Moog et al.

Microb Cell Fact (2019) 18:171

https://doi.org/10.1186/s12934-019-1220-z Microbial Cell Factories

RESEARCH Open Access

Using a marine microalga as a chassis

for polyethylene terephthalate (PET)
degradation
Daniel Moog1,2*, Johanna Schmitt1, Jana Senger3, Jan Zarzycki3, Karl‑Heinz Rexer4, Uwe Linne2,5, Tobias Erb2,3
and Uwe G. Maier1,2

Abstract 
Background:  The biological degradation of plastics is a promising method to counter the increasing pollution of our
planet with artificial polymers and to develop eco-friendly recycling strategies. Polyethylene terephthalate (PET) is a
thermoplast industrially produced from fossil feedstocks since the 1940s, nowadays prevalently used in bottle packag‑
ing and textiles. Although established industrial processes for PET recycling exist, large amounts of PET still end up in
the environment—a significant portion thereof in the world’s oceans. In 2016, Ideonella sakaiensis, a bacterium pos‑
sessing the ability to degrade PET and use the degradation products as a sole carbon source for growth, was isolated.
I. sakaiensis expresses a key enzyme responsible for the breakdown of PET into monomers: PETase. This hydrolase
might possess huge potential for the development of biological PET degradation and recycling processes as well as
bioremediation approaches of environmental plastic waste.
Results:  Using the photosynthetic microalga Phaeodactylum tricornutum as a chassis we generated a microbial cell
factory capable of producing and secreting an engineered version of PETase into the surrounding culture medium.
Initial degradation experiments using culture supernatant at 30 °C showed that PETase possessed activity against PET
and the copolymer polyethylene terephthalate glycol (PETG) with an approximately 80-fold higher turnover of low
crystallinity PETG compared to bottle PET. Moreover, we show that diatom produced PETase was active against indus‑
trially shredded PET in a saltwater-based environment even at mesophilic temperatures (21 °C). The products resulting
from the degradation of the PET substrate were mainly terephthalic acid (TPA) and mono(2-hydroxyethyl) terephthalic
acid (MHET) estimated to be formed in the micromolar range under the selected reaction conditions.
Conclusion:  We provide a promising and eco-friendly solution for biological decomposition of PET waste in a
saltwater-based environment by using a eukaryotic microalga instead of a bacterium as a model system. Our results
show that via synthetic biology the diatom P. tricornutum indeed could be converted into a valuable chassis for
biological PET degradation. Overall, this proof of principle study demonstrates the potential of the diatom system for
future biotechnological applications in biological PET degradation especially for bioremediation approaches of PET
polluted seawater.
Keywords:  Polyethylene terephthalate, PETase, Plastic pollution, Plastic degradation, Diatoms

*Correspondence: [email protected]‑marburg.de
1
Laboratory for Cell Biology, Philipps University Marburg,
Karl‑von‑Frisch‑Str. 8, 35032 Marburg, Germany
Full list of author information is available at the end of the article

© The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creat​iveco​mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided 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 Creative Commons Public Domain Dedication waiver (http://creativecommons.org/
publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Moog et al. Microb Cell Fact (2019) 18:171 Page 2 of 15

Background secreting cell factories with potential application in bio-

Plastic is an extremely useful material with a wide range logical PET recycling by now (see e.g., [13, 17, 18]). How-
of applications and seemingly no longer indispensable ever, certain disadvantages, such as the dependence on
for our daily life. However, the millions of tons of plas- PET substrate presence or the addition of costly carbon
tic waste produced each year have become a major eco- sources (at industrial scale) into the culture/bioreactor
logical issue on our planet in the last decades, mostly for growth, exist for these bacterial systems in biotech-
due to inadequate disposal and the high durability of nology approaches that have to be overcome. Moreover,
the synthetic material [1]. The consequences of plas- I.  sakaiensis and other microorganisms used so far for
tic pollution for Earth’s ecosystems are so far unfore- PETase production are not well adapted to marine habi-
seeable, but it becomes more and more evident that tats (see, e.g., [19])—the environments in which most of
the plastic accumulating in nature is harmful for life. the plastic waste accumulates. Thus, these organisms, for
One major threat is the formation and distribution of example, are not suitable for bioremediation of PET pol-
microplastics, particles smaller than 5  mm, especially luted saltwater.
in waters such as the oceans [2]. Microplastic parti- The diatom Phaeodactylum tricornutum is a marine
cles can include harmful additives and, because of their photosynthetic single-celled eukaryote with a high
small size and physical properties, can adsorb toxic potential for biotechnological applications. P. tricornu-
compounds (e.g., heavy metals and organic pollutants) tum combines the benefits of a photosynthetic organ-
and enter the food chain at the level of small animals or ism that is easily cultivable and rapidly grows under ­CO2
even microorganisms [3–7]. consumption in a saltwater-based environment, with
Polyethylene terephthalate (PET)—a plastic material those of an established laboratory model organism for
intensively used for packaging of liquids (bottles)/food which a comprehensive genetic toolbox exists. That is,
and for production of synthetic textile fibres [8]—is pro- genes can be inserted into (or edited within) the genome
duced from crude oil resources (terephthalic acid and of the diatom via standard methods and their products
ethylene glycol) with highly increasing rates predicted to can be expressed with highest efficiency under induc-
exceed more than 70 million tons per year by 2020 (see ible conditions [20–22]. The diatom P. tricornutum is
[9] and references therein). As for many other plastics, an excellent system for expression of foreign recombi-
the steadily growing demand for PET necessitates an effi- nant proteins such as antibodies, antigens [22–25] and
cient and comprehensive global manufacturing and waste even whole enzymatic pathways [26–28]. Cultivation
management system with the goal of adverting damage of P. tricornutum is cost-efficient, cells can be grown to
from nature. Although efficient processes for industrial high densities and since photoautotrophic the organism
PET (synthesis and) recycling are established, a signifi- does not require supplementation of expensive sugars
cant fraction of PET waste is still incinerated, landfilled or other carbohydrates as carbon source into the growth
or ends up in the environment as macro-, meso-, micro-, medium if a light source is present [22]. The diatom can
and nano-particles due to improper disposal [10–12]. easily be transformed with multiple constructs and as
Recently, a bacterium named Ideonella sakaiensis 201- shown before it has the ability to efficiently secrete syn-
F6 has been isolated from PET waste sources in Japan that thetic recombinant proteins into the medium fraction
is capable of utilizing this plastic as sole carbon source [24]. These features highlight the potential of P. tricornu-
[13]. I.  sakaiensis expresses a whole enzymatic pathway tum as a model organism for synthetic biology and bio-
for PET biodegradation and uptake, with two enzymes, technology and underline the benefits of the diatom over
PET hydrolase (PETase) and mono(2-hydroxyethyl) tere- bacterial expression systems with respect to developing a
phthalic acid hydrolase (MHETase), having the ability to photosynthetic PETase production factory for biological
decompose PET into its environmentally non-hazardous PET decomposition under marine conditions.
monomers—terephthalic acid (TPA) and ethylene glycol To establish P. tricornutum as a chassis for biological
(EG). I.  sakaiensis exhibits the highest natural PET deg- degradation of PET via synthetic biology, the microalga
radation efficiency known so far and PETase as well as was transformed with the genetic elements necessary to
MHETase are improved by protein engineering continu- efficiently produce and secrete PETase into the surround-
ously (see e.g., [9, 14–16]). The key enzyme PETase is nat- ing saltwater medium. We show that PETase secreted
urally secreted by I. sakaiensis, which might adhere to the by P. tricornutum possesses PET degradation ability for
surface of PET to initiate its biodegradation [13], showing different PET substrates at varying (including meso-
the potential of PETase for biological PET degradation philic) conditions. These results highlight the potential
and bioremediation approaches. Besides I. sakaiensis, of the generated microbial cell factory for the develop-
several bacterial systems, including Escherichia and ment of effective photosynthesis-driven bioremediation
Bacillus, have been utilized to generate synthetic PETase approaches for PET.
Moog et al. Microb Cell Fact (2019) 18:171 Page 3 of 15

