Baculovirus Expression of the Fusion Protein Gene of Bovine Respiratory Syncytial Virus and Utility of the Recombinant Protein in a Diagnostic Enzyme Immunoassay

Abstract

The fusion (F) protein of bovine respiratory syncytial virus (BRSV) was expressed by using a baculovirus vector. Antigenicity was tested by immunofluorescence analysis with F-specific monoclonal and polyclonal antibodies. Antibodies to recombinant F protein raised in a rabbit neutralized BRSV and human respiratory syncytial virus infectivity when tested in a plaque reduction assay. The recombinant F protein was evaluated as a source of antigen in an enzyme-linked immunosorbent assay (ELISA), and this ELISA was compared with the virus neutralization (VN) test for detecting BRSV antibodies in 10 consecutive serum samples from four calves vaccinated with a live modified BRSV vaccine and from two nonvaccinated control calves. The ELISA with the baculovirus-expressed F protein as an antigen compared favorably with the VN test and is a rapid, sensitive, and specific method for detecting serum antibodies to BRSV.

Bovine respiratory syncytial virus (BRSV), a member of the genus Pneumovirus of the family Paramyxoviridae, is closely related to human respiratory syncytial virus (HRSV). It is a major cause of lower respiratory tract disease in calves between 1 and 3 months old (12). Seroepizootiologic studies have demonstrated that exposure of cattle to BRSV is widespread in many countries (1, 5). The envelope of BRSV contains two major glycoproteins, the attachment (G) protein and the fusion (F) protein. The G protein is involved in viral attachment (6). The F protein causes fusion of viral and cellular membranes and fusion of infected cells to surrounding cells. The F protein is synthesized as a precursor, F0 (68 kDa), which is proteolytically cleaved to yield two subunits, F1 (46 kDa) and F2 (20 kDa), that are disulfide linked (2). In BRSV-infected calves, the most immunogenic and protective viral antigen is the F glycoprotein, which evokes a strong antibody response and is a major target for cytotoxic T cells (7, 9). Antibodies to this protein can neutralize virus infectivity and prevent cell fusion (15). Therefore, there is a need for a source of the F protein free from other BRSV proteins to analyze its immunogenicity.

The baculovirus expression system provides a method for the production of large quantities of biologically active and antigenic eukaryotic proteins for both research and diagnostic applications. The F protein appears to be an ideal antigen for diagnostic purposes, as sera from BRSV-infected calves contain high levels of antibody to the F protein (11). Although the F protein of BRSV has been expressed by using a baculovirus vector (3), the detailed antigenic property and utility of the recombinant F protein as a diagnostic antigen had not yet been examined. Therefore, we expressed the F protein of BRSV strain A51908 in a baculovirus system and report the immunogenicity and utility of the recombinant F protein in a diagnostic enzyme-linked immunosorbent assay (ELISA).

Construction and characterization of the recombinant baculovirus F protein.

The construction and characterization of the recombinant baculovirus encoding the F protein of BRSV strain A51908 were accomplished by established methods (10, 14). A full-length cDNA of the BRSV F gene was originally cloned in the BamHI and XbaI sites of plasmid vector pGEM-7Z⁺ (Promega) (8). The F gene fragment was excised from a plasmid and was inserted into the BamHI and XbaI sites of the baculovirus transfer vector pVL1393 so that the F gene was under the control of the AcNPV (Autographa californica nuclear polyhedrosis virus) polyhedrin promoter. Transfer vector-recombinant DNA and linearized BaculoGold baculovirus DNA (PharMingen) were used to cotransfect Sf9 (Spodoptera frugiperda) cells according to the procedure described by the manufacturer. Briefly, Grace’s insect medium was replaced with 1 ml of transfection buffer A in a 60-mm tissue culture plate seeded with 3 × 10⁶ Sf9 cells. A mixture of 0.5 μg of linearized BaculoGold baculovirus DNA and 2 μg of recombinant plasmid transfer vector in 1 ml of transfection buffer B was added dropwise to the insect cells. Following 4 to 6 h of incubation at 27°C, the buffer was removed, the cell monolayer was washed, and fresh Grace’s insect medium supplemented with 10% fetal bovine serum was added. After 4 days, the extracellular virus was harvested, passaged three times, plaque purified, and used as a stock virus.

