|Home | About | Journals | Submit | Contact Us | Français|
Recombinant Sendai virus (rSeV) was used as a live, attenuated vaccine vector for intranasal inoculation and mucosal expression of the hemagglutinin-neuraminidase (HN) surface glycoprotein of human parainfluenza virus type 3 (HPIV3). Two vaccine candidates rSeV-HPIV3HN(P-M) and rSeV-HPIV3(F-HN) were constructed in which the HPIV3 HN open reading frame and an additional gene junction was inserted in the P-M and F-HN gene junctions of rSeV, respectively. The rSeV-HPIV3HN(P-M) virus was attenuated compared to rSeV-HPIV3(F-HN) in LLC-MK2 cells, and yet both vaccine candidates grew to similar extents in NHBE cells and in the respiratory tracts of cotton rats. These results suggest that in vitro vector growth in NHBE cells more accurately predicts virus yield in cotton rats than does growth in LLC-MK2 cells. Both vaccine vectors elicited high levels of serum neutralizing antibodies and conferred protection from HPIV3 challenge in cotton rats. Compared to vaccination with a high dose (2,000,000 PFU), intranasal inoculation with a low dose (200 PFU) resulted in a 10-fold decrease in vector growth in the nasal cavity and trachea and a 50-fold decrease in the lungs. However, low-dose vaccination resulted in only modest decreases in anti-HPIV3 antibodies in sera and was sufficient to confer complete protection from HPIV3 challenge. Varying the HPIV3 antigen insertion site and vector dose allowed fine-tuning of the in vivo growth and immunogenicity of rSeV-based vaccines, but all four vaccination strategies tested resulted in complete protection from HPIV3 challenge. These results highlight the versatility of the rSeV platform for developing intranasally administered respiratory virus vaccines.
Human parainfluenza virus types 1 to 4 (HPIV1 to HPIV4) are a leading cause of serious acute respiratory infection in young children, and the annual hospitalization rate for HPIV-associated infections in the United States is 7% (1). In the United States, the annual numbers of hospitalizations for children are estimated to be 28,900 for HPIV1, 15,600 for HPIV2, and 52,000 for HPIV3 (2). Most children are infected with HPIV3 by the age of 2 years and with HPIV1 and HPIV2 by the age of 5 years (3). HPIVs can cause 50 to 75% of cases of croup and also cause pneumonia, bronchiolitis, bronchitis, and otitis media (4). HPIV3 causes >70% of serious HPIV infections, which typically require longer and more costly hospital stays than for HPIV1 or HPIV2 infections (4). In immunocompromised patients, the case fatality rate for lower respiratory tract infections caused by HPIV3 can exceed 33% (5). There are currently no licensed parainfluenza virus-specific vaccines or drugs, and HPIV infection is treated by supportive care. Steroids are used to reduce inflammation and epinephrine to relieve airway constriction; ribavirin is not effective in treating HPIV infection in normal hosts (4). Hence, there is an urgent need to develop safe and effective vaccines against the HPIVs, especially HPIV3.
A trivalent inactivated vaccine against HPIV1, HPIV2, and HPIV3 was shown to be immunogenic but not protective in children (6, 7). Subunit vaccines consisting of purified HPIV3 fusion (F) and hemagglutinin-neuraminidase (HN) surface glycoproteins are immunogenic and protective against challenge in cotton rats but have not yet been tested in nonhuman primates or in clinical trials (4). Most current efforts on HPIV vaccine development focus on live attenuated candidates that are intranasally administered (8). Such vaccines are considered attractive because they induce immunity directly in the respiratory tract by expressing viral antigens in their native forms, and they can be infectious and moderately immunogenic in the presence of maternal antibodies (4). A temperature-sensitive HPIV3 vaccine candidate, cp45, was generated by 45 rounds of cold passage (9). Bovine parainfluenza virus type 3 (BPIV3) has been developed as a Jennerian vaccine against HPIV3 (10). In ongoing phase I and II trials, cp45 and BPIV3 are overattenuated in seropositive children but partially immunogenic in seronegative children and infants (11). A chimeric human-bovine PIV3 virus that contains human F and HN genes and bovine internal genes is attenuated and protective in nonhuman primates (12) and is being evaluated in phase I trials. Reverse genetics is also being used to introduce site-directed mutations to restrict the replication of HPIV1, HPIV2, and HPIV3 (4).
Sendai virus (SeV), the murine counterpart of HPIV1, is also being developed as a Jennerian vaccine against HPIV1 and as a respiratory vaccine vector when containing an inserted F gene from the respiratory syncytial virus (RSV) or the HN gene from HPIV3 or HPIV2 (13–16). SeV and HPIV1, like HPIV3, are from the genus Respirovirus and have an amino acid sequence homology of >75% among their six structural genes (13). Vaccination of African green monkeys with wild-type SeV causes no apparent clinical symptoms and induces antibody responses and protection from HPIV1 challenge (17). Ongoing phase I trials indicate that SeV is safe and immunogenic in seropositive adults and children (18). In the cotton rat model, recombinant SeVs (rSeVs) containing an RSV F gene or an HPIV3 HN gene inserted between the SeV F and HN genes elicit binding and neutralizing antibodies, elicit gamma interferon-producing T cells, and protect from challenge by homotypic and heterotypic strains of RSV and HPIV3 (15).
