Minireview
Biochemical Engineering Research Group, School of Biological Sciences, Dublin City Uni6ersity, Dublin 9, Ireland
b
Department of Chemical Engineering, Uni6ersity College Dublin, Belfield, Dublin 4, Ireland
Received 1 October 1996; received in revised form 3 February 1997; accepted 25 July 1997
Abstract
Higher plants are the source of a vast array of biochemicals which are used as drugs, pesticides, flavourings and
fragrances. For some of these compounds, plant cell culture can provide a potential production alternative to
traditional cultivation methods or chemical synthesis routes. Many systems have been patented and the last 20 years
have seen considerable industrial and academic interest in the development of large scale cultures to produce
pharmaceutically active, high value substances. However, the industrial application of plant cell suspension cultures
has, to date, been limited. Commercialisation has essentially been impeded by economic feasibility, arising from both
biological and engineering considerations. This paper reviews the commercial development of the technology to date
and focuses on the impact of specific engineering-related factors, in particular, the shear sensitivity of plant cell
suspension cultures. Evidence of sensitivity to hydrodynamic shear in bioreactors has generally been attributed to the
physical characteristics of the suspended cells. Recent studies indicate that shear sensitivity may not be as important,
in some cases, as initially anticipated. 1997 Elsevier Science B.V.
Keywords: Plant cells; Shear sensitivity; Bioreactor; Scale-up
1. Introduction
Higher plants are recognised as important
sources of a wide range of biochemicals, used as
drugs, pesticides, flavourings and fragrances. Tra* Corresponding author. Tel.: + 353 1 7045584; fax: + 353
1 7045412; e-mail: kieranp@ccmail.dcu.ie
40
2. Commercial development
That plant cell culture offers the tantalising, but
as yet largely unattained prospect of long-term
commercial success, is evidenced in many recent
reviews concerning the technology and strategies
for its optimisation (Zenk, 1991; Verpoorte et al.,
1991; Scragg, 1992, 1995; Buitelaar and Tramper,
1992; Kreis, 1993; Shuler, 1993; Su, 1995; DiCosmo and Misawa, 1995; Dornenburg and
Knorr, 1995; Zhong et al., 1995; Schlatmann et
al., 1996). Despite its enormous potential, industrial processes involving plant cell cultures have
been limited to a handful of applications, including the production of shikonin from Lithospermum erythrorhizon (Fujita, 1988), berberine from
Coptis japonica (Fujita and Tabata, 1987) and
ginsenosides from Panax ginseng (Ushiyama,
1991). Many other processes have been investigated and patented but, to date, few have proven
to be economically viable. Plant cell cultures have
predominantly been valued as a source of naturally occurring secondary metabolites. However,
they can also be used for biotransformations such
as the glucosylation of hydroquinone to arbutin
by Rau6olfia serpentina (Lutterbach and Stockigt,
1992). In Japan, workers at Shiseido have alternatively employed C. roseus for the production of
arbutin by a similar process (Yokoyama and
Yanagi, 1991; Inomata et al., 1991). As a source
of enzymes for the genetically engineered synthesis of natural products (Scott, 1994), they offer
great promise. And large-scale processes for the
use of cell cultures as a source of biomass (Hashimoto and Azechi, 1988) have been investigated.
41
ever, the calculated product prices vary substantially with the choice of production strategy (e.g.
batch culture with product retained intracellularly; spontaneous or induced product release,
etc.). However, economic feasibility studies are
unanimous in acknowledging productivity as a
limiting factor. For example, Drapeau et al.
(1987) estimated that a 40-fold increase in the
ajmalicine productivity of C. roseus would be
required to justify the production of this compound by cell culture methods.
With regard to secondary metabolites, the productivity of plant cell suspensions is of the order
of 10 2 10 1 g l 1 per day (Scragg, 1995). Due
to the fact that productivity depends on product
yield, the organism growth rate and prevailing
biomass levels, it is clear that there is a role for
both biologists and engineers in improving system
performance. The question of economic feasibility
could be largely resolved by the development of
stable, high-yielding strains (Berlin, 1988) and the
identification of optimal cultivation conditions.
While the high biomass levels required for economic viability necessarily limit the volume of free
medium available, in an ideal system the cells
would actively secrete the product into the suspending process fluid (Buitelaar and Tramper,
1992), thereby simplifying downstream processing. However, to achieve these objectives, a fuller
understanding of the biosynthetic pathways is essential, in addition to a clearer picture of the
interactions between growth kinetics, system morphology, cellcell interaction and product synthesis.
