Specific-pathogen-free (SPF) chicken eggs and chickens were purchased from Merial Vital Laboratory Animal Technology Co., Ltd (Beijing, China). All birds were kept in isolators at the China Agricultural University throughout the experiments and the animal rearing facilities were approved by the Administration Committee of Laboratory Animals under the auspices of the Beijing Association for Science and Technology (approval ID SYXK [Jing] 2013-0013).

The study was carried out in strict accordance with the Guidance for the Care and Use of Laboratory Animals formulated by the Ministry of Science and Technology of the People’s Republic of China. The experimental protocol, including the possibility of animal death without euthanasia, was specifically considered and approved by the Animal Welfare and Ethical Censor Committee at China Agricultural University (CAU approval number 1603–05). All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.

A chicken embryo fibroblast cell line (DF-1) was grown in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Grand Island, NY) containing 10% (v/v) fetal bovine serum (FBS) (Gibco) and maintained in DMEM containing 5% FBS. Baby hamster kidney (BHK-21) cells stably expressing T7 RNA polymerase (BSR T7 cells) were grown in DMEM containing 10% FBS and 1 mg/ml G418 (Invitrogen, Carlsbad, CA, USA). Both cell lines were grown at 37°C in an incubator (Thermo Forma, Marietta, OH, USA) under 5% CO₂. NDV strain SG10 was isolated from an outbreak of ND in chickens and identified as velogenic genotype VII virus with the intracerebral pathogenicity index (ICPI) of 1.79 and the mean death time (MDT) of 45 h (h) (Liu et al., 2015). NDV strain SG10 and the lentogenic genotype II strain LaSota (ICPI = 0, MDT > 120 h) were purified using the endpoint dilution method and propagated in 9-day-old SPF embryonated chicken eggs via allantoic cavity inoculation.

The construction of full-length antigenomic cDNA of NDV strain SG10 (rSG10) has been described previously (Figure 1) (Liu et al., 2015). To construct the recombinant LaSota cDNA clone, six fragments spanning the full-length LaSota cDNA were generated by reverse transcription-polymerase chain reaction (RT-PCR) of RNA that had been purified from the allantoic fluid of infected eggs. The six fragments were sequentially cloned into the pOK12 vector through seven restriction sites (MluI, SalI, KpnI, XbaI, HindIII, XhoI, and NotI). The leader region was preceded by a T7 RNA polymerase promoter that initiated the encoded antigenomic RNA with three G residues at its 5′ end. The trailer region was flanked by a hepatitis delta virus ribozyme sequence that cleaved after the last NDV-specific residue (Figure 1). Of the seven restriction sites in the NDV sequence, the MluI and the NotI restriction sites were introduced during RT-PCR and the others occurred naturally on the genome. In addition, in order for the restriction sites to be unique, the naturally occurring SalI, XbaI and HindIII sites in the P, M and L open reading frames (ORFs) were eliminated without changing the amino acid coding to act as genetic markers. The constructed plasmid was designated pOK-rLaSota and was confirmed to be identical to the biological virus by complete nucleotide sequencing.

The F and HN ORFs of NDV strain rSG10 were inserted in combination into a full-length antigenomic cDNA of strain rLaSota in place of the original NDV F and HN ORFs using the presence of unique restriction enzyme sites and the Seamless Assembly Cloning Kit (Invitrogen). First, the full-length antigenome of rLaSota was digested with KpnI and HindIII to generate a single DNA fragment without the F and HN genes. Second, the F and HN ORFs of rSG10 were engineered by PCR amplification as two DNA fragments each flanked by 5′ and 3′ UTRs of the respective rLaSota F or HN genes. To maintain the genome length as a multiple of six nucleotides, three extra nucleotides were introduced at the 3′UTR of the HN gene as necessary. The other three DNA fragments with homologous ends, one extending from the KpnI site to the F initiation codon, one from the F termination codon to the HN initiation codon and the other from the HN termination codon to the HindIII site, were generated by PCR amplification with compatible primers. Finally, these six DNA fragments with overlapping homologous ends of various lengths were effectively joined together using the Cloning Kit. The complete F and HN ORFs of NDV strain LaSota were inserted in combination into the SG10 backbone in place of the corresponding SG10 sequence using the same strategy. The full-length rLaSota cDNA plasmid containing the rSG10 F and HN genes was designated rLaSGFHN, whereas the full-length cDNA plasmid of rSG10 containing the rLaSota F and HN genes was designated rSGLaFHN.

