BIONEWS
Imperial and Manchester researchers used bioinformatics and two enzyme edits to make soil bacteria compounds 3–8x less toxic than the standard antifungal amphotericin B in human cells.
Polyenes like amphotericin B and nystatin remain among the most effective agents for life-threatening fungal infections, but their use is limited by significant toxicity to human cells and poor solubility in blood. A UK team now reports a way to find new polyenes hidden in bacterial genomes and reshape them into compounds that sidestep both problems, with a lead candidate that ran three to eightfold less toxic to human cells than amphotericin B in lab tests, according to a writeup in Genetic Engineering & Biotechnology News.
Researchers at Imperial College London and the University of Manchester started with a bioinformatics sweep that asked which bacteria already carry the genes to build polyenes. The answer, the team reports, is many more than the handful of soil species, mostly Streptomyces, that produce amphotericin B and nystatin. Their survey turned up dozens of bacterial lineages with the biosynthetic machinery for polyenes that have not been described before, which suggests the chemistry class is far larger than the drugs on pharmacy shelves would imply.
The second half of the pipeline is targeted redesign. A glycosyltransferase adds a second sugar, which the standard polyenes lack, and an amidotransferase swaps a charged side group (a carboxylate) for a neutral amide. Together those edits shift the molecules away from the human cell membranes polyenes normally damage, while keeping the antifungal activity intact. Mirza and her colleagues built a derivative library, used NMR to confirm each new structure, then tested the compounds against human cell lines and a mouse model of invasive aspergillosis, a lung infection that is hard to clear in immunocompromised patients.
The lead candidate, called Nys34, came out of that pipeline. In lab tests across multiple human cell lines, it ran three to eightfold less toxic than amphotericin B. In mice infected with Aspergillus, Nys34 reduced the fungal burden compared with untreated animals. Mirza, a postdoctoral research associate at Manchester, framed the result as a way to scout and tune a chemistry class that has been treated as a fixed list for decades.
The data is still preclinical. Nys34 has been tested in cell lines and mice, not in people, and there is no disclosed timeline for human trials or for an investigational new drug filing. The breadth-of-producers claim, that many bacterial species already encode polyenes, has not yet been independently replicated, and only a small number of derivatives from the pipeline have been characterized so far. The underlying paper is "Enzymatic glycosylation and amidation reshapes polyene bioactivity" (Mirza et al.).
The antifungal pipeline has been thin for years, and resistance to existing drugs has been climbing. A method that turns bacterial genomes into a searchable library of new polyenes, and a pair of enzyme edits that reshape them into usable leads, gives chemists a way into the class that does not depend on waiting for the next soil screen to land a single lucky molecule. Nys34 is the proof of concept, not a treatment, and the authors have not disclosed when or whether the lead will move into formal preclinical development. The paper is the first place any of those derivatives have been characterized side by side.