Infections with EIV have remained a serious health and economic problem in horses throughout the world.1 The immune response induced by infection with an influenza virus protects against reinfection with the same or an antigenically similar virus strain. However, influenza viruses undergo frequent antigenic changes (antigenic drift); thus, protection provided by the host's immunity may be reduced as the virus becomes more antigenically distinct.2 The introduction of gene segments or entire viruses from other host species also adds to the vast genetic and antigenic diversity of influenza viruses.3 Although horses have often been regarded as isolated or dead-end hosts for influenza cross-species transmission,3,4 the transmission (and subsequent maintenance) of an equine-lineage H3N8 virus to dogs5 in the United States highlights the fact that there is not an absolute barrier for influenza viruses to emerge from horses.
Since their first isolation in 2004, CIVs have continued to evolve genetically. Phylogenetic analyses of the HA and neuraminidase genes of contemporary canine isolates indicate that the CIVs have segregated from the equine H3 Florida lineage as a distinct sublineage.5,6 Five amino acid residues at positions 54, 83, 222, 328, and 483 of H3 HA appear to differentiate CIV isolates from contemporary equine H3N8 viruses.5,6 Although the biological importance of these amino acid substitutions remains unclear, from a host immunologic perspective, the asparagine-to-lysine mutation at position 54 and possibly the serine-to-asparagine substitution at position 83 are of particular interest because they are located in antibody-binding regions of the HA protein.7,8 The potential importance of these amino acid residues is highlighted because residue 83 can be involved in antigenic drift in human H3 viruses.9,10 Similarly, the N54K substitution is in the center of an N-linked gly-cosylation motive; consequently, the posttranslational glycosylation of the protein may be altered,11 thereby resulting in the reduced accessibility of the epitope.12,13 In light of these findings, it has been hypothesized8 that the mutations at residues 54 and 83 can facilitate viral escape from neutralization by preexisting antibodies. Asparagine at position 54 is highly conserved in equine and other noncanine H3N8 influenza viruses, including the strains used for production of equine vaccines currently marketed in the United States; therefore, we hypothesized that the canine isolates could represent an emerging disease threat to horses, provided that the canine isolates have maintained the ability to infect horses. Moreover, the finding that dogs could represent a source of infection for horses would have importance for the development and implementation of biosecurity protocols on equine farms.
Despite the fact that cross-species transmission of influenza A viruses is a relatively frequent event, such newly introduced viruses are only rarely maintained in the new host species.3 Although the properties necessary to allow a virus to form a stable lineage are not known, viral HA is considered to play a key role in influenza species specificity.14,15 Given the importance of the HA protein in limiting transmission of influenza A viruses among species, it is unclear whether the genetic divergence of the canine and equine H3 viruses, including the 5 amino acid mutations in the HA protein, has resulted in a reduction of infectivity for canine-lineage influenzas in horses.
In the study reported here, we examined whether an equine-derived canine H3N8 influenza A virus maintained its infectivity in equids by inoculating ponies via aerosol exposure with a canine influenza isolate (Ca/WY). A second group of ponies (positive control group) was inoculated with a contemporary equine H3N8 virus (Eq/CO). Finally, because efficient animal-to-animal transmission is an important requirement for virus maintenance, we also evaluated the horizontal spread of virus to influenza-naïve sentinel ponies housed in direct contact with the inoculated ponies. Growth and infection characteristics of Ca/WY and Eq/CO were also examined in MDCK cells and primary CRECs and ERECs, which have been used in other studies16–19 conducted to evaluate infection with influenza viruses.
Materials and Methods
Animals—Twenty 12- to 24-month-old ponies were purchased from a commercial source. Ponies were serologically negative for EIV as determined on the basis of results of an ELISA20 and HI assay.21 All ponies were healthy and in good body condition and were maintained in accordance with guidelines established by the Colorado State University Research and Animal Resources Committee. They were fed a diet of hay and a pelleted vitamin and mineral concentrate and were group housed outdoors in pens in 3 geographically separate locations with access to water and shelter. The study protocol was reviewed and approved for conduct by the Colorado State University Institutional Animal Care and Use Committee.
Influenza viruses—The contemporary H3N8 CIV Ca/WY (GenBank accession Nos. HQ917681 and HQ99104 and the contemporary H3N8 EIV Eq/CO (Gen-Bank accession Nos. HQ917682 and HQ993105) were isolated during outbreaks of EIV and CIV, respectively. Infection with EIV was confirmed in July 2007 at an equine boarding facility that housed 42 horses. Clinical signs of EIV infection consisting of fever, coughing, and nasal discharge were initially observed in 2 horses that had returned from a show 3 days earlier. Subsequently, clinical disease spread rapidly among the remaining horses on the property. Nasal shedding of EIV was confirmed in 4 of 32 horses tested by use of a real-time RT-PCR assay. The EIV was isolated (first passage in embryonated chicken eggs) from 2 of the 4 nasal swab specimens that had positive results when tested by use of the real-time RT-PCR assay; the virus was designated as Eq/CO. After the diagnosis was established, initial (acute) serum samples were obtained from 32 of 42 horses, and a convalescent serum sample was obtained from all 32 horses 14 days later. Testing of the paired serum samples for influenza-specific antibodies by use of an HI assay revealed seroconversion in 20 of 32 (62.5%) horses.
