Engineered Newcastle disease virus expressing the F and G proteins of AMPV-C confers protection against challenges in turkeys

Avian metapneumovirus (AMPV) infects the respiratory and reproductive tracts of domestic poultry, resulting in substantial economic losses for producers. Live attenuated vaccines appear to be the most effective in countries where the disease is prevalent. However, reversion to virulence has been demonstrated in several studies. Therefore, the development of a stable and safe next generation vaccine against the AMPV disease is needed. In the present study, we generated a recombinant Newcastle disease virus (NDV) vectoring the fusion (F) protein and glycoprotein (G) genes of AMPV subtype-C (AMPV-C) as a bivalent vaccine candidate using reverse genetics technology. The recombinant virus, rLS/AMPV-C F&G, was slightly attenuated in vivo, yet maintained similar characteristics in vitro when compared to the parental LaSota virus. Vaccination of turkeys with rLS/AMPV-C F&G induced both AMPV-C and NDV-specific antibody responses, and provided significant protection against pathogenic AMPV-C challenge and complete protection against velogenic NDV challenge. These results suggest that the rLS/AMPV-C F&G recombinant virus is a safe and effective bivalent vaccine candidate and that the expression of both F and G proteins of AMPV-C induces a protective response against the AMPV-C disease.

Previously, we generated Newcastle disease virus (NDV) LaSota vaccine strain-based recombinant viruses expressing the glycoprotein (G) of AMPV-A, B, and C viruses as bivalent vaccines 13,25 . Although these vaccine candidates were stable and safe in embryonated chicken eggs and vaccinated turkeys, they conferred only partial protection against homologous AMPV challenge in turkeys. It suggests that the G protein alone may be a weak antigen and other viral immunogenic components may be needed in combination to improve efficacy against AMPV diseases. Presence of the other major AMPV surface glycoprotein, fusion (F) protein, has demonstrated protective immunity in turkeys when expressed by either DNA vaccine or recombinant fowlpox virus 14,26 . Therefore, in the present study, we re-engineered the recombinant LaSota/AMPV-C G virus 13,25 to express the F protein of AMPV-C, in combination, in an effort to improve the vaccine efficacy against AMPV-C disease.

Generation of the rLS/AMPV-C F&G virus.
A full-length cDNA clone encoding the complete antisense genome of the NDV LaSota vaccine strain and the F and G genes of AMPV-C inserted between the F and HN genes of NDV was constructed using molecular biology techniques (Fig. 1). The resulting cDNA clone, pLS/ AMPV-C F&G, was co-transfected into HEp-2 cells, along with supporting plasmids, to rescue an infectious virus. After amplification in SPF chicken embryonated eggs, the LaSota strain-based recombinant virus vectoring the F and G genes of AMPV-C, designated as rLS/AMPV-C F&G, was purified and propagated. The fidelity of the rescued virus was confirmed by sequence analysis of the amplified viral genome (data not shown).

