Blotechnol. hog. 1993, 9,355-361
955
Effects of Oxygen/Glucose/Glutamine Feeding on Insect Cell Baculovirus Protein Expression: A Study on Epoxide Hydrolase Production Min-Ying Wang, Simon Kwong,t and William E. Bentley' Center for Agricultural Biotechnology, Maryland Biotechnology Institute, and Department of Chemical Engineering, University of Maryland at College Park, College Park, Maryland 20742
The recombinant protein yields for batch cultures of the insect cell baculovirus expression system have been significantly enhanced by oxygen, glucose, and glutamine feeding. The improvement in both volumetric and specific yields was based on influencing the metabolism of infected cells. Oxygen was absolutely required for viral replication and high protein expression in infected cells. Increases of 200% in volumetric yield and 100% in specific yield of recombinant epoxide hydrolase were achieved by controlling the dissolved oxygen (DO) level to near 35% saturation. An additional 100% increase was achieved by glucose and glutamine feeding. Results indicated that the intracellular metabolite pool was not adequate for recombinant protein overproduction. Finally, the specific protein yield, based on initial infection cell density, in high cell density spinner flasks and bioreactors of spent media with glucose and glutamine feeding was equivalent to that of freshly diluted cultures.
Introduction More than 300 heterologous proteins have been expressed in the insect cell baculovirus expression system (Agathos, 1991; Atkinson et al., 1990; Luckow and Summers, 1988). This system has several advantages, including the overproduction of functional heterologous proteins owing to proper post-translational modifications and a strong polyhedrin promoter (Cameron et al., 1990;Fraser, 1989; Luckow and Summers, 19881, the ease of cell propagation (Agathos,1991)due to the earlier development of 'mammalian cell cultures, and the minimization of environmental health and safety concern due to host-range limitations of the recombinant virus (Cameron et al., 1990; Fraser, 1989;Luckow and Summers, 1988). These factors make this system a good candidate for large-scaleindustrial protein production. Further, Lazarte et al. (1992) noted the advantages of using recombinant baculoviruses to produce membrane-bound proteins. Since the fate of infected cells is irreversible lysis, membrane-bound proteins are more easily separated and at higher concentrations than cytoplasmically expressed or secreted proteins. Recently, a mammalian membrane protein, epoxide hydrolase, was expressed in Spodoptera frugiperda (Sf9) cells infected with a recombinant vector (Lacourciere et al., 1993). Low yields were reported from l-L spinner flasks, largely due to nutrient limitation. Reporta demonstrating nutrient limitation and byproduct inhibition have been limited. Lindsay and Betenbaugh (1992) evaluated the effects of surface aeration, infection cell density, and medium conditions on recombinant protein production in 100-mL spinner flasks of Sf-9 cells. Scott et al. (1992)reported a significant increase in protein yield by maintaining dissolved oxygen tension at 50% saturation in a stirred and agitated bioreactor. However, at high cell densities, oxygen maintenance was not sufficient for increasing yield. Caron et al. (1991) restored
* Author to whom correspondence should be addressed.
+ Department of Chemical and Biochemical Engineering, University of Maryland-Baltimore County, Baltimore, MD.
8756-7938/93/3009-0355$04.00/0
recombinant protein production at higher cell densities by renewing the medium at the time of infection. They also reported that by diluting a culture with 15% fresh medium and feeding 3.5 mM glucose and 4.1 mM glutamine, they doubled the specific protein yield compared to the unfed/nondiluted control; however, the maximum obtained was one-fifteenth the maximum in an oxygenated reactor with complete media renewal. Feeding the rate-limiting nutrients is less expensive and aides in elucidating the insect cell physiology under viral infection. Studies of infected cell metabolism have reported that glucose is the primary carbon substrate and that glutamine is also essential in SF-9 cell culture (Kamen et al., 1991; Wang et al., 1993). In this article, we demonstrate the effects of oxygen, glucose, and glutamine supply on recombinant protein production and cell growth, given both synchronous and asynchronous infections. Effects of feeding these nutrients during the infection stage to the otherwise spent media under oxygen-sufficient conditions and at high cell density, in both the spinner flask and fermentor, are evaluated. Specific and volumetric productivities in fermentors compete well with small-volume, oxygenated spinner flasks with renewed media.
