Design and evaluation of a poly-epitope based vaccine for the induction of influenza A virus cross-reactive CD8 + T cell responses

Abstract

The availability of influenza vaccines that can induce broadly protective immune responses is highly desirable and could also mitigate the impact of future influenza pandemics. Ideally, these vaccines also induce virus-specific CD8 + T cells, which have been identified as an independent correlate of protection. In the present study, we explored the use of an artificial immunogen that comprises of twenty highly conserved influenza virus CD8 + T cell epitopes with an HLA coverage of 99.5% of the world population. The highly attenuated viral vector Modified Vaccinia virus Ankara (MVA) was used to deliver the artificial poly-epitope sequence (rMVA-PE) and by using T cell lines raised against individual epitopes, we confirmed that the epitopes are liberated from the artificial immunogen. For efficient antigen processing and presentation, the epitopes were separated by spacer sequences. Stimulation of peripheral blood mononuclear cells of HLA-typed blood donors with rMVA-PE resulted in the activation of influenza virus-specific T cell responses. Furthermore, immunization of humanized HLA-A2.1-/HLA-DR1-transgenic H-2 class I-/class II-knockout mice (HLA-A*02:01) with rMVA-PE induced influenza virus-specific CD8 + T cell responses. Thus, rMVA-PE proved to be immunogenic both in vitro and in vivo and constitutes a promising vaccine candidate for the induction of cross-reactive CD8 + T cell responses that could afford protection against antigenically distinct influenza A viruses (IAV) of various subtypes and species, and is currently considered for further clinical testing.

Keywords: CD8 + T cells; Conserved epitope; Cross-reactivity; Influenza virus; Universal vaccine.

Fig. 1

Stepwise selection of human MHC class I epitopes from the Immune Epitope Database (IEDB) for inclusion in the poly-epitope construct. 972 epitopes were retrieved from the IEDB using search terms linear peptide, IAV (ID: 11320), assay for T cell ligand, MHC Ligand, MHC restriction: class I, Host: human. IEDB accessed on: September 2020. Subsequently, 661 epitopes were located in the internal proteins PB1, PB2, PA, NP, and M1. Among these, 483 epitopes were 8–11 amino acids in length. Only epitopes were considered that induced IFN-γ responses in PBMC and that were highly conserved in human, swine and avian IAV (conserved in > 50% of all strains in the NCBI Influenza Virus database, see Table 1). From these epitopes, 20 were selected accounting for 99.5% HLA coverage of the world population.

Fig. 2

Construction and genetic characterisation of rMVAs. (a) Schematic diagram of the MVA genome with the major deletion sites I to VI. The site of deletion III served for insertion of the gene of interest (GOI) sequence. GOI was controlled by the virus-specific promoter PmH5 and inserted via homologous recombination between MVA DNA sequences (flank-1 and flank-2) adjacent to deletion site III in the MVA genome and copies cloned in the MVA vector plasmid pIIIH5redK1L. Expression of the red fluorescent marker protein mCherry was used during plaque purification. Repetition of short flank-1 derived DNA sequences (del) served to remove the marker gene by intragenomic homologous recombination (marker gene deletion). Construction of the artificial poly-epitope gene (PE) is shown in yellow. The epitopes of interest were fused to ubiquitin (blue) and separated by spacer sequences (black). The addition of an HA-tag (green) serves for easy monitoring of gene expression. For control purposes also the M1 gene from IAV A/Porto Rico/8/34 was cloned similar to the artificial poly-epitope immunogen. (b) Proper insertion of the genes of interest into MVA and intragenomic deletion of the marker gene mCherry was confirmed by PCR analysis using viral DNA as template and deletion III site-specific oligonucleotide primers. PCR amplified a characteristic 1.5 kb DNA product from rMVA with M1 (rMVA-M1) genomic DNA and 1.7 kb polyepitope with rMVA-PE genomic DNA. The expected 0.762 kb DNA fragment was obtained from wt-MVA DNA, which used as a positive control for PCR. Uninfected CEFs and H₂O were used as negative controls. A full picture of the agarose gel shown in the supplementary document. (c) Immunostaining to confirm protein expression. CEFs were infected with the respective rMVAs or wt-MVA or left untreated. Cells were subsequently stained with antibody preparations against the HA-tag (Anti-HA) or VACV protein A27L (Anti-Vaccinia) as indicated. (d) Multiple-step replication analysis of recombinant MVAs (rMVA-PE and rMVA-M1) and wt-MVA in CEFs and A549 cells. After infection at a moi of 1, cell cultures were collected at the indicated time points and infectivity was determined by titration on CEFs. rMVAs and wt-MVA replicated efficiently in CEF cells but failed to replicate in A549 cells.

