30% of measured cell disruption, which ranged from 5 to 100%. For turbulent flow in a
capillary good agreement was also obtained between measured data and model
predictions, although the model still underestimated the cell death rate by 15%. Energy
dissipation rates were about 10^3 m^2 .s−^3 corresponding to eddy lengths of 5.9μm.
The technique is also very useful for comparing the fragility of different cells under
different conditions. Thus, Zhang et al. (1992a, 1993c) showed that cells in the
exponential phase of a batch culture are less fragile than cells in the death phase, which is
in accordance with the data obtained by Petersen et al. (1991) and Ramirez and
Mutharasan (1992). Furthermore, they showed that the cell fragility decreased in the
presence of Pluronic F68 (Zhang et al. 1992b). There was a difference between long-term
exposure, where 0.05% was sufficient and short-term exposure, where an effect was only
seen above 0.1%. This is also in accordance with observations by Ramirez and
Mutharasan (1991). Moreover, it is not difficult to imagine that a relation exists between
plasma-membrane fluidity and the bursting membrane tension.
In conclusion, shear stresses at which cell damage occurs are in the order of 1 N·m−^2
and higher. The protective additive Pluronic F68 strengthens the cell probably through
adsorption to the cell membrane. The shear stresses as measured in shear devices are,
however, average shear stresses measured in shear devices for exposure times longer than
minutes. The actual stresses cells are exposed to in these devices are somewhat higher
than the average bulk shear stresses. Furthermore, exposure times are a factor 100–
10,000 longer than those occurring in a bioreactor. Direct measurements of cell strength
may be more informative. However, in this case additional models are required to
connect the fragility of the cells to hydrodynamics of a bioreactor.
Cell-Bubble InteractionBavarian et al. (1991) were the first authors to show attachment of insect cells to air
bubbles. For successful cell attachment to air bubbles the free-energy change upon
attachment must be negative and contact between cell and bubble is required. Wu et al.
(1997) showed that cells attach to a water-hexadecane interface, thus demonstrating that a
cell contains hydrophobic parts that in principle can also interact with air-liquid
interfaces. Furthermore, they showed that addition of methyl-cellulose and Pluronic led
to a dramatic decrease in adsorption to the hexadecane. In conjunction with the finding
that Pluronic adsorbs to cells and that no stable droplets were formed at the interface,
they deduced that the decrease in hydrophobicity of the cells is due to interaction of the
cells with Pluronic and not due to adsorption of Pluronic to the hexadecane-water
interface. Jordan et al. (1994) have experimentally shown that surfactants like Pluronics
adsorb also to the bubble surface. Adsorption was due to convection rather than diffusion.
Contact of cells with a bubble surface that contained no surfactants resulted in direct cell
lysis, whereas contact with partially saturated covered bubble surface resulted in stable
attachment of the cells to the surface. When the bubbles were completely saturated by
adding 0.1% Pluronic or S% serum no interaction was observed. The fact that surfactants
also adsorb to cells was not taken into account in this study.
Absolom et al. (1983) developed a model to describe adhesion of bacteria to surfaces
based on changes in free enthalpy that in turn were related to interfacial tensions.
Likewise, Chattopadhyay et al. (1995b) presented a thermodynamic approach to predict
Multiphase bioreactor design 464