To study the dynamics of SeV infection in living animals noninvasively by bioluminescence imaging, several luciferase-containing SeV vectors were evaluated in mice, including rSeV-luc(P-M) and rSeV-luc(F-HN). These vectors contain a firefly luciferase reporter gene inserted between the phospho (P) and matrix (M) or F and HN genes of recombinant SeV, respectively (19). Insertion of the foreign gene relatively upstream in the rSeV-luc(P-M) virus decreased replication in vitro compared to downstream insertion in the rSeV-luc(F-HN) virus. However, the attenuated rSeV-luc(P-M) virus had a delayed and yet robust growth in the nasopharyngeal cavities of mice, eliciting a robust serum antibody response. Similarly, low-dose 70-PFU inoculation of a nonattenuated rSeV virus also had delayed yet robust growth in the nasopharyngeal cavity and elicited a high response of serum antibodies and protective immunity (19). Overall, the murine upper respiratory tract (URT) supported robust growth of rSeV even when the virus was attenuated or inoculated at a relatively low dose.
To investigate the extent to which an attenuated rSeV-based HPIV3 vaccine grows and promotes protective immunity in the respiratory tracts of cotton rats and to establish a preclinical model for parainfluenza virus infection and immunity, we constructed and evaluated two vaccine candidates rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) in which the HPIV3 HN gene was inserted between the P and M or F and HN genes of rSeV, respectively. Both rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) grew to robust levels, induced binding and neutralizing antibodies, and promoted protective immunity in cotton rats even when inoculated at a relatively low dose of 2 × 102 PFU, highlighting the versatility of the rSeV vaccine platform.
LLC-MK2 and Vero cells were obtained from the American Type Culture Collection (ATCC). Normal human bronchial epithelial (NHBE) cells (Lonza, Walkersville, MD) from a single 4-year-old healthy male donor were expanded, cryopreserved, and cultured in an air-liquid interface (ALI) system (20). Briefly, cells were plated in 0.33-cm2 transwell inserts (Corning, NY) and allowed to reach confluence. To establish the ALI, the apical surface of the cells was exposed to a humidified 95% air–5% CO2 environment. The apical surfaces of the NHBE cells were washed twice with phosphate-buffered saline (PBS; pH 7.2) to remove excess mucus secretion, and the basolateral medium was changed every 2 to 3 days for a minimum of 5 weeks in culture. The virus rSeV-HPIV3HN(F-HN) has been generated previously (15). To generate the virus rSeV-HPIV3HN(P-M), the HPIV3 HN gene insert was subcloned into a previously generated Enders strain-based pSeV viral genome plasmid that has a unique NotI restriction site mutated between the SeV P and M genes (19). In both viruses, the HPIV3 HN gene was the C243 strain from the ATCC, product VR-93. Infectious rSeV-HPIV3HN(P-M) was rescued by reverse genetics using the pSeV viral genome plasmid as described previously (15). Briefly, Vero cells were infected with a UV-inactivated T7 RNA polymerase-expressing recombinant vaccinia virus vTF7.3 (21) and transfected with the pSeV viral genome plasmid and pTF1 plasmids that express the SeV N, P, and L genes under a T7 promoter in the presence of Lipofectamine (Life Technologies, Grand Island, NY). Cell lysates were inoculated into 10-day-old embryonated chicken eggs, and allantoic fluid containing virus was collected 72 h postinoculation. The virus stock was amplified another round in chicken eggs and plaque purified in LLC-MK2 before another round of amplification in eggs to yield a stock of P1 virus. The infectivity of the P1 stocks of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) was measured by plaque titration in LLC-MK2 cells, and the sequences of the P1 virus stocks were confirmed by reverse transcription-PCR (RT-PCR) and sequencing.
P1 stocks of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) were serially passaged for 8 rounds in 10-day-old embryonated chicken eggs. During each passage, each egg was inoculated with 100 μl of PBS with magnesium and calcium (PBS+) containing 1,000 PFU of virus and was incubated at 35°C for 72 h, at which time the allantoic fluid containing virus was harvested. After the eighth passage, individual clones of the P8 virus stock were isolated from plaques on LLC-MK2 cells (cf. below). Single clones were picked and incubated overnight in 400 μl of PBS+ supplemented with 0.1% gentamicin before being inoculated into individual wells of a six-well plate that contained confluent monolayers of LLC-MK2 cells. To sequence the clones, RNA was extracted after 16 h of infection and used for reverse transcription-PCR (RT-PCR) and sequencing as described previously (15). To assess the abilities of individual clones to express the HPIV3 HN protein, radioimmunoprecipitation and SDS-PAGE analysis were performed after 16 h of infection.
Virus stocks were serially diluted with PBS+ supplemented with 0.1% gentamicin before being inoculated into individual wells of a six-well plate that contained confluent monolayers of LLC-MK2 cells. After 1 h at room temperature, supernatants were aspirated and the infected cells were overlaid with minimal essential medium (MEM) supplemented with 0.45% bicarbonate solution, 1× MEM amino acid solution (Invitrogen), 1× MEM vitamin solution (Invitrogen), 8 mM l-alanyl-l-glutamine (GlutaMAX; Invitrogen), 0.1% gentamicin, 0.9% agar, and trypsin at a final concentration of 0.002 mg/ml. After incubation at 33°C (5% CO2) for 4 to 7 days, a second overlay was performed using MEM supplemented in the absence of trypsin with 0.45% bicarbonate solution, 1% MEM amino acids, 1% MEM vitamins, 8 mM l-alanyl-l-glutamine (GlutaMAX), 0.1% gentamicin, 20% fetal bovine serum (FBS), 0.0035% neutral red, and 0.9% agar. The cells were then incubated at 33°C (5% CO2) for 1 to 2 days.
