Abstract
New or improved vaccines against viruses such as influenza, parainfluenza types 1–3, measles, dengue, and respiratory syncytial virus would prevent an enormous burden of morbidity and mortality. Vaccines or vaccine candidates exist against these viral diseases, but all could potentially be improved if the immunogenicity of the vaccine could be enhanced. We found that the immunogenicity in primates of a live-attenuated vaccine candidate for parainfluenza virus type 3, an enveloped RNA virus that is an important etiologic agent of pediatric respiratory tract disease, could be enhanced by expression of granulocyte–macrophage colony-stimulating factor (GM-CSF) from an extra gene inserted into the genome of a cDNA-derived virus. Expression of GM-CSF by the live attenuated recombinant virus did not per se affect the level of pulmonary viral replication in rhesus monkeys after topical administration, which was 40-fold lower than that of WT parainfluenza virus type 3. Despite that, the expressed extra gene augmented the virus-specific serum antibody response to a level that was (i) 3- to 6-fold higher than that induced by the same virus with an unrelated RNA insert of equal length and (ii) equal to the response induced by nonattenuated WT virus. In addition, topical immunization with the attenuated virus expressing GM-CSF induced a greater number of virus-specific IFN-γ-secreting T lymphocytes in the peripheral blood of monkeys than did immunization with the control virus bearing an unrelated RNA insert. These findings show that the immunogenicity of a live-attenuated vaccine virus in primates can be enhanced without increasing the level of virus replication. Thus, it might be possible to develop live-attenuated vaccines that are as immunogenic as parental WT virus or, possibly, even more so.
Live-attenuated vaccines offer a very potent means to prevent human viral diseases, and a number are in routine use, such as vaccines for polio, measles, mumps, rubella, adenovirus, and varicella. Historically, live-attenuated cowpox or, subsequently, vaccinia virus has been used since 1796 to immunize against and finally eradicate smallpox as a naturally occurring disease, and the use of a live-attenuated poliovirus vaccine has brought polio close to eradication. In general, live-attenuated vaccines are safe and efficacious. In at least two cases, namely respiratory syncytial virus (RSV) and measles viruses, the use of a live attenuated vaccine offers a particular safety advantage because it does not induce the immune-meditated disease potentiation associated with an inactivated vaccine (1). The immune response to a live-attenuated vaccine is both local and systemic, involves innate and adaptive immunity, and stimulates both humoral and cell-mediated components.
However, one obstacle to the development of live-attenuated vaccines is the difficulty in achieving a satisfactory level of attenuation without severely compromising immunogenicity. The strength of the immune response is, as a first approximation, a function of the amount of antigen expressed, with the exception of very high or very low antigen doses. For example, a recombinant human parainfluenza virus type 3 (HPIV3) mutant in which expression of the accessory C gene was ablated by reverse genetics was highly attenuated in hamsters and induced a serum antibody titer that was 28-fold lower than that induced by its WT HPIV3 parent (2). Another recombinant HPIV3 vaccine candidate that contained a number of attenuating point mutations, designated rcp45-456, induced a 48-fold lower serum antibody titer in hamsters than its less attenuated progenitor, rcp45 (3). Finally, the attenuated RSV vaccine candidate cpts248/404 induced a titer of RSV-neutralizing serum antibodies in chimpanzees that was 9-fold lower than that induced by WT RSV infection (4).
Although various components of the innate and acquired arms of the immune system contribute to restricting the replication of viruses such as RSV, HPIV3, measles viruses, and poliovirus, serum neutralizing antibodies play a major role in the long-term protection conferred by vaccination or natural infection. This is illustrated, for example, by the protective effect against many viruses provided for young infants by maternally derived serum antibodies. This also was directly demonstrated in mice with RSV, where immunization with a vector-expressed antigen that stimulated RSV-specific CD8+ cytotoxic T cells induced resistance to virus challenge that quickly waned over a period of several months, whereas parenteral immunization with viral antigens that induced durable RSV-neutralizing serum antibodies conferred long-term protection (5). Similar observations have been made with HPIV3 in hamsters (6). More recently, the effectiveness of serum antibody in immunoprophylaxis was demonstrated by the clinical efficacy of Synagis (palivizumab), a humanized RSV-neutralizing monoclonal antibody that is used for passive immunoprophylaxis of infants at high risk for severe RSV disease (7). Therefore, it seems likely that a strategy that provides for the stimulation of a higher neutralizing antibody titer would improve the efficacy of a live-attenuated vaccine.
