|Home | About | Journals | Submit | Contact Us | Français|
Adenoviral vectors (AdV) have received considerable attention for vaccine development because of their high immunogenicity and efficacy. In previous studies, it was shown that DNA immunization of mice with codon-optimized expression plasmids encoding the fusion protein of respiratory syncytial virus (RSV F) resulted in enhanced protection against RSV challenge compared to immunization with plasmids carrying the wild-type cDNA sequence of RSV F. In this study, we constructed AdV carrying the codon-optimized full-length RSV F gene (AdV-F) or the soluble form of the RSV F gene (AdV-Fsol). BALB/c mice were immunized twice with AdV-F or AdV-Fsol and challenged with RSV intranasally. Substantial levels of antibody to RSV F were induced by both AdV vaccines, with peak neutralizing-antibody titers of 1:900. Consistently, the viral loads in lung homogenates and bronchoalveolar lavage fluids were significantly reduced by a factor of more than 60,000. The protection against viral challenge could be measured even 8 months after the booster immunization. AdV-F and AdV-Fsol induced similar levels of immunogenicity and protective efficacy. Therefore, these results encourage further development of AdV vaccines against RSV infection in humans.
Human respiratory syncytial virus (RSV) is a highly infectious member of the paramyxovirus family causing upper and lower respiratory tract infections in humans. Serious acute RSV infections, including fatal cases of bronchiolitis and pneumonia, occur particularly in premature infants, immunocompromised adults, and patients with pre-existing chronic lung diseases or underlying heart defects (11, 12, 14, 39, 46, 56). In young children, RSV is the most common respiratory tract pathogen, accounting for approximately 50% of hospitalizations due to lower respiratory tract infections (21). In population-based surveillance studies for hospitalization in Europe, RSV was identified in 42 to 45% of enrolled children younger than 2 years with lower respiratory tract infections, and the rate of hospitalization due to RSV-induced respiratory illnesses was estimated at 3 to 6% among industrialized nations (45). Children with severe RSV infections suffer from oxygen deficiency with cyanosis and require intensive medical care. Furthermore, RSV infection in childhood is suspected to be a risk factor for development of asthma (36, 41, 43, 59). The urgent need for an RSV vaccine is further demonstrated by a study showing that levels of disease burden, mortality, and morbidity caused by RSV infections in the elderly are comparable to those induced by nonpandemic influenza A infections (11). However, the immunization of children with a formalin-inactivated (FI) RSV vaccine in the 1960s resulted in a more severe clinical illness, with two fatal cases, than in nonvaccinated infants following RSV infection, pointing out the difficulties in developing a safe and efficacious RSV vaccine (7, 29). It was shown previously that the enhanced disease severity and the development of pulmonary eosinophilia are mainly attributable to an excessive Th2-polarized immune response (15, 35, 57). Furthermore, the lack of high-affinity antibodies after poor Toll-like receptor stimulation has been suggested to be a key factor of the enhanced disease induced by FI RSV vaccination and subsequent RSV infection shown recently (8). However, the enhanced disease induced by FI RSV could partially be reversed by the chemical reduction of the carbonyl groups produced by prior treatment with aldehyde (34).
Passive transfer of a neutralizing monoclonal antibody directed against RSV F (palivizumab) results in significant reduction of hospitalization rate due to RSV infection in children and preterm infants (16, 25), making RSV F a promising vaccine candidate for active immunization. Besides being a target for neutralizing antibodies, RSV F additionally contains cytotoxic-T-cell epitopes (1, 37). Moreover, RSV F based DNA vaccines induced encouraging immune responses of a balanced Th1/Th2 type in mice, as pulmonary eosinophilia and disease-enhancing effects were not observed after viral challenge (4, 5, 19, 31, 52). Additionally, RSV F is highly conserved between the two antigenic subgroups of RSV, which allows generation of cross-reactive antibodies after immunization (26).