Radioimmunoprecipitation experiments were done as described previously (22) to examine the abilities of the recombinant SeV vectors to express the HPIV3 HN, SeV HN, and SeV F proteins. Briefly, monolayers of LLC-MK2 cells were infected at a multiplicity of infection (MOI) of 5 PFU/cell with rSeV-HPIV3HN(P-M), rSeV-HPIV3HN(F-HN), SeV wild-type (WT), or HPIV3 WT and then incubated at 33°C in Dulbecco modified Eagle medium (DMEM; 10% fetal calf serum [FCS], 1% l-glutamine) for 16 h. Infected LLC-MK2 cells were then washed twice with PBS+ and maintained in methionine- and cysteine-free medium for 30 min before labeling for 15 min with 100 μCi of [35S]Promix (Amersham Pharmacia Biotech) in 1 ml of DMEM lacking methionine and cysteine and containing 20 mM HEPES buffer (pH 7.3). Cells were washed once with PBS+ and chased with 3 ml of DMEM containing 2 mM methionine, 2 mM cysteine, and 20 mM HEPES buffer (pH 7.3) for 30 min. Samples were lysed with ice-cold radioimmunoprecipitation assay (RIPA) buffer (0.15 M NaCl, 9.25 mg of iodoacetamide/ml, 1.7 mg of aprotinin/ml, 10 mM phenylmethylsulfonyl fluoride). The lysate was centrifuged at 67,000 × g in an Optima L-90k ultracentrifuge (Beckman Coulter). The supernatant was incubated for 20 h at 4°C with 25 μl of rabbit polysera that were generated to KLH-conjugated peptides from the cytoplasmic tails of the HPIV3 HN, SeV F, or SeV HN proteins (15, 22). Immune complexes were adsorbed to protein A-Sepharose Cl-4B (GE Healthcare) for 1 h at 4°C. Samples were washed three times with radioimmunoprecipitation assay (RIPA) buffer containing 0.3 M NaCl, three times with RIPA buffer containing 0.15 M NaCl, and once with 50 mM Tris buffer (0.25 mM EDTA, 0.15 M NaCl [pH 7.4]). The samples were then fractionated on 12% NuPAGE Bis-Tris polyacrylamide–SDS gels (Invitrogen). Protein bands were visualized using a Typhoon 9200 variable mode imager (GE Healthcare) and densitometry was performed as described previously (19).
Monolayers of LLC-MK2 cells in 12-well dishes (100% confluent) were washed with PBS+ and then infected with 0.1 ml of recombinant virus at an MOI of 0.01 PFU/cell suspended in PBS+. One hour after infection at room temperature, the supernatant was aspirated, and cells were washed with PBS+. Then, 1 ml of DMEM supplemented with 10% FBS and 1% glutamine was added to cells, and cells were incubated at 33°C (5% CO2). Supernatants were collected at appropriate time intervals over 96 h. For growth curves in NHBE cells, viruses were diluted in 0.5% BSA-BEBM (Lonza) to equivalent titers as determined by plaque assays on LLC-MK2 cells. The apical surfaces of NHBE cells were washed three times with PBS to remove excess mucus secretion prior to infection. Fully differentiated NHBE cells were infected at an MOI of 0.01 PFU/cell. Viruses were allowed to adsorb for 1 h at 37°C on the apical side and removed by aspiration. Cells were washed three times with PBS to remove free infectious virus particles. Viruses released apically were harvested at the indicated times postinfection by the apical addition and collection of 300 μl of 0.5% BSA-BEBM allowed to equilibrate at 37°C for 30 min. Samples were stored at −80°C until assayed for virus titers. Aliquots of supernatants were stored at −80°C until virus titers were determined by the plaque assay.
Groups of adult female Sigmodon hispidus cotton rats (Harlan Sprague Dawley, Indianapolis, IN) were anesthetized in an isoflurane chamber and subsequently inoculated intranasally with 100 μl (50 μl was pipetted in each nostril and inhaled by the animal) of PBS+ containing either 2 × 102 or 2 × 106 PFU of virus rSeV-HPIV3HN(P-M), rSeV-PIV3HN(F-HN), rSeV-WT, or HPIV3 C243-WT. PBS+ was used as a mock-infected control. In experiments performed to measure growth of the SeV vaccines during primary infection, cotton rats were euthanized after 3 days of infection so that the lungs, trachea, and nasal cavities could be removed. The lungs and trachea were homogenized for 30 s on ice using an IKA T10 disperser at setting 4 (IKA Works, Inc., Wilmington, NC) and suspended in PBS to total volumes of 2 and 1 ml, respectively. Nasal turbinates were crushed in a 1-ml vial using a syringe plunger to remove attached tissue and then were filtered through a 70-μm-pore-size cell strainer (Fisher Scientific) and resuspended in PBS to a total volume of 2 ml. Viral loads in tissue homogenates were determined by plaque titration assays and are expressed as the total amount of PFU per entire tissue (nasal cavity, trachea, or lungs). In experiments designed to evaluate the immunogenicities and protective capacities of the vaccines, sera were collected 28 days after vaccination and animals were challenged 35 days after vaccination with 2 × 106 PFU of HPIV3 C243-WT. Cotton rats were euthanized 3 days after HPIV3 challenge (day 38 of the experiment) so that the lungs, trachea, and nasal cavities could be removed and homogenized as described above. HPIV3 viral loads in tissue homogenates after challenge were determined by plaque titration assays and are expressed as the total amount of PFU per entire tissue (nasal cavity, trachea, or lungs).
