Detection of REBOV-NP and -GP antibodies in swine using IFA

Swine sera were analyzed for the presence of REBOV-NP and -GP antibodies in IFA specific to REBOV-NP and -GP, respectively. In the IFA, none of the 49 swine sera collected in Japan showed a positive reaction (data not shown), and so they were considered to be REBOV-NP and -GP antibody negative. In the IFA specific to REBOV-NP, antibody positive swine sera showed characteristic granular staining patterns in the cytoplasm (Figure 1A), which were indistinguishable from those of REBOV-infected cynomolgus monkey sera [18] and REBOV-NP immunized rabbit sera (data not shown). Antibody-negative swine sera showed no reaction (Figure 1B). In the IFA specific to REBOV-GP, antibody positive swine sera showed characteristic cellular surface staining patterns (Figure 1C), which were indistinguishable from those of REBOV-infected cynomolgus monkey sera and REBOV-GP immunized rabbit sera (data not shown). Antibody-negative swine sera showed no reaction (Figure 1D). In Bulacan, 104 (71.2%) and 115 (78.8%) of the 146 swine sera showed positive reactions in the NP- and GP-specific IFA, respectively. In Pangasinan, each 54 (78.3%) of the 69 swine sera showed positive reactions in the NP- and GP-specific IFA. In total, 158 (73.5%) and 169 (78.6%) of the 215 swine sera collected at the affected farms were REBOV-NP and -GP antibody positive in the IFA, respectively (Table 1). On the other hand, none of the 98 swine sera collected in Tarlac in the Philippines showed any positive reaction (Table 1).

Detection of REBOV neutralizing antibodies in swine

The VSV-pseudotype bearing REBOV-GP efficiently infected Vero E6 cells which are known to be susceptible to REBOV infection. The infectious titer of the VSV-pseudotype reached 3.6 x 10⁶ IU/mL when measured on Vero E6 cells. Infection with the VSV-pseudotype was neutralized with rabbit serum to REBOV-GP (data not shown). Swine sera were analyzed in the NT using the VSV-pseudotype. In the NT, none of the 49 swine sera collected in Japan neutralized the infection with the VSV-pseudotype on Vero E6 cells at serum dilutions of 1 in 100. These were therefore considered to be REBOV NT antibody negative. In the affected farms, 108 (74.0%) of the 146 swine sera in Bulacan and 46 (66.7%) of the 69 swine sera in Pangasinan showed positive reactions in the NT, respectively at a serum dilution of 1 in 100. In total, 154 (71.6%) of the 215 swine sera in the affected farms showed NT antibody positive (Table 1). NT titers of the sera were then obtained for the 34 NT antibody positive sera in the affected farms. NT titers ranged between 100 and 12,800 with average of 790 and median of 400 (data not shown). In Tarlac, none of the 98 swine sera showed any positive reaction (Table 1).

Detection of REBOV-NP and -GP specific antibodies in swine using IgG-ELISA

IgG antibodies to the REBOV-NP and -GP were detected by recombinant REBOV protein-based IgG-ELISA. The OD values to the NP and GP of each serum sample at a dilution of 1 in 100 were plotted (Figure 2A and B), and ROC and TG-ROC curves were also drawn (Figure 2C to F). Since the OD values of negative samples were very low, the cut off values defined as the values at intersection points of the sensitivity curve and specificity curve were 0.077 and 0.104 for REBOV-NP and -GP, respectively. At the cut off values, sensitivity and specificity of the assay are over 95% in both IgG-ELISAs (Figure 2E and F). In Bulacan, 115 (78.8%) and 119 (81.5%) of the 146 swine sera showed positive reactions in the NP- and GP-specific IgG-ELISA, respectively. In Pangasinan, 62 (89.9%) and 46 (66.7%) of the 69 swine sera showed positive reactions in the NP- and GP-specific IgG-ELISA, respectively. In total, 177 (82.3%) and 165 (76.7%) of the 215 swine sera collected at the affected farms were REBOV-NP and -GP antibody positive in the IgG-ELISA, respectively (Table 1). Conversely, 1 of the 98 swine sera collected in Tarlac in the Philippines showed a positive reaction in the GP specific IgG-ELISA and 1 of the 49 swine sera collected in Japan showed a positive reaction in the NP- and GP-specific IgG-ELISA (Table 1). However, these results are considered to be false positives because the OD values were close to the cut off level defined by the assay, additionally all the samples collected in Tarlac and Japan tested using the alternative serological assays were negative.

Agreement between the serological assays detecting REBOV antibodies

We next examined the agreement between the serological assays used in the present study. The data obtained in each assay were binarized to either positive or negative result. These binarized data were then statistically analyzed by pair-wise comparisons in nonparametric one-way ANOVA. This analysis showed that there were no significant differences between the serological assays (p > 0.05, data not shown). The result supports the contention that all of the five serological assays used in the present study posses similar discriminatory capacity. Thus we are confident that many REBOV antibody positive samples showed a positive reaction in all of the five assays.

