β-sheet Engineering of IsPETase for PET Depolymerization

Songfeng Gao; Lixia Shi; Hongli Wei; Pi Liu; Wei Zhao; Lanyu Gong; Zijian Tan; Huanhuan Zhai; Weidong Liu; Haifeng Liu; Leilei Zhu

doi:10.1016/j.eng.2024.10.015

Engineering ›› 2025, Vol. 47 ›› Issue (4) :192 -206. DOI: 10.1016/j.eng.2024.10.015

Research Environmental Engineering—Article

research-article

β-sheet Engineering of IsPETase for PET Depolymerization

Songfeng Gao ^a^,^b^,^#
, Lixia Shi ^a^,^b^,^c^,^#
, Hongli Wei ^b^,^d^,^#
, Pi Liu ^b^,^d
, Wei Zhao ^b^,^d
, Lanyu Gong ^b^,^d
, Zijian Tan ^b^,^d^,^e
, Huanhuan Zhai ^b^,^d
, Weidong Liu ^b^,^d^,^*
, Haifeng Liu ^e^,^*
, Leilei Zhu ^b^,^d^,^*

Author information +

History +

PDF (4004KB)

Abstract

The enzymatic depolymerization of polyethylene terephthalate (PET) offers a sustainable approach for the recycling of PET waste. Great efforts have been devoted to engineering PET depolymerases on the substrate binding cleft and the surrounding loops/α-helices on the surface. Here, we report the systematic engineering of whole β-sheet regions in the core of IsPETase (a PETase from Ideonella sakaiensis) via a fluorescent high-throughput screening assay. Twenty-one beneficial substitutions were obtained and iteratively recombined. The best variant, DepoPETase β, with an increase in the melting temperatures (T_m) of 22.9 °C, exhibited superior depolymerization performance and enabled complete depolymerization of 100.5 g of untreated post-consumer PET (pc-PET; 0.26% W_enzyme/W_PET enzyme loading) in liter-scale bioreactor at 50 °C within 4 d. Crystallization and molecular dynamics simulations revealed that the improved activity and thermostability of DepoPETase β were due to enhanced hydrogen bonds and salt bridges in the β-sheet region, a more tightly packed structure of the core sheets and the surrounding helix, and improved binding of PET to the active sites. This study not only demonstrates the importance of engineering strategy in the β-sheet region of PET hydrolases but also provides a potential PET depolymerase for large-scale PET recycling.

Graphical abstract

Keywords

Directed evolution / High-throughput screening / PETase / PET depolymerization / Thermostability / β-sheet engineering

Cite this article

Download citation ▾

Songfeng Gao, Lixia Shi, Hongli Wei, Pi Liu, Wei Zhao, Lanyu Gong, Zijian Tan, Huanhuan Zhai, Weidong Liu, Haifeng Liu, Leilei Zhu. β-sheet Engineering of IsPETase for PET Depolymerization. Engineering, 2025, 47(4): 192-206 DOI:10.1016/j.eng.2024.10.015

登录浏览全文

4963

注册一个新账户忘记密码

1. Introduction

Polyethylene terephthalate (PET), synthesized from the building blocks terephthalic acid (TPA) and ethylene glycol (EG), is one of the most abundant polyester plastics [1], [2], [3], [4]. Owing to its excellent properties, such as its durability, light weight, and waterproofness, PET is widely used in the textile, packaging, electric appliance, and medical fiber industries and almost 70 million tonnes of PET are manufactured annually worldwide [1], [2], [3], [4]. Large amounts of post-consumer PET (pc-PET) waste are generated and accumulate in nature, causing severe environmental pollution [5], [6]. In addition, PET waste is one of the main sources of microplastics that have been introduced into the food chain through marine animals or mammals, posing a serious threat to animal and human health [7], [8], [9], [10]. The depolymerization and recycling of pc-PET waste is a sustainable strategy for the utilization of resources and reducing our dependence on petroleum resources. Compared with chemical depolymerization approaches, environmentally friendly biodepolymerization of PET has attracted extensive attention [6], [11], [12], [13], [14].

PET hydrolases, such as lipase [15], [16], thermophilic cutinase [17], [18], [19], [20], [21], [22], [23], [24], [25], and mesophilic IsPETase [26], have been identified and extensively studied in recent years [6], [27], [28], [29], [30], [31], [32]. Great efforts have been devoted to sourcing cutinase-like or IsPETase-like enzymes from various databases to increase the number and diversity of PET hydrolases [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52]. These enzymes can catalyze the hydrolysis of PET into mono(2-hydroxyethyl) terephthalic acid (MHET), TPA, and EG [49], [53], [54], [55]. Among them, a variant of leaf compost cutinase, LCC^ICCG [23] and a recently developed PET hydrolase, PES-H1^L92F/Q94Y [33] showed superior depolymerization performance at temperatures ranging from 60–72 °C [49]. Moreover, IsPETase can hydrolyze PET more specifically at ambient temperature [56], [57], [58]. However, the insufficient activity and robustness of IsPETase limit the development of industrial-scale enzymatic PET depolymerization processes under mild conditions. The (semi)rational design of IsPETase has been extensively carried out since the first crystal structure of PETase was resolved in 2017 [57], [59]. Several highly improved variants were obtained, such as ThermoPETase [60], DuraPETase [61], FAST-PETase [62], and DepoPETase [50], [59], [63], [64], [65], [66], [67], [68], [69], [70] (Fig. 1). Moreover, great progress has also been made in the directed evolution of IsPETase. The first directed evolution workflow employing ultra-performance liquid chromatography (UPLC) for the engineering of IsPETase yielded a thermostable variant, HotPETase (melting temperatures (T_m) = 82.5 °C), which enabled 48.1% depolymerization of the semicrystalline PET section of PET/PE composite films in 24 h with an enzyme loading of 0.29 mg·g⁻¹ at 60 °C [71]. However, until recently, the engineered regions of IsPETase were mainly focused on the substrate-binding cleft and its surrounding areas, which are composed of loops and α-helices on the surface (Fig. 1) [60], [61], [62], [64], [65], [71], [72], [73], [74]. The buried β-sheet region and its vicinity in IsPETase have been less explored, except for the four positions F229Y [65], S282C [73], M154G [71], and F201I [72]. Considering that the folding pattern of IsPETase consists of a core of nine β-sheets flanked by seven helical structures on either side, the stability of the core skeleton of IsPETase is largely determined by the strength of the interaction between the core sheets and the surrounding helical structure [56]. Moreover, the overall structural stability of IsPETase affects the integrity of the catalytic triad located on the protein surface [75]. Therefore, engineering the β-sheet region could be beneficial for improving the stability and activity of IsPETase. Additionally, there have been successful attempts to improve activity or stability by engineering the β-sheet regions of other enzymes [76], [77], [78].

In this study, we screened 58 site saturation mutagenesis (SSM) libraries covering the whole β-sheet region of IsPETase. Beneficial substitutions were identified and recombined by employing a sequential recombination strategy as well as a K-means clustering algorithm-guided combination strategy. The resulting variants exhibited significantly increased melting temperatures and excellent depolymerization performances toward different pc-PET materials under mild conditions. Moreover, complete depolymerization of untreated pc-PET waste was achieved with the best variant in a liter-scale reactor.

2. Materials and methods

2.1. Chemicals and materials

All the chemicals used in this study were of analytical grade or higher and were obtained from Sinopharm Yuanye Biological (China), Aladdin (China), or Yuanye Biological (China) unless stated otherwise. The amorphous PET films (ES301445) were obtained from Goodfellow GmbH (UK). The packaging box of drumsticks was purchased from a local supermarket (China). The lid of soda cup was purchased from Kentucky Fried Chicken (KFC; China). The plasmid was constructed by inserting the gene PETase with the signal peptide pelB between the EcoRI and XhoI restriction sites of pET22b (+). TransStart FastPfu DNA Polymerase was obtained from TransGen Biotech (China), and QuickCut Dpn I was obtained from Takara Bio, Inc. (China). Oligonucleotide synthesis was carried out by Genewiz (China) or Tsingke Biotechnology Co., Ltd. (China). DNA sequencing was conducted by Genewiz (China).

2.2. Synthesis and characterization of MHET-OH

First, 1.82 g of 2-hydroxyterephthalate (TPA-OH), 0.5 mL of sulfuric acid, and 15 mL of EG were added to a three-necked round bottom flask. The mixture in the flask was stirred and refluxed at 20 °C for 48 h. Subsequently, the supernatant was removed by reducing pressure, and 3-hydroxy-4-((2-hydroxyethyl) carbonyl) benzoic acid (MHET-OH) was separated from the residues by column chromatography. The synthesized MHET-OH was characterized by proton nuclear magnetic resonance (¹H-NMR).

