Amorphous PET film (GF-PET, crystalline 6.7%) was purchased from Goodfellow Ltd. (Shanhai, China, thickness of 250 µm). Post-consumer PET waste (PCW-PET, crystalline 16%) samples were prepared from PET flakes following the pretreatment method described previously (Tournier et al., 2020). High-crystalline PET (Hc-PET, crystalline 40%) was purchased from Dopont Ltd. (Shanghai, China). All PET samples were cut into pieces then immersed in liquid nitrogen for 1 min before micronization. The percentage crystallinity of the sample is measured by the DSC method and calculated (Tournier et al., 2020). MHET is synthesized based a previously described method (Kaabel et al., 2021). All other chemicals and reagents were of analytical grade and purchased from Sahn chemical technology Co. Ltd. (Shanghai, China). Restriction enzymes, DNA polymerase were purchased from TaKaRa Co., Ltd. (Dalian, China). Escherichia coli strains were routinely cultivated aerobically in LB medium at 37°C. All constructs in this study were generated with the expression vector pET-29a for recombinant expression in E. coli BL21 (DE3). The gene encoding LCC^ICCG cutinase, CBM, PBM and HFB4 were synthesized based on previous studies (Ribitsch et al., 2013; Ribitsch et al., 2015; Tournier et al., 2020) and that for ChBD was derived from the genome of Chitinolyticbacter meiyuanensis SYBCH1 (Zhang et al., 2018). LCC^ICCG cutinase and ChBD expression strain E. coli BL21 (pET29a-LCC^ICCG) and E. coli BL21 (pET29a-ChBD) were constructed in previous experiments (Zhou et al., 2020).

The genes were constructed using the method of splicing-by-overlap extension (Su et al., 2006). Synthetic oligonucleotides encoding sequences of peptide linkers was introduced between the lcc and binding module genes. To express the fusion enzymes, the cleavage sites for the restriction enzymes Nde I and Xho I were introduce into the 5′ end of lcc and the 3′end of binding module genes, respectively (Figure 2A). The gene fragments encoding GFP was then inserted into the multi clone site (MCS) by one step cloning for further investigations.

E. coli BL21 (DE3) harboring individual plasmids including pET29a-LCC^ICCG, pET29a-LCC^ICCG-ChBD, pET29a-LCC^ICCG-CBM, pET29a-LCC^ICCG-PBM, pET29a-LCC^ICCG-HFB4 and pET29a-LCC^ICCG-ChBD-GFP was used to inoculate 5 ml LB medium containing kanamycin (50 μg/ml) for incubation at 37°C and 200 rpm for 12 h. Then, a 2 ml seed culture solution was removed to inoculate 200 ml LB medium with kanamycin (50 μg/ml) for incubation at 37°C and 200 rpm for 4 h. When the OD value reached 0.6, IPTG was added at a final concentration of 1 mM to induce the recombinant protein expression by further incubating at 18°C for 24 h. The bacterial cells were harvested by centrifugation at 12,000 rpm at 4°C for 10 min. The obtained cell pellets were washed twice by phosphoric buffer (50 mM, pH 8.0) and resuspended with 50 mM phosphoric buffer (pH 8.0). Then, the cells were disrupted by a scientz-II D ultrasonic generator and cell debris was removed by centrifugation at 12,000 rpm and 4°C for 20 min. The recombinant enzyme with a C-terminal His6-tag was purified using Ni²⁺-NTA resin with a Biomolecular Liquid Chromatography System (AKTA design, GE Healthcare, US). The target proteins were eluted with buffer NPI-250 (50 mM NaH₂PO₄, 300 mM NaCl, 250 mM imidazole, pH 8.0). The purified protein was collected and concentrated by ultrafiltration tube to remove imidazole. Esterase activity was determined using a slightly modified version of a previously described method (Alisch et al., 2004). In a cuvette, 980 μL of 50 mM PBS buffer (K₂HPO₄/NaH₂PO₄, pH 8.0) was mixed with 10 μL p-nitrophenyl butyrate (pNPB, 10 mM) dissolved in isopropanol. The hydrolysis reaction was started by adding 10 μL of appropriately diluted enzyme solution and stopped by adding 20 μL of 1% sodium dodecyl sulfonate, after which the absorbance was measured at 410 nm.

