1. PET Biodegradation

1.1 Plasmid Construction & Protein Expression

To address the challenges associated with the degradation of PET by free enzymes, namely the accumulation of intermediate products and difficulties in post-reaction recovery, we prepared a functionalized magnetic carrier, K5C@Fe3O4, via a one-step biomimetic mineralization method. This carrier was designed for the sequential immobilization of SpyTag-TfCa-SnoopTag and SnoopCatcher-Tfh. Towards this end, we constructed three plasmids: piGEM25_01 for the expression of K5C, piGEM25_02 for SpyTag-TfCa-SnoopTag, and piGEM25_03 for SnoopCatcher-Tfh. The results from PCR (Figure 1D) and subsequent sequencing confirmed the successful construction of all three plasmids. We first expressed K5C and successfully purified it by leveraging the thermosensitive property of its ELP (Elastin-like polypeptide) tag via the Inverse Transition Cycling (ITC) method. Subsequently, the SpyTag-TfCa-SnoopTag and SnoopCatcher-Tfh proteins were also expressed. The SDS-PAGE analysis, shown in Figures 1E and 1F, confirmed the soluble expression of each protein.

Figure 1: Plasmid profile of piGEM25_01(A), piGEM25_02(B), piGEM25_03(C), 1% agarose gel electrophoresis results of PCR resutls(D), 15% SDS-PAGE resutls (E and F). In panel D, M: Trans 5K DNA Marker; Lane 1: pET-30(a); Lanes 2-3: pET-30(a)-SpyTag-Tfca-SnoopTag; Lane 4: pET-30(a)-SnoopCacther-Tfh; Lane 5: pET-30(a). In panel E, M: protein marker; Lanes 1-4 depict the purification of ELP-SC via ITC: Lane 1: Insoluble pellet after lysis of cells expressing ELP-SC; Lane 2: Product after the first ITC purification cycle; Lane 3: Highly purified ELP-SC after the second cycle; Lane 4: Soluble impurities removed by centrifugation. In panel F, Lane 2: Crude lysate of cells expressing SpyTag-TfCa-SnoopTag; Lane 4 : Crude lysate of cells expressing SnoopCatcher-Tfh.

1.2 Optimization of the Immobilized Dual-Enzyme System

To enhance the immobilization efficiency of SpyTag-TfCa-SnoopTag, the preparation conditions for the magnetic carrier were optimized. Figure 2A reveals that 3.0 M NaOH is the optimal concentration for the mineralization reaction. In this study, the concentrations of the products, terephthalic acid (TPA) and mono(2-hydroxyethyl) terephthalate (MHET), were determined using High-Performance Liquid Chromatography (HPLC). A comparison was made between two strategies: sequential immobilization (immobilizing TfCa first, followed by Tfh) and co-immobilization (mixing both enzymes with the carrier in a single step). Figure 2B indicates that the sequential immobilization method yielded significantly higher degradation efficiency, with the concentration of the sole product, TPA, successfully detected at 1.2 mM. To further evaluate the recyclability of the system, the magnetic particles were recovered post-reaction and reused in a new round of PET degradation. As shown in Figure 2C, the system retained 61% of its initial degradation activity after 10 cycles of reaction and recovery, demonstrating its excellent operational stability and potential for repeated use.

Figure 2: Biochemical characterization of the immobilized dual-enzyme system. (A) Effect of NaOH concentration during biomineralization on the activity of the subsequently immobilized enzyme, determined by HPLC measurement of TPA produced from MHET degradation. (B) Comparison of PET degradation activity between two immobilization strategies: Group A, sequential immobilization (TfCa then Tfh); Group B, co-immobilization (all components mixed at once). (C) Reusability assessment of the immobilized system over 10 reaction cycles. The relative activity was measured after each cycle, with the activity of the first cycle defined as 100%. Error bars in (B) and (C) represent the standard deviation from three independent experiments.

