Cytotechnology 4: 271-278, 1990. 9 1990 Kluwer Academic Publishers. Printed in the Netherlands.
Multistage production of Autographa californica nuclear polyhedrosis virus in insect cell cultures Martina K16ppinger, Georg Fertig, Elisabeth Fraune* and Herbert G. Miltenburger Technical University, Institute of Zoology, Cell Biology Laboratory, Schnittspahnstr. 3, D-6100 Darmstadt, FRG; *B. Braun Melsungen AG, P.O. Box 120, D-3508 Melsungen, FRG Received 30 January 1990; accepted in revised form 27 July 1990
Key words: baculovirus, in vitro production, perfusion cultivation Abstract
The aim of our study was to establish an efficient system for the in vitro production of the insect pathogenic Autographa californica nuclear polyhedrosis virus in a Spodoptera frugiperda cell line. We optimized cultivation conditions for cell proliferation as well as for virus replication in a 1.5 litre stirred tank bioreactor. Cell and virus propagation were found to be optimal at a constant oxygen tension of 40%. In order to provide sufficient nutrients during virus synthesis filtration and perfusion devices were connected to the bioreactor. A virus production procedure in a repeated batch mode by using a two stage bioreactor system is described. Stage I was optimized for cell production and stage II for virus production.
Abbreviations: Ac-NPV - Autographa californica Nuclear Polyhedrosis Virus; BV - Baculovirus; MOI - Multiplicity Of Infection; ECV - Extracellular Virus
Introduction
Insect pathogenic baculoviruses (BV) are used for biological insect pest control (Carter, 1984; Huber, 1986) and recently as potent expression vectors for the production of foreign proteins in insect cells (Luckow et al., 1988; Miller, 1988). Thus effective systems for the large scale BV production are required. We investigated the Autographa californica nuclear polyhedrosis virus, which has a biphasic life cycle: after replication in the nuclei of infected cells the viruses are either released by budding to form extracellular non-occluded virus (ECV) or they are embedded in proteinaceous
high molecular weight occlusion bodies called polyhedra, which are composed of multiple 29 kD monomeric polyhedrin molecules. The extracellular virus is responsible for secondary infection of uninfected cells, whereas the polyhedra remain in the cellular nucleus until the cells are lysing. Polyhedra are infective for larvae only but not for insect cells in vitro, so they can be used effectively in insect pest control. When the BV system is applied for the production of recombinant proteins, the genes of these proteins are commonly under control of the polyhedrin promotor sequences, and they are expressed instead of the polyhedrin gene. In order to obtain large quantities of BV and
272 recombinant proteins effective production systems are required. Production of BV in insect larvae in vivo is laborious and yields an impure product, contaminated with microorganisms and insect proteins. In contrast, in vitro production of BV in insect cell cultures allows (i) to standardize production conditions and (ii) to obtain a product o f higher purity, facilitating subsequent purification. However, difficulties in mass production of BV in vitro are due to fastidious nutritional and physical requirements for insect cell propagation (Miltenburger et al., 1980; Tramper et al., 1986) as well as for virus synthesis (Weiss e t a L, 1985). Therefore, we developed a relatively simple and effective in vitro production procedure under optimized conditions. As parameters for quality control of cultivation conditions, we determined the rates of polyhedra and extracellular virus production.
Materials and methods
(MiniKros; Microgon) to exchange spent medium for fresh medium prior to infection of cells (Fig. la). While the cells were recirculated via peristaltic pumping through this system (100 ml/ min), spent medium was removed. Thereafter, cells were resuspended in fresh medium and then transferred to the fermenter in stage II where virus production took place. In some experiments perfusion of medium was carried out during the virus production process. The cell retaining filtration device was an internal microporous tubing system (B. Braun Melsungen AG) with 1 ~tm pore size (Fraune et al., 1988). The medium for perfusion was supplemented with 5% FCS. A dosing pump in combination with a level control system was used to determine the perfusion rate. At the time of infection with Ac-NPV, the density of the Sf9 cells was about 1.5 x 106 cells/ml with a viability of more than 90% in all experiments. We infected the cells with a viral inoculum between 1 and 10 MOI, which resulted in a synchronic infection.
