Biotechnol. Prog. 1994, 10, 314-319
314
Clonal Variation in the Spodoptera frugiperda IPLB-SF21-AE Insect Cell Population Murali K. Pasumarthy and David W. Murhammer. Department of Chemical and Biochemical Engineering, The University of Iowa, Iowa City, Iowa 52242
Clones have been isolated from the heterogeneous Spodoptera frugiperda IPLB-SF21AE insect cell population. Five of these clones, in addition to the parent cell line and the SF9 cell line (another clonal isolate of the parent cell line), have been compared in regards to morphology, growth, budded virus synthesis, and recombinant protein synthesis. No significant differences in cell morphology were found among these cell lines. There was, however, a significant difference in the average cell size, with diameters ranging from 9.30 f 0.184 to 11.11f 0.22 pm and from 9.17 f 0.05 to 11.25 f 0.24 pm for cells growing in Excell 401 serum-free medium in spinner flask cultures and in TNM-FH medium supplemented with 10% FBS in tissue flask cultures, respectively. While no significant differences in the growth rates were found in TNM-FH medium containing 10% calf serum, significant differences were found in Excell 401 serum-free medium, with population doubling times ranging from 38.5 f 6.6 to 64.5 f 6.4 h in spinner flask studies. Significant differences in expression levels of Escherichia coli @-galactosidase(@-gal)were also found in both 12-well plates and spinner flasks. In the 12-well plate studies, the peak levels of @-galactosidaseobtained by these cell lines ranged from 0.332 f 0.091 t o 0.805 f 0.117 mg/106 cells and from 0.580 f 0.130 to 1.458 f 0.132 mg/106 cells in Excell 401 and Hyclone Hy-Q serum-free media, respectively. In the spinner flask studies, peak expression levels ranged from 0.128 f 0.053 to 0.573 f 0.215 mg/106cells in Excell 401 serum-free medium. Significant differences were also found in the expression levels of budded virus, which ranged from 64 f 29 to 1125 f 521 plaque-forming units (pfu)/cell and from 67 f 31 to 233 f 95 pfu/cell for the wildtype and recombinant (@-gal)Autographa californica nuclear polyhedrosis viruses, respectively.
Introduction The insect cell/baculovirus expression system is widely used in the synthesis of heterologous proteins, mainly due to its potentially high expression levels and its ability to properly perform most posttranslational modifications in a manner comparable to mammalian cells (Luckow, 1991). These posttranslational modifications include the cleavage of signal sequences, targeting to the nucleus and the cell surface, phosphorylation, formation of disulfide-linked oligomeric complexes, palmitylation, and N-linked and 0-linked glycosylation. Furthermore, baculoviruses are not pathogenic to vertebrates or plants, and they do not employ transformed cells. Both the baculovirus expression vector and the host insect cell line are important considerations in the use of the insect cell/baculovirus expression system. While extensive research has been devoted to the development of the baculovirus expression vector (Luckow and Summers, 1988; Luckow, 1991; O’Reilly et al., 1992), only limited research has been devoted to the development of host insect cell lines. Most research regarding host insect cell lines has involved the comparison of protein expression levels indifferent celllines (Wickhamet al., 1992;Wickham and Nemerow, 1993;Hink et al., 1991;Oganah et al., 1991; Betenbaugh et al., 1991; King et al., 1991). These studies demonstrated a wide range of expression levels between different cell lines. In addition, recent studies have suggested that some cell lines are superior to others in their ability to process proteins (Wickham and Nemerow, 1993). 8756-7938/94/3010-0314$04.50/0
In addition to the investigation of different cell lines, isolation of clones from insect cell lines may also be a method that can be used to find cell lines with desirable characteristics, e.g., high expression levels. This hypothesis is based on evidence that suggests that insect cell lines are generally heterogeneous. This evidence includes variations in ploidy, the method of cell line isolation, and variations in cell properties. Karyotyping of lepidopteran cells, which include the cell lines susceptible to baculovirus infection, has demonstrated that the cells are heteroploid with a modal chromosome number of approximately 100, as compared to predominantly diploid dipteran insect cell lines (Ennis and Sohi, 1976; Hink, 1979). This heterogeneity demonstrates