The contemporary CIV was isolated that same year. In mid February 2007, there was an outbreak of respiratory disease in dogs at a humane shelter. The outbreak involved all 27 dogs housed at the shelter at the time. Clinical signs included fever, lethargy, coughing, and nasal discharge. Nasal swab specimens and serum samples were obtained from 18 of 27 dogs. Paired acute and convalescent serum samples were available for 13 dogs and were tested for CIV-specific antibodies by use of an HI assay. Twelve of 18 nasal swab specimens had positive results for CIV when tested by use of a real-time RT-PCR assay, and virus was isolated from 3 dogs (first passage in embryonated chicken eggs) and designated as Ca/WY. Of the paired serum samples tested, all 13 (100%) revealed seroconversion.
Genetic characterization and growth characteristics of viruses—Both Ca/WY and Eq/CO were passaged 3 times in MDCKa cells grown in MEM,b as described elsewhere.21 To rule out introduction of spurious mutations during cell culture passage, the full-length protein that codes regions of all 8 gene segments of Eq/CO and Ca/WY from allantoic fluid (before cell culture passage) and third-passage MDCK stocks (used to inoculate ponies) were amplified by use of an RT-PCR assay. Sequences of the amplified genes were determined via direct cycle sequencing.c Sequence comparisons at the nucleotide and deduced amino acid levels were made by use of clustal analysisd and commercially available software.e Phylogenetic relationships among the HA and neuraminidase genes of the virus isolates and selected reference strains were estimated from their nucleotide sequences by use of maximum parsimony via bootstrap analysis with a commercially available software program.f Phylogenetic analyses of the HA and neuraminidase genes confirmed that Ca/WY clustered with the canine isolates and Eq/CO clustered with the contemporary equine viruses, which placed them into the canine and equine sublineages of the equine H3 Florida lineage.5,6 Moreover, analysis of the amino acid sequences of the HA genes of Ca/WY and Eq/CO verified the 5 amino acid substitutions that differentiate the equine and canine H3 consensus sequences.5,6
To characterize virus growth, 1-step growth curves were obtained for MDCK cells22 and for primary CRECs and ERECs grown at an air-fluid interface. Briefly, MDCK cells were grown in 35 × 10-mm tissue culture plates and inoculated with Ca/WY or Eq/CO at a multiplicity of infection of 10 TCID50/cell. After adsorption for 1 hour, the inoculum was removed, and 2 mL of MEM containing 0.5% bovine serum albumin,g penicillin-streptomycin,h amphotericin B,i and tolylsulfonyl phenylalanyl chloromethyl ketone-treated trypsinj (1 μg/mL) was added. Supernatants were harvested 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 20, and 24 hours after inoculation and frozen at −80°C until analysis via a quantitative real-time RT-PCR assay. Isolation, culture, and estimation of purity of primary CRECs and ERECs were performed as described elsewhere.16,23 Briefly, CRECs and ERECs were isolated from fresh canine and equine trachea, respectively, by enzymatic digestion with 1.4% pronasek and 0.1% DNase Il in calcium- and magnesium-free MEM for 48 hours. After digestion, epithelial cells were harvested and incubated on an uncoated Petri dish for 2 hours to reduce fibroblast contamination. Purity of epithelial cells was estimated via immunocytochemical staining by use of a pan-anti-cytokeratin antibodym (5 μg/mL) and analyzed via flow cytometric analysis. Primary CRECs and ERECs were seeded into type IV collagenn—coated cell culture wellso and cultured at the air-fluid interface in Dulbecco MEM-F12,h 2% of a serum substitute,p penicillin-streptomycin, and amphotericin B until confluent. Primary cells were washed with Dulbecco MEM-F12 to remove mucin, and wells were inoculated in duplicate with Ca/WY or Eq/CO at a multiplicity of infection of 10. After adsorption for 2 hours, inoculum was removed and replaced with fresh maintenance media. Alliquots of media were collected 3, 4, 5, 6, 8, 12, 16, and 24 hours after inoculation and stored at −80°C until analysis via a real-time PCR assay. Number of copies of the influenza matrix gene in cell culture media was determined by use of a quantitative real-time RT-PCR assay, as described elsewhere.24,25 In addition, presence of virus antigen was determined via immunocytochemical staining by use of a mouse anti-nucleoprotein monoclonal antibodyq (monoclonal antibody 68D2), as described elsewhere.21
Experimental design—Ponies were assigned (by use of a stratified randomization procedure) to 2 groups (5 ponies/group) for inoculation with viruses. A third group of 4 ponies served as a sham-inoculated negative control group. Subsequently, 3 noninoculated ponies were added to each of the virus-inoculation groups to serve as sentinel animals. For the duration of the study, each group was housed separately to prevent cross-contamination. Physical examination and clinical scoring were conducted on all ponies throughout the course of the study (daily from 2 days before until 14 days after inoculation [virus- and sham-inoculated ponies] and daily from 2 days before until 21 days after group introduction [sentinel ponies]). Clinical scoring was performed as described elsewhere.26 Briefly, ponies were observed for 20 min/d to assess general appearance (attitude, food intake, and respiration were quantified on a binomial scale of 0 or 1, with 0 indicating a clinically normal animal and 1 indicating an abnormal finding), coughing (0 indicating no coughing during observation, 1 indicating coughing once during observation, and 2 indicating coughing ≥ 2 times during observation), and nasal discharge (0 indicating no discharge, 1 indicating serous discharge, 2 indicating mucopurulent discharge, and 3 indicating profuse mucopurulent nasal discharge); rectal temperature was also measured.
