Ethylene Glycol‐Guided Enhancement of Bis(2‐Hydroxyethyl) Terephthalic Acid as a Primary Product of Enzymatic Poly(Ethylene Terephthalate) Depolymerization
Tobias S. Radmer, Thore B. Thomsen, Leon T. Krinn, Anne S. Meyer
TL;DR
This study shows how adding ethylene glycol can boost the production of a key recycling product during enzymatic breakdown of PET plastic.
Contribution
The novel finding is that ethylene glycol enhances BHET yield by altering enzyme behavior on PET surfaces.
Findings
Adding 27–29% ethylene glycol maximizes BHET production using the LCCICCG enzyme.
Ethylene glycol reduces the lag phase of enzymatic reactions and lessens the impact of PET crystallinity.
EG affects enzyme adsorption, increasing substrate affinity and directing enzyme activity to the liquid phase.
Abstract
Recycling of enzymatically depolymerized poly(ethylene terephthalate) (PET) involves polycondensation of bis(2‐hydroxy‐ethyl) terephthalic acid (BHET)—a degradation product of enzymatic PET hydrolysis. The recycling process is simplified when more BHET is generated by the enzymatic reaction. Here, we report how ethylene glycol (EG) addition can maximize BHET formation using leading PET hydrolases, LCCICCG, and PHL7. EG at any level above 2–5% vol/vol was found to decrease the steady‐state enzymatic degradation rates while enhancing the relative production of BHET. For LCCICCG, the highest measured BHET levels (product fraction approaching 0.5) were attained at EG levels of ∼27–29% and reaction temperature ∼62.5°C. EG shortened the enzymatic reaction lag‐phase and lowered the lag‐phase increase with PET crystallinity. EG works by perturbing the adsorption, including nonproductive…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8
FIGURE 9
FIGURE 10
FIGURE 11- —Villum Fonden10.13039/100008398
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
Topicsbiodegradable polymer synthesis and properties · Polymer crystallization and properties · Microplastics and Plastic Pollution
Introduction
1
Poly(ethylene terephthalate) (PET) is a semi‐crystalline synthetic polymer consisting of alternating ethylene glycol (EG) and terephthalic acid (TPA) units. It has become one of the most commonly used plastic materials with a global production of more than 80 million tons annually, of which a vast majority is used in single‐use applications such as single‐use beverage bottles, and predicted to continue to grow [1], thus generating enormous amounts of PET waste [1, 2]. Only ∼14% of PET waste is currently recycled, and a significant amount of PET ends up as plastic pollution in nature [3, 4]. It is therefore imperative that recycling technologies are further developed.
Recent developments show that engineered PET hydrolases can deconstruct PET efficiently for recycling of PET at scale [5]. Successful enzymatic recycling involves resynthesis of PET to a material of comparable quality to virgin petrochemical PET. For this to happen, enzymatic depolymerization is considered to require degradation of PET into its monomers, TPA and EG. To resynthesize PET, these monomers undergo an esterification process to bis(2‐hydroxyethyl) terephthalic acid (BHET) moieties, which are combined by polycondensation to PET polymers [6]. However, in practice, the enzymatic PET degradation process with PET hydrolases progresses via production of a mixture of soluble oligomeric (mono‐, di‐, tri‐, and tetra‐aromatic) fragments of varying composition during the course of reaction, depending on the type of enzyme, PET pretreatment, and reaction conditions [7, 8]. This complicates the delivery of a product stream enabling efficient repolymerization and incentivizes the design of the PET hydrolase reaction to yield a more uniform product profile.
Using EG as an additive to the enzymatic reaction medium has shown potential to alter the product profile of enzymatic depolymerization [9]. With EG in the reaction medium, a larger portion of dissolved products consist of esterified monomers, mono(2‐hydroxyethyl) terephthalic acid (MHET) and BHET, with up to 88% BHET, suggesting that PET hydrolases can catalyze a glycolysis reaction in the presence of EG [9]. Glycolysis signifies a PET depolymerization reaction that produces bis(2‐hydroxyethyl) terephthalate (BHET) and EG. With BHET as the main product of the PET hydrolase catalyzed depolymerization, the resynthesis is simpler as the initial esterification step of the PET resynthesis process may be omitted, which could dramatically reduce the cost of enzymatic PET recycling. However, only limited research has been done on the influence of EG on enzymatic PET depolymerization at different reaction conditions.