Cell culture and virus propagation Sample analysis The Spodoptera frugiperda cell line Sf9 (ATCC Cat. No. 1711) was used in the described experiments. Cells were cultured in TC100 medium (Gibco) supplemented with 10% fetal calf serum (FCS; Biochrom) and 1 mg/ml Neomycinsulfat (Biochrom): The cells were grown in suspension in spinner flasks (Bellco) and in a stirred tank reactor (Biostat MC; B. Braun Melsungen AG) with 1.5 1 working volume. The stirring speed was 50 rpm and the cultivation temperature 27~ Bubble-free aeration was achieved by a silicone tubing system (Miltenburger et al., 1980). The oxygen tension (pO2) was controlled by a gas mixing system in combination with a pO 2 controller. The pH value was kept constant at 6.3 and different p O 2 values were tested during the experiments. A two stage fermenter system was established, stage I for cell production and stage II for virus production (Fig. 1 a+b). The cell production fermenter was equipped with an external hollow fiber filtration device with 0.2 p,m pore size
Cell density was measured using a hemocytometer and viability was determined by trypan blue dye exclusion. Population doubling times and growth rate were calculated from the exponential phase of the growth curve. The glucose and lactate concentration was measured using an automatic analyzer (YSI; model 2000). For quantification of virus production, the percentage of cells containing polyhedra as well as the number of polyhedra per cell were counted by using a light microscope. If a cell contained more than 10 polyhedra, the number was set to 15, since more than 10 polyhedra per cell could not be counted reliably. The titer of extracellular virus was determined by the 50% tissue culture infective dose titration assay (TCIDs0 assay), which has been described in detail by Summers (1987).
273
a
9 I Hollow fiber filtration module medium reservoir
spent medium
Biostat MC
b level control system i i
9
medium reservoir
t
i i i
i~ _ microporous C
I~ f i l t r a t i ~
module
spent
medium
Biostat MC
Fig. 1. Design of bioreactor for cell and virus production. (a) Bioreactor stage I for cell production. The reactor was equipped with an external hollow fiber filtration module for removal of depleted medium. (b) Bioreactor stage II for virus production. The reactor was equipped with a microporous filtration module for perfusion cultivation.
274 Results
Bioreactor stage H: virus production
Bioreactor stage 1: cell cultivation
We studied the influence of different oxygen tensions on virus production in the bioreactor. As shown in Table 2, we found that an oxygen tension of 20% reduced the yield of polyhedra by more than 50% as compared to an oxygen tension of >40%. Thus, virus production was performed at an oxygen tension of 40% in all experiments. In the next studies we investigated the influence of medium exchange on virus production. Since medium change by centrifugation is laborious in the mass culture of insect cells, medium was changed by using a hollow fiber filtration device connected to the bioreactor (Fig. la). When the cells had reached a density of approximately 1.5 • 106/ml they were infected either directly, or after 90% of medium had been replaced. Afterwards, polyhedra and extracellular virus production as well as glucose consumption was monitored. As summarized in Fig. 3 replacement of medium before infection resulted in higher glucose levels (Fig. 3a) and in an increased amount of extracellular virus (Fig. 3b).
Sf9 cells were propagated in the Biostat MC reactor in batch cultures under different oxygen tensions (20%, 40%, 60% and 80% pO2). The cultures were started with seeding densities of 1-2 x 105 cells/ml. As shown in Table 1, the highest cell density and shortest population doubling times were achieved at an oxygen tension of 40%. This condition was chosen for subsequent experiments. Cell growth in the bioreactor was compared to cell growth in spinner flasks (0.5 and 1.0 1). As demonstrated in Fig. 2a, suboptimal growth was observed in spinner flasks. The concentration of lactate increased during cultivation in spinner flasks, whereas lactate remained constant in the bioreactor (Fig. 2b). From these data, we concluded that the lower cell proliferation observed in spinner flasks might be due tO oxygen limitation resulting in anaerobic glucose metabolism and concomitant lactate production.
Table 1. Cultivation of Sf9 cells in the 1.5 1 bioreactor Biostat MC in TC100 medium with 10% FCS under various oxygen tensions (pH 6.3, 50 rpm, 27~
pO 2 (%)
Doubling time (h)
Cell density (cells/ml)
Growth rate (d-1)
20 40 60 80
22.5 21.5 25:0 25.4
1.95 2.30 1.84 1.55
0.031 0.032 0.028 0.027
x • x x
106 106 106 106
Table 2. Polyhedra production in Sf9 cells 4 days post infection with Ac-NPV in the 1.5 1 bioreactor Biostat MC under different pO 2 conditions. The cells were cultivated in TC100 medium with 10% FCS and infected at a density 1.5 • 106 cells/ml (pH 6.3, 50 rpm, 27~
pO 2 (%)
Infected cells (%)
Polyhedra per cell
Polyhedra per 1 1 (• 109)
20 40 60 80
42 92 79 86
6.3 10.0 10.4 9.4
4.0 12.5 12.3 12.7
275
cells/ml
b
g/I lactate
10
1.0
0.8
0.6 10 0.4
O.
