Microplastics (MPs) are becoming a growing concern for environmentalists, with society’s integration of single-use plastics into daily life directly resulting in an accumulation of MPs in ecosystems across the world. One such ecosystem is our own human bodies. MPs and their nano-scale counterparts, nanoplastics (NPs), have been shown to accumulate in the liver, brain, and lungs, posing massive potential concerns for human health (Kim et. al, 2025). These health effects are being increasingly studied, and although the mechanisms and severity of MP accumulation are not yet fully understood, our team has approached the issue with the intent to prevent any potentially toxic side effects.
Figure 1: Chemical Structure of Polystyrene Monomer
The primary bioengineering challenges with polymers such as polyethylene, polypropylene, and polystyrene (PS) are their nonpolar structure and minimal biodegradability, causing them to be highly inert and persistent in the environment (Cai et. al, 2023). However, developments in de novo peptide design and phage display have increased the number of options for combatting MPs (Skariyachan et. al, 2022). As a result, iGEM teams have engineered various bacterial chassis to secrete/display polyethylene terephthalate hydrolases (PETases) and/or plastic binding peptides (PBPs).
The Design subteam’s preliminary research quickly led us to understand that MPs must travel to vital organs through the bloodstream. Most MPs reach the bloodstream by entering the body through oral ingestion, travelling through the gastrointestinal tract, bypassing the intestinal barrier and finally infiltrating the bloodstream (Li et. al, 2024). Considering this, our engineering team’s cycles were planned around mitigating MP interaction with the intestinal mucosae to prevent endocytosis and phagocytosis, among other things.
Figure 2: Mechanisms of Absorption of Microplastics in the Small Intestine (Prata, 2023)
Our ideation process for combatting MP accumulation has ultimately culminated in our proposed engineered probiotic, Plastipeutics. Essentially, we’ve designed an Escherichia Coli Nissle 1917 (EnC) capable of surface display of fusion proteins individually composed of an anchor protein, a flexible linker protein, and a PBP. This biological device not only allows EnC to bind MPs, but to aggregate multiple of them around itself, increasing the MPs' effective volume and limiting their capacity to pass through the intestinal epithelium.
Figure 3: Graphic Design of Plastibinder Unit
Created in https://BioRender.com
In starting the design process, our team’s first initiative was to brainstorm high level designs, or “big picture ideas,” for our solution. With the initial requirement that the small intestine would be our point of intervention, we formulated three designs based on different synthetic biology applications.
- Engineering macrophages present in Peyer's Patches (a component of the enteric immune system), to produce MP-degrading enzymes for degrading phagocytosed MPs.
- Engineering a probiotic bacteria to secrete MP-degrading enzymes into the mucosa/lumen.
- Engineering a probiotic to secrete and/or display PBPs to bind and aggregate plastics.
The first option, engineering an immune cell, seemed fairly intuitive. M-cells, epithelial cells foundational to the enteric immune system, are designed to transcytose antigens from the intestinal lumen into Peyer’s patches to be degraded. Modelling our solution after the COVID-19 vaccine, we theorized using a lipid nanoparticle to both transport mRNA safely through the GI tract and to target the mRNA delivery to M-Cells, which are primarily present in the ileum.
Figure 4: Pathway for Ingested Lipid Nanoparticle to Peyer's Patch
Created in https://BioRender.com
Alternatively, utilizing an engineered probiotic seemed viable, as it would allow us to address bloodstream entrance points beyond M-cells. The initial concept involved engineering the probiotic strain EnC to colonize the ileum and secrete PETases for degrading ambient MPs. We chose to model this design around PETase because of its thorough usage in literature and iGEM. However, our third design substitutes anchored PBPs in place of secreted PETase, making the EnC a microplastic binder that could potentially aggregate multiple MPs together.
Figure 5: Pathway for Probiotic to Ileum
Created in https://BioRender.com
With these initial designs established, our team began to break down our assumptions and develop constraints. Many of our considerations were with respect to the environment of the small intestine. For example, we wanted to ensure our solution functioned in the mucus-filled space and that it would account for all the ways MPs move through the lumen. Also, we noted the need for an enteric coating to bypass the stomach, and accounted for the ileum’s neutral pH (7.4) and human body temperature (37°C) in our models. Furthermore, we factored in the size scales of MPs, NPs, peptides, bacteria, epithelial cells, and the lumen into our modelling and decisions.