1.3 Conclusion & Outlook

By integrating ELP-based thermosensitive purification, biomimetic mineralization, and specific self-assembling covalent protein immobilization, this study successfully constructed a magnetically recyclable, immobilized dual-enzyme system for the efficient degradation of PET. The sequential immobilization strategy was identified as the optimal assembly method for this system. In the future, the system's performance can be further enhanced by optimizing the dual-enzyme ratio and modifying the structure of the linker peptides.

2. Upcycling of EG to PEAs

To further upgrade and utilize the EG produced in the previous step, we conducted the following experiments. This diagram systematically illustrates the complete synthetic pathway designed in this study, converting ethylene glycol (EG)—a byproduct of PET degradation—into high-value polyethylene succinate (PEA).

Figure 3: Complete metabolic pathway for the conversion of ethylene glycol (EG) to poly(ester amide) (PEA) in engineered E. coli. fucO, lactaldehyde reductase; aldA, aldehyde dehydrogenase A; glcDEF, glycolate dehydrogenase; glcB, malate synthase G; gcl, glyoxylate carboligase; hyi, hydroxypyruvate isomerase; glxK, glycerate 2-kinase 2; glxR, tartronate semialdehyde reductase 2; garR, Glyoxylate Reductase; phaA: acetyl-CoA acetyltransferase; phaB: acetoacetyl-CoA reductase; phaC: polyhydroxyalkanoate synthase; panD: aspartate decarboxylase; act: β-alanine CoA-transferase; pct540: propionate CoA-transferase mutant; EG, Ethylene glycol; GLA, glycolaldehyde; GA, Glycolate; GLO, Glyoxylate; TSA, (2R)-tartronate semialdehyde; (OH)-PYR, hydroxypyruvate; GLR, D-glycerate; 2PG, 2-phospho-D-glycerate; MAL, malate

2.1 Engineering E. coli for EG Utilization and Detection

Since wild-type E. coli cannot utilize EG as a carbon source, to enable BL21(DE3) to efficiently utilize EG, we first established a positive baseline using the previously reported four-gene combination gcl, hyi, glxR, glxK (constructed as piGEM25_05). Building upon this, based on the minimal critical node recommendations proposed by modeling, we retained gcl and glxK, introduced the isoenzyme garR to replace glxR, and evaluated the extended value of the upstream fucO and aldA. Three expression combinations were constructed in parallel for comparison: piGEM25_05 (gcl, hyi, glxR, glxK, literature baseline control), piGEM25_04 (gcl, garR, glxK, model-driven minimal combination), piGEM25_06 (fucO, aldA, upstream extension validation module). The rationale and logic for gene selection are detailed in the Model section. The plasmid map is shown in (Figure 4A-C). PCR amplification and electrophoresis validation confirmed that the target band size matched expectations (Figure 4D). Sequencing results further confirmed the successful construction of this recombinant plasmid.

Figure 4: Plasmid profile of piGEM25_04(A), piGEM25_05(B), piGEM25_06(C), 1% agarose gel electrophoresis results of PCR resutls (D) . In panel D, M: Trans 5K DNA Marker; Lane 1:pCDFDuet-1(linearized); Lane 2:gcl; Lane 3:garR; Lane 4:glxR; Lane 5:glxK; Lane 6:hyi; Lane 7:PRSFDuet-1(linearized); Lane 8: fuco; Lane 9:aldA;

In this experiment, the functional differences in ethylene glycol (EG) metabolism among three distinct gene combinations were systematically compared. As shown in Figure 5, both engineered strains EG04-BL21(DE3) (expressing piGEM25_04) and EG05-BL21(DE3) (expressing piGEM25_05) effectively utilized EG as the sole carbon source for growth. Specifically, after 72 hours of cultivation, EG04-BL21(DE3) reached an OD600 of 0.218 with an EG degradation rate of 28%; EG05-BL21(DE3) exhibited an OD600 of 0.24 with a 24% EG degradation rate. While both groups demonstrated comparable metabolic efficiency, the three-gene combination employed in EG04-BL21(DE3) achieved streamlined optimization of the metabolic pathway. Furthermore, the co-transformed strain EG04-06-BL21(DE3) (simultaneously expressing piGEM25_04 and piGEM25_06) and the blank control EV-BL21(DE3) failed to grow in medium where EG served as the sole carbon source.

