Assessment of IsPETase-Assisted Depolymerization of Terephthalate Aromatic Polyesters and the Effect of the Thioredoxin Fusion Domain

Featured Application

Potential use of enzymes in the development of bioprocesses for the degradation and recycling of residual plastic fibers.

Abstract

Terephthalate polyesters such as poly(ethylene terephthalate) (PET) have been massively produced over the last few decades due to their attractive properties in multiple applications. However, due to their limited biodegradability, they have accumulated in landfills and oceans, posing an environmental threat. Enzymatic recycling technologies are predicted to generate long-term socioeconomic benefits. In the present work, we compared the IsPETase (from Ideonella sakaiensis 201-F6) activity on a series of polyesters, including poly(butylene) terephthalate (PBT), poly(hexamethylene) terephthalate (PHT) and Akestra™, with PET. The IsPETase showed remarkable activity toward PET (39% degradation of the original polyester) that was higher than that toward Akestra™ (0.13%), PBT (0.25%) and PHT (0.13%) after 72 h. Thus, based on experimental data and computational analysis, we report insights into IsPETase activity on a series of terephthalate-based polyesters. Aside from that, the fusion domain (Trx) effect in the production and activity of a recombinant Trx-IsPETase is reported.

1. Introduction

From every piece of plastic ever created, around 80% still exist today [1]. Of the 6.3 billion tons of fossil fuel-derived plastic (FFP) waste produced to date, only 9% has been recycled, 12% has been incinerated, and 79% has been deposited in landfills or the natural environment [2,3]. Global plastics production reached approximately 360 million tons in 2018, with roughly 50% disposed of after a single use [4]. In Europe, 62 million tons of plastics were produced in 2018, where 9.4 million tons of plastic post-consumer waste were collected to be recycled (a 92% increase since 2006) [5]. The poor plastic waste management seen worldwide has provoked extended micro- and nano-plastic pollution in marine environments, mainly caused by the fragmentation of plastics. The ubiquitous presence of microplastics has been reported in freshwater, pristine, remote, and even atmospheric systems. The abundance of nanoplastics reported in the food chain is very concerning for wildlife and human health [6].

Polyesters constitute a major class of plastics, and they are widely used as beverage bottles, packaging, and textiles. Currently, the polyesters on the market are dominated by those containing aromatic units (e.g., terephthalate) in their backbones, which are essential for their superior thermal and mechanical properties [7]. Terephthalic acid (TPA) or its derivative dimethyl terephthalate (DMT) are widely used as precursors for the synthesis of aromatic polyesters. For instance, the most used polyester, poly(ethylene) terephthalate (PET), is synthesized by polymerizing TPA or DMT with ethylene glycol. It is commonly used for synthetic fibers in the textile industry and packaging [8,9,10]. Poly(butylene) terephthalate (PBT) is another aromatic polyester synthesized from TPA (or DMT) and 1,4-butanediol which is widely used in electrical engineering as plug connectors, household devices like irons, and sportswear due to its physical properties. Recently, a more flexible aromatic polyester, poly(hexamethylene) terephthalate (PHT) has attracted growing attention due to its potential in semiconductor materials [11,12,13]. Another recent trend in the development of aromatic polyesters in the industry is toward more durable high-T_g materials, such as Akestra^TM (Perstorp AB), PETG (Sumitomo), and Tritan^TM (Eastman Co.) [14,15,16].

Biodegradation is considered eco-friendly and, in the future, an economically favorable recycling method compared with landfills, chemical recycling, and thermal treatment [17,18]. Enzymatic degradation of plastics could aid in the biological recycling of these waste products [19], reducing energy use, by-product release, and greenhouse gas emissions compared with virgin polyester manufacturing [20]. Enzymatic recycling technologies are predicted to generate long-term socioeconomic benefits, such as added economic value and a number of jobs distributed along the supply chain [20]. As polyesters constitute 60% of the total global synthetic fiber production, there is growing concern for efficient biodegradation of recalcitrant synthetic polymers like PET and other aromatic terephthalate polyesters [18,21].