In this study, we investigate the effect of EG on the enzymatic PET depolymerization reaction using three different thermostable, gold standard PET hydrolyzing enzymes in individual reactions: LCC_ICCG_, an engineered variant of leaf‐branch compost cutinase (LCC) [10], DuraPETase, an engineered variant of the Ideonella sakaiensis PETase [11], and Polyester Hydrolase Leipzig 7 (PHL7) [12], a PET hydrolyzing enzyme known to have a different electrostatic surface potential pattern and a different attack pattern on PET than LCC_ICCG_ and DuraPETase [13]. The overall objective of the present study was to gain knowledge of the effects of EG on various factors of the reaction system, including enzyme stability, kinetics, substrate parameters and the products themselves, and determine the optimal reaction conditions for BHET formation during enzymatic PET degradation.
Results and Discussion
2
Effects of Ethylene Glycol on Substrate and Enzyme Properties
2.1
To assess the effect of EG on the glass transition temperature (T g) of the PET substrate, a series of experiments were conducted with PET samples incubated at varying EG contents at 53°C and 62.5°C. The data revealed that the bulk T g of the incubated PET samples increased linearly with the EG content, and that the incubation at the higher temperature resulted in lower bulk T g values and a similar slope (Figure 1A).
(A) Bulk T g of PET samples soaked overnight in mixtures of water and EG at 53°C or 62.5°C. (B) Evolution of the estimated surface T g of PET samples soaked for 1 h at temperatures ranging from 56°C to 72°C in defined mixtures of water and EG. EG = ethylene glycol; PET = poly(ethylene terephthalate).
This observation aligns with previous data showing that bulk T g is decreased by the absorption of water, and that the bulk T g is lowered to a higher extent at elevated temperatures [14].
The surface T g of the PET was also affected by an increase in EG levels (Figure 1B) but to a smaller extent than bulk T g. The surface T g was measured by assessing the decrease of the enthalpy relaxation (aging) peak area by differential scanning calorimetry (DSC) during incubation of PET discs at different temperatures in defined mixtures of water and EG (Supplementary Materials Figure S1). The surface T g data (Figure 1B) support the interpretation that PET absorbs water, and that this absorption lowers T g, while the effective surface T g of the samples increases with EG content. No changes were seen in the substrate crystallinity level (X C) or MAF (data not shown), but a weight increase was observed in all samples during incubation, with the weight increase being smallest at high EG levels (data not shown). It is therefore likely that EG does not directly affect the T g of PET. Instead, EG most likely decreases the plasticizing effect that water has on PET through lowering of the water activity.
When the denaturation temperature, T m, of each of the three enzymes (LCC_ICCG_, DuraPETase, and PHL7) was measured in response to an increasing EG content, it was observed that the denaturation temperature of all three enzymes was negatively affected by EG (Figure 2). During these experiments, it was also observed that increased levels of EG apparently decreased the thermal robustness of the enzymes differently at low and high pH, and that the PHL7 was less affected than the other two enzymes (Figure 2). The data also show that the decrease in T m with increasing EG content is small. This small effect is therefore not expected to have any significant influence on the robustness of the enzymes during the relatively short reactions in this study.
Response of protein denaturation temperature, T m, of LCCICCG, PHL7, and DuraPETase to increased EG content, measured at two different pH levels. EG = ethylene glycol.
Effect of EG on BHET
2.2
Spontaneous hydrolysis of BHET progressed relatively fast in water and was slower with higher EG levels (Figure 3). The degradation rate data thus showed a clear dose–response effect of EG addition, with the spontaneous hydrolysis being essentially retarded at 40% EG (Figure 3) during a 16‐hour period at room temperature.
Evolution of BHET purity with time in samples initially containing pure BHET in a mixture of water and ethylene glycol (EG). BHET = bis(2‐hydroxyethyl) terephthalate.
Enzyme Kinetics Response to EG (LCCICCG)
2.3
To assess the effect of EG levels on enzyme activity, kinetics experiments with LCC_ICCG_ were conducted at a range of EG contents in the enzymatic reaction medium. Increased EG content generally decreased the enzymatic rate of PET degradation, and EG levels at 25% and above thus halted the reaction (Figure 4A).
Michaelis–Menten enzyme kinetics with LCCICCG at a range of ethylene glycol (EG) contents. Substrate inhibited Michaelis–Menten functions are fitted to the measured data. (A) Conventional Michaelis–Menten kinetics at 5 nM enzyme dosage. Functions were fitted with a shared K M of 2.4 ± 0.75 PET discs and a shared K i of 10.6 ± 5.3 PET discs (Individual K M, K i, and k cat values are given in Table S1, Supplementary Materials). (B) Inverse Michaelis–Menten kinetics, varying the enzyme concentration at a substrate concentration of one 6 mm PET disc. PET = poly(ethylene terephthalate).
