Selective Modification of the Product Profile of Biocatalytic Hydrolyzed PET via Product‐Specific Medium Engineering

Abstract

Over the past years, enzymatic depolymerization of PET, one of the most widely used plastics worldwide, has become very efficient leading to the end products terephthalic acid (TPA) and ethylene glycol (EG) used for PET re‐synthesis. Potent alternatives to these monomers are the intermediates BHET and MHET, the mono‐ and di‐esters of TPA and EG which avoid total hydrolysis and can serve as single starting materials for direct re‐polymerization. This study therefore aimed to selectively prepare those intermediates through reaction medium engineering during the biocatalytic hydrolysis of PET. After a comparative pre‐screening of 12 PET‐hydrolyzing enzymes, two of them (LCC^ICCG, IsPETase^wt) were chosen for detailed investigations. Depending on the reaction conditions, MHET and BHET are predominantly obtainable: (i) MHET was produced in a better ratio and high concentrations at the beginning of the reaction when IsPETase^wt and 10 % EG was used; (ii) BHET was produced as predominant product when LCC^ICCG and 25 % EG was used. TPA itself was nearly the single product at pH 9.0 after 24 h due to the self‐hydrolysis of MHET and BHET under basic conditions. Using medium engineering in biocatalytic PET‐hydrolysis, the product profile can be adjusted so that TPA, MHET or BHET is predominantly produced.

The product profile of enzymatically hydrolyzed PET can be modified by medium engineering and thereby adapted to a desired product. TPA, MHET or BHET can be forced as the predominant product using a basic pH (blue), 25 % ethylene glycol (EG) and IsPETase^wt (green) or ≥25 % EG and LCC^ICCG (pink), respectively.

Introduction

Synthetic plastics are an indispensable part of daily life and are widely used due to their advantageous properties such as lightness, resistance and inertness to several influences and chemicals, high barrier properties, and cost‐effective production.[ ¹ , ⁵ ] Consequently, plastics are widely distributed around the world, reaching about 700 million tons produced per year in 2030, which is around 80 kg per human. ^[6] However, the drawbacks of plastics are also well known, including that they are non‐ or hardly decomposable in nature, or that some countries have poor waste and recycling management. ^[7] In fact, only 10–20 % of plastics are currently recycled, and about one‐half are released into the environment.[ ⁶ , ⁸ ] This leads to an increasing accumulation in the natural and specifically marine environment, with approximately tens of millions of tons entering the oceans per year.[ ⁸ , ⁹ ] Different forms of plastics intensify these problems, while nano plastic has already been detected in humans, ^[11] far from civilization in polar areas, ^[13] even in leaves and thus in forests. ^[14] Poly(ethylene terephthalate) (PET) is one of the most distributed synthetic plastics,[ ⁷ , ¹⁵ ] consisting of repeating terephthalic acid (TPA) and ethylene glycol (EG) units. The depolymerization of PET has been intensively investigated, with various chemical processes including solvolysis (glycolysis, hydrolysis, methanolysis, aminolysis, ammonolysis, phosphorolysis etc.) being the most commonly and versatilely used option.[ ⁵ , ¹⁵ , ¹⁶ ] Often very high temperatures of several hundred °C and high loads of chemicals are required. ^[18] Both entail the drawbacks of high (energy) costs, and low ecological and environmentally friendly aspects. The general aim is the complete depolymerization of PET to its basic building blocks TPA and EG, which can be isolated afterwards and used for the re‐synthesis of recycled PET.[ ⁵ , ²⁰ ]

