Expression of CCL20 and granulocyte-macrophage colony-stimulating factor, but not Flt3-L, from modified vaccinia virus ankara enhances antiviral cellular and humoral immune responses

Abstract

While modified vaccinia virus Ankara (MVA) is currently in clinical development as a safe vaccine against smallpox and heterologous infectious diseases, its immunogenicity is likely limited due to the inability of the virus to replicate productively in mammalian hosts. In light of recent data demonstrating that vaccinia viruses, including MVA, preferentially infect antigen-presenting cells (APCs) that play crucial roles in generating antiviral immunity, we hypothesized that expression of specific cytokines and chemokines that mediate APC recruitment and activation from recombinant MVA (rMVA) vectors would enhance the immunogenicity of these vectors. To test this hypothesis, we generated rMVAs that express murine granulocyte-macrophage colony-stimulating factor (mGM-CSF), human CCL20/human macrophage inflammatory protein 3alpha (hCCL20/hMIP-3alpha), or human fms-like tyrosine kinase 3 ligand (hFlt3-L), factors predicted to increase levels of dendritic cells (DCs), to recruit DCs to sites of immunization, or to promote maturation of DCs in vivo, respectively. These rMVAs also coexpress the well-characterized, immunodominant lymphocytic choriomeningitis virus nucleoprotein (NP) antigen that enabled sensitive and quantitative assessment of antigen-specific CD8(+) T-cell responses following immunization of BALB/c mice. Our results demonstrate that immunization of mice with rMVAs expressing mGM-CSF or hCCL20, but not hFlt3-L, results in two- to fourfold increases of cellular immune responses directed against vector-encoded antigens and 6- to 17-fold enhancements of MVA-specific antibody titers, compared to those responses elicited by nonadjuvanted rMVA. Of note, cytokine augmentation of cellular immune responses occurs when rMVAs are given as primary immunizations but not when they are used as booster immunizations, suggesting that these APC-modulating proteins, when used as poxvirus-encoded adjuvants, are more effective at stimulating naïve T-cell responses than in promoting recall of preexisting memory T-cell responses. Our results demonstrate that a strategy to express specific genetic adjuvants from rMVA vectors can be successfully applied to enhance the immunogenicity of MVA-based vaccines.

FIG. 1.

hCCL20, mGM-CSF, hFlt3-L, and LCMV-NP are expressed from their respective rMVA vectors in vitro and in vivo. DF-1 cells were infected with MVA, MVA-NP, MVA-NP-hCCL20, MVA-NP-mGM-CSF, or MVA-NP-hFlt3-L and cells (A) or supernatants (B) were harvested at 24 hpi for analysis of LCMV-NP expression (A) or cytokine expression (B). (A) Infected cells were permeabilized and stained with a primary anti-LCMV-NP mouse monoclonal antibody (1:100), followed by incubation with a phycoerythrin-conjugated secondary anti-mouse IgG antibody (1:400), and were analyzed by flow cytometry. The negative controls are the MVA-infected cells, and the positive controls are the MVA-NP-infected cells. (B) Expression of hCCL20 and mGM-CSF in supernatants from cells infected with MVA, MVA-NP, MVA-NP-hCCL20, or MVA-NP-mGM-CSF was measured using cytokine-specific Quantikine ELISA kits. ND, not detected. (C and D) Mice were immunized i.p. with MVA-NP-mGM-CSF (C) or MVA-NP-hCCL20 (D), and sera were analyzed at 1, 6, and 24 h following immunization for mGM-CSF (C) or hCCL20 (D) expression using Quantikine ELISA kits. Data were compiled from at least two independent experiments; symbols represent values determined from individual mice. Neither mGM-CSF nor hCCL20 was detected in sera from mice immunized with control MVA or mock-immunized with PBS (not shown). (E) Supernatants from MVA-NP- or MVA-NP-hFlt3-L-infected cells were incubated with SDS buffer, resolved by SDS-polyacrylamide gel electrophoresis, and prepared for immunoblot analysis. hFlt3-L was detected using a biotinylated anti-Flt3-L antibody (0.2 μg/ml) followed by incubation with neutravidin-HRP conjugate (1:10,000). The positive control is a purified hFlt3-L, and the negative control is the supernatant from cells infected with MVA-NP.

FIG. 2.

