Team:Kyoto/Results
Result
Overview
We aimed to create a device that makes PET fibers precipitate. The device named "SONOBE" is a protein capsule that has plastic-binding proteins on its surface. In order to create it, several components of the device are purified. They are assessed if each component works as designed. 1st component: plastic-binding proteins bound to many PET materials. They make our device strongly stick to PET fibers. 2nd component: Encapsulin did form the protein capsule. It can work as a biological polymer. 3rd component: SpyCatcher/SpyTag conjugated plastic-binding protein and Encapsulin. Through these, our device "SONOBE" assembles. We assessed our device with PET fibers, then finally succeed to observe "SONOBE" make PET fibers precipitate, and aggregate.
1. Purification of Recombinant Proteins
Key Achievement
1) Four plastic-binding proteins were successfully purified. 2) Thermotoga martima Encapsulins were purified by newly improved purification method.
As the first step, we cloned 10 genes for our device into pET-11a plasmids. The expression plasmids were introduced into BL21(DE3) and expressed by T7 promoter/ T7 RNAP system. As each recombinant protein used here carries 6x His tag (see Design), Ni-NTA agarose was used for the purification. The imidazole eluates were visualized and confirmed by SDS-PAGE followed by CBB staining.
3 plastic-binding proteins (SpyC->PETase, SpyC->engineered PETase, and Hydrophobin) out of 10 proteins tested were not purified successfully. However, as shown in Fig.1a and 1b, other 7 proteins, including 4 plastic-binding proteins ( SpyC->BaCBM2, SpyC->LCI KR-2, SpyC->TA2, and SpyC->CenA) and 3 Encapsulin derivatives (SpyTag inserted TmEncapsulin, TmEncapsulin, and sfGFP/TmEncapsulin) were successfully produced and purified to apparent homogeneity.
It is noteworthy that this is the first report of recombinant TmEncapsulin purification by His tag / Ni-NTA system.
Previously, it was reported that TmEncapsulin with the C-terminus 6x-His tag was hardly purified by Ni-NTA [1]. We thought this is because the C-terminus of TmEncapsulin is not sufficiently exposed to the surface of the capsule. To improve this situation, we inserted HA tag as a linker between the C-terminus of TmEncapsulin and 6x-His tag. As we expected, this modification enabled us to purify TmEncapsulin with Ni-NTA beads. This purification protocol is much simpler than the classical one, which requires two rounds of ultracentrifugation and size separation by gel filtration column.
The concentrations of purified proteins are listed in Table 1.
Fig. 1a SDS-PAGE gel of purified protein
Fig. 1b SDS-PAGE gel of purified protein
Table. 1 The concetrations of Purified proteins
2. Recombinant Plastic-binding Proteins Efficiently Bind to the Surface of Plastics
Key Achievement
1) Plastic-binding proteins bind to film, cloth, and fiber which are made of PET.2) Binding amounts of plastic-binding proteins were quantified.
a) PET film assay
We purified four plastic-binding proteins. We next examined those proteins to find out which protein sticks to PET most strongly. In the paper which shows BaCBM2’s PET binding ability, PET film was used for affinity assay [2]. So, we tried to compare our proteins with each other by the film dot blotting shown in the reference.
As shown in Fig.2, the negative control protein, SpyCatcher (SPYC), did not stain PET film at all. In contrast, all the plastic-binding proteins tested here strongly stained the PET film. Although BaCBM2 and CenA might be darker than the other two proteins, as stains spread, we could not quantify their signals. This blot spreading might be due to the plastic-binding proteins’ fast binding rate. The proteins in excess liquid could have bound to the neighbor area of the film in the first wash step.
Fig.2 Plastic-binding protein binding to PET film
A 3µL of protein solution dropped on PET film with serial dilution, then left for 20min. Then, the film was washed in TBST for 5min x3, then placed with Anti-His-tag-HRP conjugated for 1h. After washing, ECL substrate was added, then chemiluminescence was imaged by LAS-3000. The exposure time is 6min.