In the kinetics modelling, a global fit with a shared K m across all EG contents resulted in the best fit (highest R^2^), which confirmed that the presence of EG did not affect the K m, whereas the initial rate (and V max) decreased with increased EG contents.
At all EG levels, an apparent decrease in product release rate was observed at a substrate concentration above approximately 2 × K M, when fitting the data points obtained (Figure 4A) to a Michaelis–Menten function describing substrate inhibition:
This apparent inhibition effect has previously been observed in aqueous systems without EG [11], but it is most likely not only a substrate inhibition effect. The substrate inhibition effect may thus be exacerbated as a consequence of measuring the enzymatic rates based on product release. Soluble product release during enzymatic degradation requires multiple adjacent bond cleavages within the same PET polymer. As the surface area of the insoluble substrate increases in the form of the number of PET discs (Figure 4A), the probability of adjacent bond cleavages, and thus the likelihood of soluble product release to occur decreases, which manifests as a lower rate. This phenomenon occurred to the same extent at every level of EG (Figure 4A).
In contrast, the seemingly similar effect observed with increasing enzyme concentration but with constant substrate level (Figure 4B) might be the result of completely different phenomena, namely either that the concentration of enzymes exceeds the molar concentration of accessible surface sites, and/or that unproductive adsorption of the enzyme to the substrate surface occurs. Such unproductive adsorption will cause a steric hindrance to proper enzymatic attack on the cleavage sites, thereby decreasing the product release rate as the enzyme addition levels increase.
The phenomenon of a limited number of accessible surface attack sites agrees with the theory of interfacial enzyme catalysis [15, 16]. In this case, the initial rate, v i, is described to be a function of the inverse maximal reaction rate, ^inv^ V max, and the inverse Michaelis‐constant, ^inv^ K _ M _, respectively [15], according to the following equation with E 0 being the enzyme concentration:
According to this approach, the density of the attack sites on the PET substrate surface is calculated to be represented by the term Γ_max_ [15, 16]. Indeed, in the dose–response experiments with increased enzyme addition levels, inverse Michaelis–Menten kinetics modeling affirmed that increased enzyme concentration at a fixed substrate level (one PET disc) showed a profound increase in ^inv^ K M with EG, whereas the apparent ^inv^ V max did not change (Figure 4B).
The rate at low enzyme concentrations was highest at low EG (Figure 4B), and the onset of the rate decrease occurred at lower enzyme levels with lower EG. Evidently, with higher EG, the product release rate continued to increase with increasing enzyme concentrations above the enzyme concentration supporting ^inv^ V max at 0% EG (Figure 4B); this effect occurs as the ^inv^ K M increases with EG.
Similar effects of high EG levels on PET hydrolase activity have been reported by assessment of endpoint product levels [11, 17]. In medium engineering studies [17] using LCC_ICCG_ and EG addition levels of 10%, 25%, and 50% v/v, the total product concentrations of TPA, MHET, and BHET after 24 h of LCC_ICCG_ catalyzed hydrolysis of (bis(benzoyloxyethyl)terephthalate) (a PET model substrate) significant reduction of the total product levels were observed with 50% EG, whereas the levels obtained with 10% and 25% EG were similar [17]. EG cannot be considered to increase the enzymatic rate with increased enzyme dosage, as the overall enzymatic rates recorded with the same enzyme dosage clearly decreased with EG (Figure 4B). Instead, EG may hinder the unproductive adsorption of enzyme protein to the PET surface, which in turn essentially results in an increased Γ_max_. It thus appears that EG decreases the polarity of the reaction medium which affects the substrate affinity.
Enzyme Distribution
2.4
To test the theory that EG affects the substrate affinity of PET hydrolases by altering the reaction medium polarity, an assay was performed to compare the activity of enzyme adsorbed to the disc and suspended in the medium (Figure 5). The data verify that as EG content is increased, there is less enzyme on the PET substrate and more in the medium (Figure 5).
Distribution of LCCICCG between the substrate surface and liquid reaction medium. The distribution was measured as the activity of enzyme transferred with a PET disc after preincubation (Absdisc) versus enzyme left in the original medium (Abssolution). PET = poly(ethylene terephthalate).
A plausible explanation is that higher EG content confers a lower binding power (or a higher dissociation coefficient) between the nonpolar enzyme and the PET substrate due to the lower polarity of the aqueous reaction medium with EG present. The suggested mechanism is thus that the EG induces either a slower association rate or a faster dissociation—a changed overall affinity—between the enzyme and the substrate. In case the polarity induced change in affinity is correct, the intersect of the linear fit with the X axis at ∼27–29% EG in Figure 5 reflects the point where the medium has a similar polarity to PET, and the enzyme is not adsorbed to the PET.
