The first step was the production of plasmids encoding enzymes PETase (BBa_K4484000) and MHETase (BBa_K4484001). Additionally, surface displayed versions were designed by linking the transporter proteins AIDA-I (BBa_K4484003) and YeeJ (BBa_K4484007) to the desired enzyme sequences. Each of these plasmids were then transformed into E. coli C41(DE3) and protein expression were induced. SDS-PAGE and Western blot (His Tag targeting antibody) were used to confirm the presence of proteins at the expected size as in Figure 1. While issues appeared such as multiple bands which indicated the possibility of undesired protein truncation, the constructs were generally successful as evidenced by bands at expected sizes. Additionally, although none of the AIDA-I constructs showed a band on the Western blot (we are still trouble shooting on this topic), the AIDA-I-PETase constructs (BBa_K4484005) later showed activity on the BHET plate and AIDA-I-FAST-PETase constructs (BBa_K4484006) showed activity on a PET film, demonstrating the successful expression of functional proteins. The constructs YeeJ-FAST-PETase (BBa_K4484010) and AIDA-I-FAST-PETase (BBa_K448406) were also designed and expressed with similar results to these earlier constructs.

Figure 1. Example Western blot for various purified and surface displayed proteins.

In order to verify the activity of our constructs, we used a common activity assay involving measuring the concentration of degradation products through HPLC after degrading PET/intermediates. For PETase, reaction solutions were made by mixing enzymes or whole cell samples in the reaction buffer with an amorphous PET film. In the case of MHETase, a solution of MHET was made as it is not active on PET on its own. The HPLC measurements of MHETase samples are shown in Figure 2. Each of the constructs shows a disappearance of the peak corresponding to MHET and appearance of a TPA peak, which implies the surface displayed enzymes are active. Meanwhile, E. coli cells expressing MHETase intracellularly also showed MHET degradation, giving the potential for using E. coli with intracellular MHETase for plastics degradations.

Figure 2. HPLC measurements of OD 20 whole cell MHETase constructs after 168 hours of reaction at 25 °C. Reaction medium is 250 μM MHET in 50 mM Bicine-NaOH pH 8.5. (All same scale)

For the assays of PETase, the HPLC measurements are shown in Figure 3. Both surface display constructs of PETase seem to show little activity compared to the purified enzyme. This was hypothesized to be from low surface display efficiency of enzymes resulting in negligible activity. New constructs with surface displayed FAST-PETase were made and expressed to evaluate if increasing the enzyme activity would allow for measurable HPLC results. The FAST-PETase measurements shown in Figure 4 display mixed results as YeeJ-PETase (BBa_K4484008) shows negligible activity while AIDA-I-FAST-PETase shows measurable activity based on the appearance of peaks corresponding to degradation products TPA and MHET. The intensity is relatively lower compared to the purified enzyme which may indicate a relatively lower surface display efficiency limiting overall activity. However, in our AIDA-I-FAST-PETase samples stored at 4 °C, loss of cell viability was observed within 48 hours of expression. Hence, it is possible that the observed activity is because of cell lysates containing FAST-PETase. Future experiments will focus on ruling out this possibility.

Figure 3. HPLC measurements of pure PETase and OD 30 whole cell PETase constructs after reaction for 168 hours at 25 °C. Reaction medium is 50 mM Bicine-NaOH pH 8.5. (All same scale)

Figure 4. HPLC measurements of Pure FAST-PETase and OD 30 whole cell FAST-PETase constructs after 213 hours of reaction at 25 °C. Reaction medium is 50 mM Bicine-NaOH pH 8.5.

To begin developing our assay, we first had to synthesize the substrate material fluorescein dibenzoate (FDBz). Thin liquid chromatography was used to confirm product purity in the collected fractions which were dried by rotary evaporation. After product purification, it was analyzed by H1 NMR for which the results can be seen in Figure 5. Comparing literature available online as well as simulated NMR spectra, the spectra appears to align with expectations indicating the successful synthesis of FDBz. However, there are some questions of purity as the NMR spectra of FDBz and fluorescein are relatively close and some fluorescence was observed when the FDBz product is placed in aqueous solution, though it is unclear if this is due to fluorescein impurities or automatic hydrolysis.

