Optimal Conditions for the Expression of Glycoprotein E2 of Classical Swine Fever Virus using Baculovirus in Insect Cells
Optimal Conditions for the Expression of Glycoprotein E2 of Classical Swine Fever Virus using Baculovirus in Insect Cells
International Journal of Industrial Entomology. 2014. Dec, 29(2): 207-213
Copyright © 2014, Korean Society of Sericultural Science
  • Received : December 26, 2014
  • Accepted : December 29, 2014
  • Published : December 31, 2014
Export by style
Cited by
About the Authors
Sung Min, Bae
Seung Hee, Lee
Won Suk, Kwak
Yong Oh, Ahn
Tae Young, Shin
Soo Dong, Woo

The structural proteins of classical swine fever virus (CSFV) consist of nucleocapsid protein C and envelope glycoprotein E rns (E0), E1 and E2. Among them, E2, the most immunogenic of the CSFV glycoproteins, induces a protective immune response in swine. In this study, to determine the optimal expression conditions of glycoprotein E2 using baculovirus system, we investigated the influence of insect cells and media to the expression of recombinant E2. Recombinant virus containing glycoprotein E2 coding gene was constructed with bApGOZA DNA. Expression of the glycoprotein E2 was analyzed by SDS-PAGE and Western blot analysis using anti-CSFV E2 monoclonal antibodies. Expression of glycoprotein E2 in Sf21 cells was first observed after 3 days and reached a maximum on the 5th day after infection. Furthermore, the highest levels of glycoprotein E2 expression were observed at multiplicity of infection (MOI) of 5. When three different insect cell lines (Sf21, High-Five and Se301) were tested, High-Five cells showed the highest production. In addition, four different serum-free and serum-supplemented media, respectively, were tested for the expression of glycoprotein E2 and the budded virus (BV) titers. As a result, serum-supplemented medium provided the best conditions for protein production and the BV yield.
The baculovirus expression vector system (BEVS) is an effective and very popular used method for the production of recombinant proteins in insect cells. The baculovirus most often used as an expression vector is Autographa californica nucleopolyhedrovirus (AcNPV). Historically, the most widely used hosts for AcNPV were the established insect cell lines IPLB-Sf21-AE (Sf21) and Sf9, originally derived from Spodoptera frugiperda ovaries ( Vaughn ., 1977 ), and BTI-TN-5B1-4 (High-five), an insect cell line derived from Trichoplusia ni eggs, provided higher levels of foreign protein production ( Wickham ., 1992 ). However, levels of several recombinant protein expressions under the polyhedrin promoters are substantial, but still do not approach the levels of polyhedrin protein expression. A basic aspect of the BEVS which has been largely ignored is the importance of the baculovirus-cell culture systems in recombinant protein expression. Thus, protein production by baculovirus is affected by several factors, including the cell line ( Wickham ., 1992 ; McIntosh and Grasela, 1994 ; Hink ., 1991 ), cell density ( Reuveny ., 1993 ), and medium replacement at the time of infection ( Linday and Betenbaugh, 1992 ; Lazarte ., 1992 ) or replacement and supplementation of medium with additional additives including serum ( Reuveny ., 1993 ; Bedard ., 1994 ), the passage number of virus ( Wickham ., 1991 ), and multiplicity of infection (MOI) ( Licari and Bailey, 1991 ). In addition to the above factors, the composition of the medium also greatly impacts culture productivity ( Jesionowski ., 1997 ).
Classical swine fever virus (CSFV), featuring high fever and hemorrhagic lesions, is a highly contagious disease threatening the pig industry of the world ( Terpstra, 1991 ). CSFV is a member of the Pestivirus genus of the family of Flaviviridae ( Francki ., 1991 ). The genome is a single strand RNA of positive sense, approximately 12,300 nucleotides in length. It has a non-translated region at either end (5’NTR and 3’NTR), encompassing a single open reading frame encoding a large protein that is cleaved into smaller fragments ( Paton , 2000 ). Structural components of the CSFV virion include the capsid (C) protein and glycoproteins E rns (E0), E1, and E2 ( Thiel ., 1991 ). Glycoprotein E2 among them is the most immunogenic glycoprotein, inducing neutralizing antibodies and protection against lethal CSFV challenge (Risatti et al ., 2007). Therefore, vaccines based on glycoprotein E2 protect swine from CSF and induce high titers of neutralizing Abs ( van Rijn , 1999 ). Recently, cDNA sequences encoding different versions of glycoprotein E2 of CSFV have been inserted into the gX locus of pseudorabies virus (PRV) ( van Ziji ., 1991 ). The recombinant viruses contain the sequence of glycoprotein E2 without and with a C-terminal transmembrane region (TMR). This subunit vaccine suggested that pigs immunized with glycoprotein with TMR develop high levels of neutralizing antibodies against CSFV and are protected against challenge with a lethal dose of CSFV, whereas pig inoculated with glycoprotein E2 without TMR develop low levels of neutralizing antibodies and are partly protected upon challenge ( van Ziji ., 1991 ). However, subunit vaccines are capable of eliciting a protective immune response only when a large amount of antigen is applied.
In this study, with the aim of determining an effective cell line, p.i. time, MOI and cell culture media using BEVS, we analyzed the optimal expression conditions of recombinant glycoprotein E2. We constructed a recombinant AcNPV which contains a CSFV glycoprotein E2 with TMR gene. The glycoprotein E2 production in cells adapted to serum-free media and media with serum also was evaluated in three insect cell lines.
Materials and Methods
- Insect cell lines and media
The basal insect cell line used was a strain of S. frugiperda cell line (Sf21) in SF900II (Gibco, USA) serum-free medium. Other cell lines were a strain of S. exigua cell line (Se301) in IPL-41 medium (Welgene, Republic of Korea) with 5% fetal bovine serum (FBS) and T. ni cell line (High-Five) was maintained in Express Five serum-free medium (Gibco, USA) with L-glutamine. All cells were maintained at 27˚C in T-flasks in a non-humidified incubator.
The insect media used were serum-free and serumsupplemented media. The serum-free media were SF900II (Gibco, USA), SF900III (Gibco, USA), Insect-Xpress (BioWhittaker, USA) and MAX-XP (BD, USA). The serum-supplemented media were TC-100 (Welgene, Republic of Korea), IPL-41 (Welgene, Republic of Korea), Grace’s insect medium (JBI, Republic of Korea) and Grace’s insect medium supplemented (Gibco, USA). 5% of FBS was added.
- Cloning and construction of recombinant virus
Viral RNA was extracted from CSFV-infected cell cultures with Viral Gene-spin TM kit (iNtRON Biotechnology, Republic of Korea) as recommended by the manufacturer. The viral RNA was used as a template for cDNA synthesis using RNA LA PCR Kit (TaKaRa, Japan). The glycoprotein E2 gene was amplified from cDNA with primers CSFV E2-F (5’-GAATTCATAATGCGGCTAGCCTGCAAGGAAGATT-3’) and CSFV-E2-R (5’-CTGCAGACCAGCGGCGAGTTGTTCTGTTAGAA-3’). The purified PCR products were cloned into the T&A cloning vector (RBC Bioscience, Taiwan), and the plasmids were named as pTA-E2 ( Fig. 1 ). The plasmids were digested with EcoR I (under lined) and Pst I (under lined), and then cloned into the transfer vector pBacPAK9 (Clontech, USA), to get the recombinant transfer plasmid pBacPAK9-E2. Sf21 cells were co-transfected with a mixture of purified transfer vector, bApGOZA DNA ( Je ., 2001 ), and Cellfectin II (Invitrogen, USA), according to the manufacturer’s instruction. Five days after adding the Cellfectin II-DNA complexes to the cells, the medium containing viruses released by the transfected cells was transferred to a sterile container and stored at 4˚C. A standard plaque assay procedure ( O’Reilly ., 1992 ) was used to obtain viral plaques from dilutions of the media harvested from the co-transfections. Individual plaques were examined under a microscope and scored for polyhedra production.
PPT Slide
Lager Image
Schematic representation of the recombinant virus generation.

The strategy for the construction of recombinant baculovirus is schematically demonstrated. Nucleotide sequencing analysis confirmed that the inserted fragment contained the entire coding sequence for viral glycoprotein E2. After PCR cloning, the plasmid was digested with EcoRI and PstI and ligated into the transfer vector pBacPAK9. The recombinant transfer vector was co-transfected into Sf21 cells along with bApGOZA DNA (). Recombinant virus was purified by plaque assay and amplified in Sf21 cells.

- Recombinant virus titration
Recombinant virus titer was determined by an end-point dilution assay. In this method, 96-well plates containing 100 μL of cellular suspension at a density of 2×10 5 cells/mL were incubated for 1 h at 27˚C. Dilutions of the virus samples were performed serially in SF900II medium, from 1:10 3 to 1:10 8 , and 100 μL were used to infect Sf21 cells monolayer after removing the medium from plate wells. Eight replicates for each dilution were performed in the same plate, and two independent plates were infected for each viral sample. SF900II medium was used as negative control. Plates were screened after seven days for occlusion bodies under a microscope. The 50% tissue-culture infectious dose (TCID 50 ), i.e., the dilution which is sufficient to infect half of the cells, was calculated using equations described elsewhere ( King and Possee, 1992 ). This value was then converted to plaque-forming units (or number of infectious particles) by the relationship: PFU=0.69×TCID 50 ( King and Possee, 1992 ).
- SDS-PAGE and Western blot analysis
Sf21 cells were infected with the wild-type AcNPV and recombinant virus in a 25 cm 2 flask (1×10 6 cells) at various MOIs. After incubation at 27˚C, cells were harvested at every day for 5 days. For SDS-PAGE ( Laemmli, 1970 ) of cell lysates, wild-type AcNPV and recombinant virus-infected cells were washed twice with phosphate buffered saline (PBS). The lysate was prepared by incubating the cells with lysis buffer (20 mM Tris-HCl pH 7.5, 50 mM NaCl, 5% Glycerol, 0.1% Triton X-100) containing protein inhibitor cocktail (Sigma-Aldrich, USA) for 30 min on ice followed by sonication and mixed with protein sample buffer and boiled. The total cellular lysates were subjected to 12% SDS-PAGE. For Coomassie stains, gels were washed with deionized water and stained with BioSafe Coomassie. For Western blot analysis, 12% SDS-PAGE was performed as described above. Proteins of cellular lysates were blotted on-to a nitrocellulose membrane (Pall Corp., USA). After transfer, the membrane was incubated for 60 min with gentle agitation in 5% (w/v) non-fat dry milk in TBST buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.05% Tween 20). The blot was incubated with anti-CSFV E2 monoclonal antibody (1:640) (JENO Biotech, Republic of Korea) in TBST for 1 h and washed. Subsequently, the membrane was incubated with anti-mouse horseradish peroxidase conjugate for 30 min at room temperature. After repeated washing, the immune reactive bands were visualized with the ECL Western Blotting Detection System (Elpis Biotech, Republic of Korea).
Results and Discussion
- Glycoprotein E2 production in Sf21 cells
First, for the comparison of glycoprotein E2 productivity at different MOIs with Sf21 cells, 1×10 6 cells/ml were infected with the recombinant baculovirus, rB9CSFV-E2, at a MOI of 1, 5 or 10. The effect of MOI on infectivity and yield is as important as the condition of cells and the time of infection ( Kim , 2007 ). Expression of the glycoprotein E2 was analyzed by SDS-PAGE and Western blot with the specific CSFV-E2 monoclonal antibody. The highest amount of recombinant protein was expressed at a MOI of 5, with no further increase at a MOI of 10 ( Fig. 2A ). It had an apparent molecular weight around 35 kDa. To compare the productivity of glycoprotein E2 on the time of infection, Sf21 cells were harvested at every day for five days. The glycoprotein E2 appeared to start about three days post-infection, and reached a peak at 5 days. Presumably, the glycoprotein E2 seems to be accumulated during the expression time ( Fig. 2B ).
PPT Slide
Lager Image
The glycoprotein E2 expression in Sf21 cells by time course and MOIs .

The cells were infected with recombinant virus for various times (2~5 days) (panel A) and with various MOIs of recombinant virus (panel B). Proteins were separated on a 12% SDS-PAGE, transferred to nitrocellulose membranes for Western blot analysis and reacted with anti-CSFV E2 monoclonal antibody.

- Glycoprotein E2 production using different insect cell lines
The design of this study was to optimize the baculovirus mediated expression of the glycoprotein E2 as different levels of protein expression have been reported depending of the insect cell line used for infection ( Davis ., 1993 ). To compare the productivity of the glycoprotein E2 in three different insect cell lines, the protein expression of these cells were examined. Cells were infected with virus at MOI of 5 and harvested at five days post-infection. The highest amount of glycoprotein E2 was produced in High-Five cells, followed by Sf21 cells ( Fig.3 ). However, the expression of the glycoprotein E2 was not detected in SDS-PAGE analysis (data not shown). Previous studies on the production of recombinant protein in various insect cell lines demonstrated cells derived from embryonic tissues of cabbage looper, T. ni (High-Five), were often a better choice for production of recombinant protein than Sf21 cells ( Davis ., 1993 ). These data corroborate previous studies which have shown that there are major differences between cell lines in their ability to produce recombinant protein in the BEVS ( Hink , 1991 ; Wickham ., 1992 )
PPT Slide
Lager Image
Expression of glycoprotein E2 in Sf21, High-five and Se301 cells.

Four days post-infected insect cells were harvested, washed twice with PBS, and lysed in lysis buffer. Proteins were separated on a 12% SDS-PAGE, transferred to nitrocellulose membranes for Western blot analysis and reacted with anti-CSFV E2 monoclonal antibody.

- Optimization of glycoprotein E2 production using different culture media
The productivity of the glycoprotein E2 in different culture media (four serum-free and four serum-supplemented media) was compared. The basal culture medium of Sf21 cells were a SF900II serum-free medium and different culture media were adapted before the experiment. However, Sf21 cells were not grown to IPL-41 and Grace’s insect medium; so, these used after direct transfer from basal media without adaptation. Eight different culture media-adapted cells were infected with virus at MOI of 5 and harvested at five days post-infection. As a result, Insect-Xpress insect medium in serum-free medium ( Fig. 4A ) and TC-100 insect medium in serum-supplement medium ( Fig. 4B ) were the best conditions for the expression of the glycoprotein E2. These results suggest that expression of glycoprotein E2 in Sf21 cells was influenced by media composition. Insect cell culture media contain sugars, salts, amino acids, vitamins, organic acids, lipids, and trace elements. The influence of various nutrients may be beneficial for increasing recombinant protein yields ( Ikonomou ., 2003 ). While the mechanism by which it promotes recombinant protein production remains unknown. This conclusion was further supported by the experiments where influence of nutrients on productivity in culture media. In addition, the glycoprotein E2 was expressed at the highest level in serum-supplemented media than serumfree media. These results indicate that FBS promote recombinant protein production in Sf21 cells ( Nishikawa ., 2003 ; Hink ., 1991 ). Consequently, the recombinant protein expression was higher because a greater proportion of the culture media.
PPT Slide
Lager Image
Expression of glycoprotein E2 in various culture media.

Panel A, serum-supplement media; lane 1, GIBCO Grace’s insect medium supplemented with 5% FBS; lane 2, WelGENE Inc. TC-100 with 5% FBS; lane 3, JBI Grace insect medium with 5%; lane 4, WelGENE Inc. IPL-41 with 5% FBS; panel B, serum-free media; lane 5, GIBCO SF900-II; lane 6, GIBCO SF900-III; lane 7, BioWhittaker Insect-Xpress; lane 8, BD BaculoGold MAX-XP.

- The productivity of budded virus in different culture media
The recombinant virus-infected Sf21 cells in eight different culture media are presented as budded virus (BV) titers in Table 1 . Both initial infection rate and BV titers are correlated with the production of baculovirus-based recombinant proteins. To elucidate the relation between the virus and the recombinant protein productivities, BV productivity in various cell culture media was evaluated. The initial infection rate of virus in eight different insect cell culture media was determined at 1 hour post-infection. As can be seen from the results presented in Table 1 , the initial infection rate of virus did not show any significant difference in serum-free or supplement media. At five days postinfection, however, as determined by titration on Sf21 cells in SF900II media, 750 times more budded virus was produced in Grace’s insect media supplemented with 5% FBS, suggesting that serum-supplement media accelerate a higher rate of BV production (9.2×10 6 pfu/mL and 6.9×10 9 pfu/mL, respectively). Both BV yield and recombinant protein production were apparently increased in serum-supplement media. This was corresponding to the previous results. Insect cells have been widely cultured in basal media supplemented with vertebrate serum, in particular with a concentration of 5% to 20% of FBS, which has been shown to support cell growth, high BV titers, and recombinant protein production ( Wu ., 1989 ; Matindoost ., 2014) . Most BV productivity had direct proportion with the recombinant protein production ( Table 1 , Fig. 4 ). However, the recombinant protein production was influenced not only the BV productivity but also the kind of used media.
Baculovirus productivity in Sf21 cells under different culture media conditions
PPT Slide
Lager Image
Baculovirus productivity in Sf21 cells under different culture media conditions
In conclusion, our study suggests that the High-Five cell line and serum-supplement TC-100 insect medium are optimal for the production of the CSFV glycoprotein E2 in Sf21 cells. These results would provide useful information on the determination of insect cells and insect media in production of recombinant proteins using the BEVS.
Following are results of a study on the “Leaders INdustry-university Cooperation” Project, supported by the Ministry of Education.
Ayres MD , Howard SC , Kuzio J , Lopez-Ferber M , Possee RD 1994 The complete DNA sequence of Autographa californica nuclear polyhedrosis virus Virology 202 (2) 586 - 605    DOI : 10.1006/viro.1994.1380
Bedard C , Kamen A , Tom R , Massie B 1994 Maximization of recombinant protein yield in the insect cell/baculovirus system by one-time addition of nutrients to high-density batch cultures Cytotechnology 15 129 - 138    DOI : 10.1007/BF00762387
Davis TR , Wickham TJ , McKenna KA , Granados RR , Shuler ML , Wood HA 1993 Comparative recombinant protein production of eight insect cell lines In Vitro Cell. Dev. Biol. 29A 388 - 390
Francki RIB , Fauquet CM , Knudson DL , Brown F 1991 Fifth report of the International Committee on Taxonomy of Viruses Springer Verlag Vienna 223 - 233
Hink WF , Thomsen DR , Davidson DJ , Meyer AL , Castellino FJ 1991 Expression of three recombinant proteins using baculovirus vectors in 23 insect cell lines Biotechnol. Prog. 7 (1) 9 - 14    DOI : 10.1021/bp00007a002
Hulst MM , Westra DF , Wensvoort G , Moormann RJM 1993 Glycoprotein E1 of hog cholera virus expressed in insect cells protects swine from hog cholera J. Virol. 67 5435 - 5442
Ikonomou L , Schneider YJ , Agathos SN 2003 Insect cell culture for industrial production of recombinant proteins Appl. Microbiol. Biotecnol. 62 1 - 20    DOI : 10.1007/s00253-003-1223-9
Je YH , Chang JH , Roh JY , Jin BR 2001 Generation of Baculovirus Expression Vector Using Defective Autographa californica Nuclear Polyhedrosis Virus Genome Maintained in Escherichia coli for Occ+ Virus Production Int. J. Entomol. 2 155 - 160
Jesionowski GA , Ataai MM 1997 An Efficient Medium for High Protein Production in the Insect Cell/Baculovirus Expression System Biotechnol. Prog. 13 355 - 360    DOI : 10.1021/bp970054l
Kim JS , Choi JY , Roh JY , Lee HY , Jang SS , Je YH 2007 Production of recombinant polyhedral containing cry1Ac fusion protein in insect cell lines J. Microbiol. Biotechnol. 17 (5) 739 - 744
King LA , Possee RD 1992 The Baculovirus Expression System. A Laboratory Guide Chapman & Hall London
Lazarte JE , Tosi P , Nicolau C 1992 Optimization of the production of full-length rCD4 in baculovirus-infected Sf9 cells Biotechnol. Bioeng. 40 214 - 217    DOI : 10.1002/bit.260400204
Licari P , Bailey JE 1991 Factors influencing recombinant protein yields in an insect cell-baculovirus expression system: multiplicity of infection and intracellular protein degradation Biotechnol. Bioeng. 37 238 - 246    DOI : 10.1002/bit.260370306
Laemmli UK 1970 Cleavage of structural proteins during assembly of the head of bacteriophage T4 Nature 227 680 - 685    DOI : 10.1038/227680a0
Lindsay DA , Betenbaugh MJ 1992 Quantification of cell culture factors affecting recombinant protein yields in baculovirus infected insect cells Biotechnol. Bioeng. 39 614 - 618    DOI : 10.1002/bit.260390605
Matindoost L , Hu H , Chan LC , Nielsen LK , Reid S 2014 The effect of cell line, phylogenetics and medium on baculovirus budded virus yield and quality Archives of virology 159 (1) 91 - 102    DOI : 10.1007/s00705-013-1789-1
McIntosh AH , Grasela JJ 1994 Expression of β-galactosidase and luciferase in insect cell lines infected with a recombinant AcMNPV In Vitro Cell. Dev. Biol. 30A 275 - 278
Nishikawa N , Yamaji H , Fukuda H 2003 Recombinant protein production by the baculovirus-insect cell system in basal media without serum supplementation Cytotechnology 43 3 - 10    DOI : 10.1023/B:CYTO.0000039894.27256.0f
O’Reilly DR , Miller LK , Luckow VA 1992 Baculovirus Expression Vectors - A Laboratory Manual W. H. Freeman and Company New York
Paton DJ , McGoldrick A , Greiser-Wilke I , Parchariyanon S , Song JY , Liou PP , Stadejek T , Lowings JP , Bjorklund H , Belak S 2000 Genetic typing of classical swine fever virus Vet. Microbiol. 73 137 - 157    DOI : 10.1016/S0378-1135(00)00141-3
Radford K , Cavegn C , Bertrand M , Bernard A , Reid S , Greenfield P 1997 The indirect effects of multiplicity of infection on baculovirus expressed proteins in insect cells: secreted and nonsecreted products Cytotechnology 24 73 - 81    DOI : 10.1023/A:1007962903435
Reuveny S , Kim YJ , Kemp CW , Shiloach J 1993 Effects of temperature and oxygen on cell growth and recombinant protein production in insect cell cultures Appl. Microbiol. Biotechnol. 38 619 - 623
Reuveny S , Kim YJ , Kemp CW , Shiloach J 1993 Production of recombinant proteins in high-density insect cell cultures Biotechnol. Bioeng. 42 235 - 239    DOI : 10.1002/bit.260420211
Terpstra C 1991 Special review series: Hog cholera: an update of present knowledge Br Vet J. 147 397 - 406    DOI : 10.1016/0007-1935(91)90081-W
Thiel HJ , Stark R , Weiland E , Rumenapf T , Meyers G 1991 Hog cholera virus: Molecular composition of virions from a pestivirus J. Virol. 65 4705 - 4712
van Rijn PA , van Gennip HGP , Moormann RJM 1999 An experimental marker vaccine and accompanying serological diagnostic test both based on envelope glycoprotein E2 of classical swine fever virus (CSFV) Vaccine 17 433 - 440    DOI : 10.1016/S0264-410X(98)00215-1
van Zijl M , Wensvoort G , De Kluyver E , Hulst M , Van der Gulden H , Gielkens A , Gielkens A , Berns A , Moormann R 1991 Live attenuated pseudorabies virus expressing envelope glycoprotein E1 of hog cholera virus protects swine against both pseudorabies and hog cholera J. Virol. 65 (5) 2761 - 2765
Vaughn JL , Goodwin RH , Thompkins GJ , McCawley P 1977 The establishment of two insect cell lines from the insect S frugiperda (Lepidoptera: Noctuidae) In Vitro 13 (4) 213 - 217    DOI : 10.1007/BF02615077
Vlak J , Keus R , Mizrahi A. 1990 Viral vaccines Wiley-Liss New York Baculovirus expression vector system for production of viral vaccines 91 - 128
Wickham TJ , Davis T , Granados RR , Hammer DA , Shuler ML , Wood HA 1991 Baculovirus defective interfering particles are responsible for variations in recombinant protein production as a function of multiplicity of infection Biotechnol. Lett. 13 483 - 488    DOI : 10.1007/BF01049204
Wickham TJ , Davis T , Granados RR , Shuler ML , Wood HA 1992 Screening of insect cell lines for the production of recombinant proteins and infectious virus in the baculovirus expression system Biotechnol. Prog. 8 (5) 391 - 396    DOI : 10.1021/bp00017a003
Wong TK , Peter CH , Greenfield PF , Reid S , Nielsen LK 1996 Low multiplicity infection of insect cells with a recombinant baculovirus: the cell yield concept Biotechnol. Bioeng. 49 659 - 666
Wu J , King G , Daugulis AJ , Faulkner P , Bone DH , Goosen MFA 1989 Engineering aspects of insect cell suspension culture: a review Appl. Microbiol. Biotecnol. 32 (3) 249 - 255