2024 AIChE Annual Meeting

(611c) Location of Virus Entrapment during Virus Filtration

Authors

Wickramasinghe, R. - Presenter, University of Arkansas
Xu, W., University of Arkansas
Shirataki, H., Asahi Kasei Biomedical
Qian, X., University of Arkansas-Fayetteville
Validation of virus contamination is essential during the downstream purification of biopharmaceutical products. In the case of protein based therapeutics, virus filtration is routinely used to validate virus clearance. Adventitious virus clearance is frequently validated using minute virus of mice (MVM) a FDA recommended model parvovirus 18-22 nm in size. Virus filters are run in dead end mode. They are designed to achieve 4 or more log (10,000 fold) virus clearance in the permeate relative to the feed. In addition, more than 95% product recovery in the permeate is required. Often the product of interest may have a hydrodynamic radius in excess of 10 nm leading to a very challenging separation. In practice the performance of the virus filters has been shown to be affected by the product species.

In this study the effect of a monoclonal antibody (mAb) as well as BSA on the location of capture of MVM in three different parvovirus retentive filters has been studied. All experiments were run under constant flux conditions. The membranes were challenged with 1,000 L m-2 of feed. The concentration of aggregates for the mAb was less than 1% but 15 % (dimer and higher oligomers) for BSA. The MVM and BSA were labelled with a fluorescent dye. Laser scanning confocal microscopy was used to locate the position of virus retention on the membrane. The presence of mAb or BSA shifted the location of MVM retention toward the permeate side of the membrane. The effect was more significant for BSA. BSA aggregate retention occurred closer to permeate side of the membrane compared to MVM. The inclusion of a process pause also led to virus retention close to the permeate side of the membrane due to back diffusion of the virus particles. All three virus filter showed virus clearance in excess of 4 log under all conditions. BSA and mAb recovery was over 99%. The result indicate the effect of product aggregates as well as interactions between product aggregates and MVM on virus clearance. Finally we have developed a ‘multilayer’ model of the virus filtration membrane in order to describe the experimental findings.