INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Vesicular stomatitis virus (VSV) is the prototype of the family Rhabdoviridae and infects a broad range of animals, including cattle, horses, and swine (12). VSV infection of livestock is associated with significant disease, including vesicular lesions around the mouth, hoofs, and teats, and with loss of beef and milk production (3, 38). Although rarely fatal, VSV infection mimics early symptoms caused by a notorious veterinary pathogen of horses and cattle, foot-and-mouth disease virus (38). VSV infection of livestock occurs periodically within the United States, is associated with one of two serotypes of VSV (Indiana, VSV_I, or New Jersey, VSV_NJ), and is usually not widespread. However, major VSV epizootics occur approximately every 10 to 15 years within the United States, with the last two having occurred in 1982-1983 and 1966 (29, 39). In 1997, 183 confirmed cases of VSV_I were reported in four states (36). Arthropod vectors, such as phlebomite sandflies (8, 27) and Aedes mosquitoes (34), have been found to harbor VSV_I in nature and likely participate in the spread of the virus from animal to animal and possibly from animals to humans.

Naturally occurring human infections with VSV are rare. However several cases of VSV infection have been reported for individuals directly exposed to infected livestock (6) and for researchers directly exposed within laboratory environments (12, 13, 25). Most VSV infections are thought to be asymptomatic in humans; however, febrile illness, including chills, myalgia, and nausea, has been reported in infected individuals (5, 10, 12, 13). A single case of encephalitis associated with VSV_I infection in a 3-year-old boy was reported in 1988 (28).

The seroprevalence of VSV antibodies within the general human population is extremely low except in limited regions of enzooticity in Georgia (VSV_NJ) (11) and Central America (VSV_I and VSV_NJ) (1). Seroprevalence is also moderately high in individuals with a high risk for exposure (i.e., laboratory workers who work with the virus directly and veterinarians and ranchers exposed to infected livestock). An example of high seroprevalence of VSV antibodies within a human population (94% seropositive) was reported for a single rural Panamanian community where VSV infection was enzootic in the surrounding wildlife (35). The surrounding rural Panamanian communities had seroprevalence rates ranging from 3 to 54% in adult lifetime residents, and urban Panamanian populations had no seroprevalence of VSV antibodies (35). The seroprevalence of VSV antibodies has also been examined in high-risk groups for VSV exposure during epizootics. A report on human infection during the 1965 epizootic included a study of 41 individuals at high risk for exposure to VSV (6). Eight of the 41 individuals (20%) had serological evidence of VSV infection. This and other studies (11, 29, 39) indicate limited seroprevalence of antibodies to VSV in the categories with the highest risk of exposure and demonstrate an absence of antibodies to VSV in the general population. The low VSV seropositivity in the general population and the lack of serious pathogenicity in humans are advantages in the potential use of recombinant VSV-vectored vaccines in humans.

In addition to the low seroprevalence of VSV antibodies in the general population, other characteristics of VSV suggest that recombinant VSVs expressing foreign viral glycoproteins would be very good vaccine candidates. VSV elicits strong humoral and cellular immune responses in vivo (3, 37, 41) and naturally infects at mucosal surfaces. Mucosal immunization is a less invasive route of immunization and has been shown to elicit both mucosal and systemic immunity (17, 21, 22). Additionally, recombinant VSVs are able to accommodate large inserts and multiple genes in their genomes. This ability to incorporate large gene inserts in replication-competent viruses offers advantages over other RNA virus vectors, such as those based on alphaviruses (2, 20, 40) and poliovirus (23). Furthermore, like alphaviruses, VSV grows to very high titers in vitro, and VSV infection shuts down host cell protein synthesis. This inhibition of host cell protein synthesis allows rapid identification of viral proteins in infected cell lysates and facilitates rapid purification of large amounts of virus and viral proteins.

VSV recombinants would have advantages over other live recombinant vaccine vectors as well. Compared to large, complex genomes of viruses from the Poxviridae family, the VSV genome is relatively simple, more fully understood, and easier to manipulate. VSV's single-stranded RNA genome does not undergo reassortment and therefore lacks the potential of reassorting with wild-type viruses in vivo. Furthermore, VSV replicates within the cytoplasm of infected cells and does not undergo genetic recombination.