We recently showed that vaccination with codon-optimized RSV F expression plasmids induced improved humoral immune responses in mice compared to vaccination with wild-type cDNA expression plasmids (52). Consequently, viral load was reduced 13-fold in mice immunized with full-length RSV F and 170-fold in mice immunized with the soluble form of RSV F following RSV challenge in comparison to nonimmunized mice. Based on these results, we inserted the codon-optimized open reading frame (ORF) of both full-length RSV F and its soluble form into a replication-deficient adenoviral serotype 5 vector (AdV), generating AdV-F and AdV-Fsol, respectively, to further enhance the immunogenicity and efficiency of the delivered RSV F transgenes. AdVs were chosen because these viral vectors have been extensively studied and have proven their potential as vaccine vectors in multiple successful preclinical studies (reviewed in references 47, 24, and 51). AdVs are also potent inducers of both humoral and cellular immune responses in animal models and in humans (48, 49, 55). Furthermore, convenience of vector design, ease of handling and a robust antigen expression make AdVs a promising vaccine delivery platform. Another main advantage is their natural tropism for mucosal surfaces, which makes adenoviral vaccines convenient for the purpose of immunization against respiratory pathogens that preferentially initiate infection at the mucosal site (40).
However, AdV vaccines expressing the wild-type RSV F protein were tested in several animal models without achieving convincing protection against RSV challenge (13, 22, 23). This might be due to poor RSV F expression levels caused by premature polyadenylation, which could be overcome by codon optimization (53). Hence, here we used the codon-optimized RSV F based AdVs AdV-F and AdV-Fsol and evaluated their potential as RSV vaccines, showing greatly improved vaccine efficacy.
Replication-deficient AdVs carrying the codon-optimized RSV F gene (GenBank database entry EF566942) encoding the full-length (AdV-F) or the soluble form (AdV-Fsol) were constructed according to the AdEasy adenoviral vector system (Qbiogene, Heidelberg, Germany) (20). The coding sequence for the soluble form of RSV F (Fsol) (amino acids 1 to 524 of full-length RSV F, according to GenBank database entry EF566942) was constructed as described previously by deleting 156 nucleotides coding for 52 amino acids of the C terminus of the full-length F (53). AdVs containing the codon-optimized RSV F or RSV Fsol gene were constructed by excising the coding sequences from previously described DNA expression plasmids with XhoI and HindIII (pFsyn and pFsynED, respectively) (53) and subsequently subcloning these fragments into a derivative of p-ShuttleTetO2(pS-DP-tetO2) downstream of a bidirectional tetracycline-dependent promoter (30, 38). The corresponding recombinant adenoviral plasmids were obtained by homologous recombination of pS-DP-tetO2 with recombinant AdV backbone using electroporation into electrocompetent BJ5183 bacteria and subsequent purification of recombinant clones. Recombinant adenoviral DNA was isolated and transfected into T-REx-293 cells (Invitrogen, Karlsruhe, Germany) using calcium phosphate coprecipitation as described previously to generate the recombinant AdV-F and AdV-Fsol (9). Since T-REx-293 cells stably express the tetracycline repressor, the expression of the inserted RSV F constructs downstream of the Tet operator containing a cytomegalovirus promoter was suppressed. Hence, the T-REx-293 cell line is a derivative of 293 cells that can be used as a complementing packaging cell line for the propagation of recombinant AdVs. As a control, AdV was produced by the same method using pS-DP, encoding ovalbumin (OVA). For purification and concentration, the Vivapure Adenopack 100 was used according to the protocol provided by the manufacturer (Vivascience, Sartorius, Germany). Titer of the vector preparations was determined by photometric DNA measurement (recorded as optical particle units) and by titration using serial tenfold dilutions on HEK293 cells to determine 50% tissue culture infective doses and performing immunocytochemical staining as described below to detect RSV F-transferring adenoviruses (measured as gene-transducing units [GTU]). Replication deficiency was verified by infecting A549 cells with a high dose (at least 108 GTU/ml) of AdV-F and AdV-Fsol. While an increase in titer was found after amplification in HEK293 cells, titers were below the level of detection of 10 GTU/ml after one or two amplification cycles in A549 cells.
