Packaging and Prefusion Stabilization Separately and Additively Increase the Quantity and Quality of Respiratory Syncytial Virus (RSV)-Neutralizing Antibodies Induced by an RSV Fusion Protein Expressed by a Parainfluenza Virus Vector

ABSTRACT

Human respiratory syncytial virus (RSV) and human parainfluenza virus type 3 (HPIV3) are major pediatric respiratory pathogens that lack vaccines. A chimeric bovine/human PIV3 (rB/HPIV3) virus expressing the unmodified, wild-type (wt) RSV fusion (F) protein from an added gene was previously evaluated in seronegative children as a bivalent intranasal RSV/HPIV3 vaccine, and it was well tolerated but insufficiently immunogenic for RSV F. We recently showed that rB/HPIV3 expressing a partially stabilized prefusion form (pre-F) of RSV F efficiently induced “high-quality” RSV-neutralizing antibodies, defined as antibodies that neutralize RSV in vitro without added complement (B. Liang et al., J Virol 89:9499–9510, 2015, doi:10.1128/JVI.01373-15). In the present study, we modified RSV F by replacing its cytoplasmic tail (CT) domain or its CT and transmembrane (TM) domains (TMCT) with counterparts from BPIV3 F, with or without pre-F stabilization. This resulted in RSV F being packaged in the rB/HPIV3 particle with an efficiency similar to that of RSV particles. Enhanced packaging was substantially attenuating in hamsters (10- to 100-fold) and rhesus monkeys (100- to 1,000-fold). Nonetheless, TMCT-directed packaging substantially increased the titers of high-quality RSV-neutralizing serum antibodies in hamsters. In rhesus monkeys, a strongly additive immunogenic effect of packaging and pre-F stabilization was observed, as demonstrated by 8- and 30-fold increases of RSV-neutralizing serum antibody titers in the presence and absence of added complement, respectively, compared to pre-F stabilization alone. Analysis of vaccine-induced F-specific antibodies by binding assays indicated that packaging conferred substantial stabilization of RSV F in the pre-F conformation. This provides an improved version of this well-tolerated RSV/HPIV3 vaccine candidate, with potently improved immunogenicity, which can be returned to clinical trials.

IMPORTANCE Human respiratory syncytial virus (RSV) and human parainfluenza virus type 3 (HPIV3) are major viral agents of acute pediatric bronchiolitis and pneumonia worldwide that lack vaccines. A bivalent intranasal RSV/HPIV3 vaccine candidate consisting of a chimeric bovine/human PIV3 (rB/HPIV3) strain expressing the RSV fusion (F) protein was previously shown to be well tolerated by seronegative children but was insufficiently immunogenic for RSV F. In the present study, the RSV F protein was engineered to be packaged efficiently into vaccine virus particles. This resulted in a significantly enhanced quantity and quality of RSV-neutralizing antibodies in hamsters and nonhuman primates. In nonhuman primates, this effect was strongly additive to the previously described stabilization of the prefusion conformation of the F protein. The improved immunogenicity of RSV F by packaging appeared to involve prefusion stabilization. These findings provide a potently more immunogenic version of this well-tolerated vaccine candidate and should be applicable to other vectored vaccines.

INTRODUCTION

Human respiratory syncytial virus (RSV) and human parainfluenza virus type 3 (HPIV3) are nonsegmented negative-sense RNA viruses of the family Paramyxoviridae. They are, respectively, the first and second most common viral causes of severe acute respiratory infections in infants and young children (1–3). Despite decades of effort, licensed vaccines and antiviral drugs suitable for routine use are not available for either virus. Major challenges to developing pediatric vaccines against RSV and HPIV3 include the immaturity of the immune system during infancy, immunosuppression by maternal antibodies, inefficient immune protection at the superficial epithelium of the respiratory tract, and vaccine-induced enhanced disease associated with inactivated or subunit RSV and HPIV3 vaccines in virus-naive recipients (4, 5). Vaccines based on live-attenuated RSV strains or on a PIV vector expressing the RSV fusion (F) protein have been shown in experimental animals and clinical studies to be free of priming for enhanced disease in virus-naive children (6, 7), and thus they are being developed as pediatric intranasal (i.n.) vaccines. Immunization performed i.n. with replication-competent RSV or PIV3 strains elicits both local and systemic immune responses and partially evades the immunosuppressive effects of maternal antibodies. Substantial progress has been made toward developing suitable vaccine candidates (7–10).

The PIV3 genome (Fig. 1A) consists of six genes arranged in the following order: nucleoprotein (N), phosphoprotein (P), matrix protein (M), fusion glycoprotein (F), hemagglutinin-neuraminidase (HN), and polymerase (L). A replication-competent recombinant attenuated chimeric bovine/human PIV3 (rB/HPIV3) strain, consisting of the BPIV3 backbone with the F and HN genes replaced with those of HPIV3, was developed previously as a live-attenuated i.n. pediatric HPIV3 vaccine (11, 12). Addition of a supernumerary gene expressing the RSV fusion (F) glycoprotein, the major RSV protective antigen, resulted in a candidate bivalent vaccine for RSV and HPIV3 (13, 14). A version of the rB/HPIV3 vector with wild-type (wt) RSV F inserted in the second gene position (between the N and P genes), called MEDI-534, was previously evaluated in seronegative infants and children (7). Importantly, the vaccine appeared to be suitably attenuated; however, while all vaccine recipients seroconverted against HPIV3, only half developed detectable RSV-neutralizing serum antibodies as analyzed by a microneutralization assay in the absence of added complement. There was also a loss of RSV F protein expression by some of the vector particles in a substantial proportion of nasal washes from vaccinees (15). This indicated the need for further development to improve the immunogenicity and stability of RSV F in this vaccine candidate.

FIG 1.

rB/HPIV3 vectors expressing versions of the RSV F protein modified to contain CT and TMCT domains from the vector BPIV3 F protein. (A) Gene inserts expressing the indicated versions of RSV F were cloned into the second gene position (between the N and P genes) in rB/HPIV3 as previously described (28). HEK/opt encodes a wt RSV F protein (dark blue) (29) containing HEK assignments 66E and 101P (indicated by asterisks; see Results for an explanation) and was expressed from a codon-optimized ORF (GeneArt) (29). B3CT and B3TMCT are derivatives of HEK/opt modified to contain the CT and TMCT domains of BPIV3 F (red). Constructs with the designation DS contain an introduced disulfide bond (S155C and S290C mutations; indicated by stars) that stabilizes RSV in the pre-F conformation (29). (B) The C-terminal sequences of BPIV3 F and the B3CT and B3TMCT forms of F are aligned to that of wt RSV F (42, 53, 54) to indicate their predicted TM and CT domains, based on hydrophobicity and sequence content.

The RSV F protein mediates fusion of the viral envelope with the host cell membrane during entry, and it also mediates fusion between infected cells and neighboring cells later in infection to form syncytia. RSV F is synthesized as a precursor, F₀, that assembles into trimers. Each protomer is cleaved by a host protease into F₂ and F₁ subunits that remain disulfide linked. A transmembrane domain (TM) and a cytoplasmic tail (CT) are located at the C-terminal end of the F₁ subunit. During viral assembly, the CT is thought to interact with internal virion proteins to mediate packaging of the F protein into virions (16–18). Evidence from several paramyxoviruses implicated TM and CT in modulating fusogenic activity, stability, and protein folding (17, 19).

Newly formed RSV F trimers are in a lollipop-shaped prefusion (pre-F) form. An unknown triggering event induces a major, irreversible conformational rearrangement of the RSV F trimer into a stable, cone-shaped postfusion (post-F) form. It is this rearrangement that drives fusion. The pre-F form contains a number of unique epitopes that are highly efficient at virus neutralization but are lost following this rearrangement, in particular the antigenic site Ø on the apex of the pre-F trimer, which is recognized by the monoclonal antibodies (MAbs) D25, 5C4, and AM22 (20–22). Other pre-F-specific epitopes, such as those recognized by MAbs AM14 and MPE8, have also been identified (23–25). There are two antigenic sites (II and IV) preserved on post-F surfaces, but they are less efficient at neutralization than pre-F-specific sites (26, 27). The majority of RSV F-specific antibodies induced by natural infection bind to the pre-F form and account for most of the neutralizing activity in human serum. However, the pre-F form is metastable and is prone to flip prematurely into the post-F conformation, especially when it is not stabilized by interaction with the matrix protein (M). RSV F can be stabilized substantially in the pre-F conformation by structure-based engineering, such as the introduction of a disulfide bond (DS; S155C and S290C mutations) (20).

The present study is part of a systematic evaluation of strategies to enhance the immunogenicity of RSV F expressed from the rB/HPIV3 vector (28, 29). In the present study, we engineered the RSV F protein so that it would be packaged efficiently into the rB/HPIV3 vector particle. The effect on immunogenicity of this packaging modification was evaluated on its own and in combination with the previously described pre-F DS stabilization (20, 29). In general, the effect of packaging a heterologous glycoprotein into a vector particle had not been well characterized, and was the major interest of the present study. The packaging and pre-F stabilization each resulted in substantial increases in the quantity and quality of RSV-neutralizing antibodies in experimental animals. Combining the packaging modification with pre-F stabilization had a strongly additive effect that was particularly evident in nonhuman primates. This provides a much more immunogenic form of this important vaccine candidate, which can now return to clinical testing (7).