Materials and Methods Cells and Media. Spodoptera frugiperda (Sf-9) cell stocks (ATCC No. CRL 1711) were maintained a t 28 "C in T-flasks (25 cm2, Corning) in a VWR incubator as reported by Summers and Smith (19871, with either TNMFH medium (Hink, 1970) supplemented with 10% fetal bovine serum (FBS)or Ex-cell 401 serum-free medium (Godwin et al., 1989). Cells were routinely passaged every 3 days without antibiotics. Cells that were cultured with Ex-cell401 medium in spinner flasks (Bellco, No. 196700250 (170 X 85 mm, 250 mL)) were transferred from T-flasks after viable cell counts using trypan blue. Gentamicin (50 bg/mL) and amphotericin (2.5 bg/mL) were routinely added to suspension cultures as described by Summers and Smith (1987). Results reported for spinner
0 1993 American Chemical Society and American Institute of Chemical Engineers
Biotechnol. Prog.., 1993, Vol. 9, No. 4
I
16
17
t
+.. ...
.+.
....
14
6 %
'+ .
E
= 3r 12
+
8
4
=c 10
"0
-
.-0
9 8
L
t P 5
'
5
I,
.,
L
C 0 al
'*..
"t
Glucose Lactate
,r
.
.m -100
i'
t
4
2 0 I
'."I
0
I
40
I
I
80
I
I
120
I
I \ I
160 Time (hours)
I
200
I
I
240
I
I
280
I
I
I o
320
Figure 1. Batch culture of Sf-9 cells grown without DO or pH control. No viral infection.
flask cultures were from 50-mL flasks with a 30-mL working volume (Bellco, No. 1967-00050 (115 X 55 mm, 50 mL)). A 3-L Applikon fermentor with a l-L working volume was used for the glucose/glutamine feeding experiments. This fermentor was equipped with a pitchedblade, low-shear impeller and automatic pH/DO control via regulation of the headspace gas composition. The oxygen flow rate was manually adjusted to assist in maintaining DO above 10% air saturation, and the agitation rate was 51 rpm. The inoculum was from 250mL spinner flasks. A BiofloIII fermentor (NewBrunswick Scientific, Inc.) with a 0.8-L working volume stirred at 65 rpm was used for experiments on the effects of oxygen on cell growth and viral replication. Cells were first cultured in Ex-cell 401 medium in T-75 flasks, and several were combined for an inoculum. Viruses and Viral Infections. The recombinant baculovirus, vEHX-6, was generated by homologous recombination and then plaque-purified as described by Summers and Smith (1987). The recombinant virus, vEHX-6, incorporates the @-galactosidase gene under control of the p10 promoter so that X-gal can be added to monitor 8-galactosidase expression and virus infection on the basis of the developing blue color. The stock virus solution was provided by Dr. V. Vakharia (Center for Agricultural Biotechnology, MBI, University of Maryland at College Park), and virus titer was determined by the end-point dilution method. The infections were performed by adding different volumes of virus solution after a viable cell count. Thus, the time post infection commenced from the addition of the virus solution. The multiplicity of infection (MOI), which is the ratio of recombinant virus titer to viable cell number, for fermentor experiments was 5, versus 10for spinner flasks. Samples (2 mL) were taken from spinner flasks and fermentors during the infection process, divided into two tubes, and centrifuged for 20 min at 12 000 rpm. The supernatanta were separated from cell pellets and stored at -20 "C until measurement of glucose and lactate. The cell pellet was resuspended in 250pL of 10mM KHzP04 (pH 7.4) buffer solution (EHBS) and stored a t -20 "C until assayed for activity. Analytical. Total cell counts were performed with a VWR hemacytometer, and viability was determined by trypan blue dye exclusion using a 0.04 % solution (Sigma). Cell pellets resuspended in 250 p L of EHBS were sonicated on ice for 10 s with a microtip and a 30% pulsed duty
cycle. Since epoxide hydrolase is membrane-bound, the spectrophotometric activity assay was performed on these resuspended pellet samples as described by Armstrong et al. (1980). Glucose and lactate were measured using a YSI Model 27 analyzer (YellowSprings Instruments), and ammonia was determined by an enzyme-based assay kit (Sigma, No. 170-UV). Lactate dehydrogenase (LDH) activity was measured from the supernatant and was quantified on an IBI Biolyzer (Eastman Kodak) with dry chemistry films, both kindly shared by Dr. Govind Rao (University of Maryland at Baltimore). LDH activity reflects the extent of cell lysis, since LDH is not secreted by Sf-9 cells.
