ABSTRACT
Canine influenza is a respiratory disease of dogs caused by canine influenza virus (CIV). CIV subtypes responsible for influenza in dogs include H3N8, which originated from the transfer of H3N8 equine influenza virus to dogs; and the H3N2 CIV, which is an avian-origin virus that adapted to infect dogs. Influenza infections are most effectively prevented through vaccination to reduce transmission and future infection. Currently, only inactivated influenza vaccines (IIVs) are available for the prevention of CIV in dogs. However, the efficacy of IIVs is suboptimal, and novel approaches are necessary for the prevention of disease caused by this canine respiratory pathogen. Using reverse genetics techniques, we have developed a live-attenuated CIV vaccine (LACIV) for the prevention of H3N8 CIV. The H3N8 LACIV replicates efficiently in canine cells at 33°C but is impaired at temperatures of 37 to 39°C and was attenuated compared to wild-type H3N8 CIV in vivo and ex vivo. The LACIV was able to induce protection against H3N8 CIV challenge with a single intranasal inoculation in mice. Immunogenicity and protection efficacy were better than that observed with a commercial CIV H3N8 IIV but provided limited cross-reactive immunity and heterologous protection against H3N2 CIV. These results demonstrate the feasibility of implementing a LAIV approach for the prevention and control of H3N8 CIV in dogs and suggest the need for a new LAIV for the control of H3N2 CIV.
IMPORTANCE Two influenza A virus subtypes has been reported in dogs in the last 16 years: the canine influenza viruses (CIV) H3N8 and H3N2 of equine and avian origins, respectively. To date, only inactivated influenza vaccines (IIVs) are available to prevent CIV infections. Here, we report the generation of a recombinant, temperature-sensitive H3N8 CIV as a live-attenuated influenza vaccine (LAIV), which was attenuated in mice and dog tracheal, explants compared to CIV H3N8 wild type. A single dose of H3N8 LACIV showed immunogenicity and protection against a homologous challenge that was better than that conferred with an H3N8 IIV, demonstrating the feasibility of implementing a LAIV approach for the improved control of H3N8 CIV infections in dogs.
INTRODUCTION
Influenza A viruses (IAVs) are enveloped viruses that belong to the Orthomyxoviridae family and contain a genome that comprises eight single-stranded negative-sense RNA segments that encode 10 to 14 proteins (1). The hemagglutinin (HA) and the neuraminidase (NA) glycoproteins are the major antigenic determinants of IAV and are essential for receptor binding and fusion and virion release, respectively (2). IAV HA and NA glycoproteins within infected organisms and populations are driven to evolve antigenic variants via immunological pressure and in humans and some other hosts positive selection of viruses occurs gradually in a process known as antigenic drift (3). The antigenic diversity of glycoproteins is used to further classify IAVs, of which there are 18 HA and 11 NA subtypes (4, 5). In addition, antigenically distinct isolates can also exist within the same subtype, referred to as drifted variants. IAVs exist mainly in the wild aquatic fowl reservoir (6–9), and only a small number of mammalian hosts are currently recognized as sustaining transmission of IAVs.
Canine influenza is a contagious respiratory disease of dogs caused by two IAVs: the H3N8 equine-origin influenza virus that transferred to dogs in the United States around 1999 (10) and the avian virus-like H3N2 that transferred to dogs in Asia around 2005 (11). In 2015 an outbreak of H3N2 canine influenza virus (CIV) occurred in the United States that was due to a virus similar to those detected in dogs in Asia (12). The H3N2 CIV has also been isolated from cats in a shelter in South Korea (13, 14). These CIVs represent new threats to canine health in the United States and worldwide, since the virus may be spread through the racing track circuit, as was the case of the H3N8 strain (10, 15), while both viruses are spread widely within and among animal shelters and kennels (10, 16, 17). CIVs are still relatively new viruses and because of the low levels of infection and immunity among the broader population most dogs are susceptible to infection. Most dogs infected by CIVs show only a mild respiratory illness, but severe outcomes are also observed (18).
The recent emergence of CIVs (H3N8 and H3N2 CIVs) has increased the host range of IAVs. The continuous circulation of CIVs in dog populations creates opportunities for exposure of humans and other animals. Since dogs are susceptible to mammalian (equine-origin H3N8 CIV) and avian (avian-origin H3N2 CIV) IAVs, they may have the potential to act as “mixing vessel” hosts for new IAV strains with potential for human infection. Reassortments between H3N2 CIVs and human pandemic H1N1 IAV have been reported (19, 20), and the introduction of novel, antigenically distinct glycoproteins (HA and NA) into the backbones of human IAVs have been associated with human pandemics (21).
Vaccination is accepted as an effective strategy for the prevention of influenza infections (22, 23). To date, three types of influenza virus vaccines have been approved by the U.S. Food and Drug Administration for human use: recombinant viral HAs, inactivated influenza vaccines (IIVs), and live-attenuated influenza vaccines (LAIVs) (22, 24–27). In dogs, only IIV against both H3N8 and H3N2 CIVs are commercially available. However, we have recently reported the generation of recombinant H3N8 CIVs containing truncated or a deleted nonstructural 1 (NS1) protein as potential LAIVs candidates for the treatment of H3N8 CIV infections (28).
