|Home | About | Journals | Submit | Contact Us | Français|
The clinical symptoms caused by infection with influenza A virus vary widely and depend on the strain causing the infection, the dose and route of inoculation, and the presence of preexisting immunity. In most cases, seasonal influenza A viruses cause relatively mild upper respiratory tract disease, while sometimes patients develop an acute severe pneumonia. Heterosubtypic immunity induced by previous infections with influenza A viruses may dampen the development of clinical symptoms caused by infection with influenza A viruses of another subtype, as is the case during influenza pandemics. Here we show that ferrets acquire protective immunity after infection of the upper respiratory tract with a seasonal influenza A(H3N2) virus against subsequent infection with influenza A(H1N1)pdm09 virus inoculated by the intranasal route. However, protective heterosubtypic immunity was afforded locally, since the prior infection with the A(H3N2) virus did not provide protection against the development of pneumonia induced after intratracheal inoculation with the A(H1N1)pdm09 virus. Interestingly, some of these animals developed more severe disease than that observed in naïve control animals. These findings are of interest in light of the development of so-called universal influenza vaccines that aim at the induction of cross-reactive T cell responses.
The spectrum of clinical symptoms caused by influenza virus infections varies widely and depends on the strain causing the infection, the route of inoculation, and the immune status of the host (1). Infections of humans with seasonal influenza A viruses of the H3N2 and H1N1 subtypes typically comprise upper respiratory tract (URT) infections and are responsible for annual outbreaks of influenza that are associated with excess morbidity and mortality in the population (2). Occasionally, antigenically distinct influenza viruses, including viruses of novel subtypes for which aquatic birds are the reservoir, cause zoonotic infections of humans (3). Since antibodies to these novel viruses are not present in the human population, these novel viruses may have pandemic potential. For example, highly pathogenic avian influenza viruses (HPAIV) of the H5N1 subtype have caused human cases of infection since 1997, often with a fatal outcome (4). It recently became clear that relatively few mutations are required for these viruses to become transmissible between mammals via respiratory droplets (5, 6), which underscores their pandemic potential. Another example is the swine-origin influenza viruses of the H1N1 subtype [A(H1N1)pdm09] that caused the pandemic of 2009. Antibodies to this virus were present only in the elderly (7), resulting in a considerable disease burden and mortality, especially in younger age groups (8).
HPAIV of the H5N1 subtype and A(H1N1)pdm09 viruses have other properties in common. Both are more pathogenic in ferrets than seasonal influenza viruses (9, 10), have a stronger tropism for the lower respiratory tract (LRT) than that of seasonal influenza viruses, and may cause severe pneumonia (10).
As indicated above, the immune status of the host can influence the pathogenesis and outcome of influenza virus infection. During the 2009 pandemic, most elderly individuals were protected because they had developed antibodies after infection with influenza A(H1N1) viruses that circulated before 1957 (7). These antibodies cross-reacted with the A(H1N1)pdm09 viruses. Apart from virus-neutralizing antibodies directed to the hemagglutinin, other arms of the immune system can also afford some protection against infection with influenza virus (11).
It has been well documented that infection with an influenza A virus of a certain subtype can induce protective immunity against infection with an influenza A virus of another subtype, i.e., so-called heterosubtypic immunity (12, 13). This type of immunity has been demonstrated mainly in animals, including mice (13–15), ferrets (16, 17), and guinea pigs (18), using various combinations of influenza A virus subtypes for priming and challenge infections. Heterosubtypic immunity can be long lasting, as demonstrated in ferrets (19). Furthermore, infection with unrelated respiratory viruses does not induce heterosubtypic immunity against influenza A virus infection, indicating that heterosubtypic immunity is dependent on virus-specific adaptive immune responses (20).