Figure 5: Rational engineering of BL21 (DE3) chassis cells utilizing ethylene glycol as the sole carbon source.Error bars indicate standard error (n = 3)

2.2 Ethylene glycol（EG）detection

2.2.1 UV-Vis Spectral measurement results

To conduct real-time monitoring of the concentration during the EG assimilation process, we have developed an EG detection system for this purpose，The changes in the spectra were detected using a UV-Vis spectrophotometer. As shown in Figure 6A, the increase in EG concentration led to an enhancement in the intensity of the characteristic absorption peak (SPR peak) of the gold nanoparticles at 525 nm, but no spectral shift was observed. Figure 6B shows that the addition of different concentrations of EG to the gold nanoparticle solution resulted in color changes that were indistinguishable to the naked eye. These indicates that the size and shape of the gold nanoparticles did not change significantly during the experiment, and no notable particle aggregation occurred. This further confirms that the change in peak intensity is primarily attributed to the alteration of the dielectric environment (refractive index), rather than changes in the intrinsic properties of the nanoparticles. As shown in the Figure 6C, there was a linear relationship between A525 value and EG concentration in the range of 0-150mM, with a linear regression correlation coefficient of 0.988. This result indicated that UV-Vis absorption method is promising for the quantification analysis of EG concentration.

Figure 6A: The UV-Vis spectra obtained in the presence of different concentrations of EG.

Figure 6B: Color changes of the gold nanoparticle solution with increasing EG concentration (0, 10, 30, 50, 100 mM).

Figure 6C: Linear fitting curve of absorbance at 525 nm versus EG concentration detected by UV-vis spectrophotometer.

2.2.2 Based on the detection results from the AS7341 sensor

The prepared 800 µL AuNPs solution was mixed with 1200 µL of EG standards at varying concentrations in an LB medium system and introduced into a cuvette for sensor detection. The AS7341 sensor was activated to sequentially illuminate the sample with six LEDs aligned opposite the light-transmitting hole of the cuvette holder. Light passing through the sample was captured by the AS7341 spectral sensor, which acquired absorption spectral data at a wavelength of 525 nm. Subsequently, using the spectral data, a calibration plot was constructed to establish a linear relationship between the absorbance at 525 nm and EG concentration. As shown in Figure 7, the A525 value exhibited a linear correlation with EG concentration over the range of 0–150 mM, with a linear regression correlation coefficient of 0.9734, indicating that the method we designed here was accurate and feasible for the detection of EG.

Figure 7: the linear fitting curve of the absorbance value at 525 nm versus the EG concentration.

2.3 Synthesis & Detection of PEAs

2.3.1 Design and Construction of the PEA Synthesis Pathway

After resolving the issue of wild-type E. coli's inability to utilize EG as a carbon source, we attempted to achieve the conversion of EG into biodegradable polyester amide (PEAs) within E. coli. This required first constructing a synthetic pathway from natural amino acids to PEAs in BL12 (DE3). We designed and constructed the piGEM25_07 plasmid to synthesize polymer monomers (sodium RS-3-hydroxybutyrate and 3-aminopropionic acid) (Figure 8A) and the piGEM25_08 plasmid to activate and polymerize these monomers (Figure 8B). As shown in Figure 8C, PCR results and subsequent sequencing confirmed the successful construction of both recombinant plasmids. Expression of the target protein was then investigated. SDS-PAGE results in Figure 8D demonstrate the successful soluble expression of the key enzyme gene.

