Lab-evolved bacterial enzyme is up to 1,800 times more efficient

Making a drug or a food flavouring often means oxidising alcohols, a step chemists usually perform with harsh oxidants at high temperatures, which leaves toxic waste behind. Living organisms offer a gentler route. Fungi produce galactose oxidases, enzymes that carry out the same reactions on alcohols and simple sugars under mild conditions, using oxygen. Until now, the best-characterised versions came from fungi. A team at the Instituto de Tecnologia Química e Biológica António Xavier of NOVA University Lisbon (ITQB NOVA), in Portugal, has now described the first galactose oxidase found in bacteria, which it calls PsGalOx.

The enzyme was hard to work with for years. "For several years, the enzyme proved exceptionally difficult to produce in a soluble form and therefore to characterize," says Lígia Martins, who leads the Microbial & Enzyme Technology laboratory at ITQB NOVA. What changed was AlphaFold, an AI system whose structure predictions helped the team see the enzyme's unusual architecture and identify changes that could improve its production. The bacterial version carries two nearly identical sugar-binding modules next to its catalytic module, a feature the best-known fungal versions do not have. Tomás Frazão, a doctoral student and one of the paper's first authors, says the two copies, which share about 80% sequence identity, have distinct jobs: one enhances sugar binding, the other mainly contributes to protein stability. The team thereby challenges the common assumption that such modules are interchangeable.

The researchers then applied directed evolution, which recreates natural selection in the lab: they introduced mutations and screened thousands of bacterial colonies for better variants. The similar modules recombined into single hybrid modules, and these variants produced more soluble protein. The team also found that higher activity did not necessarily mean the enzyme had become a better catalyst; much of the improvement came from cells making more of it, and recognising that was key to further progress. The final variant is 40 to 1,800 times more efficient at processing a range of sugars and alcohols.

For chemists and manufacturers, the gain is a potentially cleaner way to make intermediates that feed food, drug and materials production, since the reaction runs under mild conditions with oxygen. Who benefits, and when, is not yet settled. The team plans to keep engineering PsGalOx to act on "substrates of practical interest", aiming at a versatile intermediate for food, pharmaceutical and materials applications, and it wants to map the evolutionary diversity of bacterial oxidases. No industrial partner and no commercial date have been announced.

The limits are clear. Every result comes from laboratory work, the 40 to 1,800 range spans many different substrates rather than one product, and the enzyme has not yet been reported in an industrial process. The next test is whether the same gains hold on the sugars and alcohols that manufacturers actually need.