2.3. Construction of the high-throughput screening assay

The fluorescence spectra of MHET-OH and TPA-OH were obtained following a previously reported procedure [63]. At an excitation wavelength (λ_Ex) of 320 nm, the fluorescence emission intensities at emission wavelengths (λ_Em) ranging from 350–850 nm of 2.5 mmol·L⁻¹ MHET-OH and TPA-OH were measured. The wavelength at which the TPA-OH product exhibited prominent fluorescence while the MHET-OH substrate resulted in the lowest background fluorescence was selected as the optimal emission wavelength. Excitation was then induced with various excitation wavelengths in the range of 230–400 nm at the optimal emission wavelength to determine the optimal excitation wavelength. Subsequently, MHET-OH acted as the substrate to detect the hydrolytic activity of the crude enzyme pET22b-PETase. pET22b-PETase (15–50 μL) was added to a 96-well black multi-tier plate (MTP) and supplemented to 150 μL with Na₂HPO₄-NaH₂PO₄ buffer (0.1 mol·L⁻¹, pH 8.0). Then, 50 μL of 10 mmol·L⁻¹ MHET-OH was added to start the reaction. The control reaction was prepared by replacing the crude enzyme with a supernatant from the expression vector pET22b. The fluorescence intensity was monitored at λ_Ex = 320 nm and λ_Em = 400 nm at 30 °C for 240 min [63].

2.4. Construction of β-sheet-SSM libraries of IsPETase

The SSM libraries were generated with the standard polymerase chain reaction (PCR) protocol using the plasmid pET22b-PETase wild-type (WT) gene [53] as a template. The degenerate codons NNK and MNN (5’→3’) were used for forward and reverse primer design, respectively. The PCR system and procedure were prepared following the instructions in the manual. After the DNA templates of the PCR products were digested with QuickCut Dpn I (Takara Bio Inc., China), the products were transformed into Escherichia coli (E. coli) BL21-Gold (DE3) competent cells. The positive clones were chosen for further investigation.

2.5. Cultivation and expression of β-sheet-SSM libraries of IsPETase in 96-well plates

The cultivation and expression of the IsPETase β-sheet-SSM libraries were conducted in 96-well MTP plates as previously described [53], except that the crude enzyme was collected by centrifugation at 4 °C for 20 min.

2.6. Screening of β-sheet-SSM libraries of IsPETase in 96-well MTP plates with the MHET-OH assay

Screening of β-sheet-SSM libraries of IsPETase was conducted with the developed substrate MHET-OH. The screening for improved hydrolytic activity was performed as described previously [63], except that bis (2-hydroxyethyl) 2-hydroxyterephthalate (BHET-OH) was replaced with MHET-OH (2.5 mmol·L⁻¹). The activity of PETase was represented by the linear increase in the fluorescence intensity per minute (Flu·min⁻¹). The activity ratio of variants to WT represents relative activity. For thermostability screening, 120 μL of Na₂HPO₄–NaH₂PO₄ buffer (0.1 mol·L⁻¹, pH 8.0) and 30 μL of the expression supernatant were mixed in a 96-well black MTP and then heated at 47/50 °C for 10 min. After cooling at room temperature for 10 min, 2.5 mmol·L⁻¹ MHET-OH was injected to initiate the reaction [53], [63]. Subsequently, the same procedure as above was applied for the activity test in this case. The residual activity (%) was represented by the ratio of the activity with heat treatment to that without heat treatment.

2.7. Rescreening of IsPETase variants in shake flasks with the MHET-OH assay

Cultivation and expression in shake flasks were carried out as previously reported [53]. The relative activity and residual activity were measured with the MHET-OH assay according to the above procedure.

2.8. Purification of IsPETase WT and variants

The expression of PETase WT and variants, the collection and concentration of crude enzyme, and the purification were performed as described previously [63]. Sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was used to determine the purity of the target protein. A BCA protein assay kit (Genstar, China) was used to quantify the concentration of the target enzyme solution.

2.9. Determination of apparent melting temperature

The apparent melting temperatures (T_m) of PETase WT and its variants were measured with a protein thermal shift dye kit (Thermo Fisher Scientific, USA). The reaction system was prepared according to the manufacturer’s instructions. An ABI 7500 FAST real-time polymerase chain reaction system (Applied Biosystems, USA) was used. A charge-coupled device (CCD) was used for monitoring the fluorescent signal changes resulting from protein heat denaturation. The first derivative curve was used for identifying the T_m values.

2.10. Long-term depolymerization of amorphous Goodfellow amorphous PET film

One square piece of 0.8 cm × 0.8 cm commercial Goodfellow amorphous PET (gf-PET) film was prepared and soaked with 60 µL of purified enzyme (0.5 mg·mL⁻¹) in 2940 µL of Na₂HPO₄–NaH₂PO₄ buffer (0.1 mol·L⁻¹, pH 8.0) [63]. Subsequently, the depolymerization reaction was performed at 37 and 50 °C for 7 d. The reaction solution was removed at 1, 5, and 7 d and boiled for 10 min to terminate the reactions. Each sample was diluted to the linear detection range of TPA and MHET and then filtered through a 0.22 µm filter for HPLC analysis [63]. Triplicates for each experiment were used to calculate the average.

2.11. Scanning electron microscopy analysis of the surface morphology and thickness of the gf-PET film

After 10 days of treatment with the WT and variant PETases, the gf-PET films were sequentially washed with 1% SDS solution, ethanol, and distilled water for 20 min. After drying, the surface morphology and thickness were observed by a SU8010 scanning electron microscope (SEM; Hitachi, Japan) operating at a beam-accelerating voltage of 5 kV under high vacuum.

2.12. Complete depolymerization assay using gf-PET and untreated pc-PET

Square gf-PET and two types of pc-PET films with a size of 0.8 cm × 0.8 cm were prepared and soaked in 36 μL of 0.5 mg·mL⁻¹ PETase WT or its variants in 2964 μL of glycine-NaOH buffer (0.1 mol·L^-1, pH 9.0). The depolymerization reaction was performed at 50 °C. Fresh enzymes and buffers were added every day. The reaction mixture of each day was diluted to the linear detection range of TPA and MHET for quantitative analysis. All experiments were conducted in triplicate.

2.13. Comparison of depolymerization performance of DepoPETase β (V20), FAST-PETase, and DepoPETase toward gf-PET films at 50–60 °C.

Round gf-PET films (0.6 cm in diameter) were prepared and soaked with DepoPETase β (V20), FAST-PETase and DepoPETase (60 μL, 0.5 mg·mL ⁻¹) in glycine-NaOH buffer (2940 μL, 0.1 mol·L⁻¹, pH 9.0). The reactions were carried out at 50, 55, and 60 °C. Fresh enzyme and solution buffer were replenished after 24 and 72 h. The samples were collected every day. The collected samples were diluted to the linear detection range of TPA and MHET before quantitative analysis. All experiments were conducted in triplicate.

2.14. Complete depolymerization of pc-PET waste in a 3 L bioreactor

A total of 100.5 g of pc-PET (PET packaging from a local supermarket) was prepared as rectangular flakes (about 0.5 cm × 1 cm) and then added to a 3 L bioreactor (New Brunswick BioFlo 115, Eppendorf, Germany) containing glycine-NaOH buffer (2 L, 0.1 mol·L⁻¹, pH 9.0) [63]. In total, 260 mg of purified DepoPETase β (0.26% W_enzyme/W_PET, W: weight) was added in batches during the depolymerization process. Temperature regulation at 50 °C was performed by water bath immersion, and constant agitation at 500 r·min⁻¹ was controlled by a single marine impeller. The pH of the reaction solution was adjusted to 8.6 (±0.2) by adding NaOH solution (2 mol·L⁻¹) via a peristaltic pump. The reaction solution was collected at different time points and diluted to the linear detection range of TPA and MHET for HPLC analysis.

2.15. HPLC analysis

The quantitative analysis of the products released from the PET samples was conducted as previously reported [63]. The mobile phase consisted of buffer A (Na₂HPO₄–NaH₂PO₄ buffer, 20 mmol·L⁻¹, pH 7.0; 70%) and buffer B (acetonitrile; 30%). The elution conditions were a linear gradient of 30% to 70% buffer B. TPA and MHET were detected at 260 nm [63]. Calibration curves for TPA and MHET were constructed by plotting the corresponding peak areas versus the concentration of standard solutions, the coefficient of determination (R), R² values of the plots were at least 0.998. The samples were diluted if the concentration of the products exceeded the range of the calibration curves.

2.16. Determination of the crystallinity of the PET samples via differential scanning calorimetry

The crystallinity of the PET materials was detected with a differential scanning calorimetry instrument (DSC; DSC 250, TA Instruments, USA), and the crystallinity was calculated as described previously [63]. The results are shown in Table S1 in Appendix A.

2.17. Consensus analysis

An amino acid sequence alignment was performed by ClustalW toward sequences with 40%–100% identity with IsPETase, which were acquired from the national center for biotechnology information (NCBI) database^† [61]. Subsequently, the sequence alignment was visualized as a logo using the WebLogo webpage^‡ [79]. Multiple sequence alignment and the crystal structure of PETase WT (PDB ID: 5XJH) [80] were used to construct evolutionary conservation profiles of the β-sheet regions of PET hydrolases via the online tool ConSurf^§ [81], [82], [83]. Structure-mapping was performed by the PyMOL program (China) [84].

