Investigations into multimodal chromatographic selectivity using in silico designed Fab fragment variants and column modeling

Abstract

Downstream bioprocessing is continuously faced with challenges requiring the development of more efficient and robust bioseparation technologies. Multimodal chromatography has the potential to provide improved selectivity and efficiency over traditional single mode chromatography. Despite a growing interest in multimodal materials, the lack of fundamental understanding associated with the multiple interactions present in these materials is limiting their implementation in industrial bioprocessing. In this thesis, a multi-tool framework will be presented for improving the understanding of selectivity of complex biomolecules in multimodal chromatographic systems.

The use of in silico protein surface property analyses combined with molecular biology techniques, protein expression and chromatographic evaluations represents a previously undescribed and powerful approach for studying multimodal selectivity with complex biomolecules. Importantly, the studies open new avenues of employing in silico protein surface characterization techniques and column modeling to aid in process design for product variant separations using multimodal chromatography.

The Fab variants were employed to investigate the effects of changing the mobile phase conditions and multimodal ligand design. Using arginine as a mobile phase modifier, Fab variant elution was promoted at a lower ionic strength. In contrast, the presence of guanidine amplified the selective trends of the multimodal system and created a window of selectivity for the separation of Fab variants which could not be separated in the system without the modifier. A novel set of multimodal chromatographic ligands were employed to examine the effects of changing the ligand structure on Fab variant retention and selectivity. Interestingly, the multimodal chromatographic ligands exhibited different selectivity trends and were capable of differentiating between different Fab classes and related variants.

Unique libraries of Fab fragment variants were designed in silico to examine the relationship between protein surface properties and selectivity in multimodal chromatographic systems. Protein surface property characterization tools were employed to identify potential multimodal ligand binding regions on the Fab fragments and to evaluate the impact of mutations on surface charge and hydrophobicity. To broaden the analysis, a diverse set of Fab variants with various CDR loops and/or framework regions was included in the analyses. Forty different Fab variants were then generated by site-directed mutagenesis, transient expression in HEK293 cells and purification by affinity chromatography. Gradient column experiments were carried out with the resulting Fab variants in various chromatographic systems to enable detailed investigations into the relative importance of local protein surface properties on protein retention and selectivity. The retention data demonstrated that the multimodal chromatographic systems provided different selectivity than traditional single mode chromatography.

The utility of multimodal chromatography in large-scale bioprocessing was investigated through the development of a column simulation model. This modeling approach was successfully implemented to describe highly selective multimodal purification processes with model proteins and complex industrial protein mixtures in concert with mobile phase modifiers as the eluents.