Granulocyte–macrophage colony-stimulating factor (GM-CSF) is a major stimulatory cytokine for the development and function of dendritic cells and macrophages (8). We recently showed that expression of GM-CSF from a transcriptional cassette inserted into the genome of RSV activated and dramatically increased the number of pulmonary lymphoid and especially myeloid dendritic cells and macrophages when inoculated into the respiratory tract of mice. Expression of GM-CSF reduced the level of replication of the recombinant RSV containing the gene, but nonetheless the modified RSV induced antibody titers similar to those stimulated by WT RSV infection (9). In the present study, we studied in a primate model the impact of GM-CSF expression on the immunogenicity of a live-attenuated vaccine candidate for HPIV3, an important agent of pediatric respiratory tract disease worldwide (10).
Materials and Methods
Construction of Antigenomic cDNAs and Recovery of Recombinant Viruses.
We previously described the construction and recovery of a fully infectious recombinant chimeric virus called B/HPIV3, which consists of bovine PIV3 (BPIV3) in which the HN and F genes have been replaced by their HPIV3 counterparts (10). Here, the full-length antigenomic cDNA of B/HPIV3 used for recovery was modified by the insertion of a cDNA encoding human GM-CSF. We note that human and rhesus monkey GM-CSF are identical in length and have eight amino acid differences (GenBank accession nos. M10663 and AY007376). The human GM-CSF cDNA (generously provided by Elizabeth Jaffee, Johns Hopkins University School of Medicine, Baltimore) was modified by PCR to be flanked on the upstream end by a BPIV3-specific gene-end signal, intergenic trinucleotide, and gene-start signal, and to be flanked on both ends by NotI sites. This cDNA was inserted into a unique NotI site that had been previously engineered into the B/HPIV3 antigenomic cDNA at the octanucleotide TCCAGATC at positions 3674–3681, located in the downstream noncoding sequence of the P gene (11). In the resulting construct, the octanucleotide was replaced by the following sequence: GCGGCCGCAAGTAAGAAAAACTTAGGATTAAAGAACGTTATG … TGAGCGGCCGC (translational initiation and stop codons of GM-CSF are in bold and the three dots between them represent the ORF; BPIV3-specific gene-end and -start sequences are single and double underlined, respectively; the NotI restriction endonuclease sites are shown in italics). In this configuration, the GM-CSF insert is under the control of a separate set of BPIV3 gene-start and gene-end signals and is the third gene in the B/HPIV3 genome (Fig. 1). For B/HPIV3 bearing the chloramphenicol acetyl transferase (CAT) gene insert (B/HPIV3/CAT), a 432-bp fragment of the CAT cDNA in reverse orientation, followed by a stop-codon, was engineered as described above for GM-CSF and inserted in its place. Both recombinant viruses were recovered in HEp-2 human epithelial cells by cotransfection of the antigenomic cDNA with N, P, and L support plasmids as described (11). In this recovery strategy, the antigenomic RNA and N, P, and L proteins expressed from the cotransfected plasmids assemble into a functional nucleocapsid/polymerase complex and launch a productive infection (16). Viruses were propagated in LLC-MK2 monkey kidney cells, and GM-CSF concentrations in cell culture medium were determined using Quantikine human GM-CSF Colorimetric Sandwich ELISA (R & D Systems).
Fig 1.
Schematic representation of the RNA genomes of B/HPIV3/GM-CSF and B/HPIV3/CAT. BPIV3 genes are represented by white bars, HPIV3 by gray bars, and the inserted genes (human GM-CSF or bacterial CAT sequence in reverse orientation) by hatched bars. Gene-start and gene-end sequences are indicated by small black bars upstream and downstream, respectively, of each gene. The viral gene designations are as follows: N, nucleocapsid protein; P, phosphoprotein; M, internal virion matrix protein; F, fusion glycoprotein; HN, hemagglutinin-neuraminidase glycoprotein; L, polymerase protein. Extragenic leader (Le) and trailer (Tr) sequences are depicted at 3′ and 5′ ends of the genomic RNAs.
Replication and Immunogenicity in Vivo.
Juvenile rhesus monkeys (2–3 years old, body weight 3–5 kg) were obtained from LABS of Virginia (Morgan Island, SC) and confirmed to be seronegative for PIV3 by a hemagglutinin inhibition (HAI) assay using HPIV3 strain JS as indicator. Monkeys were inoculated with B/HPIV3/GM-CSF or B/HPIV3/CAT (eight monkeys per group) by simultaneous intranasal and intratracheal inoculation with 105 tissue culture 50% infectious dose (TCID50) of the respective virus in a 1 ml inoculum at each site. On day 58, four monkeys from each group and four additional (nonimmunized) monkeys were challenged with WT HPIV3. On days indicated in Table 1, nasopharyngeal swabs and tracheal lavages were obtained as described (12), flash-frozen, and stored at −70°C until all samples were available for titration. In addition, for the monkeys that were challenged with WT HPIV3 on day 58, nasopharyngeal swabs were obtained on days 62–66 (daily), and tracheal lavages on days 62, 64, and 66. Virus was quantitated by plaque assay on LLC-MK2 cell monolayers in 96- or 24-well plates as described (2). Monkey serum samples were obtained on days 0, 28, and 56 and tested for antibodies to HPIV3 by HAI assay (13). Clinical observations were performed daily on days 0, 1–10, and 13 after immunization, and on days 62–66 after challenge on day 58.
Table 1.
Replication of the recombinant viruses in the respiratory tract of Rhesus monkeys
| Virus
|
URT† | LRT† | ||
|---|---|---|---|---|
| Mean peak virus titer | Mean sum of virus titer | Mean peak virus titer | Mean sum of virus titer | |
| B/HPIV3/GM-CSF | 3.1 ± 0.3 | 16.8 ± 2.2 | 2.9 ± 0.3 | 8.7 ± 0.7 |
| B/HPIV3/CAT | 2.9 ± 0.3 | 14.7 ± 1.9 | 2.6 ± 0.4 | 7.1 ± 0.8 |
| WT HPIV3 | 4.7 ± 0.3 | 23.0 ± 1.0 | 3.4 ± 0.3 | 10.6 ± 4.0 |
URT, upper respiratory tract; LRT, lower respiratory tract.
Monkeys (eight animals per group) were inoculated intranasally and intratracheally with 105 TCID50 in a 1-ml inoculum per site. Virus titers for WT HPIV3-immunized animals (10 animals per group) are taken from separate experiments performed under the same conditions.
Log10 TCID50 per ml mean value ± SE. Mean peak virus titer, the mean of the peak titers for individual animals irrespective of the day; mean sum of virus titers, for each animal, the titers for days 1–10 (URT) or days 2, 4, 6, 8, and 10 (LRT) were combined, and the mean was then calculated for each experimental group.
T Lymphocyte Response.