2.18. Energy calculations

Energy calculations with FoldX [85] were performed with the crystal structure of IsPETase (PDB ID: 5XJH). All 58 residues located on the β-sheet region of PETase were mutated to 19 other proteinogenic amino acids in silico [61]. The relative folding free energy changes (ΔΔG^Fold) were calculated based on the equation below[61]:

Δ Δ G^{F o l d} = Δ G_{s u b s t i t u t i o n}^{F o l d -} Δ G_{W T}^{F o l d}

The difference in the free energy between the folded and unfolded structures represents the ΔG^Fold [61], which was predicted by the FoldX algorithm [85]. The standard parameters were set except for temperature, which was set at 50 °C, and the average value of five calculations was used.

2.19. Crystallization, data collection, structure determination, and refinement

All crystallizations of the PETase variants were carried out using the vapor-diffusion method (25 °C). One microliter of PETase V16 (IsPETase^{D283R/V84L/N233K/F229Y/R280E/F201I/E204Q}) or PETase V20 (IsPETase^{D283R/V84L/N233K/F229Y/R280E/F201I/R53Q/D186H}) solution (30 mg·mL⁻¹ in 25 mmol·L⁻¹ tris(hydroxymethyl)aminomethane hydrochloride (Tris-HCl), pH 7.5, 150 mmol·L⁻¹ NaCl) was mixed with an equal volume of reservoir solution in sitting drop crystallization plates and equilibrated against reservoir solution (100 μL). The following crystallization conditions were used for PETase V16: 0.1 mol·L⁻¹ N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (HEPES), pH 7.5, 8% v/v ethylene glycol, 10% w/v PEG 8000, and 0.2 mol·L⁻¹ (NH₄)₂SO₄. For PETase V20 the following conditions were applied: 0.2 mol·L⁻¹ potassium fluoride (KF) pH 7.2, and 20% w/v PEG 3350. The optimal crystals reached suitable sizes for X-ray diffraction data collection within 5 d. X-ray diffraction datasets were tested and collected at beamlines BL02U1 (BL17U), BL10U2, BL17B, BL17UM, BL18U1, and BL19U1 of the National Facility for Protein Science in Shanghai (NFPS) at the Shanghai Synchrotron Radiation Facility (SSRF). Protein crystals were soaked in cryoprotectant solution and mounted onto a cryoloop before data collection at 100 K. All of the diffraction images were processed by using HKL2000 [86], and the structures were solved by molecular replacement with Phaser [87] in the program suite Phenix [88], and the structure of the IsPETase from Ideonella sakaiensis 201-F6 (PDB ID: 5XG0) [57] was used as the search model. Model building and further rounds of structure refinements were performed by using the program phenix.refine [89] and Coot [90]. Five per cent of the reflections were randomly selected for the calculations of R_free, a statistical quantity for assessing crystal structure accuracy [91]. Table S5 in Appendix A shows the diffraction data collection and structure refinement statistics. PyMOL was used to prepare all of the figures [84].

2.20. Molecular docking

Molecular docking was performed using Schrödinger 2018 software (Schrödinger, USA). The crystal structure of PETase WT (PDB ID: 5XJH) [80] was used as the receptor protein. 4-Poly PET (4PET) was docked into three systems: WT at binding Site 1 (WTB1), WT at binding site 2 (WTB2), and R53Q at binding Site 2 (Q53B2). Binding Site 1 is the catalytic site (centered at S160). Binding Site 2 is the binding region where residue R53 is located and is centered on residue F55. The length of the binding box was set to 35 × 35 × 35 (Å). Induced fit docking in Schrödinger was used to dock the 4PET, receptor, and ligand van der Waals scaling was set to 0.30, and the side chains of residues approximately 5.0 Å from the ligand pose were refined. Up to 20 docking poses were generated for each system, and these poses were clustered by interaction analysis. The best docking pose of each cluster was refined with the binding pose metadynamics. Finally, the best docking pose of each system was selected for molecular mechanics-generalized born surface area (MM-GBSA) binding energy calculations with Amber20 (USA).

2.21. Molecular dynamics simulation

All variants were built using the mutate module of Schrödinger software. The molecular dynamics (MD) simulation was performed with Amber20 software. The Protein.ff19SB and Gaff2 force fields were used to construct the simulation systems. For the 4PET binding energy calculation, a 50 ns MD simulation was performed at 310 K for all three systems (WTB1, WTB2, and Q53B2). MM-GBSA calculations were performed to evaluate the binding energy.

For structural stability analysis, 1000 ns MD simulations were performed for all eight systems (R53Q, V84L, F201I, F229Y, N233K, R280E, D283R, and DepoPETase β). The first 100 ns of the simulation was set at 310 K to allow the variant a certain amount of time to achieve the conformational changes caused by the amino acid substitutions, and the next 900 ns of the simulation was set at 370 K. Hydrogen mass repartitioning (HMR) simulation with 4 fs integration time was used to examine the changes in structural stability at high temperatures. We initiated energy minimization steps to resolve any unfavorable interactions within the initial structures. Throughout this step, the protein backbone was restrained using harmonic force constants of 20 kcal·mol ⁻¹. This minimization step comprised 1500 steps of steepest descent minimization, succeeded by 1500 steps utilizing the conjugate gradient approach. Subsequent to energy minimization, the system underwent gradual heating to 310 K over 500 ps of molecular dynamics with a constraint of 5 kcal·mol⁻¹, employing a time step of 2 fs for integration. Following the heating phase, the structures were equilibrated under constant number of particles, pressure, and temperature (NPT) conditions for 500 ps without any constraints. Periodic boundary conditions were established throughout the system with a TIP3P water model [92]. For the long-range components of the electrostatic interactions, we utilized the particle mesh Ewald (PME) method in conjunction with the Amber software suite [93]. In the molecular dynamics simulations, bonds involving hydrogen were constrained using the SHAKE algorithm [94]. Temperature control was achieved through the Langevin thermostat, which has a collision frequency of 2 ps⁻¹, while pressure was regulated by an anisotropic Monte Carlo barostat [95]. The simulations were expedited using the graphics processing unit (GPU)-accelerated version of PMEMD [96] on NVIDIA GeForce RTX 30 Series graphics cards.

2.22. Root mean square fluctuation calculation

Root mean square fluctuation (RMSF) values were calculated using an HMR MD trajectory of 900 ns (370 K) for each system, with 225 000 frames per system. To approximate the reference structure of the backbone for comparison, the initial structure of each system was used for the reference structure of the average coordinate dataset calculation. The initial structures are those conformations obtained after the minimization step using AMBER.

3. Results and discussion

3.1. Establishment of a new high-throughput screening system based on MHET-OH

We previously demonstrated that the BHET-OH assay based on the BHET-OH, an analog of BHET, can be used for the engineering of PETase [63]. Following a similar principle, we designed and synthesized another new compound, MHET-OH, which has an additional hydroxyl group on the aromatic ring compared to MHET (Fig. S1 in Appendix A), for the systematic engineering of the β-sheet region of PETase. MHET-OH is cleaved at the ester bond by PETase, releasing TPA-OH, which is thereby quantified for assessing the activity of PETase (Fig. 2(a)). The advantage of the MHET-OH assay over the BHET-OH assay is that only one ester bond of MHET-OH is cleaved for the release of fluorescent TPA-OH instead of double hydrolytic cleavage when using BHET-OH as substrate. Therefore, the MHET-OH assay allows for more prompt monitoring of the hydrolysis procedure of PETase.

We first investigated the excitation and emission spectra of the MHET-OH substrate and TPA-OH product. As shown in Fig. 2(b), the maximum emission wavelength of MHET-OH shifted by 30 nm to the right compared to that of TPA-OH when excited at 320 nm. Since the fluorescence intensity of TPA-OH can be clearly distinguished from that of MHET-OH at 400 nm, the excitation (λ_Ex) and emission (λ_Em) wavelengths used for the detection of TPA-OH were 320 and 400 nm, respectively. The fluorescent signal response shows a linear correlation (R² = 0.98) with increasing concentrations of TPA-OH and MHET-OH ranging from 0 to 2.5 mmol·L⁻¹ at λ_Ex = 320 nm and λ_Em = 400 nm (Fig. S2 in Appendix A). Subsequently, we detected the activity of WT PETase toward MHET-OH by monitoring the continuous increase in the fluorescence signal. As shown in Fig. S3 in Appendix A, the fluorescence intensity increases as the hydrolysis reaction proceeds, confirming the applicability of the developed MHET-OH assay. We further measured the residual activity of PETase WT after heat treatment at 45, 47, and 50 °C. The WT exhibited 63.3%, 38.2%, and 5.95% residual activity after being incubated at 45, 47, and 50 °C for 10 min, respectively (Fig. S4 in Appendix A). Heat treatment at 47 °C is suitable because it enables a suitable room for the improvement of thermostability, while 45 °C is slightly gentle, and 50 °C is too harsh for PETase WT. Therefore, heat treatment at 47 °C was conducted for thermostability screening.

