Landscape profiling of PET depolymerases using a natural sequence cluster framework

Abstract

Enzymes capable of breaking down polymers have been identified from natural sources and developed for industrial use in plastic recycling. However, there are many potential starting points for enzyme optimization that remain unexplored. We generated a landscape of 170 lineages of 1894 polyethylene terephthalate depolymerase (PETase) candidates and performed profiling using sampling approaches with features associated with PET-degrading capabilities. We identified three promising yet unexplored PETase lineages and two potent PETases, Mipa-P and Kubu-P. An engineered variant of Kubu-P outperformed benchmarks in terms of PET depolymerization in harsh environments, such as those with high substrate load and ethylene glycol as the solvent.