Engineering biomolecular interactions to enhance selectivity in chromatographic separations

Abstract

An initial comparison between two homologous mixed-mode cation-exchange ligands revealed that differences in the geometric presentation of a hydrophobic aromatic moiety significantly altered the elution behavior of selected proteins, specifically the selectivity of these ligands towards aromatic-type hydrophobic regions present on the protein surface. This selective effect was demonstrated to be insensitive to variations in surface ligand density or the presence of fluid phase modifiers.

This work demonstrates that the fundamental principles of biomolecular interactions can be used to engineer molecular selectivity into novel mixed-mode and peptide affinity ligands and should be applied to both simplify the selection of operating conditions and maximize the separation potential of these chromatographic operations.

Studies of increasingly complex peptide-based ligands showed that the proper presentation of complementary chemical properties to a targeted protein surface could be used to design ligands with enhanced affinity and selectivity for a target protein. Using a combination of high-throughput techniques consisting of microarray screening and 96-well plate experiments, a high-affinity peptide ligand was selected from the initial library (designed in silico) and rapidly scaled up to a small-scale column that could separate a target protein from harvested cell culture fluid.

Further investigations into a homologous series of mixed-mode cation-exchange ligands revealed that ligand chemistry and geometry had additional influences on the adsorptive properties of these complex molecules. Using the combined retention data of ion-exchange and mixed-mode ligand datasets, a QSAR model was constructed that folded ligand properties with protein surface properties to accurately predict protein retention behavior across this spectrum of chromatographic ligands and could predict the retention of these proteins on novel mixed-mode resins in silico.

The emergence of novel mixed-mode and custom affinity chromatographic media presents an opportunity to reduce the number of process steps and improve the economics of purification processes by taking advantage of the improved affinity and selectivity of these chemically-complex ligands. This project utilized a fundamental investigation of protein-ligand interactions across the spectrums of affinity and selectivity to establish design principles for engineering a desired set behavior into chromatographic separations.

A high-throughput screen of ionic salts, amino acids and small organic solutes as eluents revealed that the hydration state these co-solutes was instrumental in altering the balance of electrostatic and hydrophobic interactions in multimodal systems, thus creating selective conditions for the recovery of product and removal of aggregate species in linear gradient and weak partitioning column separations that were able to remove aggregate impurities while maintaining a high recovery of the monomer product species.