Optimization of BHET Production
2.5
The dissolved products of depolymerization reactions with LCC_ICCG_ were measured at a range of reaction temperatures and EG contents. The results were logarithmically normalized and used to fit a response model and estimate the optimal conditions for BHET production (Figure 6).
Contour plots of surface models fitted to enzymatic product concentration obtained at a range of temperatures and ethylene glycol contents at 200 nM LCCICCG at pH 8.5. (A) Total dissolved product concentrations, (B) BHET concentrations. The black point in each graph marks the reaction condition that yields the highest BHET concentration. BHET = bis(2‐hydroxyethyl) terephthalate.
Both surface models follow a similar diagonal pattern, although the optimal conditions for achieving highest BHET concentration is at a lower temperature, ∼64°C, and a higher EG content of ∼22% (Figure 6B) than the total product yield optimum which was modeled to be at ∼74–75°C and EG ∼ 5%.
Since the data showed that the BHET fraction increased with the EG content and at temperatures above ∼70°C, i.e. above the realistic reaction conditions, no reaction optimum could be found for the BHET fraction of the released products (Figure 7).
Contour plot of BHET fraction of total dissolved products observed in the optimization experiment behind Figure 6. BHET = bis(2‐hydroxyethyl) terephthalate.
A very high BHET purity up to 99% was observed at conditions with minimal enzyme activity, and excluded from the models in Figure 6, but these findings indicate a preference for releasing BHET as the initial dissolved product, which concurs with findings recently published by Heinks et al. [17].
To assess whether the observed trends are unique to LCC_ICCG_ or generally applicable for other PET hydrolases, an optimization experiment was conducted with two other PET hydrolases, namely PHL7 and DuraPETase,
A similar pattern was observed with PHL7 (Figure 8), and DuraPETase (Supplementary Materials, Figure S2 and S3) showed a similar effect of EG, as total product formation was decreased by increased EG levels, while BHET concentrations were increased.
Contour plots of empirical surface models fitted to enzymatic product concentration obtained at a range of temperatures and ethylene glycol contents at 200 nM PHL7 at pH 8.5. (A) Total dissolved product concentrations (logarithmic scale mM); (B) BHET concentrations (logarithmic scale mM). The dark points in the diagrams mark the condition with the highest observed BHET concentration. BHET = bis(2‐hydroxyethyl) terephthalate.
It is tempting to infer that EG alters the catalytic mechanism, such as the mixed glycolysis which has been suggested previously (but not proven experimentally) to explain the altered product profile [7]. However, these findings of a similar effect on different enzymes indicate that it may be the reaction conditions rather than the mechanism which is responsible for the increased formation of BHET as EG is increased up to 20‐25%, while total product formation declines.
The optimum for BHET yield therefore represents a ‘sweet spot’ between high enzymatic PET degradation rate (low EG/high temperature) and a low enzymatic‐ and spontaneous hydrolysis rate of BHET (low temperature/high EG).
This interpretation is corroborated by the findings that overall, EG was observed to gradually decrease TPA production and introduce a delay in TPA formation, while the MHET production rate was not significantly affected until 30% EG. The data thus infer that the apparently decreased TPA production is likely a result of a decreased BHET (and MHET) hydrolysis rate as EG levels increase. This interpretation is in agreement with findings of Heinks et al. that MHET levels are significantly higher in longer reactions containing EG, in accord with the observation that conversion of MHET to TPA is decreased [17].
When monitoring product profile progressions, the relative levels of BHET in the product mixture increased proportionally with EG (Figures 7 and 9), and BHET concentrations followed a classical limited growth function by time. Of the three sampled conditions (10, 20 and 30% EG), 20% EG resulted in the highest BHET concentration (Supplementary Materials Figure S4).
BHET fraction of total dissolved monomeric products in LCCICCG reactions at 61.5°C with 10, 20, and 30% ethylene glycol. BHET = bis(2‐hydroxyethyl) terephthalate.
The BHET optimum thus appears to be a compromise between enzyme activity and BHET stability. In turn, it is tempting to infer that it is possible to compensate for the decreased activity by increasing the enzyme concentration. An optimization assay was therefore performed at two additional enzyme concentrations, and the optimal conditions for high BHET levels were found at roughly the same temperature of 60–65°C, while the optimal EG level increased with enzyme dosage (Supplementary Materials Figures S5 and S6). The finding that the optimal temperature was similar at different enzyme concentrations supports a global sweet spot between enzymatic product formation and product hydrolysis.
