Access to this full-text is provided by Springer Nature.
Content available from Nature Communications
This content is subject to copyright. Terms and conditions apply.
Article https://doi.org/10.1038/s41467-022-34908-z
Biodegradation of highly crystallized
poly(ethylene terephthalate) through cell
surface codisplay of bacterial PETase and
hydrophobin
Zhuozhi Chen
1,4
, Rongdi Duan
1,4
, Yunjie Xiao
1,4
,YiWei
1,4
, Hanxiao Zhang
1
,
Xinzhao Sun
1
,ShenWang
1
, Yingying Cheng
1
,XueWang
1
, Shanwei Tong
1
,
Yunxiao Yao
1
, Cheng Zhu
1
,HaitaoYang
1,2,3
, Yanyan Wang
1
&
Zefang Wang
1,2
The process of recycling poly(ethylene terephthalate) (PET) remains a major
challenge due to the enzymatic degradation of high-crystallinity PET (hcPET).
Recently, a bacterial PET-degrading enzyme, PETase, was found to have the
ability to degrade the hcPET, but with low enzymatic activity. Here we present
an engineered whole-cell biocatalyst to simulate both the adsorption and
degradation steps in the enzymatic degradation process of PETase to achieve
the efficient degradation of hcPET. Our data shows that the adhesive unit
hydrophobin and degradation unit PETase are functionally displayed on the
surface of yeast cells. The turnover rate of the whole-cell biocatalyst toward
hcPET (crystallinity of 45%) dramatically increases approximately 328.8-fold
compared with that of purified PETase at 30 °C. In addition, molecular
dynamics simulations explain how the enhanced adhesion can promote the
enzymatic degradation of PET. This study demonstrates engineering the
whole-cell catalyst is an efficient strategy for biodegradation of PET.
Poly(ethylene terephthalate) (PET) is among the most used plastics
around the world1,2. However, large amounts of nondegradable PET
waste has rapidly accumulated in ecosystems worldwide and has
become a global issue3. Chemical, physical, and biologicalmethods of
recycling PET waste have been performed to counter this severe
environmental problem4,5. Among these recycling methods, enzymatic
hydrolysis of PET can provide a green route for recycling PET and
exhibits several advantages, such as being environmentally friendly,
saving energy, and minimizing waste6–8. In the last decade, a variety of
PET hydrolases have been found, including esterases, lipases, and
cutinases9–12. These enzymes can depolymerize amorphous or low-
crystallinity PET to a certain extent. A recent landmark study showed
that a genetically engineered LCC enzyme could achieve a minimum
90% depolymerization of the postconsumer colored-flake PET waste
within 10 h at 72 °C;9however, most of the reported PET hydrolases
can hardly degrade the high-crystallinity PET (hcPET) that is routinely
used for manufacturing bottles and textiles at mild temaperatures6,13.
In 2016, a research group from Keio University identified a novel
PET-hydrolyzing enzyme named PETase from the bacterium Ideonella
sakaiensis 201-F614. One of the most important characteristics of
Received: 22 December 2021
Accepted: 10 November 2022
Check for updates
1
School of Life Sciences, Tianjin Key Laboratory of Function and Application of Biological Macromolecular Structures, College of Precision Instrument and
Opto-electronics Engineering, Key Laboratory of Systems Bioengineering (Ministry of Education), Tianjin University, Tianjin 300072, China.
2
Tianjin Interna-
tional Joint Academy of Biotechnology and Medicine, Tianjin 300457, China.
3
Shanghai Institute for Advanced Immunochemical Studies and School of Life
Science and Technology, Shanghai Tech University, Shanghai 201210, China.
4
These authors contributed equally: Zhuozhi Chen, Rongdi Duan, Yunjie Xiao,
Yi Wei. e-mail: yanyanwang@tju.edu.cn;zefangwang@tju.edu.cn
Nature Communications | (2022) 13:7138 1
1234567890():,;
1234567890():,;
Content courtesy of Springer Nature, terms of use apply. Rights reserved
PETase is its ability to degrade hcPET at a low temperature (30 °C); due
to this property, PETase is a promising candidate for a practical
methodof recycling PET. However, the industrial applicationof PETase
faces a substantial problem, i.e., the enzymatic activity of PETase
toward hcPET remains low. Some pioneering studies have been con-
ducted to address this issue. Numerous enzyme mutants have been
rationally designed based on the structure of PETaseand exhibited an
increase in enzymatic activity toward low-crystallinity PET (lcPET)15–17.
Only a few PETase mutants can improve the degradation capacity of
hcPET to some extent9,18–22. Although previous studies demonstrated
that redesigning PETase was a promising route to enhance its enzyme
performance, the current enzymatic activity of PETase and its variants
remains insufficient to realize the industrial application of hcPET
degradation.
How can the degradation efficiency of PETase be further
increased? The enzymatic degradation of PET is a two-step process in
which the enzyme binds to the polymer substrate and subsequently
catalyzes hydrolytic cleavage23,24. So far, most studies on PETase have
focused on the second catalytic step15,16,18,19,22,25–27, and the possible
impact of the adsorption step on the degradation efficiency of
PETase has received little attention19,28–33. Recently, we constructed a
whole-cell biocatalyst by displaying PETase on the surface of yeast
cells34. Considering that the whole-cell biocatalyst showed certain
advantages over purified PETase, we continued to optimize the sur-
face display system to further increase its degradation efficiency
toward hcPET. By analyzing the crystal structure of PETase, it was
clear that PETase lacked apparent substrate binding motifs, such as
the carbohydrate-binding modules, which are generally observed in
glycoside hydrolases15,16. Several binding modules have been repor-
ted to enhance the enzymatic degradation of polymers by
increasing enzyme adsorption31,35. Therefore, it is possible to increase
the degradation efficiency of the whole-cell biocatalyst by introdu-
cing an external adhesive unit into the surface display system to
control the process of PETase-displaying cells adsorbing on the PET
surface.
Hydrophobins (HFBs), a type of small secreted protein produced
by filamentous fungi, play different roles in fungal growth and devel-
opment, such as theproduction anddispersion of spores, formation of
aerial hyphae, and stabilization of fruiting body structures36–39.In
addition to their roles in aerial growth and reproduction, HFBs can
mediate fungal attachment to hydrophobic surfaces40. For example,
hydrophobic conidiospores thatare dispersed by wind or insects easily
adhere to water-repellent biotic or abiotic substrates41.TheSC3
hydrophobin is involved in attaching Schizophyllum commune hyphae
to hydrophobic surfaces, such as Teflon.These interesting phenomena
were attributed to the self-assembly of HFBs at the interface between
the cell wall and the hydrophobic substrate42.Duringtheself-assembly
process, a hydrophobic patch on the hydrophobin surface could bind
to hydrophobic surfaces through strong hydrophobic interactions43–46.
Due to its unique amphipathic protein structure, hydrophobin plays a
natural role in manipulating the cell surface hydrophobicity of fila-
mentous fungi. Furthermore, hydrophobin has been demonstrated to
increase the attachment of several enzymes (including PET-degrading
enzymes) onto different substrates in vitro47–49.Thespeculated
mechanism that underlies these observations is that hydrophobin can
wet the surface of PET, and as a result, enzymes can more easily con-
tact and attack the hydrophobin-modified PET surface29,49,50.Inspired
by this appealing property of hydrophobin, we proposed that hydro-
phobin could be used to modify the surface hydrophobicity of PETase-
displaying yeast cells to facilitate their attachment on the hydrophobic
PET surface and ultimately enhance the degradation efficiency of the
whole-cell biocatalyst. To test our idea, we developed a codisplay
system by simultaneously displaying PETase and hydrophobin on the
yeast surface. We used the class II hydrophobin I from Trichoderma
reesei (HFBI) as an example, which is a small, amphiphilic globular
protein that readily self-assembles at hydrophilic and hydrophobic
interfaces51,52.
In this work, the data confirms that both PETase and HFBI are
functionallydisplayed on theyeast cell surface. The displayed HFBI can
profoundly increase the hydrophobicity of the yeast cells; thus, as
expected, the attachment of codisplayed cells onto the PET surface is
improved. This codisplay system shows ~328.8-fold higher degrada-
tion efficiency than that of the native PETase toward the hcPET (crys-
tallinity of 45%) at 30 °C. Thecorresponding conversion level for hcPET
is ~10.9% (depolymerization rate of 20.92 mg
products
d−1mg
enzyme
−1)at
30 °C within 10 days. Our study provides a rational organization of
different functional units on the microbial surface for enhanced bio-
catalytic activity, which could find more applications in biocatalysis,
biosensing, and bioenergy.
Results and discussion
HFBI and PETase were functionally codisplayed on the surface of
yeast cells
In our codisplay system, hydrophobin HFBI and PETase should play
different roles based on their unique protein structures. HFBI is
thought to regulate the adsorption of yeast cells on the substrate PET.
PETase is responsible for degrading the substrate PET. Figure 1ashows
the structures of the two functional proteins (PETase and HFBI) that
are codisplayed on the yeastsurface. The fuchsia part in the surface of
HFBI represents a hydrophobic patch, which provides strong hydro-
phobic interactions and is how HFBI binds to various hydrophobic
surfaces53. For PETase, the portion circled in the structure is the active
site, which is much broader than those of the other PET hydrolases;
this active site may explain why PETase can accommodate a large
substrate, suchas PET, at moderate temperatures15,16.Figure1bshowsa
schematic diagram of our codisplay system. The anchoring proteins
GCW51 and GCW61 were used to displayPETase andHFBI on the yeast
surface, respectively. A flexible linker (GGGGSGGGGS) was employed
to link the corresponding anchoring protein with PETase and HFBI
separately to prevent interference from occuring between each
anchoring protein and its displaying target54. To explore the structural
difference between wild-type PETase and linker-attached PETase, the
crystal structures of those two proteins were determined at 2.0 Å
(Fig. 1a, Supplementary Fig. 1a, and Supplementary Table 1) and 1.5 Å
(Fig. 1c and Supplementary Table 1), respectively. We compared the
structure of wild-type PETase and linker-attached PETase and found
that the tertiary structures of these two proteins were quite similar,
with an overall RMSD of 0.352Å (Supplementary Fig. 1b). In the
structure of linker-attached PETase, we can only see a part of the linker
(6 amino acids) in the form of an irregular structure, indicating that the
(GGGGS)
2
linker is flexible. This result is consistent with previous
findings that the GGGGS sequence is the most widely used flexible
linker that can appropriately separate the functional domains55.The
catalytic centers of these two proteins were compared as well. As
shown in Fig. 1d, the two active-site pockets were almost identical. The
above results revealed that the C-terminal fused linker did not induce
obvious structural changes in PETase. To validate our structural find-
ings, we also performed molecular dynamics (MD) studies for the
codisplayed PETase and HFBI (Fig. 1e and Supplementary Fig. 2).In the
simulation, the (GGGGS)
2
linker showed high flexibility, and no unique
structure was observedwhen the linker-attached PETase was displayed
by the yeast cell-wall protein GCW51. Moreover, this flexible linker
separated PETase from GCW51 in space, meaning that the two func-
tional units did not affect each other.
Theoretically, our codisplay system should perform dual func-
tions that mimic the two-step process of PET degradation23,24.Thefirst
function, which is derived from the HFBI part, is the ability to self-
assemble and facilitate the binding of yeast cells onto the hydrophobic
PET surface. The second function, which is inherited from PETase, is to
enzymatically degrade the highly crystallized substrate PET.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 2
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Supplementary Fig. 3 shows the Western blot (WB) results of both
PETase and HFBI atdifferent inductiontimes. Positive bandson the WB
film indicated that both proteins were successfully expressed in the
recombinant yeast. To further demonstrate that HFBI and PETase were
successfully displayed on the yeast cell surface, an immunolocalization
assay was performed. A flag tag was introduced to the N-terminus of
PETase as a response to the detection of anti-flag antibodies. For the
displayed HFBI, a monoclonal antibody against HFBI was used in the
immunofluorescence assay. As shown in Fig. 1f, strong fluorescent
signals of PETase and HFBI were observed on the yeast cell surfaces.