A 3µL of protein solution dropped on PET film with serial dilution, then left for 20min. Then, the film was washed in TBST for 5min x3, then placed with Anti-His-tag-HRP conjugated for 1h. After washing, ECL substrate was added, then chemiluminescence was imaged by LAS-3000. The exposure time is 6min.
b) PET cloth assay
Next, we quantified how much protein bound to PET fiber with fluorescence. Two of the purified plastic-binding proteins were fused with sfGFP in our construct, so these two proteins was able to be seen as fluorescent signals.We bought white PET-made T-shirts and cut it into pieces. The lysate of E.coli which expresses each protein was used for this experiment. The concentrations of fluorescent proteins were measured by SDS-PAGE and CBB staining, and an equal amount of proteins were used for each assay. Diluted proteins were spotted on a piece of PET cloth. The cloth was incubated for 20 min at room temperature, then washed by TBST for 5 min 3 times. Finally, fluorescent proteins were photographed.
Fig.3a shows the cloth before wash, and 3b shows the same cloth after wash. As shown in figures, sfGFP was completely washed out. In sharp contrast, the both plastic-binding proteins strongly stuck to PET cloth. The intensity of each dot was quantified with ImageJ.
As shown in Fig.4, about 65% of LCI KR-2 and about 75% of TA2 were still observed on PET cloth after wash, showing these proteins are really strong PET cloth binders.
Fig.3a Cloth dot blot by fluorescent plastic-binding proteins before washing
The dilution collection of each protein was dropped on PET cloth, then left for 20min. The protein fluorescent was imaged by LAS-3000. The exposure time is 10sec. Saturated signals were painted red.
The dilution collection of each protein was dropped on PET cloth, then left for 20min. The protein fluorescent was imaged by LAS-3000. The exposure time is 10sec. Saturated signals were painted red.
Fig.3b Cloth dot blot by fluorescent plastic-binding protein after washing
The dilution collection of each protein was dropped on PET cloth, then left for 20min. The cloth was washed in TBST for 5min x3, then protein fluorescent was imaged by LAS-3000. The exposure time is 10sec. Saturated signals were painted red.
The dilution collection of each protein was dropped on PET cloth, then left for 20min. The cloth was washed in TBST for 5min x3, then protein fluorescent was imaged by LAS-3000. The exposure time is 10sec. Saturated signals were painted red.
Fig.4 The percentage of protein retention on PET cloth
The whole signal of 37.5ng dot was used for this quantification, percentages of protein retention were calculated by comparing intensity before and after washing.
The whole signal of 37.5ng dot was used for this quantification, percentages of protein retention were calculated by comparing intensity before and after washing.
Based on the data shown above, we concluded that our fluorescent plastic-binding proteins highly stably bind to PET cloth. In order to demonstrate this conclusion clearly, we drew a picture by two different GFP inks; sfGFP alone and GFP-LCI KR-2. When washed by tap water, sfGFP is washed out, while the stable PET binder GFP-LCI KR-2 remains.
Fig.5 Fluorescent “Konkon” appears on PET cloth
We drew mascot “Konkon” with LCI KR-2 protein solution, then the cloth was soaked in sfGFP solution. In this movie, the cloth is washed with tap water.
We drew mascot “Konkon” with LCI KR-2 protein solution, then the cloth was soaked in sfGFP solution. In this movie, the cloth is washed with tap water.
c) PET fiber assay
We showed two fluorescent plastic-binding proteins bind to PET cloth very tightly. Next, we demonstrated proteins’ binding in a more realistic target: PET fiber. In cloth, fibers are close to each other, so they might create a hydrophobic environment between them. In the fiber form, they are surrounded by water, so plastic-binding proteins might behave in a different way. Thanks to human practice activity, the textile company “T” gave us plain PET fiber and we used this in this experiment (see Human Practice).In this experiment, PET fibers were soaked in water or protein solutions, then washed in TBST for 5 mins three times. The concentrations of protein solutions were 2000 ng/µL. sfGFP and other sfGFP-fused protein’s fluorescence were observed by a fluorescence microscope in 460 nm exciting light.
Clearly, when LCI KR-2 or TA2 were fused to sfGFP, the PET fiber was brightly stained by fluorescence. Water control did not show any fluorescence, indicationg that no autofluorescence was observed with these fibers. sfGFP control also showed no signals, meaning that sfGFP binding was mediated by the plastic-binding domain in the sfGFP fusion proteins.
