Expression of three recombinant proteins using baculovirus vectors in

Recombinant Autographa California baculoviruses expressing genes for pseudorabies virus glycoprotein (gp50T), human plasminogen (HPg), and ...
0 downloads 0 Views 1MB Size
Download PDF
Biotechnol. hog. 1991, 7, 9-14

9

Expression of Three Recombinant Proteins Using Baculovirus Vectors in 23 Insect Cell Lines W. F. Hink,*ptD. R. Thomsen,i D. J. Davidson,§A. L. Meyer; and F. J. Castellinos Department of Entomology, The Ohio State University, 1735 Neil Avenue, Columbus, Ohio 43210, Upjohn Company, Kalamazoo, Michigan 49001, and Department of Chemistry, University of Notre Dame, Notre Dame, Indiana 46556

Recombinant Autographa California baculoviruses expressing genes for pseudorabies virus glycoprotein (gp50T), human plasminogen (HPg), and @-galactosidase@-gal) were used to infect 23 cell lines or strains. T h e objectives were t o compare amounts of recombinant proteins expressed in the cell lines, compare yields from clones and parent lines, investigate the effects of long-term culture in serum-free medium on production, and determine if some lines yield gp50T with different glycosylation patterns. For HPg, IZD-MB0503 had the highest yield and four other lines (IPLB-TN-R2, IPLBSF-1254, IPLB-LdEIta, and CM-1) had levels above that of SF-9 cells. For gp50T, four lines (IPLB-HvT1, IPLB-SFBlAE, IPLB-SF21AE-15, and IPLB-SF-1254) had higher amounts than SF-9 cells. Some lines yielded gp50T with molecular mass about 1000 daltons larger than t h a t from SF-9 cells, which suggests increased oligosaccharide processing. Equally high levels of P-gal were expressed in three lines (SF-9, IZD-MB0503, and BCIRL-PX2-HNV3). T h e major conclusion is t h a t no single cell line produced highest yields for all three recombinant proteins. Four lines were cultured in serumfree medium for 31-34 passages and then infected with the three recombinant viruses. For most cell line-recombinant combinations, the yields in serum-free medium were equal t o or better than those in serum-supplemented medium. Medium composition had a much stronger effect on foreign gene expression than on susceptibility of cells to wild-type virus.

Introduction The baculovirus expression vector-insect cell culture system is used extensively to produce proteins of agricultural and medical importance with uses of microbial pesticides, pharmaceuticals, and vaccines (Luckow and Summers, 1988). At least 130 proteins from viruses, bacteria, plants, invertebrates, and mammals have been expressed, and most retain biological activities similar to their authentic counterparts (Luckow, 1990). Nearly all published work has been with Autographa californica nuclear polyhedrosis virus (AcNPV) baculovirus recombinants and the Spodoptera frugiperda (SF-9 or IPLBSF-21AE) insect cell lines. However, other recombinant baculovirus-cell culture or live insect systems have been described (Maedaet al., 1985;Iatrou and Meidinger, 1989). Efforts to increase the level of expression of foreign genes in baculovirus expression systems have been directed at optimizing the locations of the foreign genes in relation to transcriptional and translational signals within the polyhedrin gene. Little attention has been directed at the influence cells have on protein expression, and this is the first report on the relative abilities of numerous insect cell lines to produce recombinant proteins. Three different types of genes were inserted into AcNPV: a structural gene from a virus, a bacterial enzyme, and a gene coding for a mammalian proenzyme. This was t The

Ohio State University. Upjohn Co. 5 University of Notre Dame. f

8756-7938/91/3007-0009$02.50/0

done to produce a more general or broadly applicable conclusion about expression in the cell lines tested. Gp50 is a structural envelope glycoprotein of pseudorabies virus and contains only 0-linked oligosaccharides (Thomsen et al., 1990). The gene was modified to delete sequences coding for the hydrophobic transmembrane and hydrophilic cytoplasmicdomains. This results in secretion of the gene product into the culture medium, permitting analysis without interference by cellular proteins. Escherichia coli P-galactosidase was tested because the enzymatic activity is easily quantified. I t also allowed for a direct comparison of expression levels with results from other groups who have used this gene in the analysis of expression levels in the baculovirus expression system. HPg is a zymogen that exists in the circulatory system and is activated by cleavage to produce human plasmin. I t is a single-chain glycoprotein containing 791 amino acids. Expression of HPg in mammalian systems has been difficult because of the ubiquitous presence of intracellular plasminogen activator, which inactivates HPg (Whitefleet-Smith et al., 1989).

Materials and Methods Cell Lines and Media. Table I gives details on the 23 cell lines and strains evaluated. They are from 12 different species of insects within the order Lepidoptera. Lines were maintained, infected, and incubated in the designated serum-supplemented media. Except for IZD-MB0503,the lines were grown in the medium in which they were established or that is currently in use in the laboratory of the scientist who started the lines.

