Epitope repeat protein (ERP) gene encoding eight repeats of an IDE sequence containing ETQFLDLMRAVANSMR (C-terminal IDE aa 444–459) on NDV NP separated by tetra-glycine linker was synthesized commercially using codon optimization for baculovirus expression. The recognition enzyme sites of EcoRI and HindIII were built in the upstream and downstream of coding sequence, respectively. The synthetic gene cloned in vector backbone PMK-RQ (KanR) was received as 5 µg of lyophilized plasmid preparation. The conceptual translated sequence was submitted to I-TASSER server for homology modeling [17] and top five predicted structures were visualized with PyMOL software [18].

Synthetic gene was subcloned into pFastBac HT b vector using EcoRI and HindIII restriction sites. The recombinant plasmid (50 ng) was transformed in Escherichia coli DH B10 cells. Transformed cells were plated onto the Luria-Bertani agar containing kanamycin (50 µg/mL), gentamicin (7 µg/mL), tetracycline (10 µg/mL), Bluo-gal (100 µg/mL), and isopropylthio-β-galactoside (IPTG, 40 µg/mL) and incubated at 37℃ for 36 to 48 hours. The high molecular weight bacmid DNA was isolated from the overnight cultures by alkaline lysis purification according to the manufacturer's manual of Bac-to-Bac baculovirus expression system (Invitrogen, Carlsbad, CA, USA).

Spodoptera frugiperda 9 (Sf9) cells were cultured at 27℃ in Sf-900 II serum free medium (Invitrogen), supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Invitrogen), 50 U/mL penicillin and 50 µg/mL streptomycin. Sf9 cells were transfected with recombinant bacmid DNA using Cellfectin II, a cationic lipid for the transfection of the baculovirus particles according to the manufacturer's instructions. Briefly, for each transfection, 2 mL of Grace's Insect Medium, unsupplemented (without antibiotics and serum) was added in each well, 8×10⁵ cells per well were seeded in a 6-well plate and allowed to attach for 2 hours. The bacmid DNA and Cellfectin II (8 µL of reagent) were diluted separately in 100 µL of Grace's medium, unsupplemented (without antibiotics and serum), then mixed and incubated for 30 minutes at room temperature to form lipid-DNA complexes. The cells were washed with fresh medium, and incubated with lipid-DNA complex at 27℃ for 5 hours. The transfection solution was removed and 2 mL supplemented Sf-900 II SFM containing 10% FBS was added. Transfected Sf9 cells were incubated at 27℃ for 72 hours for baculovirus production. Recombinant baculovirus production was monitored daily by visualization of the cytopathic effects. Three to four days after transfection, recombinant baculovirus was harvested from the cell culture medium and stored at 4℃. Recombinant viruses were identified by polymerase chain reaction using gene vector specific primers (Invitrogen). The resulting baculovirus was passaged three times by infecting more Sf9 cells.

SF9 suspension cultures (2×10⁶ cells/mL) grown in 2 L of Sf900 II medium supplemented with 1% of FBS were infected with recombinant baculovirus at multiplicity of infection of 5 plaque-forming unit/cell. Cells were grown for 72 hours and cell pellets were washed in phosphate buffered saline (PBS). As recombinant ERP (rERP) carried 6× His tag at N-terminal, it was purified by using Ni-NTA spin columns under denaturant condition with 8 M urea according to Ni-NTA spin handbook (Qiagen, Hilden, Germany). The collected cells were lysed in 10 mL buffer B (10 mM NaH₂PO₄, 300 mM NaCl, 8 M urea, pH 8.0) and centrifuged the lysate at 10,000×g for 30 minutes and collecting supernatant. Loading up to 600 µL of the cleared lysate supernatant containing the 6× His-tagged protein onto a pre-equilibrated Ni-NTA spin column, centrifuged 2 minutes at 700×g and washing the column with 600 µL buffer C (10 mM NaH₂PO₄, 300 mM NaCl, 8 M urea, pH 6.3) and centrifuged 2 minutes at 700×g, the recombinant protein was finally eluted with buffer E (10 mM NaH₂PO₄, 300 mM NaCl, 8 M urea, pH 4.3). The recombinant protein was allowed to refold using refolding buffer (50 mM Tris pH 7.5, 0.5 M NaCl, 0.3% CHAPS, 1 mM DTT, 5% glycerol). rERP was dialyzed against refolding buffer (40 mM CHAPS and 10 mM DTT).

