Background
Influenza A virus pandemics continue to occur sporadically in the human population. Moreover, the emergence of a novel avian-origin reassortant influenza A (H7N9) virus in China with several characteristic features of mammalian influenza viruses poses concern for public health [
1]. Currently, vaccination using inactivated influenza virus preparations remains the primary method of prevention, especially for pre-pandemic vaccines. In particular, A/PR/8/34 (PR8)-based pre-pandemic vaccine candidates, which contain the internal gene segments of PR8 virus and the HA and NA genes from avian influenza viruses with pandemic potential, have been generated and in some cases improved with respect to their growth characteristics [
2,
3]. However, while vaccine-elicited antibodies will fully protect against homologous virus infection, eventually they offer only marginal protection against heterologous drift virus infection.
Although heterosubtypic immunity in humans appears to be weak, there is evidence that cross-reactive cell-mediated immune responses contribute to disease control [
4‐
8]. Thus, vaccine-induced primary and memory responses by T cells against the relatively conserved internal antigens of influenza could play an important role in reducing the severity of disease caused by antigenic variants that could arise [
9‐
13]. Although non-replicating influenza virus-based vaccines are known to be poor inducers of cytotoxic T cell (CTL) responses, advances in the field of adjuvants, delivery systems, and novel vaccination strategies hold promise for improving their potential effectiveness [
14‐
17]. Inactivated whole virus vaccines with intact membrane fusion activity are known to facilitate cross-presentation to CD8+ T cells by antigen presenting cells (APC), and are more immunogenic than split vaccines [
18,
19]. Moreover, gamma-ray-inactivation of whole-virus vaccines has been reported to preserve antigenicity, as well as cellular and humoral immunogenicity, to a greater extent than current viral inactivation methods [
20]. Despite the important role that multiple antigenic determinants of internal viral proteins may have in protection against infection, little is yet known about which of these antigens would better induce and/or recall memory CD8+ T cells after in vivo administration of an inactivated whole virus vaccine.
Because of different HLA haplotypes and MHC Class I-dependent CD8+ T cell functions, the quality and quantity of a T cell response can vary greatly between individuals. HLA-A*0201 is an allele expressed by nearly half of the world’s population and several studies have been performed to determine the HLA-A2-restricted CD8+ T cell responses following natural influenza virus infection [
21,
22]. In the present study, we investigated the HLA-A*0201-restricted epitope specificities of CD8+ T cells in HLA-A2.1 transgenic (AAD) mice responding to a non-replicating H7N3 whole virus, and revealed the breadth of recall responses following infection of these mice with heterologous PR8 virus. We also determined whether these vaccine-induced virus-specific CD8+ T cells, which target the internal conserved viral proteins, affected virus replication in the lungs of the mice.
Methods
HLA-A2.1 transgenic mice
AAD mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). These mice with the C57BL/6J genetic background express an interspecies hybrid class I molecule, composed of the alpha 1 and alpha 2 domains of the human HLA-A*0201 allele and the alpha 3, cytoplasmic and transmembrane domains of the mouse H-2D
d class I molecule [
23]. Female mice were used at 8–10 weeks of age; genetic authenticity was confirmed by performing PCR for the transgene. All mice were maintained at Istituto Superiore di Sanità following institutional guidelines and were used under approved protocols.
Peptides
The H2-Db restricted NP366 peptide and thirteen peptides that bind the HLA-A2.1 molecule and are conserved among diverse influenza subtypes were synthesized by Primm (Italy). Peptide purity was >90 % in all cases, and the identity of the peptides was verified by using spectrometry. The peptides were dissolved in DMSO at 10 mg/ml and stored at -20 °C.
Generation of PR8-NPN370Q virus
The mutant recombinant virus PR8-NP
N370Q, in which the asparagine at the fifth position of the epitope was replaced with glutamine (N370Q), was generated by using plasmid-driven reverse genetics, as described by Neumann et al. [
24]. The presence of the NP mutation was confirmed by sequencing PCR-amplified cDNA.