0 1991 American Chemical Society and American Institute of Chemical Engineers

Biotechnol. Prog., 1991, Vol. 7, No. 1

10

Table I. Cell Lines Infected with Baculoviruses cell line designation scientific name

culture medium GTC-100' TNM-FHb GTC-100 TNM-FH TC199-MKd BML-TC/lOe BML-TC/lO f IPL-4lg M&Mh M&M M&M M&M M&M TNM-FH TC199-MK TNM-FH TNM-FH IPL-41 IPL-41 TNM-FH TNM-FH TNM-FH TNM-FH

reference Seiburth and Maruniak, 1988

Anticarsia gemmatalis C Choristoneura fumiferana Granados and Naughton, 1976 Estigmene acrea Lynn et al., 1988 Heliothis uirescens McIntosh and Ignoffo, 1981 Hliothis zea Lynn et al., 1988 Lymantria dispar Lynn et al., 1988 Lymantria dispar Miltenberger et al., 1976 Mamestra brassicae Mitsuhashi and Shozawa, 1985 Mamestra brassicae i Mamestra brassicea Inoue and Mitsuhashi, 1985 Mamestra brassicae Inoue and Mitsuhashi, 1985 Mamestra brassicae Inoue and Mitsuhashi, 1985 Mamestra brassicae Eide et al., 1975 Manduca sexta Chen et al., 1983 Plutella xylostella Gelernter and Federici, 1986 Spodoptera exigua Spodoptera exigua j Spodoptera frugiperda Vaughn et al., 1977 Spodoptera frugiperda k Spodoptera frugiperda Summers and Smith, 1987 Spodoptera frugiperda Vaughn et al., 1977 Rochford et al., 1984 Trichoplusia ni Hink, 1970 Trichoplusia ni a Granados et al., 1986. Hink, 1970. S. S. Sohi, Forest Pest Management Institute, personal communication, 1986. McIntosh et al., 1973. Mitsuhashi and Maramorosch, 1964. J. Mitsuhashi,TokyoUniversity p Gardiner and Stockdale, 1975. f Alsocalled TC-100. Weiss et al., 1981. of Agriculture and Technology, personal communication, 1989. j B. Federici, University of California at Riverside, personal communication, 1989. J. L. Vaughn, USDA-Beltsville, personal communication, 1989. UFL-AG-286 IPRL-CF-1 RTI-EAA IPLB-HvTl BCIRL-HZ-AM1 IPLB-LdEIta IPLB-LdEItf IZD-MB0503 NIAS-MaBr-92 NIAS-MB-25 SES-MaBr-1 SES-MaBr-3 SES-MaBr-4 CM-1 BCIRL-PXP-HNV3 UCR-SE-1 CCR-SE-la IPLB-SF-21AE IPLB-SF-21AE-15 SF-9 IPLB-SF-1254 IPLB-TN-R2 TN-368

The serum-free medium (ISFM) was developed by Maiorella et al. (1988) and used for production of a recombinant protein in SF-9 cells. Gene expression was similar to that observed in serum-supplemented medium. The medium is protein-free and is composed of IPL-41 powder (Weiss et al. 1981), a lipid-ethanol emulsion in Pluronic polyol F68, and an ultrafiltrate of yeastolate. Detailed stepwise procedures for making ISFM are given in Inlow et al. (1989). Four cell lines were maintained in ISFM and, a t the time of this study, had been in ISFM for 5-6 months through 31-35 subcultures. The procedure for infecting the cell lines was designed to eliminate as many variables as possible and modified from Hink et al. (1979). Lines were in exponential growth when exposed to virus, cell density was 7.5 x l o 5 cells/ mL, the volume of virus inoculum was the same for all lines, and no spent media was carried over into the infected cell cultures. To further minimize variations in the condition of cells, a given line was infected with all four viruses a t the same time by using a common pool of cells in exponential growth. Densities of viable cells to be infected were determined by four replicate counts in a hemocytometer. Volumes of cell suspensions required to give a total of 3.75 x lo6 cells were centrifuged a t 360g for 10 min, supernatants were removed, and virus was added at a multiplicity of infection (moi) of 20 plaque-forming units (pfu)/cell. The viral inocula were left on the cells for 1 h at 28 "C with initial swirling to resuspend the cells and repeated agitation every 15min. The viral inocula were removed by centrifugation, and cells were resuspended in 5.0 mL of medium to give 7.5 x IO5 cells/mL. These 5.0-mL cell suspensions were divided into 2.5-mL aliquots and transferred to 35 x 10 mm plastic petri dishes. Infected cells were harvested by removing them with rubber scrapers after 48- and 72-h incubations a t 28 "C. The cell suspensions were centrifuged immediately after collection, the supernatants were removed from cell pellets, and both fractions were frozen. Supernatants were used in the HPg and gp5OT assays and cell pellets for p-galactosidase. Recombinant Baculoviruses. The gp50 gene (Petrovskis et al., 1986) was cloned into the BamHI restriction

