Expression of feline recombinant interferon-γ in baculovirus and demonstration of biological activity

Veterinary Immunology and Immunopathology 64 (1998) 97±105

Expression of feline recombinant interferon-g in baculovirus and demonstration of biological activity D.J. Argyle*, M. Harris, C. Lawrence, K. McBride, R. Barron, C. McGillivray, D.E. Onions Division of Small Animal Clinical Studies and Department of Veterinary Pathology, University of Glasgow Veterinary School, Bearsden Road, Glasgow G61 1QH, UK Accepted 18 March 1998

Abstract We have previously reported the cloning of the coding sequence for feline-specific interferon-g. Here, we describe the expression of this sequence in a baculovirus system and demonstrate the biological activity of the recombinant protein. The coding sequence for feline interferon was directionally cloned into the baculovirus transfer vector pAcCL29-1. Transfer vector and linearized wild-type AcMNPV (BacPAK6) were used to co-transfect Sf9 cells by calcium phosphate coprecipitation. Subsequently, wild-type and recombinant viruses were separated by plaque assay. Recombinant plaques were expanded and a master stock of virus is produced. Production of biologically active interferon-gamma from infected Sf9 cells was demonstrated using a standard cytopathic effect reduction assay, utilising vesicular stomatitis virus (VSV), and an MHC class II induction assay. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Feline; Interferon-gamma; Baculovirus

1. Introduction Interferon-gamma is a pleiotropic cytokine which is recognised as an important modulator of the immune response (Pace et al., 1985). It is produced by lymphocytes in response to antigenic or mitogenic stimulus and its major cellular target is the * Corresponding author. Tel.: +44 141 330 5700; fax: +44 141 942 7215; e-mail: [email protected] Abbreviations: CPER: Cytopathic effect reduction assay; LRU: Laboratory reference unit; WHO: World health organisation; VSV: Vesicular stomatitis virus 0165-2427/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 2 4 2 7 ( 9 8 ) 0 0 1 2 7 - 5

98

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

macrophage where it plays an important role in augmenting antigen presentation by the up-regulation of MHC expression (Farrar and Schreiber, 1993 and Pestka et al., 1987). Due to its central role in orchestrating the immune response it has become an attractive candidate for use as a vaccine adjuvant (Heath and Playfair, 1992) and as a possible immunotherapeutic agent for the treatment of cancer and infectious diseases (O'Garra, 1989 and Borden and Sondel, 1990). The cloning and sequencing of the feline interferon-gamma gene has been reported (Argyle et al., 1995). This paper reports the expression of the feline sequence in a baculovirus system to generate a fully functional recombinant protein. The production of this protein will allow us to explore this cytokine as a potential vaccine adjuvant or immunotherapeutic agent for diseases of Felidae. 2. Materials and methods 2.1. Production of recombinant virus A schematic representation of the method used to engineer the recombinant virus is shown in Fig. 1. Initially, the feline interferon-g cDNA coding sequence was subcloned into the transfer vector pAcCL29-1 (Kitts and Possee, 1993). One positive clone (designated pAcCL29-1-ifn) was grown in culture and purified on a caesium chloride gradient. A stock of wild-type AcMNPV (BacPak6) (King and Possee, 1992) was amplified in Sf9 cells and subsequently isolated on a sucrose gradient. Infectious viral DNA was restriction digested with Bsu36I and this linearized virus was co-transfected with pAcCL29-1-ifn into Sf9 cells using a calcium phosphate co-precipitation method

Fig. 1. The generation of recombinant baculovirus expressing feline interferon-gamma.