In fact, our research led us to understand that the size of MPs and NPs are integral to the nature of their interaction with epithelial cells. For example, MPs larger than 150μm will not be transcytosed, serving as an upper bound for our design, and a lower bound can be derived from the fact that MPs smaller than 4μm are less likely to accumulate in tissue (Li et. al, 2024). Although the mechanistic relationship between MPs and the intestinal epithelia is not fully understood, compartmentalizing interactions such as tissue penetration, transcytosis, and endocytosis allowed us to better understand what size ranges pose the largest threat.
With input from our advisors, our team decided to further develop the macrophage design. From our interview with Jingjiao Guan, we determined that, to implement our solution, we’d need to ensure that the mRNA vector wasn’t phagocytosed, PETase would be produced in a lysosome, PETase would survive in the acidic lysosome, and PETase would function properly in the phagolysosome upon MP phagocytosis.
Figure 6: M-cell Modification Process
Created in https://BioRender.com
At this point, a couple of obstacles stood in our way. From our interview, we’d learned that macrophages are notoriously challenging to transfect. Also, the mechanisms of phagocytosis are highly complex, and attempting to express PETase in phagolysosomes would be challenging since visualizing internal macrophage activity is very challenging. Considering these factors, our team chose to pivot away from immune cell engineering.
With this, we began pursuing a probiotic capable of degrading microplastics because of the extensive research done with PETase enzymes; we believed that attempting to apply PETase in intestinal conditions would prove fruitful. However we later came upon a key flaw with this design: PET catabolysis by most strains of PETase produces byproducts that are toxic to enterocytes (Syren et. al, 2024). This dissuaded us from attempting any PETase-based solutions, as- unless our system was capable of quickly degrading MPs into monomers- the potential toxicity would be even more harmful than MPs themselves. With that said, made our final shift to a plastic-binding probiotic.
Figure 7: Design Requirements for Degradation/Aggregation Probiotic
Created in https://BioRender.com
With our overall design selected, our first specification was ensuring that our PBPs were capable of functioning in the ileum. Our modelling team utilized a Propka model to ensure that our PBPs would have the proper net charge at the pH of the ileum. We noticed that the Histag was actually the limiting factor for suitable pH, which we can leverage in future designs where purification isn’t necessary.
In selecting our chassis, we defaulted to EcN due to its overwhelming characterization among probiotic bacteria strains. However, it’s important to note that EcN has limited secretion pathways, which may necessitate our shift to another chassis later on (Lynch et. al, 2023).
In deciding the best MP to model our engineering decisions around, we initially thought of polyethylene terephthalate since our degradation design centered around PETase. However, after shifting to a PBP mechanism, our team realized that PS was advantageous because most research on the health effects of MPs employs PS as the model particle. This means that, although the health implications of PS are well understood, those results can’t be extrapolated to other MPs (Jasinski et. al, 2024). Furthermore, we lacked the tools to create our own microplastics with micron-scale precision, so we decided to purchase plastic microplasticles instead. Since PS MPs are much less expensive than other polymers MPs, we were further compelled to model our processes after it.
The goal of our first cycle was to prove that our PSBPs bind to MPs. To this end, we began by searching for polystyrene binding peptides (PSBPs), identifying multiple broader categories in literature including deactivated enzymes, nanobodies, and simple polypeptides. Deactivated enzymes were more appealing when modelling around PET because we could utilize deactivated PETase, but since PS was our new model MP, we strayed away from that option. Nanobodies were very compelling due to their adaptability, but the lack of literature on PS-binding nanobodies dissuaded us from pursuing them. Thus, we chose to employ polypeptides, as they allowed us more flexibility in design and modelling.
For this cycle, we selected PSBPs (
Beyond selecting PSBPs, we considered other coding sequences for the fusion protein, including reporters, surface display/anchor proteins, and linker proteins. The anchor and linker proteins were part of our high level design, but we realized that, for proving binding, they were not essential. As a result, we omitted the anchor protein and, to see if the linker impacted binding/expression, we created generator units with and without linkers. Our chosen linker (BBa_K4586019) is a serine glycine linker from the parts registry.