The BRSV F protein was expressed at high levels in Sf9 cells with a recombinant baculovirus vector. The baculovirus-expressed F protein was immunoprecipitated (4) with polyclonal antibodies raised against BRSV strain A51908, and the proteins were fractionated on a sodium dodecyl sulfate (SDS)–12.5% polyacrylamide gel and stained with Coomassie blue (Fig. 1). The baculovirus-expressed F protein was similar in size to the authentic BRSV F protein and had three polypeptides: F0 (68 kDa), F1 (46 kDa), and F2 (20 kDa). The presence of F0 indicated that a fraction of the baculovirus-expressed F protein was not cleaved by the insect cell proteases during posttranslational processing. Similar results were obtained with baculovirus-expressed HRSV and BRSV F proteins and were shown to be due to poor recognition of the cleavage site by insect cell proteases (3, 16).

FIG. 1.

Expression of F protein by recombinant baculovirus and confirmation of the authenticity of recombinant F protein. Lanes: A, molecular weight standards; B, recombinant baculovirus-infected cell lysate immunoprecipitated with BRSV-specific antiserum; C, BRSV-infected BTu cell lysate immunoprecipitated with BRSV antiserum; D, recombinant baculovirus-infected cell lysate immunoprecipitated with baculovirus-expressed F-specific antiserum; E, recombinant baculovirus-infected cell lysate; F, BRSV-infected BTu cell lysate; G, wild-type AcNPV-infected cell lysate; H, molecular weight standards. Numbers are molecular weights, in thousands.

The reactivities of baculovirus-expressed F protein with seven F-specific monoclonal antibodies to different epitopes on the F protein and polyclonal antibodies were examined by indirect-immunofluorescence tests. Immunofluorescence analysis was performed as previously described (13). All F-specific monoclonal antibodies and polyclonal antibodies recognized the baculovirus-expressed F proteins but did not recognize wild-type AcNPV baculovirus-infected insect cells (data not shown). This indicated that the baculovirus-expressed F protein was antigenically similar to the authentic BRSV F protein produced in mammalian cells.

Anti-F serum.

To assess the ability of the F protein produced by the recombinant baculovirus to induce antibodies which could react with the F protein present in BRSV-infected cells, an antiserum in rabbits was produced by using the F protein obtained from SDS-polyacrylamide gel electrophoresis-separated infected Sf9 cells. In brief, the infected Sf9 cell lysate was immunoprecipitated with polyclonal antibodies raised against BRSV strain A51908 and the proteins were resolved by SDS–12.5% polyacrylamide gel electrophoresis. After electrophoresis, the proteins were visualized by soaking the gel in ice-cold 100 mM KCl solution. The exact location of the recombinant F protein was determined by comparison with the relative mobility of the F protein of BRSV-infected bovine nasal turbinate (BTu) cell lysate and protein standards (Promega). The protein band was cut from the gel, homogenized in phosphate-buffered saline, and stored at −70°C. Each rabbit received one subcutaneous injection of approximately 200 μg of either F protein or wild-type AcNPV cell protein in Freund’s incomplete adjuvant on days 10, 20, and 30. Blood was collected 5 days after the last injection, and serum samples were stored at −20°C. The ability of the antiserum to recognize F protein was tested by immunoprecipitation. Preimmune serum and antiserum to the wild-type AcNPV-infected Sf9 cell proteins did not react with any viral or cellular proteins in BRSV-infected BTu cell lysates, whereas antiserum to the F protein specifically precipitated the BRSV F protein. The anti-F protein serum reacted in an ELISA with 12 BRSV strains and HRSV strain A2 (data not shown). The anti-F protein serum also gave bright immunofluorescence on acetone-fixed BRSV-infected BTu cells but not on uninfected BTu cells, which served as a negative control. This indicated that antiserum to the baculovirus-expressed BRSV F protein could be used as a diagnostic reagent for detection of BRSV antigen.

Neutralization assay using anti-F serum.