IIVs are administered intramuscularly and elicit humoral immunity by inducing the production of neutralizing antibodies that target epitopes on HA (26, 29). On the other hand, LAIVs more closely mimic the natural route of viral infection and elicit both cellular and humoral immune responses (24), providing better immunogenicity and protection (22, 28, 30).
In mammals IAV is a respiratory pathogen that replicates in the cooler (33°C) upper respiratory tract, in addition to replicating in the warmer (37°C) conditions of the lower respiratory tract (31). This temperature difference has allowed for the development of cold-adapted (ca), temperature-sensitive (ts), attenuated (att) viruses that replicate in the upper respiratory tract but do not damage the lower respiratory tract due to the elevated temperatures restricting replication (32). For human viruses these ca, ts, and att properties have been mapped to five amino acid residues located in three viral proteins of A/Ann Arbor/6/60 H2N2 (A/AA/6/60): polymerase basic 2 (PB2) N265S; polymerase basic 1 (PB1) K391E, D581G, and A661T; and nucleoprotein (NP) D34G (33, 34). The mechanisms of attenuation are not fully understood but most likely involve multiple steps in the replication cycle of the virus (32). Importantly, when the ts signature of A/AA/6/60 was introduced into influenza A/Puerto Rico/8/34 H1N1 (PR8) or A/California/04/09 H1N1 (pH1N1) viruses, a similar ts phenotype was observed in tissue culture cells and in the mouse model of infection (35–37).
In order to develop a LAIV for the treatment of CIV H3N8 infections, we introduced the ts, ca, att mutations identified in the A/AA/6/60 LAIV into CIV H3N8 (referred to henceforth as LACIV H3N8) using reverse genetics (38). LACIV H3N8 replicated efficiently in vitro at 33°C but not at 37 or 39°C. Compared to CIV H3N8 wild-type (WT), LACIV H3N8 was attenuated ex vivo and in vivo but was able to induce protective immunity in mice against H3N8 WT upon a single intranasal (i.n.) dose, demonstrating its feasibility as a safe, immunogenic, and protective LAIV candidate.
RESULTS
Generation and characterization of H3N8 LACIV.We introduced four ts mutations identified in previous studies into the PB2 and PB1 genes of H3N8 CIV (33, 39) (Fig. 1). No mutation was introduced into the viral NP since H3N8 CIV NP already contains a G at position 34.
Effect of temperature on the polymerase activity of H3N8 LACIV. (A) Schematic representation of segments 1 (PB2) and 2 (PB1) of WT (black) and LACIV (gray) H3N8 CIV. Amino acid substitutions N265S (PB2) and K391E, E581G, and A661T (PB1) to generate the H3N8 LACIV are indicated. (B) Minigenome activity. MDCK cells (12-well plate format, 3 × 105 cells/well, triplicates) were transiently cotransfected with 250 ng of ambisense pDZ expression plasmids encoding the minimal requirements for viral genome replication and gene transcription (PB2, PB1, PA, and NP), together with 500 ng of a vRNA-like expression plasmid encoding Gaussia luciferase (Gluc) under the control of the canine polymerase I promoter (cpPol-I Gluc), and 100 ng of a pCAGGS Cypridina luciferase (Cluc) plasmid to normalize transfection efficiencies. After transfection, cells were placed at 33, 37, or 39°C, and viral replication and transcription was evaluated 24 h later by luminescence (Gluc). Gluc activity was normalized to that of Cluc. The data represent means and SD. Normalized reporter expression is relative to that in the absence of pDZ NP plasmid. Data were represented as relative activity considering WT H3N8 polymerase activity at each temperature as 100%. *, P < 0.05 (Student t test).
To determine whether the mutations introduced into the PB2 and PB1 genes confer a ts phenotype to the H3N8 CIV polymerase, we performed a minigenome assay. Both WT and LACIV H3N8 resulted in similar Gaussia luciferase (Gluc) expression levels at 33°C (Fig. 1B), but Gluc expression was reduced at higher temperatures (37 and 39°C) in cells transfected with the H3N8 LACIV plasmids. This shows that those mutations resulted in a ts phenotype when introduced in the H3N8 CIV, as previously described for A/AA/6/60 H2N2 and other influenza viruses (35–37).
We next generated an H3N8 LACIV using plasmid-based reverse genetic approaches (40, 41) and evaluated the viral replication kinetics in MDCK cells infected at a low (0.001) multiplicity of infection (MOI) and compared them to the WT H3N8 CIV (Fig. 2). At 33°C, both WT and LACIV H3N8 grew with the same kinetics and reached similar high titers (107 FFU/ml) at 48 to 72 h postinfection (p.i.) (Fig. 2A). At higher temperatures (37 and 39°C), the WT H3N8 CIV replicated at levels similar to those observed at 33°C, while H3N8 LACIV titers were reduced ca. 2 to 3 logs at 37°C (Fig. 2B) or was not detected at 39°C (Fig. 2C). These results demonstrate that mutations introduced in PB2 and PB1 conferred a ts phenotype to H3N8 CIV.
Characterization of H3N8 LACIV in vitro: MDCK cells (12-well plate format, 3 × 105 cells/well, triplicates) were infected (MOI of 0.001) with WT (black diamonds) and LACIV (gray squares) H3N8 CIVs and incubated at 33°C (A), 37°C (B), and 39°C (C). TCS were collected at 12, 24, 48, 72, and 96 h p.i., and the viral titers were determined by immunofocus assay (FFU/ml). Data represent the means and SD of the results determined in triplicate. Dotted lines indicate the limit of detection (200 FFU/ml). *, P < 0.05 (Student t test).