As an alternative to the established techniques, enzymatic hydrolysis of the ester bond in PET got increasing attention over the past years, since enzymes capable of hydrolyzing PET were studied in detail.[ ⁷ , ²³ ] A hydrolase from Ideonella sakaiensis (IsPETase^wt) was discovered in 2016, showing higher but still low activity towards amorphous and, to an even lesser extent, highly crystalline PET compared to few other enzymes at that time and ambient temperatures. ^[25] However, instead of TPA and EG, mono(2‐hydroxyethyl) terephthalate (MHET) was formed as the main product. ^[25] In the same period, it was shown that the final hydrolysis step (i. e., MHET to TPA and EG) is very slow, and thus the MHET was considered as an inhibitor for PET depolymerization catalyzed by IsPETase^wt and several other enzymes.[ ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ ] This inhibition was thus regarded as a rate‐limiting step, leading to reduced reaction rates, lower product concentrations and thus lower efficiencies. This was first tackled by searching enzymes capable of efficiently hydrolyzing MHET (commonly referred to as MHETases) in combination with PET‐hydrolyzing enzymes (often referred to as PETases). One of the most prominent MHETase is from Ideonella sakaiensis (IsMHETase), which has been used, e. g., together with IsPETase^wt and variants thereof to depolymerize PET more efficiently. ^[30] In contrast, intensive efforts were made to find PETases that are less influenced by MHET inhibition, e. g. by genome mining or via protein engineering of existing enzymes. ^[32] Several enzymes have been greatly improved by enzyme engineering in terms of lower inhibition, higher activities and higher stabilities at elevated temperatures required for efficient PET depolymerization, including TfCut2 ^[29] and improved IsPETase^wt variants such as DuraPETase^N233C/S282C, ^[35] HotPETase, ^[36] and FastPETase. ^[37] On the other hand, a huge variety of more or less efficient PETases has been identified, which in turn have been the subject of intensive enzyme engineering. The latter revealed the highly efficient variant LCC^ICCG, a cutinase discovered in leaf‐branch compost, that has already been proven in an industrial setting and is to be applied on industrial scale (50,000 tons/year capacity) soon. ^[38]

The reverse production of new PET is typically performed by condensation of TPA and EG to bis(2‐hydroxyethyl) terephthalate (BHET) by esterification with the release of water, which is then polycondensated to PET releasing EG. ^[39] Conclusively, the hydrolysis intermediate BHET is a more appropriate starting material for PET resynthesis, similar to PET recycling via glycolysis, if enzymatic PET hydrolysis could result in BHET as the only final product. Additionally, MHET could also be used as an alternative starting material for PET‐resynthesis, ^[41] releasing water instead of EG during polycondensation of BHET. This study therefore aims to investigate – starting from a preliminary enzyme screening of 12 different PETases – the influence of different reaction conditions on the product profile of enzymatic PET‐hydrolysis (Scheme 1). 3PET (bis(benzoyloxyethyl) terephthalate) was used as a well‐defined model substrate for the initial studies. The most promising results were then verified with Nano‐PET, as a more realistic PET‐substrate.

Scheme 1.

Enzymatic hydrolysis of PET. PET is hydrolyzed by various literature‐known PET‐hydrolyzing enzymes to terephthalic acid (TPA), ethylene glycol (not shown for reasons of simplicity), mono(2‐hydroxyethyl) terephthalate (MHET) and bis(2‐hydroxyethyl) terephthalate (BHET) in a certain ratio depending on the enzyme used and the total conversion extent. MHET and BHET can be completely hydrolyzed to TPA, which is then used for re‐synthesis of PET via a condensation step to BHET. This study aims to target BHET and MHET final alternative products that can also be used for re‐synthesis of PET in a more eco‐friendly manner.

Results and Discussion

Pre‐Screening of Various Literature‐known PET‐Hydrolyzing Enzymes

A total of 12 enzymes were selected for this study based on their reported high (improved) activity against different PET substrates and/or stability under high temperatures (Table S1). However, the setup performed in each study (type of PET‐substrate, reaction conditions, buffer, etc.) is decisive for the activities, so they can vary between each study and must be always compared to benchmark enzymes.[ ⁴² , ⁴³ ] Consequently, several studies used a well‐known hydrolase such as LCC^ICCG to compare newly discovered, characterized or engineered hydrolase.[ ³⁶ , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ ] Therefore, in the first step, an enzyme screening was carried out with a larger variety of PETases under defined reaction conditions for direct comparison of their performances. 3PET was used as a well‐defined model substrate within this study to enable a certain comparability with other studies.[ ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ ] All enzymes were applied under standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) and the resulting concentrations of TPA, MHET and BHET were determined after 1, 8 and 24 h by HPLC (Figure 1). Importantly, these conditions were not optimal for each enzyme, however, these were used for comparison since high temperatures are required for efficient PET hydrolysis.

Figure 1.

Pre‐screening of various PET hydrolyzing enzymes. 12 enzymes were analyzed regarding their product profile after 1, 8 and 24 h reaction time during the hydrolysis of the model substrate 3PET. The reactions were performed under standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) with 0.35 mm (=0.161 mg/mL) 3PET and 3.3 μg/mL (≈0.1 μm) enzyme. TPA is depicted in blue, MHET in green and BHET in pink. All reactions were performed in triplicates, and the error bars represent the standard deviation. The background due to self‐hydrolysis of 3PET under certain conditions (Figure S2) was always subtracted. The four enzymes showing no activity in the performed setup for unknown reasons (i. e., DmPETase, MoPE, DuraPETase^N233C/S282C, TfCUT2) are not shown here. IsPETase^wt, and LCC^ICCG (outlined in black) were chosen for further investigations.

