The plasmids encoding rRSV-Long/A2cp, rRSV-Long/A2cpts, and rRSV-Long/A2cptsΔSH were obtained by a stepwise assembly of the synthesized cDNA segments. The locations of cp and ts mutations are shown in Fig. 1A. The full-length antigenomic cDNAs of pBRB-RSV-rLong/A2cp and pBRB-RSV-rLong/A2cpts were both expected to be 18, 815 bp in size, while that of pBRB-RSV-rLong/A2cptsΔSH was expected to be 18, 485 bp in size. The corresponding lengths were confirmed by DNA sequencing (data not shown).
For the recovery of rRSVs, the plasmid containing RSV antigenomic cDNA was co-transfected together with four plasmids encoding helper proteins to BHK/T7-9 cells, and the recovered rRSVs were subsequently blindly passaged in Vero cells. The rescued rRSVs were identified by immunoplaque assay as shown in Fig. 1B and by RT-PCR (data not shown). These results demonstrated that we successfully rescued the rRSVs bearing the anticipated cp, cpts, and cptsΔSH mutations.
The titers of rRSV-Long/A2cp increased rapidly during the initial passages from passage 1 (P1) to P3 (P < 0.01) and remained constant after P3 as shown by immunoplaque assay and RT-qPCR. In contrast, the titers of rRSV-Long/A2cpts and rRSV-Long/A2cptsΔSH increased steadily until P4 (P < 0.01), and fluctuated marginally after P4 (Fig. 2A, 2B). The sequencing of the DNA fragments enclosing the individual rRSV point mutations determined the genetic stability of the rescued viruses. All the introduced att mutations were stable, and no reversion was detected. The sequencing results for passages from P1 to P9 are shown in Supplementary Table S2. To further characterize rRSVs, the growth kinetics of rRSVs and wtRSV in HEp-2 cells at 33 ℃ were assessed and compared between the two groups. The viral titers began to increase from 24 h post-inoculation and ultimately plateaued at 72 h post-inoculation for all the viruses. wtRSV achieved the titer of 4.5 × 107 pfu/mL, rRSV-Long/A2cp 2.8 × 106 pfu/mL, rRSV-Long/A2cpts 2.2 × 105 pfu/mL, and rRSV-Long/A2cptsΔSH 1.1 × 105 pfu/mL, as shown in Fig. 2C. All the three rRSVs exhibited approximately 10- or 100-fold reduced growth kinetics compared with that of the parent strain. Altogether, all rRSVs bearing the att mutations were constructed successfully, showed a markedly attenuated phenotype in vitro, and their proliferation rates were significantly lower when compared with that of the wtRSV parent.
Figure 2. In vitro characterization of recombinant human respiratory syncytial viruses (rRSVs). The replication titers during serial passages of rRSVs were monitored by immunoplaque assay (A) and by RT-qPCR (B) since passage 1 (P1). The growth curves for rRSVs and wild-type RSV (wtRSV) were established and compared (C). Each virus was harvested every other 24 h post-infection and titers were assayed by immunoplaque assay. All results are representative of three independent experiments. Data are shown as mean ± SD. *P < 0.05, **P < 0.01.
The ts phenotype of the rRSVs was examined by determining the efficiency of plaque formation by inoculation of tenfold viral dilutions at various temperatures in HEp-2 cells placed in TC24-well plates and comparing the results for wtRSV and rRSVs. The plates were incubated for 5 days at the specified temperatures in CO2 incubators calibrated to ± 1 ℃ and the average viral titers were measured by immunoplaque assay in HEp-2 cells at the corresponding temperatures (Table 1). Both rRSV-Long/A2cpts and rRSV-Long/A2cptsΔSH exhibited reduced viral titers by more than 2log10 at 37 ℃ compared to the permissive temperature of 32 ℃ and were therefore considered ts at ≥ 37 ℃. As expected, the wtRSV as well as the precursor virus rRSV-Long/A2cp were not ts and showed no statistical titer reduction from 32 to 40 ℃.