MATERIALS AND METHODS

Cells and viruses.

Rhesus monkey LLC-MK2 cells (ATCC CCL-7), African green monkey Vero cells (ATCC CCL-81), and baby hamster kidney (BHK) BSR-T7/5 cells constitutively expressing T7 RNA polymerase (30) were maintained as described previously (29). All rB/HPIV3 vectors were propagated at 32°C in LLC-MK2 cells. wt RSV was propagated at 37°C in Vero cells. The unmodified wt RSV F sequence was that of the recombinant strain A2 produced from our previously described reverse genetic system (31), whose sequence has been reported elsewhere (GenBank accession no. KT992094) (32). The RSV challenge virus was also strain A2. HPIV3 and BPIV3 sequences were from strains JS (accession no. Z11575) and Kansas/15626/84 (accession no. AF178654), respectively.

The following three rB/HPIV3 constructs used in the present study were previously described (29): (i) non-HEK/non-opt, (ii) HEK/opt, and (iii) DS RSV F. (i) The non-HEK/non-opt construct (not shown) expresses the native, unmodified wt RSV F open reading frame (ORF) and was used only in the experiments for Fig. 5 to 7, as a close facsimile of MEDI-534. (ii) The HEK/opt construct (Fig. 1A) expresses an F ORF that was codon optimized for human expression (GeneArt; Life Technologies) and contains two HEK mutations (66E and 101P) that make it identical at the amino acid level to early-passage strain A2 (29, 33): this F protein is the wild type and is the backbone for all of the constructs depicted in Fig. 1A. (iii) The DS construct expresses a version of the HEK/opt F protein that contains the DS mutations S155C and S290C (20, 29) (Fig. 1A). In the present study, the versions of the HEK/opt insert that were modified to contain the CT or TMCT region of BPIV3 F (called B3CT and B3TMCT, respectively) (Fig. 1) were made synthetically (GeneArt; Life Technologies). Flanking AscI sites and BPIV3 regulatory and noncoding gene sequences were added by PCR, using an Advantage HF 2 PCR kit (Clontech, Mountain View, CA) and the following primers: forward, ATCATGGCGCGCCAAGTAAGAAAAACTTAGGATTAATGGACCTGCAGGATGGAACTGCTGATCCTGAAGGC; and reverse, GAGATGGCGCGCCGCTAGCTATCACTGCCGGTTTGTCAGGATG (the AscI restriction site is underlined, and the ORF translation initiation and termination codons are shown in bold). The resulting B3CT and B3TMCT inserts were subcloned into the pCR4-TOPO vector (Life Technologies) as described previously (29). To create the DS/B3CT and DS/B3TMCT constructs (Fig. 1), the DS mutations (S155C and S290C) were introduced into the B3CT and B3TMCT subclones in pCR4-TOPO by using a QuikChange Lightning Multi site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA): for the S155C mutation, the primers GGCGTGGCCGTGTGCAAGGTGCTGCACC and GGTGCAGCACCTTGCACACGGCCACGCC were used; and for the S290C mutation, the primers CAGAGCTACTCCATCATGTGCATCATCAAAGAAGAGG and CCTCTTCTTTGATGATGCACATGATGGAGTAGCTCTG were used. The sequences of the RSV F inserts in pCR4-TOPO were confirmed, and the inserts were cloned into the full-length rB/HPIV3 cDNA at the 2nd gene position, using the AscI site (Fig. 1A). For a detailed diagram (including the flanking regulatory and noncoding sequences in the RSV F insert), see the previously described rB/HPIV3-F2 construct (see Fig. 1 of reference 28): all of the constructs in the present study conformed to this example. The genome nucleotide length for each construct conformed to the “rule of six” (34), and the phasing of the RSV F insert in the second position was identical to the phasing of the P gene at the same position. Phasing of the vector genes was maintained as their native phasing (35).

FIG 5.

Protection of immunized hamsters against wt RSV challenge. Hamsters (n = 6 per virus) that were immunized with the indicated vectors or wt RSV (from the results shown in Table 2) were challenged i.n. on day 30 postimmunization with 10⁶ PFU of wt RSV in a 0.1-ml inoculum. At 3 days postchallenge, hamsters were sacrificed, and nasal turbinates (A) and lungs (B) were collected. Tissue homogenates were prepared, and virus titers were determined by plaque assay on Vero cells at 37°C. Each symbol represents the RSV titer of an individual animal. Mean viral titers for the groups are shown as horizontal lines, with the numerical values indicated, and the statistical significance of differences between mean titers was determined by one-way ANOVA with the Tukey-Kramer test, using Prism software, and is indicated as follows: ns, not significant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; and ****, P ≤ 0.0001. The detection limit of the assay was 2.7 log₁₀ PFU/g of tissue and is indicated with a dotted line. The data for rB/HPIV3, HEK/opt, DS, and wt RSV were published previously (29) and were included in this figure for comparison.

FIG 7.

RSV- and HPIV3-neutralizing serum antibodies induced in rhesus macaques immunized with rB/HPIV3 vectors expressing RSV F. Sera were collected from the immunized monkeys at 0, 14, 21, and 28 days as part of the experiment described for Fig. 6. At 28 days postimmunization, all monkeys were challenged by the combined i.n. and i.t. routes with 10⁶ PFU of wt RSV in a 1-ml inoculum per site, and sera were collected on days 35 and 56. (A) Mean titers of HPIV3-neutralizing serum antibodies determined with HPIV3 PRNT₆₀ in the presence of complement. (B) Mean titers of RSV-neutralizing serum antibodies determined with RSV PRNT₆₀ in the presence of complement. (C) Titers of RSV-neutralizing serum antibodies at day 28 postimmunization determined by RSV PRNT₆₀ in the absence of complement. Each symbol represents an individual rhesus monkey, and the mean titer for each group is indicated as a solid horizontal line, with the mean value shown beside the line. For all panels, the statistical significance of differences in mean PRNT₆₀ values between pairs of groups at each time point was determined by the pairwise Student t test, and P values are indicated by asterisks (**, P ≤ 0.01; and ***, P ≤ 0.001). The detection limit of each assay is shown with a dotted horizontal line.

rB/HPIV3-RSV-F viruses were recovered in BHK BSR-T7/5 cells constitutively expressing T7 RNA polymerase as described previously (28). Recovered virus was passaged twice on LLC-MK2 cells at 32°C, and the resulting stock was used for all infections. Viral sequences were confirmed in their entirety (except for the 30 and 120 nucleotides at the 3′- and 5′-terminal ends, respectively, due to the positioning of the sequencing primers) by sequencing the RT-PCR products amplified from the viral RNA template. This demonstrated that the viruses were free of adventitious mutations at levels detectable by this method.

Fluorescence double-staining plaque assay.

As previously described (29), plaques on Vero cell monolayers were immunostained with a mixture of primary antibodies (i.e., RSV F-specific MAbs [36] and HPIV3-specific hyperimmune rabbit serum [28]) followed by secondary antibodies (i.e., IRDye800 CW-conjugated goat anti-rabbit and IRDye680 LT-conjugated goat anti-mouse). Plates were scanned with an Odyssey infrared imaging system (LiCor) at wavelengths of 800 nm and 680 nm. Fluorescent plaques were pseudocolored to appear green (800 nm) and red (680 nm) to indicate the expression of HPIV3 proteins and the RSV F protein, respectively; plaques expressing both RSV F and HPIV3 proteins appeared yellow when the two colors were merged (680 and 800 nm).

Analysis of RSV F expression by Western blotting.

Vero cells in 12-well plates (∼2 × 10⁵ cells per well) were infected with rB/HPIV3-RSV-F strains at a multiplicity of infection (MOI) of 10 50% tissue culture infective doses (TCID₅₀) per cell or with wt RSV at 3 PFU per cell and then incubated at 32°C. At 48 h postinfection (p.i.), the monolayers were washed once with ice-cold phosphate-buffered saline (PBS), lysed with 200 μl of 1× LDS buffer with 1× reducing agent (Life Technologies), and then boiled at 95°C for 5 min. Twenty-microliter aliquots of reduced, denatured lysate were loaded onto 4 to 12% NuPAGE gels (Novex-Life Technologies), with NuPage antioxidant reagent (Novex-Life Technologies) added in the running buffer in the cathode chamber. Membranes were probed with murine anti-RSV F MAb (ab43812; Abcam, Cambridge, MA), rabbit anti-HPIV3 hyperimmune antiserum (28), and goat polyclonal anti-glyceraldehyde-3-phosphate dehydrogenase (anti-GAPDH) antibody (G8795; Sigma-Aldrich, St. Louis, MO), followed by IRDye 800CW-conjugated donkey anti-mouse and IRDye680-conjugated donkey anti-goat infrared dye secondary antibodies (LiCor, Lincoln, NE). Membranes were scanned and the fluorescence intensities of protein bands were quantified using an Odyssey infrared imaging system.