site of the transfer vector pAC373 by addingBamH1 linkers to the NdeI (amino terminus) site and the Mae111 (carboxy terminus) site of the gp50 gene. Prior to insertion, a phase termination codon was inserted into the gp50 gene at the Sau3A site. This insertion creates a truncated form of gp50 gene product that lacks the hydrophilic cytoplasmic domain and hydrophobic transmembrane domain. The plasmid containing the truncated gp50 gene is designated PAC-gp50T. Insertion of the truncated gp50 gene into the virus AcNPV-E2 and isolation of the recombinant virus was done by using standard baculovirus expression vector technology (Summers and Smith, 1987). The HPg vector was constructed by cloning the BamHI/ NaeI fragment of the entire cDNA for HPg, excised from p119PN127.6 (McLean et al., 1987),into the BamHI/SmaI sites of the viral transfer vector, pAV6 (Whitefleet-Smith et al., 1989). The pAV6HPg plasmid was cotransfected with AvMNPV wild-type DNA in S. frugiperda cells. The progeny virus from the coinfection were purified by selecting occlusion body negative (OB-) plaques. The virus was further purified by two additional rounds of plaque assays. The $-galactosidase-containing virus was constructed by inserting the @galgene carried on the plasmid pAc360@-gel(Luckow and Summers, 1988) into the wild-type virus AcNPV-E2. Recombinant virus was isolated by addition of 5-bromo-4-chloro-3-indolyl(3-D-galactopyranoside (Xgal) to the overlays during plaque purification and isolation of the blue plaques as previously described (Summers and Smith, 1987). Plasminogen Assays. Enzyme-linked immunosorbent assays (ELISAs) were performed on recombinant HPg from culture media by procedures previously published (Whitefleet-Smith et al., 1989). Protein samples were separated by NaDodS04/PAGE (Laemmli, 1970) on 10% (w/v) polyacrylamide gels. The protein bands were transferred to Immobilon-P membranes (Burnette, 1981). The exact transfer conditions were 4 "C in 25 mM TrisHC1,200 mM glycine, and 15c( (v/v) methanol buffer, pH 8.3, at 20 Vfor 12 h (Davidson et al., 1990). The membranes were then incubated at 37 "C for 1 h in 1'; (w/v) gelatin (Bio-Rad EIA grade) in TBS (blocking buffer). The

Biotechnol. Prog., 1991, Vol. 7, No. 1

11

Table 11. Yields of Recombinant Proteins from Insect Cell Lines cell line designation UFL-AG-286 IPRL-CF-1 BTI-EAA IPLB-HVT1 BCIRL-HZ-AM1 IPLB-LdEIta IPLB-LdEItf IZD-MB0503 NIAS-MaBr-92 NIAS-MB-25 SES-MaBr-1 SES-MaBr-3 SES-MaBr-4 CM-1 BCIRL-PX2-HNV3 UCR-SE-1 UCR-SE-la IPLB-SF-21AE IPLB-SF-21AE-15 SF-9 IPLB-SF-1254 IPLB-TN-R2 TN-368

HPg production, pg/mL 48 h 72 h 0.8 f 0.03 0.6 f 0.04 0 0 0.4 f 0.01 0.4 f 0.01 0.0 f 0.05 1.0 f 0.10 0

0

1.4 f 0.01 1.0 f 0.10 2.6 f 0.30 0 0 0.1 f 0.04 0 0 0.4 f 0.06 0.4 f 0.03 0.8 f 0.10 0.5 k 0.01 0.8 f 0.01 0.1 f 0.03

1.5 f 0.05 1.1 f 0.10 2.6 f 0.32 0 0 0.1 f 0.06 0 0.2 f 0.02 1.5 f 0.09 0.4 f 0.03 0.7 f 0.20 0.5 f 0.06 0.7 f 0.02 0.3 f 0.07 0.8 f 0.10 1.8 f 0.22 2.1 f 0.30 1.4 f 0.25