Figure 8: Plasmid profile of piGEM25_07(A), piGEM25_08(B), 1% agarose gel electrophoresis results of PCR resutls (C)and 10% SDS-PAGE resutls (D). In panel C, M: Trans 5K DNA Marker; Lanes 1:pGro7; Lanes 2:phaC; Lanes 3:Pct540; Lanes 4:Act; Lanes 5:pTrc99a; Lanes 6:phaA; Lanes 7:phaB; Lanes 8: panD; In panel D, M: protein marker ; Lanes 1 and 2 is the supernatant and pellets of the lysed E. coli cell expressing piGEM25_08 Proteins(Control group, without LA) ; Lanes 3 and 4 is the supernatant and pellets of the lysed E. coli cell expressing piGEM25_08 Proteins (Experimental group, with LA induction) ; Lanes 5 and 6 is the supernatant and pellets of the lysed E. coli cell expressing piGEM25_07 and piGEM25_08 Proteins (Control group, without IPTG and LA) ; Lanes 7 and 8 is the supernatant and pellets of the lysed E. coli cell expressing piGEM25_07 and piGEM25_08 Proteins (Experimental group, with IPTG and LA induction);

2.3.2 Synthesis of PEA and Identification by Nile Red Staining

To validate whether the PEA synthesis pathway constructed in BL12 (DE3) can synthesize PEA, we engineered two strains: PEA01-BL12 (DE3) (transformed with only plasmid piGEM25_08, requiring exogenous monomer addition for PEA synthesis) and PEA02-BL12 (DE3) (co-transformed with piGEM25_07 and piGEM25_08). A control group, EV-BL21(DE3) (transformed with piGEM25_08 alone, without inducer), was established. After 72 hours of cultivation, PEA01-BL12 (DE3), PEA02-BL12 (DE3), and EV-BL21(DE3) were cultured for 72 hours. After incubation, samples with different OD600 values (OD600=0.3, 0.5, 1.0) to characterize the presence of PEAs within E. coli cells. As shown by the bacterial cell colors in Figure 9A-C, the control strain EV-BL21(DE3) was not stained by Nile red, while PEA01-BL12 (DE3) and PEA02-BL12 (DE3) were stained. The stained PEA02-BL12 (DE3) sample was then analyzed using a fluorescence spectrophotometer with excitation at 530 nm and emission measured at 610 nm. As shown in Figure 9D, compared to the control group, PEA02-BL12 (DE3) exhibited a strong fluorescence emission peak at 610 nm after Nile Red staining. The Nile Red staining experiment demonstrates that the de novo synthesis pathway for PEA constructed in BL21(DE3) successfully achieved PEA synthesis.

Figure 9: Nile Red staining results of the control group(A), PEA01-BL12 (DE3)(B), and PEA02-BL12 (DE3)(C) and fluorescence spectrum analysis(D). In panel A, Control group (EV-BL21(DE3)) bacterial samples at OD600 = 0.3, 0.5, and 1.0 (left to right). Cells were cultured in MR medium with 0.4 mM IPTG induction, 30°C for 96h. In panel B, PEA01-BL21(DE3) samples at OD600 = 0.3, 0.5, and 1.0 (left to right). Cultured in MR medium with 1.5mg/ml LA induction, 30°C for 96h. In panel C, PEA02-BL21(DE3) samples at OD600 = 0.3, 0.5, and 1.0 (left to right). Cultured in MR medium with 0.4 mM IPTG and 1.5mg/ml LA induction, 30°C for 96h, supplemented with 3HB and 3AP monomers. In panel D, fluorescence emission spectra of Nile Red-stained samples (excitation at 530 nm, emission measured at 610 nm). All samples were prepared by adding 3 μL of Nile Red stock solution to 1 mL of PEA suspension, followed by vortex mixing and incubation in the dark at 25 °C with shaking at 200 rpm for 20 min. After staining, samples were centrifuged (12,000 rpm, 1 min) to remove unbound dye, resuspended in PBS, and analyzed by fluorometry.