Western blot analysis of RSV F packaged in virus particles.

LLC-MK2 cells were infected with rB/HPIV3 vectors at an MOI of 0.01 TCID₅₀ per cell, and Vero cells were infected with wt RSV at an MOI of 0.01 PFU per cell. The rB/HPIV3 vector-infected LLC-MK2 cells were incubated at 32°C for 4 days, and the wt RSV-infected Vero cells were incubated at 37°C for 6 days before the medium supernatant was harvested. The medium supernatant of infected cells was clarified by low-speed centrifugation at 300 × g for 5 min and then subjected to ultracentrifugation in discontinuous 30%-60% (wt/vol) sucrose gradients. Purified viruses were pelleted at 5,000 × g, washed twice with 1× PBS, and lysed in RIPA buffer. Protein concentrations in the lysates were quantified with a Pierce bicinchoninic acid (BCA) kit (ThermoFisher Scientific, Grand Island, NY). Approximately 0.5 μg of protein of each viral lysate was processed for Western blotting as described above and analyzed using an RSV F-specific MAb (ab43812), an HPIV3 HN-specific antipeptide (amino acids 3 to 19) serum, and the anti-HPIV3 hyperimmune rabbit antiserum described above (28). The HPIV3 F protein was detected with affinity-purified polyclonal antibodies generated by hyperimmunizing rabbits with the purified recombinant HPIV3 F protein ectodomain. Secondary antibodies and infrared scanning were used as described above.

Immunoelectron microscopy.

Immunoelectron microscopy analysis was carried out as previously described (37), with a few modifications. Sucrose gradient-purified virus particles were inactivated with 2% paraformaldehyde in 1× PBS. Ten-microliter droplets of virus suspension were applied to freshly glow-discharged 200-mesh Formvar/carbon-coated nickel grids (Electron Microscopy Sciences, Hatfield, PA) for 10 min at room temperature. Immune labeling steps were performed in a BioWave microwave processor, equipped with a ColdSpot circulating water load, at approximately 170 W and 24°C. After adsorption, grids were washed twice for 1 min each in PBS, blocked for 2 min in 1× PBS containing 2% globulin-free bovine serum albumin (BSA; Sigma-Aldrich), and probed using a microwave cycle of 2 min on, 2 min off, and 2 min on (2:2:2) with a mouse anti-RSV F MAb (131-2A; Millipore) at a 1:50 dilution in blocking buffer. Grids were washed twice for 1 min each in blocking solution and then labeled using a 2:2:2 microwave cycle with goat anti-mouse IgG antibody conjugated to 5-nm-diameter gold particles (Ted Pella, Inc.). The grids were washed for 1 min each in two changes of blocking buffer, two changes of PBS, and three changes of deionized water and then negatively stained with methylamine tungstate (NanoProbes, Inc., Yaphank, NY). The samples were examined using a model H7500 transmission electron microscope (Hitachi High Technologies, Schaumburg, IL) at 80 kV. All images were acquired by using an XR100 digital camera system (Advanced Microscopy Techniques, Danvers, MA).

Hamster studies.

Hamster studies were approved by the NIH Institutional Animal Care and Use Committee (IACUC). Six-week-old Golden Syrian hamsters (Harlan Laboratories) were confirmed to be seronegative for HPIV3 and RSV by using HPIV3-specific hemagglutination inhibition (HAI) and RSV neutralization assays, respectively, and were immunized i.n. with 10⁵ TCID₅₀ of rB/HPIV3 vectors and 10⁶ PFU of wt RSV (A2 strain) in a volume of 0.1 ml. To assess viral replication, infected hamsters were sacrificed on day 3 and day 5 (6 animals per virus per day), in separate studies. Nasal turbinates and lungs were collected and homogenized in L15 medium at a ratio of 1:10 (wt/vol), and viral titers in the homogenates were determined by plaque assay on Vero cells at 32°C. In a separate study for immunogenicity and protective efficacy, additional hamsters in groups of 6 per virus were inoculated in the same way, and sera were collected on day 28 p.i. Titers of RSV- and HPIV3-neutralizing antibodies in sera were quantified by a 60% plaque reduction neutralization test (PRNT₆₀) in the presence or absence of guinea pig complement (Lonza) as described previously (29). On day 30 p.i., hamsters were challenged i.n. with 10⁶ PFU of wt RSV per hamster in a volume of 0.1 ml. Hamsters were sacrificed 3 days after challenge. The titers of replicated wt RSV in the nasal turbinates and lungs were quantified by plaque assay on Vero cells as described above.

Evaluation of selected viruses in rhesus macaques.

The nonhuman primate study was approved by the NIH IACUC. Rhesus macaques which were seronegative for HPIV3 and RSV by HPIV3 HAI and RSV neutralization assays, respectively, were inoculated by the combined i.n. and intratracheal (i.t.) routes, with 10⁶ TCID₅₀ of rB/HPIV3 vectors given via each route. Five, five, and four monkeys were infected with the non-HEK/non-opt, DS, and DS/B3TMCT vectors, respectively. On day 28, all monkeys were infected i.n. and i.t. with 10⁶ PFU of wt RSV. Nasopharyngeal swab and tracheal lavage samples were collected daily and every other day, respectively, during the first 2 weeks after infection with rB/HPIV3 vectors and also after wt RSV challenge. Titers of rB/HPIV3 vectors in collected samples were determined by TCID₅₀ assays in LLC-MK2 cells at 32°C, and titers of the challenge wt RSV strain were determined by plaque assay on Vero cells at 37°C. Sera were collected on days 0, 14, 21, 28, 35, and 56. The RSV- and HPIV3-neutralizing serum antibody titers were determined by PRNT₆₀ in the presence or absence of guinea pig complement as described above. The NIH guide and the Public Health Service policy on the humane care and use of laboratory animals served as the guidelines for the care and use of hamsters and monkeys in this study.

Octet competition assays to measure pre-F-specific antibodies in sera.

Sera were assessed for binding to the pre-F or post-F protein in the presence or absence of competing soluble post-F by bilayer interferometry (BLI), using an Octet Red384 instrument (ForteBio). Briefly, the RSV pre-F or post-F protein was loaded onto HIS1K biosensors (ForteBio) via polyhistidine tags for 300 s in 1% BSA in PBS. Typical capture levels were between 2.0 and 2.3 nm, and variability within a row of 8 tips did not exceed 0.2 nm. The loaded biosensor tips were then equilibrated for 60 s in 1% BSA in PBS. The biosensors were then incubated for 300 s with diluted serum samples (diluted 1:20 in 1% BSA in PBS) in the presence or absence of 4 μg untagged post-F. The percentage of pre-F or post-F binding retained in the presence of competing soluble post-F was calculated relative to the binding in the absence of competition. Octet assays were performed at 30°C.

Statistical analysis.

Viral titers (PFU per milliliter) and viral neutralizing titers (PRNT₆₀ titers) were log transformed for data analysis to conform to the equal variance assumption of analysis of variance (ANOVA). Unpaired one-way ANOVA followed by Tukey's multiple-comparison test was used to compare the mean for each group (n = 6) with the mean for every other group in the hamster study. The testing α level was 0.05. P values of <0.05 were considered to be statistically significant. For the rhesus monkey study, the unpaired t test was used to determine the statistical significance of the differences in mean values of neutralizing titers between pairs of groups on days 21 and 28. The test was two-tailed, and P values of <0.05 were considered to be statistically significant. Percentages of pre-F binding and log₂ PRNT₆₀ values for sera were fitted with Pearson's linear regression model, with n, P, and correlation coefficient (R) values indicated on graphs. Statistical analysis was performed with GraphPad Prism 6.0.

RESULTS

Generation of rB/HPIV3 vectors expressing an RSV F protein containing the CT or TMCT domains from the BPIV3 F protein.

We created versions of the RSV F protein, called B3CT and B3TMCT (Fig. 1A), in which the cytoplasmic tail (CT) or the transmembrane and cytoplasmic tail (TMCT) domains (Fig. 1B) were replaced with those of the BPIV3 F protein to allow compatibility with the internal proteins of the rB/HPIV3 vector. The wt RSV F parent that was used for all of the constructs, HEK/opt, was codon optimized and contains the assignments 66E and 101P (called HEK) that make the protein identical at the amino acid sequence level to an early-passage RSV (called HEK-7) of the same RSV A2 strain (29, 33). Parallel constructs (DS/B3CT and DS/B3TMCT) were made that contain the B3CT and B3TMCT domains plus S155C and S290C mutations that introduce a disulfide bond (DS) that substantially stabilizes the pre-F conformation (20, 29). The modified RSV F genes were engineered to be flanked by BPIV3 gene start and gene end transcription signals (of the P and N genes, respectively) and were inserted between the N and P genes of the rB/HPIV3 vector (Fig. 1A). rB/HPIV3 vectors expressing the HEK/opt and DS forms of RSV F (Fig. 1A) were constructed previously (29). The modified RSV F proteins and the modified genes and vectors that encode them are named according to the modifications to F, i.e., B3CT, B3TMCT, DS, B3CT/DS, and B3TMCT/DS.