0.2 f 0.1 1.5 f 0.20 1.3 f 0.30

0.6 f 0.08

&gal production, units 48 h 72 h 467 2467 0

0

40 26 667 1100 30 600 0

650 30 500 0 39 500 7833 83 833 0

0

0

87 0

1070 0 1867 23 333 83 000 50 500 2 1 367 3867 7867 84 667 40 333 34 667 3867

5000 0

0

2566 74 034 84 767 16 367 1034 2767 59 534 24 800 27 267 2234

blocking buffer was replaced with fresh blocking buffer containing 4 pg/mL of monoclonal rabbit anti-HPg (Whitefleet-Smith et al., 1989), and the membrane was incubated at room temperature for 2 h with mixing. The filter was washed with three changes of 0.05% (v/v) Tween 20 in TBS at room temperature over a 15-min period. It was next incubated with goat anti-rabbit IgG-Alkaline phosphatase conjugate in blocking buffer for 2 h at room temperature with mixing, and then washed as above. Positive bands were visualized after incubations, a t room temperature, with the substrate solution [16.5 mg of nitro blue tetrazolium/0.5 mL of 70% (v/v) aqueous DMF/8.5 mg of bromochloroindolyl phosphate in 1mL of HzO, which is added to 50 mL of 0.1 M Tris-HC1/0.1 M NaC1/0.005 M MgC12, buffer, pH 9.51. Assay for j3-Galactosidase. The procedure for measuring the activity of @-galactosidase (@-gal)was from Miller (1972). The frozen cell pellets were resuspended in 1.5 mL of demineralized water and then diluted in Z buffer so that a faint yellow color developed upon addition to o-nitrophenyl galactoside (ONPG). Chlqroform and 0.1"; SDS were added to the cells in Zbuffer, and this mixed and incubated for 5 min at 28 O C . ONPG was added, the mixture shaken and then incubated for 15 min, and the reaction was stopped by addition of Na2C03. Cell debris was pelleted by centrifugation and OD420 measured on the supernatants. Units of @-galactosidase = (1000) (OD420)/ [time of reaction (min)] [sample volume (mL)]. Evaluation of Wild-Type Virus Infection. Percentage infection with wild-type virus wag determined after incubating the cells for 96 h. Cells were suspended by shaking or scraping, and the percentage of cells containing polyhedra was determined with hemocytometer counts on four samples from each culture. PAGE/Western Blot. Secreted gp50T in cell culture supernatants was analysed by SDS-polyacrylamide gel electrophoresis (Laemmli, 1970) and Western blotting (Towbin et al., 1979). Samples (40 pL) of each supernatant were run on a 157" DATD-cross-linked polyacrylamide gel. Following electrophoresis, proteins were blotted onto 0.22-pm nitrocellulose (Schleicher and Schuell, BA83) by using a semidry electroblotter (Bjerrum and SchaeferNielsen, 1986). Unoccupied sites on the nitrocellulose were blocked by sequential 30-min incubations in 0.1 % Tween-

gp5OT production, re1 intensity of bands 0 0

+ +++ 0 ++ 0 ++ 0 0 0 0 0

+ + ++ + +++ +++ ++ +++ + +

% infection

4 0

17 43 0 78 70 78 0 0.8 0 0

0.2 65 95 92 91 4 83 90 91 73 19

20 in phosphate-buffered saline and 0.1 % Brij-58 in Trissaline (pH 7.4). The blots were incubated overnight a t 4 " C in the primary antibody solution, consisting of serum from a rabbit vaccinated with a recombinant vaccinia virus producing gp50 (Marchioli et al., 1987), diluted 1/200 in NETG buffer (150 mM NaCl, 5 mM EDTA, 48 mM TrisHC1, and 0.25 % gelatin, pH 7.4) containing 0.05 7% NP40. The blots were washed to remove excess antibody solution: two brief washes with a 1/100 dilution of wash solution (NETG, 0.5% Triton X-100, and 0.1% SDS), two 5-min washes with full-strength wash solution, and two additional brief washes with 1/100 wash solution. Alkaline phosphatase conjugated goat anti-rabbit antibody (Kirkegaard and Perry Labs) was diluted 1/500 in NETG-NP40 buffer, and the blots were incubated in this solution for 1 h a t room temperature. The blots were washed as above and incubated for 30 min in alkaline phosphatase substrate solution (veronal acetate buffer, pH 9.6, containing 0.1 mg/mL nitro blue tetrazolium chloride, 0.05 mg/mL of the p-toluidine salt of 5-bromo-4-chloro-3-indolyl phosphate, and 4 mM magnesium chloride). The color reaction was stopped by washing the blots in distilled water, and the blots were air dried. The molecular weights of the blotted proteins were determined by comparison to molecular weight markers (rainbow markers, Amersham) run on each gel. To compare the efficiency of Western blotting of proteins among several gels, a gp50T standard, produced by transfected Chinese hamster ovary cells (A. L. Meyer, unpublished work) was run on each gel.

Results HPg Production. IZD-MB0503 had the highest yield of all lines at 48 and 72 h postinfection (Table 11). The amount present at 48 h was significantly higher than any others at this time, which suggests that this line is somehow more efficient a t more rapid gene expression or secretion of the protein into the medium. Numerous other lines had higher concentrations than SF-9 at 48 and 72 h. These include IPLB-TN-R2,IPLB-SF-1254, IPLB-LdEIta, and CM-1. All lines (except TN-368) that yielded 1.4 pg of HPg/ mL or above at 72 h had relatively high levels of percent infection with wild-type virus. With IZD-MB0503, there is a direct comparative relationship between wild-type

