Henry Lai, PhD, and Samuel G Franklin, PhD, Bio-Rad Laboratories, Inc., 2000 Alfred Nobel Drive, Hercules, CA 94547 USA
Immunoglobulins (IgGs) from a variety of sources have been purified using the conventional process composed of a protein A chromatography step along with ion exchange chromatography and hydrophobic interaction chromatography steps (Jiskoot et al. 1989, Godfrey et al. 1993, Shadle et al. 1995, Ford et al. 2001). Protein A chromatography is normally applied as the first step in the process, yielding antibodies with very high purity in a single step. However, the disadvantages of protein A chromatography are: 1) antibody isomers are not well separated from different species; 2) protein A leakage requires additional chromatographic steps for protein A removal; and 3) protein A resins are substantially more expensive than ion exchange and ceramic hydroxyapatite supports.
Here we present an alternative process for antibody purification consisting of UNOsphere S chromatography and CHT ceramic hydroxyapatite chromatography. This process is simpler than the process utilizing protein A and avoids the leakage of a contaminating ligand (Jiskoot et al. 1989).
UNOsphere S support is a strong cation exchanger made from acrylamido and vinylic copolymers. It exhibits high protein binding capacity and low column backpressure at high linear velocity. Thus, it is a suitable capture resin in the downstream purification process. CHT ceramic hydroxyapatite support, with Ca2+ ions and PO4 3- ions in its spherical and ceramic structure, provides excellent resolution and selectivity for protein isolation.
Methods and Results
The BioLogic DuoFlow workstation was used to develop the processes. An analytical column (0.7 x 5 cm) was loaded with UNOsphere S support for small-scale process development. A preparative column (2.2 x 20 cm) was used in process scale-up. All SDS-PAGE was performed using Ready Gel precast gels, the Criterion cell, and the PowerPac 3000 power supply. Ceramic hydroxyapatite (Type I, 10 m and 20 m) media were used for this study.
Small-Scale Process Development
UNOsphere S chromatography was used as the capture step for murine IgG1 isolation. The binding capacity of UNOsphere S support for murine IgG1 was optimized at pH values from 7.0 to 4.0. Decreasing the pH in the cell culture increased the binding capacity of UNOsphere S support for murine IgG1 (data not shown). The murine IgG1 was stable for at least 6 hr in buffer at pH 4.0 (data not shown), so the cell culture was adjusted to pH 4.0. The cell culture was then loaded onto the UNOsphere S column as described in Figure 1. The murine IgG1 was eluted in the NaCl gradient. The fractions were analyzed by SDS-PAGE (Figure 2). It was found that Phenol Red and some contaminating proteins were removed in the flow-through pool by UNOsphere S chromatography.
For ceramic hydroxyapatite chromatography, the murine IgG1 fractions were pooled and diluted 5-fold with 1 mM sodium phosphate buffer, pH 6.8. The murine IgG1 elution pool from the UNOsphere S column could not be loaded directly on the CHT supports bec ause the ionic strength of the pooled solution was too high to allow murine IgG1 to bind. The diluted sample was loaded onto the CHT column at 300 cm/hr. Some contaminants were removed by a NaCl gradient. The murine IgG1 was eluted by a phosphate gradient (Figure 3).
The fraction pool was analyzed by SDS-PAGE (Figure 4). We estimated that >95% purity of murine IgG1 was achieved in this purification process.
A 2.2 x 20 cm column was packed with 60 ml of UNOsphere S to a bed height of 16 cm and run as described (Figure 5). Some contaminants were removed, and the murine IgG1 could be isolated in the process at this scale. The results indicated that the UNOsphere S purification process could be scaled up at least 30 times.
The murine IgG1 pool from the UNOsphere S column was further purified by CHT chromatography as described in Figure 6. Murine IgG1 was eluted in a phosphate gradient applied after some contaminants were removed by a NaCl gradient.
SDS-PAGE analysis showed that the scaled-up CHT purification process yields murine IgG1 of similar purity to that in small-scale purification (Figure 7). The murine IgG1 was isolated by phosphate gradient; however, some minor bands were observed in lane 3 when the gel was overloaded with 40 g of a murine IgG1 elution pool. These bands suggest that there are still some degradation products or contaminants in the purified murine IgG1. Average recovery in this scale-up purification process was about 80%. Data from flow cytometry analyses suggests that the murine IgG1 solution from our purification process yields very similar immunospecificit y to processes using protein A chromatography (data not shown).
Special thanks to Paul Yazaki, PhD, Beckman Research Institute of the City of Hope National Medical Center, Duarte, CA, for his numerous contributions to this project.
Ford CH et al., Affinity purification of novel bispecific antibodies recognising carcinoembryonic antigen and doxorubicin, J Chromatogr B Biomed Sci and Appl 754, 427435 (2001)
Godfrey MAJ et al., Assessment of the suitability of commercially available SpA affinity solid phases for the purification of murine monoclonal antibodies at process scale, J Immunol Methods 160, 97105 (1993)
Jiskoot W et al., Two-step purification of a murine monoclonal antibody intended for therapeutic application in man. Optimisation of purification conditions and scaling up, J Immunol Methods 124, 143156 (1989)
Shadle PJ et al., Antibody purification, US patent 5,429,746 (1995)
back to top