3.2. Engineering of the β-sheet region of PETase for enhanced hydrolytic activity and thermostability

SSM libraries of 58 amino acid residues that form the β-sheet region and its vicinity to PETase were constructed. Two hundred clones were screened for each SSM library using the MHET-OH assay (Fig. S5 in Appendix A). After the first screening of about 12 000 clones and rescreening of 157 clones, 21 amino acid substitutions located at ten positions with improved hydrolytic activity or thermostability were identified, as shown in Fig. 3(a). Among them, 14 substitutions (R53A/Q/H, S54T, T56E, E204Q, N233S/M/F/G/L, and R280Y/E/F) exhibited enhanced activity toward MHET-OH at 30 °C, ranging from 1.1–1.7 fold that of the PETase WT. Moreover, 19 thermostable substitutions, including R53A/Q/H, S54T, T56E, V84L, F201I, E204Q, F229Y, N233K/S/M/G/F/L, R280S/A/E, and D283R, exhibited increased T_m values of 1.3–10.0 °C compared with that of the WT (46.7 °C; Fig. 3(b)). Among them, the substitutions S54T, T56E, F201I, and R280S/A maintained 38.1%–64.5% activity, while the WT maintained 28.5% after heat treatment at 47 °C for 10 min, and the substitutions R280E, V84L, E204Q, F229Y, N233K/S/M/G/F/L, and D283R maintained 12.4%–85.2% of their activity, while the WT almost completely lost its activity after heat treatment at 50 °C for 10 min (Fig. S6 in Appendix A). Subsequently, we analyzed the lethality rates of each SSM library to evaluate the effects of these 58 positions of the β-sheet region on the activity and thermostability of PETase. As shown in Fig. 3(c), the 58 amino acid positions were divided into three groups according to the lethality rate of the SSM library. Group 1 (red triangles) comprised amino acid positions with lethality rates lower than 50% both with and without heat treatment. Most of these positions showed similar levels of lethality with and without heat treatment, except for E204, E231, and A152. The lethality rate of Group 1 was relatively low compared to that of Group 2 and Group 3, indicating that most of the residues of Group 1 are flexible for amino acid substitutions and are less likely to negatively affect enzyme activity or thermal stability. Moreover, 60% of the identified beneficial positions, including R53, T56, E204, F229, N233, and R280, belonged to Group 1. Most of the residues in Group 2 (yellow circles) exhibited a less than 50% lethality rate before heat treatment but a far greater than 50% lethality rate after heat treatment, indicating that most of these residues are more conserved in terms of thermostability and less conserved in terms of activity. Notably, two identified substitutions (S54T and D283R) with improved thermal stability belong to Group 2. Group 3 (green squares) includes the positions with a high lethality rate (> 60%) before heat treatment and similar or even higher lethality values after heat treatment, which implies that most of the residues in Group 3 are dominantly conserved in activity and less conserved in thermostability. The identified substitutions F201I and V84L belong to Group 3, and their activities are 8.6% and 29.8% lower, respectively than that of the WT, while their T_m values are 4.3 and 3.8 °C higher than that of the WT, respectively. The sequence consensus analysis (Figs. 3(d) and Fig. S7 in Appendix A) illustrates the variability of the residues of the β-sheet region, which is consistent with the lethality rate analysis. Moreover, FoldX analysis of these 58 positions of the β-sheet region showed that 57% of our identified substitutions showed reduced ΔG^Fold, while 43% showed increased ΔG^Fold, indicating partial consistency between the experimental data and FoldX analysis (Fig. S8 in Appendix A). Overall, these results provide comprehensive insight into the mutability and contribution of the amino acid residues of the β-strand region of PETase, which is valuable for expanding the field of PETase engineering.

After the identification of single amino acid substitutions of the β-sheet region, iterative recombination of the 21 substitutions at ten positions was performed to yield further improved variants with a sequential recombination strategy (Fig. 4(a)) and K-means clustering algorithm-guided combination strategy in parallel (Supplementary Text and Fig. S9 in Appendix A). As the thermostability further increased in the recombination step, the screening pressure was gradually increased by increasing the heat treatment temperature (from 47 to 60 °C) or time (10–25 min) in the heat inactivation step to assess the combinatorial variants. The sequential recombination generated several excellent variants for further characterization, and the recombination process is shown in Fig. 4(b). The four most stabilizing substitutions, N233K, D283R, F229Y, and V84L, were first selected for a systematic recombination study, generating all 11 possible combinational variants, V1–V11 (Table S1 in Appendix A). Although V11 exhibited about 50% reduced hydrolytic activity, it (V84L/F229Y/N233K/D283R) has an increased T_m value of 17.1°C and retained about 94% of its activity after heat inactivation at 47 °C for 15 min or even more rigorous conditions (55 °C for 15 min; Table S2 in Appendix A). It is advisable to select variants with higher thermostability as templates for subsequent accumulation to circumvent enzyme unfolding [61]. Therefore, V11 was considered a template for the second round of recombination. The remaining 10 substitutions (R53A/H, S54T, T56E, F201I, E204Q, and R280F/A/S/E) were combined with V11 and resulted in the best hit, V12 (V11 + R280E), which exhibited slightly decreased thermostability but 100% recovery of activity in comparison with V11. Subsequently, V12 was selected as the new template for third-round recombination, combined with the substitutions at the remaining five positions, R53A/H/Q, S54T, T56E, F201I, and E204Q. In the third round, variant V13 (V12+F201I) showed the highest thermostability (95.0% of residual activity after heat treatment at 60 °C for 10 min) but reduced activity by higher than 50% compared to V12, and variants V14 (V12 + S54T) and V15 (V12 + R53H) exhibited slightly improved thermostability (by 9.1%) and hydrolytic activity (by 18.8%) compared with that of V12, respectively. Considering both the thermostability and activity, V13, V14, and V15 served as templates for the next stage. In the fourth round, the substitutions E204Q, R53Q/A/H, T56E, and S54T were introduced into V13, generating six V13-derived variants that exhibited slight decreases in residual activity (1.6% to 14.8%) after heat treatment (60 °C for 20 min) and up to 34.3% greater activity than that of V13. Two variants, V16 (V13 + E204Q), which has slightly decreased thermostability and activity, and V17 (V13 + R53Q), which has 34.3% greater activity than that of V13, were chosen for further characterization in this study. However, introducing substitutions (R53A/Q, T56E, and E204Q) to V14 significantly reduced the thermostability (12.5% to 39.4% after heat treatment at 60 °C for 20 mins). The introduction of the substitutions T56E, S54T, and E204Q to V15 increased the residual activity by 1.8–2.3 fold. In particular, variants V18 (V15 + T56E) and V19 (V15 + S54T) showed 100% recovered activity. Moreover, the T_m values of representative combinatorial variants (V11, V13, V14, V16–V19) in the key stages of recombination increased by 16.9–18.6 °C compared to those of the WT (Fig. 4(c)), which is mostly consistent with the results of the heat inactivation experiments shown in Fig. 4(b). Subsequently, V17, which had the greatest increase in the T_m value (ΔT_m = 18.6 °C), was combined with the substitution D186H, which has been reported previously to increase thermostability and hydrolytic activity [60], [61], [72], [73]. The MHET-OH assay indicated that the resulting variant V20 showed 100% recovery activity as did the WT at 30 °C and 30.7% greater residual activity than did V17 after heat treatment at 60 °C for 25 min. The T_m value of V20 increased by 4.3 °C compared to that of V17 (Figs. 4(b) and (c)).

To cover as many recombinations as possible, K-means clustering algorithm-guided [86] recombination was performed in parallel to sequential recombination (Fig. S9 in Appendix A). As shown in Fig. S9(a), the 21 single substitutions were grouped into three clusters (Cluster 1: D283R, N233K/S/L/M/G/F, and E204Q; Cluster 2: V84L, F201I, R53A, and R280A/Y/F/S/E; Cluster 3: R53Q/H, S54T, T56E, and F229Y) by a set of parameters, including experimentally verified relative activity, residual activity, the expression level of substitutions, interactions between substitutions (such as hydrogen bonds and salt-bridge interactions), the location of substitutions as determined by the distance between C_α atoms of the substitutions and the catalytic triads, whether it is near the PET-binding pocket and potential effects on the structure (Table S3 in Appendix A). However, after five rounds of recombination, all the resulting variants exhibited significantly lower thermostability (residual activity reduced by 17.9% to 81.5%) than V11, which was obtained from the sequential recombination strategy performed in parallel. Therefore, the recombination strategy based on the K-means algorithm was terminated (Fig. S9(c)). The variants obtained with the sequential recombination strategy were subjected to further characterization.

