Rescue of Newcastle Disease Virus from Cloned cDNA: Evidence that Cleavability of the Fusion Protein Is a Major Determinant for Virulence

Abstract

A full-length cDNA clone of Newcastle disease virus (NDV) vaccine strain LaSota was assembled from subgenomic overlapping cDNA fragments and cloned in a transcription plasmid between the T7 RNA polymerase promoter and the autocatalytic hepatitis delta virus ribozyme. Transfection of this plasmid into cells that were infected with a recombinant fowlpoxvirus that expressed T7 RNA polymerase, resulted in the synthesis of antigenomic NDV RNA. This RNA was replicated and transcribed by the viral NP, P, and L proteins, which were expressed from cotransfected plasmids. After inoculation of the transfection supernatant into embryonated specific-pathogen-free eggs, infectious virus derived from the cloned cDNA was recovered. By introducing three nucleotide changes in the cDNA, we generated a genetically tagged derivative of the LaSota strain in which the amino acid sequence of the protease cleavage site (GGRQGR↓L) of the fusion protein F0 was changed to the consensus cleavage site of virulent NDV strains (GRRQRR↓F). Pathogenicity tests in day-old chickens showed that the strain derived from the unmodified cDNA was completely nonvirulent (intracerebral pathogenicity index [ICPI] = 0.00). However, the strain derived from the cDNA in which the protease cleavage site was modified showed a dramatic increase in virulence (ICPI = 1.28 out of a possible maximum of 2.0). Pulse-chase labeling of cells infected with the different strains followed by radioimmunoprecipitation of the F protein showed that the efficiency of cleavage of the F0 protein was greatly enhanced by the amino acid replacements. These results demonstrate that genetically modified NDV can be recovered from cloned cDNA and confirm the supposition that cleavage of the F0 protein is a key determinant in virulence of NDV.

Newcastle disease is a serious avian disease with worldwide distribution that can cause severe economic losses in the poultry industry (2). The causative agent of the disease, Newcastle disease virus (NDV) or avian paramyxovirus type 1, is a member of the Paramyxoviridae and has been assigned to the genus Rubulavirus in the subfamily Paramyxovirinae (31). However, recently we presented evidence suggesting that NDV is not a member of the genus Rubulavirus but should be assigned to a new genus within the subfamily Paramyxovirinae (11). Similar to other Paramyxoviridae, NDV contains a nonsegmented single-stranded RNA genome of negative polarity (24). The RNA is 15,186 nucleotides (nt) in size (11, 23, 38) and contains six genes which encode the nucleocapsid protein (NP), phosphoprotein (P), matrix protein (M), fusion protein (F), hemagglutinin-neuraminidase (HN), and large polymerase protein (L) (30). In addition to these gene products, additional proteins (designated V and W protein) may be produced by an RNA-editing event that occurs during transcription of the P gene (48). NDV lacks the gene encoding the small hydrophobic protein that is occasionally present in members of the genus Rubulavirus (24).

NDV strains can be classified as highly virulent (velogenic), intermediate (mesogenic), or nonvirulent (lentogenic) on the basis of their pathogenicity for chickens (5). The molecular basis for pathogenicity of NDV is mainly determined by the amino acid sequence of the protease cleavage site of the F protein and by the ability of cellular proteases to cleave the F protein of different pathotypes (34, 35). Cleavage of the precursor glycoprotein F0 to F1 and F2 by host cell proteases is required for progeny virus to become infective (14, 34, 41). As a result, lentogenic viruses can replicate only in areas with trypsin-like enzymes such as the respiratory and intestinal tracts, whereas virulent viruses can replicate in a range of tissues and organs resulting in fatal systemic infection (34, 35). Lentogenic NDV strains such as the Hitchner B1 (18) and LaSota (15) strains are widely used as live vaccines against Newcastle disease. However, despite their widespread use, NDV live vaccines may still cause disease signs, depending upon environmental conditions and the presence of complicating infections. To improve the efficacy and safety of current NDV live vaccines, we set out to develop a system that would allow the genetic modification of NDV. Genetic modification of nonsegmented negative-strand RNA viruses has recently become possible (reviewed in references 10 and 36). This process, which is often referred to as reverse genetics, involves the in vivo transcription, by means of T7 RNA polymerase, of cDNA-encoded antigenomic RNA and the simultaneous expression, from cotransfected plasmids, of those viral proteins that are required for replication and transcription of the viral RNA. To date, infectious virus has been generated from cloned full-length cDNA of rabies virus (45), vesicular stomatitis virus (25, 50), Sendai virus (13, 20), measles virus (40, 44), human respiratory syncytial virus (9), rinderpest virus (4), human parainfluenza virus type 3 (12, 19), simian virus 5 (17), and bovine respiratory syncytial virus (7).

