Wumu Xu, Xianghong Qian, Hironobu Shirataki, Daniel Strauss, S. Ranil Wickramasinghe
Virus filtration is used for validation of virus clearance in the manufacture of monoclonal antibodies. Requirements for this size exclusion process include 10,000-fold reduction of virus particles and at least 95% recovery of the monoclonal antibody in the permeate. The size difference between the rejected parvovirus particles, 25 nm, and monoclonal antibody, 10-12 nm, is about two. Membranes are designed with pore sizes around the average size of the parvovirus. Given the polydispersity in membrane pore size and virus particles, filter performance is dependent on membrane properties, feed and operating conditions. Compromised performance usually occurs due to product related foulants such as irreversible and reversible aggregates. Three commercial asymmetric hollow fiber membranes were investigated. The location of minute virus of mice entrapment within the membrane was determined by laser scanning confocal microscopy in the presence and absence of a monoclonal antibody or bovine serum albumin that forms irreversible aggregates. Experiments were conducted under constant flux for 12 hours and constant pressure for 12.5-40 hours. Membrane throughput was 900 L m -2 and effective parvovirus removal (LRVs above 4) was verified under all conditions tested. Lower permeate fluxes and longer filtration times lead to virus capture deeper within the membrane. Moreover, product-membrane interaction at low fouling conditions led to the migration of virus particle deeper within the membrane. Product aggregates, similar in size to virus particles led to displacement of virus particles deeper within the membrane and broadening of the entrapment zone. These results provide unique insights into virus filter performance. • Visualized virus capture location in membrane by laser scanning confocal microscopy • Lower fluxes lead to virus displacement deeper into membrane • Presence of monoclonal antibody leads to virus displacement deeper into membrane • Presence of aggregates similar in size to virus lead sot broadening of capture zone • Membrane polymer affects location of virus capture zone