The rB/HPIV3-RSV-F vectors were rescued and grew efficiently to titers of >10⁷ TCID₅₀/ml in Vero cells (the cell substrate for vaccine manufacture). Nucleotide sequencing showed that each of the recovered viruses was free of detectable adventitious mutations. Each virus preparation was evaluated by a fluorescence double-staining plaque assay to detect expression of the RSV F protein together with HPIV3 antigens (see Materials and Methods). This showed that each virus had uniform plaque morphology and that 99 to 100% of the plaques expressed the RSV F protein (not shown).

Incorporation of RSV F into rB/HPIV3 virion particles.

Equal amounts (0.5 μg protein) of sucrose gradient-purified virus preparations were analyzed by Western blotting to quantify packaging of the RSV F protein and the vector N, HN, and F proteins. The wt RSV F protein was not packaged efficiently into rB/HPIV3 particles (Fig. 2A and B, lanes 2). In contrast, the B3CT and B3TMCT modifications increased the efficiency of packaging of RSV F 19- to 20-fold (Fig. 2A and B, lanes 3 and 4): indeed, the efficiency of packaging was equivalent (per microgram of virion protein) to that of wt RSV virions (Fig. 2A, compare lanes 3 and 4 to lane 5). The further inclusion of the DS mutations (DS/B3CT and DS/B3TMCT) did not affect the packaging efficiency (Fig. 2A and B, lanes 6 and 7). Efficient packaging of RSV F into vector particles did not interfere with the packaging of vector HN (Fig. 2A, lanes 3, 4, 6, and 7) but reduced the packaging of the vector F protein by 50 to 60% (Fig. 2A and C, lanes 3, 4, 6, and 7).

FIG 2.

Packaging of RSV F into rB/HPIV3 virion particles. (A to C) Sucrose gradient-purified preparations of the indicated rB/HPIV3 vectors or wt RSV (0.5 μg of protein per sample) were analyzed by Western blotting using antibodies specific to RSV F, HPIV3 HN, HPIV3 F, and BPIV3 N, as described in Materials and Methods. (A) Western blot images. (B) Quantification of RSV F packaging efficiency, based on the ratio of RSV F to BPIV3 N, calculated from the data in panel A. All ratios were normalized against that of HEK/opt, which was set at 1. Packaging efficiency values are indicated at the top of the bars. The packaging efficiency of wt RSV F was not calculated (NA) because of the lack of a BPIV3 N value. (C) Quantification of HPIV3 F packaging efficiency. The ratio of total HPIV3 F (F₀ + F₁) to BPIV3 N was calculated and normalized against that of the empty rB/HPIV3 vector, which was set at 1. Packaging efficiency values are indicated at the top of the bars. (D to I) Immunoelectron microscopy analysis of RSV F in rB/HPIV3 virions. Sucrose gradient-purified virus preparations of the indicated rB/HPIV3 vectors and wt RSV were fixed with paraformaldehyde and incubated with an RSV F-specific mouse MAb, followed by a goat anti-mouse IgG specific secondary antibody conjugated with 5-nm-diameter gold particles, and subjected to transmission electron microscopy. A representative virion image for each virus is shown, and arrows indicate representative virion-associated gold particles. (D) Filamentous wt RSV particle exhibiting abundant gold particles; (E) spherical empty rB/HPIV3 vector particle that remained unlabeled; (F) HEK/opt virion with only trace amounts of gold particles; (G to I) B3CT, B3TMCT, and DS/B3TMCT virions, respectively, showing abundant labeling with gold particles. The RSV F protein detected on the latter vector particles (G to I) had an abundance and distribution pattern similar to those of wt RSV (D).

Transmission electron microscopy (TEM) of immunogold-labeled, sucrose gradient-purified virus preparations was employed to visualize RSV F in virion particles (Fig. 2D to I). Consistent with the Western blot results, little RSV F was detected on the virion particles of the HEK/opt construct (Fig. 2F), whereas substantially more RSV F was observed with the B3CT, B3TMCT, and DS/B3TMCT constructs (Fig. 2G to I). The RSV F protein detected on these vector particles had an abundance and distribution pattern similar to those of wt RSV (Fig. 2D).

Replication of rB/HPIV3 vectors and expression of RSV F in infected Vero cells.

Multicycle replication of rB/HPIV3 vectors was evaluated in Vero cells, which are the cell substrate for vaccine manufacture (Fig. 3A and B). Each of the vectors expressing RSV F was slightly more restricted for replication than the empty rB/HPIV3 vector, but there was little difference between the RSV F-expressing vectors (Fig. 3A and B). B3CT in particular replicated somewhat slower during exponential phase (i.e., days 1 and 2), but it eventually reached high titers that were not significantly different from those of HEK/opt and B3TMCT (Fig. 3A). B3TMCT replicated somewhat faster initially (i.e., days 1 and 2), and thus it reached its peak titer earlier (on day 3) than the other RSV-F-bearing constructs (after day 4). The kinetic difference between B3CT and B3TMCT was not observed when DS mutations were introduced: the titers of all three DS-based viruses were indistinguishable (Fig. 3B). This showed that efficient packaging of RSV F proteins bearing the BPIV3 F CT and TMCT domains had minimal or no discernible effect on vector replication compared to that of vectors expressing wt-like RSV F. Similar results were obtained with LLC-MK2 cells (not shown). The latter cells are competent to make type I interferons in response to viral infection, whereas Vero cells are not, and thus this factor did not substantially affect the in vitro replication of these vectors.

FIG 3.

Replication, intracellular protein expression, and syncytium formation in Vero cells by rB/HPIV3 vectors expressing various forms of RSV F. (A and B) Multicycle replication of rB/HPIV3 vectors in Vero cells at 32°C. Vero cells were infected with the following vectors at an MOI of 0.01 TCID₅₀ per cell, in triplicate: empty rB/HPIV3, HEK/opt, B3CT, and B3TMCT (A) or empty rB/HPIV3, DS, DS/B3CT, and DS/B3TMCT (B). A portion of the medium (0.5 ml out of 2 ml) was collected from the infected cell culture and replaced with fresh medium every 24 h during a period of 6 days. Viral titers in medium supernatants were determined by TCID₅₀ assays in LLC-MK2 cells at 32°C. (C) Western blot analysis of the expression of RSV F by vectors and wt RSV in Vero cells at 32°C. Vero cells were infected at an MOI of 10 TCID₅₀ per cell with the indicated vectors or 3 PFU per cell with wt RSV and then incubated at 32°C. Cell lysates were harvested at 48 h p.i., and expression of RSV F was analyzed by Western blotting. Analysis of BPIV3 N and GAPDH provided controls for vector expression and gel loading, respectively. Cleaved (F₁) and uncleaved (F₀) forms of RSV F are indicated. (D to L) Syncytium formation in Vero cell monolayers infected with empty rB/HPIV3 (D), HEK/opt (E), B3CT (F), B3TMCT (G), DS (H), DS/B3CT (I), or DS/B3TMCT (J), at an MOI of 10 TCID₅₀ per cell, or with wt RSV (K) at an MOI of 3 PFU per cell, or mock infected (L). Images were acquired at 48 h p.i.

Vero cells were infected with the rB/HPIV3-RSV-F vectors, lysed, and subjected to Western blot analysis to compare the efficiencies of expression of RSV F. This analysis showed that the total levels of expression of RSV F (F₀ + F₁) were comparable for the various vectors (Fig. 3C, lanes 2 to 7). The three DS forms appeared to have somewhat more uncleaved F₀ than the non-DS versions, suggesting that the DS mutations modestly reduced the efficiency of cleavage, but with an inconsequential difference.

Inspection of infected cell monolayers revealed minimal syncytium formation by the empty rB/HPIV3 vector (Fig. 3D), whereas monolayers infected with wt RSV exhibited large syncytia (Fig. 3K). Expression of wt RSV F by rB/HPIV3 also induced large syncytia (not shown), but this was greatly reduced when the HEK mutations were added to create an RSV F protein identical to that of early-passage RSV (HEK/opt) (Fig. 3E), as previously discussed (29). In contrast, monolayers infected with B3CT exhibited a hyperfusogenic phenotype (Fig. 3F), suggesting that this modification increased the sensitivity of RSV F to triggering. This hyperfusogenic phenotype was not observed with B3TMCT (Fig. 3G), indicating that inclusion of the TM domain counteracted the effect of B3CT alone. Versions of the RSV F protein bearing the DS mutations did not induce evident syncytia (Fig. 3H to J), as would be expected for mutations that stabilized the pre-F conformation.

Temperature sensitivity phenotype of rB/HPIV3 vectors.