A more detailed examination of the optimal EG level by enzyme concentration was therefore performed at the optimal temperature (Figure 10). The optimal EG level increases with enzyme concentration as expected, as the higher enzyme concentration compensates for decreased PET affinity. The optimal EG level for maximal BHET level by enzyme concentration approaches a ceiling of ∼28%, i.e., the affinity threshold observed in the distribution assay (Figure 5). As BHET is hydrolyzed slower at high EG, the BHET yield increases with EG. However, as the substrate affinity decreases with EG content, additional enzyme is required at the higher EG levels to reach the highest possible BHET concentration level. However, going above the ∼ 28% EG ceiling is not beneficial, as most of the enzyme will remain in solution in the reaction medium as the affinity drops. Consequently, the amount of enzyme reacting with dissolved products rather than insoluble PET increases. Increasing the enzyme load while at the maximum EG level indeed seemed to increase the rate of enzymatic BHET hydrolysis (red arrows, Figure 10B) more than the enzyme catalyzed rate of PET degradation (green arrows, Figure 10B), de facto reducing the BHET level achieved (Figure 10B).
(A) The ethylene glycol content yielding the highest BHET concentrations versus the LCCICCG concentration at reactions at 62.5°C and pH 8.5. (B) The corresponding highest BHET concentrations observed under these reaction conditions for each enzyme concentration. The dotted line marks a turning point at which increasing the enzyme concentration no longer increases BHET concentrations. Green arrows illustrate relative enzyme catalyzed BHET formation via PET degradation; red arrows illustrate relative levels of enzymatic BHET degradation (size of arrows are meant as an illustration, not to scale). The specific optimum for BHET production by LCCICCG was thus found to be 62.3 ± 0.9°C, pH 8.50, 380 nM LCCICCG and 27% v/v ethylene glycol, marked in these figures by a blue point. BHET = bis(2‐hydroxyethyl) terephthalate; PET = poly(Ethylene Terephthalate).
Therefore, the optimal condition for maximizing BHET yield occurs right before this EG limit is reached, as evident at the specific reaction condition of 380 nM LCC_ICCG_ and 27% EG (Figure 10B). Interestingly, the EG ‘ceiling’ of 27‐29% coincides with the point where the PET surface T g was found to be 63.5°C (Figure 1), which was above the reaction temperature of 62.5°C. Hence, it cannot be ignored that this EG ceiling of ∼27% could also be related to EG inducing a slight increase in surface T g.
Duration of the Lag Phase
2.6
To assess whether EG affects the duration of the lag phase in enzymatic degradation of more crystalline PET, continuous depolymerization reactions with a range of PET crystallinities and EG contents were measured in a modified microplate (Figure 11).
The length of the observed lag phase for LCCICCG catalyzed product release versus sample crystallinity (X C) at various ethylene glycol (EG) contents. The high X C value of 19% for the 0% EG sample in parentheses is an estimate.
As observed in previous studies [13, 16], the lag phase duration was linearly correlated with the degree of PET crystallinity (X C), and EG significantly reduced this effect, for example by 3.5 times at 30% EG (Figure 11). This observation infers that sterically blocked cleavage sites contribute to the lag phase due to unproductive enzyme adsorption to the substrate surface [13, 18]. As EG decreases the PET affinity of the enzyme, unproductive adsorption is likely also decreased. We therefore suggest that the observed reduction in the lag phase duration with increasing EG is influenced by the EG causing less unproductive adsorption in accord with the observations that less enzyme adsorbs to the PET disc versus enzyme in solution with increasing EG (Figure 5).
Methods
3
Sample pH Prediction
3.1
As the pH of a Glycine‐NaOH buffer is affected by both EG and temperature, a model was created to predict the sample pH by EG. The dataset contained 59 evenly distributed measurements and covered 0–60%v/v EG, 20–75°C, and pH 8.6–10.6 of 1 M stock at room temperature. R ^2^ of the prediction model was 0.9787, and it was used to adjust 1 M Glycine‐NaOH buffers at room temperature to reach the desired pH at specific sample conditions.
PET Substrate Preparation
3.2
‘Amorphous’ discs, measured to have a crystallinity, X C, of 8.2 ± 1.8%, were cut from a 1 mm thick PET sheet (cat. No. ES30301 from Goodfellow Cambridge Ltd, Huntingdon, UK) using a Ø 6 mm stationary hole puncher as described earlier [19].
The crystallinity of the PET samples was modified by annealing three discs in a 2 mL Eppendorf tube in a heating block at 115°C for a specified duration of time, followed by quenching in an ice bath for 30 s [19].