Additionally, no fluorescent signal was detected inside the cells. Fur-
thermore, WB analysis of different cell fractions (whole cells, cell walls,
and protoplasts) was performed to detect the cellular location of
linker-attached PETase. As shown in Supplementary Fig. 4, immune
petase linker
linkerhfbI
gcw51
gcw61
PETase HFBI
Co-display yeast cellHFBI PETase
aGGGGSGGGGS
b
Linker
PETase-linker
c
S160
H237 D206
Catalytic triad
d
Linker
100 ns
HFBI
PETase
GCW51
Linker
e
PETaseHFBIMerge
24 48 72 96
Induction time (h)
f
24 48 72 96
0
1
2
3
4
5
6
7
8
9
Induction time (h)
Fluorescence intensity
g
h
0h
24h
48h
72h
96h
41 42 43 44 45
Retention time (min)
TPA
MHET
BHET
1234
Ratio (HFBI:PETase)
0
400
800
1200
1600
2000
2400
24 48 72 96
Induction time (h)
Turnover rate ( x 10-5 sec -1)
i
PETase
HFBI
3.3
3.9
1.5
1.0
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved
signals of PETase mainly originated from the cell wall faction, apart
from the whole-cell fraction. No obvious signals were detected from
the protoplast fraction, regardless of the codisplay and display system.
All the above results confirmed that both proteins were displayed on
the yeast cell surface, as expected. In addition, the intensity of fluor-
escence was increased with increasing induction time, and the amount
of surface-displayed HFBI was greater than that of the surface-
displayed PETase when the induction time was less than 96 h
(Fig. 1g). These results indicated that induction time played an
important role in the levels of PETase and HFBI expression on the cell
surface34.
After we obtained the recombinant yeast cell that codisplayed
PETase and HFBI, we were eager to know whether the displayed PETase
maintained its enzymatic activity when it was codisplayed with
hydrophobin HFBI. Figure 1h shows the degradation products released
by the recombinant yeast cell that codisplayed PETase and HFBI by
using hcPET as the substrate. The major product released by the dis-
played PETase was mono(2-hydroxyethyl) terephthalic acid (MHET),
and trace amounts of terephthalic acid (TPA) and bis(2-hydroxyethyl)-
TPA (BHET) were detected at the same time. The ratio between the
liberated products of displayed PETase and that of the native PETase
was almost the same (Supplementary Fig. 5d). Together, these results
indicated that the displayed PETase was biologically active and
exhibited the same reaction mechanism as that of native PETase,
cleaving the polymer preferentially at similar positions14,56.
The turnover rate of the recombinant yeast cell was calculated by
quantitatively analyzing the high-performance liquid chromatography
(HPLC) results from Fig. 1h. The standard curve of each released pro-
duct was calculated to accurately measurethe amount of each product
(Supplementary Fig. 5a–c). A semiquantitative WB method was then
used to estimate how much PETase was displayed in the yeast cells. We
first generated a standard curve of wild-type PETase by performing
grayscale analysis with the WB results (Supplementary Fig. 6a, b).
Subsequently, yeast cells (at 48 h induction) were subjected to the
same WB analysis to detect the displayed PETase (Supplementary
Fig. 6c). According to the standard curve, 151.2ng of PETase in a total
of 2 × 107displayed yeast cells. For the Codisplay system, 2 × 107cells
contained 35.2 ng of displayed PETase. As shown in Fig. 1iandSup-
plementary Fig. 7, the turnover rate of the recombinant yeast cell was
dramaticallyaffected by the induction time and the ratiobetween HFBI
and PETase that wasdisplayed onthe cell surface. The ratio reached a
maximum value at the induction time of 48 h when the ratio between
HFBI and PETase displayed on the cell surface also reached its max-
imum. The above results revealed that the amount of HFBI on the cell
surface exhibited an obvious impact on the enzymatic activity of the
displayed PETase57.
In our codisplay system, PETase is theoretically responsible for
degrading the substrate PET. To confirm that surface-displayed HFBI
does not degrade PET as well, three surface display systems were
constructed, including the PETase (S160A)/HFBI codisplay system,
linker-GCW51 (removing PETase)/HFBI codisplay system and HFBI
single-display system, which are shown in Supplementary Fig. 8a-c.
Enzymatic analysis of each surface display system revealed that none
of the displayedsystems constructedabove exhibited PET degradation
activity, as expected (Supplementary Fig. 8d, e), suggesting that
PETase performed an enzymatic function rather than hydrophobin on
the surface of yeast cells.
Surface-displayed HFBI can increase the adsorption of codis-
played cells on PET surfaces by improving their surface
hydrophobicity
In the above study, we confirmed the surface expression of hydro-
phobin HFBI. Microbial adhesion to hydrocarbons (MATH) and water
contact angle (WCA) measurements were performed to verify whe-
ther the displayed HFBI could induce alterations in the surface
hydrophobicity of the recombinant yeast cells. MATH is a classical
technique for determining the surface hydrophobicity of various
cells58,59. This method involves examining the adhesion of cells to
liquid hydrocarbons (e.g., n-butanol, p-xylene), as this adhesion is
directly associated with the hydrophobic surface properties of cells
(Fig. 2a)60. As shown in Fig. 2b, all yeast cells in the control group
(without induction) remained in the lower aqueous phase, indicating
that the cell surface properties were hydrophilic. As induction time
increased, the yeastcells that codisplayed PETase and HFBI gradually
moved into the upper oil phase, causing a concomitant loss in tur-
bidity throughout the lower aqueous phase. The relative hydro-
phobicity of the cell surface defined in the Methods section
continued to increase with the induction time and reached a max-
imum at 96 h (Fig. 2c). To further demonstrate that the increase in
cell surface hydrophobicity was mainly due to the surfaced-displayed
HFBI, a MATH assay was conducted by using yeast cells that dis-
played only PETase.Supplementary Fig. 9 shows that yeast cells with
only PETase remained in the lower aqueous phase within the entire
induction time. Together, these results demonstrated that the dis-
played HFBI was responsible for the increase in the codisplayed cell
hydrophobicity.
WCA is another commonly used method to measure the surface
hydrophobicity of tested cells61. After performing high-density cell
modifications, the WCA of the hcPET film can be determined by
examining the cell surface properties. As shown in Fig. 2d, e, the hcPET
film displayed a typical hydrophobic property (WCA is 85°). After
modifications were performedwith yeast cellsthat codisplayed PETase
andHFBI(48hinduction),theWCAofthehcPETfilm slightly changed
from 85° to 73°, indicating the codisplayed yeast cells exhibited a
hydrophobic surface property. To clarify the contribution of the dis-
played HFBI to the hydrophobicity of the codisplayed yeast cells, the
WCA of yeast cells displaying only PETase was also determined. The
hydrophilic property (WCA is 37°) of the PETase-displayed cell con-
firmed that the HFBI part played an essential role in determining the
cell surface hydrophobicity. Together, these results verified that by
exposing its hydrophobic patch on the protein surface, the displayed
HFBI can indeed increase the hydrophobicity of the codisplayed
cells43–46.
To determine whether the effect of displayed HFBI on cell
hydrophobicity can be transformedinto an effect on the attachment of
yeast cells on the PET surface, we observed the adsorption of the
Fig. 1 | Protein expression and function of P. pastoris GS115/PETase-HFBI at
different induction times. a X-ray crystal structure of the amphiphilic hydro-
phobin HFBI (the hydrophobic patch is marked in red) and PETase (blue indicates
positive charge, red indicates negative charge,and the dotted line indicates active
pocket position). bSchematic diagram of the codisplay system. cThe overall
structure of PETase-linker (the red arrow is marked as Linker). dCatalytic triad
comparison of wild-type PETase with PETase-linker. eMD simulations of the
codisplay system over 100 ns. Each domain is assigned a unique color.
fFluorescence microscopy of immunostained P. pastoris cells expressing PETase
and HFBI ontheir surface under different induction times. Cells were labeledwith a
primary rabbit anti-FLAG antibody followed by a fluorescently labeled secondary
goat anti-rabbit antibody and a primary mouse anti-HFBI antibody followed by a
fluorescently labeled secondary goat anti-mouse antibody. The scale bar is 5μm.
The experiment is repeated three times independently, with similar results
obtained. One representative is shown. gThe intensity of the average fluorescence
of tested cells was determined with ImageJ software. hHPLC analysis of the pro-
ducts released from the hcPET film degraded by PETase and HFBI codisplayed on
the yeast cell. iQuantitative analysis of HPLC results and the protein expression
ratio. The turnover rate was used to evaluate the enzyme activity of GS115/PETase-
HFBI codisplay cells. n=3 independent experiments. Data were presented as mean
values± SD. Source data for panels (g–i) are provided as a Source Data file.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 4
Content courtesy of Springer Nature, terms of use apply. Rights reserved
codisplay cells on the hcPET surface under different conditions. As
shown in Fig. 2f, the codisplaycells coveredalmost all the surface areas
of hcPET. However, under the same binding condition for the control
sample-yeast cells that displayed only PETase, a sharply reduced
attachment to the PET surface was observed. The similar adsorption
differences between the codisplay cells and cells displaying only
PETase could also be observed when the same low concentration of
yeast cells was used in the binding assays (Fig. 2g). We quantified the
aqueous phase
(Yeast suspensions)
oil phase
(N-butanol)
shake oil phase
aqueous phase
Yeast cells
transfer
a
024 48 72 96
Codisplay
b
Contact
angel Codisplay cells
PET film
Control(PET) Codisplay Display GS115
Cell type
de
0
20
40
60
80
Control Codisplay Display GS115
Cell type
Contact angle (°)
85°
73°
37°
29°
Control
Codisplay
Control
Codisplay
fg
Induction time (h)
0244872
96
Induction time (h)
0
20
40
60
80
100
Cell surface
hydrophobicity (%)
c
Fig. 2 | Detection of the cell surface hydrophobicity. a Schematic diagram of the
MATH experiment.bImage the MATH experiment;cMe asurement result of the cell
surface hydrophobicity. n= 3 independent experiments. Data were presented as
mean va lues ± SD. dSchematic diagram of the WCA experiment and image of the
WCA experiment; eMeasurement result of WCA.f,gThe adsorption of the control
sample-yeast cells displaying only PETase and the codisplay cells on the hcPET
surface. The experiments are repeated three times independently, with similar
results obtained. One representative is shown. Source data for panels (c,e)are
provided as a Source Data file.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 5
Content courtesy of Springer Nature, terms of use apply. Rights reserved
yeast cell numbers adsorbed on the hcPET surface for all tested sam-
ples (Supplementary Fig. 10). It was quite clear that the binding num-
bers of the HFBI-displaying yeast cells on the hcPET surface were
approximatelytwice that of the control group ateach tested condition.
The adsorption differences between the codisplay cells and cells dis-
playing only PETase on the surface lcPET could also be observed,
regardless of the concentration of yeast cells in the binding assays
(Supplementary Fig. 11a, b). The number of HFBI-displaying yeast cells
adsorbed on the lcPET surface was approximately twice that of the
control group under each tested condition (Supplementary Fig. 11c, d).