Figure 5. H1 NMR for FDBz synthesis product.

Tests with Pure Enzymes

When evaluating the assays, we first used purified enzymes to evaluate the effectiveness of the test and various methods of analysis. Figure 6 below shows the result of one such assay with varying concentrations of PETase. Fluorescence values correspond to measurements taken at the end of three hours relative to negative controls. The approximately linear correlation between the fluorescence values and enzyme concentration confirms the ability to measure enzyme activity reliably.

Figure 6. Fluorescence relative to negative control across various purified PETase concentrations after 3 hours of reaction with 250 μM FDBz.

We also showed that the assay could display the increased efficiency of different enzymes using FAST-PETase, a modified PETase optimized using artificial intelligence. Figure 7 illustrates the significant increase in fluorescence of FAST-PETase compared to regular PETase which corresponds to the expected increase in activity.

Figure 7. Relative fluorescence of PETase and FAST-PETase after 3 hours of reaction with 250 μM FDBz (0.5 μM purified enzyme, 37°C)

Tests with Surface Displayed Enyzmes

After using purified enzymes, experiments with both surface displayed E. coli whole cell and PETase whole cell lysate samples (without purification) were performed. The results from these experiments typically looked similar to Figure 5, which shows an inverse relationship between fluorescence and cell concentration. This phenomenon is mostly caused by the unexpected rise in fluorescence with plain E. coli cells and lysate over time which is generally comparable with that of enzyme-containing samples. The reason behind this unexpected rise has not been explained and is hoped to be addressed in future work.

Figure 8. Fluorescence of PETase lysate at various concentrations (OD prior to cell lysis) after 3 hours of reaction with 250 μM FDBz, 37°C.

BHET Plates

BHET is a compound similar to PET that is a trimer made up of one TPA molecule and two ethylene glycol molecules. As such, PETase shows activity on this molecule while organisms lacking the enzymes will be unable to degrade it. Plates for qualitative evaluation of PET degrading activity were developed by creating a suspension of BHET in agar media. Early experiments evaluated varying concentrations of BHET to determine a standardized procedure. At high concentrations, the BHET would quickly form large crystal structures when cooling, which raised concerns about accuracy given the heterogeneous distribution. A BHET concentration of 20 mM was decided as the standard as it resulted in a uniform turbidity throughout the solution and plate. Next, we then began experiments to confirm the sensitivity of the assay by placing various purified enzymes and controls on the plate, for which an example result is shown in Figure 9. The BHET plates were considered successful as they showed visual changes in the spots of active enzymes and no visual change for enzymes inactive on PET. Further experiments investigated surface displayed enzymes on the plates, with an example shown in Figure 10. We were able to confirm that only PETase was able to degrade the BHET as evidenced by the change in appearance of the plate while MHETase and control portions showed no visual difference. These experiments also revealed that this assay could not distinguish between intracellular and surface displayed proteins as cells expressing intracellular PETase would create positive visual changes.

Figure 9. Example BHET plate assay containing purified PETase, FAST-PETase, and MHETase

Figure 10. Example BHET plate spots under 2x magnification showing (from left to right) purified PETase (1), YeeJ-PETase (4), AIDA-I-PETase (7), YeeJ-MHETase (19).

One final assay we performed was comparing the effectiveness of standard PETase against FAST-PETase. Multiple samples of both enzymes were placed on a 20 uM plate and recorded over approximately 30 minutes. The FAST-PETase quickly creates clear zones in the locations much more quickly than standard PETase, indicating its significantly higher efficiency as expected, though the activity generally cannot be quantified.

Figure 11. Before and after 30 minutes of reaction of purified FAST-PETase and regular PETase on a 20 mM BHET plate.

Directed Evolution

Another experiment performed was the mutation of the PETase insert sequence through error-prone PCR for the purpose of showing the ability to develop new enzymes with improved activity and thermal stability. Two mutant sequences were developed and expressed before being tested using a BHET plate assay. Based on the image in Figure 12, both mutants created no visual changes in the plate and as such appear to have caused a loss of function in the enzyme.

Figure 12. Mutant PETase enzymes on a BHET plate showing loss of function compared to the original PETase enzyme.