3.3. Depolymerization performance of PETase variants toward gf-PET

After obtaining the variants with significantly improved properties, we further investigated the depolymerization performance of the key variants, including V11, V13, V16, V17, and V20, toward commercial amorphous PET film (gf-PET, from the supplier Goodfellow) in a 3 mL reaction system without controlling the pH at 37 °C and 50 °C. Although V17 showed a 0.4 °C higher T_m than V16 (Fig. 4(c)), its depolymerization performance was lower than that of V16, as shown in Fig. 5. As shown in Fig. 5(a), the concentration of the total released products mediated by the PETase variants increased by 5.9–12.5 fold at 37 °C and by 139.2–179.4 fold at 50 °C compared with that of the PETase WT after 7 d of incubation. Notably, at 50 °C, the five variants showed 3.5–6.9 fold increased activity compared with their activity at 37 °C, while the WT exhibited only 26% of its activity at 37 °C. Overall, V20 exhibited the highest PET depolymerization performance. These results further verified the remarkably enhanced thermal resistance of the identified variants. Moreover, clear distinctions between the surfaces of gf-PET films treated with the PETase variants and WT were observed even at the visual level. The surface morphology of the gf-PET films treated with the variants V16 and V20 was rougher and less transparent than that of the films treated with the WT both at 37 and 50 °C (Fig. S10 in Appendix A). Scanning electron microscopy (SEM) was carried out to analyze the surface morphology in detail. As shown in Fig. 5(b), severe erosion occurred on the surface of the V16/V20-treated gf-PET films with significantly increased and distensible holes at both 37 and 50 °C, while some tiny pores on the surface of the WT-treated gf-PET films were observed at 37 °C but not at 50 °C. Moreover, the thickness of the enzyme-treated gf-PET films was measured by SEM. The thickness of the gf-PET treated with V16 and V20 was reduced by 30.6% at 37 °C and by 25.6%–57.5% at 50 °C, respectively, compared with that of the buffer-only control, while only a 4.3% reduction in thickness was observed at 37 °C for the WT (Table S4 in Appendix A).

3.4. Depolymerization of untreated pc-PET wastes

Apart from the commercial PET samples commonly used in the literature, it is crucial to explore the depolymerization performance of the variants for untreated pc-PET wastes. Enzymatic depolymerization of the amorphous gf-PET film and two types of pc-PET, including a packaging box of drumsticks from a local supermarket with 7.6% crystallinity and the lid of soda cup from KFC with 4.8% crystallinity (Table S5 in Appendix A), was carried out in a 3 mL reaction system at 50 °C under the optimized reaction conditions. All PET films were cut into squares (0.8 cm × 0.8 cm). In addition, the depolymerization performances of the purified variants V16 and V20 were compared with that of the reported well-characterized FAST-PETase (Fig. S11 in Appendix A), which showed superior depolymerization performance [62]. As shown in Figs. 6(a)–(c) [62] and Fig. S12 in Appendix A, variants V16, V20, and FAST-PETase all achieved complete depolymerization of the gf-PET film (6.2 g·L⁻¹), the drumsticks packaging box (6.9 g·L⁻¹) and the lid of soda cup (8.8 g·L⁻¹) within 4–7 d, while the PETase WT did not achieve complete depolymerization under the same conditions. Moreover, the complete depolymerization time of variant V16 was relatively longer than that of V20 and FAST-PETase. V20 showed comparable activity to that of FAST-PETase and produced 364.1, 305.3, and 72.3 fold more products than PETase WT when depolymerizing gf-PET, the drumstick packaging box, and the lid of the soda cup, respectively. Notably, further comparison of the depolymerization performance showed that V20 exhibited greater activity than FAST-PETase [62] and DepoPETase [63] at 55 and 60 °C, as shown in Fig. S13 in Appendix A. After 2 d of incubation, although the total amount of depolymerization products released by V20, FAST-PETase, and DepoPETase at 55 °C was similar, V20 exhibited a 1.7-fold greater initial reaction rate than FAST-PETase and DepoPETase. At 60 °C, V20 released 6.6-fold and 4.4-fold more products than did FAST-PETase and DepoPETase, respectively, after 4 d of incubation. Moreover, V20 showed a 3.5 °C higher T_m than DepoPETase, suggesting that V20 exhibits superior thermal stability. Interestingly, when we increased the substrate loading from 5 mg/3 mL to 150 mg/3 mL, V20 functioned slightly better at 50 °C than at 55 °C. The possible reason could be that a higher substrate loading requires hydrolysis for a longer time and that 50 °C is more ideal for DepoPETase β, even though it exhibited a higher initial reaction rate at 55 °C than at 50 °C (Fig. S14 in Appendix A). The results above inspired us to further investigate the application of the variant PETase V20, which was named DepoPETase β.

To further demonstrate the applicability of DepoPETase β in the PET waste depolymerization process, we performed the liter-scale depolymerization of untreated pc-PET (packaging of snacks from a local supermarket, 100.5 g) with the pH regulated automatically in a 3 L bioreactor at 50 °C. The initial reaction volume was 2 L. As shown in Figs. 6(d) and (e), complete depolymerization of the PET flakes was achieved by DepoPETase β in 96 h with an enzyme loading of 0.26% (W_enzyme/W_PET). The amount of released product increased linearly in the initial 72 h, and 83% of the products were obtained. After 96 h, although the sum of the MHET and TPA did not increase further, the ratio of TPA to MHET changed dynamically. The proportion of TPA increased from 79.2% at 96 h to 96.6% at 144 h, indicating that DepoPETase β further hydrolyzed the hydrolysis product MHET into TPA at the late stage of depolymerization, which can simplify the subsequent purification and recycling of depolymerization products for practical application. These results demonstrated that DepoPETase β is a promising catalyst for enzymatic PET waste depolymerization.

3.5. Molecular mechanism of the enhanced property of DepoPETase β

To elucidate the molecular mechanism of the enhanced PET depolymerization performance of the variants, the crystal structures of variants V16 (PDB ID: 8H83) and DepoPETase β (PDB ID: 8J5N) were solved with resolutions of 1.93 and 1.95 Å, respectively (Table S6). According to the crystal structure and MD analysis, DepoPETase β showed significantly decreased RMSF values in loop B (116–119), loop C (185–190), loop E (237–244), and loop F (276–281) compared to those of the WT, as well as a minor decrease in other regions (Fig. 7). The significantly decreased RMSF value indicates that DepoPETase β has overall greater structural stability than the WT. Particularly, DepoPETase β exhibits significant improvements in stability in loop E, where the catalytic residue H237 is located, and in loop C, which forms the substrate binding pocket. The enhanced stability of loop E and loop C will improve the structural integrity of the catalytic triad, thereby improving the depolymerization performance of PETase.

We performed MD of the single substitutions R53Q, V84L, F201I, F229Y, N233K/F, R280E, D283R, and DepoPETase β to analyze the structural changes caused by each amino acid substitution. As shown in Fig. 7(c), when Asp283 of β9 was substituted with Arg, the nearby single salt bridge (D283–R285) located on β9 became a double salt bridge composed of E231–R283 between β8 and β9 and D220–R285 between α5 and β9. In addition, a hydrogen bond was formed between N212 of α5 and R283 of β9. These changes stabilized loop E containing the catalytic residue H237 (Fig. S15(a) in Appendix A) while also eliminating the electrostatic repulsion between E231 of β8 and D283 of β9, resulting in a triple benefit. When Phe at 229 of β8 is substituted with Tyr, E231 can form hydrogen bonds with both Y229 and S282. The introduced hydrogen bonds stabilized E231 and the hydrogen bond network of S282. The construction of this hydrogen bond network led to adjustments in the relative positions of sheets β7, β8, and β9, especially at the beginning of β9, where S282 is located. The distance between the beginning of β9 and the end of β7 decreased from 9.0 Å in the WT to 7.8 Å in the variant DepoPETase β (Fig. 7(d)). The shortened distance was beneficial for the stability of loop E (connected with β8) and loop F (connected with β9), which was reflected by their significantly reduced RMSF values (Fig. S15(b)). As shown in Fig. 7(e), the substitution R280E utilizes a water molecule to form a bridging hydrogen bond with N246. Moreover, R280E can also form a hydrogen bond with T279. This newly introduced hydrogen bond network enhances the interaction between loop E and loop F, thereby increasing the structural stability of loop E. This can be observed from the RMSF analysis shown in Fig. S15(c), in which there is a moderate reduction in loop E and only a slight reduction in loop F. The large and rigid hydrophobic residue F201 located in β7, together with the surrounding hydrophobic residues, which are mainly composed of Leu and Ile, prevent α6 from fully contacting β7. In contrast, substitution F201I allowed α6 to move closer to β7, resulting in tighter packing of PETase (Fig. 7(f) and Fig. S15(d)). With respect to the V84L substitution, there were two voids around residue V84 in the WT, which affected the strength of the interaction between β4 and α3, leading to insufficient stability of loop B. By substituting Val with Leu at residue 84, the two gaps disappeared (Fig. 7(g)). Compared with that of the WT, the RMSF of residues 113-119 of variant V84L in loop B decreased, indicating stronger structural stability (Fig. S15(e)). Substitution of N233K adjacent to the edge of β8 resulted in a salt bridge between E204 and K233, neutralizing the charges, maintaining the stability of the hydrogen bond between E231 and S282, and resolving all unfavorable effects, such as electrostatic repulsion between E204 and E231 (Fig. 7(h)). N233K was also reported by Lu et al. [62] during our independent study period. Moreover, we observed that substitutions to hydrophobic residues at 233, such as N233F/L/M, can also have a stabilizing effect (Fig. 3(b)). Taking N233F as an example, we can see that F233 occupies the space above S282, pushing E204 away and allowing E204 to form hydrogen bonds with N205 and G234. Although the increase in interaction caused by the substitution of N233F is not as significant as that caused by the substitution of N233K, N233F is still an effective substitution for reducing electrostatic repulsion.