Almost all enzymes produced MHET predominantly at the beginning (1–8 h), while the concentration of MHET decreased and of TPA increased over the reaction time. This is recognized for many PET hydrolyzing enzymes,[ ²⁶ , ³⁶ , ⁴⁷ , ⁵⁴ ] most likely because MHET inhibits these enzymes to a certain extent, causing MHET hydrolysis to be the rate‐limiting step.[ ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ ] The different reaction rates from MHET to TPA can be clearly seen when visualizing the MHET/TPA and TPA/MHET ratios over time (Figure S1). Notably, DmPETase, ^[44] MoPE, ^[55] and DuraPETase^{N233C/S282C[56]} did not show activity in this study in contrast to the literature for unknown reasons and TfCUT2 ^[57] was active only at reduced temperatures (40 °C, data not shown).

Based on this screening, two enzymes were chosen for further analysis: 1) LCC^ICCG as an efficient reference enzyme widely used for comparison in several studies[ ³⁶ , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ ] and the most prominent one for industrial use ^[43] ; and 2) IsPETase^wt with very high MHET/TPA ratios and high product concentrations at the beginning.

Selective Modification of the Product Profile by Medium Engineering

In the next step, the effect of different reaction conditions on the product profile was analyzed in reactions with the two chosen enzymes. As this study aims to produce the desired intermediates MHET and BHET as primary final products, the hydrolysis step from BHET/MHET to TPA and EG should be limited as much as possible. Therefore, a reduction of the water content in the reactions was targeted by increasing the concentration of an exemplary non‐toxic organic solvent (DMSO) and EG (10, 25, and 50 %). It has been shown occasionally that the product profile generated by PETases can be altered when EG was used as cosolvent.[ ⁵⁸ , ⁵⁹ ] In addition, it has been previously shown that DMSO and EG affect the solubility BHET and TPA, which might be used for in situ removal and would allow a re‐use of unspent co‐solvents in solution. ^[60] However, it should be noted that DMSO may be exchangeable, depending on the intended downstream process. In addition to solvents, the reaction pH can also have an impact on the product profile as it has recently been shown. ^[66] Therefore, the standard pH (7.5) was compared to an acidic (5.0) and a basic (9.0) pH. As before, the TPA, BHET and MHET concentrations were determined after defined reaction times (1, 8, 24 h), whereby the standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) were used for comparison with the varying conditions. In general, all conditions affected the product profile to varying degrees, which also depends on the enzyme applied (Figure 2, Table 1, Figure S2‐S6).

Figure 2.

Total product concentrations for 3PET hydrolysis. The total product concentrations of TPA, MHET and BHET (not benzoic acid as artificial byproduct, see Figure S7) for the hydrolysis of 3PET by IsPETase^wt and LCC^ICCG under different conditions (as indicated) are shown after 24 h reaction time. The hydrolysis was performed using 0.35 mm (=0.161 mg/mL) 3PET, 3.3 μg/mL (≈0.1 μm) enzyme, and standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) unless the varying condition stated. For reactions with different concentrations of DMSO and EG, the standard pH of 7.5 was used. All reactions were performed in triplicates, and the error bars represent the standard deviation. The background due to self‐hydrolysis of 3PET under certain conditions (Figure S2) was subtracted and the total product concentrations for the entire reaction course are shown in Figure S6.

Table 1.

Effect of different conditions on the product profile of 3PET hydrolysis after 1 h (except for pH 9.0 after 24 h).

	IsPETasewt			LCCICCG
	$\frac{\mathrm{MHET}}{\mathrm{TPA}}$	$\frac{\mathrm{BHET}}{\mathrm{TPA}}$	$\frac{\mathrm{TPA}}{\mathrm{MHET}}$	$\frac{\mathrm{MHET}}{\mathrm{TPA}}$	$\frac{\mathrm{BHET}}{\mathrm{TPA}}$	$\frac{\mathrm{TPA}}{\mathrm{MHET}}$
pH 7.5	8.3	0.1	0.1	4.7	0.1	0.2
10 % DMSO	12.1	0.3	0.1	7.5	0.4	0.1
25 % DMSO	5.1	0.3	0.2	5.0	0.4	0.2
50 % DMSO	0.5	0.1	1.9	0.5	0.1	1.9
10 % EG	12.9	2.2	0.1	11.2	5.9	0.1
25 % EG	10.8	5.4	0.1	9.1	14.4	0.1
50 % EG	6.5	7.1	0.2	4.5	14.7	0.2
pH 5.0	2.4	0	0.4	2.4	0.1	0.4
pH 9.01	0.05	0	14.3	0.1	0	20.3