The NP, P, and L ORFs were reciprocally exchanged between the LaSota and SG10 antigenome cDNAs using the relevant restriction enzyme sites noted above and the Seamless Assembly Cloning Kit (Invitrogen). Seven chimeras based on a full-length antigenomic cDNA of rLaSota was constructed in the same strategy: rLaSGNP, rLaSGP, rLaSGL, rLaSGNPP, rLaSGNPL, rLaSGPL, and rLaSGNPPL, as well as a parallel set of seven constructs based on a full-length antigenomic cDNA of rSG10: rSGLaNP, rSGLaP, rSGLaL, rSGLaNPP, rSGLaNPL, rSGLaPL, and rSGLaNPPL. All the exchanged regions in the full-length cDNAs were sequenced by dideoxynucleotide sequencing to confirm the presence of the desired genes.

The recombinant parental and chimeric viruses were recovered by co-transfection of each NDV chimeric full-length cDNA plasmid, along with helper plasmids expressing the NP, P, and L proteins, into BSR T7/5 cells, as previously described Liu et al. (2015). The respective helper plasmids of either LaSota or SG10 were used along with their respective full-length cDNAs to prevent potential heterologous recombination in the transfected cells. Four days after transfection, the cell culture supernatants and cell monolayers were harvested and used to recover the chimeric NDVs by injection into the allantoic cavities of 9-day-old SPF embryonated chicken eggs. Recovery of the virus was confirmed by hemagglutination (HA) assay using 1% chicken red blood cells (RBCs). Total RNA was extracted from NDV-positive allantoic fluid with TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. The rLaSota parents were completely sequenced, while only the exchanged genes and flanking sequences of the chimeric derivatives were sequenced. No adventitious mutations were detected. Cells transfected with rLaSota chimeric cDNAs were maintained in medium supplemented with 5 μg/ml N-tosylphenylalanine chloromethyl ketone (TPCK)-treated trypsin (Sigma-Aldrich, St. Louis, MO, USA) as a source of exogenous protease (Madansky and Bratt, 1981).

The pathogenicity of the chimeric viruses was determined by the MDT test in 9-day-old SPF embryonated chicken eggs and by the ICPI test in 1-day-old SPF chicks as described elsewhere (Cevik et al., 2004). Briefly, for the MDT test, a series of 10-fold dilutions of infected allantoic fluid was made in sterile phosphate buffered saline (PBS), and each dilution (0.1 ml) was inoculated into five 9-day-old eggs via the allantoic cavities route and incubated at 37°C. The eggs were examined once every 8 h for 7 days, and the time at which mortality of each embryo was first observed was recorded. The highest dilution yielding 100% mortality was considered to be the minimum lethal dose. The MDT was determined as the mean time (in hours) for the minimum lethal dose of virus to kill all the inoculated embryos. The MDT has been used to classify NDV strains as velogenic (taking less than 60 h to kill), mesogenic (taking 60–90 h to kill) or lentogenic (taking more than 90 h to kill). For the ICPI tests, for each virus 0.05 ml of a 1:10 dilution of fresh infective allantoic fluid was inoculated into 10 1-day-old SPF chicks via the intracerebral route. The birds were monitored for clinical symptoms and mortality once every 24 h for 8 days. At each observation, the birds were scored as follows: 0 if normal, 1 if sick and 2 if deceased. The ICPI is the mean of the score per bird per observation over the 8-day period. The pathotype definitions by the ICPI are as follows: virulent strains, 1.50–2.00; moderately virulent strains, 0.70–1.50; avirulent strains, 0.00–0.70.