Previously we reported that intranasal vaccination of mice with recombinant VSV expressing influenza virus hemagglutinin (HA) protein (VSV-HA) was able to protect mice from lethal intranasal influenza virus challenge (30). Prior to challenge, vaccinated mice had high levels of serum neutralizing antibodies to influenza virus A/WSN. After challenge, the mice were completely protected from death, weight loss, pneumonia-associated pathology, and influenza virus replication in the lungs. Although the mice were completely protected from influenza virus challenge, there was pathogenesis associated with the initial VSV-HA vaccination. Initial pathogenesis was indicated by inactivity, ruffled coats, weight loss, and VSV-HA replication in the lungs (assayed 48 h after inoculation). The vector-associated pathogenesis remained a significant concern in the development of a nonpathogenic VSV-vectored vaccine. In addressing this concern, we reported the further attenuation of recombinant VSVs containing truncations within the cytoplasmic tail of the glycoprotein (VSV-CT1 and VSV-CT9) (30). Here we report on the complete attenuation of a VSV-CT1 recombinant expressing influenza virus HA (CT1-HA), its immunogenicity, and its efficacy in protecting mice from a lethal influenza virus challenge. Additionally, we report the construction, recovery, and characterization of a recombinant VSV containing a complete substitution of the VSV G protein with influenza virus HA (G-HA). We also examine and report the absence of vector-associated pathogenesis and the immunogenicity and protective efficacy of recombinant G-HA in the mouse model system.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of plasmids and recovery of recombinant viruses. (i) pBSIIKS()HA. The HA gene (influenza virus A/WSN HA protein) was PCR amplified from pVSV-HA (16) with the following primers (restriction sites are underlined): XmaI HA(+), 5'-GAGCCCGGGAAAATGAAGGCAAAAC-3' (containing an XmaI restriction enzyme sequence followed by 5' sequence from the HA gene), and XhoI ssHA(), 5'-CCACTCGAGATCGATCTCTG TTAGTTTTTTTCATACCTCAGATGCATATTCTGC-3' (containing an XhoI restriction enzyme site followed by reverse complements of the VSV transcriptional START/STOP sequence and the 3' HA gene sequence). The ~1.7-kb HA PCR fragment was digested with restriction enzymes XmaI and XhoI, agarose gel purified, electroeluted, and ligated with T4 DNA ligase into pBSIIKS() vector (product no. 212208; Stratagene, La Jolla, Calif.) which had been digested with XmaI and XhoI. C600 competent bacteria were transformed with ligation reactions and plated on Luria-Bertani medium agar plates containing ampicillin at 100 µg/ml. Colonies were screened by PCR for the 1.7-kb HA insert. Positive colonies were expanded in 250- to 500-ml Luria-Bertani medium-ampicillin cultures, and plasmids (pBSIIKS()HA) were purified from these cultures with a Plasmid Maxi kit (product no. 12163; Qiagen Inc., Santa Clarita, Calif.). Expression of HA was confirmed by indirect-immunofluorescence screening of BHK cells transfected with pBSIIKS()HA and infected with vTF7-3, recombinant vaccinia virus expressing T7 polymerase (7).

(ii) Recombinant CT1-HA. Plasmid pVSVCT-1XMN (31) (containing the VSV antigenome with an additional transcriptional STOP/START sequence and an XmaI restriction enzyme site downstream of the CT1-G) was digested with restriction enzymes XmaI and NotI and purified. The 1.7-kb HA fragment was PCR amplified from pBSIIKS()HA plasmid with primer XmaI HA(+) and primer NotI HA(), 5'-CTCGCGGCCGCTCAGATGCATATTCTGC-3' (containing a NotI restriction enzyme site and reverse complement sequence of the 3' HA gene). The HA-PCR fragment was digested with restriction enzymes XmaI and NotI, purified, and ligated to the XmaI and NotI sites of the pVSVCT-1XMN vector. Plasmid pVSVCT1-HA was isolated and used in the standard VSV recovery system (19) to generate the recombinant CT1-HA virus.

(iii) Recombinant VSVG-HA. Plasmid pVSVMXA2XN2 was digested with restriction enzymes XmaI and XhoI to generate a VSVG vector fragment which was purified from the excised G fragment. The HA gene, trailed by the VSV transcriptional STOP/START sequence, was excised from pBSIIKS()HA by digestion with restriction enzymes XmaI and XhoI, purified, and ligated to the VSVG vector fragment. A plasmid confirmed for the HA insert and VSV G deletion (pVSVG-HA) was used in a modified VSV recovery system previously described by Schnell, et al. (33) to generate the recombinant VSVG-HA virus.

(iv) Recombinant VSVG. Plasmid pVSV-XMN was digested with restriction enzymes MluI and NheI to generate a VSVG vector fragment. The VSVG fragment was made blunt with T4 DNA polymerase in the presence of dATP, dCTP, dGTP, and dTTP and ligated with T4 DNA ligase. Transformations, PCR screenings, and expansions of positive colonies (those lacking VSV G) were performed as described above. A plasmid confirmed for the VSV G deletion (pVSVG) was used in a modified VSV recovery system previously described by Schnell et al. (33) to generate the recombinant VSVG virus.

