Counter-intuitive enhancement of degradation of polyethylene terephthalate through engineering of lowered enzyme binding to solid plastic.

Degradation of solid polyethylene terephthalate (PET) by leaf branch compost cutinase (LCC) produces various PET-derived degradation intermediates (DIs), in addition to terephthalic acid (TPA), which is the recyclable terminal product of all PET degradation. Although DIs can also be converted into TPA, in solution, by LCC, the TPA that is obtained through enzymatic degradation of PET, in practice, is always contaminated by DIs. Here, we demonstrate that the primary reason for non-degradation of DIs into TPA in solution is the efficient binding of LCC onto the surface of solid PET. Although such binding enhances the degradation of solid PET, it depletes the surrounding solution of enzyme that could otherwise have converted DIs into TPA. To retain a subpopulation of enzyme in solution that would mainly degrade DIs, we introduced mutations to reduce the hydrophobicity of areas surrounding LCC's active site, with the express intention of reducing LCC's binding to solid PET. Despite the consequent reduction in invasion and degradation of solid PET, overall levels of production of TPA were ~3.6-fold higher, due to the partitioning of enzyme between solid PET and the surrounding solution, and the consequent heightened production of TPA from DIs. Further, synergy between such mutated LCC (F125L/F243I LCC) and wild-type LCC resulted in even higher yields, and TPA of nearly ~100% purity.

Rational mutagenesis of Thermobifida fusca cutinase to modulate the enzymatic degradation of polyethylene terephthalate.

Thermobifida fusca cutinase (TfCut2) is a carboxylesterase (CE) which degrades polyethylene terephthalate (PET) as well as its degradation intermediates [such as oligoethylene terephthalate (OET), or bis-/mono-hydroxyethyl terephthalate (BHET/MHET)] into terephthalic acid (TPA). Comparisons of the surfaces of certain CEs (including TfCut2) were combined with docking and molecular dynamics simulations involving 2HE-(MHET)3, a three-terephthalate OET, to support the rational design of 22 variants with potential for improved generation of TPA from PET, comprising 15 single mutants (D12L, E47F, G62A, L90A, L90F, H129W, W155F, ΔV164, A173C, H184A, H184S, F209S, F209I, F249A, and F249R), 6 double mutants [H129W/T136S, A173C/A206C, A173C/A210C, G62A/L90F, G62A/F209I, and G62A/F249R], and 1 triple mutant [G62A/F209I/F249R]. Of these, nine displayed no activity, three displayed decreased activity, three displayed comparable activity, and seven displayed increased (~1.3- to ~7.2-fold) activity against solid PET, while all variants displayed activity against BHET. Of the variants that displayed increased activity against PET, four displayed more activity than G62A, the most-active mutant of TfCut2 known till date. Of these four, three displayed even more activity than LCC (G62A/F209I, G62A/F249R, and G62A/F209I/F249R), a CE known to be ~5-fold more active than wild-type TfCut2. These improvements derived from changes in PET binding and not changes in catalytic efficiency.