that cells within a given lepidopteran cell line are not generally genetically identical. The method by which insect cell lines are usually isolated is also consistent with having heterogeneous cell populations. For example, the Spodoptera frugiperda IPLB-SF21-AE insect cell line originated from a population of cells that was isolated from S. frugiperda ovarian tissue (Vaughn et al., 1977) and then adapted to growth in medium free of insect hemolymph (Gardiner and Stockdale, 1975); Le., the cells were not cloned. Since cells within these heterogeneous populations are probably not genetically identical, differences in properties such as morphology, growth rate, viral productivity, and recombinant protein expression levels may exist. Previous results have shown clonal variations in many properties within heterogeneous insect cell populations. Distinct morphological differences within the Spodoptera exigua UCR-SE-1 population have been observed, with
0 1994 American Chemical Society and American Institute of Chemical Engineers
Biotechnol. Prog., 1994, Vol. 10, No. 3
both epithelial-like and spindle-shaped cells coexisting (Gelernter and Federici, 1986). The spindle-shaped cells support replication of both S. exigua and Autographa catifornica nuclear polyhedrosis viruses, while the epithelial-like cells only support replication of the A. californica nuclear polyhedrosis virus (AcNPV). Variations of cellular properties within the heterogeneous Trichoplusia ni TN368 have also been observed, including growth rates (Volkman and Summers, 1975, 1976; Brown and Faulkner, 1975), attachment levels (Billimoria and Carpenter, 1983),and the ability to replicate AcNPV (Volkman and Summers, 1975,1976). Similar variations have been observed among clonal populations of various Heliothis zea cell lines (Corsaro and Fraser, 1987;Lenz et al., 1991). In regards to recombinant protein expression levels, Hink et al. (1991) found significant variations in the production levels of Escherichia coli 0-galactosidase, human plasminogen, and pseudorabies gp50T surface protein within both the S. frugiperda IPLB-SF21-AE and S. exigua UCR-SE-1 heterogeneous populations. Unfortunately, only two IPLB-SF21-AE clones (SF9 and IPLBSF21-AE-15) and one UCR-SE-1 clone (UCR-SE-la), in addition to the parent cell lines, were investigated. These previous results clearly demonstrate the potential for improving recombinant protein expression levels by cloning heterogeneous insect cell populations. In the present research, a variety of clones isolated from the heterogeneous S. frugiperda IPLB-SF21-AE insect cell population are compared in regards to their morphology, growth, budded virus synthesis, and recombinant protein expression in order to determine whether cloning can lead to cell lines posssessing characteristics (e.g., expression levels) superior to the parent population.
Materials and Methods Cell Lines, Viruses, and Media. The Spodoptera frugiperda SF9 and IPLB-SF21-AE insect cell lines were obtained from Max D. Summers (TexasA&M University) and W. Fred Hink (Ohio State University), respectively. The viral constructs for wild-type and recombinant (941 @-gal)Autographa californica nuclear polyhedrosis viruses (AcNPV) were obtained from Invitrogen (San Diego, CA) and Max D. Summers, respectively. Excell 401 insect cell medium was obtained from JRH Biosciences (Lenexa, KS), while the fetal bovine serum (FBS), bovine calf serum (CS), and Hy-Q serum-free medium were obtained from HyClone (Logan, UT). The TNM-FH medium was prepared using chemicalspurchased from Sigma (St.Louis, MO), as described by Summers and Smith (1987). Cell densities and sizes were determined using a Coulter multisizer, and viabilities were determined using the trypan blue exclusion technique. Cells were maintained in 25-cm2 tissue culture flasks (Corning) in TNM-FH medium containing 10% serum (either FBS or CS) utilizing standard protocols (Summers and Smith, 1987). Cloning. Midexponential phase S. frugiperda IPLBSF21-AE cells growing in 25-cm2 tissue culture flasks containing TNM-FH medium supplemented with 10% calf serum were diluted with "conditioned" medium to a concentration of 1 cell per 200 pL. The conditioned medium was obtained from cells in the midexponential growth phase and was supplemented with antibiotics (Sigma) to a final concentration of 50 pg/mL streptomycin and 50 units/mL penicillin. The diluted cell suspension was then added to 96-well plates (200 pL/well). After the cells were allowed 1-2 h to attach, the wells were investigated under the microscope, and those containing only one cell were marked. Any cell population arising