The kinetic parameters for the enzymatic hydrolysis of pNPB were calculated based on the reaction rates determined in the initial reaction phase during first 3 min reaction. The kinetic parameters for hydrolysis of amorphous PET film which was cut into pieces then immersed in liquid nitrogen for 1 min and micronization. obtained in the initial phase during a reaction time of 30 min was used to determine the reaction rates. Further, this amount was subtracted from the sum of the total soluble hydrolysis products (TPA, MHET and BHET) obtained to calculate the reaction rates.

The purified enzyme (LCC^ICCG-ChBD-GFP) and the pre-treated PET substrate were mixed thoroughly for 10 min, and centrifuged at 12,000 rpm for 10 min at 4°C. The adsorption efficiency of the binding module to the plastic substrate was calculated based on the protein concentrations determined in the supernatant before and after mixing and incubation. Experiments on this adsorption efficiency were carried out at different incubation temperatures (4, 20, 40, and 65°C) to monitor the maximum binding efficiency of different enzyme variants at the temperature where the plastic degradation was performed.

Protein concentrations were determined by the Bradford method (Campion et al., 2017). The amount of adsorbed ChBD-GFP was determined using the fluorescence intensity detected by Hitachi F7000 fluorescence spectrophotometer similarly as describe before (Cha and Kwon, 2018). The adsorption rate was calculated with the formula below: Adsorption   rate = ( 1 − GFP   concentration   in   supernatant Initial   GFP   concentraion ) × 100 %

The enzymatic degradation of PET materials was carried out in a 500 ml reactor. The reaction began with the addition of 0.5 μmol of enzyme to 200 ml buffer (100 mM potassium phosphate buffer pH 8) containing 120 mg of PET substrate, which was then incubated at 65°C under stirring at 100 rpm for 12 h. The enzymatic depolymerization efficiency was calculated using the sum amount of BHET, MHET and TPA released into the reaction supernatants as measured by HPLC. For this calculation, the average molecular weight of one repeating unit of the PET polymer of 192.2 g/mol was used (Yan et al., 2021).

The samples were filtered through 0.22 μm filters and analyzed by a Shimadzu CMB-20 A HPLC system coupled with a C18 column (InerSustain, 4.6 × 250 mm, 5 μm). The C18 column was eluted with solvent (18% acetonitrile, 1% formic acid and 81% water, pH 2.5) in 0–25 min. The samples were monitored at 254 nm.

The gene encoding LCC^ICCG-ChBD, LCC^ICCG-CBM, LCC^ICCG-PBM and LCC^ICCG-HFB4 were cloned into the expression vector pET29a. The peptide S2 was used as a linker between the PET hydrolase and the binding module. An additional construct pET29a-LCC^ICCG-ChBD-GFP with a gene encoding the green fluorescent protein (GFP) at a 3’ downstream position to ChBD was also created for fluorescent detection of the recombinant proteins (Figure 2A). SDS-PAGE in Figure 2B revealed that the LCC^ICCG variant, as well as the fusion proteins were obtained as highly purified recombinant, respectively.

The kinetic parameters for the enzymatic hydrolysis of pNPB were determined for the fusion proteins and compared to those obtained with LCC^ICCG (Table 1). The K _m values obtained with the fusion proteins were significantly higher, while the corresponding k _cat /K _m value was approximatively 20% less than that of LCC^ICCG. But instead, the kinetic parameters with amorphous PET film as substrate shows that the fusing proteins LCC^ICCG-ChBD have low K _m values (131.9 μM) compared to LCC^ICCG (171.8 μM) mearing better substrate affinity to amorphous PET film. At the same time, LCC^ICCG-ChBD have about 2 times k _cat /K _m than the cutinase without binding module (Table 2).

