pubs.acs.

org/acscatalysis Research Article

One-Pot Chemo-bioprocess of PET Depolymerization and Recycling

Enabled by a Biocompatible Catalyst, Betaine
Dong Hyun Kim,# Dong Oh Han,# Kyu In Shim, Jae Kyun Kim, Jeffrey G. Pelton, Mi Hee Ryu,
Jeong Chan Joo, Jeong Woo Han, Hee Taek Kim, and Kyoung Heon Kim*
Cite This: ACS Catal. 2021, 11, 3996−4008 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: Poly(ethylene terephthalate) (PET) has been widely

Downloaded via UNIV BODENKULTUR WIEN on April 27, 2021 at 16:14:00 (UTC).

used in various industries due to its unique physical properties.

However, PET causes major environmental problems globally due to
its low degradability and recycling rate. Since it is nearly impossible to
replace PET with other materials, an efficient approach for PET
recycling is necessary for a circular economy. Herein, for a paradigm
shift toward the approach for resource recovery of PET components,
we developed an integrated process for depolymerizing PET and
converting PET monomers to high-value products in a one-pot
process. The key of our approach is the use of the biocompatible
catalyst betaine in a glycolysis process that enables whole PET
glycolysis slurry as a substrate to be directly applied to further
bioprocesses. Based on the density functional theory (DFT) analysis,
betaine effectively catalyzed PET depolymerization by two strong
hydrogen interactions between betaine, EG, and PET as well as by the synergetic effect between the anion and cation groups of
betaine. Through the glycolysis of PET with betaine and the optimized enzymatic hydrolytic process for the PET glycolysis slurry,
PET was depolymerized to terephthalate (TPA, 31.0 g/L, 62.8%, mol/mol) and ethylene glycol (EG, 11.7 g/L, 63.3%, mol/mol) at
high titers and high yields. This process was further applied to the bioconversion of TPA and EG present in the PET hydrolysate to
protocatechuic acid (PCA) and glycolic acid (GLA), respectively. This one-pot chemo-bioprocess integrating chemical glycolysis,
enzymatic hydrolysis, and bioconversion for PET depolymerization and recycling was suggested to be highly applicable to the
upcycling of waste PET.
KEYWORDS: PET recycling, betaine, depolymerization, glycolysis, PETase, MHETase, bioconversion, density functional theory analysis

■ INTRODUCTION
Plastics have been widely used in almost every industry
Conventional recycling strategies are classified as energy
recovery, physical recycling, and chemical recycling.6 Energy
including clothing, food, medicine, and even automotive owing recovery involves the incineration or pyrolysis of PET,7,8 and
physical recycling involves the grinding and remelting of PET
to their desirable physical properties such as durability,
into secondary plastics, which are both considered down-
flexibility, waterproof, and lightness. However, the use of
cycling processes as lower-value products are produced.
plastics is associated with serious environmental problems
Chemical recycling of PET involves depolymerization to
owing to the considerably low rate of natural degradation and
TPA and EG by cleaving the ester bonds between TPA and
recycling.1 Approximately 6,300 million metric tons of plastic
EG followed by repolymerization of TPA and EG to produce
wastes had been generated in 2015, and among them, 60% had
virgin PET or other polymers.6,8 Since chemical recycling is
been sent to landfills or oceans and only 9% had been
aimed at the reproduction of virgin plastics from monomers,
recycled.1 However, it is nearly impossible to replace plastics
generally, separation and purification steps for obtaining high-
with other materials throughout the industrial fields. Therefore,
purity monomers should be required during the chemical
it is essential to increase the recycling rate of plastic wastes to
alleviate plastic-derived environmental problems. To accom-
plish this, much effort is going into recycling various plastics Received: September 13, 2020
these days.2−4 In particular, poly(ethylene terephthalate) Revised: February 26, 2021
(PET), which is a polyester comprising terephthalate (TPA) Published: March 23, 2021
and ethylene glycol (EG) linked via ester bonds, is one of the
most extensively used plastics globally.5 Many efforts to recycle
PET have been performed through various approaches.

© 2021 American Chemical Society https://doi.org/10.1021/acscatal.0c04014

3996 ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Table 1. Comparison of the Activities of PET-Degrading Enzymes against Various PET-Derived Substrates
titer of TPA (or collective titer of
substrate enzyme source pH temp. (°C) time (h) TPA, MHET, and BHET) reference
whole slurry of PET IsPETaseS121E/D186H/R280A, Ideonella sakaiensis 8.0 40 12 186.7 mM this study
glycolysis IsMHETaseW397A
oligomers of PET IsPETaseS121E/D186H/R280A, I. sakaiensis 8.0 40 3 9.9 mM this study
glycolysis IsMHETaseW397A
oligomers
PET granule IsPETaseS121E/D186H/R280A, I. sakaiensis 8.0 40 24 61.7 μM this study
IsMHETaseW397A
PET film IsPETaseWild I. sakaiensis 7.0 30 18 (0.3 mM) Yoshida et
al.10
PET (drinking bottle IsPETaseY58A I. sakaiensis 9.0 30 20 21 nM (54.0 nM) Liu et al.38
fabrics)
PET film IsPETaseR280A I. sakaiensis 9.0 30 36 15.7 μM (31.9 μM) Joo et al.39
PET film IsPETaseS121E/D186H/R280A I. sakaiensis 9.0 40 72 37.6 μM (120.9 μM) Son et al.25
PET coupon IsPETaseW159H/S238F I. sakaiensis 7.2 30 96 0.8 mM (1.2 mM) Austin et al.40
PET film IsPETaseI179F I. sakaiensis 8.5 30 48 6.4 mM Ma et al.41
PET film hydrolase (TfH) Thermobif ida f usca 7.0 30 18 (<0.1 mM) Yoshida et
al.10
PET film cutinase (LCC) leaf-branch compost 7.0 30 18 (<0.1 mM) Yoshida et
al.10
PET film cutinase (FsC) Fusarium solani (<0.1 mM) Yoshida et
al.10
amorphous PET cutinase (FsC) F. solani 8.0 30 96 0.9 mM Vertomme et
al.42
PET fabric hydrolase (TfuQ132A/T101A) T. f usca 7.5 60 48 19.3 mM Silva et al.43
PET flake from a cutinase (HiC), lipase (CALB) Humicola insolens, 7.0 60 to 37a 24 13.6 mM de Castro et
recycling plant Candida antarctica al.12
amorphous PET cutinase (HiC), lipase (CALB) H. insolens, C. 7.0 60 to 37a 24 60.0 mM de Castro et
antarctica al.12
amorphous PET cutinaseF243I/D238C/S283C/Y127G leaf-branch compost 8.0 72 10 1.0 M Tournier et
(LCC) al.11
amorphous PET film cutinase (LCC) leaf-branch compost 8.0 70 120 215.6 mM Tiso et al.44
UV-treated cutinase (LCC) leaf-branch compost 8.0 70 24 67.9 mM Falkenstein et
amorphous PET al.45
filmb
amorphous PET film cutinase (LCC) leaf-branch compost 8.0 70 24 48.4 mM Falkenstein et
al.45
a
The enzymatic PET degradation was performed at 60 °C for 3 h using HiC and then at 37 °C for 21 h using CALB. bUV irradiation of the PET
films was performed for 14 days.

recycling of PET, resulting in a process with a low economic IsPETase showed higher activity against highly crystalline PET
efficiency. than other PET-degrading enzymes.10 However, even though
An approach for resource recovery of PET components from many efforts have been performed to enhance the activities of
PET waste, where the PET monomeric components, TPA and IsPETase and IsMHETase through protein engineering, titers
EG, can be used as feedstocks to produce value-added of TPA, MHET, and BHET produced from the biological
materials by microbes, can provide a solution to the sustainable depolymerization of crystalline PET by using only enzymes are
use of PET, enabling a circular economy. Recently, TPA and still significantly low (54 nM−6.4 mM) (Table 1). Recently,
EG were shown to be converted to high-value products such as the engineered leaf-branch compost cutinase (LCC) was
gallic acid, pyrogallol, catechol, muconic acid, and vanillic acid reported, which is stable at temperatures higher than 84 °C
via protocatechuic acid (PCA) derived from TPA and glycolic that is the glass transition temperature of PET, and 90% of
acid (GLA) derived from EG through bioconversion using PET was shown to be depolymerized by using only LCC.11
microbial cells in our previous study.9 However, that study just However, in that study, amorphous PET or low-crystalline
showed a proof of concept to produce various chemicals, and PET pretreated by extrusion and micronization to increase the
the conversion yields and titers of final products were very low. amorphous region of PET was used.11 While commercially
To successfully achieve this kind of resource recovery process, available PET has a high crystallinity of about 30%,12,13 LCC
the construction of an efficient PET degradation process still could not depolymerize the crystalline region of PET.11
should be a prerequisite. To construct an environmentally In an approach of the resource recovery of PET components
friendly and green depolymerization process of PET, many from waste PET, the depolymerization problem of PET can be
efforts to use various enzymes and improve their activities for analogously perceived as that of lignocellulose for the
PET depolymerization have been exerted. In particular, PET enzymatic production of glucose from cellulose in the
hydrolase (PETase) and mono(2-hydroxyethyl) terephthalate lignocellulose biorefinery. Although various enzymes related
(MHET) hydrolase (MHETase) were discovered in Ideonella to lignocellulose depolymerization including cellulases, xyla-
sakaiensis, a bacterium that efficiently degrades and assimilates nases, and β-glucosidases have been extensively studied and
PET in nature,10 and these two enzymes have received much developed, it is still impossible to efficiently depolymerize
attention as ideal enzymes for PET degradation because lignocellulose only by using enzymes. To efficiently degrade
3997 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 1. Overall scheme of one-pot chemical glycolysis and enzymatic hydrolysis of PET. Stage 1: chemical glycolysis of PET using betaine as a
novel catalyst. Stage 2: enzymatic hydrolysis using the whole slurry derived from PET glycolysis as the substrate without any separation and
purification steps involving IsPETaseMut and IsMHETaseMut. Stage 3: whole-cell bioconversion using the enzymatic hydrolysate of PET glycolysis
slurry as the substrate without intermediate steps for separation and purification steps to a high-value product, PCA.

