Protein A chromatography is ubiquitous to antibody purification. The high specificity of Protein A for binding the Fc-region of antibodies and related products enables unmatched clearance of process impurities like host cell proteins, DNA, and virus particles. A recent development is the commercialization of research-scale Protein A membrane chromatography products that can perform capture step purification with short residence times (RT) on the order of seconds. This study investigates process-relevant performance and physical properties of four Protein A membranes: Purilogics Purexa™ PrA, Gore® Protein Capture Device, Cytiva HiTrap™ Fibro PrismA, and Sartorius Sartobind® Protein A. Performance metrics include dynamic binding capacity, equilibrium binding capacity, regeneration-reuse, impurity clearance, and elution volumes. Physical properties include permeability, pore diameter, specific surface area, and dead volume. Key results indicate that all membranes except the Gore® Protein Capture Device operate with flow rate-independent binding capacities; the Purilogics Purexa™ PrA and Cytiva HiTrap Fibro™ PrismA have binding capacities on par with resins, with orders of magnitude faster throughput; and dead volume and hydrodynamics play major roles in elution behavior. Results from this study will enable bioprocess scientists to understand the ways that Protein A membranes can fit into their antibody process development strategies.
Protein A chromatography is the preferred unit operation for purifying Fc-based proteins. Convective chromatography technologies, like membrane adsorbers, can perform the purification rapidly and improve throughput dramatically. While the literature reports the preparation of Protein A membrane adsorbers utilizing traditional coupling chemistries that target lysine or thiol groups on the Protein A ligand, this study demonstrates a new approach utilizing copper-free dibenzocyclooctyne (DBCO)-azide click chemistry. The synthetic pathway consists of three main steps: bioconjugation of Protein A with a DBCO-polyethylene glycol (PEG) linker, preparation of an azide-functionalized membrane surface, and click reaction of DBCO-Protein A onto the membrane surface. Using polyclonal human immunoglobulins (hIgG) as the target molecule, Protein A membranes prepared by this synthetic pathway showed a flowrate-independent dynamic binding capacity of ~10 mg/mL membrane at 10% breakthrough. Fitting of static binding capacity measurements to the Langmuir adsorption isotherm showed a maximum binding (qmax) of 27.48 ± 1.31 mg/mL and an apparent equilibrium dissociation constant (Kd) of value of 1.72 × 10−1 ± 4.03 × 10−2 mg/mL. This work represents a new application for copper-less click chemistry in the membrane chromatography space and outlines a synthetic pathway that can be followed for immobilization of other ligands.
The development of convective technologies for antibody purification is of interest to the bioprocessing industries. This study developed a Protein A membrane using a combination of graft polymerization and copper(I)-catalyzed alkyne–azide click chemistry. Regenerated cellulose supports were functionalized via surface-initiated copolymerization of propargyl methacrylate (PgMA) and poly(ethylene glycol) methyl ether methacrylate (PEGMEMA300), followed by a reaction with azide-functionalized Protein A ligand. The polymer-modified membranes were characterized using attenuated total reflectance Fourier-transform infrared spectroscopy (ATR-FTIR), gravimetric analysis, and permeability measurements. Copolymer composition was determined using the Mayo–Lewis equation. Membranes clicked with azide-conjugated Protein A were evaluated by measuring static and dynamic binding (DBC10) capacities for human immunoglobulin G (hIgG). Copolymer composition and degree of grafting were found to affect maximum static binding capacities, with values ranging from 5 to 16 mg/mL. DBC10 values did not vary with flow rate, as expected of membrane adsorbers.