| Summary: | <p>Abstract</p> <p>Background</p> <p>Mycolactones are immunosuppressive and cytotoxic polyketides, comprising five naturally occurring structural variants (named A/B, C, D, E and F), produced by different species of very closely related mycobacteria including the human pathogen, <it>Mycobacterium ulcerans</it>. In <it>M. ulcerans </it>strain Agy99, mycolactone A/B is produced by three highly homologous type I polyketide megasynthases (PKS), whose genes (<it>mlsA1</it>: 51 kb, <it>mlsA2</it>: 7.2 kb and <it>mlsB</it>: 42 kb) are found on a 174 kb plasmid, known as pMUM001.</p> <p>Results</p> <p>We report here comparative genomic analysis of pMUM001, the complete DNA sequence of a 190 kb megaplasmid (pMUM002) from <it>Mycobacterium liflandii </it>128FXT and partial sequence of two additional pMUM replicons, combined with liquid chromatography-tandem mass spectrometric (LC-MS/MS) analysis. These data reveal how PKS module and domain differences affecting MlsB correlate with the production of mycolactones E and F. For mycolactone E these differences from MlsB in <it>M. ulcerans </it>Agy99 include replacement of the AT domain of the loading module (acetate to propionate) and the absence of an entire extension module. For mycolactone F there is also a reduction of one extension module but also a swap of ketoreductase domains that explains the characteristic stereochemistry of the two terminal side-chain hydroxyls, an arrangement unique to mycolactone F</p> <p>Conclusion</p> <p>The mycolactone PKS locus on pMUM002 revealed the same large, three-gene structure and extraordinary pattern of near-identical PKS domain sequence repetition as observed in pMUM001 with greater than 98.5% nucleotide identity among domains of the same function. Intra- and inter-strain comparisons suggest that the extreme sequence homogeneity seen among the mls PKS genes is caused by frequent recombination-mediated domain replacement. This work has shed light on the evolution of mycolactone biosynthesis among an unusual group of mycobacteria and highlights the potential of the <it>mls </it>locus to become a toolbox for combinatorial PKS biochemistry.</p>
|
|---|