The two chosen enzymes were applied in reactions with different conditions and analyzed regarding their product profile after 1, 8 and 24 h reaction time. All reactions were performed in triplicates and under standard conditions (0.1 m phosphate buffer pH 7.5, 60 °C, 0.35 mm (=0.161 mg/mL) 3PET, 3.3 μg/mL (≈0.1 μm) enzyme) unless the varying condition stated. MHET/TPA, BHET/TPA and TPA/MHET ratios are listed for each reaction condition for the reaction time of 1 h, except for pH 9.0 after 24 h. Significantly higher TPA (blue), MHET (green) and BHET (purple) ratios are highlighted. Additional information can be found in Figure S2–S6.

Low DMSO concentrations (10–25 %) resulted in lowered hydrolysis towards TPA and similar or increased concentrations of MHET, except for IsPETase^wt, which produced significantly lower concentrations of MHET at 25 % DMSO. However, specifically at the beginning of the reaction (1 h), TPA and MHET amounts were generally both decreased, but the MHET/TPA ratios increased. Higher DMSO concentration (25–50 %) led to reduced concentrations of all products, with IsPETase^wt showing the lowest, probably due to the generally lower stability of this enzyme.[ ³⁶ , ⁵⁶ ]

Similarly, lower concentrations of EG (10–25 %) resulted in reduced TPA product concentrations, whereas 25 % EG almost eliminated the production of TPA. The MHET concentration was often slightly lowered at the beginning, but significantly higher or similar after 24 h at both EG concentrations. The MHET/TPA ratios are significantly increased at all reaction times and are generally higher compared to DMSO. As already shown in the literature with a few examples using HiC,[ ⁵⁸ , ⁵⁹ ] increasing EG concentrations went along with increasing BHET amounts, while BHET was otherwise only detected in negligible amounts. The BHET/TPA ratios were highest when using 50 % EG. The total product concentrations were reduced with increasing EG concentrations, although they did not decrease as significantly as with DMSO. Interestingly, and in contrast to DMSO, 25 % EG did not affect the total product yield, and in some cases even increased it slightly.

An acidic pH (5.0) resulted in significantly reduced TPA and MHET concentrations, showing a lower activity of PETases at this pH, as supported by literature reports.[ ⁴⁴ , ⁴⁸ , ⁵⁴ , ⁶⁷ ] In contrast, a basic pH (9.0) initially increased TPA and MHET concentrations, which might be attributed to enhanced enzyme activity combined with self‐hydrolysis of 3PET under basic conditions (Figure S2). The TPA concentration increased over the reaction time, while the MHET concentration decreased, leading to TPA as the only product after 24 h in all cases. This is probably due to the self‐hydrolysis of MHET and BHET to TPA at basic pH (Figure S8) so that the enzymatically produced MHET and BHET are being rapidly chemically hydrolyzed to TPA simultaneously with (enhanced) enzymatic hydrolysis. The basic conditions further improved the enzymatic activity as the product (i. e., MHET), which in most cases inhibits the enzyme, is removed continuously. Interestingly, IsPETase^wt generated the highest TPA concentrations under these conditions, whereas it had been engineered over the last years to overcome MHET inhibition to produce high TPA amounts.

In summary, low concentrations (10–25 %) of EG and DMSO reduced the TPA amounts and thus forced the hydrolysis to stop at the product MHET, with EG performing more efficiently. The basic pH resulted in TPA being the single product.

Transfer of Product‐Specific Reaction Conditions to Nano‐PET as Substrate

3PET is a model substrate that has been used in several studies to assess the activity of PET‐hydrolyzing enzymes.[ ⁴⁸ , ⁴⁹ , ⁵¹ , ⁵² , ⁵³ ] To verify our results, PET powder with a crystallinity of >50 % was changed into Nano‐PET, which was then used as a more realistic substrate (Figure 3–4, Figure S9‐S10, Table 2). Importantly, after preparation of the nanoparticles, the crystallinity is mostly eliminated, and highly amorphous particles are present. ^[70] Nano‐PET offers the advantages of a high substrate surface area, provides good accessibility for the enzyme, and enables a stable suspension that can be distributed with high accuracy. Similar amounts of Nano‐PET compared to 3PET were used and the most promising conditions yielding predominantly one product were applied (i. e., 10 and 25 % DMSO and EG in addition to pH 9.0; each compared to standard conditions).

