Design of Project Collection
In the course of our project we generated 27 level 0 constructs, 25 level 1 constructs and 35 level 2
constructs by employing the modular cloning system (MoClo), which is based on Golden Gate Assembly
(Weber et al., 2011; Crozet et al., 2018). This allows us to provide iGEM with a total of 87 MoClo
constructs. The level 2 constructs, shown in Figure 1, were transformed into Chlamydomonas
reinhardtii (Chlamydomonas). We worked with the Chlamydomonas strains UVM4 in the CC-4350 background
and its sister strain in the CC-4533 background, hereafter referred to as Clip. UVM4 is a suitable
expression chassis and easy to transform (Neupert et al., 2009). The Clip sister strain is also a
suitable expression chassis, but more robust that UVM4 and therefore also suitable for growth in
To realize the biodegradation of polyethylene terephthalate (PET) by Chlamydomonas, we designed different MoClo constructs encoding the enzymes MHETase and PETase. PETase cleaves the polymer PET into its monomers MHET. MHETase then cleaves MHET into the basic modules ethylene glycol (EG) and terephtalic acid (TPA). We introduced three point mutations W159H, S238F and R280A into wild-type PETase (WT-PETase) to generate a mutant form (MUT-PETase). These changes have been reported to improve the activity of the enzyme (Austin et al., 2018; Joo et al., 2018). To secret these enzymes into the medium, we used three secretion signals, namely of carbonic anhydrase (cCa), arylsulfatase (ARS) and the gametic lytic enzyme (GLE) and removed the original secretion signal of Ideonella sakaiensis (Yoshida et al., 2016). To increase the amount of secreted enzyme, we also equipped the enzymes with a C-terminal glycomodule composed of tandem serine and proline repeats of 20 units (SP20-tag) (Ramos-Martinez et al., 2017).
Suitability of Chlamydomonas as PET Degrader
We first tested, whether the breakdown products of PET are toxic for Chlamydomonas. For this, we cultivated Chlamydomonas in the presence of 5 mM and 10 mM of breakdown products TPA and EG and determined growth rates. As shown in Figure 2, the growth rates were not significantly different to those of cultures lacking PET breakdown products and we can conclude that they are not toxic.
Furthermore, we could show that Chlamydomonas cells spontaneously attach to PET particles (Figure 3), which is of great advantage to increase the efficiency of secreted PETase and MHETase on microplastic particles. With these results we could demonstrate that Chlamydomonas is well-suited for degrading PET in a photobioreactor, since it shows spontaneous attachment to PET, is easy to engineer thanks to the MoClo system, and the PET breakdown products show no toxicity.
As a first step to equip Chlamydomonas with plastic-degrading enzymes MUT-PETase and MHETase, we transformed the UVM4 strain with construct L2A which targets the enzymes with a C-terminal 3xHA tag to the cytosol. As shown in Figure 4, both enzymes accumulated in transformants, but to different levels. This is typically resulting from position effects. Here we could observe a mass shift for both enzymes compared with the Ideonella sakaiensis forms of ~5 kDa for MUT-PETase and ~10 kDa for MHETase. These are most likely caused by the 3xHA tag and post-translational modifications.
Troubleshooting of the Secretion Process
Next, we wanted to have both enzymes secreted and to this end generated construct L2C that differs from L2A by the presence of sequences equipping MUT-PETase and MHETase with the N-terminal secretion signal from carbonic anhydrase (cCA). As shown in Figure 5a, we could detect MHETase in the medium, but no MUT-PETase. We reasoned that MUT-PETase might not be secreted and looked for it in cell proteins. Indeed, MUT-PETase accumulated to high levels in the cell body (Figure 5b). We next employed immunofluorescence to localize MUT-PETase in the cell and detected it in the ER (Figure 5c).