Results could be detected in the control (wild type). Interest-

Expression and secretion of PETase in the algal system ingly, when compared to the protein standard the sig-
The gene sequence encoding I.  sakaiensis PETase nal for ­PETaseR280A-GFP (clone 24) appeared at more
(improved/engineered version: ­ PETaseR280A [14]) was than 70  kDa while the calculated molecular mass was
adapted to the codon usage of P.  tricornutum and 57.7  kDa, which indicated post-translational modifica-
expressed as fusion with gfp in the diatom to test whether tion of the recombinant protein in P. tricornutum (see
the product ­(PETaseR280A-GFP) is correctly and efficiently below). For the two remaining clones, the GFP fusion
synthesized. Although the ­ PETaseR280A-GFP construct protein could be detected in the pellet fraction only, indi-
was expressed with the endogenous bacterial signal cating that the P­ ETaseR280A-GFP protein is expressed, but
peptide (SP), GFP fluorescence was detected via confo- cannot be secreted by the diatom efficiently (Additional
cal laser scanning microscopy (CLSM) in the ER and file 1: Figure S1).
most likely other compartments of the secretory path- Since both, the relatively bulky GFP moiety and the
way (Fig.  1). In addition, secretion of ­PETaseR280A-GFP non-eukaryotic bacterial signal peptide of Ideonella
into the medium was investigated. For this approach, PETase, might represent factors reducing an efficient
three different P­ ETaseR280A-GFP expressing clones were secretion and/or enzymatic function of the PETase fusion
analyzed for the presence of the enzyme in the culture protein by the diatom, we substituted the endogenous SP
medium via concentration of the proteins in the medium with the SP of P. tricornutum alkaline phosphatase (AP)
fraction, using SDS-PAGE and Western Blot (see “Meth- [29] and replaced the 27.5  kDa GFP by a 1  kDa FLAG-
ods”). To this end, a total volume of 50 ml supernatant/ tag (DYKDDDDK). After the modified gene was trans-
medium fraction of a diatom culture was used. Besides formed into the alga, several clones grew on selection
the concentrated proteins in the medium fraction, a medium and three of them were further analyzed via
total protein extract was obtained from the cell pellet to Western Blot as described above. As depicted in Fig.  2,
analyze both fractions for the presence of the expressed ­PETaseR280A-FLAG was secreted into the medium by
­PETaseR280A-GFP fusion protein. As shown in Additional P.  tricornutum most likely by means of the signal pep-
file 1: Figure S1, at least one of the three clones express- tide of the diatom AP protein as a targeting signal. No
ing the ­PETaseR280A-GFP fusion construct (clone 24) was FLAG-specific signal was detected in the medium frac-
able to secrete the recombinant protein detected in both tion of the wild type control. As only weak signals for
the medium and cell pellet fraction, whereas no signal ­PETaseR280A-FLAG were present for clone 2 and 3 in the

Fig. 1  Expression and localization of PETase-GFP in the diatom P. tricornutum. ­PETaseR280A-GFP (see schematic of fusion protein) was expressed
successfully in the diatom. Confocal laser scanning microscopy showed that the recombinant protein localized in the ER and most likely other
compartments of the secretory pathway. Secretion of the fusion protein could not be analyzed via this method. The lower part of the figure shows a
wild type control in which no recombinant protein is expressed. Only plastid autofluorescence but no GFP signal was detectable. SP signal peptide,
GFP green fluorescent protein, TL transmitted light, PAF plastid autofluorescence, Merge overlay of GFP and PAF
Moog et al. Microb Cell Fact (2019) 18:171 Page 4 of 15

cell pellet fraction, secretion of the recombinant proteins precipitated with TCA and separated via SDS-PAGE fol-
by the diatom clones occurred with high efficiency. To lowed by a Coomassie-staining. The staining revealed
test if the positive signals in the medium fraction were that corresponding bands to both dominant signals
actually a result of secretion and not due to lysis of algal observed in the Western Blot were present in the SDS-gel
cells, we performed a control experiment using an anti- (Additional file 1: Figure S2). A subsequent mass spectro-
body against alpha-tubulin, a component of the cytoskel- metric (MS) analysis shed light on the nature of the two
eton of the eukaryotic cell, which has been established as bands. The identity of both bands could be unambigu-
a suitable control protein for secretion analyses of recom- ously assigned to AP_SP-PETaseR280A-FLAG. Whereas
binant proteins before [24]. Signals for alpha-tubulin analysis of the upper band (< 55  kDa, Additional file  1:
could be observed exclusively in the cell pellet and not Figure S2) resulted in detection of 7 unique peptides and
the medium fraction, confirming that cells remained a coverage of the protein sequence of 28%, for the lower
intact and no substantial cell lysis took place during the band (> 40  kDa, Additional file  1: Figure S2) 8 unique
experiment (Fig. 2). peptides were identified via mass spectrometry covering
As already observed for P ­ETaseR280A-GFP, we again 33% of the protein sequence (see Additional file 2: Tables
detected a putative mass shift for the expressed protein S1 and S2). These results indicate that the FLAG-tag pro-
on the Western Blot. The calculated molecular mass of tein detected in the medium was indeed AP_SP-PETa-
the AP_SP-PETaseR280A-FLAG is 30.4  kDa, those of the seR280A-FLAG. In order to investigate if the observed
expected processed form (SP removed) 28.5  kDa. The mass shift was caused by post-translational modifications
dominant signals observed on the Western Blot were two of the enzyme, we exemplarily tested N-linked glycosyla-
bands between 40 and 50–55  kDa, which was approxi- tion of the secreted protein. To this end, the supernatant
mately 10–25  kDa higher than the predicted molecular of a 500  ml culture of AP_SP-PETaseR280A-FLAG clone
mass (Fig.  2). To investigate the nature of the observed 2 was concentrated to a volume of 250 µl and a fraction
signals, the proteins of the medium fraction of a 500 ml was treated with PNGase F before it was separated on an
culture expressing AP_SP-PETaseR280A-FLAG (clone SDS-gel and analyzed via Western Blot (see “Methods”).
2) were concentrated via 10  kDa cutoff filter units, As shown in Additional file  1: Figure S3, the signals for

Fig. 2  Secretion analysis of PETase-FLAG. a Schematic of the expressed recombinant protein AP_SP-PETaseR280A-FLAG. b Western Blot after
SDS-gel separation of the cell pellet (10 µg of total protein) and medium fractions (total precipitated protein fraction) of 50 ml cultures (induced
at ­OD600 = 0.4) expressing AP_SP-PETaseR280A-FLAG. Detection of recombinant proteins was conducted using an antibody against the FLAG-tag
(α-FLAG). As control for intracellular proteins, an alpha-tubulin antibody (α-Tubulin) was used. Wild type medium and cell pellet fractions as well
as a FLAG positive control lysate (Rockland, FLAG+) served as control protein fractions. AP_SP-PETaseR280A-FLAG clone 2 showed the highest
expression and secretion efficiency (middle), whereas complete secretion of the recombinant protein was only achieved by clone 1 (left). As shown
by the control via alpha-tubulin detection (right), presence of ­PETaseR280A-FLAG in the medium fraction was not due to cell lysis. A signal in the
range of the calculated molecular mass of AP_SP-PETaseR280A-FLAG (30.4 kDa) could only be observed for clone 2 (left and middle). The dominant
signals detected by the FLAG-tag antibody appeared at molecular masses of approximately 40 and 50–55 kDa in AP_SP-PETaseR280A-FLAG clone 1,
2 and 3. Calculated molecular masses: AP_SP-PETaseR280A-FLAG: 30.4 kDa; FLAG-tag, 1 kDa; ­PETaseR280A-FLAG: 28.5 kDa; FLAG+, 60 kDa. AP alkaline
phosphatase, SP signal peptide, WT wild type, AP_# AP_SP-PETaseR280A-GFP clone #. Numbers beside/on the Western Blots indicate molecular
masses of the marker (PageRuler™ Prestained 10–180 kDa Protein Ladder) bands in kDa
Moog et al. Microb Cell Fact (2019) 18:171 Page 5 of 15

