1 Stability of FTD-BMP-4

1.1 Why Do We Do Stability Simulation?

Bone Morphogenetic Protein-4 (BMP-4) is a well-known growth factor involved in regulating osteogenesis and tissue regeneration. However, due to its complex structure, BMP-4 contains numerous intra- and inter-chain disulfide bonds, which are crucial for maintaining its stability and bioactivity. Expressing BMP-4 in bacterial systems such as E. coli is challenging, as it often results in misfolded proteins or inclusion body formation due to improper disulfide bond formation.

To enhance its targeting and release properties, BMP-4 has been fused with the Fibrinogen-like Domain (FTD), a tag known for its strong collagen-binding affinity. However, this fusion introduces additional structural complexity, raising concerns about whether the FTD may negatively impact the thermal stability of BMP-4, potentially leading to denaturation or loss of function under physiological conditions. Evaluating the thermal stability of the FTD-BMP-4 fusion is critical, as any instability may compromise its efficacy in tissue engineering applications.

To address this challenge, molecular dynamics (MD) simulations offer a powerful tool to predict how the fusion of BMP-4 with FTD influences the protein's stability under thermal stress. This computational approach allows for in-depth analysis of conformational changes, disulfide bond stability, and protein folding dynamics, providing valuable insights before experimental validation. By simulating different environmental conditions, this study aims to assess whether the FTD tag stabilizes or destabilizes BMP-4, offering guidance for future tag optimizations and reducing the cost of trial-and-error in wet lab experiments.

Here are three key reasons for conducting this molecular dynamics (MD) simulation experiment to predict the thermal stability of FTD-BMP-4:

• Structural Complexity of BMP-4: BMP-4 is a complex protein with several intra- and inter-chain disulfide bonds, which play a crucial role in maintaining its proper folding and biological activity. Expressing BMP-4 in a prokaryotic system like E. coli often leads to challenges in forming correct disulfide bonds, resulting in misfolding or the formation of inactive inclusion bodies. By introducing an external domain like FTD, there is a potential risk of disrupting BMP-4’s already delicate structure, further compromising its stability. MD simulations allow us to evaluate whether the fusion with FTD stabilizes or destabilizes BMP-4 at a molecular level, minimizing the risk of thermal denaturation before performing costly experimental procedures.
• Predictive, Low-Cost Approach: Conducting in vitro stability experiments on multiple fusion constructs can be labor-intensive, time-consuming, and expensive, especially when working with proteins like BMP-4 that are challenging to express and purify. MD simulations provide a cost-effective alternative for predicting the structural stability and behavior of the FTD-BMP-4 fusion under different environmental conditions (e.g., temperature changes). If the fusion appears unstable in the dry lab, alternative tags can be considered (e.g., different CBDs), reducing the time and cost of trial-and-error in the wet lab.
• Optimization of Fusion Tags: Predicting the thermal stability of FTD-BMP-4 using MD simulations is critical for ensuring that the FTD tag itself does not negatively affect the structural integrity of BMP-4. Since the goal of the FTD tag is to enhance BMP-4’s functionality by directing it to collagen matrices, it’s essential to confirm that it does not inadvertently lead to protein misfolding or instability. If the simulations reveal instability, other potential fusion tags can be tested in silico until the optimal tag for BMP-4's stability and functionality is found. This computational prediction ensures that only the most promising constructs are moved forward into the wet lab, optimizing the development pipeline.

Fig1.1 cell factor’s fusion protein structures simulated by alphafold3

1.2 Experimental Method

To design a molecular dynamics (MD) simulation using GROMACS to verify the thermal stability of BMP-4 when fused with FTD, here’s an organized step-by-step approach.

1.2.1 Preparation of Protein Structure

• Input Sequences: Begin by determining the primary sequences of the FTD domain and BMP-4.
• Structure Prediction: Using AlphaFold 3, predict the 3D structure of the fusion protein (FTD-BMP-4). AlphaFold provides highly accurate protein structure predictions based solely on the primary sequence, giving a reliable starting point for molecular dynamics simulations.
• Fusion Model: Ensure the fusion between the FTD domain and BMP-4 is biologically relevant, preserving functional domains. We also performed sequence alignment to confirm the junction between the two domains does not disrupt key functional sites.
• Structure Validation: Post-prediction, assess the quality of the predicted structure by checking stereochemical properties (Ramachandran plots).
• Energy Minimization: Once the structure is confirmed, perform energy minimization to relax any steric clashes or structural strain before proceeding with the MD simulation.