Here we report the generation of infectious NDV entirely from cloned cDNA, and we demonstrate that this procedure can be used to generate genetically modified NDV strains. We show that the virulence of NDV can be changed dramatically by modifying the protease cleavage site of the F0 protein. By using cDNA-derived NDV strains that differ only in the amino acid sequence of the protease cleavage site of the F0 protein, we proved that cleavability of the F0 protein is a key determinant for virulence of NDV.

MATERIALS AND METHODS

Cells and viruses.

Primary chicken embryo fibroblasts (CEF) cells and CER cells (47) were grown in Glasgow modification of Eagle’s medium (GMEM)-Eagle’s minimum essential medium (1:1; both from ICN) containing 5% fetal calf serum. QM5 cells (3) were grown in medium 199 supplemented with 10% tryptose phosphate broth and 10% fetal calf serum. BHK-21 cells (ATCC CCL-10) were grown in GMEM containing 5% fetal calf serum. NDV strain LaSota (ATCC VR-699) was plaque purified by three rounds of plaque purification on CEF cells. Virus from the third round of plaque purification (designated clone E13-1) was grown in 9- to 11-day-old embryonated specific-pathogen-free (SPF) eggs. The fowlpox recombinant virus fpEFLT7pol (6) (hereafter called FPV-T7), which expresses T7 RNA polymerase, was a kind gift from Michael Skinner and was grown on primary chicken embryo liver cells.

Construction of transcription vector.

A low-copy-number transcription vector, designated pOLTV5, was constructed from plasmid pOK12 (49) and transcription vector 2.0 (a generous gift of Andrew Ball) (37) by using standard molecular biological techniques (43). Plasmid pOK12 was digested with PvuII, and the DNA fragment containing the replication origin and the kanamycin resistance gene was isolated. This DNA fragment was ligated to an Eco47III-AflII fragment (the AflII site was made blunt by using Klenow DNA polymerase I) from transcription vector 2.0. From the resulting plasmid, an XbaI-NheI fragment was deleted to eliminate as much unique restriction sites as possible. Transcription vector pOLTV5 contains the T7 DNA-dependent RNA polymerase promoter followed by unique StuI and SmaI restriction sites, the autocatalytic ribozyme from hepatitis delta virus, and the transcription termination signal from bacteriophage T7. DNA fragments cloned between the StuI and SmaI restriction sites can be transcribed either in vitro or in vivo by using T7 RNA polymerase. After transcription, the 5′ end of the resulting transcripts contains two G residues encoded by the plasmid. Due to the autocatalytic activity of the hepatitis delta virus ribozyme, the 3′ end of the transcripts corresponds to the exact terminal nucleotide of the cloned DNA fragment (37).

Construction and cloning of full-length NDV cDNA.

Large (4- to 7-kb) subgenomic cDNA fragments spanning the entire NDV genome were generated by means of high-fidelity reverse transcription-PCR (RT-PCR) as described previously (11). Plasmids containing the 3′- and 5′-terminal NDV sequences (11) were used to generate DNA fragments that contained the exact 3′- or 5′-terminal end of NDV by means of PCR. The latter fragments were joined in an overlap PCR and cloned in pOLTV5 between the StuI and SmaI sites. Full-length NDV cDNA was assembled in this plasmid by joining subgenomic cDNA fragments at overlaps by using shared restriction sites according to the strategy shown in Fig. 1. The resulting plasmid was designated pNDFL (Fig. 2).

FIG. 1.

Full-length NDV cDNA was assembled from subgenomic overlapping cDNA fragments that were generated by high-fidelity RT-PCR as described by de Leeuw and Peeters (11). The cDNA fragments were joined at shared restriction sites and assembled in transcription plasmid pOLTV5 in which the 3′- and 5′-terminal NDV sequences were previously cloned between the StuI and SmaI sites (see text for details). The genetic map of NDV is shown at the top, and the horizontal lines below the genetic map indicate positions of the individual cDNAs. The bottom line shows the final composition of the full-length cDNA.

FIG. 2.

(A) Circular map of the pNDFL plasmids, which consist of full-length NDV cDNA cloned between the StuI and SmaI restriction sites of transcription plasmid pOLTV5 (see text for details). Relevant genetic elements and restriction sites are shown. T7, RNA polymerase promoter; rbz, hepatitis delta virus ribozyme; 3′, 3′-terminal leader sequence of NDV; 5′, 5′-terminal trailer sequence of NDV. (B) Nucleotide sequence in plasmid pNDFL of the T7 RNA polymerase promoter, the 3′- and 5′-terminal ends of NDV strain LaSota, and part of the flanking hepatitis delta virus ribozyme. The positions of the transcription start of T7 RNA polymerase and the cleavage site of the ribozyme are indicated by arrows. The sequence of NDV is boxed. (C) Nucleotide and deduced amino acid sequences of the region of the F gene that specifies the protease cleavage site of the F0 proteins in plasmids pNDFL and pNDFLtag. The nucleotide changes introduced to modify the amino acid sequence of the cleavage site are boxed (see also Fig. 3). The resulting amino acid changes are underlined and shown in bold. The arrows indicate the position of proteolytic cleavage. (D) Nucleotide sequence in pNDFLtagHPA of the synthetic 12-mer linker (boxed) which was used to insert a unique HpaI restriction site (bold) between nt 109 and 110 (XmnI site; GAANN↓NNTTC) of NDV.