The rB/HPIV3 vectors were evaluated for temperature sensitivity (Table 1). We previously showed that the empty rB/HPIV3 vector has a shutoff temperature (see Table 1, footnote c, for a definition) of 40°C (28), and this was also observed in the present study (Table 1). We also previously showed (28) that insertion of unmodified wt RSV F into the second gene position made the rB/HPIV3 vector 1°C more temperature sensitive, with a shutoff temperature of 39°C (28). In the present study, the shutoff temperature for the HEK/opt construct (which differs from wt RSV F by codon optimization and the HEK assignments) was reduced 1°C, to 38°C (Table 1), which might reflect an effect of increased F expression due to codon optimization and HEK assignments (29). The introduction of the DS mutations into this backbone had no effect. In comparison, the B3CT and B3TMCT constructs had a shutoff temperature of 37°C, which was returned to 38°C when the DS mutations were introduced. Thus, the insertion of an efficiently expressed F ORF, with or without the DS mutations, reduced the shutoff temperature 2°C (to 38°C), the introduction of the foreign TM or TMCT domains in the absence of the DS mutations reduced the shutoff temperature by an additional 1°C (to 37°C), and the latter decrease was reversed by the addition of the DS mutations.

TABLE 1.

Temperature sensitivity of rB/HPIV3 vectors on LLC-MK2 cell monolayers^a

Virus	Titer at permissive temp of 32°C (log10 PFU/ml)b	Titer (log10 PFU/ml) at the indicated temp (°C)c
		35	36	37	38	39	40
rB/HPIV3	6.3	6.4	6.3	6	5.7	5.4	2.2
HEK/opt	6.9	6.5	6	5.4	3.7	3.3	<1.7
B3CT	7.3	6.7	5.6	4.9	<1.7	<1.7	<1.7
B3TMCT	7.7	6.8	6.1	4.9	<1.7	<1.7	<1.7
DS	7.2	6.8	6.3	5.9	3.7	<1.7	<1.7
DS/B3CT	6.8	6.6	5.9	5.2	<1.7	<1.7	<1.7
DS/B3TMCT	6.5	5.9	5.3	5	3.6	<1.7	<1.7

^aPlaque titration assays were carried out on LLC-MK2 cell monolayers as described previously (42). Plaques were identified by using rabbit anti-HPIV3 hyperimmune serum.

^bVirus titers are means for two repeats in one experiment.

^cNumbers in bold indicate the values at the shutoff temperatures for plaque formation. The shutoff temperature was defined as the lowest temperature at which there was a ≥100-fold reduction in plaque number compared with the plaque number at 32°C. The empty rB/HPIV3 vector had a shutoff temperature of 40°C. This was reduced to 38°C with HEK/opt, indicating an effect of the F insert. The further introduction of the B3CT and B3TMCT domains resulted in a reduction to 37°C. The further insertion of the DS mutations into these packaged forms returned the shutoff temperature to 38°C. Thus, the addition of an RSV F insert and modifying RSV F to be packaged each resulted in a modest reduction in viral fitness, which was partly offset by pre-F stabilization.

Replication and genetic stability of rB/HPIV3 vectors in hamsters.

Hamsters were infected i.n. with the various vectors at a dose of 10⁵ TCID₅₀ or with wt RSV at 10⁶ PFU (Table 2). Viral titers in the nasal turbinates and lungs at days 3 and 5 (6 animals per virus per day) were determined by plaque assay. Compared with the empty vector, vectors bearing an RSV F insert were significantly more attenuated in the nasal turbinates and lungs, with the effect being substantially greater in the lungs. In general, the four vectors with efficient packaging of RSV F (i.e., B3CT, B3TMCT, DS/B3CT, and DS/B3TMCT) were more attenuated than vectors in which packaging of RSV F was minimal (HEK/opt and DS constructs), an effect that was most evident in the more permissive nasal turbinates. Thus, enhanced packaging of RSV F appeared to be attenuating in vivo. The DS mutations alone or combined with B3TMCT had little effect on replication, but addition of the DS mutations to B3CT resulted in a modest but significant increase in replication, likely due to countering the hyperfusogenic phenotype of B3CT.

TABLE 2.

Replication of rB/HPIV3 vectors and wt RSV in the upper and lower respiratory tracts of hamsters

Immunizing virusa	Titer (mean log10 PFU/g ± SEM)b,c
	Nasal turbinates		Lungs
	Day 3	Day 5	Day 3	Day 5
rB/HPIV3	5.4 ± 0.0 (A)	5.3 ± 0.1 (A)	4.9 ± 0.3 (A)	5.2 ± 0.1 (A)
HEK/opt	2.7 ± 0.2 (B)	5.1 ± 0.1 (A)	1.9 ± 0.1 (B)	2.4 ± 0.2 (BC)
B3CT	1.7 ± 0.1 (C)	2.2 ± 0.3 (B)	≤1.7 (B)	≤1.7 (B)
B3TMCT	1.8 ± 0.1 (CD)	2.4 ± 0.2 (BD)	1.7 ± 0.1 (B)	≤1.7 (B)
DS	2.3 ± 0.2 (BC)	3.9 ± 0.2 (CD)	≤1.7 (B)	2.6 ± 0.3 (CD)
DS/B3CT	2.5 ± 0.2 (BD)	4.1 ± 0.1 (C)	1.7 ± 0.1 (B)	2.7 ± 0.3 (C)
DS/B3TMCT	1.8 ± 0.1 (C)	3.2 ± 0.2 (D)	≤1.7 (B)	1.7 ± 0.1 (BD)
wt RSV	4.9 ± 0.2	NA	4.5 ± 0.4	NA

^aHamsters were inoculated intranasally with 10⁵ TCID₅₀ of rB/HPIV3 vectors and 10⁶ PFU of wt RSV (A2) in a 0.1-ml inoculum.

^bNasal turbinates and lungs were collected on day 3 and day 5 p.i. (6 animals per virus per day), and viral titers were determined by plaque assay on Vero cells at 32°C. The detection limit for the plaque assay was 1.7 log₁₀ PFU/g of tissue.

^cMean viral titers were assigned to different groups (A, B, C, and D) by the Tukey-Kramer test. Within each column, mean titers with different letters are statistically significantly different (P < 0.05). Titers with two letters are not significantly different from those with either letter. NA, the wt RSV titer was not determined on day 5.

The stability of RSV F expression by all vectors after in vivo replication was evaluated by a fluorescence double-staining plaque assay (Table 3). Among 91 specimens with detectable virus, 84 had plaques with 100% expression of the RSV F protein, and the remaining had ≥88% expression (Table 3). The presence of plaques negative for expression of RSV F was not greater on day 5 than on day 3, suggesting that there was not a progressive loss with time. Overall, these results showed that most of the constructs maintained stable expression of RSV F in vivo. Thus, somewhat surprisingly, efficient incorporation of RSV F into the vector particles did not appear to select for a virus in which its expression was silenced.

TABLE 3.

Stability of RSV F expression by rB/HPIV3 vectors after replication in the respiratory tract of hamsters^a

Virus	Day 3			Day 5
	Hamster no.b	% RSV F expressionc		Hamster no.b	% RSV F expressionc
		Nasal turbinates	Lungs		Nasal turbinates	Lungs
HEK/opt	481	100	NA	397	100	NA
	482	100 (n = 5)	NA	398	100	NA
	483	100	NA	399	100	100
	484	100	NA	400	100	100 (n = 4)
	485	100 (n = 2)	100 (n = 1)	401	100	100
	486	100 (n = 7)	100 (n = 3)	402	100	100
B3CT	499	100 (n = 1)	NA	415	100 (n = 5)	100 (n = 1)
	500	NA	NA	416	100 (n = 8)	NA
	501	NA	NA	417	NA	NA
	502	NA	NA	418	100 (n = 1)	100 (n = 1)
	503	NA	NA	419	100 (n = 3)	NA
	504	NA	NA	420	NA	100 (n = 1)
B3TMCT	505	100 (n = 1)	NA	421	100 (n = 9)	NA
	506	NA	NA	422	100 (n = 3)	NA
	507	100 (n = 4)	NA	423	100 (n = 1)	NA
	508	100 (n = 1)	NA	424	100 (n = 4)	NA
	509	100 (n = 1)	NA	425	100 (n = 8)	NA
	510	100 (n = 3)	NA	426	100	100 (n = 1)
DS	523	NA	100 (n = 1)	439	98	NA
	524	100	NA	440	100	100
	525	100 (n = 2)	100 (n = 1)	441	100	100
	526	100	NA	442	100	100 (n = 1)
	527	100	NA	443	100	100
	528	100 (n = 1)	100 (n = 1)	444	95	100 (n = 3)
DS/B3CT	529	100	100 (n = 2)	445	100	100
	530	100 (n = 5)	NA	446	99	100 (n = 1)
	531	100 (n = 7)	100 (n = 1)	447	100	88
	532	NA	NA	448	100	92
	533	100	100 (n = 1)	449	89	100 (n = 2)
	534	100 (n = 3)	NA	450	100	100
DS/B3TMCT	535	100 (n = 2)	NA	451	100 (n = 5)	100 (n = 1)
	536	NA	NA	452	100	NA
	537	100 (n = 3)	NA	453	100	100 (n = 1)
	538	NA	NA	454	95	NA
	539	100 (n = 3)	NA	455	100 (n = 6)	100 (n = 1)
	540	100 (n = 2)	NA	456	100	100 (n = 2)

^aThe percentage of PFU in the inoculum that expressed RSV F was 100% for all virus constructs, as determined by a double-staining plaque assay on Vero cells as described in Materials and Methods.