With our coding sequences selected, we designed a PSBP generator skeleton to insert them into for expression. Our generator’s overall skeleton was guided by the NEB PURExpress® In Vitro Protein Synthesis Instruction Manual’s template DNA sequence, as we planned to their in vitro and in vivo T7-based protein expression systems.
Figure 8: Required elements for template DNA (NEB, 2025, pg. 4)
Staying within the template DNA’s restrictions, we included a 5’ UTR (
Figure 9: SBOL Designs for PSBP Generator Units with and without Linker Protein
We inserted our generator device into a Twist kanamycin resistance backbone (
Figure 10: Snapgene Plasmid Map of BBa_25BD5FF2
After ordering the synthetic plasmids from Twist Bioscience, we were informed that two of our three vectors failed to synthesize. After receiving the full report from Twist, we realized the issue was that our 5’ UTR had ten consecutive adenine repeats. Learning this, we researched 5’ UTR options and found that a Poly-A spacer of five base pairs long would also be sufficient (Takahashi et. al, 2013) so we used a truncated 5’ UTR (
Chronologically, this is when we shifted our focus from PET to PS. Also, this was when we chose to order generators with and without linkers, which allowed our assays to test PSBP-linker fusion proteins and pure PSBPs as experimental groups.
Figure 11: Snapgene Plasmid Maps for BBa_258USD40 and BBa_251QKY3O
The generators with and without linkers for five different PSBPs and a negative control were synthesized by TWIST Bioscience. This left us with five PSBP generators, five PSBP + linker generators, our negative control vector, and the PET binding peptide generator from our prior order. See our parts page for more information!
Our team attempted expression of 11 binding units, PET Binding Unit B and PS Binding Units A through PS Binding Units J. We tried two methods of expression, referred to as plans A and B. Plan A involved using the NEBExpress Cell-free E. coli Protein Synthesis System (NEB #E5360S), using our designed plasmids that contained the coding sequence to express PS Binding Unit A, C, E, F, and G. We ran plan B in parallel, in which we transformed chemically competent T7 Express E. coli cells using PS Binding Unit Generator A through PS Binding Unit Generator J (See the Parts section of our Wiki!) From these transformed cells, we plated them on LB-KAN in 1:1 and 1:10 dilutions and allowed colonies to grow overnight at 37°C. The following morning, we moved the plates to the fridge to prevent satellite colonies from growing. We then made 25mL overnight cultures in LB media supplemented with Kanamycin. We took 5mL per construct of the overnight culture and made glycerol stocks for stable storage of transformed T7 express cells for minipreps for more DNA for future use. The remaining 20mL was split in half for small scale expressions — 10mL for IPTG induction and 10mL for non-induced controls. The cell cultures were grown at 37° C until an optical density, OD600, between 0.4-0.8 was reached. Then, IPTG or water (negative control) was added to reach a final concentration of 0.4 mM. Afterwards, 8mL of culture was centrifuged down in respective Eppendorf tubes to produce bacterial cell pellets which we then froze at -80° C, for future use.
When ready, the team defrosted the cell pellets, lysed them using sonication and ran the noninduced and induced cell cultures in addition to the in-vitro expressed PS Binding Units A, C, E, F, and G in a 16% tricine gel. This gel percentage was chosen due to this percentage having peptide resolution down to 1kDa. This gel was run for half the time to ensure the ~3kDa peptides didn’t run off as the ladder only went down to 10kDa and was stained with a silver nitrate. Silver nitrate as a stain is very sensitive as it can detect protein concentrations as low as 0.1 ng/band (Jiang et al). This was followed by Coomassie blue dye to stain anything silver nitrate could not. The team observed highly visible bands in the 10-150kDa range, but the team was unable to locate a band in the expected 2.5-3kDa range, which is a characteristic of our peptides.
As a result of this setback, we hypothesized that our peptide was aggregating to itself or other proteins, experienced possible degradation, or it bound to the walls of the plastic pipettes. The result of our contribution, which verified that the T7 Max promoter (BBa_K4207000) works and does increase expression, made the possibility of our device not expressing at unlikely. Therefore, to test our hypotheses of whether degradation or aggregation was the problem, the team completed another 16% tricine gel running PS Binding Units A, H, and J in the gel. However, we decided to purify the cell lysate and added it to our test. We ran this 16% tricine gel with non-purified and purified induced and non-induced cells containing the DNA of the binding units mentioned above (fig.12,13), in addition to an anti-His western blot (fig.14) to confirm if aggregation took place. In both tests we observed a strong ladder indicating a successful positive control, but we did not observe any bands in the 2.5-3 kDa region where we expected our peptide.