Cloning and expression of NP, P, and L genes.

DNA fragments containing the NP, P, and L genes were generated by means of high-fidelity RT-PCR (11) and cloned in the eukaryotic expression vector pCIneo (Promega).

To clone the NP gene, primers 365 (5′-GTGTGAATTCCGAGTGCGAGCCCGAAG-3′) and 892 (5′-TTGCATGCCTGCAGGTCAGTACCCCCAGTC-3′) were used for PCR using Pwo DNA polymerase (Boehringer Mannheim), and the resulting DNA fragment was digested with EcoRI and cloned in pCIneo between the EcoRI and SmaI sites. Expression of the NP gene was verified in a transient expression assay by using an immunoperoxidase monolayer assay (IPMA) (51) using monoclonal antibody (MAb) 38 (42).

To clone the P gene, primers pRT1 (5′-CAAAGAATTCAGAAAAAAGTACGGGTAGAA-3′) and p2 (5′-GCAGTCTAGATTAGCCATTCACTGCAAGGCGC-3′) were used for PCR, and the resulting DNA fragment was digested with EcoRI and XbaI and cloned in pCIneo between the EcoRI and XbaI sites. Expression of the P gene was verified in an IPMA by using MAb 688 (42).

The L gene was recovered from cDNA clone L7a (Fig. 1) by digestion with SacII and SalI. Before digestion with SalI, the SacII site was made blunt by treatment with T4 DNA polymerase. The resulting fragment was cloned in pCIneo between the blunted (Klenow DNA polymerase) NheI site and the SalI site. The 5′ untranslated region between the T7 promoter and the ATG start codon of the L gene contained two out-of-frame ATG codons that might interfere with correct expression of the L protein. Therefore, by means of PCR mutagenesis, a derivative was constructed in which the first ATG was missing and in which the second ATG was changed to AAG.

Cotransfections.

CEF cells or QM5 cells were grown overnight to 60 to 80% confluence in six-well culture dishes and infected at a multiplicity of infection (MOI) of 1 with FPV-T7 for 1 h at 37°C. Subsequently, the cells were cotransfected with 0.25 μg of plasmid pNDFL, 0.4 μg of pCIneoNP, 0.2 μg of pCIneoP, and 0.2 μg of either pCIneoL or pCIneo (negative control) by using 9 μl of FuGene6 (Boehringer Mannheim) for CEF cells or 8 μl of LipofectAMINE (GibcoBRL/Life Technologies) for QM5 cells. After incubation for 4 h (CEF cells) or 16 h (QM5 cells), the transfection mixture was replaced by culture medium containing 5% allantoic fluid. Three to six days later, the culture supernatant was harvested, passed through a 0.20-μm-pore-size filter, and inoculated into the allantoic cavities of 9- to 11-day-old embryonated SPF eggs.

Modification of the protease cleavage site of the F0 protein.

The sequence encoding the protease cleavage site of the F0 protein was modified by means of PCR mutagenesis. To this end, the F gene was assembled from two overlapping PCR fragments. The first PCR fragment was generated by using primers NDV5F (5′-ACGGGCTAGCGATTCTGGATCCCGGTTGG-3′) and F5R (5′-AAAGCGCCGCTGTCTCCTCCCTCCAGATGTAGTCAC-3′). The residues shown in bold are changes introduced in the primer to change the amino acid sequence of the protease cleavage site of the F0 protein from that of the NDV LaSota strain (GGRQGR↓L) to that of the consensus cleavage site for virulent NDV strains (GRRQRR↓F). The second PCR fragment was generated by using primers F3F (5′-GGAGGAGACAGCGGCGCTTTATAGGCGCCATTATTGG-3′) and IV09 (5′-CTCTGTCGACACAGACTACCAGAACTTTCAC-3′). The two overlapping PCR fragments (the overlap is underlined in the primer sequences) were joined in a second PCR by using primers NDV5F and IV09. The resulting fragment, which contains the entire open reading frame of the F gene and encodes a virulent protease cleavage site, was cloned in pCIneo, yielding pCIneoF^wt. The StuI-NotI fragment (nt 4646 to 4952) from pCIneoF^wt was used to replace the corresponding fragment in subgenomic cDNA clone L21a (Fig. 1), which was subsequently used to generate full-length cDNA. The resulting plasmid was designated pNDFLtag (Fig. 2).