315
from one of these marked wells was considered a clonal population. We found that the cells would not propagate in the absence of conditioned medium. Adaptation of Cells to Serum-Free Medium and Suspension Cultures. All of the cell lines were adapted to Excell 401 serum-free medium and suspension cultures prior to obtaining the growth rates and protein expression levels. The adaptation to serum-free medium was conducted in 25-cm2tissue culture flasks by diluting the cell suspension with an equal volume of Excell 401 serum-free medium at each subculture. Therefore, the serum concentration was reduced by 50% at each subculture. This procedure was repeated over 10 times for each cell line. The cells were then grown in 75-cm2tissue culture flasks and transferred to 50-mL spinner flasks. These cells were then subcultured a t least three times in the spinner flasks prior to obtaining the growth and protein expression data. Growth Studies. Growthxates were obtained in 50mL spinner flasks (Bellco Model 1967-00050)maintained at 28 "C and 100 rpm in Excell 401 serum-free medium and TNM-FH medium supplemented with 10% calf serum. The cell density was determined twice daily, and each of these densities was based on three cell counts. The population doubling times (PDT) were calculated from a least-squares fit of the growth curve data in the exponential growth region. Protein Expression in Suspension Cultures. Cells adapted to Excell 401 serum-free medium and suspension growth were infected when they reached a density of approximately 106 cells/mL (midexponentialgrowth phase) at a multiplicity of infection (MOI) of 10with recombinant baculovirus (941 &gal). The spinner flask cultures were maintained at 28 OC and 100 rpm throughout the experiments. Every 24 h the cell density and viability were determined, and samples were taken for future 0-gal quantification. Samples were prepared for @-galquantification by removing 1mL of cell suspension and pelleting the cells by centrifugation at lOOg for 10 min. The supernatant was used for determining the extracellular levels. The cell pellet was used to determine the intracellular levels and was prepared by rinsing twice with 0.01 M PBS and lysing the cells with a Virsonic 300 sonicator operated at 10% maximum power for 30 s. Samples were stored at -85 "C until the assays were performed. Protein Expression in StationaryCultures. Clones were grown in 75-cm2tissue culture flasks in TNM-FH medium supplemented with 10% FBS. Cells in the exponential growth phase were isolated from these tissue culture flasks and diluted with medium to a density of 106 cells/mL. The resulting cell suspension was then used to seed 12-well plates a t a density of lo6 cells/well (1 mL/ well). Preliminary experiments in the 12-well plates indicated that more than 95 % of cells attached after a 2-h incubation for all clones. The cells were infected with recombinant 941@-galAcNPV at an MOI of 10. The plates were incubated a t 28 O C for 2 h to allow for uptake of virus, and then the medium was removed and replaced with either 1mL of fresh Excell 401 serum-free medium or 1 mL of fresh Hy-Q serum-free medium (Broussard and Summers, 1989). The plates were incubated at 28 OC. Samples were taken every 24 h and prepared for 0-gal quantification, as in the case of the suspension cultures (one well was used for each sample). Budded Virus Production. Exponential growth phase cells in TNM-FH medium supplemented with 10% CS in 75-cm2 tissue culture flasks were placed in 15-mL centrifuged tubes, where they were infected at an MOI of 10 with either wild-type AcNPV or recombinant 941 @-gal
316
Table 1. Average Cell Diameter of Clones Obtained from the Spodoptera fiueiperaa IPLB-SF21-AEParent Cell Linea diameter (rrm) clone Excell 401 TNM-FH + 10% FBS 1 10.42 f 0.05 (n = 15) 10.39 f 0.04 (n = 13) 2 9.35 f 0.87 (n = 6) 9.17 f 0.05 (n = 7) 9.39 f 0.12 (n = 13) 9.59 f 0.08 (n = 7) 3 4 9.30 f 0.18 (n = 15) 9.77 f 0.16 (n = 7) 5 9.78 f 0.06 (n = 16) 9.86 f 0.06 (n = 7) SF9 11.25 f 0.24 (n = 7) 11.11f 0.22 (n = 15) SF21 10.70 f 0.17 (n = 16) 10.44 f 0.09 (n = 7) 0 The diameters were determined at various times in midexponential growth phase in Excell 401 serum-free medium in 50-mL spinner flasks and in TNM-FH medium supplemented with 10% FBS in 75-cm2tissue culture flasks. The f values represent the 95% confidence levels based on n data points.
