Invasive fungal infections kill on a scale that rivals the better-known bacterial superbugs, and the mainstay drugs, amphotericin and nystatin, are decades old, toxic and increasingly challenged by resistance. Scientists at Imperial College London and the University of Manchester have now discovered new members of the polyene antifungal family and, just as importantly, a cleaner biological route to improving them, reported in Nature.
Using genome mining and a bioinformatics pipeline, the team identified bacteria capable of producing polyenes structurally distinct from existing antifungals, solved their structures by nuclear magnetic resonance and characterised the enzymes responsible. They then exploited enzymatic glycosylation and amidation to build a library of derivatives, several of which showed greater antifungal activity, lower toxicity and better solubility than the parent drugs.
The lead compound, Nys34, reduced fungal burden in a mouse model of invasive aspergillosis without substantial signs of toxicity, and appears to act differently from amphotericin, which could make it valuable against amphotericin-resistant pathogens. Because the enzyme-based process avoids lengthy multistep chemical synthesis, it may also prove cleaner and more scalable, a practical consideration that often decides whether promising antimicrobials ever reach patients.