Introduction of a unique HpaI restriction site in full-length NDV cDNA.

Plasmid pNDFLtag (Fig. 2) was digested with AatII (Fig. 2A), and the smaller of the two fragments, which contained almost the entire pOLTV5 vector and a small part (226 nt) of the 3′ end of NDV, was circularized by using T4 DNA ligase. The resulting plasmid (designated pAAT2) was digested with XmnI, ligated to a nonphosphorylated synthetic 12-mer linker (CCTGTTAACAGG), and recircularized by using T4 DNA ligase. The resulting plasmid (pAAT2HPA), which contained a unique HpaI site (underlined), was digested with AatII and ligated with the larger AatII fragment from pNDFLtag to regenerate full-length NDV cDNA. The resulting plasmid was designated pNDFLtagHPA (Fig. 2A and D).

RT-PCR and sequence analysis.

The nucleotide sequence of the region of the F gene that encodes the protease cleavage site was analyzed by sequencing an RT-PCR fragment obtained by using primer 3UIT (5′-ACCAAACAGAGAATCCGTGAGTTA-3′) for reverse transcription and primers P4731+ (5′-AAGCTCCTCCCGAATCTGCC-3′) and P5020− (5′-GCGGCAATGCTCTCTTTAAG-3′) for PCR. The resulting PCR fragment was purified by using a High Pure PCR purification kit (Boehringer Mannheim) and sequenced by using a PRISM kit (Perkin-Elmer) and an Applied Biosystems ABI310 automatic sequencer. The presence of the unique HpaI restriction site in strain NDFLtagHPA was analyzed by performing an RT-PCR followed by digestion with HpaI and agarose gel electrophoresis. Primer 3UIT was used for reverse transcription, and primers 3UIT and P312− (5′-GCATCTTCGCTAACAGCAATCC-3′) were used for PCR.

Pulse-chase labeling and radioimmunoprecipitations.

Radioimmunoprecipitations were performed as described previously (39). Briefly, BHK-21 cells were infected with NDV at an MOI of 10 for 3 h at 37°C, after which time the medium was replaced by medium containing 1% of the original amount of methionine (starvation medium). After 2 h, this medium was replaced by starvation medium containing 50 μCi of [³⁵S]methionine and the cells were incubated at 37°C for 30 min (pulse). Subsequently, the medium was replaced by medium containing an excess of unlabeled methionine, and the cells were further incubated for 30 min (chase). Cell lysates were prepared by lysing the cells in phosphate-buffered saline containing 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate (PBSTDS), and the NDV F protein was precipitated from the lysates by incubation with MAb 8E12A8C3 followed by incubation with protein A-Sepharose. The precipitates were boiled in sample buffer and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis in 10% polyacrylamide gels. Labeled proteins were visualized by means of autoradiography.

HI tests and pathogenicity tests.

Hemagglutination inhibition (HI) tests were performed according to European Community directive 92/66/EC. The intracerebral pathogenicity index (ICPI) was determined in day-old chickens as described by Alexander (1).

RESULTS

Construction of full-length NDV cDNA.

To construct a full-length cDNA clone of NDV strain LaSota, overlapping cDNA clones spanning the entire NDV genome were generated and joined at shared restriction sites according to the strategy shown in Fig. 1. The generation, cloning, and sequencing of the subgenomic cDNA fragments have been described previously (11). The full-length NDV cDNA was assembled in transcription plasmid pOLTV5, a low-copy-number plasmid that was constructed from plasmid pOK12 (49) and transcription plasmid 2.0 (37). Plasmid pOLTV5 contains the T7 DNA-dependent RNA polymerase promoter followed by unique StuI and SmaI restriction sites, the autocatalytic ribozyme from hepatitis delta virus, and the transcription termination signal from bacteriophage T7. The resulting plasmid, in which the full-length NDV cDNA was cloned between the T7 RNA promoter and the hepatitis delta virus ribozyme (i.e., between the StuI and SmaI sites), was designated pNDFL (Fig. 2A and B). The nucleotide sequence of the NDV cDNA in pNDFL differed from the consensus sequence of NDV strain LaSota (11) at the following positions: nt 1755, G (A); nt 3766, A (G); nt 5109, G (A); nt 6999, T (C), nt 7056, G (A); nt 9337, G (A); nt 9486, A (T); nt 10195, T (C); and nt 13075, A (G) (consensus sequence in parentheses). These differences are probably the result of misincorporation during reverse transcription and/or PCR amplification. The nucleotide differences result in three amino acid changes: F protein, R¹⁸⁹ (Q); HN protein S²⁰⁰ (P); and L protein, N³⁶⁹ (I) (consensus between parentheses). Transcription of pNDFL by T7 RNA polymerase generates antigenomic RNA that contains two extra G residues at the 5′ end (Fig. 2B). Since it has been shown for a number paramyxoviruses that the presence of extra G residues did not prohibit the rescue of infectious virus from cDNA (13, 17, 19, 25, 45), we assumed that also in NDV these two extra G residues would not interfere with replication.