Enzymes with PET-depolymerizing activity have been found to be mainly from the carboxylic ester hydrolases family (EC 3.1.1.-). The first study in 2005 reported 50% degradation of low-crystallinity (9%) PET after three weeks by using the hydrolase TfH from the actinomycete Thermobifida fusca at 55 °C, adding fresh enzyme every week [22]. Since then, other thermostable esterases have shown effective PET depolymerizing activity in shorter periods [23]. More recently (in 2020), an engineered form of leaf and branch compost cutinase (LCC) has attracted attention for its outstanding PET depolymerizing activity, degrading up to 90% of amorphized and micronized PET in 10 h at 72 °C [24]. At lower temperatures, in 2016, a novel mesophilic bacterium Ideonella sakaiensis 201-F6 was found to use PET as its main carbon and energy source [19]. A key enzyme, IsPETase (EC 3.1.1.101), was identified as being responsible for PET depolymerization, which showed more activity than others when compared at mild temperatures (27–40 °C) [19,25,26]. Several IsPETase structural studies were published [17,27,28,29], which set the basis for mechanistic studies [30,31] as well as for the development of a series of mutants with increased activity [32] or thermostability [33,34,35,36,37]. Both thermostable and mesostable PET-active enzymes constitute a valuable toolbox for diverse applications. Mesostable PETases could be used in processes where high temperatures are not convenient. However, practical applications will most likely need the integration of waste plastic pretreatment processes (for example, micronization, amorphization, or others) and the use of engineered enzymes or whole cells.

Several commercial polyesters underwent biodegradation studies [38,39,40]. However, to our knowledge, apart from PET, little is known about the biodegradation of aromatic polyesters. IsPETase has also shown some activity against a novel synthetic AB-type indole-base polyester [16]. In this work, IsPETase activity against various aromatic polyesters with terephthalate bound to diols of different lengths and rigidity is assessed. Semicrystalline forms of PET, PBT, PHT, and Akestra™, which is amorphous, are used as potential substrates for IsPETase, and molecular docking is performed to explore the structural implications.

On the other hand, the production of recombinant PETases is an efficient type of technology that can be scalable to industrial levels. However, one of the main drawbacks is the production of a soluble enzyme, and a significant fraction of IsPETase produced in Escherichia coli can precipitate as inclusion bodies, losing their activity and negatively affecting the yield. A common strategy to enhance the solubility of recombinant proteins is the use of fusion partners, such as thioredoxin (Trx) [41]. Therefore, the Trx domain effect both in production and IsPETase (Trx-IsPETase) activity is also evaluated in this work.

This article focuses on understanding the fundamental aspects of IsPETase and Trx-IsPETase activities; more practical aspects will be addressed in the future. The dissolution–precipitation of the polymers used here is not related to their recycling but with their preparation in a similar physical form, such as powders, for enzymatic studies. The pretreatment to reduce the degree of crystallinity, as mentioned before, of PET and other polyesters facilitate the enzyme activity, even for thermostable enzymes, which have the advantage of working closer to the PET glass transition point [23].

2. Materials and Methods

2.1. Production and Purification of Enzymes

IsPETase and Trx-IsPETase were produced in Escherichia coli Rosetta-gami^TM2 (DE3) according to standard protocols [15,16]. The gene encoding the IsPETase from I. sakaiensis (GenBank accession code GAP38373.1) without a signal peptide (first 29 amino acids from the N-terminal) was chemically synthesized (GenScript USA Inc., Piscataway, NJ, USA) with codons optimized for E. coli and cloned into the expression vectors pET28b(+) and pET32b(+). The latter added a thioredoxin fusion domain (Trx) in the N-terminal of the IsPETase. The obtained plasmids, pET28b(+)::IsPETase and pET32b(+)::Trx-IsPETase, were introduced into E. coli Rosetta-gami^TM2 (DE3) (Novagen, U.S) by heat shock transformation. The recombinant strains were grown at 37 °C in an LB medium supplemented with 35 μg/mL kanamycin or 100 μg/mL ampicillin as selection factors for the cells harboring pET28b(+)::IsPETase or pET32b(+)::Trx-IsPETase plasmids, respectively. The recombinant strains were cultivated in 300 mL of the LB medium at 37 °C until OD_600nm = 0.6–1.0 was reached. Then, the production of IsPETase or Trx-IsPETase was induced by adding 1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and overnight incubation at 20 °C. IsPETase and Trx-IsPETase were purified from cell extracts by Immobilized Metal Affinity Chromatography (IMAC). A HisTrap™ Fast Flow column installed in an ÄKTA start protein purification system (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) was equilibrated with a binding buffer (100 mM Tris-HCl, 500 mM NaCl, pH 7.4). Then, the cell extract was loaded. After a washing step, the bound proteins were eluted with an elution gradient using a buffer (100 mM Tris-HCl, 500 mM NaCl, 500 mM imidazole, pH 7.4). The protein purity was analyzed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE), while the concentration was determined spectrophotometrically at 280 nm.