It was observed from the crystal structures of DepoPETase β that D283R forms a new salt bridge, while V84L fills the gap that allows penetration of water molecules. Both F229Y and R280E establish hydrogen bond networks. These results are similar to those observed from the MD simulation results of single substitutions. However, N233K and F201I of the crystal structure of DepoPETase β differ from the MD simulation results of single substitutions (Fig. S16 in Appendix A). In the simulation system involving the N233K single-point substitution, N233K forms a salt bridge directly with E204 or indirectly through a water-mediated salt bridge. In DepoPETase β, K233 has no interaction with the nearby negatively charged residues E204 and E280, which we believe is inconsistent with the improvement in stability. After MD simulation of DepoPETase β, it can be observed that N233K forms a hydrogen bond with the main-chain carbonyl of T279, which strengthens the hydrogen bond network consisting of K233, N246, T279, and E280 and a water molecule. The substitution F201I did not cause visible main-chain conformational changes in the DepoPETase β crystal structure.

The substrate binding surface of PETase displays an overall positive charge, which mainly exists in two groove regions, one of which is the catalytic site of the substrate, and the other groove is located on the β1 side, both of which contain strongly positive charge regions. With respect to the R53Q substitution located on β1, the MD simulation revealed that R53 had no potential hydrogen bond or salt bridge interactions with other parts of PETase. Docking the 4PET substrate at the two groove regions of PETase WT showed that the MM-GBSA binding energy of 4PET at the catalytic site was (−36.15 ± 3.17) kcal·mol⁻¹, while the binding energy of 4PET at the β1 binding site reached (−32.79 ± 3.16) kcal·mol⁻¹. The similar binding energy at the two sites led to potential competition between the two binding sites. Substitution R53Q disrupted the salt bridge interaction between R53 and the carboxyl group of 4PET, which was reflected by the decreased binding energy of 4PET to the β1 site ((−29.63 ± 3.48) kcal·mol⁻¹) (Fig. 7(i)). The variation in binding energy reduces the competitive binding of the β1 site, which can increase the probability of substrate binding at the catalytic sites and thereby enhance the activity. By comparing the simulation results of the single substitutions and the crystal structure of DepoPETase β, we can see that the simulated structural changes caused by D283R, V84L, F229Y, and R280E are consistent with the crystal structure of DepoPETase β.

Overall, among the nine β-sheets, β1-2, and β8-9 are located on the surface of PETase, and their residues are partially exposed. The stability of these two regions relies on the construction of strong hydrogen bonds and salt bridges. The substitutions D283R, F229Y, R280E, and N233K contributed to enhancing the number of hydrogen bonds and salt bridges in this region. β3-7 is in the core of PETase. Therefore, increasing the hydrophobic interactions by the substitution D186H and a more tightly packed structure caused by F201I and V84L was beneficial for stabilizing the β3-7 region.

4. Conclusions

In summary, this study provides the first systematic engineering study of the whole β-sheet region of IsPETase with a high-throughput fluorescence assay. Iterative recombination of the identified beneficial substitutions resulted in the combinatorial variant DepoPETase β, which significantly improved depolymerization performance compared with that of the WT at mild temperatures. Moreover, complete depolymerization of untreated PET waste in a bioreactor was achieved by DepoPETase β, demonstrating its potential application in the enzymatic PET depolymerization process. Further mechanistic analysis of DepoPETase β revealed that the beneficial substitutions located in the outer β-sheets constructed strong hydrogen bonds and salt bridges, while the beneficial substitutions located in the inner β-sheets supported a more tightly packed structure, providing new insights into PET hydrolases. The engineering strategy for the β-sheet region is likely applicable to other PETase hydrolases that share a scaffold similar to that of PETase.

CRediT authorship contribution statement

Songfeng Gao: Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Lixia Shi: Visualization, Validation, Methodology. Hongli Wei: Writing – original draft, Investigation, Data curation. Pi Liu: Visualization, Investigation, Data curation. Wei Zhao: Investigation, Data curation. Lanyu Gong: Investigation, Data curation. Zijian Tan: Visualization. Huanhuan Zhai: Investigation. Weidong Liu: Visualization, Supervision, Investigation. Haifeng Liu: Writing – review & editing, Supervision, Conceptualization. Leilei Zhu: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was funded by the National Key Research and Development Program of China (2023YFC3903300), the Tianjin Synthetic Biotechnology Innovation Capacity Improvement Project (TSBICIP-IJCP-003, TSBICIP-KJGG-009-0203, and TSBICIP-BRFI-005), and the Innovation Fund of Haihe Laboratory of Synthetic Biology (22HHSWSS00018).

Appendix A. Supplementary material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2024.10.015.

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Welle F.Twenty years of PET bottle to bottle recycling—an overview.Resour Conserv Recycling 2011; 55(11):865-875.

[2]	Nicholson SR, Rorrer NA, Carpenter AC, Beckham GT.Manufacturing energy and greenhouse gas emissions associated with plastics consumption.Joule 2021; 5(3):673-686.

[3]	Bornscheuer UT.Feeding on plastic.Science 2016; 351(6278):1154-1155.

[4]	Webb HK, Arnott J, Crawford RJ, Ivanova EP.Plastic degradation and its environmental implications with special reference to poly(ethylene terephthalate).Polymers 2013; 5(1):1-18.

[5]	Li Q, Hai R, Lin A.Human health risk assessment on contaminants in recycled plastic bags packaged foods.Environ Sci Technol 2010; 33(11):181-185.

[6]	Tournier V, Duquesne S, Guillamot F, Cramail H, Taton D, Marty A, et al.Enzymes’ power for plastics degradation.Chem Rev 2023; 123(9):5612-5701.

[7]	Chae Y, Kim D, Kim SW, An YJ.Trophic transfer and individual impact of nano-sized polystyrene in a four-species freshwater food chain.Sci Rep 2018; 8(1):274.

[8]	Hahladakis JN, Velis CA, Weber R, Iacovidou E, Purnell P.An overview of chemical additives present in plastics: migration, release, fate and environmental impact during their use, disposal and recycling.J Hazard Mater 2018; 344:179-199.

[9]	Pabortsava K, Lampitt RS.High concentrations of plastic hidden beneath the surface of the Atlantic ocean.Nat Commun 2020; 11(1):4073.

[10]	Jambeck JR, Geyer R, Wilcox C, Siegler TR, Perryman M, Andrady A, et al.Plastic waste inputs from land into the ocean.Science 2015; 347(6223):768-771.

[11]	Ellis LD, Rorrer NA, Sullivan KP, Otto M, McGeehan JE, Roman-Leshkov Y, et al.Chemical and biological catalysis for plastics recycling and upcycling.Nat Catal 2021; 4(7):539-556.

[12]	Wei R, Tiso T, Bertling J, O K’Connor, Blank LM, Bornscheuer UT.Possibilities and limitations of biotechnological plastic degradation and recycling.Nat Catal 2020; 3(11):867-871.

[13]	Coviello CG, Lassandro P, Sabb MFà, Foti D.Mechanical and thermal effects of using fine recycled PET aggregates in common screeds.Sustainability 2023; 15(24):16692.

[14]	Lerna M, Foti D, Petrella A, Sabb MFà, Mansour S.Effect of the chemical and mechanical recycling of PET on the thermal and mechanical response of mortars and premixed screeds.Materials 2023; 16(8):3155.

[15]	Brzozowski AM, Savage H, Verma CS, Turkenburg JP, Lawson DM, Svendsen A, et al.Structural origins of the interfacial activation in Thermomyces (humicola) lanuginosa lipase.Biochemistry 2000; 39(49):15071-15082.

[16]	Parisi D, Riley C, Srivastava AS, McCue HV, Johnson JR, Carnell AJ.PET hydrolysing enzymes catalyse bioplastics precursor synthesis under aqueous conditions.Green Chem 2019; 21(14):3827-3833.

[17]	Ronkvist AM, Xie WC, Lu WH, Gross RA.Cutinase-catalyzed hydrolysis of poly(ethylene terephthalate).Macromolecules 2009; 42(14):5128-5138.

[18]	Muller RJ, Schrader H, Profe J, Dresler K, Deckwer WD.Enzymatic degradation of poly(ethylene terephthalate): rapid hydrolyse using a hydrolase from T. fusca.Macromol Rapid Commun 2005; 26(17):1400-1405.

[19]	Then J, Wei R, Oeser T, Barth M, Belisario-Ferrari MR, Schmidt J, et al.Ca²⁺ and Mg²⁺ binding site engineering increases the degradation of polyethylene terephthalate films by polyester hydrolases from Thermobifida fusca.Biotechnol J 2015; 10(4):592-598.

[20]	Kawabata T, Oda M, Kawai F.Mutational analysis of cutinase-like enzyme, Cut190, based on the 3D docking structure with model compounds of polyethylene terephthalate.J Biosci Bioeng 2017; 124(1):28-35.

[21]	Oda M, Yamagami Y, Inaba S, Oida T, Yamamoto M, Kitajima S, et al.Enzymatic hydrolysis of PET: functional roles of three Ca²⁺ ions bound to a cutinase-like enzyme, Cut190*, and its engineering for improved activity.Appl Microbiol Biotechnol 2018; 102(23):10067-10077.