0
0 r-i
24
48
72 h~
.o
,
9
0.51 spinner flask 1.01 spinner flask 1,51 bioreactor
0 ~ []
:,
'o! 2
6 ur
0,51 spinner flask 1.01 spinner flask 1.51 bioreactor
Fig. 2. Cultivation of Sf9 cells in 0.5 and 1.0 1 spinner flasks and in the 1.5 1 bioreactor Biostat MC in TC100 medium with 10% FCS. The parameters in the bioreactor were pH 6.3, 40% PO2, 50 rpm, 27~
(a) growth curve; (b) lactate concentration.
Table 3. Polyhedra production in Sf9 cells 4 days post infection with Ac-NPV in the 1.5 1 bioreactor Biostat MC under different modes of medium supply. Cells were cultivated in TC100 medium with 10% FCS and infected after they had reached a density of 1.5 • 106 ceUs/ml. TC100 medium with 5% FCS was used for perfusion (pH 6.3, 50 rpm, 27~ pO 2 (%)
Medium change prior to infection
Perfusion
Infected cells (%)
Polyhedra per cell
Polyhedra (x 109)/1
Total amount of ECV (TCID 50 x 1012)
40 40 40
+ +
+
64 88 94
9.4 9.5 11.6
9.0 12.5 16.2
2.8 5.6 15.8"
* Note that in perfusion culture three times more medium containing extracellular virus (ECV) was obtained than in non-perfused cultures.
276 0 v~ [].
0
without medium change with medium change with perfusion
mg glucose/ml
a
TC
without medium change with medium change
IDI,1081mi
b
40
1.0
3O
~I
0,
20
O, 1
0.
0
24
48
~ - - ~ _ 0
72 96 hours post infection C
% infected cells
,i,,
24
I 48
,
I , I > 72 96 hours post infection
d
polyhedra/cell
1
100
-
90 ~80 ~-
1
60
40 - 30220 ] 10-0
I 0
0 []
I.. 24
withoul medium change with medium change with perfusion
I 48
I , 1. [:::::> 72 96 hours post infection
I
0
1
f
24
1
48
J
I
a
I
72 96 hours post infection
0
without medium change with medium change I~ ..... with perfusion
Fig. 3. Comparison of virus production in the bioreactor Biostat MC under different cultivation conditions (O cultivation without medium replacement, 9 cultivation with medium replacement by hollow fiber filtration, [] cultivation with medium perfusion with TC100 medium supplemented with 5% FCS). Cells were cultivated at 40% pO2, pH 6.3, 50 rpm and 27~ in TC100 medium with 10% FCS. (a) Glucose concentration; (b) Extracellular virus concentration (not shown for peffusion cultivation); (c) Percentage of infected cells; (d) Number of polyhedra per cell.
277 Also, a higher percentage of infected cells (Fig. 3c) and more polyhedra per cell were obtained (Fig. 3d). In summary, medium exchange resuited in an increase in the overall polyhedra production of 28%: 9 x 109 polyhedra/1 as compared to 12.5 x 109 polyhedra/1 (see Table 3). The rapid decline in glucose concentration indicated high metabolic activity and nutrient consumption of cells during virus synthesis (Fig. 3a). In order to supply nutrients continuously, perfusion of medium was started 3 hours after infection with a medium dilution rate of 0.3 (d-l), which was increased to 0.6 (d-1) 21 hours post infection. In this way glucose concentration could be maintained between 0.6 and 0.7 mg/ml (Fig. 3a). Extracellular virus did not accumulate in the reactor vessel, because the virus passed the filtration membrane and was diluted in the filtrate. Thus perfused and non-perfused cultures were not comparable with respect to accumulated virus and therefore we did not include the concentration of extracellular virus in Fig. 3b. Using the perfusion system the number of polyhedra per nucleus was increased (Fig. 3d). This resulted in a higher overall production of polyhedra by 23% as compared to non-perfused cultures (see Table 3). Also the total yield of extracellular virus, shown in Table 3, was obviously increased in the perfused culture. After having established optimized parameters for cell and virus production we developed a multistage BV production system which is schematically presented in Fig. 1. Cells are grown to a density of about 1.5 • 106 cells/ml (usually within 4 days) in fermenter stage I. Then fresh medium is provided by the use of a hollow fiber filtration device. Afterwards 90% of the cell suspension is transferred to fermenter stage II and cells are infected with Ac-N-PV; The remaining 10% of the cell suspension in stage I is used as inoculum for further cell production. Virus production starts in fermenter II. Four days after infection cells and polyhedra are harvested for further purification. 5% of the culture volume is left in the bioreactor as inoculum for infection of
new cells. The described system enables us to grow cells permanently in stage I while simultaneously producing virus in stage II. The process was carried out continuously in our laboratory for several weeks with serial production cycles and without decrease in virus yield.