Biological characterization of the rLS/AMPV-C F&G virus.
To determine if the addition of the AMPV-C F gene affected viral pathogenicity and growth dynamics, the characteristics of the rLS/AMPV-C F&G virus were examined in vitro and in vivo using the MDT and ICPI tests and several titration assays. As shown in Table 1, the recombinant viruses appear to be slightly attenuated in day-old chickens with a lower ICPI (0.0) compared to the parental LaSota strain. The titers of the recombinant viruses grown in embryonated eggs or DF-1 cells, measured by EID 50 , TCID 50 , and HA, were comparable to the titers of the parental LaSota strain ( Table 1). Replication of the rLS/AMPV-C F&G virus was slightly delayed in the early stages (first 36 hours) of infection, but after time, was able to replicate to similar titers compared to the parental LaSota virus (Fig. 2). The open reading frame of the AMPV-C F gene, amplified from viral genomic RNA, was cloned into the pIRES-hrGFP-2a vector (pIRES, Clontech, Mountain View, CA) downstream of the IRES sequence. Subsequently, the IRES and the AMPV-C F ORF sequences were amplified and cloned into the pLS/AMPV-C G vector downstream of the NDV F ORF using an In-Fusion PCR Cloning Kit (Clontech). The NDV Gene Start and Gene End signal sequences and the AMPV-C F and G ORF sequences are underlined. The direction of the T7 promoter is indicated by a bold black arrow. HDVRz and T7Φ represent the site of the Hepatitis delta virus ribozyme and the T7 terminator sequences, respectively. Expression of the F protein by rLS/AMPV-C F&G. Expression the AMPV-C G protein from the rLS/ AMPV-C G, the parental virus of rLS/AMPV-C F&G, was detected previously 13 . Here we detected expression of the AMPV-C F protein from rLS/AMPV-C F&G infected DF-1 cells by IFA using chicken anti-AMPV-C F polyclonal antibody and the FITC-labeled goat anti-chicken IgG (Fig. 3b). To localize the AMPV-C F expression in relation to the NDV backbone, NDV HN protein expression was also detected by IFA using a mouse anti-NDV HN monoclonal antibody (Mab) and Alexa Fluor 568 conjugated goat anti-mouse IgG (Fig. 3c). After merging both fluorescent images, AMPV-C F and NDV HN protein expression co-localized to the same cells (Fig. 3d), which corresponded to typical NDV induced viral cytopathic effects (CPE) observed under bright-field microscopy (Fig. 3a). This confirms that the AMPV-C F protein was expressed during recombinant virus replication.
Immune response and protection against AMPV-C challenge. To assess the efficacy of each recombinant vaccine, turkeys were vaccinated with each recombinant virus and challenged with AMPV-C. No visible clinical signs were observed in turkeys in any treatment group post vaccination. Sixty percent (60%) of SPF turkeys that were immunized with the rLS/AMPV-C F&G virus produced AMPV-C-specific antibodies at 14 DPV and 100% at 28 DPV as measured by ELISA (Table 2). In contrast, only 40-50% of SPF turkeys vaccinated with the rLS/AMPV-C G virus developed AMPV-C-specific antibodies at 14 and 28 DPV (Table 2). From 3 DPC, 100% of the SPF turkey poults vaccinated with the LaSota control exhibited typical clinical signs of AMPV-C infection when challenged (Fig. 4), showing nasal exudates when squeezed (score 1), nasal discharge (score 2), and/or frothy eyes (score 3). The infected birds showed peak clinical signs between 6-7 DPC that gradually disappeared by 10 DPC. In the rLS/AMPV-C G vaccinated groups, 80% of turkeys showed clinical signs with mean scores of 1.  (Table 2). Challenge virus shedding from turkeys vaccinated with either rLS/ AMPV-C G or rLS/AMPV-C F&G was also detected on 3 and 5 DPC. By 7 DPC, the number of vaccinated birds shedding the challenge virus was decreasing, where 75% of turkeys in the rLS/AMPV-C G vaccinated groups and 40% of the birds in the rLS/AMPV-C F&G groups were positive for viral RNA (Table 2). This indicates that the rLS/AMPV-C F&G vaccine may induce better protection against challenge compared to the rLS/AMPV-C G vaccine.
Immune response and protection against NDV challenge. As shown in Table 3, turkeys immunized with the LaSota vectored and non-vectored vaccines induced a NDV-specific humoral response as measured by HI assay. All vaccinated turkeys were protected against a lethal NDV challenge without showing any signs of disease. In contrast, all of the birds in the unvaccinated control group displayed signs of disease with conjunctivitis and severe depression from 2 to 4 DPC and 100% mortality at 5 DPC (Table 3).