A prior infection with a seasonal human influenza A virus of the H3N2 subtype also afforded a certain degree of protection against infection with A(H1N1)pdm09 virus (17, 21, 22). Infection with an influenza A virus does not induce sterilizing immunity but can mitigate the clinical signs caused by a subsequent infection with a (pandemic) virus of another subtype. There is limited and circumstantial evidence that heterosubtypic immunity also exists in humans (23, 24). For example, it was found that during the pandemic of 1957, which was caused by an influenza A virus of the H2N2 subtype, individuals who had experienced an A(H1N1) virus infection in the previous years were less likely to develop severe disease (25). Various arms of the immune system contribute to heterosubtypic immunity (12). Virus-neutralizing antibodies directed against the globular head region of the influenza A virus hemagglutinin play a minor role, but it has been demonstrated that cross-reacting and cross-protecting antibodies against the stem region of the hemagglutinin can afford heterosubtypic immunity, especially against infection with an influenza A virus with a hemagglutinin of the same group (26–30). In addition, cross-protecting antibodies against the ectodomain of the M2 protein have been identified (31). Another correlate of protection in heterosubtypic immunity is the presence of virus-specific CD4+ and CD8+ T cells directed against shared epitopes located in the relatively conserved internal viral proteins, such as the nucleoprotein and the M1 protein (32, 33). Indeed, adoptive transfer of T cells induced by influenza A(H3N2) virus infection afforded recipient mice protection against A(H1N1)pdm09 virus (21). In contrast to antibodies against the stem region of the hemagglutinin, virus-specific CD4+ and CD8+ T cells directed against conserved inner proteins are able to afford protection against all subtypes of influenza A virus, regardless of the hemagglutinin group.
One outstanding issue that needs to be addressed is the extent to which heterosubtypic immunity induced by previous infection with an influenza A virus that has a hemagglutinin of a different group can afford protection against influenza virus infection at distant sites away from the site of primary infection. In most studies, the same route of inoculation is used for priming and challenge infections with viruses of different subtypes (16, 17, 22). For example, priming with a seasonal influenza A virus and subsequent challenge infection of ferrets with A(H1N1)pdm09 viruses were performed by intranasal inoculation (17, 22, 34). Following intranasal inoculation of influenza A(H1N1)pdm09 virus, ferrets develop primarily disease of the URT (9), with limited dissemination of infection to the LRT.
This corresponds with reports of severe viral pneumonia in humans during the 2009 pandemic (35–37). Inoculation of A(H1N1)pdm09 virus via the intratracheal route causes moderate to severe pneumonia in ferrets, which is a good model for the LRT disease seen in humans (10).
In the present study, we assessed the protective effect of a primary infection of the URT with seasonal influenza A(H3N2) virus against subsequent infection of the URT and LRT with A(H1N1)pdm09 virus in the ferret model. We found that intranasal inoculation of A(H3N2) virus afforded local protection against URT disease after intranasal inoculation with A(H1N1)pdm09 virus but failed to afford protection against viral pneumonia after intratracheal inoculation with A(H1N1)pdm09 virus.
Influenza virus A/Brisbane/010/2007 (A/Brisbane/010/07; H3N2) was propagated in the allantoic cavity of 11-day-old embryonated chicken eggs. Allantoic fluid was harvested at 2 days postinoculation (p.i.), cleared by low-speed centrifugation, and stored at −80°C. Influenza virus A/Netherlands/602/2009 [A(H1N1)pdm09; H1N1] was propagated in confluent Madin-Darby canine kidney (MDCK) cells (9). After cytopathic changes were complete, culture supernatants were harvested, cleared by low-speed centrifugation, and stored at −80°C. Virus titers in MDCK cells were determined as described previously (38). All experiments with influenza A viruses were performed under biosafety level 2 (BSL-2) conditions, and experiments with animals were performed under BSL-3 conditions.
