Investigation of protein adsorption in hydrophobic interaction chromatographic (HIC) systems

Hydrophobic interaction chromatography (HIC) has been shown to be an important separation mode for the purification of biomolecules. Several factors which influence protein adsorption in HIC systems, such as protein properties, stationary phase chemistry, type and concentration of salt were investigated in this thesis. The work presented in this dissertation includes an investigation of protein adsorption/recovery behavior at low salt conditions, various classes of protein adsorption isotherm behavior, a study of the protein adsorption mechanism using quantitative structural property relationship models and the development of a methodology for predicting protein retention time in any given gradient condition in HIC systems.; The adsorption and recovery of proteins under low salt conditions in HIC systems was investigated using both high-throughput screening (HTS) and computational methods. Proteins were classified based on various combinations of adsorption and recovery behavior and molecular descriptors associated with classes of adsorption were determined. In addition, a range of protein isotherms on various HIC resin systems were determined and the effects of salt concentration, resin chemistry and protein properties on the isotherms were examined. The resulting isotherms exhibited unique patterns of adsorption behaviors. The establishment of unique classes of adsorption behavior may shed light on our understanding of the behavior of proteins in HIC systems. The protein retention mechanism was also investigated using the preferential interaction theory in concert with QSPR modeling. In addition to successfully predicting the water release values, which is an important thermodynamic parameter, the selected descriptors provide insights into the protein physicochemical properties which influence protein affinity in HIC system. Finally, the a priori prediction of protein gradient adsorption behavior in HIC systems was achieved by using protein isocratic linear elution data in concert with quantitative structure property relationship (QSPR) modeling. The ability to predict protein chromatographic behavior directly from protein crystal structures may have significant application for a range of bioprocessing applications.