The ability of anti-F protein serum to neutralize the virus infectivity was tested by a 80% plaque reduction test (Table 1). Briefly, various twofold dilutions (1:2 through 1:1,024) of anti-F protein serum were mixed with 100 μl of tissue culture medium containing approximately 100 PFU of either HRSV strain A2 or BRSV strain FS1, incubated for 2 h at 37°C, and then inoculated onto BTu cells in six-well plates. After an adsorption period of 2 h, the virus-antibody mixture was removed and the cell monolayer was overlaid with minimal essential medium containing 0.8% methylcellulose and 6% fetal bovine serum. Seven days after incubation, virus plaques were counted after staining with crystal violet. Negative and positive control experiments were done with normal rabbit serum and with polyclonal antiserum, respectively. There were averages of 133 BRSV and 131 HRSV plaques in the wells not treated with the anti-F serum. In BRSV-infected wells treated with 1:2 to 1:32 dilutions of anti-F serum, there was more than an 80% reduction of BRSV plaques, whereas in HRSV-infected wells treated with anti-F serum, dilutions of 1:2 to 1:8 gave more than an 80% reduction of HRSV plaques. This indicates that 1:32 and 1:8 dilutions of anti-F serum gave almost complete neutralization (≥80%) of BRSV and HRSV infectivity, respectively. It appears that the baculovirus-expressed BRSV F protein contained important neutralizing epitopes and can induce antibodies in animals that not only neutralize BRSV infection but also can cross-neutralize HRSV infection to a lesser extent.

TABLE 1.

Neutralization of RSV with anti-F serum^a

Serum dilution	BRSV strain FS1		HRSV strain A2
	No. of plaques	% Reduction	No. of plaques	% Reduction
2	0	100	12	91
4	5	96	19	85
8	11	92	26	80
16	16	88	32	76
32	25	81	37	72
64	37	72	46	65
128	49	63	60	54
256	77	42	83	37
512	91	32	102	22
1,024	105	21	118	10
None (preimmune serum)	133	0	131	0

^aFor neutralization assays, twofold dilutions of anti-F serum were incubated with 100 PFU of BRSV strain FS1 or HRSV strain A2 for 2 h at 37°C. The mixture was added to BTu cell monolayers on six-well plates. Plaques were counted after 7 days. A ≥80% reduction of plaques is considered complete neutralization of virus infectivity. 

Vaccination of calves and serological testing.

Since the antibody response to BRSV is predominantly directed against the F protein and the F gene is highly conserved (>95%) among BRSV strains (7), the baculovirus-expressed F protein was used as an antigen in an ELISA for the detection of BRSV antibodies. Ten consecutive serum samples collected on days 0 (day of weaning), 7, 14, 28, 35, 42, 56, 84, 112, and 140 from six calves were tested in this assay. The ELISA results were compared with those of virus neutralization (VN) tests to investigate the consistency of the immunity level as measured by these two methods (Fig. 2). In this study, six purebred Angus calves from the Wye herd (Wye Research and Educational Center, Queenstown, Md.) were used. Calves were weaned at an average age of 205 days. All calves were found to be BRSV free by both virus isolation and VN tests. Four calves (vaccinated group; calves 63, 75, 84, and 90) were inoculated intramuscularly with an attenuated BRSV vaccine strain as recommended by the manufacturer (Pfizer Animal Health). A booster vaccination was given 4 weeks later. The two remaining calves served as nonvaccinated controls. On the day of weaning, the nasal mucosa of each calf was swabbed for virus isolation. Nasal swabs were swirled in a transport medium, and the eluates were inoculated onto BTu cells for virus culturing. The specimens were considered negative if no cytopathic effects (syncytium formation) developed within two subpassages.

FIG. 2.

ELISA and VN test results obtained with serum samples from four vaccinated calves. Calves were inoculated intramuscularly on day 0 with an attenuated BRSV vaccine, and a booster vaccination was given after 4 weeks. The ELISA and the VN test were performed as described in the text.

Serum was collected from each calf for the determination of ELISA titers and titers of VN antibody against BRSV on 10 occasions: days 0 (day of weaning), 7, 14, 28, 35, 42, 56, 84, 112, and 140. All serum samples were heated to 56°C for 30 min and stored at −20°C. For the ELISA, 96-well Immulon-2 plates (Dynatech) were coated by overnight treatment at 4°C with approximately 100 ng of the F protein in 0.1 M sodium bicarbonate (pH 9.5) per well. After adsorption of the antigen, the plates were saturated with blocking buffer (5% skim milk powder in TBS-T [144 mM NaCl in 25 mM Tris-HCl {pH 7.6}–0.1% Tween 20]) for 30 min at room temperature. Serum samples at a single test dilution of 1:10 in blocking buffer (100 μl) were incubated in antigen-coated wells for 2 h at 37°C. After a wash with TBS-T, a 1:150 dilution in blocking buffer (100 μl) of affinity-purified horseradish peroxidase-conjugated goat anti-bovine immunoglobulin M and immunoglobulin G antibodies (Kirkegaard & Perry Laboratories) was added for 1 h at 37°C. ABTS [2,2′-azino-di-(3-ethylbenzothiazoline sulfonate)] substrate (100 μl) was added after a wash with TBS-T. The A₄₁₀ was recorded with an automated spectrophotometer (Titertek Multiscan; Flow Laboratories, Vienna, Va.). The cutoff value for a positive test was taken as the mean absorbance plus 3 standard deviations for a panel of five control serum samples with no detectable virus neutralization antibodies.