Healthy young adult female ferrets (Mustela putorius furo) between 6 and 12 months of age were purchased from a commercial breeder. Ferrets were tested for the presence of serum antibodies against recent influenza A/H1N1 and A/H3N2 viruses by hemagglutination inhibition (HI) assay. Seronegative ferrets were assigned to one of four experimental groups of six ferrets each. Ferrets of groups 1 (H3N2-H1N1it) and 3 (H3N2-H1N1in) were inoculated intranasally with 1 × 106 50% tissue culture infective doses (TCID50) of influenza virus A/Brisbane/010/07 (H3N2) in a total volume of 0.25 ml phosphate-buffered saline (PBS), while ferrets of groups 2 (PBS-H1N1it) and 4 (PBS-H1N1in) were inoculated with PBS. One week before inoculation with PBS or influenza A/Brisbane/010/07 virus, a temperature logger (DST Micro-T Ultra small temperature logger; Star-Oddi, Reykjavik, Iceland) was placed in the peritoneal cavity of each ferret. The loggers recorded the body temperature of the animals every 10 min. Mean body temperatures for each ferret were calculated in increments of 24 h.
Four weeks after inoculation with the influenza A/H3N2 virus, ferrets of groups 1 and 2 were inoculated intratracheally with 1 × 106 TCID50 of influenza A(H1N1)pdm09 virus in 3 ml PBS, while ferrets of groups 3 and 4 were inoculated intranasally with 1 × 106 TCID50 of influenza A(H1N1)pdm09 virus in 0.5 ml PBS, as described previously (9, 10). All inoculations and surgery to place temperature loggers were performed under anesthesia with ketamine-medetomidine (reversed with atipamezole). During infection experiments, ferrets were monitored daily for the presence of clinical signs. Before and 2, 4, 6, and 7 days after infection with influenza virus, ferrets were anesthetized with ketamine to determine their body weight and to collect nasal and pharyngeal swabs. Four and 7 days after inoculation with influenza A(H1N1)pdm09 virus, or earlier if ferrets became moribund and reached humane endpoints, three animals of each group were weighed and euthanized by exsanguination while under anesthesia with ketamine and medetomidine. Necropsies were performed according to standard procedures, and samples of the lungs (all lobes of the right lung and the accessory lobe) and nasal turbinates were collected to assess infectious virus titers and to study histopathological changes in these organs. The weight of the lungs of each ferret was determined after necropsy and expressed as the weight relative (%) to the body weight of the ferret at the day of necropsy.
During the experiment, ferrets were housed in groups and received food and water ad libitum. An independent animal ethics committee (Dier Experimenten Commissie) approved the experimental protocol before the start of the experiments.
Serum samples were collected before and 28 days after infection with influenza virus A/Brisbane/010/07 (H3N2). Serum samples were stored at −20°C until use. Sera were tested for the presence of antibodies against influenza viruses A/Brisbane/010/07 (H3N2) and A(H1N1)pdm09, using an HI assay performed with four hemagglutinating units of virus and 1% turkey erythrocytes and a virus microneutralization assay (VN assay) performed with 100 TCID50 of the respective virus, as described previously (39, 40).
Influenza A viruses A/Brisbane/010/07 (H3N2) and A(H1N1)pdm09 were propagated in MDCK cells and purified and concentrated by isopycnic density centrifugation. Subsequently, the viruses were inactivated by dialysis against PBS containing 0.1% formaldehyde for 4 days with continuous stirring at room temperature. After inactivation, antigens were dialyzed against PBS. The purity of the antigens was tested by SDS-PAGE, and inactivation was confirmed by failure of passaging on MDCK cells. The protein concentration was determined using a bicinchoninic acid (BCA) protein assay kit (Thermo Scientific Pierce Protein Biology Products). The presence of virus-specific T cells in peripheral blood mononuclear cells (PBMC) of ferrets 28 days after inoculation with influenza A/H3N2 virus was tested using a T cell proliferation assay as described previously (16, 41). In brief, cryopreserved PBMC were thawed, washed, and then labeled with 0.3 μM carboxyfluorescein diacetate succinimidyl ester (CFSE; Invitrogen, Breda, Netherlands) for 5 min. Cells were subsequently incubated in the presence of 50 ng of either inactivated influenza virus A/Brisbane/010/07 (H3N2) or A(H1N1)pdm09. Cells incubated with PBS and phytohemagglutinin (PHA) were included as negative and positive controls, respectively. After 6 days of incubation (with 4 days in the presence of the supernatant of concanavalin A-stimulated ferret lymph node cells), cells were stained with fluorescently labeled monoclonal antibodies directed against CD3 and CD8 and with a Live/Dead marker. Cells were analyzed by flow cytometry using a FACS-CantoII flow cytometer in combination with FACS-Diva software (BD, Alphen a/d Rijn, Netherlands). Antigen-specific proliferation of T cells was calculated by subtracting the mean number of CFSElow CD3+ CD8− or CFSElow CD3+ CD8+ cells of the medium controls from the mean number of CFSElow CD3+ CD8− or CFSElow CD3+ CD8+ cells stimulated with either antigen.