LACIV H3N8 is attenuated in vivo in mice.Since the H3N8 LACIV presented defects in replication at higher (37 and 39°C) temperatures, we next investigated whether H3N8 LACIV was also attenuated in mice. No signs of infection were detected after infection with WT H3N8 CIV. Therefore, CIV titers in the lungs of infected (105 PFU) mice were determined on days 2 (n = 3) and 4 (n = 3) p.i. and used as a measure of viral attenuation (Fig. 3). Notably, virus titers in the lungs were only detected in mice inoculated with WT H3N8 CIV, and no virus was detected in mice infected with the H3N8 LACIV. These results indicate that H3N8 LACIV is also attenuated in vivo.
Attenuation of H3N8 LACIV in vivo. Female 5- to-7-week-old C57BL/6 WT mice (n = 6) were infected i.n. with 105 PFU of WT or LACIV H3N8 CIVs. Three mice were sacrificed at days 2 (black) and 4 (gray) p.i., and the lungs were harvested for virus titrations using an immunofocus assay (FFU/ml). Data represent the means and SD. Dotted line indicate limit of detection (200 FFU/ml). ND, virus not detected.
Intranasal vaccination with H3N8 LACIV induces protective immunity against WT H3N8 CIV challenge.To evaluate the immunity generated by the H3N8 LACIV, mice (n = 6) were vaccinated i.n. with 103 PFU of H3N8 WT or LACIV, mock vaccinated with phosphate-buffered saline (PBS), or vaccinated intramuscularly (i.m.) with 100 μl of Nobivac, a commercial IIV against H3N8 CIV (Fig. 4). Humoral immune responses were evaluated in mice sera collected 2 weeks later (Fig. 4A). Total H3N8 CIV antibody responses were characterized by enzyme-linked immunosorbent assay (ELISA) using cell lysates from mock- or H3N8 CIV-infected MDCK cells (41). Mice vaccinated with the H3N8 LACIV elicited high serum IgG titers against parental H3N8 CIV, whereas the antibody titers of mice vaccinated with Nobivac were lower than those in the H3N8 LACIV or WT vaccinated mice (Fig. 4A). In addition, hemagglutination inhibition (HAI) assays were performed to examine the presence of anti-HA neutralizing antibodies on sera from vaccinated mice (Fig. 4B), showing that HAI titers against CIV H3N8 were higher in mice vaccinated with the H3N8 LACIV than those observed in mice vaccinated with the H3N8 IIV.
Immunogenicity of H3N8 LACIV: Female 5- to-7-week-old C57BL/6 WT mice (n = 6) were vaccinated i.n. with 103 PFU of WT or LACIV H3N8 CIVs. Mice mock (PBS) vaccinated or vaccinated i.m. with 100 μl of an H3N8 CIV IIV (Nobivac) were used as internal controls. (A) Induction of humoral responses. At 14 days postvaccination, mice were bled, and sera were evaluated for the presence of total IgG antibodies against H3N8 CIV proteins using cell extracts of MDCK-infected cells by ELISA. MDCK mock-infected cell extracts were used to evaluate the specificity of the antibody response. OD, optical density. Data represent the means ± the SD of the results for six individual mice. *, Nobivac versus LACIV; **, WT versus LACIV; ***, WT versus Nobivac (P < 0.05 using the Student t test). (B) HAI titers. Sera from immunized mice were evaluated by HAI using four HAU of WT H3N8 CIV and 2-fold serial dilutions of the indicated sera. ND, not detected. *, WT versus LACIV or Nobivac; or *, LACIV versus Nobivac (P < 0.05 using the Student t test).
To further examine the immunogenicity of LACIV, we evaluated whether the virus can induce a localized CD8 T cell response (Fig. 5). To this end, mice (n = 4) were immunized as described above and, at 10 days p.i., lung and spleen samples were collected and single-cell preparations made for flow cytometric analysis (42). Live CD8 T cells were further gated for a H3N8 CIV-specific population (42, 43). The results showed that vaccination with H3N8 LACIV induced elevated levels of NP and PA-specific lung CD8 T cells, similar to those induced by WT virus in both lungs (Fig. 5A) and the spleen (Fig. 5B). Importantly, animals vaccinated with Nobivac did not show a T cell response either in the lung (Fig. 5A) or the spleen (Fig. 5B), highlighting important differences in induced immunity between H3N8 LACIV and the H3N8 IIV.
CD8 T cell response induced by H3N8 LACIV. Female 5- to-7-week-old C57BL/6 WT mice (n = 4) were vaccinated i.n. with 103 PFU of WT or LACIV H3N8 CIVs. Mice mock (PBS) vaccinated or vaccinated i.m. with 100 μl of an H3N8 CIV IIV (Nobivac) were used as internal controls. At 10 days p.i., the lungs (A) and spleen (B) were extracted, and the cells were prepared for flow cytometry. Live CD8 T cells specific for NP or PA tetramers were counted. The data represent the means ± the SD of the results for four individual mice. *, WT versus LACIV, Nobivac, or PBS; or *, LACIV versus Nobivac or PBS (P < 0.05 using the Student t test).