Figure 3.

Screening of different reaction conditions during Nano‐PET hydrolysis by all two chosen enzymes. Product concentrations of TPA (blue), MHET (green) and BHET (pink) under different conditions (as indicated) were determined after 1, 8 and 24 h reaction time and shown in comparison to the hydrolysis under standard conditions (std.) (0.1 m phosphate buffer, pH 7.5, 60 °C, light bars) at each time point. For all conditions, 0.15 mg/mL Nano‐PET and 3.3 μg/mL (≈0.1 μm) enzyme were applied. All reactions were performed in triplicates, and the error bars represent the standard deviation. The background due to self‐hydrolysis of Nano‐PET under certain conditions (Figure S12) was subtracted. The reaction conditions yielding predominantly TPA (blue), MHET (green) or BHET (pink) in high concentrations compared to standard conditions are circled and summarized in the table.

Figure 4.

Total product concentrations for Nano‐PET hydrolysis. Total product concentrations of TPA, MHET and BHET identified the hydrolysis of Nano‐PET by IsPETase^wt and LCC^ICCG under different conditions (as indicated) after 24 h reaction time. The hydrolysis was performed using 0.15 mg/mL Nano‐PET, 3.3 μg/mL (≈0.1 μm) enzyme, and standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) unless the varying condition stated. All reactions were performed in triplicates, and the error bars represent the standard deviation. The background due to self‐hydrolysis of Nano‐PET under certain conditions (Figure S11) was subtracted and the total product concentrations for the entire reaction course is shown in Figure S10.

Table 2.

Effect of different reaction conditions on the product profile of Nano‐PET hydrolysis after 1 h (except for pH 9.0 after 24 h).

	IsPETasewt			LCCICCG
	$\frac{\mathrm{MHET}}{\mathrm{TPA}}$	$\frac{\mathrm{BHET}}{\mathrm{TPA}}$	$\frac{\mathrm{TPA}}{\mathrm{MHET}}$	$\frac{\mathrm{MHET}}{\mathrm{TPA}}$	$\frac{\mathrm{BHET}}{\mathrm{TPA}}$	$\frac{\mathrm{TPA}}{\mathrm{MHET}}$
pH 7.5	8.2	0.8	0.1	5.2	0.2	0.2
10 % DMSO	13.4	1.5	0.1	11.1	3.3	0.1
25 % DMSO	22.1	4.4	0.1	26.8	10.2	0.05
10 % EG	15.9	3.7	0.1	16.6	9.5	0.1
25 % EG	37.0	25.3	0.1	86.0	123.4	0.01
pH 9.0	0.1	0.4	12.7	0.1	0.7	13.8

The two chosen enzymes were applied in different reaction conditions and analyzed regarding their product profile after 1, 8 and 24 h reaction time. All reactions were performed in triplicates and under standard conditions (0.1 m phosphate buffer pH 7.5, 60 °C, 0.15 mg/mL Nano‐PET, 3.3 μg/mL (≈0.1 μm) enzyme) unless the varying conditions stated. MHET/TPA, BHET/TPA and TPA/MHET ratios are listed for each reaction condition for the reaction time of 1 h, except for pH 9.0 after 24 h. Significantly higher TPA (blue), MHET (green) and BHET (pink) ratios are highlighted. Additional information can be found in Figure 3 and Figure S9.

Low DMSO concentrations (10 %) resulted in significantly lowered product concentrations of TPA over time in all cases, while MHET remained similar. High DMSO concentrations suppressed TPA formation even more and almost stopped it in the case of LCC^ICCG and IsPETase^wt, whereas the MHET amount was similar or significantly increased after 24 h, but decreased at the beginning. Interestingly, BHET was produced at both DMSO concentrations in low but substantial concentrations. As before, increasing DMSO concentrations led to lower product concentrations.

When using EG instead of DMSO, very comparable results were observed, while typically higher product concentrations were observed for EG at both concentrations. In the case of MHET, significantly higher MHET/TPA ratios were observed, with the highest when using 25 % EG. Increasing the EG concentrations resulted in substantially higher BHET concentrations than before. The highest BHET concentration and BHET/TPA ratio were observed when LCC^ICCG was used in reactions with 25 % EG. In conclusion, EG was the superior solvent compared to DMSO for the preparation of MHET and BHET and both could be produced as predominant products with high concentrations. At 25 % EG, the increase in MHET/TPA ratio and decrease in TPA/MHET ratio was highest in all cases compared to standard conditions. Importantly, EG could also serve as a reactant that enables the esterification of TPA to MHET or MHET to BHET.