Tissue samples of ferrets were collected, snap-frozen on dry ice, and stored at −80°C until further processing. Tissue samples were weighed, subsequently homogenized in Hanks' balanced salt solution containing 0.5% lactalbumin, 10% glycerol, 200 U/ml penicillin, 200 μg/ml streptomycin, 100 U/ml polymyxin B sulfate, 250 μg/ml gentamicin, and 50 U/ml nystatin (ICN Pharmaceuticals, Zoetermeer, Netherlands) with a FastPrep-24 homogenizer (MP Biomedicals, Eindhoven, Netherlands), and then centrifuged briefly. After collection, nasal and pharyngeal swabs were stored at −70°C in the same medium as that used to homogenize lung samples. Triplicate 10-fold serial dilutions of swab samples and quadruplicate 10-fold serial dilutions of tissue samples were used to inoculate MDCK cells as described previously (38). HA activity of the culture supernatants collected 5 days following infection of MDCK cells was used as an indicator of infection. The titers were calculated according to the Spearman-Karber method and are expressed as log10 TCID50 per gram tissue or per milliliter (for swabs) (42).
Tissues from ferrets euthanized after inoculation with influenza A(H1N1)pdm09 virus were examined macroscopically for the presence of lesions, and lungs (after inflation with 10% neutral buffered formalin) were fixed with 10% neutral buffered formalin. After fixation and embedding in paraffin, tissues were sectioned at 4 μm, and tissue sections were examined by staining with hematoxylin and eosin (HE). Semiquantitative analysis of pulmonary lesions 4 and 7 days after inoculation was performed as described previously (43). In brief, four lung sections from each ferret were scored for the size and severity of inflammatory foci by examining five arbitrarily chosen fields in each section by light microscopy with a 10× objective. Each field was scored for the size of the inflamed area (1, smaller than or equal to the area of a 10× objective; 2, larger than the area of a 10× objective and smaller than or equal to the area of a 2.5× objective; and 3, larger than the area of a 2.5× objective) and the severity of inflammation (1, mild; 2, moderate; and 3, marked), and the average cumulative value for the size and severity of inflammation per field was used as the histopathology score for each ferret.
In addition, the presence and size of peribronchiolar cuffs in each ferret were determined by counting the number of lymphocyte cell layers around 4 bronchioles per lung section. Using an immunoperoxidase method, serial lung tissue sections were also stained with a monoclonal antibody directed against the nucleoprotein of the influenza A virus for the detection of virus-infected cells in the lungs (44).
After inoculation with influenza virus A(H1N1)pdm09, the presence of statistically significant differences in weight loss and virus titers in nasal and pharyngeal swabs between ferrets of the H3N2-H1N1it and PBS-H1N1it groups and between ferrets of the H3N2-H1N1in and PBS-H1N1in groups was assessed using the Mann-Whitney U test only when data for more than 3 animals per group were available. Differences were considered significant when P values were <0.05.