Discussion
Previously, we generated and evaluated NDV LaSota strain-based recombinant viruses expressing the AMPV G protein as bivalent vaccines, which provided partial protection against homologous AMPV challenge in turkeys 13,25 . In the present study, we further engineered the recombinant LaSota/AMPV-C G virus to express the other major surface glycoprotein of AMPV-C, the F protein, using an IRES to promote the foreign gene expression within a native NDV transcription unit 27 . This recombinant virus, rLS/AMPV-C F&G, was slightly attenuated in young chickens and delayed the onset of infection in cell culture when compared to the parental LaSota strain. It was  stable and did not show any apparent biological changes when measured by MDT and virus titration after 10 passages in SPF chicken embryos. These results demonstrate the safety, stability, and possible application of this recombinant vaccine candidate in young turkeys, which is the most vulnerable population for NDV and AMPV diseases 4,28,29 . The F and G proteins are two major surface glycoproteins of AMPV and believed to play important roles in virus infection, pathogenicity and immunogenicity. Expression of the F or G protein as a vaccine antigen from DNA vaccine, fowlpox or NDV vector conferred only partial protection in homologous challenged turkeys 13,14,25,26 . A combination of the F, G, and/or other antigenic conponents of the virus may be needed to improve the immune response and confer protection. In this study, we showed that the NDV recombinant vectoring both the AMPV-C F and G protein genes, rLS/AMPV-C F&G, induced a stronger immune response in turkeys against AMPV-C challenge than the rLS/AMPV-C G virus expressing the G protein alone. All turkeys vaccinated with the rLS/ AMPV-C F&G virus developed AMPV-C specific antibody by 28 DPV, whereas only 50% of the rLS/AMPV-V G virus-vaccinated turkeys seroconverted by this time. Turkeys vaccinated with the F&G virus showed less clinical signs and shed less challenge virus than birds in the control groups at 7 DPC, demonstrating the improved efficacy induced by the presence of both AMPV-C F and G antigens. However, less than 50% the turkeys vaccinated with the rLS/AMPV-C F&G virus still showed mild transient clinical signs between 5-7 DPC. This incomplete protection conferred by the rLS/AMPV F&G virus warrants the need to futher develop a recombinant AMPV-C vaccine candidate and program, perhaps by either relocating the AMPVgenes closer to the 3′ end of the NDV genome for increased expression of the antigens or employing a prime and boost vaccination stragegy.  In summary, in the present study we successfully generated a NDV LaSota strain-based recombinant virus vectoring the F and G protein genes of AMPV-C. Turkeys vaccinated with this recombinant virus were protected against a lethal velogenic NDV challenge and partially protected against pathogenic AMPV-C challenge. Compared with the rLS/AMPV-C G virus, the addition of the F gene in the new recombinant vaccine candidate, rLS/AMPV-C F&G, induced a stronger immune response in turkeys, indicated by reduced clinical signs and viral shedding after challenge. The results suggest that the rLS/AMPV-C F&G may be a promising bivalent vaccine candidate against NDV and AMPV-C disease, yet needs further development before use in large clinical trials on a commercial scale.

Methods
All methods used in this study were carried out in accordance with the guidelines and protocols approved by . The virulent AMPV-C virus stock (AMPV-CO Tr) was prepared from the tracheal tissues of AMPV-CO strain-infected SPF turkeys and titrated in SPF turkeys to determine 50% infective dose (ID 50 ) as described previously 16 . Viral RNA from the NDV-infected chicken embryo AF or DF-1 cells was extracted using the TRIzol-LS reagent according to the manufacturer's instructions (Life Technologies). Total cellular RNA from tracheal tissues was extracted using a MagMAX ™ AI/ND Viral RNA Isolation kit (Life Technologies) following the manufacturer's procedures.

Construction of recombinant LaSota cDNA clones containing the F and G genes of AMPV-C.
The backbone of the previously generated infectious LaSota/AMPV-C G clone, pLS/AMPV-C G 13 , was used to construct a recombinant cDNA clone containing the F gene of AMPV-C. To facilitate the expression of the AMPV-C F gene, an internal ribosomal entry site (IRES) sequence was inserted between the open reading frames (ORF)s of the NDV F and AMPV-C F genes (Fig. 1) as described previously 27 . Briefly, the cDNA encoding the AMPV-C F ORF was subcloned downstream of the IRES sequence in pIRES-hrGFP-2a vector (pIRES, Clontech, Mountain View, CA). Subsequently, the IRES and the AMPV-C F ORF sequences were amplified and subcloned downstream of the NDV F ORF in the pLS/AMPV-C G vector using the In-Fusion PCR Cloning Kit (Clontech). The resulting recombinant clone contains the AMPV-C F gene as the 2 nd ORF in the NDV F gene transcription unit and the AMPV-C G gene as an independent transcription unit between the NDV F and HN gene. The full-length cDNA clone was designated pLS/AMPV-C F&G and amplified in Stbl2 cells at 30 °C for 24 hours and purified using a QIAprep Spin Miniprep kit (Qiagen).
Virus rescue and propagation. Rescue of the recombinant, infectious LaSota/AMPV-C F&G virus was performed by transfecting the full-length cDNA clone, pLS/AMPV-C F&G, and supporting plasmids into HEp-2 cells as described previously 30 . The rescued virus was amplified by inoculating 100 µl of the infected cell lysate into the allantoic cavity of 9-day-old SPF chicken embryos and incubating the embryos at 37 °C. After 4 days of incubation, the AF was harvested and the presence of the rescued virus was detected by hemagglutination (HA) assay 31 . HA-positive AF was diluted in PBS and amplified in chicken embryos three times. The AF was harvested from the infected embryos, aliquoted, and stored at −80 °C as a stock. The nucleotide sequence of the rescued virus was determined by sequencing the RT-PCR products amplified from the viral genome as described previously 13 .
Virus titration, pathogenicity, and growth dynamics assays. Titers of the LaSota, rLS/AMPV-C G, and rLS/AMPV-C F&G viruses were analyzed and compared by the standard HA test in a 96-well microplate, the 50% tissue culture infectious dose (TCID 50 ) assay on DF-1 cells, and the 50% egg infective dose (EID 50 ) assay in 9-day-old SPF chicken embryos 31 . Pathogenicity of the recombinant virus was also determined using the standard mean death time (MDT) and intracerebral pathogenicity index (ICPI) assays 31 . mixture of FITC-labeled goat anti-chicken IgG (H + L) (SouthernBiotech, 1:1000 dilution) and Alexa Fluor 568 conjugated goat anti-mouse IgG (Life Technologies, 1:1000 dilution) for 1 hour at 37 °C. Immunofluorescence of the infected cells was visualized and photographed using an inverted fluorescence microscope at 100X magnification with filter combinations for FITC or Alexa Fluor 568 (Nikon, Eclipse Ti, Melville, NY).