^bSpecimens were from the experiment performed for Table 2.

^cThe percentages of PFU isolated from nasal turbinate and lung specimens that expressed RSV F were determined by a double-staining plaque assay. The data for HEK/opt and DS were published previously (29) and are included here for comparison. NA, no plaques were obtained from the sample due to a low titer. For any sample with fewer than 10 plaques, the total number of plaques (n = x) is given.

Neutralizing serum antibody responses and protective efficacy of rB/HPIV3 vectors in hamsters.

A second set of hamsters was immunized as described above, serum samples were collected at day 28, and i.n. challenge with wt RSV (10⁶ PFU) was performed on day 30. RSV- and HPIV3-neutralizing serum antibody titers were quantified by virus neutralization assays with (RSV and HPIV3) or without (RSV) added guinea pig complement (Fig. 4). The assay performed with added complement is commonly used for RSV and HPIV3 neutralization assays because it allows sensitive detection of virus-specific antibodies, since complement can confer viral lysis and steric hindrance capabilities to antibodies that otherwise might not be neutralizing in vitro (38). In contrast, the assay performed without complement is a more stringent test for efficiently neutralizing (“high quality”) antibodies (29, 38).

FIG 4.

Levels of serum neutralizing antibodies induced by rB/HPIV3 vectors and wt RSV in hamsters. Hamsters (n = 6) were inoculated intranasally with 10⁵ TCID₅₀ of rB/HPIV3 vectors and 10⁶ PFU of wt RSV (A2) in a 0.1-ml inoculum. Serum samples were collected at day 28 postinoculation. RSV-neutralizing antibody titers were determined by a 60% plaque reduction neutralization test (PRNT₆₀) with (A) or without (B) guinea pig complement. (C) HPIV3-neutralizing antibody titers were determined by PRNT₆₀ with complement. The detection limit of each assay is indicated by a dotted line. The colored symbols represent individual animal titers. The mean titer for each group is shown as a vertical bar, with error bars representing standard errors of the means (SEMs), and the value of the mean titer is shown above the bar. Mean titers were assigned to different groups (A, B, C, and D) by the Tukey-Kramer test. Mean titers with different letters are statistically different (P < 0.05). Titers with two letters are not significantly different from those with either letter.

In the RSV neutralization assay performed with complement, each of the rB/HPIV3 vectors expressing RSV F induced a substantial titer of RSV-neutralizing serum antibodies (Fig. 4A, lanes 2 to 7). The mean antibody titers of the DS and DS/B3TMCT constructs were the highest among the vectors (Fig. 4A, lanes 5 and 7), while that of B3CT was the lowest (lane 3). Wild-type RSV (Fig. 4A, lane 8) induced a significantly higher mean titer of RSV-neutralizing antibodies than those of any of the tested vectors. This was not surprising given that wt RSV replicated to a higher titer than those of the vectors (Table 2) and also expresses G protein, the second RSV neutralization antigen.

In the RSV neutralization assay performed in the absence of complement, RSV-neutralizing serum antibodies were not detected for HEK/opt and B3CT (Fig. 4B, lanes 2 and 3). In contrast, B3TMCT, DS, DS/B3CT, and DS/B3TMCT (Fig. 4B, lanes 4 to 7) induced mean titers of high-quality RSV-neutralizing antibodies that exceeded those induced by wt RSV, although the differences were not significant (Fig. 4B, lane 8). DS/B3CT had improved immunogenicity compared to that of B3CT, suggesting an effect of pre-F stabilization mitigating the hyperfusogenic phenotype of B3CT, but it was still somewhat lower than that of DS, B3TMCT, and DS/B3TMCT. Overall, in both assays, DS/B3TMCT was the most immunogenic vector (Fig. 4A and B, lanes 7), suggesting a partially additive effect of B3TMCT and the DS mutations in enhancing immunogenicity. All of the rB/HPIV3-RSV-F vectors induced similar titers of HPIV3-neutralizing (i.e., vector-specific) serum antibodies (Fig. 4C, lanes 2 to 7), which were all significantly lower than that with the empty vector (Fig. 4C, lane 1), reflecting their greater in vivo attenuation.

The immunized hamsters were challenged with wt RSV i.n. to evaluate the protective efficacy of the rB/HPIV3 vectors (Fig. 5). Compared with the nonprotective empty rB/HPIV3 vector, each of the vectors expressing RSV F was significantly protective in the nasal turbinates, except for B3CT. Protection by the various vectors was somewhat greater in the lungs, although B3CT was less protective than the others. DS/B3TMCT was the only vector that, like wt RSV, conferred complete protection in the lungs of all six hamsters and complete protection in the nasal turbinates of five of six hamsters. This equivalence to wt RSV in terms of protection is noteworthy because wt RSV replicated to 50- to 630-fold higher titers than those of DS/B3TMCT (Table 2), contains both neutralization antigens F and G, and contains all of the other RSV proteins as potential antigens for cellular immunity, whereas the vector expresses only the RSV F protein. The high level of efficacy of DS/B3TMCT suggested that the effects of the TMCT and DS modifications were additive to some extent.

Evaluation of selected rB/HPIV3 vectors in rhesus monkeys.

The following three rB/HPIV3-RSV-F vectors were chosen for further evaluation in rhesus monkeys: (i) DS/B3TMCT, which was the most immunogenic and protective vector in the hamster study described above; (ii) DS, which was the most immunogenic and protective vector in hamsters in a previous study (29); and (iii) non-HEK/non-opt (not shown in Fig. 1A or discussed above), which expresses wt RSV F that has not been codon optimized, does not contain the HEK changes, and has not otherwise been modified (29) and thus is very similar to the MEDI-534 vector that was previously evaluated in seronegative infants and young children (7).

Rhesus monkeys were immunized by the combined i.n. and i.t. routes, with 10⁶ TCID₅₀ of vector per route. Replication in the upper respiratory tract (URT) and the lower respiratory tract (LRT) was assessed by titration of viruses in nasopharyngeal swabs and tracheal lavage fluids, respectively (Fig. 6). On day 28, all rhesus monkeys were challenged by the combined i.n. and i.t. routes with 10⁶ PFU of wt RSV per site to evaluate protective efficacy. Sera were collected at various time points after immunization and challenge to determine the titers of RSV-neutralizing antibodies (Fig. 7).

FIG 6.

Replication of rB/HPIV3 vectors in the respiratory tract of rhesus macaques. Rhesus monkeys were infected by the combined i.n. and i.t. routes with the indicated rB/HPIV3 vectors at a dose of 10⁶ TCID₅₀ in a 1-ml inoculum per route. Groups of five, five, and four monkeys per group were infected with non-HEK/non-opt, DS, and DS/B3TMCT, respectively. Nasopharyngeal swabs and tracheal lavage fluids were collected on the indicated days. Viral titers in collected samples were determined by TCID₅₀ assay on MK2 cells at 32°C. The mean titers for the groups at each time point are plotted, and error bars represent SEMs.

Following inoculation with the rB/HPIV3 vectors, shedding was detected in the respiratory secretions for up to 10 to 12 days (Fig. 6). DS and non-HEK/non-opt replicated with kinetics and titers similar to those of each other in the URT and LRT, while DS/B3TMCT was 100 to 1,000 times more attenuated in the URT and was 10 to 100 times more attenuated in the LRT, where shedding was barely detected (Fig. 6A and B). The reduced replication of DS/B3TMCT was not due to increased temperature sensitivity, since its in vitro shutoff temperature was the same as that for the DS construct (Table 1). Thus, the B3TMCT modification was highly attenuating in this nonhuman primate, presumably due to incorporation of the foreign RSV F protein into the rB/HPIV3 particles and the resulting reduced incorporation of the vector F protein (Fig. 2A to C).

Vectors were evaluated for the stability of RSV F expression after replication in the rhesus respiratory tract (from the experiment shown in Fig. 6) by use of a double-staining plaque assay (Table 4). Nasopharyngeal swab samples collected at days 4, 5, and 6 (the days of peak replication) (Fig. 6A) were tested, but no tracheal lavage fluid samples were tested because of the highly restricted replication (Fig. 6B). Viral plaques were available for 39 of the 42 relevant samples (Table 4). In 28 samples, all of the viral plaques expressed the RSV F protein. In the remaining 11 samples, ≥91% of the plaques were positive for the expression of RSV F (Table 4). The DS/B3TMCT vector did not exhibit increased instability of RSV F expression compared to the other vectors, indicating that efficient packaging of RSV F did not seem to select for a virus in which its expression was silenced. In addition, there was no evident progressive increase in the number of RSV-negative plaques between days 4, 5, and 6. Rather, RSV-negative plaques appeared to be sporadic, such that they might be detected in a given animal on one day and not be detected in the same animal on a subsequent day.