To solve this problem, we generated constructs L2H, L2I, and L2J for the expression of MUT-PETase alone equipped with the secretions signals from carbonic anhydrase (cCA), gamete lytic enzyme (GLE), and arylsulfatase (ARS), respectively. While a module encoding the cCA signal is present in the MoClo kit (Crozet et al., 2018), those for GLE and ARS were generated by us. Arylsulfatase is essential for the mineralization of sulfate by hydrolyzing sulfate esters under conditions of sulfate deprivation (deHostos et al., 1988). Gamete lytic enzyme is a metalloprotease that mediates digestion of the cell wall during gametogenesis (Kinoshita et al., 1992). While we detected no MUT-PETase in the medium when it was equipped with the cCA signal, weak signals were detected when it contained secretion signals GLE and ARS (Figure 6).
Enhancing Secretion and Identification of Secreted Proteins
In the meanwhile, we had learned about a report where repeats of serine and proline attached to the C-terminus of secreted proteins strongly increases the yield of secretion (Ramos-Martinez et al, 2017). That’s why we designed constructs L2M, N and O to express MUT-PETase and HMETase with repeats of 20 serine and proline residues (SP20-tag). As shown in Figures 7 and 8, this strongly improved the secretion efficiency for both enzymes. Proline in these repeats is hydroxylated and thereby converted to hydroxyproline, which subsequently gets O-glycosylated. This engineering technology of hydroxyproline-rich glycoproteins exploits the glycosylation code of cell wall proteins and leads to their efficient export into the medium (Kieliszewski and Shpak, 2001). Accordingly, we observed a strong increase in the apparent molecular mass for both enzymes that is likely due to heavy protein glycosylation (Figure 8).
Our detection of MUT-PETase and MHETase was based on an antibody against the 3xHA tags at their C-termini. Due to the glycosylation of the proteins, we detected multiple protein bands and could not distinguish between the two enzymes. Therefore, we identified the proteins in these bands by mass spectrometry. We could detect eight specific peptides for MHETase and three specific peptides for MUT-PETase and therefore could verify that both enzymes are secreted into the medium (Figure 9).
To verify effective secretion of both enzymes equipped with the SP20 tag, we analyzed their presence in the medium and in the cell bodies in three tranformants generated with constructs L2M, N, and O (Figure 8) differing only by the secretion signals for MUT-PETase. As shown in Figure 10, hardly any protein was detected in cell bodies while we detected the bulk of the proteins in the medium. The highest amounts of secreted, glycosylated PETase were found in transformant N6 expressing MUT-PETase with the ARS secretion signal.
To estimate the quantity of secreted enzymes we performed quantitative immunoblot blot analysis. Transformants N6 and M5 had the highest amounts of secreted MUT-PETase and MHETase (Figure 11).
Since MHETase accumulates at a higher level than MUT-PETase, we wanted to secret both enzymes equally. Therefore, we transformed clone N6 (Figure 8) with construct L2AI (12a). Thereby, we could demonstrate an equally strong secretion of MHETase and MUT-PETase from the transformants 4, 9 and 10. Furthermore we observed secretion of MUT-PETase in all transformants, excluding transformant 1.
On a long run we need to grow Chlamydomonas engineered with plastic degrading enzymes in a bioreactor. Since the UVM4 strain is too fragile, we had to use another strain background and for this used the more robust Clip sister line of UVM4. We transformed Clip with the constructs L2M and L2N and could also in this strain background demonstrate successful secretion of both enzymes, with MHETase accumulating to much higher levels than MUT-PETase (Figure 12).