Discussion We present data on the optimization of BV production in stirred tank bioreactors. Since oxygen supply is a major requirement in the scaling up of in vitro cell culture systems, we kept a constant pO2 by silicone tubing oxygenation. For cell proliferation an oxygen tension of 40% was found to be optimal. With respect to polyhedra production we found less than half of the production rate at an oxygen tension of 20% compared to 40%. These findings agree with those of other authors, who studied the oxygen requirement for growth of insect cells (Hink et al., 1976; Maiorella et al., 1988) and virus synthesis (Street et al., 1978; Weiss et al., 1985). Our results indicate that sufficient and adequate nutrient supply is a prerequisite for the virally induced accelerated metabolic activity of cells. Using the hollow fiber filtration module, depleted medium could be removed within 20 min and fresh medium could be provided prior to infection offering the possibility of infecting cells simultaneously at a high cell density. By perfusion cultivation under constant glucose levels a further enhancement of polyhedra production by 23% was achieved, although three times more medium had to be used. More details about the critical nutritional requirements for virus replication are needed, which could finally be helpful in designing a special low cost perfusion medium. The parallel use of two bioreactors is a relatively simple procedure for semi-continuous BV production. This is an alternative to the semicontinuous method described by Hink (1982) and the fully continuous procedure described by Kompier and coworkers (1988).
278
Acknowledgements We wish to thank P. Czech for excellent technical assistance. This work has been supported by the federal ministry for research and technology, Bonn.
References 1. Carter JH (1984) Viruses as pest control agents. Biotechnology and Genetic Engineering Reviews 1: 375-419. 2. Fraune E, Fenge C, Kuhlmann W (1988) Development of large scale perfusion cell culture reactors. Paper presented at the BIO Symposium, Tokyo. 3. Hink WF, Strauss EM (1976) Growth of the Trichuplusia ni (TN-368) cell line in suspension culture. In: Kurstak E, Maramorosch K (eds) Invertebrate tissue culture. Applications in medicine, biology and agriculture, pp. 297-300. Academic Press, New York. 4. Hink WF (1982) Production of Autographa californica nuclear polyhedrosis virus in cells from large scale suspension cultures. In: Kurstak E (ed) Microbial and Viral Pesticides pp. 493-506. Marcel Dekker, New York. 5. Huber J (1986) Use of Baculoviruses in Pest Management Programs. In: Granados RR and Federici BA (ed) The biology of Baculoviruses Vol II, pp. 181-202 CRC Press Inc., Boca Raton, Florida. 6. Kompier R, Tramper J, Vlak JM (1988) A continuous process for the production of baculovirus using insect cell cultures. Biotechnology letters 10: 849-854.
7. Luckow VA, Summers MD (1988) Trends in the development of baculovirus expression vectoi's. Bin/Technology 6: 47-55. 8. Maiorella B, Inlow D, Shaugher A, Harano D (1988) Large scale insect cell culture for recombinant protein production. Bio/Technology 6: 1406-1410. 9. Miller LK (1988) Baculoviruses as gene expression vectors. Ann. Rev. Microbiol. 42: 177-199. 10. Miltenburger HG, David P (1980) Mass production of insect ceils in suspension. Develop. Biol. Standard. 46: 183-186. 11. Streett DA, Hink WF (1978) Oxygen consumption of Trichoplusia ni (TN-368) insect cell line infected with Autographa califomica nuclear polyhedrosis virus. J: Invertebr. Pathol. 32: 112-113. 12. Summers MD, Smith GE (1987) A manual of methods for Baculovims vectors and insect cell culture procedures. Texas Agricultural Experiment Station Bulletin no. 1555. 13. Tramper J, Williams JB, Joustra D, Vlak JM (1986) Shear sensitivity of insect cells in suspension. Enzyme Microb. Technol. 8: 33-36. 14. Weiss SA, Pepiow D, Smith GC, Vaughn JL, Dougherty E (1985) Biotechnical aspects of a large scale process for insect cells and baculoviruses. In: Kurstak C (ed) Techniques in the life sciences, Vol C1, pp. C110/1-16. Elsevier Scientific Publishers Ireland Ltd.