Indirect immunofluorescence assays (IFA). Immunofluorescence assays were performed on baby hamster kidney cells (BHK-21; American Type Culture Collection) or BHK-G cells (expressing VSV G [33]) as previously described (32). The cells were either transfected with pBSIIKS()HA plasmid (10 µg/plate) and infected with vTF7-3 (multiplicity of infection [MOI] = 10) or infected with supernatants from VSV recovery assays. The cells were fixed in 3% paraformaldehyde, washed in phosphate-buffered saline (PBS) and incubated with primary antibody (mouse monoclonal antibody 523/6 directed to influenza virus A/WSN HA protein; from Robert Webster) at 1:100 dilutions in PBS-glycine with 5 mg of bovine serum albumin/ml added. A second incubation was done in the presence of secondary antibody (fluorescein isothiocyanate-conjugated goat anti-mouse antibody) at 1:50 in PBS-glycine-bovine serum albumin.

Metabolic labeling of cells and recombinant viruses. BHK or BHK-G cells (in 3.5-cm-diameter dishes) were infected with recombinant viruses (MOI = 1 to 10) and incubated at 37°C in Dulbecco's modified Eagle medium (with L-glutamine, sodium pyruvate, high glucose, and high bicarbonate [3.7 g/liter]) (product no. 56 499; JRH Biosciences, Lenexa, Kans.) (DMEM) containing 5% fetal bovine serum (FBS) and 100 U of penicillin-streptomycin (PS)/ml for 1 h. The medium was aspirated and replaced with 1 ml of DMEM-5% FBS-PS, and the plates were incubated for an additional 3 h. The cells were washed twice with PBS, and 1 ml of labeling medium (10% DMEM, 90% DMEM minus methionine and cysteine, 100 µCi of ³⁵S translabel [product no. NEG772, Easy Tag EXPRESS protein labeling mix; NEN Life Sciences, Boston, Mass.]) was added to each plate. For radiolabeled cell extracts, the cells were incubated for 1 h, washed twice in PBS, and lysed in 500 µl of ice-cold detergent lysis buffer (1% Nonidet P-40, 0.4% deoxycholate, 66 mM EDTA, and 10 mM Tris-Cl, pH 7.4) for 30 min on ice. The lysates were transferred to cold 1.5-ml Eppendorf tubes and centrifuged at 16,000 × g for 2 min at 4°C. The supernatants were transferred to fresh, cold 1.5-ml Eppendorf tubes and stored at 20°C. For radiolabeled viruses, the cells were incubated at 37°C overnight. The supernatants, containing radiolabeled virus, were collected in 15-ml conical tubes and clarified by spinning at 1,250 × g for 10 min at room temperature. The clarified supernatants were layered onto 4 ml of 20% sucrose in Beckman polyallomer centrifuge tubes (13 by 51 mm) and centrifuged at 38,000 rpm for 1 h at 4°C in a Beckman SW50.1 rotor in a Beckman L8-M ultracentrifuge. The pellets were resuspended in 250 to 500 µl of Tris-EDTA, pH 7.6, and stored at 20°C until analyzed. Sample buffer containing 2-mercaptoethanol was added to the purified viruses to a 1× final concentration, and samples were boiled for 3 to 5 min before being loaded onto sodium dodecyl sulfate (SDS)-polyacrylamide gels (10% acrylamide) for analysis.

Radiolabeled immunoprecipitation assays. Primary antibody was added to 200 to 300 µl of radiolabeled cell extracts at 1:100 dilutions, and SDS was added to 0.2% final concentration. The antibody-extract solutions were incubated at 37°C for 45 min. Samples were precipitated by adding 30 µl of fixed Staphylococcus aureus (Pansorbin; Calbiochem) and incubating the mixture at 37°C for 30 min. Precipitates were pelleted by centrifugation, washed in 1.5 ml of RIPA buffer (1% Nonidet P-40, 1% deoxycholate, 0.1% SDS, 150 mM NaCl, and 10 mM Tris-Cl, pH 7.4) three times, and pelleted again. The pellets were resuspended in 30 to 40 µl of sample buffer containing 2-mercaptoethanol, boiled for 3 to 5 min, and pelleted again. The supernatants were carefully removed and loaded onto SDS-10% polyacrylamide gels.