Generation of infectious NDV from full-length cDNA.

To generate infectious NDV entirely from cloned cDNA, CEF or QM5 cells were infected with FPV-T7, a fowlpox recombinant that expresses T7 RNA polymerase (6). Subsequently, the cells were transfected with plasmid pNDFL together with plasmids encoding the NDV NP, P, and L proteins that are essential for replication of NDV. As a negative control, the plasmid encoding the L protein (pCIneoL) was replaced by the empty vector plasmid pCIneo in parallel cotransfection experiments. After cotransfection, the cells were incubated for 3 to 6 days in medium containing 5% allantoic fluid. The addition of allantoic fluid is necessary because CEF or QM5 cells lack the proteases required for efficient cleavage of the F0 protein of lentogenic NDV strains. Cleavage of the F0 protein is absolutely required for cell-to-cell spread and for the generation of infectious virus (14, 34, 41). After 3 days of incubation, we performed an immunological staining of the fixed monolayers by using a MAb against the NDV F protein. The results showed that cells that expressed the F protein were present in monolayers which had been cotransfected with pNDFL and the NP, P, and L support plasmids (data not shown). This finding indicated that NDV genome replication and gene expression were occurring in these cells. Expression of the F protein was not detected when pCIneoL had been replaced by pCIneo in the cotransfection experiments.

To recover infectious virus, the supernatant of transfected monolayers was injected into the allantoic cavities of 9- to 11-day-old embryonated SPF eggs. Four days later, the allantoic fluid was harvested, analyzed in a hemagglutination assay, and passaged further in eggs. The results showed that only allantoic fluid from eggs inoculated with the supernatant of cells transfected with a combination of pNDFL and the NP, P, and L support plasmids yielded a positive reaction in the hemagglutination assay (data not shown). The allantoic fluid was subsequently analyzed in an HI test by using NDV HN-specific MAbs 7B7, 5A1, 7D4, and 4D6 (26, 29). The results showed that the NDV strain that was recovered from the inoculated eggs showed the same reaction pattern as the original LaSota strain (Table 1). The virus that was recovered from the inoculated eggs was designated NDFL to distinguish it from the original LaSota strain.

TABLE 1.

HI titers of antisera and MAbs

Strain	Titera					ICPIb	Pathotypec
	NDV serum	MAb 7B7	MAb 7D4	MAb 5A1	MAb 4D6
LaSota E13-1	256	320	5,120	40	10,240	0.31	Lentogenic
NDFL	128	2,560	10,240	1,280	10,240	0.00	Lentogenic
NDFLtag	256	640	10,240	20	10,240	1.28	Mesogenic
NDFLtagHPA	128	2,560	2,560	640	10,240	1.18	Mesogenic
Hitchner B1	256	80		320	10,240	0.11	Lentogenic
Herts	128	2,560				1.86	Velogenic
Texas GB	128	10,240			10,240	1.64	Velogenic

^aExpressed as the reciprocal of the highest serum dilution that caused inhibition of hemagglutination. Results for control serum and APMV3 serum were in all cases negative. 

^bDetermined as described by Alexander (1). 

^cLentogenic strains, ICPI < 0.7; mesogenic strains, ICPI 0.7 to 1.4; velogenic strains, ICPI > 1.4. 

Generation of genetically modified NDV from full-length cDNA.

To show unambiguously that the cotransfection system could be used to recover infectious virus from cloned full-length NDV cDNA, genetic tags were introduced in plasmid pNDFL. First, the amino acid sequence of the protease cleavage site in the F0 protein was changed from that of the LaSota strain (GGRQGR↓L) to the consensus sequence of virulent NDV strains (GRRQRR↓F) (41) by means of PCR mutagenesis (for details, see Materials and Methods). The resulting plasmid was designated pNDFLtag (Fig. 2A and C). By changing the amino acid sequence of the proteolytic cleavage site of the F gene, which is an important determinant for the virulence of different NDV isolates (14, 34, 41), we aimed at introducing both a genetic tag as well as a phenotypic tag. We made another derivative of pNDFL which contained, in addition to the modified F gene, a unique restriction site as the result of the introduction of a small oligonucleotide into the NDV genome. It has been shown that efficient replication of several paramyxoviruses is dependent on the genome length being a multiple of six nucleotides (so-called rule of six) (8, 22). The genome length of NDV LaSota (15,186 nt) is a multiple of six, which suggests that also replication of NDV is dependent on the rule of six. Therefore, a synthetic 12-mer oligonucleotide linker that contained an HpaI restriction site was introduced in the XmnI site between nt 109 and 110 of NDV in plasmid pNDFLtag. The resulting plasmid was designated pNDFLtagHPA (Fig. 2A and D).