Results and Discussion Effect of Oxygen on Sf-9 Cell Growth and Recombinant Protein Production. An experiment was run in a BiofloIII fermentor with surface aeration and without pH and DO control, mimicking a l-L spinner flask. Figure 1 depicts the profiles of cell concentration, pH, and DO in a batch culture. The DO dropped to zero after approximately 170 h. Simultaneously, the cells entered the stationary phase with a maximum concentration of 1.45 X 106 cells/mL. The stationary phase lasted about 20 h, while the DO continued to remain at zero and the nutrient concentrations were in ample supply. Subsequently, the viable cell density decreased, which suggests that cell growth was strongly inhibited under oxygen limitation. This is in agreement with Scott et al. (1992), who demonstrated that subculturing an aerated fermentor culture in oxygen-limited spinner flasks resulted in a lower final cell yield. An indication of reduced oxygen supply and energy production efficiency is the steady accumulation of lactate, which reached 4 mM by the end of the fermentation. We have included this experiment because the length of the stationary phase will be discussed relative to the subsequent infected cultures. Two batch cultures were conducted to examine the effects of DO during the viral replication and protein production phases. The first batch culture was conducted in a fermentor without DO control (Figure 21, while the other was conducted with DO regulated near 35% (Figure 3). Virus infection was performed for the fermentor without DO control by adding 15 mL of virus solution (MOI 5) at 91 h, after the cells had attained a concentration of 6.15 X 106 cells/mL. Shortly after infection, the DO
357
Blotechnol. Rog..,1993, Vol. 9, No. 4 ~~
I
I'
-
......+.+.....+...'+"..+
14 t'
12
10
-
Viable cells + Glucose 0
Lactate DO p~
infection -100
-
- B- 50 8 -
40
120
80
160
200
240
280
200
240
280
0
Time (hours) 0.36
2 0.3
I 8
k.c
0,
'0
sm $j0.2
-I
0.14
40
.
80
.
120 160 Time (hours)
Figure 2. Batch culture of Sf-9 cells grown without DO control. (a) Glucose, DO, lactate, and viable cell concentrationsand pH are plotted versus time. Viral infection (MOI= 5) is denoted by the arrow. (b)Epoxide hydrolase activity,lactate dehydrogenase activity, and cell viability for the same experiment are shown over time.
dropped to zero and remained at zero for another 90 h. The short spike in DO was attributed to the volumetric change by the addition of virus solution as noted by Kamen et al. (1990). The viable cell concentration remained near 6.55 X 105 cells/mL and then started to decrease simultaneously with an increase in DO after 190 h. The onset of cell lysis, however, is marked by both the increment in LDH activity and the decrement in cell viability near 150 h (Figure 2b). Note that, after the DO dropped to zero, the lactate attained a maximum of 2.5 mM, reflecting the inadequacy of oxidative phosphorylation to supply the required ATP. A long stationary phase in this culture (- 100 h) may indicate that viral replication was blocked due to oxygen deficiency. The growth kinetics for the batch culture with DO control is shown in Figure 3a. Virus infection was also performed at 90 h when the cells had attained a concentration of 6.55 X 105 cells/mL. Unlike the oxygenuncontrolled culture, the viable cell concentration continued to increase to a maximum of 1.05 X lo6 cells/mL. The increased demand in oxygen immediately following infection was met in this experiment, and the DO was maintained above 10% and averaged 35% for the re-
mainder of the experiment. The continuous addition of oxygen likely provided the requirement for continued cell growth. Note that the lactate concentration remained below 1 mM for the whole experiment, while glucose concentration dropped almost 45 % ,from 14to 8mM after infection. The profiles of cell viability, lactate dehydrogenase (LDH) activity, and epoxide hydrolase activity are shown in Figure 3b. The maximum cell concentration was maintained for only 20 h and started to decrease at 135 h. Note also that cell lysis (LDH activity) increased immediately following infection and the cell viability dropped shortly thereafter. Further, since the viable cell number increased significantly during this time, there was a secondary infection of the remaining cells,demonstrating that viral replication was quite active. This experiment confirmed our previous hypothesis that infected cells live longer in an oxygen-limited (low agitation) environment, even longer than uninfected cells when in an oxygenlimited stationary phase (see Figures 1and 2). Since cell division and cellular DNA synthesis are blocked in infected cells, cellular regulation becomes controlled by the viral DNA replication, transcription, and translation activities (Fraser, 1989;Knudson and Tinsley, 1974). Consequently,
-
Biotechnol. h g . , 1993, Vol. 9, No. 4
16
aE
14
-
-..-
W . . ? . . "
w
Viable cells
+
Glucose
I 0
'*.,
" "
.