Swine serum specimens

A total of 215 swine sera were collected from two REBOV affected pig farms, located in Pangasinan and Bulacan in 2008 (Figure 3). Of these, 146 sera were collected from swine at the farm in Bulacan, and 69 were collected from those in Pangasinan. Swine samples in the affected farms were collected under quarantine of the Philippines. The sera were kept frozen at the Research Institute for Tropical Medicine (RITM) in the Philippines until use. Ninety-eight swine sera were collected from July to September 2010 from swine aged between 2 and 20 months (median of 4.5 months) in Tarlac in the Philippines, where no swine epizootic has been documented. The swine specimens at Tarlac were collected and used under approval of IRB (No. 2009-018) of RITM, and informed consent was obtained from the farm owners. Forty-nine swine sera collected at slaughterhouse in 2006-7 in Japan, kindly supplied by Dr. T-C Li at the National Institute of Infectious Diseases, were used as REBOV antibody negative control sera. Swine sera were inactivated at 56°C for 30 minutes, and then the gamma globulin fractions were precipitated with ammonium sulfate and reconstituted in phosphate buffered saline (PBS).

Immunofluorescence assay (IFA) specific to REBOV-NP and -GP using HeLa cells expressing the recombinant protein

The entire cDNA of REBOV-GP ORF was amplified from the swine lymph node specimen used for amplification of the cDNA of NP by RT-PCR using the primers REBOV-GP/F (5’-CGA AGC TTC GAA CAT GGG GTC AGG ATA TCA ACT-3’) and REBOV-GP/R (5’-CGA AGC TTC AAC ACA AAA TCT TAC ATA TAC AAA G-3’) (the Hind III site is underlined). The complete nucleotide sequence of the amplicon was determined to be identical to that of Reston 08A (GenBank accession number FJ621583), indicating that the specimen was collected from the same farm as that of sample group A from which Reston 08A was isolated [17]. The Hind III fragment of the amplicon was subsequently cloned into pKS336 [28], and the pKS336 plasmid with REBOV-GP, pKS336-pREBOV-GP, was used for transfection. HeLa cells were transfected with pKS336-pREBOV-GP using a FuGENE HD (Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s instructions. The cells transfected with the plasmids were selected with 3 μg/mL of blasticidin S-hydrochloride (Invitrogen, CA, USA) in Dulbecco’s modified Eagle medium (DMEM) supplemented with 5% fetal calf serum (FCS). The selected cells were analyzed for the expression of REBOV-GP in IFA using anti-REBOV-GP rabbit serum. The cells expressing REBOV-GP were spotted on multiwell slide glasses, fixed in acetone, and used as antigens for IFA specific to REBOV-GP. The IFA specific to REBOV-NP was performed as described previously [18], with slight modifications. Mock-transfected HeLa cells were used as negative control antigens in the IFA. The antigen slides were incubated with serially diluted swine sera under humidified conditions at 37°C for 1 hour. The antigen slides were washed in PBS, then reacted with rabbit anti-pig IgG conjugated with fluorescein isothiocianate (FITC) (Bethyl, TX, USA) at a dilution of 1 in 100 at 37°C for 1 hour. The slides were washed in PBS, covered with cover glasses, and examined for staining patterns under a fluorescence microscope fitted with appropriate barrier and excitation filters for FITC visualization. The antibody titer in the IFA was determined as the reciprocal of the highest dilution showing positive staining.

Generation of vesicular stomatitis virus (VSV)-pseudotype bearing REBOV-GP

Generation of VSV-pseudotype bearing REBOV-GP was performed as described previously [29–31]. Briefly, 293 T cells were transfected with pKS336-pREBOV-GP and cultured for 24 hours, and then the cells were infected with VSVΔG* (kindly supplied by Prof. M.A. Whitt, University of Tennessee) [30, 31] and cultured for 24 hours. The culture supernatants were then collected, filtered through a 0.22 μm-pore-size filter, and stored at -80°C until use. The infectious titer (infectious unit, IU) of VSV-pseudotype on Vero E6 cells was determined by counting the number of Green Fluorescent Protein (GFP)-expressing cells.

Neutralization test (NT)

The sera were diluted twofold from 1 in 100 with DMEM containing 5% FCS and 1,000 IU of VSV-pseudotype bearing REBOV-GP. The mixture was incubated for 1 hour at 37°C, then inoculated onto Vero E6 cells seeded on 96-well plates. The numbers of VSV-pseudotype infected cells were determined by counting of the number of GFP-positive cells according to the methods described previously [23, 29, 30]. NT titers of the tested sera were defined as the reciprocals of the highest dilutions at which more than 50% inhibition of infectivity was observed. Serum samples were considered to be NT antibody negative when less than 50% inhibition of infectivity was observed at a dilution of 1 in 100.

