Abstract
Current influenza vaccines containing primarily hypervariable haemagglutinin and neuraminidase proteins must be prepared against frequent new antigenic variants. Therefore, there is an ongoing effort to develop influenza vaccines that also elicit strong and sustained cytotoxic responses against highly conserved determinants such as the matrix (M1) protein and nucleoprotein (NP). However, their antigenic presentation properties in humans are less defined. Accordingly, we analysed MHC class I and class II presentation of endogenously processed M1 and NP in human antigen presenting cells and observed expansion of both CD8+- and CD4+-specific effector T lymphocytes secreting gamma interferon and tumour necrosis factor. Further enhancement of basal MHC-II antigenic presentation did not improve CD4+ or CD8+ T-cell quality based on cytokine production upon challenge, suggesting that endogenous M1 and NP MHC-II presentation is sufficient. These new insights about T-lymphocyte expansion following endogenous M1 and NP MHC-I and -II presentation will be important to design complementary heterosubtypic vaccination strategies.
-
Supplementary material is available with the online version of this paper.
Introduction
Current trivalent inactivated influenza vaccines (TIV) mainly induce a humoral response against hypervariable haemagglutinin (HA) and neuraminidase (NA) surface antigens. Hence, production of new vaccines is required for every new influenza strain. Considering the ongoing threat of an influenza pandemic (WHO, 2009), vaccines targeting better conserved antigens are required. The influenza A matrix protein (M1) and nucleoprotein (NP) share more than 90 % amino acid sequence identity even between distant influenza A subtypes (Heiny et al., 2007; and see Supplementary Table S1, available in JGV Online). Moreover, many human M1 and NP T-cell epitopes are essential for the virus’ fitness (Berkhoff et al., 2006) and are thus well conserved (Bui et al., 2007; Lee et al., 2008). In contrast, HA and NA vary in up to 40 % of their amino acid sequence. These differences are mainly seen in the extracellular globular head of the HA glycoprotein, which contains the receptor-binding site (Schweiger et al., 2002) whose inhibition is critical to prevent cell entry of the virus. Hence, HA glycoprotein is less likely to induce heterosubtypic cross-reactive immunity.
Viral proteins must be processed by infected cells or antigen presenting cells (APCs) to elicit a cellular immune response. Although such a response does not confer sterile immunity to influenza, it has been shown to mediate influenza virus clearance in animal models (Thomas et al., 2006) and in humans (McMichael et al., 1983). In animal models, a cellular immune response has long been associated with heterosubtypic protection against various influenza A strains (Furuya et al., 2010; Taylor & Askonas, 1986), including H5N1 (Epstein et al., 2002; Price et al., 2009; Zhirnov et al., 2007) and the 2009 pandemic H1N1 (Skountzou et al., 2010). It has also been established that M1 and NP (Kreijtz et al., 2008; Lee et al., 2008) are the main targets of the human immune cellular response against influenza, while HA and NA are the main protective targets for the humoral immune response. Hence, recent reports strongly suggest that M1 and NP could be very relevant targets for an influenza pan-specific vaccine.
A wide range of conserved MHCs of class I and class II influenza A NP- and M1-specific epitopes has been characterized (Jameson et al., 1999; Kreijtz et al., 2008; Lee et al., 2008). However, little is known about how influenza-protein processing in human APCs stimulates expansion of T lymphocytes, particularly when these proteins are endogenously expressed, which would most probably be the case in T-cell-stimulating vaccines. Most studies on epitope identification with influenza-specific human T cells have exploited peptide libraries covering the influenza genome (Kreijtz et al., 2008; Lee et al., 2008), exogenously loaded proteins (Gschoesser et al., 2002) or influenza-infected cells (Gschoesser et al., 2002; Jameson et al., 1999; Kreijtz et al., 2008). In the latter two cases, some of the processed proteins could originate from an exogenous source potentially involving the endosomal pathway, while use of peptide libraries excludes endogenous processing pathways. In contrast, vaccines inducing M1- and NP-specific T-cell expansion, such as the live-attenuated virus vaccines (LAIV) most likely involve endogenous expression of these proteins. Endogenously expressed M1 is known to be presented by MHC-I and cross-presented by MHC-II in human APCs (Jaraquemada et al., 1990; Nuchtern et al., 1990). However, the magnitude and importance of MHC-II cross-presentation of endogenous M1 is unclear (Schmid et al., 2007), and no information is available regarding NP MHC-II cross-presentation. It is also unclear whether MHC-II cross-presentation can stimulate a robust in vitro expansion of specific T lymphocytes.
