▼ DESIGN ▼

Overview

Polyethylene terephthalate (PET) is a widely used plastic that persists in aquatic environments due to its chemical stability. Traditional recycling methods are energy-intensive and inefficient. Our goal is to develop a biologically contained, modular system that accelerates PET degradation using engineered E. coli expressing surface-displayed PETase variants.^[1]

Project Design

Genes used	Part ID
HotPETase	BBa_25OAK6TB
CsgA-(G4S)2-HotPETase	BBa_252IGL39
HotPETase-(G4S)2-BaCBM2	BBa_25IHX9R7
CsgA-(G4S)2-HotPETase-(G4S)2-BaCBM2	BBa_251QH76Y

Units

HotPETase

Overview

HotPETase is a thermally stabilized and more efficient successor to the pioneering enzyme IsPETase, which was discovered in the plastic-eating bacterium Ideonella sakaiensis.

While the discovery of IsPETase proved that nature could evolve an enzyme to break down PET plastic, its practical application was limited by its low stability and rapid loss of function at the elevated temperatures where PET becomes malleable. HotPETase was specifically engineered to overcome this weakness, retaining its hydrolytic capabilities at temperatures above 70°C.^[2] This enhanced capability positions HotPETase as a promising tool for the efficient depolymerization of plastic waste, paving the way for a truly circular plastic economy.

HotPETase is the simplest unit of our project, consisting solely of the engineered enzyme.

HotPETase is an engineered enzyme created through directed evolution to efficiently break down polyethylene terephthalate (PET) plastic. Unlike its natural predecessor IsPETase, which is relatively fragile, HotPETase is thermally stable and operates effectively at temperatures above 70°C, retaining IsPETase's ability to hydrolyze PET but with significantly improved robustness.^[3]

This is a critical advantage because PET plastic becomes amorphous at 70°C, making it much easier to facilitate enzymatic depolymerization into its monomeric building blocks

Structure

There are many structural features present in HotPETase that allow it to break down PET polymers effectively. Like its parent enzyme IsPETase, HotPETase is a serine hydrolase. It features a catalytic triad composed of three amino acids: serine, histidine, and aspartate (Ser160, His237, Asp206), forming a proton-relay network at its active site.

Moreover, the active site contains key residues like Trp185 that interact with the aromatic rings of the PET substrate. The side chain of Trp185 is highly flexible, which allows the enzyme to accommodate and position the rigid PET polymer, resulting in enhanced substrate binding and flexibility.^[4]

Mechanism

PET degradation follows a two-step serine hydrolase mechanism. The first step is acylation, where the serine nucleophile attacks the carbonyl carbon of the PET ester bond, leading to the cleavage of the polymer and the formation of a covalent acyl-enzyme intermediate.

The second step, deacylation, is when the water molecules hydrolyze this intermediate, releasing the soluble product (MHET, BHET, TPA) and regenerating the free enzyme.^[3]

CsgA-HotPETase

Overview

The CsgA-HotPETase consists of the CsgA protein and HotPETase joined by a (G4S)2 linker. CsgA-HotPETase is a fusion construct that enhances the PET plastic degradation capabilities of the HotPETase. It utilizes the surface-display abilities of CsgA, a curli fiber subunit from E. coli, with the thermostable enzymatic activity of HotPETase, anchoring the enzyme onto the CsgA curli fibers across the bacterial surface.

