Figure 1: Molecular docking and molecular dynamics simulations of Tf C and 2HE-(MHET)3. (A) Surface representation of Tf C docked with 2HE-(MHET)3, showing catalytic triad residues in orange and potential regions selected for altering PET binding affinity shown in pink. Residues inTf C participating in interaction with 2HE-(MHET)3obtained from (B) Molecular docking and (C) Molecular dynamics simulations performed for 75 ns.
3.2 Activity assessments of various Tf C mutants upon PET and BHET
Figure 2A compares activities of mutants and Tf C against commercially-sourced PET (film). G62A was ~1.7-fold more active than Tf C, corroborating its reported ~2.0-fold higher activity [Wei et al., 2016]. D12L, and E47F, designed to enhance surface hydrophobicity at regions distal to the substrate-binding site, showed activity levels similar to that ofTf C. W155F, H184S, H184A, and F209S, designed to create more space at or near PET binding sites, and H129W, designed to increase hydrophobicity (in imitation of PETase), showed drastic reductions in activity against PET, indicating that mutations in the active site impact catalysis negatively, through changes in binding site geometry that are not tolerated, despite conservation of overall enzyme fold. Importantly, a previous study showed higher (rather than lower) activity in H129W [Furukawa et al., 2019]. Since Tf C contains S136 as a vicinal residue of H129W, and S136 happened to be T136 in ourTf C clone, we mutated T136 to S136 to create H129W/T136S which, however, also showed lower (rather than higher) activity, in contradiction of the earlier report. We propose that H129 influences catalysis in a previously-unsuspected manner.
W155F and H184S were non-active against commercial-sourced PET films of high crystallinity (~250 microns thickness). Both mutants, however, showed activity against thin PET films of low crystallinity (~10 microns thickness), made through dissolution of PET granules (each of ~30 mg) and evaporation of the solvent (HFIP), on the walls of a micro-centrifuge tube. Interestingly, using such thin films, A173C showed ~40% improvement over Tf C. However, using commercially-sourced PET films, A173C was 30% less active thanTf C, suggesting differences in how these enzymes see PET backbones in films of different thickness and crystallinity. A173 is situated just below the ‘catalytic triad’ residues D176 and H208 (Supporting Information Figure S2B). The A173C mutation could be anticipated to decrease space in the binding cleft, facilitating better catalysis upon thin PET films containing accessible PET chains (Supporting Information Figure S4) but giving rise to poorer catalysis with crystalline or semi-crystalline (thicker) commercial PET films (Figure 2A) at the reaction temperature of 60 ˚C. A173C/A210C, and A173C/A206C, created to introduce a disulphide bond below the catalytic cleft (mimicking such a bond in PETase) led to loss of activity, suggesting distortion of Tf C’s binding site. Deletion of V164 in a loop connecting β-strands below the active site also led to loss of activity.
L90, situated distal to the binding site, was mutated to A90 (less hydrophobic), and also to F90 (more hydrophobic), analogous to F125 in LCC. L90A showed reduced activity upon PET (data not shown). In contrast, G62A/L90F showed a ~2.3-fold increase in activity, compared to Tf C. G62A/F209I (corresponding to F243I variant of LCC [Tournier et al., 2020]), showed a ~3.17-fold improvement. The variant G62A/F249R showed the highest hydrolytic potential against commercial-sourced PET films, with a ~3.5-fold improvement over Tf C. G62A/F209I and G62A/F249R are anticipated to show reduced PET binding, due to the replacement of phenylalanine by a less hydrophobic residue, or a charged residue. This may allow a higher fraction of the Tf C population to remain in solution and, therefore, to act upon degradation intermediates accumulating in solution, so as to function synergistically with the bound enzyme fraction of Tf C (invading and degrading solid PET), to generate TPA. This could explain the enhancement of activity in these two variants, through a reduction in non-specific PET binding, or faster dissociation from PET, without any alterations in the active site.
Most mutants showing loss of activity against commercial PET film also showed reduced catalytic degradation of BHET. In Figure 2B, H129W shows the least activity with BHET, with H129W/T136S showing improved degradation of BHET into TPA. L90A and A173C/A206C showed BHET hydrolysis comparable to Tf C. This suggests that the inability of these mutants to degrade PET owes to compromised PET-binding, and not to compromise hydrolysis of ester bonds. The G62A variant along with four other binding-altered variants made using G62A as background (i.e., G62A/L90F, G62A/F209I, G62A/F249A and G62A/F249R) showed ~1.6 to ~2.0-fold improvements in conversion of BHET into TPA. The mutants showed similar relative trends in their ability to hydrolyse BHET over 12 h (Figure 2B) or 4 h (Supporting Information Figure S5). F209I showed maximum enhancement of BHET conversion, ~2 fold higher than Tf C after 12 h of incubation.