Immunogenicities of the rSeV-HPIV3HN viruses in cotton rats were compared to those of HPIV3 and SeV by measuring the enzyme-linked immunosorbent assay (ELISA) and neutralization titers of sera collected 28 days after vaccination as described previously (15). For ELISAs in which anti-HPIV3 HN-specific binding antibody activity was detected, HPIV3 was grown in LLC-MK2 cells, and supernatants were concentrated using a Millipore Amicon filter unit. Concentrates of HPIV3 virus were lysed in disruption buffer (0.5% Triton X-100, 0.6 M KCl, 0.05 M Tris [pH 7.8]), diluted 1:3,000 with PBS, and coated on 96-well ELISA plates. Lysates of WT SeV were similarly plated and prepared. After overnight incubation, plates were blocked with PBS containing 3% bovine serum albumin (BSA; Sigma-Aldrich). Serum samples from cotton rats were serially diluted with PBS and incubated on plates for 2 h at 37°C. Plates were then washed and incubated with rabbit anti-cotton rat antibody (kindly provided by Greg Prince, Virion Systems, Rockville, MD) for 30 min at room temperature. Plates were washed with PBS and incubated with anti-rabbit IgG-horseradish peroxidase conjugate (diluted 1:3,000 in PBS, 1% BSA; Bio-Rad, Hercules, CA) for 30 min at room temperature, washed again with PBS, and incubated with ABTS [2,2′azinobis(3-ethylbenzthiazolinesulfonic acid); Southern Biotechnology Associates, Inc., Birmingham, AL]. Spectrophotometric absorbance was read at 405 nm. Neutralization capacities of sera were measured by serially diluting with PBS and mixing with ca. 10 50% tissue culture infective doses of HPIV3. Viruses were either the homologous C243 strain obtained from the ATCC or heterologous 4-04, 5-97, and 8-94 isolates obtained from the Department of Pathology at St. Jude Children's Research Hospital. LLC-MK2 cell monolayers in 24-well plates were then overlaid with the HPIV3-serum mixtures and incubated for 1 h at 33°C before adding DMEM supplemented with glutamine, antibiotics, and 5% FCS. The LLC-MK2 cultures were then incubated for 4 days at 33°C (5% CO2). From each well, 100 μl of supernatant was mixed with 100 μl of 0.5% turkey erythrocytes and hemagglutination assays were performed and evaluated as described previously (15).
Statistical analyses were performed using GraphPad Prism 5. ELISA and neutralization titers were compared using an unpaired, two-tailed Student t test with a 95% confidence level.
To investigate how the site of insertion of the foreign gene affects the replication and immunogenic potential of a SeV-based vaccine, we used reverse genetics to create rSeV-HPIV3HN(P-M), in which the HPIV3 HN gene and an additional gene junction were inserted between the P and M genes of rSeV Enders strain (Fig. 1). The rSeV-HPIV3HN(F-HN) vaccine candidate was generated and found to elicit protective immunity in cotton rats previously (15), but the genetic stability and capacity of the virus to grow in cell culture or in cotton rats was not determined in that study. The rSeV-HPIV3HN(P-M) vaccine candidate was generated for the present study after discovering that an analogous rSeV-luciferase(P-M) reporter virus grew to high titers and expressed high levels of luciferase in the URT, while simultaneously being attenuated in the lungs of mice (19). To determine the extent to which a P-M or an F-HN antigen insertion site alters the production capacity of an rSeV-HPIV3HN vaccine, groups of six chicken eggs were inoculated with 1,000 PFU of rSeV-HPIV3HN(P-M) or rSeV-HPIV3HN(F-HN), and infected eggs were incubated for 24, 48, or 72 h at 35°C. At each time point, allantoic fluid was harvested and plaque titrated in LLC-MK2 cells. Both vaccine candidates produced the highest titers after 72 h of incubation, and the average titers of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) at this time point were 1 × 108 PFU/ml and 9 × 108 PFU/ml, respectively. Thus, both viruses were capable of growing to high titers, and the production capacity of rSeV-HPIV3HN(F-HN) was ~9-fold that of rSeV-HPIV3HN(P-M) under standard conditions.
The genetic stabilities of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) were determined by passaging them 8 times in embryonated chicken eggs. Allantoic fluid from the P8 eggs was plaque titrated in LLC-MK2 cells, and 10 plaques from each virus were selected and propagated in duplicate six-well plates of LLC-MK2 cells for 16 h. From one set of six-well plates, the cells were harvested and RNA was extracted for RT-PCR and cDNA sequencing. From the other set of six-well plates, the HPIV3 HN protein was radioimmunoprecipitated and analyzed by SDS-PAGE. All 10 clones each of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) retained a functional insert, as assessed by RT-PCR and sequencing (five representative clones from each shown in Fig. 2). All ten clones from each virus also expressed the HPIV3 HN protein, as measured by radioimmunoprecipitation. These results indicated that both viruses were viable vaccine candidates with regard to genetic stability and production capacity and could be further evaluated for their virologic properties in vitro.
