IMAC capture of recombinant protein from unclarified mammalian cell feed streams

Abstract

Fusion-tag affinity chromatography is a key technique in recombinant protein purification. Current methods for protein recovery from mammalian cells are hampered by the need for feed stream clarification. We have developed a method for direct capture using immobilized metal affinity chromatography (IMAC) of hexahistidine (His6) tagged proteins from unclarified mammalian cell feed streams. The process employs radial flow chromatography with 300-500 μm diameter agarose resin beads that allow free passage of cells but capture His-tagged proteins from the feed stream; circumventing expensive and cumbersome centrifugation and/or filtration steps. The method is exemplified by Chinese Hamster Ovary (CHO) cell expression and subsequent recovery of recombinant His-tagged carcinoembryonic antigen (CEA); a heavily glycosylated and clinically relevant protein. Despite operating at a high NaCl concentration necessary for IMAC binding, cells remained over 96% viable after passage through the column with host cell proteases and DNA detected at ∼ 8 U/mL and 2 ng/μL in column flow-through, respectively. Recovery of His-tagged CEA from unclarified feed yielded 71% product recovery. This work provides a basis for direct primary capture of fully glycosylated recombinant proteins from unclarified mammalian cell feed streams.

Keywords: Chinese hamster ovary cells; IMAC; downstream processing; integrated processing; radial flow chromatography; recombinant protein.

Figure 1

Radial flow chromatography methods. A: Axial and radial flow chromatography. Frits are located at the outer and inner circumferences of the radial column. Bed height (H), flow direction (F). B: Process flow diagram of integrated radial flow chromatography control.

Figure 2

Analysis of unclarified feed streams containing different densities of CHO cells before and after passage through a scale down RFC with large bead resin (n = 3). A: Percentage cell viability of unclarified feed stream in pre‐column, flow through and washing conditions. B: qPCR quantification of host cell DNA in pre‐column and flow through conditions. C: Quantification of supernatant protease activity in unclarified feed stream and maximum cell lyses samples.

Figure 3

Particle size distribution of crude feed streams applied to a scale down radial‐flow column packed with large bead resin. Feed stream cell density; (A) media only, (B) 0.5 × 10⁶ cells/mL, (C) 1 × 10⁶ cells/mL, (D) 2 × 10⁶ cells/mL, (E) 3 × 10⁶ cells/mL, (F) 8.5 × 10⁶ cells/mL. Blue indicates pre‐column sample, red flow through, and green PBS wash.

Figure 4

FPLC and SDS–PAGE separation of eluted fractions of a model His‐tagged single‐chain variable fragment from conditioned media buffered to IMAC binding conditions. Red numbers indicate 1 mL fraction collection.

Figure 5

Design of experiments based optimization of buffer composition for IMAC capture of His‐tagged CEA. A: Pareto Chart indicates significant factor interactions above 1 value limit (black line). Blue indicates negative effects, orange positive effects and hollow significant effects (Factor labels; a = pH, b = NaCl, c = imidazole). B: Factor interaction of NaCl and pH (Red = 500 mM NaCl, Black = no NaCl addition). C: 2D, and (D) 3D surface response curve of His‐tagged CEA capture.

Figure 6

Cell survival and column flow characteristics of feed stream under optimized buffer conditions. A: CHO cell survival in buffered media with increasing NaCL concentration over 60 min. B: Protease activity in unclarified feed stream and maximum cell lyses samples after column flow buffered with 500 mM NaCl, 20 mM imidazole, and pH 8.5. C: Host cell DNA in pre‐column, flow through, and wash samples. D: PSD of feed stream from column buffered with 500 mM NaCl, 20 mM imidazole, and pH 8.5.

Figure 7

Comparison of crude and processed feed applied to a RFC system packed with large diameter resin. A: Fed‐batch shake flask growth of CHO cells expressing His‐tagged CEA. B: Viable cells/mL before and after passage through column. C: Protease activity in crude and processed feed streams. D: Percentage recovery of His‐tagged CEA from crude and processed feed streams.

Figure 8

Primary capture of His‐tagged CEA from 3 L fed‐batch bioreactor fermentation. A: Fed‐batch fermentation of CHO in SUT bioreactor. B: Time course of CEA expression in the bioreactor. C: Anti‐CEA Western blot of column eluate. D: Anti‐His Western blot of column eluate. All elutions and positive control were diluted to 1 μg/mL prior to western blotting.