Differential Scanning Calorimetry of PET Samples
3.3
DSC was performed on a TGA/DSC 1 system (Mettler Toledo, Greifensee, Switzerland) in two phases. In phase 1, the temperature was raised from 20 to 85°C at a rate of 40°C/min, held for 1 minute and lowered from 85 to 20°C at ‐40°C/min. By holding the sample above T g for a short period, the effect of enthalpy relaxation is reversed, removing the resulting ‘ageing peak’. In phase 2, the sample was heated from 20 to 270°C at 10°C/min. X C, X MAF, X RAF, and T g were calculated from the results of phase 2. Note: As described in the case of surface T g determination (see below), phase 1 was omitted, and the area of the aging peak was recorded.
Substrate Soaking
3.4
Two amorphous PET discs (total weight 66.16 ± 0.52 mg) were incubated in 500 µL reaction mixture containing 100 mM glycine‐NaOH buffer at pH 9 with a range of 0‐80% EG by volume, contained in 2 mL Eppendorf tubes at 53°C or 62.5°C overnight (16 h). The discs were padded dry and analyzed by DSC.
Surface T
g
3.5
An amorphous disc was incubated at a range of temperatures for 1 h in 500 µL samples containing 400 mM glycine‐NaOH buffer adjusted to pH 8.5 and with varying EG contents from 0 to 60%v/v. The disc was then patted dry using a paper towel, and the aging peak was measured by DSC on a Pyris 1 Calorimeter (Perkin Elmer, Waltham, Massachusetts, USA). To remove the aging peak the sample was rapidly preheated (at 40°C/min.) to 85°C, held for 1 min., and then rapidly cooled (at −40°C/min.) to 20°C. The sample was then heated to 270°C at 10°C/min., and the DSC curve collected. T g, ΔH m, ΔH cc and Δc p were obtained and used to estimate X c, X MAF and X RAF [19, 20].
Denaturation Temperature
3.6
The thermal unfolding program of the Prometheus Panta (NanoTemper Technologies GmbH, München, Germany) was used to determine the denaturation temperature, T m, of enzyme samples in 400 mM Glycine‐NaOH buffer adjusted to pH 9.0 at 65 and 85°C, respectively.
Enzyme Distribution
3.7
Four samples were incubated simultaneously at 62.5°C, containing 400 mM Glycine‐NaOH buffer adjusted to pH 8.5 and an EG content between 0 and 30%v/v in Protein LowBind Eppendorf tubes (Sarstedt AG & Co. KG, Nümbrecht, Germany). Samples 1 and 4 contained an initial enzyme concentration, while samples 2 and 3 contained no enzyme. An amorphous PET disc was added to sample 1, and after a 5‐minute preincubation, the disc was removed using tweezers and patted dry using paper towel. The preincubated disc was then quickly added to sample 2, while fresh discs were added to samples 1 and 4. All samples were incubated for 90 min, and the absorbance was measured by UV‐Vis spectrometry in UV‐transparent microplates (Greiner Bio‐One, Kremsmünster, Austria) using an Epoch2 plate reader (BioTek, Winooski, Vermont, USA). The absorbance of sample 3 was subtracted from all samples as background, and the ratio Abs_2_/Abs_1_ was defined as the distribution of enzyme between the substrate surface and reaction medium.
Ratios were included only for conditions in which both samples had an absorbance lower than the respective sample 4 to ensure that the substrate was not saturated.
Enzyme Kinetics
3.8
1000 µL samples in 2 mL Protein LowBind Eppendorf tubes (Sarstedt AG & Co. KG, Nümbrecht, Germany) were incubated at 62.5°C with EG contents ranging from 0 to 30%v/v. Four 20 µL samples were pipetted over a reaction time between 1.5 and 2.5 h, and the absorbance at 240 nm was measured by UV‐Vis spectrometry in UV‐transparent microplates (Greiner Bio‐One, Kremsmünster, Austria) using an Epoch2 plate reader (BioTek, Winooski, Vermont, USA). From these measurements, the soluble product release rates, measured in TPA_eq_, were calculated, and substrate inhibited Michaelis–Menten functions were fitted using Origin Pro 15. To account for any background from EG oxidation products, controls of each EG content were always included and subtracted from the sample absorbance.
For conventional Michaelis–Menten kinetics, samples with 5 nM enzyme were incubated with substrate loadings from 0 to 15 ‘amorphous’ discs. To achieve loadings of 0.5 and 0.25 discs, respectively, whole discs were cut in half and quarters using scissors to fragments weighing 16 ± 0.5 and 8 ± 0.5 mg, respectively. For the inverse Michaelis–Menten kinetics, one PET disc was incubated per enzyme level with a range of enzyme concentrations from 0 to 1000 nM.