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

99

(stratagene). Parental and recombinant viruses were separated by plaque assay as described by King and Possee (1992). Recombinant viruses were subjected to two rounds of plaque purification and subsequently used to generate master and working stocks of recombinant virus. 2.2. Generation of recombinant viral stocks High titre working stocks of virus were generated on Sf9 cells to enable efficient, high multiplicity of infection during the production cycle. Viral titre was determined by plaque assay (King and Possee, 1992) and was found to be 51010 PFU/ml. This was designated as the master stock of virus. A working stock of virus was generated in a similar way and plaque assay demonstrated this to be 3108 PFU/ml. 2.3. Production of recombinant protein Prior to determining the optimum culture conditions for maximal protein production, a time course study was performed (data not shown). Subsequently, a set protocol for protein production was established to minimise any variations between supernatant batches. For maximum protein production, cells were infected with 10PFU/cell recombinant virus in logarithmic growth (cell viability >98%) and at a concentration of 5105 cells/ml in spinner culture. The supernatant was harvested for 5 days postinfection and was clarified by centrifugation. The supernatant was filtered through a 0.2u filter and stored at ÿ208C. 2.4. Cell lines The F422 cells are an FeLV positive cell line derived from a naturally occurring lymphosarcoma (Rickard et al., 1969). The cells were cultured in RPMI medium supplemented with 10% FCS and penicillin/streptomycin. FEA cells are a continuous feline embryonic cell line and were maintained in plasticware at 378C in 95% air, 5% CO2. Spodoptera frugiperda (Sf9) cells (Invitrogen) are the host for the propagation of wildtype or recombinant Autographica californica multiple nuclear polyhedrosis virus (AcMNPV). Cultures were maintained in plasticware at 288C in TC100 medium (Gibco) supplemented with 10% FCS and penicillin/streptomycin. There is no requirement to maintain cells in a CO2 incubator. For the most part, cells were grown as adherent monolayer cultures until sub-confluence. For protein production, the cells were grown as suspensions cultures in spinner culture flasks in TC100 complete medium. For maximum protein production the virus was added at 10 pfu/cell. 2.5. Wild-type AcMNPV BacPAK6 wild-type AcMNPV was used in this experiment (Kitts and Possee, 1993). BacPAK6 has two Bsu36I sites which flank a b-galactosidase gene, thus recombinant BacPAK6 viral plaques stain blue with X-gal.

100

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

2.6. Vesicular stomatitis virus A temperature sensitive mutant of VSV (VSVtsE2) was used in these experiments. VSVtsE2 is derived from the New Jersey strain of VSV and contains two independent mutations within the NS gene (Rae and Elliott, 1986). 2.7. Assay for feline rIFN-g activity The anti-viral activity of feline recombinant interferon-gamma was assayed by inhibition of cytopathic effects of vesicular stomatitis virus (VSV) in feline embryonic fibroblast (FEA) cells in 96 well culture plates (CPER or Cytopathic Effect Reduction Assay). Cells were added to wells at a concentration of 2104 cells/well in 100 ml of medium. VSV virus was added at a concentration of 4 pfu/cell and the assay was carried out in triplicate. Staining was carried out by the addition of 100 ml of Amido Blue Black (0.05% solution in 9% acetic acid with 0.1 M sodium acetate) to each well and leaving them for 1 h. The cells were fixed for 1 h by the addition of 100 ml of formalin acetate (10% formaldehyde solution in 9% acetic acid with 0.1 M sodium acetate). Dye elution was carried out by the addition of 100 ml of 0.38% sodium hydroxide to each well. Optical density at 595 nm was measured spectrophotimetrically. The staining, fixing and elution protocols were as described by Balkwill, 1991. The inhibition titres of feline rIFN-g were expressed in laboratory reference units defined as the reciprocal of the dilution at which 50% protection against virus was obtained as measured by the optical density of FEA cell dye uptake in the wells of the microtitre plates. 2.8. Class II histocompatibility antigen analysis F422 cells were grown in tissue culture and used at a concentration of 3105 cells/ml. Two-fold serial dilutions of the interferon preparation were made in 12 well tissue culture plates. Wells were also assigned to virus and cell controls. One ml of the cell suspension was added to each well and the cells were allowed to incubate at 378C/5% CO2 for around 48 h. Following incubation, the cells in each well were divided in two to give a test (‡) and negative control (ÿ) for the subsequent flow cytometry analysis. The test cells were labelled with MHC class II anti-cat monoclonal antibodies (IgG1) and the negative control cells were labelled with CD8 anti-cat monoclonal antibodies (IgG1). The MHC class II antigen is inducible on F422 cell lines (Willette, unpublished results), whereas the cells are CD8 negative. The anti-cat MHC class II monoclonal antibody was used to demonstrate any changes in MHC class II expression on the cell line. The anti-cat CD8 monoclonal antibody was used to measure non-specific binding of mouse monoclonal antibody. The cell controls had been incubated in the absence of interferon-gamma. The virus controls were cells which had been incubated with Sf9 cell supernatants containing wild-type PAK6 baculovirus at a 1:32 dilution. Both of these controls were labelled with anti-cat MHC class II and anti-cat CD8 monoclonal antibodies in the same manner as the cells incubated with the interferon preparations. Sheep anti-mouse IgG FITC conjugate (Sigma) was used to demonstrate fluorescence by flow cytometry. For flow cytometry (EPICS cell sorter (Coulter Electronics, Luton)), isotype matched controls were used to set an analysis gate such that <1% were positive.