Here, we sought to characterize the MHC-I and -II presentation of endogenous influenza M1 and NP antigens in a T-cell-inducing vaccine-like context for the first time. We performed in vitro T-cell sensitization assays, by stimulating normal (healthy) donor (ND) peripheral blood mononuclear cells (PBMCs) with autologous APCs expressing endogenous M1 or NP antigens from a DNA plasmid and by characterizing the ensuing CD8+ and CD4+ T-lymphocyte response. Considering the capacity of endogenously expressed M1 to be presented by MHC-II, we questioned whether NP shared similar properties. Finally, we investigated whether the level of MHC-II cross-presentation was sufficient for M1 and NP CD4+ T-cell expansion, which plays a key role in cellular immune responses.
Results
Efficient presentation of MHC-I and -II epitopes derived from M1 and NP to human CD8+ and CD4+ T lymphocytes
Initially, we evaluated whether endogenous wild-type (WT) influenza M1 and NP proteins are efficiently presented by human MHC-I and -II. We electroporated CD40-activated B lymphocytes (CD40-B cells) with M1- and NP-encoding plasmids (Fig. 1a) and co-cultured them with previously generated autologous M1- or NP-specific CD8+ or CD4+ T-cell clones. CD40-B cells are valuable model APCs because they can be expanded from a limited quantity of PBMCs, while possessing antigen-presenting characteristics comparable to dendritic cells (Lapointe et al., 2003; Schultze et al., 1997). In these cells, endogenously expressed M1 and NP proteins were both recognized by CD8+ (Fig. 1c, black bars) and by CD4+ T-cell clones (Fig. 1d, black bars). To our knowledge, this represents the first report of NP MHC-II cross-presentation. Observed M1 MHC-II cross-presentation is in agreement with previous work (Jaraquemada et al., 1990; Nuchtern et al., 1990).
MHC-I and -II presentation of endogenously expressed influenza M1 and NP. Scheme of (a) WT M1 or NP and (b) M1 or NP cloned with gp100 MHC-II mobilization sequences in pcDNA3.1 plasmid. (c, d) CD40-B cells (APCs) were electroporated with plasmids encoding WT proteins (black bars), proteins cloned in fusion with the gp100 MHC-II mobilization sequences (grey bars) or an irrelevant mock (Dickkopf-1) protein (white bars), and presented to CD8+ (c) or CD4+ (d) T-cell clones specific to each antigen. IFN-γ secretion was assessed by ELISA. The data are representative of two independent experiments. (e) Expression of the M1 and gp-M1 plasmids was evaluated by Western blot of transfected HEK-293T cells (β-actin as control). The data are representative of three independent experiments. (f) Intracellular staining and flow cytometry analysis of NP and gp-NP expression from transfected HEK-293T. The data are representative of two independent experiments.
We next questioned whether this MHC-II presentation to CD4+ T cells could be enhanced to favour a better cellular immune response. We fused the M1 and NP genes with previously characterized gp100 MHC-II mobilization sequences (Fig. 1b; referred to as gp-M1 and gp-NP). These sequences enhance the MHC-II presentation of endogenous proteins by targeting them to endosomal compartments without disrupting presentation of MHC-I epitopes (Lepage & Lapointe, 2006).