Two days before inoculation, nasal swab specimensr and serum samples were obtained. On day 0 (day of inoculation), each of the 5 ponies in virus-inoculated groups was sedated by IV administration of detomidines (5 μg/kg) and butorphanol tartratet (0.05 mg/kg). Sedated ponies in one group were inoculated via aerosol inhalation with 107 TCID50 Eq/CO, whereas sedated ponies in the other group were inoculated via aerosol inhalation with 107 TCID50 Ca/WY, as described elsewhere.27 The 4 negative control ponies were sham inoculated. On day 2 after inoculation, 3 influenza-seronegative ponies (ie, sentinel ponies) were introduced to each of the virus-inoculated groups and housed with them throughout the remainder of the study.
Nasal swab specimens were obtained from each pony (daily from 2 days before until 14 days after inoculation [virus- and sham-inoculated ponies] and daily from 2 days before until 21 days after group introduction [sentinel ponies]). Swab specimens were placed in 1 mL of viral transport medium containing PBS solution, 0.5% bovine serum albumin, penicillin-streptomycin, nystatin,u and gentamicinv; vials were stored at −80°C until further analysis. Blood samples for serologic testing were collected from viral-inoculated and sham-inoculated ponies on days 7, 14, and 21. Serum for sentinel ponies for days 7, 14, and 21 was obtained 2 days after serum was obtained for the viral-inoculation ponies.
Evaluation of virus shedding—Number of copies of the matrix gene in nasal swab specimens was determined by use of an established quantitative real-time RT-PCR assay.24,25 Briefly, for each nasal swab specimen, RNA was extracted from 140 μL of viral transport medium by use of a commercial RNA extraction kitw conducted in accordance with the manufacturer's instructions. One-tube real-time RT-PCR assay was performed by use of the following cycling conditions: 10 minutes at 52°C, 5 minutes at 95°C, and 45 cycles of 10 seconds at 95°C and 45 seconds at 68.4°C. Each nasal swab specimen was assayed in duplicate. Water and transport medium were included as negative control samples in each assay. A positive control sample consisted of 101 TCID50 A/Ca/Colorado/224986/06 virus, (GenBank accession Nos. HQ917678 and HQ993101) in viral transport medium. Purified full-length matrix gene RNA was used as a standard for quantification of the number of copies of the influenza virus matrix gene. Matrix gene RNA was transcribed from the T7 promoter by use of a large-scale RNA production kitx conducted in accordance with the manufacturer's instructions. The RNA generated was treated with RNase-free DNase I and tested for purity via gel electrophoresis and PCR assay before it was used in the real-time RT-PCR assay. Purified RNA was suspended in RNase-free water, quantified with a spectrophotometer, and stored in 10-μL aliquots at −80°C. To determine the minimum detection level for the real-time RT-PCR assay, in vitro transcribed RNA was serially diluted in RNase-free water to yield dilutions that ranged from 107 to 100 copies of matrix gene RNA/mL. To evaluate interassay variation, Ct values of 10 RNA standard curves performed on different days were determined. The mean, SEM, and coefficient of variation were calculated. Minimum detection level of RNA was 103 copies of the matrix gene/reaction for the real-time RT-PCR assay. Amplification of dilutions of the RNA transcripts revealed linearity over a range of 6 orders of magnitude. The mean ± SEM Ct value corresponding to each dilution of matrix gene RNA was 15.9 ± 0.13 at 108 copies, 19.5 ± 0.14 at 107 copies, 22.9 ± 0.1 at 106 copies, 26.3 ± 0.11 at 105 copies, 29.9 ± 0.13 at 104 copies, and 33.8 ± 0.3 at 103 copies. Matrix gene RNA at 102, 101, and 10° copies was undetectable or detectable at higher Ct values (> 36). Coefficients of variation were between 1.5% and 4.7% for the Ct values.
HI assay—The HI assays were performed as described elsewhere.21 Briefly, sera were pretreated with receptor-destroying enzymey and incubated overnight at 37°C. Following enzyme inactivation, 2-fold serial dilutions of sera were mixed with 4 hemagglutination units of Eq/CO and Ca/WY, respectively. The assays were developed by adding 0.5% (vol/vol) chicken RBCs, and HI antibody titers were interpreted as the reciprocal of the highest dilution that caused complete inhibition of agglutination.