Infected T-REx-293 cells were lysed 48 h following infection. Protein separation was performed under nonreducing conditions (without 2-mercaptoethanol) on sodium dodecyl sulfate-8 to 10% polyacrylamide gels. After blotting onto nitrocellulose membranes, proteins were incubated at 4°C over night with monoclonal antibody against RSV F (18F12) (2). After washing, the membrane was incubated with horseradish peroxidase-coupled secondary antibody (Dako, Hamburg, Germany), and detected proteins were visualized by enhanced chemiluminescence (Biozym, Hamburg, Germany).
For detection of RSV F protein expression in cells infected with AdV-F or AdV-Fsol, the cells were fixed with ethanol 48 h following infection and were incubated with monoclonal antibody against RSV F (18F12) (2). After washing, the cells were incubated with horseradish peroxidase-coupled secondary antibody (Dako), and the RSV F protein was visualized by using AEC solution (Sigma, Taufkirchen, Germany). In order to detect the soluble form of the RSV F protein, its secretion was blocked by addition of 10 mM monensin (final concentration, 2 μM) (Sigma) 24 h before the staining.
AdV vaccines were produced as described above. For application of plasmid DNA vaccines (pFsynED), the DNA was purified using an EndoFree plasmid gigakit from Qiagen according to the manufacturer's instructions. The concentration of endotoxin (<0.1 endotoxin unit per vaccine dose) was measured in each preparation by a QCL-1000 chromogenic LAL endpoint assay according to the instruction manual (Lonza, Verviers, Belgium).
Specific-pathogen-free animals (6-week-old female BALB/c mice) were obtained from Charles River (Sulzfeld, Germany) and maintained under specific-pathogen-free conditions in isolated ventilated cages in accordance with national law and institutional guidelines. For immunization, AdVs or DNA vaccines were diluted in 50 μl phosphate-buffered saline (PBS) and were administered in various doses subcutaneously (injection into both hind paws), intramuscularly (injection into the quadriceps of the hind limb), or intranasally (application to the nostrils). Intramuscular and intranasal applications were performed under slight anesthesia with 50 mg of Ketamine (CP-Pharma, Burgdorf, Germany) and 10 mg of Xylazine (Bayer, Leverkusen, Germany) per kg of body weight. All vaccination studies were performed according to a homologous prime-boost vaccination scheme, with animals being given boosters 4 weeks after the first immunization with the same vaccine. Mice in the control groups received the AdV-OVA control or were left untreated (not immunized). Mice were challenged intranasally 3 weeks after the last immunization under Ketamine/Xylazine anesthesia (as described above) with 107 PFU of RSV strain Long in 50 μl PBS as described previously (52). Five days after infection, mice were sacrificed, and bronchoalveolar lavage fluids (BALs) were collected from lungs by applying and removing 1 ml PBS twice. For recovery of lung homogenates, lungs were removed and then processed with a homogenizer in 1 ml PBS to obtain a single-cell suspension. The supernatants from BALs and lung homogenates were collected and stored at −80°C for determination of RSV load at later time points. The Animal Care and Use Committee of the Ruhr-University Bochum approved all described studies.
Stocks of the RSV Long strain were prepared as described previously by infection of HEp2 cells at a low multiplicity of infection (52). Defective interfering particles from our stock were removed by two subsequent limiting dilution steps as described previously (17). Virus titers were determined by infection of HEp2 cells with serially diluted virus stocks and by detection of virus-infected cells using immunocytochemical staining with monoclonal antibody (3C4) against the P protein (58). For large-scale RSV preparation, HEp2 cells were infected with 1 ml virus stock at a dose of 106 PFU/ml in 175-cm2 flasks. After complete cell lysis, cells were scraped off the flask and centrifuged for 10 min at 1,500 × g. Supernatants were pooled, filtered through a 0.45-μm-pore-size sterile filter, and ultracentrifuged at 50,000 × g through a 10% sucrose cushion for 2 h at 4°C. The pellet was resuspended in 10% sucrose containing PBS and stored at −80°C.
