maximum allowed levels of these contaminants in the final product. Host cell DNA

might be very sticky and some viruses might attach to either host cell debris or host

cell DNA complicating clarification steps (depth filtration, centrifugation) after

virus harvesting. Salt concentrations/osmolality of media as well as choice of

membrane material and cut-off of the respective filters will need to be screened

thoroughly as each medium and cell line will give a different background and

change of harvest time point will immediately change the cell broth composition.

In order to determine the optimal harvest time point, the ratio of total virus titer to

contaminants should be considered. For attenuated vaccines and viral vectors, a high

infectious titer is necessary to achieve high potency, hence, an early harvest time point

should be targeted. Certain viruses show a low stability, which is characterized by a

steep decrease in infectivity over time. One possible countermeasure is a multiple

harvest strategy, in which the virus is harvested and stored at lower temperature and

new medium is added to the bioreactor. This strategy is often used for adherent cells

and slowly propagating lytic viruses. Moreover, continuous harvesting with sub-

sequent cooling to prevent degradation could also be applied (see chapter 6).

Specific aspects of the intracellular virus replication cycle on process perfor-

mance can also not be neglected. Some viruses bud from the (apical) cell membrane

during virus release and, hence, carry a lipid bilayer as an envelope (see Table 5.9).

Such enveloped viruses are often less stable at higher temperature, sensitive to

lower pH values and fast degraded by contact with detergents; all resulting in in-

fectivity losses. For a few viruses (e.g., MVA), a considerable number of virus

particles remain within the cell. For reaching maximum virus yields, freeze-thaw

cycles or the use of high-pressure homogenizers is recommended to disrupt the cell

membrane and to release the virions.

Another factor to consider during viral vaccine production is the effect of shear

stress (e.g., agitation, pumping) and aeration (e.g., O2 and CO2). High shear forces

can hinder virus binding to the cells or can lead to early cell death and with that to

lower virus titers. Cells go either into apoptosis due to virus infection or into ne-

crosis due to shear stress. For some viruses, such as IAV, the right timing of

apoptosis induction is important for virus release and with that virus yield. It is thus

not as simple as just trying to avoid cell death for as long as possible, to keep cells

productive. It will always be a combination of parameters and events that will result

in higher virus yields. In principle, apoptosis is hallmarked by DNA fragmentation,

plasma-membrane blebbing, and creation of apoptotic bodies (fragmentation) [85].

Different methods (e.g., imaging flow cytometry, NMR spectroscopy, proteomic

approaches) are now established to further investigate cellular bottlenecks during

viral vaccine production that might result in low virus yields (see also chapter 8).

This understanding might help to identify high-producer cells and to optimize

production processes.

Finally, the vaccine type (live-attenuated, inactivated, vector) also has a sig-

nificant impact on design and optimization of virus production processes. Many

attenuated vaccine strains show a lower replication rate and, thus, often reduced

virus yields. In contrast, manufacturing processes for inactivated vaccines that

comprise infectious and non-infectious virus particles often display very high titers.

For production of pathogenic viruses without an option to vaccinate employees

118

Bioprocessing of Viral Vaccines