We believe that the similar surface hydrophobicity of hcPET and lcPET
resulted in similar experimental results (Supplementary Fig. 12) by
WCA. To further verify the role of HFBI in the codisplay system, we
designed a codisplay system in which only hydrophobin was removed
and the other parts remained unchanged. Supplementary Fig. 13a
shows that both liker-GCW61 and PETase were successfully displayed
on the surface of yeast cells. Then, we tested the enzymatic degrada-
tion performance against hcPET and lcPET, and we found that the
codisplay system without hydrophobin exhibited a greatly reduced
degradation ability toward both hcPET and lcPET compared with that
of the normal codisplay system (Supplementary Fig. 13b, c). In addi-
tion, the adsorption capacity of this system on the PET substrate was
reduced to approximately half of the original value (Supplementary
Fig. 14). These results clearly revealed that the adhesive unit HFBI was
particularly important for the codisplay system to exhibit a high
degradation capacity to PET. Taken together, all these results con-
firmed that hydrophobin HFBI displayed on the yeast cell surface
introduced the extra binding capacity to the yeast cell on the PET
surface.
The codisplay system exhibited extraordinary enzymatic activ-
ity against PET
We then explored the optimal conditions of the codisplay cells for
hcPET hydrolysis and compared the results with those of purified
PETase. Figure 3a and Supplementary Fig. 15a show that temperature
exhibited an obvious influence on the turnover rates for both the
codisplay cells and native PETase. The codisplay cells exhibited an
enhanced enzymatic activity compared with that of native PETase at
each tested temperature condition. The optimalreaction temperature
was 40°C, similar to native PETase, indicating that the codisplay cells
remained in a heat-labile degradation system14. Since high temperature
can promote the chain mobility of PET to accelerate its enzymatic
degradation, a heat stabilizing unit could be introduced into the cur-
rent surface display system to further improve its degradation effi-
ciency in the future18.Figure3b and Supplementary Fig. 15b shows the
effect of pH on the turnover rates of the codisplay cells. The maximum
turnover rate of the codisplay cells was achieved at pH 9, similar to
native PETase. At other pH conditions, the codisplay cells exhibited
much higher turnover rates than those of native PETase. For example,
both yeast cells displaying only PETase (Supplementary Fig. 16) and
native PETase were almost inactive at pH 10. However, the codisplay
cells maintained more than 50% of their total enzyme activity, indi-
cating that their alkaline tolerance was significantly improved. The
surface-displayed HFBI may contribute to this result since HFBI is very
stable at extreme pH conditions and can even resist protease degra-
dation in vitro44,62–64. The increased tolerance of the codisplay cells to
alkaline environments can facilitate the PET pretreatment process65.
We then explored whether the cell number of codisplay cells
influenced their turnover rate. Figure 3c shows that the turnover rate
for the codisplay cells was at its maximum (2146.5 × 10−5sec−1)and
remained constant when the protein concentration was less than
1.3 nM. When the protein concentration exceeded 1.3 nM, the con-
version rates decreased rapidly. This result revealed that there the
displayed PETase on the yeast surface might be optimal, which cor-
responded to the maximum PET surface-reaching of the displayed
enzyme molecules. Additional displayed PETase molecules sterically
hindered the PET surface and appeared inactive, leading to a
decrease in the turnover rate of the codisplay cells. The native
PETase and the display cells faced similar situations as the codisplay
system, and the turnover rate reached maximum values (6.5 × 10−5
and 227.9 × 10−5sec−1) when the protein concentrations were 370 nM
and 11.3 nM, respectively (Supplementary Fig. 16c and Supplemen-
tary Fig. 17). Finally, we found that the maximum turnover rate of the
codisplay cells was strikingly 328.8-fold higher than that of native
PETase (Fig. 3d).
Then, we performed both SEM and optical microscopy to observe
the morphological changes in the codisplay cell-treated hcPET film. At
the same time, we used native PETase-treated films as controls. As
shown in Fig. 3e, there was almost no surface erosion on the PETase-
treated hcPET film, suggesting that the enzymatic activity of native
PETase against hcPET was rather low. In contrast, apparent cracks and
erosion on the PET surface were observed by SEM when the yeast cells
that codisplayed HFBI and PETasewere applied to hcPET. Microscopic
observation of cross-sections of the PET films was also performed to
further evaluate the degradation results of our codisplay system. As
shown in Fig. 3f-g and Supplementary Fig. 18, almost no change in the
thickness of the PETase-treated hcPET film was observed. A small but
obvious reduction in thickness (~3.2%) was observed when our codis-
played yeast cells degraded the hcPET film.
To further verify the functionality of the codisplay system, the
enzymatic activity of the whole-cell biocatalyst toward a commercially
available lcPET (PET-GF, crystallinity of 6.3%) was also measured.
Considering that the degradation substrate was changed, we opti-
mized the degradation conditions of both the codisplay cells and
purified PETase for lcPET. As shown in Supplementary Fig. 19 and
Supplementary Fig. 20, several reaction conditions, including tem-
perature, pH, and cell number, caused obvious effects on the turnover
rates of the codisplay cells and purified PETase. The optimal reaction
temperature remained at 40 °C (Supplementary Fig. 19a), and the
maximum turnover rate of the codisplay cells was achieved at pH 9,
similar to native PETase (Supplementary Fig. 19b). These results were
similar to thoseobtained by using hcPET as the degradation substrate.
Theturnoverrateforthecodisplaycellsreacheditsmaximum
(2909.3 × 10−2sec−1) when the normalized PETase concentration was
0.65 nM. In contrast, the turnover rate of purified lcPET reached a
maximum value (5.6 × 10−2sec−1) when the protein concentration was
181.4nM (Supplementary Fig. 19c). Accordingly, the maximum turn-
over rate of the codisplay cells against lcPET was 519.5-fold higher than
that of the purified PETase (Supplementary Fig. 19d).
Supplementary Fig. 21 shows the morphological changes in the
codisplay cell-treated lcPET film, as observed by SEM and optical
microscopy. Obvious erosion spots were observed on the surface of
lcPET, which is consistent with the results of hcPET degradation,
indicating that the results were statistically significant. In addition, it
can be seen from the SEM results that the PET films degraded by the
codisplay system were full of round pits with a size of ~5 microns,
which was basically consistent with the size of yeast cells. These results
revealed that yeast cells adsorbed on the PET surface and then exerted
enzyme activity, as seen in the adsorption experiment (Supplementary
Fig. 10 and Supplementary Fig. 11). The purified PETase contained small
corrosion spots locally and was very small. This result indicated that
the codisplay system could efficiently degrade lcPET and hcPET.
In addition, we tested the crystallinity of lcPET and hcPET before
and after degradation at different temperatures using the codisplay
system, as shown in Supplementary Fig. 22. For lcPET, it can be seen
that the crystallinity increases with increasing conversion level, indi-
cating that PETase degraded amorphous PET in the codisplay system.
However, the crystallinity of the PET film treated with the system
without enzyme at different temperatures did not change obviously
compared with that of the control group, indicating that the
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 6
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Codisplay
PETase
20 30 40 50
0
3
6
9
12
1000
2000
3000
4000
5000
Turnover rate (x 10-5 sec-1)
a
0
2
4
6
8
10
600
1200
1800
2400
2345 67 8910
Codisplay
PETase
pH
Turnover rate (x 10-5 sec-1)
b
Codisplay
PETase
0
1
2
3
4
5
6
7
1800
2000
2200
2400
Turnover rate (x 10-5 sec-1)
Protein concentration (nM)
612
18 1000 2000 3000
3701.30
123456
0
600
1200
1800
2400
1.30
c
e
Commercial hcPET
Control PETase Codisplay
f
Control PETase
Codisplay
Commercial hcPET
(Crystallinity 45%)
0
1
2
3
4
5
6
1800
2000
2200
2400
Codisplay PETase
Turnover rate (x 10-5 sec-1)
0
1
300
320
340
Relative activity (Fold)
2146
329
6.5
1
d
0
50
100
150
200
250
175.4 175.2 170.1
Control
PETase
Codisplay
Film thickness (μm)
g
Fig. 3 | Optimization of the degradation system and visualization of hcPET film
degradation. a Effect of temperature, bpH, and cprotein concentration on PET
hydrolysis. n= 3 independent experiments. Data were presented as mean values ±
SD. dComparison of the turnover rate of the codisplay cells and purified PETase
under the optimal conditions using PET as a substrate. n=3 independent experi-
ments. Data were presented as mean values ± SD. eSEM image of commercial
hcPET film before and after incubation with PETase and codisplay cells. (Scale bar:
5μm). fMicroscopic observation of a cross-section of a commercial hcPET film
before and after incubation with PETase and codisplay cells. (Scale bar: 100μm).
gMeasurement result of the cross-section of thecommercial hcPET film. Data were
presented as mean values ± SD. All the experiments are repeated three times
independently, with similar results obtained. One representative is shown for (e)
and (f). Source data for panels (a–e)and(g) are provided as a Source Data file.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved
temperature (range of 20–50 °C) exhibited little effect on the crystal-
linity of PET. The change in crystallinity observed in the experiment
was mainly caused by enzyme digestion. The same results can also be
observed in hcPET samples, except that the crystallinity decreases with
increasing PET degradation, which suggested that the codisplay sys-
tem exhibited no selectivity for PET hydrolysis. At the same time, the
change in temperature did not affect the crystallinity of the hcPET film,
indicating that the change in crystallinity of PET was also caused by
PETase.
Finally, we compared the turnover rate based on the total pro-
ducts of MHET, TPA, and BHET for the purified PETase, yeast cells
that displayed only PETase, and those with codisplay at the optimal
condition using lcPET (Supplementary Fig. 23a) and hcPET (Supple-
mentary Fig. 23b) as substrates. The conversion levels of hcPET and
lcPET within 18h at 30 °C were ~3.0% (depolymerization rate of 3.27
mg
products
h−1mg
enzyme
−1) and 55% (depolymerization rate of 178.15
mg
products
h−1mg
enzyme
−1), respectively (Supplementary Fig. 24).
The codisplay whole-cell biocatalyst system is robust and
efficient
It is expected that the codisplay cells will be used as whole-cell bio-
catalysts on an industrial scale; therefore, several properties related
to industrial applications of the codisplay cells were evaluated,
including thermal stability, reusability, chemical or solvent stability,
storage conditions, and long-term enzymatic activity. Figure 4a
shows that the relative turnover rate of the codisplay cells remained
at 100% when the cells were placed at 30 °C for 7 days, indicating that
the enzyme activity was almost unchanged during this incubation
time. Free PETase lost 40% of its original hydrolytic activity after
1 day of incubation at 30°C, and its enzyme activity was completely
lost on the fifth day. These results revealed that the codisplay cells
exhibited considerably higher thermostability66,67. This is one intrin-
sic advantage of the surface display system68,69. Sanna et al. reported
that native HFBI did not degrade at 30 °C for 18 h62. Therefore, the
excellent thermostability of HFBI might play a role in the thermo-
stability of the codisplay cells. Then, the reusability of the codisplay
cells was investigated.
As shown in Fig. 4b, the codisplay cells retained 85% of the ori-
ginal turnover rate after three rounds of use and 50% after seven
cycles. There are two possible reasonsfor the decline in the turnover
rate. The first reason is that cells are inevitably lost in the recycling
process. The second reason is that the degradation products might
occupy the catalytic pocket of the displayed PETase and inhibit its
activities to some extent70. Next, we determined whether the pre-
sence of organic solvents or detergents would affect the turnover
rate of codisplay cellsdue to the pretreatment process of PET, which
often involves these chemicals. Figure 4c shows that the turnover
rate toward hcPET retained 98, 82, and 72% when the codisplay cells
were incubated in the tested solutions that contained 0.1% Triton X-
100, 10% methanol, and 10% ethanol, respectively. Compared with
the yeast cells that displayed PETase, the codisplay cells showed a
higher turnover rate at each test chemical solution34. For example,
the turnover rate ofyeast cells that displayed PETase remainedat just
50% in 0.1% Triton X-100 solution34. The chemical stability of the
codisplay cells is beneficial to the PET pretreatment processes, such
as cleaning71. It is well known that hydrophobin can maintain its
functional activity under extreme conditions42. Hence, the chemical
stability of hydrophobin HFBI may be related to the stability of the
codisplay cells in the chemical solutions. Finally, the turnover rate of
the whole-cell biocatalyst was evaluated before and after lyophiliza-
tion, which is a dehydration process that facilitates the storage and
transport of codisplay cells72. Supplementary Fig. 25 shows that the
turnover rate of the codisplay cells toward PET remained nearly 100%
after freeze-drying, suggesting that the whole-cell biocatalyst
retained almost total enzymatic activity after the dehydration
process. Therefore, lyophilization is considered a storage option for
the future use of whole-cell biocatalysts.