Expanding Experiments into a Bioreactors
To determine the optimal conditions for secreting our enzymes, we analyzed secretion and growth at different light and temperature regimes. Thanks to the advice of Prof. Dr. Zimmermann (University of Leipzig, Institute of Biochemistry) we learned that at PETase is more active and PET is better accessible at higher temperatures. That’s why we tested secretion and growth not only at 25 °C but also at 33 °C. UVM4 grew faster at 33°C and 170 µE light intensity but reached a lower final cell density that when grown at 25°C and 80 µE light intensity (Figure 14A). Also Clip grew faster at 33°C and 170 µE light intensity but reached the same final cell density as when grown at 25°C and 80 µE light intensity (Figure 15A). Transformants with constructs L2N or L2M both accumulated more MHETase and less MUT-PETase when grown at 33°C and 170 µE light intensity, while they both accumulated less MHETase and more MUT-PETase when grown at 25°C and 80 µE light intensity (Figures 14b and 15b). Hence, growth conditions obviously have a strong impact on the amounts of secreted proteins.
After the successful production and secretion of MUT-PETase and HMETase in Clip strains, we wanted to test the suitability of the engineered strains for growth in a photobioreactor. We did this in parallel in two 1.9-L photobioreactors run in turbidostat mode, in which the cell density was kept at different thresholds. We could show that both enzymes accumulated in the culture medium about 65 hours after inoculation. Surprisingly, between 120 h and 243 h of cultivation, levels of secreted enzymes declined for yet unknown reasons. Further experiments are required to elucidate the mechanisms involved here.
First Activity Measurements
After the successful production and secretion of MUT-PETase and MHETase via transgenic Chlamydomonas cells, we wanted to demonstrate the functionality of these enzymes. For this purpose, we incubated the secreted enzymes with PET, BHET and MHET under different conditions. Since PETase degrades PET and BHET into MHET and MHETase degrades MHET into EG and TPA, the resulting products can be detected by reversed phase high-performance liquid chromatography (HPLC). The phenol moiety in BHET, MHET, and TPA allows detection of these compounds by UV light. Since such a moiety is missing in EG, EG is not detectable by reversed phase HPLC.
We first wanted to test, whether we could detect degradation of PET, BHET, and MHET by HPLC and for this needed the purified PETase and MHETase enzymes. We received an appropriate expression vector from Dr. rer. nat. Gottfried J. Palm, University of Greifswald, for the production of hexa-histidine-tagged MHETase. We produced the recombinant proteins MHETase and PETase containing three point mutations in E.coli and purified them via Ni-NTA affinity chromatography (Figure 17). For their expression, we used the five E.coli strains Rosetta-gami, Origami, ArcticExpress, ER2566 and SHuffle T7 but only succeeded with E.coli SHuffle T7.
After the successful purification of MUT-PETase and MHETase from E.coli, we tested PETase activity by incubating the enzyme with PET film and measuring putative degradation products BHET, MHET, and TPA via reversed-phase HPLC (Figure 18). Since we could recognize peaks for all three degradation products after lowering the scale, we can conclude that PETase is active on PET film when produced in E.coli. This is supported by the finding that no degradation products were detected when the protein had been thermally denatured before incubation with the PET film. Moreover, we can state that our analytical platform is well-suited for the detection of PETase activity.
Activity of Secreted Enzymes from Chlamydomonas
The next step was to analyze the activity of the secreted MUT-PETase and MHETase produced in Chlamydomonas. For this, we had to extract the enzymes from Chlamydomonas transformant M5, which secretes both enzymes into the culture medium. We did this by affinity purification using anti-HA magnetic beads and used the parent strain UVM4 as control (Figure 19). We then measured the activity of the eluted proteins against BHET by reversed-phase HPLC. A peak for MHET was detected in the assay with eluates of the M5 strain and from the parent strain. This indicates an activity for the conversion of BHET to MHET coming from the buffer or Chlamydomonas. However, in the assay performed with eluates from the M5 strain we detected a peak for TPA, which was absent in the assay with eluates from the parent strain. This indicates that there is MHETase activity. When comparing the TPA peaks in plots with the same x-axis scale between E.coli-derived and Chlamydomonas-derived enzymes, they are even comparable.