Whole blood was collected into Vacutainer EDTA tubes (BD Biosciences) on days 0, 6, 10, and 21 and centrifuged in Ficoll (ICN) for 30 min at 400 × g and 20°C. The band of peripheral blood mononuclear cells (PBMC) was collected, washed twice with Hanks' balanced salt solution (HBSS) containing 2% FBS (Invitrogen), and pelleted (10 min at 200 × g and 4°C). Cells were resuspended in RPMI-1640 medium (Invitrogen) containing 10% FBS and 10% DMSO and stored in liquid nitrogen. The monkey IFN-γ enzyme-linked immunospot (ELISPOT) assay (Utrecht University, Utrecht, The Netherlands) was performed according to the manufacturer's recommendations. Briefly, the cells were thawed and washed, and 2 × 106 cells in RPMI medium 1640 containing 5% FBS, 2 mM glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, and 25 mM Hepes buffer (all from Invitrogen) were stimulated by co-incubation with 5 × 106 TCID50 B/HPIV3 overnight at 37°C. After this stimulation, the cells were washed twice in RPMI medium 1640, containing the same additions as above, and resuspended in the same medium, and 400,000 cells per sample were added to ELISPOT 96-well plates that were precoated overnight with antibodies specific to monkey IFN-γ. The plates were incubated at 37°C for 5 h, then the cells were washed away. Biotinylated detector antibodies were then added and the plates incubated at 4°C overnight. Plates were washed, incubated with anti-biotin antibody double-labeled with gold particles and a φ-reductase for 1 h at 37°C, and washed again. The spots of silver were then developed with activator solution and counted under magnification. Statistical analysis was done by a Student's t test.
Results
Construction of the Recombinant PIV3 Vaccine Candidate.
The attenuated HPIV3 vaccine candidate under study was B/HPIV3, a recombinant chimeric virus developed by reverse genetics in which the F and HN protective antigen glycoprotein genes are derived from HPIV3 and all of the other genes are from BPIV3 (ref. 11; Fig. 1). BPIV3 is a closely related bovine counterpart of HPIV3 that is attenuated in primates because of a natural host range restriction (11). B/HPIV3 was designed to combine the attenuated backbone of BPIV3 with the major antigenic determinants of HPIV3 and is currently in preparation for clinical trials (11).
For the present study, B/HPIV3 was further modified by the addition of a transcription cassette encoding human GM-CSF (which has eight amino acid substitutions compared with its rhesus monkey counterpart) inserted between the P and M genes (Fig. 1). In this transcription cassette, the ORF for GM-CSF was engineered to be flanked by BPIV3-specific gene-start and gene-end signals such that the inserted cassette would be expressed as a separate mRNA from the recombinant B/HPIV3 genome. Because addition of a transcription cassette can potentially decrease the efficiency of PIV3 replication in vivo (14), a control virus was constructed with a noncoding gene insert of similar length. Specifically, a portion of CAT ORF (432 nucleotides) in reverse orientation was modified to be flanked by gene-start and gene-end sequences and was inserted into the same gene junction. For both viruses, the constructs were designed so that the total length of the antigenomic cDNA was a multiple of 6 and the first nucleotide of the of N and L gene-start sequences was in the second position of the respective hexamer, whereas those of the other genes were in the first hexamer position, thereby complying with conserved features of parainfluenza viruses that are thought to be optimal for viral replication and gene expression (15).
Both viruses were recovered in cell culture from transfected cDNA as described (16). The viruses were designated B/HPIV3/GM-CSF and B/HPIV3/CAT, and were biologically cloned and passaged in monkey kidney LLC-MK2 cells. The presence of the GM-CSF and CAT inserts in the viral genomes was confirmed by RT-PCR, and the GM-CSF insert was analyzed by sequencing with the result that adventitious mutations were not detected (data not shown). To determine the level of expression of GM-CSF by virus-infected cell culture, LLC-MK2 cells in six-well plates were infected at a multiplicity of infection of 2 tissue culture 50% infectious dose (TCID50) units per cell with either virus. The concentration of GM-CSF in the cell culture media 24 h postinfection was 1.25 μg/ml for B/HPIV3/GM-CSF and <1 ng/ml for B/HPIV3/CAT. The presence of the GM-CSF insert had no effect on the efficiency of B/HPIV3 replication in vitro (data not shown), a property that is important for vaccine production. This result is consistent with the finding that recombinant PIV can accept inserts of at least 50% of its natural genome length with little or no effect on its ability to replicate in vitro (17).
Expression of GM-CSF Does Not Restrict the Replication of B/HPIV3 in Monkeys.