Viruses and inoculum. Recombinant viruses were grown and titers were determined as previously described for VSV-HA (30). The plaque-purified recombinant CT1-HA was grown and its titers were determined on BHK cells, and the plaque-purified recombinants VSVG-HA and VSVG were grown on BHK-G cells (33) and passaged once through BHK cells, and titers were determined on BHK-G cells. All recombinants were thawed and diluted with DMEM to appropriate titers immediately prior to inoculation. Influenza virus A/WSN/33 (H1N1) (1.4 × 10⁶ PFU/ml), used for challenges and neutralization assays, was grown in MDBK cells in DMEM-10% FBS-PS, and titers were determined by standard plaque assay on MDBK cells with 1% agarose-1× DMEM (serum free)-2 µg · ml¹ acetylated trypsin (product no. T-6763; Sigma, St. Louis, Mo.) overlays. The WSN virus was thawed immediately prior to challenge, diluted 1:10 in DMEM, and administered in a 50-µl total volume (~7 × 10⁴ PFU/mouse). Viruses or DMEM were administered intranasally as described below.

Inoculation of mice. Five- to 6-week-old female BALB/c mice from Charles River Laboratories were housed in filter-isolette cages upon arrival. The mice were inoculated no earlier than 4 days after arrival. Prior to inoculation (day 0), the mice were lightly anesthetized with Methoxyflurane (Mallinckrodt Veterinary, Inc., Mundelein, Ill.) and marked by ear punch. Twenty-five microliters of inoculum were delivered intranasally by 200-µl pipette to the anesthetized mice, and the mice were weighed in a plastic beaker to ±0.02 g. Boosts were administered in an identical fashion with viruses of equal kind, titer, and volume on day 21, unless otherwise indicated. Challenge was also administered as described but in a total volume of 50 µl per mouse on day 35, unless otherwise indicated. The mice were observed and weighed unanesthetized on a daily basis.

Neutralization assays. Blood samples were collected from anesthetized mice by retro-orbital bleeds and allowed to clot at room temperature. The clots were removed, and samples were centrifuged in a TOMY MTX-150 centrifuge (TMA-11 fixed-angle rotor) at 4°C for 10 to 15 min at 5,500 rpm. The clarified sera were transferred to sterile Eppendorf tubes and heat inactivated at 56°C for 45 to 60 min. The heat-inactivated sera were diluted with PBS in serial twofold dilutions in 96-well plates. Generally, 50 µl of serum was mixed with 50 µl of PBS and 50 µl was transferred for serial dilutions. For analysis of titers of neutralizing antibodies to influenza virus, equal volumes (50 µl) of influenza virus A/WSN (~200 PFU) were then added to the diluted sera and mixed. The 96-well plates containing virus and sera were incubated at 37°C for 30 to 45 min. Approximately 2,000 MDBK cells in DMEM-20% FBS-PS (100-µl total volume) were added to each well. The plates were incubated at 37°C, 5% CO₂, for 2 to 3 days. In each assay, the sera were diluted and analyzed in duplicate, and each assay was repeated at least once. Neutralization titers are those dilutions which correspond to complete inhibition of virus-associated cytopathic effect (CPE). Neutralization titers are reported as averages from all assays. For neutralization to VSV, the assays were identical except ~200 PFU of recovered wild-type VSV (VSVrwt) was added to each well instead of influenza virus A/WSN and ~200 to 500 BHK cells in DMEM-10% FBS-PS were added to each well instead of MDBK cells.

Recovery of viruses from tissue. Two days after inoculation or 2 days after challenge, two mice from each group were heavily anesthetized. Blood was collected via cardiac puncture and transferred into K₂EDTA-treated 2-ml Microcontainer tubes (product no. 365974; Becton Dickinson, Franklin Lakes, N.J.) and stored at 4°C. The lungs were excised bilaterally, rinsed externally in sterile PBS, and placed in sterile, preweighed 2-ml Nunc cryotube vials. The cryotube vials were immediately dropped in liquid nitrogen and later transferred to 80°C storage. Plasma samples were prepared by centrifugation as described earlier for serum samples. The lungs were thawed, suspended in 9 volumes (wt/vol) of DMEM-2.5% FBS-PS, and homogenized in Dounce homogenizers. To assay for recoverable virus from initial infections with recombinant viruses, BHK or BHK-G cells ~80% confluent in 6-well plates were infected with 40 µl of plasma or 200 µl of the 10% lung suspensions per well, incubated for 1 h at room temperature, and washed in PBS. Two milliliters of DMEM-5% FBS-PS was added to each well, and the plates were incubated at 37°C and observed daily for 3 days for virus-associated CPE. The procedure was the same for assaying for recoverable virus from challenge except MDBK cells were used instead of BHK or BHK-G cells and DMEM-10% FBS-PS was added after the PBS wash. The remainder of the 10% lung suspensions were stored at 80°C for further analysis.