Sera were harvested by tail vein puncture and analyzed for immunoglobulin G1 (IgG1) and IgG2a RSV F-specific antibody levels using conventional enzyme-linked immunosorbent assay (ELISA) methods. Briefly, ELISA plates (Maxisorp; Nunc, Wiesbaden, Germany) were coated with heat-inactivated (30 min, 56°C) RSV with 107 PFU/ml overnight at 4°C in carbonate coating buffer (pH 9,5). After three washes with PBS containing 0.05% Tween 20 (PBS-T), the plates were blocked with PBS-T containing 5% fat-free milk powder for 1 h at room temperature. Mouse sera were diluted 1:10 for IgG2a and 1:100 for IgG1 ELISA in PBS-T with 2% fat-free milk, added to each emptied well, and incubated for 1 h at room temperature. After the plates were washed as before, IgG subclass-specific antibodies conjugated with alkaline phosphatase (BD Biosciences, Heidelberg, Germany) were added to each well and incubated for 1 h at room temperature. The plates were washed as before and pNPP substrate solution (Sigma) was added. The reaction was stopped with 1 M sodium hydroxide solution, and absorbance at 405 nm was determined using an ELISA reader (Tecan, Kirchheim, Germany).
Neutralizing-antibody titers were determined by using recombinant green fluorescent protein (GFP)-expressing RSV (rgRSV) (18) in a 96-well neutralization format assay. After serial twofold dilution of mouse sera in Hanks' balanced salt solution (HBSS), complement was inactivated by incubation at 56°C for 30 min. The rgRSV was diluted to a final dose of 50 to 60 GFP-positive plaques per well and was incubated with the diluted mouse sera or without serum as a control for 1 h at 37°C. HEp2-cells (1 × 104) were added to each well and incubated for 48 h. The number of GFP-positive plaques per well was determined by fluorescence microscopy. The highest serum dilution inhibiting rgRSV infection by more than 50% in comparison to the negative control was taken as the neutralizing-antibody titer. The detection limit of neutralizing antibody was set at a 1:6 serum dilution.
Viral RNA was isolated from 140 μl BAL with a QIAamp viral RNA minikit (Qiagen, Hilden, Germany) or from 100 μl lung homogenate supernatant using Trizol reagent (Invitrogen) according to the manufacturer's instructions. The RNA was eluted in 50 μl elution buffer. The RNA in the isolate from the lung homogenate was subsequently quantified using the Molecular Probes RiboGreen RNA procedure (Invitrogen). For one-step RT-PCR, 5 μl of isolated RNA was analyzed by using a QuantiTect probe RT-PCR kit from Qiagen with SYBR green (sense primer, RSA-1 [5′-AGATCAACTTCTGTCATCCAGCAA]; antisense primer, RSA-2 [5′-GCACATCATAATTAGGAGTATCAAT]; modified from reference 54). An RNA standard was prepared by cloning the PCR product into the pCR2.1-TOPO vector using the TOPO TA cloning kit (Invitrogen) and subsequent in vitro transcription by T7 RNA polymerase. In vitro-transcribed RNA was quantified using the Molecular Probes RiboGreen RNA quantitation kit (Invitrogen), which allowed subsequent determination of copy numbers in the RNA standard. Tenfold serial dilutions of the RNA standard were included in each PCR assay. Amplification and detection were performed in a Rotor-Gene (Corbett Life Science, Sydney, Australia) under the following conditions: 1 h at 50°C, 15 min at 95°C, and 45 cycles of 15 s at 95°C and 1 min at 60°C. The specificity of the PCR products was verified by melting point analysis from 45 to 95°C.
