Studies on engineering glycosylation of polyene macrolide antibiotics

Glycosylated polyene macrolides are primarily of interest because of their potent antifungal activity. This group includes amphotericin B, a broad-spectrum antibiotic used to treat life-threatening systemic mycoses. However, amphotericin B is notoriously toxic and generally only given as a drug of last resort. Synthetic chemistry and synthetic biology have been used in attempts to develop safer analogues. Polyenes are complex natural products synthesized by actinomycete bacteria. Hundreds of these compounds have been chemically characterized. Most consist of a polyketide scaffold modified with a single aminodeoxysugar, mycosamine. Chemical studies have shown that replacement of mycosamine with 2-deoxymycosamine or 2-epimycosamine can reduce the toxicity of amphotericin B. Addition of extra sugars can also improve water-solubility and pharmacological properties. This project aimed to develop glycosylation engineering further as a means of generating new polyene analogues. Several new fronts were opened up. The work can be summarized as having five related strands. (1) The extending glycosyltransferase NypY is known to add a hexosyl sugar to the mycosamine sugar of amphotericin B. A recombinant hexahistidine-tagged form of NypY was expressed in Streptomyces lividans and partially purified in a soluble form. This opens the way for further exploitation of this enzyme. (2) The naturally occurring aromatic heptaene 67-121C has mannose attached to the mycosamine sugar. A simple chemical method was used to convert this disaccharide to branched trisaccharide or tetrasaccharide chains. These new analogues are of interest as anti-Leishmania drug candidates. (3) Mining of actinomycete genome sequences uncovered nine silent biosynthetic gene clusters for polyenes that may contain additional sugars. Five of these clusters contained genes for biosynthesis and attachment of aminodeoxysugars related to vicenisamine and forosamine. These are of interest for further investigation because positive charges generally increase activity of membrane-active biocides. Pseudonocardia endophytica was found to contain a complete biosynthetic gene cluster for nystatin A3.

(4) Nystatin A3 is a degenerate heptaene that contains an L-digitoxose deoxysugar attached to C-35 of the macrolactone ring. The macrolactone of the pentaene selvamicin contains methyl-L-digitoxose at the corresponding position, C-27. Genes for putative digitoxose-specific glycosyl transferases PenSV and SelSV were introduced into the amphotericin-producer S. nodosus, along with genes for biosynthesis of dTDP-L-- digitoxose and the related sugar dTDP-L--mycarose. No new glycoanalogues of amphotericin B were obtained but refinement of this approach could lead to success in the future.

(5) The amphotericin polyketide synthase was engineered in an attempt to make a ring- contracted pentaene analogue that might act as a better acceptor for SelSV and other glycosyltransferases that modify C27 of pentaenes. This was partially successful in that it contributes new knowledge of how to engineer polyketide synthases to make new antibiotics.

All five of these areas can be taken further to advance the development of improved drugs.