We next evaluated the ability of H3N8 LACIV to induce protective immunity. Mice (n = 6) were vaccinated i.n. with 103 PFU of H3N8 WT or LACIV, vaccinated i.m. with 100 μl of the IIV Nobivac, or mock vaccinated with PBS. Two weeks later mice were challenged with 105 PFU of WT H3N8 CIV and viral titers in the lungs of infected mice (n = 3/group) were determined 2 and 4 days after challenge (Fig. 6). Mock-vaccinated mice showed high viral titers in the lungs at days 2 and 4 p.i., whereas mice immunized with H3N8 WT CIV and with LACIV showed no detectable virus in the lungs at those times (Fig. 6). Mice vaccinated with the H3N8 IIV showed high viral titers at day 2 but no detectable virus at day 4 p.i. (Fig. 6).
Protection efficacy of H3N8 LACIV against homologous viral challenge. Female 5- to-7-week-old C57BL/6 WT mice (n = 6) were vaccinated i.n. with 103 PFU of WT or LACIV H3N8 CIVs. Mice mock (PBS) vaccinated or vaccinated i.m. with 100 μl of an H3N8 CIV IIV (Nobivac) were used as internal controls. At 15 days postvaccination, mice were challenged i.n. with 105 PFU of WT H3N8 CIV. To evaluate viral replication, mice were euthanized at days 2 (n = 3, black) and 4 (n = 3, gray) postchallenge, and the lungs were harvested, homogenized, and used to quantify viral titers by immunofocus assay (FFU/ml). The dotted line indicates the limit of detection (200 FFU/ml). ND, virus not detected. Data represent the mean ± the SD. *, P < 0.05 using the Student t test.
H3N8 LACIV is attenuated in canine tracheal explants compared to H3N8 WT CIV.To compare LACIV and WT H3N8 CIV pathogenicity and replication efficiency at the site of infection within dogs (Fig. 7), we inoculated dog tracheal explants with each virus and compared histological lesions (Fig. 7A), viral NP expression (Fig. 7B), changes in ciliary function (Fig. 7C), and viral replication (Fig. 7D) at different times (days 1, 3, and 5) p.i. The H3N8 WT CIV induced major histological changes in dog tracheal explants, with thinning and desquamation of the epithelium, loss of cilia (Fig. 5A), and significant reduction of ciliary function (Fig. 7C) between days 1 and 5 p.i. Histological damages induced by H3N8 LACIV were delayed and reduced compared to WT H3N8 CIV, since the epithelium maintained its normal thickness until day 3 p.i. (Fig. 7A), and the ciliary function (Fig. 7C) was only significantly reduced from day 3 p.i. Viral kinetics (Fig. 7D) and NP expression (Fig. 7B) were comparable between the two viruses, although only WT H3N8 CIV was detectable at day 1 p.i. (Fig. 7D). Overall, these results indicate that H3N8 LACIV pathogenicity is attenuated in canine tracheal explants compared to its WT counterpart.
Comparison of WT and LACIV H3N8 CIV infection phenotypes in canine tracheal explants. (A) Histological features of dog tracheal explants infected with 200 PFU of H3N8 WT CIV or H3N8 LACIV or mock-infected with infection media. Lesions are shown in sections stained with H&E. (B) CIV H3N8-infected cells were detected by immunostaining for the viral NP, and positive cells are stained in brown. For both panels A and B, pictures are representatives of three independent experiments, and the black horizontal bars represent 20 μm. (C) Graphical representation of the bead clearance average time of CIV- or mock-infected dog tracheal explants for three independent experiments. Data represent the means ± the SD. ns, P > 0.05 (D1, LACIV versus Mock); *, P < 0.05 (D1, WT versus Mock); **, P < 0.01 (D3, LACIV versus Mock); ***, P < 0.001 (D3, WT versus Mock); ****, P < 0.0001 (D5, LACIV and WT versus Mock). (D) Average viral replication of H3N8 LACIV and H3N8 WT CIV in canine tracheal explants from three independent experiments. Data represent the means ± the SD. The dotted line indicates the limit of detection (20 FFU/ml).
H3N8 LACIV provides limited protection against heterologous H3N2 CIV challenge.We next evaluated whether H3N8 LACIV could induce protective immunity against a heterologous H3N2 CIV challenge (Fig. 8). Mice (n = 6) were vaccinated i.n. with 103 PFU of WT or LACIV H3N8, mock vaccinated with PBS, or vaccinated i.m. with 100 μl of the H3N8 IIV Nobivac or a commercial H3N2 IIV (Zoetis). Antibodies against H3N2 CIV were evaluated by ELISA using cell lysates from mock- or H3N2 CIV-infected MDCK cells as antigens (Fig. 8A). In addition, antibodies against the HA (Fig. 8B) or NA (Fig. 8C) proteins of H3N2 CIV were also evaluated. When the cell lysate was used as an antigen, antibodies against H3N2 CIV were detected in sera of mice vaccinated with WT H3N8 CIV and, to a lower extent, in mice vaccinated with H3N8 LACIV, although the levels were lower than those against H3N8 CIV (Fig. 8A). Similarly, using the recombinant proteins (HA or NA) as antigens to perform the ELISA, only antibodies against H3N2 CIV HA were detected in samples of animals immunized with WT H3N8 CIV (Fig. 8B). However, no antibodies were detected against the NA protein (Fig. 8C). No detectable IgG antibodies against H3N2 CIV were detected in mice vaccinated with the H3N8 IIV Nobivac either using the cell extracts (Fig. 8A) or the recombinant proteins (Fig. 8B and C). The H3N2 CIV IIV induced higher IgG antibodies against H3N2 CIV in all cases, as expected (Fig. 8A to C).