To compare M1 and NP expression levels with or without gp100 MHC-II enhancing sequences in human cells, their expression in human embryonic kidney (HEK)-293T cells was assessed by Western blotting (M1) and intracellular staining (NP). Both M1 (27.9 kDa) and gp-M1 (39.9 kDa) were produced at comparable levels (Fig. 1e). The faster-migrating band in the gp-M1 lane represents the gp-M1 fusion protein after gp100 signal peptide (SS) cleavage. The proportions of NP- and gp-NP-transfected cells were also similar for both proteins, as were their expression level as determined by the mean fluorescence intensity (MFI) determined by flow cytometry (Fig. 1f). Thus, both WT (M1 or NP) and MHC-II-enhanced (gp-M1 or gp-NP) proteins were produced at similar levels. Accordingly, M1 and gp-M1 were equally recognized by a CD8+ T-cell clone specific to the well-characterized and conserved HLA-A2 restricted M158–66 epitope (Bednarek et al., 1991; Touvrey et al., 2009) (data not shown). Overall, gp100 MHC-II mobilization sequences did not seem to affect M1 or NP expression levels.
Consistent with the previously reported enhancement of MHC-II presentation associated with the gp100 sequences (Lepage & Lapointe, 2006), gp-M1 and gp-NP resulted in higher gamma interferon (IFN-γ) secretion by CD4+ T-cell clones compared with their WT counterparts (Fig. 1d, grey bars), while MHC-I presentation to CD8+ T cells was slightly down-modulated (Fig. 1c, grey bars). These results are also in line with M1 MHC-II enhanced presentation with autophagosome-associated protein Atg8/LC3 sequences (Schmid et al., 2007). It was unclear, however, if enhanced MHC-II presentation improved CD4+ and CD8+ T-cell expansion.
Expansion of human T lymphocytes by in vitro T-cell sensitization with endogenously expressed M1 and NP
To evaluate if enhanced MHC-II presentation improved CD4+ and CD8+ T-cell expansion, we next stimulated PBMCs with autologous CD40-B cells electroporated with M1, gp-M1, NP or gp-NP plasmids. Bulk T-cell cultures were restimulated according to the same procedure on day 7. T cells were tested for specificity to their relevant antigens on day 21 of the expansion protocol, 14 days after antigen stimulation to obtain sufficient cell counts.
Recognition assays from T-cell expansions were performed with three representative HLA-A2+ NDs. All donors developed M1- and NP-specific T-cell lines when stimulated with M1 and NP, with or without enhancement of MHC-II presentation, as determined by IFN-γ secretion (Fig. 2a, b). Addition of antibodies blocking MHC-I and -II presentation of APCs revealed that most of M1 and NP T-cell lines were composed of CD8+ T cells as IFN-γ secretion was inhibited by MHC-I-specific antibodies (black arrows). A similar IFN-γ response by CD8+ T cells was observed in three additional NDs (Supplementary Fig. S1a, c, d). Among these T-cell lines, at least one responded to the defined M158–66 epitope (Supplementary Fig. S1d), which is consistent with reports of the M158–66 epitope being an important part of a relatively broad influenza T-cell epitope response in HLA-A2+ NDs (Boon et al., 2002).
Characterization of human T lymphocytes expanded by in vitro sensitization with endogenously expressed M1 and NP. NP- and M1-specific T lymphocytes were expanded from PBMCs with autologous APCs expressing WT proteins or optimized versions of MHC-II presentation as described in Methods. The specificity of expanded T-cell lines was assessed by co-culture with APCs expressing M1, gp-M1 (a), NP or gp-NP (b), and IFN-γ secretion was assessed by ELISA. Where indicated, MHC restriction was identified by the addition of MHC-I (black arrow) or -II (white arrow) -blocking antibodies from ND#1, #2 and #3 T-cell lines grown as described in Methods. (c, d) The specificity of MHC-blocking antibodies was assessed with M1-specific CD8+ and NP-specific CD4+ T-cell clones for ND#1 and #2 (c) and ND#3 (d). Bars represent the mean±sd of duplicate cultures.
Furthermore, M1 and NP stimulations of bulk T cells resulted in the generation of CD4+ T cells secreting significant amounts of IFN-γ, since MHC-II blocking antibodies interfered with IFN-γ secretion by M1- (ND#2 and #3), NP- (ND#3) and gp-NP- (ND#2) grown T-cell lines (Fig. 2a, b, white arrows). Similar results were also obtained with gp-M1 expanded T-cell lines with two other NDs (Supplementary Fig. S1b, c). Moreover, a single-cell line may contain a heterogeneous population of both CD4+ and CD8+ T lymphocytes thereby explaining the effect of both MHC-I and -II-blocking antibodies on certain cell lines (M1-grown, ND#2 and #3). However, expansion of CD4+ T cells was independent of the presence of the MHC-II enhancing sequences. The efficacy of the MHC-blocking antibodies was controlled by the blockade or strong inhibition of IFN-γ secretion by previously isolated CD8+ or CD4+ T-cell clones (Fig. 2c, d).