3. Engineering considerations
From an engineering perspective, the primary
challenges lie in the area of process scale-up. Plant
cell suspensions can now be almost routinely cultivated in small-scale configurations. Moreover, in
addition to the commercial processes mentioned
earlier, there are a number of examples of largescale, albeit non-commercial systems. Nicotiana
tabacum cultures have been successfully cultivated
in a 20 m3 stirred tank reactor (Noguchi et al.,
1977). Westphal (1990) reported on long-term cul-
42
Table 1
Summary of biocatalyst characteristics
Biocatalyst
Shape
Size (mm)
Cell wall
Aggregates
Plant cells
Spherical
Cylindrical
l: 100500
d: 2050
Yes
Yes
20 100
101
Animal cells
Spherical
d: 1020
No
No
20
101
Bacteria
Spherical
Cylindrical
Spiral
d: B1
l: B5
Yes
Yes/No
Yeasts
Spherical
d: 510
Yes
Yes/No
10
102
Moulds
Mycelial
d: 510
l: B100
Yes
Yes/No
10 20
102
0.5 10
103
3.1.1. Aggregation
Plant cells are significantly larger and slower
growing than most microbial organisms. Individual plant cells have a typically characteristic
length of the order of 101 102 mm and may be
spherical to cylindrical in shape. Aggregation is
common, largely due to the failure of cells to
separate after division, although the secretion of
extracellular polysaccharides (ECP), particularly
in the later stages of batch growth, may contribute to increased aggregation (Taticek et al.,
1991). The aggregation phenomenon has been
used in the development of self-immobilisation
methods (Prenosil et al., 1987; Hegglin et al.,
1990). Aggregates, comprising up to 102 cells, may
be many millimetres in diameter (Tanaka, 1982)
and typically exhibit a tendency to settle. Aggregation patterns, variously studied using image
analysis (Kieran et al., 1995) and sieving (Mavituna and Park, 1987) techniques, vary significantly between cell lines and also as a
consequence of culture age and cultivation conditions. For example, N. tabacum, which is frequently
described
as
highly
aggregated
(Hashimoto and Azechi, 1988; Hooker et al.,
1989; 1990) exists, under certain conditions as
unbranched chains of up to 50 cells (DeJong et
al., 1967). Zenk et al. (1975) reported that 60% of
a Morinda citrifolia culture existed as single cells
or cells of two chains, whereas, in a study by
Kieran et al. (1993) the corresponding figure lay
between approximately 10% (lag and stationary
3.1.2. Rheology
Aggregation, aggregate interactions, high
biomass concentrations (e.g. up to 70 g l 1 on a
dry weight basis (Matsubara and Fujita, 1991))
and, in some cases, ECP secretions, result in high
whole-broth viscosities for plant cell suspensions.
There have been no comprehensive on-line rheological studies. Data have been collected for a
variety of plant cell systems using conventional
viscometers, although, as with other microbial
suspensions, care must be taken to avoid sedimentation effects (Scragg et al., 1986) and/or aggregate disruption (Rosenberg, 1987). Using the
concept of the apparent viscosity of a fluid (Metzner and Otto, 1957) rotational devices fitted with
purpose-built impellers have also been used for
the rheological characterisation of plant cell suspensions: for example, helical ribbon impellers,
designed to satisfy the need for good suspension
mixing at the widest possible range of laminar
flow conditions (Jolicouer et al., 1992; Kieran,
1993). A summary of relevant studies is presented
in Table 2 and it is apparent that the majority of
suspension cultures investigated exhibit non-Newtonian, shear-thinning characteristics. Many systems also show evidence of a yield stress.
Thixotropic behaviour has been observed, although only in isolated samples (Wagner and
43
Vogelmann, 1977). In common with many microbial suspensions, the apparent viscosity is found
to be strongly dependent on biomass concentration (Tanaka, 1982; Zhong et al., 1992a; Kieran,
1993), although the data reported in the literature
do not, in general, refer to the high density suspensions required for economically viable largescale processing. The influence of the morphology
of the suspended cells and/or aggregates on the
apparent viscosity of the suspension requires further investigation (Zhong et al., 1992a; Curtis and
Emery, 1993); it should also be noted that the use
of biomass concentration (on a dry weight basis)
as a correlating factor for viscosity can mask
effects attributable to variations in individual cell
size and water content over the course of the
growth cycle.
44
Table 2
Rheological characterisation of plant cell suspension cultures
System
Characterisationb
Viscometric device
Reference
Papa6er somniferum
Nicotiana
tabacum
Batch cultivation
Semi-continuous
Nicotiana
tabacum
Perilla frutescens
B14.5
Newtonian
= 9.0
Pseudoplastic,
n= 0.6
Newtonian
Datura stramonium
=7.9
513
Brookfield-type
B20
Pseudoplastic,
n :0.7
Bingham plastic
Brookfield-type
Casson plastic
Zhong et al.
