To construct recombinant NDV expressing p72 (rNDV/p72), the open reading frame of the p72 gene was inserted into a genomic cDNA clone of NDV between the P and M genes (Figure 1A). The recombinant NDV was rescued from the full-length antigenomic cDNA clones as described in Materials and Methods. The correct sequence of the p72 gene was confirmed by PCR and sequencing. The stability of the p72 gene inserted in the NDV vector was assessed by passaging rNDV/p72 ten times through SPF 9-day-old embryonated chicken eggs. Sequence analysis of the p72 gene recovered after passaging showed that the integrity of the gene was conserved and that p72 was stably maintained without any mutations. Western blotting indicated that the three major NDV proteins (HN, NP, and M) were detected in lysates from cells infected with both rNDV and rNDV/p72. As expected, p72 was only expressed in lysates from cells infected with rNDV/p72 (Figure 1B).
Figure 1. Generation of recombinant Newcastle disease virus (NDV) expressing the p72 gene of African swine fever virus (rNDV/p72). (A) The p72 gene was flanked by the NDV gene-start (GS) and gene-end (GE) signals and inserted into the intergenic region between the P and M genes. (B) Production of p72 by rNDV or rNDV/p72. Baby hamster kidney (BHK-21) cells were infected with either rNDV or rNDV/p72 at a multiplicity of infection (MOI) of 1 and the cell lysates were collected 24 h post-infection for analysis by western blotting. NDV proteins were detected using anti-NDV antibodies directed against the HN, NP, or M proteins (upper panel). p72 (73 kDa) was detected using p72 monoclonal antibodies (lower panel). Lane 1: rNDV/p72, lane 2: Modified rNDV (control).
To evaluate whether the insertion of the p72 gene altered the pathogenicity of the NDV strain, we compared the pathogenicity of rNDV/p72 with its vector in chicks and mice. The pathogenicity of rNDVs in poultry was evaluated by the MDT assay in SPF embryonated chicken eggs and by the ICPI in 1-day-old chicks (Table 1). The MDT for both rNDV and rNDV/p72 exceeded 90 h, and, in fact, rNDV/p72 was more attenuated than the parental virus ( > 120 h for rNDV/p72 compared to 96 h for rNDV). The ICPI values of rNDV and rNDV/p72 were 0.53 and 0.00, respectively. Chicks infected with rNDV/ p72 had no significant clinical symptoms during the 8-day period of the ICPI test.
Virus MDT┼ ICPI╪ rNDV 96 h 0.53 rNDV/p72 > 120 h 0.00 Note: ┼Mean death time (MDT) pathotype deï¬nition: virulent strains, < 60 h; moderately virulent strains, 60 90 h; avirulent strains, > 90 h; ╪Intracerebral pathogenicity index (ICPI) is the criterion to classify the virulence of NDV isolates recommended by Office International Des Epizooties (OIE). Pathotype deï¬nition: virulent strains, 1.5 to 2.0; moderately virulent strains, 0.7 to 1.5; avirulent strains, 0.0 to 0.7.
Table 1. Pathogenicity of recombinant NDV and recombinant NDV expressing African swine fever virus protein 72 in embryonated chicken eggs and chicks.
The pathogenicity of rNDV and rNDV/p72 in mammals was evaluated in mice inoculated via the i.m. and i.n. routes. All of the mice survived the test. There were no marked body weight differences between the mice infected with viruses and the control group after either i.m. (Figure 2A) or i.n. (Figure 2B) inoculation. Lungs were collected on day 3 p.i. and the viral titer determined in SPF chicken embryos. The viral titer was below the limit of detection at this time point. These results suggested that the insertion of the p72 gene did not change the pathogenicity of rNDV in mice.
Figure 2. Weight changes in mice inoculated intramuscularly (i.m.) in the thigh muscle of a hind leg (A) or intranasally (i.n.) (B) with 108 50% egg infectious dose (EID50) in 100 μL (i.m.) or 30 μL (i.n.) of rNDV or rNDV/p72. The mice in each group (n = 5) were observed and weighed daily for 14 days. Body weight changes for each group are shown as ratios to the weight at day 0, which was set as 100.