Together, these results indicate that both CD8+ and CD4+ T cells can be simultaneously expanded by exposure to endogenous M1 and NP in an in vitro T-cell sensitization assay, with or without enhancement of MHC-II presentation, based on IFN-γ secretion from the bulk T-cell population. However, in addition to IFN-γ, other Th1 cytokines are required for optimal antiviral activity. Indeed, multi-cytokine secretion is intimately linked to robust immune cellular responses.
Effector cytokine secretion of M1- and NP-specific CD8+ and CD4+ T cells
Accordingly, we investigated the quality of M1 and NP CD8+ T cells generated by in vitro T-cell sensitization at the single-cell level. More specifically, we evaluated whether the intrinsic endogenous MHC-II presentation of M1 and NP antigens was sufficient for the generation of CD4+ T-cell-mediated help that could improve the quality of expanded CD8+ T cells. We, therefore, analysed multi-cytokine expression [IFN-γ, tumour necrosis factor (TNF) and interleukin (IL)-2] by M1- and NP-specific T cells after stimulation with their cognate antigen using intracellular cytokine staining (ICS). An example of the gating strategy is shown in Supplementary Fig. S2. While no IL-2 expression was detected in any of the CD8+ T-cell lines tested (data not shown), 0.5–4 % of T cells from each line produced both TNF and IFN-γ (Fig. 3a–c). Up to 6 % of T cells were single IFN-γ producers, and up to 3 % produced TNF only (Fig. 3a–c). Furthermore, CD8+ T-cell clones expanded with either gp-M1- or NP-expressing APCs presented the degranulation marker CD107a on their cellular surface after antigen-specific stimulation (Supplementary Fig. S3a). Macrophage inflammatory protein 1 beta (MIP-1β), an important chemokine for the recruitment of memory T cells and other immune cell types, was also detected in most M1- and NP-specific T-cell lines (Supplementary Fig. S3b).
Effector cytokine secretion by CD8+ T lymphocytes specific to M1 or NP, with or without enhancement of MHC-II presentation. T-cell lines were stained intracellularly for TNF, IFN-γ and IL-2 after 6 h of stimulation by CD40-B cells electroporated with relevant (cognate) or mock antigens. However, no IL-2 was detected in any of the CD8+ T-cell lines and was excluded from the analysis. (a, b) The percentage of CD3+/CD8+ T lymphocytes producing both TNF and IFN-γ was analysed in cell lines from three NDs cultured with CD40-B electroporated with M1 and NP, with or without gp100 MHC-II mobilization sequences. The background of mock-stimulated CD8+ T-cell lines was subtracted from the antigen-specific signal (see Supplementary Fig. S3b, available in JGV Online). Cytokine secretion from ND#1 (a) and #2 (b) was assessed with the same target as used for expansion, while cytokine secretion for ND#3 (c) was assessed with targets deprived of gp100 mobilization sequences.
Consistent with IFN-γ quantification (Fig. 2), MHC-II-enhanced presentation of M1 and NP did not result in increased CD4+ T-cell responders (Fig. 4). Comparable proportions of TNF+ or TNF+/IFN-γ+ CD4+ T cells were obtained in M1- and NP-specific T-cell lines expanded with either MHC-II-enhanced or WT sequences. In contrast to CD8+ T cells, low IL-2 levels were occasionally detected in CD4+ T cells (Fig. 4b).
Effector cytokine secretion by CD4+ T lymphocytes specific to M1 or NP, with or without enhancement of MHC-II presentation. T-cell lines from three NDs were cultured, stimulated and stained as in Fig. 3. The percentage of CD3+/CD4+ T cells producing TNF, IFN-γ or IL-2 was analysed after co-culture with APCs expressing M1 and NP, with or without gp100 MHC-II mobilization sequences. The background of mock-stimulated CD4+ T-cell lines was subtracted from the antigen-specific signal (Supplementary Fig. S2b). Cytokine production from ND #1 (a) and #2 (b) T-cell lines was assessed with the same target as used for expansion, while cytokine secretion for ND#3 (c) was assessed with targets deprived of gp100 mobilization sequences.