The growth kinetics of rLaSota and rSG10 and their chimeric viruses were determined under multiple-cycle growth conditions in DF-1 cells and in SPF chicken embryos. Duplicate wells of six-well plates containing DF-1 cells were infected with virus at a multiplicity of infection (MOI) of 0.01. After 1 h of adsorption, the cells were washed with DMEM and then covered with DMEM containing 5% FBS and placed at 37°C under 5% CO₂ in an incubator. The medium of the cells infected with rLaSota or its chimeric viruses was supplemented with 5 μg/ml TPCK-treated trypsin. Supernatant samples were collected and replaced with an equal volume of fresh medium at 12 h intervals until 72 h post-infection (hpi). Virus titers in the collected supernatants were quantified via limiting dilution in DF-1 cells using the endpoint method of Reed and Muench (1938) and expressed as the 50% tissue culture infective dose (TCID₅₀).

To analyze the growth kinetics of chimeric viruses in chicken embryos, the allantoic cavities of 9-day-old SPF chicken embryos were inoculated with 10⁶ TCID₅₀ of each virus per embryo. Three embryos were chilled at regular 12 h intervals until 72 hpi, allantoic fluid samples were harvested, and the titers (TCID₅₀) of virus in the collected samples were determined by the method described above.

Two viruses, rSGLaNPPL and rLaSGNPPL, and their parent strains (rSG10 and rLaSota) were used to determine the pathogenicity in chickens. Four-week-old SPF white Leghorn chickens were assigned randomly to five groups of 20 birds each (10 for sampling and 10 for clinical observation) and each group was inoculated by the oculonasal route with either one of the four viruses or with 0.9% NaCl as the negative control. The target dose of the inoculum was 10⁶ 50% egg infectious dose (EID₅₀) of virus per bird. The birds were monitored daily for clinical signs for 14 dpi. Two birds were euthanized daily from days 1–5 pi for detection of gross lesions, and samples were collected into two parts. One part (trachea, lung, brain, proventriculus, cecal tonsil, liver, spleen, and duodenum) was used for virus titration and the other part (brain, spleen, lung, proventriculus, trachea and cecal tonsil) was fixed in 10% buffered formalin for histopathology. Necropsies were carried out, and external and internal abnormalities were recorded.

For virus titration, the tissue samples were homogenized in PBS containing antibiotics and virus titers were measured in DF-1 cells. The cleared tissue homogenates were serially diluted 10-fold and inoculated in DF-1 cells. At 72 hpi, cells were fixed with cold methanol and incubated with hyper-immune anti-NDV serum at 37°C for 1 h. Then the cells were washed three times with PBS and incubated with the secondary antibody, goat anti-chicken IgG-FITC (Sigma–Aldrich), at 37°C for 1 h. The virus titers were determined as TCID₅₀ per gram using the Reed and Muench method (Reed and Muench, 1938).

For histopathology, tissue samples were fixed in 10% neutral buffered formalin for 48 h. All sampled tissues were routinely embedded in paraffin and 5-μm sections were cut for hematoxylin and eosin staining and examined for lesions using light microscopy. The tissue sections were scored according to the severity of the observed lesions. The absence of injury was classified as -, while mild, moderate and severe injuries were classified as +, ++, and +++, respectively.

To detect whether viral RNA synthesis correlated with the intrinsic activity of the polymerase-associated proteins, two synthetic DNAs containing the T7 RNA polymerase promoter and the firefly luciferase (FLuc) gene (in an antisense orientation) flanked at the 5′ side by the trailer region and at the 3′ side by the leader region (LaSota and SG10), followed by the hepatitis delta virus ribozyme and the transcription termination signal, were synthesized and cloned into plasmid pGL3-Basic with KpnI and XbaI (Nishio et al., 2001). The constructed minigenome plasmids in which the number of nucleotides from the start of the trailer sequence to the end of the leader sequence was a multiple of six, were designated pLaSota-FLuc and pSG10-FLuc.

The BSR T7 cells were seeded in 24-well culture plates. The cells were cotransfected with either pLaSota-FLuc or pSG10-FLuc (340 ng) and helper plasmids expressing NP (340 ng), P (170 ng), and L (170 ng) proteins originating either from the LaSota strain or from the SG10 strain as well as a Renilla luciferase reporter gene (RLuc, 10 ng). At 20–22 hpi, the cells were washed in PBS once, and 100 μL passive lysis buffer (Promega, Madison, WI, USA) was added. Cells were vigorously mixed for 15 min, and 20 μL of lysate from each well were used for dual luciferase assay by dual luciferase assay kit (Promega). The relative luciferase activity unit was defined as the ratio of FLuc to RLuc activities (Dortmans et al., 2010). Six separate experiments were performed, with luciferase expression being measured in triplicate in each experiment.