TABLE 4.

Stability of RSV F expression by rB/HPIV3 vectors after replication in the respiratory tract of rhesus macaques^a

Virus	Monkey ID	% RSV F expression in nasopharyngeal swabsb
		Day 4	Day 5	Day 6
Non-HEK/non-opt	RH37033	100	100	100
	RH37051	91	93	93
	RH37057	97	100	100
	RH37063	100	92	100
	RH37380	100	100	100
DS	RH37034	100	94	100
	RH37073	100	100	100
	RH37078	93	100	96
	RH37103	100	100	100
	RH37760	100	95	96
DS/B3TMCT	RH38487	NA	NA	NA
	RH37324	100	100	100
	RH37140	100	93	100
	RH37360	100	100	100

^aThe percentage of PFU in the inoculum that expressed RSV F was 100% for all virus constructs, as determined by a double-staining plaque assay on Vero cells as described in Materials and Methods.

^bThe percentages of PFU isolated from nasopharyngeal swabs that expressed RSV F were determined by a double-staining plaque assay. NA, plaques were not obtained from the sample due to a low titer.

The HPIV3-neutralizing (i.e., vector-specific) serum antibody response was evaluated (Fig. 7A). The response to DS/B3TMCT was delayed compared to the responses to the other two constructs, consistent with its increased attenuation, although the titers reached similar levels by days 35 and 56.

Evaluation of the RSV-neutralizing serum antibody response on days 14, 21, and 28 (i.e., prechallenge) showed that each of the three rB/HPIV3-RSV-F vectors induced substantial titers in vitro in the presence of added complement (Fig. 7B). Surprisingly, despite being highly attenuated, DS/B3TMCT induced significantly higher titers of RSV-neutralizing serum antibodies than DS (8-fold on day 28) and non-HEK/non-opt (3.3-fold on day 28), which induced responses similar to each other (Fig. 7B). Also, the response to DS/B3TMCT was detectable earlier than those for the other two constructs.

In addition, RSV-neutralizing antibodies in sera collected at day 28 were assayed in the absence of complement (Fig. 7C). This showed that the complement-independent neutralizing antibody titer induced by DS/B3TMCT (∼1:290 on day 28) was 100- and 30-fold higher than those induced by non-HEK/non-opt and DS, respectively. The very high serum antibody response to DS/B3TMCT compared to that to DS in rhesus monkeys suggests that either the TMCT mutation has a significantly stronger effect than the DS mutation in this primate species or the combination of TMCT and DS mutations was synergistic in enhancing immunogenicity.

Protective efficacy was evaluated by i.n. challenge with wt RSV at day 28. However, none of the animals shed challenge RSV at a level detectable by plaque assay (not shown). Thus, each of the vectors induced sufficient immunity to completely block shedding of infectious virus in this animal model, reflecting the semipermissive nature of RSV replication in nonhuman primates. Following the challenge, the titers of RSV-neutralizing serum antibodies increased (Fig. 7B). However, it was not clear whether this reflected boosting due to the RSV challenge or was due to the original immunization.

Quantification of pre-F-specific serum antibodies and correlation with complement-independent RSV-neutralizing activity.

We assessed the relative proportions of pre-F-specific antibodies in sera of hamsters and rhesus monkeys immunized with the various constructs (from the experiments for Fig. 4 and 6). As shown in Fig. 8, hamster (A and B) and rhesus (C and D) sera were assayed for binding to the pre-F (A and C) or post-F (B and D) protein in the presence of an excess of competing soluble post-F protein. The purpose of including competing post-F was to deplete antibodies specific to post-F and to shared epitopes on pre- and post-F, thus facilitating detection of antibodies to epitopes unique to pre-F.

FIG 8.

Pre-F-specific serum antibodies induced in hamsters and rhesus monkeys by rB/HPIV3 vectors and their correlation with titers of high-quality RSV-neutralizing antibodies. Hamster and rhesus monkey sera from the experiments performed for Fig. 4 and 7 were assayed by biolayer interferometry (BLI) for the ability to bind to the biosensor-tagged pre-F or post-F protein in the presence or absence of competing soluble post-F protein. Binding activity was calculated as follows: (binding in the presence of competing post-F)/(binding in the absence of competing post-F) × 100. (A and C) Percentages of pre-F-specific antibody binding retained in the presence of competing post-F in hamster sera (A) and rhesus monkey sera (C). (B and D) Percentages of post-F-binding antibodies retained in the presence of competing post-F in hamster sera (B) and rhesus monkey sera (D). Note that the “postfusion” construct included in panels A and B is an rB/HPIV3 vector expressing post-F that was not shown in Fig. 4 but was part of the same experiment and was described previously (29). (E and F) Correlations of the percentages of pre-F-binding antibodies from panels A and C with RSV-neutralizing antibody titers, assayed in the absence of complement, in hamster sera (E) and rhesus monkey sera (F) from the experiments performed for Fig. 4 and 7. The fitting line was generated with a linear regression model. Pearson's correlation coefficient (Pearson R), the sample size (N), and the P value are indicated on the graph. The sera used for panels A, B, and E comprised 45 of the 48 specimens representing the vectors in Fig. 4A and B; three sera (two for B3CT and one for the postfusion construct) were omitted due to low pre-F-binding titers.

As shown in Fig. 8A, sera of hamsters immunized with vector constructs containing the B3TMCT and/or DS mutations (i.e., B3TMCT, DS, DS/B3CT, and DS/B3TMCT) had significantly larger proportions of pre-F-specific antibodies than sera of hamsters immunized with wt RSV and vector constructs without the B3TMCT or DS mutations (i.e., HEK/opt, B3CT, and the postfusion construct). (The postfusion construct expresses a postfusion form of the RSV F protein that was described previously [29], and these sera were from animals that were immunized in parallel with the other constructs used for Fig. 4.) In rhesus monkeys, DS induced a larger proportion of pre-F-specific antibodies than non-HEK/non-opt, and DS/B3TMCT induced a substantially larger proportion of pre-F-specific antibodies than DS (Fig. 8C). The difference between the last two constructs was substantially greater than that observed for hamsters (Fig. 8A), suggesting that the effect of the TMCT modification was substantially greater in the nonhuman primates (Fig. 8C). For all serum samples, low levels of post-F binding were retained, indicating that post-F competition was effective (Fig. 8B and D). These data suggested that DS and B3TMCT independently and additively stabilized RSV F in the pre-F conformation, leading to the induction of more pre-F-specific antibodies.

The proportion of pre-F-specific antibodies in serum was found to correlate directly with the reciprocal RSV log₂ PRNT₆₀ titers measured in the absence of complement for hamsters (r = 0.61; P < 0.0001; n = 45; Pearson's correlation) and rhesus monkeys (r = 0.92; P < 0.0001; n = 13; Pearson's correlation) (Fig. 8E and F). This indicated that an important component of serum capable of neutralizing RSV in vitro was the presence of a large proportion of pre-F-specific antibodies.

DISCUSSION

An rB/HPIV3 vector expressing an unmodified version of the RSV F ORF (MEDI-534) was previously shown to be well tolerated in seronegative infants, but the immunogenicity and genetic stability of the RSV F insert needed improvement (7). In the present study, we investigated (i) whether we could engineer the RSV F protein to be packaged efficiently into the rB/HPIV3 particles, (ii) whether this would improve the immunogenicity of the vector-expressed RSV F protein, (iii) how this compared with stabilization of RSV F in the DS pre-F conformation, and (iv) whether the combination could provide even further improvement.

Expression of a heterologous viral glycoprotein by a viral vector, such as a negative-strand virus, may result in its packaging into the vector particle, but that is unpredictable. In some cases, the heterologous glycoprotein might be packaged without need for modification, although the efficiency may be reduced (37, 39–41). In other cases, the heterologous glycoprotein might need to be modified in order to be packaged, such as by incorporating a vector-compatible CT (22, 42, 43). Also, packaging of a heterologous glycoprotein may result in attenuation of the vector. Additional effects may occur: for example, evidence from several paramyxoviruses implicated the TM, CT, and other domains in modulating fusogenic activity, stability, and protein folding (17, 19).

Replacing the native CT or TMCT domains of the RSV F protein with their BPIV3 F counterparts (B3CT and B3TMCT constructs) resulted in the RSV F protein being packaged into the rB/HPIV3 particles with an efficiency equivalent, per microgram of virion protein, to that of RSV virions. This resulted in reduced packaging of the vector F protein, presumably due to competition reflecting the shared CT or TMCT domains, while packaging of the vector HN protein was unaffected. Surprisingly, this did not significantly impair vector growth in cell culture compared to that of constructs expressing forms of RSV F that were not efficiently packaged. However, packaging of RSV F did confer increased restriction of replication in vivo. This was particularly evident in rhesus monkeys, where inclusion of the TMCT domains resulted in reductions in vector replication of 100- to 1,000-fold in the URT and 10- to 100-fold in the LRT. Conversely, there was no indication that packaging of the RSV F protein into the vector particles enhanced their infectivity and replication. Replacement of the CT domain alone resulted in a hyperfusogenic phenotype that conferred a modest amount of restriction in vitro and in vivo. This was not observed with B3TMCT.