RP‐HPLC Analysis
3.9
Reaction samples were diluted in a filtered mixture of water, methanol, and ethylene glycol to a final content of 5%v/v methanol and 20%v/v ethylene glycol to stop the reaction and slow spontaneous product hydrolysis. Samples were passed through an Accucore C18 150*4.6 mm column with a particle size of 2.6 µm was using a Vanquish LC system (Thermo Fisher Scientific Inc., Massachusetts, USA) at 1 mL/min at 40°C in 24%v/v acetonitrile and 5.6 mM formic acid [13]. (Examples of HPLC chromatograms are shown in the Supplementary Materials, Figures S7 and S8).
Optimal Reaction Conditions
3.10
1000 µL samples containing 200 nM LCC_ICCG_ and one amorphous PET disc in 400 mM glycine‐NaOH buffer were incubated at 900 RPM for 2 h in a Thermomixer C (Eppendorf, Hamburg, Germany). The samples contained a range of EG contents in duplicates and were incubated at a range of temperatures. The glycine‐NaOH buffers were adjusted no more than 90 min in advance from 1 M glycine‐NaOH pH 9.00 using 6 M NaOH to the pH which would result in a sample pH of 8.50 at the individual reaction conditions. At the end of the incubation, 20 µL samples were transferred to 180 µL stop solution (6% methanol, 20% EG, 74% ultra‐pure water), and the samples were analyzed by HPLC.
To measure product profiles of continuous reactions, simultaneously running duplicate or triplicate reactions were alternately sampled for immediate HPLC analysis while removing no more than four 20 µL samples per individual tube.
Reactions with Crystalline Substrate
3.11
Following the method described by Thomsen et al. [11], UV‐Star microplates (Greiner, Kremsmünster, Austria) were modified by removing the walls separating two diagonally adjacent cells to connect two diagonally adjacent wells using a rotary tool with a round 4 mm carving burr, leaving 0.5–1 mm around the opening (molten plastic burrs were removed).
300 µL samples containing 400 mM glycine‐NaOH buffer, 200 nM LCC_ICCG_ and EG ranging from 0 to 30% v/v were initiated by adding PET discs annealed for 0–20 min. Sealing tape (Sarstedt AG & Co. KG, Nümbrecht, Germany) was applied to the plate, and it was then incubated in a ThermoMixer C with a microplate insert and ThermoTop (Eppendorf, Hamburg, Germany), set to 62°C and 500 RPM. Prior to measurements, the tape was removed and absorbance at 240 nm was recorded in an Epoch 2 reader (BioTek, Winooski, Vermont, USA). The tape was then re‐applied, and the plate was incubated until the next measurement was taken. The mean absorbance of relevant control samples was subtracted from the respective samples at each time point. The lag phase duration (t lag) was determined by multiple linear regression as a function the catalytic rate (product formation versus time) and X C using SAS JMP15 (SAS institute, Cary, NC, USA) as outlined previously [18].
Conclusion
4
Previous experimental studies on enzymatic PET degradation have shown that when EG is present as a reaction solvent, the product profile of the enzymatic action contains relatively more esterified monomers, MHET and BHET [9, 17]. This outcome has been suggested to be the result of simultaneous hydrolysis and glycolysis during the enzymatic PET treatment. However, the mechanistic understanding of this phenomenon in relation to the enzyme reaction, including the influence on enzyme kinetics, as well as the optimal EG levels and reaction conditions for achieving maximal BHET, have remained elusive. By systematically assessing the influence of EG levels on the enzymatic depolymerization of PET, the present study revealed that the presence of EG in the reaction medium affected the productive and nonproductive adsorption of LCC_ICCG_ to the PET surface. A universal workable limit of EG for maximizing the production of bis(2‐hydroxyethyl) terephthalic acid, i.e. BHET, was identified at a reaction solvent concentration of 27‐29% EG.
EG was found to affect both the enzyme kinetics, the response to substrate crystallinity, and spontaneous product hydrolysis. In combination, these effects allowed for attaining optimal reaction conditions for BHET production via enzymatic PET degradation, which in this work was at a sweet spot of 62.3 ± 0.9°C, pH 8.5, 380 nM LCC_ICCG_ and 27% EG. The work thus demonstrates how EG as a solvent can control the product formation of enzymatic PET degradation and explains the mechanisms behind the effect.