AcNPV. The infected cells were incubated at 28 OC for 1-2 h. The infected cells were then centrifuged at lOOg for 15 min, resuspended in a volume of Excell 401 serumfree medium required to obtain a cell density of lo6cells/ mL, and then placed in 12-well plates (1mL/well). The plates were incubated at 28 "C. The supernatants were collected at 96 h postinfection and stored at 4 OC until assayed for budded virus. Assays. The @-galconcentration was determined using the procedure described by Miller (1972a). Briefly, samples were diluted in Z-buffer (0.06 M Na~HP04,0.04 M NaH2P04,O.Ol M MgS04, and 0.05 M P-mercaptoethanol) and equilibrated in a 28 OC bath. o-Nitrophenyl galactosidewas added to these tubes. After sufficient color formation (typically about 10-15 minutes), the reaction was stopped by adding 1 M sodium carbonate solution, and the absorbance at 420 nm was determined, the intensity of which is proportional to the amount of o-nitrophenol released. One unit of 0-gal is defined as the amount of the enzymethat releases 1nmol of o-nitrophenol per minute at 28 "C and pH 7.0. One milligram pure 0-galactosidase corresponds to 300 000 units (Miller, 1972b). The budded virus titer was determined using the LCID50 method (Hughes and Wood, 1986; Summers and Smith, 1987) in conjunction with the Spearman-Karber analysis method (Hughes and Wood, 1986).
Results and Discussion Cell Morphology and Size. No significant variations in cell morphology were observed among the clones, i.e., all of the clones appeared rounded. This is in contrast to the differences observed among the clones of S. exigua UCR-SE-1 and H. zea IPLB-HZ-1075 cell lines (Gelernter and Federici, 1986; Corsaro and Fraser, 19871, where considerable variations in cell morphologies were found between clones. The average cell sizes were obtained from the size distribution given by the Coulter multisizer. The average cell diameters of the SF21clones varied from 9.30 f 0.18 to 11.11 f 0.22 pm and from 9.17 f 0.05 to 11.25 f 0.24 pm in Excell 401 medium and TNM-FH medium supplemented with 10% FBS, respectively (Table 1). Cell Growth. No significant variations in the population doubling times (PDT) were found among the clones grown in TNM-FH medium supplemented with 10% calf serum, with a typical doubling time of approximately 30 h (Table 2). This result was not surprising since the cells were grown long-term in this same medium prior to cloning; thus, any clones that grew significantly faster than the others should have had sufficient time to become dominant within the culture. The results obtained in serumcontaining medium are contrasted with those obtained in
Biotechnol. Prog., 1994, Vol. 10, No. 3
Table 2. Population Doubling Times (PDT) of Clones Obtained from the Spodoptera frugiperda IPLB-SF21-AE Parent Cell Line in SO-mL Spinner Flasks in TNM-FH Medium Containing 10% CS and in Excell 401 Serum-Free Medium.
PDT (h) TNM-FH + 10% CS Excell 401 1 35.1 f 15.9 (n = 4) 43.4 f 5.4 (n = 4) 2 30.7 f 3.5 (n = 4) 46.9 f 5.8 (n = 5) 31.6 f 5.5 (n = 4) 3 51.2 f 5.3 (n = 4) 4 30.9 f 9.1 (n = 4) 64.5 f 6.4 (n = 5) 38.0 f 17.5 (n = 4) 5 39.9 f 6.7 (n = 8) 28.3 f 2.3 (n = 7) SF9 43.2 & 6.0 (n = 5) SF21 35.6 f 9.2 (n = 5) 38.5 f 6.6 (n = 4) The population doubling times were calculated from a leastsquares fit of the growth curve data in the exponential growth region. The f values represent the 95% confidence levels based on n data points. clone
Excell 401 serum-free medium, in which the population doubling times varied from 38.5 f 6.6 to 64.5 f 6.4 h (Table 2). This variation is not unexpected since the cells were cloned prior to adaptation to serum-free medium, and there is no reason to believe that a cell's ability to grow in serumcontaining medium is directly related to its ability to grow in serum-free medium. The results obtained in serum-containing medium are consistent with those of Volkman and Summers (1975, 1976), who found little variation in the growth rates of Trichoplusia ni TN368 clones in TNM-FH medium supplemented with FBS. Although no previous research has compared the growth rates of SF21 clones in either serum-containing or serum-free medium, there have been studies in which the growth rate of the SF9 cell line has been determined. The population doubling time of 28.3 f 2.3 h found in the present research utilizing TNM-FH medium supplemented with 10% CS is somewhat larger than the doubling times ranging from 16 to 20 h found for the SF9 cell line in TNM-FH medium supplemented with 5-1596 FBS (Murhammer and Goochee, 1988)and the 18 h