Firstly, our team created the GLP-1 part (BBa_K5083000) for the first time, tested and characterized its expression in Escherichia coli. We added the PelB signal peptide to its N-terminus, successfully secreting GLP-1 extracellularly (BBa_K5083001). This achievement can assist future teams interested in continuing GLP-1 research.

Secondly, we established a workflow that uses artificial intelligence tools to generate and evaluate de novo protein designs. This process incorporates various AI models for protein design and integrates traditional molecular dynamics simulations. It can inspire future teams working on artificial protein design.

Lastly, we improved the existing part-BBa_K4987003, the cold-inducible promoter pCspA, and validated its functionality in E. coli DH5α. Under low-temperature conditions, we successfully used the pCspA promoter to induce the expression of lysozyme. In summary, we successfully validated and utilized the cold-inducible promoter pCspA (BBa_K4987003) in E. coli DH5α.

GLP-1 was first created by our team in the iGEM parts registry (BBa_K5083000). GLP-1 is a polypeptide composed of 23 amino acids, and when it binds to the GLP-1 receptor in vivo, it can induce insulin secretion by pancreatic beta cells. It also has functions in delaying gastric emptying and reducing gastric acid secretion.

Our team successfully expressed GLP-1 in E. coli BL21 and Nissle 1917, and we successfully introduced the PelB signal peptide to the N-terminus of GLP-1 for secretion.

We successfully amplified the GLP-1 sequence and the GLP-1 sequence with the PelB signal peptide at the N-terminus via PCR (Fig 1.A). The amplified fragments were ligated into the vector and transformed into E. coli BL21. The protein expression was validated by ELISA and Western Blot. ELISA results showed that GLP-1 was expressed intracellularly at a concentration of 62 pg/mL, and the PelB secretion signal facilitated efficient secretion of GLP-1 extracellularly, with the extracellular concentration reaching 60 pg/mL (Fig 1.B). Western Blot results indicated that without the signal peptide, GLP-1 was expressed solely intracellularly, while with the addition of the PelB secretion signal, most of the GLP-1 was secreted extracellularly (Fig 1.C).

In conclusion, we successfully amplified the GLP-1 fragment and enabled its secretion to the extracellular space using the PelB signal peptide.

Our team developed a comprehensive workflow for the artificial generation of proteins and antibodies:

First, we need to select an appropriate length for the protein or antibody and inform the model about the target sites and hotspots that need to be retained and prioritized.

Next, we used the RFdiffusion protein design model, specifically the "Practical Considerations for Binder Design" section developed by David Baker’s group, to design the backbone of the protein/antibody. We identified residues 167, 168, and 624 from DPP-4 as hotspots for this design. The selected length for the design ranged from 10 to 100 amino acids. After generating multiple conformations, we selected the most suitable protein/antibody structures for subsequent side-chain generation. Finally, from the 100 artificially generated sequences, one optimal sequence was chosen for further experimentation.

Next, we used the dl_binder_design section of David Baker’s protein side-chain design model, ProteinMPNN, to design the complete sequences. For the 20,000 generated full amino acid sequences, we employed ESMfold for structural modeling, filtering sequences with a pLDDT score >80 and RMSD <15 for further analysis. This screening process resulted in 13 sequences selected for subsequent evaluation.

Then, we performed molecular dynamics (MD) simulations on the 13 selected sequences. The simulations consisted of a 100 ns equilibration phase followed by 100 ns of long-term MD simulations, totaling 200 ns per sequence.

Finally, we compared the binding energy of GLP-1 to DPP-4 with that of DPP-4 and its known protein inhibitors. The results showed that the protein/antibody we designed exhibited superior affinity for DPP-4 compared to the natural inhibitors.

Our team's goal is to establish a comprehensive artificial protein design workflow to assist other teams in initiating their designs more efficiently. We also welcome experts in this field to provide valuable feedback on our workflow, helping future teams improve their de novo protein/antibody design processes.

Our team utilized the cold-inducible promoter pCspA (BBa_K4987003), which is available in the iGEM parts registry, to induce the expression of the lytic proteins T4 holin and T4 lysozyme, with BBa_B0015 as the terminator. The gene fragments were cloned into the plasmid pET23b, and the constructed plasmid was transformed into E. coli DH5α.

This diagram illustrates the gene circuit we constructed, featuring the cold-inducible promoter, T4 holin, and T4 lysozyme sequences. We successfully amplified the sequences for the cold-inducible promoter and the lytic proteins T4 holin and T4 lysozyme (Fig 6).

Next, we validated the functionality of the cold-inducible promoter. Our experiments demonstrated that the optimal growth temperature for E. coli DH5α is 37°C, while growth slows significantly as the temperature decreases (Fig 7.A). We then introduced the cold-inducible promoter CspA and linked it to a red fluorescent protein to verify its functionality.

The results showed that under low-temperature conditions, the OD600 value of BL21 after 12 hours was only 0.5, whereas at 37°C, the OD600 reached 2.3. However, the Fluorescence/OD600 value at low temperature was 278, significantly higher than the Fluorescence/OD600 value of 37°C (Fig 7.B). This indicates that the cold-inducible promoter CspA can effectively induce expression under low-temperature conditions.

Finally, after introducing the suicide gene, the engineered strain containing the lysozyme sequence remained in a low-density state at 16°C, with the OD600 value consistently around 0.3 (Fig 7.C), demonstrating the functionality of this suicide system.

In conclusion, we improved the existing part pCspA (BBa_K4987003), validated its functionality in E. coli DH5α, and successfully induced the expression of lysozyme through the cold-inducible promoter pCspA (BBa_K4987003), effectively inducing bacterial cell death.

Our team successfully created the GLP-1 part (BBa_K5083000) for the first time and achieved its expression in E. coli . After adding the PelB secretion signal peptide, we facilitated its secretion (BBa_K5083001). Additionally, we established a workflow utilizing artificial intelligence tools to design and evaluate proteins, integrating traditional molecular dynamics simulations, aimed at inspiring future research. Furthermore, we validated the functionality of the cold-inducible promoter pCspA (BBa_K4987003) in E. coli DH5α, successfully inducing the expression of lysozyme at low temperatures, thereby demonstrating the effectiveness of its suicide system. Overall, our achievements provide significant reference and support for related research teams.