Statistical analysis—Generalized estimating equations were used to analyze overall mean differences in the numbers of copies of the matrix gene in nasal swab specimens, HI antibody titers, clinical scores, and rectal temperatures between the Ca/WY-and Eq/CO-inoculated groups and between the viral-inoculated groups and the sham-inoculated group. Mean differences were adjusted on the basis of day and clustered on repeated measures for each outcome in the analysis. Clinical scores were ranked before analysis. All other outcome variables were logarithmically transformed, when necessary, to meet the major assumptions, including linearity and normality. The HI antibody titers with a value of 0 were converted to a value of 1 in the viral-inoculated groups for data transformation and statistical analyses. Comparison of the number of copies of the matrix gene for Ca/WY in ERECs and CRECs at 24 hours after inoculation was performed by use of a 1-way ANOVA. Numbers of copies of the matrix gene for Ca/WY in CRECs and Eq/CO in ERECs were compared by use of a Student t test. Number of copies of the matrix gene was logarithmically transformed prior to both analyses. For all analyses, values of P < 0.05 were considered significant. All statistical analyses were performed by use of commercially available software.z
Results
In vitro growth characteristics—No differences were detected in growth kinetics and number of copies of the matrix gene for Ca/WY and Eq/CO at 24 hours after inoculation in MDCK cell cultures (data not shown). Similarly, no significant (P = 0.5) differences were found when comparing growth kinetics and number of copies of the matrix gene for Eq/CO in ERECs and Ca/WY in CRECs. Correspondingly, visual inspection of ERECs and CRECs inoculated with Eq/CO and Ca/WY, respectively, and immunocytochemically stained revealed that the viruses were able to infect virtually every available cell derived from their respective host species (Figure 1). In contrast, Ca/WY had a low-infectivity phenotype in ERECs, which was paralleled by a significantly (P = 0.005) lower number of copies of the matrix gene in ERECs, compared with the number of copies of the matrix gene in CRECs (data not shown).
Photomicrographs of immunocytochemical staining of virus-infected preparations of representative cell cultures 24 hours after inoculation of Ca/WY into ERECs (A) and CRECs (B) or inoculation of Eq/CO into ERECs (C) and of noninoculated control ERECs (D). Cells were fixed for immunocytochemical analysis 24 hours after inoculation and stained with anti-neurophysin monoclonal antibody 68D2, biotinylated secondary antibody, avidin-horseradish peroxidase, and aminoethyl carbazole chromogen. Red staining indicates virus-infected, influenza neurophysin-positive cells. Bar = 9 μm.
Citation: American Journal of Veterinary Research 72, 8; 10.2460/ajvr.72.8.1071
Clinical responses to inoculation—Clinical responses of ponies to inoculation differed substantially between the 2 viruses. More specifically, all ponies inoculated with Eq/CO developed severe mucopurulent nasal discharge (duration of 7 to 11 days) and coughing (duration of 12 to 14 days; Figure 2). In addition, all 5 ponies were pyrectic (> 38.6°C), with a duration of 2 to 7 days (Figure 3). In contrast, in the Ca/WY-inoculated group, only 1 of 5 ponies developed mild to moderate mucopurulent nasal discharge (duration of 2 days). Moreover, none of the ponies in the CIV-inoculated group developed a cough or were pyrectic. The Eq/CO-inoculated ponies had significantly (P < 0.001) more disease, as assessed on the basis of clinical scores and rectal temperatures, than did the Ca/WY- or sham-inoculated ponies. Clinical scores and rectal temperatures did not differ significantly between the Ca/WY- and sham-inoculated ponies.
Mean ± SEM clinical scores for 5 ponies inoculated with Eq/CO (black diamonds) and 3 sentinel ponies housed with the Eq/CO group (white diamonds) and 5 ponies inoculated with Ca/WY (black triangles). Mean scores for 3 sentinel ponies housed with the Ca/WY group (white triangles) and 4 sham-inoculated (negative control) ponies (data not shown) remained at 0 throughout the study. Day of inoculation was designated as day 0. Sentinel ponies were not introduced into the groups until day 2. Clinical scoring was based on assessment of general appearance (attitude, food intake, and respiration were quantified on a binomial scale of 0 or 1, with 0 indicating a clinically normal animal and 1 indicating an abnormal finding), coughing (0 indicating no coughing during observation, 1 indicating coughing once during observation, and 2 indicating coughing ≥ 2 times during observation), and nasal discharge (0 indicating no discharge, 1 indicating serous discharge, 2 indicating mucopurulent discharge, and 3 indicating profuse mucopurulent nasal discharge). Clinical scores for the Eq/CO-inoculated ponies were significantly (P < 0.001) higher than the scores for the Ca/WY- or sham-inoculated ponies.
