Author contributions: Anli Zhang, Yongquan He, Yang Huang and Di Tian conducted the study, Anli Zhang and Chao Qiu analyzed the data, Di Tian, Yanmin Wan and Yanqin Ren provided H7N9 viruses, Xiaoyan Zhang coordinated the study, Anli Zhang and Jianqing Xu wrote the manuscript, Jianqing Xu conceived and designed the study.
Competing interests statement: The authors declare that they have no competing financial interests.
Citation: Zhang AL, He YQ, Huang Y, et al. A pseudovirus-based hemagglutination-inhibition assay as a rapid, highly sensitive, and specific assay for detecting avian influenza A (H7N9) antibodies. Infect Dis Trans Med, 2015; 1(01):12-15
A novel avian influenza A (H7N9) virus infection in humans emerged in China in early 2013, causing acute respiratory distress syndrome, particularly in elderly subjects [1, 2, 3]. Over 160 deaths were reported by July 2014 (Chinese Center for Disease Control and Prevention, unpublished data). As the H7N9 virus has become endemic in wild and domestic birds [4, 5], transmission from birds to humans will likely continue in the future; thus, monitoring H7N9 infections in the human population will be critical for public health decision making. Both hemagglutination-inhibition (HI) and microneutralization (MN) assays are commonly used to detect antibodies in serum, and they have been extensively employed in seroprevalence investigation . Because the H7N9 virus has to be handled in a BSL3 laboratory, HI and MN tests are currently restricted to a few laboratories. Here, we report a safe, sensitive, and specific pseudovirus-based HI assay for detecting H7N9 antibodies that can be performed in a BSL2 laboratory.
Pseudovirus was packaged as previously described [7, 8]. Briefly, 8 × 106 human embryonic kidney 293T cells were seeded in a 10-cm dish overnight. The medium was refreshed with 1% Dulbecco’ s modified Eagle’ s medium (DMEM) prior to transfection. Cells were co-transfected with 5 g of the lentivirus vector pNL4-3-Luc R-E-, 2.5 g of pVKD-HA, and 2.5 μ g of pVKD-NA using TurboFect (Thermo Fisher Scientific, Waltham, MA, USA). The coding sequences of hemagglutinin (HA) (GenBank ID: KC853228) and neuraminidase (NA) (GenBank ID: KC853231) were amplified from the A/Shanghai/4664T/2013(H7N9) strain by reverse transcription polymerase chain reaction (RT-PCR). The pseudovirus-containing supernatant was harvested after 48h incubation and digested for 30 min at 37℃ with 100 l of trypsin-EDTA (Life Technologies, Grand Island, NY, USA) per 900 l of supernatant. Then, 300 g/ml of trypsin inhibitor (AMRESCO, Solon, OH, USA) was added to the digested supernatant, which was stored in aliquots in a -80℃ freezer until used in the following assay. The pseudovirus was 2-fold serially diluted and mixed with 1% horse red blood cells (HRBCs, kindly provided by Shanghai Equestrian Sports Field) to determine the HA titers.
We then used the pseudovirus to test the HA titers of 80 serum samples, including 40 sera collected from RT-PCR-confirmed H7N9 influenza A-infected patients hospitalized in Shanghai Public Health Clinical Center and 40 sera from H7N9-negative volunteers who had not reported a cough or fever during the previous year. All sera were tested for nonspecific agglutinins; if agglutination occurred in wells with a serum dilution of 1:20 or higher, then the serum was adsorbed with packed HRBCs to remove nonspecific agglutinins according to the recommendations of the US Centers for Disease Control. Sera were then treated with three volumes of receptor-destroying enzyme (RDE) (Denka Seiken, Tokyo, Japan) for 18h at 37℃, followed by 30min incubation at 56℃ to inactivate the RDE. Sera were 2-fold serially diluted, starting at a 1:10 dilution. Pseudovirus was adjusted to 8 HA units per 50 l, and then 25 l of pseudovirus was added to the diluted serum and incubated for 30 min. Finally, 50 l of 1% HRBCs were added and incubated for 1h to let the HRBCs settle. The result was read by tilting the plates at 60° , at which point the settled HRBCs coalesced into a teardrop shape, indicating complete hemagglutination inhibition. The serum pseudovirus HI (PHI) titer is the reciprocal of the serum dilution in the last well exhibiting complete hemagglutination inhibition. The HI assay with live H7N9 virus (A/Shanghai/4664T/2013) was tested in parallel with the same procedure, except that the experiment was performed in a BSL3 facility. The neutralizing activities against the H7N9 virus were also validated by the MN assay. In a 96-well plate, 2-fold serially diluted sera, starting at a 1:10 dilution, were incubated with 200 median tissue culture infective doses (TCID50s) of the H7N9 viruses in a final volume of 100 l at 37℃ for 1h, and then the mixture was added to Madin-Darby canine kidney (MDCK) cells that had been cultured overnight. Pathological changes in the MDCK cells were determined 72h later using a microscope.
