Expression and cellular immunogenicity of a transgenic antigen driven by endogenous poxviral early promoters at their authentic loci in MVA

Abstract

CD8(+) T cell responses to vaccinia virus are directed almost exclusively against early gene products. The attenuated strain modified vaccinia virus Ankara (MVA) is under evaluation in clinical trials of new vaccines designed to elicit cellular immune responses against pathogens including Plasmodium spp., M. tuberculosis and HIV-1. All of these recombinant MVAs (rMVA) utilize the well-established method of linking the gene of interest to a cloned poxviral promoter prior to insertion into the viral genome at a suitable locus by homologous recombination in infected cells. Using BAC recombineering, we show that potent early promoters that drive expression of non-functional or non-essential MVA open reading frames (ORFs) can be harnessed for immunogenic expression of recombinant antigen. Precise replacement of the MVA orthologs of C11R, F11L, A44L and B8R with a model antigen positioned to use the same translation initiation codon allowed early transgene expression similar to or slightly greater than that achieved by the commonly-used p7.5 or short synthetic promoters. The frequency of antigen-specific CD8(+) T cells induced in mice by single shot or adenovirus-prime, rMVA-boost vaccination were similarly equal or marginally enhanced using endogenous promoters at their authentic genomic loci compared to the traditional constructs. The enhancement in immunogenicity observed using the C11R or F11L promoters compared with p7.5 was similar to that obtained with the mH5 promoter compared with p7.5. Furthermore, the growth rates of the viruses were unimpaired and the insertions were genetically stable. Insertion of a transgenic ORF in place of a viral ORF by BAC recombineering can thus provide not only a potent promoter, but also, concomitantly, a suitable insertion site, potentially facilitating development of MVA vaccines expressing multiple recombinant antigens.

Figure 1. Schematic of transgene insertion at endogenous promoter driven locus of MVA, using B8R as an example.

(A) Surrounding open reading frames (ORFs) in the MVA genome are indicated by white arrows, with B8R highlighted in grey. Black “p” above black bar indicates predicted B8R early promoter core region (see Table 2), overlapping with the B7R ORF. The left homology arm (LHA) and right homology arm (RHA) sequences (white boxes), each 50 bp in length, were added by PCR to the ends of a cassette comprising a model antigen, tPA-Pb9-rLuc8PV (narrow grey arrow) and the bacterial selectable marker GalK (black arrow), with its bacterial promoter (small black triangle). The LHA was designed to place the initiation codon of tPA-Pb9-rLuc8PV in the same position as that of the B8R ORF. Crossed lines indicate homology arm recombination events between targeting amplicon and MVA-BAC. (B) After recombineering of this 2.4 kb targeting amplicon into MVA-BAC to replace B8R, tPA-Pb9-rLuc8PV was placed under control of the B8R promoter.

Figure 2. Activity of endogenous promoters compared to p7.5 and SSP in vitro.

BHK cells were infected with 1 pfu/cell of recombinant MVA carrying tPA-Pb9-rLuc8PV under the control of the indicated promoters. Renilla luciferase activity in the culture supernatant was quantified at 8 h post-infection. Cells were either untreated (grey bars) or exposed to 40 µM AraC during and after infection (open bars), to inhibit post-replicative gene expression. Data shown are the mean and standard deviation of duplicates after subtraction of signal at 1 h post-infection and are representative of two independent experiments.

Figure 3. Timecourse of endogenous promoter activities compared to p7.5 and SSP in vitro.

BHK cells were “spinoculated” (see Materials and Methods) with 1 pfu/cell of recombinant MVA carrying tPA-Pb9-rLuc8PV under the control of the indicated promoters. Renilla luciferase activity was quantified in the inoculum, the culture supernatant at various time points post-infection, and in the cell lysate at 24 h post-infection. Cells were either untreated (A) or exposed to 40 µM AraC during and after infection (B), to inhibit post-replicative gene expression. Data shown are the mean and standard deviation of duplicates and are representative of two independent experiments.

Figure 4. Cellular immunogenicity of recombinant MVA antigen (tPA-Pb9-rLuc8PV) driven by endogenous promoters compared to p7.5 and SSP in single-shot (A, B) or heterologous prime-boost (C) vaccination regimens.

For single-shot (A, B), BALB/c mice were immunized i.m. with 10⁶ pfu rMVA) and splenic CD8⁺ T cell responses to Pb9 peptide were determined 7 days later by intracellular cytokine staining and flow cytometry (A) or IFN-γ ELIspot (B). For prime-boost (C), BALB/c mice were immunized i.m. with 10⁸ infectious units of AdCh63-tPA-Pb9-rLuc8PV and 56 days later received 10⁶ pfu rMVA. At day 70 (14 days post-boost), splenic CD8⁺ T cell responses to Pb9 peptide were determined by IFN-γ ELIspot. Circles represent the responses of individual mice, with lines indicating the mean and the error bars showing SEM. See text for statistical analysis. The data shown are representative of two independent experiments.

Figure 5

(A) Timecourse of activity of mH5 promoter compared to p7.5 in vitro. BHK cells were “spinoculated” (see Materials and Methods) with 1 pfu/cell of recombinant MVA carrying tPA-Pb9-rLuc8PV under the control of the indicated promoters. Renilla luciferase activity was quantified in the inoculum and then in the culture supernatant at various time points post-infection. Cells were either untreated or exposed to 40 µM AraC during and after infection as indicated, to inhibit post-replicative gene expression. Data shown are the mean and standard deviation of duplicates. The data shown are representative of two independent experiments. (B) Cellular immunogenicity of recombinant MVA antigen (tPA-Pb9-rLuc8PV) driven by mH5 compared to p7.5 in a single-shot vaccination regimens. BALB/c mice were immunized i.m. with 10⁶ pfu rMVA and splenic CD8⁺ T cell responses to Pb9 peptide were determined 7 days later by intracellular cytokine staining and flow cytometry. Circles represent the responses of individual mice, with lines indicating the mean and the error bars showing SEM. See text for statistical analysis. The data shown are representative of two independent experiments.

Figure 6. Growth rates of rMVA expressing tPA-Pb9-rLuc8PV under the control of endogenous MVA promoters in comparison to MVA lacking tPA-Pb9-rLuc8PV but containing the same GFP marker gene (nrMVA).

BHK cells were infected with viruses at 1 pfu/cell (A), 0.5 pfu/cell (B) or 0.25 pfu/cell (C) and fluorescence of the viral GFP marker gene was quantified every 6 minutes for 36 h using a BMG FluoSTAR equipped with 37°C and 5% CO₂ incubation. Thick lines show the mean of two replicates and adjacent thin lines of the same colour represent the standard deviation. The data shown are representative of two independent experiments. Fluorescence intensity is expressed in arbitrary units.

Figure 7. Genetic stability of rMVA expressing tPA-Pb9-rLuc8PV under control of endogenous promoters.

Viruses were subjected to ten serial passages in CEFs, titred, and inoculated onto BHK cells at 0.001 pfu/cell. After 2 days, the cells were harvested and individually sorted into the wells of a 96-well plate using the CyCLONE attachment of a MoFlo flow cytometer. Two days later, renilla luciferase activity in the cell lysates was determined after scoring of wells as positive (+) or negative (−) for the viral GFP marker gene, indicating infection in the well. A cut-off of three standard deviations above the geometric mean of the GFP (dashed line labelled 3σ) was used to score GFP⁺ and GFP⁻ wells luciferase positive (Luc⁺) or negative (Luc⁻). Wells in which cell monolayers were lost during processing were excluded. Raw data for the pB8R recombinant (A) and well scores for all viruses (B) are shown.