We further extracted intracellular polyesteramides (PEAs) and observed them under fluorescence microscopy after staining with Nile Red. As shown in Figure 10, compared to the control group without inducer addition, the PEAs extracted from the experimental group exhibited specific staining with Nile Red and emitted significant fluorescence under the microscope. This further validated our successful synthesis of PEAs using the whole-cell system.

Figure 10: Fluorescence microscopy of intracellular PEAs stained with Nile Red. The experimental group （A）shows strong fluorescence, whereas the negative control （B）exhibits minimal background signal.

2.4 Conclusion & Outlook

In this study, bioinformatics analysis identified key metabolic genes involved in EG utilization by Escherichia coli. Compared to the previously reported four-gene combination gcl–hyi–glxR–glxK, the gcl-garR-glxK combination maintains (and slightly enhances) EG utilization efficiency while reducing one gene module. Subsequent efforts may optimize the expression levels of these three genes and employ bioinformatics methods to identify more efficient metabolic pathways, thereby enhancing EG utilization in E. coli. Furthermore, to ultimately achieve efficient carbon flux conversion into intracellular polyesteramides (PEAs), we constructed a non-native metabolic pathway within E. coli, successfully achieving intracellular PEAs synthesis. However, current yields remain low. Future work will focus on optimizing gene expression within the pathway and refining extraction steps to enhance PEAs yield and extraction efficiency. Additionally, our project has successfully developed a simple, convenient, and specific EG detection method. The entire process requires no complex pretreatment; simply mixing the sample with the sensing solution yields results within minutes, significantly outperforming traditional chromatographic methods. We believe that in the near future, optical detection, signal processing, and display functions could even be integrated into a portable device to better serve societal needs.

3. Upcycling of TPA to PCA

Having addressed the valorization of EG, one of the PET degradation products, we turned our attention to upcycling the other major byproduct, terephthalic acid (TPA). This effort was motivated by a 2024 project from our university's UESTC-iGEM team, which encountered very low vanillin yields when converting TPA—one of the two main products of PET hydrolysis—into vanillin. We hypothesized that a key limiting factor was the insufficient concentration of protocatechuic acid (PCA), a precursor to vanillin. To address this, our project focused on optimizing the expression conditions for the enzymes in the TPA-to-PCA pathway. We first designed plasmid piGEM25_09 to express TphA1, TphA2, TphA3, and TphB, which collectively catalyze the conversion of TPA to PCA. In a key departure from the previous project, our expression protocol included the supplementation of FeCl₃, L-cysteine, and ZnSO₄. The SDS-PAGE analysis in Figure 11B confirmed the successful soluble expression of these target proteins in the E. coli BL21(DE3) host. We then measured the concentration of the resulting PCA using HPLC and investigated the effect of cell density on TPA conversion efficiency. As shown in Figure 12D, at an OD600 of 60, the PCA titer reached 1250 μM.

15% SDS-PAGE results of the pellets and supernant of empty vector and the lysed E. coli cell harbouring piGEM25_09 plasmid — Figure 11: 15% SDS-PAGE results of the pellets and supernant of empty vector (panel B: lane 1 and lane 2) and the lysed E. coli cell harbouring piGEM25_09 plasmid (panel B: lane 3 and lane 4).

3.2 HPLC Analysis of TPA-to-PCA Conversion

Figure 12: HPLC analysis for the bioconversion of TPA to PCA. Standard curves obtained from HPLC for (A) mono-(2-hydroxyethyl) terephthalate (MHET), (B) terephthalic acid (TPA), and (C) protocatechuic acid (PCA). (D) Effect of bacterial optical density (OD₆₀₀) on the final concentration of PCA produced from TPA by the engineered E. coli.

3.3 Conclusion & Outlook

We have successfully upcycled TPA into PCA using engineered E. coli that co-expresses multiple enzymes. Through the optimization of expression conditions for TphA1, TphA2, TphA3, and TphB, a substantial increase in PCA yield was achieved. Looking ahead, our future work will focus on addressing the challenge of TPA transmembrane transport to further elevate PCA production.