IsPETase and Trx-IsPETase were produced in Escherichia coli Rosetta-gami^TM2 (DE3) according to standard protocols [15,16]. The gene encoding the IsPETase from I. sakaiensis (GenBank accession code GAP38373.1) without a signal peptide (first 29 amino acids from the N-terminal) was chemically synthesized (GenScript USA Inc., Piscataway, NJ, USA) with codons optimized for E. coli and cloned into the expression vectors pET28b(+) and pET32b(+). The latter added a thioredoxin fusion domain (Trx) in the N-terminal of the IsPETase. The obtained plasmids, pET28b(+)::IsPETase and pET32b(+)::Trx-IsPETase, were introduced into E. coli Rosetta-gami^TM2 (DE3) (Novagen, U.S) by heat shock transformation. The recombinant strains were grown at 37 °C in an LB medium supplemented with 35 μg/mL kanamycin or 100 μg/mL ampicillin as selection factors for the cells harboring pET28b(+)::IsPETase or pET32b(+)::Trx-IsPETase plasmids, respectively. The recombinant strains were cultivated in 300 mL of the LB medium at 37 °C until OD_600nm = 0.6–1.0 was reached. Then, the production of IsPETase or Trx-IsPETase was induced by adding 1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and overnight incubation at 20 °C. IsPETase and Trx-IsPETase were purified from cell extracts by Immobilized Metal Affinity Chromatography (IMAC). A HisTrap™ Fast Flow column installed in an ÄKTA start protein purification system (GE Healthcare Bio-Sciences AB Uppsala, Sweden) was equilibrated with a binding buffer (100 mM Tris-HCl, 500 mM NaCl, pH 7.4). Then, the cell extract was loaded. After a washing step, the bound proteins were eluted with an elution gradient using a buffer (100 mM Tris-HCl, 500 mM NaCl, 500 mM imidazole, pH 7.4). The protein purity was analyzed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE), while the concentration was determined spectrophotometrically at 280 nm.

2.2. Polymer Preparation

The Akestra^TM was a courtesy from Perstorp (intrinsic viscosity (IV)~0.64 dL/g, T_g~95 °C). The PET (Ramapet N180) was purchased from Indorama (IV~0.8 dL/g, T_g~78 °C, T_m~245 °C), and the PBT was purchased from Sigma-Aldrich (M_n~38,000 g/mol, T_m~203 °C). The PHT (M_n~26,000 g/mol, M_w~43,000 g/mol, IV~1.26, T_g~17 °C, T_m~140 °C) was synthesized according to a previously published protocol [11,12]. The PHT and Akestra™ were dissolved in chloroform and precipitated into methanol, and the resulting fine powders were dried in a vacuum oven for 12 h before use. The PET and PBT were dissolved in hexafluoroisopropanol (HFIP) and precipitated into methanol, and the resulting fine powders were dried in a vacuum oven for 12 h before use. Changes in crystallinity were assessed by differential scanning calorimetry.

2.3. Differential Scanning Calorimetry Analysis

Differential scanning calorimetry (DSC) measurements were performed using a TA Instruments DSC Q2000. The samples of PET, PBT, and PHT were studied with a heating rate of 10 °C min⁻¹ under nitrogen with a purge rate of 50 mL min⁻¹. The heating ramp was from 25 °C to 200 °C or 300 °C. It was observed that the crystallinity of the PET decreased from ~24% to ~11.3% after dissolution–precipitation, according to DSC analysis and calculated with the following equation:

$$Crystalinity (\%)=\frac{\left|∆H_m\right|-\left|∆H_C\right|}{\left|∆H_m^{°}\right|}\times 100$$

where $∆H_m$ and $∆H_C$ are the heats of melting and cold crystallization, respectively, and $∆H_m^{°}$ is the heat of melting for a 100% crystalline polymer estimated to be 140 J/g [26]. No such change was observed for the PBT and PHT after precipitation, indicating that the original crystallinity degree was not affected by dissolution–precipitation (data not shown). Considering the theoretical $∆H_m^{°}$ = 140 J/g [42], the calculated degree of crystallinity of the PBT used was 38.6%. As PHT analysis did not show a cold crystallization temperature after precipitation, it was assumed that it was as a semicrystalline form.