Figure 12: 16% Tricine Gel with Silver Stain, High and Low Contrast
Figure 13: 16% Tricine Gel with Silver and Coomassie Stain
Figure 14: 16% Tricine Gel for anti-His Western Blot
Figure 15: 16% Tricine Gel with Ponceau Stain
Our team faced the problem of no visible signs of expression in the induced, noninduced, and in-vitro lanes. We proceeded to stain with a mixture of Coomassie and silver nitrate due to evidence suggesting that peptides or proteins found that are more more basic tend to not be as efficiently stained [with silver nitrate] in comparison to acidic counterparts (Chevallet et al., 2006).
At this point, our team developed two major hypotheses regarding our results of the tricine gel. We speculated that the peptides were not expressing or that they were expressing but the peptides were aggregating, increasing their mass and appearing in a band within the 10-150kDa range. Due to our confidence in our design, and the three different ways of expression, we decided to investigate alternative ways to test our hypothesis that the peptide was expressing, but aggregating. Working with our advisor, Alana Chang, we chose to test our hypothesis for peptide expression through an anti-His western blot, due to the peptide having an engineered His-tag. Our design initially included this His-tag for the purpose of purification using nickel columns. We then ran the tricine again but with purified versions (induced and noninduced cell cultures) of PS Binding Units A, B, and H, instead of all 10 constructs. We decided to purify to attempt to remove undesired proteins in the cell lysate. We chose PS Binding Units A, H, and J due to their more favorable absorbency values read at 595 nm (the wavelength for proteins) from our Bradford assay.
Utilizing our glycerol stocks, we re-streaked PS Binding Units A, H, and J on LB-Kan plates followed by inoculating 20mL of LB media supplemented with Kanamycin with single clones derived from the plates. As done previously, we divided the cell cultures of PS Binding Units A, H, and J in half and induced half with IPTG and left the other half noninduced. Of these 2x10 mL of induced/noninduced cells, we made 4 cell pellets: induced for whole cell lysate, induced for purification, noninduced for whole cell lysate, non-induced for purification. To begin purification of the respective samples, we lysed the pellets using sonication and used the NEBExpress Ni Spin Columns Quick Start Protocol (NEB #S1427) to purify the peptides. Using our purified strains of PS Binding Units A, H, and J we ran another 16% tricine gel, transferred the peptides/proteins to a PVDF membrane and did an anti-His western blot to attempt to see the presence of our peptides followed by ponceau staining the membrane to visualize all transferred peptides. The lanes in figures 14 and 15 are of unpurified induced and noninduced whole cell lysates (BLUE), as well as purified induced and noninduced samples (RED).
Our team observed no bands present upon anti-His western blot in any of the lanes (fig.14) for the 3 constructs hinting to failure of expression of our peptides and not potential aggregation like we hypothesized from our initial silver/Coomassie-stained tricine gel. The ladder was observed in this blot as the antibody was purified using a His tag. Upon ponceau staining, we observed stronger banding in the whole cell lysate lanes (fig.15) which is expected as there should be more protein present in these lanes.
With these results, our team continued to question whether expression was successful and hypothesized that 1) expression never occurred or 2) degradation of our peptides post-expression may be occurring or 3) expression is occurring, but the peptides are being lost due to usage of plastic pipette tips and Eppendorf tubes, as the peptide was modelled to have high binding affinity to plastics. All 3 hypotheses result in the absence of our peptides in the anti-His blot. With the various unknowns of whether we experienced an expression issue, or our peptides binding to plastic equipment, the team decided to alter the design with the addition of 2 fusion proteins comprised of Super Folding Green Fluorescent Protein (BBa-k5105008) and Small Ubiquitin-like Modifier (BBa_25USM2Y8) derived from Brachypodium distachyon, to increase the mass so we can run SDS-PAGE gels instead of Tricine gels, which are more time-intensive and difficult, as well as to increase solubility respectively.
Having determined that our peptides likely failed to express, our team chose to modify the designs to maximize expression. With guidance from our advisor, Alana Chang, we added a bdSUMO tag (
Figure 16: Snapgene Linear Map for BBa_25EYHQQP
3 new designs of our biological devices were constructed: PS Binding Unit K Generator, PS Binding Unit L Generator, and PS Binding Unit M Generator. All of the devices were designed by our iGEM team and then synthesized by Twist Bioscience.