AP_SP-PETaseR280A-FLAG treated with PNGase F cor- incubated with small polyethylene terephthalate glycol
responded to a significantly lesser molecular mass than (PETG; a highly amorphous PET copolymer, see “Meth-
the untreated sample (negative control). This indicates ods”) commercial film fragments for 1  week at 30  °C.
that the recombinant protein AP_SP-PETaseR280A-FLAG This temperature is near the optimum for PETase activ-
is (N-linked) glycosylated when expressed in the diatom. ity (~ 35  °C) as reported in the literature [13, 30]. The
liquid fractions of the samples were analyzed (expected
PET degradation experiments using diatom produced was mainly MHET and to a minor degree TPA and bis(2-
and secreted PETase hydroxyethyl) terephthalic acid (BHET) as products) via
Having shown that the diatom efficiently secretes AP_SP- UHPLC (see “Methods” and below), whereas the small
PETaseR280A-FLAG into the culture medium (for analysis PETG film parts were investigated with the SEM for vis-
of production and secretion efficiency by clone 2 within ible alterations of their surface structure. As shown in
a time frame of 7  days see Additional file  1: Figure S4), Fig. 4, incubation of a small PETG film particle with 1 ml
we next investigated whether the secreted enzymes are of the medium fraction (sterile filtered, cutoff 0.22  µm)
able to degrade PET. To this end, we chose two different for 7  days at 30  °C led to similar, although more com-
experimental approaches. In the first one, cells from AP_ prehensive, structures in the surface of the PETG film as
SP-PETaseR280A-FLAG clone 1 were grown in contact to observed after the first approach (PET bottle film deg-
PET film fragments sticking upright in an f/2 (saltwater) radation by AP_SP-PETaseR280A-FLAG clone 1 on solid
agar plate, which was overflowed with 2 ml liquid f/2. The medium, Fig. 3). Again, we observed holes, furrows and
cells were cultivated under inducing conditions (medium branching canyon-like structures in the upper layer of the
supplemented with nitrate) for 2 to 6 weeks. A P. tricor- PETG film, clearly pointing to an efficient and this time
nutum wild type culture grown under similar conditions area-wide degradation process of the plastic material.
served as a negative control. PET film was removed from No obvious change in surface structure was observed
the culture plate, sputtered with gold and analyzed via for a similar PETG substrate incubated with 1 ml of the
scanning electron microscopy (SEM), to monitor PET medium fraction of a wild type culture (Fig.  4, see also
degradation by P ­ETaseR280A-FLAG (see “Methods”). Additional file 1: Figure S7). UHPLC analyses of the con-
Whereas untreated PET film and fragments incubated centrated liquid fractions in which the PETG film was
with wild type cells showed, besides a typical smooth incubated revealed the presence of mainly TPA, whereas
surface, stress marks characteristic of commercially used MHET, the product, which was expected with the high-
water bottles (occasional scratches and surface disrup- est abundance, was present in comparably low quantity
tions), the PET film incubated with AP_SP-PETaseR280A- (Fig. 4). No generation of TPA and MHET was observed
FLAG expressing clone 1 showed a completely different in the wild type control. Interestingly, when an identical
structure (Fig.  3). In these samples, certain areas of the experiment was performed using PET film from a bottle
PET film, which were in contact with P. tricornutum cells instead of commercial PETG film (above) as substrate,
on the agar plate, showed holes, dents, furrows and cavi- no significant TPA and MHET production was detect-
ties clearly visible under the SEM. The small holes were able under the selected reaction conditions, neither in
reminiscent to structures observed in earlier PET degra- the sample incubated with 1  ml of the medium fraction
dation experiments using PETase [13, 15] (see also Addi- of AP_SP-PETaseR280A-FLAG clone 1 nor in the wild type
tional file 1: Figure S5). The furrows sometimes appeared control (not shown).
in a canyon-like shape as if the missing plastic was A similar experiment was conducted with 1  ml of
washed out by a running liquid in a branched manner. In the medium fraction of the strongly expressing AP_
one particular case, we could observe a diatom cell mark SP-PETaseR280A-FLAG clone 2 (dense 500  ml culture
in the form of the fusiform morphotype of P.  tricornu- induced for 3  days; see Additional file  1: Figure S4 for
tum from which several holes and furrows had their ori- ­PETaseR280A-FLAG secretion analysis in relation to
gin (Additional file 1: Figure S6), which overall supports expression time) and bottle PET as well as commercial
functional secretion and enzymatic activity of AP_SP- PETG film. Here the results were highly comparable to
PETaseR280A-FLAG synthesized by the diatom. the observations made in the experiments with AP_SP-
In a second approach, we used liquid cultures of P. PETaseR280A-FLAG clone 1 (see Additional file  1: Fig-
tricornutum wild type and clones expressing AP_SP- ures  S8, S9 and S10). Essentially, two differences were
PETaseR280A-FLAG (clone 1 and 2) grown in volumes of detected: (i) an almost equal amount of TPA and MHET
50, 150 or 500  ml. For a proof of principle experiment, was identified in the sample where PETG film was
1 ml of the medium fraction of an induced (4 days) 50 ml incubated with 1  ml of the medium fraction of AP_SP-
culture expressing AP_SP-PETaseR280A-FLAG (clone PETaseR280A-FLAG clone 2 for 7 days at 30 °C (see Addi-
1) and 1  ml of a wild type control culture were further tional file  1: Figure S8 and Additional file  2: Table  S3)
Moog et al. Microb Cell Fact (2019) 18:171 Page 6 of 15

Fig. 3  Scanning electron microscopic analysis of PET bottle film degradation by PETase-FLAG secreted from P. tricornutum. As depicted in the
upper left part, untreated PET and PET incubated with wild type cells on an f/2 agar plate overflowed with 2 ml f/2 liquid medium showed, besides
a usually smooth surface, occasional stress marks. In contrast, PET incubated for 5 weeks with cells expressing AP_SP-PETaseR280A-FLAG (clone 1),
was lanced by holes, dents, furrows and cavities when inspected via SEM (see also Additional file 1: Figure S5). In a specific area of the PET disk a
structure (imprint) similar to the form of a P. tricornutum cell (fusiform morphotype) from which several holes and furrows originated was detected
(see also Additional file 1: Figure S6). AP alkaline phosphatase, SP signal peptide, WT wild type, AP_1 AP_SP-PETaseR280A-GFP clone 1

and (ii) low, but similar amounts of TPA and MHET points and the content of the samples was analyzed for
were detected in the supernatant (medium fraction) of the presence of PETase reaction products. The results
AP_SP-PETaseR280A-FLAG clone 2 incubated with PET of these approaches revealed that degradation of PET/
bottle film (see Additional file  1: Figure S9 and Addi- PETG film substrate by PETase-FLAG was very low or
tional file  2: Table  S3). The latter result was supported even absent under most of the selected conditions (not
by structural differences (presence of holes) in the PET shown). However, when we used shredded PET—a mix-
bottle film when inspected via SEM (Additional file  1: ture of micro- and macro-plastics up to approximately
Figure S9). A rough estimation of the formed total prod- 1 cm in size (see Additional file 1: Figure S11)—positive
ucts (MHET + TPA) of both approaches (26.32 µm from results were obtained. In this approach we performed a
PETG vs. 0.31  µM from bottle PET) indicated that the time series experiment with an around 1 week old 50 ml
turnover of PETG by P ­ ETaseR280A-FLAG is approximately culture of AP_SP-PETaseR280A-FLAG producing clone
80-fold higher when compared to bottle PET (Additional 2 and a 50  ml wild type control culture, which were
file 2: Table S3) under the selected reaction conditions. incubated with an industrially shredded PET substrate
In a further approach 50 and/or 150  ml cultures were (approx. 5  g), respectively. 1  ml samples for UHPLC
incubated with 5 × 1.5 cm PET and PETG pieces from a analysis were taken after 0, 1, 2, 3, 6, 10, and 14  days
bottle or commercial film or shredded PET particles for ­(T0–T14) and filtered (cutoff 0.22  µm). At time points
up to 14  days with and without agitation at 21 or 26  °C ­T3, ­T6, and ­T10, the cultures were supplied with new
(see “Methods”). Samples were collected at different time nitrate to steadily induce expression of the recombinant
Moog et al. Microb Cell Fact (2019) 18:171 Page 7 of 15

Fig. 4  SEM and UHPLC analysis of PETG film degradation by PETase-FLAG secreted from P. tricornutum. A small piece of PETG film was incubated
with 1 ml of supernatant (medium fraction) of a 50 ml culture expressing AP_SP-PETaseR280A-FLAG (clone 1, induced for 4 days) and wild type and
analyzed via SEM. As shown in the upper part, similar but more area-wide changes in the surface of the PETG film as observed in the solid approach
(PET bottle film degradation by AP_SP-PETaseR280A-FLAG clone 1 on solid medium, Fig. 3) were detected. The wild type control (lower left) did not
show any significant aberrations in the surface structure of the PETG film. UHPLC analysis (lower right) of the medium fractions after 1 week of
incubation with the PETG film at 30 °C revealed production of TPA and MHET in sample AP_1 (AP_SP-PETaseR280A-FLAG clone 1), which were absent
from the wild type control. PETG polyethylene terephthalate glycol, SN supernatant, AP alkaline phosphatase, SP signal peptide, WT wild type, AP1/
AP_1 AP_SP-PETaseR280A-FLAG clone 1

protein AP_SP-PETaseR280A-FLAG by clone 2. Interest- A similar approach was taken for AP_SP-PETaseR280A-
ingly, we observed a progressive increase of TPA and FLAG producing clone 1 and shredded PET (approx. 10 g)
MHET with a higher amount of MHET generated until in 150 ml culture volume. 1 ml samples for UHPLC analy-
­T3. After ­T6 the amount of produced MHET progres- sis were collected after 0, 3, 6, 10, and 14  days (­T0–T14).
sively decreased while TPA still significantly increased At the same time, the cultures were supplied with nitrate
(Fig.  5, Additional file  2: Table  S3). The level of MHET (see above). As shown in Additional file 1: Figure S11, we
decreased to a minimum at T ­ 10, which did not perceiv- obtained very similar results to those shown in Fig. 5. At
ably change until ­T14, although the total concentration of ­T3 MHET was more abundant than TPA, whereas at T ­6
products (TPA + MHET) increased almost linearly until the situation reversed, the TPA level increased progres-
­T14 (Additional file  2: Table  S3). No production of TPA sively until T
­ 14 and MHET basically completely vanished
or MHET was detected in the wild type control (Fig. 5). at ­T10 (Additional file 1: Figure S11). The results of these
The identity of the formed products was confirmed by UHPLC experiments clearly show that TPA and MHET
mass spectrometry; estimated product quantifications (identity also confirmed by mass spectrometry) were effi-
in the micromolar range are shown in Additional file  2: ciently produced from shredded PET substrates when
Table S3. incubated with AP_SP-PETaseR280A-FLAG producing
Moog et al. Microb Cell Fact (2019) 18:171 Page 8 of 15