After inoculation with influenza virus A/Brisbane/010/07 (H3N2), ferrets developed mild to moderate clinical signs, including decreased appetite and weight loss (mean weight loss of 4% at day 2 p.i.). Furthermore, a rise in body temperature was observed, with peaks on days 2 (mean body temperature ± standard deviation [SD], 39.3°C ± 0.2°C) and 6 (mean body temperature ± SD, 38.4°C ± 0.2°C) p.i. In PBS-inoculated ferrets, no clinical signs or rise in body temperature was observed (Fig. 1A).
Nasal and pharyngeal swabs collected before and 2, 4, 6, and 7 days after inoculation were tested for the presence of infectious virus. In all A/H3N2 virus-inoculated animals, virus was detected on days 2 and 4 after inoculation, and it was detected in 9 of 12 ferrets on day 6 p.i. (Fig. 1B). On day 7 p.i., all animals tested negative. The nasal and pharyngeal swabs collected before inoculation and from PBS-inoculated animals also tested negative for the presence of virus.
Serum samples and PBMC were collected from ferrets of all groups 4 weeks after inoculation with influenza A(H3N2) virus. The presence of virus-specific T cells was evaluated using a T cell proliferation assay. The mean number of CFSElow CD3+ CD8− cells after stimulation with influenza A/H3N2 virus was higher in ferrets inoculated with influenza A/H3N2 virus (mean ± SD, 1,901 ± 1,848) than in PBS-inoculated control ferrets (485 ± 571). In addition, the presence of cross-reactive T cells was tested by stimulating PBMC with inactivated influenza virus A(H1N1)pdm09. Again, the mean number of CFSElow CD3+ CD8− cells was higher in A/H3N2 virus-infected animals (1,141 ± 1,771) than in mock-infected animals (601 ± 605) (Fig. 1C). Virus-specific proliferation of CD3+ CD8+ cells was hardly detectable (data not shown). In addition, all ferrets developed antibodies to influenza virus A/Brisbane/010/07 (H3N2), as measured by HI (mean titer ± SD, 1,011 ± 799) and VN (mean titer ± SD, 1,216 ± 365) assays. These antibodies did not cross-react with influenza virus A(H1N1)pdm09 in either assay (data not shown).
Four weeks after inoculation with influenza A(H3N2) virus or PBS, ferrets were inoculated with influenza virus A(H1N1)pdm09 by the intranasal route (groups PBS-H1N1in and H3N2-H1N1in) or the intratracheal route (groups PBS-H1N1it and H3N2-H1N1it). Following inoculation, all ferrets developed clinical signs, from day 1 after inoculation onwards. Clinical signs included anorexia, weight loss, lethargy, and labored breathing. However, intratracheally inoculated ferrets developed more severe clinical signs than those in ferrets inoculated by the intranasal route.
No significant differences in weight loss were observed between ferrets of groups H3N2-H1N1it and PBS-H1N1it after intratracheal inoculation with influenza virus A(H1N1)pdm09 (Fig. 2A and andD).D). One ferret in group H3N2-H1N1it had to be euthanized for ethical reasons 6 days after inoculation, as it displayed severe clinical signs. In contrast to groups H3N2-H1N1it and PBS-H1N1it, differences were observed between groups H3N2-H1N1in and PBS-H1N1in. The animals that were primed by infection with A(H3N2) influenza virus displayed a reduced loss of body weight upon intranasal inoculation of influenza virus A/Netherlands/602/09 compared to unprimed PBS control animals (Fig. 2G and andJ).J). On day 4 p.i., this difference was statistically significant (P = 0.015).
Upon challenge infections, we also monitored the body temperature. After intratracheal inoculation with A(H1N1)pdm09 influenza virus, body temperatures peaked at day 1 to 2 p.i., and no clear differences were observed between primed and unprimed animals. After intranasal inoculation of naïve ferrets with A(H1N1)pdm09 virus, a sharp rise in body temperature was observed on day 2 p.i. (mean body temperature ± SD, 39°C ± 0.1°C), while in ferrets of group H3N2-H1N1in this rise in body temperature was less pronounced (mean body temperature ± SD, 38.1°C ± 0.4°C) (Fig. 2B, ,E,E, ,H,H, and andKK).