time found in TNM-FH medium supplemented with 10% FBS (Ogonah et al., 1991). The difference between the growth rates found in serum-supplemented media in the present research (CS) and in previous research (FBS) could be due to the superior growth characteristics of the FBS-supplemented medium. The population doubling time obtained in the present research of 43.2 f 6.0 h for the SF9 cell line in Excell 401 serum-free medium is significantly larger than the population doubling times of 20-28 h (Murhammer, 1989),-2330 h (Betenbaugh et al., 19911, and 24 h (Ogonah et al., 1991) found by previous researchers in Excell 400 serumfree medium. The differences between the growth rates of the SF9 cell line in Excell serum-free media found in the present study (Excell 401) and in previous studies (Excell 400)cannot easily by explained in terms of growth media differences. Actually, one would expect to find a higher growth rate in the present research, in contrast to what was actually found, since Excell 401 medium is reported by the manufacturers to be designed for increased cell growth. A possible explanation for the observed differences in growth rates is that the SF9 cells used in the present research are not genetically identical to those used in the previous research. This hypothesis assumes that the cells' genetic makeups can change with passage and that this change is also a function of the subculturing technique. In support of this hypothesis, Ennis and Sohi (1976) demonstrated that the modal chromosome number of a Malacosoma disstria insect cell line decreased with passage. In addition, Hilwig and Eipel (1978/79) found
317
Biotechnol. Prog., 1994, Vol. 10, No. 3
48hpi 72hpi 96hpi EZI 1 2 0 h p i
1
2
3
4
5
sf9
sf21
Clone Figure 1. E. coli @-galactosidase(@-gal)expression levels at various times postinfection (PI) of clones obtained from the Spodoptera frugiperda IPLB-SF21-AE parent cell line in 50mL spinner flasks in Excell 401 serum-free medium. The cells were infected in midexponential growth phase at a multiplicity of infection of 10. The values given are based on four spinner flasks and represent the total @-galconcentration (intracellular + extracellular). The error bars represent the 95% confidence levels.
that the ploidy of a S. littoralis insect cell line varied with cell age and culture conditions. We are currently investigating this hypothesis by determining the karyotypes of SF9 cells obtained from a variety of laboratories. The karyotypes of the SF21 clones and their genetic stabilities are also under investigation. Recombinant Protein Synthesis in Suspension Culture. Significant variations in @-galexpressionlevels were found among the SF21 clones in the spinner flask studies utilizing Excell 401 serum-free medium (Figure 1). The best producer of @-galwas clone 2 (0.573 f 0.215 mg/106cells),while clone 3 produced the lowest level (0.128 f 0.053 mg/106 cells). The total (i.e., intracellular and extracellular) concentration of @-galincreased with time postinfection (PI),reaching peak levels at 96 h PI for all of the cell lines except SF21. The decreased levels at 120 h PI were probably due to a combination of the release of intracellular @-galinto the medium following cell lysis and the degradation of the resulting extracellular @-galby proteases released from lysed cells. This hypothesis is consistent with results obtained by Oganah et al. (1991), in which maximum P-gal activity was followed by a reduction in total activity for both SF9 and Trichoplusia ni TN368 cell lines in suspension cultures. Recombinant Protein Synthesis in Stationary Culture. There was also considerable variation of @-gal expression levels among the SF21 clones in stationary cultures utilizing Excell 401 serum-free medium (Figure 2). Clone 2 and the SF9 cell line were the best @-gal producers (0.805 f 0.117 and 0.805 f 0.116 mg/106 cells, respectively), while clone 3 was again the poorest @-gal producer (0.332 f 0.091 mg/106cells). The @-galexpression levels were considerably higher than those obtained in spinner flasks for all of the cell lines. A probable explanation for these higher expression levels was that the replacement of the medium with fresh medium following viral infection eliminated nutrient depletion, which could otherwise occur (Kamen et al., 1991). In addition, Lindsay and Betenbaugh (1992) found that replacement with fresh medium increased recombinant protein expression levels in suspensioncultures. It should