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

101

Five thousand events were collected in LIST mode and percentage positive were calculated relative to isotype matched negative controls.

3. Results 3.1. Protein production In the absence of specific anti-sera to feline recombinant interferon-gamma, monitoring of protein production in the baculovirus system was carried out using the bioassays described. 3.2. In-vitro assays Supernatants from baculovirus cultures were examined for biological activity in-vitro using both a standard VSV protection assay (CPER, Cytopathic Effect Reduction Assay) and an MHC class II up-regulation assay. VSV CPER assays were performed on the baculovirus supernatants using FEA cells as host cells. The results of a CPER assay for two independent interferon-gamma samples are shown in Table 1 and demonstrated graphically in Fig. 2. The interferon preparation was quantified as the reciprocal of the dilution of interferon which gives 50% protection. This value was expressed in terms of laboratory reference units (LRU). Interferon-gamma is highly species specific and thus it is impossible to standardise these units in terms of international units (as described by WHO). A titration of the interferon preparation demonstrated a consistent specific activity of 8.5106 Laboratory reference units (LRU)/Litre of baculovirus culture media. The MHC class II induction assay was performed on the baculovirus interferon preparation derived from spinner culture. The results of the flow cytometry analysis are shown in Table 2. The figures in this table are the mean values from duplicate experiments and represent the percentage of cells positive for MHC class II. The results are shown graphically in Fig. 3. All of the cells which were labelled with the anti-cat CD8 monoclonal antibody showed negligible fluorescence. Cell controls and wild-type virus controls showed base line fluorescence when labelled with the anti-cat MHC class II monoclonal antibody indicating that there is some expression of MHC class II antigen on these cells in the absence of interferon-gamma. However, incubation of the cells with interferon dilutions clearly enhances the expression of this antigen on these cells. Table 1 CPER assay for two independent feline interferon-g samples. The readings are shown as optical density measurements (595 nm). The O.D. reading for cell and virus controls were 2.888 and 0.994, respectively

Sample 1 Sample 2

1:10

1:100

1:103

1:104

1:105

2.732 2.256

2.559 2.279

2.405 2.059

1.817 2.006

1.563 1.250

102

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

Fig. 2. Graphical representation of the CPER assays for interferon samples 1 and 2 (Table 1). Table 2 Flow cytometry analysis of feline rIFN-g induced MHC class II expression on F422 cells. Figures represent the number of cells positive for MHC class II and are mean values from duplicate experiments. The percentage of nonspecifically positive cells stained with the isotype matched control was in the range 0.8±1.8

MHC class II CD8 marker

1:4

1:8

1:16

1:32

1:64

1:128

1:256

31.1 neg.

38.9 neg.

44.9 neg.

63.55 neg.

69.9 neg.

68.85 neg.

69.3 neg.

4. Discussion The cloning of the feline-specific coding sequence for interferon- gamma afforded the opportunity to produce recombinant protein. Research has demonstrated that interferon-gamma is species specific and therefore this work demonstrates an advance in feline immunology and will allow the potential of this cytokine to be explored in the treatment of feline diseases. The first reports of the use of baculovirus as expression systems were published by Smith et al. (1983) and Pennock et al. (1987) who used AcMNPV to produce b-interferon and b-galactosidase, respectively, in SF9 cells. Maeda et al. (1985) also expressed b-interferon using BmNPV system in the silkworm. In the succeeding years a wide variety of recombinant proteins have been expressed using this system (reviewed by King and Possee, 1992), including feline interferon-alpha (Sakurai et al., 1992) and human interleukin-2 (Smith et al., 1985). In Smith's baculovirus system the interferon was secreted at levels of 5106 U/ml and the protein was glycosylated. The data indicated that the signal peptide at the amino terminal end of the