The results showed that the PHI titers were significantly correlated with both the MN titers (Fig. 1A, Spearman’ s r=0.96, P< 0.001, n=80) and HI titers (Fig. 1B, Spearman’ s r=0.88, P< 0.001, n=80). The HI titer was also correlated with the MN titer (Fig. 1C, Spearman’ s r=0.89, P< 0.001, n=80). To compare their accuracy, we used receiver operating characteristic (ROC) curves to test the sensitivities and specificities of the PHI, HI, and MN assays. Differences in sensitivity and specificity using different thresholds are shown in Figure 2. When using a PHI or MN cut-off titer of 10, the sensitivity and specificity reached 1.0. For the HI assay, the best cut-off point was at a titer of 10, with a sensitivity of 0.75 and a specificity of 1.0. The area under the ROC curve (AUC) was usually used as a summary index of overall diagnostic accuracy; the AUCs of both the PHI and MN assays were 1.0, indicating perfect discrimination. The AUC of the HI assay was 0.75, which indicates poor sensitivity, as previous studies proved [9, 10]. To further confirm its specificity, we tested the H7-based PHI assay on sera from subjects immunized with the 2009 H1N1 vaccine strain, as well as on sera from mice immunized with an H7 strain or various combinations of H1, H2, H3, and H5 strains. Only sera from SH13 H7N9-immunized mice showed a titer above 80, whereas all other sera from 2009 H1N1-immunized subjects or from sera from mice immunized with various combinations of H1, H2, H3, and H5 strains displayed titers below 10 (Table 1 and Table 2). These data suggest that the detection of H7-specific antibodies by the H7-based PHI assay is highly specific.
Discussion and Conclusions: Here, we developed a safe, sensitive and specific method for detecting H7N9 antibodies. Conventional HI or MN assays require propagating live H7N9 viruses, which is expensive and has to be performed in a BSL3 laboratory, thereby restricting their use to a few laboratories. However, the pseudovirus can be mass produced and safely handled in a BSL2 laboratory. Thus, a pseudovirus-based HI assay is convenient and cost-effective for researchers conducting H7N9-related, large-scale serum screening.
The MN assay is broadly used to determine host neutralizing antibodies against H7N9 infection, and its sensitivity and specificity have been confirmed in previous studies [6, 11]. The results of the PHI assay were highly correlated with those of the MN assay, but much less so with those of the HI assay. This may be because the HI assay is less sensitive as it detects antibodies directed against the receptor binding site of HA, whereas the MN assay measures the total neutralizing antibodies that are capable of blocking the entry of influenza viruses. The reason why the PHI assay is more sensitive than the HI assay remains unknown. One possibility is that pseudovirus is less effective in agglutinating HRBCs; thus, they are more susceptible to neutralization. Unlike pseudovirus-based neutralization  or MN assays, the PHI assay does not require cell culture and virus replication procedures. Moreover, it can be rapidly performed in a BSL2 laboratory, and it is resistant to interference from antiviral drugs. Because all sera from H7N9-infected patients were collected 10 days after disease onset, the sensitivity and specificity of the PHI assay for samples collected within 10 days of disease onset, in comparison with the MN assay, remains unknown.
In summary, the PHI assay is a rapid and highly sensitive and specific assay that can be performed in a BSL2 laboratory, and is therefore an appropriate assay for most laboratories. This assay could be used alone in surveillance and seroprevalence studies or in combination with the MN assay in assessing neutralization activities.
This work was supported by the Program for Emergency Response to H7N9 Outbreak (2013QLG003), Shanghai Municipal Commission of Health and Family Planning, and the 973 National Key Basic Research Project (2014CB542502), Ministry of Science and Technology of the P.R. of China.
The authors have declared that no competing interests exist.