Virus Mean virus titera (log10 pfu/ml ± SD) at the indicated temperature (℃) Mean titer in micec 32 35 36 37 38 39 40 Nasal wash (log10 total pfu ± SD) Lung (log10 pfu/g tissue ± SD) wtRSV 6.9 ± 0.2 6.9 ± 0.1 6.8 ± 0.1 7.1 ± 0.1 7.1 ± 0.1 6.8 ± 0.1 6.8 ± 0.1 3.1 ± 0.1 4.4 ± 0.2 rRSV-Long/A2cp 6.1 ± 0.1 6.1 ± 0.1 6.0 ± 0.1 6.0 ± 0.1 5.9 ± 0.1 5.8 ± 0.1 5.7 ± 0.1 2.6 ± 0.1 3.1 ± 0.1 rRSV-Long/A2cpts 6.0 ± 0.1 5.2 ± 0.1 4.7 ± 0.1 < 1.0b < 1.0 < 1.0 < 1.0 2.3 ± 0.1 2.0 ± 0.1** rRSV-Long/A2cptsΔSH 5.4 ± 0.1 4.3 ± 0.2 4.2 ± 0.1 < 1.0b < 1.0 < 1.0 < 1.0 2.1 ± 0.1 1.9 ± 0.1* an = 3 replicates (at each temperature).
bShut-off temperature is defined as the restrictive temperature at which a 100-fold or greater reduction compared to the titer observed at the permissive temperature of 32 ℃ and the lowest shut-off temperatures for each virus are italic.
cGroups of five mice were administered 1 × 106 pfu of the indicated virus intranasally under light anesthesia on day 0 and sacrificed on day 4. Virus titer was determined in the nasal washes (log10 total pfu) and lung tissues (log10 pfu/g tissue).
*The significant difference of RSV replication titers between nasal wash and lung tissue. All the results were shown as the representive of three independent experiments. Data were shown as mean ± SD. *P < 0.05, **P < 0.01.
Table 1. Characterization of the temperature sensitivity and attenuation (att) phenotypes of recombinant RSVs (rRSVs) in vitro and in vivo.
We also intranasally infected mice with 1 × 106 pfu of rRSVs and measured viral titers in samples obtained from nasal washes and lung tissue homogenates on days 2, 4, 6, and 8 post-infection by using immunoplaque assay and RT-qPCR (Fig. 3A-3D). The results showed that the titers of all three rRSVs in the nasal wash specimens at day 4 were 0.52-1.06 log10 total pfu, lower than that of wtRSV (P < 0.01-P < 0.001), and the titers in the lung tissues at day 4 and day 6 were 0.51-2.51 and 1.55-3.99 log10 pfu/g tissue, lower than the respective ones of wtRSV (P < 0.001). The number of RNA copies of the three rRSVs in the nasal wash specimens at day 4 was 0.32-0.88 log10 RSV copies/μg RNA, lower than that of wtRSV (P < 0.01-P < 0.001); in addition, in the lung tissues at days 4 and 6 the three rRSVs exhibited 1.00-2.55 and 1.20-3.74 log10 RSV copies/μg RNA, respectively, lower than the corresponding number of RNA copies of wtRSV (P < 0.001). These results are consistent with the corresponding immunoplaque test results. Among the three rRSVs, the two rRSVs possessing ts mutations displayed a higher level of attenuation than rRSV-Long/A2cp in the lung tissues (P < 0.001).
Figure 3. The attenuation (att) phenotype of recombinant human respiratory syncytial viruses (rRSVs) assayed in intranasally infected BALB/c mice. Viral titers in samples obtained from nasal washes were analyzed by immunoplaque test (A) and RT-qPCR (B); viral titers in lung tissues were analyzed by immunoplaque test (C) and RT-qPCR (D). Data are shown as mean ± SD. **P < 0.01, ***P < 0.001. LOD: limit of detection.