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

103

Fig. 3. Flow cytometry analysis of interferon induced MHC class II expression on F422 cells. Following incubation of the cells with the interefron prearation, there is marked up-regulation of MHC class II expression compared to isotyped matched controls.

primary translation product was removed and the interferon had no anti-viral effect on the baculovirus. The high levels of expression of foreign genes reported for baculovirus systems, the fact that foreign genes undergo post- translational modification, and the ability to scale up protein production encouraged the use of this system for expression of feline recombinant interferon-gamma. The ultimate aim for the project was to take the recombinant protein towards clinical application. While glycosylation is unimportant for the biological activity of the interferon protein, it may be important in terms of extending the biological half-life of the protein. This is obviously an important factor when considering therapeutic potential. Thus, baculovirus expression could potentially yield a product with high specific activity, potentially greater in-vivo half-life than yeast or E.coli derived proteins, and be easily scaled up to producing large amounts of protein for clinical use. The expression of feline recombinant interferon-alpha(tau) in the baculovirus/silkworm system has been reported by Ueda et al. (1993). Interestingly, this group demonstrated that their interferon preparation protected feline cells (Crandell feline kidney-CRFK) against challenge with feline herpesvirus (FHV) or feline calicivirus (FCV). It was found in our system that it was difficult to demonstrate reproducible protection to challenge with FCV of FEA cells incubated with recombinant interferon-gamma (data not shown). However, interferon-alpha is a far more potent anti-viral cytokine than interferon-gamma (Farrar and Schreiber, 1993) and this may explain the poor reproducibility of the antiviral assay based upon FCV in our system compared to the assays carried out using feline recombinant interferon-alpha. The MHC class II induction assay appeared to be as sensitive as the anti-viral assays, especially at low concentrations, for the detection of biologically active recombinant protein. This is in accordance with reported assay performances being up to 10 times

104

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

more sensitive than the anti-viral assays (Current Protocols in Immunology, 1994). This is an easy assay to perform and may be more practical on a day to day basis considering the need for containment facilities for the use of VSV. Interestingly, at high concentrations of interferon (1:4±1:32), while MHC class II was still up-regulated, the degree of up-regulation was less than those cell incubated with lower dilutions. This may be because at higher concentrations there is a higher concentration of TC100 medium which may be altering the pH or nutrient concentrations and thus the cells will not be growing under optimum conditions. The MHC class II induction assay seem to be very sensitive and up-regulation is still seen at dilutions of interferon preparations of 1:4096 (data not shown) Although it is difficult to quantify the baculovirus derived interferon, other than in terms of laboratory units, it can be concluded that recombinant interferon-gamma can be stably expressed in the baculovirus system and that the protein product appears to be biologically active. It is difficult to assess accurately the level of expression from the baculovirus system. During the plaque purification stages, recombinant plaques were chosen which appeared to give the highest level of expression in the crude anti-viral assays. These apparent differences in levels of expression of different phenotypes may not actually represent variations in the phenotype but may reflect differences in experimental conditions. However, it has been advised that several alternative recombinant isolates are screened in case there are differences in the expression phenotype, although this is a rare occurrence (Bishop, 1992). Usually the expression level of a foreign gene does not vary from experiment to experiment, provided the infection conditions are optimal (cell viability of >98%, cells in exponential growth, use of high titred virus stocks to establish one-step growth). It has been described that the first 23 amino acids of the coding sequence for interferon-gamma represent a leader sequence which is cleaved from the mature protein. No protein sequence analysis is available for the feline recombinant protein, but the demonstration of biological function indicated that the protein had undergone posttranslational processing, i.e. there was secretion of a functional protein which indicated that the signal sequence is being cleaved and homodimers formed. In conclusion, this work demonstrates the stable expression of biologically active feline recombinant interferon-gamma in a baculovirus system. The up-regulation of MHC class II expression on cells allows for an assay system for interferon-gamma, and also reinforces the conceptual idea that interferon-gamma may be a useful adjuvant for viral vaccines. Clinical trials using interferon in this manner have been performed in mouse models and in human immuno-compromised patients (Quiroga et al., 1990) with encouraging results. Work is currently in progress to assess the potential of feline recombinant interferon-gamma has a viral vaccine adjuvant in cats. Acknowledgements This work was supported by the Wellcome Trust and Q1 biotechnology. The pAcCL29-1 vector and BacPAK6 virus was kindly provided by Dr. Jones at the NERC Institute of Virology and Environmental Microbiology, Oxford. The primary VSVtsE2

