A critical issue in the design of bio-culture reactors is that cells are sensitive to a wide range of influences, which can, either singly or in combination, reduce cell productivity and viability.
Influences that might disturb cells include physical damage due to agitation, inadequate nutrient distribution, temperature control or waste disposal, population dynamics, diseases and foreign invasions.
In addition, biological studies show that cell productivity can be affected by low levels of prolonged shear, but cells may also be capable of recovering from, and even being stimulated by, certain levels of mechanical stress.
Traditional methods of reactor design rely on combinations of biochemical insight, engineering calculation and time-consuming trial and error, often contributed by different specialists.
With respect to damage due to agitation, the traditional approach has been to experimentally determine survival rates at various shear levels, and to then try to design reactors that do not exceed them.
As reactor size increases so does the power of the agitators, however, increasing the potential for mechanical stress as well as damage. Crucially, empirical approaches fail to consider how the cumulative effects of stress and recovery as cells move around a reactor might affect productivity.
We have achieved this by first modelling the flows inside different reactor designs, and then tracking the shear exposure history of cells as they move around the reactor. These hypothetical shear history traces are then used to drive a biological damage model, including recovery, to produce an estimate of the cell performance in that particular reactor.
In our analysis, we have studied a cylindrical reactor with two impellors in which the angle of the impellor blades is the main variable, i.e. a pair of radial Rushton impellors or a pair of axial impellors in the same tank (Figure 1).
Our model allows us to quantify differences between these systems in terms of peak shear values, the frequency of their occurrences and the general shear stress durations experienced by tracer cells. This reveals markedly different shear histories in the two bio-reactors (Figure 2).
Figure 3 shows the results of combining this data with a shear damage-recovery model. In the example, cells in the radial Rushton impellor-driven reactor recover from periodic shear stress, thanks to intervening periods in environments of relative calm, while cells in the double axial reactor suffer escalating shear stress. We are thus in a position to make specific estimates of cell productivity and viability in different bio-reactor designs.
This approach, which we have also used to examine the fluctuating nutrient environment, allows us to integrate dedicated single-cell stress and other micro-response data with macroscopic reactor design elements. This provides a route to rapid development, optimised scale-up and efficient exploitation of valuable bio-cultures.