Overview
Environmental protection has always been a popular topic in iGEM projects, especially environmental pollution, which has attracted much attention in recent years. In the course of this project, we've amassed a substantial repository of experiential knowledge that holds potential value for forthcoming iGEM teams. This project focuses on the reconstruction of the ancestral sequence of the PETase (ASR-PETase) and machine learning to design promising ASR-PETase variants to achieve the closed-loop recycling of PET as well as exploring future applications (e.g., open-loop recycling of PET, degradation of microplastics in vivo).
In addition, we use software such as AutoDock 4 version (v4.2.6) and Gromacs 2022.5 to explore the microscopic mechanisms of enzyme catalysis.
Wet Lab Contribution
New Data
PETase is a PET hydrolase that can depolymerize PET into monomers. However, the activity of PETase is limited by temperature ranges. To overcome this challenge, we used profile-HMM algorithm for protein sequences in silico and obtained the ancestral sequence of PETase (ASR-PETase) by FireProt (https://loschmidt.chemi.muni.cz/fireprotweb/) according to evolutionary nodes (Fig. 1).
We recognize the advantages of artificial intelligence (AI) through the application of models. It can extract complex, large amounts of unstructured data into abstract, high-level representations through neural networks. AI can reuse features, and as the number of layers deepens, it can acquire more abstract features, with higher capabilities and flexibility. We believe that AI can help improve the accuracy and efficiency of protein modification in the future. AI technology is based on the continuous learning ability of massive data and the intelligent exploration ability in the unknown space, which effectively meets the needs of the current synthetic biology engineering trial-and-error platform. Therefore, we believe that AI technology has great potential in the exploration of complex biological features and the design of living systems.
New Model
Protein language model is the migration application of various language models in the field of biochemistry. By inputting protein sequences and learning the underlying biochemical properties, secondary and tertiary structures and functional inherent laws in the sequences, the tasks of predicting protein structure, predicting protein function and generating new sequences can be completed. By learning from known protein data, ESM-1v discerns pivotal sites intricately tied to mutations, thereby enabling an efficient prognostication of the repercussions these mutations have on protein function.
Through the calculation of the model, the heat map of the replacement of sequence sites can be obtained. Based on the heat map, points that can be replaced are able to be obtained (Fig. 2).
We use the ESM-1v model and provide examples of its use to the public. ESM-1v model is a kind of protein language model. The Zero-Shot algorithm adopted by ESM-1v model enables us to predict protein mutations even when there are few experiments or the relevant information is missing. The process we provide for the use of the model could provide new ways to help future iGEM teams predict proteins.
We recognize the advantages of artificial intelligence through the application of models. It can extract complex, large amounts of unstructured data into abstract, high-level representations through neural networks. Artificial intelligence can reuse features, and as the number of layers deepens, it can acquire more abstract features, with higher capabilities and flexibility. We believe that AI can help improve the accuracy and efficiency of protein modification in the future. AI technology is based on the continuous learning ability of massive data and the intelligent exploration ability in the unknown space, which effectively meets the needs of the current synthetic biology engineering trial-and-error platform. Therefore, we believe that AI technology has great potential in the exploration of complex biological features and the design of living systems.
New Tool
To deeply analyze the microscopic mechanism of PET biodegradation, MD simulation was employed to demonstrate the driving force behind the increased degradation capacity. We used AutoDock 4 version (v4.2.6) for ligand-protein docking and GROMACS 2022.5 software for molecular dynamics simulations of ligand-protein complexes. We used VMD for WIN 64, version v1.9.4 (June 29, 2021) and PyMOL (TM) 2.5.5 - Incentive Product software to visualize the obtained data, which helped us analyze the mechanism of the increased enzyme activity.
The software we used helped us analyze the mechanism of enzyme catalysis from a microscopic perspective, so we recommend these softwares to future iGEM teams. We hope that they can also provide a new perspective for future iGEM teams to design variants, improve design efficiency, and shorten the experiment cycle.
New Information
First, PET waste was degraded into the homogeneous terephthalic acid (TPA) by a two-enzyme system, and then TPA was synthesized into dimethyl terephthalate (DMT) under H2SO4/CH3OH chemical conditions. Finally, DMT was directly synthesized into PET by condensation reaction with ethylene glycol (EG) under the chemical condition of Ti(Oi-Pr)4 (Fig. 3).
Through the PET repolymerization we have learned from the literature and confirmed, we hope that they can also provide a way for the future iGEM team. PETase and monomer TPA can not only be applied to the recycling of waste PET, but also replace the traditional PET degradation methods, thus reducing environmental pollution and reducing the oil consumption required for the production of initial TPA, and extending the recycling industry chain. It has a positive impact on a variety of scenarios.
At the same time, our wet experiment can provide inspiration for other teams working on PET plastics. The PET degradation products can not only achieve a closed cycle, but also can achieve an upgrade cycle. Considering that microplastics have been detected in human blood, we will make an inactivated injection to achieve the removal of microplastics in the body in the future. Waste plastics can not only be used as a raw material for PET production, but also as a raw material for synthetic graphite. It can also act as a reactant to generate potassium diformate and hydrogen.
We hope that these ideas will help to fully realize the recycling of PET and sustainable development in the future.
Human Practice Contribution
According to the content of our project, the team members produced popular children's science picture books on the themes of environmental protection and marine plastic waste. At the same time, we also made online micro-lessons about plastics. Our team organized international summit on environmental protection and sustainable development. We had set up a WeChat group chat for teams participating in Bioremediation circuit in China, and planed to keep the group chat for easier communication in the future. Our aim is to communicate our understanding of synthetic biology in simple language through diverse platforms. Through this approach, we aspire to enlighten a broader audience about both synthetic biology and the process of plastic degradation. It also conveys the idea of sustainable development (Fig. 4).
Reference
1. Huang, B.; Fan, T.; Wang, K.; Zhang, H.; Yu, C.; Nie, S.; Qi, Y.; Zheng, W.-M.; Han, J.; Fan, Z.; Sun, S.; Ye, S.; Yang, H.; Bu, D., Accurate and efficient protein sequence design through learning concise local environment of residues. Bioinformatics 2023, 39 (3), 122.
2. Zhang, T.; He, Q.; Xu, Z.; Suo, F.; Zhang, C.; Hu, Q., Effects of substrate on growth and lipid accumulation of Tribonema sp. FACHB-1786. Chinese journal of biotechnology 2020, 36 (11), 2478-2493.
3. Liu, Y.; Zhang, L.; Wang, W.; Zhu, M.; Wang, C.; Li, F.; Zhang, J.; Li, H.; Chen, Q.; Liu, H., Rotamer-free protein sequence design based on deep learning and self-consistency. Nature Computational Science 2022, 2 (8), 526-526.
4. Li, A.; Sheng, Y.; Cui, H.; Wang, M.; Wu, L.; Song, Y.; Yang, R.; Li, X.; Huang, H., Discovery and mechanism-guided engineering of BHET hydrolases for improved PET recycling and upcycling. Nature Communications 2023, 14 (1), 4169.
5. Ko, S.; Kwon, Y. J.; Lee, J. U.; Jeon, Y.-P., Preparation of synthetic graphite from waste PET plastic. Journal of Industrial and Engineering Chemistry 2020, 83, 449-458.
6. Zhou, H.; Ren, Y.; Li, Z.; Xu, M.; Wang, Y.; Ge, R.; Kong, X.; Zheng, L.; Duan, H., Electrocatalytic upcycling of polyethylene terephthalate to commodity chemicals and H2 fuel. Nature Communications 2021, 12 (1), 4679.
