Development of a novel Newcastle disease virus (NDV) neutralization test based on recombinant NDV expressing enhanced green fluorescent protein

Two cell lines were used in this study, DF-1 (derived from chicken fibroblasts) and Vero (monkey kidney cells), which were purchased from ATCC (Manassas, VA, USA). Both cell lines were maintained in Dulbecco’s modified Eagle medium (DMEM) F12 (HyClone) supplemented with 5% heat-inactivated fetal bovine serum (FBS), 2.5% chicken serum (ChkS) (Sigma–Aldrich), 100 U/mL of penicillin and 100 μg/mL of streptomycin at 37 °C in an atmosphere of 5% CO₂.

Based on the published nucleotide sequence of lentogenic NDV strain in the GenBank (Accession No. Y18898) [22], a full-length clone of NDV (pFLC-LS1) was assembled in pCI-modified vector (pCI/−SnaBI) from three overlapping fragments. Figure 1 shows sequential subcloning of three fragments into the vector, each having unique restriction sites, namely NheI-EagI, BssHII-SgrAI, and SnaBI-SapI, respectively. Fragments-1 and -2 were chemically synthesized (GenScript, New Jersey, USA) that contain NDV sequences, and fragment-3 was taken from previously recovered full-length pNDV-PE18 [23], which was derived from LaSota strain. Fragment-1 (4387 nt) includes sequences (from 5′ to 3′) of a self-cleaving hammerhead ribozyme (HHRz), the N, P and partial M genes, between the NheI and EagI restriction sites. The HHRz sequence 5′-CTG ATG AGT CCG TGA GGA CGA AAC TAT AGG AAA GGA ATT CCT ATA GTC-3′ was included immediately upstream of the Leader sequence of the virus. Fragment-2 (5865 nt) consists of partial M, F, HN and partial L protein genes and the hepatitis delta virus anti-genome ribozyme (HDVRz). The HDVRz sequence 5′-GGG TCG GCA TGG CAT CTC CAC CTC CTC GCG GTC CGA CCT GGG CAT CCG AAG GAG GAC AGA CGT CCA CTC GGA TGG CTA AGG GAG AGC CA-3′ was inserted immediately downstream of the Trailer sequence of the virus [24]. Fragment-3 (4989 nt) contains the remaining L gene sequence (4769 nt upstream) between SnaBI and SapI restriction sites. All nucleotide changes due to creation or deletion of restriction sites in the clone are shown in Table 1. Finally, the resulting plasmid pFLC-LS1 (19,319 bp) was obtained which contains the plus sense genome sequence of NDV (15,186 nt). DNA of this plasmid was completely sequenced to verify its integrity by Sanger sequencing (Macrogen Inc., Korea).

Supporting plasmids were generated from the full-length clone pFLC-LS1 vector using primers Nfor, Nrev, Pfor, Prev, Lfor and Lrev (Table 2). Amplified open reading frames (ORFs) of the N, P and L protein genes were cloned into the pCI vector. The resulting plasmids pCI-N, pCI-P and pCI-L were used for the recovery of recombinant viruses. The ORFs of N and P proteins were amplified by PCR using gene-specific primers that contain EcoRI and NotI restriction sites, whereas the L ORF was amplified using Lfor and Lrev primers that contain SpeI and NotI restriction sites, respectively. The L ORF was cloned between NheI and NotI, which resulted in the destruction of NheI/SpeI restriction site. All forward primers contained the Kozak sequence GCCACC immediately upstream of the start codon (ATG).

To insert the eGFP gene in the NDV full-length clone, fragment-A was chemically synthesized (GenScript, New Jersey, USA). Fragment A flanked by AsiSI and PciI restriction sites contains the HHRz ribozyme sequence, Leader sequence, eGFP gene cassette flanked by the gene-start and gene-end of the N gene and right after the gene-end a second gene-start that will be part of the N cassette as shown in Fig. 2. To achieve high levels of expression, the eGFP cassette was inserted into the most 3′-proximal locus [25] between AsiSI and BbvCI sites of the clone by ligation with fragment-B (Fig. 2). Fragment-B contains the N and P genes flanked by PciI and BbvCI restriction sites. This fragment was obtained by PCR amplification with primers NDV116_F and NDV3211_R (see Table 2), using the pFLC-LS1 as a template. Both fragments were cloned simultaneously in pFLC-LS1 between AsiSI and BbvCI sites (Fig. 2). The resulting plasmid was designated as pFLC-LS1-1eGFP which contains the NDV genome and eGFP gene that is 16,182 nt long, accomplishing the rule of six. DNA of this plasmid was completely sequenced to verify its integrity by Sanger sequencing (Macrogen Inc., Korea).