Fig. 6 Plastic-binding proteins also bind to PET fiber
PET fibers were soaked in each protein solution, then washed in TBST for 5 min x 3 times. Fluorescence was observed in 460 nm exciting light and imaged with 250 ms exposure time. Magnification is 10x.
PET fibers were soaked in each protein solution, then washed in TBST for 5 min x 3 times. Fluorescence was observed in 460 nm exciting light and imaged with 250 ms exposure time. Magnification is 10x.
As shown in Fig.7a and 7b, BaCBM2 binds the most to PET fiber. According to the references, BaCBM2 and CenA is a polyethylene terephthalate (PET)-binding protein, LCI KR-2 is a polypropylene (PP) binding protein, and TA2 is a polyurethane (PU) binding protein. Therefore, this result is consistent with the reported observation. It is of interest that LCI KR-2 (PP binding) and TA2 (PU binding) proteins also showed a moderate but measurable amount of PET-binding.
Fig.7a SDS-PAGE gel for quantification of proteins bound to PET fiber
20 cm of PET fibers were soaked in protein solutions, then washed in TBST for 5 min three times. Washed fibers were soaked in 50µL of 2x SDS sample buffer. Bound proteins were eluted with boiling. SDS-PAGE for 40min in 200V. CBB-stained.Lower band in lane 5 is BaCBM2. Upper band in lane 5 is LCI KR-2. Lower band in lane 6 is CenA. Upper band in lane 7 is TA2.
20 cm of PET fibers were soaked in protein solutions, then washed in TBST for 5 min three times. Washed fibers were soaked in 50µL of 2x SDS sample buffer. Bound proteins were eluted with boiling. SDS-PAGE for 40min in 200V. CBB-stained.Lower band in lane 5 is BaCBM2. Upper band in lane 5 is LCI KR-2. Lower band in lane 6 is CenA. Upper band in lane 7 is TA2.
Fig.7b BaCBM2 bind most to PET fiber
SDS-PAGE’s gel band intensity quantified with ImageJ. The y-axis shows amounts of protein which bind to 20cm PET fiber.
SDS-PAGE’s gel band intensity quantified with ImageJ. The y-axis shows amounts of protein which bind to 20cm PET fiber.
3. Encapsulin properly make spherical polymer
Key Achievement
1) TmEncapsulin properly formed a spherical polymer.
As shown in the section.1, TmEncapsulin was successfully purified with Ni-NTA beads to apparent homogeneity. We next assessed the size of the protein polymer.
The TmEncapsulin purified and eluted from Ni-NTA was loaded on the top of 10%-60% linear sucrose gradient. The samples were ultracentrifuged in 100,000g for 18 hours at 4℃ with SW41(Beckman) and fractionated on a 96-well plate with BioComp. Absorbance 260 nm was monitored.
In Fig. 8, the blue line shows A260 of E. coli lysate (control). As shown in the figure, bacterial ribosomes are observed as peaks in the indicated position. The red line shows the A260 profile of eluted TmEncapsulin. Around the 60th fraction, a peak was clearly observed. As TmEncapsulin polymer’s size is 20 nm and the 70S bacterial ribosome’s size is also about 20 nm, this peak around 60th fraction looks exactly the TmEncapsulin spherical polymer. Interestingly, the first drop of the fractionation (the top of the fraction) was around 10 (signals between 1-9 can be attributed to air bubbles), showing that most of the A260 signal in the purified TmEncapsulin were collected in 70S-80S area, with almost no accumulation of monomer form.
In order to confirm that the 60th fraction’s peak result from TmEncapsulin, we examined the fraction with SDS-PAGE. In Fig.9, lane1 is TmEncapsulin expressed E. coli lysate, lane 2 is purified protein, and lane 3 is 60th fraction. As shown in lane 3, the 60th fraction properly has Encapsulin. Taken together, we concluded that our TmEncapsulin conserves the spherical structure.
The TmEncapsulin purified and eluted from Ni-NTA was loaded on the top of 10%-60% linear sucrose gradient. The samples were ultracentrifuged in 100,000g for 18 hours at 4℃ with SW41(Beckman) and fractionated on a 96-well plate with BioComp. Absorbance 260 nm was monitored.