Fig. 3

In vitro antigenicity test with rMVA constructs. (a) NP₄₄₋₅₂ (CTELKLSDY)-specific CD8 + T cells incubated with rMVA- or wt-MVA-infected HLA-A*01 transgenic A549 cells at moi of 1 and 3. The CD8 + T cells response were measured by IFN-γ ELISpot assay. Only stimulation with rMVA-PE activated specific T cells. NP₄₄₋₅₂ peptide loaded cells were used as a positive control. wt-MVA infected and uninfected transgenic A549-A*01 cells were used as negative controls, as well as T cells only. (b) Graphical representation of ELISpot data produced with CD8 + T cells directed against the four different peptides, PB1₃₀₋₃₈ (YSHGTGTGY), PB1_591−599 (VSDGGPNLY), M1₅₈₋₆₆ (GILGFVFTL) and NP₄₄₋₅₂ (CTELKLSDY) that were incubated with HLA-matched transgenic A549 cells infected (moi 3) with wt-MVA or rMVAs or loaded with peptide. The number behind the peptide represents the position of the peptide in a poly-epitope construct. Data are expressed as spot-forming unit (SFU) per 10⁶ CD8 + T cells after subtraction of values obtained with uninfected HLA-A*01 transgenic A549 cells. Data were obtained from two independent experiments performed in duplicate (n = 4). (c) Response of PB1_591−599 (VSDGGPNLY)-specific CD8 + T cells after stimulation with autologous monocytes infected with rMVAs or wt-MVA infected (moi 3) or loaded with peptide measured by IFN-γ ELISpot assay. Activation of PB1_591−599 -specific T cells was observed with rMVA-PE but not with wt-MVA or MVA-M1. Number behind the peptide represents the position of the peptide in a poly-epitope construct. Data are expressed as spot-forming unit (SFU) per 10⁶ CD8 + T cells after subtraction of values obtained with uninfected monocytes. Data derived from two independent experiments performed in duplicate (n = 4). Means ± standard deviation (SD) are shown. ∗∗, p < 0.01; ∗∗∗, p < 0.001; ∗∗∗∗, p < 0.0001; ns, not significant. p values of all shown significance indicators were determined by Two-way ANOVA and Tukey’s multiple comparison test.

Fig. 4

In vitro antigenicity test with rMVA constructs using CD8 + T cells expanded after stimulation with IAV. IAV-specific CD8 + T cells obtained from two blood donors (HLA-A*02:01 and A*01:01, respectively) were incubated with corresponding HLA- transgenic A549 cells infected with the indicated rMVAs or wt-MVA and their response was measured by IFN-γ ELISpot assay. M1₅₈₋₆₆ and PB1_591−599 loaded A549 cells were included as positive controls. Data are expressed as spot-forming unit (SFU) per 10⁶ CD8 + T cells after subtraction of values obtained with uninfected A549 cells. Data derived from two independent experiments performed in duplicate (n = 4). Means ± SD are shown. ∗∗, p < 0.01; ∗∗∗∗, p < 0.0001; ns, not significant. p values were determined by Two-way ANOVA and Tukey’s multiple comparison test.

Fig. 5

In vitro immunogenicity of rMVA-PE. In vitro immunogenicity of rMVA-PE was assessed by stimulating PBMCs of HLA-A*01:01 and A*02:01 positive blood donors with rMVA-PE, wt-MVA or IAV Ned98 (moi of 1) as indicated. After 12 days, the frequency of peptide-specific CD8 + T cells was determined by IFN-γ ELISpot assay after re-stimulation with corresponding HLA- transgenic A549 cell loaded with HLA-A*01:01 peptides NP₄₄₋₅₂ (CTELKLSDY) (a, d) and PB1_591−599 (VSDGGPNLY) (b, e) or HLA-A*02:01 peptide M1₅₈₋₆₆ (GILGFVFTL) (c, f). The number behind the peptide represents the position of the peptide in a poly-epitope construct. Means ± SD are shown (n = 3). ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗∗, p < 0.0001. p values were determined by Two-way ANOVA and Tukey’s multiple comparison test.

Fig. 6

In vivo immunogenicity of rMVA-PE. (a) Experimental protocol. HLA-A2.1-/HLA-DR1-transgenic H-2 class I-/class II-knockout mice (n = 6) were immunized intramuscularly with 10⁷ PFU of wt-MVA, or rMVA-PE, or with vaccination buffer (Mock) on days 0 and 21 (A). On day 35, splenocytes were isolated and tested by ELISpot assay to detect wt-MVA and rMVA-PE-induced CD8 + T cell responses (b, c). Splenocytes were incubated with A*02:01 restricted influenza virus-specific peptide M1₅₈₋₆₆ (b) which is in the 5th position in the poly-epitope vaccine or MVA-specific peptide A6L₆₋₁₄ (c). Data are expressed as spot-forming unit (SFU) per 10⁶ splenocytes after subtraction of background control. Frequencies of IFN-γ producing CD8 + T cells were also determined by intracellular cytokine staining and flow cytometry (d, e). Splenocytes were stimulated with the HLA-A*02:01 restricted peptide M1₅₈₋₆₆ or MVA-specific peptide A6L. The data points represent individual mice (n = 6). The graphs represent the mean ± SD. n.s. not significant (p > 0.05); **, p < 0.01 (Mann-Whitney test).