Biotechnol. Prog., 1991, Vol. 7, No. 1

12

7

8

10

12

26

CHO 27 gp50T

68 -

92.5

46

-

30

-

21.5

-

Figure 1. Western blot of gp50T from various cell lines: 7 is IPLB-HvT1,8 is BCIRL-HZ-AM1,lO is IPBL-LdEIta, 12 is IZDMRO503,26 is SF-9,27 is IPLR-SF-1254,and CHO is a Chinese hamster ovarv cell line transfected to produce gp5OT.

virus replication and the levels of expression of two of the foreign genes (HPg and @-gal). Lynn and Hink (1980) evaluated AcNPV replication in IZD-MB0503, IPLBSF1254, RTT-EAA, and TN-368, and of the four lines, IZD-MBO503 produced nonoccluded virus more rapidly and developed more polyhedral occlusion bodies (POBs). Since the two recombinant proteins were expressed sooner (compare levels a t 48 h among the four lines) and also reached higher concentrations in IZD-MB0503, the recombinant and wild-type viruses must interact in similar ways with these four cell lines. Another way of looking a t the situation is that, for these four cell lines, the relative numbers of POBs produced are directly related to the levels of expression of two of the three foreign genes. For some lines, there appears to be no common direct correlation between percent cells infected with wild-type virus and yields of HPg, i.e., BCIRL-PX2-HNV3 had 95% infection and low (0.4 pg/mL) HPg while TN-368 had 1gC; infection and relatively high (1.4 pg/mL) HPg. @-GalactosidaseProduction. The three highest producers were SF-9, IZD-MBO503, and BCIRL-PX2-HNV3, with no significant differences between them. The levels of &gal increased between 48 and 72 h in all lines except UCR-SE-1. This line had the highest amount of 48 h but, possibly due to degradation, the level of 72 h was below the other three. The yield of @-galfrom SF-9 cells a t 48 h is similar to that reported by Broussard and Summers (1989) a t 50 h postinfection. gp5OT Production. Relative yields of gp50T, a t 72 h postinfection, from all lines were compared to that amount produced by SF-9 cells in serum-supplemented medium. The bands on the Western blots were visually compared and SF-9 was given a value of ++. Higher levels were scored as +++ and lower amounts as + (Table I). Several lines yielded more gp50T than SF-9; these were IPLBHvT1, IPLB-SF21AE, IPLB-SF21AE-15, and IPLB-SF1254. There is probably no more than a 2-fold difference in the amounts produced by the various lines. Three lines (IPLB-HvT1, IPLB-LdEIta, and IPLB-SF1254)produced gp50T forms that were approximately 1000 daltons larger than that from SF-9 cells (Figure 1). These

larger molecular weights are probably due to increased glycosylation. gp50T from BCIRL-HZ-AMI is smaller by about 10 000 daltons, probably due to proteolytic cleavage. Lines that yielded no gp50T also produced little or no @-gal,but one of these produced significant amounts of HPg (IPLB-LdEItf). Comparative Yields from Cells in Serum-Freeand Serum-SupplementedMedia. With SF-9 cells, there is no consistent pattern for relative yields of recombinant proteins in the two media (Table 111). The HPg yields were about 2 times higher in serum-free medium, while with @-galthe opposite occurred, i.e., better production in serum-supplemented medium. There were no differences in gp50T values and susceptibility of cells to wild-type virus. IPLB-SF-21AE cells produced more HPg in serumsupplemented medium, while higher levels of @-galwere present in serum-free medium. The gp50T recombinant was expressed a t the highest level in serum-supplemented medium with no expression in serum-free medium. The TN-368 cells yielded equal amounts of HPg and gp5OT in the two media. However, HPg was produced sooner postinfection in serum-free medium. As with IPLBSF-21AE cells, @-galwas more concentrated in serum-free medium. The BTI-EAA cells produced more HPg and gp50T in serum-free medium. Others have also observed that expression of recombinant proteins in SF-9 cells was influenced by media composition. Expression of human erythropoietin was elevated when infected cells were incubated in a medium with 4 % fetal bovine serum (FBS) and 20% of the usual amounts of lactalbumin hydrolysate and yeastolate used in routine culture of SF-9 cells (Quelle et al., 1989). Broussard and Summers (1989) obtained higher levels of @-galactosidase by incubating infected cells in a medium with 0.5% FBS after growing the cells in the presence of FBS a t the normal concentration of 10%. Our experimental plans and those of Maiorella et al. (1988) were different in that, instead of just transferring infected cells to reduced FBS, cells were grown for many passages and the infected cells incubated in ISFM in the total absence of FBS. Maiorella et al. (1988) obtained equal expression of recombinant human macrophage colony stimulating factor in ISFM and medium with 10% FBS. Our major conclusion is that four different insect cell lines retain the capacity to express foreign genes after prolonged culture in ISFM and, with most lines and recombinants, the yields in ISFM are better than or equal to those in serumsupplemented medium. For any given line, there is no significant difference in percent infection with wild-type virus in the two media. However, there are significant differences in yields of recombinant proteins, which suggests that medium composition may affect transcriptional or translational events more strongly in recombinant viruses than in wild-type viruses. If one assumes that percent infection with wild-type virus can be used as an indicator of the number of cells that might have been infected with the recombinants, then on the basis of product yield (gp50T and HPg) per infected cell, all three lines in Table I11 are better than SF-9. This is reasonable since recombinant viruses replicate like wildtype viruses in cultured cells (Maeda, 1989)and the kinetics of appearance of a recombinant protein were similar to those of polyhedrin in wild-type AcNPV infected cells (Barber et al., 1990). Calculation of potential gp50T and HPg yields for IPLB-SF-21AE and TN-368 cells by