Address for offprints: Martina K16ppinger, Technical University, Institute of Zoology, Cell Biology Laboratory, Schnittspahnstr. 3, D-6100 Darmstadt, FRG
Multistage production of Autographa californica nuclear polyhedrosis virus in insect cell cultures Martina K16ppinger, Georg Fertig, Elisabeth Fraune* and Herbert G. Miltenburger Technical University, Institute of Zoology, Cell Biology Laboratory, Schnittspahnstr. 3, D-6100 Darmstadt, FRG; *B. Braun Melsungen AG, P.O. Box 120, D-3508 Melsungen, FRG Received 30 January 1990; accepted in revised form 27 July 1990
Key words: baculovirus, in vitro production, perfusion cultivation Abstract
The aim of our study was to establish an efficient system for the in vitro production of the insect pathogenic Autographa californica nuclear polyhedrosis virus in a Spodoptera frugiperda cell line. We optimized cultivation conditions for cell proliferation as well as for virus replication in a 1.5 litre stirred tank bioreactor. Cell and virus propagation were found to be optimal at a constant oxygen tension of 40%. In order to provide sufficient nutrients during virus synthesis filtration and perfusion devices were connected to the bioreactor. A virus production procedure in a repeated batch mode by using a two stage bioreactor system is described. Stage I was optimized for cell production and stage II for virus production.
Abbreviations: Ac-NPV - Autographa californica Nuclear Polyhedrosis Virus; BV - Baculovirus; MOI - Multiplicity Of Infection; ECV - Extracellular Virus
Introduction
Insect pathogenic baculoviruses (BV) are used for biological insect pest control (Carter, 1984; Huber, 1986) and recently as potent expression vectors for the production of foreign proteins in insect cells (Luckow et al., 1988; Miller, 1988). Thus effective systems for the large scale BV production are required. We investigated the Autographa californica nuclear polyhedrosis virus, which has a biphasic life cycle: after replication in the nuclei of infected cells the viruses are either released by budding to form extracellular non-occluded virus (ECV) or they are embedded in proteinaceous
high molecular weight occlusion bodies called polyhedra, which are composed of multiple 29 kD monomeric polyhedrin molecules. The extracellular virus is responsible for secondary infection of uninfected cells, whereas the polyhedra remain in the cellular nucleus until the cells are lysing. Polyhedra are infective for larvae only but not for insect cells in vitro, so they can be used effectively in insect pest control. When the BV system is applied for the production of recombinant proteins, the genes of these proteins are commonly under control of the polyhedrin promotor sequences, and they are expressed instead of the polyhedrin gene. In order to obtain large quantities of BV and
272 recombinant proteins effective production systems are required. Production of BV in insect larvae in vivo is laborious and yields an impure product, contaminated with microorganisms and insect proteins. In contrast, in vitro production of BV in insect cell cultures allows (i) to standardize production conditions and (ii) to obtain a product o f higher purity, facilitating subsequent purification. However, difficulties in mass production of BV in vitro are due to fastidious nutritional and physical requirements for insect cell propagation (Miltenburger et al., 1980; Tramper et al., 1986) as well as for virus synthesis (Weiss e t a L, 1985). Therefore, we developed a relatively simple and effective in vitro production procedure under optimized conditions. As parameters for quality control of cultivation conditions, we determined the rates of polyhedra and extracellular virus production.
Materials and methods
(MiniKros; Microgon) to exchange spent medium for fresh medium prior to infection of cells (Fig. la). While the cells were recirculated via peristaltic pumping through this system (100 ml/ min), spent medium was removed. Thereafter, cells were resuspended in fresh medium and then transferred to the fermenter in stage II where virus production took place. In some experiments perfusion of medium was carried out during the virus production process. The cell retaining filtration device was an internal microporous tubing system (B. Braun Melsungen AG) with 1 ~tm pore size (Fraune et al., 1988). The medium for perfusion was supplemented with 5% FCS. A dosing pump in combination with a level control system was used to determine the perfusion rate. At the time of infection with Ac-NPV, the density of the Sf9 cells was about 1.5 x 106 cells/ml with a viability of more than 90% in all experiments. We infected the cells with a viral inoculum between 1 and 10 MOI, which resulted in a synchronic infection.
