(106a) Investigating Virus Retention Mechanisms During Parvovirus Filtration

Authors: 
Bakhshayeshi, M. - Presenter, The Pennsylvania State University
Zydney, A. - Presenter, The Pennsylvania State University
Kuriyel, R. - Presenter, PALL Life Sciences
Mehta, A. - Presenter, Genentech, Inc.
Hohwald, S. - Presenter, Genentech, Inc
Jackson, N. - Presenter, PALL Life Sciences
van Reis, R. - Presenter, Genentech, Inc.


Virus filtration provides a robust, size-based method for virus removal that compliments other unit operations to achieve the high levels of viral clearance required for the production of therapeutic proteins. Parvovirus filters provide significant removal of viruses as small as 20 nm, which is only a few times larger than the size of most therapeutic proteins. Commercially available parvovirus filters provide more than 4 LRV (log reduction value) in virus titer, however, several recent studies have reported a significant decline in virus retention during the course of filtration. The objective of this study was to examine the fundamental mechanisms governing virus retention during parvovirus filtration and to identify the factors leading to the observed reduction in LRV. Experiments were performed with Pall Ultipor DV20® virus filters made from a hydrophilic PVDF. Bacteriophage pp7 and nanocolloid gold particles were used to study the retention characteristics. The location of the gold particles within the filter were studied by transmission electron microscopy, while the pp7 were labeled with Alexa Fluor® 488 so that the filters could be imaged by confocal scanning laser fluorescence microscopy. Data were obtained both in the presence and absence of protein to determine the role of membrane fouling on the observed decline in virus retention. Experiments performed with highly purified phage suspensions showed a small decline in LRV over time, even in the complete absence of any flux decline. In addition, the LRV for filters that were previously fouled with IgG showed no loss in initial phage retention. These results clearly demonstarte that the LRV decline for the DV20 was not directly related to membrane fouling, in contrast to previous hypotheses presented in the literature. Confocal images indicate that the majority of the phage was captured within the membrane interior, with the concentration profile varying during the filtration. Similar results were obtained with the colloidal gold particles using TEM. The capture phenomena was dependent on the ratio of the particle to pore size, with very different behavior for different flow orientations due to the variation in pore size through the depth of the DV20 filter. These results provide important insights into the factors controlling parvovirus retention and could provide a framework for the development of more robust parvovirus filters for biotechnology applications.

Topics: