Influenza D Virus Infection in Feral Swine Populations, United States

Influenza D virus (IDV) has been identified in domestic cattle, swine, camelid, and small ruminant populations across North America, Europe, Asia, South America, and Africa. Our study investigated seroprevalence and transmissibility of IDV in feral swine. During 2012–2013, we evaluated feral swine populations in 4 US states; of 256 swine tested, 57 (19.1%) were IDV seropositive. Among 96 archived influenza A virus–seropositive feral swine samples collected from 16 US states during 2010–2013, 41 (42.7%) were IDV seropositive. Infection studies demonstrated that IDV-inoculated feral swine shed virus 3–5 days postinoculation and seroconverted at 21 days postinoculation; 50% of in-contact naive feral swine shed virus, seroconverted, or both. Immunohistochemical staining showed viral antigen within epithelial cells of the respiratory tract, including trachea, soft palate, and lungs. Our findings suggest that feral swine might serve an important role in the ecology of IDV.

Swine were introduced into what is now the United States in the 15th century. Since that time, populations of free-ranging swine have spread to ≈40 states. These swine are escaped domestic animals, imported wild boar, or hybrids of the two, and they now number ≈5 million ( Figure  1, panel A) (16)(17)(18)(19). Feral swine transmit diseases that are swine-specific (feral and domestic) as well as diseases that can be transmitted to domestic species (cattle, sheep, goats, horses, and dogs) and wild mammals; some of the more important diseases include porcine circovirus-2, pseudorabies virus, Brucella suis, and influenza A viruses (IAVs) as well as vesicular diseases (16,17,(20)(21)(22)(23)(24)(25). Feral swine have been shown to have contact with domestic swine in transitional and commercial settings (17,20). Moreover, feral swine also frequently interact with free-ranging cattle near shared water sources (16). Of particular concern, feral swine populations are increasing and pose potential threats to domestic swine and human public health (26).
Little is known regarding seroprevalence of IDV in feral swine. In this study, we conducted serologic surveillance to estimate seroprevalence of IDV in the feral swine population in the United States. We also conducted infection experiments to determine the pathogenesis and transmission of IDV in feral swine.  (Table  1). North Carolina, Oklahoma, and Texas have large domestic swine populations; Texas and Oklahoma have large cattle populations; and Hawaii, Oklahoma, and Texas have large feral swine populations. Of the 256 serum samples, 118 were from male feral swine, 135 were from female feral swine, and 3 were from feral swine of unknown sex; in addition, 32 of the samples were from juveniles (<2 mo of age), 43 from subadults (2-12 mo of age), and 181 from adults (>1 y of age) ( Table 2).

Hemagglutination Assays
We performed hemagglutination (HA) and HA inhibition (HAI) assays as previously described (11). In brief, we treated serum samples with receptor-destroying enzyme (Denka Seiken Co., Tokyo, Japan) at 37°C for >18 hr, followed by heat inactivation at 55°C for 30 min. We diluted inactivated serum to a final concentration of 1:10 with 1× phosphate-buffered saline. We added turkey red blood cells to the serum (concentration 1:20) at 4°C for 30 min and then centrifuged the serum at 13,000 rpm for 1 min to pellet the red blood cells. We conducted the HA and HAI assays with 0.5% turkey red blood cells at 4°C against a testing IDV; we considered samples with an HAI titer >1:40 as IDV seropositive. We tested all serum samples against influenza D/46N and D/13N.

Infection Experiments
We trapped a total of 26 feral swine over the course of 100 trap nights in Mississippi and transported them to the US Department of Agriculture's Mississippi Field Station under state permits (nos. 894, 896, and 908). All 26 feral swine tested seronegative for pseudorabies, brucellosis, and IDVs (D/13N and D/46N).
We randomly separated the 26 feral swine into 3 groups: virus-inoculated animals (n = 12), contact animals (n = 8), and control animals (n = 6). We used 12 pens to house the 26 feral swine; 4 pens contained 1 virus-inoculated animal plus 1 contact animal, 4 pens contained 2 virus-inoculated animals plus 1 contact animal, and 3 pens contained 2 control animals. Pens housing control feral swine were in an animal room separate from pens housing contact and inoculated feral swine. We inoculated each animal for the virus-inoculated group intranasally with 1 mL of D/46N (10 6 50% tissue culture infective dose [TCID 50 ]) and each control animal with 1 mL of phosphate-buffered saline. At 2 days postinoculation (dpi), we moved 1 contact animal into the pen housing 1 or 2 virus-inoculated feral swine. At 0, 3, 5, 7, 9, 11, and 21 dpi, we collected nasal washes, rectal swab samples, and blood from all animals. We stored nasal washes and rectal swab samples at -80°C and serum samples at -20°C.