D.J. Argyle et al. / Veterinary Immunology and Immunopathology 64 (1998) 97±105

105

virus stock was kindly provided by Dr. Richard Elliott of the MRC Institute of Virology, Glasgow. References Argyle, D.J., Smith, K., McBride, K., Fulton, R., Onions, D.E., 1995. Nucleotide and predicted peptide sequence of feline interferon-g. DNA Sequence 5, 169±171. Balkwill, F.R., 1991. Cytokines, a practical approach, ch. 23. IRL Press, pp. 324±326. Bishop, D.H.L., 1992. Baculovirus expression vectors. Seminars in Virology 3, 253±264. Borden, E.C., Sondel, P.M., 1990. Lymphokines and cytokines as cancer treatment. Cancer 65, 800±814. Current Protocols in Immunology (1994). In: Colligan, J.E., Kruisbeek, A.M., Margulies, D.H., Sevach, E.M., Strober, W., (Eds.), Pub. Wiley, INC. Farrar, M.A., Schreiber, R.D., 1993. The molecular cell biology of interferon-gamma and its receptor. Ann. Rev. Immunol. 11, 571±611. Heath, A.W., Playfair, J.H.L., 1992. Cytokines as immunological adjuvants. Vaccine 10(7), 427±434. King, L.A., Possee, R.D., (1992). The Baculovirus Expression System: A Laboratory Guide. Published by Chapman and Hall. Kitts, P.A., Possee, R.D., 1993. A method for producing recombinant baculovirus expression vectors at high frequency. Bio. Techniques 14, 810±817. Maeda, S., Kawai, T., Obinata, M., Fujiwara, H., Horiuchi, T., Saeki, Y., Sato, Y., Furasawa, M., . Nature 315, 592±594. O'Garra, A., 1989. Interleukins and the immune system 2. The Lancet 1(8645), 1003±1005. Pace, J.L., Russel, S.W., LeBlane, P.A., Murasko, D.M., 1985. Comparative effects of various classes of mouse interferons on macrophage activation for tumour cell killing. J. Immun. 134, 977±981. Pennock, G.D., Shoemaker, C., Miller, L.K., . Molecular and Cell Biology 4, 399±406. Pestka, S., Langer, J.A., Zoon, K.C., Samuel, C.E., 1987. Interferons and their actions. Ann. Rev. Biochem. 56, 727±777. Quiroga, J.A., Castillo, I., Porres, J.C., 1990. Recombinant gamma-interefron as adjuvant to hepatitis-B vaccine in haemodialysis patients. Hepatology 12, 661±663. Rae, B.P., Elliott, R.M., 1986. Characterisation of the mutations responsible for electrophoretic mobility differences in the NS proteins of vesicular stomatitis virus New Jersey complementation group E mutants. J. Gen. Virology 67(12), 2635±2643. Rickard, C.G., Post, J.E., deNoronha, F., Barry, L.M., 1969. A transmissible virus induced lymphocytic leukaemia of the cat. J. Natl. Cancer. Inst. 42, 987±1014. Sakurai, T., Ueda, Y., Sato, M., Yanai, A., 1992. Feline interferon production in silkworm by recombinant baculovirus. J. Vet. Med. 54(3), 563±565. Smith, G.E., Ju, G., Ericson, B.L., Moschera, J., Lahm, H., Chizzonite, R., Summers, M.D., 1985. Modification and secretion of human interleukin-2 produced in insect cells by a baculovirus expression vector. Proc. Natl. Acad. Sci. USA 82, 8404±8408. Smith, G.E., Summers, M.D., Fraser, M.J., 1983. Production of human beta interferon in insect cells infected with a baculovirus expression vector. Molecular and Cell Biology 3(12), 2156±2165. Ueda, Y., Sakurai, T., Yanai, A., 1993. Homogeneous production of feline interferon in silkworm by replacing single amino acid code in signal peptide sequence of recombinant baculovirus and characterisation of the product. J. Vet. Med. Sci. 55(2), 251±258.