- Jb
-
a?
80
40
120
160
200
280
240
0
320
Time (hours)
0
Viability
r
.
*
i
i
0.,
..*
'0.
0.14 1
0
0
I
50
LDH Epoxide Hydrolase
I
-67J
8
+
m'
17
a
+; .I
,
.
'
40'
,
.
l
.
80
.
f
1
120
I
1
I
160
I
200
I
I
240
I
I
280
'.. 1
.'..e 1
I
2o
320
Time (hours)
Figure 3. Batch culture of Sf-9cella grown with DO control. (a)Glucose, DO, lactate,and viable cell concentrationsand pH are plotted versus time. Viral infection (MOI= 5 ) is denoted by the arrow. (b) Epoxide hydrolase activity, lactate dehydrogenase activity, and cell viability for the same experiment are shown over time. if viral-associated activity is attenuated by the lack of oxygen,the cells remain in a suspended or quiescent state. As previously discussed, the viral DNA synthesis was slowed down due an insufficient oxygen supply, and perhaps this reduction in viral DNA synthesis provided for the extended integrity of the cell membrane. While DO was controlled near 35% during the post infection period, the epoxide hydrolase activity reached 0.044 unit/mL (Figure 3b), which was 200% higher than the value in the oxygen-uncontrolled experiment (Figure 2b). The specific productivity increased by 100% as well (Table I). This increase in protein production is accomplished only by continued glucose uptake after infection. In the uncontrolled experiment, infected cells consumed little glucose (- 1mM) after the DO dropped to zero. On the other hand, both the uninfected and infected cells in the DO-controlledexperiment continued to uptake glucose and consumed almost 1g/L after infection for cell growth and viral replication. This definitively demonstrates that infected cells require oxygen and extracellular glucose to maintain metabolic activity, ensuring energy and biosynthetic precursor supply. Endogenous metabolism, in an oxygen-limited environment, cannot supply precursors and
energy (via lactate formation) in order to synthesize recombinant protein. Glucose/GlutamineFeeding. Spinner Flasks. Three experiments were performed to evaluate glucose and glutamine feeding in spinner flasks. A higher MOI of 10 was used to establish synchronous infection, so that the substrate utilization is attributed to the infected cells rather than the uninfected cells. The use of 50-mL spinner flasks with a 30-mL working volume prevents oxygen limitation. The cells were grown in a 250-mLspinner flask with an 80-mL working volume until their density reached 3.85 X 106 cells/mL, and they were then divided and infected in three smaller spinner flasks a t an MOI of 10 and a cell density of 3.3 X 106cells/mL. One of the spinner flasks was diluted 3:l with fresh medium as a control. Previously, this method was reported as resulting in the highest specific yield attainable in spinner flasks (Caron et al., 1990; Lindsay and Betenbaugh, 1992). Thus, we will refer to this spinner flask as an oxygen- and substratesufficient control. One of the spinner flasks was immediately fed 0.5 mL of 200 mM glutamine and 80 pL of 2 M glucose. By feeding both glutamine and glucose immediately upon infection,their concentrations are raised
358
Biotechnoi. hog., 1993, Vol. 9, No. 4
-5
- 4
-3
-2
-1
I
t
I
I
I
I
20
0
I
I
I
I
60
40
80
I
I
I
100
10
I
120
Time (hr.)