Our team will attempt in-vitro expression of PS Binding Unit A, B, and D using the NEBExpress Cell-free E. coli Protein Synthesis System (NEB #E5360S). This will be followed by purification using NEBExpress Ni Spin Columns (NEB #S1427). We then will utilize the SpectraMax iD5 for a kinematic assay over 10 hours to verify expression. The results will be presented at the judging session.
Testing will be continued after the wiki freeze with our new iteration of binding units. First, we plan to cleave the 12x his-tag and bdSUMO tag by using a SUMO protease solution, leaving a fusion protein composed of a glycine serine linker, a PS binding peptide, and the superfolded GFP. We had primarily considered centrifugation and vacuum filtration, with centrifugation affording more control while vacuum filtration was more cost and time-efficient. We ultimately chose to do the vacuum filtration for the binding assay. To seal the Buchner funnel and filter, pressure will be applied to the surface for 30 seconds while the vacuum is on. 250 µL of HEPEs buffer solution will be added to a 15mL test tube with 9.8 mL of DI water making approximately 10 mL of a buffer solution with a pH of 7.4. Separate 10 µL, 30 µL, 50 µL, 100 µL, and 150 µL peptide stock of each PS Binding Unit will be added to a 10 mL buffer solution to test binding characteristics. To detect the yield of the PS Binding Units that bound to the microplastics a Bradford assay will be used. The result of this test will be presented at the judging session if the build process of the PS Binding Units is successful. However, based on our modelling, the vacuum filtration binding assay shows promise. This is supported in section 8 of the modeling page, which created a Markov State Machine (MSM) model that computes the converged probability of the number of peptides bound at the surface interface given unfavourable probabilities of binding and unbinding and the total number of peptides.
Once we’re able to consistently express our peptides and prove they bind to microplastics, future engineering cycles would focus on working towards a functional, safe, and effective product. To realize a functional product, we’d need to slowly build up to implementing our solution in the human small intestine. This would involve proving that surface display of the fusion protein is occurring, testing our biological system in ileum-like conditions, and finding how long and the concentration at which our chassis can colonize the gut microbiome. Making our product safe would primarily involve adding a safety module/kill switch to prevent the bacteria from developing antimicrobial resistance once excreted. Increasing the efficacy of our product can be done in many ways. Accurately quantifying and characterizing microplastics requires specific technology and skills our team lacks, which hinders our ability to accurately interpret wet lab results for binding assays. In the dry lab, our modelling team hopes to determine association and dissociation constants between our peptides and PS, ultimately towards computationally modelling MP-PSBP binding dynamics in the gut and allowing further optimize our peptides characteristics.
"Plastipeutics" represents a novel approach to mitigating the internal threat of microplastic poisoning. By engineering a probiotic to display PBPs, we aim to create a system that aggregates microplastics in the gut to be excreted. Our first engineering cycle has successfully established that our computationally derived peptides bind to microplastics. With a firm understanding of focuses for future cycles, we are well-positioned to test and refine this system into a viable therapeutic solution.
Alshehri, A.S., Bergman, M.T., You, F. and Hall, C.K. (2024). Biophysics-Guided Uncertainty-Aware Deep Learning Uncovers High-Affinity Plastic-Binding Peptides. Digital Discovery, [online] (2). doi:https://doi.org/10.1039/d4dd00219a.
Chevallet, M., et al. (2006) Silver staining of proteins in polyacrylamide gels. PMC. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC1971133/ (Accessed: 3 October 2025)
Deich, C., Cash, B., Sato, W., Sharon, J., Aufdembrink, L., Gaut, N.J., Heili, J., Stokes, K., Engelhart, A.E. and Adamala, K.P. (2023). T7Max Transcription System. Journal of Biological Engineering, 17(1). doi:https://doi.org/10.1186/s13036-023-00323-1.