Plasmids pNDFLtag and pNDFLtagHPA were used to generate virus by using the cotransfection system described above. Infectious viruses, designated NDFLtag and NDFLtagHPA, respectively, were recovered from the allantoic fluid of embryonated eggs. In an HI test, all MAbs including 7D4, which is specific for the LaSota strain (26, 29), showed the same reaction pattern with the newly generated viruses as with the original LaSota strain (Table 1). The nucleotide sequence of the region encoding the protease cleavage site of the F protein of NDFLtag was determined by sequencing an RT-PCR fragment that covered the corresponding genome sequence. The results showed that the expected nucleotide sequence was present in the F gene of NDFLtag (Fig. 3). The presence of the unique HpaI restriction site in NDFLtagHPA was verified by means of RT-PCR followed by HpaI digestion. As shown in Fig. 4, the PCR fragment from strain NDFLtagHPA was cleaved by HpaI and generated two fragments of the expected sizes. These results unambiguously show that NDFLtag and NDFLtagHPA were derived from plasmids pNDFLtag and pNDFLtagHPA, respectively, and demonstrate that genetically modified NDV can be generated from cloned full-length NDV cDNA.

FIG. 3.

Nucleotide sequence of the region of the F gene that specifies the protease cleavage site of the F0 protein of strains NDFL (top) and NDFLtag (bottom). DNA fragments containing the relevant regions were generated by RT-PCR and sequenced by using a PRISM kit (Perkin-Elmer) and an Applied Biosystems ABI310 automatic sequencer. The three nucleotide replacements, which were introduced in pNDFLtag by means of PCR mutagenesis, are boxed (cf. Fig. 2C).

FIG. 4.

Identification of the HpaI restriction site in strain NDFLtagHPA (cf. Fig. 2D). DNA fragments containing nt 1 to 312 of strains E13-1, NDFL, NDFLtag, and NDFLtagHPA were generated by means of RT-PCR. The PCR products were either left undigested (−) or subjected to HpaI digestion (+) and run on a 1.5% agarose gel in the presence of ethidium bromide. The marker lanes (M) to the left and right contain PstI-digested lambda DNA.

The protease cleavage site of the F protein of NDV is a key determinant for virulence.

It is generally assumed that the amino acid sequence of the protease cleavage site of the F0 protein is an important determinant for virulence of NDV strains (14, 34, 41). However, since other differences in the genomic sequence of different NDV strains may contribute to virulence (27, 28), definite proof is still lacking. The availability of strains NDFL and NDFLtag, which differ only in the amino acid sequence of the protease cleavage site of the F0 protein, offered the unique opportunity to test this supposition. The virulence of the original LaSota strain (clone E13-1) and of the newly generated strains NDFL, NDFLtag, and NDFLtagHPA was examined by determining the ICPI in day-old chickens. The results showed that the ICPI of strains NDFLtag and NDFLtagHPA were 1.28 and 1.18, respectively (out of a maximum possible value of 2.0), which is far above the values for strains NDFL (ICPI = 0.00) and clone E13-1 (ICPI = 0.31) (Table 1). These results demonstrate that, as expected, the virulence of NDV is largely (but not exclusively) determined by the amino acid sequence of the protease cleavage site of the F protein.

To show that the alteration in amino acid sequence of the protease cleavage site indeed affected cleavage of the F0 protein, BHK-21 cells were infected with strains NDFL and NDFLtag, viral proteins were labeled by means of pulse-chase labeling, and the F protein was subjected to radioimmunoprecipitation. The results showed that cleavage of the F0 protein of strain NDFL was much less efficient than cleavage of the F0 protein of strain NDFLtag (Fig. 5).

FIG. 5.

Proteolytic cleavage of the F0 proteins of NDV strains NDFL and NDFLtag. Cleavage of the F0 proteins was examined by means of pulse-chase labeling of infected cells followed by radioimmunoprecipitation of the F protein. BHK-21 cells were infected at an MOI of 10 for 3 h at 37°C, after which time the medium was replaced by medium containing 1% of the original amount of methionine (starvation medium). After 2 h, this medium was replaced by starvation medium containing 50 μCi of [³⁵S]methionine and the cells were incubated at 37°C for 30 min (pulse). Subsequently, the medium was replaced by medium containing an excess of unlabeled methionine and the cells were further incubated for 30 min (chase). Cell lysates were prepared by lysing the cells in PBSTDS, and the F protein was precipitated from the lysates by incubation with MAb 8E12A8C3 followed by incubation with protein A-Sepharose. The precipitates were analyzed by denaturing sodium dodecyl sulfate-polyacrylamide gel electrophoresis in 10% polyacrylamide gels, and labeled proteins were visualized by means of autoradiography. The positions of the precursor protein F0 and the cleavage product F1 are indicated. The small F2 subunit is not visible on this autoradiogram because it lacks methionine residues.

