The discovery
In 2016, Yoshida et al. published a landmark paper in Science describing Ideonella sakaiensis 201-F6 — a bacterium isolated from a PET-contaminated recycling site in Sakai, Japan, that could grow using PET as its primary carbon and energy source.
The bacterium secretes two enzymes working in sequence: PETase hydrolyses PET into MHET (mono(2-hydroxyethyl) terephthalic acid), and MHETase further breaks MHET into terephthalic acid (TPA) and ethylene glycol — the two monomers from which PET is synthesised. Those monomers can, in principle, be recovered and repolymerised to make virgin-quality PET.
Enzymes catalysing specific hydrolysis reactions are covered in our enzyme glossary entry — PETase is a particularly striking example of how the active site shape determines substrate specificity.
Engineering for speed
Wild-type PETase is far too slow for industrial application. Two engineering efforts dramatically improved it:
- 2018 — Portsmouth/NREL mutant: A structural study of PETase's active site led researchers to create a mutant with a wider active site cleft, improving activity by ~20%. Notably, the improved mutant was found accidentally while attempting to understand the natural enzyme.
- 2022 — FAST-PETase (UT Austin): Machine learning was used to predict five mutations that together increased activity dramatically. FAST-PETase can depolymerise a plastic bottle to >95% monomers within 24 hours at 50°C — compared to weeks for the wild-type enzyme.
For how machine learning is transforming chemistry and materials science more broadly, see our quantum computing and chemistry article.
Industrial reality
French biotechnology company Carbios has built a pilot-scale enzymatic recycling plant and reported depolymerisation of 97% of PET within 10 hours using their optimised enzyme system at 72°C. They completed a 40,000-litre demonstration reactor in 2021.
The recovered TPA is chemically identical to petroleum-derived TPA and can be used without downgrading — a significant advantage over mechanical recycling, where degraded polymer chains reduce quality with each cycle.
However, PET represents only ~8% of global plastic production. The vast majority of plastic waste — polyethylene, polypropylene, polystyrene — lacks efficient enzymatic routes. Research into enzymes for these plastics is at an early stage.