AlphaFold predictions helped ITQB NOVA engineer a bacterial galactose oxidase
ITQB NOVA researchers engineered PsGalOx, which they call the first bacterial galactose oxidase, into a variant 40 to 1,800 times more efficient. The team says much of the gain in directed evolution came from cells making more protein; drug substrates are still ahead.
The Scientist · Science desk
What happened
- Unlike the best-characterized fungal versions, the bacterial enzyme carries two nearly identical sugar-binding modules in addition to its catalytic module.
- Doctoral student Tomás Frazão says the two modules share about 80% sequence identity yet differ in function: one boosts sugar binding and the other mainly stabilizes the protein.
- In directed evolution, the team screened thousands of bacterial colonies, and the similar modules recombined into single hybrid modules that gave more soluble protein.
Why it matters
- Who has to decide decision Groups evolving hard-to-express enzymes have to measure protein amount next to activity, or they will credit catalysis for what expression delivered.
- What it sets up precedent Near-identical binding modules that split sugar binding from stability give engineers a counterexample to treating such modules as interchangeable.
- What's now possible capability A bacterial enzyme that variants now produce in larger soluble amounts gives the group a scaffold for its planned survey of bacterial oxidases.
The headline range hides a spread. A variant 40 to 1,800 times more efficient across a range of sugars and alcohols [9] spans a factor of 45 between the two ends of that range [14]. The release does not define "efficient", name the substrates at either end, or state the starting enzyme's baseline, so the figure cannot yet be set against what a process chemist would need.
The more portable result concerns what the screen measured. A colony that looks more active may hold a better catalyst or simply more enzyme. According to the researchers, much of the improvement during evolution was the second [8], and they say telling the two apart was key to improving the enzyme further [11]. In our view, that separation is the part other groups can reuse, whatever becomes of PsGalOx itself.
AlphaFold's role gets one sentence from Lígia Martins, who leads the Microbial & Enzyme Technology laboratory at ITQB NOVA [13]. "A major turning point came with the emergence of AlphaFold, an AI system whose structural predictions helped us understand the enzyme's unusual architecture and identify changes that could improve its production," Martins said [4]. That credits the predictions with understanding and with candidate changes. The hybrid sugar-binding modules came out of the directed evolution, which the release calls unexpected [7]. How much of the final gain came from changes the predictions suggested, and how much from evolved ones, the release leaves open. The paper's title names both module recombination and second-shell substitutions [10].
The case for milder chemistry rests on the enzyme class. Fungal galactose oxidases act on alcohols and simple sugars under mild conditions using oxygen [2]. The conventional route uses harsh oxidants and elevated temperatures and generates toxic waste [1]. Martins said the team's plan is to "continue engineering PsGalOx to act on substrates of practical interest, creating a versatile intermediate for food, pharmaceutical and materials applications" [15]. That plan does not say whether the engineered variant keeps its gains on a molecule a drug maker would actually oxidize.
The account here rests on the institute's release, carried by Phys.org [12]. The "first" claim is the team's own [3].
What to watch
- Whether the paper reports kinetic constants and a starting-variant baseline for each substrate behind the 40-to-1,800-fold range.
- Any PsGalOx variant tested on a pharmaceutical, flavor or fragrance substrate, as Martins says the team plans.
- Results from the planned survey of bacterial oxidases on how specificity evolved.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+15
- Incentives
- Not scored
- Confidence55
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Producing pharmaceuticals, food ingredients, and flavors and fragrances often requires oxidizing alcohols, and conventional chemistry typically relies on harsh oxidants and elevated temperatures and generates toxic waste.
- [2]
Galactose oxidases, produced by living organisms including fungi, can carry out alcohol oxidation on alcohols and simple sugars under mild conditions, using oxygen.
- [3]
ITQB NOVA researchers describe PsGalOx as the first galactose oxidase in bacteria.
- [4]
Martins said: "A major turning point came with the emergence of AlphaFold, an AI system whose structural predictions helped us understand the enzyme's unusual architecture and identify changes that could improve its production,"
- [5]
Unlike the best-characterized fungal galactose oxidases, the bacterial counterparts carry two nearly identical sugar-binding modules in addition to the catalytic module, according to the study published in Chem & Bio Engineering.
- [6]
Tomas Frazao, a doctoral student at ITQB NOVA and one of the paper's first authors, says the duplicated modules share approximately 80% sequence identity but have distinct functions: one enhances binding to sugars, while the other primarily contributes to protein stability.
- [7]
The team used directed evolution, introducing mutations and screening thousands of bacterial colonies; the evolution revealed something unexpected, as the highly similar sugar-binding modules recombined into single hybrid modules, and the resulting variants produced more soluble protein.
- [8]
The researchers realized that increased activity during evolution did not necessarily mean the enzyme had become better at speeding up reactions; much of the improvement resulted from cells producing more of it.
- [9]
The team developed a variant that was 40 to 1,800 times more efficient at processing a range of sugars and alcohols.
- [10]
The paper is Taborda et al., 'Engineering a Modular Bacterial Galactose Oxidase through CBM Recombination and Second-Shell Substitutions', Chem & Bio Engineering (2026), DOI 10.1021/cbe.6c00167.
- [11]
The researchers say the distinction between higher expression and better catalysis was key to further improving the enzyme.
- [12]
The Phys.org article is marked as provided by ITQB NOVA.
- [13]
Lígia Martins leads the Microbial & Enzyme Technology laboratory at ITQB NOVA.
- [14]
The upper end of the reported 40-to-1,800-fold range is 45 times the lower end.
- [15]
Martins said the team's plan is to "continue engineering PsGalOx to act on substrates of practical interest, creating a versatile intermediate for food, pharmaceutical and materials applications."
- [16]
The researchers want to explore the evolutionary diversity of bacterial oxidases, which Martins says will help them understand how specificity evolved and identify robust, versatile scaffolds for developing new biocatalysts.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- phys.orgFirst bacterial galactose oxidase expands toolkit for greener drug and food ingredient production
1 article · October 15, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- Protein structure predictionFollow
- Enzyme EngineeringFollow
- BiocatalysisFollow