Rhesus monkeys were infected by simultaneous intranasal and intratracheal inoculation of 105 TCID50 of B/HPIV3/GM-CSF or B/HPIV3/CAT. Virus replication in the upper and lower respiratory tract was determined by titration of nasopharyngeal swab samples (URT) obtained daily from days 1 through 10, and tracheal lavage samples (LRT) on days 2, 4, 6, 8, and 10 postinfection (Table 1). There was not a statistically significant difference between B/HPIV3/GM-CSF and B/HPIV3/CAT with regard to either the mean of the sum of the daily titers or the mean peak titers in the URT and LRT. The mean of the sum of the daily titers takes into account the amount of virus shed on each day and represents the total virus load during the infection, whereas the mean peak titers reflects the highest titer achieved in each animal. Comparison of the kinetics of virus replication (data not shown) indicated that there was no significant difference in the time of termination of infection for virus expressing GM-CSF, suggesting that the cytokine did not stimulate innate and adaptive immune effectors involved in resolving viral infection.
We also compared replication of both the recombinant viruses with that of WT recombinant HPIV3 (WT HPIV3) from separate experiments, which were performed in an identical manner (Table 1). Both the mean sum of virus titers and the mean peak virus titer in URT, but not in LRT, were significantly lower in animals infected with B/HPIV3/GM-CSF or with B/HPIV3/CAT than in monkeys infected with WT HPIV3. Mean peak virus titer of B/HPIV3/GM-CSF in URT was 40-fold lower than that that of WT HPIV3 (P < 0.01).
After challenge of four monkeys per group with WT HPIV3 on day 58, no virus or low concentrations of the virus (mean peak titer of 101.7 TCID50 per ml or less) were detected in URT and LRT of the animals immunized with B/HPIV3/GM-CSF or B/HPIV3/CAT. In nonimmunized animals, mean peak virus titers (TCID50 per ml) in the URT and LRT was 103.7 and 104.6, respectively. Clinical observations of the animals found no rhinorrhea, cough, sneeze, or other disease signs after immunization with either recombinant virus or after challenge of the immunized animals with WT HPIV3.
Expression of GM-CSF Increases the Immunogenicity of the Virus.
Serum samples were obtained from monkeys described above immediately before and 28 and 56 days postimmunization. Titers of HPIV3-specific serum antibodies were determined by HAI assay. As a control, sera obtained from monkeys infected with 105 TCID50 of WT HPIV3 on day 28 postinfection were titrated in the same HAI assay (Fig. 2). Infection with B/HPIV3/GM-CSF resulted in a significantly higher virus-specific serum antibody titer compared with B/HPIV3/CAT, with the difference in the experiment shown in Fig. 2 being 6.2-fold on day 28 (P < 0.001) and 3.1-fold on day 56 (P < 0.02). Indeed, the immunogenicity of B/HPIV3/GM-CSF equaled, or slightly exceeded, that of WT HPIV3, despite the attenuated replication of the former virus.
Fig 2.
Reciprocal HPIV3-specific serum antibody titers determined by HAI assay for monkeys immunized with the indicated virus. Eight monkeys per group were immunized with B/HPIV3/GM-CSF, B/HPIV3/CAT, or WT HPIV3. Dots represent the serum HAI antibody titers for each monkey on days 28 and 56; bars represent the means. Before immunization, all animals were confirmed to be negative for HPIV3-specific serum antibodies (HAI titers were <1:4).