Overall, our results indicate that, in the context of in vitro T-cell sensitization, the intrinsic MHC-II presentation of endogenously expressed M1 and NP antigens is sufficient for Th1 cytokine secretion of specific CD4+ and CD8+ T cells. The secretion of Th1 cytokines further suggests an effector phenotype of the expanded M1- and NP-specific T lymphocytes.
Phenotype of M1- and NP-specific CD8+ and CD4+ T cells
To pursue phenotypic characterization of IFN-γ-secreting M1- and NP-specific T cells, their surface marker profiles were analysed. Most cells lacked the naive marker CD45RA, the secondary lymphoid tissue ‘homing’ receptor CD62L (Fig. 5), and the IL-7 receptor (CD127) (Supplementary Fig. S4), consistent with an effector T-cell phenotype. Again, there were no major differences between the phenotype of CD8+ T cells from either MHC-II-enhanced or WT M1- and NP-grown T-cell lines (Fig. 5a), and a similar phenotype was observed in CD4+ T cells (Fig. 5b).
Surface effector phenotype of T lymphocytes specific to M1 or NP, with or without enhancement of MHC-II presentation. T-cell lines from NDs were cultured, stimulated and stained as described in Methods. (a) The percentage of CD45RA−/CD62L− effector CD8+ T cells from three NDs was analysed after co-culture with APCs expressing M1 or NP, with or without gp100 MHC-II mobilization sequences. Cytokine secretion and surface marker expression from ND#1 and #2 were assessed with the same target as used for expansion, while cytokine secretion for ND#3 was assessed with targets deprived of gp100 mobilization sequences for ND#3. (b) The same analysis performed in (a) was used for CD45RA−/CD62L− effector CD4+ T cells of M1- or NP-specific T-cell lines from ND#1 and #2.
Discussion
A robust cellular immune response against conserved influenza M1 and NP could provide heterosubtypic immunity to influenza. Current TIVs do not induce a good cellular immune response. However, available modified LAIV (Mueller et al., 2010) or DNA vaccines in development (Kim & Jacob, 2009; Moss, 2009) are expected to trigger strong humoral and cellular immune responses. These influenza vaccines will most likely involve endogenous expression of antigens in APCs. Hence, this study focused on the characterization of MHC-I and -II presentation of endogenously expressed M1 and NP conserved influenza proteins by model APCs which expanded both CD8+- and CD4+-specific T lymphocytes from human PBMCs in in vitro T-cell sensitization assays. Furthermore, although MHC-II presentation was increased by the addition of gp100 MHC-II mobilization sequences, the quality of expanded M1- and NP-specific CD8+ and CD4+ T cells was comparable between MHC-II-enhanced or WT proteins expressed by APCs.
In line with previous reports that identified a variety of MHC-I and -II T-cell epitopes in M1 and NP recognized by most individuals' PBMCs (Gschoesser et al., 2002; Jameson et al., 1999; Kreijtz et al., 2008; Lee et al., 2008), we expanded both M1- and NP-specific T lymphocytes from PBMCs of different HLA-A2+ NDs. We have also identified some MHC epitopes recognized by these T cells (Doucet et al., 2010). However, many of the previous studies on influenza antigen presentation have focused on clonal T-cell populations that may have been altered during isolation. In contrast, we speculate that T-cell lines expanded in the current study more closely mimic in vivo polyclonal expansion. Accordingly, we obtained around 5–6 % of specific T lymphocytes, which is consistent with the proportion of antigen-specific T cells expanded in vivo.