Splenocytes were isolated from animal spleens in HBSS using a cell strainer (BD Biosciences). Red blood cells were lysed in 2 ml ACK lysis buffer for 2 min, and the reaction was stopped in 8 ml of HBSS. Cells were collected in R10 medium (RPMI with 10% fetal calf serum and 0.1% β-mercaptoethanol), and 3 × 106 cells per well were plated in 96-well round-bottom plates. CD8+ T cells were stimulated by an RSV F-specific immune dominant peptide (F82-98 [DKYKNAVTELQLLMQ]) for 16 h with 2 μg/ml peptide per well at 37°C, and cytokines were blocked intracellularly by 2 μM monensin (28). Splenocytes stimulated with phorbol myristate acetate (30 ng/ml) and ionomycin (500 ng/ml) induced 16.6% (±8.7%; average for all animals analyzed) interferon-positive CD8+ cells and served as a positive control (data not shown). As a negative control, cells were cultured in R10 medium alone. All antibodies were obtained from BD Biosciences. After stimulation, cells were washed with PBS/bovine serum albumin/azide, and Fc(II/III) receptors were blocked with a mixture of anti-CD16 and anti-CD32 antibody. Thereafter, cells were labeled with an anti-CD8-allophycocyanin (APC) antibody. Cells were washed with PBS and fixed in 2% paraformaldehyde, followed by permeabilization with 0.5% saponin in PBS/bovine serum albumin/azide buffer. Intracellular gamma interferon (IFN-γ) was detected with a phycoerythrin-conjugated anti-IFN-γ antibody. Cells were analyzed in a FACSCalibur instrument using CellQuestPro software (BD Biosciences) and reanalyzed with WinMDI 2.9.
Statistical analyses were performed with GraphPad Prism version 4.02 (GraphPad Software, Inc.). Mean immune responses among groups of mice were compared by analyses of variance followed by pairwise multiple comparison using Tukey's test. For the determination of the arithmetic mean values and standard deviations of viral loads, logarithmically transformed values were used.
Since codon optimization of RSV F-encoding DNA vaccines enhances protective efficacy, AdVs with codon-optimized RSV F ORFs were constructed to analyze the protection of mice against a subsequent RSV challenge. Therefore, we used an E1/E3 deletion-containing AdV to insert the full-length and soluble forms of the codon-optimized RSV F ORFs to generate AdV-F and AdV-Fsol, respectively (Fig. (Fig.1A).1A). Since RSV F is cytotoxic and induces apoptosis (10), the RSV F cDNA was placed under the control of a tetracycline-dependent promoter repressing transgene expression in T-REx-293 cells. Using conventional HEK293 cells without repression, the AdV-F and AdV-Fsol vectors could not be expanded (data not shown). AdV-F and AdV-Fsol vector stocks, produced in T-REx-293 cells, resulted in efficient expression of full-length and soluble RSV F, respectively (Fig. (Fig.1C).1C). Using a monoclonal antibody specific for RSV F in Western blot analysis under nonreducing conditions, two oligomeric RSV F bands could be detected. The lower band (approximately 160 kDa) comigrates with the F protein expressed by RSV-infected HEp2 cells (Fig. (Fig.1B).1B). In comparison, the soluble form of the RSV F protein was detected at approximately 150 kDa, which is consistent with the migration of the soluble form of the ectodomain of RSV F protein observed previously (53). In immunocytochemical assays, cells transduced by AdV encoding full-length RSV F showed a plaque-like structure which may result from syncytium formation due to RSV F expression. The soluble form of RSV F could be detected only after its secretion had been blocked by monensin treatment in single cells in the absence of syncytium formation (Fig. (Fig.1C1C).
To test the immunogenicity and efficacy of the AdV vaccine encoding the full-length RSV F, groups of six BALB/c mice were immunized twice (Fig. (Fig.2).2). Initially, the immune response of AdV-F was compared to the DNA vaccine inducing the most pronounced efficacy (pFsynED) in our previous studies (52). Therefore, the immune response after immunization with 1 × 108 GTU of AdV encoding full-length RSV F was compared to a standard DNA immunization with pFsynED, both given by the subcutaneous route. As controls, one group received an AdV encoding an irrelevant protein (OVA) and one group was left untreated. Two weeks after the last immunization, RSV-binding antibodies of subclasses IgG1 and IgG2a as well as RSV-neutralizing antibodies were detected. Levels of IgG1 antibodies induced by AdV-F vaccination were comparable to those induced by the pFsynED DNA vaccine. In contrast, the amount of IgG2a and the neutralizing-antibody titer were elevated 13-fold, on average, after vaccination with AdV-F (Fig. (Fig.3A).3A). Nonimmunized animals and AdV-OVA-immunized animals showed neither RSV-specific IgG1 and IgG2a antibodies nor RSV-neutralizing antibodies.