1

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

To determine if attenuated VSV vectors could be employed for vaccine applications, we generated three new VSV recombinants. The first, designated VSVCT1-HA (CT1-HA), encodes a VSV G protein with 28 of its 29 cytoplasmic amino acids deleted (31) and expresses an additional protein, the influenza virus A/WSN HA (WSN HA), between the truncated G gene and the L gene (Fig. 1). Because the VSV-CT1 mutant is greatly attenuated (30, 31), we anticipated attenuation of the CT1-HA recombinant also. A second recombinant, designated VSVG-HA (G-HA), has a complete deletion of the VSV G gene and expresses the WSN HA gene from the site of the G deletion between the VSV M and L genes (Fig. 1). We anticipated that this virus would not be able to propagate in culture without VSV G supplied in trans. A VSVG (G) virus with VSV G deleted was prepared as a control. Our standard recovery system was used to make the first recombinant (19), and a modified recovery system in which VSV G protein is provided in trans was used to recover the two recombinants with VSV G deleted (33).

View larger version (11K): [in this window] [in a new window]   FIG. 1.   Recombinant VSV-influenza virus A/WSN constructs. A schematic representation of recombinant viruses indicating gene order is shown 3' to 5' on the negative-strand genomic RNA. Each intergenic region contains a transcriptional STOP/START signal recognized by the VSV polymerase (32). Restriction enzyme sequences used for constructing cDNAs of the recombinants are also shown.

Growth characteristics of the CT1 and G recombinants. Upon recovery of recombinant CT1-HA, G-HA, and G viruses, we examined the growth of each virus in culture. CT1-HA, containing a truncated VSV G protein, could be grown and passaged on BHK cells, but both G-HA and G viruses had to be maintained on BHK-G cells induced for G expression (33). Infectivity of G-HA and G viruses was lost after two low-multiplicity passages through BHK cells. However, it is possible to passage G-HA and G recombinants through BHK cells once and recover a low titer of infectious virus from the supernatant. Although the levels of G protein in the first-passage virus were below the limits of detection in IFA, radiolabeled immunoprecipitation assays, and SDS-polyacrylamide gel electrophoresis (PAGE) analysis of radiolabeled cell extracts and purified viruses, these recombinants must contain a residual amount of G protein, since neutralization of the infectivity was accomplished with antibody to VSV G protein.

CT1-HA and G-HA recombinants express HA. Recombinant viruses were examined initially for protein expression in infected cells by IFA. BHK cells infected with recombinant CT1-HA or G-HA expressed HA at the cell surface as detected by IFA with either polyclonal rabbit serum raised to influenza virus A/WSN (16) or mouse monoclonal antibody raised to WSN HA. BHK cells infected with recombinant G expressed VSV N but lacked HA expression as expected. BHK cells infected with G-HA and G lacked detectable VSV G at the cell surface, while VSV G was readily detected on CT1-HA-infected cells (data not shown).

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Immunoprecipitations of proteins from 35S-labeled cell extracts. Cells infected with recombinant VSVs were metabolically labeled, lysed, and immunoprecipitated with mouse monoclonal antibodies specific for influenza virus A/WSN HA protein (A, lanes 2 to 8) or for VSV G protein (A, lane 1, and B). VSVG and VSVG-HA were from BHK-G cell extracts; all other samples were from BHK cell extracts. VSVrwt and VSV-HA recombinants were run on all gels for comparison of immunoprecipitated proteins. Mock, immunoprecipitations of uninfected BHK cell lysates.

Recombinant G-HA incorporates large amounts of HA protein into virions. Previous studies showed that the influenza virus HA protein, the major influenza virus spike glycoprotein in virions, was incorporated into virions of VSV-HA, although at lower efficiency than with VSV G (16). To determine if the G truncation or the absence of VSV G affected the relative amount of HA incorporation into VSV virions, cells infected with VSV-HA, CT1-HA, G-HA, or G were labeled with [³⁵S]methionine and [³⁵S]cysteine overnight. The cell supernatants were collected when all cells showed signs of virus-associated CPE. Viruses were purified from these supernatants, and the radiolabeled proteins were analyzed by SDS-PAGE. Signal intensities of nucleocapsid and phosphoprotein (N-P) were quantitated for each sample on a phosphorimager after background subtraction, and a second gel was run with volumes of each sample which were normalized for the content of N-P. The phosphorimage of this second SDS-PAGE gel is shown in Fig. 3.