4 5 sf9 sf21 Clone Figure 2. E. coli @-galactosidase(@-gal) expression levels at various times postinfection (PI) of clones obtained from the Spodoptera frugiperda IPLB-SF21-AE parent cell line in 12well plates in Excell 401 serum-free medium. The cells were infected in midexponential growth phase at a multiplicity of infection of 10. The values given are based on four wells and represent the total @-galconcentration (intracellular + extracellular). The error bars represent the 95 % confidence levels. 1
2
3
also be noted that the time at which the maximum @-gal activity was obtained in the stationaryculture experiments was generally later than the time at which the maximum activity was obtained in the spinner flask studies. It is likely that this behavior resulted from delayed cell lysis in stationary cultures for reasons that are not clear at the present time. Significant variations in @-galexpression levels were also found among the SF21 clones in stationary cultures utilizing Hy-Q serum-free medium (Figure 3). The best producers of @-galwere the SF9 cell line and clone 2 (1.458 f 0.132 and 1.148 f 0.186 mg/106cells, respectively), while the poorest @-galproducer again was clone 3 (0.580 f 0.130 mg/106 cells). The expression levels obtained in Hy-Q medium were considerably higher (in some cases more than 2-fold higher) than those obtained in Excell 401 serum-free medium for all of the cell lines (Figure 2). The reason for this large difference is unknown at the present time. We are currently investigating the expression of glycoproteinsin these clones in order to determinewhether protein glycosylation also varies among the SF21 clonal populations. BuddedVirus (BV) Synthesis. Significant variations in wild-type and recombinant 941 @-galBV production were found among the SF21 clones in stationary culture experiments (Table 3). Clone 2 and the SF21 parent cell line were the best producers of the wild-type virus (1125 f 512 and 1054 f 242 pfu/cell, respectively), while clone 2 and the SF9 cell line were the best producers of recombinant virus (233 f 95 and 210 f 26 pfu/cell, respectively). Clone 4 was the poorest producer of both wild-type and recombinant virus (64 f 29 and 67 f 31 pfu/mL, respectively). These variations in BV synthesis among the clones are consistent with those found among clones isolated from other heterogeneous insect cell populations, including Trichoplusia ni TN368 (Volkman and Summers, 1975,1976)and two Heliothis zea cell lines (Lenz et al., 1991). The variation in budded virus synthesis levels among the clones is not unexpected due to both the heteroploid nature of lepidopteran cells, Le., cells cannot be regarded as genetically identical, and the fact that the
Biotechnol. Prog., 1994, Vol. 10, No. 3
318
the experiments (with the exception of @-galproduction in Hy-Q medium (Figure 3),in which clone 2 is the second best producer). Comparison of the correspondingexpression levels among the other clones demonstrates that high BV expressionlevels do not necessarily correspond to high recombinant protein expression levels. This was also demonstrated by King et al. (1991) in experiments where BV and recombinant protein (@-gal)expressionlevels were compared between Mamestra brassicae MB0507 and SF21 cell lines. From a mechanistic standpoint, there is no reason to expect a correlation between BV and recombinant protein synthesis, since recombinant protein expression driven by the polyhedrin promoter occurs independent of BV synthesis.
1.6 G . Y 3
0
Conclusions 1
2
3
4
5
sf9
sf21
Clone Figure 3. E. coli 6-galactosidase (6-gal) expression levels at various times postinfection (PI) of clones obtained from the Spodoptera frugiperda IPLB-SF-21AE parent cell line in 12well plates in Hy-Q serum-freemedium. The cells were infected in midexponentialgrowth phase at a multiplicity of infection of 10. The values given are based on four wells and represent the total @-galconcentration(intracellular+ extracellular). The error bars represent the 95% confidence levels. Table 3. Budded Virus Production in Clones Obtained from the Spodoptera frugiperda IPLB-SF21-AE Parent Cell Line in 12-Well Plates in Excel1 401 Serum-Free Mediuma
clone 1 2 3 4 5
SF9 SF21
virus produced (pfu/cell) wild type recombinant @-gal) 377 f 159 84 f 19 1125 f 521 233 f 95 556 f 104 161 f 84 64-f 29 67 f 31 460 f 172 113 f 29 374 f 78 210 f 48 1054 f 242 196 f 26
aThe cells were infected in midexponential growth phase at a multiplicity of infection of 10. The values given are based on four wells and represent the viral titer obtained at 96 h postinfection. The f values represent the 95% confidence levels.