Upstream insertion of a secreted alkaline phosphatase or luciferase reporter gene into the P-M gene junction of rSeV has been shown to result in greater virus attenuation and greater foreign gene expression than insertion into the downstream F-HN gene junction (19, 23). This effect is presumably due to the polarity of transcription in combination with alterations in the ratios of expressed viral genes (19, 23–27). To compare the levels of HPIV3 HN expression after insertion of the foreign gene into the P-M and F-HN gene junctions of the rSeV vector, LLC-MK2 cells were infected with an MOI of 5 PFU/cell of rSeV-HPIV3HN(P-M), rSeV-HPIV3HN(F-HN), rSeV-WT, or HPIV3-WT. The levels of expression of the HPIV3 HN protein, along with the SeV F and HN proteins, were measured by radioimmunoprecipitation and SDS-PAGE (Fig. 3). The levels of expression of the HPIV3 HN protein vectored from rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) were 83 and 52%, respectively, compared to infection with HPIV3-WT (Fig. 3B). The HPIV3HN(F-HN) virus and rSeV-WT had similar levels of expression of SeV F and HN proteins; there was a substantial decrease in SeV F protein expression and an increase in SeV HN expression in rSeV-HPIV3HN(P-M), but this was not statistically significant. Overall, both rSeV-HPIV3HN viruses expressed the foreign gene product and viral proteins, and HPIV3 HN protein expression was significantly lower than that of HPIV3-WT for rSeV-HPIV3HN(F-HN) but not rSeV-HPIV3HN(P-M).
We next investigated how insertion of the HPIV3 HN gene into the P-M and F-HN gene junctions influenced the replication kinetics of the rSeV vector in vitro. LLC-MK2 or normal human bronchial epithelial (NHBE) cells were inoculated with rSeV-HPIV3HN(P-M), rSeV-HPIV3HN(F-HN), or rSeV-WT at an MOI of 0.01 PFU/cell, and virus titers in supernatants collected at various time points were measured by plaque titration. In LLC-MK2 cells, multistep growth curves were measured at 37°C and 33°C (Fig. 4A and andB).B). At both temperatures, both viruses containing the HPIV3 HN gene had delayed and lower growth than rSeV-WT. rSeV-HPIV3HN(P-M) grew to a maximum titer that was 10- to 50-fold lower than for rSeV-HPIV3HN(F-HN). Just as both viruses containing the HPIV3 HN gene were attenuated in LLC-MK2 cells, the two vaccine candidates also had delayed and lower growth compared to rSeV-WT in NHBE cells at 33°C compared to rSeV-WT (Fig. 4C). Both rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) grew to titers similar to each other and ~100-fold lower than rSeV-WT in NHBE cells. Thus, rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) were similarly attenuated for virus growth compared to rSeV-WT in NHBE cells, while rSeV-HPIV3HN(P-M) was attenuated to a larger extent than rSeV-HPIV3HN(F-HN) in LLC-MK2 cells and embryonated chicken eggs.
To study the influence of the HPIV3 HN antigen insertion site and vector dose on the growth of an rSeV-based vaccine in an animal model, groups of three to five cotton rats were inoculated intranasally with rSeV-HPIV3HN(P-M), rSeV-HPIV3HN(F-HN), or rSeV-WT at a low dose of 2 × 102 PFU/animal or a high dose of 2 × 106 PFU/animal. After 3 days of infection, the cotton rats were euthanized so that nasal cavity, tracheal, and lung tissues could be recovered from the animals and in vivo viral loads could be measured by plaque titration. Overall, viral loads were measured for six groups (three viruses at two doses) in three tissues (nasal cavity, trachea, and lungs). The experimental design allowed comparison of in vivo vaccine virus growth based on the (i) HPIV3 antigen insertion site, (ii) the inoculated virus dose, and (iii) infected respiratory tract tissue.
rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) had similar viral loads compared to each other in the same tissue after inoculation with the same dose (Fig. 5). Compared to the rSeV-WT vector, the two vaccine candidates were also similarly attenuated in vivo, just as they were in NHBE cells. After inoculation with either the low or the high dose, the rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) viruses had viral loads ~50-fold lower than rSeV-WT in the trachea (Fig. 5B) and ~10-fold lower in the nasal cavities (Fig. 5A) and the lungs (Fig. 5C).
Inoculation with the low dose of virus resulted in a lower virus load in vivo for a given virus and tissue. However, there was greater reduction in viral load at the low dose in the lungs than the nasal cavities and tracheas for all three viruses: an ~50-fold reduction in viral load in the lungs but only an ~10-fold reduction in the nasal cavities or tracheas.
Three days after intranasal vaccination, the load of each of the three viruses was 10- to 100-fold higher in the nasal cavity than the trachea or lungs, indicating that the rSeV-WT vector and the rSeV-HPIV3HN vaccine candidates had a greater propensity for growth, and/or to avoid clearance in the URT under the given inoculation conditions. Three days after inoculation with the high dose, rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) reached a load of nearly 107 PFU in the nasal cavity but only approximately 104 and 105 PFU/ml in the trachea and lungs, respectively.
Taken together, these results reveal that (i) rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) had similar growth phenotypes in cotton rats, but viral loads were lower than rSeV-WT, (ii) lower viral load due to low-dose inoculation was more pronounced in the lungs than the trachea and nasal cavity, and (iii) rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) grew to higher loads in the nasal cavity than in the trachea and lungs.
To study the influence of the HPIV3 HN antigen insertion site and vector dose on the immunogenicity and protective capacity of rSeV-based vaccines, a vaccination-challenge experiment was performed with groups of three to five cotton rats and repeated once. Animals were vaccinated intranasally on day 0, sera were collected on day 28, animals were challenged with 2 × 106 PFU/animal of HPIV3 strain C243 on day 35, and respiratory tissues were harvested on day 38. Experimental groups were vaccinated with rSeV-HPIV3HN(P-M), rSeV-HPIV3HN(F-HN), or rSeV-WT at doses of 2 × 102 or 2 × 106 PFU/animal. Control groups were vaccinated with HPIV3 strain C243 at a dose of 2 × 106 PFU/animal or with PBS.
