Many studies have focused on the identification of optimal aggregate size distributions
for metabolite production (e.g. Hulst et al., 1989; Hanagata et al., 1993). However,
Keβler et al. (1999) recently demonstrated that, on the basis of active biomass alone (i.e.
excluding stored carbohydrates), specific ajmalicine production in suspensions of
Catharanthus roseus was independent of aggregate size, for aggregate diameters of less
than 250 μm. What is required for effective optimisation of any given cell line is the
identification and maintenance of conditions supporting optimal productivity.
Broth RheologyBy comparison with non-filamentous microbial systems, plant cell suspensions are
commonly perceived as highly viscous fluids. However, it is important to note that plant
systems are typically operated at significantly higher biomass levels than microbial
suspensions. Curtis and Emery (1993) found that the apparent viscosities of plant cell
suspensions were similar to those of yeast suspensions (Reuβ et al., 1979), at cell volume
fractions of up to 0.3. Although Theological properties are strongly line-dependent and
may also vary with cultivation conditions, data collected for a variety of suspensions give
a good picture of typical trends. Relevant studies are summarised by Kieran et al. (1997).
Analysis of the available data indicates that many suspensions exhibit non-Newtonian,
shear-thinning characteristics. At higher biomass concentrations, some suspensions show
evidence of a yield stress (e.g. Wagner and Vogelmann, 1977; Zhong et al., 1992b).
Thixotropic behaviour has been occasionally observed in some cell lines (e.g. Wagner
and Vogelmann, 1977), but the data are extremely limited.
Broth rheology is strongly dependent on biomass concentration. For Cudrania
tricuspidata, C. roseus and N. tabacum suspensions, Tanaka (1982) observed that broth
apparent viscosity varied with biomass concentration (DW) raised to the power of 6.5.
More recently, Curtis and Emery (1993) correlated broth viscosity with biomass (FW)
concentration, expressed in terms of cell volume fraction. Because viscosity depends on
the fractional volume of solid material present in the suspension, correlation with cell dry
weight is less appropriate during the later stages of a batch growth cycle (Zhong et al.,
1992b) when, due to cellular expansion, fresh and dry weight profiles diverge.
In addition to biomass effects, increased broth viscosity is often attributed to the
presence of ECPs in the suspending fluid. In cases where the ECP is the product of
interest, high polysaccharide concentrations can significantly increase broth viscosity
and, in turn, reduce mass transfer coefficients. Takei et al. (1995) reported mass transfer
limitation in suspensions of Polianthes tuberosa, which are employed by Kao for ECP
production. The viscosity of a 4 g L−^1 aqueous ECP solution was 550 cP. Mineral salt
supplements reduced solution viscosity and the consequent increase in mass transfer rates
facilitated enhanced ECP production, to a level of 6.5 g L−^1 ; using a fed-batch system,
the current ECP productivity is in excess of 20 g L−^1 month−^1 (Otsuji, 1999).
In general, however, only negligible or modest changes in filtrate viscosity have been
recorded over the course of the growth cycle for many cell lines. Early studies of N.
tabacum (Kato et al., 1978) revealed pseudoplastic behaviour, with the broth apparent
viscosity increasing by a factor of 27, while that of the cell-free broth varied only
between 0.9 cP and 2.2 cP. Figure 14.2 shows broadly similar trends for suspensions of
(a) M. citrifolia (Cusack and Kieran, 1998) and (b) Papaver somniferum (Curtis and
Bioreactor design for plant cell suspension cultures 423