To investigate whether the expression of the p72 gene affected the growth of the parental NDV strain, we evaluated the growth kinetics of each virus in BHK-21 cells. The cells were infected at an MOI of 1 (Figure 3A) or 0.01 (Figure 3B) in the presence of exogenous protease. Both of the viruses replicated efficiently in the BHK-21 cells, and had similar replication kinetics. The virus titers ranged from 103.5 EID50 to 107.5 EID50 when the initial inoculation was at the MOI of 1 (Figure 3A), and 102.5 EID50 to 106EID50 when the initial inoculation was at the MOI of 0.01 (Figure 3B). rNDV and rNDV/p72 remained at high titers 72 hpi.
Figure 3. In vitro growth kinetics of parental (rNDV) and recombinant (rNDV/p72) viruses in BHK-21 cells following infection at an MOI of 1 (A) or 0.01 (B). Exogenous protease was provided in the infected cells. The viral titers were determined in specific pathogen free (SPF) chicken embryos. Bars represent the mean SD of three independent experiments (n = 3).
Next, the levels of NDV specific antibodies were assessed to characterize the humoral response to rNDV and rNDV/p72. The NDV specific antibody activity was measured using an HI assay. None of the serum samples in the PBS group contained NDV specific antibody at any time. In contrast, the serum from all of the mice inoculated with rNDV or rNDV/p72 contained high levels of antibody after immunization (Figure 4). In addition, rNDV/p72 elicited similar levels of NDV specific serum antibodies as the parental virus (p > 0.05) (Figure 4).
Figure 4. Induction of NDV specific antibodies in 4-week-old mice in response to immunization with rNDV/p72. Mice were inoculated via the intramuscular route at 2 week intervals. Serum samples were collected 1 week after the 3rd and 4th immunizations. NDV specific antibodies were measured by a hemagglutination inhibition assay using chicken erythrocytes. The mean and standard error of antibodies for each group are shown.
Having established that rNDV/p72 elicited a strong humoral response to the NDV vector, we assessed the level of p72 specific antibody production. None of the mice had detectable anti-ASFV p72 antibody before immunization. Serum IgG antibody specific to p72 was evaluated by ELISA. All of the mice immunized with rNDV/p72 developed p72 specific antibodies 1 week after the third or fourth immunization. Mice that were immunized four times had higher p72-specific antibody levels than mice immunized three times (p < 0.05). No p72 specific IgG was detected in control mice immunized with PBS or with rNDV (Figure 5A). The IgG subclasses induced by p72, i.e. IgG1 or IgG2a, were determined by ELISA (Figure 5B, 5C). The p72-specific IgG1 and IgG2a antibody titers followed the same pattern as the total serum IgG. Higher levels of IgG1 than IgG2a were observed at both time points; the difference in levels was equivalent to a Th2/Th1 ratio of 6.
Figure 5. Antibody titers in mice after immunization with rNDV/p72. Mice were inoculated via the intramuscular route four times at 2 week wecks intervals. The titers of p72-specific total IgG (A) and the subtypes IgG1 (B) and IgG2a (C) were determined by ELISA against purified p72 protein. Serum samples were collected 1 week after the 3rd and 4th immunizations. The mean and standard error of antibodies for each group are shown (*p < 0.05, **p < 0.01, and *** p < 0.001).
To investigate the cellular response in mice induced by immunization with rNDV/p72, cytokine production (IFN-γ and IL-4) and T-cell proliferation in response to p72 stimulation were assessed in splenocytes collected 2 weeks after the last immunization. Cytokine production was assessed using the ELISPOT assay (Figure 6A) and the proliferative response was measured with the CCK-8 assay (Figure 6B). Mice immunized with rNDV/p72 showed high levels of p72-specific IFN-γ (135 SFC/106 secreting cells) and IL-4 (108 SFC/106 secreting cells), while mice immunized with the rNDV vector had background levels of cytokine production (p < 0.05) (Figure 6A). In addition, higher levels of T-cell proliferation in response to p72 were observed in mice immunized with rNDV/p72 compared to rNDV-immunized mice (p < 0.05) (Figure 6B).
Figure 6. The p72-specific cellular response after immunization with rNDV/p72. (A) Splenic lymphocytes from mice immunized with rNDV/p72 or rNDV were stimulated with purified p72 and analyzed for cytokine production (IFN-γ and IL-4) by enzyme-linked immunospot (ELISPOT). The mean numbers of spot-forming cells (SFC)/106 cells are shown and bars represent the standard deviation of SFCs in three parallel wells (n = 3). (B) Splenic lymphocytes were used to test T-cell specific proliferative responses using the Cell Counting Kit-8 assay. Proliferative response was determined by the O.D. value. The mean and standard error of the O.D. values for each group are shown (*p < 0.05).