Taylor-Vortex bioreactors for enhanced mass transport

Abstract

The route by which engineered tissue products are manufactured in vitro commonly utilises dynamic tissue culture bioreactors in which cell culture media and proliferating cells interact. Very often, the cells are supported on biodegradable polymer constructs that assist in the development and formation of the 3-D architecture of the tissue. It has been observed experimentally that a core of necrotic tissue often forms within these constructs, and this has been attributed to poor mass transport of nutrients and oxygen from the culture media (Freed et al. 1996; Vunjak-Novakovic et al., 1998, Obradovic et al., 1999). Mixing within the liquid media provides a more uniform concentration of nutrients and oxygen and facilitates their delivery to cells through enhanced convective transport. The degree of mixing can be problematic, however, as cells have a thin permeable membrane and are sensitive to shear forces. As the mixing intensity increases, the size of circulating eddies decreases, which leads to an increase in the shearing forces that are generated within the fluid, and on the cells and constructs that it contains (Papoutsakis, 1991). Exposure to critically high shear may result in irreparable damage to the cell membrane, leading to cell death, as can prolonged exposure to sub-critical shear stresses. Nevertheless, shear stress has an important role to play in the stimulation of mechanotransduction pathways, which can, in turn, influence the physiology of the cell (Enfors et al., 2001). Ideal mixing in the bioreactor is difficult to achieve in practice, therefore a compromise between shear and mass transport must be made (Vunjak-Novakovic et al., 1999). For a functional tissue sample to be engineered in vitro, a uniform physico-chemical environment must be maintained over time for normal morphogenesis of the cells (Spier, 1995). Individual cells might be considered to grow within their own microenvironment, where they experience dynamic physical and chemical changes, which must be controlled within acceptable limits. It follows that the bioreactor can be considered as a unit that is comprised of these microenvironments (Tramper, 1995). In practical terms, it is not possible to maintain this degree of uniformity, even in a very small bioreactor, but this remains a long-term goal. Traditional spinner flask reactors suffer considerably from the fact that they cannot generate a uniform flow and mass transport environment (Vunjak-Novakovic et al., 1996). Well-mixed high shear regions exist around the rotating impeller whereas laminar regions, or even dead zones exist far from the impeller, exacerbated at larger scales of reactor (Leib et al., 2001). Techniques to overcome this can be employed but have a tendency to make the bioreactor design more complex and, importantly, more difficult to scale correctly such that the transport phenomena are replicated. The long-term term goal of tissue engineering is to commercialise the in-vitro production of functional tissues and for this to be realised scaling of the production process is of paramount importance (Langer and Vacanti, 1993; Lewis, 1995; Naughton, 1998). The Taylor-vortex bioreactor is a promising alternative to achieve many of these goals and comprises two coaxial cylinders, either of which can rotate. Media, cells and constructs occupy the annular space between the cylinders. The system has a simple geometry making it relatively easy to scale. The flow regimes that can be achieved within this type of bioreactor are diverse from laminar through to highly turbulent flows, with many unique regimes in between as a consequence of the geometry. Importantly, these regimes are uniform throughout the bioreactor with no dead zones and there is a high degree of control over the maintenance and transition between regimes. This review chapter addresses the design aspects of a Taylor-vortex bioreactor and describes approaches used to control and quantify the flow and mass transport environment with a view to developing mass transport correlations for scaling purposes. © 2005 Springer.