Proteins extracted from infected Sf9 cells were fractionated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis under reducing conditions. The separated proteins were blotted onto an Immun-Blot polyvinylidene fluoride membrane (Bio-Rad, Richmond, CA, USA) using a wet transfer system (Bio-Rad). The membrane was blocked with 5% skim milk in PBS containing 0.1% Tween 20 (PBST) at room temperature for 1 hour. After washing 3 times with PBST, the blocked membrane was subsequently incubated with NDV chicken antiserum (diluted 1:100) for 1 hour, rinsed in PBST, and incubated with alkaline phosphatase-conjugated goat anti-chicken immunoglobulin (diluted 1:1,000, Pierce, Rockford, IL, USA) for 1 hour. Protein bands were visualized by enhanced chemiluminescence using NBT/BCIP solution (Sigma-Aldrich, St. Louis, MO, USA).

Optimal dilutions of rERP and sera were determined by a checkerboard titration test with NDV positive and negative sera previously confirmed by hemagglutionation inhibition (HI) test. The antigen was coated in 96-well ELISA plates (PolySorb, Nunc, Roskilde, Denmark) ranging in concentration from 2 µg/mL to 0.0625 µg/mL in 50 mM carbonate/bicarbonate buffer (pH 9.6). Reference positive and negative sera were both diluted serially from 1:100–1:6,400 and tested to determine the optimal serum dilution. The dilutions that gave the optimum difference in absorbance at 450 nm between positive and negative sera were selected to test the sera panel. The working dilution of goat anti-chicken HRP-IgG (Sigma-Aldrich), the reaction temperature, time and other conditions also were optimized.

Ten NDV negative serum samples from specific pathogen free (SPF) chickens were used to determine a cutoff value for the ELISA assay. The mean negative serum optical density (OD) value plus three standard deviations (SD) was used as the cutoff value. Each sample was repeated in triplicate wells and the mean value was calculated. The experiment was repeated two more times. In addition to SPF chicken serum, different hyper-immune chicken sera to NDV, H9N2 avian influenza virus (AIV), infectious bronchitis virus (IBV), and infectious bursal disease virus (IBDV) were tested for specificity of rERP using an indirect ELISA assay. To test the sensitivity of the rERP, the chicken anti-NDV serum was diluted serially (1:100–1:6,400) and the serum with each dilution was used to react with the expressed rERP in an indirect ELISA.

A total of 30 chicken sera comprising three groups with each group of 10 were used in the study. These sera were kept at the OIE reference laboratory for Newcastle disease, Animal and Plant Quarantine Agency (APQA), Korea. First group were sera taken from mock-infected SPF chicken (NDV antibody negative). Second and third groups were sera taken 14 dpi from SPF chickens vaccinated with IDE deleted recombinant Newcastle disease virus (rNDV) [4] and NDV LaSota strain [19], respectively, which kindly supplied by the OIE reference laboratory for ND, Korea. All vaccinated sera were proved to be serologically positive by a HI test using NDV antigen. All sera were available with the laboratory. Microtiter plates were coated with 100 µL of 1 µg/mL rERP protein in 50 mM carbonate/bicarbonate buffer pH 9.6, and incubated overnight at 4℃. Plates were washed three times with PBST and blocked with 150 µL per well PBST with 5% skim milk powder solution at 37℃ for 2 hours. The chicken anti-NDV serum (1:400) was added to each. After 1-hour incubation at 37℃, plates were washed three times with PBST. One hundred microliters horseradish peroxidase-conjugated goat anti-chicken IgG (Pierce) at 1:2,000 dilution was added to each well and plates were incubated at 37℃ for 1 hour. Then plates were washed three times with PBST and 100 µL TMB (3,3,5,5-tetramethylbenzidine) was added to each well. After 10-minute incubation at 37℃, reactions were stopped with 2 M H₂SO₄ and the OD at 450 nm was measured with microplate reader (Sunrise, Tekan, Männedorf, Switzerland).