For the VN test, serum samples were serially diluted twofold in microtitration plates. BRSV strain A51908 (100 50% tissue culture infective doses) was added to each serum dilution. After incubation for 1 h at 37°C, cells were added in amounts sufficient to form a monolayer. The plates were incubated for 5 days. Cytopathic effects were then examined microscopically. The reciprocal of the highest serum dilution that completely inhibited cytopathic effects was recorded as the VN titer. A titer of >4 was judged sufficient to consider the calf BRSV seropositive. The mean ELISA absorbance value for five VN-negative, prevaccination serum samples was 0.186 (standard deviation, 0.03). Serum samples were considered positive for BRSV when the A₄₁₀ was >0.293 (mean ± 3 standard deviations). The results of the ELISA and the VN test for the vaccinated calves are presented in Fig. 2. There was agreement between the ELISA and the VN test. All VN-positive serum samples from vaccinated calves were positive in the ELISA. Serum samples collected on days 0 and 7 were negative in both the ELISA and the VN test. An antibody response was detected for the first time by the ELISA on day 14 in all four vaccinated calves. The VN test first detected antibodies in calves 84 and 90 on days 14 and 21, respectively, but failed to detect antibodies in two other calves (63 and 75) until day 35. In all vaccinated calves, there was a significant rise in antibody titer after booster vaccination on day 28. The antibody titers in postvaccination sera were higher between 35 and 84 days in both tests. The antibody titers dropped sharply after day 56 in all vaccinated calves and were still detectable on day 140 in all the vaccinated calves. All serum samples from the two nonvaccinated calves were negative for BRSV antibodies in both the ELISA and the VN test (data not shown). The VN test gives an estimate of the level of protection against BRSV infection, but it is unsuitable for screening large numbers of serum samples for epidemiological studies or for early and rapid detection of antibodies to BRSV. The ELISA is sensitive and rapid in detecting BRSV antibodies produced early in the immune response and is suitable for whole-herd testing. The ELISA with F as an antigen is specific, as it measured specific antibody response to the F protein of BRSV.

Generally, a live attenuated BRSV vaccine is recommended for calves at 6 months of age, with a booster 3 or 4 weeks after the first dose. Because maternal antibodies suppress serum and mucosal antibody responses of all isotypes, serodiagnostic testing by ELISA may be less sensitive for diagnosing BRSV infection in calves younger than 3 months of age. Antibodies from naturally BRSV-infected animals cannot be distinguished from antibodies from vaccinated animals in an ELISA, because the antibodies from both groups of animals recognize the same F protein of BRSV. However, in vaccinated animals, the efficacy of the vaccine to induce an adequate immune response and the level of antibody to the F protein can be evaluated by use of an ELISA. The ELISA is also useful in the diagnosis of BRSV infection in seronegative calves and nonvaccinated animals.

The baculovirus-expressed BRSV F protein not only provides large quantities of pure antigen free from other BRSV proteins for use in an ELISA but also offers the advantage of measuring specific antibody response to the F protein of BRSV. Since the F protein is the major immunogenic and protective viral antigen, which induces both cell-mediated immunity and a high level of protective neutralizing antibodies early in the infection, the detection of a specific antibody response to the F protein indicates the level of protective antibodies produced against BRSV infection. It also indicates a recent vaccination or a past clinical infection.

We conclude that the availability of large quantities of baculovirus-expressed BRSV F protein offers an opportunity to study antigenic and immunogenic characteristics of the F protein as well as to use this recombinant protein as a diagnostic reagent in ELISAs. Since F protein is an important target of the host immune response to RSV infection, the recombinant BRSV F protein has the potential for use as a subunit vaccine against BRSV infection.