View larger version (33K): [in this window] [in a new window]   FIG. 3.   Sucrose-purified recombinant virions. Cells infected with recombinant viruses were metabolically labeled with 35S translabel. Supernatants were collected, and viruses were purified by centrifugation through 20% sucrose. The volumes loaded were normalized for N-P protein content. Lane 1, 1% of total VSV-HA virus from a 3.5-cm-diameter dish of BHK cells; lane 2, 18% of total CT1-HA virus from a 3.5-cm-diameter dish of BHK cells, lane 3, 14% of total VSVG-HA virus from a 3.5-cm-diameter dish of BHK-G cells; lane 4, 13% of total VSVG virus from a 3.5-cm-diameter dish of BHK-G cells.

HA is found in the virion envelope. To demonstrate that HA was present specifically in the virion envelope and to examine the morphology of the recombinant virions lacking VSV G protein, purified VSVrwt and G-HA were adsorbed onto carbon-coated grids and incubated with HA monoclonal antibody followed by labeling with gold-conjugated goat anti-mouse immunoglobulin G as described precisely by Schnell et al. (33). The grids were negatively stained with 2% phosphotungstic acid and examined on a Zeiss EM910 electron microscope. Strong labeling of HA proteins on the virion surface of recombinant G-HA was clear in greater than 70% of virions, and all virions had the characteristic bullet shape expected for VSV (Fig. 4B). Although the background labeling is high, it is clear that no label is specifically associated with the surface of VSVrwt in the control (Fig. 4A).

View larger version (89K): [in this window] [in a new window]   FIG. 4.   Immunogold labeling of sucrose-purified recombinants. Electron micrographs of recombinant VSVrwt (A) and VSVG-HA (B) bound to carbon-coated grids, labeled with mouse monoclonal antibody to HA and secondary gold-conjugated antibody, and negatively stained.

Recombinants CT1-HA and G-HA are attenuated in mice. Once we had demonstrated the incorporation of HA into the recombinant CT1-HA and G-HA virions, we examined their potential as attenuated vaccines for protection from influenza virus in a mouse model system. Six-week-old female BALB/c mice were inoculated intranasally with recombinant viruses at 5 × 10⁴ PFU/mouse. The mice were observed and weighed daily, and weight loss was used as an indication of pathogenesis as previously reported (30). The mice inoculated with recombinant VSV-HA began losing body weight the day after inoculation, as seen previously (30). These mice lost 17% of their initial body weight before they began regaining weight (Fig. 5A). Weight loss corresponded with decreased activity and poor grooming, all of which indicate vector-associated pathogenesis. In contrast, the mice receiving CT1-HA, G-HA, or G recombinants did not lose significant amounts of weight (Fig. 5) compared to mice receiving medium alone. Also, no change in activity or grooming was observed for these mice in comparison to mice receiving medium alone. The lack of significant weight loss and the lack of decreased activity and grooming indicate that the CT1-HA and G-HA vectors are attenuated in the mouse model. The mice were inoculated a second time with identical titers of virus 21 days after the initial inoculation. Weights, activity, and grooming were unaffected in all groups in the days following the second inoculation.

View larger version (29K): [in this window] [in a new window]   FIG. 5.   Attenuated recombinants provide protection from lethal influenza virus challenge. Average daily weights of mice inoculated with VSV recombinants and challenged with influenza virus A/WSN are shown. (A) Mice were inoculated intranasally with VSV-HA (n = 5) or CT1-HA (n = 5) at 5 × 104 PFU/mouse on day 0. The mice were boosted with an equal amount of inoculum on day 21. The mice were challenged with a lethal dose of influenza virus A/WSN on day 35. (B) Conditions were the same as for panel A, except the mice were inoculated and boosted with VSVG (n = 4) or VSVG-HA (n = 5) at 5 × 104 PFU/mouse. The error bars indicate ±0.5 × standard deviation.

Recombinant viruses are immunogenic in mice. To examine the immunogenicity of the recombinants, mice were bled prior to initial inoculations, second inoculations, and challenges, and surviving mice were bled after challenge. Serum samples were pooled, unless otherwise indicated, and screened for neutralizing antibodies to VSV and influenza virus A/WSN. Serum neutralizing antibody titers correspond to titers which provided complete inhibition of virus-associated CPE when virus was preincubated with serum before the addition of susceptible cells (Table 3). All sera collected before inoculation had no titers of antibody to either VSV or influenza virus. Eighteen days after initial inoculation, mice receiving VSV-HA and CT1-HA had titers of neutralizing antibody to both VSVrwt and influenza virus in the serum. Mice inoculated with G-HA and G had no measurable titers of neutralizing antibody to either VSVrwt or influenza virus. Thirty-two days after the initial inoculation, which was also after a second inoculation (boost), mice receiving the G-HA recombinant had low-level titers of neutralizing antibody to influenza virus. No titers of antibody to VSVrwt were detected in G-HA- or G-inoculated mice. No titers of antibody to influenza virus were detected in G-inoculated mice. Titers of neutralizing antibody to influenza virus in the sera of mice surviving challenge reached 1:4,096.