In Fig. 8, the blue line shows A260 of E. coli lysate (control). As shown in the figure, bacterial ribosomes are observed as peaks in the indicated position. The red line shows the A260 profile of eluted TmEncapsulin. Around the 60th fraction, a peak was clearly observed. As TmEncapsulin polymer’s size is 20 nm and the 70S bacterial ribosome’s size is also about 20 nm, this peak around 60th fraction looks exactly the TmEncapsulin spherical polymer. Interestingly, the first drop of the fractionation (the top of the fraction) was around 10 (signals between 1-9 can be attributed to air bubbles), showing that most of the A260 signal in the purified TmEncapsulin were collected in 70S-80S area, with almost no accumulation of monomer form.
In order to confirm that the 60th fraction’s peak result from TmEncapsulin, we examined the fraction with SDS-PAGE. In Fig.9, lane1 is TmEncapsulin expressed E. coli lysate, lane 2 is purified protein, and lane 3 is 60th fraction. As shown in lane 3, the 60th fraction properly has Encapsulin. Taken together, we concluded that our TmEncapsulin conserves the spherical structure.
Fig. 8 TmEncapsulin polymer appears as the peak
TmEncapsulin expressed E. coli lysate and purified protein solution was loaded on 10%-60% sucrose linear gradient / 20 mM Tris 7.5, 50 mM NaCl, then centrifuged in 100,000g for 18 hours at 4℃ with SW41(Beckman). The solution was fractionated on a 96-well plate with BioComp. At the same time, 260nm absorption was measured.
TmEncapsulin expressed E. coli lysate and purified protein solution was loaded on 10%-60% sucrose linear gradient / 20 mM Tris 7.5, 50 mM NaCl, then centrifuged in 100,000g for 18 hours at 4℃ with SW41(Beckman). The solution was fractionated on a 96-well plate with BioComp. At the same time, 260nm absorption was measured.
Fig. 9 SDS-PAGE of the fraction
Lane1 is TmEncapsulin expressed E. coli lysate, lane 2 is purified protein, and lane 3 is 60th fraction. SDS-PAGE for 40min in 200V. CBB-stained.
Lane1 is TmEncapsulin expressed E. coli lysate, lane 2 is purified protein, and lane 3 is 60th fraction. SDS-PAGE for 40min in 200V. CBB-stained.
4. Protein conjugation thorough SpyCatcher/SpyTag system
Key Achievement
1) Plastic-binding protein and protein capsule can be conjugated in vitro.2) Kinetics of SpyCatcher/SpyTag bond formation was monitored.
As we showed above, our two components: 1. Plastic-binding protein and 2. Encapsulin worked exactly as planned. Next, we conjugated these two components through SpyCatcher/SpyTag system. SpyCatcher and SpyTag form an isopeptide bond between them when they are mixed. Each plastic-binding protein is fused with SpyCatcher, and Encapsulin has SpyTags inserted on its surface (see Design).
The equal amount of SpyCatcher-Plastic-binding protein (SpyC-PBP) solution and SpyTag inserted TmEncapsulin (SpyTmEnc) solution were mixed and incubated for 16h at room temperature. Samples were taken and assessed with SDS-PAGE.
These results show we successfully conjugated several proteins to Encapsulin by SpyTag-SpyCatcher system in vitro. This means that any protein with SpyCatcher can be efficiently and easily displayed on the surface of the protein capsule.
Next, we measured the time development of SpyCatcher-SpyTag bond formation. An equal amount of SpyCatcher protein (SpyC) and SpyTag inserted TmEncapsulin (SpyTmEnc) were mixed and incubated at room temperature. At different time points, 10 min, 30 min, 60 min, 180 min, 360 min, 1200 min, the reaction was stopped by adding 2x SDS sample buffer. Mixed samples were assessed with SDS-PAGE. The intensities of the conjugated bands were quantified.
As shown in Fig. 11, conjugated bands become evident gradually (labeled with arrow). Signals were quantified and summarized in Fig. 12. The reaction looks saturated after 360 minutes, even though substrates still remain a lot. This might be explained by water evaporation while incubation. Otherwise, it is possible that a bound protein prevents another protein from binding to near sites on a capsule cage. If it is the case, it might limit the number of binding proteins on a capsule.