Immunogenicity and protection efficacy of H3N8 LACIV against heterologous H3N2 CIV challenge. Female 5- to-7-week-old C57BL/6 WT mice were vaccinated i.n. with 103 PFU of WT and LACIV H3N8 CIVs. Mice mock (PBS) vaccinated or vaccinated i.m. with 100 μl of the H3N8 (Nobivac) and an H3N2 CIV IIV (Zoetis) were used as internal controls. (A to C) Antibody cross-reactivity against the heterologous CIV H3N2. At 14 days postvaccination, mice were bled, and sera were evaluated by ELISA for total IgG antibodies against H3N2 CIV proteins using cell extracts of MDCK-infected cells (A). Mock-infected MDCK cell extracts were used to evaluate the specificity of the antibody response. OD, optical density. Data represent the means ± the SD of the results for six individual mice. *, Nobivac versus LACIV; **, WT versus LACIV; or ***, WT versus Nobivac (P < 0.05 using the Student t test). Specific antibody response against recombinant HA (B) and NA (C) proteins from H3N2 CIV were evaluated by ELISA. Data represent the means ± the SD of the results for pooled serum samples. *, WT versus LACIV (P < 0.05 using the Student t test). (D) Protection efficacy of H3N8 LACIV against heterologous H3N2 CIV challenge. At 15 days postvaccination, mice were challenged i.n. with 105 PFU of WT H3N2 CIV. To evaluate WT H3N2 CIV replication, mice were sacrificed at days 2 (n = 3, black) and 4 (n = 3, gray) postchallenge, and the lungs were harvested, homogenized, and used to evaluate the presence of virus by immunofocus assay (FFU/ml). The dotted line indicates the limit of detection (200 FFU/ml). ND, virus not detected. Data represent means ± the SD. *, P < 0.05 using Student t test.
The lower level of cross-reactive immunity against H3N2 CIV upon vaccination with the H3N8 LACIV was confirmed after challenge i.n. with H3N2 CIV 2 weeks postvaccination (Fig. 8D). Mock-vaccinated mice showed high H3N2 CIV titers that were undistinguishable from those seen in animals vaccinated with the H3N8 CIV IIV Nobivac. In contrast, mice vaccinated with the H3N2 CIV IIV showed reduced or undetectable titers at days 2 and 4 postchallenge, respectively. Although we observed similar H3N2 CIV titers at day 2 postchallenge, viral titers at day 4 p.i. in mice vaccinated with the H3N8 LACIV were ∼100 times lower than those obtained in the mock-vaccinated group. These results suggest that H3N8 LACIV can induce limited cross-reactive immune responses and heterologous protection, most probably mediated by a T cell response, against H3N2 CIV but that the efficacy is lower than that obtained with the H3N2 IIV.
DISCUSSION
We report here a novel LAIV prepared using plasmid-based reverse genetics techniques, which may be used for the prevention of H3N8 CIV. We generated a recombinant H3N8 CIV containing the mutations responsible for the ts phenotype of the human A/AA/6/60 H2N2 LAIV, resulting in ts H3N8 CIV (LACIV) that was highly attenuated in vivo and ex vivo compared to its WT counterpart. Our H3N8 LACIV was able to confer, upon a single i.n. immunization in mice, complete protection against challenge with WT H3N8 CIV. This demonstrates the feasibility of using the ts H3N8 LACIV as a safe, immunogenic, and protective vaccine to control H3N8 CIV in dogs. However, the H3N8 LACIV showed limited immunogenicity and protection efficacy against the heterologous H3N2 CIV, suggesting the need of a different LACIV for the treatment and control of H3N2 CIV.
The ts, ca, and att A/AA/6/60 H2N2 LAIV has been licensed for human use since 2003, and is used as a master donor virus for the generation of both seasonal or potentially pandemic human LAIV by creating reassortant viruses containing the six H2N2-derived internal viral RNA (vRNA) segments (PB2, PB1, PA, NP, M, and NS) and the two glycoprotein-encoding vRNAs (HA and NA) from a virus that antigenically matches the strains predicted to circulate in the upcoming influenza season (seasonal vaccine) or potentially pandemic strains (pandemic vaccine) (44, 45). Five mutations (PB2 N265S; PB1 K391E, D581G, and A661T; and NP D34G) are responsible for the ts phenotype of the A/AA/6/60 H2N2 LAIV, and those mutations also impart a strong ts phenotype and attenuation to other viral strains, such as PR8 (36, 46) and pH1N1 (37).
Intranasal immunization is a desirable delivery method for providing optimal immunity to IAV because it leads to the generation of a mucosal immune responses, creating an immune barrier at the site of potential infection (47), as well as systemic humoral responses, cellular immunity (45, 48–51). Similar to infection with WT IAV, LAIV immunization also leads to recruitment of influenza-specific CD8 T cells to the lungs (42, 50, 52–54), which provides immunity against heterologous influenza challenge (42, 52). Thus, a LAIV rather than an IIV is desired for the control of IAV infections (22, 23, 27, 55).