The stimulation of HPIV3-binding antibodies 28 days after vaccination was measured by ELISAs using HPIV3-coated plates (homotypic C243 strain) and peripheral blood sera collected from cotton rats (dilutions were 1:1,000; 1:10,000; and 1:100,000). In cotton rats vaccinated with 2 × 106 PFU of rSeV-HPIV3HN(P-M) or rSeV-HPIV3HN(F-HN), serum anti-HPIV3 binding antibody levels were comparable to those generated after vaccination with 2 × 106 PFU of HPIV3 (Fig. 6A). Lowering the dose of the rSeV-HPIV3HN viruses to 2 × 102 PFU also resulted in relatively high levels of serum anti-HPIV3-binding antibodies, albeit at levels lower than those obtained after vaccination with a 2 × 106 PFU dose. For example, vaccination with the lower dose of rSeV-HPIV3HN(F-HN) resulted in a statistically significant decrease in HPIV3-binding antibody response (P = 0.03) compared to vaccination with the higher dose. Comparing low-dose vaccination of the two vaccine candidates to each other, rSeV-HPIV3HN(P-M) elicited a 25% higher level of HPIV3-binding antibodies than rSeV-HPIV3HN(F-HN), although the difference was not statistically significant (P = 0.16).
To determine the abilities of the viruses to generate immunogenic responses to the rSeV backbone, the levels of serum anti-SeV binding antibodies were also measured by ELISA (Fig. 6B). Robust levels of serum anti-SeV binding antibodies were generated after vaccination with the rSeV-HPIV3HN viruses, although levels were lower than those generated after vaccination with the empty rSeV-WT vector. Low levels of cross-reactive serum-binding antibodies were detected: anti-HPIV3 binding antibodies after vaccination with rSeV-WT (Fig. 6A) and anti-SeV binding antibodies after vaccination with HPIV3 (Fig. 6B). The weak cross-reactive response has been reported previously and is not surprising as the respiroviruses SeV and HPIV3 share sequence and antigenic similarities (15).
Next, we measured the ability of the vaccines to elicit anti-HPIV3 serum-neutralizing responses. Vaccination with 2 × 106 PFU of rSeV-HPIV3HN(P-M) or rSeV-HPIV3HN(F-HN) was at least as potent as that with 2 × 106 PFU of HPIV3 in inducing serum-neutralizing activity against the homologous HPIV3 C243 strain virus (Fig. 7A). Vaccination with the low 2 × 102 PFU dose of the rSeV-HPIV3HN viruses also stimulated robust serum neutralizing activity, but levels were lower than those induced by high-dose vaccination [P values of 0.048 and 0.001 for rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN), respectively]. At a given dose, the rSeV-HPIV3HN(P-M) virus elicited a higher level of serum-neutralizing activity against the homologous C243 isolate than rSeV-HPIV3HN(F-HN) (P < 0.05 for both high- and low-dose inoculation), which is consistent with rSeV-HPIV3HN(P-M) expressing a higher level of the HPIV3 HN protein in vitro (Fig. 3B).
The abilities of the sera to neutralize heterologous isolates of HPIV3 were also tested using clinical isolates of HPIV3 (8-94, 5-97, and 4-04) that had been isolated during several different seasonal outbreaks by the Department of Clinical Pathology at St. Jude Children's Research Hospital (15). All three clinical isolates of HPIV3 were neutralized by sera collected from cotton rats that were vaccinated with either of the rSeV-HPIV3HN viruses or HPIV3 (Fig. 7B to toD).D). Vaccination with a higher dose of virus resulted in a higher level of serum-neutralizing activity against the heterologous HPIV3 viruses, just as it had been for the homologous C243 strain. Overall, results from the serum-neutralization analyses showed that (i) a higher dose of vaccine was associated with a higher neutralizing response and (ii) rSeV-HPIV3HN(P-M) elicited a higher neutralizing response than rSeV-HPIVHN(F-HN) against the homologous C243 isolate, but not against the heterologous clinical isolates.
After demonstrating that the rSeV-HPIV3HN vaccines were immunogenic, we investigated their abilities to induce responses that protect cotton rats from infection after an HPIV3 challenge. On day 35, animals were challenged intranasally with 2 × 106 PFU of HPIV3 C243 strain and euthanized 3 days later to collect respiratory tissues for plaque titration of the HPIV3 challenge virus. As shown in Fig. 8, cotton rats vaccinated with either the low or high dose of rSeV-HPIV3HN(P-M) or rSeV-HPIV3HN(F-HN) were completely protected from the challenge by HPIV3, showing no detectable growth of the challenge virus in the nasal cavity, trachea, or lungs. Vaccination with 2 × 106 PFU of HPIV3 also resulted in complete protection from the HPIV3 challenge throughout the respiratory tract. Vaccination with the empty rSeV-WT vector conferred partial protection compared to the PBS control group, as reported previously (15). Thus, both low and high doses of the two rSeV-HPIV3HN vaccine candidates stimulated a response sufficient to fully protect from HPIV3 challenge in cotton rats 5 weeks after vaccination.