Virally expressed hCCL20 or mGM-CSF enhances LCMV-NP_118-126-specific CD8⁺ T-cell responses following single-dose rMVA vaccination in mice. Groups of mice were immunized i.p. with MVA-NP, MVA-NP-hCCL20, MVA-NP-mGM-CSF, MVA-NP-hFlt3-L, or PBS, and their spleens were harvested at 7, 15, and 30 days after immunization for LCMV-NP_118-126 tetramer (A and B) and IFN-γ ICS (C and D) analyses. (A) Representative LCMV-NP_118-126-specific CD8⁺ T-cell responses at 7, 15, and 30 dpi from individual mice within each experimental group. LCMV-NP_118-126-specific CD8⁺ T cells were detected via flow cytometry following surface staining of splenocytes with an allophycocyanin-conjugated LCMV-NP_118-126 tetramer. The frequencies of LCMV-NP_118-126-specific CD8⁺ T cells are expressed as percentages of total CD8α⁺ splenocytes. The negative control is the naïve group, and the positive control is the MVA-NP group. (B) Frequencies of LCMV-NP_118-126-specific CD8⁺ splenocytes determined by MHC tetramer analysis compiled from three independent experiments are shown. The symbols denote values determined for individual mice, and the group means are denoted by the horizontal bars. At each time point, a nonparametric ANOVA (Kruskal-Wallis test) was performed to compare all immunization groups. Planned intergroup comparisons (P values shown) were performed via Dunn's multiple comparisons test only when the overall ANOVA was significant (P ≤ 0.05). NS, not significant (i.e., P > 0.05). (C) Representative LCMV-NP_118-126-specific CD8⁺ T-cell responses at 7, 15, and 30 dpi from individual mice within each experimental group. Interferon γ-producing LCMV-NP_118-126-specific CD8⁺ T cells were detected via flow cytometry following stimulation with the LCMV-NP_118-126 peptide. The frequencies of LCMV-NP_118-126-specific CD8⁺ T cells are expressed as percentages of total CD8α⁺ splenocytes. The negative control is the naïve group, and the positive control is the MVA-NP group. (D) Frequencies of LCMV-NP_118-126-specific CD8⁺ T-cell responses determined by IFN-γ ICS assays compiled from three independent experiments are shown. The symbols denote values determined for individual mice, and the group means are denoted by the horizontal bars. P values were calculated as described above for panel B. NS, not significant.

FIG. 3.

Virally expressed hCCL20 or mGM-CSF enhances MVA-specific CD8⁺ T-cell and antibody responses following single-dose vaccination in mice. Mice were immunized i.p. with MVA, MVA-NP-hCCL20, MVA-NP-mGM-CSF, or PBS. (A) Representative MVA-specific CD8⁺ and CD4⁺ T-cell responses at 7 dpi from individual mice within each experimental group. IFN-γ-producing MVA-specific CD8⁺ and CD4⁺ T cells were detected via flow cytometry following stimulation with MVA-infected A20 cells. The frequencies of MVA-specific CD8⁺ and CD4⁺ T cells are expressed as percentages of total CD8α⁺ or CD4⁺ splenocytes, respectively. The negative control is the naïve group, and the positive control is the group immunized with MVA. (B) Frequencies of MVA-specific CD8⁺ and CD4⁺ T cells (note different scales) were determined at 7 dpi in two independent experiments. The symbols denote values determined for individual mice, and the horizontal bars denote the group means. Nonparametric ANOVAs (Kruskal-Wallis tests) were performed to compare CD8⁺ or CD4⁺ T-cell responses between all immunization groups. (C) MVA-specific serum antibody titers were determined by ELISA at 7 days (n = 10 mice per group) and 30 days (n = 5 mice per group) following immunization with the indicated virus. The median titers, range of values, and P values for pairwise comparisons between groups of mice immunized with MVA-NP-mGM-CSF or MVA-NP-hCCL20 versus MVA using a two-tailed Wilcoxon rank sum test are shown.

FIG. 4.

Cytokine-expressing rMVA vectors do not elicit enhanced boosting of memory CD8⁺ T-cell responses. Mice were immunized with plasmid pCMVNP to prime LCMV-NP-specific CD8⁺ T-cell responses and were boosted 30 days later with MVA-NP, MVA-NP-hCCL20, MVA-NP-mGM-CSF, MVA-NP-hFlt3-L, or PBS (Mock). Frequencies of LCMV-NP-specific CD8⁺ splenocytes were measured at 7 days postboost by MHC-I-tetramer analysis and are expressed as percentages of total CD8α⁺ splenocytes. (A) Representative LCMV-NP_118-126-specific CD8⁺ T-cell responses from individual mice within indicated experimental groups. (B) Summary of LCMV-NP_118-126-specific CD8⁺ T cells measured in two independent experiments. Data points represent values derived from individual mice, and the bars denote the group means.