Beyond Geometric Pre-Reaction States: A Bottleneck-Guided Framework for PET Hydrolase Engineering

Enzymatic depolymerization of poly(ethylene terephthalate) (PET) is a promising route toward closed-loop plastic recycling, yet the catalytic factors that become limiting once an enzyme already samples reaction-ready geometries remain unclear. Here, we compare two parent-variant pairs, FAST-PETase/FAST-Y61Q and LCC-ICCG/LCC-A2 (H218Y/N248D), using replicated molecular dynamics (MD) simulations, pre-reaction-state (PRS) analysis, noncovalent-interaction characterization, quantum-mechanical cluster calculations, and distortion/interaction activation-strain (DIAS) analysis. Y61Q remodels the second-shell environment and the conformational ensemble of the PET ethylene-glycol linker, thereby increasing the population of Bürgi–Dunitz compatible attack geometries. By contrast, LCC-A2 and LCC-ICCG exhibit similar populations of the two-coordinate geometric PRS, whereas LCC-A2 shows more favorable oxyanion-hole organization, Ser-His proton-transfer readiness, substrate polarization, and transition-state interactions, together with a lower calculated acylation barrier. These results support a bottleneck-guided view of PET-hydrolase engineering: when reaction-ready geometries are sparsely populated, mutations can act primarily by improving geometric preorganization; once such geometries are already abundant, further gains can arise from improving the chemical quality of the PRS ensemble. This framework distinguishes PRS quantity from PRS quality and provides a mechanistic basis for selecting stage-appropriate descriptors in PET-hydrolase design.