|M.Sc Student||Zadok Israel|
|Subject||Evaluation of Protein-Imprinted Hydrogel Performance Using|
Molecular Dynamics Simulation
|Department||Department of Chemical Engineering||Supervisor||Professor Simcha Srebnik|
|Full Thesis text|
My research focuses on understanding the mechanisms involved, in the combination of proteins in artificial polymers as "molecularly imprinted materials" - a general term for techniques used to create materials with selective "memory" and functionality. This "memory" effect in the material could be used for diagnostics, separation, catalysis, artificial antibodies, targeted drug delivery. Due to the large size and nature of proteins and the need for functionality and bio-compatibility of the application, imprinting of proteins in hydrogels felicitates the flexibility, porosity and volume required for non-diagnostic applications.
Hydrogels are large flexible polymer networks, capable of retaining large amounts of water - similar to a living tissue. A combination of external conditions, temperature, salinity, pH, etc. can cause the hydrogel to collapse and release its contents. Incorporating therapeutic agents into the gel can transform it into a delivery system.
The behavior of proteins in hydrogels has been studied both experimentally and theoretically. The diffusivity of Cytochrome C and Lysozyme, two proteins of similar weight Isoelectric point, in non-imprinted hydrogel, differs by an order of magnitude. Imprinted behavior also differs, while cytochrom c is adsorbed on non-specific sites on lysozyme imprinted hydrogels. Lysozyme has negligible adsorption on cytochrome imprinted hydrogels.
Two formulation of hydrogel were simulated, using coarse grained model for the protein, gel and the solvent. One formulation contained only acrylamide and bis-acrylamide as crosslinker. The second contained in addition acidic and basic monomers, argued by the original article to enhance the imprinting effect.
The parameters for the force field were adjusted to fit the model to data. The system was simulated using molecular dynamics and used a simplified polymerization model to enable cross-linking. Hybrid simulation, combining molecular dynamics and Grand Canonical Monte Carlo, was used to ensure the proper water content of the gel.
Simulated gels exhibit similar swelling behavior to experimental gels, but higher compressibility and Poisson's ratio than experimental gels - indicating a less elastic behavior. This could be a result of simulation size smaller than needed to average over high and low density regions of the system.
Potential energy difference of proteins in imprinted and non-imprinted gels was measured. The difference for proteins in imprinted sites is 4-5 times more negative than in non-imprinted. Improved formulation lowers the potential energy by ~15[kcal/mol] for lysozyme. Cytochrome C displays a different behavior lowering the energy for imprinted sites, while showing no apparent order for non-imprinted sites. In some CYC samples the energy difference is positive, indicating the protein is more stable in water.
The equilibrium constant for proteins in imprinted sites was evaluated using Kirkwood-Buff integrals. The results indicate a strong, practically irreversible binding of the protein to the site - this is supported by experimental data. Gibbs free energy values for lysozyme are higher than cytochrome c.
Future research directions include: better formulation of the problem using Kirkwood-Buff terminology, which would enable the estimation of partial chemical potentials for different types of monomers. Another potential direction is a steered molecular dynamics method which will include rotational forces on the protein.