Biotechnol. Prog., 1991, Vol. 7, No. 1

13

Table 111. Product Yields from Cells in Serum-Free and Serum-Supplemented Media gp50T, re1 intensity of bands 72 h

++

SF-9 cells in SSM” SF-9 cells in SFMb SF-21AE cells in SSM SF-21AE cells in SFM TN-368 cells in SSM TN-368 cells in SFM EAA cells in SSM EAA cells in SFM a

++ +++ 0 + + + ++

HPg, d m L 48 h 72 h 0.2 0.8 0.8 1.8 0.8 0.7 0.1 0.2 1.4 0.6 1.2

1.4

0.4

0.4 0.8

1.0

P-gal, units 48 h 72 h 59 534 84 667 2367 16 200 1034 4000 7267 13 500 2234 3867 20 900 40 650

% infection

90 89 4

5

19 21

17 20

SSM, serum-supplemented medium. b SFM, serum-free medium

increasing infection to 90 % (that of SF-9 cells) would give theoretical values much above those of SF-9 cells. Comparison of Yields from P a r e n t a n d Cloned Cell Strains. IPLB-SF-21AE is the parent line from which two cloned cell strains, used in this study, were derived [SF-9 (Summers and Smith, 1987) and IPLB-SF-21AE15 (J.Vaughn, USDA-Beltsville, personal communication, 1989)l. The parent produced significantly more HPg than either clone a t 48 h. However, a t 72 h there was no difference in HPg between parent and SF-9. The parent yielded more gp50T than SF-9, but SF-9 produced more P-gal than the parent. IPLB-SF-21AE-15was lowest for HPg, better than the parent for P-gal, and equal for gp50T. Another example is the parent line (UCR-SE-1)and a clone from it (UCR-SE-la) (B. Federici, University of California at Riverside, personal communication, 1989). 0-gal and gp50T were higher in the parent and yields of HPg were similar. From these results it appears that cloning is an approach that should be considered in efforts to maximize yields of recombinant proteins. However, it must be kept in mind that a clone that expresses a particular recombinant well may be inappropriate for another. Discussion Comparative yield data suggest that many of these cell lines should be considered as candidates for production of recombinant proteins. However, posttranslational modifications may be more important than yields when evaluating lines. Our data show that lines have different glycosylation properties and future work will probably reveal diversity in other modifications such as assembly into complexes, phosphorylation, and signal peptide cleavage. Before any of the lines that yield comparatively high levels of proteins can be considered for large-scale production, other concerns must be addressed. Growth rates and cell densities must be acceptable in suspension culture. Cells should grow well in serum-free medium, and gene expression should not be suppressed. There is little published information of this nature for many of the better lines. However, large-scale suspension culture of IPLBSF-21AE and TN-368 cells is discussed by Weiss et al. (1985) and Hink and Strauss (1980), respectively. A clone of the overall best producer line, IZD-MB0503, grew well in suspension in spin flasks of 1.5-2 L and a fermentor at a volume of 10 L (Miltenburger and David, 1980). The line could be cultured continuously in suspension by periodic harvesting and addition of fresh medium with maximum cell densities of 3 x lo6 cells/mL and a population doubling time of 17 h. Numerous factors such as medium nutrients, pH, osmotic pressure, 0 2 tension, cell density, stage of growth, and moi are known to influence the infection of cultured

cells with baculoviruses and the expression of foreign genes. In this study, for a given line, all these variables were constant and the only variable was the recombinant itself. The wide variations in levels of expression of foreign genes in specific cell lines are unpredictable, with some genes expressed at elevated levels in one line but low levels in others and vice versa. For example, IPLB-TN-R2 had comparatively high HPgand low gp50T, while IPLB-HVT1 had low HPg and high gp5OT. It is obvious that both the cell substrate and recombinant interact to influence gene expression when other factors are held constant. When all yield results are considered, there are two lines that produced relatively high levels of all three recombinant proteins. IZD-MB0503 produced the most HPg with P-gal and gp50T levels equal to SF-9 cells. IPLBSF-1254 was third highest for HPg and P-gal and better than SF-9 for gp50T.