Immunofluorescence assay (IFA). Expression of the AMPV-C F protein in infected DF-1 cells was exam-
Immunization and challenge experiment. To evaluate the efficacy of rLS/AMPV-C F&G vaccine against AMPV-C and NDV challenge, two animal experiments were performed. In experiment 1, SPF turkeys were used in the BLS-2E animal facilities at SEPRL to evaluate the protection against virulent AMPV-C challenge. In experiment 2, SPF turkeys were used in the BSL-3E animal facilities at SEPRL to evaluate the protection against lethal NDV challenge. All turkeys were housed in Horsfal isolators (Federal Designs, Inc., Comer, GA) with ad libitum access to feed and water. At the termination of the experiments, all birds were humanely euthanized in accordance with SEPRL's Institutional Animal Care and Use Committee approved animal use protocols. Experiment 1. Sixty one-day-old SPF turkey poults were randomly divided into six groups of 10 birds. Each bird in groups 1 and 2 was inoculated with 50 µl of the LaSota vaccine (10 7 TCID 50 /ml) via intranasal (IN) and intraocular (IO) routes as vaccine vector controls for a total of 100 µl. Birds in groups 3 and 4 were vaccinated with 100 µl of rLS/AMPV-C G (1.0 × 10 7 TCID 50 /ml) and birds in groups 5 and 6 were vaccinated with 100 µl of rLS/ AMPV-C F&G (1.0 × 10 7 TCID 50 /ml) per bird via IN/IO routes. At 14 days post-vaccination (DPV), the birds in groups 1, 3, and 5 were challenged with100 µl of pathogenic AMPV-C (1 × 10 3 ID 50 /ml) via IN/IO routes. At 28 DPV, the birds in groups 2, 4, and 6 were challenged with the same dose of pathogenic AMPV-C via IN/IO routes. Immediately before challenge, blood samples were collected from each bird for detection of serum antibody responses. The challenged birds were monitored daily for clinical signs of AMPV disease for 14 days. Typical clinical signs of the AMPV disease, nasal exudates when squeezed (score 1), nasal discharge (score 2), and/or frothy eyes (score 3), were assessed using the scoring system of Cook et al. 17 . At 3, 5, and 7 days post-challenge (DPC), intra tracheal swabs were collected from each bird for detection of challenge virus shedding.  32 . Serum samples were collected immediately before challenge for NDV antibody detection using the standard hemagglutination inhibition (HI) assay 31 . After challenge, the birds were monitored daily for clinical signs of Newcastle disease and mortality for two weeks.
Detection of immunoresponse and challenge virus shedding. The antibody response against NDV was determined by performing a standard HI assay using LaSota virus as antigen 17,31 . The antibody response to AMPV-C was examined by an enzyme-linked immunosorbent assay (ELISA) as described previously 16 . Briefly, Sucrose gradient purified AMPV-C virus was used as antigen. Turkey sera were diluted (1:100) and individually tested in triplicates. Negative and AMPV-C positive sera were included in each plate of the ELISA test as controls. Turkey sera were deemed positive when the mean chemiluminescence relative light unit (RLU) value was greater than the mean RLU plus 2× standard deviation of the three dilutions (1:100-1:400) of the negative turkey serum pool. Viral shedding from turkey tracheal tissue following challenge with AMPV-C was detected by RT-PCR with AMPV-C N gene-specific primers as described previously 13,16 .