All statistical analyses were performed using the unpaired t-test in GraphPad Prism Software Version 6.0 (GraphPad Software Inc., San Diego, CA, USA). Statistically significant differences between experimental groups were determined using the analysis of variance (ANOVA) method. The significance was considered as follows: significant, p = 0.01–0.05 (^∗); very significant, p = 0.001–0.01 (^∗∗); extremely significant, p = 0.0001–0.001 (^∗∗∗).

A cDNA clone encoding the antigenome of strain LaSota (rLaSota) was constructed from six cDNA segments that were synthesized by RT-PCR from virion-derived genomic RNA (Figure 1). The virulence of rLaSota was evaluated by the ICPI test in 1-day-old SPF chicks. The ICPI value of rLaSota was 0, which was the same as its biological parent LaSota (date not shown). These results showed that the cDNA-derived strain rLaSota and its parent had the same biological characteristics.

To determine the role of the viral surface glycoproteins in NDV virulence, the complete ORFs of the F and HN genes in combination were reciprocally exchanged between full-length antigenomic cDNAs of the velogenic genotype VII NDV strain rSG10 and the low-virulence genotype II strain rLaSota. Sequence analysis of the chimeric cDNAs confirmed the intended gene exchanges and the absence of any undesired mutations. The two recovered recombinant viruses were named rSGLaFHN and rLaSGFHN.

To evaluate the contributions of the polymerase-associated proteins (NP, P, and L) to the virulence of genotype VII NDV, the genes encoding these proteins were exchanged between strains rLaSota and rSG10 using reverse genetics. Fourteen chimeric viruses were recovered. Seven chimeric viruses were derivatives of strain rLaSota in which the NP, P and L genes were replaced individually or in combination with those of rSG10. Seven chimeric viruses were derivatives of rSG10, in which the NP, P and L ORFs were replaced individually or in combination with those of rLaSota. All chimeric viruses were amplified by RT-PCR and sequenced, which confirmed the correct structure of each exchange and a lack of adventitious mutations.

To determine the contributions to genotype VII NDV virulence of the F and HN protein genes in combination, the virulence of the two chimeric viruses along with their respective parental viruses was evaluated by determining the ICPI and MDT tests. The respective MDT values of parental rSG10 and rLaSota were 57 h and >120 h in embryonated eggs, and they had ICPI values of 1.86 and 0.00, respectively (Table 1). These values were consistent with the designations of rSG10 as velogenic and rLaSota as avirulent. Replacement of the envelope genes (F and HN) of the rSG10 strain in combination with those of the rLaSota strain (rSGLaFHN) resulted in a significant decrease in virulence, the MDT and ICPI values being >120 h and 0.86, respectively. Derivatives of rLaSota bearing the envelope protein genes (F and HN) of rSG10 (rLaSGFHN) in combination showed an obvious increased virulence with MDT and ICPI values of 79 h and 1.49 respectively (Table 1). The results indicated that exchange of these envelope-associated protein genes in combination between the low-virulence strain LaSota and the highly virulent genotype VII strain SG10 had a marked effect on the viral virulence.

We further examined the roles of the polymerase-associated proteins genes (NP, P and L) in NDV virulence. Derivatives of rSG10 bearing the polymerase-associated protein genes of rLaSota had respective MDT/ICPI values as follows: rSGLaNP, 72 h/1.78; rSGLaP, 52 h/1.89; rSGLaL, 83 h/1.68; rSGLaNPP, 58 h/1.90; rSGLaNPL, 86 h/1.48; rSGLaPL, 84 h/1.40; and rSGLaNPPL, 89 h/1.32 (Table 2). Derivatives of rLaSota bearing the polymerase-associated protein genes of rSG10 had respective MDT/ICPI values as follows: rLaSGNP, >120 h/0.00; rLaSGP, >120 h/0.00; rLaSGL, >120 h/0.00; rLaSGNPP, >120 h/0.34; rLaSGNPL, >120 h/0.14; rLaSGPL, >120 h/0.12; and rLaSGNPPL, >120 h/0.63 (Table 2). The results indicated that the individual replacement of the polymerase-associated genes between the SG10 strain and LaSota strain did not affect the virulence of NDV except for the rSGLaL. The simultaneous replacement of the three polymerase-associated genes (NP + P + L) had an evident effect on viral virulence, as shown by a decrease in the ICPI value from 1.86 for rSG10 to 1.32 for rSGLaNPPL and an increase in the ICPI value from 0.00 for rLaSota to 0.63 for rLaSGNPPL (Table 2).

