(43f) Evaluation of Anion Exchange Membranes for the Removal of Trace Impurities from Mammalian Cell Culture Derived Protein Solutions
AIChE Annual Meeting
Monday, November 17, 2008 - 10:10am to 10:30am
The single largest cost incurred during development/manufacturing of a biopharmaceutical protein product (such as a Monoclonal Antibody or ?mAb?) occurs during the purification process. Purification of a mAb usually involves three chromatography steps: a capture step from cell culture medium (such as Protein A affinity); an intermediate purification step to remove product-associated impurities and other major contaminates; and a polishing step to remove trace impurities such as virus, DNA, and other low concentration impurities. Owing to the negative charge present on the surface of DNA and most viruses, as well as many Host Cell Proteins (HCP), at neutral pH, anion-exchange (AEX) chromatography is employed often as the predominant polishing step for such impurities, operating under conditions that the engineered protein of interest does not bind but ?flows thru.? Historically, the biopharmaceutical industry has used strong quaternary amine (Q) AEX resins in their processes for removal of trace impurities. The advantage of membrane adsorbers over traditional resins are convective flow to all the binding ligands enabling binding at higher flow rates (not limited by pore diffusion), low pressure drops across the membrane, one-time use disposability, and increased virus access to binding ligands (virus particles only bind to resin bead surface, too large to diffuse into pores). Therefore, we seek to evaluate the potential of two commercially available AEX membrane adsorbers (Sartorius Sartobind® Q and Pall Mustang® Q) and one beta AEX membrane from Millipore (ChromaSorb? Prototype) to remove specific impurities: Minute Virus of Mice (MVM), HCP, and DNA. Results to date, with MVM as the only impurity, show maximum MVM removal factors of > 3.5 log for all membranes. The Sartobind Q shows a significant (p < 0.05) effect on LRV (log virus removal) with pH and NaCl concentration changes, while the Mustang Q shows a significant effect on LRV with changes in NaCl concentration only. Interestingly, the Millipore ChromaSorb? Prototype membrane showed no significant effect on LRV with changes in pH, NaCl conc., or flow rate. Ongoing/future work will demonstrate the trace impurity removal capabilities of all three membranes with and without the presence of a ?protein of interest? present, i.e. a monoclonal antibody. In addition, efforts will be made to develop a mathematical model to explain the removal potential of the membranes under different solution conditions and differing levels of binding solutes.