Citation: American Journal of Veterinary Research 72, 8; 10.2460/ajvr.72.8.1071
Mean ± SEM rectal temperatures for 5 ponies inoculated with Eq/CO (black diamonds) and 3 sentinel ponies housed with the Eq/CO group (white diamonds) and for 5 ponies inoculated with Ca/WY (black triangles) and 3 sentinel ponies housed with the Ca/WY group (white triangles). Rectal temperatures for the Eq/CO-inoculated ponies were significantly (P < 0.001) higher than the temperatures for the Ca/WY- or sham-inoculated ponies. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 72, 8; 10.2460/ajvr.72.8.1071
Virus shedding—The duration of viral shedding and number of copies of the matrix gene detected in nasal secretions were distinctly different between the canine and equine H3N8 viruses (Figure 4). In the Eq/CO-inoculated ponies, the influenza virus matrix gene was detected in nasal secretions in all 5 ponies starting as soon as 2 days after inoculation. Virus shedding in nasal secretions was detectable for up to 8 days, with ≥ 104 copies of the matrix gene for at least 5 days in all 5 ponies. In contrast, for the pones inoculated with Ca/WY, the matrix gene was not detected in nasal swab specimens obtained from 4 ponies. Only 1 Ca/WY-inoculated pony had nasal swab specimens with positive results when tested by use of the real-time RT-PCR assay (≤ 104 copies of the matrix gene for 2 consecutive days [days 5 and 6] after inoculation). Overall, the mean number of copies of the matrix gene in the nasal swab specimens was significantly higher in ponies inoculated with Eq/CO than in ponies inoculated with Ca/WY. In contrast, there were no significant differences in the number of copies of the matrix gene detected in nasal secretions of ponies inoculated with Ca/WY and the sham-inoculated control ponies.
Mean ± SEM number of copies of the matrix gene for virus shed in nasal secretions for 5 ponies inoculated with Eq/CO (black diamonds) and 3 sentinel ponies housed with the Eq/CO group (white diamonds), 5 ponies inoculated with Ca/WY (black triangles) and 3 sentinel ponies housed with the Ca/WY group (white triangles), and 4 sham-inoculated (negative control) ponies (asterisks). Minimum detection level of the real-time RT-PCR assay was 103 copies of the matrix gene (horizontal dashed line). See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 72, 8; 10.2460/ajvr.72.8.1071
HI antibody responses—The pattern of systemic antibody responses generally mimicked the differences in the severity of clinical signs and extent of viral shedding in nasal swab species found among the inoculation groups. Prior to inoculation, all ponies were serologically negative for EIV and CIV (Table 1). Seven days after inoculation, all 5 ponies inoculated with Eq/CO had detectable, low, virus-specific antibody titers; by day 14 and day 21, these titers had increased substantially (≥ 1:1,024). In contrast, none of the ponies inoculated with Ca/WY developed detectable antibody titers by day 7. Moreover, only 2 of 5 Ca/WY-inoculated ponies developed a low antibody response by day 14 (1:64 and 1:16, respectively) and day 21 (1:64 and 1:4, respectively). The overall mean HI antibody responses in ponies inoculated with the contemporary equine isolate were significantly higher, compared with the responses for the ponies inoculated with the canine isolate.
Serum antibody response of ponies after aerosol inoculation with influenza virus as measured by use of an HI assay.
| Group* | Before inoculation | Day 7 | Day 14 | Day 21 |
|---|---|---|---|---|
| Sham inoculated (n = 4) | 0 | 0 | 0 | 0 |
| Eq/CO (n = 5)† | 0 | 32.0 ± 11.0 | 2,560.0 ± 494.6 | 1,356.8 ± 656.7 |
| Eq/CO sentinels (n = 3)†‡ | 0 | 2.7 ± 3.8 | 768.0 ± 161.9 | 853.3 ± 358.0 |
| Ca/WY (n = 5) | 0 | 0 | 31.2 ± 35.8 | 10.4 ± 9.5 |
| Ca/WY sentinels (n = 3)‡ | 0 | 0 | 0 | 0 |
Values reported are mean 6 SEM titers; titers were interpreted as the reciprocal of the highest dilution that caused complete inhibition of agglutination
Ponies were inoculated with Eq/CO or Ca/WY. Sham-inoculated ponies were a negative control group. Day of inoculation was designated as day 0. Sentinel ponies were not introduced into Eq/CO and Ca/WY groups until day 2.
Values differ significantly (P < 0.05) among days.
Blood samples for serologic testing were collected from viral-inoculated and sham-inoculated ponies on days 7, 14, and 21; serum from sentinel ponies for days 7, 14, and 21 was obtained 2 days after the corresponding serum samples were obtained from the viral-inoculation ponies.
Virus transmission to sentinel ponies—To examine the potential for horizontal transmission of virus, 3 influenza-seronegative ponies were introduced to the EIV- and CIV-inoculated groups to serve as sentinel animals. Clinical signs, virus shedding, and antibody responses of the sentinel ponies housed with the Eq/CO-inoculated group closely mirrored the responses observed in the 5 Eq/CO-inoculated ponies. Specifically, all 3 ponies exposed to the Eq/CO-inoculated ponies had muco-purulent nasal discharge, coughing, and pyrexia as early as 2 days after introduction to the group (Figures 2 and 3). In addition, viral shedding in nasal secretions was detected in all 3 sentinels for the Eq/CO group on day 3 after introduction and persisted for up to 7 days (Figure 4). All 3 sentinels had influenza-specific seroconversion by day 14 (Table 1). In contrast, none of the sentinel ponies in the Ca/WY-inoculated group developed clinical abnormalities or shed detectable amounts of virus in nasal secretions. Lack of biologically relevant virus transmission was further confirmed by the lack of an immune response in the 3 sentinel ponies for the Ca/WY group.
Discussion
The objectives for the study reported were to evaluate whether a contemporary CIV isolate could infect, replicate, cause clinical disease, and spread among ponies and in primary respiratory epithelial cells grown in cell culture. Although CIV was first identified subsequent to the interspecies transfer of an equine-lineage H3N8 virus to dogs,5 the canine viruses have since evolved genetically, which has resulted in segregation of the canine genes from the equine H3 Florida lineage.5,6 On the basis of this genetic diversion, we hypothesized that CIV could represent a disease threat to horses, provided the viruses had maintained the ability to infect horses. Results of the present study indicated that the contemporary canine H3N8 influenza virus Ca/WY was ineffective at causing infection or replication in ponies. Moreover, virus inoculation did not result in clinical disease or spread of virus to naïve in-contact sentinel ponies. By comparison, the contemporary Eq/CO isolate was efficient for infection, replication, and transmission between ponies. Although the latter finding was not unexpected, on the basis of the epidemiological findings for the outbreak during which Eq/CO was isolated and results of similar experiments conducted with EIV by our research group24 and by other investigators,20,26,27 the nearly complete lack of infectivity for the canine isolate in horses was interesting. The isolate Ca/WY is clearly not an inherently replication-defective virus. This notion is supported by the rapid spread of virus among dogs at an animal shelter and the efficient growth of the virus in MDCK cells and primary CRECs. In mammals, the primary targets of influenza viruses are airway epithelial cells, and these cells are increasingly being used to study influenza virus-host interactions. In fact, several studies16–19 have revealed that primary respiratory epithelial cells represent a suitable in vitro system for the investigation of species-specific infection characteristics and the host range of influenza A viruses. In the present study, we extended our findings of the in vivo infection characteristics of Ca/WY and Eq/CO to an in vitro culture system of primary respiratory cells. Although both viruses were able to infect and replicate comparably in primary airway epithelial cells derived from their respective host species, the viruses had distinct differences in infection and replication efficiencies in primary ERECs. These results closely paralleled the infection characteristics observed in ponies, in particular the data for viral shedding in nasal secretions, which suggests that primary ERECs and CRECs could represent a feasible method for the evaluation of virus and host factors that control species-specific replication characteristics of influenza in horses.
The first step of infection with influenza viruses is dependent on the interaction of the viral HA with cellular sialic acid residues. As such, the viral HA is thought to be a major contributor to the host range for the influenza viruses.15,28 Investigators have found CIV isolates (including Ca/WY) contain 5 amino acid differences in their HA protein that are not present in the most closely related equine H3 HAs (including Eq/CO), and it has been hypothesized that these amino acid substitutions are a result of mutational adaptation of the virus to its new host species.5,6 Therefore, it is possible that the amino acid differences in the HA protein could account for at least part of the observed differences in infectivity and replication efficiency of Ca/WY and Eq/CO. For example, the isoleucine-to-threonine substitution at position 328 was near the cleavage site of the H3 HA protein.7 Because the peptide structure connecting the HA1 and HA2 subunit has been found to determine tissue tropism of avian influenza virus,29,30 it is possible that diminishing the ability of equine cellular proteases to cleave the CIV HA resulted in the reduced infectivity of Ca/WY in ponies. Similarly, the asparagine-to-threonine substitution at position 483 results in the loss of a glycosylation site in the HA2 subunit,5,6 which has been found to affect the interaction of HA with the host cell receptor and the release of progeny virus from the host cell.31 In addition, the substitution of leucine for tryptophan at position 222 is remarkable because it represents a nonconservative change adjacent to the receptor binding pocket.5,6 Because species specificity of influenza viruses is partly determined by the binding preference of the HA protein to cellular sialic acid species,15,28 modulation of receptor binding specificity as a result of the leucine-to-tryptophan substitution at position 222 could provide an attractive explanation for the inability of Ca/WY to infect and to replicate in ponies.
Finally, although HA is recognized as a key factor controlling the species specificity of influenza A viruses, it is clear that the other 7 influenza virus gene segments also contribute to the host range for influenza viruses. Therefore, it is possible that the critical viral factors that impact infection and replication of Ca/WY in ponies may include equine-to-canine amino acid substitutions that are present in gene segments other than HA.