Furthermore, both CD8+ and CD4+ M1- and NP-specific T cells were expanded. Classically, the MHC-II presentation of endogenously processed proteins to CD4+ T cells is considered limited and restricted to exogenous proteins (Rush et al., 2002; Van den Bosch et al., 2006; Voo et al., 2002). MHC-I and -II influenza antigenic cross-presentation has, however, been reported. Soluble NP can be cross-presented by MHC-I in PBMCs (Gschoesser et al., 2002), whereas endogenous WT M1 can be cross-presented by MHC-II (Jaraquemada et al., 1990; Nuchtern et al., 1990). Thus, we report similar NP MHC-II antigenic cross-presentation for the first time. It is also possible that M1 and NP were endocytosed from dying-electroporated APCs as a secondary classical MHC-II presentation mechanism. Nonetheless, there was MHC-II presentation to previously expanded CD4+ T cells (Fig. 1f). In this case, the short period of time between APCs’ electroporation and their co-culture with the CD4+ T cells should not allow for dead APCs to be taken up by other APCs for classical MHC-II antigenic processing.
Hence, the precise mechanism of endogenous M1 and NP MHC-II cross-presentation is unclear. In the case of M1, it does not involve the classical endosomal secretory pathway (Jaraquemada et al., 1990; Nuchtern et al., 1990), the proteasome, or the transporter associated with antigen processing (TAP), but does instead depend on lysosomal proteases (Guéguen & Long, 1996). Considering the nuclear and cytoplasmic localization of these two proteins, autophagy, involved in influenza A replication, could be a plausible MHC-II cross-presentation mechanism (Schmid et al., 2007; Zhou et al., 2009). Accordingly, M1 MHC-II presentation was enhanced by autophagosomal targeting sequences (Schmid et al., 2007).
MHC-II antigenic cross-presentation of influenza antigens is of particular interest since the anti-influenza CD4+ helper T cells are crucial for specific cellular immune responses (Janssen et al., 2003; Maecker et al., 1998; Swain et al., 2006). Our in vitro experimental setting did not allow long-term T-cell culture and therefore prevented the evaluation of the role of CD4+ T cells in long-term CD8+ memory T-cell expansion. Nevertheless, expanded CD4+ T cells could directly participate in the peripheral anti-influenza immune response by secreting inflammatory cytokines/chemokines (Nakanishi et al., 2009; Swain et al., 2006). Indeed, mouse memory lung CD4+ T cells have recently been shown to direct enhanced protection from influenza virus infection through effector functions such as IFN-γ secretion (Teijaro et al., 2010).
Multifunctionality is a hallmark of an efficient antiviral T-cell response (Akondy et al., 2009; De Rosa et al., 2004; Miller et al., 2008; Seder et al., 2008) and is less characterized in human influenza-specific T cells (Lee et al., 2008; Touvrey et al., 2009), particularly after stimulation with multiple potential epitopes. In all donors, TNF- and/or IFN-γ-secreting M1- and NP-specific CD8+ T lymphocytes were detected, indicating the presence of Tc1 influenza-specific effector CD8+ T cells implicated in influenza virus clearance and protection (Baumgarth & Kelso, 1996; Deng et al., 2004; Hikono et al., 2006).
Furthermore, M1- and NP-specific T-cell lines secreted MIP-1β. C-C chemokine receptor 5 (CCR5), MIP-1β’s receptor, is transiently upregulated on CD8+ memory T cells after influenza infection and has been shown to be crucial for virus control in mice (Kohlmeier et al., 2008). M1- and NP-specific T cells also degranulated upon antigen-specific stimulation, suggesting potential cytotoxic activity that is also critical in influenza virus clearance, although the presence of lytic enzymes was not assessed.
Altogether, our data suggest sufficient MHC-I and -II presentation of M1 and NP proteins for multifunctional effector T-cell expansion, at least upon short-term stimulation. Notably, NP-specific mouse CD8+ T cells also present an effector phenotype 10 days following NP DNA vaccine immunization or influenza infection (Supplementary Fig. S5).
We also observed a comparable effector phenotype of M1- and NP-specific T cells expanded with or without MHC-II enhanced presentation, as determined by CD45RA and CD62L expression. However, in contrast to other studies carried out on influenza M158–66-specific T cells (Touvrey et al., 2009), we did not detect CD127, a marker present on both naive and memory T cells, on M1- and NP-specific effector (CD45RA−/CD62L−) T cells. The lack of this marker may be explained by the early stage of effector T cells which did not yet re-express detectable levels of CD127 or by subtypes of memory T cell not expressing CD127 (Bachmann et al., 2005; Boettler et al., 2006; Touvrey et al., 2009). We also used IL-2 to expand T cells to sufficient numbers for our experiments, which could have favoured an effector phenotype.