[22]	Sulaiman S, Yamato S, Kanaya E, Kim JJ, Koga Y, Takano K, et al.Isolation of a novel cutinase homolog with polyethylene terephthalate—degrading activity from leaf-branch compost by using a metagenomic approach.Appl Environ Microbiol 2012; 78(5):1556-1562.

[23]	Tournier V, Topham CM, Gilles A, David B, Folgoas C, Moya-Leclair E, et al.An engineered PET depolymerase to break down and recycle plastic bottles.Nature 2020; 580(7802):216-219.

[24]	Ding ZD, Xu GS, Miao RJ, Wu NF, Zhang W, Yao B, et al.Rational redesign of thermophilic PET hydrolase LCCICCG to enhance hydrolysis of high crystallinity polyethylene terephthalates.J Hazard Mater 2023; 453:131386.

[25]	Li QB, Zheng Y, Su TY, Wang Q, Liang QF, Zhang ZD, et al.Computational design of a cutinase for plastic biodegradation by mining molecular dynamics simulations trajectories.Comput Struct Biotechnol J 2022; 20:459-470.

[26]	Yoshida S, Hiraga K, Takehana T, Taniguchi I, Yamaji H, Maeda Y, et al.A bacterium that degrades and assimilates poly(ethylene terephthalate).Science 2016; 351(6278):1196-1199.

[27]	Chen ZZ, Wang YY, Cheng YY, Wang X, Tong SW, Yang HT, et al.Efficient biodegradation of highly crystallized polyethylene terephthalate through cell surface display of bacterial PETase.Sci Total Environ 2020; 709:136138.

[28]	Chen K, Hu Y, Dong XY, Sun Y.Molecular insights into the enhanced performance of ekylated PETase toward PET degradation.ACS Catal 2021; 11(12):7358-7370.

[29]	Liu K, Xu ZP, Zhao ZY, Chen YX, Chai YT, Ma L, et al.A dual fluorescence assay enables high-throughput screening for poly(ethylene terephthalate) hydrolases.ChemSusChem 2023; 16(5):e202202019.

[30]	Feng SS, Yue Y, Zheng MN, Li YW, Zhang QZ, Wang WX.IsPETase and IsMHETase-catalyzed cascade degradation mechanism toward polyethylene terephthalate.ACS Sustain Chem Eng 2021; 9(29):9823-9832.

[31]	Zheng MN, Li YW, Dong WL, Zhang WX, Feng SS, Zhang QZ, et al.Depolymerase-catalyzed polyethylene terephthalate hydrolysis: a unified mechanism revealed by quantum mechanics/molecular mechanics analysis.ACS Sustain Chem Eng 2022; 10(22):7341-7348.

[32]	Ma Y, Yao MD, Li BZ, Ding MZ, He B, Chen S, et al.Enhanced poly(ethylene terephthalate) hydrolase activity by protein engineering.Engineering 2018; 4(6):888-893.

[33]	Pfaff L, Gao J, Li ZS, Jaeckering A, Weber G, Mican J, et al.Multiple substrate binding mode-guided engineering of a thermophilic pet hydrolase.ACS Catal 2023; 12(15):9790-9800.

[34]	Erickson E, Gado JE, Avilán L, Bratti F, Brizendine RK, Cox PA, et al.Sourcing thermotolerant poly(ethylene terephthalate) hydrolase scaffolds from natural diversity.Nat Commun 2022; 13(1):7850.

[35]	Sagong HY, Son HF, Seo H, Hong H, Lee D, Kim KJ.Implications for the PET decomposition mechanism through similarity and dissimilarity between PETases from Rhizobacter gummiphilus and Ideonella sakaiensis.J Hazard Mater 2021; 416:126075.

[36]	Nakamura A, Kobayashi N, Koga N, Iino R.Positive charge introduction on the surface of thermostabilized PET hydrolase facilitates PET binding and degradation.ACS Catal 2021; 11(14):8550-8564.

[37]	Eiamthong B, Meesawat P, Wongsatit T, Jitdee J, Sangsri R, Patchsung M, et al.Discovery and genetic code expansion of a polyethylene terephthalate (PET) hydrolase from the human saliva metagenome for the degradation and bio-functionalization of PET.Angew Chem Int Ed 2022; 61(37):e202203061.

[38]	White MFM, Wallace S.A new PETase from the human saliva metagenome and its functional modification via genetic code expansion in bacteria.Angew Chem Int Ed 2023; 135(12):e202216963.

[39]	Hong H, Ki D, Seo H, Park J, Jang J, Kim KJ.Discovery and rational engineering of PET hydrolase with both mesophilic and thermophilic PET hydrolase properties.Nat Commun 2023; 14(1):4556.

[40]	Li A, Sheng Y, Cui H, Wang M, Wu L, Song Y, et al.Discovery and mechanism-guided engineering of BHET hydrolases for improved PET recycling and upcycling.Nat Commun 2023; 14(1):4169.

[41]	Zhang S, Hu Q, Zhang YX, Guo H, Wu Y, Sun M, et al.Depolymerization of polyesters by a binuclear catalyst for plastic recycling.Nat Sustain 2023; 6(8):965-973.

[42]	Velasco-Lozano S, MÀGalm és, Olazabal I, Sardon H, López-Gallego F, et al.Mechanistic studies of a lipase unveil effect of pH on hydrolysis products of small PET modules.Nat Commun 2023; 14(1):3556.

[43]	Hayes HC, Luk LYP.Investigating the effects of cyclic topology on the performance of a plastic degrading enzyme for polyethylene terephthalate degradation.Sci Rep 2023; 13(1):1267.

[44]	Kalathil S, Miller M, Reisner E.Microbial fermentation of polyethylene terephthalate (PET) plastic waste for the production of chemicals or electricity.Angew Chem Int Ed 2022; 61(45):e202211057.

[45]	Ye M, Li Y, Yang Z, Yao C, Sun W, Zhang X, et al.Ruthenium/TiO₂-catalyzed hydrogenolysis of polyethylene terephthalate: reaction pathways dominated by coordination environment.Angew Chem Int Ed 2023; 62(19):e202301024.

[46]	Liu Y, Zhang C, Feng J, Wang X, Ding Z, He L, et al.Integrated photochromic–photothermal processes for catalytic plastic upcycling.Angew Chem Int Ed 2023; 62(38):e202308930.

[47]	Gopal MR, Dickey RM, Butler ND, Talley MR, Nakamura DT, Mohapatra A, et al.Reductive enzyme cascades for valorization of polyethylene terephthalate deconstruction products.ACS Catal 2023; 13(7):4778-4789.

[48]	Gong X, Ma F, Zhang Y, Li Y, Wang Z, Liu Y, et al.In situ construction of an intramolecular donor–acceptor conjugated copolymer via terephthalic acid derived from plastic waste for photocatalysis of plastic to hydrogen peroxide.ACS Catal 2023; 13(18):12338-12349.

[49]	Arnal G, Anglade J, Gavalda S, Tournier V, Chabot N, Bornscheuer UT, et al.Assessment of four engineered pet degrading enzymes considering large-scale industrial applications.ACS Catal 2023; 13(20):13156-13166.

[50]	von G Haugwitz, Han X, Pfaff L, Li Q, Wei HL, Gao J, et al.Structural insights into (tere)phthalate-ester hydrolysis by a carboxylesterase and its role in promoting PET depolymerization.ACS Catal 2022; 12(24):15259-15270.

[51]	Palm GJ, Reisky L, Böttcher D, Müller H, Michels EAP, Walczak MC, et al.Structure of the plastic-degrading Ideonella sakaiensis MHETase bound to a substrate.Nat Commun 2019; 10(1):1717.

[52]	Tarazona NA, Wei R, Brott S, Pfaff L, Bornscheuer UT, Lendlein A, et al.Rapid depolymerization of poly(ethylene terephthalate) thin films by a dual-enzyme system and its impact on material properties.Chem Catal 2022; 2(12):3573-3589.

[53]	Shi LX, Liu HF, Gao SF, Weng YX, Zhu LL.Enhanced extracellular production of IsPETase in Escherichia coli via engineering of the pelB signal peptide.J Agric Food Chem 2021; 69(7):2245-2252.

[54]	Falkenstein P, Zhao Z, Di A Pede-Mattatelli, Künze G, Sommer M, Sonnendecker C, et al.On the binding mode and molecular mechanism of enzymatic polyethylene terephthalate degradation.ACS Catal 2023; 13(10):6919-6933.

[55]	Zheng MN, Li YW, Xue R, Dong WL, Zhang QZ, Wang WX.Hydrolases catalyzed nanosized polyethylene terephthalate depolymerization: new insights from QM/MM analysis.J Clean Prod 2022; 377:134429.

[56]	Chen CC, Han X, Ko TP, Liu W, Guo RT.Structural studies reveal the molecular mechanism of PETase.FEBS J 2018; 285(20):3717-3723.