Recombinant baculoviruses expressing recombinant REBOV proteins

cDNA of REBOV nucleoprotein (NP) open reading frame (ORF) from a swine lymph node specimen collected at the farm in Bulacan was amplified by reverse transcription polymerase chain reaction (RT-PCR), and the nucleotide sequence of the amplicon was determined to be identical to that of Reston 08A. The cDNA was subcloned into pGEM-Teasy (Promega, WI, USA) and used for the following PCR template. PCR was performed using the plasmid clone to add BamHI linker with primers REBOV-NP/F (5’-GGG CTA GCG GAT CCA AGT CGA TAT GGA TCG TGG GAC C-3’) and REBOV-NP/R (5’-TTG CGG CCG CGG ATC CCT GAT GGT GCT GCA AGA TTG-3’) (the restriction site is underlined). A BamHI-digested fragment of the plasmid was then subcloned into pAcYM1-C-His plasmid, a derivative of pAcYM1 plasmid [32] carrying eight histidine coding sequences just downstream of the BamHI site, to construct pAcYM1-His-pREBOV-NP. A recombinant baculovirus, Ac-His-pREBOV-NP, was generated using a previously described method [33]. The recombinant baculovirus, which expresses the ectodomain of REBOV-glycoprotein (GP) with a histidine-tag at its carboxyl terminus [22], was used to prepare recombinant REBOV-GP for IgG-ELISA. A baculovirus (Ac-ΔP) that lacks the polyhedorin gene was used to prepare a negative-control antigen in insect cells.

Expression and purification of recombinant NP and GP of REBOV in baculovirus

Tn5 insect cells were infected with Ac-His-pREBOV-NP and Ac-His-REBOV-GP and then incubated at 26°C for 72 hours and 48 hours, respectively. The cells were washed in PBS, lyzed in PBS containing 1% NP40 and 8 M urea, and then clarified by centrifugation at 8,000 rpm for 10 min. The supernatant fraction was collected, and recombinant REBOV-NP and REBOV-GP were purified using a Ni2 + -resin purification system (QIAGEN, Düsseldorf, Germany) according to the manufacturer’s instructions. The expression and purification of REBOV-NP and REBOV-GP were confirmed in 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie Brilliant Blue R-250 staining Tn5 cells infected with Ac-ΔP were processed similarly and used as negative control antigen in the ELISAs. The purified antigens and negative control antigen were kept at -80°C until use.

IgG-ELISA

Swine sera were analyzed for the presence of antibodies to REBOV-NP and REBOV-GP in IgG-ELISA, essentially as described previously [22, 34, 35]. Briefly, half of the wells of the ELISA plates (Falcon, NJ, USA) were coated with a predetermined optimal quantity (approximately 100 ng/well) of each purified recombinant protein, and the rest of the wells were coated with the negative control antigens. After washing the plates three times in PBS containing 0.05% Tween-20 (T-PBS; SIGMA-ALDRICH, MO, USA), each well was incubated with 200 μL of T-PBS containing 5% skim milk (MT-PBS; Yukijirushi, Hokkaido, Japan) for 1 hour at 37°C. After washing the plates three times in T-PBS, the antigen-coated and negative-control-antigen–coated wells were inoculated with the test samples (100 μL/well) in MT-PBS at a dilution of 1 in 100, incubated for 1 hour at 37°C, and washed in T-PBS. Then, each well was incubated with a Protein A/G conjugated with horseradish peroxidase at a dilution of 1 in 1,500 (PIERCE, IL, USA) in MT-PBS for 1 hour at 37°C. The plates were washed three times, and 100 μL of ABTS solution (Roche Diagnostics, Mannheim, Germany) was added to each well. The plates were incubated for 30 min, and the optical density (OD) was measured at 405 nm with a reference at 490 nm. The adjusted ODs for each tested sample were calculated by subtracting the ODs of the negative-control-antigen–coated wells from those of the corresponding REBOV-antigen–coated wells.

Statistical methods

Sensitivity, specificity and predictive values for positive and negative tests were calculated by standard methods. Sensitivity and specificity were defined as the probability that the target assay result was positive when the IFAs specific to REBOV-NP and -GP and NT showed positive and the probability that the target assay result was negative when the IFAs and NT showed negative, respectively. Receiver operating characteristics (ROC) and two graph-ROC (TG-ROC) curves were analyzed using Stat Flex software (Artech Co. Ltd., Osaka, Japan) [36, 37]. To examine the agreement between the serological assays used in the present study, we employed Friedman test with Dunn's post-hoc test [38, 39] using GraphPad Prism using pair-wise comparisons in nonparametric one-way ANOVA.