Since the emergence of H3N8 CIV in 1999 in the United States and the H3N2 CIV in Asia (in 2005) and the United States (in 2015), CIVs have been maintained mainly in animal shelters and kennels as those populations allow ready transmission of the virus (10, 16, 17, 56, 57). Strains of the H3 subtype of IAV infect a number of mammalian hosts, including humans, pigs, horses, dogs, cats, and seals, as well as poultry (6, 11, 57–61). Naturally occurring H3N1 virus carrying the HA gene of an avian-like H3N2 CIV and the other seven segments of the human pH1N1 has been reported in dogs in Korea (20), suggesting that dogs could act as an intermediate host for genetic reassortment of IAV, including those that might infect humans. However, no transmission of H3N8 or H3N2 CIV transmission from dogs to humans has been reported to date. It may be possible to eradicate both H3N8 and H3N2 CIVs from the dog population through infection control, as well as by using vaccination approaches. CIV LAIVs represent a better option for efficient CIV control and probably eradication since they induce better and faster antiviral immunity.
The H3N8 LACIV generated here was ts and attenuated in mice and induced protective immune responses against challenge with homologous H3N8 CIV WT, and the responses were stronger than those obtained with a commercial H3N8 CIV IIV. Its replication and pathogenesis were also restricted in canine tracheal explants, and we are currently evaluating the safety, immunogenicity, and protection efficacy of our H3N8 LACIV in dogs, the real target population.
To achieve protection in dogs, animals are vaccinated i.m. with 1 ml of the CIV H3N8 IIV (Nobivac) using a two-dose regime (62). The average weight of a male or female mouse is ∼20 g, whereas the average weight of a female or male beagle dog is ∼10 kg. Thus, in principle mice should be vaccinated with 500 times less (just 2 μl) of the CIV H3N8 IIV than dogs. However, in our experiments, mice were immunized i.m. with 100 μl of the CIV H3N8 IIV, a 50× higher dose than on a weight basis. In addition, to evaluate the amount of antigen in the CIV H3N8 IIV, we performed an hemagglutination assay using the commercial CIV H3N8 IIV or our CIV H3N8 LAIV. The assay showed that mice vaccinated with the CIV H3N8 IIV were inoculated with approximately 106 viral particles/mouse of inactivated virus, 1,000 times more than the dose (i.e., 103) of the H3N8 CIV LAIV. The LAIV still elicited better antibody responses and protection compared to the CIV H3N8 IIV.
Segment eight of IAV encodes the NS1 viral protein, which controls the adaptive immune responses by inhibiting the interferon-antiviral response of the host (63). Therefore, NS1 is a virulence factor that offers an attractive target for the development of attenuated viruses as LAIVs. In fact, IAVs harboring a truncated-NS1 have been shown as promising vaccines candidates (28, 64–71). In a recent work, we have generated recombinant H3N8 CIVs containing truncated (NS1-126, NS1-99, or NS1-73) or deleted (ΔNS1) NS1 proteins and tested them as potential LAIVs against CIV H3N8 infections (28). The recombinant NS1 mutant H3N8 CIVs were attenuated in vivo (mice) and in vitro (dog tracheal explants) but were able to confer complete protection against challenge with WT CIV H3N8 (28). Moreover, the immunogenicity and protection efficacy of our NS1 mutant H3N8 CIVs was also better than that observed with an H3N8 CIV IIV (28). Future research should determine which one of these attenuation strategies (NS1 mutant or ts H3N8 CIVs) is more efficient for their implementation as a LAIV for the prevention and control of H3N8 CIV in dogs.
MATERIALS AND METHODS
Cells and viruses.Human embryonic kidney 293T cells (293T; ATCC CRL-11268) and Madin-Darby canine kidney cells (MDCK; ATCC CCL-34) were grown at 37°C with 5% CO2 in Dulbecco modified Eagle medium (DMEM; Mediatech, Inc.) supplemented with 10% fetal bovine serum (FBS) and 1% PSG (penicillin, 100 U/ml; streptomycin 100 μg/ml; l-glutamine, 2 mM) (40).
Recombinant wild-type (WT) and live-attenuated (LACIV) H3N8 CIVs were generated using A/canine/NY/dog23/2009 H3N8 plasmid-based reverse genetics techniques (72) and grown in MDCK cells at 33°C. Influenza A/Ca/IL/41915/2015 H3N2 was also grown in MDCK cells at 33°C. For infections, virus stocks were diluted in PBS, 0.3% bovine albumin BA, and 1% PS (PBS/BA/PS). After viral infections, cells were maintained in DMEM with 0.3% BA, 1% PSG, and 1 μg/ml TPCK (tolylsulfonyl phenylalanyl chloromethyl ketone)-treated trypsin (Sigma) (38).
Plasmids.To generate H3N8 LACIV, the PB2 and PB1 genes were subcloned in a pUC19 plasmid (New England BioLabs), and then ts mutations (PB2 N265S and PB1 K391E, D581G, and A661T) were introduced by site-directed mutagenesis. The presence of introduced mutations was confirmed by sequencing. Mutated PB2 and PB1 viral segments were subcloned from pUC19 into the ambisense pDZ plasmid for virus rescue. To test the ability of WT and LACIV H3N8 polymerases to replicate and transcribe at different temperatures (33, 37, and 39°C) using a minigenome assay, we engineered a pPolI plasmid containing the canine RNA polymerase I (Pol-I) promoter and the mouse Pol-I terminator separated by SapI endonuclease restriction sites (cpPol-I). The canine Pol-I promoter was obtained by PCR from MDCK cells (73). Then, the Gaussia luciferase (Gluc) reporter gene containing the 3′ and the 5′ noncoding regions of the viral NP (v)RNA was cloned into the cpPol-I plasmid to generate the cpPol-I Gluc reporter plasmid. All plasmids were confirmed by sequencing (ACGT Inc.). Primers for the generation of the different plasmids are available upon request.