In this study, we investigated how the inoculated dose of a SeV-based vaccine vector and the intergenic site of insertion of the foreign antigen influence replication, immunogenicity, and protective capacity in a cotton rat model. On the basis of findings from a previous study that addressed the in vivo dynamics of the SeV vector in mice using firefly luciferase as the inserted foreign gene (19), we generated 2 live vaccine viruses rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) in which an HPIV3 HN gene was inserted into the P-M or F-HN gene junction of rSeV, respectively. Consistent with results from previous studies using rSeV reporter viruses (19, 23), we found that the relatively upstream insertion site in the rSeV-HPIV3HN(P-M) construct resulted in higher foreign gene expression but lower vector growth in cultured LLC-MK2 cells than with the rSeV-HPIV3HN(F-HN) construct. Surprisingly, though, rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) grew to similar titers in vitro in NHBE cells and in vivo throughout the respiratory tracts of cotton rats. Just as the two live viruses grew to similar levels when vaccinated intranasally, they also stimulated similar homologous and cross-reactive immunogenic responses in peripheral blood and gave complete protection from the HPIV3 challenge 5 weeks after vaccination. A two-tier dose escalation was incorporated into the experimental design to enable a simultaneous comparison of low- (2 × 102 PFU) and high-dose (2 × 106 PFU) vaccination with the rSeV-HPIV3HN constructs. The high-dose inoculation resulted in an ~10-fold increase in vector growth in the nasal cavity and trachea and an ~50-fold increase in vector growth in the lungs. An increase in vector dose was associated with modest increases in the levels of HPIV3-binding and HPIV3-neutralizing activity by sera, and yet even low-dose vaccination resulted in complete protection from HPIV3 challenge. Overall, the results demonstrate the versatility of the rSeV platform for the construction of effective, live intranasal vaccines and suggest that even low doses of rSeV-based vaccines may be immunogenic and protective in naive subjects during dose-escalation studies.
Viral transcription by SeV initiates at a single promoter near the 3′ end of the genome and proceeds from the 3′ to 5′ end of the genome (N < P < M < F < HN < L), resulting in a gradient of viral protein expression, with expression of the N protein highest and that of the L protein lowest (28). This sequential mechanism of transcription is shared by the other members of the order Mononegavirales, such as Newcastle disease virus (NDV) (29) and PIV5 (26) (Paramyxoviridae), the rhabdovirus vesicular stomatitis virus (24, 25) (Rhabdoviridae), the filovirus Ebola virus (30) (Filoviridae), and the Borna disease virus (31) (Bornaviridae). For nonsegmented, negative-strand RNA viruses the intergenic site of insertion of a foreign gene into the viral genome determines the level of expression of the foreign gene and the magnitude and ratio of expression of viral genes. In turn, the magnitude and ratio of expression of viral genes helps determine the kinetics and extent of virus replication (32).
To better understand the importance of the foreign gene insertion site in developing a vectored paramyxovirus vaccine, we compared the virologic and immunologic properties of rSeVs into which a foreign vaccine gene was inserted into the P-M or F-HN gene junction. Similar to findings on the in vitro properties of rSeVs expressing reporter genes (19, 23), our results show that insertion of the HPIV3 HN gene upstream between the P and M genes versus downstream between the F and HN genes results in increased HPIV3 HN antigen expression and decreased virus growth in LLC-MK2 cells and embryonated chicken eggs. In general, the two SeV-based HPIV3 vaccines grew to similar levels in the cotton rat respiratory tract, promoting similarly high levels of systemic and mucosal immune responses and protecting from HPIV3 challenge as efficiently as vaccination with HPIV3. It should be noted that upstream insertion of the HPIV3 HN antigen between the P and M genes stimulated higher levels of serum neutralizing antibodies against homologous HPIV3 virus than downstream insertion between the F and HN genes, consistent with rSeV-HPIV3HN(P-M) having higher HPIV3 HN protein expression than rSeV-HPIV3HN(F-HN).
From the standpoint of vaccine development, the in vivo fitness and immunogenicity of the rSeV-HPIV3HN(P-M) virus highlight the versatility of rSeV as a respiratory vaccine vector. On the other hand, the similar efficacy of the two vaccine candidates also raises the question about why relative upstream positioning of the foreign antigen in the rSeV-HPIV3HN(P-M) construct does not appear to attenuate vector growth in vivo compared to rSeV-HPIV3HN(F-HN). Although tissue titers of rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) reached similar levels 5 days after inoculation, the rSeV-HPIV3HN(P-M) virus might have had delayed replication earlier in infection. In the case of the canine distemper virus (genus Morbillivirus), decreased F protein expression decreases recognition of the virus by the ferret immune system, thereby promoting earlier virus clearance (33). Lower SeV F protein expression by rSeV-HPIV3HN(P-M) could also result in lower immune recognition initially in cotton rats, allowing the virus to reach growth levels comparable to rSeV-HPIV3HN(F-HN). On the other hand, both rSeV constructs had similar multistep replication kinetics in NHBE cells, which may provide a better model for the cotton rat respiratory tract than LLC-MK2 cells or embryonated chicken eggs. The efficient production of infectious virus may be more sensitive to the foreign gene insertion site and altered viral gene expression in the context of LLC-MK2 cells than respiratory tract tissues. Our findings on the rSeV vector are similar to those showing that rHPIV3-MeHA(P-M), a recombinant HPIV3 virus with the measles virus HA gene inserted into the P-M gene junction, has delayed growth in LLC-MK2 cells compared to rHPIV3-MeHA(HN-L) and yet grows to similar levels in the upper and lower respiratory tracts of hamsters (34). However, this does not suggest that any paramyxovirus vector attenuated in vitro by foreign gene insertion will remain fully fit for growth in vivo. For example, a recombinant SeV with enhanced green fluorescent protein inserted upstream of the N gene has significantly reduced replication kinetics and virulence in mice (35).
