Introduction


This modeling work establishes how PETase from Ideonella sakaiensis interacts with PET oligomers while translating those insights into design choices for construct composition and validation; specifically, the IsPETase-Amy3S Coding Part Codon Optimised for Nicotiana benthamiana (BBa_252Q5ZWV).

🎯 Modeling Goals

  1. Understand PETase–PET Affinity: Determine whether PETase binds PET 1-mer, 2-mer, and 3-mer with sufficient affinity and favourable interaction profiles to justify the project’s enzyme choice and secretion signal strategy.
  2. Translate Data into Design: Convert docking-derived trends (binding affinity, hydrophobic contacts, hydrogen bonds) into practical design implications and planned bench validation steps.
  3. Ensure Scientific Rigor: Document assumptions, limitations, and next steps to meet iGEM expectations that a model either guides design or simulates behaviour used in decision-making.

Scope


The modelled construct is based on our simple part featuring PETase from I. sakaiensis and α-amylase 3 signal peptides (BBa_252Q5ZWV), which perform extracellular excretion. By testing its feasibility, we reduce costly lab cycles of cloning, transformation, and plant expression.

Methods Summary


Molecular docking was performed for PET monomer, dimer, and trimer against PETase to extract binding affinity measured in kilocalories per mole (kcal/mol) as well as hydrophobicity and hydrogen-bond maps. In addition, the docking simulation also presents residue-level interaction maps which annotate van der Waals contacts, pi stacking, carbon-hydrogen bonds, conventional hydrogen bonds, and alkyl-pi-donor interactions for each oligomeric substrate bound in the active-site cavity.

The outputs of our model experiment include numeric affinities and qualitative visualisations for interaction patterns and surface hydrophobicity, forming the foundation for our downstream engineering use.

Key Results: Binding the Plastic Challenge


As we explored the molecular dance between Ideonella sakaiensis PETase and PET oligomers, our docking simulations uncovered how this enzyme grips—or gradually releases—its plastic partners.

Table 1. PETase binding affinity with PET oligomers
Polymer Binding Affinity (kcal/mol)
Mono -5.3
Di -5.0
Tri -4.6

Binding affinities (kcal/mol) decreased in magnitude with oligomer size: mono −5.3, di −5.0, tri −4.6, indicating progressively looser binding as polymer length increases. This decrease is likely due to the increase in substrate size.

Visual Insights

The subsequent graphics depict our PETase-PET monomer docking simulation.

Monomer-CPX Interaction Map

Figure 1. Monomer-CPX

PETase-PET hydrogen bond map

Figure 2. Monomer-CPX-Hydrogen Bond.

PETase-PET hydrophobicity surface

Figure 3. Monomer-CPX-Hydrophobicity

The subsequent graphics depict our PETase-PET dimer docking simulation.

Dimer-CPX Interaction Map

Figure 1. Dimer-CPX

PETase-PET hydrogen bond map

Figure 2. Dimer-CPX-Hydrogen Bond

PETase-PET hydrophobicity surface

Figure 3. Dimer-CPX-Hydrophobicity

Finally, the following graphics depict our PETase-PET trimer docking simulation.

Dimer-CPX Interaction Map

Figure 1. Trimer-CPX

PETase-PET hydrogen bond map

Figure 2. Trimer-CPX-Hydrogen Bond

PETase-PET hydrophobicity surface

Figure 3. Trimer-CPX-Hydrophobicity

Qualitative Analysis

PET oligomers consistently docked into the PETase active site with benzene rings stacking and aligning to more hydrophobic regions, suggesting dominant Van der Waals contributions to stabilisation.

In addition, hydrogen bonding increases in quantity alongside oligomer size. However, this increase did not offset the decline in overall affinity, consistent with steric fit limits in the catalytic pocket for longer stacks.

Interpretation: The modelling analyses indicates that prioritising cleavage pathways yielding shorter PET fragments are advantageous, as PETase demonstrates higher catalytic affinity for monomers and dimers compared to trimers within the modelled binding pocket. This finding therefore supports the use of α‑amylase 3 signal peptides, since extracellular secretion increases exposure to PET fragments where short oligomers dominate, consistent with the hydrophobic binding trends. Together with our earlier observation that PET is primarily located in the extracellular domain, these results led to the incorporation of BBa_252Q5ZWV (the modelled construct which contains both PETase and α‑amylase 3 signal peptides) into all constructs used in our plant transformation experiments.

Despite results which contain detailed statistics and visual representations, docking treats substrates and enzymes largely as rigid or semi-flexible structures, which can under-represent conformational dynamics and PET crystallinity effects. Additionally, solvent and long-timescale rearrangements are only roughly modelled and not fully resolved. To fully understand turnover behaviour, further work needs to be done with molecular dynamics or kinetic modelling.

With that said, the conclusions of this experiment are positioned as directional guidance for engineering rather than definitive activity predictions.

As resources permit, we may apply restrained molecular dynamics to sample induced fit and solvent effects, improving confidence intervals around binding and orientation hypotheses.

To perform molecular docking simulation, the software PyRx is used. Specifically, the AutoDock Vina developed by Trott and Olson (2010) within PyRx performed the docking simulations.

To run PyRx, BBa_252Q5ZWV is converted to three-dimensional protein files via AlphaFold (Jumper et al., 2021).

Lastly, docking data are interpreted with the help of Discovery Studio Visualiser (Dassault Systèmes, 2025).

All Protein Data Bank files are accessible through this link.


References:
  • Trott, O., & Olson, A. J. (2010). AutoDock Vina: Improving the Speed and Accuracy of Docking with a New Scoring Function, Efficient Optimization, and Multithreading. Journal of Computational Chemistry, 31(2), 455–461. https://doi.org/10.1002/jcc.21334
  • Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., … Hassabis, D. (2021). Highly accurate protein structure prediction with AlphaFold. Nature, 596(7873), 583–589.
  • Dassault Systèmes. (2025). BIOVIA Discovery Studio Visualiser.