We here report a novel PET hydrolase originating from a soil microbial genome sequence. This enzyme, bbPET0069, exhibits characteristics resembling a cutinase-like Type I PET-degrading enzyme but lacks disulfide bonds. Notably, bbPET0069 displayed remarkable synergy withCandida antarcticalipase B (CALB), demonstrating rapid and efficient PET degradation. To improve the PET degradation activity of bbPET0069, we employed a three-dimensional (3D) structural modeling to identify mutation sites around its substrate binding domain combined with a protein language model for effective mutation prediction. Through three initial rounds of directed evolution, we achieved a significant enhancement in PET degradation with CALB, resulting in a 12.6-fold increase compared to wild-type bbPET0069 without CALB. We confirmed its PET degradation activity in PET nanoparticles and films, and our proposed approach enabled efficient PET degradation to terephthalic acid monomers up to 95.5%. Our approach, which integrates a two-enzyme system with protein engineering, demonstrates the potential for enhancing the activity of emerging PET-degradation enzymes, which may possess unique attributes.Graphical AbstractA novel PET hydrolase, bbPET0069, was identified from a soil microbial genome. bbPET0069 and CALB showed remarkable synergy in PET degradation. Using surface feature analysis, PET degradation activity of bbPET0069 was significantly improved. This combination of a two-enzyme system and surface feature analysis holds promise for enhancing emerging PET-degradation enzymes.