Growth and injury of freely suspended animal cells in an agitated and surface-aerated bioreactor
Kunas, Kurt T.
McIntire, Larry V.
Doctor of Philosophy
The mechanism of animal cell damage due to agitation and aeration in a stirred bioreactor and the effect of media components on cell damage were investigated. Hybridoma cells ATCC CRL-8018 were grown in controlled bioreactors of 1-2 liter working volume, each agitated with a 7 cm-diameter impeller having pitched blades 1.8 cm wide by 7 cm tall. The experimental agitation speed was imposed at a consistent point in the exponential phase of batch growth in each culture. The apparent growth rate and cell yields on substrates and metabolites were used to assess the cell response. In bioreactor liquid volumes of 1.0-1.2 liters in the presence of a gas headspace, severe cell damage was caused at agitation speeds of 190-220 rpm as 1-3 mm bubbles were entrained at the vortex tip, moved through the bulk liquid, then burst at the liquid surface. Cell damage in this agitation range was not caused by turbulent eddies in the bulk fluid or merely by the presence of entrained bubbles, since cell growth was possible up to 700 rpm in the presence of large numbers (millions) of 50-300 $\mu$m bubbles when no gas headspace was present in the reactor. When bubble entrainment was eliminated in a completely filled reactor, reduced cell growth was found at 700-800 rpm, which corresponds to a Kolmogorov eddy size close to the cell size. When cell damage was caused by vortexing and bubble entrainment, increasing the serum concentration over a range of 1-10% decreased cell damage. The effect of serum was modeled with a saturation-type expression. The protection mechanism of serum was not due to a change in the medium's surface tension or viscosity. Increasing the viscosity over a range of 0.9-2.1 centipoise by the addition of 0-3% dextran increased the severity of cell damage but did not influence the agitation speed at which damage occurred. Over a wide range of values, modeling the apparent growth rate as a difference between an unchanging, intrinsic growth rate and a variable death rate was consistent with the experimental results. Also, the ratios of the apparent cell yields on glucose and lactate to the apparent growth rate remained constant, which is consistent with the above model.
Chemical engineering; Microbiology; Biochemistry