Fig. 5  UHPLC analysis of the medium fraction of shredded PET incubated with PETase-FLAG producing clone 2. The experiment was performed
in 50 ml f/2-medium containing AP_SP-PETaseR280A-FLAG expressing clone 2 (AP2) and approximately 5 g of shredded PET. At ­T0 the cultures were
adjusted to an ­OD600 of 0.4 and expression of the recombinant protein (AP_SP-PETaseR280A-FLAG) was induced with nitrate. Samples of 1 ml were
taken at individual time points (T, 1 day = 24 h) and fresh nitrate was supplemented to the cultures at ­T3, ­T6 and ­T10. The standards used for UHPLC
analyses are shown on the lower right. Note that the concentrations of standard compounds are not equal. BHET bis(2-hydroxyethyl) terephthalic
acid, MHET mono(2-hydroxyethyl) terephthalic acid, TPA terephthalic acid, WT wild type, AP2 AP_SP-PETaseR280A-GFP clone 2

clones. In contrast, the wild type culture of P. tricornutum such as PET is often expensive and can involve harmful
did not produce any PET degradation products in con- chemicals and a high demand of temperature, energy and
trol experiments. BHET, which besides MHET is another time [31]. Thus, an eco-friendly biological degradation of
potential product of the PETase reaction [13], was not synthetic PET polymers into reusable monomers (TPA
produced in any detectable amounts in neither of the cul- and EG) via bioremediation is desired. Bioremediation
tures during our experiments. is also one of the most promising solutions to counter-
act plastic pollution on Earth. Micro- and nano-plastics,
Discussion once present in the environment, because of their small
If correctly disposed plastic is not waste but instead a sizes are very difficult to remove from nature. Especially
valuable resource that can be part of a closed (circular) waterbodies are highly polluted with plastic waste, and
recycling process. The recycling of used plastic materials Earth’s oceans are accumulating more plastics year in and
Moog et al. Microb Cell Fact (2019) 18:171 Page 9 of 15

year out. Many plastic materials are highly durable and in P.  tricornutum [33–35]. We exemplarily tested the
resistant to biological degradation through e.g., micro- post-translational modification of AP_SP-PETaseR280A-
organisms. The recently discovered PETase is a bacterial FLAG via N-glycosylation by PNGase F (cleaves N-linked
hydrolase possessing the ability for PET degradation [13]. oligosaccharides) treatment and observed that the mass
Recent studies on PETase demonstrate the potential of increase of the specific protein band was indeed lesser
the enzyme for possible applications in biological PET when potential N-linked glycans were removed enzy-
degradation and recycling (see e.g., [13–18, 30, 32]). In matically (Additional file 1: Figure S3). This clearly indi-
this work, we expand the possible range of applications cates that at least a part of the observed mass increase
for PETase via generating a synthetic microbial cell fac- of the recombinant protein was due to N-glycosylation
tory in the form of a marine microalga capable of produc- taking place in the diatom cell. This observation also cor-
ing and secreting functional PETase that can efficiently relates with the presence of predicted N-glycosylation
decompose PET in a saltwater-based environment. sites in the AP_SP-PETaseR280A-FLAG sequence (see
By means of expression and localization studies of Additional file 1: Figure S12) and the results of the mass
GFP fusion proteins as well as Western Blot analyses, spectrometric analyses in which certain peptides (result-
our studies show that the marine diatom P. tricornutum ing from tryptic digestion), including predicted N-gly-
is a suitable chassis for the production of recombinant cosylation sites, likely due to deviating masses could not
­PETaseR280A-FLAG and its secretion into the medium be identified (Additional file 2: Tables S1 and S2). How-
fraction (Figs. 1, 2, Additional file 1: Figure S1). The lat- ever, to explain the huge increase in molecular mass of
ter is possible either by expressing PETase with the origi- ­PETaseR280A-FLAG, further post-translational modifica-
nal bacterial SP of I. sakaiensis PETase, or as a synthetic tions, such as, for example, O-glycosylation of the protein
construct in which the SP of a well-known secreted fac- by P. tricornutum, have to be considered (and postulated)
tor of the diatom, alkaline phosphatase (AP) [29], was as well. Whether the post-translational modifications of
used. Western Blot analyses of several PETase-GFP and ­PETaseR280A-FLAG in P.  tricornutum have an influence
-FLAG producing P.  tricornutum clones revealed that on its catalytic activity or substrate specificity is so far
the constructs with a FLAG-tag were secreted much unknown and will be (in parallel to their identity) investi-
more efficiently than those containing a GFP-tag (Fig.  2 gated in more detail in future studies.
and Additional file 1: Figure S1), which is not surprising In any case, the results of our studies indicate that
and might be due to the massive increase in theoretical ­PETaseR280A-FLAG—an engineered version of I. sakaien-
molecular mass (GFP: + 27.5  kDa; FLAG: + 1  kDa) and sis PETase with increased activity [14]—produced and
size (bulky barrel-like structure of GFP) of the expressed secreted by P.  tricornutum possesses catalytic activity
GFP fusion protein probably hindering secretion by the against PET as well as PETG copolymer. This could be
diatom cell. With AP_SP-PETaseR280A-FLAG clone 2 we shown by several PET/PETG degradation experiments
obtained a P. tricornutum mutant cell line capable of the with diatom cells growing on solid f/2-saltwater agar as
highest production and secretion rates of the recombi- well as in liquid cell culture (Fig.  3, 4, 5 and Additional
nant enzyme upon induction of expression when com- files 1, 2). The proof of principle of enzyme functionality
pared to other generated clones (Fig. 2). was provided in the course of a PET bottle film decom-
As evident from several Western Blots, a mass shift position approach on solid agar (Fig. 3) in which partial
of the diatom-produced recombinant PETase protein degradation of the substrate was achieved after several
was observed during our experiments (Fig.  2 and Addi- weeks of incubation. The observed structural changes
tional file 1: Figure S1). Whereas the original I. sakaien- (holes and cavities) in the PET surface were comparable
sis PETase protein has a molecular mass of 30.1 kDa, the to those reported in earlier PET degradation experiments
theoretical values for the here synthesized recombinant using PETase [13, 15], strongly indicating that the dia-
versions are 57.7 for ­PETaseR280A-GFP and 30.4  kDa for tom produced ­PETaseR280A-FLAG enzyme was capable of
AP_SP-PETaseR280A-FLAG. However, when immunode- decomposing bottle PET.
tection with an antibody specific for the FLAG-tag was As we had generated several promising diatom mutants
performed signals with higher molecular masses in the efficiently producing and secreting ­ PETaseR280A-FLAG
range of at least 10–25 kDa more could be detected. Via into the medium (Fig.  2), we aimed at scaling up the
mass spectrometry, these bands could be identified as approach and reducing reaction time and thus changed
­PETaseR280A-FLAG unequivocally (Additional file  1: Fig- to a liquid saltwater environment (cell culture) for fur-
ure S2, Additional file  2: Tables S1 and S2). As already ther studies. Initially, PET as well as PETG decomposi-
shown for endogenous as well as secreted recombinantly tion was tested at 30 °C—temperature conditions nearly
expressed proteins, typical eukaryotic post-translational optimal for PETase [13, 30], but not for P.  tricornutum
modifications in the form of N-glycosylation can occur cell growth [36]. For this, we used only 1 ml of a diatom
Moog et al. Microb Cell Fact (2019) 18:171 Page 10 of 15