Next, we applied the codisplay cells to degrade commercial PET
bottles. Figure 4d shows that the whole-cell biocatalyst can efficiently
degrade different highly crystallized PET bottles. According to a pre-
vious method13, we calculated the conversion level of the codisplay
cells against the highly crystallized PET bottles used in our study. As
showninTable1, the conversion levels were 3.1, 3.1, and 2.8% for
Nestle, Coca-Cola, and Pepsi PET bottles, respectively. In Yoshida’s
paper, the conversion level of native PETase against hcPET was
~0.0097%14. Compared to native PETase, the codisplay cells exhibited
much higher degradation efficiency in the highly crystallized PET
bottles. All were consistent with the conversion levels for hcPET and
lcPET. The conversion level of the codisplay system was 3 and 55%
toward hcPET and lcPET, respectively, which was consistent with pre-
vious microscopic observations. We also summarized and calculated
the conversion levels of several reported PETase mutants using the
same procedure in the Methods section (Table 1). Clearly, the codis-
play system showed the best conversion level under 18 h for both high-
and low-crystallinity PET. Finally, to evaluate the long-term function-
ality of the designed whole-cell biocatalyst, we used the codisplay
system to degrade hcPET for 10 days. As shown in Fig. 4e, f, the total
products and relative conversion levels of hcPET in both the codisplay
system and the display system increased with time and reached a
maximum around the ninth day, which was 10.9 and 1.2%, respectively.
However, the degradation activity of wild-type PETase for hcPET was
very low, and the conversion level was only 0.003%. These results
showed that by establishing the codisplay system, the stability of
PETase increases.
Although the conversion level of codisplay cells against the
commercial bottles was the highest ever reported for any degradation
system using PETase to our knowledge, there is still much room for
improvement. For example, we can screen and test more adhesive
units to further improve the binding of the yeast cell onto the PET
surface. By doing so, we hope that the codisplay system canbe applied
in the large-scale biological recycling of PET in the future.
Molecular insights into the two-step degradation of PET by the
codisplay system
To explore how membrane-anchored HFBI/PETase engaged PET
chains and how these interactions facilitated the subsequent cleavage
of PET, all-atom MD simulations were performed. MD simulation has
been adopted by several researchers to investigate the interactions
between wild-type or mutant PETase and their substrate PET in soluble
states11,73–77. In the preequilibrated system (0 ns), randomly distributed
4PET did not interact with the active center of PETase or the hydro-
phobic region of HFBI (Fig. 5a, Supplementary Fig. 26, and Supple-
mentary Movies 1, 2). Over time, 4PET gradually aggregated near the
hydrophobic amino acids of HFBI at ~40–50 ns. At ~60–70 ns, 4PET
completely attached to the hydrophobic patch of HFBI, which occur-
red mainly through the interaction of hydrophobic residues, and
remained mostly bound for the rest of the trajectory. During this
period, 4PET began to gather near the active center of PETase; even-
tually, a number of 4PET molecules were gathered at ~100 ns. Calcu-
lation of the distances between 4PET and the other two proteins
agreed with the possible role of HFBI in the local enrichment of PETs
(Supplementary Fig. 27). From the simulated trajectories, we postu-
lated that our codisplay system completed PET degradation through
two-step processes involving adsorption and hydrolysis, wherein the
hydrophobin HFBI favorably binds to the PET chains and grabs the
chains firmly (Fig. 5b); then, PETaseinteracts with PET and performs a
hydrolysis function (Fig. 5c).
Next, to delineate the interaction between PETase-linker-
GCW51 and hcPET and to elucidate whether PETase-linker-GCW51
follows a conformational selection or induced-fit mechanism, we
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 8
Content courtesy of Springer Nature, terms of use apply. Rights reserved
a
0
10
20
30
40
50
60
70
80
90
100
Relative turnover rate (%)
1234567
Days
0
10
20
30
40
50
60
70
80
90
100
Relative turnover rate (%)
Codisplay
PETase
1234567
Cycles
Codisplay
b
0
2
4
6
8
10
12
0123456789
10 11
Native PETase
Display
Codisplay
hcPET
Relative degradation (%)
Times (d)
f
c
0
20
40
60
80
100
Relative turnover rate (%)
Control TritonControl Methanol Ethanol
Chemical/Solvent conditions
0
200
400
600
800
1000
Released MHET (nmoles)
d
Nestlé Coca-Cola Pepsi Unknown
966 948 878
3
Commercial PET bottles
e
0
1
2
3
4
012345678910
Times (d)
Total products (μmoles)
Native PETase
Display
Codisplay
Fig. 4 | The stability and functionality of the codisplay system. a The thermo-
stability of GS115/PETase-HFBI and purifiedPETase. bEffect of recycling timesby P.
pastoris GS115/PETase-HFBI. cEffect of chemical/solvent conditions on P. pastoris
GS115/PETase-HFBI and GS115/PETase-GCW51. dDegradation of commercial PET
bottles with the whole-cell biocatalyst. The first to third columns of data were
measured by using the codisplay whole-cell biocatalyst, and the fourth column of
data was obtained from ref.14. The total pr oducts(e) and relative conversion levels
(f) of the hcPET degraded by purified PETase, P. pastoris GS115/PETase-GCW51 and
GS115/PETase-HFBI for a long reaction time. All the experiments are repeated three
times. Data were presented as mean values± SD. Source data are provided as a
Source Data file.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved
performed induced fit molecular docking (IFD) analysis of
PETase-linker-GCW51, in which a PET-tetramer represented the
polymer substrate (4PET)73.Thefive highest-scoring docking
poses were subjected to analysis with respect to the spatial
arrangement of the residues of the catalytic triad and the con-
formation of the reactive part of the oligomer substrate73.We
found the best-predicted docking pose while taking the pro-
ductivity of the catalytic triad into account, as shown in Fig. 5c.
The reacting carbonyl carbon of the substrate was bound with the
scissile ethylene glycol moiety in the canonical gauche con-
formation (Ψ
gauche
) in a twisted chain conformation, which was
consistent with how that wild-type PETase binds to 4PET73.
In this study, we developed a codisplay system to mimic the
natural two-step gradation process of PET by a whole-cell bioca-
talyst. Among the codisplay systems, the hydrophobin HFBI dis-
played on the yeast cell surface was proposed to function as an
adhesive unit to enhance the attachment of codisplay cells to the
PET surface, which would achieve the aim of efficient degradation
of hcPET by the whole-cell biocatalyst. Our data showed that the
displayed HFBI causes the codisplay cells to exhibit a hydro-
phobic surface property. A twofold increase in the binding
numbers of the hydrophobin-producing yeast to hydrophobic
PET was observed under the optimal catalytic conditions.
Hydrophobic interactions between yeast cells and their substrate
PET may play a major role in attachment. An enzymatic assay
confirmed that the enzymatic activity of the codisplay cells was
~328.8 times higher than that of the purified PETase. Our results
not only provide an efficient strategy for efficiently biodegrading
hcPET but also demonstrate the plasticity of the surface display
system. By introducing different functional modules into the
surface display system, its performance can be greatly improved.
In the future, more functional units, such as MHETase, that can
hydrolyze MHET to PET reagents can be introduced into the
whole-cell biocatalyst system to further enhance its performance.
Methods
Strains and culture conditions
The microbial strains and plasmids used in this study are listed in
Supplementary Table 2. Pichia pastoris (P. pastoris)strainGS115and
Escherichia coli (E. coli)DH5αwere stored in our laboratory. E. coli
DH5αcells were used in pPICZαA plasmid construction and were
incubated at 37 °C in Luria broth (LB) low salt medium (1% w/v tryp-
tone,0.5%w/vyeastextract,and0.5%w/vNaCl)supplementedwith
100 μg/mL zeocin and were also used in pPIC9 plasmid construction
and were incubated at 37 °C in LB medium (1% w/v tryptone, 0.5% w/v
yeast extract, and 1% w/v NaCl) supplemented with 100 μg/mL ampi-
cillin. P. pastoris yeast strains were cultured at 30 °C in the following
media: YPD (1% w/v yeast extract, 2% w/v peptone, and 2% w/v glucose)
for subcultivation, BMGY (1% w/v yeast extract, 2% w/v peptone,
100 mM potassium phosphate pH 6.0, 1.34% w/v yeast nitrogen base
(YNB), and 1% v/v glycerol) for cell growth, and BMMY (same as BMGY
but 1% v/v glycerol was substituted for 1% v/v methanol) for recom-
binant protein production.
Amplification of target genes
The codon-optimized PETase gene sequence (Supplementary
Sequence 1) and the T. reesei HFBI gene sequence (GenBank KU173825)
were synthesized by the BGI Group (China). All primers used for plasmid
construction, which were synthesized by GENEWIZ (China), are listed in
Supplementary Table 3. The gcw51 gene (NCBI accession no.
XM_002493737.1) was amplified by polymerase chain reaction (PCR)
from the genomic DNA of P. pastoris GS115 using primers 51-F/51-R,
which contained an overlap area at the 5′-terminus and an EcoRIsiteat
the 3′-terminus. The petase gene was amplified using the PCR method
with primers P-F/P-R, containing an Xho Isite,aflag tag at the 5′-termi-
nus, and an overlap area at the 3′-terminus. The gcw61 gene (NCBI
accession no. XM_002494287.1) was amplified by PCR from the genomic
DNA of P. pastoris GS115 using primers 61-F/61-R, containing a Not Isite
at the 3′-terminus and an overlap area at the 5′-terminus. The hfb I gene
was amplified using the same method with primers hf-F/hf-R, which
contained an EcoRIsiteatthe5′-terminus and an overlap area at the 3′-
terminus. All PCR products were gel-purified, and then the fusion frag-
ments PETase-GCW51 and HFBI-GCW61 were successfully built through
overlap PCR. The petase-linker-gcw51 and petase-linker genes were
cloned from the already constructed GS115/PETase-HFBI using the
restriction sites Nde IandXho I by the primers p51-F/p51-R and PL-F/PL-R,
respectively.Then,thegenewasclonedintothepET-21a(+)vectorusing
the same restriction sites.