Immunization of mice and challenge experiments
Groups of AAD mice were injected intraperitoneally (i.p.) with 500 HAU (approximately 3.5 μg of viral protein in a 0.5-ml volume) of either live or heat-inactivated (HI) low pathogenic A/turkey/Italy/214845/02 virus (HI-H7N3). Heat inactivation of small volumes of virus suspension was performed in a water bath at 56 °C for 30 min and confirmed by the complete loss of infectivity of MDCK cells. Intranasal (i.n.) immunization was performed by anesthetizing mice with Avertin (2,2,2-tribromoethanol) and then inoculating them with 102.4 TCID50 of PR8 virus, resuspended in 40 μl of PBS. For the viral challenge, mice in groups of 7 were i.n. infected with 3 LD50 (50 % lethal dose), corresponding to 103.7 TCID50 in a volume of 40 μl, of PR8-NPN370Q virus. Naïve mice were included as a negative control. The mice were weighted daily and monitored for mortality for two weeks. To determine viral lung titers, mice were sacrificed 3 and 7 days post-infection (p.i.), and lung homogenates were titrated in MDCK cells. The limit of viral detection was 0.5 log10 TCID50/ml.
IFN-γ ELISPOT assays
Spleens of mice (5–7 mice per group) were collected at the indicated time points and assayed for antigen-specific IFN-γ-producing cells by using an IFN-γ ELISPOT assay. In the challenge experiments, inflammatory cells were also collected from the lymph nodes draining the respiratory tracts (mediastinal lymph nodes, MLN), lungs and bronchoalveolar lavages (BAL) of infected mice as previously described, and then directly subjected to the ELISPOT assay [
25,
26]. Single-cell suspensions from the lymphocytic populations were cultured with the indicated synthetic peptides or DMSO in anti-IFN-γ-coated plates at 37 °C for 36–40 h. Colored spots representing IFN-γ-releasing cells are reported as the number of spot-forming cells (SFC) per 10
6 cells.
Serology
Serum samples were collected from mice vaccinated with HI-H7N3 immediately before and 7 days after challenge, and tested for the presence of influenza-specific antibodies by using an hemagglutination–inhibition (HI) assay and a micro-neutralization (MN) assay, as described previously [
27]. Sera were also tested in ELISA, using purified whole inactivated viruses [
28].
Statistics
The statistical significance of the differences between groups of animals was determined by using the Wilcoxon-Mann-Whitney test; P < 0.05 was considered significant.
Discussion
In the present study, we evaluated CD8 + T cell responses to internal viral proteins elicited by an inactivated H7N3 virus in AAD mice. We found that CD8+ T cell responses to both the immunodominant epitopes and to most of the highly conserved subdominant epitopes could be detected in AAD mice upon inoculation with live virus, whereas immunization with HI-H7N3 virus induced CD8+ T cell responses mainly to the immunodominant epitopes. We also found that the vaccine-induced T cell responses appeared to reduce the lung viral load in mice challenged with the PR8-NPN370Q virus, which lacks the H2-Db-NP366 epitope.
Most influenza-specific memory T cells that are detectable in humans are directed mainly against the remarkably conserved epitopes of the internal viral proteins, and have been reported to accumulate either as resident or central memory T cells in the lungs and lymphoid tissues, respectively, following recovery from influenza virus infections [
37‐
41]. Therefore, the rapid boosting of these cross-reactive T cells in the respiratory tract by a pre-pandemic vaccine could be extremely beneficial to provide protection against severe influenza virus infections [
42]. Here, the CD8 + T cell responses to the HLA-A2-restricted influenza epitope M1
58 paralleled those specific to the H2-D
b-NP
366 epitope in AAD mice, following immunization with live virus. Interestingly, the level of responses specific to the H2-D
b-PA
224 epitope was lower in AAD mice than in C57BL/6 mice. This low reactivity may reflect a reduced level of antigen-specific T cell precursors in AAD mice, compared with the higher frequency of naïve precursors for H2-D
b-PA
224 than H2-D
b-NP
366 in C57BL/6 mice [
43]. Alternatively, it may be due to superior competition of T cells specific to the H2-D
b-NP
366 and M1
58 epitopes from the most abundant internal proteins during antigen presentation.
