ENGINEERING

We employed two different iterations of the Design-Build-Test-Learn (DBTL) cycle to pick the best mutations and zone in on the ones to use in the wet lab.

Iteration 1: in-silico mutant MHETase generation

Iteration 2: in-silico to the wet lab

ITERATION 1

Dry Lab

1.1 Design: The design process began with a paper by Cui et al., who adopted an approach on finding stabilizing mutations in enzyme using the ddG parameter. They named this strategy GRAPE and used it to enhance the performance of PETase, an enzyme capable of degrading PET plastics.
Meanwhile, Lu et al [2]. introduced a neural network machine learning model, MutCompute, to predict mutations for creating improvised enzymes. They leveraged the lessons learnt from Cui’s work to create FAST-PETase, a more efficient version of the enzyme. We designed a similar approach for MHETases to explore possible mutations that could make it more stable compared to the wild type such that it can act with FAST-PETase.

1.2 Build:Inspired by these studies, our team decided to merge the best of both approaches to find better mutant MHETases. We started by using the MutCompute machine learning model to predict mutations. Our goal was straightforward-identify mutations that could not only improve the enzyme’s activity, but also enhance its stability. We selected the mutations with the lowest ddG values (-1.0 KJ/mol as the baseline) for further examination, aiming for a stabilizing mutation that would improve the enzyme’s performance.

1.3 Test: The sorted variants were then tested. To evaluate how well the mutant enzyme interacted with its substrate MHET, we used AutoDock Vina for molecular docking. The force field utilized was Charmm27. For our in-silico studies, we used the crystal structure PDB ID: 6QGA as it had been previously used in studies for figuring out the mechanism of the enzyme [3]. Some variants showed higher docking scores than the wild type. We performed molecular dynamics simulation and MM/PBSA on the proving mutants and using parameters like RMSD, RMSF and free energy calculations of the protein-ligand complex, we monitored the effect of mutations.

1.4 Learn: From the results that we got from the above performed in-silico analysis iterations, our team , with the aid of the dry lab research scholar Mr. Rudra Mishra, came to a conclusion that the single mutant S491A and a double mutant S196A, T434S are better performing variants of the MHETase enzyme, proving via in-silico study that they perform better than its wild type..

Fig 1.1 Docking Scores

Fig 1.2 RMSD of protein backbone

Fig 1.3.1 Free energy calculations for the S491A

Fig 1.3.2 Free energy calculations for S196A, T434

Fig 1.3.3 Free energy calculations for WT

ITERATION 2

Wet Lab

2.1 Design

• Our first aim was to find a suitable chassis for our gene of interest, based on its length.
• We studied various subclasses of PET vectors for expression in the Escherichia coli BL21 (DE3) strain.
• We came to the conclusion that pET-22b(+) vector was the best choice as it has a pelB sequence (signaling peptide which helps to translocate our protein of interest to periplasmic space for extracellular production) and 6X Histidine tag for purification through nickel affinity chromatography. It also has a desired promoter sequence for the expression. Mr. Gomathinayagam, a research assistant at our lab, aided us in finalizing the protocols.
• All the reagents including qPCR and colony PCR kits, Escherichia coli strains (DH5ɑ and BL21 (DE3)), antibodies, IPTG, Plasmid DNA isolation kit, protein purification kit were selected after extensive literature review.

Vector map of pET-22b(+)

FASTPETase gene sequence-870bp

MHETase gene sequence-1809bp

FASTPETase cloned in pET-22b(+) with EcoR1 and Xho1 restriction sites

Wild Type MHETase cloned in pET-22b(+) with EcoR1 and Xho1 restriction sites

S491A single mutant of MHETase

S196A, T434S double mutant MHETase

2.2 Build

• For the construction of our vector, we ordered our gene sequence with our plasmid, as the option was available in Genscript.This eliminated potential errors in restriction digestion and ligation. Our genes of interest (FASTPETase and MHETase) along with the promoter-PelB-GOI-6X His Tag- terminator were inserted in the MCS region of our plasmid.
• Troubleshooting was done for frameshift and misreading of base pairs while performing cloning simulation in SnapGene. Our mutant MHETase gene was also codon optimized for E.coli.

S491A single mutant MHETase cloned in pET-22b(+) with EcoR1 and Xho1 restriction sites

S196A, T434S double mutant MHETase cloned in pET-22b(+) with EcoR1 and Xho1 restriction sites

For cloning studies, we designed primers using Snapgene and Primer3 software for various sequences.

2.3 Test Competent cells were prepared for E. coli DH5ɑ and E. coli BL21 (DE3) strains respectively and stored in glycerol stock at -80’C. Transformation of wild type MHETase and FASTPETase were done in the cloning vector, E. coli DH5ɑ strain. We confirmed successful transformants using antibiotic selection (ampicillin) and plated them again to confirm false postives , this was done with the aid of Mr.Bharath and Ms. Janani, the research scholars present in our lab. Our GOIs were confirmed by colony PCR and qPCR using our primers and had successful amplification peaks. Ms. Srivarshini Shankar helped us set up the qPCR recipe. Mr. Venkat and Ms.Huldah aided us in interpreting the colony PCR results. We grew the cloned colonies in LB broth overnight, isolated plasmid and stored some positive colonies at -20’C. We proceeded with cloning them in E. coli BL21(DE3) strain for expression studies and selected the transformants. Colony PCR and qPCR were performed to check the presence of GOI and positive results were obtained.

2.4 Learn After the results that we got during the test, we had to troubleshoot for errors and steps and learnt the exact way of approaches, protocols and interpretations. At first the results weren’t proper - the qPCR results were not as expected, transformation efficiency was low. Integrating expert inputs, helped us get our desired results in subsequent runs. Dr. Gothandam KM suggested a way to overcome evaporation during colony PCR, by using a longer , more dense gel and to run it for a shorter period of time.

References

1. Yinglu Cui, Yanchun Chen, Xinyue Liu et al. Computational Redesign of a PETase for Plastic Biodegradation under Ambient Condition by the GRAPE Strategy. ACS Catalysis 2021 11 (3), 1340-1350. DOI: 10.1021/acscatal.0c05126

2. Lu, H., Diaz, D.J., Czarnecki, N.J. et al. Machine learning-aided engineering of hydrolases for PET depolymerization. Nature 604, 662–667 (2022). https://doi.org/10.1038/s41586-022-04599-z

3. Alexandre V. Pinto, Pedro Ferreira, Rui P. P. Neves et al. Reaction Mechanism of MHETase, a PET Degrading Enzyme. ACS Catalysis 2021 11 (16), 10416-10428. DOI: 10.1021/acscatal.1c02444

4. Shi, L., Liu, H., Gao, S., Weng, Y., & Zhu, L. (2021). Enhanced extracellular production of is PETase in Escherichia coli via engineering of the pelB signal peptide. Journal of Agricultural and Food Chemistry, 69(7), 2245-2252.

5. Yoshida, S., Hiraga, K., Taniguchi, I., & Oda, K. (2021). Ideonella sakaiensis, PETase, and MHETase: From identification of microbial PET degradation to enzyme characterization. In Methods in enzymology (Vol. 648, pp. 187-205). Academic Press.