Literature Cited Barber, G. N.; Clegg, J. C. S.; Lloyd, G. Expression of the Lassa virus nucleocapsid protein in insect cells infected with a recombinant baculovirus: application to diagnostic assays for Lassa virus infection. J. Gen. Virol. 1990, 71, 19-28. Bjerrum, 0. J.; Schafer-Nielsen, C. In Electrophoresis ’86: Proceedings of t h e 5 t h Meeting of the International Electrophoresis Society; Dunn, M. J., Ed.; VCH Verlagsgesgeltschaft: Weinheim, Germany, 1986; pp 315-327. Broussard, D. R.; Summers, M. D. Effects of serum concentration and media composition on the level of polyhedrin and foreign gene expression by baculovirus vectors. J . Inuertebr. Pathol. 1989, 54, 144-150. Burnette, W. H. “Western Blotting”: Electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 1981, 112, 195-203. Chen, Q.;McIntosh, A. H.; Ignoffo, C. M. Establishment of a new cell line from the pupae of the diamond-back moth, Plutella xylostella (Lepidoptera: Plutellidae). J . Cent. China Teach. C ~ l l 1983, . 27, 99-107. Davidson, D. J.; Higgins, D. L.; Castellino, F. J. Plasminogen activator activites of equimolar complexes of streptokinase with variant recombinant plasminogens. Biochemistry 1990, 29, 3585-3590. Eide, P. E.; Caldwell, J. M.; Marks, E. P. Establishment of two cell lines from embryonic tissue of the tobacco hornworm, Manduca sexta (L.). I n Vitro 1975, 11, 395-399. 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 xigua (Lepid0ptera:Noctuidae) that supports replication of nuclear polyhedrosis viruses from Spodoptera exigua and Autographa californica. J . Inuertebr. Pathol. 1986, 48, 199-207.

14

Granados, R. R.; Naughton, M. In Inuertebrate Tissue Culture, Applications i n Medicine, Biology, and Agriculture; Kurstak, E., Maramorosch, K., Eds.; Academic: New York, 1976; pp 379-389. Granados, R. R.; Derksen, A. C. G.; Dwyer, K. G. Replication of the Trichoplusia n i granulosis and nuclear polyhedrosis viruses in cell cultures. Virology 1986, 152, 472-476. Hink, W. F. Established insect cell line from the cabbage looper, Trichoplusia ni. Nature 1970, 266, 466-467. Hink, W. F.; Strauss, E. M. In Invertebrate Systems I n Vitro; Kurstak, E., Maramorosch, K., Dubendorfer, A., Eds.; Elsevier/ North-Holland Biomedical: New York, 1980; pp 27-33. Hink, W. F.; Strauss, E. M.; Ramoska, W. A. Propagation of Autographa californica nuclear polyhedrosis virus in cell culture: methods for infectingcells. J . Inuertebr. Pathol. 1977, 30, 185-191. Iatrou, K.; Meidinger, R. G. B o m b y x mori nuclear polyhedrosis virus-based vectors for expressing passenger genes in silkmoth cells under viral or cellular promoter control. Gene 1989, 75, 59-71. Inlow, D.; Shauger, A.; Maiorella, B. Insect cell culture and baculovirus propagation in protein-free medium. J . Tissue Cult. Methods 1989, 12, 13-16. Inoue, H.; Mitsuhashi, J. Further establishment of continuous cell lines from larval fat bodies of the cabbage armyworm, Mamestra brassicae (Lepid0ptera:Noctuidae). A p p l . E n t o mol. Zoo(. 1985, 20, 496-498. Laemmli, L.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970,227, 680-686. Luckow, V. A. In Recombinant D N A Technology and A p p l i cations; Ho, C. Prokop, A., Bajpai, R., Eds.; McGraw Hill: New York, 1990; in press. Luckow, V. A.; Summers, M. D. Trends in the development of baculovirus expression vectors. Biotechnology 1988, 6, 4755. Lynn, D. E.; Hink, W. F. Comparison of nuclear polyhedrosis virus replication in five lepidopteran cell lines. J . Inuertebr. Pathol. 1980, 35, 234-240. Lynn, D. E.; Dougherty, E. M.; McClintock, J. T.; Loeb, M. In Invertebrate and Fish Tissue Culture; Kuroda, Y., Kurstak, E., Maramorosch, K., Eds.; Springer Verlag: New York, 1988; pp 239-242. Maeda, S. Expression of foreign genes in insects using baculovirus vectors. A n n u . Reu. Entomol. 1989, 34, 451-472. Maeda, S.; Kawai, T.; Obinata, M.; Fujiwara, H.; Horiuchi, T.; Saeki, Y.; Sato, Y.; Furusawa, M. Production of human tu-interferon in silkworm using a baculovirus vector. Nature 1985, 315, 592-594. Maiorella, B.; Inlow, D.; Shauger, A.; Harano, D. Large-scale insect cell-culture for recombinant protein production. Biotechnology 1988, 6 , 1406-1410. Marchioli,C. C.;Yancey,R.J.;Petrovskis,E.A.;Timmins, J. G.; Post, L. E. Evaluation of pseudorabies virus glycoprotein gp50 as a vaccine for Aujesky’s disease in mice and swine: Expression by vaccinia virus and Chinese hamster ovary cells. J . Virol. 1987, 61, 3977-3982. McIntosh, A. H.; Ignoffo, C. M. Replication and infectivity of the single-embedded nuclear polyhedrosis virus, Baculouirus h e liothis, in homologous cell lines. J . Inuertebr. Pathol. 1981, 37, 258-264. McIntosh, A. H.; Maramorosch, K.; Rechtoris, C. Adaptation of an insectcellline (Agallia constricta) inmammaliancellculture medium. I n Vitro 1973, 8 , 375-378.