Figure 4. Total and viable cell densities and glucose concentration for three spinner flask cultures of Sf-9 cells infected at t = 0 and MOI = 10 glucose/glutamine fed culture (w, total cells; 0 , viable cells; m, -, glucose); nonfed high cell density control (A,total cells; A, viable cells, A,-, glucose); low cell density, freshly diluted culture (V,total cells; V, viable cells; glucose not shown). Table I. Overall Performance Comparison between Batch Cultures with and without DO Control and Glucose/ Glutamine Feeding DO experiments: fermentor 1 2 oxygen level not controlled 35% 11.08 maximum viable cells (106 cells/mL) 7.25 5.5 glucose uptake after infection (mM) 1.0 44.65 14.65 total protein activity (units) 4.0 X 109 2.0 x 109 specific yield (units/l06 cells) feeding experiments: spinner flasks maximum viable cells (106 cells/mL) glucose uptake after infection (mM) total protein activity (units) specific yield (units/l06 cells) feeding experiments: fermentor maximum viable cells (I@cells/mL) glucose uptake after infection (mM) total protein activity (units) specific yield (units/l06 cells)
glucose/ glutamine fed unfed 40 41
control 11
10.8
4.5
3.8
fed culture unfed culture diluted culture
0"
0
20
40
I
I
60
I
I
80
I
100
'
120
Xme (hr.)
6.3 4.76
X
3.5 1.65 109 2.89 X 109 5.0 X 109
glucose/ glutamine fed unfed 46 40 ~~~
3.5
14.1 77 2.32
X
control 10 4
2.7 1.0 109 1.86 X 109 3.34 X 109
to approximately that of fresh media (4.1 mM glutamine and 5.5 mM glucose). Thereafter, since glucose is easily and rapidly measured, daily glucose feeding maintained an ample glucose supply (see Figure 4). Glutamine was not measured, but on the basis of the results of Kamen et al. (19911, Godwin et al. (19891, and Caron et al. (19901, we suspect that glutamine was not a limiting factor in these experiments. The last spinner flask was a control high cell density culture, without glucose/glutamine feeding. The cell density, protein activity, and pH are depicted in Figures 4 and 5. Even at an MOI of 10, the cell growth was not totally inhibited by the viral infection (Figure 4). The cell densities in two flasks increased from the time of infection until 12 h post infection (hpi), which indicates that all cells were infected in the three flasks only after
"-2'---:IIJ,, 5.9
0 0
20 20
40 40
60 60 Time (hr.)
80 80
100 100
120 , 120
Figure 5. (a)Epoxide hydrolase activity in three cultures infected at MOI = 10 (Figure 4). (b) pH for the same three experiments. 12 hpi. A t 48 hpi, the viabilities for the dilution, fed, and
unfed cultures, were 0.77, 0.81, and 0.79, respectively (Figure 4). After that time the cell viability continually decreased (not shown). The epoxide hydrolase activity (Figure 5a) continued to increase for the culture with fresh medium until 118hpi. In contrast, in the high cell density cultures, epoxide hydrolase activity increased more rapidly initially and then decreased a t 88 hpi for the glucose/ glutamine fed culture and at 48 hpi for the unfed culture.
Biotechnol. frog., 1993, Vol. 9, No. 4 l m l GIu.(PM)
a
1'"
Time (hr)
0.1 5
E
3 0.08 .2
5 a
A
3 Q)
0.04
P
W
0.02
0
Time (hr) Figure 6. (a) Viable cell density and glucose concentration for three cultures of Sf-9 cells infected at t = 155 h and MOI = 5: glucose/glutamine fed fermentor culture ( 0 ,viable cells; B, -, glucose);nonfed high cell density spinner flask control (A,viable cells; A,-, glucose); low cell density, freshlydiluted spinner flask culture (v,viable cells; glucose not shown). (b) Cell viability and epoxide hydrolase activity for the same three cultures as a.
A significant improvement in protein production from glucose and glutamine feeding was seen as early as 12 hpi. Coincidentally, the glucose concentration in the unfed culture had dropped dramatically by this time (Figure 4). The increase in activity continued well into the production phase, and since the viable cell concentration in the fed culture was significantly lower after 48 hpi, the increase in activity was due primarily to an increase in specific productivity (Table I). In all cultures with high productivity, we noticed a decrease in activity late in the production phase. We suspect that the epoxide hydrolase activity may be subject to proteolysis, but this is presently under investigation. Also, note that the viable and total cell densities for these two high cell density cultures were roughly similar immediately following viral infection and feeding (Figure 4), with a slightly higher specific growth rate for the fed culture. This is discussed and more clearly demonstrated in the following fermentor experiments. Finally, the pH of these three flasks exhibited different trends during the infection process (Figure 5b). In the unfed cell culture, the pH increased from 6.0 to 6.9, which
is consistent with previous reports suggesting that, with low glucose concentration, the cells utilize amino acids (e.g., glutamine) as a carbon source and secrete ammonia (Kamen et al., 1991). In contrast, the pH of the fed culture remained relatively constant, which illustrates a balance between the utilization of sugars and amino acids. Finally, the pH decreased slightly for the dilution culture, which may be caused by lactate or alanine secretion, the end products at higher residual glucose concentration. In our previous study involving infected cells resuspended in fresh media, we found that lactate and ammonia were depleted from the media during the first 50-60 h post infection and both increased thereafter, although lactate increased to a greater extent. This is consistent with the resuspended culture in this work. Glucose/Glutamine Feeding: Fermentor. The cell growth curve for a fermentor experiment is shown in Figure 6a. Cells were infected with an MOI of 5 at 153 h (2.9 X lo6 cells/mL). A 40-mL aliquot was taken after addition of the virus solution, 30 mL of which was subcultured in a 50-mL spinner flask and designated as the unfed culture.