Structure

CsgA is a gene-encoded protein that consists of 22 amino acids at the N-terminus, which is cleaved off during secretion. It is then followed by five repeating subunits (R1-R5), with each being 19 to 23 amino acids long, forming the amyloid core.^[5] Lastly, there is a C-terminal segment for interacting with CsgG pores and CsgB nucleator. R1 and R5 are responsible for amyloid fibril formation while R2-R4 repeats assist in maintaining the amyloidogenicity of CsgA. It is speculated that the stable steric zipper structures formed by the spine segments of the R1 and R5 repeats allow the core of the CsgA fibrils to be structured and result in the cross-β architecture of the fibrils^[6][7], which stabilizes and structures the biofilm. This allows for curli fiber formation to occur under a variety of conditions.^[6][8]

Mechanism

The signal sequence of CsgA is found at the N-terminus of the CsgA, which acts as a molecular address tag in order to be directed towards the biogenesis pathway in E. coli. Once the CsgA is folded and guided through the CsgG pores, the signal sequence is cleaved off and leaves CsgA ready for assembly. Afterwards, the CsgA is nucleated by CsgB, which helps accelerate CsgA aggregation^[9][10][11] to form the curli fibers. CsgF then localizes CsgB,^[12] positioning the CsgB correctly on the outer membrane for better spatial organisation and to attain full functionality. CsgF also mediates the protease resistance of CsgB, which correlates to its ability to nucleate CsgA.^[7][13]

Interaction with HotPETase

As mentioned, the main purpose of the CsgA gene is for curli fiber formation. The (G4S)2 linker genetically fuses the HotPETase to the C-terminal end of CsgA, allowing the construct to be incorporated into the extracellular curli matrix. When the CsgA undergoes nucleation, the entire fusion protein is incorporated as a result and remains catalytically active on the bacterial surface,^[14] maximising the HotPETase's surface exposure and more importantly, effectively transforming the E. coli cell into a whole-cell biocatalyst. In addition, this feature of CsgA enables a higher likelihood of PET binding to the HotPETase, resulting in a greater rate of degradation.

BaCBM2-HotPETase

Overview

The BaCBM2-HotPETase consists of the BaCBM2 protein and HotPETase joined by a (G4S)2 linker. BaCBM2 is a gene-encoded carbohydrate-binding module (CBM) that was originally found in Bacillus anthracis. The gene's primary function in our project is to enhance the enzyme's affinity to PET,^[15] tightly anchoring the enzyme to PET to increase the local concentration of the enzyme on the substrate surface and thereby significantly boosting the overall degradation efficiency.

Although it is called a carbohydrate-binding module, BaCBM2 has been shown to bind strongly to synthetic polymers, with a particularly high affinity for crystalline PET surfaces.^[15] We plan to utilize this protein in our project to minimize the spatial gap between HotPETase and PET, improving the contact between the enzyme and the PET. This can significantly accelerate binding and subsequently the overall hydrolysis rate.

Structure

The BaCBM2 protein has a planar, platform-like binding surface, revealed by molecular dynamics simulations. This surface features 3 tryptophan residues, namely Trp9, Trp44, Trp63, to form an aromatic triad.^[15] Trp9 and Trp44 assist in initiating contact and anchoring the interface between the PET surface, while Trp63 alongside Asn64 extends the contact surface of the interaction. Trp63 also allows the CBM to orient the triad parallel to the surface of the PET to better facilitate enzyme-substrate alignment.^[15]

Mechanism

This aromatic triad is crucial for binding. Trp9 and Trp44 interact with the phenyl rings in the PET polymer through hydrophobic and π-π stacking interactions. These interactions are further stabilized by additional hydrogen bonds formed by surrounding polar amino acids located near the aromatic triad.^[15] The optimum ratio of hydrophobic to polar interactions at the interface determines the strength of the CBM's binding to PET.^[15][16]

The critical role of tryptophan residues was confirmed experimentally via tryptophan quenching measurements,^[15] where the addition of PET nanoparticles to the CBM increased quenching of the tryptophan fluorescence, indicating a direct interaction between them.

The binding of PET also induces a conformational change to the BaCBM2 itself, allowing Trp63 and Asn64 to access the interface. As the aromatic triad is arranged on a flat, planar surface, when the CBM aligns parallel to the PET surface, simulations^[15] reveal that these tryptophan residues can simultaneously engage in π–π stacking with the phenyl rings of PET, which increases binding strength and stability.

