Understanding the fundamental biophysics behind protein-NP interactions would be the essential of designing and engineering protein-NP systems for better performance in applications such as biosensors and NP drug carriers. To address this point, we developed a coarse grain model, which evolves from a well parameterized and validated surface potential for protein-(flat) surface interactions I developed as part of my Ph.D research. The main advantage of this method is that using a single-bead-per-residue model it quantitatively accounts for the hydrophobic interactions of residues and the NP surface and the dependency of protein stability on the curvature of NPs. As validated by the reference experimental study, the coarse grain model successfully reproduces the protein GB1 structure , orientation, and binding affinity on the NP that are consistent with experimental observations such as circular dichroism (CD) and fluorescence spectroscopy. Further, this method has been successfully applied to study the effect of NP on the nucleation of protein fibrillation, which is one of the main concerns of the potential toxicity of NP in vivo. The accelerating or inhibiting effects on the protein fibrillation is depending on the surface hydrophobicity and curvature of the NP and the inherent protein stability, which can be captured by this coarse grained simulation method.
Furthermore, denaturants such as urea and guanidinium chloride (GdmCl) are widely used to study protein folding/denaturing processes in solution and at interfaces such as protein-NP complexes. Denaturants especially have their advantages in studying protein folding over heating due to the convenience in experiments. To extend the ability of the coarse grained simulation system for the direct comparison to experiments, we implemented a model that simulates protein folding in urea and GdmCl. In this work, by using the similar idea and rescaling the parameters of O'Brien's molecular transfer model, we implement a Tanford's transfer free energy method for the KB Go-like model. More importantly, instead of reconstructing the all-atom structure as the MTM, we employ an optimized fast solvent accessible surface area estimator to directly calculate the residue-based SASA based on the Cα information only. Therefore, the free energy contribution of the solvent is able to be measured and included in the simulation sampling on the fly. The results show great consistency in the predicted change of protein melting temperature by this simulation method and the experimentally measured values. Different thermally stable intermediates of protein folding in denaturants are captured by this simulation method, which may be out of limit of experimental measurements.
Optimization and expansion of this simulation system is on-going, which aims to form an integrated modeling system that accurately and efficiently predicts proteins' stability (both thermal and in denaturants), binding affinity, aggregation rate, structure, and orientation at various NP surfaces and solvent conditions. Therefore, together with collaborations with experimental methods, this simulation system has a strong potential in leading to the better design of NP-protein complexes for biosensors and drug delivery systems.
See more of this Group/Topical: Meet the Faculty Candidate Poster Session – Sponsored by the Education Division