The use of equine influenza pseudotyped lentiviruses for serological screening

Background: The H3N8 subtype of equine influenza virus continues to produce outbreaks worldwide, with resultant economic and agricultural impact. The surface hemagglutinin (HA) glycoprotein elicits neutralizing antibody responses following a natural infection or vaccination. Unfortunately, the standard assays used for influenza serology present certain practical issues, such as interlaboratory variability, complex protocols, and the necessity to handle certain virus strains in high biological containment facilities. To address this, avian and human influenza HA pseudotyped retroviruses have been used successfully in antibody neutralization assays. In this study, we have, for the first time, generated equine influenza pseudotyped lentiviruses to detect HA-directed neutralizing antibodies in equine sample sera. Materials and Methods: Stocks of H3N8 A/equine/Sussex/89 were grown in 10- day-old embryonated eggs, with viral RNA extracted and HA amplified using RT-PCR and custom primers. The product was cloned into the pI.18 expression vector. Pseudotyped viruses were then produced through co-transfection of HEK293T cells with plasmids expressing the HA, HIV gag-pol, and firefly luciferase reporter genes and harvesting of virus from the supernatant. To produce infective pseudotype particles, it was necessary to also cotransfect a plasmid encoding the TMPRSS2 endoprotease to cleave the HA. High-titer pseudotype virus (PV) was then used in PV antibody neutralization assays (PVNAs); a luminometer was used to quantify PV infectivity (luciferase expression). Sera from vaccinated (n = 21) and influenza-naïve (n = 2) equines were tested. These samples were also screened using the single radial hemolysis (SRH) assay. Correlation and significance were determined by Pearson analysis using GraphPad Prism software. Results: Pseudotype viruses were generated successfully, with a titer of 1 x 109 relative luminescence units per mL. The TMPRSS2 protease was essential for this process, with no significant PV generation in the absence of the protease expression plasmid. The PVs thus produced were used in PVNAs to distinguish all test sera from vaccinated versus nonvaccinated equines. All vaccinated equines exhibited PVNA antibody titers of IC50 >1100 (positive control IC50: >40,000; negative sera: <80). Antibody was detected by SRH in 18/20 of the vaccinated equines (range: 61-207 mm2). There was a 65% correlation between results from the 2 assays (r = 0.65, P = .002). Conclusion: H3N8 equine influenza pseudotyped lentiviruses have been generated for the first time. These were successfully used to detect anti-HA neutralizing antibodies in serum samples. We were also able to use the pseudotype virus neutralization assay to distinguish sera from vaccinated versus influenza-naïve (by SRH) equines. The correlation between PVNA and SRH assays was good (65%), with the pseudotype assay exhibiting slightly more sensitivity. Future work will require more extensive testing of the PVNA using a larger number of serum samples to assess sensitivity/specificity and inter/intralaboratory variability and to define a protective antibody titer.