Blotechnol. hog., 1993, Vol. 9, No. 4
The remaining 10mL was diluted 3-foldwith fresh medium as the high specific protein production control. After part of the culture was removed to spinner flasks, the culture in the fermentor was fed with 14mL of 200 mM glutamine and 2.9 mL of 2 M glucose. Subsequently, the glucose concentration in the fermentor was maintained by occasional addition of 2 M glucose (see Figure 6a). In agreement with previous reports (Licari and Bailey, 1991),a lower MOI in this set of experiments did not totally inhibit cell growth (Figure 6a). The cells in the fermentor and in the unfed culture continued to grow until 40 hpi. However, the viable cell number in the diluted culture began to decrease after 16 hpi (Figure 6b). Also note that both the viable and total cell densities for the fed culture in the fermentor were higher than those in the unfed culture. This is consistent with the spinner flasks and is attributed to the influence of glucose and glutamine on the Sf-9 specific growth rate. The cell densities and MOI for both high cell density cultures were the same at the time of infection, yet the fed culture increased in cell number more rapidly. Also consistent with animal or hybridoma cell growth, this demonstrates that Sf-9 cell growth is dependent on glucose and glutamine concentrations, and subsequent models of cell growth and product expression at these cell densities should reflect this dependence. The epoxide hydrolase activity (Figure 6b) for the diluted culture increased from the beginning of infection to 77 hpi (230 h). The two higher cell density cultures increased more rapidly as expected, but the glucose and glutamine fed culture ultimately reached a 50% higher yield than the unfed culture (Table I and Figure 6b). The rate of increase in activity was similar between the two cultures. More importantly, the epoxide hydrolase activity in the fed culture increased even though the viability had dropped to76.2 5% by210 h. Again, theglucose and glutamine feed enhanced the specific productivity of the Sf-9 cells. However, the specific yields from the fermentor experiment were lower than those from the spinner flask experiment, although the benefits from feeding were again demonstrated by comparing the fed to the unfed cultures. The reason for this decrease was unknown, although it was not linked to the fermentor since the subcultured cells were in spinner flasks, but may have been a cell agefpassage effect as noted by Caron et al. (1992). When comparing the overall performance of the effects of oxygen,glucose, and glutamine feeding, the total protein activity of the DO-controlled culture was 3 X that of the uncontrolled culture. Further, the glucose/ glutamine fed culture was 2-fold higher in yield than the DO-controlled culture. Ultimately, the epoxide hydrolase yield was 40-60 mg/L (Lacourciere et al., 1993).
Conclusions This study demonstrates that oxygen, glucose, and glutamine must be available in ample supply, not only for recombinant protein synthesis but also for viral replication. The addition of glucose and glutamine enhanced cell growth and density (volumetric yield) as well as specific productivity in high cell density batch cultures. Glucose/ glutamine fed-batch spinner flask cultures resulted in similar yields of recombinant protein as cultures with replenished fresh media. Finally, oxygen limitations found in larger spinner flasks were diminished in sparged and agitated bioreactors.
Acknowledgment The authors acknowledge the contributions of Dr. Vik Vakharia, who constructed vEHX-6 and provided the viral
361
infection solution, and Dr. Richard Armstrong, who provided substrates for epoxide hydrolase activity assays. Also, Dr. Govind Rao graciously provided the IBI Biolyzer for LDH activity. Financial support of this work was provided in part by the Center for Agricultural Biotechnology (Ph.D. fellowship for M.-Y.W.).