Vector construction and yeast transformation
The fusion fragment petase-gcw51 and the vector pPIC9 were digested
with Xho IandEcoR I at 37 °C for 1 h, respectively. Then, the gene and
vector were ligated in the ligation system at 16 °C overnight to obtain
the recombinant plasmid pPIC9/petase-gcw51. The fusion fragments
hfbI-gcw61, linker-gcw61 and the vector pPICZαA were digested with
EcoRIandNot I at 37 °C for 1h, respectively. Then, the two genes and
Table 1 | Comparison of the degradation abilities of PETase and its mutants toward PET films
Degradation system PET film
(crystallinity)
Reaction
conditions
Total products
(µmoles)
Conversion level
(%)a
Normalized Conversion
level (18 h)
References
Codisplay 45% 30 °C 18 h 0.92 3.0 3.0% Our work
Nestle 30 °C 18 h 0.97 3.1 3.1% Our work
Coca-Cola 30 °C 18 h 0.95 3.1 3.1% Our work
Pepsi 30 °C 18 h 0.88 2.8 2.8% Our work
S121E/D186H/R280A 41.79% 30 °C 72 h 0.018 0.041 0.015% 18
41.79% 30 °C 72 h 0.037 0.085 0.04% 18
S121E/D186H/S242T/N246D 41.79% 37 °C 20 day 0.24 0.55 0.021% 20
IsPETase 30% 30 °C 18 h 0.0045 0.0097 0.0097% 14
DuraPETase 30% 37 °C 10 day 1.55 15 ± 1 1.38% 17
S238F/W159H 14.8 ± 0.2% 30 °C 96 h 0.62 1.43 0.27% 11
Codisplay Gf-PET 6.3% 30 °C 18 h 25.05 55 55% Our work
TS-PETase Gf-PET 6.3% 30 °C 6 day 4.2 9.61 1.20% 21
R280A Unknown 30 °C 18 h 0.0055 0.013 0.013% 15
aThe calculation method is from Table S2 in Fusako Kawai et al. 2019.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved
vector were ligated in the ligation system at 16 °C overnight to obtain
the recombinant plasmids pPICZαA/hfbI-gcw61 and pPICZαA/lin-
ker-gcw61.
To display PETase and HFBI on the cell surface, the recombinant
plasmids pPIC9/petase-gcw51 were then linearized with Stu I, and
pPICZαA/hfbI-gcw61 and pPICZαA/linker-gcw61 were then linearized
with Sac I. Then, the linearized products were subsequently trans-
formed into the host strain GS115 by electroporation.
Finally, the genome of the positive transformants selected from
YPD plates containing 100 μg/mL Zeocin as the selective marker was
used as a template, and PCR was carried out using the target gene
primers and AOX primers to screen three kinds of true positive
transformants, GS115/PETase-HFBI, GS115/PETase, and GS115/
PETase-ΔHFBI.
Western blot and immunofluorescence microscopy analysis
The purified Flag-tagged PETase protein was used to prepare protein
standard curves. A nanodrop was used to accurately measure the
PETase concentration. Proteins (10, 50, 200, and 300ng) were accu-
rately obtained for SDS-PAGE and transferred to PVDF Immobilon-P
transfer membranes (0.45 μM pore size, Sigma-Aldrich) under 100V
for 1 h for use in immunoblot analyses. The antibodies used for Wes-
tern blotting were primary rabbit anti-FLAG antibody (ABclonal, cata-
log# AE092, 1:2000), primary mouse HFBI antibody (storage in our
laboratory, 1:1000), primary mouse anti-His antibody (SUNGENE BIO-
TECH, catalog# LK8001, 1:1000), goat anti-rabbit IgG antibody (SUN-
GENE BIOTECH, catalog# LK2001, 1:2000), and goat anti-mouse
antibody (SUNGENE BIOTECH, catalog# LK2003, 1:2000). All serial
Western blot data were representative of at least three biological
experiments. The chemiluminescent substrate (Thermo Fisher Scien-
tific) was added for imaging detection (GelDoc XR+, Bio-Rad, USA).
Then, ImageJ was used to quantify the grayscale of the lane and draw
the standard curve.
To determine the soluble and nonsoluble ratio of PETase
expressed in the yeast strain, we performed Western blot analysis on
whole cells, cell walls, and protoplasts. Before the samples were pre-
pared for SDS-PAGE, the OD
600
value of the yeast solution was mea-
sured, and the cell counts were all calculated as the same as 2 ×108.
For the whole-cell proteins, the cells were disrupted in 50 μL
buffer A (20 mM Tris/HCl, pH 7.5,200 mM NaCl) by a bead beater (SI-
D258, Scientific Industrial) at 4°C. Then, 5× SDS-PAGE loading buffer
(E153-01, GenStar, China) was added to a final concentration of 1×. After
being boiled at 100 °C for 10 min, the samples were centrifuged at
13,000×gfor 15 min, and the supernatants were collected and resolved
by SDS-PAGE.
For the cell wall proteins, the cells were disrupted in 50μLbuffer
A (20 mM Tris/HCl, pH 7.5, 200 mM NaCl) by a bead beater (SI-D258,
Scientific Industrial) at 4 °C. Then, 150 μL of 1% (v/v) Triton X-100 was
added to extract cell wall proteins at 4 °C for 30 min78,79. 5× SDS-PAGE
loading buffer (E153-01, GenStar, China) was added to a final con-
centration of 1×. After being boiled at 100 °C for 10 min, the samples
were centrifuged at 13,000×gfor 15 min, and the supernatants were
collected and resolved by SDS-PAGE.
For protoplast proteins, the cells were resuspended in 600 μL
sorbitol buffer, and 25 U lyticase (Tiangen)was added and incubated at
4 °C for 30 min. Samples were centrifuged at 1500×gfor 10 min, and
the cell pellets were collected80. Subsequently, the cell pellets were
disrupted in 5 0 μL buffer A (20 mM Tris/HCl, pH 7.5, 200 mM NaCl) by
a bead beater (SI-D258, Scientific Industrial) at 4 °C. Then, 5× SDS-PAGE
loading buffer (E153-01, GenStar, China) was added to a final con-
centration of 1×. After being boiled at 100 °C for 10 min, the samples
were centrifuged at 13,000×gfor 15 min, and the supernatants were
Fig. 5 | Proposed two-step mechanism of PET degradation by the codisplay
system. a MD simulation of the dynamic process of PET adsorption by the codis-
play system. bHydrophobic interaction between PET substrate and HFBI. cIFD
analysis of the conformational fitting for the codisplay system during hydrolysis.
Each domain is assigned a unique color. The active sites are marked with blue
spheres. 4PET are marked with sticks.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 11
Content courtesy of Springer Nature, terms of use apply. Rights reserved
collected and resolved by SDS-PAGE. Then, the grayscale of the lane
was used to quantify and calculate the proportion of protein in each
cell component.
To quantify the relationship between the number of cells and
the concentration of PETase in the codisplay system and display
system, we used 2 × 107yeast cells after 48 h of induction for SDS-
PAGE electrophoresis and quantitative analysis by Western blot-
ting. According to the standard curve, for the display system, the
PETase expression amount of 2 × 107cells was equivalent to
151.2 ng. For the Codisplay system, the PETase expression amount
of 2 × 107cells was equivalent to 35.2ng. PET degradation
experiments in this study were carried out by taking corre-
sponding multiples of 2 × 107yeast cells, and the concentration of
PETase was calculated according to the proportion.
The antibodies used for immunofluorescence microscopy analy-
sis were FITC goat anti-mouse IgG (H+ L) (Abclonal, catalog# AS001,
1:200) and rhodamine (TRITC) goat anti-rabbit IgG (H + L) (Abclonal,
catalog# AS040, 1:200). To confirm the expression and localization of
PETase-GCW51 and HFBI-GCW61 fusion proteins on the yeast cell
surface, the yeast cells were visualized on a confocal microscope
(UltraView Vox, PerkinElmer, USA).
Protein expression and purification
PETase-linker and PETase-linker-GCW51 were expressed in E. coli
BL21(DE3)inLuriabroth(LB)at16°Cfor16to18hwith0.5mMiso-
propyl β-D-1-thiogalactopyranoside (IPTG). The purification proce-
dures are described as follows:
Bacteria expressing proteins were harvested and resuspended in a
lysis buffer containing 20 mM Tris-HCl pH 7.5, 0.3M NaCl, and 10%
glycerol and lysed by high-pressure homogenization. After cen-
trifugation (16,000×g, 30 min at 4 °C), the supernatant was loaded
onto a nickel nitrilotriaceticacid (Ni-NTA)column (GE Healthcare). The
column was washed using lysis buffer supplemented with 30mM
imidazole and eluted using lysis buffer supplemented with 300mM
imidazole. Finally, the protein was purified by gel-filtration chroma-
tography (Superdex 75 10/300 GL, GE Healthcare) using a buffer
containing 20 mM Tris·HCl pH 7.5, 0.3 M NaCl.
Crystallization, data collection, and structure determination
All crystals were grown by the microbatch-underoil method unless
otherwise specified. PETase was crystallized at 16 °C by mixing 1 μLof
protein (15 mg/mL) with 1 μL of crystallization buffer containing 0.2 M
calcium chloride dihydrate, 0.1 M HEPES sodium pH 7.5, 28% v/v
polyethylene glycol 400. The crystal of PETase-linker was grown at
16 °C from a mixture of 1 μL protein (15 mg/mL) and 1 μL crystallization
buffer containing 10% w/v PEG 8000, 100 mM MES/sodium hydroxide
(pH 6.0) and 200 mM zinc acetate. The crystals were cryoprotected by
Parabar 10312 (previously known as paratone oil). X-ray diffraction
data were collected on beamline BL19U1 at the Shanghai Synchrotron
Radiation Facility at 100K and at a wavelength of 0.97861 Å. Data
integration and scaling were performed using HKL3000. The wild-type
PETase and PETase-linker structures were all solved by molecular
replacement using the structure of IsPETase (PDB ID: 5XJH)15 as a
search model through the PHASER program from the CCP4 package.
Model building and refinement wereperformed using PHENIX (version
1.14) and COOT (version 0.8.9).
Enzyme activity assay for PET
To evaluate the hydrolytic activity of PETase and the codisplayed
recombinant P. pastoris, the hcPET film (Good Fellow, crystallinity, 45%,
thickness, 0.175 mm, diameter, 6 mm) and the lcPET film (Goodfellow
Cambridge,PET-GF,thecrystallinityof6.3%,0.25mmthick,diameter,
6 mm) were used as the substrates for degradation assays with the
purified PETase enzyme and the displayed system. Before the reaction,
the PET film was soaked separately in 0.5% Triton X-100, 10 mM Na
2
CO
3
,
and distilled water (each was performed at 50 °C for 30min at 550 rpm),
and then air-dried for the reaction. Subsequently, the PET film was
placed into a tube with 300 μL buffer containing 50 mM glycine-NaOH
(pH9.0)for18hat30°Cwith370nMpurified enzyme and corre-
sponding yeast cells. To optimize the induction time, purified PETase-
and GS115/PETase-HFBI-displayed yeast cells were induced for 0, 24, 48,
72, and 96 h for the enzyme activity assay. After removing the enzyme-
treated PET film from the reaction mixture,theenzymereactionwas
terminated by heating at 85 °C for 10 min. The reaction mixture samples
were then centrifuged at 12,000×gfor 5 min. The supernatant of each
sample was further analyzed by high-performance liquid chromato-
graphy (HPLC) to quantify the PET monomers released from PET
depolymerization. To compare the PET hydrolytic activity of the dis-
played system with WT PETase across a range of pH values (2.0–10.0) at
20 and 50 °C, a similar experimental setup was used.
Then, the enzymatic activity of the codisplayed PETase was
determined by calculating the turnover rate of the codisplayed PETase
in each experiment. The turnover rate was calculated by normalizing
the amount (moles) of the total products (MHET, BHET, TPA) gener-
ated in each experiment to the amount (moles) of PETase present in
each experiment, and then the resultant ratio was further divided by
the enzymatic degradation time (18 × 3600 s). All the following enzy-
matic activities of codisplayed or native PETasewere determined in the
same way.
For the long-term enzyme activity reaction, the PET film was
placedintoatubewith300μL buffer containing 50 mM glycine-NaOH
(pH 9.0) at 30 °C with 370 nM enzyme. Then, samples were taken every
24 h for 10 consecutive days, and the amount of the three products was
detected by HPLC to calculate the conversion level. A similar experi-
mental setup was used for the displayed system.