The kinetics of the PIV3-specific T lymphocyte response were determined by quantifying virus-specific T lymphocytes in peripheral blood before and on days 6, 10, and 21 postimmunization. The cells were cryopreserved and the number of IFN-γ-secreting T cells per 400,000 peripheral blood mononuclear cells (PBMC) was determined by enzyme-linked immunospot (ELISPOT) assay after stimulation in vitro with the parental virus, B/HPIV3. Control experiments with lymphocytes from uninfected animals showed that the stimulation was specific to and completely dependent on prior infection of the animals with B/HPIV3, and that the in vitro stimulation was insufficient alone to induce detectable positive cells (data not shown). Other control experiments with lymphocytes from B/HPIV3-infected animals showed that the in vitro stimulation was essential for the detection of positive cells (data not shown). There was no response to specific stimulation with either virus before immunization, and (as shown in Fig. 3) <20 virus-specific IFN-γ-secreting T cells per 400,000 appeared by day 6. By day 10, the number of IFN-γ-secreting T lymphocytes rose to between 40 and 140, and by day 21, the number decreased to between 18 and 90 per 400,000. Comparison of monkeys immunized with B/HPIV3/GM-CSF or B/HPIV3/CAT demonstrated that the number of virus-specific T lymphocytes in the GM-CSF group was 1.7-fold higher (P = 0.10) on day 10 and 2.5-fold higher (P < 0.05) on day 21.
Fig 3.
Numbers of peripheral blood T lymphocytes isolated after immunization with B/HPIV3/GM-CSF (group A) or B/HPIV3/CAT (group B) that secrete IFN-γ in response to specific in vitro stimulation with B/HPIV3. The number of IFN-γ-secreting T lymphocytes per 400,000 cells was determined by enzyme-linked immunospot. Individual monkeys are indicated by dots, and means are shown by bars.
Discussion
The results presented here indicate that attenuated recombinant vaccine viruses can be modified to induce an elevated antibody response in primates. We previously showed that infection of mice with a recombinant RSV expressing GM-CSF resulted in a dramatic increase in the number of pulmonary dendritic cells and macrophages (9). However, our experience is that immune responses to human viruses in mice can be extremely robust and typically are much greater than responses observed in non-human primates or humans (18). Also, the semipermissive nature of mice for the replication of human pathogens such as RSV and PIV3 limits their usefulness for evaluating live attenuated vaccines. Therefore, it was important to extend the murine study to primates. RSV does not replicate efficiently in monkeys, whereas HPIV3 does, and therefore the present study was performed with an attenuated HPIV3 derivative, B/HPIV3, that was constructed in previous work and is a promising vaccine candidate for HPIV3 that will be evaluated clinically in further work. We have previously shown that insertion of an extra gene comparable in length to that of GM-CSF into the genome of B/HPIV3 resulted in a small, inconsistent decrease in virus replication in the upper respiratory tract of rhesus monkeys and did not affect replication in the lower respiratory tract or the immunogenicity of the virus (19).
The titer of virus-specific serum antibodies induced in rhesus monkeys by B/HPIV3/GM-CSF was 3.1- to 6.2-fold higher than that induced by B/HPIV3/CAT and was at least equal to that induced by WT HPIV3. This result is remarkable considering the 40-fold lower mean peak virus titer in the URT of B/HPIV3/GM-CSF-infected animals compared with those infected with WT HPIV3. An increase of this magnitude in human vaccines arguably would augment the efficacy of a PIV or RSV vaccine and probably be an important improvement for any live-attenuated vaccine.
PIV3, RSV, and other respiratory viruses replicate in the epithelium of the respiratory tract that contains dendritic cells and macrophages, i.e., potent antigen-presenting cells (20). Capture of respiratory viral antigens by dendritic cells probably occurs at the site of primary viral replication and is followed by migration of antigen-loaded cells to regional lymph nodes for presentation of antigen to lymphocytes. GM-CSF plays a central role in the activation of dendritic cells, including antigen capture, processing, and presentation in the context of MHC I and II, B7, and other co-stimulatory molecules (21). In the present study, GM-CSF, produced transiently but abundantly along with viral antigens, presumably acted locally to promote influx, maturation, and activation of dendritic cells and macrophages and to improve antigen presentation to lymphocytes, as demonstrated in our previous study in mice with the recombinant RSV expressing GM-CSF (9). The luminal pulmonary epithelium is the main site of PIV3 replication; therefore, the main stimulatory effect of GM-CSF on the number of virus-specific lymphocytes would be expected to occur in regional lymph nodes, e.g., the tracheobronchial and mediastinal lymph nodes. Nonetheless, we were able to detect an increased number of virus-specific T lymphocytes in the peripheral blood mononuclear cells (PBMC) of monkeys immunized with B/HPIV3/GM-CSF. Because GM-CSF expressed by recombinant RSV was found to stimulate pulmonary CD4+ T cells in mice and because we observed an increased antibody response in primates vaccinated with an attenuated vaccine virus expressing GM-CSF in the present study, we assume that a significant population of antigen-specific T lymphocytes in animals immunized with B/HPIV3/GM-CSF is CD4+ T helper cells.