The equal amount of SpyCatcher-Plastic-binding protein (SpyC-PBP) solution and SpyTag inserted TmEncapsulin (SpyTmEnc) solution were mixed and incubated for 16h at room temperature. Samples were taken and assessed with SDS-PAGE.
In Fig. 10, several kinds of combinations of proteins were shown. In lane 4 and 5, SpyTmEnc is loaded with or without SpyC. Only in lane 5, which is mixed with SpyC, the upper band appeared. The molecular weight of each protein is SpyC: 15.37k, SpyTmEnc: 37.04k, so the conjugated protein should be 52.41k. We concluded that the upper band is the conjugated protein. Likewise, as shown in lane 7 and 9, SpyC-PBPs are successfully conjugated to SpyTmEnc. As the negative control, we tested TmEncapsulin without SpyTag. As expected, TmEnc and SpyC did not produce conjugated protein as shown in lane 3.
Fig.10 SpyTmEnc can be conjugated with different kind of protein
3µL of SpyCatcher-Plastic-binding protein (SpyC-PBP) solution and 3µL of SpyTag inserted TmEncapsulin (SpyTmEnc) solution was mixed, then placed for 16h at room temperature. Then 6µL of 2x SDS sample buffer was added. 10µL of each sample was loaded. SDS-PAGE for 30min in 200V. CBB-stained.
3µL of SpyCatcher-Plastic-binding protein (SpyC-PBP) solution and 3µL of SpyTag inserted TmEncapsulin (SpyTmEnc) solution was mixed, then placed for 16h at room temperature. Then 6µL of 2x SDS sample buffer was added. 10µL of each sample was loaded. SDS-PAGE for 30min in 200V. CBB-stained.
As shown in Fig. 11, conjugated bands become evident gradually (labeled with arrow). Signals were quantified and summarized in Fig. 12. The reaction looks saturated after 360 minutes, even though substrates still remain a lot. This might be explained by water evaporation while incubation. Otherwise, it is possible that a bound protein prevents another protein from binding to near sites on a capsule cage. If it is the case, it might limit the number of binding proteins on a capsule.
Fig. 11 Time development of SpyCatcher-SpyTag bond formation
3µL of SpyCatcher protein (SpyC) solution and 3µL of SpyTag inserted TmEncapsulin (SpyTmEnc) solution was mixed, then placed for several times at room temperature (10 min, 30 min, 60 min, 180 min, 360 min, 1200 min). The concentrations of protein solutions are SpyC: 500ng/µL SpyTmEnc: 1000ng/µL. The reaction was stopped by adding 2x SDS sample buffer. Mixed samples were assessed with SDS-PAGE for 30 min in 200V. CBB-stained.
3µL of SpyCatcher protein (SpyC) solution and 3µL of SpyTag inserted TmEncapsulin (SpyTmEnc) solution was mixed, then placed for several times at room temperature (10 min, 30 min, 60 min, 180 min, 360 min, 1200 min). The concentrations of protein solutions are SpyC: 500ng/µL SpyTmEnc: 1000ng/µL. The reaction was stopped by adding 2x SDS sample buffer. Mixed samples were assessed with SDS-PAGE for 30 min in 200V. CBB-stained.
Fig. 12 Quantification of conjugated band
Conjugated bands’ intensity was quantified with ImageJ. Orange dots show averages value of three experiments. Blacklines show standard deviations. The time point 60min was deleted because it included negative value.
Conjugated bands’ intensity was quantified with ImageJ. Orange dots show averages value of three experiments. Blacklines show standard deviations. The time point 60min was deleted because it included negative value.
5. Mixing "SONOBE" with fibers
Key Achievement
1)Proteins actually made fibers precipitate.2)The device increased the size of microfibers.
In previous sections, we showed that every component worked properly, and all components are combined into our device: "SONOBE". Next, we assessed the interaction
between our device and microfibers.