Minigenome assays.MDCK cells (12-well plate format, 5 × 105 cells/well, triplicates) were cotransfected in suspension using Lipofectamine 2000 with 250 ng of each of the H3N8 WT or LACIV ambisense pDZ PB2, PB1, PA, and NP plasmids, together with 500 ng of the cpPol-I Gluc plasmid. A mammalian expression pCAGGS plasmid encoding Cypridina luciferase (Cluc; 100 ng) was also included to normalize transfection efficiencies (74). Cells transfected in the absence of the pDZ NP plasmid were used as negative control. At 24 h posttransfection, the Gluc and Cluc expression levels were determined using a luciferase assay kit (New England BioLabs) and quantified with a Lumicount luminometer (Packard). The fold induction over the level of induction for the negative control (the absence of NP) was determined. The mean values and standard deviations (SD) were calculated and statistical analysis was performed using a two-tailed Student t test using Microsoft Excel software.
Virus rescue.Virus rescues were performed as previously described (40, 75). Briefly, cocultures (1:1) of 293T/MDCK cells (6-well plate format, 106 cells/well, triplicates) were cotransfected in suspension, using Lipofectamine 2000 (Invitrogen), with 1 μg of the eight-ambisense H3N8 WT CIV (pDZ-PB2, -PB1, -PA, -HA, -NP, -NA, -M, and –NS) plasmids. To rescue the H3N8 LACIV, WT PB2 and PB1 pDZ plasmids were substituted by those containing PB2 and PB1 H3N8 LACIV. At 12 h posttransfection, the transfection medium was replaced with postinfection (p.i.) medium containing DMEM supplemented with 0.3% BSA, 1% PSG, and 0.5 μg/ml TPCK-treated trypsin (Sigma). Tissue culture supernatants (TCS) were collected 3 days posttransfection, clarified, and used to infect fresh monolayers of MDCK cells (6-well plate format, 106 cells/well, triplicates). At 3 days p.i., recombinant viruses were plaque purified and scaled up using MDCK cells at 33°C (40). Virus stocks were titrated by standard plaque assay (PFU/ml) in MDCK cells at 33°C (40).
Virus growth kinetics.Multicycle growth analyses were performed by infecting confluent monolayers of MDCK cells (12-well plate format, 5 × 105 cells/well, triplicates) at an MOI of 0.001. Viral titers in TCS collected at various times p.i. were determined by immunofocus assay (fluorescent forming units [FFU]/ml) in MDCK cells as previously described (40). Briefly, confluent MDCK cells (96-well plate format, 5 ×104 cells/well, triplicates) were infected with 10-fold serial dilutions of H3N8 WT or LACIV. At 12 h p.i., the cells were fixed and permeabilized (4% formaldehyde and 0.5% Triton X-100 in PBS) for 15 min at room temperature. After being washed with PBS, the cells were incubated in blocking solution (2.5% BSA in PBS) for 1 h at room temperature and then incubated with 1 μg/ml of an anti-NP monoclonal antibody (HB-65; ATTC) for 1 h at 37°C. After a washing step with PBS, the cells were incubated with FITC-conjugated secondary anti-mouse antibody (Dako) for 1 h at 37°C. The mean values and SD were calculated using Microsoft Excel software.
Animal experiments.Adult (5- to 7-week-old) female WT C57BL/6 mice were purchased from the National Cancer Institute (NCI) and maintained in the animal care facility at the University of Rochester under specific-pathogen-free conditions. Animal experiments were approved by the University Committee of Animal Resources and complied with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Research Council (76). Mice were anesthetized intraperitoneally with 2,2,2-tribromoethanol (Avertin; 240 mg/kg [body weight]) and then inoculated i.n. with 30 μl of the indicated amounts of H3N8 WT or LACIV or H3N2 WT. Alternatively, 100 μl of a commercially available inactivated H3N8 CIV vaccine (Nobivac; Merck Animal Health) or inactivated H3N2 CIV vaccine (Zoetis) were inoculated i.m. Virus replication was determined by measuring viral titers in the lungs of infected mice at the indicated days p.i. To that end, three mice in each group were euthanized by administration of a lethal dose of Avertin and exsanguination, and the lungs were collected and homogenized. Virus titers were determined by immunofocus assay (FFU/ml) as indicated above. Mouse sera were collected by submandibular bleeding 24 h prior to viral challenges and evaluated for the presence of influenza virus antibodies by ELISAs and neutralizing antibodies by HAI assays.
Evaluation of T cells response in lung and spleen.Mice (n = 4) were immunized as described above: 103 PFU of H3N8 WT or LACIV, mock vaccinated with PBS, or vaccinated with 100 μl of Nobivac (IIV).