We concluded from the study reported here that Ca/WY, a contemporary equine-derived canine H3N8 influenza virus, has virtually lost the ability to infect, replicate, and cause clinical disease in ponies or to spread among ponies. This was further supported by the significantly lower values for clinical infection scores and copy numbers of the matrix gene for Ca/WY in ERECs, compared with results for the Eq/CO isolate. However, although these results appear to support the existence of a barrier to CIV infection in horses, it remains to be determined whether other canine viruses have similar restrictions for infectivity of equids.
ABBREVIATIONS
| Ca/WY | A/Canine/Wyoming/86033/07 virus |
| CIV | Canine influenza virus |
| CREC | Canine respiratory epithelial cell |
| Ct | Threshold cycle |
| EIV | Equine influenza virus |
| Eq/CO | A/Equine/Colorado/10/07 virus |
| EREC | Equine respiratory epithelial cell |
| HA | Hemagglutinin |
| HI | Hemagglutination inhibition |
| MEM | Minimal essential medium |
| MDCK | Madin-Darby canine kidney |
| RT | Reverse transcriptase |
American Type Culture Collection, Manassas, Va.
Eagle MEM, Gibco, Invitrogen, Carlsbad, Calif.
BigDye terminator cycle sequencing ready reaction kit, Perkin-Elmer Applied Biosystems, Foster City, Calif.
Clustal analysis, Kyoto University Bioinformatics Center, Kyoto, Japan. Available at: genome.jp/tools/clustalw. Accessed May 10, 2011.
Lasergene, DNASTAR Inc, Madison, Wis.
PAUP 4.0, Macintosh beta version 10, Sinauer Associates Inc, Sunderland, Mass.
Fisher Scientific, Fair Lawn, NJ.
Gibco, Invitrogen, Carlsbad, Calif.
Biowhittaker, Cambrex Bioscience, Walkersville, Md.
Worthington Biochemical Corp, Lakewood, NJ.
Roche Applied Science, Indianapolis, Ind.
Sigma-Aldrich Chemical Co, St Louis, Mo.
Anti-cytokeratin (pan) antibody, Zymed, Invitrogen, Carlsbad, Calif.
Human placental collagen, type IV, Sigma-Aldrich Chemical Co, St Louis, Mo.
Transwell-clear permeable supports, Costar, Corning, Fisher Scientific, Fair Lawn, NJ.
Ultroser G, Pall Life Sciences, Pall Corp, Cergy, France.
Provided by Martha W. McGregor and Yoshihiro Kawaoka, Department of Pathobiological Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, Wis.
Sterile Dacron polyester-tipped applicators, Hardwood Products Com LLC, Guilford, Me.
Dormosedan (10 mg/mL), Pfizer, Kalamazoo, Mich.
Torbugesic (10 mg/mL), Fort Dodge, Fort Dodge, Iowa.
Nystatin suspension (mycostatin), Sigma-Aldrich Chemical Co, St Louis, Mo.
Gentamicin reagent solution, Gibco, Invitrogen, Carlsbad, Calif.
QIAamp viral RNA mini kit, QIAGEN, Hilden, Germany.
RiboMAX large-scale RNA production systems kit, Promega, Madison, Wis.
Denko Seiken Co, Tokyo, Japan.
STATA, Macintosh version 10.1, StataCorp LP, College Station, Tex.
References
- 1.↑
Mumford EL, Traub-Dargatz JL, Carman J, et al. Occurrence of infectious upper respiratory tract disease and response to vaccination in horses on six sentinel premises in northern Colorado. Equine Vet J 2003; 35: 72–77.
- 2.↑
Mumford JA, Jessett D, Dunleavy U, et al. Antigenicity and immunogenicity of experimental equine influenza ISCOM vaccines. Vaccine 1994; 12: 857–863.
- 3.↑
Webster RG, Bean WJ, Gorman OT, et al. Evolution and ecology of influenza A viruses. Microbiol Rev 1992; 56: 152–179.
- 4.
Tumova B. Equine influenza—a segment in influenza virus ecology. Comp Immunol Microbiol Infect Dis 1980; 3: 45–59.
- 5.↑
Crawford PC, Dubovi EJ, Castleman WL, et al. Transmission of equine influenza virus to dogs. Science 2005; 310: 482–485.
- 6.
Payungporn S, Crawford PC, Kouo TS, et al. Influenza A virus (H3N8) in dogs with respiratory disease, Florida. Emerg Infect Dis 2008; 14: 902–908.
- 7.↑
Skehel JJ, Wiley DC. Receptor binding and membrane fusion in virus entry: the influenza hemagglutinin. Annu Rev Biochem 2000; 69: 531–569.
- 8.↑
Von Grotthuss M, Rychlewski L. Influenza mutation from equine to canine. Science 2006; 311: 1241–1242.
- 9.
Plotkin JB, Dushoff J. Codon bias and frequency-dependent selection on the hemagglutinin epitopes of influenza A virus. Proc Natl Acad Sci USA 2003; 100: 7152–7157.