Our goal is to make microfibers precipitate, so to begin with, we mixed microfibers with the device. In order to get microfibers of a consistent quality, we cut the fibers which was provided from the company“T”into smaller fibers: about 1mm or less (About compant "T",see Human Practice). The device was prepared by mixing SpyTag-inserted TmEncapsulin (lysate) and SpyC->sfGFP->LCI KR-2 (lysate). Fibers were added to water and mixed well in a 50mL beaker.
To make sure that our device would really work for precipitation, we also tested it with BSA solution and LCI KR-2 solution (with no Encapsulin).
Finally, with the great help by Sanyo Chemical Industries (see Human Practice),
we examined if our device would actually aggregate microfibers. In this experiment, we measured the size of microfibers' particle with laser diffraction by Microtrak MT3000Ⅱ. Microfibers and devices were added to circulating water in the machine. In this machine, microfibers' particle diameters were measured, and the frequency of each diameter particle was output.
This result suggests some of 10 to 20 µm particle aggregate into larger particles by adding the device. The diameter of the fiber which we used here is about 15 µm. These results can be interpreted as that the microfibers were actually aggregated into bundles when adding our device.
As shown in Fig.13, on the left side, most of the fibers were floating on the water surface. On the right side, 0.5mL of the device was added. In this case, fibers on the water surface precipitated clearly.
Fig. 13 Fibers on the water surface precipitated with "SONOBE"
Fibers were mixed with water. Some fibers were floating when they were mixed just with water (Left). When the device was added, most of the fibers precipitated (Right).
Fibers were mixed with water. Some fibers were floating when they were mixed just with water (Left). When the device was added, most of the fibers precipitated (Right).
As shown in Fig.14, fibers also precipitated when adding LCI KR-2. In the control experiment, we show that BSA did not enhance the precipitation in contrast. We interpreted the precipitation shown in Fig.13 simply depends on LCI KR-2. In these experiments, the surface of fibers might be covered with LCI KR-2, then turned into hydrophilic. This could also be an alternative way to precipitate microfibers. Although we couldn't exactly confirm the effect of Encapusulin in this assay, we clearly showed that the PET-binding proteins solely can contribute to the precipitation of microfibers.
Fig. 14 Fibers precipitated simply with LCI KR-2
Fibers were mixed with water. Some fibers was floating when they were just mixed with water and BSA (Left). When the LCI KR-2 was added, almost all of the fibers precipitated (Right).
Fibers were mixed with water. Some fibers was floating when they were just mixed with water and BSA (Left). When the LCI KR-2 was added, almost all of the fibers precipitated (Right).
In Fig.15, the distribution of diameter is shown. The blue line shows the result when only fibers were added. Particles around 10 to 20 µm show up most frequently. The red line shows the result when fibers and the device were added. Compared to the blue line, the peak in 10 to 20 µm decreased. Particles more than 20 µm also increased.
Fig. 15 The sizes of microfiber particles increased
The distribution of microfibers’ particle size are shown. The blue line shows the result when only fibers were measured. The red line shows the result when the mixture of fibers and the device were measured.
The distribution of microfibers’ particle size are shown. The blue line shows the result when only fibers were measured. The red line shows the result when the mixture of fibers and the device were measured.
Summary
Every component of our device was purified and assessed. All components worked as designed and we succeeded in creating our device "SONOBE". After we completed the device "SONOBE", we mixed them with microfibers and they actually made fibers precipitate. It requires further experiment to assess this effect, but we have suggested one possible way to solve the microplastic problem.
References
1 Moon, H., Lee, J., Min, J., and Kang, S. (2014).Developing genetically engineered Encapsulin protein cage nanoparticles as a targeted delivery nanoplatform.
Biomacromolecules 15, 3794–3801. 2 Weber, J., Petrović, D., Strodel, B., Smits, S.H.J., Kolkenbrock, S., Leggewie, C., and Jaeger, K.E. (2019).
Interaction of carbohydrate-binding modules with poly(ethylene terephthalate).
Appl. Microbiol. Biotechnol. 103, 4801–4812. 3 Bilal DEMİREL1, Ali YARAȘ, Hüseyin ELÇİÇEK (2011).
Crystallization Behavior of PET Materials.
BAÜ Fen Bil. Enst. Dergisi Cilt 13(1) 26-35.