2.4. Enzymatic Reactions

The reactions were prepared by soaking ~20 mg of Akestra, which is an amorphous polymer, semicrystalline PET, PBT, or PHT in 2 mL of a 50-mM phosphate buffer with a pH of 7.2, 10% (v/v) dimethyl sulfoxide (DMSO), and 0.09–0.3 mg of freshly prepared enzyme [19,27,28,43]. All reactions were prepared in triplicate, including a negative control (reaction mixture without enzyme). The reactions, including the negative controls, were incubated at 37 °C with shaking at 200 rpm for a variable time from 7 to 240 h [16,17,19,27,28,32,43] for determining the apparent saturation curve for Txr-IsPETase and for 72 h for assessing the depolymerization activity against a series of polymers.

2.5. HPLC Analysis

Samples (0.5 mL of each reaction and the controls) were retrieved and diluted at a 1:1 ratio with DMSO, filtered (200-nm PTFE membrane), and transferred to HPLC vials for further analysis. An Ultimate 3000 RS (Dionex) equipped with a UV/Vis detector (SPD-20A) and a hydrophobic C18 column (Kinetex^® 1.7 µm XB-C18 100 Å, LC Column 50 × 2.1 mm) was used to analyze the components of each sample. Mobile phases A (20%) and B (80%) consisted of acetonitrile and an aqueous solution containing 0.1% formic acid, respectively. The flow rate was fixed at 0.4 mL/min for 3 min of analysis per sample. Analytes containing TPA were detected at 260 nm. The HPLC method and data were processed with Chromeleon v. 6.8 software. The TPA and equivalents in the TPA were quantified using a standard curve of TPA and dissolved in DMSO in a range from 830 µg/L to 2 g/L.

2.6. LCMS Analysis

The separation was performed as described in Section 2.4 using a similar Ultimate 3000 RS HPLC-system (Thermo Fisher Scientific) coupled with an LTQ Velos Pro Ion trap mass spectrom (Thermo Fisher Scientific) using a heated electrospray ionization source (HESI-II). The samples were analyzed in full scan mode between m/z 100 and 400 in negative mode while employing a spray voltage of −3 kV.

2.7. Effec of the Enzymatic Activity on PET Crystalinity

The residual PET samples from enzymatic reactions and controls, after 9 days of incubation, were washed with detergent vortexing, rinsed with copious amounts of ultrapure water, and dried under airflow for 48 h. The final dry weights were measured and compared to the initial weights (~20 mg). The dried samples were subjected to DSC analysis (Section 2.3) to determine the thermal parameters.

2.8. Molecular Modeling of Trx-IsPETase

The structure of the Trx-IsPETase fusion protein was modeled using the atomic coordinates of the crystallized IsPETase (PDB: 5XJH) and thioredoxin domain (Trx) (PDB: 5E4W). All water molecules and ligands were removed from the crystallized structures. The interdomain linker structure was predicted by ab initio protein assembly using the QUARK [44,45] program. These three modules were manually assembled in YASARA [46], with the Trx domain in the N-terminal and the catalytic IsPET domain in the C-terminal, both connected by the interdomain linker. The resulting fusion protein was energetically optimized.

2.9. Molecular Dynamic Simulations

The modeled Trx-IsPETase and IsPET (PDB: 5XJH) were subjected to molecular dynamics simulations using GROMACS [47,48]. Each system included one protein solvated in TIP3P water molecules containing 0.15% CaCl₂ ions into a box of a 10-Å extension from the protein. The calculations were performed by applying the OPLS-AA/L force field [49], periodic boundaries, 8-Å cut-off of short-range electrostatic and van der Waals forces, and long-range forces calculated by the particle mesh Ewald method. The simulations consisted of three steps: energy minimization, equilibrations, and production, as described in a previous publication [50]. After 500 ns of production, the root-mean-square deviations (RMSDs) of the alpha carbons and root-mean-square fluctuation (RMSF) were calculated. The graphical analysis and molecular pictures were performed using UCSF Chimera software v. 1.15 [51].

