Chemicals are going green as today’s factories move away from harsh substances and resource-heavy methods to manufacture everyday products. Enzymes like galactose oxidases (GOases) are powering this transition with their ability to efficiently turbocharge industrially- useful chemical reactions.
“As biocatalysts, GOases can help oxidise certain alcohols into vital intermediates for many chemical and pharmaceutical products,” said Ee Lui Ang, Department Head (Strain Engineering) at the A*Star Singapore Institute of Food and Biotechnology Innovation (A*Star SIFBI). “While chemical oxidation reactions normally involve high temperatures, potent oxidants and toxic byproducts, GOases allow the reactions to take place under mild atmospheric conditions using benign reagents, making them relatively greener.”
However, naturally-occurring GOases have their practical limits: they’re well-adapted to processing sugars and other small molecules, but they struggle to crack more complex molecules under industrial conditions, which restricts their usefulness when making medicines and agrochemicals.
Ang set out to discover GOases suited to these challenging molecules, working with A*STAR colleagues from the A*Star Bioinformatics Institute (A*Star BII) and the A*Star Institute of Sustainability for Chemicals, Energy and Environment (A*Star ISCE²). These included Wan Lin Yeo, SIFBI Lead Research Officer; Dillon Tay, ISCE2 Senior Scientist; Hao Fan, BII Senior Principal Investigator; Sebastian Maurer-Stroh, BII Executive Director; and Yee Hwee Lim, ISCE2 Director (Chemical Biotechnology and Catalysis).
The team’s approach was twofold: using directed evolution, they sped up the natural selection process by generating comprehensive libraries of GOases with varying mutations, then screening them for the most promising candidates. Tapping into resources from Singapore’s National Supercomputing Centre, they also developed computational models to predict which mutations would likely affect GOase performance in manufacturing processes.
“Directed evolution has been incredibly successful at developing industrial enzymes, but it’s still labour and time-intensive; the odds of finding the right enzyme can be one-in-tens of thousands,” said Ang. “We wanted to significantly accelerate this process through improved predictions that would help us design smarter enzyme libraries and find the hits more efficiently.”
Ang noted that most enzyme engineering projects focus on beneficial mutations, discarding more than 99 percent of the remaining data. However, the team’s model took neutral and negative mutations into account. “We used the full set of wet lab data from our high-throughput screening—good and bad mutations included—to develop in silico models that can more accurately predict GOase activity and stability," said Ang.
Their approach generated over 80 new, improved GOase variants able to handle commonly-used bulky benzylic and alkyl secondary alcohols in pharmaceutical production. The team reported that some of these mutants showed up to 2,400-fold increased activity over existing versions.
Combined with existing computational tools like YASARA and FoldX, the team’s models could also accelerate enzyme engineering for other industrial processes. With patents pending and supported by industry partners, Ang and colleagues are refining their models for other enzyme types.
“We are also transferring our combined experimental and modelling workflow to other enzyme classes of interest to further refine the method and validate its generalisability,” Ang added.
The A*STAR-affiliated researchers contributing to this research are from the A*Star Singapore Institute of Food and Biotechnology Innovation (A*Star SIFBI), A*Star Bioinformatics Institute (A*Star BII) and A*Star Institute of Sustainability for Chemicals, Energy and Environment (A*Star ISCE²).
