Cells, virus and vectors
HEK293T cells were purchased from the American Type Culture Collection (ATCC) (Manassas, VA, USA) and cultured in Dulbecco’s Modified Eagle’s Medium (Life Technologies Corp, USA) supplemented with 10% fetal bovine serum (FBS, Gibco USA) and penicillin/streptomycin at 37 °C in 5% CO2. The NDV strain (LaSota) viral stock was propagated in 9- to 11-day-old SPF chicken embryos . This representative strain belonging to Class II was attenuated one. The allantoic fluid with high HA titers was clarified by centrifugation at 6000 × g for 20 min at 4 °C, applied to 10-50% sucrose gradient, then centrifuged at 110,000 × g for 5 h for purification at 4 °C as described in a previous study . The purified NDV particles (1 mg mL-1) were suspended in phosphate buffer saline (PBS, pH 7.2) and were used as the coating antigen to screen the nanobodies. In addition, the NDV Class II virulent strain F48E9 (Genotype IX, GenBank accession number: MG456905), Class II virulent strain sx10 (GenBank accession number: KC853020), and Class I strain QH-1 (GenBank accession number: KT223818) were used to determine the ability of the developed assay for detecting different NDV strains. To construct the VHH library, the pMECS vector was kindly provided by Professor Muyldermans and used as described in the previous procedures . To express the nanobody and produce fenobody in the bacterial system, the expression vector pET-28a (Novagen, USA) was used. To produce RANbody, the pCMV-N1-HRP vector was constructed using the pCMV-N1-EGFP (Clontech, Japan) as a backbone based on previous descriptions .
Bactrian camel immunisation and library construction.
A healthy 4-year-old male Bactrian camel was immunized through the subcutaneous route with the commercial inactivated vaccine containing the inactive NDV (LaSota) strain (PULIKE biological engineering company, Luoyang, China). The Bactrian camel was injected a total of four times, which was performed at two-week intervals. The titration of the antibody against NDV in the serum samples was detected with an indirect ELISA based on a previous description , except the purified NDV particles (200 ng/well) were employed as the coating antigen. Five days after the last immunisation, the lymphocytes were isolated from the collected blood by Leucosep® tubes (Greiner Bio-One, Germany). Then, total cellular RNA was extracted, and the cDNA was synthesized using the oligo-dT12-18 primer with the SuperScript™ II Reverse Transcriptase based on the instructions. The VHH sequences from the cDNA pool were amplified in two rounds using nested PCR (Additional file 1: Table S1), and a VHH phage display library was constructed based on the procedures as previously described . After identification using the bacterial PCR with the primer pairs MP57 and GIII (Additional file 1: Table S1), the library was stored at -80 °C in LB medium supplemented with 20% w/v glycerol until used.
Screening of specific nanobodies against NDV
Phage rescue and biopanning were performed as described previously . In biopanning, 1 µg purified NDV particles (LaSota strain) were used as the coating antigen. After three rounds of biopanning, the enrichment of specific phage particles was evaluated with polyclonal phage ELISA. Then, 96 colonies were randomly picked, induced with isopropyl-β-D-thiogalactoside (IPTG, 1 mM), and their periplasmic extracts were tested by indirect ELISA for the presence of NDV-specific nanobodies. All positive clones contained VHH genes were sequenced and grouped according to their complementary determining regions 3 (CDR3) sequence. In addition, the binding activities of different nanobodies were evaluated with indirect ELISA using different dilutions of periplasmic extracts as the first antibody.
Selection of the NDV-specific nanobodies for producing fenobodies and RANbodies
To select the nanobodies for subsequently producing fenobodies and RANbodies, a capture ELISA was designed. Briefly, the ELISA plate was coated with the different periplasmic extracts containing the NDV-specific nanobodies and incubated 12-14 h at 4 °C. Each nanobody was coated into the two wells. Purified NDV particles (1 µg/well, positive control) and PBS (negative control) were separately added to the two wells of each nanobody after they were blocked and washed. Then, the monoclonal antibody against NDV (QIANXUN Biotech Company, Guangzhou, China) was added in 1:1000 dilutions and incubated for 1 h at 37 °C. After washed again, the HRP-goat anti-mouse IgG antibody (TransGen Biotech Company, Beijing, China) was added, and the reaction was colored with tetramethylbenzidine (TMB) [A: 205 mmol L-1 potassium cirate (pH 4.0); B: 41 mmol L-1 tetramethyl benzidine; A:B (v/v) = 39:1]. After the reaction was stopped using 3 mol L-1 H2SO4, the optical density at 450 nm (OD450nm value) was read using an automatic ELISA plate reader. Then, the coated nanobodies emerging from the positive OD450nm value/negative OD450nm value (P/N) > 3.0 were selected for producing fenobodies and RANbodies.