(1992a)
Ballica et al.
(1992)
Ballica and Ryu
(1993)
Kieran (1993)
Catharanthus
527
roseus
Cudriana tricuspi- 515
data
Catharantus
roseus
Nicotiana
tabacum
a
b
Pseudoplastic,
n: 0.8
Pseudoplastic,
0.1BnB0.9
Pseudoplastic
Jolicouer et al.
(1992)
Tanaka (1982)
n: 0.53
culture is inhibited; reduced productivity observed at higher gassing rates is variously attributed to the stripping of CO2 and essential
volatiles from the system (Kato et al., 1975;
Ducos and Pareilleux, 1986) or shear-related effects (Ballica and Ryu, 1993). Using a gas recirculation bioreactor for scale-up studies
involving C. roseus, Schlatmann et al. (1993)
confirmed the importance of dissolved gaseous
components for system performance and concluded (Schlatmann et al., 1994) that loss of an
unidentified essential volatile factor was responsible for the reduced ajmalicine synthesis observed on scale-up from a shake flask to an
aerated bioreactor.
Aeration of bioreactors can lead to foaming
and in plant cell suspension cultures this prob-
45
46
Table 3
Methods used for the assessment of shear-related effects in plant cell suspension cultures
System response
Parameter measured
Reference
Reduction in viability
Growth rate
Regrowth potential
Membrane integrity
Dye exclusion
Dual isotope labelling
Dielectric permitivity
pH variation
Protein release
Total organic carbon
Secondary metabolite release
Change in metabolism
Ho et al. (1995)
Rosenberg (1987), Takeda et al. (1994), Zhong et al.
(1994)
Takeda et al. (1994)
Hooker et al. (1989); Zhong et al. (1994)
Tanaka et al. (1988)
Aggregate size/shape
Expansion index
47
Table 4
Defined flow-field shear experiments using plant cell suspension cultures
Cell suspension
Apparatus
Flow regime
Damage indicator
Reference
Morinda citrifolia
Laminar/turbulent
Submerged jet
Turbulent
Dye exclusion
Morphology
Dye exclusion
Morphology
Daucus carota
Couette viscometer
Laminar
Morphology
Regrowth
Cell lysis
Nicotiana
tabacum
Couette-type
Transitional/turbulent
Viability
Hooker et al. (1989)
Cell lysis
Metabolite production
nar flow conditions in a viscometric device, Vogelmann et al. (1978) suggested a critical shear stress
of between 80 and 200 N m 2 for M. citrifolia;
using regrowth potential as an indicator of system
response, a critical shear stress of 50 N m 2 was
identified for suspensions of Daucus carota
(Rosenberg, 1987). Recent studies (Dunlop and
Namdev, 1993; Kieran et al., 1995; MacLoughlin
et al., 1997) have pointed to the use of energy
dissipation as a correlating factor for shear-related damage. However, given the variety of aggregate morphologies exhibited by plant cell
systems, no single mechanism for cell damage has
been conclusively identified. Because these devices
are generally operated under non-sterile conditions, system response is most frequently monitored in terms of loss of viability and accordingly,
more subtle, non-lytic effects may be overlooked
(Namdev and Dunlop, 1995).
There are a number of important conclusions to
be drawn from these studies which have significant consequences for the success of commercial
systems in conventional bioreactors. Interestingly,
it appears that plant cell suspensions are more
robust than initially anticipated and that significant losses of viability are not, in general, to be
expected under normal operating conditions in a
conventional STR. However, other less dramatic
shear-related effects may be observed, including
Respiration activity
4. Conclusions
This paper has considered only a limited number of engineering-related issues which impinge
48
upon the commercialisation of plant cell suspension technology. The overall focus of related research is the achievement of large-scale cultivation
of high-yielding cell lines and efficient product
recovery, through the integration of biologicallyand engineering-based approaches. Process modelling has been successfully used as a tool for the
design, analysis and optimisation of many biological systems, including plant cell suspension cultures (Hooker and Lee, 1992; van Gulik et al.,
1993), but there is scope for development of models which take account of the fact that suspensions may not be homogeneous with respect to
metabolite synthesis. In the area of process control, a large-scale operation of plant systems is
hampered by difficulties associated with the online analysis of parameters such as cell or aggregate size distributions which may play an
important role in metabolite productivity. Different operating strategies (e.g. batch, two-stage,
continuous, etc.) may be required for cell lines
exhibiting different metabolite synthesis kinetics
(e.g. growth associated and non-growth associated) and which either store the product of interest intracellularly or, preferably, secrete it into the
suspending fluid. Non-lethal methods for inducing
product release and the use of two-phase systems
(Buitelaar and Tramper, 1992) can facilitate continuous processing without loss of biomass. The
product can be concentrated and by removing
feed-back inhibition effects, productivity may be
improved. Immobilisation (Hulst and Tramper,
1989) offers the possibility of a low-shear environment, enhanced cell cell contact, biomass re-use
and, if product secretion can be achieved, a facility for product removal. Some of these advantages have already been demonstrated using
self-immobilised plant cell systems in fluidised-bed
reactors (Dubuis et al., 1993; Khlebnikov et al.,
1995). Despite the obvious attractions of these
integrated processing approaches, problems of
both economic and technical feasibility must be
resolved before their application on an industrial
scale can be considered. Through concerted research efforts in the areas discussed in this paper
and in those highlighted above, wide-scale commercialisation of plant cell suspension technology
may soon be feasible.