Figure 3 showed the time courses of adsorption rates of PET hydrolases with and without ChBD determined fluorimetrically (Figure 3B). The GFP fluorescence and LCC activity showed a decent linear relationship within 24 h (Figure 3A). After 80 min of incubation at 4°C, the maximum adsorption rate of 22.3% was obtained with LCC^ICCG-ChBD. By contrast, the saturated adsorption rate with LCC^ICCG was approximately 7% which has been achieved already after 30 min of incubation (Figure 3C). As shown in Figure 3D, the adsorption time courses with LCC^ICCG-ChBD were investigated at various temperatures. At 40°C, the maximum adsorption efficiency of 30.2% was achieved after 70 min of incubation. This is the optimal binding temperature for the chitin binding domain used in this study according to previous publications (Zhou et al., 2020). With increasing incubation temperature to 65°C or 72°C, the maximum adsorption rate decreased slightly to 27% or 23.5%, which was achieved after 2 h incubation and considered beneficial for the enzymatic plastic degradation carried out at this temperature.

Figure 4 showed the time course study of the GF PET degradation by the cutinase variant with and without various binding modules in terms of the release of HPLC detectable hydrolysis products. In all samples, mono (2-hydroxyethyl) terephthalate (MHET) is the most abundant product followed by terephthalic acid (TPA), and bis (2-hydroxyethyl) terephthalate (BHET). During the total incubation time up to 12 h, the levels of released TPA and MHET increased continuously. After 12 h reaction at 65°C, LCC^ICCG yielded 2.12 mM hydrolysis products composed of 0.76 mM TPA, 1.25 mM MHET and 0.11 mM BHET (Figure 4A). LCC^ICCG-ChBD showed a higher activity against the GF-PET, releasing 0.92 mM TPA, 1.5 mM MHET and 0.32 mM BHET after 12 h incubation, accounting for 29% more yield in total product amount than that without binding module (Figure 4A). In comparison, the highest amount of degradation products composed of 1.23 mM TPA, 1.56 mM MHET and 0.25 mM BHET was yielded with LCC^ICCG-CBM, and is 14% higher than with LCC^ICCG-ChBD (Figure 4A). The yield of degradation products with LCC^ICCG-PBM and LCC^ICCG-HFB4 was comparable or slightly lower than that with LCC^ICCG, indicating that the fusion of PBM and HFB4 to LCC^ICCG might have weakly impaired its catalytic activity. As shown in Figure 4B, depolymerization degrees of 87.5 and 98.5% were achieved with LCC^ICCG-ChBD and LCC^ICCG-CBM, respectively, which are 19.6 and 30.6% higher than that obtained with the enzyme without binding modules (Figure 4B).

PET substrates with different crystallinity, such as GF-PET (6.7%), Pcw-PET (16%) and Hc-PET (40%), were used to investigate the effect of crystallinity on the degradation performance catalyzed by various cutinase versions. Figure 5 indicated that the effectiveness of enzymatic hydrolysis decreased significantly as polymer crystallinity increased. LCC^ICCG, in particular, was nearly impossible to break down Hc-PET exhibiting only 0.029 mM hydrolysis products released after 12 h. Interestingly, the fusion enzymes LCC^ICCG-ChBD and LCC^ICCG-CBM showed markedly increased depolymerization performance against Hc-PET by releasing 0.335 and 0.215 mM hydrolysis products, respectively, which are 11.6 and 7.1 times higher than without binding modules. The superior activity of LCC^ICCG-ChBD and LCC^ICCG-CBM was noticeable against all three PET substrates with different crystallinity. Furthermore, ChBD outperform other binding modules in facilitating the breakdown of Hc-PET materials.

The degradation of GF-PET film with LCC^ICCG-ChBD was carried out at 65 °C for 6 h. As shown in Figure 6A, GF-PET is cut into 2*2 cm pieces, and Figure 6B shows the residual plastic fragments after enzymatic breakdown. Figure 6C shows a SEM image of a plastic sheet prior to the degradation, with the surface appearing smooth. A SEM image of a GF-PET film after degradation with increased surface roughness is shown in Figure 6D.