As in the experiments with the model compound 3PET, a basic pH (9.0) led to a significant reduction of MHET in favor of TPA, which is close to the single product after 24 h. Again, this can be attributed to the more efficient pH‐dependent chemical hydrolysis of MHET to TPA, but not to the hydrolysis of Nano‐PET (as in the case of 3PET), which is not prone to alkaline induced self‐hydrolysis (Figure S11).

It is noteworthy that the concentration of each hydrolysis product was generally higher in the case of Nano‐PET compared to 3PET. This might be due to the limited MHET/BHET/TPA molecules releasable in the case of 3PET compared to Nano‐PET, as only one molecule of MHET, BHET, or TPA can be released in the case of the 3x benzoic acid substrate 3PET. In addition, the solubility and thus accessibility of these substrates (i. e., PET‐source and intermediates/oligomers) can also have an influence in the respective solvent. ^[71] It was shown that self‐hydrolysis was significantly more pronounced for 3PET under specific reaction (e. g., 50 % EG or pH 9.0) conditions demonstrating the limitation of the model substrate. Apart from this, the results were comparable, suggesting that the conditions themselves influence the reaction, e. g., by solubilizing the intermediate MHET or reducing the hydrolysis rate of MHET or BHET due to the reduced concentration of water. In case of high EG concentrations, re‐synthesis of MHET and BHET intermediates from TPA instead of slower enzymatic hydrolysis may also be possible, which should be investigated in more detail in further studies.

Conclusions

PET is usually enzymatically depolymerized completely to TPA and EG, which are then isolated to re‐synthesize PET. As the conventional PET polycondensation begins anyway with BHET, which can be derived from the synthesis of TPA and EG or directly from PET glycolysis, this study aims to target MHET and BHET as the final products of biocatalytic PET hydrolysis by medium engineering for a greener PET resynthesis method.

Compared to DMSO, EG was the better solvent (and reagent) of choice as significantly higher MHET/TPA or BHET/TPA ratios were achieved with similar or higher total product concentrations. In particular, i) MHET was produced as a predominant product with the highest concentrations (>400 μm) and MHET/TPA ratios (16×) at the beginning of the reaction when IsPETase^wt and 10 % EG was used; ii) BHET was produced as a predominant product with the highest concentrations (about 200 μm) and BHET/TPA ratios (123×) at the beginning of the reaction when LCC^ICCG and 25 % EG was used; iii) TPA was produced as nearly single product with high concentrations (about 300 μm) and TPA/MHET ratios (13‐14×) at the end of the reaction with all enzymes at pH 9.0. Importantly, the increased production of MHET and BHET was most significant at the beginning of the reaction (1–8 h), while longer reaction times (24 h) resulted in increasing TPA concentrations due to MHET and BHET hydrolysis, albeit at a slower rate compared to standard conditions. Combined with in situ removal techniques, MHET and BHET should be constantly isolated and thus efficiently produced.

Experimental Section

General Information

All chemicals and solvents used were purchased from commercial distributors of analytical grade: Merck (Darmstadt, Germany), Carl Roth (Karlsruhe, Germany), Thermo Fisher Scientific (Waltham, MA, USA), TCI Europe (Zwijndrecht, Belgium), Th.Geyer (Renningen, Germany), BLD Pharmatech (Reinbek, Germany) and used as received. The plasmids for all enzymes were provided by the respective scientists mentioned in Table S1, except for Turbo‐PETase and Cut190**SS/L136F, which were synthesized by BioCat GmbH (Heidelberg, Germany).

Chemical Synthesis of MHET

MHET was used for hydrolysis experiments and as a HPLC standard and thus chemically synthesized according to the recently published adapted protocol. ^[72] Briefly, 8.4 mmol KOH and 8.7 mmol BHET were dissolved in ethylene glycol (540 mL per approx. 80 g substrates) and heated at 120 °C for 3 h. The solution was transferred to a separating funnel and water (900 mL per 540 mL EG) was added. After washing the solution three times with chloroform (150 mL per 540 mL EG), the aqueous solution was cooled to 4 °C and then acidified to pH 3.0 to precipitate MHET. The solution was filtered, the filtrate was washed with water and thereby discolored. After drying the filtrate, the white product (MHET) was milled and characterized by ¹H− and ¹³C‐NMR (Figure S12) and HPLC analysis. According to HPLC analysis, the resulting MHET powder consisted of 83.9±1.7 % MHET, 13.4±0.9 %TPA and 2.6±1.2 % BHET.