Dhoriyani, J., Bergman, M. T., Hall, C. K., & You, F. (2025). Integrating biophysical modeling, quantum computing, and AI to discover plastic-binding peptides that combat microplastic pollution. PNAS nexus, 4(1), pgae572. https://doi.org/10.1093/pnasnexus/pgae572
Emenike, E.C., Okorie, C.J., Ojeyemi, T., Egbemhenghe, A., Iwuozor, K.O., Saliu, O.D., Okoro, H.K. and Adeniyi, A.G. (2023). From oceans to dinner plates: The impact of microplastics on human health. Heliyon, [online] 9(10), p.e20440. doi:https://doi.org/10.1016/j.heliyon.2023.e20440.
INSTRUCTION MANUAL PURExpress ® In Vitro Protein Synthesis. (2025). [online] NEB. Available at: https://www.neb.com/en/-/media/nebus/files/manuals/manuale6800_e3313_e6840_e6850.pdf?rev=60ba1e078b384d21bf0d5bbfd00495b9&hash=9D592FD6C6E23B68E8C5A4408E25B171
Jasinski, J., Matthias Völkl, Wilde, M.V., Valérie Jérôme, Fröhlich, T., Freitag, R. and Scheibel, T. (2024). Influence of the polymer type of a microplastic challenge on the reaction of murine cells. Journal of Hazardous Materials, 465, pp.133280–133280. doi:https://doi.org/10.1016/j.jhazmat.2023.133280.
Jiang, Y., Zheng, L., Lin, L., Lin, S., Xu, K., Deng, S. & Zhang, Q. (2022) Modification in Silver Staining Procedure for Enhanced Protein Staining. BioMed Research International, 2022, Article ID 6243971. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC9122722/ ( Accessed: 4 October 2025)
Kim, D., Kim, D., Kim, H.-K., Jeon, E., Sung, M., Sung, S.-E., Choi, J.-H., Lee, Y., Kang, K.-K., Lee, S. and Lee, S. (2025). Organ-specific accumulation and toxicity analysis of orally administered polyethylene terephthalate microplastics. Scientific Reports, [online] 15(1). doi:https://doi.org/10.1038/s41598-025-91170-1.
Li, Y., Chen, L., Zhou, N., Chen, Y., Ling, Z. and Xiang, P. (2024). Microplastics in the human body: A comprehensive review of exposure, distribution, migration mechanisms, and toxicity. Science of The Total Environment, [online] 946, pp.174215–174215. doi:https://doi.org/10.1016/j.scitotenv.2024.174215.
Lynch, J.P., González-Prieto, C., Reeves, A.Z., Bae, S., Powale, U., Godbole, N.P., Tremblay, J.M., Schmidt, F.I., Ploegh, H.L., Kansra, V., Glickman, J.N., Leong, J.M., Shoemaker, C.B., Garrett, W.S. and Lesser, C.F. (2023). Engineered Escherichia coli for the in situ secretion of therapeutic nanobodies in the gut. Cell Host & Microbe, 31(4). doi:https://doi.org/10.1016/j.chom.2023.03.007.
Prata, J. C. (2023) ‘Microplastics and human health: Integrating pharmacokinetics’, Critical Reviews in Environmental Science and Technology, 53(16), pp. 1489–1511. doi: 10.1080/10643389.2023.2195798.
Skariyachan, S., Taskeen, N., Kishore, A.P. and Krishna, B.V. (2022). Recent advances in plastic degradation – From microbial consortia-based methods to data sciences and computational biology driven approaches. Journal of Hazardous Materials, 426, p.128086. doi:https://doi.org/10.1016/j.jhazmat.2021.128086.
Takahashi, S., Furusawa, H., Ueda, T. and Okahata, Y. (2013). Translation Enhancer Improves the Ribosome Liberation from Translation Initiation. Journal of the American Chemical Society, 135(35), pp.13096–13106. doi:https://doi.org/10.1021/ja405967h.
Wikipedia contributors. (2025, September 2). Polystyrene. In Wikipedia, The Free Encyclopedia. Retrieved 20:32, September 22, 2025, from https://en.wikipedia.org/w/index.php?title=Polystyrene&oldid=1309133961
Ximena Lopez-Lorenzo, Hueting, D., Bosshard, E. and Per-Olof Syrén (2024). Degradation of PET microplastic particles to monomers in human serum by PETase. Faraday Discussions, [online] 252. doi:https://doi.org/10.1039/d4fd00014e.
AI Statement: We’d like to acknowledge the use of Google’s Gemini in drafting of this page, although the final product was composed by the FSU iGEM Engineering team.