In the present study, vaccination with the low or high dose of either of the live attenuated rSeV-HPIV3HN vaccine candidates resulted in URT-biased growth of the vector, robust production of binding and neutralizing antibodies in peripheral blood, and complete protection from HPIV3 challenge throughout the cotton rat respiratory tract. Moreover, both vaccine candidates were genetically stable after 8 passages and grew to high titers in embryonated chicken eggs, yielding at least 109 PFU per egg, which corresponds to ~500 U of a 2 × 106 PFU vaccine produced in a single egg. Thus, rSeV appears to have attractive and versatile properties as a vaccine vector against HPIV3, at least in the cotton rat model.
Enzootic lab strains of SeV, such as the Enders strain used here, can be pathogenic in mice if delivered in high doses directly into the lungs (19). However, such pathogenicity is dependent on mouse strain (19, 36–38) and limited to murine species, likely because the ability of SeV to counteract interferon-mediated innate immunity is specific to mice and deficient in other species such as humans (39). Unmodified wild-type SeV elicits protective immunity against HPIV1 without causing adverse events in nonhuman primates (17, 40) and is associated with no adverse events after intranasal inoculation into seropositive humans (18). Just as SeV is a promising Jennerian vaccine candidate against HPIV1, rSeV too is being developed as a vaccine vector against HPIV3, as reported here and in our earlier study (15), and other respiratory paramyxoviruses such as HRSV (13, 14, 41) and HPIV2 (16). Intranasal vaccination with the SeV-vectored HRSV vaccine, in which the HRSV F gene was inserted into the F-HN gene junction of rSeV, elicits protective immunity without causing pathology in both cotton rats (14) and African green monkeys (42). Analogous experimental trials in African green monkeys with the HPIV3 vaccine candidates described in our study are needed to compare rSeV-HPIV3HN(P-M) and rSeV-HPIV3HN(F-HN) for immunogenicity, protective capacity, and potential pathology in an animal model more closely related to humans.
In addition to SeV, several other respiratory paramyxoviruses have also been investigated as Jennerian vaccines and/or vaccine vectors in animal models and, in some cases, human trials. BPIV3 and bovine respiratory syncytial virus (BRSV) are protective against HPIV3 and HRSV, respectively, in cotton rats and monkeys (10, 43). However, these Jennerian vaccines appear to be unsuitable for effective protection in humans because of limited cross-protection in humans by BPIV3 and poor protection in chimpanzees by BRSV (44). Live attenuated HPIV3 and HRSV candidates have also been investigated, including host range mutants, cold-passage mutants, and temperature-sensitive mutants (44). More recently, reverse genetics has been used to generate a chimeric BPIV3 virus containing the HPIV3 F and HN envelope genes (12, 45) and a chimeric BPIV3 virus containing the HPIV1 and HPIV2 envelope genes (46). A BPIV3/HPIV3 chimeric virus, including an additional RSV G or F gene is immunogenic in rhesus monkeys (47) and does not cause any adverse events in adults and seropositive children, although the vaccine is minimally immunogenic in seropositive individuals (48, 49). Recombinant PIV5 expressing a foreign influenza virus HA gene is also under development (50, 51). NDV has been evaluated as a veterinary vaccine vector against various pathogens, including HPIV3 (52), avian influenza viruses (53), avian metapneumovirus (54), avian paramyxovirus type 3 (55), and Nipah virus (56).
Despite a growing number of paramyxovirus-vectored vaccines, some of which have shown promise in clinical trials, none has yet emerged as clearly effective in seronegative infants (57). Recombinant SeV remains an attractive candidate as a respiratory vaccine vector because of its high production capacity in chicken eggs, its restricted host range, its immunogenicity and protective capacity in animal models, and its apparent safety to date in seropositive adults and children. Previous studies have shown that a single, mixed vaccine cocktail of rSeV(F-HN) vaccines can protect cotton rats against challenge from multiple human pathogens, including HPIV1, HPIV2, HPIV3, and RSV (15, 16). Our finding that that the attenuated rSeV-HPIV3HN(P-M) vaccine was highly immunogenic and protective in cotton rats raises the possibility of engineering a single rSeV with multiple inserted foreign genes to target multiple pathogens such as HPIV1, HPIV3, and RSV from the same vector. Moreover, the finding that vaccination with 2 × 102 PFU of rSeV-HPIV3HN viruses was immunogenic and protective in cotton rats suggests that even low doses may be effective in dose-escalation trials. Overall, rSeV appears to be a versatile paramyxovirus vaccine vector, and the high immunogenicity and protective capacity conferred by the rSeV-HPIV3 viruses in cotton rats suggest that SeV-based HPIV3 vaccine candidates are attractive candidates for further testing in nonhuman primates.
We thank Julia Hurwitz and Allen Portner for providing viruses and other reagents and for helpful discussions. We thank Robert Sealy for assistance with cotton rat sera collection and, along with Bart Jones, for expert technical assistance on ELISA and neutralization titer assays. We thank Randall Hayden and the ATCC for providing HPIV3 virus isolates. We thank Bernard Moss for the vTF7.3 vaccinia virus. We thank the Animal Resources Center for supporting animal experiments. We thank the Hartwell Center for Bioinformatics and Biotechnology for DNA synthesis and sequencing. We thank Vani Shanker of Scientific Editing for proofreading the manuscript.
This project was supported by grant R01AI083370 from the NIH/NIAID and by ALSAC and the Children's Infection Defense Center at St. Jude Children's Research Hospital.
Published ahead of print 20 March 2013