Dry Lab Overview

Explore our computational work, including enzyme modeling, docking, and mutagenesis.

Enzyme Selection

We chose Thermobifida fusca (TfCut2) for our project because it is one of the most well-studied enzymes in literature known to effectively degrade both PET and PBAT. We chose TfCut with a docking-affinity-based approach, choosing the strain with the most consistent results when reacted with PET and PBAT. For this we chose to use Autodock Vina. However, since PET and PBAT are long-chain polymers, docking their monomeric units does not accurately reflect real enzyme-substrate interactions. Therefore, we tested multiple ligand models—including PET dimers, called PETduo and trimers, called PETtri (Beijing_United, 2022) to better represent the true structural complexity of these plastics. For PBAT, since there are no prior iGEM docking models, we designed and tested a range of full-length PBAT fragment models. Through performing cross-docking simulations of these ligands with various plastic-degrading enzymes, we found that TfCut2-KW3, which is known to grow at high temperatures. This consistently produced comparatively favorable docking results, demonstrating its suitability as an enzyme for degradation. .

Docking Simulations

Docking using UCSF Chimera involves preparing both receptor and ligand molecules and running AutoDock Vina from within the Chimera interface. First, you load your receptor and ligand structures into Chimera (commonly in `.pdb` or `.mol2` formats). Using the **Dock Prep** tool under *Tools → Structure Editing*, you prepare both molecules by adding hydrogens, assigning charges, and fixing incomplete atoms. Once prepared, go to *Tools → Surface/Binding Analysis → AutoDock Vina* to set up the docking. In this window, you select the receptor and ligand, define the search space by adjusting a grid box around the binding site, and set docking parameters such as number of modes and exhaustiveness. Make sure to provide the path to the Vina executable during setup. Once the docking run starts, Chimera will call AutoDock Vina and display the docked poses as separate models. You can explore these poses interactively in the Chimera viewer and analyze binding affinities and interactions using the **ViewDock** tool. It is recommended to save the docking results and examine key poses closely for hydrogen bonding and steric compatibility. Chimera also allows you to generate high-quality images of docking results for presentation or publication. This integrated workflow makes Chimera a powerful interface for running and visualizing docking studies with AutoDock Vina.

  • Receptor: AlphaFold-predicted TfCut2 structures (.cif format)
  • Ligands: PETduo, PETtri, and custom PBAT fragments
  • Docking Box: Centered on the active site, optimized for catalytic interaction

Software Tools Used

  • AlphaFold Beta Server: Alpha Fold is an AI system developed by Google DeepMind that predicts a protein’s 3D structure. By inputting the amino acid sequence of an enzyme, such as TfCut2 or its mutants, AlphaFold generates a .cif file that contains predicted models of the 3D structure of the enzyme. This structural model is crucial for understanding how mutations affect enzyme shape and secondary structures, especially when used in further analysis or docking.
  • UCSF Chimera:UCSF Chimera is a program for the interactive visualization and analysis of molecular structures and related data, including density maps, trajectories, and sequence alignments. We used it as a molecular visualization tool as well as a molecular dynamics simulator. It helps us run the docking simulation.
  • PyMOL– Pymol is used to visualize the 3D structure of biological molecules. Our goal is to visualize the binding socket in which TfCut2 binds with PET and PBAT, and show specific structures, such as catalytic triad and disulfide bridge, on the protein.
  • PyMOL Instructions
  • EpHod – EpHod is an AI-powered tool designed to predict the optimal pH at which the enzyme performs best.EpHod is composed of two main machine learning pipelines: RLATtr which is a neural network that looks at patterns in the sequence, and SVR which is a regression model that double checks the results. EpHod takes in a FASTA format sequence, along with some other parameters, we were able to evaluate the optimal pH conditions for enzyme activity.

Multiple Sequence Alignment (MSA) & BLAST

Multiple sequence alignment (MSA) involves aligning three or more biological sequences—such as proteins or nucleic acids—of similar lengths to identify regions of similarity. In our project, we used MSA to analyze TfCut2 sequences from various species identified through BLAST searches. The goal was to identify conserved regions, which remain unchanged across species and may play a critical role in the enzyme’s structure or catalytic function. These conserved regions are especially important when studying TfCut2's ability to degrade PET (polyethylene terephthalate), as they may include active site residues or binding domains essential for breaking down plastic. By contrast, non-conserved regions vary more between species and are likely to be involved in less critical functions, possibly allowing the enzyme to adapt to different environments. Understanding these differences helps us identify which parts of TfCut2 are essential for PET degradation, and can guide efforts to engineer more efficient or stable versions of the enzyme for plastic biodegradation.

Data Analysis in R

Using RStudio and packages like dplyr and ggplot2, we processed docking results to identify the most promising TfCut2 mutations.

  • Scatter plots: Residue vs. Docking Score
  • Heatmaps: Mutation × Substrate performance
  • Filtering by cutoff: Only mutants with consistent high binding for PET and PBAT selected