The bacteria Ideonella sakaiensis have enzymes that break down plastic, but their performance is poor. Zurich researchers are trying to improve its efficiency with a robotic selection system and artificial intelligence.
It was Japanese researchers who discovered this enzyme called PETase four years ago in the soil of a PET recycling facility in Osaka. But she’s working too slowly to tackle mountains of plastic.
Rebecca Buller’s team at the Zurich University of Applied Sciences (ZHAW) decided to apply the method of directed evolution – awarded the Nobel Prize for chemistry in 2018 – to increase its efficiency.
Zurich scientists have already selected several hundred variants of the enzyme in the laboratory by feeding them with microplastics and observing which are the fastest, ZHAW said in a press release on Wednesday.
The winners will now serve as the basis for a new evolutionary round which will this time be robotized. An automated platform will make several thousand variants of the enzyme per week, which would take ten times longer for a human.
The process will also be assisted by artificial intelligence algorithms responsible for making proposals for optimizing the structure of enzymes on the basis of data collected in the laboratory. This makes it possible, for example, to determine which amino acids are the most important for the degradation of PET.
The team thus hopes to create “super-enzymes” capable, for example, of purifying drinking water from microplastics. An application in PET recycling is also conceivable, according to Ms Buller. Researchers from ETH Zurich are also contributing to this work.