culture synthesizing P ­ ETaseR280A-FLAG and incubated thus suggest that so far only shredded PET can be effi-
the fraction with commercial PETG or bottle PET film. ciently degraded by diatom produced PETase-FLAG
While the approach with commercial highly amorphous in cultures with living cells, while degradation of PET/
PETG copolymer film proved to be highly successful PETG film might require conditions more tailored to
(Fig.  4, Additional file  1: Figures  S7, S8), only little PET the enzyme’s needs than to those of the algal cell cul-
degradation could be observed for bottle PET incubated ture. However, with respect to the overall potential of
with ­PETaseR280A-FLAG containing f/2-medium (Addi- the diatom produced PETase for degradation of differ-
tional file  1: Figure S9, Additional file  2: Table  S3). A ent PET substrates in culture, especially factors such as
rough estimation of the formed products of the degrada- the growth rate of P. tricornutum in connection to the
tion reaction catalyzed by P ­ ETaseR280A-FLAG (Additional PETase expression level as well as enzyme activity and
file 2: Table S3) under the selected conditions suggested stability (half-life) will have to be analyzed in more detail
an approximately 80-fold higher turnover of PETG than and adjusted in future studies. Based on the stability of
bottle PET. These results indicate that the diatom-pro- PETase-FLAG produced by the algae (accumulation in
duced ­PETaseR280A-FLAG possesses different specificities the medium fraction and stable for at least 7 days with-
for different PET substrates, which, among other things, out appearance of recognizable protein degradation
might be due to the degree of crystallinity of the different products; see Additional file 1: Figure S4), one possibility
polymer materials that has an influence on their biodeg- would be to use only the saltwater supernatant contain-
radability (see e.g., [37]). However, the results show that ing the enzyme that can be heated up to the temperature
not only PET but also PETG copolymer can be degraded optimum of PETase for a more efficient biodegradation
by the ­ PETaseR280A-FLAG produced and secreted by of PET/PETG film (see Fig. 4).
P. tricornutum. As shown by the PET degradation experiments using
When we scaled up our liquid PET/PETG degrada- industrially shredded material as a substrate (Fig.  5),
tion experiments to diatom culture size (50–500 ml), we MHET was produced first as a major component of PET
again recognized differences in PET turnover depend- degradation by PETase-FLAG accompanied by a lower
ing on the form of the substrate. Although the liquid increase of TPA until ­T3 (Fig.  5). At ­T6 this situation
approach at P. tricornutum culturing conditions (see changed in that the amount of TPA present in the sample
“Methods”) using industrially shredded PET as substrate abruptly passed the MHET level by an increase more than
was highly successful (see Fig.  5 and Additional file  1: threefold (see also Additional file 2: Table S3). From ­T6 to
Figure S11), PET and PETG film (bottle and commercial) ­T10 MHET then decreased to a minimum level that did
substrates were not or rather poorly degraded by the not change significantly in the last measurement ­(T14).
diatom produced and secreted ­PETaseR280A-FLAG. This A very similar observation was made for an independent
observation is most likely due to several circumstances experiment using a different clone (AP_SP-PETaseR280A-
causing suboptimal reaction conditions for the enzyme. FLAG clone 1) in which MHET basically disappeared
First, the majority of experiments were performed at at ­T10, while the TPA level was still slightly increasing
21  °C, the optimal growth temperature of P. tricornu- (Additional file 1: Figure S11). This observation might be
tum, or 26  °C, which is an almost critical temperature explained by two potential scenarios: in the first one the
point for diatom growth/survival. Both temperatures are here synthesized ­ PETaseR280A-FLAG functions similar
suboptimal for PETase activity, which has its tempera- to PETase reported in earlier studies producing MHET
ture optimum near 35  °C [13, 30]. Second, agitation of and TPA as main products, not being able to further
the culture, which is necessary for optimal growth of P. decompose generated MHET [14]. The almost complete
tricornutum, might also have impaired PETase effectiv- absence of MHET from T ­ 10 might be due to PETase-
ity, especially to attach to the rather smooth surface of independent turnover or removal of the substance by a
PET/PETG film substrates. Third, the composition of so far unknown factor. In putative scenario two, the engi-
the f/2-medium in which the diatoms grow might have neered version of I. sakaiensis PETase ­(PETaseR280A [14]),
influenced PETase function, although it was shown ear- produced in this experiment might be able not only to
lier that salt can increase activity of the enzyme [30]. degrade PET into TPA under the chosen conditions, but
Another factor might be the general nature of the sub- also further decompose MHET into TPA. However, this
strate. Whereas shredded PET possesses an enhanced, is in contrast to previous observations that suggested that
rough surface providing ample contact sites and starting the original I. sakaiensis PETase is not able to decompose
points for PETase activity, larger pieces of PET film with MHET [13, 14]. In addition, while having maintained
a smooth surface (and a potentially higher polymer crys- constant induction of P ­ ETaseR280A-FLAG expression by
tallinity) are most likely much harder to be efficiently supplementation of the culture with fresh nitrate at sev-
attacked by PETase-FLAG (see above). Our findings eral time points, the less increasing (Fig.  5, Additional
Moog et al. Microb Cell Fact (2019) 18:171 Page 11 of 15

file 2: Table S3) or rather stagnant level (Additional file 1: Methods

Figure S11) of TPA from T ­ 10 together with the absence of Plasmid construction, transfection of diatoms and confocal
MHET suggest that the PET degradation gradually comes microscopy
to a halt. The reason for this might be the decay of the The gene encoding an engineered version of I. sakaien-
culture caused by a lack of nutrients, such as phosphate, sis PETase (IsPETaseR280A) [14] was synthesized by Syn-
vitamins and trace elements that have been used up by bio Technologies (USA) according to the codon usage of
the growing diatoms. As a consequence, PETase produc- P. tricornutum with a 5′ EcoRI and a 3′ BamHI restric-
tion is shut down, which might explain the reduced TPA tion site and cloned together with egfp (BamHI/HindIII)
increase/static TPA level and the continuing absence into the nitrate inducible pPha-NR shuttle vector (NCBI
of MHET. Moreover, as mentioned earlier, it cannot be accession number: JN180663). To generate a version of
excluded that potential post-translational modifica- PETase efficiently secreted by the diatom, the petase gene
tions of ­PETaseR280A-FLAG (N-glycosylation, etc.) might was modified at the 5′- and 3′-ends. The gene sequence
have altered the enzymatic activity or substrate specific- encoding the original bacterial signal peptide was
ity of the enzyme enabling the further decomposition of replaced by a gene region encoding the signal peptide
MHET into TPA. A potentially extended substrate range of P.  tricornutum alkaline phosphatase (Phatr2: 49678,
of the diatom-produced PETase will be investigated in first 30  aa), a protein that is secreted into the extracel-
detail in future experiments. lular environment by the diatom [29]. At the 3′-end, gfp
was replaced by a FLAG-tag (DYKDDDDK) encoding
Conclusions sequence, followed by a stop codon and a HindIII restric-
Taken together, our studies provide the proof of princi- tion site. Both modifications were induced via oligonu-
ple that P­ ETaseR280A-FLAG produced by a marine dia- cleotide primer sequences synthesized by Sigma-Aldrich
tom is functional for PET degradation under (mesophilic specific to petase that were extended for the particular
marine) growth conditions of the model system. In detail gene region (5′: GAA​TTC​ATG​AAA​TTC​TCT​ACT​GCC​
we showed efficient production and secretion of the GTT​GTA​TCA​C TC​ATA​ACC​GTC​G CA​C CA​C TG​GTC​
recombinant proteins, enzyme functionality with respect GTC​G GC​G CC​C AA ​ACT​A AT​C CT ​TAC​G CT​C GCGG;
to different PET substrates (PET and PETG film and 3′: AAG​CTT​ATT​TAT​CAT​CAT​CGT​CTT​TGT​AAT​CGG​
shredded PET) under varying conditions highlighting the
­ 5® High-Fidelity 2X Master Mix (NEB).
AGC​AAT​TAG​CGG​TAC​GGA​AAT​CG) in PCR reac-
enormous potential for further experiments and applica- tions using the Q
tions with respect to biological PET degradation. These Generated artificial gene sequences were cloned into the
include the design and development of photobioreac- pJet1.2/blunt plasmid using the CloneJET PCR Cloning
tors for PET bioremediation as well as the development Kit (Thermo Fisher Scientific) and validated via sequenc-
of efficient closed- or open-loop recycling strategies for ing (Macrogen), before they were cloned into pPha-NR
TPA (and EG) to synthetize new PET from its own deg-
coli culture was then isolated via the ­NucleoBond® Xtra
(see above). High purity plasmid of a 50  ml Escherichia
radation products up to the point of further metabolic
engineering of the microalgal metabolism in order to Midi/Maxi kit (Macherey–Nagel) and transformation of
generate cells capable of completely metabolizing PET P. tricornutum wild type cells was performed as described
and use it as a carbon source. Moreover, the results of our previously [38]. To analyze the expression and locali-
studies might be instrumental and pave the way for self- zation of GFP fusion proteins, confocal laser scanning
contained applications helping to decrease the currently microscopy (CLSM) was performed. For this, expression
enormous plastics pollution on our planet, especially of the GFP fusion proteins was induced by incubation
with respect to saltwater environments—i.e., the oce- of the transformed cells for 24  h in liquid f/2-medium
anic ecosystems—(closed bioremediation systems). An containing 0.9 mM ­NaNO3 as a nitrogen source (activa-
important benefit of the model system used in this study tion of the nitrate reductase promoter). The cells were
for PETase production is that the diatom is a marine, analyzed using a Leica TCS SP2 confocal laser scanning
photosynthetic organism. Its habitat is marine water—the microscope equipped with a HCX PL APO 40×/1.25–
place where the majority of non-recycled plastic waste 0.75 Oil CS objective. GFP fluorescence was excited at
(macro- and micro-plastic) finally ends up on our planet. 488  nm by a 65  mW Argon laser, while excitation was
Thus, the here generated microbial cell factory might not detected between 500 and 520 nm. In parallel, the auto-
only be useful for climate friendly PET recycling, but also fluorescence of the plastid was excited at 488 nm as well
an application in delimited sewage plant-like bioreactor and its emission was detected between 625 and 720 nm.
systems for oceanic microplastic decomposition might be Obtained images were processed digitally with Leica LAS
conceivable. AF lite and ImageJ [39].
Moog et al. Microb Cell Fact (2019) 18:171 Page 12 of 15