In AAD mice immunized with a single dose of inactivated H7N3 vaccine, CD8+ T cell responses were mainly elicited by immunodominant epitopes. Several factors contribute to the differences between live and inactivated viruses in the CD8+ T cell response to the subdominant epitopes. In particular, the provision of additional antigens in the form of endogenously synthesized viral proteins, following infection with a live virus, plays a key role. Our results are consistent with those of other studies showing that epitopes derived from more abundant viral proteins contribute to immunodominant responses, although the inherent features of the antigens and their capacity to be cross-presented by APC also likely play a role [
32,
44‐
46]. Likewise, others have reported that CTL induction by cross-presentation of Lymphocytic Choriomeningitis virus antigens, that were constitutively expressed in tumor cell lines and injected in mice, was driven by immunodominant but not subdominant epitopes [
47]. To further investigate the immunogenicity of the whole inactivated virus vaccine, we used HI-H7N3 in a prime-boost approach with live preparations of serologically distinct PR8 virus. We detected and efficiently recalled, upon subsequent viral infection, mainly CD8+ T cell responses that were restricted to the immunodominant epitopes H2-D
b-NP
366 and M1
58, and the overlapping M1
59. Importantly, the M1
58-specific CD8+ T cells were efficiently recruited to the respiratory tract by subsequent i.n. infection with PR8-NP
N370Q virus, and were likely responsible for the reduction in lung viral load. Although we cannot exclude that CD4+ and CD8+ T cells with other specificities may help reduce viral infection, the lack of serum cross-reactive antibodies further corroborates the most likely protective effect of M1
58-specific CD8+ T cells. Nevertheless, no changes were observed in mortality of these mice receiving a single dose of a parenterally administered vaccine as compared to unvaccinated mice. Our results agree with other research findings showing that a subsequent influenza virus infection in the respiratory tract is required for reactivation of i.p. primed influenza virus-specific CD8+ T cells and access to the lung airways [
38,
48]. Indeed, activation and maintenance of T cell populations in the airways are dependent upon the route of infection and prolonged presentation of viral antigens in the lymph nodes draining the respiratory tract. Thus, new immunization strategies that induce broadly cross-reactive antibodies and T-cell responses at mucosal surfaces could greatly improve influenza vaccine efficacy. Although mouse models do not closely reflect human influenza virus infection, yet our data suggest that a single immunization with a whole inactivated influenza virus vaccine might also be able to elicit or recall cross-protective immune responses specific to immunodominant NP and M epitopes in humans, and thus, to some extent, counteract an infection by an heterologous virus. In this context, it would be interesting to explore the immunogenic pattern and magnitude of the response elicited in the presence of adjuvants or following immunization of these mice with live, attenuated influenza vaccine (LAIV). Previous studies have shown quantitative and qualitative differences in T cell responses to LAIV and inactivated vaccines. In particular, viral replication of LAIV in the respiratory mucosa, especially of young children with low levels of pre-existing immunity, provides better protection against infection, likely due to a broader array of dominant and subdominant T cell responses as well as beneficial induction of immune-effectors at the site of infection [
49,
50].
The M1
58 epitope is highly conserved among multiple influenza viral strains and subtypes, and it is able to elicit CTL responses in both HLA-A*02- and HLA-C*08-positive individuals [
29,
51]. Although functional constraints on influenza virus CTL epitopes may limit escape from CTL-mediated immunity, epitope mutations may occur and, in some cases, these variant epitopes can still be recognized, to some extent, by cross-reactive T cells [
52‐
54]. Indeed, CTL escape variants have been shown to arise following independent amino acids mutations either at the anchor residues or in the T cell receptor contact residues of CTLs specific to the NP of influenza A viruses isolated during circulation in human populations [
55,
56]. For these reasons, it is strongly desirable to have more epitopes recognized by the immune system to fight a viral infection. Furthermore, we cannot exclude the possibility that inactivated whole virus-based vaccines of avian origin would cross-react better than PR8-based vaccines against new emerging zoonotic influenza viruses in human hosts. Although viral internal proteins are highly conserved, the 10 %–15 % sequence dissimilarity in the NP of different influenza subtypes, for example, may be responsible for the loss of epitopes of unknown HLA restrictions. Therefore, in the context of inactivated vaccines against highly pathogenic influenza strains that could arise from animal reservoirs, the recall of dominant CD8+ T cell responses targeting the M and NP proteins most closely related to the emerging virus strains could be of great value in vaccine effectiveness. Further studies are warranted to better characterize the relationship between pre-existing CD8+ T-cell immunity across the multiple ethnicities established in the human population and epitope conservancy among the internal proteins of potentially pandemic influenza strains [
57].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Designed the experiments: MRC and BG; Performed the experiments: BG, GDM, ES, MF; Analyzed the data: BG, GDM, ID and MRC; Wrote the paper: BG, MRC. All authors read and approved the final manuscript.