We also found that the replication titers of each rRSV bearing ts mutations were significantly higher in the samples obtained from nasal washes than in those obtained from lung tissue homogenates (P < 0.05 or P < 0.01), as shown in Table 1. Consistent with their in vitro ts phenotypes, the two rRSVs harboring ts mutations displayed shut-off temperatures equal to 37 ℃ and are therefore more competent to multiply in the nasal cavity environment, characterized by lower temperatures (Table 1).
To determine both the genetic and phenotypic stability of rRSVs, rRSV-Long/A2cpts was passaged in vitro at non-permissive temperatures to induce mutation, in line with the classical theory of the survival of the fittest. Briefly, rRSV-Long/A2cpts was serially passaged twice at 37 ℃, and then subjected to two further passages at 39 ℃ and one passage at 40 ℃ (Fig. 4). After each passage, one of the duplicate plates was immunostained with anti-RSV antibodies. Initially, after the expansion at 37 ℃, all the wells exhibited positive RSV immunostaining. After P4 at 39 ℃, more than 80% of the wells had positive RSV immunostaining, suggesting the presence of the temperature-sensitive intermediate (tsi) viruses at this temperature. At 40 ℃, the control wells containing wtRSV all had positive immunostaining. In sharp contrast, only 20%-30% of the wells containing rRSV-Long/A2cpts had positive RSV immunostaining. To investigate the nt changes at the ts markers, we analyzed the sequence of 1-2 kb cDNA fragments spanning from the start of the M2 gene through the L gene. RT-PCR was performed on five randomly chosen potential tsi revertants from RSV-positive wells to detect nt changes at the 248 and 404 ts sites of the L and M2 genes by sequence analysis. The characteristics of the biologically derived tsi viruses from the passaging of rRSV-Long/A2cpts at 39 ℃ are listed in Supplementary Table S3. The tsi strain were characterized by nt changes causing the reversion to the wt nt or amino acid at the 248 or the 404 ts markers and the partial loss of temperature sensitivity.
Figure 4. Passaging of rRSV-Long/A2cpts at 37 ℃ and 39 ℃. To determine the frequency of intermediate temperature sensitive (tsi) viruses, two plates of HEp-2 cells were inoculated at multiplicity of infection (MOI) = 0.1 and supplied with 200 μL of medium. At 5 days post-infection new uninfected HEp-2 cells in 96-well plates were inoculated with 100 μL/well of supernatant from the previous virus passage. Following a 1-h incubation, the inoculum was removed and the cells were fed with 200 μL of medium. Blind passaging was performed in duplicate and at progressively higher temperatures. For each passage, one plate was immunostained and a duplicate plate was used to seed the next passage. Six wells of HEp-2 cells infected with wild-type RSV (wtRSV) were used as a positive control.
To study the immunogenicity of the rRSVs, mice were immunized intranasally with 1 × 106 pfu of rRSVs or wtRSV. Mice inoculated with PBS were used as negative control. Serum samples were collected 21 days after immunization, and the levels of RSV-specific serum IgG were detected by ELISA (Fig. 5A). The mice in both immunized groups produced significant RSV-specific IgG responses compared to the negative control (P < 0.001), but there was no difference between the IgG levels induced by the three rRSVs (P > 0.05). Compared with wtRSV, similar immunogenicity was observed for rRSV-Long/A2cp and rRSV-Long/A2cptsΔSH (P > 0.05). However, rRSV-Long/A2cpts was characterized by a slightly lower level of immunogenicity than that of wtRSV (P < 0.05).
Figure 5. RSV-specific immune responses and neutralizing antibody responses following infection with recombinant human respiratory syncytial viruses (rRSVs) and wild-type RSV (wtRSV). BALB/c mice were immunized with either rRSV-Long/A2cp, rRSV-Long/A2cpts, rRSV-Long/A2cptsΔSH, or wtRSV at 1 × 106 pfu/mouse or PBS (50 μL/mouse), via intranasal (i.n.) route, or FI-RSV at 1.875 μg/mouse via intramuscular (i.m.) route at day 0. Serum antibody titers against RSV were examined by ELISA (A), and the levels of induced serum neutralizing antibodies were assessed by immunoplaque assay (B) at day 21 post-immunization. The results for the neutralizing antibodies were expressed as the reciprocal of the highest serum dilution fold providing 50% inhibition of plaque formation. The induced isotype of RSV-specific IgG levels in BALB/c mice after immunization with rRSVs were assessed by ELISA (C), and ratios of IgG2a and IgG1 were calculated (D). nc: negative control group. Data are shown as mean ± SD. ns: no significance, *P < 0.05, **P < 0.01, ***P < 0.001.