Safety is an important issue for developing live chimeric viruses as potential vaccine candidates. GM-CSF encoded by the recombinant PIV3 is produced locally at the site of infection (pulmonary epithelium), and transiently during the 7–14 days of virus replication. The possibility exists that the expression of GM-CSF might result in local or systemic toxicity, or that the influx of immune-mediating cells into the lung might cause disease. There were no signs of overt disease in the monkeys in the present study after immunization or challenge, although future studies should investigate this further by direct pulmonary histological examination. Recombinant GM-CSF is already in clinical use for the treatment of certain neutropenic patients, both adults and children (22), and it is currently being evaluated for prophylaxis and treatment of neonatal septicemia (23). In addition, GM-CSF was demonstrated to be effective for the treatment of certain HIV-infected patients by the reducing virus load (24). It has been hypothesized that increased expression of GM-CSF in the respiratory tract might be associated with atopic disease (25), but it is possible that its expression will have a beneficial or neutral effect. Although the use of any new vaccine strategy raises safety issues, these can be addressed in additional studies in primates before the initiation of clinical studies.
The expression of GM-CSF from a recombinant vaccine virus provides a built-in adjuvant delivered at the site of virus replication. Gene inserts in nonsegmented negative-strand RNA viruses, at least, have been found to be surprisingly stable, and the small GM-CSF insert should pose no additional complication to vaccine production. Improved immunogenicity in live attenuated vaccines should provide a higher and more durable level of protection. RSV, the PIVs, and other viruses contribute to an enormous burden of morbidity and mortality worldwide that could be reduced by improved vaccines that are relatively simple to manufacture and administer.
Acknowledgments
We thank Kim-Chi Tran, Ernie Williams, and Fatemah Dawoodi for technical assistance and Dr. Elizabeth Jaffee for providing the human GM-CSF cDNA.
Abbreviations
GM-CSF, granulocyte–macrophage colony-stimulating factor
PIV3, parainfluenza virus type 3
RSV, respiratory syncytial virus
HPIV3, human PIV3
BPIV3, bovine PIV3
CAT, chloramphenicol acetyl transferase
TCID50, tissue culture 50% infectious dose
HAI, hemagglutinin inhibition
URT, upper respiratory tract
LRT, lower respiratory tract
References
- 1.Polack F. P., Auwaerter, P. G., Lee, S. H., Nousari, H. C., Valsamakis, A., Leiferman, K. M., Diwan, A., Adams, R. J. & Griffin, D. E. (1999) Nat. Med. 5, 629-634. [DOI] [PubMed] [Google Scholar]
- 2.Durbin A. P., McAuliffe, J. M., Collins, P. L. & Murphy, B. R. (1999) Virology 261, 319-330. [DOI] [PubMed] [Google Scholar]
- 3.Skiadopoulos M. H., Surman, S. R., St. Claire, M., Elkins, W. R., Collins, P. L. & Murphy, B. R. (1999) Virology 260, 125-135. [DOI] [PubMed] [Google Scholar]
- 4.Crowe J. E., Bui, P. T., Davis, A. R., Chanock, R. M. & Murphy, B. R. (1994) Vaccine 12, 183-790. [DOI] [PubMed] [Google Scholar]