Literature Cited Agathos, S. N. ProductionScale Insect Cell Culture. Biotechnol. Adv. 1991, 9, 51-68.
Armstrong, R. N.; Levin, W.; Jerine, D. M. Hepatic Microsomal Hydrolase. J. Biol. Chem. 1980,255 (lo), 4698-4705. Atkinson, A. E.; Weitzman, M. D.; Obosi, L.; Beadle, D. J.; King, L. A. Baculovirus as Vectors for Foreign Gene Expression in Insect Cells. Pestic. Sci. 1990,28, 215-224. Cameron, I. R.; Bishop, D. H. L. Insect Cell Culture Technology in Baculovirus Expression Systems. TIBTECH 1989, 7,6670.
Caron, A. W.; Kamen, J.; Massie, B. High-Level Recombinant Protein Productionin Biorectors Using the BaculovirusInsect Cell Expression System. Biotechnol. Bioeng. 1990,36,11331140.
Fraser, J. J. Expression of Eucaryotic Genes in Insect Cell Cultures. In Vitro 1989, 25 (3), 225-235. Godwin, G.; Belisle, B.; DeGiovanni, A.; KoNer, J.; Gong, T.; Wojchowski, D. Ex-cell 400TM, for Serum-Free Growth of Insect Cells and Expreesionof RecombinantProteins. In Vitro 1989,25 (31, 17a.
Hink, W. F. Established Insect Cell Line from the Cabbage Looper, Trichoplusia ni. Nature (London) 1970, 226, 466467.
Kamen, A. A.; Tom, R. L.; Caron, A. W.; Chavarie, C.; Kamen, J. Culture of Insect Cells in a Helical Ribbon Impeller Bioreactor. Biotechnol. Bioeng. 1991, 38, 619-628. Knudson, D. L.; Tinsley, T. W. Replication of a Nuclear PolyhedrosisVirus in a ContinuousCellCultureof Spodoptera frugiperda: Purification, Assay of Infectivity, and Growth Characteristics of the Virus J. Virol. 1974, 14 (4), 934-944. Lacourciere, G. M.; Vakharia, V. N.; Tan, C. P.; Cobb, D. I.; Edwards, G. H.; Moos, M.; Armstrong, R. N. Interaction of Hepatic Microsomal Epoxide Hydrolase Derived from a Recombinant BaculovirusExpression Systemwith an Azaarene Oxide and an Aziridine Substrate Analogue. Biochemistry, in press. Lazarte, J. E.; Tosi, P.-F.; Nicolau, C. Optimization of the Production of Full-LengthrCD4 in Baculovirus-Infected Sf9 Cells. Biotechnol. Bioeng. 1992,40,214-217. Licari,P.; Bailey, J. E. Factors Influencing Recombinant Protein Yields in an Insect Cell-Baculovirus Expression System: Multiplicity of Infection and Intracellular ProteinDegradation. Biotechnol. Bioeng. 1991, 37, 238-246.
Lindsay, D. A.; Betenbaugh,M. J. Quantificationof Cell Culture FactorsAffecting Recombinant Protein Yields in BaculovirusInfected Insect Cells. Biotechnol. Bioeng. 1992,39,614-618. Luckow, V. A.; Summers, M. D. Trends in the Development of BaculovirusExpression Vectors. Biol Technology 1988,6,4755.
Scott,R. I.; Blanchard,J. H.; Ferguson, C. H. R. Effects of oxygen on recombinantprotein production by suspension cultures of Spodoptera frugiperda (Sf-9)insect cells. Enzyme Microb. Technol. 1992,14, 798-804.
Summers, M. D.; Smith, G. E. A Manual of Methods for Baculovirus Vectors and Insect Cell Culture Procedure. Tex. Agric. Exp. Stn., [Bull.] 1987, No. 1555.
Wang, M. Y.; Vakharia, V.; Bentley, W. E. Expression of Epoxide Hydrolase in Insect Cells: A Focus on the Infected Cell. Biotechnol. Bioeng., in press. Weiss, S. A.; Orr, T.; Smith, G. C.; Kalter, S. S.; Vaughn, J. L.; Dougherty, E. M. Quantitative Measurement of Oxygen Consumption in Insect Cell Culture Infected with Polyhedrosis Virus. Biotechnol. Bioeng. 1982, 26, 1145-1154. Accepted February 18, 1993.