Efficacy of recombinants in generating protective immunity. In an initial experiment to examine the efficacy of recombinant VSVs expressing influenza virus HA in protecting mice from a lethal influenza virus challenge, mice were inoculated with recombinant virus at 5 × 10⁴ PFU/mouse. The mice were challenged 21 days after this single, initial immunizing inoculation. Fifty microliters of a 1:2 dilution of influenza virus A/WSN in DMEM (~3.5 × 10⁵ PFU/mouse) was administered for challenge. Mice which had been inoculated with CT1-HA or VSV-HA, which also had titers of neutralizing antibody to influenza virus A/WSN in the serum 18 days after inoculation, were completely protected from lethal challenge as indicated by 100% survival. Mice inoculated with the G-HA recombinant, which lacked titers of neutralizing antibody in the serum at day 18 after initial inoculation, were partially protected from lethal challenge; only two of five mice survived.

Efficacy of boost. Because the preceding experiments did not examine the direct efficacy of a second inoculating dose, a third set of experiments was conducted. Three groups of eight mice each were inoculated with VSV-HA, CT1-HA, or G-HA (10⁴ PFU/mouse). Eighteen days after inoculation sera were collected and assayed for titers of neutralizing antibody to influenza virus A/WSN. On day 21 after initial inoculation four mice from each group were inoculated with DMEM and the other four mice from each group were boosted with an equal titer of initial inoculum. Sera were collected on day 32 and assayed for neutralizing antibodies to influenza virus A/WSN. The serum neutralizing antibody titers are reported in Table 4 and indicate that boosting offers no significant advantage over single inoculations for CT1-HA or VSV-HA immunizations. Antibodies raised to the G protein encoded by these vectors probably prevent infection and could explain the lack of boosting responses. In contrast, mice receiving two inoculations of the G-HA virus had higher concentrations of serum neutralizing antibodies to influenza virus A/WSN than mice receiving a single inoculation at day 32. These data indicate that boosting is efficacious in the instance of G-HA immunization, when the vector does not encode G protein.

Attenuation of CT1-HA to confer systemic immunity on mice. The ability of CT1-HA to provide complete protection from lethal challenge following a single inoculation in the absence of vector-associated pathogenesis led to further characterization of the vector's immunogenicity. Thirty-six mice were divided into six groups of six mice each and were inoculated with 10, 100, or 1,000 PFU/mouse of either CT1-HA or VSV-HA. Sera were collected and screened 2 and 4 weeks after inoculation. No mice receiving 10 PFU of CT1-HA virus had neutralizing antibody titers to influenza virus A/WSN at 4 weeks postinoculation (Table 5). In contrast, mice receiving 10 PFU of VSV-HA had various levels of serum neutralizing antibodies to influenza virus A/WSN at 4 weeks postinoculation (Table 5). The neutralizing antibody titers of mice receiving 100 PFU of CT1-HA are comparable to those obtained in mice receiving 10 PFU of VSV-HA. This pattern of a 10-fold-greater amount of CT1-HA necessary to elicit a neutralizing antibody response similar to that generated by VSV-HA is consistent throughout the data (Table 5). This demonstrates that CT1-HA is attenuated in immunogenicity in the mouse model. Neutralizing titers at 2 weeks postinoculation were less than those at 4 weeks postinoculation for all mice and therefore are not shown.

Long-term immunity in CT1-HA- and VSV-HA-inoculated mice. The longevity of the humoral immune response induced by these vectors has also been examined. Mice inoculated with a single dose of VSV-HA or CT1-HA (10⁴ PFU/mouse) have now been observed for 4 months after an initial intranasal inoculation. These mice still have high levels of serum neutralizing antibodies to influenza virus. CT1-HA-inoculated mice have neutralizing antibody titers ranging from 1:512 to 1:2,048, and VSV-HA-inoculated mice have neutralizing titers ranging from 1:1,024 to 1:4,096. These mice have not yet been challenged with influenza virus, but in all previous experiments these levels of antibody have correlated with 100% protection and little or no weight loss after challenge.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We report here the successful attenuation of VSV vectors (VSV-CT1 and VSVG) which produce no measurable signs of pathogenesis as shown by weight loss, inactivity, or poor grooming. The attenuated CT1-HA vector was not detected in the lungs or blood of mice examined at 6, 12, 24, and 48 h after inoculation and yet was able to elicit a strong humoral immune response after a single intranasal inoculation. In mice inoculated with the CT1-HA recombinant, humoral immunity is raised to both the VSV G protein and the foreign viral protein (HA). The humoral immune response is also systemic, since serum neutralizing antibody titers are present after a mucosal inoculation. Furthermore, this vector is 100% effective in providing complete protection from a subsequent lethal influenza virus challenge. No significant weight loss is measured in mice after challenge, no change in activity or grooming is observed, 100% of the mice survive challenge, and no influenza virus is recovered from lungs or blood 2 days after challenge.