Chemical Synthesis of 3PET

3PET was used as a model substrate in this study and chemically synthesized according to an published adapted protocol. ^[73] Benzoyl chloride and 2‐chloroethanol in an excess of 1.2 was heated at 110 °C for 24 h under a reflux condenser. After removal of residual 2‐chlorethanol under reduced pressure (5 mbar) at 60 °C, 1 mol TPA per 2 mol product was dissolved in dimethylformamide (DMF, 500 mL per 50 g of product) and added to the product together with triethylamine (in an equal molar amount as the product) and heated at 140 °C for 24 h under a reflux condenser. After removal of DMF under reduced pressure (5–10 mbar) at 60 °C, the brown residue was dissolved in toluol (equal volume as DMF), the suspension was filtered, and the filtrate was washed with toluol to dissolve the product out of the filtrate. Toluol was removed under reduced pressure (5–25 mbar) at 60 °C, and the resulting yellowish residue was dissolved in methanol (double the volume of DMF). The product was filtered, and the filtrate was washed with methanol and thereby discolored. After drying the filtrate, the white product (3PET) was mortared and characterized by ¹H− and ¹³C‐NMR‐spectroscopy (Figure S13). This substrate is only artificial, as it will not be generated naturally, as this variant lacks the two terminal carboxyl groups. Consequently, artificial, non‐natural hydrolysis products can be generated (i. e., benzoic acid (BA), 2‐hydroxyethyl benzoate (HEB), 2‐(benzoyloxy)ethyl(2‐hydroxyethyl)terephthalate, 4‐((2‐(benzoyloxy)ethoxy)carbonyl)benzoic acid), which have also been partially recognized in other studies.[ ⁴⁸ , ⁴⁹ , ⁵¹ , ⁵² , ⁵³ ] BA is ultimately always produced and is thus one of the main products in addition to the naturally produced TPA, MHET and BHET (Figure S7). The other products were not detected in this and most other studies, indicating their quick degradation (Figure S15).

Production of Nano‐PET

The model substrate Nano‐PET was prepared as described in the literature. ^[74] Briefly, 1 g of PET was dissolved in 10 ml of hexafluoroisopropanol (HFIP) (ratio of 1 : 10 (PET:HFIP)) until fully dissolved. The resulting solution was then added dropwise to 100 ml of water (ratio 1 : 10 (PET/HFIP:H₂O)), under vigorous stirring, to form a suspension. This suspension was subsequently filtered through a paper filter. Finally, a rotary evaporator was used to remove the HFP, which was collected for reuse. An additional step involved transferring the final solution to a graduated cylinder and allowing it to settle for 2 hours. At high suspension densities, two layers formed, with the upper layer being more stable. This upper layer was decanted and remained stable for up to several months. After evaporating a fixed volume of Nano‐PET, a concentration of 1.8 mg/ml was detected.

Transformation, Gene Expression and Protein Purification

All enzymes used in this study were produced as follows: E. coli Shuffle T7 and E. coli BL21(DE3) were transformed with the respective plasmid (Table S1) by chemical transformation. Briefly, cells were incubated with 10 ng plasmid for 30 min on ice, 2 min at 42 °C and again 5 min on ice. Afterwards, the cells were plated on LB‐agar plates using the respective antibiotic at 37 °C. One colony was used to inoculate a preculture, which was incubated overnight at 37 °C and used to inoculate the expression culture with an optical density at 600 nm (OD₆₀₀) of 0.1. Cells were cultured until an OD₆₀₀ of 0.6–0.8 was reached and protein expression was induced by adding the respective IPTG concentration (Table S1). Protein expression was performed at 20 °C for 16 h. Cells were then harvested by centrifugation and frozen until use. For enzyme purification, the frozen cell pellets were dissolved in 5 % elution buffer (50 mm Tris pH 8.0, 150 mm NaCl, with varying % imidazole, whereas the percentage describes the percentage amount of imidazole and 100 % corresponds to 300 mm). After the addition of 0.15 mg/mL lysozyme and 0.015 mg/mL DNAseI, the suspension was incubated for 20 min at room temperature and 15 min at 4 °C, after which the cells were lysed on ice using ultrasonication. Insoluble fragments were removed by centrifugation, the supernatant was filtered (0.22 μm filter) and applied to HisTrap^TM HP columns (Ni2+‐NTA affinity chromatography) using syringes. Purification was then utilized with a step gradient (5, 10, 15, 70, 100 % elution buffer) using a peristaltic pump, eluting the enzymes with the 70 % elution buffer. The enzyme solutions were buffer exchanged (50 mm Tris pH 8.0, 50 mm NaCl), concentrated using Amicon^® Ultra‐15 centrifuge filter units, and stored at −20 °C.