Cellular preparations.At 10 days p.i., the lungs and spleens were perfused with PBS, removed, and separated into right and left lobes. Lung tissue was dissociated in C tubes by the GentleMACS (Miltenyi Biotek) using the Lung01 program. Samples were incubated in 5 ml (2 μg/ml) of Collagenase II in RPMI plus 8% FBS at 37°C for 30 min, with gentle agitation every 10 min. After digestion, samples were further dissociated using the Heart01 program. Cell suspensions were filtered through 70-μm-pore size filters prior to 75:40 Percoll (GE Healthcare) discontinuous gradient separation. The top layer, containing fat and other debris, was removed by aspiration. The cell layer was harvested and washed prior to counting and staining. Single-cell suspensions were prepared from collected spleens by disruption in RPMI plus 8% FBS. Counting was achieved through trypan blue exclusion on a hemocytometer.
Flow cytometry.Single cell suspensions were stained in PBS containing 1% FBS, purified CD16/32 (clone 2.4G2), NP and PA tetramers (43), and the following antibodies: TCRβ-PerCPCy5.5, CD8α-FITC, CD4-BV421, CD44-APCCy7, and CD62L-PECy7. Cells were subsequently stained for viability using Live/Dead Aqua (Invitrogen). All antibodies were obtained from eBioscience, BD Biosciences, or BioLegend. PA and NP tetramers were obtained from the NIH tetramer core facility (Atlanta, GA). Cells were analyzed by an LSRII (BD Biosciences) in the University of Rochester Flow Cytometry core facility and analyzed using FlowJo software (Tree Star).
ELISAs.ELISAs were performed as previously described (40) by coating 96-well plates at 4°C for 16 h with lysates from mock- or H3N8 or H3N2 WT CIV-infected MDCK cells or with H3N2 CIV HA (250 ng per well [IRR catalog no. FR-1478]) or NA (250 ng per well [IRR catalog no. FR-1479]). After blocking with 1% BSA for 1 h at room temperature, the plates were incubated with 2-fold serial dilutions (starting dilution of 1:50) of mouse sera for 1 h at 37°C. After incubation, the plates were washed with H2O and incubated with a horseradish peroxidase-conjugated goat anti-mouse IgG (1:2,000; Southern Biotech) for 1 h at 37°C. Reactions were then developed with tetramethylbenzidine (TMB) substrate (BioLegend) for 10 min at room temperature, quenched with 2N H2SO4, and read at 450 nm (Vmax kinetic microplate reader; Molecular Devices).
HAI assays.To evaluate the presence of H3N8 CIV neutralizing antibodies, mous sera were treated with receptor-destroying enzyme (RDE; Denka Seiken) and heat inactivated for 30 min at 56°C. The sera were then serially 2-fold diluted (starting dilution of 1:50) in 96-well V-bottom plates and mixed 1:1 with 4 hemagglutinating units (HAU) of WT H3N8 CIV for 30 min at room temperature. The HAI titers were determined by adding 0.5% turkey red blood cells to the virus-antibody mixtures for 30 min on ice, as previously described (40). The geometric mean titers and SD from individual mice (n = 6) were calculated from the last well where hemagglutination was inhibited.
Canine tracheal explants preparation and virus titrations.Three dog tracheas were harvested from healthy beagles (Charles River Laboratories) that had been used as negative controls in unrelated studies. Briefly, tracheas were collected aseptically immediately upon euthanasia and transported in prewarmed medium as previously described (18). Tracheas were washed five times for a total period of 4 h and maintained at 33°C, 5% CO2, and 95% humidity between washes. The connective tissue was removed, and the trachea was then open lengthwise. Each tracheal ring was divided in four 0.25-cm2 explants and transferred with the epithelium facing upward onto an agarose plug covered by a sterile filter. The explants were kept for a total of 6 days at 33°C, 5% CO2, and 95% humidity.
Tracheal explants were infected after a period of 24 h postpreparation (designed as day zero) with 200 PFU of WT or LACIV H3N8 or mock infected with culture medium. Inoculated explants were sampled for virus quantification, bead clearance assays, and histology at days 0, 1, 3, and 5 p.i. Viral replication was evaluated by plaque assays on MDCK cells.
Estimation of bead clearance time.The ciliary function of tracheal explants was evaluated as previously described (18) by placing 5 μl of polystyrene microsphere beads (Polysciences, Northampton, United Kingdom) on the explant apical surfaces and measuring the time to displace the beads.
Histological analysis and immunohistochemistry.After collection, the explants were fixed in 10% buffered formalin for a minimum of 48 h before paraffin embedding and sectioning. Subsequently, 4-μm sections of paraffin embedded tissue were either stained with hematoxylin and eosin (H&E) for histopathological evaluation or immunostained for the viral NP using standard procedures as previously described (18). For NP immunostaining, the Dako supervision system was used according to the manufacturer's instructions, along with a monoclonal mouse anti-NP (clone HB65; dilution, 1:500). Slides were counterstained with Mayer's hematoxylin. Histological images were captured with cellD software (Olympus).
ACKNOWLEDGMENTS
This research was partially funded by the University of Rochester Technology Development Fund to L.M.-S. and C.R.P. C.C. was supported by the Horserace Betting Levy Board, and P.R.M. was supported by the Medical Research Council of the United Kingdom (grant MC_UU_120/14/9).
FOOTNOTES
- Received 8 November 2016.
- Accepted 5 December 2016.
- Accepted manuscript posted online 7 December 2016.
- Copyright © 2017 American Society for Microbiology.