In hamsters, the B3TMCT construct induced high titers of antibodies that efficiently neutralized RSV in vitro in the presence or absence of complement. Antibodies that neutralize RSV in vitro in the absence of complement, which we termed “high-quality antibodies,” have been suggested to be the most relevant for protection (38), and in the present study they were associated with increased protection against RSV challenge in vivo. Furthermore, these high-quality antibodies correlated with specificity for the pre-F protein. The B3TMCT construct was very similar to the DS construct in the ability to induce complement-independent antibodies with a large proportion of binding to the pre-F protein, suggesting that the enhanced immunogenicity conferred by B3TMCT can likely be attributed to its potential pre-F-stabilizing activity.

The induction of high-quality antibodies in hamsters with constructs bearing B3TMCT or DS mutations alone or in combination (B3TMCT, DS, and DS/B3TMCT) was striking compared to that with the HEK/opt construct (expressing an RSV F protein that was unmodified except for the HEK assignments), which replicated to a significantly higher titer yet did not induce detectable high-quality antibodies. In addition, the titers of high-quality antibodies induced in hamsters by the B3TMCT, DS, and DS/B3TMCT constructs equaled or exceeded that induced by wt RSV, even though wt RSV replicated to titers significantly higher than those of these vectors and had the added contribution of neutralizing antibodies induced by the G glycoprotein (the other major RSV neutralization antigen). The B3TMCT, DS, and DS/B3TMCT constructs also induced much higher levels of pre-F-specific antibodies in hamsters than those induced by wt RSV. These observations illustrate the dramatically improved immunogenicity of the engineered RSV F proteins.

Although the B3CT and B3TMCT versions of RSV F were packaged into rB/HPIV3 virions with similar efficiencies, the B3CT construct induced serum antibodies that neutralized RSV in vitro in the presence of complement, but not in its absence, and that did not efficiently bind to the pre-F protein. Adding the DS mutations to B3CT (DS/B3CT) conferred the ability to induce high-quality antibodies, although the titer remained significantly lower than those with B3TMCT and DS/B3TMCT. These differences could not be attributed to differences in vector replication or levels of RSV F expression. Rather, the inability of B3CT to induce high-quality RSV-neutralizing antibodies likely reflects its hyperfusogenic phenotype, resulting in a reduction in the amount of F protein present in the pre-F conformation, as opposed to the increased pre-F obtained with B3TMCT and DS stabilization.

In rhesus monkeys, the DS construct replicated similarly to non-HEK/non-opt (expressing an unmodified wt F protein) and induced a similar titer of antibodies that neutralized RSV in the presence of complement. However, the DS construct also induced high-quality neutralizing antibodies, whereas non-HEK/non-opt did not, as well as a substantial increase in the proportion of antibodies binding to the pre-F protein. Despite the DS/B3TMCT construct being 100- to 1,000-fold more attenuated than the other two constructs in rhesus monkeys, it induced substantially higher titers of RSV-neutralizing antibodies detected in the presence or absence of complement and a substantially larger proportion of antibodies specific to the pre-F protein. In addition, the neutralizing serum antibody response in rhesus monkeys was induced more rapidly by DS/B3TMCT than by the other two candidates, a favorable characteristic for a pediatric vaccine, which ideally should induce protection as early in life as possible. Thus, the effect of the DS and TMCT mutations together was much greater in rhesus monkeys than the effect of the DS mutations alone. This might mean that there was a synergistic effect of the combination or that the effect of the TMCT domains was disproportionately greater than that of the DS mutations in rhesus monkeys. Further studies will be necessary to distinguish between these possibilities. The effect of combining the DS and B3TMCT modifications compared to the DS mutations alone was much greater in rhesus monkeys than in hamsters, and presumably is more predictive of results in humans.

The mechanism for the effect of the DS mutations seems clear, namely, stabilization (via an introduced disulfide bond) of the F protein in the pre-F conformation that is highly immunogenic for RSV-neutralizing antibodies. Although the B3TMCT construct was not specifically engineered for increased stability of the pre-F protein, it did induce high titers of pre-F-specific antibodies. Therefore, substantial stabilization of pre-F appeared to occur with B3TMCT. It may be that the inclusion of the B3TMCT domains or packaging of the F protein into vector particles increased its stability in the pre-F conformation. Also, an RSV F protein that is packaged and released in vector particles likely would be taken up earlier and more efficiently by antigen-presenting cells than unpackaged F protein released by cell destruction, and therefore would be more likely to avoid triggering and loss of the pre-F conformation. Another mechanism might also be involved. Specifically, the high efficiency of uptake of B3TMCT-containing RSV F into rB/HPIV3 particles suggests that it might be incorporated in native high-density, repetitive, multimeric arrays that enhance immunogenicity (44–46). Thus, the improvements in immunogenicity of B3TMCT may have involved both stabilization of the pre-F conformation and changes in antigen delivery and presentation.

A wide array of RSV vaccine candidates are presently in various stages of preclinical and clinical development (http://sites.path.org/vaccinedevelopment/files/2016/07/RSV-snapshot-July_13_2016.pdf). As noted in the introduction, RSV vaccines based on PIV vectors or attenuated RSV strains are the only ones known to be safe in RSV-naive recipients (6, 7). Another PIV-vectored RSV vaccine presently under development is based on Sendai virus (murine PIV1) expressing the unmodified RSV F protein; like rB/HPIV3, Sendai virus is attenuated by host range differences (47). With regard to live-attenuated RSV strains, a number of candidates are presently in clinical trials and represent several different mechanisms of attenuation, specifically, temperature-sensitive mutations, which in some cases have been stabilized against deattenuation, combined with deletion of the SH surface protein gene (8, 36); deletion of the transcription/replication regulatory protein M2-2, which appears to be very promising (10); and deletion of the interferon antagonist NS2 gene in combination with deletion of a codon in the L gene (48). Other attenuated RSV strains in preclinical development represent strategies including codon deoptimization of the NS1 and NS2 ORFs (49, 50), codon-pair deoptimization of various RSV ORFs (51), and the use of less fusogenic F protein (49). An advantage of an attenuated RSV strain is that it provides all or nearly all of the RSV antigens and thus should induce a more complete immune response than a vector expressing one or several RSV ORFs. On the other hand, a vaccine based on infectious RSV faces obstacles inherent to RSV, including nonrobust replication in vitro, large filamentous particles, and physical instability, which complicate vaccine manufacture and supply and might limit vaccine availability. In contrast, a PIV-vectored RSV vaccine grows more efficiently in cell culture and forms particles that are more discrete and do not share the instability of RSV, and thus would be much easier to manufacture and deliver, particularly to resource-limited countries. A PIV-vectored RSV vaccine also provides bivalent protection against the PIV vector and RSV. Also, a PIV-vectored RSV vaccine can express the RSV F protein stabilized in the pre-F conformation, which provides the possibility of antibody responses that are strongly skewed toward highly neutralizing antibodies and may prime recipients for improved responses throughout life. In addition, a PIV-vectored RSV vaccine should be insensitive to restriction by RSV-specific antibodies from proposed passive RSV immunoprophylaxis strategies, such as maternal immunization or postpartum antibody administration.

As noted, the rB/HPIV3-RSV-F construct (MEDI-534) that was previously evaluated in infants and young children did not efficiently induce high-quality RSV-neutralizing serum antibodies. In the present study, we confirmed that a construct that is very similar to MEDI-534 (the non-HEK/non-opt construct) similarly was very poorly immunogenic for high-quality RSV-neutralizing serum antibodies. In contrast, the DS/B3TMCT construct, in particular, induced high titers of RSV-neutralizing serum antibodies in rhesus monkeys with or without complement, even though it was disproportionately highly restricted in replication. Titers of high-quality RSV neutralizing antibodies in hamsters induced by DS/B3TMCT exceeded those induced by wt RSV, and presumably also by live-attenuated RSV vaccine candidates. DS/B3TMCT and the other constructs also appeared to have substantial stability of RSV F expression in vitro, in hamsters, and in rhesus monkeys. Therefore, the DS/B3TMCT construct is an excellent candidate to be returned to the clinic, and we are presently preparing clinical trial material for evaluation in infants and young children. Also, since the addition of the B3TMCT domain proved to be highly attenuating, we are developing new constructs that express the TMCT RSV F insert from other PIV backbones that are less attenuated so that the final constructs will replicate ∼100- to 1,000-fold more efficiently than the B3TMCT construct from the present study and should have substantially increased immunogenicity. Also, additional mutations that provide increased stabilization of the pre-F conformation, such as the use of cavity-filling mutations (Cav1) S190F and V207L (20–22), as well as other mutations (52), have been reported and continue to be developed. Therefore, these newer stabilized versions may make chimeric PIV vector vaccines even more immunogenic, and this is presently being investigated.