Enzymatic Reactions

These were performed in a total volume of 1.6 mL in 2 mL reaction vessels with 3PET (0.35 mm=0.161 mg/mL) or Nano‐PET (0.15 mg/mL) as substrate. 100 nm (≈3.3 μg/mL) enzyme and standard conditions (0.1 m phosphate buffer, pH 7.5, 60 °C) were applied, except the condition varied as mentioned in the respective sections (i. e., pH 5.0 (0.1 m sodium acetate buffer) or 9.0 (0.1 m bicine buffer) instead of pH 7.5; addition of final 10, 25, 50 % DMSO or EG; controls without enzyme). At each reaction time point (1, 8, 24 h), 500 μL were acidified with 9 μL 37 % HCl (final pH of 1–2), heated at 95 °C for 10 min and centrifuged to remove the enzyme. 450 μL of the supernatant were then diluted with an equal volume of the analytic solution (0.1 m phosphate buffer, 20 % DMSO, pH 5.0) to a final volume of 900 μL (final pH of 1.5–4) for HPLC measurements. Notably, the HPLC solutions had to be acidified to pH below 4.0, as otherwise, TPA behaved differently (Figure S14). All reactions were performed in triplicates.

MHET/BHET Stability Experiments

250 μm BHET or MHET were dissolved in 0.1 m phosphate buffer and the pH was adjusted to 5.0, 7.5 or 9.0. After an incubation time of 30 min at room temperature, the solutions were treated analogously to the enzyme reactions (i. e., acidified, heated, centrifuged, diluted) and analyzed by HPLC measurements.

HPLC Measurements

Quantification of TPA, MHET and BHET was realized via the Chromaster® HPLC system (VWR, Darmstadt, Germany) whereas the retention times of the analytes were determined through measurement of commercial TPA and BHET and the synthesized MHET. Since TPA had the highest purity (>99 %), and TPA, MHET and BHET share the same absorption behavior due to the same number of absorbing structures at the specific wavelength, all compounds were quantified on the base of a linear calibration curve from TPA in a range from 3–250 μm calibration curve. The hydrolyzed products were separated with an Eclipse XDB−C18 (4.6×12.5 mm; 5 μm) guard column and a Zorbax Eclipse XDB−C18 (4.6×250 mm; 5 μm) column (Agilent Technologies, Waldbronn, Germany) run under isocratic conditions at 35 °C. The eluent was composed of 25 % acetonitrile and 75 % (water+0.1 % formic acid) (v/v) and was pumped with a flow rate of 0.8 mL/min. Each run lasted 16 min for which 50 μL were injected and the analytes were detected with an UV detector at 241 nm. Negative effects of the cosolvents on the detectability of the analytes were not observed.

NMR Measurements

NMR spectra were recorded on an AV III Bruker‐BioSpin (¹H: 400,13 MHz; ¹³C: 100,62 MHz) or Avance Neo Bruker BioSpin (¹H: 600,13; ¹³C: 150,9). ¹H and ¹³C shifts were referenced to internal solvent resonances and reported in parts per million relative to TMS. The solutions (50–60 mg/mL) were prepared in DMSO‐d6.

Determination of Protein Concentrations

To determine the total protein concentration of enzyme solutions, the Pierce™ Coomassie Protein Assay Kit from ThermoFisher Scientific (Bradford Assay, article number 23200) was used with bovine serum albumin as the calibration standard.

SDS‐PAGE

Protein samples after enzyme purification were incubated with SDS sample buffer (50 mm Tris‐HCl pH 6.8, 12 % glycerol, 4 % SDS, 0.01 % bromophenol blue) and 100 mm DTT at 95 °C for 5 minutes. Bio‐Rad mini gels were used for SDS‐PAGE with 5 % stacking gels and 12 % running gels, and electrophoresis was performed at 150 V per gel. After completion, the gels were washed with water, stained with colloidal Coomassie staining solution (20 % ethanol, 10 % ammonium sulphate, 3 % orthophosphoric acid, 0.1 % Coomassie G‐250) for 24 h and destained with water.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Heinks T., Hofmann K., Last S., Gamm I., Blach L., Wei R., Bornscheuer U. T., Hamel C., von Langermann J., ChemSusChem 2025, 18, e202401759. 10.1002/cssc.202401759

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.