Calculation of TCID 50
We serially diluted homogenate supernatants from tissue samples and nasal washes from 10 -1 to 10 -6 and titrated them in HRT-18G cells. We tested each sample in triplicate and calculated the TCID 50 by using the Reed-Muench method (27).

Immunohistochemical Staining
We performed IDV immunohistochemical staining as previously described (28). In brief, we heated slides at 65°C overnight. We deparaffinized and retrieved slides in an antigen retrieval solution at pH 6.1 (Dako, Carpinteria, CA, USA) by using a decloaking chamber. We used Tris-buffered saline (TBS) with 0.5% Tween to wash slides and used 3% H 2 O 2 to quench peroxidase activity; we then blocked slides with 10% normal goat serum (Invitrogen, Carlsbad,
Previous serologic surveillance showed that ≈4.9% of feral swine had been exposed to IAVs (25). We explored whether an opportunity exists for feral swine to be exposed to both IAV and IDV. Our results show that 13 (5.1%) of the 256 serum samples were IAV positive (Table 1, Table 2).
To evaluate the characteristics of IDV infection in feral swine, we inoculated D/46N virus intranasally into 12 feral swine. We chose D/46N because this virus was shown to cause infection and transmission as well as a substantial increase in neutrophil tracking in tracheal epithelia of the infected calves, and we intended to compare the pathogenesis in cattle with that in feral swine (28). The IDV-inoculated swine showed no clinical signs or changes in body temperature. Viral titration of nasal washes showed that, at 3 dpi, 7 of 12 D/46N-inoculated swine shed virus with a maximum titer of 2.199 log 10 TCID 50 /mL, and that, at 5 dpi, 6 of the 8 remaining virus-inoculated swine shed virus with a maximum titer of 2.366 log 10 TCID 50 /mL. None of the remaining 5 virusinoculated swine shed virus at or after 7 dpi (Figure 2; online Technical Appendix Table 3). No virus was detected in any rectal swab samples from these experimentally infected feral swine.
HAI results indicated that 7 (63.6%) of 11 virus-inoculated animals seroconverted at 5 dpi and all 8 remaining virus-inoculated animals seroconverted at 7 dpi (online Technical Appendix Table 4). We did not detect virus in any fecal swab samples from virus-inoculated swine, nor did we detect virus in nasal washes or fecal swab samples from the control feral swine, which remained seronegative throughout the study.
The viral titrations of feral swine tissues demonstrated viral replication in the upper and lower respiratory tract as well as the soft palate. At 5 dpi, viral titers were highest in the trachea sections (2.699-2.366 log 10 TCID 50 /mL) and lowest in the left and right caudal lung and soft palate (0.699-2.199 log 10 TCID 50 /mL 0 ) (online Technical Appendix Table 3). At 7 dpi, we detected no virus in nasal swab samples; however, the highest (maximum) viral titer (3.866 log 10 TCID 50 /mL) was in the soft palate, and the lowest viral titers (0.699 log 10 TCID 50 /mL) in the lower trachea (online Technical Appendix Table 3).
Three of the 8 contact animals exposed to IDVinfected feral swine had detectable viral shedding: 1 animal shed virus at 5 days postexposure (titer 2.032 log 10 TCID 50 /mL), and 1 animal each shed virus at 7 and 9 days postexposure (both had a titer equal to the detection limit). At 19 days postexposure, 1 of the 3 remaining contact feral swine seroconverted, with an HAI titer of 1:40, indicating that IDV can be transmitted among feral swine (Table 3;  online Technical Appendix Table 4). Nearly half of the contact animals exposed to IDV-infected feral swine were infected with IDV ( Figure 2). We did not detect virus in any rectal swab samples from the contact feral swine. None of the contact animals showed any clinical signs or change in body temperature.