Cell culture and virus propagation Sample analysis The Spodoptera frugiperda cell line Sf9 (ATCC Cat. No. 1711) was used in the described experiments. Cells were cultured in TC100 medium (Gibco) supplemented with 10% fetal calf serum (FCS; Biochrom) and 1 mg/ml Neomycinsulfat (Biochrom): The cells were grown in suspension in spinner flasks (Bellco) and in a stirred tank reactor (Biostat MC; B. Braun Melsungen AG) with 1.5 1 working volume. The stirring speed was 50 rpm and the cultivation temperature 27~ Bubble-free aeration was achieved by a silicone tubing system (Miltenburger et al., 1980). The oxygen tension (pO2) was controlled by a gas mixing system in combination with a pO 2 controller. The pH value was kept constant at 6.3 and different p O 2 values were tested during the experiments. A two stage fermenter system was established, stage I for cell production and stage II for virus production (Fig. 1 a+b). The cell production fermenter was equipped with an external hollow fiber filtration device with 0.2 p,m pore size
Cell density was measured using a hemocytometer and viability was determined by trypan blue dye exclusion. Population doubling times and growth rate were calculated from the exponential phase of the growth curve. The glucose and lactate concentration was measured using an automatic analyzer (YSI; model 2000). For quantification of virus production, the percentage of cells containing polyhedra as well as the number of polyhedra per cell were counted by using a light microscope. If a cell contained more than 10 polyhedra, the number was set to 15, since more than 10 polyhedra per cell could not be counted reliably. The titer of extracellular virus was determined by the 50% tissue culture infective dose titration assay (TCIDs0 assay), which has been described in detail by Summers (1987).
273
a
9 I Hollow fiber filtration module medium reservoir
spent medium
Biostat MC
b level control system i i
9
medium reservoir
t
i i i
i~ _ microporous C
I~ f i l t r a t i ~
module
spent
medium
Biostat MC
Fig. 1. Design of bioreactor for cell and virus production. (a) Bioreactor stage I for cell production. The reactor was equipped with an external hollow fiber filtration module for removal of depleted medium. (b) Bioreactor stage II for virus production. The reactor was equipped with a microporous filtration module for perfusion cultivation.
274 Results
Bioreactor stage H: virus production
Bioreactor stage 1: cell cultivation
We studied the influence of different oxygen tensions on virus production in the bioreactor. As shown in Table 2, we found that an oxygen tension of 20% reduced the yield of polyhedra by more than 50% as compared to an oxygen tension of >40%. Thus, virus production was performed at an oxygen tension of 40% in all experiments. In the next studies we investigated the influence of medium exchange on virus production. Since medium change by centrifugation is laborious in the mass culture of insect cells, medium was changed by using a hollow fiber filtration device connected to the bioreactor (Fig. la). When the cells had reached a density of approximately 1.5 • 106/ml they were infected either directly, or after 90% of medium had been replaced. Afterwards, polyhedra and extracellular virus production as well as glucose consumption was monitored. As summarized in Fig. 3 replacement of medium before infection resulted in higher glucose levels (Fig. 3a) and in an increased amount of extracellular virus (Fig. 3b).
Sf9 cells were propagated in the Biostat MC reactor in batch cultures under different oxygen tensions (20%, 40%, 60% and 80% pO2). The cultures were started with seeding densities of 1-2 x 105 cells/ml. As shown in Table 1, the highest cell density and shortest population doubling times were achieved at an oxygen tension of 40%. This condition was chosen for subsequent experiments. Cell growth in the bioreactor was compared to cell growth in spinner flasks (0.5 and 1.0 1). As demonstrated in Fig. 2a, suboptimal growth was observed in spinner flasks. The concentration of lactate increased during cultivation in spinner flasks, whereas lactate remained constant in the bioreactor (Fig. 2b). From these data, we concluded that the lower cell proliferation observed in spinner flasks might be due tO oxygen limitation resulting in anaerobic glucose metabolism and concomitant lactate production.
Table 1. Cultivation of Sf9 cells in the 1.5 1 bioreactor Biostat MC in TC100 medium with 10% FCS under various oxygen tensions (pH 6.3, 50 rpm, 27~
pO 2 (%)
Doubling time (h)
Cell density (cells/ml)
Growth rate (d-1)
20 40 60 80
22.5 21.5 25:0 25.4
1.95 2.30 1.84 1.55
0.031 0.032 0.028 0.027
x • x x
106 106 106 106
Table 2. Polyhedra production in Sf9 cells 4 days post infection with Ac-NPV in the 1.5 1 bioreactor Biostat MC under different pO 2 conditions. The cells were cultivated in TC100 medium with 10% FCS and infected at a density 1.5 • 106 cells/ml (pH 6.3, 50 rpm, 27~
pO 2 (%)
Infected cells (%)
Polyhedra per cell
Polyhedra per 1 1 (• 109)
20 40 60 80
42 92 79 86
6.3 10.0 10.4 9.4
4.0 12.5 12.3 12.7
275
cells/ml
b
g/I lactate
10
1.0
0.8
0.6 10 0.4
O.