Biotechnol. Prog., 1991, Vol. 7, No. 1

McLean, J. W.; Tomlinson, J. E.; Kuang, W.; Eaton, D. L.; Chen, E. Y .;Gless, G. M.; Scanu, A. M.; Lawn, R. M. c-DNA sequence of human apolipoprotein (a) is homologous to plasminogen. Nature 1987, 330, 132-137. Miller, J. H. Experiments i n Molecular Genetics; Cold Spring Harbor Laboratory: Cold Spring Harbor, NY, 1972; pp 352355. Miltenburger, H. G.; David, P. Mass production of insect cells in suspension. Deu. B i d . Stand. 1980, 46, 183-186. Miltenburger, H. G.; David, P.; Wulf, H. In Proceedings of the 1st International Colloquium of Inuertebrate Pathology; Queen’s University: Kingston, Ontario, 1976; pp 393-394. Mitsuhashi, J.;Maramorosch, K. Leafhopper tissue culture: embryonic, nymphal, and imaginal tissues from aspectic insects. Contrib. Boyce T h o m p s o n Inst. 1964, 22, 435-460. Mitsuhashi, J.; Shozawa, A. Continuous cell lines from larval hemocytes of the cabbage armyworm, Mamestra brassicae. Deu. Growth Differ. 1985, 27, 599-606. Petrovskis, E. A.; Timmins, J. G.; Armentrout, M. A.; Marchioli, C. C.; Yancey, R. J.; Post, J. E. DNA sequence of the gene for the pseudorabies virus gp50, a glycoprotein without N-linked glycosylation. J . Virol. 1986, 59, 216-223. Quelle, F. W.; Caslake, L. F.; Burkert, R. E.; Wojchowski, D. M. High-level expression and purification of a recombinant human erythropoietin produced using a baculovirus vector. Blood 1989, 74, 652-657. Rochford, R.; Dougherty, E. M.; Lynn, D. E. Establishment of a cell line from embryos of the cabbage looper, Trichoplusia ni (Hubner). I n Vitro 1984, 20, 823-825. Seiburth, P. J.; Maruniak, J. E. Growth characteristics of a continuous cell line from the velvetbean caterpillar, Anticarsiagemmatalis Hubner (Lepid0ptera:Noctuidae). I n Vitro Cell. Deu. Biol. 1988, 24, 195-198. Summers, M. D.; Smith, G. A. A manual of methods for baculovirus vectors and insect cell culture procedures. Ten. Agric. E x p . S t n . Bull. 1987, B1555, 1-56. Thomsen, D. R.; Post, L. E.; Elhammer, A. P. Structure of O-glycosidically linked oligosaccaride synthesized by the insect cell line SF-9. J . Cell. Biochem. 1990, 43, 67-79. Towbin, H.; Staehelin, T.; Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and applications. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 4350-4354. Vaughn, J. L.; Goodwin, R. H.; Tompkins, G. J.; McCawley, P. The establishment of two cell lines from the insect Spodoptera frugiperda (Lepid0ptera:Noctuidae). I n Vitro 1977,13,213217. Weiss, S. A.; Smith, G. C.; Kalter, S. S.; Vaughn, J. L. Improved method for the production of insect cell cultures in large volume. I n Vitro 1981, 17, 495-502. Weiss, S. A,; Peplow, D.; Smith, G. C.; Vaughn, J. L.; Dougherty, E. In Techniques i n the L i f e Sciences, Cell Biology Vol. C1; Kurstak, E., Ed.; Elsevier Scientific: New York; 1985; Vol. C1, pp C110/1-C110/16. Whitefleet-Smith, J.; Rosen, E.; McLinden, J.; Ploplis, V. A.; Fraser, M. J.; Tomlinson, J. E.; McLean, J. W.; Castellino, F. J. Expression of human plasminogen cDNA in a baculovirus vector-infected insect cell system. Arch. Biochem. Biophys. 1989,271, 390-399. Accepted October 31, 1990. Registry No. Pg, 9001-91-6; @-gal,9031-11-2.