- 5.Connors M., Collins, P. L., Firestone, C. Y. & Murphy, B. R. (1991) J. Virol. 65, 1634-1637. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tao T., Davoodi, F., Cho, C. J., Skiadopoulos, M. H., Durbin, A. P., Collins, P. L. & Murphy, B. R. (2000) Vaccine 18, 1359-1366. [DOI] [PubMed] [Google Scholar]
- 7.Sorrentino M. & Powers, T. (2000) Pediatr. Infect. Dis. J. 19, 1068-1071. [DOI] [PubMed] [Google Scholar]
- 8.Witmer-Pack M. D., Olivier, W., Valinsky, J., Schuler, G. & Steinman, R. M. (1987) J. Exp. Med. 166, 1484-1498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bukreyev A., Belyakov, I. M., Berzofsky, J. A., Murphy, B. R. & Collins, P. L. (2001) J. Virol. 75, 12128-12140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chanock R. M., Murphy, B. R. & Collins, P. L. (2001) in Fields Virology, eds. Knipe, D. M., Howley, P. M., Griffin, D. E., Lamb, R. A., Martin, M. A., Roizman, B. & Straus, S. E. (Lippincott–Raven, Philadelphia), Vol. 1, pp. 1341–1379. [Google Scholar]
- 11.Schmidt A., McAuliffe, J., Huang, A., Surman, S., Bailly, J., Elkins, W., Collins, P., Murphy, B. & Skiadopoulos, M. (2000) J. Virol. 74, 8922-8929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Crowe J. E., Collins, P. L., London, W. T., Chanock, R. M. & Murphy, B. R. (1993) Vaccine 11, 1395-1404. [DOI] [PubMed] [Google Scholar]
- 13.van Wyke Coelingh K. L., Winter, C. & Murphy, B. R. (1985) Virology 143, 569-582. [DOI] [PubMed] [Google Scholar]
- 14.Skiadopoulos M. H., Surman, S. R., Durbin, A. P., Collins, P. L. & Murphy, B. R. (2000) Virology 272, 225-234. [DOI] [PubMed] [Google Scholar]
- 15.Kolakofsky D., Pelet, T., Garcin, D., Hausmann, S., Curran, J. & Roux, L. (1998) J. Virol. 72, 891-899. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Durbin A., Hall, S., Siew, J., Whitehead, S., Collins, P. & Murphy, B. (1997) Virology 235, 232-232. [DOI] [PubMed] [Google Scholar]
- 17.Skiadopoulos M. H., Surman, S. R., Riggs, J. M., Orvell, C., Collins, P. L. & Murphy, B. R. (2002) Virology 297, 136-152. [DOI] [PubMed] [Google Scholar]
- 18.Collins P. L., Purcell, R. H., London, W. T., Lawrence, L. A., Chanock, R. M. & Murphy, B. R. (1990) Vaccine 8, 164-168. [DOI] [PubMed] [Google Scholar]
- 19.Schmidt A. C., Wenzke, D. R., McAuliffe, J. M., St. Claire, M., Elkins, W. R., Murphy, B. R. & Collins, P. L. (2002) J. Virol. 76, 1089-1099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sertl K., Takemura, T., Tschachler, E., Ferrans, V. J., Kaliner, M. A. & Shevach, E. M. (1986) J. Exp. Med. 163, 436-451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sallusto F. & Lanzavecchia, A. (1994) J. Exp. Med. 179, 1109-1118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schaison G., Eden, O. B., Henze, G., Kamps, W. A., Locatelli, F., Ninane, J., Ortega, J., Riikonen, P. & Wagner, H. P. (1998) Eur. J. Pediatr. 157, 955-966. [DOI] [PubMed] [Google Scholar]
- 23.Rondini G. & Chirico, G. (1999) Curr. Opin. Hematol. 6, 192-197. [DOI] [PubMed] [Google Scholar]
- 24.Brites C., Gilbert, M. J., Pedral-Sampaio, D., Bahia, F., Pedroso, C., Alcantara, A. P., Sasaki, M. D., Matos, J., Renjifo, B., Essex, M., et al. (2000) J. Infect. Dis. 182, 1531-1535. [DOI] [PubMed] [Google Scholar]
- 25.Holt P. G. (2001) Thorax 56, 419-420. [DOI] [PMC free article] [PubMed] [Google Scholar]