0
0 r-i
24
48
72 h~
.o
,
9
0.51 spinner flask 1.01 spinner flask 1,51 bioreactor
0 ~ []
:,
'o! 2
6 ur
0,51 spinner flask 1.01 spinner flask 1.51 bioreactor
Fig. 2. Cultivation of Sf9 cells in 0.5 and 1.0 1 spinner flasks and in the 1.5 1 bioreactor Biostat MC in TC100 medium with 10% FCS. The parameters in the bioreactor were pH 6.3, 40% PO2, 50 rpm, 27~
(a) growth curve; (b) lactate concentration.
Table 3. Polyhedra production in Sf9 cells 4 days post infection with Ac-NPV in the 1.5 1 bioreactor Biostat MC under different modes of medium supply. Cells were cultivated in TC100 medium with 10% FCS and infected after they had reached a density of 1.5 • 106 ceUs/ml. TC100 medium with 5% FCS was used for perfusion (pH 6.3, 50 rpm, 27~ pO 2 (%)
Medium change prior to infection
Perfusion
Infected cells (%)
Polyhedra per cell
Polyhedra (x 109)/1
Total amount of ECV (TCID 50 x 1012)
40 40 40
+ +
+
64 88 94
9.4 9.5 11.6
9.0 12.5 16.2
2.8 5.6 15.8"
* Note that in perfusion culture three times more medium containing extracellular virus (ECV) was obtained than in non-perfused cultures.
276 0 v~ [].
0
without medium change with medium change with perfusion
mg glucose/ml
a
TC
without medium change with medium change
IDI,1081mi
b
40
1.0
3O
~I
0,
20
O, 1
0.
0
24
48
~ - - ~ _ 0
72 96 hours post infection C
% infected cells
,i,,
24
I 48
,
I , I > 72 96 hours post infection
d
polyhedra/cell
1
100
-
90 ~80 ~-
1
60
40 - 30220 ] 10-0
I 0
0 []
I.. 24
withoul medium change with medium change with perfusion
I 48
I , 1. [:::::> 72 96 hours post infection
I
0
1
f
24
1
48
J
I
a
I
72 96 hours post infection
0
without medium change with medium change I~ ..... with perfusion
Fig. 3. Comparison of virus production in the bioreactor Biostat MC under different cultivation conditions (O cultivation without medium replacement, 9 cultivation with medium replacement by hollow fiber filtration, [] cultivation with medium perfusion with TC100 medium supplemented with 5% FCS). Cells were cultivated at 40% pO2, pH 6.3, 50 rpm and 27~ in TC100 medium with 10% FCS. (a) Glucose concentration; (b) Extracellular virus concentration (not shown for peffusion cultivation); (c) Percentage of infected cells; (d) Number of polyhedra per cell.
277 Also, a higher percentage of infected cells (Fig. 3c) and more polyhedra per cell were obtained (Fig. 3d). In summary, medium exchange resuited in an increase in the overall polyhedra production of 28%: 9 x 109 polyhedra/1 as compared to 12.5 x 109 polyhedra/1 (see Table 3). The rapid decline in glucose concentration indicated high metabolic activity and nutrient consumption of cells during virus synthesis (Fig. 3a). In order to supply nutrients continuously, perfusion of medium was started 3 hours after infection with a medium dilution rate of 0.3 (d-l), which was increased to 0.6 (d-1) 21 hours post infection. In this way glucose concentration could be maintained between 0.6 and 0.7 mg/ml (Fig. 3a). Extracellular virus did not accumulate in the reactor vessel, because the virus passed the filtration membrane and was diluted in the filtrate. Thus perfused and non-perfused cultures were not comparable with respect to accumulated virus and therefore we did not include the concentration of extracellular virus in Fig. 3b. Using the perfusion system the number of polyhedra per nucleus was increased (Fig. 3d). This resulted in a higher overall production of polyhedra by 23% as compared to non-perfused cultures (see Table 3). Also the total yield of extracellular virus, shown in Table 3, was obviously increased in the perfused culture. After having established optimized parameters for cell and virus production we developed a multistage BV production system which is schematically presented in Fig. 1. Cells are grown to a density of about 1.5 • 106 cells/ml (usually within 4 days) in fermenter stage I. Then fresh medium is provided by the use of a hollow fiber filtration device. Afterwards 90% of the cell suspension is transferred to fermenter stage II and cells are infected with Ac-N-PV; The remaining 10% of the cell suspension in stage I is used as inoculum for further cell production. Virus production starts in fermenter II. Four days after infection cells and polyhedra are harvested for further purification. 5% of the culture volume is left in the bioreactor as inoculum for infection of
new cells. The described system enables us to grow cells permanently in stage I while simultaneously producing virus in stage II. The process was carried out continuously in our laboratory for several weeks with serial production cycles and without decrease in virus yield.