DISCUSSION

In this report we describe the generation of infectious NDV entirely from cloned full-length cDNA. This was accomplished by using an approach in which cDNA-encoded antigenomic RNA is synthesized by means of T7 RNA polymerase in cells that simultaneously express the viral replication proteins NP, P, and L from cotransfected plasmids. A similar approach has been used to generate infectious virus from cloned full-length cDNA of a number of other nonsegmented negative-strand RNA viruses (4, 7, 9, 12, 13, 17, 19, 20, 25, 40, 45, 50). In accordance with the results of others (13, 17, 19, 25, 45), we observed that the addition of two extra G residues at the 5′ end of the antigenomic RNA by T7 RNA polymerase did not prohibit the formation of infectious virus. Analysis by means of rapid amplification of the 5′ end of viral RNA of the cDNA-derived strains showed that the two G residues were lost during replication (data not shown).

The cotransfection system that we developed for NDV seems to be very efficient. By using approximately 1 μg of total DNA (0.25 μg of pNDFL, 0.4 μg of pCIneoNP, 0.2 μg of pCIneoP, and 0.2 μg of pCIneoL) per 3.5-cm-diameter culture dish, we can generate tens of infective centers in transfected monolayers (data not shown). Inoculation of a fraction (5 to 10%) of the transfection supernatant into embryonated eggs results in virus multiplication that can be easily quantitated by means of a hemagglutination assay. The addition of allantoic fluid to the culture medium of transfected cells was required for the successful recovery of infectious NDV. Allantoic fluid of embryonated eggs contains specific proteases that are absent from cultured cells and which are needed to cleave the F0 protein of lentogenic NDV strains (34, 35). Cleavage of the F0 protein into the F1 and F2 subunits is required to liberate a highly hydrophobic fusion domain at the N terminus of F1 that is required for the membrane fusion activity of the F protein (41). We expected that the addition of allantoic fluid to the transfection supernatant would not be necessary since the virus would be activated after inoculation of the transfection supernatant into the allantoic cavity of embryonated eggs. However, attempts to rescue infectious virus from transfection supernatants that lacked allantoic fluid were unsuccessful. This observation suggests that cleavage of the F protein is required for virus release. However, since the amount of virus that is released from the transfected monolayer will be much larger in the presence of allantoic fluid, this observation may merely indicate that successful recovery of virus from inoculated eggs is dependent on the amount of infectious virus in the inoculum.

The viral RNA that was used to synthesize cDNA was derived from plaque isolate E13-1 of NDV vaccine strain LaSota. Clone E13-1 was obtained after three rounds of plaque purification and was passaged in embryonated eggs before viral RNA was isolated. Nevertheless, the nucleotide sequence of the NDV cDNA in plasmid pNDFL differed at nine positions from the sequence that was determined for clone E13-1 (11). These differences are probably the result of misincorporation during reverse transcription and/or PCR amplification, not of strain variations. Despite these differences, the cDNA could be used to generate infectious virus (designated strain NDFL). When we tested the virulence of NDFL by intracerebral inoculation of day-old chickens, it proved to be virtually nonvirulent (ICPI = 0.00 [Table 1]). Since virulence was lower than that of the parent virus clone E13-1 (ICPI = 0.31), it seems that one or several of the nucleotides that differ between NDFL and clone E13-1 are responsible for the difference in virulence. In this respect, it is worthwhile mentioning that plaques produced on CEF cells by strain NDFL were somewhat smaller in size than plaques produced by strain E13-1 (Fig. 6). This difference in plaque size was more pronounced when the cells were grown under an agarose overlay then when they were grown under a liquid overlay. Furthermore, NDFL and E13-1 reached similar titers when they were grown in embryonated eggs (1.4 × 10⁹/ml for E13-1 and 1.2 × 10⁹/ml for NDFL, respectively), indicating that replication of strain NDFL is not impaired in embryonated eggs. Since NDFL is nonvirulent, it will be interesting to test its capacity as a live vaccine against Newcastle disease.

FIG. 6.

Plaques produced by NDV strains E13-1 (A) and NDFL (B) on CEF cells. CEF cells were infected with virus and incubated for 5 days under a 1% agarose overlay containing 5% allantoic fluid (11). Plaques were visualized by immunological staining (51) by using a MAb against the NDV F protein.