Nasal and pharyngeal swabs collected on days 2 and 4 (n = 6) and on days 6 and 7 (n = 3) p.i. were tested for the presence of infectious virus. Virus titers detected in nasal swabs collected from ferrets of group H3N2-H1N1it were lower than those in swabs from group PBS-H1N1it on days 2, 4, 6, and 7, although the differences were not statistically significant. In addition, virus titers in pharyngeal swabs from ferrets in group H3N2-H1N1it were lower than those for group PBS-H1N1it on days 2, 6, and 7 p.i.
These differences in virus titers were larger between A(H3N2) virus-primed and naïve ferrets after intranasal challenge with A(H1N1)pdm09 virus. The virus titers of A(H3N2) virus-primed animals were significantly lower in nasal swabs on days 2 and 4 p.i. and in pharyngeal swabs on day 4 p.i. No virus was detected in nasal and pharyngeal swabs collected from the remaining ferrets of group H3N2-H1N1in on days 6 and 7 p.i., while the nasal swabs of all animals in group PBS-H1N1in tested positive at these time points, as did the pharyngeal swabs collected on day 6 p.i. (Fig. 2C, ,F,F, ,I,I, and andLL).
Nasal turbinates and lung specimens taken on days 4 and 7 p.i. tested positive for the presence of virus for all ferrets except for one from the H3N2-H1N1in group, which was euthanized on day 7. In general, the mean virus titers in nasal turbinates of ferrets in groups PBS-H1N1it and PBS-H1N1in were higher than those in nasal turbinates of the A(H3N2) virus-primed animals in groups H3N2-H1N1it and H3N2-H1N1in (Table 1). Lungs of all ferrets tested positive on day 4 p.i., except for one ferret of the H3N2-H1N1in group. On day 7 p.i., lungs of two ferrets of the H3N2-H1N1it group and the PBS-H1N1it group and one ferret of the PBS-H1N1in group tested positive for the presence of virus, but none of the lungs of the ferrets of the H3N2-H1N1in group tested positive. Mean lung virus titers were similar for the intratracheally inoculated groups, while the mean virus titers detected in ferrets of group H3N2-H1N1in were lower than those of ferrets in group PBS-H1N1in (Table 1).
After euthanasia of ferrets at 4 and 7 days p.i., lungs of ferrets were examined for macroscopic lesions. Dark-red firm foci were observed in the lungs of a proportion of the ferrets, and the area of affected foci was estimated. The area of affected lung tissue was larger in intratracheally inoculated ferrets than in intranasally inoculated ferrets, which corresponded with the higher relative lung weights in intratracheally inoculated ferrets. In addition, the area of affected lung tissue was larger in unprimed control ferrets than in H3N2 virus-primed ferrets at 4 days p.i. (mean values of 57% and 40% for ferrets of groups PBS-H1N1it and H3N2-H1N1it, respectively, and 25% and 11% for ferrets of groups PBS-H1N1in and H3N2-H1N1in, respectively). At 7 days p.i., the area of affected lung tissue was somewhat larger in H3N2 virus-primed animals (mean values of 73% and 11% for ferrets of groups H3N2-H1N1it and H3N2-H1N1in, respectively, and 57% and 5% for ferrets of groups PBS-H1N1it and PBS-H1N1in, respectively). The observed differences in the area of affected lung tissue between groups correlated with differences in relative lung weight and in the histopathology score (Table 1). On histological examination of all intratracheally inoculated ferrets, a multifocal or coalescing, moderate or severe necrotizing bronchointerstitial pneumonia was observed, which was characterized by the presence of inflammatory cells in the lumina and walls of alveoli and bronchioles on both days 4 and 7 after inoculation. In addition, necrosis of the alveolar and bronchiolar epithelium was observed. At day 4 p.i., the infiltrate