase gene, because the amino acid sequence deduced from ORF-B showed high homology to a putative midecamycin 4-O-methyltransferase encoded on mdmC.