High-quality plasmid DNA was obtained from the pFLC-LS1, pFLC-LS1-1eGFP and three supporting plasmids using the Plasmid Midi Kit (Qiagen). Vero cells were grown to 80% confluency in a 12-well plate and then co-transfected with 1 μg of pCI-N, 0.5 μg of pCI-P, 0.2 μg of pCI-L and 2 μg of pFLC-LS1 or pFLC-LS1-1eGFP, using the Lipofectamine™ LTX & plus Reagent (Invitrogen, USA) according to manufacturer’s instructions. Briefly, plasmids were diluted in a 400 μl of Opti-MEM medium (Invitrogen, USA). Next, Lipofectamine was added and incubated for 15 min at room temperature. Vero cells were washed with Dulbecco’s Phosphate Buffered Saline (DPBS) and then the plasmid-Lipofectamine mixture was added to the monolayer. After 4 h of incubation at 37 °C, transfected cells were washed and then maintained in DMEM containing 5% FBS at 37 °C and 5% CO₂. Next day, allantoic fluid (AF) was added to a final concentration of 5% and cells were incubated for 4 more days. Cell supernatant was harvested, pelleted at high speed and inoculated into 8-days old SPF chicken embryonated eggs and incubated for 4 days. AFs were harvested, clarified, and aliquoted and stored at −80 °C. Recovery of the viruses was first confirmed by hemagglutination assay (HA). For parental virus (rLS1), the presence of NDV-specific protein was evaluated by immunofluorescence and confirmation of genetic markers was performed by RT-PCR and restriction enzyme digestions. To confirm the recovery of rLS1-1eGFP virus, infected DF-1 cells were observed under a fluorescence microscope.

DF1 cells infected with the rLS1 virus at a multiplicity of infection (MOI) of 0.01 for 48 h. After discarding the supernatant, the cells were washed with DPBS 3 times (as all washing steps in this section) and fixed with 4% paraformaldehyde in DPBS for 15 min at room temperature. Fixed cells were washed, permeabilized with 1% SDS in DPBS for 15 min at room temperature and washed again. Permeabilized cells were blocked with 5% Bovine Serum Albumin (BSA) (Sigma-Aldrich) in DPBS for 30 min at room temperature. Cells were incubated for 2 h with a monoclonal antibody against the NDV ribonucleoprotein (RNP) (cat. n° ab138719, Abcam, USA) at a final concentration of 3.85 μg/mL in a solution of 5% BSA in DPBS. After washing, the cells were incubated for 1 h with a goat polyclonal antibody anti-mouse IgG labeled with Alexa Fluor 594 (cat n° ab150116, Abcam) at a final concentration of 1 μg/mL in a solution of 5% BSA in DPBS, then washed. Nuclei were stained with DAPI for 5 min. Cells were examined under the Observer.A1 fluorescent microscope (Carl Zeiss, Germany). The fluorescent signal images were taken at 400X magnification with the AxioCam MRc5 camera (Carl Zeiss, Germany).

The identity of genetic markers in recombinant NDVs was confirmed after RT-PCR amplification from genomic RNA and sequencing of the DNA fragments. Two genetic markers evaluated were; creation of the BssHII site in the M gene and destruction of the BbvCI site in the L gene (see Table 1). Briefly, viral RNA was extracted from AFs stocks using the QIAamp Viral RNA mini kit (Qiagen) and cDNA was generated using the ProtoScript® II First Strand cDNA Synthesis Kit (New England Biolabs Inc). The PCR reactions were performed with primers M + 4100 [26] and NDV4394_R to amplify the region containing the BssHII site, and primers NDV14035_F and NDV15003_R for BbvCI site that was destroyed (see Table 2). Restriction analysis of the PCR products was carried out on a 2% agarose gel.

DF-1 cells were seeded into 12-well plates at density 1.5 × 105 cells in 1 mL of DMEM F12 per well, a day before the assay. Next day, serial dilutions of 1:10 were made by mixing 25 μL of virus with 225 μL of serum free DMEM F12. After washing DF-1 cells monolayer with DPBS, dilutions (0.2 mL/well) were added into wells and incubated for 1 h at 37 °C in an atmosphere of 5% CO₂ for viral absorption. Later, the inoculum was removed and each well was covered with 1 mL of an overlay medium consisting of DMEM supplemented with 0.5% agarose, 5% AF and 30 mM MgCl₂ [27]. The overlay medium was allowed to solidify by placing the plates on a level surface at room temperature for 15 min, the plates were then incubated for 5 days at 37 °C and 5% CO₂. The plates were fixed and stained with a mixture of 0.2% crystal violet (Sigma-Aldrich) and 3.2% paraformaldehyde overnight at room temperature [28]. The plates were rinsed with tap water, dried, viewed and photographed for EliSpot (EliSpot Reader versión 7.0). The NDV titers were reported as plaque-forming units per milliliter (PFU/mL).