The growth kinetics of 14 chimeric viruses in which the NP, P, and L genes were replaced individually or in combination between the low-virulence strain LaSota and the highly virulent genotype VII NDV strain SG10, along with their respective parental viruses, were compared using multicycle growth curves in DF-1 cells. When the growth of strains rLaSGNP, rLaSGP and rLaSGL was compared with that of the parental rLaSota, no significant growth differences were observed (p > 0.05) except for rLaSGNP which showed an obvious increase at 60 hpi (p < 0.05) (Figure 2A). However, the chimeric viruses rLaSGNPL and rLaSGNPPL, in which the NP + L or NP + P + L genes of the rLaSota virus in combination were replaced with those of the rSG10 virus, showed much higher virus titre than the parental rLaSota virus at 36 h and 48 hpi (p < 0.05) (Figure 2B). These results indicated that, among the polymerase-associated protein, the NP and L proteins had an obvious influence on the level of virus replication and this influence was greatest when induced by the interaction of homologous NP, P, and L proteins.

Some constructs of rSG10 bearing rLaSota-derived NP, P, and L genes, alone or in combination, showed a decreased replication compared with parental rSG10. The chimera rSGLaNP showed the lowest replication from 24 hpi to the end of the observation period (p < 0.05 or 0.01), while the chimera rSGLaL showed the lowest replication at 48 hpi (p < 0.05) (Figure 2C). The chimeric virus rSGLaNPL, in which the SG10 NP and L genes in combination were replaced with those of the LaSota virus, showed decreased replication at 24 hpi (p < 0.05), while another chimeric virus rSGLaNPPL also showed significantly decreased replication from 24 hpi to the end of the observation period (p < 0.05) (Figure 2D).

The in vivo growth characteristics of the 14 chimeric viruses and their parental viruses were further evaluated in 9-day-old SPF chicken embryos (Figure 3). For chimeric viruses with LaSota backbone, rLaSGNP, rLaSGP and rLaSGL grew at the same rate as their parental rLaSota virus. The other chimeric viruses, of which the rSG10 virus-derived NP + L, NP + P, or NP + P + L genes in combination were replaced with those of the rLaSota virus, showed a slightly increased growth compared to the parental rLaSota (Figures 3A,B). This indicated that the presence of the polymerase-associated genes (NP, P, and L) of rSG10 virus individually in the rLaSota backbone did not alter the replication of chimeric viruses, while the combinations had a slight effect in 9-day-old chicken embryos. For the chimeric viruses with SG10 backbone, rSGLaL showed decreased replication compared to the parental rSG10 virus at 48 hpi, while the embryos infected with rSG10, rLaSGNP and rLaSGP had all died at this time. The chimeric viruses rSGLaNPL and rSGLaNPPL showed a significantly decreased replication at 24 hpi (p < 0.05) (Figures 3C,D). These results indicated that changing the NP or P gene did not alter the replication of the chimeric viruses compared with that of their parental viruses, except that rSGLaL, rSGLaNPL and rSGLaNPPL showed decreased replication at an early or later stage of infection.