In the next experiment, we analyzed the dose dependency of the AdV-F vaccine with respect to the induction of humoral immune responses. Four 10-fold serial dilutions of AdV-F ranging from 2 × 105 to 2 × 109 GTU were given subcutaneously (Fig. (Fig.3B).3B). Sera collected after a single immunization two weeks after priming were tested in the same analysis to control induction of immune responses after priming. The amount of IgG1 antibodies was relatively small in all applied doses of AdV vaccine, whereas the amount of IgG2a antibodies clearly depended on the vaccine dose. Surprisingly, only after priming were the titers of neutralizing antibodies dependent on the dose of AdV-F, while after boosting the induced antibody levels were comparable. Even with the lowest dose (2 × 105 GTU per animal), an efficient neutralizing-antibody titer in the blood of immunized animals was achieved (log2 7.7). All the other immunized groups showed similar induction of neutralizing antibodies after boosting.
In the third experiment, the influence of the route of administration on the humoral immune response was analyzed. Mice were immunized with 1 × 108 GTU by the subcutaneous, the intramuscular, and the intranasal routes (Fig. (Fig.3C).3C). Compared to the previous experiments, the IgG1 and IgG2a antibody titers following vaccination via the subcutaneous route could be reproduced nicely. The intramuscular vaccination induced higher IgG1 and lower IgG2a antibody levels than the subcutaneous immunization. The highest IgG1 and the lowest IgG2a antibody levels were induced by the intranasal route, which also induced neutralizing-antibody titers of up to 1:900. The neutralizing-antibody titer was therefore fivefold higher than following subcutaneous application and eightfold higher than following intramuscular application of the vaccine.
Three weeks after boost immunization animals were challenged with RSV intranasally. Five days after challenge, when the viral load peaked in the lungs, animals were sacrificed and BALs were collected. Cells were isolated from BALs, stained, and analyzed by microscopy to measure the amount of eosinophils after RSV challenge. In all challenged animals used in this study, the amount of eosinophils was not elevated (data not shown). Additionally, viral RNA was extracted from BAL and was analyzed by quantitative RT-PCR for determination of RSV load in animal lungs. In the first experiment, RSV-RNA copy numbers were reduced 50-fold in the lungs of the DNA immunized group compared to the nonimmunized group and to the group immunized with AdV-OVA (Fig. (Fig.4A).4A). Vaccination with the AdV vaccines suppressed the viral load below the level of detection. All differences of the copy numbers were statistically significant. Since the measured viral load in the BAL was below the detection limit in the AdV immunized group, we decided to isolate whole RNA from lung homogenates to expand the detection range. The viral copy numbers were determined by quantitative RT-PCR and normalized by the concentration of total cellular RNA. In the dose escalation experiment, subcutaneous immunization with all doses of AdV-F used led to a similar reduction, by a factor of approximately 250-fold (Fig. (Fig.4B).4B). The reduction in the subcutaneously immunized group in the route experiment was comparable to that in the dose escalation experiment (Fig. (Fig.4C).4C). The highest efficacy of 50,000-fold reduction of viral load was achieved by intranasal application of the vaccine, whereas intramuscular immunization resulted in the lowest efficiency.
We and others have shown previously that immunogenicity could be enhanced by using the soluble form of the F protein of RSV instead of the full-length form (31, 52). Therefore, we performed a side-by-side comparison of the vaccines encoding the full-length and soluble forms of the protein. The induction of RSV F-specific IgG1 and IgG2a antibodies was comparable between immunized groups (Fig. 5A and B). Neutralizing-antibody titers were slightly higher using the soluble form of RSV F after a second immunization but not after the first (Fig. (Fig.5C).5C). Both AdV vaccines reduced the viral load after challenge to below the level of detection, indicating a >60.000-fold reduction (Fig. (Fig.5D5D).