Monolayer cultures of DF-1 cells were seeded at 50–60% confluence in 12-well plates and infected with rLS1 or rLS1- eGFP viruses at a (MOI) of 0.05. Cells were cultured with DMEM containing 1% FBS and 5% AF with 5% CO₂ at 37 °C. Supernatants were collected 12, 24, 36, 48, 60 and 72 h post-infection (h.p.i.). Collected supernatants were quantified in DF-1 cells by plaque assay as described above.

Serums samples obtained from 57 chickens were used for the experiments. Thirty seven serum samples were collected from field vaccinated chickens from different farms. Six commercially available chicken sera (Charles River Laboratories), corresponding to sera anti-Marek’s disease virus (MDV), anti-Avian adenovirus type-1, anti-Infectious Laryngotracheitis virus (ILTV), anti-Infectious bronchitis virus (IBV), anti-NDV, another anti-NDV serum (cat. n° ab34402, Abcam), and a serum from a SPF chicken was included as negative control. The remaining 13 serum samples were obtained from SPF chickens inoculated with one dose of LaSota vaccine at 1 day of age and their sera taken at the 4th week. All serum samples were heat inactivated (56 °C for 30 min) and stored at −20 °C [29]. Detailed information about the serum samples is listed in Additional file 1: Table S1.

Conventional NT and eGFP-NT were carried out to quantify nAbs in chicken sera. For both assays, serum samples were titrated in duplicate, using 96-well flat bottom nonpyrogenic polystyrene culture plates (Corning, NY, USA). A day before the assay, 1 × 10⁴ DF-1 cells per well were seeded in 0.1 mL of DMEM F12 supplemented with 5% FBS. The serum samples were serially diluted by 2-fold (starting from 1:2) with DMEM F12 serum free medium, supplemented with FA and FBS at final concentrations of 5% and 1%, respectively. Serum dilutions were mixed with 100 PFU of virus (rLS1 or rLS1-1eGFP, respectively). Each dilution was evaluated in duplicate. For conventional NT, DF-1 cells were infected with mixtures of virus-serum. After 4 days of incubation at 37 °C in 5% CO₂, cells were washed with DPBS. Then cells were fixed and stained with a mixture of 0.2% crystal violet and 3.2% paraformaldehyde for 15 min at room temperature. NDV nAb titers were determined as the reciprocal of the highest dilutions that both replicates presented a clear CPE.

For eGFP-NT, once DF-1 cells were infected with mixtures of virus-serum, cells were incubated for 48 h and observed under a fluorescence microscope and NDV nAb titers were determined as the reciprocal of the highest dilutions that did not express the eGFP in any of the duplicates. Wells with one or more fluorescent foci were considered positive [30].

HI was performed according to the OIE terrestrial manual [31]. Briefly, 25 μL of PBS per well were dispensed in clear 96-well V-bottom plates, then 25 μL of serum was placed on the first well, 2-fold dilutions of 25 μL of the serum suspension were made across the entire plate. Then, 25 μL of diluted virus containing 2⁴ hemagglutination units (HAU) of rLS1 virus was added to each well and incubated for 30 min at room temperature. Next, 25 μL of chicken red blood cells (RBCs) at 1% were added to each well and incubated for 40 min at room temperature. The titration was determined as the highest dilution of serum causing complete HI.

ELISAs against NDV (IDEXX laboratories, USA) were performed on all serum samples at room temperature, according to manufacturer’s instructions. Briefly, 100 μl of each serum sample (diluted 1:500 in DPBS), and 100 μl of the negative and positive control samples were dispensed into duplicate wells and incubated for 30 min. The plates were then washed thrice and incubated with 100 μl of conjugate per well for 30 min. Wells were washed thrice and then 100 μl per well of substrate solution was added, incubated in the dark for another 30 min, after which 100 μl stop solution was immediately added. The plates were read using an Epoch 2 microplate reader (Biotek, USA) at 450 nm. Data obtained were analyzed and sample to positive ratio (S/P) calculated.

The t-student test was performed to compare the mean titer of each time of the growth kinetics curve. Linear regressions were performed for the analysis of correlation between NT and eGFP-NT, HI and ELISA titers. All statistical analysis were performed in GraphPad Prism 6.01 (GraphPad Software Inc., San Diego, CA).