2.10. Molecular Docking

The receptor was prepared from the IsPETase crystallographic structure (PDB: 6EQE). The ligands studied were PET monomer (bis-(hydroxyethyl) terephthalate (BHET)), PBT monomer (bis-(hydroxybuthyl) terephthalate (BHBT)), PHT monomer (bis-(hydroxyhexyl) terephthalate (BHHT)), and an Akestra^TM fragment (Figure 1). All ligands were modeled using the Avogadro program [52]. The molecular geometries were optimized through molecular mechanics using the MMFF94 force field, with 5000 steps, a steepest descent algorithm, and a convergence of 10e⁻⁷. All ligands were docked using Autodock Vina [53] implemented in YASARA [46]. Visual analysis was performed with USCF Chimera v. 1.15 [51].

3. Results and Discussion

3.1. Enzyme Production

Recombinant IsPETase and Trx-IsPETase were successfully produced as soluble enzymes in E. coli Rosetta-gami^TM2 (DE3). The SDS-PAGE analysis showed proteins of 31.5 and 43.5 kDa, which is consistent with the theoretically calculated molecular weights of IsPETase and Trx-IsPETase, respectively (Figure 2A). IsPETase had a production yield of 79 mg/L of culture, with an estimated purity higher than 90%, while Trx-IsPETase had a yield of 93 mg/L and purity of ~80%. A larger elution gradient was needed to achieve an acceptable purity compared with the Trx-IsPETase (Figure 2B). Furthermore, the purified Trx-IsPETase eluted in more concentrated fractions than the IsPETase, as shown by the shape of the peak in the purification chromatograms (Figure 2B). It is worth mentioning that IsPETase is sensitive to precipitation, which limits its production. To overcome this drawback, several strategies were evaluated, such as secretion systems, expression hosts, and the Trx fusion domain. The Trx domain together with the host strain E. coli Rosetta-gami^TM2, conferred the highest production level compared with other cytoplasmic and even some extracellular systems (

Table 1. Production of recombinant IsPETase in different systems.

Host	Plasmid	Production (mg/L)	Reference
Cytosolic systems
E. coli Rosetta-gamiTM2	pET28b(+)::IsPETase	79	This work
E. coli Rosetta-gamiTM2	pET32b(+)::Trx-IsPETase	93	This work
E. coli BL21 (DE3)	pET21b(+)-Is-PETase	no detected	[54]
E. coli		4	[43]
Extracellular systems
E. coli BL21 (DE3)-T1R	pET22b-SPMalE:IsPETase	6.2	[43]
E. coli BL21 (DE3)-T1R	pET22b-SPLamB:IsPE-Tase	3	[43]
E. coli BL21 (DE3)	pPel-IsPETase (pET22b)	650	[55]
Bacillus subtilis	integrative vector p2JM103	80	[26]

).

3.2. Product Profile and Activity of IsPETase and Trx-IsPETase

The products of both enzymes against PET were evaluated by HPLC. DMSO was used as a co-solvent for the depolymerization products. TPA, MHET, and BHET have low solubility in water. All controls containing the polymers, buffer, and DMSO did not show any detectable degrading effect of the DMSO under the same experimental conditions as the enzymatic reactions. Furthermore, dissolved products are required for analysis by HPLC.

The chromatograms showed the same product profiles (Figure 3A). With terephthalic acid (TPA) eluted at ~0.61 min, other peaks at ~0.76 and ~0.96 min were identified by HPLC-MS (Figure 3B) as mono-(2-hydroxyethyl) terephthalate (MHET) and bis-(2-hydroxyethyl) terephthalate (BHET), respectively. According to the relative area of the peaks, MHET was the main product, followed by TPA, while BHET was present in a significantly smaller fraction.

Trx-IsPETase showed slightly higher activity than IsPETase (Figure 3C). All resulting peaks from HPLC analysis mentioned in this section were regarded as terephthalic acid equivalents (TPAeq), assuming the same molar extinction coefficient ε = 17,000 M⁻¹cm⁻¹ conferred by the aromatic ring. Thus, the calculated TPAeq produced could be determined. The total amount of degraded polymer could be calculated using the TPAeq and the repeating molecular weight (192 g/mol). The results of the depolymerization are shown in Table 1. The degree of depolymerization after 72 h by IsPETase was 23%, and for Trx-IsPETase, it was 32%. The enzyme/PET ratio was slightly higher for Trx-IsPETase (0.005) than IsPETase (0.004). However, the specific activity, defined here as the mg of product per mg of enzyme and time (mg degradation products/mg enzyme h⁻¹ or h⁻¹), gave a higher value for Trx-IsPETase (0.7 h⁻¹) than IsPETase (0.9 h⁻¹) (

Table 2. PET depolymerization activity of IsPETase and Trx-IsPETase. Degradation products consist of the sum of the TPA, MHET, and BHET quantified.

Enzyme	Enzyme/PET Ratio (mg/mg)	Degradation Products (mg/L)	Relative Production (%)			Depolymerization PET (%)	Specific Activity (h−1)
			TPA	MHET	BHET
IsPETase	0.004	2304	24.2	73.7	2.1	23	0.7
Trx-IsPETase	0.005	3230	31.5	66.7	1.8	32	0.9

). Although the differences were not remarkable, the student’s t-test gave a statistically significant higher product formation for Trx-IsPETase than IsPETase (Figure 3C).

The dynamic stability of Trx-IsPETase was determined by monitoring the formation of the product for 240 h. The degradation products increased in concentration during the first 72 h until reaching saturation (Figure 4). The saturation was attributed to the inactivation of the enzyme and not to the completion of the reaction, since the substrate conversion reached ~35%. The apparent inactivation rate constant ($a_S^{app}$) was determined according to first-order kinetics with the coexistence of a sufficient substrate (20 mg PET per 2 mL) [56]:

$$\frac{\left[P\right]}{\left[E\right]}={{\displaystyle \int }}_0^t{k}_S^{app}dt= \frac{k_{S,0}^{app}}{a_S^{app}}\left(1-e^{\left(-a_S^{app}t\right)}\right)$$

where $\left[P\right]$ is the degradation products concentration, $\left[E\right]$ is the enzyme concentration, and $k_{S,0}^{app}$ and $k_s^{app}$ are the initial enzyme activity in a given time ($t$). The integrate form describes the apparent saturation curve (Figure 4). Thus, $a_S^{app}$ yielded 0.09 h⁻¹, and $k_{S,0}^{app}$ yielded $6.2 \frac{\mathrm{mg} \mathrm{Degradation} \mathrm{products}}{\mathrm{mg} \mathrm{Enzyme} \times \mathrm{h}}$.

We noticed that the enzymatic reactions apparently showed an impact on the crystallinity (or re-crystallization rate) of the residual PET, as indicated by the DSC results (Figure 5). The DSC analysis showed a cold crystallization peak at 122 °C in all three enzymatically treated PET samples, while the control did not (Figure 5). The measured crystallinity of the residual PET after enzymatic reaction (~22%) or the control (30.5%) was higher than that of the initial PET powder sample (~11%). This suggests that PET, prepared by precipitation, could re-crystallize under the reaction conditions. Thus, assuming that the Trx-IsPETase primarily attacks amorphous regions, it may retard the re-crystallization rate. We should emphasize that the observed change of crystallinity of the PET was likely a complex consequence including the cold re-crystallization process under the reaction conditions, selectivity of the enzymatic degradation of the crystalline and amorphous regions, and the possible influence of the enzyme on the rate of re-crystallization. More thorough investigations on the crystallization processes will be needed to allow for deeper insight into the impact of enzymes on the crystallization rate and crystallinity of PET.

On the other hand, the residual weight of the enzymatically treated PET corresponded to 34 ± 5% degradation at 235 h, consistent with the percentage of degradation calculated by the HPLC method at 72 h (Table 2), considering that the reactions reached saturation at the maximum product formation at 72 h.

