Influenza A virus (IAV) is capable of infecting a wide variety of avian and mammalian species, including swine. The control of IAV in swine populations is complicated by the fact that the virus is endemic in contemporary herds and may circulate in any age group. The detection of IAV has historically been based on testing individual pig nasal swab (virus detection) or serum (antibody detection) specimens. While individual pig sampling is adequate for the diagnosis of clinical IAV infections, the collection of adequate numbers of individual pig samples is too costly and labor-intensive for routine influenza surveillance or large-scale ecological studies. Therefore, the general question addressed in this dissertation is whether oral fluid specimens could be used to surveil IAV infections as an alternative to individual animal sampling. More specifically, the aim of this research was to evaluate an IAV oral fluid antibody enzyme-linked immunosorbent assay (ELISA) for the detection of IAV nucleoprotein (NP) antibody and its use in surveillance of swine populations. This question was addressed in the logical series of experiments described below.

The initial objective was to determine whether diagnostic levels of IAV NP antibodies could be detected in swine oral fluid specimens by adapting the serum ELISA protocol to the oral fluid matrix (Chapter 3). The NP antibody ELISA was selected because the NP is highly conserved among IAV subtypes. The procedure for performing the NP blocking ELISA on oral fluid was modified from the serum testing protocol by changing sample dilution, sample volume, incubation time, and incubation temperature. The detection of NP antibody was evaluated using pen-based oral fluid samples (n = 182) from pigs inoculated with either influenza A virus subtype H1N1 or H3N2 under experimental conditions and followed for 42 days post inoculation (DPI). NP antibodies in oral fluid were detected from DPI 7 to 42 in all inoculated groups, i.e., the mean sample-to-negative (S/N) ratio of influenza-inoculated pigs was significantly different (p < 0.0001) from uninoculated controls (unvaccinated or vaccinated-uninoculated groups) through this period. Oral fluid vs. serum S/N ratios from the same pen showed a correlation of 0.796 (Pearson correlation coefficient, p < 0.0001). The results showed that oral fluid samples from influenza virus-infected pigs contained detectable levels of NP antibodies for ≥ 42 DPI.

The availability of serum and oral fluid NP ELISAs provided the tools necessary to describe the kinetics of IAV NP antibody (IgM, IgA, and IgG) in serum and oral fluid specimen from animals of defined IAV infection status (Chapter 4). A significant oral fluid IgM response was only detected in unvaccinated groups. The maximum oral fluid IgM response in these groups was detected at DPI 8, after which it rapidly declined. Oral fluid IgA was detected in both vaccinated and unvaccinated groups on DPI 6. Levels of oral fluid IgA remained relatively stable through DPI 42. Oral fluid IgG responses in both vaccinated and unvaccinated groups were detected by DPI 8 and remained stable through DPI 42. IgM responses in serum and oral fluid were highly correlated in unvaccinated groups (r = 0.810), as were serum and oral fluid IgG responses in both unvaccinated (r = 0.839) and vaccinated (r = 0.856) groups. In contrast, the correlation between serum and oral fluid IgA was weak (r ~ 0.3), regardless of vaccination status. The results from this study demonstrated that NP-specific IgM, IgA, and IgG antibody were detectable in serum and oral fluid and their ontogeny was influenced by vaccination status, the time course of the infection, and specimen type.

The feasibility of IAV surveillance in the field was evaluated using pre-weaning oral fluid samples from litters of piglets in four ~12,500 sow, IAV-vaccinated, breeding herds (Chapter 5). All four herds were considered endemically infected with IAV based on historic diagnostic data. Oral fluid samples were collected from 600 litters prior to weaning and serum samples from their dams after weaning. Litter oral fluid samples were tested for IAV by virus isolation, quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR), RT-PCR subtyping, and sequencing. Commercial NP ELISA kits and NP isotype-specific assays (IgM, IgA, and IgG) were used to characterize NP antibody in litter oral fluid and sow serum. All litter oral fluid specimens (n = 600) were negative by virus isolation. Twenty-five oral fluid samples were positive by qRT-PCR, based on screening (Laboratory 1) and confirmatory testing (Laboratory 2). No hemagglutinin (HA) and neuraminidase (NA) gene sequences were obtained, but matrix (M) gene sequences were obtained for all qRT-PCR-positive samples submitted for sequencing (n = 18). Genetic analysis revealed that all M genes sequences were identical (GenBank accession no. KF487544) and belonged to the triple reassortant influenza A virus M gene (TRIG M) previously identified in swine. The proportion of IgM- and IgA-positive samples was significantly higher in sow serum and litter oral fluid samples, respectively (p < 0.01). Consistent with the extensive use of IAV vaccine, no difference was detected in the proportion of IgG- and blocking ELISA-positive sow serum and litter oral fluids. This study supported the use of oral fluid sampling as a means to conduct IAV surveillance in pig populations and demonstrated the inapparent circulation of IAV in piglets.