(484a) Understanding Protein Transport in Polymer-Grafted Ion Exchangers through Multi-Scale Modeling

Authors: 
Carta, G., University of Virginia
Shirts, M. R., University of Virginia
Basconi, J., University of Virginia

When separating proteins using ion-exchange chromatography, matrices with pores with charged polymer grafts have exhibited increased binding capacity, faster adsorption kinetics, and a different diffusion properties compared to traditional macroporous matrices. However, these enhancements depend strongly on the protein properties and concentration as well as the buffer salt concentration. We use multi-scale modeling to study the factors which affect protein transport in polymer-grafted and macroporous materials, which in general depends on both the equilibrium adsorption and diffusion behavior of proteins in the different phases.

We have developed coarse-grained molecular models suitable of proteins diffusing through representative pores of both polymer-grafted and macroporous materials for ion-exchange chromatography. We use a hypothesis-driven, top-down approach, which allows us to access simulations of hundreds of protein and multiple polymer chains out to multiple microsecond time scales.

The kinetic and equilibrium information from these MD simulations is also employed in a simple rate model to study the macroscopic protein concentration profiles in the different materials over time. We apply this multi-scale modeling approach to different types of proteins and adsorbed phases, as understanding how protein transport rates are affected by various protein and material properties should provide insight into the general mechanism by which charged polymers can enhance chromatographic performance.

In simulations of lysozyme interacting with agarose-based matrices with charged dextran polymers, we obtain good agreement with experimental results for equilibrium binding capacities and effective transport rates, as well as their dependence on system parameters such as polymer graft density and surface topology. We observe faster adsorption kinetics and a different adsorption front for the polymer-grafted material as compared to the macroporous material, which is qualitatively consistent with experiments. We also obtain good qualitative agreement with larger proteins such as BSA and IgG, with experimental differences being explainable in terms of molecular details such as volume exclusion and charge distribution. In all cases, the geometry of the surface drastically affects the kinetics of the system, suggesting that idealized smooth surfaces may not be appropriate for modeling surface adsorption in chromatographic materials.