consisted mainly of neutrophils and macrophages, while at day 7 p.i., mainly macrophages and lymphocytes were observed. In addition, various amounts of edema fluid, fibrin, and erythrocytes were observed (Fig. 3C, ,D,D, ,G,G, and andH).H). The peribronchiolar and perivascular cellular infiltrates were most prominent in lungs of ferrets of the H3N2-H1N1it group, as determined by calculation of the number of lymphocyte cell layers around bronchioles (Table 1; Fig. 3A, ,B,B, ,E,E, and andF).F). In lungs of intranasally inoculated animals, mild or moderate bronchitis and bronchiolitis were observed that were characterized by the presence of inflammatory cells in the lumina and walls of the bronchi and bronchioles and necrosis of the bronchiolar epithelium (Fig. 3I, ,J,J, ,M,M, and andN).N). In the alveoli, only focal zones with infiltrates of inflammatory cells were present, mainly adjacent to bronchi and bronchioles (Fig. 3K, ,L,L, ,O,O, and andP).P). The presence of pathological changes at day 4 p.i. correlated with the presence of virus-infected cells as demonstrated by immunohistochemistry and virus titers in the lungs. In alveoli of intratracheally inoculated ferrets, many virus-infected cells were present, while only a few virus-infected cells were present in alveoli of intranasally inoculated ferrets. Virus-infected cells were present in bronchiolar walls of ferrets of all groups.
In the present study, the protective potential of heterosubtypic immunity induced after infection of the URT with a seasonal influenza A(H3N2) virus was assessed against infection of the URT or the lungs with influenza A(H1N1)pdm09 virus in ferrets. To infect the URT, the challenge virus was inoculated via the intranasal route, while for infection of the lungs as the primary site of virus replication, the virus was inoculated intratracheally. Infection of the URT with seasonal influenza viruses typically results in a self-limiting disease of relatively mild severity. This was also seen after infection of ferrets with influenza virus A/Brisbane/010/07 (H3N2). It is interesting that infection with a seasonal A(H3N2) virus was restricted to the URT even after intratracheal inoculation (45). After inoculation, the virus replicated in the URT, as demonstrated by the presence of infectious virus in nasal and pharyngeal swabs at 2 to 6 days p.i. The animals developed fever, and virus-specific antibodies and T cells were induced in response to infection, the latter of which cross-reacted with A(H1N1)pdm09 virus.
Upon intranasal inoculation of A(H1N1)pdm09 virus, animals also developed fever comparable to that observed during infection with the seasonal A(H3N2) virus. In contrast, intratracheal inoculation of A(H1N1)pdm09 virus caused acute necrotizing pneumonia characterized by diffuse alveolar damage, as described previously (10). The latter was used as a model for human cases of acute pneumonia observed during the pandemic of 2009 (46, 47).
To assess the protective capacity of immunity induced by infection with a seasonal influenza A(H3N2) virus, ferrets were infected with the influenza virus A/Brisbane/010/07 4 weeks prior to intranasal or intratracheal inoculation with the A(H1N1)pdm09 virus A/Netherlands/602/09. The outcome of infection was compared with that for unprimed control animals. Priming with the seasonal A(H3N2) virus induced protective immunity against intranasal inoculation with the A(H1N1)pdm09 virus. Primed animals displayed fewer clinical signs and had lower virus titers in the upper and lower respiratory tracts than those of control animals. In contrast, after inoculation via the intratracheal route, no differences were observed between the A(H3N2)-primed and control animals.
