Toward Material-Specific Peptides From in Silico Guided Peptide Engineering to Target PET in the Presence of PS.

Synthetic polymers are indispensable but pose major recycling challenges, particularly in mixed plastic waste streams. Enzymatic recycling offers a promising route to specific polymer depolymerization into oligomers and monomers. This requires controlling molecular recognition at polymer interfaces. Material-binding peptides (MBPs) can enhance catalytic degradation by directing catalysts to specific polymer surfaces, yet strategies for engineering polymer specificity remain limited. Here, we introduce a computational-experimental workflow for designing polymer-selective MBPs. Molecular dynamics simulations identified aromatic tyrosine residues as key contributors to polystyrene (PS) binding in the MBP MacHis, while polyethylene terephthalate (PET) interactions were unaffected. Guided by these insights, targeted tyrosine-to-glutamate substitutions were introduced, yielding variants with an 85% reduction in PS binding while maintaining PET affinity. Steered molecular dynamics simulations confirmed the reduced PS interaction, predicting a comparable 69% decrease in binding free energy. This design principle generalized to the structurally distinct peptides LCI and Tachystatin A2, where analogous substitutions reduced PS binding by 67% or 74%, respectively. Practical application was shown by fusing peptide variants to the LCCICCG enzyme and characterizing PET degradation in mixed plastics. This workflow provides a strategy for engineering polymer-selective binding peptides and enables programmable molecular recognition at synthetic polymer interfaces.