Cell cultivation and PETase secretion analysis via Western end, the supernatant (medium fraction) of a 500 ml cul-
Blot ture expressing the recombinant protein was filtered
Phaeodactylum tricornutum (strain Bohlin, UTEX646) and concentrated (see above) to a volume of 250  µl in
cells were grown either in liquid or on solid saltwater f/2- 50  mM sodium phosphate buffer (pH 7.5). 18  µl of the
medium in constant light (24 h at 8000–10,000 Lux). The concentrated protein fraction was incubated with 1 µl of
f/2-medium was consisting of the following ingredients: PNGase F enzyme under denaturating reaction condi-
1.66% (w/v) Tropic Marin (Dr. Biener GmbH) salt, 2 mM tions according to the manufacturer’s protocol. As a neg-
Tris/HCl (pH 8.0), 36 µM N ­ aH2PO4, either 1.5 mM N ­ H4Cl ative control, an identical amount of recombinant protein
or 0.9 mM ­NaNO3 was used as a nitrogen source, as well was subjected to the same treatment replacing PNGase
as vitamins and trace elements (see [40]). The pH of liq- F by reaction buffer. Final analysis of the protein fraction
uid f/2-medium was adjusted to 8.0. For solid f/2-medium, was performed via Western Blot (see above).
1.3% (w/v) agar agar (Roth) was added. Selection and cul-
tivation of transformed clones was conducted by using
Zeocin (InvivoGen) of a final concentration of 75  µg/ Plastic degradation experiments
ml as a supplement to the medium. Liquid cultures were PET degradation was analyzed in a solid and a liquid
grown under constant agitation (120–150 rpm). To analyse approach. For the solid approach, PET from a conven-
the secretion of recombinant proteins into the medium, tional bottle (Fiji water, 500 ml, specific polymer prop-
50 ml P. tricornutum culture was grown for approximately erties unknown) was cut into small pieces less than
1  week in f/2-medium containing 1.5  mM N ­ H4Cl and 1 cm2 in area. P. tricornutum clones expressing AP_SP-
adjusted to an optical density of 0.3 to 0.4 (­OD600). Induc- PETaseR280A-FLAG were grown on f/2 agar plates con-
tion of protein expression was achieved by harvesting the taining nitrate (­NO3−, inducing conditions) in direct
cells (1500×g, 5  min., 21  °C) followed by a medium shift contact to PET, which was sticking upright in the solid
using f/2 containing 0.9  mM ­NaNO3 and further culti- medium. The plate was overflowed with 2  ml nitrate-
vation for 24  h to 7  days. The cells were pelletized again containing liquid f/2-medium to generate an aqueous
as before, washed with PBS or f/2-medium once and the environment for enzymatic catalysis and incubated
supernatant (medium fraction) was filtrated (0.22 µm pore for 2 to 6  weeks under continuous light at 21  °C (see
above). A P.  tricornutum wild type culture grown in
with centrifugal filter units ­(Amicon® Ultra-15) withhold-
size). Subsequently the medium fraction was concentrated
direct contact to PET under similar conditions served
ing proteins larger than 10 kDa (cutoff). Samples were con- as a negative control. The experimented was finished
centrated to a volume less than 1000 µl and proteins within by removing the PET fragments from the plate and fol-
the concentrated medium fraction were precipitated with lowed by further processing of the material for SEM
10% (v/v) trichloracetic acid (TCA). After two to three analysis (see below).
wash steps using 80% (v/v) ice-cold acetone proteins were In the liquid approach, P.  tricornutum clones trans-
dried under vacuum and dissolved in 15  µl SDS loading formed with the gene encoding AP_SP-PETaseR280A-
dye containing 8 M urea and 1% (v/v) β-mercaptoethanol. FLAG were grown in conical flasks containing 50, 150
From the pellet fraction a total protein extract was or 500  ml f/2-medium with N ­ H4+ (non-inducing con-
generated by resuspending the cells with 200  µl alka- ditions) under agitation for approximately 1  week. As
line lysis buffer containing 1.7  M NaOH, 7.5% (v/v) a control, a wild type culture was used. Both cultures
β-mercaptoethanol and protease inhibitor cocktail. After were adjusted to the same optical density ­ (OD600
incubation for 30 min on ice, the proteins were precipi- between 0.3 and 0.6) and cultivated further in f/2-
tated as described above and dissolved in 100 µl of SDS medium containing N ­O3− (inducing conditions)
loading dye. 10  µg protein of the pellet fraction and the together with 5 to 10 g of shredded PET (specific poly-
total volume of the medium fraction (15  µl) were sepa- mer properties unknown), which was kindly provided
rated via SDS-PAGE and transferred to a nitrocellulose by ALPLA-Werke Lehner GmbH & Co KG (Gemünden,
membrane using Western Blot. For immunodetection Germany) at no cost. The shredded PET substrate was
antibodies against GFP (Rockland, 1:3000), α-tubulin consisting of small pieces mostly less than 1 cm2 in size
(Sigma-Aldrich, 1:2000) and FLAG-tag (DYKDDDDK, with a considerable fraction of microplastic (≤ 5  mm)
Rockland, 1:2000) were used. and sterilized by a wash with 70% Ethanol for ≥ 12  h.
The ethanol was removed by drying the PET under vac-
uum for at least 5 h. During incubation of the PET with
PNGase F assay the AP_SP-PETaseR280A-FLAG expressing cells and the
N-glycosylation of expressed recombinant proteins wild type control, samples of 1  ml volume were taken
was analyzed using PNGase F (NEB) treatment. To this from the supernatant at different time points at which
Moog et al. Microb Cell Fact (2019) 18:171 Page 13 of 15

also fresh nitrate was added to the culture. The sam- (Agilent) equipped with an electrospray ionization source
ples were filtrated (0.22 µm pore size), frozen in liquid set to negative ionization mode. The analytes were sepa-
nitrogen and stored at − 80 °C before they were further rated on a RP-18 column (50 mm × 2.1 mm, particle size
analyzed via ultra high performance liquid chromatog- 1.7 µm, Kinetex EVO C18, Phenomenex) using a mobile
raphy (UHPLC, see below). phase system comprised of 0.1% formic acid in water
In addition, 1 ml supernatant of P. tricornutum clones (A) and acetonitrile (B) with the following gradient con-
expressing AP_SP-PETaseR280A-FLAG or wild type was dition at 40  °C and a flow rate of 250  µl/min: 0  min 5%
incubated for 1 to 2  weeks at 30  °C with small pieces B; 1  min 5% B, 6  min 95% B; 6.5  min 95% B; 7  min 5%
(less than 1  cm2 in size) of PET-based copolyester from
a commercial film ­(LUMEX®A, PETG SPECTAR​® Resin,
B. For the MS, capillary voltage was set at 3.5  kV and
nitrogen gas was used as nebulizing (20 psig), drying
Polycasa), a highly amorphous material consisting of (13  l/min, 225  °C) and sheath gas (12  l/min, 40  °C). MS
polyethylene terephthalate glycol (PETG), with a thick- data were acquired with a scan range of 50–1100  m/z.
ness of 0.8  mm (specific physical properties unknown), LC–MS data were analyzed using MassHunter Qualita-
or bottle PET film (see above). The PET fragment was tive Analysis software (Agilent) using the following exact
removed and prepared for SEM (see below), whereas the mass traces: TPA [M−H]− = 165.0193  m/z and MHET
liquid fraction was frozen in liquid nitrogen and stored at [M−H]− = 209.0455 m/z.
− 80 °C prior UHPLC analysis (see below).
Supplementary information
Supplementary information accompanies this paper at https​://doi.
Scanning electron microscopy of PETase treated PET org/10.1186/s1293​4-019-1220-z.