Based on previous results, the pathogenicity of two chimeric viruses (rSGLaNPPL and rLaSGNPPL) and their parental viruses (rSG10 and rLaSota) were evaluated further in 4-week-old SPF chickens by inoculating each bird with 10⁶ EID₅₀ of virus through the natural route of infection. No obvious clinical signs were observed in any of the birds inoculated with rLaSota, and slight punctate hemorrhages and catarrhal exudates were observed in the throat and trachea at days 4–5 pi (Table 3). Birds inoculated with rLaSGNPPL exhibited slight depression, punctate hemorrhages and catarrhal exudates in the throat and trachea, and hemorrhages and enlargement of the thymus at 4–5 dpi. The pathogenicity of the rSGLaNPPL and its parental virus differed greatly, the birds infected with rSGLaNPPL developing mild illness with moderate lesions and no mortalities, whereas the birds infected with rSG10 developed severe illness with severe lesions by days 3–5 pi and a 100% mortality by days 5.

Two chickens from each group were sacrificed daily on days 1–5 pi for virus titration by indirect immunofluorescence in DF-1 cells (Figure 4). The results showed that rLaSota could replicate in certain organs, such as trachea and occasionally in brain or duodenum. While rLaSGNPPL showed increased replication ability, and the virus replicated to moderate or high titers in many of the sampled tissues, such as proventriculus, brain, lung, duodenum and cecal tonsil and especially in trachea, the replication ability showed significant differences (p < 0.001) compared to rLaSota. For rSG10 and rSGLaNPPL, all sampled tissues in the rSG10-inoculated birds showed high viral titers from 3 dpi. The virus titers in most of the sampled tissues were significantly lower in rSGLaNPPL-inoculated birds than in birds inoculated with the parental rSG10 (p < 0.001), indicating a decreased replication ability of rSGLaNPPL compared with the parental virus.

Two chickens from each group were sacrificed at days 4 pi for histopathological analysis (Figure 5; Table 4). For the chimera rLaSGNPPL and its parental virus rLaSota, no obvious histological changes were observed in sampled tissues except for focal congestion in the lungs of the group inoculated with rLaSGNPPL (empty arrows) (Figure 5). The parental rSG10 caused the following moderate to severe histological changes in all of the sampled tissues: lymphocytic infiltration of the lamina propria (black arrows); submucosal congestion of the tracheal mucosa (empty arrows); hemorrhage, RBC infiltration (black arrows) and congestion (empty arrows) in the lung; severe mucosal epithelial shedding in the proventriculus (empty arrows); congestion of the lamina propria (black arrows); necrosis and shedding of lymphocytic cells in the mucosa of cecal tonsil (empty arrows); increased and gathered microglial cells in the brain (black arrows); and multifocal confluent coagulative necrosis (black arrows) in the spleen (Figure 5). However, the chimera rSGLaNPPL exhibited decreased virulence with mild or moderate lesions in the trachea, lung, proventriculus and cecal tonsil, as well as no obvious changes in the brain and spleen compared with the parental rSG10 virus.

To further detect whether viral RNA transcription and replication correlated with the intrinsic activity of the polymerase-associated proteins, two NDV minigenome systems (pLaSota-FLuc and pSG10-FLuc) were constructed and a dual luciferase assay was carried out.

For the pLaSota-Fluc minigenome, using the SG10 helper plasmids, NP or P proteins individually, no significant differences were observed in comparison (p > 0.05) with that of the LaSota replication proteins only. The FLuc activity was reduced in the presence of L, NP + L, and P + L proteins from SG10 (p < 0.001). However, the two combinations, NP + P and NP + P + L of SG10, caused a highly significant increase in FLuc activity compared with the native LaSota strain (p < 0.01) (Figure 6A). For the second pSG10-FLuc minigenome, when we replaced the SG10 helper plasmids, NP, P, L, NP + P, NP + L, P + L, NP + P + L with their LaSota strain counterparts, they showed a similar influence on the expression of the reporter gene, resulting in a significant decrease in replication of the FLuc gene (p < 0.001) (Figure 6B). These results revealed that the activity of the SG10 replication complex was higher than that of the LaSota replication complex, and the combined action of all three homologous replication proteins was required to reach the highest level of activity. The two minigenomes demonstrated different levels of reporter gene expression when driven by the LaSota replication proteins and those of the SG10 virus. This indicated that the nucleotide differences in the leader or trailer sequences that code for the genomic and antigenomic promoter (63.8% homologous) of LaSota and SG10 had a great influence on the efficiency of RNA synthesis.