Preparation of fenobodies against NDV
The fenobodies were expressed according to the previous descriptions with some modifications . The complete gene encoding ferritin based on the sequences of P. furiosus was provided by GENWIZ Biotech Company. According to the above obtained sequences encoding the nanobody against NDV and ferritin, the two primer pairs (Additional file 1: Table S1) were designed. One pair was used to amplify the selected VHH genes, and the other was utilized to amplify the truncated gene encoding from the 1 to 146 amino acid (aa) region of ferritin. Meanwhile, some overlapping gene sequences were designed in the reverse and forward primers to amplify the nanobodies and ferritin, respectively (Additional file 1: Table S1). Then, overlapping PCR was used to amplify the fusion gene, in which nanobodies replaced helix ε and loop (147-174 aa) of ferritin by GS linker (GGGS)3. Subsequently, to construct the recombinant plasmid (pET-28a-fenobody), the fusion genes were digested with the enzymes Nde I and BamH I (TaKaRa, Japan) and then were ligated into the pET-28a vector digested with the same two enzymes. After identification, the fenobody was expressed in the transformed Transetta (DE3) E. coli (TransGen Biotech, China) by adding 0.1 mmol L-1 IPTG at 16 °C. Then, the pellets were resuspended in lysis buffer (NaH2PO4·2H2O 7.80 g, 50 mmol L-1, NaCl 29.22g, 500 mmol L-1, imidazole 0.68 g, 10 mmol L-1, pH 8.0). After sonication on ice, the soluble fenobody with His tags was loaded onto a Ni-NTA-6FF Column (Smart-Lifesciences, Changzhou, China). In addition, an anti-H9N2 nanobody was designed to produce fenobody as the negative control. After purification, SDS-PAGE and Western blot assays were employed to analyze the expression and purification of the fenobody, while the purified fenobodies were frozen at -20 °C in 1 mmol PMSF and 0.02% w/v NaN3. In addition, the fenobodies were negatively dyed with uranium acetate and observed at TEM (JEM-1400) whether fenobody self-assembled into 24 subunit nanocage. To verify the fenobody binding to the NDV particles, indirect ELISA was used with the purified NDV particles as the coated antigen. The procedure for the indirect ELISA is characterized below.
Affinity and half-life extension test of fenobody
To compare the affinity with the NDV particles and half-life extension of the fenobody with univalent or traditional nanobody, traditional nanobodies were expressed and purified based on a previous description . Briefly, the VHH genes encoding the different nanobodies were directly ligated into the pET-28a vector. The nanobodies were expressed in E.coli BL21 (DE3) and purified with the Ni-NTA resin according to the manufacturer’s instruction. Then, the purified traditional nanobodies were analyzed with SDS-PAGE.
To compare the affinity of the fenobody and traditional nanobody, a capture ELISA were designed and performed based on a previous characterisation with some modifications [24,30]. Briefly, after the same amounts of fenobodies and traditional nanobodies were coated into the ELISA plates, NDV particles (1 µg/well), anti-NDV monoclonal antibodies, and HRP labeled goat anti-mouse antibodies were then added to the plates one by one. To accurately calculate the parameters of the binding reaction using GraphPad Prism 5 software, the above capture ELISA first employed the fenobody (used in the developed sandwich ELISA) and the corresponding nanobody as the coating antigens with the amount of 800 ng/well. Then, the same reagents were added into the wells, except that 100 µl allantoic fluid containing NDV with HA titers (212 to 20) was used instead of NDV particles (1 µg/well).
To analyze the half-life extension, the fenobody and traditional nanobody were both labeled with FITC based on the descriptions by Fan K et al.  and Fisher et al . The FITC-labeled fenobody and FITC-labeled traditional nanobody were intravenously injected into female BALB/c mice via the tail vein. Then, the blood was collected from the tail vein at different time points, including 20, 30, 40, 60, 120, 240, 360, 720, 1200, and 1440 min. The fluorescence of the FITC-protein in the blood was determined on VICTOR™ X Series Multilabel Plate Readers (PerkinElmer, USA). Then, the values were analyzed using the Origin software.
Preparation of RANbodies against NDV
To produce RANbodies against NDV, the method described by Sheng at al was followed using HRP as the reporter . Briefly, the modified vector pCMV-N1-HRP and positive phagemid pMECS containing the genes encoding nanobodies were digested with both Pst I and Not I enzymes. Then, the nanobody genes were ligated into the vector pCMV-N1-HRP using a DNA Ligation Kit according the manufacturer’s instructions (TaKaRa, Japan). The positive recombinant plasmids were confirmed by sequencing and used for transfection. Then, the HEK293T cells were transfected with the positive plasmids. After 72 h of transfection, the medium containing the secreted RANbodies was harvested, and supplemented with 0.02% w/v NaN3 for direct use. In addition, to avoid the exogenous pollution, a nanobody against H9N2 was also selected to produce RANbody as the negative control.
The expressions of RANbodies in the HEK293T cells and in the medium were determined using indirect immunofluorescence assay (IFA) and Western blot assay, respectively. The two assays both employed anti-His monoclonal antibody as the first antibody. The FITC and HRP-labeled goat anti-mouse antibodies were separately used as the second antibodies for the IFA and Western blot assay, respectively. In addition, the specific binding with NDV and titers of RANbodies in the medium were identified by direct ELISA using purified NDV particles as the coated antigen.
Indirect ELISA was used to analyze the specific binding of periplasmic extracts from 96 clonal E.coli, fenobodies, and RANbodies with NDV. Briefly, the 96-well Maxisorp microtiter plates (Nucn-Immunoplate, Roskilde, Denmark) were coated with the purified NDV particles (LaSota strain, 200 ng/well) and incubated at 4 °C overnight. The purified H9N2 avian influenza virus (AIV) particles were used as the negative control. After three more washings with PBS containing 0.5% w/v Tween-20 (PBS’T), the periplasmic extracts, fenobodies, and RANbodies were separately added into the wells and incubated for 1 h at RT. After washing three times, anti-His tag monoclonal antibodies and HRP-goat anti-mice IgG antibodies were subsequently added into the wells containing periplasmic extracts and fenobodies. After another three washes, TMB (100 μL/well) was added and incubated in the dark for 15 min at RT. For the RANbodies (nanobodies fusion with HRP), the TMB was directly added into the wells without the first and second antibodies to initiate a color reaction. The color reaction was stopped with 3 mol L-1 H2SO4 (50 μL/well), and the OD450nm values were read using an automatic ELISA plate reader (BIO-RAD).
Development of the fenobody and RANbody-based sandwich ELISA
To develop the sandwich ELISA, the fenobodies and RANbodies were separately used as the capture and detection reagents, respectively. To obtain the best pair, the ELISA plate was coated with the 800 ng/well of different fenobodies and incubated at 4 °C overnight. After the plate was washed three times with PBS’T, the purified NDV particles (LaSota strain, 1 µg/well), as the positive (P) control, and H9N2 AIV, as the negative (N) control, were separately added for 1 h after blocking with 5% w/v skim milk in PBS’T for 1 h. After washed again, 100 µL of different RANbodies was added into the well and incubated for 1 h at RT. After a final washing, TMB was added to produce the color reaction, and the OD450nm values were read after the reaction was stopped with 3 mol L-1 H2SO4. The best pair of nanobodies was selected at the highest P/N value.
Secondly, the optimal amounts of capture fenobody and detecting RANbody for the sandwich ELISA were determined using a checkerboard titration. Different amounts of the fenobody (100, 200, 400, 800 and 1000 ng/well) and different dilution ratios of RANbodies (1:1, 1:10, 1:100, 1:1000, and 1:10,000) were used in the sandwich ELISA. Same amounts of purified NDV particles (1 µg/well) and H9N2 AIV were used. Then, the optimal amounts of fenobody and RANbody were determined when the P/N value was the highest.
Further, the incubation times between fenobody and NDV and between NDV and RANbody were optimized. The incubation time of fenobody capturing NDV was tested for 20, 40, 60, 80, and 100 min, while the incubation time of RANbody and NDV was examined for 30, 60, and 90 min. Using H9N2 AIV as the negative control and checkerboard titration, when the P/N ratio was highest, the two optimal times were determined.