Mean ELISA titer and SDs of each group were calculated using Microsoft Excel (Microsoft, Redmond, WA, USA). Differences in ELISA absorbance between groups were analyzed by the one-way analysis of variance (ANOVA). A p-value of <0.05 was considered statistically significant.

The construct encoding ERP on NDV NP was designed as shown in Fig. 1A. The codon optimized gene (486 bp) was designed to have 8 repeats of 16 amino acids (ETQFLDLMRAVANSMR) mimicking C-terminal IDE (aa 444–459) on NDV NP. rERP gene insert was excised from the plasmid after digestion with EcoRI and HindIII enzymes. Resulting fragment (486 bp) was subcloned into the pFastBac HT b vector using above mentioned restriction sites. Recombinant baculovirus containing the recombined ERP gene insert (bac-rERP) was generated by transposition of recombinant vector into the SF9 cells by transfection. The codon optimized gene was commercially synthesized from GeneArt (Regensburg, Germany) and subcloned into baculovirus transfer vector. Alignment of obtained nucleotide and deduced amino acid sequences of the rERP gene insert revealed 100% match with the designed construct (data not shown). I-TASSER server predicted 5 models, all of which showed optimal spacing in epitopes separated by linkers (Fig. 1B). Graphic visualization of tertiary structure of top predicted models for rERP protein suggests that IDE reactive antibodies would be freely accessible to epitopes separated by tetra-glycine linkers in tertiary dimensional space.

Bac-to-Bac expression system was used to produce a baculovirus encoding N-terminally 6× His tagged rERP protein under the transcriptional control of polyhedrin promoter. Recombinant baculovirus bac-rERP infected cells were lysed under denaturing conditions and rERP was purified using His tag purification. The protein was allowed to refold using refolding buffer. To identify the expressed rERP, the purified protein was analyzed by western blot assay. The chicken anti-NDV serum raised by infecting SPF chickens with La Sota strain of NDV [19] was used as the primary antibody in the analysis. The results showed purified rERP with molecular weight of approximately 20 kDa reacted specifically (lanes 1 to 6 in Fig. 2) to the NDV positive serum. No specific band was detected with negative controls (lanes N1 and N2 in Fig. 2), which indicated that the rERP is expressed correctly, and has good reaction ability with chicken specific anti-NDV serum.

The optimal concentration of the rERP (1 µg/mL) and the dilution rate of sera (1:400) for indirect ELISA assays to evaluate the specificity and sensitivity of the expresses rERP were determined. To set up a cutoff value for indirect ELISA assay, 10 SPF chicken serum samples were analyzed. The mean OD₄₅₀ value of these samples as measured by ELISA was 0.231 with a SD of 0.00022. For a 99% confidence interval, the cutoff value determined was 0.232 (mean OD of SPF sera samples+3 SD). Based on the cutoff value, the reactivity of rERP with chicken anti-AIV, anti-IBV, anti-IBDV, and normal sera was measured. The result showed that OD450 nm values of non-NDV sera were lower than cutoff value (Fig. 3A). The minimum dilution titer of chicken anti-NDV serum detected was 1:1,600 according to cutoff value of 0.232 (Fig. 3B).

A battery of 30 chicken sera samples divided in ten chickens per group in three was tested. Samples in groups of mock infected chicken sera and chicken vaccinated with the IDE deleted rNDV showed OD values below cut-off value. The OD value for samples from wildtype NDV LaSota vaccinated chicken group were in the range of 0.255 to 1.77, which was above determined cut-off value (Fig. 4). The indirect ELISA showed good repeatability through the multiple experiments.