Discussion We present data on the optimization of BV production in stirred tank bioreactors. Since oxygen supply is a major requirement in the scaling up of in vitro cell culture systems, we kept a constant pO2 by silicone tubing oxygenation. For cell proliferation an oxygen tension of 40% was found to be optimal. With respect to polyhedra production we found less than half of the production rate at an oxygen tension of 20% compared to 40%. These findings agree with those of other authors, who studied the oxygen requirement for growth of insect cells (Hink et al., 1976; Maiorella et al., 1988) and virus synthesis (Street et al., 1978; Weiss et al., 1985). Our results indicate that sufficient and adequate nutrient supply is a prerequisite for the virally induced accelerated metabolic activity of cells. Using the hollow fiber filtration module, depleted medium could be removed within 20 min and fresh medium could be provided prior to infection offering the possibility of infecting cells simultaneously at a high cell density. By perfusion cultivation under constant glucose levels a further enhancement of polyhedra production by 23% was achieved, although three times more medium had to be used. More details about the critical nutritional requirements for virus replication are needed, which could finally be helpful in designing a special low cost perfusion medium. The parallel use of two bioreactors is a relatively simple procedure for semi-continuous BV production. This is an alternative to the semicontinuous method described by Hink (1982) and the fully continuous procedure described by Kompier and coworkers (1988).
278
Acknowledgements We wish to thank P. Czech for excellent technical assistance. This work has been supported by the federal ministry for research and technology, Bonn.
References 1. Carter JH (1984) Viruses as pest control agents. Biotechnology and Genetic Engineering Reviews 1: 375-419. 2. Fraune E, Fenge C, Kuhlmann W (1988) Development of large scale perfusion cell culture reactors. Paper presented at the BIO Symposium, Tokyo. 3. Hink WF, Strauss EM (1976) Growth of the Trichuplusia ni (TN-368) cell line in suspension culture. In: Kurstak E, Maramorosch K (eds) Invertebrate tissue culture. Applications in medicine, biology and agriculture, pp. 297-300. Academic Press, New York. 4. Hink WF (1982) Production of Autographa californica nuclear polyhedrosis virus in cells from large scale suspension cultures. In: Kurstak E (ed) Microbial and Viral Pesticides pp. 493-506. Marcel Dekker, New York. 5. Huber J (1986) Use of Baculoviruses in Pest Management Programs. In: Granados RR and Federici BA (ed) The biology of Baculoviruses Vol II, pp. 181-202 CRC Press Inc., Boca Raton, Florida. 6. Kompier R, Tramper J, Vlak JM (1988) A continuous process for the production of baculovirus using insect cell cultures. Biotechnology letters 10: 849-854.
7. Luckow VA, Summers MD (1988) Trends in the development of baculovirus expression vectoi's. Bin/Technology 6: 47-55. 8. Maiorella B, Inlow D, Shaugher A, Harano D (1988) Large scale insect cell culture for recombinant protein production. Bio/Technology 6: 1406-1410. 9. Miller LK (1988) Baculoviruses as gene expression vectors. Ann. Rev. Microbiol. 42: 177-199. 10. Miltenburger HG, David P (1980) Mass production of insect ceils in suspension. Develop. Biol. Standard. 46: 183-186. 11. Streett DA, Hink WF (1978) Oxygen consumption of Trichoplusia ni (TN-368) insect cell line infected with Autographa califomica nuclear polyhedrosis virus. J: Invertebr. Pathol. 32: 112-113. 12. Summers MD, Smith GE (1987) A manual of methods for Baculovims vectors and insect cell culture procedures. Texas Agricultural Experiment Station Bulletin no. 1555. 13. Tramper J, Williams JB, Joustra D, Vlak JM (1986) Shear sensitivity of insect cells in suspension. Enzyme Microb. Technol. 8: 33-36. 14. Weiss SA, Pepiow D, Smith GC, Vaughn JL, Dougherty E (1985) Biotechnical aspects of a large scale process for insect cells and baculoviruses. In: Kurstak C (ed) Techniques in the life sciences, Vol C1, pp. C110/1-16. Elsevier Scientific Publishers Ireland Ltd.
Address for offprints: Martina K16ppinger, Technical University, Institute of Zoology, Cell Biology Laboratory, Schnittspahnstr. 3, D-6100 Darmstadt, FRG