Validation of the developed sandwich ELISA
To determine the cut-off value of the sandwich ELISA, 150 negative samples, including tracheal and cloacal swabs (n = 45), allantoic fluid (n = 25), cell culture medium (n = 25), chicken sera (n = 25) and tissue samples (n = 30), were tested. Generally, NDV is detected from these above samples in clinical trials. The tracheal and cloacal swabs were washed with 20-50 µL PBS, then tissue samples were grinded, freeze-thawed three times, and centrifuged using the supernatant. The cut-off value was the mean of the OD450nm values of 150 negative samples + 3 times standard deviations (SD).Different amounts (0 to 1000 ng) of the purified NDV particles and different HA titers (28 to 20) of allantoic fluid containing NDV were both detected with the sandwich ELISA to determine the limited viral amount of the assay. Based on the OD450nm values of sandwich ELISA and HA titres of allantoic fluid, the correlation between sandwich ELISA and HA test was calculated using Microsoft Excel. The correlation between the sandwich ELISA and amounts of purified NDV particles was also calculated based on the OD450nm values and different amounts of NDV.
To determine the specificity of the sandwich ELISA, other poultry disease viruses, including H9N2 AIV, H5N1 AIV, H7N9 AIV, infectious bronchitis virus (IBV), infectious bursal disease virus (IBDV), fowl adenovirus (FADV), and J subgroups of avian leukosis virus (ALV-J) were also detected using the assay.
In addition, to determine if the developed sandwich ELISA can be used to detect different NDV strains, reference strains F48E9 (Class II virulent strain), QH-1 (Class I strain), and clinical isolate sx10 (Class II virulent strain) were selected. Meanwhile, to evaluate whether the dead NDV was detected by the developed sandwich ELSIA or not, the different NDV strains were inactivated with 5% formaldehyde at 4 °C for 7 days. Then, the inactivated NDV strains were detected with the developed sandwich ELISA.
Application of the developed sandwich ELISA for detecting NDV in chicken tissue samples
To verify that the developed sandwich ELISA can be used to detect NDV from different tissue samples, an animal experiment was designed, and different tissue samples were collected from the chickens infected with NDV.
Twenty 30-day-old specific-pathogen-free (SPF) chickens were randomly divided into two groups. One group (n = 10) was infected with NDV strain F48E9 stock containing 25 HA titers using the nasal route. The second group was inoculated with PBS as the negative control. After challenged, two chickens each were necropsied at 3, 4, 5, 7, and 10 days post inoculation (dpi). A total of 340 swab samples (170 for each group) were collected from the thymus, pancreas, proventriculus, liver, spleen, kidney, small intestine, large intestine, cecal tonsil, feces, brain, trachea, lung, throat swab, tracheal swab, bursa, and cloacal tissues. Then, the same amounts of collected tissue samples were grinded. After freeze-thawing three times, 100 µL suspensions of each tissue sample were used to detect NDV using the developed sandwich ELISA.
Comparisons of the developed sandwich ELISA with other commercial methods
As described above, all 340 tissue samples from the animal experiment and four different NDV strains (LaSota, F48E9, sx10, and QH-1) were further used to detect NDV with a commercial monoclonal antibody-based sandwich ELISA kit (Yoyong Biotechnology Company, Guangzhou, Chain) and a commercial immune colloidal gold strip (Yoyong Biotechnology Company, Guangzhou, Chain). The coincidence rates of the developed sandwich ELISA with the monoclonal antibody-based sandwich ELISA kit and immune colloidal gold strip were calculated using Microsoft Excel’s CORREL function. In addition, the 367 samples from the clinical chickens, including 189 tracheal and cloacal swabs and 178 tissue samples, were also tested with the three assays, and the coincidence rates were calculated.
In addition, the above positive samples detected by the developed sandwich ELISA were used to inoculate the SPF chicken embryos by allantoic cavity route. Then, the allantoic fluid samples were collected and detected using the HA test. The coincidence rates of the two assays were also calculated using the Microsoft Excel’s CORREL function.
All experiments were repeated at least three times. Kappa index values were calculated to estimate the coincidence between the developed sandwich ELISA and the monoclonal antibody-based commercial ELISA kit, commercial immune colloidal gold strip, and HA test. These calculations were performed using SPSS software (Version 20, http://www.spss.com.cn).