This average CBM-PET potential energy of BaCBM2 is also significantly lower than that of BaCBM5, which has a very hydrophobic contact surface, explaining that the increased van der Waals' forces contribute significantly to binding strength and thus the degree of immobilization as well.^[15]

Interaction with HotPETase

The general principle is to create a fusion protein where the BaCBM2 domain serves as an "anchor" that tightly binds the entire enzyme to the PET surface. This anchoring gives the catalytic domain more opportunities to break the polymer's ester bonds, significantly enhancing the degradation rate.

Related Studies

A fusion protein was constructed linking BaCBM2 to a cutinase (Tfuc) from Thermobifida fusca. This fusion protein demonstrated a 2.8-fold increase in PET film degradation efficiency compared to the cutinase alone.^[17]

Other studies have successfully fused different CBMs with high PET affinity to the PET hydrolase LCC_ICCG, resulting in an approximately 5-fold enhancement of its activity on PET powder.^[18]

BaCBM2-HotPETase-CsgA

Overview

To synergize the surface display capabilities of CsgA and the enhanced PET affinity of BaCBM2, a fusion construct of BaCBM2, HotPETase, and CsgA – fused by (G4S)2 linkers – was designed.

We plan to utilize this construct in our project to determine the combined effect of the CsgA and BaCBM2 genes and compare their performance with the individual fusion constructs, testing whether the surface display density and PET binding affinity can simultaneously be increased.

Structure

The construct follows a linear arrangement: BaCBM2 is fused to the N-terminus of HotPETase, which ensures that the PET-binding surface of BaCBM2 exposed on the N-terminus is unaffected while avoiding steric hindrance.

CsgA is fused to the C-terminus of HotPETase, allowing the CsgA to self-assemble into curli fibers while displaying the entire construct externally. The amyloidogenic core remains intact as well.

Flexible (G4S)2 linkers are used between BaCBM2 – HotPETase and HotPETase – CsgA.

Mechanism

The functionality of the fused construct incorporates both mechanisms of the CsgA-HotPETase and BaCBM2-HotPETase. Together, the surface exposure of the enzyme is maximized by the CsgA curli fibers, which also localize the fusion construct to the bacterial surface. The construct's affinity for PET is increased by BaCBM2, anchoring the enzyme to the PET surface and enhancing the rate of hydrolysis. Lastly, the construct can remain active at elevated temperatures due to the thermostable nature of the HotPETase.

This integrated construct is expected to enhance PET degradation efficiency by combining substrate anchoring, surface localization, and thermal resilience.

(G4S)2 Linker

Overview

To maintain the structural integrity of the functional domains and avoid steric hindrance, flexible (G4S)2 linkers have been used in the fusion constructs.

Structure

The peptide sequence of (G4S)2 is GGGGSGGGGS. Due to the small side chain of glycine and the polarity and solubility of serine, the sequence forms a flexible polypeptide chain that allows the adjacent domains to move independently without clashing.

Mechanism

The separation of domains through the (G4S)2 linker is crucial as it prevents the domains from interfering with one another, also known as steric hindrance. Otherwise, the folding of the proteins involved (CsgA, BaCBM2, and HotPETase) will be altered. This ensures that each domain preserves its own functional integrity without inhibition.

In addition, the linkers reduce aggregation and misfolding during protein synthesis, allowing the protein to remain soluble and be properly folded.^[19] Here, "misfolding" refers to unintended disruptions to the tertiary structure, not the regulated formation of the CsgA fibrils.

The (G4S)2 linker is also short enough to maintain the functional domains' proximity, yet long enough to preserve their biological functions,^[20] allowing for conformational flexibility or modular rearrangement when needed.

Plasmid

Novelty

Our project's novelty lies within its sequential and innovative approach to enhance PET degradation, moving beyond the optimization of a single enzyme but including different gene combinations to enhance its ability through synthetic biology.

This project all began with HotPETase, an enzyme we have chosen because of its thermostable property. HotPETase has a significant advantage over other PETase variants.

After experimentation to confirm that HotPETase has been successfully expressed by E. coli and could effectively degrade PET polymers, we transformed a new gene combination to E. coli, which consists of a new gene BaCBM2 alongside our core gene HotPETase, forming the fusion protein Hot-PETase-BaCBM2. We hypothesised that the BaCBM2 gene could enhance the enzyme's ability to bind to the plastic surface and thereby boost efficiency. Through our wet lab experiments, we have successfully proved this concept in the end.

Our 3rd gene combination, CsgA-HotPETase, utilizes the CsgA gene, whilst the BaCBM2 gene was absent, we took this approach in order to know how each of the the individual genes (CsgA & BaCBM2 respectively) contributed to the increased performance of the HotPETase.

We chose to utilize the CsgA gene as it allows the expression of our HotPETase onto curli fibres, acting as a surface display for HotPETase for whole-cell biocatalysis of PET.

The combination of genes has eventually provided evidence to be effective in anchoring our enzyme onto E. coli, facilitating an easy recovery, reusable, and more natural approach in degrading plastic in aquatic environments.

Finally, all 3 genes were combined to form our final gene combination, which is also our target construct, CsgA-HotPETase-BaCBM2. This gene combination fuses the concepts of the previous 2 units, and experimentation was further conducted to identify possible synergies between the genes. This gene combination not only allows better binding, but also has the PETase be displayed on the curli fibres, thus achieving even greater effectiveness and efficiency.

After we have successfully engineered the modified PETase with higher degradation efficiency for PET plastic, we didn't just stop but to put the engineered enzyme in practice. We placed them in a bioreactor of our custom made 3D-printed cylindrical model connected to a water pump. Calcium alginate beads were also placed in a chamber inside the model. The purpose of the whole model was to degrade microplastics as well as to remove heavy metals from rivers and oceans with a novel approach.

We believe that this sequential approach stands out from other teams. Given the fact that we worked with iterations of gene constructs as we go, with each of which differing by only one gene with each other, the functional assessment of each gene construct step by step was made possible. No teams have taken this approach as they used the typical method of introducing all mutations simultaneously.

Protein Expression System for PET Biodegradation by Engineered E. coli

Construct design balances catalytic efficiency, substrate targeting, and biosafety: HotPETase provides a thermostable catalytic core, BaCBM2 increases hydrophobic interactions with PET, and CsgA enables immobilization through curli formation for reusability and containment.^[39][40]

Gene	Function
HotPETase	Thermostable catalytic core for PET degradation
BaCBM2	Enhances hydrophobic interactions with PET surface
CsgA	Enables immobilization through curli formation
(G4S)2 Linker	Provides flexibility and prevents steric hindrance

Flexible linkers separate domains to minimize steric interference and preserve individual folding with all coding sequences synthesized with E. coli‑preferred codons and assembled into a polycistronic or single‑ORF architecture depending on expression needs.^[41][42] Bench workflows include transformation into DH5α for plasmid propagation and BL21(DE3) for expression, induction optimization, His‑tag purification, and multi‑modal validation (SDS‑PAGE and gel electrophoresis).^[43][44]

For deployment, we use calcium‑alginate beads to remove heavy metals. Our mathematical models use Bernoulli and Darcy frameworks to guide bead packing and flow rate to optimize residence time versus throughput. (Refer to Model)

To test for functionality, we observed the mass change and absorbance change of samples containing our cells with PET. (For specific setup and data, refer to Experiments)

References

[1] 2024 Waseda Tokyo iGEM Team Design

https://2024.igem.wiki/waseda-tokyo/design/

[2] Berselli A, Menziani MC, Muniz-Miranda F. Structure and Energetics of PET-Hydrolyzing Enzyme Complexes: A Systematic Comparison from Molecular Dynamics Simulations. J Chem Inf Model. 2024 Nov 11;64(21):8236-8257. doi: 10.1021/acs.jcim.4c01369. Epub 2024 Oct 21. PMID: 39432831.