Discussion
Since it was first identified in domestic swine in 2011, IDV has been detected in a wide range of animal hosts, mainly livestock such as bovids, goats, horses, and sheep, across multiple continents, indicating that IDV is a transboundary pathogen (12)(13)(14)(15)(16)(17)(18)(19)30). In the United States, feral swine serve as an important vector between domestic and wild animals for multiple transboundary diseases, such as Brucella suis and IAV (20)(21)(22)(23)(24)(25). Our study demonstrated that the seroprevalence rate of IDV in feral swine was 19.1%, which is similar to rates (13.5%-18.3%) reported for commercial beef cattle (15) but higher than the reported rate (<10%) in domestic swine (1). For example, a serologic study that used 220 pigs (3-20 weeks old) found that only 9.5% of tested domestic swine had an IDV titer ≥1:10, with a GMT of 1:20.7 (1). The difference in the seroprevalence between domestic and feral swine might be related to the fact that feral swine are mobile and have ample opportunities to come into contact with various domestic and wild animals; thus, compared with domestic swine, feral swine could have additional opportunities to be exposed to IDV. Among the 4 states that we sampled, the state with the highest IDV seroprevalence rate in feral swine also has the largest cattle population ( Figure 1); however, whether IDV transmissions between bovids and feral swine is bidirectional is unknown. Two previous studies suggested that feral swine are likely to have indirect and direct contact with free-range bovine herds near water sources and that higher B. suis seroprevalence among free-range bovine herds was likely attributable to the bovine herd's close proximity to feral swine (16,23). Nevertheless, the seroprevalence rate reported in our study among feral swine was based on convenience samples; thus, an epidemiologic study will be needed to determine the enzootic status of IDV in the feral swine population.
Previous studies have suggested that domestic swine are major sources of IAV exposure for feral swine (24,25). IDV was shown not to cross-react with IAV (1), and our study showed that 42.3% of the IAV-seropositive feral swine had exposure to IDV, indicating that feral swine were exposed to both IDV and IAV. In addition, our results showed that the seroprevalence rate of IDV in IAVseropositive feral swine was more than twice that observed among IAV-negative feral swine. However, a larger epidemiologic study covering larger geographic areas and longer periods is needed to test the hypothesis that IDV would be more prevalent in IAV-positive than IAV-negative feral swine. The effect of IDV infections on the pathogenesis of IAV, or vice versa, remains unknown, but findings from our study and previous studies suggest that future work ought to focus on whether feral swine act as a vector for transboundary disease between domestic swine and cattle.
The results of our animal challenge study show that IDV can be transmitted among feral swine; however, the infection resulted in limited clinical signs. Cattle infected with IDV shed virus up to 9 dpi, with a peak titer of 4.417 log 10 TCID 50 /mL (28), whereas the swine infected with IDV in our study shed virus only up to 5 dpi, with a peak titer of 2.366 log 10 TCID 50 /mL (Figure 2). Such a difference might be attributable to the distinct patterns of viral distributions in the respiratory tracts of the infected animals. In cattle, IDV was predominantly distributed in the tissues of the upper respiratory tract (i.e., turbinate and trachea) (28), but in swine the distribution was predominantly in the middle and lower respiratory tract (i.e., trachea and lung) (Figure 4, panel A). Previous studies have suggested that, under laboratory conditions, IDV replicates in the upper and lower respiratory tracts of guinea pigs infected with a bovine strain of IDV (6). In addition, bovine IDV infects ferrets and can be transmitted to IDV-naive ferrets through direct contact; however, IDV cannot infect naive ferrets through a fomite contaminated with nasal drainage from IDV-infected calves (1,28). Another swine infection study indicated that no viruses were detected in the lung of domestic swine that were infected with D/swine/Oklahoma/1334/2011 (1). The genetic variations between D/ swine/Oklahoma/1334/2011and D/46N used in these studies might have led to the difference in viral tissue tropisms observed in these 2 studies.
Three IDV-inoculated animals and 1 contact animal in our study had transient viremia, a finding consistent with a previous study that found IDV in animal serum samples during IDV surveillance (31). We found IDV at moderate viral titers in the soft palate of feral swine at 3 dpi, and the viremia lasted >3 days for some animals (online Technical Appendix Table 3). The soft palate has been identified as a major site of influenza virus infection in ferrets (32). Previous studies have shown that several bacteria (e.g., Streptococcus porcinus, Streptococcus dysgalactiae, Staphylococcus aureus, Staphylococcus hyicus, Streptococcus suis, Yersinia enterocolitica, Salmonella spp., and Listeria monocytogenes) and viruses (e.g., porcine reproductive and respiratory syndrome virus and porcine circovirus-2) can be isolated from the soft palate (33). The rich distribution of lymphoid tissue in the soft palate might enable virus from the soft palate to enter the bloodstream and cause the transient viremia observed in feral swine that we observed in the experimentally infected feral swine.
The transmission ability of IDV through direct contact is similar to that in domestic swine; however, the transmission ability in feral and domestic swine seems to be less efficient than that in bovids, as suggested by viral load titers and durations of shedding (1,28). Given the limited transmissibility of IDV in feral swine, we would speculate that feral swine could have additional opportunities for exposure to IDV in addition to IDVs circulating in the feral swine populations.
In summary, our findings suggest that IDV has been circulating in the feral swine population across multiple states in the United States and that IDV can be transmitted among feral swine. Although the economic impact of IDV on commercial livestock remains unknown, our findings suggest that feral swine might be important in the ecology of IDV. Further studies are needed to understand whether other wild animals are infected by IDV and to what extent interspecies transmission contributes to IDV maintenance in domestic and wild populations.