These findings indicate that after infection of the URT with a seasonal A(H3N2) virus, heterosubtypic immunity is induced locally, and this fails to protect against infection at a distant site. Note that it was demonstrated previously that in mice inoculated with a high dose of a fast-replicating influenza A/H5N1 virus, the virus outpaced the immune response, resulting in weak virus-specific CD8+ T cell responses, high lung viral loads, and a detrimental outcome of infection (48). Also, in our model, the intratracheal inoculation of A(H1N1)pdm09 virus may have favored the virus, and virus-specific T cell responses induced by infection with the seasonal influenza virus may have been insufficient to counteract primary virus replication at a distant site, as well as the development of acute pneumonia. However, weight loss at 4 days p.i. correlated inversely with the number of A(H1N1)pdm09 virus-specific T cells measured on the day of challenge infection (Pearson's correlation coefficient [R] = 0.54), although no significant differences in T cell responses were observed between primed and unprimed animals. Also, the relatively high virus dose used to inoculate ferrets may have contributed to the failure to protect the lungs, since it has been shown that large numbers of cells become infected shortly after inoculation, leading to high virus loads (45). However, this dose was used in multiple studies in which the pathogenesis of influenza A viruses in ferrets was evaluated (9, 10).
In contrast, local protective immunity in the URT may prevent or slow down further dissemination and virus spread to the lungs. Indeed, the lung virus titers were lower in A(H3N2) virus-primed animals that were subsequently inoculated intranasally with A(H1N1)pdm09 virus.
These findings are in agreement with previous studies which demonstrated that after intranasal inoculation with a seasonal influenza A(H3N2) virus, ferrets were protected from infection with a highly pathogenic avian influenza A virus of the H5N1 subtype or with influenza A(H1N1)pdm09 virus inoculated via the intranasal route (16, 17).
Similar results were also obtained in a mouse model using intranasal inoculation for priming and administration of challenge virus (15, 21), but it should be emphasized that the mode of action and the protective mechanisms in mice may be fundamentally different from those in ferrets.
For instance, a seasonal A(H3N2) virus inoculated intranasally in a volume of 50 μl into mice typically reaches the lungs, resulting in virus replication in that organ, which in turn causes pathological changes and the formation of inducible bronchiole-associated lymphoid tissue (iBALT) (21, 49), containing local virus-specific B and T cells, which may contribute to protective immunity in the lungs to viruses of other subtypes, including H5N1 and H1N1pdm09 (16). In contrast, influenza virus A/Brisbane/010/2007 (H3N2) inoculated intranasally into ferrets did not cause pathological changes or formation of iBALT in the lungs and therefore may confer only inefficient direct protection against replication in the lungs (16). As indicated above, the presence of local immunity may restrict A(H1N1)pdm09 virus replication in the URT and subsequent virus dissemination, thus affording protection to the lungs indirectly.
Some of the animals that were primed by infection with seasonal A(H3N2) virus and subsequently challenged with A(H1N1)pdm09 virus via the intratracheal route (H3N2-H1N1it) developed more severe disease than that observed in the unprimed control animals. One of these animals had to be euthanized for ethical reasons because it reached a humane endpoint. This animal and another from this group displayed signs of severe inflammation of the lungs upon histological examination, and it cannot be excluded that priming by infection with the seasonal influenza A/H3N2 virus predisposed animals to the development of more severe disease after intratracheal infection with the A(H1N1)pdm09 virus. It is possible that the induction of cross-reactive virus-specific T cells played a role in this immunopathological process (50–52).
Several candidate universal influenza vaccines are in various stages of development (29, 34, 53–56). Those that aim at the induction of cross-protective T cell responses especially hold promise, as it has been shown in numerous animal models that the induction of virus-specific T cell responses affords protection against challenge infection (12, 57). Using a poxviral vector for the delivery of conserved influenza virus antigens, it was shown that virus-specific T cells can afford human study subjects clinical protection against intranasal challenge infection (54). Whether the induction of virus-specific T cell responses will also protect against infection of the lower respiratory tract or possibly predispose individuals to immune pathology may depend on the route of vaccination, the nature and dose of the vaccine used, and the magnitude of the T cell response induced. Further research is required to address these outstanding issues.
This study was financially supported by an EU FluPig grant (FP7-GA258084).
We thank Koert Stittelaar and Leon de Waal for excellent technical advice and Cindy van Hagen, Willem van Aert, and Ronald Boom for excellent technical assistance.
Published ahead of print 30 January 2013