- 10.
Verhoeyen M, Fang R, Jou WM, et al. Antigenic drift between the haemagglutinin of the Hong Kong influenza strains A/Aichi/2/68 and A/Victoria/3/75. Nature 1980; 286: 771–776.
- 11.↑
Shakin-Eshleman SH, Spitalnik SL, Kasturi L. The amino acid at the X position of an Asn-X-Ser sequon is an important determinant of N-linked core-glycosylation efficiency. J Biol Chem 1996; 271: 6363–6366.
- 12.
Ye Y, Si ZH, Moore JP, et al. Association of structural changes in the V2 and V3 loops of the gp120 envelope glycoprotein with acquisition of neutralization resistance in a simian-human immunodeficiency virus passaged in vivo. J Virol 2000; 74: 11955–11962.
- 13.
Zhang M, Gashen B, Blay W, et al. Tracking global patterns of N-linked glycosylation site variation in highly variable viral glycoproteins: HIV, SIV, and HCV envelopes and influenza hemagglutinin. Glycobiology 2004; 14: 1229–1246.
- 14.
Ito T. Interspecies transmission and receptor recognition of influenza A viruses. Microbiol Immunol 2000; 44: 423–430.
- 15.
Suzuki Y, Ito T, Suzuki T, et al. Sialic acid species as a determinant of the host range of influenza A viruses. J Virol 2000; 74: 11825–11831.
- 16.
Busch MG, Bateman AC, Landolt GA, et al. Identification of amino acids in the HA of H3 influenza viruses that determine infectivity levels in primary swine respiratory epithelial cells. Virus Res 2008; 133: 269–279.
- 17.
Matrosovich MN, Matrosovich TY, Gray T, et al. Human and avian influenza viruses target different cell types in cultures of human airway epithelium. Proc Natl Acad Sci USA 2004; 101: 4620–4624.
- 18.
Thompson CI, Barclay WS, Zambon MC, et al. Infection of human airway epithelium by human and avian strains of influenza A virus. J Virol 2006; 80: 8060–8068.
- 19.
Wan H, Perez DR. Amino acid 226 in the hemagglutinin of H9N2 influenza viruses determines cell tropism and replication in human airway epithelial cells. J Virol 2007; 81: 5181–5191.
- 20.↑
Nelson KM, Schram BR, McGregor MW, et al. Local and systemic isotype-specific antibody responses to equine influenza virus infection versus conventional vaccination. Vaccine 1998; 16: 1306–1313.
- 21.↑
Landolt GA, Karasin AI, Phillips L, et al. Comparison of the pathogenesis of two genetically different H3N2 influenza A viruses in pigs. J Clin Microbiol 2003; 41: 1936–1941.
- 22.↑
Kaverin NV, Webster RG. Impairment of multicycle influenza virus growth in Vero (WHO) cells by loss of trypsin activity. J Virol 1995; 69: 2700–2703.
- 23.
Bals R, Beisswenger C, Blouquit S, et al. Isolation and air-liquid interface culture of human large airway and bronchiolar epithelial cells. J Cyst Fibros 2004; 3 (suppl 2):49–51.
- 24.↑
Soboll G, Hussey S, Minke JM, et al. Onset and duration of immunity to equine influenza virus resulting from canarypoxvectored (ALVAC) vaccination. Vet Immunol Immunopathol 2010; 135: 100–107.
- 25.
Landolt GA, Karasin AI, Hofer C, et al. Use of real-time reverse transcriptase polymerase chain reaction assay and cell culture methods for detection of swine influenza A viruses. Am J Vet Res 2005; 66: 119–124.
- 26.↑
Chambers TM, Holland RE, Tudor LR, et al. A new modified live equine influenza virus vaccine: phenotypic stability, restricted spread and efficacy against heterologous virus challenge. Equine Vet J 2001; 33: 630–636.
- 27.↑
Mumford JA, Hannant D, Jessett DM. Experimental infection of ponies with equine influenza (H3N8) viruses by intranasal inoculation or exposure to aerosols. Equine Vet J 1990; 22: 93–98.
- 28.
Ito T, Couceiro JN, Kelm S, et al. Molecular basis for the generation in pigs of influenza A viruses with pandemic potential. J Virol 1998; 72: 7367–7373.
- 29.
Bosch FX, Garten W, Klenk HD, et al. Proteolytic cleavage of influenza virus hemagglutinins: primary structure of the connecting peptide between HA1 and HA2 determines proteolytic cleavability and pathogenicity of avian influenza viruses. Virology 1981; 113: 725–735.
- 30.
Rott R, Reinacher M, Orlich M, et al. Cleavability of hemagglutinin determines spread of avian influenza viruses in the chorioallantoic membrane of chicken embryo. Arch Virol 1980; 65: 123–133.
- 31.↑
Vigerust DJ, Ulett KB, Boyd KL, et al. N-linked glycosylation attenuates H3N2 influenza viruses. J Virol 2007; 81: 8593–8600.