and utilize lignocellulosic components, a chemical pretreat- to protect cells, proteins, and enzymes against environmental
ment process that can loosen and break lignocellulose stresses such as osmotic stress, high temperature, and water
structures for higher substrate accessibility enzymes is required scarcity,22 it is a biocompatible and environmentally friendly
prior to an enzymatic depolymerization process.14 In the same catalyst. Therefore, it may not inhibit PET-degrading hydro-
context, a chemical pretreatment process may be necessary as a lases in the preceding enzymatic depolymerization step, not
prerequisite for the enzymatic depolymerization of PET to negatively affect cell viability in the bioconversion step, and not
establish a more practical PET depolymerization and recycling adversely affect the environment. Also, betaine is considered a
process. more economically feasible catalyst than ILs and DESs for
In this study, by mimicking the lignocellulose biorefinery, we pretreating bulk chemicals such as waste PET. It is because the
designed and demonstrated a novel one-pot chemo-bioprocess preparation of ILs and DESs is known to be complicated and
for PET recycling that integrates not only chemical and costly.21 Therefore, betaine may enable the whole PET
enzymatic steps to overcome the low depolymerization glycolysis slurry as a substrate to be directly applied to further
efficiency of PET but also a bioconversion step to produce bioconversion processes without any separation and purifica-
value-added products directly from TPA and EG in enzymatic tion steps. We suggested a possible mechanism underlying the
hydrolysate of PET glycolysis (Figure 1). The key of the PET degradation of PET through glycolysis catalyzed by betaine by
recycling chemo-bioprocess is the use of betaine in the PET conducting a density functional theory (DFT) analysis.
glycolysis step that could be considered a pretreatment step as Furthermore, to verify whether our novel PET glycolysis step
in the lignocellulose biorefinement process. Although metal can be efficiently connected to further bioprocesses such as
ions, which catalyze PET degradation by forming the enzymatic depolymerization and bioconversion steps, we
coordination bond with the carbonyl oxygen of PET and demonstrated a one-pot chemo-bioprocess for PET recycling.
activating the carbonyl carbon of PET, are commonly used as For the one-pot depolymerization of PET, the PET glycolysis
catalysts in PET glycolysis,6 they might not only inhibit slurry without any separation steps was subjected to enzymatic
enzymatic activity in the enzymatic degradation step and cell hydrolysis using engineered mutant PETase and MHETase to
viability in the bioconversion step but also adversely affect the produce TPA and EG. Lastly, the whole PET hydrolysate
environment. Ionic liquids (ILs),15,16 organocatalysts,17 and obtained from the one-pot chemical glycolysis and enzymatic
deep eutectic solvents (DESs)18−20 have also been studied as hydrolysis was biologically and directly converted to higher-
catalysts for glycolysis, in which these catalysts have a value products using whole-cell catalysts. To our knowledge,
synergistic effect of their cations capable of interacting with this study is the first experimental demonstration to envision
the carbonyl oxygen of the ester bond in PET like the catalytic the integrated chemical glycolysis and enzymatic hydrolysis for
one-pot recycling of waste PET.

■
function of metal ions and their anions capable of interacting
with the hydrogen of the hydroxyl group in EG. However,
these catalysts have several disadvantages such as high cost, MATERIALS AND METHODS
complicated preparation methods, and poor compatibility with Materials. PET granules, BHET (bis(2-hydroxyethyl)
enzymes and cells.17,20 In this study, we suggest that using terephthalate), TPA, and PCA were purchased from Sigma-
betaine as a catalyst in the glycolysis of PET could overcome Aldrich (St. Louis, MI). EG and betaine were purchased from
these problems. Since betaine is a zwitterion having both Junsei (Tokyo, Japan) and Alfa Aesar (Ward Hill, MA),
positively and negatively charged functional groups,21 it could respectively. The PET granules used in this study contained
show the synergetic effect in PET depolymerization as a similar 30% glass particles as a reinforcing agent. The PET granules
effect of ILs or DESs. Furthermore, since betaine is were milled using a blender (FM-700SS, Hanil Electric, Seoul,
biosynthesized by the oxidation of choline in living organisms Korea) and sieved to segregate into particle sizes of 450−800
3998 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

μm. Since MHET is not commercially available, it was using an Amicon ultracentrifugal filter unit (MW cutoff value
prepared by hydrolyzing BHET with IsPETaseMut (mutated of 10,000; Millipore, Billerica, MA).
IsPETase described later in the section ″Preparation of IsMHETaseMut was expressed primarily in inclusion bodies
Recombinant PETase and MHETase″) following a method using the abovementioned expression system. Therefore, to
described previously with certain modifications.10 Briefly, 118 obtain purified IsMHETaseMut, the inclusion body was
mM BHET was incubated with 1,500 nM PETase in 200 mM unfolded using 8 M urea and refolded by dialysis. All the
sodium phosphate (pH 8.0) at 40 °C for 12 h. buffers, such as lysis, binding, and elution buffers, used for
Chemical Glycolysis of PET. For each glycolysis reaction, purifying IsMHETaseMut contained 8 M urea. The purification
3.5−11.5 g of PET granules was depolymerized in EG using a steps were the same as those for IsPETaseMut, except for the
certain quantity of betaine as a catalyst in a 100 mL reaction addition of 8 M urea. The fraction containing IsMHETaseMut
vessel in a microwave digester (Milestone, Shelton, CT) was dialyzed against a buffer containing 50 mM Tris−HCl at
equipped with a thermocouple and a magnetic stirrer at 190 °C pH 8.0. The purified recombinant enzymes were analyzed
for 30−120 min. The major product of glycolysis, BHET, was using 8% sodium dodecyl sulfate polyacrylamide gel electro-
identified through high-performance liquid chromatography phoresis (SDS-PAGE). The protein was quantified using a
(HPLC), GC/MS, and NMR analyses. In addition, BHET, bicinchoninic acid protein assay kit (Thermo Fisher Scientific,
MHET, and TPA were quantified using HPLC. Waltham, MA) using bovine serum albumin (Sigma) as the
DFT Analysis. All density DFT calculations were protein standard.
performed using the Jaguar software (Schrödinger, New PETase and MHETase Activity Assay. To investigate the
York, NY) equipped with an optimization tool.23,24 The inhibitory effects exerted by various catalysts such as betaine,
B3LYP hybrid DFT method and the basis set of 6-31G(d) Ca2+, Co2+, Cu2+, Fe2+, Mg2+, Mn2+, Ni2+, Pb2+, Zn2+, and Li+
were used. Conformational analysis was performed using the on the enzymatic activity of IsPETaseMut and IsMHETaseMut,
Conformational Search tool from the MacroModel package each enzyme quantity corresponding to 2.3 U was incubated in
with the OPLS-2005 force field to explore the possible initial the presence of 1 mM of each catalyst along with 2.5 mg of
configurations for the quantum chemical calculations. Energy BHET and MHET, respectively, in 500 μL of a buffer
convergence was set to energy change <5 × 10−5 hartree and containing 50 mM sodium phosphate (pH 8) at 40 °C for 15
root-mean-square density matrix change <5 × 10−6 hartree. min. The enzymatic reactions were terminated by adding
Moreover, the structures were optimized freely using methanol, and the supernatant obtained by centrifugation
redundant internal coordinates with the default Schlegel (25,188g, 10 min) was analyzed using HPLC.
To determine the optimal loading quantity of IsPETaseMut
guess. Detailed noncovalent interactions were confirmed
and IsMHETaseMut when BHET and MHET were used as the
using the Surface tool with a noncovalent grid density of
substrates, respectively, the initial reaction velocities at
20.0 pts/Å.
different enzyme loadings were compared. Different loadings
Preparation of Recombinant PETase and MHETase.
of IsPETaseMut and IsMHETaseMut, ranging from 50 to 400
The genes encoding IsPETaseS121E/D186H/R280A (IsPETaseMut)
nmol/g BHET (equivalently from 1.4 to 11.4 mg/g BHET)
and IsMHETaseW397A (IsMHETaseMut), which are protein-
and from 300 to 1,200 nmol/g MHET (equivalently from 18.8
engineered mutants with higher enzymatic activity and higher
to 75.4 mg/g MHET), respectively, were incubated in 50 mM
thermal stability than their wild-type counterparts IsPETase25 sodium phosphate buffer (pH 8) at 40 °C for 10 min along
and IsMHETase,26 respectively, were synthesized commercially with 10 g/L BHET and 5 g/L MHET, respectively.
by codon optimization for expression in Escherichia coli. The For the unit definition of IsPETaseMut and IsMHETaseMut,
genes for IsPETaseMut and IsMHETaseMut were individually we established that 1 unit (U) would produce 100 μmol of
cloned into the NdeI (5′ end) and XhoI (3′ end) sites of a MHET and TPA from BHET and MHET, respectively, per
pET28a vector (Novagen, Darmstadt, Germany). A poly-6- min at pH 8 and 40 °C. Next, the enzymatic reaction profiles
histidine reside was added to the N terminus of the protein to were monitored to assess if the different enzyme loadings of
aid the purification of the protein using affinity chromatog- IsPETaseMut and IsMHETaseMut, corresponding to 2.3, 4.5, and
raphy (GE Healthcare, Piscataway, NJ). Each plasmid that 9.0 U, could efficiently hydrolyze the slurry of PET glycolysis,
harbored a mutant gene was transformed into E. coli which contained 10 g/L of BHET, upon incubation with 50
BL21(DE3). Recombinant E. coli BL21(DE3) was cultured mM sodium phosphate buffer (pH 8) at 40 °C for 6 h. The
in an LB medium containing 40 μg/mL of kanamycin at 37 °C PET glycolysis slurry was solubilized in the enzymatic reaction
and 200 rpm until OD600 reached 0.5. Enzyme overexpression mixture by adding 5% (v/v) dimethyl sulfoxide.
was induced using 0.5 mM isopropyl-β-D-thiogalactopyrano- To investigate the effect of the buffering strength on the
side (IPTG) administered at 16 °C for 18 h. The cells were enzymatic hydrolysis of the PET glycolysis slurry, the enzyme
harvested by centrifugation at 4,000g for 20 min at 4 °C. For loadings of IsPETaseMut and IsMHETaseMut corresponding to
the purification of IsPETaseMut, pET28-IsPETaseMut-harboring 2.3 U were incubated at various concentrations of Bicine-
cells were resuspended in a lysis buffer (20 mM Tris−HCl, pH NaOH buffer (pH 8) ranging from 50 to 1,000 mM with the
7.4), disrupted by ultrasonication, and centrifuged at 8,000g for PET glycolysis slurry containing 10 g/L of BHET at 40 °C for
20 min at 4 °C. The supernatant was introduced into a 6 h. While monitoring the enzymatic hydrolysis process, the
HisTrap column (GE Healthcare, Piscataway, NJ) that was pH change profiles in the enzymatic reaction mixtures were
previously equilibrated with a binding buffer (20 mM sodium monitored simultaneously using a pH test paper (Johnson Test
phosphate, 500 mM NaCl, and 20 mM imidazole at pH 7.4). Papers Universal Paper, UK).
The loaded proteins were released from the HisTrap column To determine the maximal loading of the glycolysis slurry in
by eluting with an elution buffer (20 mM sodium phosphate, the enzymatic hydrolysis step, the Bicine-NaOH buffer was
500 mM NaCl, and 300 mM imidazole at pH 7.4). The buffer used instead of the sodium phosphate buffer because buffering
was exchanged, and the eluted IsPETaseMut was concentrated by the former offers greater stability than that by the sodium
3999 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