samples
PET substrates (disks and square-cut parts) were treated Additional file 1: Figure S1. Secretion analysis of PETase-GFP in P.
tricornutum cultures. Figure S2. Coomassie-staining and mass spectrom‑
with 1% SDS for at least 12 h and dried for ≥ 5 h under etry analysis of secreted PETase-FLAG in P. tricornutum cultures. Figure
vacuum. The PET samples were than attached to sam- S3. Western Blot of a PNGase F treated protein sample (18 µl) of the total
ple stubs (15 × 6  mm, Plano GmbH) and sputtered with precipitated medium fraction. Figure S4. Expression and secretion effi‑
ciency analysis of AP_SP-PETaseR280A-FLAG using Western Blot. Figure S5.
gold under vacuum using a sputter coater (Balzers Union, Scanning electron microscopic analysis of PET bottle film degradation by
Lichtenstein). The samples were analyzed using a Hitachi AP_SP-PETase-FLAG clone 1 (AP_1) secreted from P. tricornutum on a f/2
S-530 scanning electron microscope. Image processing agar plate for 5 weeks. Figure S6. Scanning electron microscopic image
of a P. tricornutum clone AP_SP-PETase-FLAG_1 (AP_1) cell imprint on PET
was performed with ImageJ/Fiji [39]. bottle film incubated on a f/2 agar plate for 5 weeks. Figure S7. Scanning
electron microscopic analysis of amorphous PETG film degradation by
Ultra high performance liquid chromatography (UHPLC) PETase-FLAG tag secreted from P. tricornutum. Figure S8. Scanning elec‑
tron microscopy and UHPLC analysis of amorphous PETG film treated with
of PETase treated PET samples 1 ml supernatant of a 500 ml culture of a P. tricornutum clone expressing
PET degradation products in the growth medium were AP_SP-PETase-FLAG (clone 2). Figure S9. Scanning electron microscopy
analyzed by UHPLC using an Infinity II system (Agi- and UHPLC analysis of PET (bottle) film treated with 1 ml supernatant of
a 500 ml culture of a P. tricornutum clone expressing AP_SP-PETase-FLAG
lent Technologies, CA, USA). Samples were either con- (clone 2). Figure S10. UHPLC with 1 ml supernatant of a 500 ml culture
centrated between six and eightfold by speedvac or ca. of a P. tricornutum clone expressing AP_SP-PETase-FLAG_2 and standard
15-fold by lyophilization and subsequent redissolvation measurements. Figure S11. PET degradation experiment (UHPLC) using
shredded PET as a substrate and clone AP_SP-PETase-FLAG_1. Figure S12.
in water before they were applied to UHPLC. Standard Predicted N-glycosylation pattern for AP_SP-PETase-FLAG by NetNGlyc 1.0.
compounds and PET degradation products (i.e., TPA and Additional file 2: Table S1. Mass spectrometry analysis results for
MHET) were monitored at 240  nm employing a diode the < 55 kDa band (see Additional file 1: Figure S2). Table S2. Mass spec‑
array detector. TPA and MHET were separated on a trometry analysis results for the > 40 kDa band (see Additional file 1: Figure
S2). Table S3. Quantification of TPA and MHET production.
reverse phase C18 column (Sonoma, 3 µm C18(2) 100 Å,
10 cm × 2.1 mm; ES industries, NJ, USA) using a mobile
phase system comprised of 50 mM phosphoric acid and Abbreviations
PET: polyethylene terephthalate; PETG: polyethylene terephthalate glycol;
methanol. Chromatographic separation was carried out TPA: terephthalic acid; EG: ethylene glycol; MHET: mono(2-hydroxyethyl) tere‑
using the following gradient conditions at 50  °C and a phthalic acid; BHET: bis(2-hydroxyethyl) terephthalic acid; SP: signal peptide;
flow rate of 250 µl/min: 0 min 20% methanol; 2 min 20% AP: alkaline phosphatase; CLSM: confocal laser scanning microscopy; UHPLC:
ultra high performance liquid chromatography; MS: mass spectrometry; SEM:
methanol, 12 min 40% methanol. Approximations of TPA scanning electron microscopy.
and MHET concentrations were calculated using the
detected peak sizes/areas. Individual product peaks were Acknowledgements
We would like to thank Thomas Heimerl, Fransziska Hempel and Lars-Oliver
collected and applied to UHPLC–MS (mass spectrom- Essen for helpful discussions and valuable technical advices supporting the
etry) to verify their identities. UHPLC–MS analysis was progress of the project. Moreover, we are grateful to ALPLA-Werke Lehner
carried out using a 6550 iFunnel Q-TOF LC–MS system GmbH & Co KG (Gemünden, Germany) for providing industrially shredded PET,
Moog et al. Microb Cell Fact (2019) 18:171 Page 14 of 15