Supporting Information
Additional supporting information can be found online in the Supporting Information Section. Supporting Fig. S1 : Enthalpy relaxation (Ageing) peak areas of incubated PET discs (after 1 h incubation) by incubation temperature at various ethylene glycol (EG) contents. Supporting Fig. S2 : Total product release rates in (TPA equivalents per minute) of DuraPETase by ethylene glycol (EG) content. Supporting Fig. S3 : BHET fraction of soluble products of PET degradation using DuraPETase by reaction time, measured at various ethylene glycol (EG) contents. Supporting Fig. S4 : Plot of the BHET concentration progressions at 10%, 20% and 30% ethylene glycol (EG) corresponding to the data shown in Figure 9. Outliers have been removed, and monomolecular growth functions have been fitted to the data. Supporting Fig. S5 : BHET concentrations by reaction temperature and ethylene glycol fraction, %V/V, after two‐hour PET degradation reactions with 380 nM (left) and 761 nM (right) LCC_ICCG_ at pH 8.5. Supporting Fig. S6 : Predicted optimal conditions for BHET formation by LCC_ICCG_ at two enzyme concentrations after a two‐hour reaction as well as the R^2^ values of the prediction models. Calculated based on the data shown in Figure S5. Supporting Fig. S7 : Example of a chromatogram of PET degradation fragments (LCC_ICCG_ treatment, no EG addition). Peak magnification of the chromatogram shown in the insert. Supporting Fig. S8 : Example of a chromatogram of PET degradation monomers (LCC_ICCG_ treatment, no EG addition). Peak magnification as per the chromatogram shown in the insert. Supporting Table S1: Kinetic parameters V_max_, k cat (min^‐1^), K M (no. of PET discs), and K i of substrate inhibition (no of PET discs); Data were calculated from linear fits of the Hanes‐Woolf transformed rate data of LCC_ICCG_ (5 nM) catalyzed PET degradation with different ethylene glycol (EG) levels v/v as shown in Figure 4 (main paper). Data are shown as average values of triplicate measurements (R^2^ of fits ranged from 0.83‐0.99). Parameters for EG levels >25% could not be determined.
Author Contributions
Tobias S. Radmer: formal analysis (lead), investigation (lead), methodology (equal), validation (equal), writing – original draft (equal), writing – review & editing (equal). Thore B. Thomsen: conceptualization (equal), investigation (equal), methodology (supporting), validation (supporting), writing – review & editing (supporting). Leon T. Krinn: investigation (supporting), methodology (supporting), writing – review & editing (supporting). Anne S. Meyer: conceptualization (lead), project administration (lead), supervision (lead), writing – original draft (equal), writing – review & editing (equal).
Funding
This study was supported by Villum Fonden (Grant 40815).
Conflict of Interest
The authors declare no conflicts of interest.
Supporting information
Supplementary Material
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1T. M. Joseph , S. Azat , Z. Ahmadi , et al., “Polyethylene Terephthalate (PET) Recycling: A Review,” Case Studies in Chemical and Environmental Engineering 9 (2024): 100673.
- 2Plastics Europe , “Plastics ‐ the Facts 2024,” accessed 2024, https://plasticseurope.org/knowledge‐hub/plastics‐the‐fast‐facts‐2024/.
- 3R. Chen , S. Deng , T. Cui , S. Duan , Q. Jia , and L. Zhang , “Progress in Recycling and Reutilization of Waste Polyethylene Terephthalate,” Progress in Rubber, Plastics and Recycling Technology 40 (2023): 77–97.
- 4W. W. Y. Lau , Y. Shiran , R. M. Bailey , et al., “Evaluating Scenarios toward Zero Plastic Pollution,” Science 369 (2020): 1455–1461.32703909 10.1126/science.aba 9475 · doi ↗ · pubmed ↗
- 5G. Arnal , J. Anglade , S. Gavalda , et al., “Assessment of Four Engineered PET Degrading Enzymes Considering Large‐Scale Industrial Applications,” Acs Catalysis 13 (2023): 13156–13166.37881793 10.1021/acscatal.3c 02922 PMC 10594578 · doi ↗ · pubmed ↗
- 6F. Javier , M. Rodríguez , and P. González , “Microwave‐Assisted Esterification Step of Poly(ethylene Terephthalate) (PET) Synthesis through Ethylene Glycol and Terephthalic Acid,” Polymer Bulletin 76 (2019): 2931–2944.
- 7V. Tournier , S. Duquesne , F. Guillamot , et al., “Enzymes’ Power for Plastics Degradation,” Chemical Reviews 123 (2023): 5612–5701.36916764 10.1021/acs.chemrev.2c 00644 · doi ↗ · pubmed ↗
- 8S. Dewi , T. Horbach , T. Neumann , K.‐E. Jaeger , W. Streit , and L. Hartmann , “Evaluation of Pretreatment Strategies and Combinations of Enzymes for Degradation of Polyethylene Terephthalate,” Journal of Polymer Science 63 (2025): 2770–2781.