phosphate buffer at pH 8. Different concentrations of the High-Performance Liquid Chromatogrphy (HPLC). To

glycolysis slurry ranging from 10 to 50 g/L of BHET were analyze the products of PET glycolysis, BHET, MHET, and
incubated with 1 M Bicine-NaOH buffer (pH 8) with the TPA, and the whole-cell conversion product, PCA, these
enzyme loadings of IsPETaseMut and IsMHETaseMut corre- products were diluted with methanol and then analyzed by
sponding to 2.3 U at 40 °C for 12 h. The pH change profiles in HPLC (Agilent 1100, Agilent Technologies, Waldronn,
the enzymatic reaction mixtures were also monitored. Germany) equipped with a UV−vis detector (G1314A) and
Bacterial Strains and Plasmids. E. coli PCA-1 used in a C18 column (OpitmaPak C18-51001546, 5 μm, 150 × 4.6
this study for converting TPA to PCA was the same as that mm). The mobile phase A comprised distilled water containing
used in our previous study.9 Specifically, the genes tphAabc and 0.1% trifluoroacetic acid (Sigma-Aldrich), and the mobile
tphB, which encode the products that convert TPA to PCA phase B comprised methanol (Honeywell, Charlotte, NC).
using DCD, originating from Comamonas sp. E6, were The flow rate was fixed at 1 mL/min. The mobile phase B was
synthesized by IDT and were digested and ligated into allowed to flow for 2 min at 5%, the concentration was
plasmids pKM212 (constructing pKM212TphAabc) and gradually changed to 57% and flown for 18 min, and the
pKE112 (constructing pKE112TphB), respectively. These detection wavelength was fixed at 254 nm. To analyze GLA,
constructed plasmids were transformed into E. coli XL1-Blue EG, and whole-cell conversion byproducts such as ethanol,
for constructing the strain PCA-1. Gluconobacter oxydans acetic acid, and glycerol, an Aminex HPX-87H column (Bio-
KCCM 40109 (Korean Culture Center of Microorganisms, Rad, Hercules, CA) was used. For the second HPLC analysis,
Seoul, Korea) was used as a host for the whole-cell the column temperature was maintained at 65 °C, and 0.005 M
bioconversion of EG to GLA.9 H2SO4 was flown through the column at a rate of 0.5 mL/min.
Whole-Cell Bioconversion of TPA to PCA. The whole- Gas Chromatography/Mass Spectrometry (GC/MS).
cell bioconversion of TPA to PCA using E. coli PCA-1 was To identify the major products of glycolysis and whole-cell
performed at 30 °C and 250 rpm in 100 mL baffled flasks conversion such as BHET and muconic acid, respectively, GC/
following a slightly modified version of a method described MS analysis was performed according to a method described
previously.27 The products were centrifuged at 25,188g for 10
earlier.9 For the seed culture, E. coli PCA-1 was cultured in 5
min at 4 °C, and 20 μL of supernatants was vacuum-dried. For
mL of the LB medium containing 50 and 40 μg/mL of
the derivatization of the reaction products, the dried samples
ampicillin and kanamycin, respectively, at 37 °C and 200 rpm.
were incubated with 10 μL of 40 mg/mL methoxyamine
The seed culture was then transferred into 1 L of the LB
hydrochloride in pyridine (Sigma-Aldrich) at 30 °C for 90 min
medium in a 2.8 L flask and incubated at 37 °C and 200 rpm and then incubated with 50 μL of N-methyl-N-trimethylsilyl-
until OD600 reached 0.5. Overexpression of the genes was trifluoroacetamide (Sigma) at 37 °C for 30 min. The
induced by administering 0.5 mM of IPTG at 16 °C for 18 h. derivatized samples were analyzed using an Agilent 7890A
The cells were harvested by centrifugation at 4,000g for 20 min GC/5975C MSD (Agilent Technologies, Wilmington, DE)
at 4 °C and washed with 100 mM sodium phosphate buffer equipped with a DB5-MS column (0.25 mm × 30 m, 0.25 μm
(pH 6.5). The harvested cells were inoculated in 8 mL of the film thickness; Agilent Technologies). Each aliquot consisting
MR medium containing 20 g/L of glycerol in 100 mL baffled of 1 μL of the derivatized samples was injected into the GC in
flasks at 30 °C and 200 rpm. The MR medium contains the the splitless mode. The temperature of the GC oven was
following components per liter: 6.7 g of KH2PO4, 4 g of programmed as follows: 100 °C for 3.5 min, increased to 160
(NH4)2HPO4, 0.8 g of MgSO4·7H2O, 0.8 of g citric acid, 10 of °C at 15 °C/min and held for 20 min, increased to 200 °C at
mg thiamine-HCl, and 5 mL of concentrated trace metal 20 °C/min and held for 20 min, and increased to 280 °C at 20
solution. The concentrated trace metal solution contained the °C/min and held for 5 min. Ionization was achieved by
following components per liter: 5.46 g of FeSO4, 1.51 g of electron impact at 70 eV, and the ion source was maintained at
CaCl2, 1.23 g of ZnSO4, 0.34 g of MnSO4, 0.64 g of CuSO4, 230 °C. The mass spectra were recorded in the range of 50 to
0.09 g of (NH4)6Mo7O24, and 0.01 g of Na2B4O7. The 500 m/z.
enzymatic hydrolysate of PET glycolysis containing 4.5 g/L of Nuclear Magnetic Resonance (NMR). The samples were
TPA as a substrate in whole-cell bioconversion was added to dissolved in DMSO for NMR analysis. All the experiments
the medium. Additionally, 5.7 g/L of reagent-grade TPA was were performed and results were recorded on a Bruker Avance
added to the same medium as a positive control for comparing II spectrometer operated at 900 MHz and 25 °C (Bruker,
the potential of whole-cell bioconversion using the enzymatic Billerica, MA). The instrument was equipped with a TCI
hydrolysate of PET glycolysis. cryoprobe and was controlled using Topspin software (version
Whole-Cell Bioconversion of EG to GLA. The whole-cell 3.2, Burker).
bioconversion of EG to GLA using G. oxydans KCCM 40109 Fourier-Transform Infrared (FT-IR) Spectroscopy. FT-
was performed according to a method described previously IR spectroscopy analysis was performed to characterize the
with slight modifications.9 For the seed culture, G. oxydans chemical structures of BHET and the depolymerized products.
KCCM 40109 was cultured in 5 mL of a medium consisting of The FT-IR spectra were recorded using a Bruker ALPHA-P
20 g/L sorbitol, 20 g/L yeast extract, 5 g/L (NH4)2SO4, 2 g/L and LPHA-T spectrometer equipped with a deuterated
KH2PO4, and 5 g/L MgSO4·7H2O at 30 °C and 200 rpm. The triglycine sulfate detector. The samples were analyzed in
seed culture was then transformed into 1 L of the same attenuated total reflectance mode with 128 scans within a
medium in a 2.8 L flask and incubated at 30 °C and 200 rpm. frequency range of 4,000−400 cm−1.
The cells were harvested according to the method described
earlier. The whole-cell bioconversion of EG to GLA was
performed at 30 °C and 200 rpm in 10 mL of the medium with
■ RESULTS AND DISCUSSION
Glycolysis of PET Using Betaine. Based on the similarity
30.6 g/L of EG. The experiments of whole-cell conversion with the saccharification of lignocellulose, wherein chemical
were performed in duplicate. pretreatment is required before the enzymatic hydrolysis of
4000 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Table 2. Glycolysis of PET at Various Conditions

