Abstract

The partition behaviour of pure IgG in aqueous two-phase systems was examined in order to investigate the effects of changes in phase properties on the partition coefficient, K. Factors such as molecular weight (M.W.) of PEG, pH, and concentration of NaCl and other salts were all found to influence K. Due to the high level of interaction between the phase components, optimal conditions were found using a factorial design. Using this methodology it was possible to find conditions which gave extremely high values of K for the IgGs (> 100). Partition behaviour of some potential contaminants, BSA and transferrin, were also studied. All the components were characterized for hydrophobicity, pI and M.W. Conditions were chosen under which the partition coefficients of the contaminants was low (pH 6, 15 % PEG 1450, 14% Phosphate and 12% NaCl). Precipitation of IgG was also minimized. These conditions were chosen for the extraction of IgG from the contaminants in a crude concentrated supernatant. The presence of culture media contaminants had a strong effect in the systems and hence on the partition of IgG which was lower than for pure IgG. The ratio of K(IgG)/K(cont) was > 25. The conditions chosen for the first forward extraction into the PEG rich phase were 15% PEG 1450, 14% phosphate 12% NaCl and pH 5.5. Virtually all IgG partitioned to the top phase and the contaminants to the bottom phase. Optimal conditions for the back extraction of the IgGs into a bottom salt phase were also found (addition of 14% phosphate with no NaCl). An industrial serum free, crude, concentrated culture supernatant of hybridoma produced murine IgG, with a relatively low level of protein contaminants (14% IgG purity) was processed in this system. After the back extraction the contaminants were reduced 18 fold giving IgG with 80% purity. A 5.9 fold purification was obtained out of a 7.3 maximum possible (at 100% pure IgG). All of the IgG was recovered. The partition behaviour of pure IgG in aqueous two-phase systems was examined in order to investigate the effects of changes in phase properties on the partition coefficient, K. Factors such as molecular weight (M.W.) of PEG, pH, and concentration of NaCl and other salts were all found to influence K. Due to the high level of interaction between the phase components, optimal conditions were found using a factorial design. Using this methodology it was possible to find conditions which gave extremely high values of K for the IgGs (> 100). Partition behaviour of some potential contaminants, BSA and transferrin, were also studied. All the components were characterized for hydrophobicity, pI and M.W. Conditions were chosen under which the partition coefficients of the contaminants was low (pH 6, 15% PEG 1450, 14% Phosphate and 12% NaCl). Precipitation of IgG was also minimized. These conditions were chosen for the extraction of IgG from the contaminants in a crude concentrated supernatant. The presence of culture media contaminants had a strong effect in the systems and hence on the partition of IgG which was lower than for pure IgG. The ratio of K IgG/K cont was > 25. The conditions chosen for the first forward extraction into the PEG rich phase were 15% PEG 1450, 14% phosphate 12% NaCl and pH 5.5. Virtually all IgG partitioned to the top phase and the contaminants to the bottom phase. Optimal conditions for the back extraction of the IgGs into a bottom salt phase were also found (addition of 14% phosphate with no NaCl). An industrial serum free, crude, concentrated culture supernatant of hybridoma produced murine IgG, with a relatively low level of protein contaminants (14% IgG purity) was processed in this system. After the back extraction the contaminants were reduced 18 fold giving IgG with 80% purity. A 5.9 fold purification was obtained out of a 7.3 maximum possible (at 100% pure IgG). All of the IgG was recovered.