Taken together, our results demonstrate that endogenously expressed influenza M1 and NP are sufficiently well presented by MHC-I and -II for in vitro short-term CD8+ and CD4+ T-cell activation. The mechanism by which influenza M1 and NP are cross-presented by MHC-II remains to be defined but could involve autophagy, among other alternative MHC cross-presenting pathways identified recently. Thus, the diversity of antigen presentation mechanisms appears to depend on the pathogen studied and could, therefore, contribute to reconfiguration and improvement of influenza vaccine strategies.
Methods
PBMCs.
PBMCs obtained from healthy individuals after informed consent were separated from heparinized donor blood as described previously (Pelletier et al., 2009). PBMCs were cryopreserved in 90 % FBS (Wisent)/10 % DMSO (Sigma), and stored in liquid nitrogen.
Generation of CD40-B cells.
CD40-B cells were generated as described previously (Lapointe et al., 2003). Briefly, recombinant soluble CD40L (1000 ng ml−1; Immunex Corporation) and recombinant human IL-4 (250 U ml−1; Peprotech) were added to PBMCs on the first day which were then cultured in complete medium, consisting of Iscove’s modified Dulbecco’s complete medium (Invitrogen) supplemented with 7.5 % human AB serum (heat-inactivated; Gemini Bio-products), 2 mM l-glutamine, 100 U penicillin ml−1, 100 g streptomycin ml−1, and 10 g gentamicin (Wisent) ml−1. Fresh complete medium containing 250 U IL-4 ml−1 and 1000 ng CD40L ml−1 was added on day 3. After the first round of proliferation (days 5–8), cells were either frozen for future use or restimulated every 2–3 days when the culture reached a density of 1.5–2×106 cells ml−1, about 40–90 % of proliferating cells being CD19+ HLA-DR+ B lymphocytes.
Plasmids.
NP and M1 from the influenza virus A/Puerto Rico/8/1934/H1N1 (PR8) strain [uniprot # P03466 (NP) and # P03485 (M1)], from which the first methionine was deleted, were cloned between the putative NH2-terminal signal sequence (first 23 residues) and the last 84 residues from gp100, which includes previously characterized MHC-II compartment mobilization sequences (Lepage & Lapointe, 2006). The resulting constructs, named gp-NP, gp-M1 and WT NP and M1 cDNA sequences, were optimized with GeneOptimizer from Geneart and cloned in pcDNA3.1(+) plasmid (Invitrogen). The plasmids were transformed into Escherichia coli one shot top 10 competent cells (Invitrogen) and prepared with Mega-Prep kit (Qiagen).
Cell lines.
HEK-293T cells, obtained from ATCC, were cultured in RPMI 1640 (Wisent) supplemented with 10 % heat-inactivated FBS (Wisent), 2 mmol l-glutamine L−1, 100 U penicillin/streptomycin ml−1 and 10 µg gentamicin (R-10) ml−1. Cells were cryopreserved in 90 % R-10/10 % DMSO, and stored in liquid nitrogen.
Western blotting.
HEK-293T cells were transfected with M1 or gp-M1 plasmids using Lipofectamine and Plus reagents (Invitrogen) according to the manufacturer’s instructions. After 24 h, protein extracts from pelleted cells were prepared, quantified and resolved by 10 % SDS-PAGE, as performed previously (Turcotte et al., 2007). An anti-M1 mouse mAb (1/200, clone GA2B; AbD Serotec) and a secondary peroxidase-conjugated goat anti-mouse antibody (1/10 000; Chemicon) were employed for Western blotting revelation.
Electroporation of APCs.