The cross-protection against subgroup A wtRSV Long and subgroup B (WV VR1400) infection was also detected and the neutralizing antibody level was measured (Fig. 5B). Similar to the IgG responses, no significant difference in the levels of neutralizing antibodies was induced by immunization with the three rRSVs (P > 0.05). Analogically, when compared with wtRSV, similar neutralizing antibody responses and cross-protective effects were induced by rRSV-Long/A2cp and rRSV-Long/A2cptsΔSH (P > 0.05). However, rRSV-Long/A2cpts was characterized by a slightly lower response level (P < 0.05).
To determine the type of immune response induced in mice by rRSV immunization, the titers of IgG2a and IgG1, RSV-specific IgG subtypes, were examined, and the IgG2a/IgG1 ratio was calculated. The results showed that both rRSV and wtRSV immunization elicited a Th1-biased immune response, different from the Th2-biased immune response induced by intramuscular immunization with FI-RSV (Fig. 5C, 5D). In addition, the FI-RSV immunized mice also showed a significantly reduced neutralizing antibody response compared with mice in the wtRSV group (P < 0.01). However, the IgG response was similar between both FI-RSV and wtRSV immunization groups.
To evaluate the efficacy of RSV vaccine candidates, changes in body weight and lung viral titers were examined in vaccinated mice after RSV challenge. The mice were challenged with wtRSV 28 days after rRSV immunization, and their body weight was monitored for 5 consecutive days (Fig. 6A). Weight loss in challenged mice immunized by wtRSV and the three rRSVs was significantly reduced from day 2 to day 5 compared to that of FI-RSV-immunized mice (P < 0.05 or P < 0.001). On the other hand, mice immunized by either rRSV-Long/A2cp or rRSV-Long/A2cptsΔSH did not display significant differences in weight loss compared with that of wtRSV (P > 0.05), but a significant difference was observed between mice immunized with rRSV-Long/A2cpts and wtRSV on day 3 and day 4 after the challenge (P < 0.05). Moreover, the immunization with rRSV-Long/A2cpts resulted in weight changes different from those induced by rRSV-Long/A2cp or rRSV-Long/A2cptsΔSH at day 3 after challenge (P < 0.05, data not shown). The slightly decreased neutralizing antibody response observed in mice after immunization with rRSV-Long/A2cpts when compared to other rRSVs, mentioned above, may be an explanation for this phenomenon. Altogether, rRSV-Long/A2cp and rRSV-Long/A2cptsΔSH exhibited the best efficacy among the three rRSVs, which was reflected by weight changes after wtRSV challenge.
Figure 6. Induction of protective immunity against human respiratory syncytial virus (RSV) infection in the vaccinated BALB/c mice. After the immunized BALB/c mice were challenged with wild-type RSV (wtRSV, 1 × 106 pfu/mouse) at day 28 post-immunization, the body weight changes were monitored daily from day 1 to day 5 following viral challenge (A). Lung homogenates were collected on day 5 post-challenge, and lung viral loads were determined by immunoplaque assay (B) and RT-qPCR (C). nc: negative control group. Data are shown as mean ± SD. ns: no significance, *P < 0.05, **P < 0.01, ***P < 0.001.