Although complete neutralization (sterilizing immunity) of the challenge virus was not achieved in mice inoculated with G-HA (low titers of influenza virus were recovered from pulmonary tissues 2 days after challenge), the mice were completely protected from weight loss and lethality associated with more severe pathogenesis when given two immunizing doses of G-HA. A third immunizing inoculation with G-HA virus may further boost the humoral immune response and provide sterilizing immunity to the challenge virus. This possibility will be examined in future assays.

The effectiveness of the VSVG-HA vector is presumably due to an ability to infect cells for at least one round of multiplication. We believe the infection is mediated primarily by residual VSV G protein derived from the BHK-G cell line, since neutralization of the G-HA virus is accomplished in vitro with G antibodies but not with HA antibodies. We believe that a single round of infection is critical for protective immunity because mice inoculated intranasally with large doses of UV-inactivated VSV-HA did not produce neutralizing antibodies to VSV or to influenza virus and the mice were not protected from subsequent influenza virus challenge. Although the G-HA recombinant was measurable at low levels in the lungs 6 to 12 h after inoculation, it was still highly attenuated, showing no sign of viremia (6 to 48 h after initial inoculation) and causing no measurable weight loss or observable change in activity after initial inoculations or boosts.

The VSVG vector also offers a unique attribute not offered by the VSV-CT1 vector. Immunization with VSVG recombinants indicates that VSVG is a reusable vector, since no neutralizing antibodies are directed to the vector itself while a specific antibody response is elicited to the foreign protein expressed from the vector. We have examined the reusable characteristic of the VSVG vector in more detail by taking mice previously inoculated with VSVG-HA, which have measurable titers of antibody to influenza virus but not to VSVrwt, and inoculating them with VSVrwt. These mice subsequently develop high titers of neutralizing antibody to VSV in the serum (data not shown) which correspond to production of antibodies to VSV G (14). We have also shown that mice which have first been inoculated with a VSV recombinant expressing VSV G (and have high VSV neutralizing antibody titers) can subsequently be inoculated with a recombinant lacking the VSV_I G protein but expressing a G protein of another vesiculovirus (Chandipura virus) and are able to raise high titers of neutralizing antibody to Chandipura virus G (data not shown). Not only is the VSVG vector reusable, but an interesting observation is that mice previously exposed to the VSV core proteins but not to G protein do not lose as much weight when subsequently inoculated with constructs containing G protein. This may be due to the time of inoculation, since older mice may be less susceptible to pathogenesis than younger mice (4, 9, 15, 18, 26), or it may be due to partial protection elicited by cell-mediated immunity (24, 41). Cell-mediated immunity to these recombinant VSVs has not yet been examined in the mouse model.

We believe there is significant potential for VSV as a highly attenuated vaccine vector for the delivery of heterologous viral proteins. Vectors such as VSV-CT1 provide strong humoral responses without pathogenesis after single inoculations at doses as low as 10³ PFU/mouse, and vectors such as VSVG, which are also highly attenuated, have potential as reusable vaccines within an individual for multiple immunizations against heterologous viruses or for viruses, such as influenza virus, which require annual reimmunization. These live attenuated recombinants have the advantage of immunizing with VSV-based vectors without the risks inherent in the heterologous virus(es) themselves. They can be developed rapidly in the native G, CT1, or G backgrounds and can be delivered via the intranasal route, providing systemic immunity. It is likely that cell-mediated immunity and mucosal antibodies (not examined in this report) also provide protection in VSV-immunized mice, since VSV elicits a strong cell-mediated response (41) and since both immunizing inoculations and challenges may be administered through the mucosal, intranasal route. Furthermore, VSV vectors expressing foreign viral glycoproteins are likely to generate long-term immunity characteristic of natural infections. Development of the VSV vector system for vaccine delivery would complement other recombinant systems currently employed and other systems which are being developed.