1.2.2 Setting Up Molecular Dynamics in GROMACS

a. Generate Topology Files

• Use GROMACS pdb2gmx command to generate topology files for the FTD-BMP-4 fusion. Choose a force field suited for protein stability simulations. We choose AMBER ff14SB. This force field is a popular choice for simulating biomolecular systems and is widely used for protein dynamics due to its well-balanced treatment of protein backbone and side chain interactions. It includes refined parameters for side chains and backbone torsions that improve the accuracy of protein folding simulations.
• Select a water model like TIP3P for solvation, which is commonly used in thermal stability studies.

b. Solvation and Ion Addition

• Create a solvent box using the gmx solvate command. Ensure the box size allows sufficient space around the protein to avoid boundary interactions.
• Neutralize the system by adding counterions (Na⁺ or Cl⁻) using the gmx genion command.

c. Energy Minimization

• Run energy minimization to remove any bad contacts in the protein or water molecules.
• Monitor potential energy to ensure it converges before proceeding to equilibration.

1.2.3 Equilibration Phases

Perform two phases of equilibration:

1. NVT (constant volume, temperature): This step ensures the system is equilibrated at the target temperature (e.g., 300 K). Run the simulation for 100-200 ps.
2. NPT (constant pressure, temperature): This phase allows the system to equilibrate in terms of pressure and volume. Run it for another 100-200 ps.

1.2.4 Molecular Dynamics Production Run

Time Scale: Run the MD simulation for a longer time, ideally 500 ns or more, to capture thermal fluctuations and structural changes over time. We run it for 600ns.

1.2.5 Analysis of Stability

1. Root Mean Square Deviation (RMSD): Calculate RMSD using the gmx rms tool to track overall structural changes over time. Compare the deviation of FTD-BMP-4 at different temperatures to see if it deviates more at higher temperatures, indicating potential instability.
2. Radius of Gyration (Rg): Measure the Rg to assess the compactness of the protein over time. If the protein unfolds or expands significantly at high temperatures, the Rg will increase.

1.2.6 Reporting and Data Interpretation

Thermal Stability Conclusion: Based on RMSD, RMSF, and secondary structure analyses, we can quantify the thermal stability of FTD-BMP-4. Specifically, stable RMSD values, minimal structural fluctuations, and other elements at high temperatures will confirm thermal stability.

1.3 Results

1.3.1 RMSD

Fig1.2 RMSD simulation results of FTD-BMP-4

The RMSD (Root Mean Square Deviation) plot provides insights into the structural stability of the FTD-BMP-4 fusion protein over time during molecular dynamics (MD) simulations.

X-axis (Time, ns): The horizontal axis represents time in nanoseconds (ns). It shows the progression of the molecular dynamics (MD) simulation over time, from 0 to 600 nanoseconds in this case. The longer the time, the more the system evolves, allowing us to observe how the protein structure changes and stabilizes.

Y-axis (RMSD, nm): The vertical axis represents the Root Mean Square Deviation (RMSD) in nanometers (nm). RMSD measures the average deviation of the protein's atomic positions from a reference structure (usually the starting structure) over time. Higher RMSD values indicate greater deviation from the initial structure, implying larger conformational changes, while lower values indicate that the structure remains closer to the initial conformation.

• Initial Fluctuation (0-50 ns): At the beginning of the simulation, there is a sharp increase in the RMSD values, indicating that the protein undergoes significant conformational changes as it adjusts from its initial structure (likely from energy minimization). This is typical during the early phase of MD simulations when the protein equilibrates to the simulation environment.
• Initial Fluctuation (0-50 ns): At the beginning of the simulation, there is a sharp increase in the RMSD values, indicating that the protein undergoes significant conformational changes as it adjusts from its initial structure (likely from energy minimization). This is typical during the early phase of MD simulations when the protein equilibrates to the simulation environment.
• Stabilization Phase (50-100 ns): Around 50 ns, the RMSD begins to stabilize, with fluctuations reducing significantly. This indicates that the protein has reached a relatively stable conformation, and the major structural adjustments have settled.
• Stable Behavior (100-600 ns): After 100 ns, the RMSD fluctuates around a stable value of approximately 1.0-1.2 nm, suggesting that the overall structure of FTD-BMP-4 is thermally stable. The consistent RMSD value over the remaining simulation time shows that no significant structural denaturation or unfolding occurs.

This result suggests that the FTD-BMP-4 fusion protein maintains its structural integrity over time, even under thermal stress. The steady RMSD values after the initial fluctuation phase indicate that the fusion does not significantly destabilize BMP-4. Thus, the FTD tag can be considered compatible with BMP-4, making it a suitable candidate for further experimental validation and potential use in tissue engineering applications, such as bone repair.

1.3.2 Rg

Fig1.3 Rg simulation results of FTD-BMP-4

The Radius of Gyration (Rg) plot provides insight into the compactness or folding behavior of the protein during the molecular dynamics (MD) simulation.