[3] Burgin, T., Pollard, B.C., Knott, B.C. et al. The reaction mechanism of the Ideonella sakaiensis PETase enzyme. Commun Chem 7, 65 (2024). https://doi.org/10.1038/s42004-024-01154-x

[4] Han, X., Liu, W., Huang, JW. et al. Structural insight into catalytic mechanism of PET hydrolase. Nat Commun 8, 2106 (2017). https://doi.org/10.1038/s41467-017-02255-z

[5] Wang, X., Hammer, N. D., & Chapman, M. R. (2008). The molecular basis of functional bacterial amyloid polymerization and nucleation. Journal of Biological Chemistry, 283(31), 21530–21539. https://doi.org/10.1074/jbc.m800466200

[6] Perov, S., Lidor, O., Salinas, N., Golan, N., Fligelman, E. T., Deshmukh, M., Willbold, D., & Landau, M. (2019). Structural Insights into Curli CsgA Cross-β Fibril Architecture Inspire Repurposing of Anti-amyloid Compounds as Anti-biofilm Agents. PLoS Pathogens, 15(8), e1007978. https://doi.org/10.1371/journal.ppat.1007978

[7] Shewmaker, F., McGlinchey, R. P., Thurber, K. R., McPhie, P., Dyda, F., Tycko, R., & Wickner, R. B. (2009). The functional curli amyloid is not based on in-register parallel Β-Sheet structure. Journal of Biological Chemistry, 284(37), 25065–25076. https://doi.org/10.1074/jbc.m109.007054

[8] Dueholm, M. S., Nielsen, S. B., Hein, K. L., Nissen, P., Chapman, M., Christiansen, G., Nielsen, P. H., & Otzen, D. E. (2011). Fibrillation of the Major Curli Subunit CsgA under a Wide Range of Conditions Implies a Robust Design of Aggregation. Biochemistry, 50(39), 8281–8290. https://doi.org/10.1021/bi200967c

[9] Dunbar, M., DeBenedictis, E., & Keten, S. (2019). Dimerization energetics of curli fiber subunits CsgA and CsgB. Npj Computational Materials, 5(1). https://doi.org/10.1038/s41524-019-0164-5

[10] Hammer, N. D., Schmidt, J. C., & Chapman, M. R. (2007). The curli nucleator protein, CsgB, contains an amyloidogenic domain that directs CsgA polymerization. Proceedings of the National Academy of Sciences, 104(30), 12494–12499. https://doi.org/10.1073/pnas.0703310104

[11] Hammar, M., Bian, Z., & Normark, S. (1996). Nucleator-dependent intercellular assembly of adhesive curli organelles in Escherichia coli. Proceedings of the National Academy of Sciences, 93(13), 6562–6566. https://doi.org/10.1073/pnas.93.13.6562

[12] 2024 Wasedo Tokyo iGEM Team Model

https://2024.igem.wiki/waseda-tokyo/model/

[13] Nenninger, A. A., Robinson, L. S., & Hultgren, S. J. (2009). Localized and efficient curli nucleation requires the chaperone-like amyloid assembly protein CsgF. Proceedings of the National Academy of Sciences, 106(3), 900–905. https://doi.org/10.1073/pnas.0812143106

[14] 2021 Yucai_SZ iGEM Team Engineering

https://2021.igem.org/Team:Yucai_SZ/Engineering

[15] Weber, J., Petrović, D., Strodel, B. et al. Interaction of carbohydrate-binding modules with poly(ethylene terephthalate). Appl Microbiol Biotechnol 103, 4801–4812 (2019). https://doi.org/10.1007/s00253-019-09760-9