infection process is not synchronous, i.e., different cells will uptake different numbers of viral particles, which in turn affects the infection kinetics (Volkman et al., 1992). Specifically, this variation could occur at any one or more of the steps involved in BV synthesis, including (1)uptake of BV by adsorptive endocytosis, (2) migration of BV through the cytoplasm to the nucleus, where entry is gained via nuclear pores, (3) BV replication, and (4) transfer of BV from the nucleus to the cytoplasmicmembrane, where viral particles bud from the membrane (O'Reilly et al., 1992). The titers of wild-type virus were significantly higher than those obtained for the recombinant virus for all cell lines except clone 4. These results are in contrast with those of King et al. (1991), who found no. significant differences in the expression levels of the wild-type and recombinant (@-gal)viruses. It is not understood why such a variation occurs in the present research, but it must involvedifferences between the wild-type and recombinant BVs in regard to the rate and/or efficiency of one or more of the steps in the BV infection process discussed above. The only definitive trend that can be observed when comparing the BV (Table 3) and recombinant protein expression levels (Figures 1-3) among the clonal populations is that clone 2 is the best producer in almost all of
Significant differences have been found among clones isolated from the Spodoptera frugiperda IPLB-SF21-AE insect cell line in regard to cell size, cell growth, recombinant protein synthesis, and budded virus synthesis. In the present study, one clone (designated as clone 2) was found to be superior to the other clones (including SF9) and the parent cell line in regard to recombinant @-galactosidase and budded virus synthesis. These results clearly demonstrate that clone selection can have a significant impact on recombinant protein expression levels; therefore, clones, in addition to a variety of established cell lines, should be investigated when optimizing recombinant protein expression levels with the insect cell/ baculovirus expression system. Cellular properties other than expression levels, however, should be considered when selecting a cell line for scale-up of recombinant protein expression, including growth in serum-free medium, the ability to grow in suspension, and the ability of the host cell to perform posttranslational modifications (e.g., glycosylation) in a manner that results in a biologically active product (Murhammer, 1991).
Literature Cited Betenbaugh,M. J.;Balog,L.; Lee, P. S. Productionof recombinant proteins by baculovirus infected gypsy moth cells. Biotechnol. Prog. 1991, 7 (5), 462-467. Billimoria,S. L.; Carpenter, W. M. TN368A an attached strain of Hink's Trichoplusiu ni (TN368) cell line. I n Vitro 1983, 19,870-874. Broussard, D. A.; Summers,M. D. Effects of serum concentration and media composition on the level of polyhedrin and foreign gene expressionby baculovirusvectors. J.Inuertebr. Pathol. 1989,54, 144-150. Brown, M.; Faulkner, P. Factors affecting the yield of virus in a cloned cell line Trichoplusia ni infected with a nuclear polyhedrosis virus. J. Inuertebr. Pathol. 1975,26, 251-257. Corsaro, B. G.; Fraser, M. J. Characterizationof clonal populations of the Heliothis zea cell line IPLB-HZ 1075. I n Vitro Cell Deu. Biol. 1987, 23, 855-862. Ennis, T. J.; Sohi, S. S. Chromosomal characterization of five lepidopterancelllines of Malucosoma disstria (Lasiocampidae) and Christoneura fumiferana (Tortricidae). Can. J . Genet. Cytol. 1976,18,471-477. Gardiner, G. R.; Stockdale, H. Two tissue culture media for production of lepidopteran cells and nuclear polyhedrosis viruses. J. Inuertebr. Pathol. 1975,25, 363-370. Gelernter, W. D.; Federici, B. A. Continuous cell line from Spodoptera exigua (Lepidoptera: Noctuidae) that supports replication of nuclear polyhedrosis viruses from Spodopteru exigua and Autographa californica. J.Inuertebr. Pathol. 1986, 48,199-207. Hilwig, I.; Eipel, H. E. Characterization of insect cell lines by DNA content. 2. Angew. Entomol. 1978/79,87, 216-220.
