Silica-based solid phases for affinity chromatography: Effect of pore size and ligand location upon biochemical productivity

Abstract

The influence of pore size and surface chemistry upon the productivity in affinity chromatography of three silica-based solid phases, Sorbsil C-200, C-500, and C-1000 (40-60 microm particle diameter and the corresponding pore diameters of 20, 50, and 100 nm), was studied using three model ligand/biomolecule systems of varying molecular masses. These studies revealed two unique parameters, biochemical productivity and maximum physical capacity, of the matrix as generically essential in the successful design and operation of productive affinity chromatography systems. Biochemical productivity, the molar ratio of the amount of product recovered per unit volume of adsorbent and ligand concentration, utilized the expected stoichiometry of binding of the two molecules to assess the efficacy of the adsorbent. This parameter, determined by equilibrium binding in batch suspensions and by saturation binding capacities and recoveries in fixed beds, yielded the optimum ligand concentration required for maximal performance. Maximum physical capacity, of the adsorbent to accommodate the biomolecules, was calculated from pore and molecular dimensions assuming that there was no steric hindrance to access. Using an immobilized human-IgG (Hu-IgG)/anti-Hu-IgG monoclonal antibody (MCAB) system, in which both the ligand and the product are of the same size (150 kDa), it was shown that the physical capacity of C-200 was only 16% of the theoretically expected amount. This capacity increased to 70 and 90% of the expected value with C-500 and C-1000, respectively, as the steric hindrance to protein penetration induced by pore dimensions decreased. The distribution of immobilized Hu-IgG within individual particles, visualized by immunofluorescence and immunogold labeling, showed that the ligand was restricted to the peripheral 3 microm of the C-200 particles (12% radius). In contrast, it was present throughout the C-1000 particles, indicating that there was no hindrance to access in this solid phase. The C-200 was suitable for use in a small ligand/biomolecule system studied (immobilized trypsin-inhibitor binding trypsin; 22.1 and 23.3 kDa, respectively) for which more than 60% of the maximum physical capacity was available for interactions. The C-500 proved satisfactory for the Hu-IgG/MCAB model system but showed steric limitations when an immobilized anti-beta-galactosidase MCAB (anti-beta-gal) was used to purify a larger product (beta-galacosidase; 460 kDa). The binding capacity and overall productivity of Hu-IgG- and anti-beta-gal-C-1000 was equivalent to that of Sepharose CL-4B. Selection of matrices with pore sizes appropriate to the dimensions of the ligand and product was, therefore, important. Finally, the Sorbsil silicas packed easily into beds and were used successfully with conventional chromatography equipment for low-pressure affinity chromatography. They therefore offer an ideal alternative to silica-based high-performance liquid affinity chromatography and soft-gel supports.