That genetic modification is a powerful tool to study the biological functions of viral gene products is demonstrated by our observation that modifying the protease cleavage site of the F0 protein can dramatically change the virulence of NDV. Amino acid sequencing of the F0 precursor proteins of many NDV strains has shown that lentogenic viruses have a single arginine (R) that links the F2 and F1 chains, whereas mesogenic or velogenic strains possess additional basic amino acids forming two pairs at the site of cleavage. Furthermore, the F2 chain of virulent strains generally starts with a phenylalanine (F) residue whereas that of nonvirulent strains generally starts with a leucine (L) (41). When the protease cleavage site of the F0 protein of strain NDFL (GGRQGR↓L) was converted to the consensus protease cleavage site of virulent NDV strains (GRRQRR↓F), this modification resulted in a dramatic increase in virulence from ICPI = 0.00 for NDFL to ICPI = 1.28 (out of a possible maximum of 2.0) for NDFLtag (Table 1). Thus, the virulence of NDFLtag is similar to that of a mesogenic strain whereas the antigenic profile is similar to that of the lentogenic parent strain LaSota (Table 1). That the genetic modification indeed affected cleavage of the F0 protein was verified by performing pulse-chase experiments followed by radioimmunoprecipitations, which showed that the efficiency of cleavage of the F0 protein was greatly enhanced by the amino acid replacements (Fig. 5). These results provide the definite proof that the efficiency of cleavage of the F0 protein is the key determinant for virulence of NDV. By using the same approach, the cleavage site of the F0 protein may be modified, at will, to any other amino acid sequence. This may lead to the generation of a series of NDV strains that display a spectrum of virulence levels in vivo.

Our results indicate that the efficiency of cleavage of the F0 protein is not the only determinant that is responsible for virulence of NDV. Velogenic NDV strains may exhibit an ICPI of as high as the maximum possible value of 2.0 (2). This indicates that, apart from the cleavage site of the F0 protein, additional nucleotide sequences in the genomic RNA of NDV contribute to virulence. For instance, differences in transcription and translation may modulate growth and cell-to-cell spread of the virus and/or cytopathogenicity (27, 28). The availability of an infectious cDNA of NDV allows for the systematic analysis of sequences that are involved in transcription and replication. This may lead to the design of new NDV vaccines that combine optimal immunogenicity with complete safety.

Apart from the modification of the cleavage site of the F0 protein, a genetic tag consisting of an HpaI restriction site was introduced in the NDV cDNA in order to facilitate the identification of infectious virus that is generated from cloned cDNA. Since replication of several paramyxoviruses (probably including NDV) is dependent on the genome length being a multiple of 6 nt (8, 22), we inserted 12 nt in the NDV genome between nt 109 and 110 (Fig. 2D). This particular position was chosen because it had been shown for Sendai virus that an insertion of a short oligonucleotide linker was tolerated at an analogous position close to the start of the NP gene but not at a position more upstream (16). Furthermore, it has been shown that the genomic and antigenomic promoters of the paramyxovirus simian virus 5 consist of two discontinuous elements, designated conserved region I, comprising the terminal 19 nucleotides, and conserved region II, comprising nt 73 to 90 from the terminal end (32, 33). The successful generation of recombinant virus NDFLtagHPA indicates that the insertion of 12 nt at a position close to but beyond the putative genomic promoter is well tolerated and apparently does not interfere with replication. The virulence of strain NDFLtagHPA (ICPI = 1.18) was close to that of its parent strain NDFLtag (ICPI = 1.28), indicating that the insertion did not greatly influence the virulence of the recombinant virus. Furthermore, plaques produced by NDFLtagHPA and NDFLtag on CEF were similar in size, and the two strains reached similar virus titers after growth in embryonated eggs.

Safety is one of the most important properties of live vaccines. However, for many live vaccines, including NDV, immunogenicity is often inversely related to virulence. Therefore, further attenuation of live vaccines without losing immunogenicity is one of the most desired alterations for which genetic modification could be used. In this respect, it has been shown that elimination of expression of the V protein of Sendai virus resulted in a markedly reduced in vivo pathogenicity for mice (21). Similar to Sendai virus, NDV generates a V protein from the P gene by a mechanism known as RNA editing (48). It is conceivable that elimination of expression of the V protein of NDV may also result in an attenuated phenotype in vivo.

Several properties make NDV an attractive vaccine vector for vaccination of poultry against respiratory or intestinal diseases. (i) NDV can be easily cultured to very high titers in embryonated eggs. (ii) Mass culture of NDV in embryonated eggs is relatively cheap. (iii) NDV vaccines are relatively stable and can be simply administered by mass application methods such as addition to drinking water or by spraying or aerosol formation. (iv) The natural route of infection of NDV is by the respiratory and/or intestinal tract, which are also the major natural routes of infection of many other poultry pathogens. (v) NDV can induce local immunity despite the presence of circulating maternal antibody. Since other paramyxoviruses have successfully been used for the incorporation and expression of foreign genes (16, 17, 46), we expect that NDV can also be used as a vaccine vector for the delivery of foreign antigens to the immune system.