APCs were electroporated with a MP-100 microporator (Digital-bio) following procedures according to the manufacturer. Briefly, CD40-B cells (1–2×106) were sedimented for 15 min at 700 r.p.m. (100g) on a 164 mm diameter SO-1T rotor, resuspended in 200 µl of resuspension buffer (Digital-bio) and mixed with 3 µg 10−6 cells of DNA. The cells were immediately electroporated with one pulse at 1700 V for 20 ms and resuspended at 1×106 cells ml−1 of Iscove’s modified Dulbecco’s complete medium containing 10 % FBS and 2 mM l-glutamine (all from Wisent), without antibiotics. GFP-electroporated CD40-B cells were used as controls for transgene expression in every experiment. GFP expression in CD40-B cells was around 30 % at the initiation of the co-culture with PBMCs (data not shown).
Expansion of M1- or NP-specific peripheral T lymphocytes and cloning of antigen-specific bulk T-cell cultures.
PBMCs were stimulated with autologous CD40-B cells electroporated with the M1, gp-M1, NP or gp-NP DNA plasmids in a 2 : 1 ratio. On day 7, bulk T cells were restimulated according to the same procedure, and 150–300 IU IL-2 (Feldan Bio) ml−1 was added to the cultures on the following day and every 2–3 days. T-cell specificity was assessed on day 21 by ELISA or by intracellular cytokine assay on the basis of cytokine secretion. T-cell lines or clones (5×104) were co-cultured with CD40-B cells (5×104) electroporated with DNA or pulsed with peptides. IFN-γ secretion was measured by ELISA as described previously (Pelletier et al., 2009). In some recognition assays, CD40-B cells were pre-pulsed for 20 min at 37 °C under 5 % CO2 with 40–80 µg ml−1 of blocking antibodies specific for either MHC-I (clone W6/32) or MHC-II (clone IV A12).
M1- or NP-specific bulk T cells expanded as mentioned earlier were cloned by limiting dilution and cultured as described previously (Lapointe et al., 2003). T-cell clone phenotypes were analysed by flow cytometry, and their specificities were evaluated as before by co-culture with M1- or NP- electroporated CD40-B cells on the basis of their IFN-γ secretion.
Recognition assays.
IFN-γ ELISAs were performed as before (Pelletier et al., 2009), while MIP-1β ELISAs were performed with the CCL4/MIP-1β Duoset kit (R&D Systems) according to the manufacturer’s instructions.
For ICS, T-cell lines were co-cultured with autologous CD40-B cells electroporated with M1, gp-M1, NP or gp-NP DNA plasmids in a 2 : 1 ratio for 1 h and for an additional 6 h in the presence of brefeldin A (5 µg ml−1; Sigma). The cells were stained with Alexa-700, Pacific Blue and Allophycocyanin (APCy)-H7-conjugated antibodies specific to human CD3, CD4 and CD8, respectively (all from BD Biosciences). For T-cell phenotype analysis, the cells were also stained with APCy, PE and FITC-conjugated antibodies specific for human CD62L, CD127 and CD45RA or corresponding isotype controls (all from BD Biosciences). Dead cells were excluded by staining with the live/dead fixable dead cell stain kits (Invitrogen) when indicated. The cells were surface stained, fixed and permeabilized with FoxP3 staining buffer set (eBioscience) according to the manufacturer’s instructions. Intracellular staining was performed with antibodies against IFN-γ (eBioscience), TNF, IL-2 (the last two from BD Biosciences) or corresponding isotype controls. Flow cytometry data were acquired using the LSRII instrument (BD Bioscience), and analysed by FlowJo software (Tree Star) with a Boolean gating strategy.
For intracellular detection of NP, HEK-293T cells were transfected as before, with NP or gp-NP with a fluorescein-conjugated antibody against influenza A NP (cat #12-030; Argene).
Acknowledgements
This work was supported by the Canadian Institutes of Health Research (CIHR; grant # PAN-83153). R. L. and J.-D. D. are recipients of scholarships from Fonds de la recherche en santé du Québec (FRSQ). N. A. holds a Donald Paty Career Development Award from the Multiple Sclerosis Society of Canada and a Chercheur-Boursier from the FRSQ. The authors thank Jessica Godin-Ethier and Mélissa Mathieu for helpful discussions, Alexandre Reuben for careful reading of this manuscript and Ovid Silva and the CRCHUM Research Support Office (Bureau d’aide à la recherche) for manuscript editing. The authors declare no competing financial interests.