3.3. Structural Studies

The obtained molecular structure of Trx-IsPETase included a Trx domain in the N-terminal, followed by an interdomain linker module and the catalytic domain in the C-terminal (Figure 6A). Two amino acids, Met1 and Ser2, were added in the N-terminal of the Trx domain, which extends until Ser111. The interdomain linker starts in Gly112 and ends in Ala169. The catalytic domain encompasses Arg170, which corresponds to the Arg34 in the IsPETase, to Ser426, which corresponds to Ser290 in IsPETase. While the Trx and catalytic domains were obtained from crystallographic structures, the interdomain linker structure was predicted using an ab initio approach. It was modeled as a structure composed of three α-helices, with the third one containing six histidines used in the purification by affinity chromatography (IMAC) (Figure 6A).

The initial structure of Trx-IsPETase was obtained in a closed-form conformation, with the Trx domain oriented toward the catalytic domain (Figure 6B). Although this structure was energetically minimized, molecular dynamics simulations showed a significant conformational change, giving an open-shaped conformation (Figure 6B,C). which resulted in a more energetically favorable structure.

3.4. IsPETase Activity on Terephthalate Aromatic Polyesters

PET, PBT, PHT, and Akestra^TM were used as substrates for IsPETase. In all cases, the main product was TPA (Figure 7), unlike previous reactions where MHET was the predominant product of PET depolymerization (Figure 3A). This could be due to the substantially higher amount of enzyme used for these substrates (0.01 mg IsPETase per mg of the substrate). The enzymatic products were quantified as TPAeq, and the percentage of depolymerization was calculated as described in Section 3.2. Thus, PET showed remarkable enzymatic depolymerization (39%) compared with the other polymers (<0.25%) (

Table 3. Depolymerization of different terephthalate polyesters after 72 h with an IsPETase/substrate ratio of 0.01 (mg/mg).

Terephthalate Aromatic Polyester	Degradation Products (mg/L)	Depolymerization (%)
PET	3934	39.34
PBT	25	0.25
PHT	12.5	0.13
Akestra™	13.3	0.13

). Interestingly, although by far, the second-best substrate was PBT, followed by PHT and Akestra™, suggesting a correlation of IsPETase activity with the length and size of the TPA linker group present on the polymer.

The ligand binding modes were studied by coupling monomers BHET, BHBT, BHHT, and an Akestra^TM fragment at subsite I of IsPETase (Figure 8). The complexes obtained were compared with the co-crystallized structure of the hydroxyethyl methyl terephthalate (HEMT)/IsPETase double mutant (PDB: 5XH3). This mutant has glycine instead of arginine 132 and alanine instead of nucleophilic serine 160 (corresponding to R103G/S131A in 5XH3) [23]. The coupled BHET, corresponding to a PET monomer, had the same pose as the co-crystallized HEMT analog (Figure 8A), indicating a reliable coupling prediction, and no further refinements were needed. Indeed, the aromatic and carbonyl groups of all the ligands overlapped very well with the corresponding groups of HEMT (Figure 8B–D). Therefore, the main interactions between the studied ligands and the subsite I were conserved. It was suggested that Trp 185 and Trp 159 form p-p interactions with the aromatic groups of the ligands at subsites I and II, respectively [17]. In the complexes obtained here, Trp 185 forms p-p interactions with all ligands. However, the length of the diol linkers in BHBT and BHHT seems unfavorable for an interaction of a second aromatic ring with Trp 159. Interestingly, one of the m-dioxane rings of the Akestra^TM fragment formed a hydrophobic interaction with Trp 159 (Figure 8D), but the complexity of its polymeric form could significantly reduce the enzyme’s access to ester linkages. Therefore, the low IsPETase activity in PBT, PHT, and Akestra^TM could be attributed to limited access or unfavorable interactions at subsite II.

4. Conclusions

The thioredoxin fusion domain confers a positive effect both in production as well as in PET depolymerization activity. Aside from that, the structural molecular modeling and dynamics showed that the Trx domain neither interfered with the active site nor induced significant conformational changes in the catalytic domain.

On the other hand, IsPETase showed remarkable PET-depolymerizing activity compared with PBT, PHT, and Akestra^TM. These results suggest that larger or bulkier TPA linkers negatively affect the interaction of polyesters with the enzyme.

The enzymatic depolymerization of terephthalate aromatic polyesters other than PET is an open field to be explored, both by discovering novel enzymes or engineering polyethylene terephthalate hydrolases.