X-axis (Time, ns):The horizontal axis represents time in nanoseconds (ns), indicating the progression of the MD simulation over time. In this case, the simulation runs for 600 ns.

Y-axis (Rg, nm):The vertical axis shows the Radius of Gyration (Rg) in nanometers (nm), which is a measure of the distribution of the atoms around the protein’s center of mass. It reflects the protein's overall size and compactness.

Low Rg values indicate that the protein is more compact or tightly folded.

High Rg values suggest that the protein is more expanded or loosely folded.

• Initial High Rg (0-50 ns): The plot shows that the Rg starts relatively high (around 2.3 nm) during the first 50 ns. This period likely reflects initial conformational adjustments as the protein begins to equilibrate in the MD simulation, possibly undergoing some unfolding or expanding before finding a more stable conformation.
• Stabilization Phase (50-600 ns): After about 50 ns, the Rg values stabilize between 2.0 and 2.15 nm, with relatively small fluctuations. This indicates that the protein adopts a more compact and stable conformation over time and does not experience major unfolding or denaturation.
• No Significant Expansion or Collapse: Throughout the remainder of the simulation, the absence of large deviations or long-term increases in Rg suggests that the FTD-BMP-4 fusion protein maintains its structural integrity, without significant unfolding or aggregation. The relatively consistent Rg values suggest that the protein retains a stable and compact structure under the conditions simulated.

The Radius of Gyration (Rg) results confirm the compactness and structural stability of the FTD-BMP-4 fusion protein during the molecular dynamics simulation. The stable Rg values, especially after the initial phase, indicate that the fusion does not lead to significant unfolding or instability, further supporting the conclusion from the RMSD analysis that FTD-BMP-4 is structurally stable.

2 Rational Design of VEGF to Enhance Its Binding to Receptors

We have carefully redesigned the existing VEGF121 part by introducing targeted mutations at two key residues, R105S and P106Y, aiming to enhance its binding affinity to VEGFR and improve its biological function. Our efforts are demonstrated through rigorous experimental validation, including a scratch assay with MC3T3-E1 cells, which shows that the mutant significantly enhances cell migration and wound healing compared to the wild-type VEGF121. Additionally, we have meticulously documented our findings, following the Best New Improved Part award requirements. This includes detailed experimental protocols, quantitative results, and clear comparisons between the mutant and wild-type proteins under identical conditions. Our thorough documentation on the Part's Registry Main Page ensures that the new part is well-supported by experimental data, showcasing its superior performance.

2.1 Targeted Mutagenesis of VEGF to Enhance Receptor Binding and Boost Angiogenesis

In clinical settings, rapid angiogenesis is essential, but achieving this requires high concentrations of VEGF, which can be costly and potentially trigger immune reactions and side effects due to excessive cell factor levels. Therefore, there is a pressing need for a molecule that can promote effective blood vessel growth at lower concentrations. To address this, we aim to engineer VEGFA through rational design, enhancing its receptor-binding efficiency. This modification is expected to result in better angiogenic outcomes with reduced cell factor concentrations, improving therapeutic efficacy.

The activity of wild type VEGF is currently limited, which presents a challenge because rapid stimulation of angiogenesis is crucial in therapeutic contexts. On a molecular level, increasing the binding affinity between VEGF and its receptor (VEGFR) could significantly enhance the speed and effectiveness of this process. By performing targeted mutagenesis on VEGFA, we aim to improve receptor interaction, thereby boosting VEGF’s bioactivity. This modification would lead to more efficient promotion of vascular growth, an essential factor in tissue regeneration and repair applications such as HEK293 cell migration studies.

2.2 Investigating the Mechanism of Interaction between VEGF and VEGFR

Thus, we began to study the interaction between VEGF and its receptor. Using VEGF121 as a model, we focused on its interaction with the extracellular domain of VEGFR (PDB: 4CKV). We applied AlphaFold3 to analyze the interaction mechanism between VEGF121 and VEGFR, allowing us to explore the structural details of their binding. This approach aims to provide insights into how rationally designed mutations in VEGF can enhance receptor binding and angiogenesis at lower cell factor concentrations.

Fig2.1 Wild-type VEGF (cyan) interacts with VEGFR (extracellular domain, green), with yellow dashed lines representing hydrogen bonds

Fig2.2 All schematic diagram of the interaction between wild-type VEGF (cyan) and VEGFR (extracellular domain, green). The yellow dotted lines represent hydrogen bonds.