The full-length NDV clone pFLC-LS1 was assembled together with three fragments through several restrictions sites, as depicted in Fig. 1. We used an RNA polymerase II promoter system to recover the virus, as previously reported for NDV [32, 33], which allowed the virus rescue in a T7 RNA polymerase independent fashion. Unique restriction sites created or destroyed by silent mutations into the full-length clone as genetic markers (see Table 1) were verified by sequencing the DNA. This full-length clone, along with the supporting plasmids, was used to transfect DF-1 cells. Three days after transfection, the cytopathic effect (CPE) was observed and recombinant rLS1 was recovered at a high titer after the first passage in chicken eggs (1 × 10⁸ PFU/mL). Immunofluorescence assay of infected cells was positive when reacted with NDV-specific monoclonal antibody against the RNP complex, confirming the presence of NDV (Fig. 3a). To verify the identity of the rLS1 virus and dismiss any possible contamination by other NDV strain i.e. LaSota, two of the genetic markers were assessed by RT-PCR and restriction enzyme digestion (Fig. 3b). Absence of BssHII site and deletion of BbvCI restriction site was verified by RT-PCR, followed by restriction enzyme digestion, using LaSota strain as a control (see Fig. 3b). In addition, sequencing of these amplified DNA fragments were performed in order to confirm the existence of genetic markers, proving that rLS1 virus was certainly derived from pFLC-LS1 plasmid.

In order to obtain high expression levels of the eGFP, we inserted the eGFP cassette into the first position within the NDV genome. The pFLC-LS1-1eGFP construct was correctly assembled, as verified by sequencing of the entire plasmid, and used for transfection of DF-1 cells. Transfection supernatant containing the rLS1-1eGFP virus was passaged in SPF embryonated eggs. The collected AF contained a titer of 4.8 × 10⁸ PFU/mL, similar to the parental virus. Chicken embryos infected with the rLS1-1eGFP observed under a transilluminator showed high levels of eGFP expression in the umbilical cord (Fig. 4a). Replication in umbilical cord stem cells have also been reported for other viruses such as Influenza, hepatitis B and herpes simplex 1 viruses [34‐36]. For NDV, since the chicken umbilical cord is connected with the chorioallantoic membrane, it is possible that umbilical cord stem cells allowed the replication of the virus at high levels and then release them into the allantoic fluid.

Comparison of growth kinetics of the recovered viruses was performed in DF-1 cells at an MOI of 0.05. Both viruses exhibited similar growth characteristics. However, rLS1-1eGFP showed significantly lower titers than rLS1 (p < 0.01) at 24 and 36 h.p.i. (Fig. 4b). The highest titers achieved in cell culture for both viruses were approximately 10⁶ PFU/ml (Fig. 4b), whereas the titers in AF collected from SPF embryonated eggs were at least 10⁸ PFU/ml (data not shown).

The correlation between eGFP expression level and the efficiency of viral replication was assessed by infecting DF-1 cells at different MOIs with rLS1-eGFP, and recording these cells every 12 h. The intensity and number of cells expressing the eGFP were increasing in a time- and dose-dependent manner (Fig. 4c). The eGFP expression can be clearly visualized as early as 24 h.p.i., although it can be detected with low expression at 12 h.p.i. These results indicates that eGFP expression can be used to monitor viral replication.

Conventional- and eGFP-NT were first evaluated by using diluted reference chicken serum samples. Reference serum samples, positive to other pathogens (IBV, MDV, Avian adenovirus type 1 and ILTV) and from a SPF chicken, were used to evaluate non-specific neutralization. These sera tested negative, as expected, for both neutralization methods (See Additional file 1: Table S1). Fluorescence in the eGFP-NT was visible as early as 18 h of incubation, and it was clearly visible at 24 h.

To assess the efficacy of rLS1-1eGFP virus to measure nAb titers, the conventional- and eGFP-NT were performed using a total of 57 serum samples. Most of the samples presented the same neutralization titer when measured with both techniques. Moreover, the correlation between the conventional NT and eGFP-NT was very high (R ² = 0.994), as shown in Fig. 5a. Thus, we conclude that rLS1-1eGFP can be used as a reporter virus to determine nAbs titers with great reliability.

Correlations between nAbs titers and the antibody titers measured by HI and ELISA were evaluated. In Fig. 5b, the relation between log2 HI and log2 NT is shown. A strong correlation (R ² = 0.816) was observed between NT and HI titers. In addition, several samples that tested positive to the HI assay, with titers corresponding to dilutions of up to 1:8, were negative for neutralization. These results are consistent with the fact that the equation of the line (Y = 0.816X + 1.238) shows that when NT is 0, in average, HI titers would be positive and greater than a titer corresponding to a dilution of 1:2. It is worth mentioning that serum samples positive to HI and negative to the NT assay came from vaccinated chickens (see Additional file 1: Table S1), suggesting that either nAbs are not circulating at that time because plasma cells become memory B cells or specific antibodies against NDV exist, but these are not nAbs.

Antibody titers measured by ELISA (S/P) also have a strong correlation (R ² = 0.791) with the log₂ NT titers (Fig. 5c). Both HI and ELISA showed greater dispersion to the conventional NT than the eGFP-NT. These results indicate that HI and ELISA are not as reliable as eGFP-NT, showing that this method can close the gap between a rapid and accurate measurement of nAbs against NDV.