[16] Møller, M. S., Bleckert, A., Jäckering, A., & Strodel, B. (2025). Unraveling the relationship between PET surfaces and their hydrolases. Trends in Biochemical Sciences. https://doi.org/10.1016/j.tibs.2025.05.002

[17] Zhang, Y., Liu, Z., Li, G., Fu, X., Zhang, Y., Wang, Z., Tian, Y., & Wu, J. (2022). [The application of carbohydrate binding module-Thermobifida fusca cutinase fusion protein in polyethylene terephthalate degradation]. PubMed, 38(1), 217–225. https://doi.org/10.13345/j.cjb.210036

[18] Dai, L., Qu, Y., Huang, J., Hu, Y., Hu, H., Li, S., Chen, C., & Guo, R. (2021). Enhancing PET hydrolytic enzyme activity by fusion of the cellulose–binding domain of cellobiohydrolase I from Trichoderma reesei. Journal of Biotechnology, 334, 47–50. https://doi.org/10.1016/j.jbiotec.2021.05.006

[19] Zhao, H. L., Yao, X. Q., Xue, C., Wang, Y., Xiong, X. H., & Liu, Z. M. (2008). Increasing the homogeneity, stability and activity of human serum albumin and interferon-α2b fusion protein by linker engineering. Protein Expression and Purification, 61(1), 73–77. https://doi.org/10.1016/j.pep.2008.04.013

[20] Chen, X., Zaro, J. L., & Shen, W. (2012). Fusion protein linkers: Property, design and functionality. Advanced Drug Delivery Reviews, 65(10), 1357–1369. https://doi.org/10.1016/j.addr.2012.09.039

[21] Studier, F., & Moffatt, B. A. (1986). Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. Journal of Molecular Biology, 189(1), 113–130. https://doi.org/10.1016/0022-2836(86)90385-2

[22] Mierendorf, R. C., Morris, B. B., Hammer, B., & Novy, R. E. (2003). Expression and purification of recombinant proteins using the PET system. Humana Press eBooks, 257–292. https://doi.org/10.1385/0-89603-485-2:257

[23] Hayat, S. M. G., Farahani, N., Golichenari, B., & Sahebkar, A. (2018). Recombinant Protein Expression in Escherichia coli (E. coli): What We Need to Know. Current Pharmaceutical Design, 24(6), 718–725. https://doi.org/10.2174/1381612824666180131121940

[24] Wu, J., & Filutowicz, M. (1999). Hexahistidine (His6)-tag dependent protein dimerization: a cautionary tale. Acta Biochimica Polonica, 46(3), 591–599. https://doi.org/10.18388/abp.1999_4131

[25] Robescu, M. S., & Bavaro, T. (2025). A comprehensive guide to enzyme Immobilization: All you need to know. Molecules, 30(4), 939. https://doi.org/10.3390/molecules30040939

[26] pET29-b+ plasmid

https://www.snapgene.com/plasmids/pet_and_duet_vectors_(novagen)/pET-29b(%2B)

[27] Amouyal, M. (2014). From adjacent activation in Escherichia coli and DNA cyclization to eukaryotic enhancers: the elements of a puzzle. Frontiers in Genetics, 5. https://doi.org/10.3389/fgene.2014.00371

[28] Lac Repressor-DNA Complex

https://www.ks.uiuc.edu/Research/Multiscale/

[29] Daber, R., Stayrook, S., Rosenberg, A., & Lewis, M. (2007). Structural analysis of LAC repressor bound to allosteric effectors. Journal of Molecular Biology, 370(4), 609–619. https://doi.org/10.1016/j.jmb.2007.04.028

[30] Becker, N. A., Peters, J. P., Lionberger, T. A., & Maher, L. J. (2012). Mechanism of promoter repression by Lac repressor–DNA loops. Nucleic Acids Research, 41(1), 156–166. https://doi.org/10.1093/nar/gks1011