titer of titer of
input EG (g) (or PET/ catalyst (mg) (or reaction collective yield of of BHET BHET MHET
catalyst PET (g) temp. (°C) EG, mol/mol) catalyst/PET, wt %) time (min) and MHET (%)a (g/L) (g/L)
betaine 11.5 190 11.14 (1:3) 57.5 (0.5) 30 34.6 256.8 24.9
10.0 190 12.92 (1:4) noneb 30 16.8 117.1 N.D.c
10.0 190 12.92 (1:4) 10.0 (0.1) 30 26.6 185.3 N.D.
10.0 190 12.92 (1:4) 50.0 (0.5) 30 52.3 330.6 27.6
10.0 190 12.92 (1:4) 100.0 (1) 30 42.4 286.9 6.8
10.0 190 12.92 (1:4) 50.0 (0.5) 60 58.7 368.6 32.8
10.0 190 12.92 (1:4) 50.0 (0.5) 90 55.1 341.2 35.1
10.0 190 12.92 (1:4) 50.0 (0.5) 120 53.8 313.8 50.4
9.0 190 14.54 (1:5) 45.0 (0.5) 30 52.7 307.6 21.6
8.0 190 15.50 (1:6) 40.0 (0.5) 30 51.6 268.3 19.1
6.0 190 19.37 (1:10) 30.0 (0.5) 30 49.8 189.8 6.7
3.5 190 22.60 (1:20) 17.5 (0.5) 30 37.2 81.3 2.9
TBD 10.0 190 12.92 (1:4) 50.0 (0.5)d 60 55.1 328.4 45.9
a
Yield of BHET and MHET = (concentration (g/L) of BHET and MHET) × (total reaction volume (L))/(theoretical maximal content (g) of
BHET and MHET from input PET). bNo catalyst was used. cNot detected. dTBD was used as a catalyst for glycolysis to compare the catalyst
performance of betaine and organic catalysts.

Figure 2. Identification of products derived from PET glycolysis using betaine as a catalyst. (a) Experimental scheme of PET glycolysis using
betaine. (b) FT-IR spectra of PET degradation catalyzed by betaine. (c) Comparison of FT-IR spectra between PET granule and oligomers in
glycolysis product. (d) Total ion chromatograms of PET glycolysis product and BHET standard obtained from GC/MS analysis. Mass spectra of
(e) PET glycolysis product and (f) BHET standard obtained from GC/MS analysis. (g) 1D 1H NMR and (h) 1D 13C NMR spectra of the soluble
fraction obtained from PET glycolysis product when BHET standard was used as the reference.

cellulose28 can be conducted, we conducted pretreatment of 1:4 (358.2 g/L) was higher than that at 1:5 (329.2 g/L)
before conducting the enzymatic depolymerization of PET. (Table 2). Analogous to the fact that high-substrate loading is
Herein, betaine and EG were used as a catalyst and solvent, advantageous for obtaining high-titer sugar from the
respectively, for glycolysis of PET under various conditions saccharification of lignocellulose, which can consequently
(Table 2). First, to determine an optimal PET loading ratio reduce the operating and capital costs,29,30 the higher PET
relative to the solvent EG, multiple ratios of PET to EG loading ratio, 1:4, was selected as the optimal ratio in this
ranging from 1:20 to 1:3 (mol/mol) were tested at 190 °C for study.
30 min using 0.5% (w/w) betaine per PET sample. As the Next, to determine the optimal catalyst loading ratio relative
PET-to-EG ratio decreased from 1:5 to 1:20, the total yield of to PET, PET glycolysis was performed by varying betaine
BHET and MHET from glycolysis was reduced from 52.7 to loadings per PET sample from 0 to 1.0 wt % under fixed
37.2% (Table 2). Although the collective yield of BHET and conditions of a PET-to-EG loading ratio of 1:4 (mol/mol),
MHET at the PET loading ratio of 1:4 (52.3%) was marginally 190 °C, and 30-min reaction time (Table 2). As the betaine
lower than that at the PET loading ratio of 1:5 (52.7%), the loading increased to 0.5 wt %, the collective yield of BHET and
collective titer of BHET and MHET at the PET loading ratio MHET increased from 16.8 to 52.3%. However, a further
4001 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 3. Relative gas-phase energies (Erel,a−f) of different interaction structures formed between PET polymer, betaine, and EG determined using
DFT analysis. Red: O atoms; blue: N atoms; white: H atoms, and gray: C atoms. The three major bonds with the greatest bond strength were
highlighted in yellow, blue, and red in the order of bond strength. (a) Erel of structure A, 41.40 kcal/mol. (b) Erel of structure B, 12.96 kcal/mol. (c)
Erel of structure C, 26.83 kcal/mol. (d) Erel of structure D, 30.18 kcal/mol. (e) Erel of structure E, 42.59 kcal/mol. (f) Erel of structure F, 0.00 kcal/
mol. Among the possible interaction structures in this study, structure F showed the lowest value of Erel. Gas-phase energies of all structures were
listed in Table S1.

increase in betaine loading to 1 wt % did not result in a glycolysis slurry was detected at the same retention time as that
corresponding increase in PET depolymerization. Glycolysis of the BHET standard (Figure 2d). Additionally, the mass
conducted with a betaine loading ratio at 0.5 wt % with a PET/ spectrum of BHET acquired by the GC/MS analysis of the
EG molar ratio of 1:4 at 190 °C for 60 min achieved the PET glycolysis slurry was confirmed to be identical to the
highest yield (58.7%) and collective titer (401.4 g/L) of BHET spectrum obtained for its standard (Figure 2e,f). To determine
and MHET among those achieved for a reaction duration with greater precision whether BHET is the major product in
ranging from 30 to 120 min (Table 2). Betaine at 0.5 wt % was the PET glycolysis slurry, we separated the soluble fraction
selected as the optimal novel catalyst concentration for PET from the PET glycolysis slurry by adding water to dissolve
glycolysis. Various types of organic catalysts have been used for BHET. Based on the NMR analysis of this sample, the 1D 1H
chemical glycolysis of PET due to the synergetic activation by and 1D 13C NMR spectra of the soluble fraction obtained from
the organic catalyst.4,31 Therefore, to compare the catalyst the PET glycolysis slurry were identical to those of the BHET
ability of betaine and organic catalysts, 1,5,7- standard (Figure 2g,h).
triazabicyclo[4.4.0]dec-5-ene (TBD), which is one of the We also conducted FT-IR analysis of the PET glycolysis
representative organic catalysts used for chemical glycolysis of oligomers in the insoluble fraction obtained from the PET
PET, was tested in this study. As a result, the enhancement glycolysis slurry after washing with water (Figure 2c). A new
effect of betaine as a catalyst on PET depolymerization was broad peak was observed at approximately 3,387 cm−1, which
similar to that of TBD at the optimal conditions. Therefore, represents the hydroxyl group peak in the FT-IR spectrum of
betaine was suggested to be an outperforming catalyst for the the PET glycolysis oligomers compared to the FT-IR spectrum
chemical glycolysis of PET in addition to its other advantages. of the PET granule. However, the hydroxyl group peak was
Identification of Products Derived from PET Glycol- hardly observed in the PET granule spectrum as the sample
ysis. Since the effectiveness of betaine as a catalyst in PET had considerably fewer free terminals of PET than those of
glycolysis has not been studied earlier, it is necessary to analyze PET oligomers. Therefore, the FT-IR analysis verified that
whether BHET is the major product in the PET glycolysis PET oligomers were generated as a result of glycolysis.
slurry obtained after glycolysis with betaine. To accomplish DFT Analysis of PET Glycolysis Using Betaine. To
this, FT-IR) GC/MS, and NMR analyses were performed investigate the mechanism underlying the degradation of PET
(Figure 2). In the FT-IR analysis, a hydroxyl group peak was by glycolysis with betaine as a novel green catalyst, we
observed in the PET glycolysis slurry after glycolysis at 3,436 performed DFT analysis in which the gas-phase energies of
cm−1, which was identical to that of the BHET standard and possible interaction structures between the PET polymer, EG,
represented the newly formed hydroxyl bond at the BHET and betaine (Figure S2) and the strengths of interaction bonds
molecule terminus (Figure 2b). (Table S1) were calculated. Furthermore, six types of possible
Furthermore, we compared the total ion chromatograms of interaction structures were suggested for the coordination
GC/MS analysis of the BHET standard to those of the PET between the molecules primarily via hydrogen bonds (Figure
glycolysis slurry. The major peak obtained for the PET 3). These hydrogen bonds could be formed between the
4002 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 4. The proposed mechanism of PET depolymerization through glycolysis catalyzed by betaine. (a) Noncovalent bond interactions in the
optimized interaction structure F. (b) Diagram of the putative mechanism of PET depolymerization catalyzed by betaine.