With only three hydrogen bonds identified between VEGF and VEGFR (K16-E10, D63-R93, Y25-L73), the relatively weak interaction is further confirmed by the AutoDock Vina score of -6.3, indicating suboptimal binding energy. Such a low binding score suggests that the current VEGF structure has limited capacity to strongly engage with its receptor. Enhancing VEGF’s binding affinity is crucial for improving its biological function, particularly for therapeutic purposes like stimulating angiogenesis. Optimizing these interactions through mutagenesis can strengthen the receptor engagement and boost angiogenic activity, making VEGF more effective in promoting rapid blood vessel formation and tissue repair. By strategically increasing hydrogen bond numbers or improving molecular stability, we can potentially create a more efficient and potent version of VEGF for clinical applications.

We ultimately decided to increase the number of hydrogen bonds rather than focusing on thermal stability for several reasons. First, enhancing hydrogen bonding directly improves the binding affinity between VEGF and VEGFR, which is crucial for stronger receptor interactions and better signaling. In contrast, improving thermal stability may only extend the protein's half-life but won't necessarily enhance its receptor-binding efficiency. Prolonged half-life may cause cytokines to spread to other sites, causing side effects. Since our goal is to boost biological activity, increasing hydrogen bond interactions provides a more direct and effective approach for improving VEGF's functionality.

2.3 Mutants

From the perspective of hydrogen bond formation, enhancing VEGF's interaction with VEGFR can be achieved by targeting amino acids in the interface region that currently cannot form hydrogen bonds. Mutations should aim to introduce amino acids with hydrogen-bonding capabilities.

Typically, residues like proline (P), which has a rigid cyclic structure that cannot form hydrogen bonds, can be mutated to tyrosine (Y) or serine (S), which possess hydroxyl groups that actively participate in hydrogen bonding. Other examples include mutating alanine (A) to glutamine (Q) or asparagine (N) to introduce polar side chains capable of forming hydrogen bonds.

By strategically selecting and mutating such residues in the binding region, we can significantly improve the VEGF-VEGFR interaction, leading to stronger receptor engagement and enhanced biological activity.

Fig3.1 Mutations lead to increased hydrogen bonding

Based on the results from multiple AlphaFold3 simulations, we found that mutating P106 to Y while also mutating R105 to S can increase the number of hydrogen bonds without compromising the stability of the structure. If only P106 is mutated to Y, the large steric hindrance from R105, which is located at the turn, causes structural instability. By mutating R105 to a smaller amino acid like S, we can maintain structural integrity while improving the interaction through additional hydrogen bonds. After the R105S and P106Y mutations, Y106 forms a hydrogen bond with E19 of VEGFR, further enhancing the binding affinity between VEGF and VEGFR. This newly introduced hydrogen bond provides a more stable interaction that could improve the biological effectiveness of VEGF in therapeutic applications.

Fig3.2 VEGF (wild type & R105S and P106Y mutation, cyan) interacts with VEGFR (extracellular domain, green), with yellow dashed lines representing hydrogen bonds

2.4 Experimental Verification

The scratch assay, also known as the wound healing assay, is a widely used method for studying cell migration. It simulates wound healing by creating a "scratch" in a cell monolayer, and the ability of cells to move into this gap is tracked over time.

To assess the wound-healing capabilities of our VEGF mutants (R105S and P106Y), we will conduct a scratch assay using MC3T3-E1 cells. Cells will be cultured in a 6-well plate until 80-90% confluency, followed by a scratch using a pipette tip. Afterward, serum-free medium containing 50 ng/mL of either wild-type or mutant VEGF will be added. Images of the scratch will be taken at 24 hours to monitor wound closure. We will compare the wound closure rates between the wild-type and mutant VEGF to evaluate enhanced cell migration.

Fig4.1 VEGF121 Mutant (R105S, P106Y) Enhances Wound Healing and Cell Migration

The data from the scratch assay show a significant improvement in both wound healing percentage and cell migration velocity for the VEGF121(R105S, P106Y) mutant compared to the wild-type VEGF121. The left graph demonstrates that the wound healing percentage for the mutant is significantly higher (p < 0.01) than for the wild-type. Similarly, the right graph indicates that the cell migration velocity is also significantly increased in the mutant (p < 0.05). These results suggest that the R105S and P106Y mutations enhance VEGF's ability to promote cell migration and wound healing.

The results of the scratch assay demonstrate that the VEGF121 mutant (R105S, P106Y) significantly outperforms the wild-type VEGF121 in both wound healing percentage and cell migration velocity. The increased wound healing percentage indicates that the mutant VEGF enhances cell migration and wound closure more effectively than the wild-type. The higher cell migration velocity in the mutant further supports this finding, suggesting that the mutations enhance the interaction between VEGF and VEGFR, likely due to increased hydrogen bonding and receptor binding affinity. This improvement could have valuable implications for therapeutic applications where rapid angiogenesis and tissue repair are needed.