310
Biotechnol. Prog., 1994, Vol. I O , No. 3
Hink, W. F. The 1979 compilation of invertebrate cell lines and culture media. In Invertebrate s y s t e m in vitro; Kurstak, E., Maramarosch, K., Dubendorfer, A., Eds.; ElseviedNorthHolland Biomedical Press: New York, 1979, pp 563-578. Hink, W. F.; Thomsen, D. R.; Davidson, D. J.; Meyer, A. L.; Castellino,F. J. Expressionof three recombinant proteins using baculovirus vectors in 23 insect cell lines. Biotechnol. Prog. 1991, 7,9-14. Hughes, P. R.; Wood, H. A. In vivo and in vitro bioassaymethods for baculoviruses. In The biology of baculoviruses, Vol. II; Granados, R. R., Federici, B. A., Eds.; CRC Press, Inc.: Boca Raton, FL, 1986; pp 1-30. Kamen, A. A.; Tom, R. L.; Caron, A. W.; Chavarie, C.; Massie, B.; Archambault, J. Culture of insect cells in a helical ribbon impeller bioreactor. Biotechnol. Bioeng. 1991, 38, 619-628. King, L. A.; Mann, S. G.; Lawrie, A. M.; Mulshaw, S. H. Replication of wild-type and recombinant Autographa californica nuclear polyhedrosis virus in a cell line derived from Mamestra brassicae. Virus Res. 1991, 19, 93-104. Lenz, C. J.; McIntosh, A. H.; Mazzacano, C.; Monderloh, U. Replication of Heliothis zea nuclear polyhedrosis virus in cloned cell lines. J. Invertebr. Pathol. 1991,57, 227-233. Lindsay, D. A.; Betenbaugh, M. J. Quantification of cell culture factors affecting recombinant protein yields in baculovirus infected cells. Biotechnol. Bioeng. 1992, 39, 614-618. Luckow, V. A. Cloning and expression of heterologous genes in insect cells with baculovirus vectors. In Recombinant DNA technology and applications; Prokop, A., Bajpai, R. K., Ho, C. S., Eds.; McGraw-Hill, Inc.: New York, 1991; pp 97-152. Luckow, V. A.; Summers, M. D. Trends in the development of baculovirus expression vectors. BiolTechnology 1988,6,4755. Miller, J. H. Assay of 8-galactosidase. In Experiments in molecular genetics; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1972a; pp 352-355. Miller, J. H. Purification of 8-galactosidase. In Experiments in molecular genetics; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1972b; pp 398-404. Murhammer, D. W. Scale-up of insect cell cultures: protective effects of Pluronic F-68. Ph.D. Dissertation, University of Houston, 1989. Murhammer, D. W. The use of insect cell cultures for recombinant protein synthesis: Engineering aspects. Appl. Biochem. Biotechnol. 1991, 31, 283-310.
Murhammer, D. W.; Goochee, C. F. Scale-up of insect cell cultures: protective effects of Pluronic F-68. BiolTechnology 1988,6, 1411-1418. Ogonah, 0.;Shuler, M. L.; Granados, R. R. Protein production (8-galactosidase) from a baculovirus vector in Spodoptera frugiperda and Trichoplusia ni cells in suspension culture. Biotechnol. Lett. 1991, 13 (4), 265-270. OReilly, D. R.; Miller, L. K.; Luckow, V. A. Baculovirus expression vectors: a laboratory manual;W. H. Freeman and Company: New York, 1992. Summers, M. D.; Smith, G. E. A manual of methods for baculovirus vectors and insect cell culture procedures; Texas A&M Experimental Station Bulletin No. 1555; Texas A&M University: College Station, TX, 1987. Vaughn, J. L.;Goodwin, R. H.; Tompkins, G. J.; McCawley, P. The establishment of two cell lines from the insect Spodoptera frugiperda (Lepidoptera; Noctuidae). In Vitro 1977,13 (4), 213-217. Volkman, L. E.; Summers, M. D. Nuclear polyhedrosis virus detection: Relative capabilities of clones developed from Trichoplusia ni ovarian cell line TN368 to serve as indicator cells in a plaque assay. J. Virol. 1975, 16 (6), 1630-1637. Volkman, L. E.; Summers,M. D. Comparative studies with clones derived from a cabbage looper ovarian cell line, TN368. In Invertebrate tissue culture: applications in medicine, biology, and agriculture; Kurstak, E., Maramarosch,K., Eds.;Academic Press: New York, 1976; pp 289-296. Volkman, L. E.; Talhouk, S. N.; Oppenheimer, D. I.; Carlton, C. A. Nuclear F-actin: a functional component of baculovirusinfected lepidopteran cells? J . Cell Sci. 1992, 103, 15-22. Wickham,T. J.; Nemerow, G. R. Optimization of growth methods and recombinant protein production in BT1-Tn-5B1-4 insect cells using the baculovirus expression system. Biotechnol. h o g . 1993, 9, 25-30. Wickham, T. J.; Davis, T.; Granados, R. R.; Shuler, M. L.; Wood, H. A. Screening of insect cell lines for the production of recombinant proteins and infectious virus in the baculovirus expression system. Biotechnol. h o g . 1992, 8, 391-396. Accepted January 7, 1994.' ~~
~
~
~~~
~
*Abstract published in Advance ACS Abstracts, April 1, 1994.