which was used as a main substrate for plastic degradation experiments in plastic-degrading Ideonella sakaiensis MHETase bound to a substrate.
this work. We thank Dr. Niña Cortina for help with the LC-MS analysis. Nat Commun. 2019;10:1717.
10. Ioakeimidis C, Fotopoulou KN, Karapanagioti HK, Geraga M, Zeri C,
Authors’ contributions Papathanassiou E, Galgani F, Papatheodorou G. The degradation
DM, UGM and TE designed the study. DM and JoS performed the transforma‑ potential of PET bottles in the marine environment: an ATR-FTIR based
tion, localization, Western Blot and plastic degradation experiments and DM, approach. Sci Rep. 2016;6:23501.
JoS and UGM interpreted the data. JaS and JZ conducted the UHPLC analyses 11. Webb HK, Arnott J, Crawford RJ, Ivanova EP. Plastic degradation and
and JaS, JZ, TE and DM interpreted the results. KHR, DM and JoS performed its environmental implications with special reference to poly(ethylene
the scanning electron microscopy experiments and interpreted the obtained terephthalate). Polymers. 2013;5:1–18.
observations. Mass spectrometry analyses were carried out by UL; UL and DM 12. Ragaert K, Delva L, Van Geem K. Mechanical and chemical recycling of
interpreted the data. DM wrote the paper. All authors read and approved the solid plastic waste. Waste Manag. 2017;69:24–58.
final manuscript. 13. Yoshida S, Hiraga K, Takehana T, Taniguchi I, Yamaji H, Maeda Y, Toyo‑
hara K, Miyamoto K, Kimura Y, Oda K. A bacterium that degrades and
Funding assimilates poly(ethylene terephthalate). Science. 2016;351:1196–9.
This work was supported by the Philipps University Marburg and the SYN‑ 14. Joo S, Cho IJ, Seo H, Son HF, Sagong HY, Shin TJ, Choi SY, Lee SY, Kim KJ.
MIKRO research center. JZ was supported by FET-Open Grant 686330 (“Future Structural insight into molecular mechanism of poly(ethylene tereph‑
Agriculture”). thalate) degradation. Nat Commun. 2018;9:382.
15. Austin HP, Allen MD, Donohoe BS, Rorrer NA, Kearns FL, Silveira RL, Pol‑
Availability of data and materials lard BC, Dominick G, Duman R, El Omari K, et al. Characterization and
All data generated or analyzed during this study are included in this published engineering of a plastic-degrading aromatic polyesterase. Proc Natl
article and its additional information files. Acad Sci USA. 2018;115:E4350–7.
16. Han X, Liu W, Huang JW, Ma J, Zheng Y, Ko TP, Xu L, Cheng YS, Chen CC,
Ethics approval and consent to participate Guo RT. Structural insight into catalytic mechanism of PET hydrolase.
Not applicable. Nat Commun. 2017;8:2106.
17. Huang X, Cao L, Qin Z, Li S, Kong W, Liu Y. Tat-independent secretion
Consent for publication of polyethylene terephthalate hydrolase PETase in Bacillus subti-
Not applicable. lis 168 mediated by its native signal peptide. J Agric Food Chem.
2018;66:13217–27.
Competing interests 18. Seo H, Kim S, Son HF, Sagong HY, Joo S, Kim KJ. Production of extracel‑
The authors declare that they have no competing interests. lular PETase from Ideonella sakaiensis using sec-dependent signal
peptides in E. coli. Biochem Biophys Res Commun. 2019;508:250–5.
Author details 19. Tanasupawat S, Takehana T, Yoshida S, Hiraga K, Oda K. Ideonella
1
 Laboratory for Cell Biology, Philipps University Marburg, Karl‑von‑Frisch‑Str. 8, sakaiensis sp. nov., isolated from a microbial consortium that degrades
35032 Marburg, Germany. 2 SYNMIKRO Research Center, Hans‑Meerwein‑Str. poly(ethylene terephthalate). Int J Syst Evol Microbiol. 2016;66:2813–8.
6, 35032 Marburg, Germany. 3 Max Planck Institute for Terrestrial Microbiology, 20. Daboussi F, Leduc S, Marechal A, Dubois G, Guyot V, Perez-Michaut C,
Karl‑von‑Frisch‑Str. 10, 35043 Marburg, Germany. 4 Department for Mycology, Amato A, Falciatore A, Juillerat A, Beurdeley M, et al. Genome engineer‑
Philipps University Marburg, Karl‑von‑Frisch‑Str. 8, 35032 Marburg, Germany. ing empowers the diatom Phaeodactylum tricornutum for biotechnol‑
5
 Gerätezentrum für Massenspektrometrie und Elementanalytik, Philipps ogy. Nat Commun. 2014;5:3831.
University Marburg, Hans‑Meerwein‑Straße 4, 35032 Marburg, Germany. 21. Nymark M, Sharma AK, Sparstad T, Bones AM, Winge P. A CRISPR/
Cas9 system adapted for gene editing in marine algae. Sci Rep.
Received: 28 June 2019 Accepted: 27 September 2019 2016;6:24951.
22. Hempel F, Maier UG. Microalgae as solar-powered protein factories.
Adv Exp Med Biol. 2016;896:241–62.
23. Hempel F, Lau J, Klingl A, Maier UG. Algae as protein factories: expres‑
sion of a human antibody and the respective antigen in the diatom
References Phaeodactylum tricornutum. PLoS ONE. 2011;6:e28424.
1. Geyer R, Jambeck JR, Law KL. Production, use, and fate of all plastics 24. Hempel F, Maier UG. An engineered diatom acting like a plasma cell
ever made. Sci Adv. 2017;3:e1700782. secreting human IgG antibodies with high efficiency. Microb Cell Fact.
2. Eriksen M, Lebreton LC, Carson HS, Thiel M, Moore CJ, Borerro JC, Gal‑ 2012;11:126.
gani F, Ryan PG, Reisser J. Plastic pollution in the World’s Oceans: more 25. Hempel F, Maurer M, Brockmann B, Mayer C, Biedenkopf N, Kelterbaum
than 5 trillion plastic pieces weighing over 250,000 tons afloat at Sea. A, Becker S, Maier UG. From hybridomas to a robust microalgal-based
PLoS ONE. 2014;9:e111913. production platform: molecular design of a diatom secreting mono‑
3. Barnes DK, Galgani F, Thompson RC, Barlaz M. Accumulation and frag‑ clonal antibodies directed against the Marburg virus nucleoprotein.
mentation of plastic debris in global environments. Philos Trans R Soc Microb Cell Fact. 2017;16:131.
Lond B Biol Sci. 2009;364:1985–98. 26. Hempel F, Bozarth AS, Lindenkamp N, Klingl A, Zauner S, Linne U, Stein‑
4. Rios LM, Moore C, Jones PR. Persistent organic pollutants carried buchel A, Maier UG. Microalgae as bioreactors for bioplastic produc‑
by synthetic polymers in the ocean environment. Mar Pollut Bull. tion. Microb Cell Fact. 2011;10:81.
2007;54:1230–7. 27. Slattery SS, Diamond A, Wang H, Therrien JA, Lant JT, Jazey T, Lee K,
5. Mato Y, Isobe T, Takada H, Kanehiro H, Ohtake C, Kaminuma T. Plastic Klassen Z, Desgagne-Penix I, Karas BJ, Edgell DR. An expanded plas‑
resin pellets as a transport medium for toxic chemicals in the marine mid-based genetic toolbox enables Cas9 genome editing and stable
environment. Environ Sci Technol. 2001;35:318–24. maintenance of synthetic pathways in Phaeodactylum tricornutum. ACS
6. Chae Y, Kim D, Kim SW, An YJ. Trophic transfer and individual impact Synth Biol. 2018. https​://doi.org/10.1021/acssy​nbio.7b001​91.
of nano-sized polystyrene in a four-species freshwater food chain. Sci 28. D’Adamo S, Schiano di Visconte G, Lowe G, Szaub-Newton J, Beacham
Rep. 2018;8:284. T, Landels A, Allen MJ, Spicer A, Matthijs M. Engineering the unicel‑
7. Law KL. Plastics in the Marine environment. Ann Rev Mar Sci. lular alga Phaeodactylum tricornutum for high-value plant triterpenoid
2017;9:205–29. production. Plant Biotechnol J. 2019;17:75–87.
8. Wei R, Zimmermann W. Biocatalysis as a green route for recycling the 29. Lin HY, Shih CY, Liu HC, Chang J, Chen YL, Chen YR, Lin HT, Chang YY,
recalcitrant plastic polyethylene terephthalate. Microb Biotechnol. Hsu CH, Lin HJ. Identification and characterization of an extracellular
2017;10:1302–7. alkaline phosphatase in the marine diatom Phaeodactylum tricornutum.
9. Palm GJ, Reisky L, Bottcher D, Muller H, Michels EAP, Walczak MC, Mar Biotechnol (NY). 2013;15:425–36.
Berndt L, Weiss MS, Bornscheuer UT, Weber G. Structure of the
Moog et al. Microb Cell Fact (2019) 18:171 Page 15 of 15

30. Liu C, Shi C, Zhu S, Wei R, Yin CC. Structural and functional characteriza‑ 36. Bojko M, Brzostowska K, Kuczynska P, Latowski D, Olchawa-Pajor M,
tion of polyethylene terephthalate hydrolase from Ideonella sakaiensis. Krzeszowiec W, Waloszek A, Strzalka K. Temperature effect on growth,
Biochem Biophys Res Commun. 2019;508:289–94. and selected parameters of Phaeodactylum tricornutum in batch cul‑
31. Salvador M, Abdulmutalib U, Gonzalez J, Kim J, Smith AA, Faulon JL, tures. Acta Biochim Pol. 2013;60:861–4.
Wei R, Zimmermann W, Jimenez JI. Microbial genes for a circular and 37. Wei R, Breite D, Song C, Gräsing D, Ploss T, Hille P, Schwerdtfeger R, Mat‑
sustainable Bio-PET economy. Genes (Basel). 2019;10:373. ysik J, Schulze A, Zimmermann W. Biocatalytic degradation efficiency of
32. Fecker T, Galaz-Davison P, Engelberger F, Narui Y, Sotomayor M, Parra LP, postconsumer polyethylene terephthalate packaging determined by
Ramirez-Sarmiento CA. Active site flexibility as a hallmark for efficient their polymer microstructures. Adv Sci. 2019;6:1900491.
PET degradation by I. sakaiensis PETase. Biophys J. 2018;114:1302–12. 38. Mix AK, Cenci U, Heimerl T, Marter P, Wirkner ML, Moog D. Identifica‑
33. Baiet B, Burel C, Saint-Jean B, Louvet R, Menu-Bouaouiche L, Kiefer- tion and localization of peroxisomal biogenesis proteins indicates the
Meyer MC, Mathieu-Rivet E, Lefebvre T, Castel H, Carlier A, et al. presence of peroxisomes in the cryptophyte Guillardia theta and other
N-glycans of Phaeodactylum tricornutum diatom and functional charac‑ ‘chromalveolates’. Genome Biol Evol. 2018. https​://doi.org/10.1093/
terization of its N-acetylglucosaminyltransferase I enzyme. J Biol Chem. gbe/evy21​4.
2011;286:6152–64. 39. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years
34. Mathieu-Rivet E, Kiefer-Meyer MC, Vanier G, Ovide C, Burel C, Lerouge of image analysis. Nat Methods. 2012;9:671–5.
P, Bardor M. Protein N-glycosylation in eukaryotic microalgae and its 40. Guillard RR, Ryther JH. Studies of marine planktonic diatoms I. Cyclotella
impact on the production of nuclear expressed biopharmaceuticals. nana Hustedt, and Detonula confervacea (cleve) Gran. Can J Microbiol.
Front Plant Sci. 2014;5:359. 1962;8:229–39.
35. Vanier G, Hempel F, Chan P, Rodamer M, Vaudry D, Maier UG, Lerouge
P, Bardor M. Biochemical characterization of human anti-hepatitis B
monoclonal antibody produced in the microalgae Phaeodactylum Publisher’s Note
tricornutum. PLoS ONE. 2015;10:e0139282. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions