WO2008146019A2

WO2008146019A2 - Method for baculovirus generation from insect cells in high-troughput plates

Info

Publication number: WO2008146019A2
Application number: PCT/GB2008/001867
Authority: WO
Inventors: Kari Juhani Airenne; Seppo Ylä-Herttuala; Hanna-Riikka KÄRKKÄINEN; Olli Laitinen
Original assignee: Ark Therapeutics Ltd
Current assignee: Ark Therapeutics Ltd
Priority date: 2007-06-01
Filing date: 2008-06-02
Publication date: 2008-12-04
Anticipated expiration: 2009-12-01
Also published as: WO2008146019A3; GB0710535D0; WO2008146019A8

Abstract

The present invention is a method of baculovirus generation from insect cells, using a high-throughput plate, wherein the insect cells are transfected or infected by a virus construct, in suspension, and wherein the plate comprises wells which have a substantially square cross-section.

Description

BACULOVIRUS GENERATION Field of the Invention

This invention relates to baculovirus generation. Background of the Invention Various on-going and finished genome projects are sources of novel open reading frames, which are likely to provide protein drug candidates or targets for medical use. The vast amount of data requires efficient high-throughput methods to turn the genetic code into functional information. Thus, the need for a simple and flexible protein production technology with a high diversity and capacity exists. Baculoviruses are potential tools for this purpose since during the last 20 years, thousands of proteins from different species have been produced in insect cells using baculovirus expression system. The versatility of baculoviruses has become even more apparent from their use as successful gene transfer vehicles in multiple vertebrate cell lines and in vivo. In addition, recent developments in the baculovirus production methods have rendered them more suitable for high-throughput screening purposes (Airenne, 2003).

While expressing recombinant proteins in insect cells or transducing vertebrate cells for gene delivery studies, amplification and titer determination of recombinant baculoviruses viruses are critical steps in obtaining accurate and consistent results. Traditionally, the titer determination is done by detecting morphological changes in infected cells by using plaque formation and or point dilution assays. However, these procedures are laborious and plaque assay, in particular, is highly dependent on technical expertise. Indeed, several improved protocols for virus titering have been described. Some of these methods are based on immunological assays, which recognize the viral proteins on the surface of infected cells. For example, Bac-PAK baculovirus rapid titer kit (BD Biosciences; Clontech) is based on the immunological assay against gp64 glycoprotein present on the virus surface. The kit reduces the time required for titer determination to 2 days, however with high cost and manual labor. With shorter time, Mulvania and colleagues detected anti-gp64 labeled cells by flow cytometry 18 hours post infection. Flow cytometry has also been used to detect labeled viral DNA as well as to detect DNA and capsid proteins simultaneously. Titer has also been determined by analyzing cell size after infection and amplifying fragments of selected viral genes by quantitative PCR. As a disadvantage, false positive results may occur when fragments of baculoviral genome are detected. Some methods are based on the expression of the GFP marker protein driven by a viral promoter in infected insect cells. However, in these systems the fluorescent marker is transferred to the baculovirus genome together with the transgene from the donor vector, an arrangement which increases the size of the donor vector, possibly leading to problematic cloning and lack of space for long human genes.

Attempts to miniaturise baculovirus systems to a 96-well format have been unsuccessful. Most of the previous studies describing virus generation in multiwell format, show formation of viruses in 24-well plates. Miniaturisation to a 96-well plate platform would enable the creation of recombinant baculoviruses in a true high-throughput format for highly efficient screening purposes, carried out currently in major pharmaceutical companies (Astra Zeneca, GlaxoSmithKline, Novartis etc.)

For example, McCaII et al, 2005, disclose transfection in suspension, but in 24-well plates.

Bahia et al, 2005, disclose a method to culture insect cells in deep 24-well plates for small-scale optimization for baculovirus-mediated protein expression experiments. It is stated that "The capacity of insect cells to grow in 48- and 96-well format was also investigated, but we were unable to obtain reliable growth of cells in these systems."

Garzia et al, 2003, show no virus generation in HTS format, but an automated method to express and purify a protein in a 96-well plate format.

Albala et al, 2000, report that baculoviruses were produced using a PCR- based method in 96-well plates. However, no details of the method have been published since 2000, which makes it impossible to evaluate the functionality of this baculovirus HTS system. If functional, the PCR step hampers automation. Summary of the Invention

The present invention is based on the discovery that baculoviruses can be generated in high-throughput plates. The present invention is a method of baculovirus generation from insect cells, using a high-throughput plate, wherein the insect cells are transfected or infected by a virus construct, in suspension, and wherein the plate comprises wells which have a substantially square cross-section. Description of Preferred Embodiment

As used herein, a "virus construct" is any baculovirus genome vector that can generate a baculovirus. One example of a suitable virus construct is a BVboostFG virus. Another example is F-Bacmid. As used herein, a "high-throughput plate" is a plate suitable for high- throughput screening (HTS). A HTS plate contains multiples of 96-wells. In the present invention, the HTS-plate is preferably a 96-well plate.

In the present invention, the insect cells are transfected or infected, in suspension. This may improve the baculovirus generation, and help make it possible in 96-well plates. Preferably, they are kept in suspension throughout the transfection process. This may be achieved by shaking the plate.

In the present invention, the plate contains wells which have a substantially square cross-section. This may aid the baculovirus generation.

In a preferred embodiment, the F-bacmid contins a fluorescent protein expression cassette. Preferably, the expression cassette is an enhanced green fluorescent protein expression cassette.

In a preferred embodiment, the insect cells are spodoptera frugiperda (sf cells) or Trichoplusia ni (Tn) cells. More preferably, the cells are sf9 or BTI-TN-5B1- 4 (High Five). A study was conducted using a novel baculoviral genome shuttle vector, F- bacmid, by introducing an enhanced green fluorescent protein (EGFP) gene under the control of polyhedrin promoter into the baculovirus genome. The EGFP cassette was inserted by MuA transposase random reaction and found to be located into the ODV-E66 gene of the baculovirus genome. This novel baculovirus vector enables the easy detection of recombinant virus generation in insect cells by fluorescence. A simple and rapid 18 hour titer determination protocol compatible with F-bacmid or any fluorescent protein expressing baculovirus involves flow cytometry and miniaturization of the baculovirus generation to 96-well format. We show that the F- bacmid viruses produce efficiently desired recombinant proteins in insect cells. Furthermore, the baculoviruses generated by the F-bacmid system also function efficiently as gene delivery vehicles in vertebrate cells. The F-bacmid vector is based on Ark's improved Tn7 mediated BVboost baculovirus generation system, further supporting the suitability of the system for recombinant protein production, high-throughput screening (HTS), functional genomics and gene delivery.

In this study baculovirus generation was shown to be successful in deep 96- well plates for the first time. Using green fluorescent we could reliably detect the positive cells and progression of the infection. Also protein production in the 96-well plate format was verified. This miniaturisation is a milestone in high-throughput screening efforts, allowing several applications like expression cassette screening for gene therapy or screening of variants of therapeutic proteins as well as optimization of expression conditions before initiation of large-scale experiments. Miniaturisation has been done so that all phases can be automated by the use of robot.

To address issues of efficient virus generation, high-throughput protein production and titering, we have developed a 96-well system for the generation of baculoviruses and created a new recombinant baculovirus (F-bacmid) carrying an EGFP expression cassette under the control of insect promoter polyhedrin. This expression cassette was introduced into the baculovirus genome by using a random MuA transposase reaction. The newly constructed F-bacmid is compatible with transposition-based baculovirus generation systems (6,7,21). Consequently, the infected insect cells can be visualized by fluorescence microscopy or flow cytometry, making the infection monitoring and titer determination substantially faster and more convenient as compared to traditional methods.

Aspects of this invention that may be critical to success are the use of substantially square wells and/or cell culture on bottles followed by transfer to plate, before transfection, and/or transfection of the cells in suspension. Rapid expression of recombinant proteins remains challenge in the post- genomic era, since most of the current protocols are time consuming and laborious, limiting the number of constructs and expression conditions which can be studied simultaneously. A high-throughput method, based on established baculovirus system, would be highly useful for this purpose, since baculovirus-insect cell system has capability for eukaryotic protein processing and efficient protein production. However, so far attempts to convert baculovirus system into 96-well format have been unsuccessful. Therefore, we studied the possibility of using 96-well plates for baculovirus generation, titering and optimization of protein production in parallel with a possibility for rapid screening in prokaryotes, invertebrates and vertebrates by using tetrapromoter vector pBVboostFG. For high-throughput purposes the cloning utilizes the site-specific recombination of bacteriophage lambda and a zero background. An EGFP expression cassette was incorporated into baculovirus genome by a random Mu-transposase approach (Figure 1). The resultant construct was named F-bacmid. The cassette was found to be integrated into the ODV-E66 gene. As the ODV-E66 is not included in the structure of a budded baculovirus, integration does not disturb any the functions of the viral vector, but enhances safety by eliminating the formation of intact occlusion derived viruses.

In this study baculovirus generation was shown to be successful in deep 96- well plates. Using green fluorescent protein we could reliably detect progression of the infection as well as protein production in the 96-well plate format. This miniaturization is a vital milestone in developing a high-throughput screening systems, allowing extensive and rapid optimization of multiple expression conditions before initiation of large-scale experiments. Miniaturisation also enables automation. Most of the previous studies describing virus generation in multiwell format, show formation of viruses in 24-well plates. We also observed that the increase of the incubation time from 3 days to 7 days in 96 well plates increased the amount of EGFP positive cells, supported by McCaII et al. Miniaturisation to a 96- well plate platform enables the creation of recombinant baculoviruses and proteins in true high-throughput format for highly efficient screening purposes.

Titer determination is essential for optimal virus amplification, protein production and transduction of vertebrate cells. In addition to plaque- and end-point dilution assays, several improved protocols have been described. It has been shown that plaque assay and end-point titering methods may give lower titers compared to methods developed more recently. Shen and colleagues obtained, on average, 3.7 times higher titers with flow cytometric titering compared to plaque assay and end-point titering. The traditional titering methods are known to be less sensitive and the difference was considered to be a combination of this and the use of SYBR Green I to stain viral DNA, making it impossible to distinguish between infective and non infective virus particles. Stoffel and Rowlen developed a flow cytometer based method to determine titer of a virus stock without infection. They detected, simultaneously, viral DNA labeled with POPO-3 and capsid protein labeled with Sypro Red. This method gave near equal titers than those reported from plaque assays by factors ranging from 2.3 to 3.7.

Since infected cells are the only indication of infective viruses, Mulvania et a/, labelled cells with anti-gp64 and phycoerythrin-conjugated secondary antibody followed by detection by flow cytometry to determinate the titers in less than 24 hours. Yet, the cost of antibodies renders this method rather expensive and tedious. The amount of infective viruses can also be detected by analyzing the size of infected cells in suspension. Janakiraman and colleagues measured cell- diameter 24 hours post infection and applied statistical modeling techniques to obtain virus titer, which were three-fold higher as compared to titers obtained with the plaque assay method. This was explained with the lower sensitivity of the plaque assay as compared to the cell-dimension measurement.

In the current study we describe a titer determination method based on a newly developed F-bacmid and detection of green fluorescence by flow cytometry from baculovirus infected cells. Infection was carried out in suspension cells, mimicking the typical virus and protein production situations. Thus, titers obtained with this method are more accurate for virus amplification and recombinant protein production. Use of EGFP cassette in the vector backbone means that there is no need to transfer the fluorescent marker together with the transfer vector, leaving other cloning/promoter sites of the donor vectors available for additional modifications like pseudotyping and capsid modifications. For improving the baculovirus properties as gene delivery tools, such modifications have been essential. Pseudotyping with VSV-G and VSV-GED has shown to significantly enhance gene delivery, avidin-display has enabled biodistribution imaging and capsid protein vp39 modifications have enabled protein delivery. In addition, transgene expression levels have been enhanced by introducing the Woodchuck Hepatitis Virus Post-transcriptional Regulatory element (WPRE) in to expression cassettes (Mahόnen, A unpublished). Since the accurate viral titer is essential for both protein production and gene therapy, we assessed the accuracy of the flow cytometric method by comparing it to end-point dilution method. Flow cytometric analysis was performed after 18 and 24 hours post infection and it was observed that 24 h incubation resulted to 4 times overestimation of the viral titer. This is likely due to secondary infections. The infection was therefore restricted to 18 hours to eliminate the possibility of viral spread due to secondary infections. For the titer calculations, the method used by Mulvania et al was modified to include only the dilutions containing 1 to 10% of EGFP positive cells. The use of less diluted viruses showing higher EGFP percentages leads to underestimation of viral titers, because the probability of one cells to be infected with multiple viruses in lower dilutions raises. If calculations were carried out by using the method described in Mulvania, titers were in average 3,05 higher than with our calculations. The titers obtained with the new method correlated well with the end-point dilution method, the latter giving on average 1.8 times higher titers than the flow cytometric analysis. This kind of variation has been reported previously. One explanation to this difference might be that during the flow cytometric titer determination, infection is carried out in suspension cells in contrast to adherent cells used in end-point dilution method. A linear regression analysis showed that the titers obtained with the flow cytometric method corresponded well with titers from the end-point dilution with an excellent correlation coefficient of 0.9924. Differences in variances between flow cytometric assay and end-point dilution method were not statistically significant, indicating that both of the methods are equal in reproducibility. Consequently it is concluded that the flow cytometric assay for baculovirus titering is a fast, accurate and reliable.

In order to confirm that F-bacmid modifications do not affect the protein production capability, we produced avidin and VEGF-D^ΔNΛC by using both F-bacmid based and standard baculoviruses. No significant differences were detected in small or large scale protein production in insect or vertebrate cells. The simultaneous expression of EGFP does not seem to affect the protein production or purification.

The presented 96-well plate format offers an efficient virus generation platform enabling high-throughput virus and protein production. The novel F- bacmid based protocol allows direct measuring of infective titer by using a simple and fast flow cytometric method. Steps of bacmid isolation, baculovirus generation and virus titering are adapted into true high throughput format so that the system can be automated. The following study illustrates the invention. Study Construction of modified baculovirus shuttle vector (bacmid)

In order to modify baculovirus shuttle vector propagated in E. coli (bacmid (Lucklow et al, 1993), the baculovirus genome (bacmid) was obtained by extracting the plasmid DNA from DHIOBac cells (Invitrogen, Carlsbad, CA, USA) and subsequently transforming DH10B cells (Invitrogen). Colonies resistant for kanamycin and sensitive to tetracycline (no helper plasmid) were selected. The helper plasmid, providing Tn7 transposition proteins, was isolated similarly with the exception that selection scheme was reversed as compared to bacmid selection. The cassette expressing EGFP was prepared by using the following procedure: Plasmid pfastbad EvolEGFP was digested with Swal and SnaBI. The cassette containing the EGFP gene under the polyhedrin promoter was isolated and blunted with T4 DNA polymerase. This was ligated with EcoRV digested pEntranceposon(cam^r) plasmid (Finnzymes, Espoo, Finland). The accomplished plasmid was digested with BgIII and the fragment containing the pPohl-EGFP cassette flanked by Mu transposon inverted terminal repeats was used in transposition reaction (Figure 1) according to manufacturer's instructions (Finnzymes). The EGFP expression cassette containing bacmids were screened for antibiotic resistance by using chloramphenicol.

Determination of EGFP expression cassette integration site

Inverted PCR was used to determine the integration site of the EGFP expression cassette in the bacmid DNA from the validated clone. The DNA was isolated from bacteria and amplified by PCR using 5'- ATACTCGTCGACAAGCTTCTCG-3' and δ'-GTATCAACAGGGACACCAGGAT-S' primers, which amplified the DNA outwards from the EGFP expression cassette. The amplified fragments were purified using Wizard^® SV Gel and PCR clean up system (Promega, Madison, Wl, USA) and cloned into pGEM^®-T Easy vectors (Promega) according to manufacturer's instructions. Correct clones were screened for by blue-white screening and restriction analysis with EcoRI. To determine the integration site of the EGFP expression cassette in the baculovirus genome the TA- cloned inserts were sequenced with primers δ'-AATACGACTCACTATAGGG-S' and δ'-CATACGATTTAGGTGACACTATAG-a¹. The obtained nucleotide sequences were fed into the NCBI-BLAST^® (National Centre of Biotechnology, USA- Basic

Local Alignment Search Tool) search program.

Construction of F-bacmid containing DH10BacΔTn7 E. coli strain

To reconstruct an E. coli strain that allows Tn7 mediated generation of recombinant baculoviral genomes without background, F-bacmid containing DH10B cells were transformed with helper plasmid and the resulting bacterial clone was further transformed by pBVboostΔamp and selected for as described in Airenne 2003. The strain was named as DH10BacΔTn7EGFP. Generation of recombinant baculoviruses Recombinant baculoviruses expressing transgene under universal promoter were prepared as described in Laitinen, 2005 by using E. coli strain DH10BacΔTn7EGFP. Cells were transformed with pBVboostFG (7) + avidin or VEGF-D^ΔNAC donor vectors, containing chicken avidin cDNA (22) and VEGF-D^ΔNΔC cDNA (23), respectively. As controls, the same donor vectors were used to generate viruses by using conventional BVboost system. Sf9 (Spodoptera frugiperda, Invitrogen) cells producing the new F-bacmid viruses were detected by fluorescence microscopy. Titer determination by flow cytometry

Titers were determined by infecting insect cells in suspension with a series of virus dilutions and determining the percentage of infected cells by flow cytometry. The virus samples were serially diluted in duplicates and 0.5 ml of each dilution was used to infect 0.5 ml of Sf9 cells (approximately 2 x 10⁶ cells/ml). Incubation was allowed to continue for 18 hours at 27°C with shaking at 270 rpm, before the samples were centrifuged for 5 minutes at 500 x g. The cells were resuspended in 2% FBS in PBS and analyzed by flow cytometry (FACSCalibur, Becton Dickinson, Franklin Lakes, NY, USA) to reveal the percentage of the cells which were infected by EGFP expressing baculoviruses. The percentage of cells emitting green fluorescence for each sample was analyzed by counting 10000 cells per measurement in triplicates. To ensure optimal infection time and to ascertain its effect on viral titer, titer was also determined 24 hours post infection. In addition, the virus titers were also determined by using end-point dilution method. Protein production efficiency of F-bacmid derived viruses in insect cells

To study protein production, 25 ml of Sf9 cells (1 x 10⁶ cell/ml) were infected with avidin and VEGF-D^ΔNΔC viruses [F-bacmid derived and standard BVboost viruses] at a multiplicity of infection (MOI) 5. Infection was accomplished with secondary virus preparation and allowed to continue for 4 days at 27^QC, shaking at 110 rpm. The infected cells were lysed and total protein amount was quantified with Coomassie Plus- The Better Bradford™ Assay Reagent (Pierce, Rockford, IL, USA). Lysates and supematants were analyzed by Western blotting using polyclonal rabbit anti-avidin and mouse anti-hVEGF-D (R&D systems, Minneapolis, MN, USA) antibodies. The concentration of VEGF-D^ΔNΔC in the supernatant was also analyzed by ELISA (Quantikine® Human VEGF-D kit, R&D Systems) according to manufacturer's instructions.

Protein production was also performed in larger scale with F-bacmid derived baculovirus expressing VEGF-D^ΔNΔC. Sf9 cells (1 x 10⁶ cells/ml, 500 ml) were infected with the virus at MOI 5. The cells were incubated at 27^QC, shaking at 110 rpm for 4 days prior to harvesting the medium for protein purification. Purification was accomplished with metal affinity chromatography (IMAC). The protein was attached to 3 ml of cobalt containing resin (BD Talon Metal Affinity Resin, BD Biosciences Clontech, Palo Alto, CA, USA) and subsequently transferred to purification columns (EconoPac, Bio-Rad, Hercules, CA, USA). The protein containing column was washed 3 times with washing buffer (50 mM Sodium phosphate, 300 mM NaCI , pH 7.0) prior to collection of 12 eluates (elution buffer; 50 mM HEPES, 20 mM NaCI, 450 mM imidazole) comprising 1 ml each. Eluate fractions 2-4 were dialyzed (Slide-A-LyzerDialysis Casette, 3.5K MWCO, Pierce) twice for 6 hours and concentrated against buffer containing 50 mM HEPES, and 20 mM NaCI, 30% PEG at pH 7.4. The concentration of the purified protein was determined using DC protein assay kit (Bio-Rad) according to manufacturer's instructions. Testing F-bacmid derived baculoviruses in mammalian cells Human hepatoma (HepG2), rat glioma (BT4C), and human ovarian cancer

(SKOV-3) cells were transduced with avidin and VEGF-D^ΔNΛC viruses (F-bacmid derived and standard BVboost viruses). HepG2, BT4C and SKOV-3 cells were plated into a 6 cm plate containing 5 x 10⁵, 6 x 10⁵, and 1.5 x 10⁵ cells, respectively, and transduced after 24 hours. It has been shown that mammalian cell cultures are not disturbed by small amounts of insect cell medium when transducing mammalian cells with unconcentrated baculoviruses. Thus, mammalian cells were transduced with unpurified secondary virus preparations in insect cell medium. These secondary virus preparations were diluted into 0.5 ml Insect X-press medium (BioWhittaker, Heidelberg, Germany), to contain equal amounts of infective virus particles (MOI 5) according to the virus titers. These dilutions were added to 2.0 ml of complete mammalian cell medium and the cells were transduced with this solution. Fresh medium was changed after 24 hours and transduction was allowed to continue for another 24 hours prior sample analysis. Both the transduced cells and the medium were analyzed by Western blotting as described previously. The concentration of VEGF-D^ΔNΔC from the supernatant was also analyzed by ELISA (Quantikine^® Human VEGF-D kit). Baculovirus generation in 96-well plate format Baculovirus generation was carried out in deep 96-well plate format by transfecting Sf9 cells in suspension with F-bacmid expressing VEGF-D^ΔNΛC. The transfection solution consisted of 40 μl of Insect X-press medium, 0.6 μl of cationic lipid reagent Cellfectin^® (Invitrogen) and 5 μl of F-bacmid DNA. The bacmid DNA was isolated with Montage BAC₉₆ Miniprep Kit according to the manufacturer's instructions (Millipore, Billericia, MA, USA) and the DNA concentrations were measured by NanoDrop (NanoDrop Technologies, Wilmington, DE, USA). Following incubation of the transfection solution at room temperature for 30 minutes, 0.5 x 10⁶ Sf9 insect cells in 0.5 ml of medium (Insect X-press) were added. After 5 hours of incubation at 27°C shaking at 270 rpm (Innova44 shaker, New Brunswick Scientific Edison, NJ, USA), an additional 0.5 ml of Insect X-press medium was added to the cells. The samples were incubated at 27°C shaking at 270 rpm for 7 days. The vigorous shaking was essential to keep the cells in suspension. Possible cross-contamination between the wells was eliminated and aeration ensured by covering the plate with breathable sealing film. Following incubation, a sample of 100 μl was collected from each well of the plate and the formation of viruses was analyzed.

The generated baculoviruses were amplified and proteins were produced by infecting in 750 μl (approximately 1 x 10⁶ cells/ml) Sf9 cells in 96-well plate, with 250 μl of primary virus supernatant (collected by centrifuging samples at 500 x g for 5 minutes). The cells were incubated at 27°C, shaking at 270 rpm for 7 days. Virus amplification was analyzed by fluorescence microscopy. The virus containing medium was collected by centrifugation as described previously, and stored at 4⁵C. Protein production was analyzed by Western blotting using polyclonal rabbit anti- avidin and mouse anti-hVEGF-D (R&D systems) antibodies.

In addition to fluorescence detection by fluorescent microscopy, virus formation was monitored daily by flow cytometry (starting from day 3 post transfection). Samples of 150 μl were removed from each well and the cells harvested by centrifugation. Samples were resuspended in 2 % FBS in PBS and the percentage of cells emitting green fluorescence was analyzed by counting 10000 cells in triplicate. Results Effect of the integration site to the function of the virus Based on the sequence analysis using NCBI-BLAST^®, the integration site was found to have a 98% identity to the ODV-E66 (a 79 kDa protein) gene (Figure 1). ODV-E66 is a structural protein found on the envelope of occlusion derived baculovirus. The ODV-E66 protein is not included in the structure of budded baculovirus (the form used in protein production and gene therapy applications). Thus, the integration site should not disturb the function of the viral vector. To support this hypothesis, F-bacmid derived viruses were produced, and it was shown that EGFP expression cassette integrated in ODV-E66 gene does not affect the titer (Table 1) or function of the virus. Titer determination by flow cytometry To determine titer, Sf9 insect cells were infected with serial dilutions of

EGFP expressing baculoviruses and the percentage of infected insect cells was determined by flow cytometry at 18 and 24 hours post infection. However, the 24 h incubation led to overestimation of virus titer due to secondary infections (Table 3). Calculations of virus titers were carried out by modifying the normalization method described by Mulvania et al. (Tables 2 and 3). First, the EGFP percentages of duplicate samples were averaged and normalized by dividing the averaged percentage by the maximum EGFP percentage (EGFP percentage of 1 :500 dilution of concentrated virus or undiluted secondary virus preparation). The value represents a corrected percentage of the EGFP expression. The corrected percentage was then multiplied by the dilution factor and by the number of cells per ml, to obtain the titer of the virus stock.

_. ... . .- averaged EGFP % of duplicate samples ... . . „ .

Titer (ifu/ml) = — x dilution factor x cells per ml maximum EGFP %

Table 1 : Virus constructs and their titers obtained with end-point dilution and flow cytometer methods. Titers are shown as averages of multiple analyses. The F-bacmid derived viruses are secondary virus preparations and the two0 last ones are concentrated viruses.

End-point SD n FACS SD n

F-bacmid- VEGF-D^⁰1 6.85E+08 4.30E+08 4 3.71E+08 1.19E+08 6

F-bacmid-VEGF-D^ΔNΔC H 1.20E+09 0.0 2 8.44E+08 4.08E+08 6

F-bacmid- VEGF-D¹^ Ql 2.57E+09 1.06E+09 2 1.77E+09 9.55E+08 8

F-bacm id- A v idin 4.25E+08 2.01E+08 4 3.12E+08 8.56E+07 2

FG-GF 1+2 1.67E+10 3.62E+O9 4 2.99E+10 2.99E+10 12

FG-EGFP 2.48E+10 2.40E+10 4 6.19E+09 1.66E+O9 6

Table 2: Titer determination of a concentrated virus after 18 hours incubation.5

Titer (ifu/ml) = CP x DF x Cells per ml = (1.41 / 19.07) x 500000 x 1.73E+06 cells/ml = 6.35E+10

Dilution EGFP % of Average Corrected Percentage (CP) Dilution factor CP x DF Titer = (CP x DF) x

Serie 1 Serie 2 = (average/average of 1 :500) x 100 (DF) cell number

1 :500 18,82 19,36 19,07 - 500 - -

1 :5000 3,75 4,47 4,09 21 ,45 % 5000 1072,36 1.85E+09

1 :50000 1,69 1,16 1,41 7,37 % 500000 36837,97 6,35E+10

1 :500000 0,1 0,19 0,13 0,66 % 5000000 32773,99 5.65E+10

1 :5000000 0,1 0,13 0,10 0,50 % 50000000 249082,33 4.30E+11

1 :50000000 0,05 0,00 0,01 0,03 % 500000000 131095,96 2.26E+11

Titer: 6,35E+10

Table 3: Titer determination of a concentrated virus after 24 hours of incubation. 5 Titer = CP x DF x Cells per ml = (0.53 / 45.35) x 5000000 x 1.73E+06 cells/ml = 1.01 E+11

Dilution EGFP % 0f Average Corrected Percentage (CP) Dilution factor CP x DF Titer = (CP x DF) x

Seriβ 1 Serle 2 = (average/average of 1:500) x 100 (DF) cell number

1 500 44,53 46,16 45,35 - 500 5 - -

1 5000 19,61 18,99 19,30 42,56 % 5000 2128,13 3.67E+09

1 50000 7,71 5,34 6,53 14,39 % 500000 71948,40 1.24E+11

1 :500000 0,71 0,35 0,53 1,17 % 5000000 58440,84 1,01 E+11

1 5000000 0,06 0,05 0,06 0,12 % 50000000 60646,16 1.05E+11

1 50000000 0,06 0,03 0,05 0,10 % 500000000 496195,83 8.56E+11

Titer: 1,01 E+11

Instead of using all dilutions for titer determination, titers were calculated from dilutions which contained 1 to 10% of EGFP positive cells following normalization (bolded in Tables 2 and 3). Using dilutions that produced higher0 EGFP percentages led to an underestimation of the virus titer and vice versa. If more than one dilution had a corrected percentage of between 1 to 10%, these corrected percentages were averaged for titer determination. To confirm the results obtained from the flow cytometric assay, viral titers were also determined with end- point dilution method based on the detection of morphological changes and EGFP5 fluorescence of cells.

The results obtained using the flow cytometric assay corresponded well with the titers from the end-point dilution combined with the EGFP fluorescence detection (Table 1 ). However, the end-point dilution method gave on average 1.8 times higher results than the flow cytometric analysis. Correlation, determined by0 linear regression analysis, between the methods was excellent (R=0.9924). Variances of the titers (Table 1 ) obtained with both methods showed statistically no significant difference, confirming that the reproducibility of the methods is equally good. Protein production efficiency in insect and mammalian cells 5 Sf9 insect cells were infected with either F-bacmid derived or standard

BVboostFG viruses containing avidin and VEGF-D^ΔNΔC expression cassettes under universal promoter. Protein production was analyzed by Western blotting and ELISA. Western blot analysis showed high levels of VEGF-D^ΔNΔC expression when Sf9 cells were infected with F-bacmid derived or normal BVboostFG viruses. These0 results were confirmed by ELISA. As in the case of VEGF-D^ΔNΔC, avidin production did not significantly differ between the F-bacmid derived and standard BVboostFG viruses, with results within the normal variation.

Protein production was also performed in larger volume (0.5 I) with F-bacmid derived baculovirus expressing VEGF-D^ΛNΔC. The concentration of purified VEGF- D^ΔNΔC was comparable to the production and purification of VEGF-D^ΔNΔC by using the traditional pBVboost system.

F-bacmid viruses resulted in efficient gene expression in human hepatoma (HepG2), rat glioma (BT4C), and human ovarian cancer (SKOV-3) cells also. Virus generation in 96-well plate format Virus generation was carried out in 96-well plate format by transfecting 0.5 x

10⁶ Sf9 suspension cells with 5 μl of isolated bacmid DNA. Seven days post transfection, aliquots of cell suspensions were collected and primary virus stock was used in virus amplification and protein production.

The polyhedrin promoted expression of EGFP enabled detection of infection by fluorescence microscopy, showing that 88-100% of the 96 wells contained viruses. The EGFP percentage of F-bacmid transfected Sf9 cells increased daily between days 3 to 7. On day 7, the percentage of positive cells per well varied between 70 to 85%. Protein production was successfully detected with Western blotting.

Reference List

Airenne et al (2003) Nucleic Acids Res., 31 , e101.

Luckow et al (1993) J. Virol., 67, 4566-4579. Laitinen et al (2005) Nucleic Acids Res., 33, e42.

Bahia et a/ (2005) Protein Expr.Pυrif., 39, 61-70.

McCaII et a/ (2005) Prote/π Expr.Pυrif., 42, 29-36.

Garzia ef a/ (2003J, Biotechniqυes, 35, 384-391

Albala ef al (2000), J. Cell Biochem, 80, 187-191 Mulvania et al (2004) BioProcessing journal, 3, 47-53.

Claims

1. A method of baculovirus generation from insect cells, using a high- throughput plate, wherein the insect cells are transfected or infected by a virus construct, in suspension, and wherein the plate comprises wells which have a substantially square cross-section.

2. A method according to claim 1 , wherein the high-throughput plate is a 96- well plate.

3. A method according to claim 1 or claim 2, wherein the insect cells are kept in suspension by aeration.

4. A method according to claim 3, wherein the aeration is by shaking the plate.

5. A method according to any preceding claim, wherein the insect cells are transfected for less than 20 hours.

6. A method according to any preceding claim, wherein the infective titer is calculated.

7. A method according to any preceding claim, wherein the virus construct is F- bacmid.

8. A method according to claim 7, wherein the F-bacmid contains a fluorescent protein expression cassette.

9. A method according to any preceding claim, wherein the insect cells are Spodoptera frugiperda (sf) cells or Trichoplusia ni (Tn) cells.

10. A method according to claim 9, wherein the sf cells are sf9 cells.

11. A method according to claim 9, wherein Tn cells are BTI-TN-5B1-4 (High Five).