Since affinity-purification of the enzymes from culture medium is tedious and expensive, we used ultracentrifuge filters for this purpose and achieved a ~20-fold concentration. We then tested the activity of MUT-PETase and MHETase in concentrated culture medium against PET (Figure 20). To this end, the concentrates from transformant M8 and the parental strain were incubated with PET film. We observed tiny peaks for BHET and TPA that were absent in the assays with concentrate from the CC-4533 parent strain or buffer alone. This indicated very weak activity of both enzymes secreted by Chlamydomonas.
After we have shown a low activity of the enzymes secreted from CC-4533 transformant M8 against PET, we repeated the experiment with BHET under the same conditions with enzymes secreted from UVM4 transformant M5 (Figure 21). Unexpectedly, we observed a small peak for MHET in the assays performed with the concentrate from the parental strain or buffer alone, which was absent in the assay with concentrate from transformant M5. In contrast, a peak for TPA was observed in the assay with concentrate from transformant M5, which was absent in the assays with concentrate from the parental strain or buffer alone. These findings indicate that there is activity for the conversion of BHET to MHET already in the buffer. However, the activity for the conversion of MHET to TPA, i.e. for MHETase, is only present in the concentrate from transformant M5. We suggest that there is a enzymatic activity against BHET, since the peak area of TPA of M5 is major than the sum of the peak areas of MHET and TPA in controls.
To verify the activity for MHETase from transformant M5, we incubated the concentrated medium from transformant M5 and from the UVM4 parent strain directly with MHET as substrate (Figure 22). Thereby, we could observe a complete conversion from MHET to TPA only in the assays with concentrates from M5 and no peaks for TPA in assays with concentrates from the parent strain. This corroborates our conclusion that the MHETase secreted from transformant M5 shows indeed a high activity.
Temperature Dependent Activity
Thanks to Prof. Dr. Zimmermann from the University of Leipzig we know that PETase shows a higher activity at higher temperature (Link zu IHP (Human practice)). We therefore tested the activity of the secreted enzymes of transformant M8 against MHET and PET at 25°C and 33°C (Figure 23). We observed activity of MUT-PETase and MHETase against PET and MHET at 25°C and 33°C. However, we couldn’t observe great differences in activity between the temperatures.
Since there was no difference in enzymatic activity between 25°C and 33°C, we compared the activity of MUT-PETase and MHETase of transformant M5 at 30°C and 40°C and used BHET as substrate (Figure 24). As already described in Figure 21, at 30°C we observed a peak for MHET in the assays performed with the concentrate from the UVM4 parent strain and buffer alone, which was missing in the assay with the concentrate from transformant M5. Instead, a peak for TPA was only observed in the assay with M5 concentrate, but not in those with buffer of concentrate from the UVM4 parent strain. At 40°C, stronger peaks for MHET were detected in all assays and the peak for TPA in the assay with M5 concentrate was smaller. These data indicate that spontaneous conversion of BHET to MHET is enhanced at 40°C versus 30°C, but MHETase activity is impaired.
After we successfully secreted MUT-PETase and MHETase and proofed the activity of both enzymes we wanted to ensure the safety of our bio-recycling method. Therefore, we used an auxotrophic strain and tested its dependence on arginine. We proofed that Chlamydomonas strain cw325 can’t grow without the presence of arginine in the medium. Consequently, it could be implemented in our system to enhance safety.
The following experiments could be carried out to optimize PET biodegradation via genetically engineered Chlamydomonas:
To enhance the secretion of both enzymes, especially of MUT-PETase, the constructs harboring the coding sequences for each enzyme should be transformed sequentially into Chlamydomonas. We would first screen for transformants expressing PETase to very high levels and then transform this strain with the MHETase construct and screen again. We would need to equip the enzymes with different tags to better distinguish between them.
The HA tag at the C-terminus of PETase might interfer with its activity. We would construct a vector containing the coding sequence for PETase without any tags and raise an antibody against recombinant PETase from E.coli to detect it in Chlamydomonas transformants.
Another approach, which already was planned but only partially realized, is to create a fusion protein of MUT-PETase and MHETase separated by a linker. We hope that this will result in an equally strong secretion of both enzymes and an improvement of their activity since the subsequent reaction can immediately take place.
To gain a deeper understanding of our enzymes we could design constructs for the production and secretion of the cutinase from Prof. Dr Zimmermann and compare its activity with that of MUT-PETase.
PET has crystalline and amorphous structures, with the amorphous structures being better accessible for degradation than the crystalline structures. Since PET exhibits amorphous structures at higher temperatures, another approach would be to engineer thermostability into both enzymes. The thermostable enzymes could be purified from the culture medium for activity assays at higher temperatures.
In order to prevent possible glycosylation near the active center, the glycosylation patterns discovered in the model could be removed by inserting point mutations. Furthermore, the activity of MUT-PETase could be compared with WT-PETase.
- Austin HP, Allen MD, Donohoe BS, Rorrer NA, Kearns FL, Silveira RL, Pollard BC, Dominick G, Duman R, El Omari K, Mykhaylyk V, Wagner A, Michener WE, Amore A, Skaf MS, Crowley MF, Thorne AW, Johnson CW, Woodcock HL, McGeehan JE, Beckham GT (2018) Characterization and engineering of a plastic-degrading aromatic polyesterase. Proc Natl Acad Sci U S A 115: E4350-E4357
- Crozet P, Navarro FJ, Willmund F, Mehrshahi P, Bakowski K, Lauersen KJ, Perez-Perez ME, Auroy P, Gorchs Rovira A, Sauret-Gueto S, Niemeyer J, Spaniol B, Theis J, Trosch R, Westrich LD, Vavitsas K, Baier T, Hubner W, de Carpentier F, Cassarini M, Danon A, Henri J, Marchand CH, de Mia M, Sarkissian K, Baulcombe DC, Peltier G, Crespo JL, Kruse O, Jensen PE, Schroda M, Smith AG, Lemaire
- SD (2018) Birth of a photosynthetic chassis: a MoClo toolkit enabling synthetic biology in the microalga Chlamydomonas reinhardtii. ACS Synth Biol 7: 2074-2086
- de Hostos EL, Togasaki RK, Grossman A (1988) Purification and biosynthesis of a derepressible periplasmic arylsulfatase from Chlamydomonas reinhardtii. J Cell Biol 106: 29-37
- Joo S, Cho IJ, Seo H, Son HF, Sagong H-Y, Shin TJ, Choi SY, Lee SY, Kim K-J (2018) Structural insight into molecular mechanism of poly(ethylene terephthalate) degradation. Nature Communications 9: 382
- Kieliszewski MJ, Shpak E (2001) Synthetic genes for the elucidation of glycosylation codes for arabinogalactan-proteins and other hydroxyproline-rich glycoproteins. Cell Mol Life Sci 58: 1386-1398
- Kinoshita T, Fukuzawa H, Shimada T, Saito T, Matsuda Y (1992) Primary structure and expression of a gamete lytic enzyme in Chlamydomonas reinhardtii: similarity of functional domains to matrix metalloproteases. Proc Natl Acad Sci U S A 89: 4693-4697
- Neupert J, Karcher D, Bock R (2009) Generation of Chlamydomonas strains that efficiently express nuclear transgenes. Plant J 57: 1140-1150
- Ramos-Martinez EM, Fimognari L, Sakuragi Y (2017) High-yield secretion of recombinant proteins from the microalga Chlamydomonas reinhardtii. Plant Biotechnol J 15: 1214-1224
- Weber E, Engler C, Gruetzner R, Werner S, Marillonnet S (2011) A modular cloning system for standardized assembly of multigene constructs. PLoS One 6: e16765
- Yoshida S, Hiraga K, Takehana T, Taniguchi I, Yamaji H, Maeda Y, Toyohara K, Miyamoto K, Kimura Y, Oda K (2016) A bacterium that degrades and assimilates poly(ethylene terephthalate). Science 351: 1196-1199