The specific activity (SA)9of the enzyme during the PET depoly-
merization reaction, in mg of equivalent products generated per h per
mg of enzyme (mg
products
h−1mg
enzyme
−1) for 18 h reactions (1), and in
mg of equivalent total products generated per d per mg of enzyme
(mg
products
d−1mg
enzyme
−1) for prolonged reactions (2), was determined
by monitoring the liberation of the total products of terephthalic acid
(TPA), mono(ethylene terephthalate) (MHET) and bis(2-hydroxyethyl)
terephthalate (BHET). TPA, MHET, and BHET were measured accord-
ing to standard curves drawn below, as prepared from commercial TPA
and BHET.
SA1=4mPET
18 × mPETase
ð1Þ
SA2=4mPET
10 × mPETase
ð2Þ
HPLC analysis of the degradation product of PET
HPLC was performed on a Waters e2695 equipped with a HyPURITY
C18 (Thermo Fisher Scientific, No 22105-254630) column
(4.6 × 250 mm). The mobile phase was methanol/18 mM phosphate
buffer (pH 2.5) at a flow rate of 0.5 mL min−1,andtheeffluent was
monitored at a wavelength of 240nm. The typical elution conditions
were as follows: 0 to 30 min, 25% (v/v) methanol, and 30 to 50min,
25–100% methanol linear gradient. The total peak areas of MHET, TPA,
and BHET were used to calculate the amount of products in each PET
hydrolysis reaction.
Standard MHET was obtained from the complete hydrolysis of
BHET, which was purchased from Sigma. The detailed process was as
follows. Four millimolar BHET was incubated with 50 nM PETase in
40 mM Na
2
HPO
4
-HCl (pH 7.0), 80 mM NaCl, and 20% (v/v) DMSO at
30 °C. After the complete hydrolysis of BHET to MHET was confirmed
by HPLC, the protein was removed from the reaction mixture with
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Amicon Ultra 10kDa (Merck Millipore), resulting in a 4mM MHET
solution.
Then, MHET, TPA, and BHET of different concentrations were
used to run HPLC, and the standard curves of the three were drawn
according to the relationship between peak area and sample loading.
Cell surface hydrophobicity assay
The hydrophobicity of yeast cell surfaces was determined by using
microbial adhesion to hydrocarbons (MATH), which was modified
from a previous protocol55. The induction time was 0, 24, 48, 72, and
96 h, and the yeast cell suspension was washed twice with PUM buffer
(containing 150 mM phosphate, potassium, and magnesium, pH 7.1).
Then, the absorbance of the cells in the PUM buffer was adjusted to
OD
600
= 4 and defined as A
1
. The cell suspension (0.75 mL) was added
to a 2 mL glass bottle, then 0.75 mL of n-butanol was added, and the
cap was tightly closed. After 30 s of an eddy, the mixture was left for
3 min to complete the two-phase separation. The absorbance of the
water phase at 600 nm was determined as A
2
. The hydrophobicity is
given as a percentage (%) = (A
1
−A
2
)/A
1
.
Contact angle measurements
Yeast cells were harvested by centrifugation, washed once with Milli-Q
water,and finally resuspended in pure ethanol at a concentration of 108
cells mL−1. A small volume of the cell suspension (250 μL) was spread
over the PET film. After drying the first layer of cells, two more layers
were added, completely covering the PET film. Contact angles were
measured on agar plates covered with cells. The measurements were
carried out in a standard contact angle apparatus (KSV Instruments
Ltd, CAM 200). The contact angles were determined automatically
withtheaidofanimageanalysissystem.
Adsorption assay of cells on PET
Each PET film (diameter, 6 mm) was placed in 2 × 106,4×10
6,2×10
7,
4×10
7, and 1.2 × 108mL−1cell suspensions and shaken at 30 °C, and the
PET film was removed after 18h for microscopic examination (BX51,
Olympus, Japan).
Optimization of the codisplay system
To optimize the pH, the reactions were conducted in 50 mM NaH
2
PO
4
-
NaOH (pH 4.0–5.0), 50 mM Na
2
HPO
4
-HCl (pH 6.0–8.0), or 50 mM
glycine-NaOH (pH 9.0–10). To optimize the temperature, the reactions
were conducted at 20, 30, 40, and 50 °C.
Standardizing the cell number to protein concentration
To quantify the displayed PETase concentration, we used Western
blotting for quantitative analysis. First, we used PETase for Western
blot grayscale analysis and plotted a standard curve, as shown in
Supplementary Fig. 6a, b. Subsequently, we used 2 × 107yeast cells
after 48 h of induction for SDS‒PAGE electrophoresis and quantitative
analysis by Western blot, as shown in Supplementary Fig. 6c. Accord-
ing to the standard curve, for the display system, the PETase expres-
sion amount of 2 × 107cells was equivalent to 151.2 ng. For the
Codisplay system, the PETase expression amount of 2 × 107cells was
equivalent to 35.2 ng. PET degradation experiments in the manuscript
were carried out by taking corresponding multiples of 2 × 107yeast
cells, and the concentration of PETase was calculated according to the
proportion.
Optimization of protein concentration for the PET hydrolysis
reaction
The PETase protein concentrations were 0.5, 1, 5, 10, 20, 40, 70, and
100 μgmL
−1. The MHET quality of the PETase reaction with PET at
10 μgmL
−1was defined as 100%, and the relative enzyme activity was
calculated. The GS115/PETase-HFBI cell concentration was standar-
dized to 2 × 106,4×10
6,2×10
7,4×10
7,1.2×10
8,2×10
8,2.8×10
8,and
4×10
8mL−1. All reactions were performed in 50 mM glycine-NaOH (pH
9.0) at 30 °C for 18 h.
Scanning electron microscopy (SEM)
Each PET film (diameter, 6 mm) was e xamined by SEM, both before and
after degradation treatment, under the same conditions. PET samples
were rinsed with 1% SDS, then rinsed with Milli-Q water and ethanol,
sputter-coated with Au, and mounted on aluminum stubs using carbon
tape. SEM imaging was performed using an FEI Quanta 400 FEG
instrument under a low vacuum operating with a gaseous solid-state
detector. Imaging was performed with a beam-accelerating voltage
of 5 keV.
Observing the thickness change of PET film before and after
degradation
To measure the change in PET film thickness, the hydrolyzed PET film
after each incubation was transferred into a 1.5mL tube with buffer
(1 mL, 100 mm Tris-HCl, pH 6.8) containing 5% w/v SDS to remove the
adsorbed protein and cells and was washed with Milli-Q water. The
washed films were cut in half to measure the cross-section and set
vertically on the stage of an inverted microscope equipped with a V10
ocular and a V10 objective. The cross-section was observed in the
bright field. The thickness was calculated from the number of pixels in
the image based on the scale bar.
Reusability and stability assays of the whole-cell biocatalyst
For the thermostability of yeast, the residual yeast activity was mea-
sured after incubation at 30 °C for 1–7 days in 50 mM glycine-NaOH
buffer (pH 9.0). The yeast was recycled after centrifugation at 3500×g
for 5 min. For the chemical/solvent condition’s stability of yeast, the
residual yeast activity wasmeasured afterincubation in 50 mM glycine-
NaOH (pH 9.0) supplemented with methanol (final concentration, 10%
v/v), ethanol (final concentration, 10% v/v), DMSO (final concentration,
10% v/v), Tween (final concentration, 0.1% v/v), and Triton X-100 (final
concentration, 0.1% v/v) for 18 h at 30°C. Then, the PET reaction was
started after washing 3 times with 50 mM glycine-NaOH (pH 9.0) at
3500×gfor 5 min.
Effect of freeze-drying on PET film hydrolysis
For the freeze-dried yeast,the freeze-dried yeast activity wasmeasured
after incubation at 4 °C overnight in 5 0 mM glycine-NaOH buffer (pH
9.0). The OD
600
value was measured to ensure that the cell count was
the same as that before freeze-drying.
Enzyme assays for commercial PET bottles
GS115/PETase-HFBI cells (induced for 48 h, total cell number is
6.2 × 109) were incubated with 6 mm diameter hcPET cut out from
different commercial PET bottles (brands: Coca-Cola, Nestlé, and
Pepsi) in 50mM glycine-NaOH buffer (pH 9.0) for 18h at 30 °C.
Growth curve measurement
The transformant GS115/PETase-HFBI was inoculated into BMGY
medium at 30 °C to an OD
600
of 2 to 6. The culture was centrifuged at
3500×gfor 5 min and resuspended in BMMY medium to an OD
600
of
1. To induce the fusion protein for expression, cells were incubated at
30 °C with 100% methanol every 24 h to a final concentration of 1%.
Pichia pastoris GS115 was used as a negative control. The OD
600
was
measured every 24h from 0 to 216 h, and the growth curve
was drawn.
Termination reaction
All PET hydrolysis reactions were terminated by diluting the
aqueous solution with 18 mM phosphate buffer (pH 2.5) containing
10% (v/v) DMSO, followed by heat treatment (85 °C, 10min). All BHET
hydrolysis reactions were terminated by diluting the aqueous solution
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved
with 16 mM phosphate buffer (pH 2.5) containing 20% (v/v) DMSO
followed by heat treatment (80°C, 10 min). The supernatant obtained
by centrifugation (13,500×g, 10 min) was analyzed by HPLC.
Calculation of the LOD and recovery of HPLC analysis
According to Supplementary Fig. 29, we determined that the LOD of
HPLC analysis was 0.029 nmol. The recovery rate was ~100% in our
study. After the PET hydrolysis reaction was terminated, there were
three locations for the major degradation product MHET. The first
location was the supernatant of the reaction buffer, the second wasthe
surface of the PET film where MHET might adsorb, and the last wasthe
inner surface of the test tube where MHET might adsorb as well.
Therefore, we measured MHET samples from those three locations.
The first sample was obtained by centrifugation. The second and third
samples were obtained by washing the PET film and test tube,
respectively. After the HPLC measurements, we found that nearly 100%
of MHET was in the supernatant of the reaction buffer (Supplementary
Fig. 16). This result was consistent with the result from Yoshida’sstudy,
which only measured MHET from the supernatant of the reaction
buffer.
Differential scanning calorimetry (DSC) measurements
Differential scanning calorimetry was performed using a DSC214
polyma (NETZSCH, Germany) using ≈6–8 mg of a dry PET sample. A
heating rate of 10 °C min−1was applied for the temperature range
from −20 to 300 °C. The glass transition temperature (T
g
), cold
crystallization temperature (T
cc
), and melting point (T
m
) of various
PET samples were obtained using the first heating scan. The initial
fraction crystallinity X
0
was calculated as follows:
X0=X1+4Hcc
4H0
mTcc
ð3Þ
X1=4Hm
4H0
mTm
ð4Þ
4H0
mTcc
=4H0
mTm
4CpTmTcc
ð5Þ
as described before81,82,whereX
0
is the initial crystallinity, X
∞
is the
complete crystallinity, ΔH
m
is the melting enthalpy, and ΔH
cc
is the
cold crystallization enthalpy. ΔH0
m
(T
m
) is the melting enthalpy of pure
crystalline PET at a melting temperature of 140 J g−1,83.ΔH0
m
(T
cc
)isthe
melting enthalpy of pure crystalline PET at the temperature of cold
crystallization, which can be calculated according to Eq. (5). ΔC
p
is the
difference in the heat capacity of amorphous and crystalline PET of
0.17 J g−1K−1,81.
AlphaFold modeling
In this study, AlphaFold2 was used to predict PETase-linker-GCW51 and
HFBI-linker-GCW61. The prediction of complexes was run twice with
different random seeds, and tenmodels were obtained. Beginning with
a visual inspection, four models were selected to check the protein
structural quality for the side chain conformations using the prime
module of Schrödinger 2021-3. Eventually, the one complex with the
highest quality score was selected for further optimization with sub-
sequent MD simulations.
Molecular dynamics simulations
The all-atom MD simulations were performed by Gromacs 2019.6 with
CHARMM36 force field84,85. The structural models of PETase and GCW51
fusion protein, HFBI and GCW61 fusion protein were generated by
AlphaFold286. The 4PET polymer chain and the corresponding para-
meter files (.top and.itp) were generated following previously published
procedures87. For the modeling of solution-state PETase, a predocked
PETase/4PET or PETase-GCW51/4PET complex was solvated with TIP3P
waters and ions (0.15 M NaCl, total of 164,478 atoms). The Nose‒Hoover
thermostat (303.15 K) and Parrinello-Rahman isotropic NPT ensembles
were adopted with h-bond LINCS constraints. Three independent tra-
jectories were performed (250 ns each) after a short 10 ns equilibrium
simulation. To model the membrane-bound PETase, PETase-GCW51 and
HFBI-GCW61 were anchored to a bilayer of lipids through glycosyl-
phosphatidylinositol modifications at protein N-termini. A typical
composition of lipids included 66 molecules of ergosterol, 32 molecules
of POPA, 64 molecules of POPC, 60 molecules of POPE, and 96 mole-
cules of POPI to mimic the yeast membrane environment and maintain
the bilayer stability (15 nm × 15 nm in the xy dimension)88,89. To mimic
PET plastic in a possible solvent-swelling state, 20 individual chains of
4PET were added to the system at random positions along with 6 nm
thickness of TIP3P waters and ions (0.15 M NaCl), the resulting system
contained a total of 343086 atoms. After equilibrium was reached,
simulations were performed using three independent trajectories for
50 ns. Representative conformations were extracted from each trajec-
tory by gmx cluster for further analysis. Quantitative analysis was also
implemented using gmx rms, rmsf, Rg tools.
Molecular docking
Induced fit docking (IFD) experiments were performed using Schrö-
dinger 2021-3, employing the Maestro graphical interface on PETase-
Linker-GCW51 generated by AlphaFold2. A PET-tetramer capped at both
ends as ethyl esters was prepared and minimized in Maestro using the
LigPrep module with the OPLS3e force field90.ThePETase-Linker-
GCW51 structure was prepared in Maestro employing the Protein Pre-
paration Wizard91. Protonation states were assigned according to the
pKa values and a pH of 7 with PropKa. Finally, the structures were
refined through restrained minimization using the OPLS3e force field to
within an RMS gradient of 0.1 kcal mol−1Å−1. Finally, a Glide redocking
step was performed using the extra precision Grid-Based Ligand
Docking with Energetics (Glide XP) algorithm. In Glide XP docking, a
better correlation between best poses and scores was obtained73.
Reporting summary
Further information on research design is available in the Nature
Portfolio Reporting Summary linked to this article.
Data availability
The coordinates for crystal structures of wild-type PETase and PETase-
linker have been deposited in the Protein Data Bank (PDB), with the
accession codes 8GU5 [https://doi.org/10.2210/pdb8GU5/pdb]and
8GU4 [https://doi.org/10.2210/pdb8GU4/pdb], respectively. The data
generated in this study are provided in the Supplementary Information
and the Source Data file provided with this paper. Data were also
available from the corresponding author upon request. Source data
are provided with this paper.
References
1. Schwabl,P.,Köppel,S.,Königshofer,P.,Bucsics,T.&Liebmann,B.
Detection of various microplastics in human stool: a prospective
case series. Ann. Intern. Med. 171,453–457 (2019).
2. Palm, G. J. et al. Structure of the plastic-degrading Ideonella
sakaiensis MHETase bound to a substrate. Nat. Commun. 10,
1717 (2019).
3. Jambeck, J. et al. Plastic waste inputs from land into the ocean.
Science 347,768–771 (2015).
4. Kawecki, D., Scheeder, P. R. W. & Nowack, B. Probabilistic material
flow analysis of seven commodity plastics in Europe. Environ. Sci.
Technol. 52,9874–9888 (2018).
5. Nomura, K. et al. Multiblock copolymers for recycling polyethylene-
poly(ethylene terephthalate) mixed waste. ACS Appl. Mater. Inter-
faces 12,9726–9735 (2020).
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 14
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6. Taniguchi, I. et al. Biodegradation of PET: current status and appli-
cation aspects. ACS Catal. 9,4089–4105 (2019).
7. Zhang, X., Fevre, M., Jones, G. O. & Waymouth, R. M. Catalysis as an
enabling science for sustainable polymers. Chem. Rev. 118,
839–885 (2017).
8. Shoda, S., Uyama, H., Kadokawa, J., Kimura, S. & Kobayashi, S.
Enzymes as green catalysts for precision macromolecular synth-
esis. Chem. Rev. 116,2307–2413 (2016).
9. Tournier, V. et al. An engineered PET depolymerase to break down
and recycle plastic bottles. Nature 580,216–219 (2020).
10. Chen, S., Su, L., Chen, J. & Wu, J. Cutinase: characteristics, pre-
paration, and application. Biotechnol. Adv. 31,1754–1767 (2013).
11. Austin, H. P. et al. Characterization and engineering of a plastic-
degrading aromatic polyesterase. Proc. Natl Acad. Sci. USA 115,
4350–4357 (2018).
12. Tokiwa, Y. & Suzuki, T. Hydrolysis of polyesters by lipases. Nature
270,76–78 (1977).
13. Kawai, F., Kawabata, T. & Oda, M. Current knowledge on enzymatic
PET degradation and its possible application to waste stream
management and other fields. Microb. Biotechnol. 103,
4253–4268 (2019).
14. Yoshida, S. et al. A bacterium that degrades and assimilates poly(-
ethylene terephthalate). Science 351, 1196–1199 (2016).
15. Joo, S. et al. Structural insight into molecular mechanism of poly
(ethylene terephthalate) degradation. Nat. Commun. 9, 382 (2018).
16. Han, X. et al. Structural insight into catalytic mechanism of PET
hydrolase. Nat. Commun. 8, 2106 (2017).
17. Cui, Y. et al. Computational redesign of a PETase for plastic bio-
degradation under ambient condition by the GRAPE strategy. ACS
Catal. 11,1340–1350 (2021).
18. Son, H. F. et al. Rational protein engineering of thermo-stable
PETase from Ideonella sakaiensis for highly efficient PET degrada-
tion. ACS Catal. 9,3519–3526 (2019).
19. Chen, K., Hu, Y., Dong, X. & Sun, Y. Molecular insights into the
enhanced performance of EKylated PETase toward PET degrada-
tion. ACS Catal. 11,7358–7370 (2021).
20. Son, H. F. et al. Structural bioinformatics-based protein engineering
of thermo-stable PETase from Ideonella sakaiensis.Enzym. Microb.
Technol. 141, 109656 (2020).
21. Zhong,J.E.Z.L.,Voigt,C.A.&Sinskey,A.J.Anabsorbancemethod
for analysis of enzymatic degradation kinetics of poly(ethylene
terephthalate) films. Sci. Rep. 11,928(2021).
22. Lu, H. et al. Machine learning-aided engineering of hydrolases for
PET depolymerization. Nature 604, 662–667 (2022).
23. Shah, A. A., Hasan, F., Hameed, A. & Ahmed, S. Biological degra-
dation of plastics: a comprehensive review. Biotechnol. Adv. 26,
246–265 (2008).
24. Ammala, A. et al. An overview of degradable and biodegradable
polyolefins. Prog. Polym. Sci. 36,1015–1049 (2011).
25. Kim, J. W. et al. Functional expression of polyethylene
terephthalate-degrading enzyme (PETase) in green microalgae.
Micro. Cell Fact. 19, 97 (2020).
26. Liu,B.etal.Proteincrystallography and site-direct mutagenesis
analysis of the poly(ethylene terephthalate) hydrolase PETase from
Ideonella Sakaiensis.Chembiochem 19,1471–1475 (2018).
27. Sadler, J. C. & Wallace, S. Microbial synthesis of vanillin from waste
poly(ethylene terephthalate). Green. Chem. 23, 4665–4672 (2021).
28. Furukawa, M., Kawakami, N., Oda, K. & Miyamoto, K. Acceleration of
enzymatic degradation of poly(ethylene terephthalate) by surface
coating with anionic surfactants. ChemSusChem 11,
4018–4025 (2018).
29. Puspitasari, N., Tsai, S.-L. & Lee, C.-K. Class I hydrophobins pre-
treatment stimulates PETase for monomers recycling of waste PETs.
Int. J. Biol. Macromol. 176,157–164 (2021).
30. Xue, R. et al. Fusion of chitin-binding domain from Chit-
inolyticbacter meiyuanensis SYBC-H1 to the leaf-branch compost
cutinase for enhanced PET hydrolysis. Front. Bioeng. Biotechnol. 9,
762854 (2021).
31. Ribitsch, D. et al. Fusion of binding domains to Thermobifida cel-
lulosilytica cutinase to tune sorption characteristics and enhancing
PET hydrolysis. Biomacromolecules 14,1769–1776 (2013).
32. Zhang, Y., Wang, L., Chen, J. & Wu, J. Enhanced activity toward PET
by site-directed mutagenesis of Thermobifida fusca cutinase-CBM
fusion protein. Carbohydr. polym. 97,124–129 (2013).
33. Dai, L. et al. Catalytically inactive lytic polysaccharide mono-
oxygenase PcAA14A enhances the enzyme-mediated hydrolysis of
polyethylene terephthalate. Int. J. Biol. Macromol. 190,
456–462 (2021).
34. Chen, Z. et al. Efficient biodegradation of highly crystallized poly-
ethylene terephthalate through cell surface display of bacterial
PETase. Sci. Total Environ. 709, 136138 (2020).
35. Gamerith, C. et al. Improving enzymatic polyurethane hydrolysis by
tuning enzyme sorption. Polym. Degrad. Stab. 132,69–77 (2016).
36. Aimanianda, V. et al. Surface hydrophobin prevents immune
recognition of airborne fungal spores. Nature 460, 1117–1121 (2009).
37. Maiolo, D. et al. Bioreducible hydrophobin-stabilized supraparticles
for selective intracellular release. ACS Nano 11,9413–9423 (2017).
38. Talbot Fungal biology: coming up for air and sporulation. Nature
398,295–296 (1999). & J., N.
39. Linder,M.,Szilvay,G.R.,Nakarisetala,T.&Penttila,M.Hydro-
phobins: the protein‐amphiphiles of Filamentous Fungi.FEMS
Microbiol. Rev. 29,877–896 (2005).
40. Linder, M. B. Hydrophobins: proteins that self assemble at inter-
faces. Curr. Opin. Colloid Interface Sci. 14,356–363 (2009).
41. Gebbink, M. F. B. G., Claessen, D., Bouma, B., Dijkhuizen, L. &
Wosten, H. A. B. Amyloids - A functional coat for microorganisms.
Nat. Rev. Microbiol. 3, 333–341 (2005).
42. Wosten, H. A. B., Schuren, F. & Wessels, J. G. H. Interfacial self-
assembly of a hydrophobin into an amphipathic protein membrane
mediates fungal attachment to hydrophobic surfaces. EMBO J. 13,
5848–5854 (1994).
43. Meister, K., Baumer, A., Szilvay, G. R., Paananen, A. & Bakker, H. J.
Self-assembly and conformational changes of hydrophobin classes
at the air-water interface. J. Phys. Chem. Lett. 7,4067–4071 (2016).
44. Wang, Z., Michael, L., Qiao, M. & Linder, M. B. Mechanisms of pro-
tein adhesion on surface films of hydrophobin. Langmuir 26,
8491–8496 (2010).
45. Hakanpaa, J. et al. Atomic resolution structure of the HFBII hydro-
phobin, a self-assembling amphiphile. J. Biol. Chem. 279,
534–539 (2004).
46. Gruner, M. S. et al. Self-assembly of class II hydrophobins on polar
surfaces. Langmuir 28,4293–4300 (2012).
47. Espino-Rammer, L. et al. Two novel class II hydrophobins from Tri-
choderma spp. stimulate enzymatic hydrolysis of poly(ethylene
terephthalate) when expressed as fusion proteins. Appl. Environ.
Microbiol. 79,4230–4238 (2013).
48. Ribitsch, D. et al. Enhanced cutinase-catalyzed hydrolysis of poly-
ethylene terephthalate by covalent fusion to hydrophobins. Appl.
Environ. Microbiol. 81,3586–3592 (2015).
49. Puspitasari, N., Tsai, S.-L. & Lee, C.-K. Fungal hydrophobin RolA
enhanced PETase hydrolysis of polyethylene terephthalate. Appl.
Biochem. Biotechnol. 193,1284–1295 (2021).
50. Puspitasari, N. & Lee, C. K. Class I hydrophobin fusion with cellulose
binding domain for its soluble expression and facile purification. Int.
J. Biol. Macromol. 193,38–43 (2021).
51. Nakari-Setälä, T., Aro, N., Kalkkinen, N., Alatalo, E. & Penttilä, M.
Genetic and biochemical characterization of the Trichoderma reesei
hydrophobin HFBI. Eur. J. Biochem. 235,248–255 (1996).
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 15
Content courtesy of Springer Nature, terms of use apply. Rights reserved
52. Oude Vrielink, A. S. et al. Suspended crystalline films of protein
hydrophobin I (HFBI). J. Colloid Interface Sci. 447,107–112 (2015).
53. Xiao, Y. et al. Dual-functional protein for one-step production of a
soluble and targeted fluorescent dye. Theranostics 8,
3111–3125 (2018).
54. Wu, X., Fraser, K., Zha, J. & Dordick, J. S. Flexible peptide linkers
enhance the antimicrobial activity of surface-immobilized bacter-
iolytic enzymes. ACS Appl. Mater. Interfaces 10,
36746–36756 (2018).
55. Chen, X., Zaro, J. L. & Shen, W. C. Fusion protein linkers: property,
design and functionality. Adv. Drug. Deliv. Rev. 65,
1357–1369 (2013).
56. Brueckner, T., Eberl, A., Heumann, S., Rabe, M. & Guebitz, G. M.
Enzymatic and chemical hydrolysis of poly(ethylene terephthalate)
fabrics. J. Polym. Sci. A 46,6435–6443 (2008).
57. Smith, M. R. et al. Elucidating structure-performance relationships
in whole-cell cooperative enzyme catalysis. Nat. Catal. 2,
809–819 (2019).
58. Zoueki, C. W., Tufenkji, N. & Ghoshal, S. A modified microbial
adhesion to hydrocarbons assay to account for the presence of
hydrocarbon droplets. J. Colloid Interface Sci. 344,492–496 (2010).
59. Kaczorek, E., Chrzanowski, Ł., Pijanowska, A. & Olszanowski, A.
Yeast and bacteria cell hydrophobicity and hydrocarbon biode-
gradation in the presence of natural surfactants: rhamnolipides and
saponins. Bioresour. Technol. 99,4285–4291 (2008).
60. Rosenberg, M. Microbial adhesion to hydrocarbons: twenty-five
years of doing MATH. FEMS Microbiol. Lett. 262,129–134 (2006).
61. Faille, C., Lemy, C., Allion-Maurer, A. & Zoueshtiagh, F. Evaluation of
the hydrophobic properties of latex microspheres and bacillus
spores. Influence of the particle size on the data obtained by the
MATH method (microbial adhesion to hydrocarbons). Colloids Surf.
B Biointerfaces 182, 110398 (2019).
62. Askolin,S.,Nakarisetala,T.&Tenkanen,M.Overproduction,pur-
ification, and characterization of the Trichoderma reesei hydro-
phobin HFBI. Appl. Microbiol. Biotechnol. 57,124
–130 (2001).
63. Wang, X. et al. Probing the self-assembly and the accompanying
structural changes of hydrophobin SC3 on a hydrophobic surface
by mass spectrometry. Biophys. J. 87,1919–1928 (2004).
64. Wang, X., Gravelandbikker, J. F., De Kruif, C. G. & Robillard, G. T.
Oligomerization of hydrophobin SC3 in solution: from soluble state
to self-assembly. Protein Sci. 13,810–821 (2004).
65. Liu,X.Y.etal.Inulinhydrolysisandcitricacidproductionfrom
inulin using the surface-engineered Yarrowia lipolytica displaying
inulinase. Metab. Eng. 12, 469–476 (2010).
66. Cooney, C. L. Bioreactors: design and operation. Science 219,
728–733 (1983).
67. Zhang, S. et al. An efficient, recyclable and stable immobilized
biocatalyst based on bioinspired microcapsules-in-hydrogel scaf-
folds. ACS Appl. Mater. Interfaces 8,25152–25161 (2016).
68. Lee, G.-Y. et al. Isomaltulose production via yeast surface display of
sucrose isomerase from Enterobacter sp. FMB-1 on Saccharomyces
cerevisiae.Bioresour. Technol. 102,9179–9184 (2011).
69. de Carvalho, C. C. C. R. Whole cell biocatalysts: essential workers
from nature to the industry. Microb. Biotechnol. 10,250–263 (2017).
70. Barth, M. et al. Effect of hydrolysis products on the enzymatic
degradation of polyethylene terephthalate nanoparticles by a
polyester hydrolase from Thermobifida fusca.Biochem. Eng. J. 93,
222–228 (2015).
71. Fan, S. et al. Controllable display of sequential enzymes on yeast
surface with enhanced biocatalytic activity toward efficient enzy-
matic biofuel cells. J. Am. Chem. Soc. 142, 3222–3230 (2020).
72. Toscani,A.,Risi,C.,Black,G.W.,Brown,N.L.&Castagnolo,D.
Monoamine oxidase (MAO-N) whole cell biocatalyzed aromatiza-
tion of 1,2,5,6-tetrahydropyridines into pyridines. ACS Catal. 8,
8781–8787 (2018).
73. Guo, B. et al. Conformational selection in biocatalytic plastic
degradation by PETase. ACS Catal. 12, 3397–3409 (2022).
74. Jerves,C.,Neves,R.P.P.,Ramos,M.J.,daSilva,S.&Fernandes,P.
A. Reaction mechanism of the PET degrading enzyme PETase stu-
died with DFT/MM molecular dynamics simulations. ACS Catal. 11,
11626–11638 (2021).
75. da Costa, C. H. S. et al. Assessment of the PETase conformational
changes induced by poly(ethylene terephthalate) binding. Proteins
89,1340–1352 (2021).
76. Boneta, S., Arafet, K. & Moliner, V. QM/MM study of the enzymatic
biodegradation mechanism of polyethylene terephthalate. J. Chem.
Inf. Model. 61,3041–3051 (2021).
77. Feng, S. et al. Is PETase- and Is MHETase-catalyzed cascade
degradation mechanism toward polyethylene terephthalate. ACS
Sustain. Chem. Eng.9,9832(2021).
78. Jo,J.H.,Im,E.M.,Kim,S.H.&Lee,H.H.Surfacedisplayofhuman
lactoferrin using a glycosylphosphatidylinositol-anchored protein
of Saccharomyces cerevisiae in Pichia pastoris.Biotechnol. Lett. 33,
1113–1120 (2011).
79. Lodder,A.L.,Lee,T.K.&Ballester, R. Characterization of the Wsc1
protein, a putative receptor in the stress response of Sacchar-
omyces cerevisiae.Genetics 152,1487–1499 (1999).
80. Wang, J.-m et al. Construction of arming Yarrowia lipolytica surface-
displaying soybean seed coat peroxidase for use as whole-cell
biocatalyst. Enzym. Microb. Technol. 135, 109498 (2020).
81. Wei, R. et al. Biocatalytic degradation efficiency of postconsumer
polyethylene terephthalate packaging determined by their polymer
microstructures. Adv. Sci. 6, 1900491 (2019).
82. Wellen, R., Rabello, M. & Canedo, E. Nonisothermal crystallization
and fusion of PET/PS blends. Part I: Effect of PS content. In 29th
International Conference of the Polymer Processing Society (2013).
83. Mehta,A.K.,Gaur,U.K.&Wunderlich,B.K.Equilibriummelting
parameters of poly(ethylene terephthalate). J. Polym. Sci. B. 16,
289–296 (1978).
84. Abraham,M.J.etal.GROMACS: high performance molecular
simulations through multi-level parallelismfromlaptopstosuper-
computers. SoftwareX 1-2,19–25 (2015).
85. Huang, J. & MacKerell, A. D. Jr CHARMM36 all-atom additive protein
forcefield: validat ion based on comparison to NMR data. J.Comput.
Chem. 34,2135–2145 (2013).
86. Jumper, J. et al. Highly accurate protein structure prediction with
AlphaFold. Nature 596,583–589 (2021).
87. Zhang, X., Vázquez, S. A. & Harvey, J. N. Vibrational energy relaxa-
tion of deuterium fluoride in d-dichloromethane: insights from dif-
ferent potentials. J. Chem. Theory Comput. 17,1277–1289 (2021).
88. Jo, S., Lim, J. B., Klauda, J. B. & Im, W. CHARMM-GUI membrane
builder for mixed bilayers and its application to yeast membranes.
Biophys. J. 97,50–58 (2009).
89. Liu, R. et al. Bomidin: an optimized antimicrobial peptide with broad
antiviral activity against enveloped viruses. Front. Immunol. 13,
851642 (2022).
90. Harder, E. et al. OPLS3: a force field providing broad coverage of
drug-like small molecules and proteins. J. Chem. Theory Comput.
12,281–296 (2016).
91. Banks, J. L. et al. Integrated modeling program, applied chemical
theory (IMPACT). J. Comput. Chem. 26,1752–1780 (2005).
Acknowledgements
We thank the staff from beamlines BL17U1, BL18U1, BL02U1, and BL19U1
at Shanghai Synchrotron Radiation Facility (China). This work was sup-
ported by grants from the National Natural Science Foundation of China
(No. 31970048) to Z.W., the National Natural Science Foundation of
China (No. 61971302) to Yanyan W., the National Natural Science Foun-
dation of China (NO. 82202518) to Y.X., the National Natural Science
Foundation of China (NO. 22007071and 22077094) to C.Z.
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 16
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Author contributions
Z.W. and Yanyan W. conceived and supervised the project; Z.W., Yanyan
W., Yi W., Z.C., Y.X., R.D., and H.Y. designed the experiments; C.Z. and
Y.Y. performed the computational simulations; Z.W., Yanyan W., Y.X., Yi
W., Z.C., R.D., C.Z., X.S., Y.C., X.W., S.T., H.Z., S.W., and H.Y. analyzed and
discussed the data; Z.W., Yanyan W., Yi W., Z.C., R.D., C.Z., and Y.X. wrote
the manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information The online version contains
supplementary material available at
https://doi.org/10.1038/s41467-022-34908-z.
Correspondence and requests for materials should be addressed to
Yanyan Wang or Zefang Wang.
Peer review information Nature Communications thanks the anon-
ymous reviewer(s) for their contribution to the peer review of this
work. Peer reviewer reports are available.
Reprints and permissions information is available at
http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to jur-
isdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as
long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if
changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not
included in the article’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted
use, you will need to obtain permission directly from the copyright
holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2022
Article https://doi.org/10.1038/s41467-022-34908-z
Nature Communications | (2022) 13:7138 17
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY 4.0
Content may be subject to copyright.