References
Alfermann, A.W., Petersen, M., 1995. Natural product formation by plant cell biotechnology. Plant Cell Tissue Organ
Cult. 43, 199 205.
Anonymous, 1994. ESCA scales up taxol. Chemical Marketing
Reporter. May 24, p. 7.
Ballica, R., Ryu, D.D.Y., 1993. Effects of rheological properties and mass transfer on plant cell bioreactor performance: production of tropane alkaloids. Biotechnol.
Bioeng. 42, 1181 1189.
Ballica, R., Ryu, D.D.Y., Powell, R., Owen, D., 1992. Rheological properties of plant cell suspensions. Biotechnol.
Prog. 8, 413 420.
Berlin, J., 1988. Formation of secondary metabolite in cultured plant cells and its impact on pharmacy. In: Bajaj,
Y.P.S. (Ed.), Biotechnology in Agriculture and Forestry:
Medicinal and Aromatic Plants I, vol. 4. Springer-Verlag,
Berlin, pp. 37 59.
Bond, P.A., Fowler, M.W., Scragg, A.H., 1988. Growth of
Catharanthus roseus cell suspensions in bioreactors: on-line
analysis of oxygen and carbon dioxide levels in inlet and
outlet gas streams. Biotechnol. Lett. 10, 713 718.
Bringi, V., Kadkade, P.G., Kane, E.J., Prince, C.L., Roach,
B., Schubmehl, B.F., 1995. High yield production of taxol
and taxane production from Taxus species cell cultures: by
cultivation in a nutrient medium cells derived from callus
and/or suspension cultures of Taxus species useful for
cancer treatment, US Patent 5 407 816.
Buitelaar, R.M., Tramper, J., 1992. Strategies to improve the
production of secondary metabolites with plant cell cultures: a literature review. J. Biotechnol. 23, 111 141.
Curtis, W.R., Emery, A.H., 1993. Plant cell suspension culture
rheology. Biotechnol. Bioeng. 42, 520 526.
DeJong, D.W., Jansen, E.F., Olson, A.C., 1967. Oxidoreductive hydrolytic enzyme patterns in plant suspension culture
cells. Exp. Cell Res. 47, 139 156.
DiCosmo, F., Misawa, M., 1995. Plant cell and tissue culture:
alternatives for metabolite production. Biotechnol. Adv.
13, 425 453.
Doran, P.M., 1993. Design of reactors for plant cells and
organs. Adv. Biochem. Eng. Biotechnol. 48, 117 168.
Dornenburg, H., Knorr, D., 1995. Strategies for the improvement of secondary metabolite production in plant cell
cultures. Enzym. Microb. Technol. 17, 674 684.
Drapeau, D., Blanch, H.W., Wilke, C.R., 1987. Economic
assessment of plant cell culture for the production of
ajmalicine. Biotechnol. Bioeng. 30, 946 953.
Dubuis, B., Pluss, R., Romette, J.L., Kut, O.M., Prenosil,
J.E., Bourne, J.R., 1993. Physical factors affecting the
design and scale-up of fluidized-bed bioreactors for plant
cell culture. In: Nienow, A.W. (Ed.), Bioreactor and Bioprocess Fluid Dynamics, BHR Group Conference Series,
vol. 5. MEP, London, pp. 89 100.
Dubuis, B., Kut, A.M., Prenosil, J.E., 1995. Pilot-scale cultures of Coffea arabica in a novel loop fluidised-bed reactor. Plant Cell Tissue Organ Cult. 43, 171 183.
49
50
51
52