The molecular phenomena underlying multimodal chromatographic separations : a molecular dynamics study

Abstract

In the past several years, protein-based therapeutics have emerged as a powerful tool for treating complex diseases, such as cancer, diabetes, and autoimmune diseases. These therapeutics have come to dominate the pharmaceutical landscape, fueling the need for more effective methods for purifying proteins. Standard single-mode chromatography techniques such as hydrophobic interaction, ion exchange, or size exclusion chromatography are often insufficient to address the existing purification challenges. To this end, a new class of chromatography, multimodal (MM) chromatography, has been demonstrated to achieve unique selectivities and difficult separations.

This thesis addresses these challenges by exploring the molecular nature of protein-multimodal ligand interactions using molecular dynamics simulations. First, we investigated these complex protein-ligand interactions by developing a representative model system consisting of two bonded molecules (the ligand) binding to a surface (the protein). In this simplified system, it was straightforward to develop an understanding of the nature of synergy as well as explore how the additional complexities affected synergy.

We then turned our attention to studying the hydration and conformational equilibria of three multimodal ligands, Capto MMC, Nuvia cPrime, and Prototype 4. We performed simulations in a variety of solvents and evaluated the relative importance of intramolecular and solvent-mediated interactions in determining ligand conformations. Upon introducing proteins to these systems, we found that the ligand conformations observed in water were fully conserved when the ligand bound to a protein surface. Additional investigations into the water structure near the ligands indicated that small changes in ligand chemistry led to significant differences in ligand solvation. These solvation differences affected the desolvation process that accompanies ligands binding to the protein surface. Consequently, we found that while all ligands bound strongly to charged amino acids, each ligand displayed a different probability of binding to hydrophobic amino acids.

In addition to studying ligands in free solution, we also explored how ligands behaved on a surface. We studied systems of increasing complexity including a single tethered ligand, a pair of ligands, and a ligand triplet. We then explored the role that ligand chemistry, surface density, and linker length played in governing the location and relative accessibility of different ligand moieties on a multimodal surface and developed an approach for quantifying pattern formation on multimodal surfaces. Finally, we quantified the strength of ligand-ligand interactions involved in the formation of a pair and a triplet of ligands and used that information in a coarse-grained Monte Carlo simulation approach to predict pattern formation in a high density ligand surface.

Finally, in the last part of this thesis, we explore how the insights developed in these molecular studies can be applied to develop an optimally separable and orthogonal set of multimodal resins. In particular, we develop a framework for quantifying separability and orthogonality and calculate these metrics for a small set of proteins and resins. We discuss how these metrics can be combined with molecular studies to arrive at an optimal set of multimodal resins.

Mulitmodal chromatography employs a synergistic combination of multiple weak interactions between immobilized ligands and proteins of interest. By using ligands having not single, but multiple interacting groups such as charged, hydrophobic, or hydrogen-bonding moieties connected in an optimal molecular architecture, multimodal chromatography can not only address previously extremely challenging separations but also dramatically simplify existing separation processes. Its use in industrial applications has been limited, however, because it is challenging to predict protein-ligand interactions, and by extension retention in chromatography systems.