On day 5 after the challenge, the lung tissues were collected from the examined mice. Lung viral titers were analyzed by immunoplaque assay and RT-qPCR (Fig. 6B, C). Immunization with wtRSV and each rRSV resulted in decreased lung viral titers when compared to those observed in mice immunized with FI-RSV and mice from the negative control group (P < 0.001). On the other hand, there were no significant differences in lung viral titers among mice immunized with the three rRSVs (P > 0.05). However, the lung viral loads were significantly higher in mice from the rRSV-Long/A2cp and rRSV-Long/A2cpts groups than in the wtRSV-immunized group (P < 0.05). Further, no observable differences existed between the lung viral titers of mice immunized with rRSV-Long/A2cptsΔSH and wtRSV after wtRSV challenge (P > 0.05). These results showed that all rRSV-immunized mice were protected against RSV infection; however, rRSV-Long/A2cptsΔSH provided the best protection.
Since the FI-RSV vaccine causes ERD in immunized children and animals, the pathological examination of lungs in primed mice after RSV challenge is an important safety index for the RSV vaccine candidates. To this purpose, we collected the mouse lung tissues for histological sections 5 days after the challenge, performed H & E staining, and evaluated the pulmonary pathology under a microscope (Fig. 7A). The alveolar walls in mice immunized with FI-RSV were thickened, and alveolar cavities were compressed to form lung parenchyma. There was also a large number of inflammatory cells infiltrating around the blood vessels and bronchi. In negative control mice, the alveolar walls appeared thickened, the alveolar septa collapsed resulting in cavitation, and inflammatory cell infiltration was present around the blood vessels and bronchi. In the wtRSV- and rRSV-Long/A2cpts-immunized mice, thickening alveolar walls, cavitation, and increased levels of inflammatory cells were also observed in the lung tissues, albeit to a less extent. On the other hand, the immunization of mice with rRSV-Long/A2cp and rRSV-Long/A2cptsΔSH resulted in clear as well as less destroyed alveoli, and only a few inflammatory cells accumulated around the bronchi.
Figure 7. Absence of pulmonary histopathology and eosinophilia in mice vaccinated with recombinant human respiratory syncytial viruses (rRSVs) following RSV challenge. After the immunized BALB/c mice were challenged with wild-type RSV (wtRSV, 1 × 106 pfu/mouse) at day 28 post-immunization, lung tissues were collected at day 5 post-challenge and pulmonary histopathology was analyzed by hematoxylin and eosin staining (H & E) (A). The H & E-stained lung sections from each mouse were scored for inflammation, including peribronchiolar inflammation (B), perivascular inflammation (C), interstitial pneumonia (D), and alveolitis (E). Distribution and intensity of eosinophils were measured and analyzed using specific stains and Image-Pro Plus (F). IOD: integrated optical density. nc: negative control group. Data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.
Lung sections from all mice were scored for inflammation around airways, blood vessels, and interstitial and alveolar spaces as described in the Materials and Methods (Fig. 7B-7E). A considerably severe (score around 4) lung histopathology was observed in mice vaccinated with FI-RSV while a less severe lung histopathology (score around 2) was characteristic for mice in the negative control group. The least severe lung histopathology (score range, 1 to 2) was observed in at least two mice in each rRSV-vaccinated group (rRSV-Long/A2cp and rRSV-Long/A2cptsΔSH) compared to the FI-RSV vaccination group (P < 0.001) and the negative control group (either P < 0.05, P < 0.01, or P < 0.001). These results showed that vaccination with two of the three rRSVs resulted in protective immunity without causing any obvious signs of ERD, while the documented FI-RSV immunization-triggered ERD after viral challenge was confirmed. Altogether, the observed pulmonary pathology exhibited similar trends to those of the relative body weight loss among the experimental groups.
In the lungs of mice immunized with FI-RSV we also observed the infiltration of eosinophils, unlike in the lungs of mice from the rRSVs and wtRSV vaccination groups (P < 0.001) (Fig. 7F). In particular, the lungs of mice immunized with rRSV-Long/A2cptsΔSH displayed the least number of eosinophils in comparison with the wtRSV vaccination group and the negative control group (P < 0.05; P < 0.01). This result is consistent with the histopathological observations, and indicates decreased inflammation owing to a better protection against RSV infection through immunization with rRSV-Long/A2cptsΔSH.