For a vaccine to be considered successful, long-lasting immunity is required. Therefore, we immunized mice twice with AdV-F and determined the antibody titer every 4 weeks up to 35 weeks after boosting (Fig. (Fig.6A).6A). The neutralizing antibodies slowly declined by less than a factor of 3 but remained at high levels until 35 weeks after the second immunization. Challenging the mice led to a reduction of viral load by a factor of 100-fold in comparison to nonimmunized control animals (Fig. (Fig.6B).6B). Although the efficacy was lower at this late challenge time point, there was still a substantial reduction of viral load.
One of the advantages of AdV vaccines is the induction of strong cytotoxic-T-cell responses. Therefore, we measured the induction of IFN-γ-producing CD8+ T cells after in vitro stimulation with RSV F peptides. Since AdV-F and AdV-Fsol reduced viral load after challenge to comparable levels, we investigated the cellular immune responses for only the AdV-Fsol vaccine. The AdV-Fsol vaccine was selected to avoid safety concerns due to the use of a fusion-competent, syncytium-inducing viral immunogen. Mice were immunized once or twice with AdV-Fsol, and levels of IFN-γ-producing CD8+ T cells were analyzed in comparison to levels in RSV-infected animals and nonimmunized control animals by intracellular cytokine staining (Fig. (Fig.7).7). Cells of each animal were either stimulated with an immune dominant RSV F peptide (F82-98) or left unstimulated. Without stimulation, the percentage of IFN-γ-producing CD8+ cells ranged from 0.2 to 1.3%. Similar values were obtained after RSV F peptide stimulation of cells isolated from unvaccinated animals, RSV-infected animals, and animals immunized only once with AdV-Fsol. In contrast, after prime-boost immunization with AdV-Fsol, the percentage of IFN-γ-producing CD8+ cells increased to 5.8 to 15.2%.
The attempt to develop a protective vaccine against RSV underwent a major setback 40 years ago, when the FI virus vaccine resulted in an enhanced disease after RSV infection and the deaths of two vaccinated children (reviewed in reference 6). In the present study, we show that immunization of mice with AdVs carrying the codon-optimized form of the fusion protein of RSV led to a long-lasting protective immune response to RSV without signs of illness after challenge or vaccination-related adverse effects. The mouse model for RSV infection is well established and often used for vaccination studies (5, 27, 42, 44). We observed here that viral load after RSV challenge is reduced by a factor of about 700-fold after two immunizations with AdV-F given via the subcutaneous route. Furthermore, even by using low doses of AdV-F of 2 × 105 GTU twice, we induced high levels of neutralizing-antibody titers (average, log2 7.7) and a reduction of viral load of more than 250-fold in the lungs after challenge compared to the load in nonimmunized animals. Protective immunity was mediated to comparable extents by both low and high vaccination doses. Compared to subcutaneous immunization, intranasally administered AdVs induced fivefold-higher neutralizing-antibody titers in serum, and viral load reduction was expanded to a factor of up to 60,000, indicating that the intranasal route is very efficient and convenient for AdV immunization against RSV in mice. These results are consistently more convincing than those obtained recently with the AdV vaccine expressing ten-amino-acid peptide repeats of the soluble core domain of RSV G (61). After a single intranasal immunization with RSV G-based AdVs, viral loads were reduced only around 20-fold compared to loads in control mice.
We suggest that the induced immune responses after vaccination are not mediated by p53-dependent apoptosis, which was recently shown to be triggered by the expression of the RSV F protein (10). The F-mediated apoptosis is shown to be induced by the full-length RSV F but not by the truncated form, whereas the protective efficacies of and the immune responses induced by the two constructs are comparable (Fig. (Fig.44).