[31] Camsund, D., Heidorn, T. & Lindblad, P. Design and analysis of LacI-repressed promoters and DNA-looping in a cyanobacterium. J Biol Eng 8, 4 (2014). https://doi.org/10.1186/1754-1611-8-4

[32] Barkley, M. D., Riggs, A. D., Jobe, A., & Bourgeois, S. (1975). Interaction of effecting ligands with lac repressor and repressor-operator complex. Biochemistry, 14(8), 1700–1712. https://doi.org/10.1021/bi00679a024

[33] Myung, S., Park, J., Han, J., & Kim, T. (2020). Development of the Mammalian Expression Vector System that can be Induced by IPTG and/or Lactose. Journal of Microbiology and Biotechnology, 30(8), 1124–1131. https://doi.org/10.4014/jmb.2003.03030

[34] Pan, H., Xie, Z., Bao, W., & Zhang, J. (2008). Optimization of culture conditions to enhance cis-epoxysuccinate hydrolase production in Escherichia coli by response surface methodology. Biochemical Engineering Journal, 42(2), 133–138. https://doi.org/10.1016/j.bej.2008.06.007

[35] Polumienko, A. L., Grigor’eva, S. P., Lukashova, L. I., Lobach, J. B., Riybaltovskaja, N. B., Lushnikov, A. A., & Domaradskij, I. V. (1985). Bacterial vectors which confer resistance to kanamycin. Biochemical and Biophysical Research Communications, 133(3), 1112–1118. https://doi.org/10.1016/0006-291x(85)91251-3

[36] Dunkle, J. A., Vinal, K., Desai, P. M., Zelinskaya, N., Savic, M., West, D. M., Conn, G. L., & Dunham, C. M. (2014). Molecular recognition and modification of the 30S ribosome by the aminoglycoside-resistance methyltransferase NpmA. Proceedings of the National Academy of Sciences, 111(17), 6275–6280. https://doi.org/10.1073/pnas.1402789111

[37] Lallemand, P., Leban, N., Kunzelmann, S., Chaloin, L., Serpersu, E. H., Webb, M. R., Barman, T., & Lionne, C. (2012). Transient kinetics of aminoglycoside phosphotransferase(3′)‐IIIa reveals a potential drug target in the antibiotic resistance mechanism. FEBS Letters, 586(23), 4223–4227. https://doi.org/10.1016/j.febslet.2012.10.027

[38] Toth, M., Frase, H., Antunes, N. T., Smith, C. A., & Vakulenko, S. B. (2010). Crystal structure and kinetic mechanism of aminoglycoside phosphotransferase‐2″‐IVa. Protein Science, 19(8), 1565–1576. https://doi.org/10.1002/pro.437

[39] Macek, B., Forchhammer, K., Hardouin, J. et al. Protein post-translational modifications in bacteria. Nat Rev Microbiol 17, 651–664 (2019). https://doi.org/10.1038/s41579-019-0243-0

[40] Duan, S., Zhang, N., Chao, T. et al. The structural and molecular mechanisms of type II PETases: a mini review. Biotechnol Lett 45, 1249–1263 (2023). https://doi.org/10.1007/s10529-023-03418-3

[41] Codon Optimisation Tool

https://www.novoprolabs.com/tools/codon-optimization

[42] Liu, Q., Schumacher, J., Wan, X., Lou, C., & Wang, B. (2017). Orthogonality and Burdens of Heterologous AND Gate Gene Circuits in E. coli. ACS Synthetic Biology, 7(2), 553–564. https://doi.org/10.1021/acssynbio.7b00328

[43] 2024 HKUST iGEM Engineering

https://2024.igem.wiki/hkust/engineering

[44] Brown, C.W., Sridhara, V., Boutz, D.R. et al. Large-scale analysis of post-translational modifications in E. coli under glucose-limiting conditions. BMC Genomics 18, 301 (2017). https://doi.org/10.1186/s12864-017-3676-8