[57]	Han X, Liu W, Huang JW, Ma J, Zheng Y, Ko TP, et al.Structural insight into catalytic mechanism of PET hydrolase.Nat Commun 2017; 8(1):2106.

[58]	Wang N, Guan F, Lv X, Han D, Zhang Y, Wu N, et al.Enhancing secretion of polyethylene terephthalate hydrolase PETase in Bacillus subtilis WB600 mediated by the SPamysignal peptide.Lett Appl Microbiol 2020; 71(3):235-241.

[59]	Liu YD, Liu ZZ, Guo ZY, Yan TT, Jin CX, Wu J.Enhancement of the degradation capacity of IsPETase for PET plastic degradation by protein engineering.Sci Total Environ 2022; 834:154947.

[60]	Son HF, Cho IJ, Joo S, Seo H, Sagong HY, Choi SY, et al.Rational protein engineering of thermo-stable PETase from Ideonella sakaiensis for highly efficient PET degradation.ACS Catal 2019; 9(4):3519-3526.

[61]	Cui YL, Chen YC, Liu XY, Dong SJ, Tian YE, Qiao YX, et al.Computational redesign of a PETase for plastic biodegradation under ambient condition by the grape strategy.ACS Catal 2021; 11(3):1340-1350.

[62]	Lu HY, Diaz DJ, Czarnecki NJ, Zhu CZ, Kim WT, Shroff R, et al.Machine learning-aided engineering of hydrolases for PET depolymerization.Nature 2022; 604(7907):662.

[63]	Shi LX, Liu P, Tan ZJ, Zhao W, Gao JF, Gu Q, et al.Complete depolymerization of PET wastes by an evolved PET hydrolase from directed evolution.Angew Chem Int Ed 2023; 62(14):e202218390.

[64]	Erickson E, Shakespeare TJ, Bratti F, Buss BL, Graham R, Hawkins MA, et al.Comparative performance of PETase as a function of reaction conditions, substrate properties, and product accumulation.ChemSusChem 2022; 15(1):e202101932.

[65]	Meng X, Yang L, Liu H, Li Q, Xu G, Zhang Y, et al.Protein engineering of stable IsPETase for PET plastic degradation by premuse.Int J Biol Macromol 2021; 180:667-676.

[66]	Chen ZZ, Duan RD, Xiao YJ, Wei Y, Zhang HX, Sun XZ, et al.Biodegradation of highly crystallized poly(ethylene terephthalate) through cell surface codisplay of bacterial PETase and hydrophobin.Nat Commun 2022; 13(1):7138.

[67]	Kosiorowska KE, Moreno AD, Iglesias R, Leluk K, Mironczuk AM.Production of PETase by engineered Yarrowia lipolytica for efficient poly(ethylene terephthalate) biodegradation.Sci Total Environ 2022; 846:157358.

[68]	Guo BY, Vanga SR, Lopez-Lorenzo X, Saenz-Mendez P, Ericsson SR, Fang Y, et al.Conformational selection in biocatalytic plastic degradation by PETase.ACS Catal 2022; 12(6):3397-3409.

[69]	Yin QD, You SP, Zhang JX, Qi W, Su RX.Enhancement of the polyethylene terephthalate and mono-(2-hydroxyethyl) terephthalate degradation activity of Ideonella sakaiensis PETase by an electrostatic interaction-based strategy.Bioresour Technol 2022; 364:128026.

[70]	Achatz S, Jarasch A, Skerra A.Structural plasticity in the loop region of engineered lipocalins with novel ligand specificities, so-called anticalins.J Struct Biol X 2022; 6:100054.

[71]	Bell EL, Smithson R, Kilbride S, Foster J, Hardy FJ, Ramachandran S, et al.Directed evolution of an efficient and thermostable PET depolymerase.Nat Catal 2022; 5(8):673-681.

[72]	Brott S, Pfaff L, Schuricht J, Schwarz JN, Bottcher D, Badenhorst CPS, et al.Engineering and evaluation of thermostable IsPETase variants for PET degradation.Eng Life Sci 2022; 22(3–4):192-203.

[73]	Zhong-Johnson EZL, Voigt CA, Sinskey AJ.An absorbance method for analysis of enzymatic degradation kinetics of poly(ethylene terephthalate) films.Sci Rep 2021; 11(1):928.

[74]	Austin HP, Allen MD, Donohoe BS, Rorrer NA, Kearns FL, Silveira RL, et al.Characterization and engineering of a plastic-degrading aromatic polyesterase.Proc Natl Acad Sci USA 2018; 115(19):E4350-E4357.

[75]	Fecker T, Galaz-Davison P, Engelberger F, Narui Y, Sotomayor M, Parra LP, et al.Active site flexibility as a hallmark for efficient PET degradation by I. sakaiensis PETase.Biophys J 2018; 114(6):1302-1312.

[76]	Wang CH, Lu LH, Huang C, He BF, Huang RB.Simultaneously improved thermostability and hydrolytic pattern of alpha-amylase by engineering central beta strands of TIM barrel.Appl Biochem Biotechnol 2020; 192(1):57-70.

[77]	Lou D, Tan J, Zhu L, Ji S, Wang B.The β-sheet core is the favored candidate of engineering SDR for enhancing thermostability but not for activity.Protein Pept Lett 2017; 24(6):511-516.

[78]	Lou D, Tan J, Zhu L, Ji S, Tang S, Yao K, et al.Engineering Clostridium absonum 7α-hydroxysteroid dehydrogenase for enhancing thermostability based on flexible site and ΔΔG prediction.Protein Pept Lett 2018; 25(3):230-235.

[79]	Crooks GE, Hon G, Chandonia JM, Brenner SE.Weblogo: a sequence logo generator.Genome Res 2004; 14(6):1188-1190.

[80]	Joo S, Cho IJ, Seo H, Son HF, Sagong HY, Shin TJ, et al.Structural insight into molecular mechanism of poly(ethylene terephthalate) degradation.Nat Commun 2018; 9(1):382.

[81]	Ashkenazy H, Abadi S, Martz E, Chay O, Mayrose I, Pupko T, et al.Consurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules.Nucleic Acids Res 2016; 44(W1):W344-W350.

[82]	Landau M, Mayrose I, Rosenberg Y, Glaser F, Martz E, Pupko T, et al.Consurf 2005: the projection of evolutionary conservation scores of residues on protein structures.Nucleic Acids Res 2005; 33(Suppl 2):S299-S302.

[83]	Ashkenazy H, Erez E, Martz E, Pupko T, Ben-Tal N.Consurf 2010: calculating evolutionary conservation in sequence and structure of proteins and nucleic acids.Nucleic Acids Res 2010; 38(Suppl 2):W529-W533.

[84]	Delano WL.The pymol molecular graphics system.Proteins Struct Funct Bioinf 2002; 30:442-454.

[85]	Guerois R, Nielsen JE, Serrano L.Predicting changes in the stability of proteins and protein complexes: a study of more than 1000 mutations.J Mol Biol 2002; 320(2):369-387.

[86]	Otwinowski Z, Minor W.Processing of X-ray diffraction data collected in oscillation mode.Methods Enzymol 1997; 276:307-326.

[87]	McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ.Phaser crystallographic software.J Appl Cryst 2007; 40(Pt 4):658-674.

[88]	Adams PD, Afonine PV, Bunkóczi G, Chen VB, Davis IW, Echols N, et al.Phenix: a comprehensive python-based system for macromolecular structure solution.Acta Crystallogr D 2010; 66(Pt 2):213-221.

[89]	Terwilliger TC, Grosse-Kunstleve RW, Afonine PV, Moriarty NW, Zwart PH, Hung LW, et al.Iterative model building, structure refinement and density modification with the phenix autobuild wizard.Acta Crystallogr D 2008; 64(Pt 1):61-69.

[90]	Terwilliger TC, Grosse-Kunstleve RW, Afonine PV, Moriarty NW, Adams PD, Read RJ, et al.Iterative-build omit maps: map improvement by iterative model building and refinement without model bias.Acta Crystallogr D 2008; 64(Pt 5):515-524.

[91]	Brünger AT.Assessment of phase accuracy by cross validation: the free R value. Methods and applications.Acta Crystallogr D 1993; 49(Pt 1):24-36.

[92]	Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW, Klein ML.Comparison of simple potential functions for simulating liquid water.J Chem Phys 1998; 79(2):926-935.

[93]	Pearlman DA, Connelly PR.Determination of the differential effects of hydrogen bonding and water release on the binding of FK506 to native and Tyr8₂→Phe82 FKBP-12 proteins using free energy simulations.J Mol Biol 1995; 248(3):696-717.

[94]	Andersen HC.Rattle: a “velocity” version of the shake algorithm for molecular dynamics calculations.J Comput Phys 1983; 52(1):24-34.

[95]	Chow KH, Ferguson DM.Isothermal–isobaric molecular dynamics simulations with monte carlo volume sampling.Comput Phys Commun 1995; 91(1–3):283-289.

[96]	Grand SL, GoTz AW, Walker RC.SPFP: speed without compromise—a mixed precision model for GPU accelerated molecular dynamics simulations.Comput Phys Commun 2013; 184(2):374-380.

RIGHTS & PERMISSIONS

THE AUTHOR