We and others have previously shown that immunogenicity and efficacy of the soluble form of RSV F were more potent than those of the full-length RSV F in DNA vaccination regimens (31, 52). Regarding viral vector vaccines, it was shown with both vaccinia virus-based vectors and chimeric parainfluenza virus type 3-expressing vectors that the soluble form of the F protein induces higher immune responses and is biased to a more Th2-like response in the absence of any vaccine-augmented disease (4, 50). In the present study, the soluble form was as efficient as the full-length construct of RSV F. Only slight differences in the amounts of IgG1 and IgG2a antibody subclasses could be detected. Viral load after both vaccinations was reduced below the level of detection. Therefore, we were unable to detect any differences in immunogenicity or viral load reduction after challenge. In addition to the strong antibody response, two immunizations with AdV-Fsol resulted in a substantial cellular immune response: we detected 5.8% and 15.2% IFN-γ-producing CD8+ T cells after stimulation with an RSV F-peptide in two independent experiments. In contrast, we found only low levels of IFN-γ-producing CD8+ T cells after stimulation in RSV-infected animals or singly immunized animals.
The preexisting immunity against AdV5 has been shown to reduce the vaccine efficacy (3, 33, 60). However, others observed that the preexisting immunity could be overcome by using larger amounts of vector or using alternate routes of administration (49, 60). Therefore, we analyzed the influence of preexisting immunity to AdV serotype 5 in our animal model by immunizing twice with AdV expressing an irrelevant protein before immunizing twice with AdV-F (data not shown). The amount of RSV F-specific antibodies and the induced titers of neutralizing antibodies against RSV as well as the reduction of viral load after challenge in preimmunized animals were similar to those in naïve mice, which might be due to the high dose (108 GTU/animal) of AdV used for vaccination. Furthermore, others have shown that the preexisting immunity against adenovirus serotype 5 did not lower the efficacy of the AdV vaccines used in a comparable mouse model (61).
In a previous study, we compared the immunogenicity and efficacy of DNA vaccination to those of an FI RSV vaccine (52). Since we did not change the experimental setup and since the DNA vaccine used in our previous study was also used in the present study, we can compare the immunogenicity of the AdV vaccine to that of the FI RSV vaccine. The AdV vaccine, particularly after intranasal immunization, seems to induce higher titers of neutralizing antibodies than the FI RSV vaccine. However, both vaccines suppressed viral load to below the level of detection, excluding conclusions on the comparative efficacy of the vaccines.
AdVs harbor several advantages for the application as a vaccine vector. AdVs induce rapid and potent immune responses due to their high transduction efficiency and therefore entail high immunogenicity directed toward the AdV vector proteins and the inserted foreign gene (reviewed in reference 32). As shown previously, the efficacy of vaccination using codon-optimized ORFs is highly advantageous over that of wild-type sequences, since the expression of the wild-type cDNA of RSV F via RNA polymerase II in eukaryotic cells is impaired by premature polyadenylation (52, 53). This study confirms the superiority of codon-optimized ORFs over viral wild-type sequences for vaccination.
In previous studies, AdVs containing the wild-type cDNA form of RSV F and RSV G were tested in dogs, ferrets, and chimpanzees (22, 23). The outcome of vaccination was disappointing, since the immunization of a chimpanzee with the tested AdV vaccines elicited only slight immune responses against RSV (22). In a subsequent study, it was shown that the AdV vaccine could induce protection against RSV in a newly established ferret model, but the different AdV serotypes tested were effective only at high doses (23). Since the human RSV strain used does not replicate in the lungs of ferrets, ferrets are not a very stringent animal model for the RSV vaccination and challenge. In a study published recently, intranasal vaccination with an AdV encoding the wild-type RSV F cDNA reduced viral load of after challenge less than 100-fold (13). In contrast to those wild-type based AdVs, our codon-optimized AdV vaccines induced high antiviral antibody titers and good CD8+-T-cell responses and potently reduced viral load after challenge (more than 60,000-fold). Thus, AdVs carrying the codon-optimized form of RSV F are promising vaccine candidates for further preclinical and clinical development.
We thank M. E. Peeples and P. Collins (NIH, Bethesda, MD) for providing the rgRSV.
This work was supported by a grant from the FoRUM program of the Ruhr-Universitaet Bochum (F467-2005).
Published ahead of print on 23 September 2009.