Figure 5. Optimization of enzymatic hydrolysis using whole slurry of PET glycolysis as a substrate. (a) Inhibitory effects of various catalysts (at 1
mM), metal ions, TBD, and betaine on the hydrolytic activities of IsPETaseMut and IsMHETaseMut with BHET and MHET as substrates,
respectively. (b) Effect of enzyme concentration on the initial velocities of IsPETaseMut and IsMHETaseMut with BHET and MHET as the
substrates, respectively. (c) Effect of different enzyme loadings of IsPETaseMut and IsMHETaseMut on the enzymatic hydrolysis of the PET
glycolysis slurry containing 10 g/L of BHET in 50 mM sodium phosphate buffer (pH 8.0). (d) Effect of different concentrations of Bicine-NaOH
buffer on the enzymatic hydrolysis of the PET glycolysis slurry containing 10 g/L of BHET. (e) HPLC analyses of the products of the enzymatic
reaction performed in 1 M Bicine-NaOH using 2.3 U of enzyme with 10 g/L of BHET. Effects of (f) different enzyme loadings and (g) substrate
loadings on the enzymatic hydrolysis of the PET glycolysis slurry containing 10 to 50 g/L of BHET in 1 M Bicine-NaOH buffer (pH 8.0). (h)
Theoretical maximal yields of TPA on initial input PET and BHET contents in the whole PET glycolysis slurry. (i) Comparison of the efficiency of
enzymatic hydrolysis of different substrates, such as PET granules, PET glycolysis oligomers, whole PET glycolysis slurry, and reagent-grade BHET.

hydrogen of the hydroxyl residue in EG and the carbonyl EG and the hydrogen in betaine and PET polymer (Figure 3,
oxygen of betaine, between the hydrogen of the hydroxyl Figures S3−S7, and Tables S2−S7). Among the different
residue in EG and the carbonyl oxygen of ester bonds in PET interaction structures, structure F had the lowest gas-phase
polymer, and between the oxygen of the hydroxyl residue in energy (Table S1). There were two major strong noncovalent
4003 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

bond interactions between BO1 and EH1 and between PO3 optimized by analyzing their initial reaction velocities with
and EH6 in structure F (Figure 3f and Table S7). These their ideal substrates, BHET and MHET, respectively (Figure
noncovalent interactions played an important role in 5b). In the case of IsPETaseMut, as the enzyme loading against
determining the strength of the binding between the PET BHET increased, the initial velocity increased and the velocity
polymer, EG, and betaine, and other noncovalent bond reached saturation at approximately 900 μM/min with 259
interactions, including those formed between the hydrogen nmol of IsPETaseMut/g BHET (equivalently 7.4 mg of
in betaine and the oxygen and carbon in the PET polymer, IsPETaseMut/g BHET) (Figure 5b). Therefore, the maximal
facilitated the efficient depolymerization of PET (Figure 4 and velocity with varying loadings of IsPETaseMut was 900 μM/min
Table S7). (equivalent to 9 U/L), which was determined to be the
Based on structure F, which had the lowest gas-phase optimal enzyme loading.
energy, we suggested a possible mechanism of PET Unlike that in IsPETaseMut, as the enzyme loading of
degradation by glycolysis using betaine (Figure 4b). First, IsMHETaseMut with MHET increased, its initial velocity
owing to the two major high-strength hydrogen bond increased linearly in the enzyme loading range tested in this
interactions between BO1 and EH1 and between PO3 and study (Figure 5b). Since BHET and the oligomers in the PET
EH6, betaine and EG could be positioned close to the PET glycolysis slurry are first hydrolyzed by IsPETaseMut to MHET,
polymer. Additionally, other noncovalent interactions between which is subsequently hydrolyzed by IsMHETaseMut to TPA
betaine and the PET polymer, including those between BH9 and EG (Figure 1), the excessive hydrolysis by IsMHETaseMut
and PO3, between BO2 and PH6, between BH11 and PO4, compared to that by IsPETaseMut is considered wasteful.
and between BH3 and PC5, could occur. Among these, owing Therefore, to match the maximal velocity of 900 μM/min with
to the interaction between the carbonyl oxygen of ester bonds IsPETaseMut, a loading of IsMHETaseMut exhibiting the same
in PET polymer and the hydrogen of the methyl group in initial velocity (equivalent to 9 U/L) can be used (Figure 5b),
betaine (i.e., between PO3 and BH9) as well as the interaction which can be optimal for the consecutive enzymatic hydrolysis
with the quaternary ammonium cation in betaine, betaine of the PET glycolysis slurry.
interacts with the carbonyl oxygen of PET, which activates the When the optimal enzyme loading of the two enzymes (9
carbonyl carbon in the PET polymer. Moreover, the strong U/L for both) was added to the PET glycolysis slurry
hydrogen interaction between BO1 and EH1, as well as the containing 10 g/L of BHET in 50 mM sodium phosphate
interaction between a negatively charged functional group in buffer (pH 8.0), the final titer of TPA after 6 h was 7.7 g/L
betaine and the hydrogen of the hydroxyl ion in EG, increases (Figure 5c). As the enzyme loadings were reduced to 4.5 and
the O−H bond length in the hydroxyl group in EG, and the 2.3 U/L, the final TPA titers decreased to 6.4 and 4.6 g/L,
hydrogen can be removed with greater ease, which respectively.
consequently increases the nucleophilicity of the oxygen in Effect of pH and Optimal Buffer on Enzymatic
the hydroxyl group in EG. Next, the activated hydroxyl oxygen Hydrolysis of BHET. The final products of the enzymatic
in EG attacks the activated carbonyl carbon in the ester bonds hydrolysis of the PET glycolysis slurry are TPA and EG. Since
in the PET polymer, and a tetrahedral intermediate is formed. TPA is a dicarboxylic acid, as the enzymatic reaction
A single side chain of PET is released upon cleavage of the acyl progresses, the pH of the reaction mixture may decrease.
oxygen bond in PET. Lastly, one side chain in PET is replaced When 4 g/L of TPA was dissolved in 50 mM sodium
with EG. These reactions may occur continuously, with the phosphate (initial pH 8.0), the pH of the TPA and buffer
formation of BHET as the final product of PET glycolysis. mixture decreased sharply to 5.5 (Figure S8). This reduction in
Therefore, betaine could be used as an effective and pH could have significantly inhibited IsPETaseMut activity (as
biocompatible catalyst in glycolysis that can be used further indicated by the reduction by 19.5% at pH 5.5) (Figure S9).
in the enzymatic degradation and bioconversion steps without When the pH change was monitored during the enzymatic
any separation steps for developing the integrated PET hydrolysis of BHET in the PET glycolysis slurry at 9.0 U/L for
recycling process. both enzymes, the pH of the enzymatic reaction mixture was
Inhibitory Effects of Betaine on PETase and reduced to less than 6 after 30 min, possibly owing to the
MHETase. Prior to the optimization of enzymatic hydrolysis generation of 5.5 g/L of TPA (Figure S10). Therefore, to
of PET glycolysis slurry using IsPETaseMut and IsMHETaseMut, improve TPA production, it was necessary to prevent pH
the inhibitory effects of betaine on these enzymes, in reduction as a consequence of TPA production.
comparison with those of metal ions and the organic catalyst To prevent pH reduction caused by TPA production, we
TBD, were investigated at the catalyst concentration of 1 mM selected an optimal buffer. Several buffers such as sodium
(Figure 5a). It was because betaine used for glycolysis was phosphate, Bicine-NaOH, and Tris−HCl buffers were
carried over in enzymatic hydrolysis of the whole slurry of PET evaluated with respect to their buffering capacities and the
glycolysis for the one-pot process. Surprisingly, unlike metal effect on enzymatic hydrolysis at pH 8.0. From pH 7.0 to 8.0,
ions commonly used in glycolysis that inhibit the two enzymes, the sodium phosphate buffer exhibited the highest relative
betaine did not exert any inhibitory effect on either enzyme activity of IsPETaseMut among the tested buffers (Figure S9).
(Figure 5a). In addition, 1 mM TBD also strongly inhibited However, since the general buffering range of sodium
both IsPETaseMut and IsMHETaseMut (Figure 5a). Therefore, phosphate buffer is between pH 5.8 and 7.4, the buffering
the whole slurry of PET glycolysis was subjected to enzymatic capacity at pH 8.0 could be lower than those of Bicine-NaOH
hydrolysis without removing betaine. and Tris−HCl buffers. Specifically, when we tested 1.25 g/L of
Optimal Loadings of PETase and MHETase in TPA dissolved in each buffer at 50 mM (initial pH 8.0), the
Enzymatic Hydrolysis of BHET and MHET. To maximize pH of sodium phosphate, Bicine-NaOH, and Tris−HCl
the enzymatic hydrolysis of the whole slurry of PET glycolysis decreased to 7.05, 7.32, and 7.47, respectively (data not
by minimizing unnecessary additions of the enzymes, the shown). Therefore, only Bicine-NaOH and Tris−HCl buffers
enzyme loadings of IsPETaseMut and IsMHETaseMut were were considered suitable based on the effect of their
4004 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 6. Whole-cell bioconversion of TPA and EG in the enzyme hydrolysate of PET glycolysis slurry to PCA and GLA, respectively. (a)
Biosynthetic route of conversion of TPA to PCA. (b) Production of PCA from TPA in the enzyme hydrolysate of PET glycolysis slurry and
reagent-grade TPA by E. coli strain PCA-1. (c) Molar yields of PCA from TPA from different sources on the basis of theoretical maximum. (d)
Biosynthetic route of conversion of EG to GLA. (e) Production of GLA from EG present in the enzyme hydrolysate of PET glycolysis slurry and
from reagent-grade EG using G. oxydans KCCM 40109. (f) Molar yields of GLA from EG from different sources on the basis of theoretical
maximum. Values and error bars represent means ± standard deviations from duplicate experiments.

concentration on enzyme hydrolysis. In the case of Bicine- PET starts at one end of the PET polymer molecule.32 In other
NaOH buffer, when 2.3 U/L of the two enzymes was used, the words, if there are multiple free ends of the molecule available
final TPA production after 6 h increased from 3.1 to 8.5 g/L as for hydrolysis initiation, PET hydrolysis can be enhanced. For
the buffer concentration increased from 50 to 500 mM (Figure instance, among various PET-derived samples, the whole slurry
5d). In addition, increasing the concentration of Bicine-NaOH or the oligomers of PET glycolysis obtained in this study
buffer to 1 M did not adversely affect the enzymatic reaction yielded considerably higher product titers in the enzymatic
with respect to the product profile (Figure 5e). However, when hydrolysis than did other PET polymer samples (Table 1).
1 M Tris−HCl buffer was used, an unknown peak was Lastly, from the PET glycolysis slurry containing 40 g/L of
generated that was dissimilar with the peaks representing BHET and 62.3 g/L of EG, the enzyme hydrolysate consisting
MHET, BHET, and TPA (Figure S11). This indicates that of 31.0 g/L of TPA and 85.5 g/L of EG was successfully
using Tris−HCl buffer at a high concentration could reduce obtained after enzymatic hydrolysis in one-pot processing.
product yield and alter the product profile. Furthermore, when PET-degrading hydrolases have been identified and
1 M Bicine-NaOH buffer was used for enzymatic hydrolysis, developed to facilitate the biodegradability of PET; however,
even a low enzyme loading such as 2.3 U/L yielded a TPA titer the product titers obtained from the enzymatic hydrolysis of
as high as 9.5 g/L, which was similar to those obtained at PET are not comparable to those required for industrial
higher enzyme loadings such as 4.5 U/L (8.7 g/L) and 9.0 U/ applications (Table 1). An exception to this was the
L (8.8 g/L) (Figure 5f). Based on the above results, Bicine- engineering of LCC to increase thermal stability and activity
NaOH buffer was selected as the optimal reaction buffer in this above the glass transition temperature of amorphous PET (i.e.,
study to reduce the necessary enzyme loadings. 84 °C),33 owing to which 1 M titer of TPA was formed as a
Optimal Substrate Loading for Enzymatic Hydrolysis product of PET hydrolysis at high temperature (72 °C) (Table
of PET Glycolysis Slurry. Lastly, to maximize TPA 1).11 However, since LCC exhibits limited activity toward
production, the effect of substrate loadings based on BHET crystalline PET, amorphous PET was used as the substrate in
contained in the PET glycolysis slurry (ranging from 10 to 50 the study mentioned, which was obtained by thermomechan-
g/L) on TPA production was investigated (Figure 5g,h). As ical processes involving extrusion and micronization at high
the BHET loading increased from 10 to 40 g/L in the PET temperatures.11 In particular, the substrate for the enzyme
glycolysis slurry, the final TPA titer gradually increased from hydrolysis obtained from commercial post-consumer was the
10.4 to 31.0 g/L after 12 h (Figure 5g). When calculating the low-value residue remaining after producing virgin PET by the
theoretical maximal yield of TPA on the basis of input BHET chemical recycling with the thermomechanical process so that
mass in the PET glycolysis slurry, the TPA yields obtained at it might be more easily degraded than other PET samples. It
10−40 g/L of BHET marginally exceeded 100% (Figure 5h). means that the PET used as the substrate for the enzymatic
This implied that, apart from BHET, there were other reaction in that study was treated by certain chemical recycling.
oligomers present in the PET glycolysis slurry that were Therefore, that study also might agree with the need for
hydrolyzed into TPA (Figure 5g). In fact, PET oligomers chemical pretreatment to increase the amorphous region of
obtained from the PET glycolysis slurry were hydrolyzed better PET. In this study, by applying the glycolysis using betaine as
and produced approximately 1.6 g/L of TPA compared to that the catalyst, which corresponds to the chemical pretreatment,
produced from PET granules (∼10 mg/L) (Figure 5i). This the efficient enzymatic degradation was successfully performed
could be attributed to the fact that the enzymatic hydrolysis of at low temperature.
4005 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

To further apply the one-pot processing of PET in the substrate for the production of high-value materials in the
complete upcycling of PET, the enzyme hydrolysate of the bioconversion process.
PET glycolysis slurry was directly used as the substrate for
whole-cell bioconversion to produce high-value materials
without any separation step.
■ CONCLUSIONS
To depolymerize PET to TPA and EG at high titers and high
Bioconversion of PET Monomers in PET Hydrolysate. yields as a one-pot process, we have integrated the glycolysis
In this study, PET depolymerization was performed in a one- process for PET using the biocompatible betaine catalyst with
pot process without any separation step. Therefore, we tested the optimized enzymatic hydrolysis process for the PET
whether the PET monomers in the PET hydrolysate could be glycolysis slurry. The synergistic effect of the cation and anion
applied to the bioconversion for producing value-added of betaine on PET glycolysis was revealed by the DFT analysis.
materials. To do this, whole-cell bioconversion was performed Both the glycolysis of crystalline PET using betaine and the
using metabolically engineered E.coli strain PCA-1 and G. enzymatic hydrolysis using IsPETase and IsMHETase resulted
oxydans KCCM 40109 to convert the PET monomers TPA in the highest titers and yields from both glycolysis and
and EG, respectively. enzymatic hydrolysis of PET compared to values reported
First, TPA was converted to PCA via 1,2-dihydroxy-3,5- previously. This integrated depolymerization process was
cyclo-hexadiene-1,4-dicarboxylate (DCD) in sequential enzy- shown to be further extended to the bioconversion of TPA
matic reactions with TPA 1,2-dioxygenase and DCD and EG to PCA and GLA, respectively, at high yields without
dehydrogenase (Figure 6a). PCA is considered a key precursor inhibitory or side reactions.

■
of various aromatic compounds because many aromatic
compounds derived from lignin were already demonstrated ASSOCIATED CONTENT
to be converted to high-value chemicals via PCA.9,34 *
sı Supporting Information
Moreover, several aromatic compounds, including gallic acid,
The Supporting Information is available free of charge at
pyrogallol, catechol, muconic acid, and vanillic acid, can be https://pubs.acs.org/doi/10.1021/acscatal.0c04014.
synthesized metabolically via PCA. Therefore, to convert TPA
to PCA, the engineered E. coli strain PCA-1 expressing tphAabc Comparison of the activities between wild-type and
and tphB, which encode TPA 1,2-dioxygenase and DCD mutant PETase and MHETase (Figure S1), abbrevia-
dehydrogenase, respectively (both originating from Comamo- tions of betaine, EG, and PET repeat unit in the DFT
nas sp. E6),9 was used as a whole-cell catalyst (Figure 6a). As a analysis (Figure S2), noncovalent bond interactions in
result, E. coli PCA-1 converted 4.5 g/L of TPA in the PET the optimized interaction structures (Figures S3−S7),
hydrolysate to 3.8 g/L of PCA with a molar yield of 90.4% pH change during the enzymatic hydrolysis of PET
(mol/mol), which was similar to that obtained using reagent- sodium phosphate buffer (Figure S8), effect of pH on
grade TPA (Figure 6b,c). The final PCA titer obtained in this IsPETaseMut activity (Figure S9), pH change of the
study was 8.8 times higher than that obtained in a previous enzyme reaction mixture (Figure S10), HPLC analysis of
study.9 enzymatic hydrolysis product of BHET(Figure S11),
Next, EG in the PET hydrolysate was converted to a high- gas-phase energies and relative gas-phase energies of
value chemical, GLA (Figure 6). GLA is the smallest α- different interaction structures (Table S1), and non-
hydroxy acid and is widely used as a dyeing and tanning agent, covalent bond interactions in the optimized interaction
flavoring agent, preservative, adhesive, and skin care agent.35 structures (Tables S2−S7) (PDF)
Furthermore, GLA can be used as a monomer of various
biodegradable polymers such as polyglycolate and poly(lactic-
co-glycolic acid) used for medical devices and drug delivery
■ AUTHOR INFORMATION
Corresponding Author
systems.36 To do this, G. oxydans KCCM 40109 was used as a Kyoung Heon Kim − Department of Biotechnology, Graduate
whole-cell catalyst. As a result, G. oxydans converted 30.6 g/L School, Korea University, Seoul 02841, Republic of Korea;
of EG in the PET hydrolysate to 31.4 g/L of GLA with a molar orcid.org/0000-0003-4600-8668; Phone: +82-2-3290-
yield of 91.6% (mol/mol), which also was similar to that of 3028; Email: [email protected]
90.1% (mol/mol) obtained using reagent-grade EG (Figure
6e,f). The final GLA titer obtained in this study was 39.3 times Authors
higher than that obtained in a previous study.9 To the best of Dong Hyun Kim − Department of Biotechnology, Graduate
our knowledge, the titers of PCA and GLA obtained in this School, Korea University, Seoul 02841, Republic of Korea
study were the highest among studies using crystalline PET Dong Oh Han − Department of Biotechnology, Graduate
polymers as a substrate. Furthermore, these results imply that School, Korea University, Seoul 02841, Republic of Korea
the microbial bioconversions were not inhibited by PET Kyu In Shim − School of Interdisciplinary Bioscience and
hydrolysate (Figure 6). Bioengineering, Pohang University of Science and Technology,
In the case of PCA, it can be further converted to monomers Pohang, Gyeongbuk 37673, Republic of Korea
or their precursors of various polymers such as lactones, Jae Kyun Kim − Department of Biotechnology, Graduate
branched monomers, aromatic dicarboxylic acids, and School, Korea University, Seoul 02841, Republic of Korea
dicarboxylic acids.34 These kinds of chemicals converted Jeffrey G. Pelton − California Institute for Quantitative
from PCA can be separated and purified by the combination Biosciences, University of California, Berkeley, California
of acidification and precipitate, extraction, or using anion 94720, United States
exchange resins. 34 In addition, GLA present in the Mi Hee Ryu − Bio-based Chemistry Research Center,
bioconversion medium can be also separated by using ion- Advanced Convergent Chemistry Division, Korea Research
exchange resins.37 Therefore, the enzyme hydrolysate from the Institute of Chemical Technology, Daejeon 34114, Republic
PET glycolysis slurry used in this study is suitable as a of Korea
4006 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Jeong Chan Joo − Bio-based Chemistry Research Center, (9) Kim, H. T.; Kim, J. K.; Cha, H. G.; Kang, M. J.; Lee, H. S.;
Advanced Convergent Chemistry Division, Korea Research Khang, T. U.; Yun, E. J.; Lee, D.-H.; Song, B. K.; Park, S. J.; Joo, J. C.;
Institute of Chemical Technology, Daejeon 34114, Republic Kim, K. H. Biological Valorization of Poly(ethylene terephthalate)
of Korea Monomers for Upcycling Waste PET. ACS Sustainable Chem. Eng.
Jeong Woo Han − School of Interdisciplinary Bioscience and 2019, 7, 19396−19406.
(10) Yoshida, S.; Hiraga, K.; Takehana, T.; Taniguchi, I.; Yamaji, H.;
Bioengineering and Department of Chemical Engineering,
Maeda, Y.; Toyohara, K.; Miyamoto, K.; Kimura, Y.; Oda, K. A
Pohang University of Science and Technology, Pohang, Bacterium that Degrades and Assimilates Poly(ethylene terephtha-
Gyeongbuk 37673, Republic of Korea; orcid.org/0000- late). Science 2016, 351, 1196−1199.
0001-5676-5844 (11) Tournier, V.; Topham, C. M.; Gilles, A.; David, B.; Folgoas, C.;
Hee Taek Kim − Bio-based Chemistry Research Center, Moya-Leclair, E.; Kamionka, E.; Desrousseaux, M.-L.; Texier, H.;
Advanced Convergent Chemistry Division, Korea Research Gavalda, S.; Cot, M.; Guémard, E.; Dalibey, M.; Nomme, J.; Cioci, G.;
Institute of Chemical Technology, Daejeon 34114, Republic Barbe, S.; Chateau, M.; André, I.; Duquesne, S.; Marty, A. An
of Korea; orcid.org/0000-0002-3941-031X Engineered PET Depolymerase to Break Down and Recycle Plastic
Complete contact information is available at: Bottles. Nature 2020, 580, 216−219.
(12) de Castro, A. M.; Carniel, A.; Nicomedes Junior, J.; da
https://pubs.acs.org/10.1021/acscatal.0c04014
Conceičaõ Gomes, A.; Valoni, É . Screening of Commercial Enzymes
for Poly(ethylene terephthalate) (PET) Hydrolysis and Synergy
Author Contributions Studies on Different Substrate Sources. J. Ind. Microbiol. Biotechnol.
#
D.H.K and D.O.H. contributed equally. 2017, 44, 835−844.
Author Contributions (13) Liu, R. Y. F.; Hu, Y. S.; Schiraldi, D. A.; Hiltner, A.; Baer, E.
K.H.K., H.T.K., and J.C.J. conceived and designed the project. Crystallinity and Oxygen Transport Properties of PET Bottle Walls. J.
Appl. Polym. Sci. 2004, 94, 671−677.
D.H.K. and D.O.H. designed and performed all the experi-
(14) Tomás-Pejó, E.; Alvira, P.; Ballesteros, M.; Negro, M. J.
ments and analyzed the data. J.K.K. provided the technical Pretreatment Technologies for Lignocellulose-to-Bioethanol Con-
assistance for the whole-cell conversion experiment. J.G.P. version. In Biofuels; Pandey, A., Larroche, C., Ricke, S. C., Dussap, C.-
performed the NMR analysis. K.I.S. and J.W.H. performed the G., Gnansounou, E., Eds.; Academic Press: Amsterdam, 2011;
DFT analysis. M.H.R. performed the FT-IR analysis. D.H.K., Chapter 7, pp. 149−176.
D.O.H., and K.H.K. wrote the initial version of manuscript, (15) al-Sabagh, A. M.; Yehia, F. Z.; Eissa, A.-M. F.; Moustafa, M. E.;
and all authors participated in writing, reviewing, and Eshaq, G.; Rabie, A.-R. M.; ElMetwally, A. E. Glycolysis of
approving the manuscript. Poly(ethylene terephthalate) Catalyzed by the Lewis Base Ionic
Notes Liquid [Bmim][OAc]. Ind. Eng. Chem. Res. 2014, 53, 18443−18451.
(16) Yue, Q. F.; Wang, C. X.; Zhang, L. N.; Ni, Y.; Jin, Y. X.
The authors declare no competing financial interest.

■
Glycolysis of Poly(ethylene terephthalate) (PET) Using Basic Ionic
Liquids as Catalysts. Polym. Degrad. Stab. 2011, 96, 399−403.
ACKNOWLEDGMENTS (17) Jehanno, C.; Flores, I.; Dove, A. P.; Müller, A. J.; Ruipérez, F.;
This work was supported by the Mid-career Researcher Sardon, H. Organocatalysed Depolymerisation of PET in a Fully
Program (2020R1A2B5B02002631) through the National Sustainable Cycle Using Thermally Stable Protic Ionic Salt. Green
Chem. 2018, 20, 1205−1212.
Research Foundation of Korea (NRF). D.H.K. acknowledges
(18) Liu, B.; Fu, W.; Lu, X.; Zhou, Q.; Zhang, S. Lewis Acid−Base
the grant support by the NRF (2020R1C1C1008196). We Synergistic Catalysis for Polyethylene Terephthalate Degradation by
acknowledge the facility support received from the Institute of 1,3-Dimethylurea/Zn(OAc)2 Deep Eutectic Solvent. ACS Sustainable
Biomedical and Food Safety at CJ Food Safety Hall, Korea Chem. Eng. 2019, 7, 3292−3300.
University. (19) Musale, R. M.; Shukla, S. R. Deep Eutectic Solvent as Effective

■ REFERENCES
(1) Geyer, R.; Jambeck, J. R.; Law, K. L. Production, Use, and Fate
Catalyst for Aminolysis of Polyethylene Terephthalate (PET) Waste.
Int. J. Plast. Technol. 2016, 20, 106−120.
(20) Wang, Q.; Yao, X.; Geng, Y.; Zhou, Q.; Lu, X.; Zhang, S. Deep
Eutectic Solvents as Highly Active Catalysts for the Fast and Mild
of All Plastics Ever Made. Sci. Adv. 2017, 3, No. e1700782.
(2) Zhu, J.-B.; Watson, E. M.; Tang, J.; Chen, E. Y.-X. A Synthetic Glycolysis of Poly(ethylene terephthalate) (PET). Green Chem. 2015,
Polymer System with Repeatable Chemical Recyclability. Science 17, 2473−2479.
2018, 360, 398−403. (21) Pätzold, M.; Siebenhaller, S.; Kara, S.; Liese, A.; Syldatk, C.;
(3) Christensen, P. R.; Scheuermann, A. M.; Loeffler, K. E.; Helms, Holtmann, D. Deep Eutectic Solvents as Efficient Solvents in
B. A. Closed-Loop Recycling of Plastics Enabled by Dynamic Biocatalysis. Trends Biotechnol. 2019, 37, 943−959.
Covalent Diketoenamine Bonds. Nat. Chem. 2019, 11, 442−448. (22) Craig, S. A. S. Betaine in Human Nutrition. Am. J. Clin. Nutr.
(4) Jones, G. O.; Yuen, A.; Wojtecki, R. J.; Hedrick, J. L.; García, J. 2004, 80, 539−549.
M. Computational and Experimental Investigations of One-Step (23) Schrödinger Release 2015−2: Jaguar, version 8.8; Schrödinger,
Conversion of Poly(carbonate)s into Value-added Poly(aryl ether LLC: New York, NY, 2015
sulfone)s. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 7722−7726. (24) Slater, J. C.; Phillips, J. C. Quantum Theory of Molecules and
(5) Sinha, V.; Patel, M. R.; Patel, J. V. PET Waste Management by Solids Vol. 4: The Self-Consistent Field for Molecules and Solids.
Chemical Recycling: A Review. J. Polym. Environ. 2010, 18, 8−25. Phys. Today 1974, 27, 49.
(6) Geyer, B.; Lorenz, G.; Kandelbauer, A. Recycling of Poly- (25) Son, H. F.; Cho, I. J.; Joo, S.; Seo, H.; Sagong, H.-Y.; Choi, S.
(ethylene terephthalate) − A Review Focusing on Chemical Methods. Y.; Lee, S. Y.; Kim, K.-J. Rational Protein Engineering of Thermo-
eXPRESS Polym. Lett. 2016, 10, 559−586. Stable PETase from Ideonella sakaiensis for Highly Efficient PET
(7) Brems, A.; Baeyens, J.; Vandecasteele, C.; Dewil, R. Polymeric Degradation. ACS Catal. 2019, 9, 3519−3526.
Cracking of Waste Polyethylene Terephthalate to Chemicals and (26) Palm, G. J.; Reisky, L.; Böttcher, D.; Müller, H.; Michels, E. A.
Energy. J. Air Waste Manage. Assoc. 2011, 61, 721−731. P.; Walczak, M. C.; Berndt, L.; Weiss, M. S.; Bornscheuer, U. T.;
(8) Rahimi, A.; García, J. M. Chemical Recycling of Waste Plastics Weber, G. Structure of the Plastic-Degrading Ideonella sakaiensis
for New Materials Production. Nat. Rev. Chem. 2017, 1, No. 0046. MHETase Bound to a Substrate. Nat. Commun. 2019, 10, 1717.

4007 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008
ACS Catalysis pubs.acs.org/acscatalysis Research Article

(27) Kim, D. H.; Yun, E. J.; Lee, S.-H.; Kim, K. H. Novel Two-Step (44) Tiso, T.; Narancic, T.; Wei, R.; Pollet, E.; Beagan, N.; Schröder,
Process Utilizing a Single Enzyme for the Production of High-Titer K.; Honak, A.; Jiang, M.; Kenny, S. T.; Wierckx, N.; Perrin, R.;
3,6-Anhydro-L-galactose from Agarose Derived from Red Macroalgae. Avérous, L.; Zimmermann, W.; O’Connor, K.; Blank, L. M. Bio-
J. Agric. Food Chem. 2018, 66, 12249−12256. Upcycling of Polyethylene Terephthalate. BioRxiv 2020.
(28) Jung, Y. H.; Kim, K. H. Acidic Pretreatment. In Pretreatment of (45) Falkenstein, P.; Gräsing, D.; Bielytskyi, P.; Zimmermann, W.;
Biomass; Pandey, A., Negi, S., Binod, P., Larroche, C., Eds.; Elsevier: Matysik, J.; Wei, R.; Song, C. UV Pretreatment Impairs the Enzymatic
Amsterdam, 2015; Chapter 3, pp. 27−50. Degradation of Polyethylene Terephthalate. Front. Microbiol. 2020,
(29) Kim, D. H.; Park, H. M.; Jung, Y. H.; Sukyai, P.; Kim, K. H. 11, 689.
Pretreatment and Enzymatic Saccharification of Oak at High Solids
Loadings to Obtain High Titers and High Yields of Sugars. Bioresour.
Technol. 2019, 284, 391−397.
(30) Modenbach, A. A.; Nokes, S. E. The Use of High-Solids
Loadings in Biomass PretreatmentA Review. Biotechnol. Bioeng.
2012, 109, 1430−1442.
(31) Fukushima, K.; Coulembier, O.; Lecuyer, J. M.; Almegren, H.
A.; Alabdulrahman, A. M.; Alsewailem, F. D.; Mcneil, M. A.; Dubois,
P.; Waymouth, R. M.; Horn, H. W.; Rice, J. E.; Hedrick, J. L.
Organocatalytic Depolymerization of Poly(ethylene terephthalate). J.
Polym. Sci., Part A: Polym. Chem. 2011, 49, 1273−1281.
(32) Zhang, L. Kinetics of Hydrolysis of Poly(ethylene tereph-
thalate) Wastes Catalyzed by Dual Functional Phase Transfer
Catalyst: A Mechanism of Chain-End Scission. Eur. Polym. J. 2014,
60, 1−5.
(33) Chen, D.; Zachmann, H. G. Glass Transition Temperature of
Copolyesters of PET, PEN and PHB as Determined by Dynamic
Mechanical Analysis. Polymer 1991, 32, 1612−1621.
(34) Johnson, C. W.; Salvachúa, D.; Rorrer, N. A.; Black, B. A.;
Vardon, D. R.; St. John, P. C.; Cleveland, N. S.; Dominick, G.;
Elmore, J. R.; Grundl, N.; Khanna, P.; Martinez, C. R.; Michener, W.
E.; Peterson, D. J.; Ramirez, K. J.; Singh, P.; VanderWall, T. A.;
Wilson, A. N.; Yi, X.; Biddy, M. J.; Bomble, Y. J.; Guss, A. M.;
Beckham, G. T. Innovative Chemicals and Materials from Bacterial
Aromatic Catabolic Pathways. Joule 2019, 3, 1523−1537.
(35) Salusjärvi, L.; Havukainen, S.; Koivistoinen, O.; Toivari, M.
Biotechnological Production of Glycolic Acid and Ethylene Glycol:
Current State and Perspectives. Appl. Microbiol. Biotechnol. 2019, 103,
2525−2535.
(36) Suggs, L. J.; Moore, S. A.; Mikos, A. G. Synthetic Biodegradable
Polymers for Medical Applications. In Physical properties of polymers
handbook; Mark, J. E., Ed.; Springer: New York, 2007; pp. 939−950.
(37) Hua, X.; Cao, R.; Zhou, X.; Xu, Y. Integrated Process for
Scalable Bioproduction of Glycolic Acid from Cell Catalysis of
Ethylene Glycol. Bioresour. Technol. 2018, 268, 402−407.
(38) Liu, B.; He, L.; Wang, L.; Li, T.; Li, C.; Liu, H.; Luo, Y.; Bao, R.
Protein Crystallography and Site-Direct Mutagenesis Analysis of the
Poly(ethylene terephthalate) Hydrolase PETase from Ideonella
sakaiensis. ChemBioChem 2018, 19, 1471−1475.
(39) Joo, S.; Cho, I. J.; Seo, H.; Son, H. F.; Sagong, H.-Y.; Shin, T. J.;
Choi, S. Y.; Lee, S. Y.; Kim, K.-J. Structural Insight into Molecular
Mechanism of Poly(ethylene terephthalate) Degradation. Nat.
Commun. 2018, 9, 382.
(40) Austin, H. P.; Allen, M. D.; Donohoe, B. S.; Rorrer, N. A.;
Kearns, F. L.; Silveira, R. L.; Pollard, B. C.; Dominick, G.; Duman, R.;
El Omari, K.; Mykhaylyk, V.; Wagner, A.; Michener, W. E.; Amore,
A.; Skaf, M. S.; Crowley, M. F.; Thorne, A. W.; Johnson, C. W.;
Woodcock, H. L.; McGeehan, J. E.; Beckham, G. T. Characterization
and Engineering of a Plastic-Degrading Aromatic Polyesterase. Proc.
Natl. Acad. Sci. U. S. A. 2018, 115, E4350−E4357.
(41) Ma, Y.; Yao, M.; Li, B.; Ding, M.; He, B.; Chen, S.; Zhou, X.;
Yuan, Y. Enhanced Poly(ethylene terephthalate) Hydrolase Activity
by Protein Engineering. Engineering 2018, 4, 888−893.
(42) Vertommen, M. A. M. E.; Nierstrasz, V. A.; van der Veer, M.;
Warmoeskerken, M. M. C. G. Enzymatic Surface Modification of
Poly(ethylene terephthalate). J. Biotechnol. 2005, 120, 376−386.
(43) Silva, C.; Da, S.; Silva, N.; Matamá, T.; Araújo, R.; Martins, M.;
Chen, S.; Chen, J.; Wu, J.; Casal, M.; Cavaco-Paulo, A. Engineered
Thermobifida fusca Cutinase with Increased Activity on Polyester
Substrates. Biotechnol. J. 2011, 6, 1230−1239.

4008 https://doi.org/10.1021/acscatal.0c04014
ACS Catal. 2021, 11, 3996−4008