A Recombinant MVA-Based RSV Vaccine Induces T-Cell and Antibody Responses That Cooperate in the Protection Against RSV Infection

Abstract

Respiratory syncytial virus (RSV) causes a respiratory disease with a potentially fatal outcome especially in infants and elderly individuals. Several vaccines failed in pivotal clinical trials, and to date, no vaccine against RSV has been licensed. We have developed an RSV vaccine based on the recombinant Modified Vaccinia Virus Ankara-BN^® (MVA-RSV), containing five RSV-specific antigens that induced antibody and T-cell responses, which is currently tested in clinical trials. Here, the immunological mechanisms of protection were evaluated to determine viral loads in lungs upon vaccination of mice with MVA-RSV followed by intranasal RSV challenge. Depletion of CD4 or CD8 T cells, serum transfer, and the use of genetically engineered mice lacking the ability to generate either RSV-specific antibodies (T11µMT), the IgA isotype (IgA knockout), or CD8 T cells (β2M knockout) revealed that complete protection from RSV challenge is dependent on CD4 and CD8 T cells as well as antibodies, including IgA. Thus, MVA-RSV vaccination optimally protects against RSV infection by employing multiple arms of the adaptive immune system.

Keywords: MVA; RSV; mode of action; protection; vaccines.

Figure 1

Effect of T-cell depletion on efficacy. BALB/c mice (n = 5) either were vaccinated (IN) twice 3 weeks apart with 1 × 10⁸ TCID₅₀ MVA-RSV or received TBS. RSV challenge (IN, 10⁶ pfu) was performed 2 weeks after the last vaccination. One day prior and 1 day after RSV challenge, mice were either treated (IP) with an anti-CD4 (CD4 depl.) or anti-CD8 (CD8 depl.) or an isotype-matched control antibody (isotype control). For long-term depletion of CD4 T cells, injection with the anti-CD4 antibody was performed 2 days before prime vaccination and thereafter twice a week. (A) RSV-specific IgG in serum was measured on Day 33 (12 days post second vaccination) by ELISA. Titers were comparable between vaccinated mice treated with the isotype control antibody and anti-CD4 or anti-CD8 antibody (p = 0.9965 for CD4-depleted mice and p = 0.5426 for CD8-depleted mice; Tukey test). (B) RSV-specific mucosal IgA and IgG in BAL fluids were measured on Day 39 (4 days post challenge) by ELISA. Mice treated with anti-CD4 or anti-CD8 antibody had comparable RSV-specific mucosal IgA (p = 0.9717 for CD4-depleted mice and p = 0.1087 for CD8-depleted mice, Tukey test) as well as mucosal IgG (p = 0.9991 for CD4-depleted mice and p = 0.8205 for CD8-depleted mice, Tukey test) to vaccinated mice treated with the isotype control antibody. No antibodies could be measured in the CD4 long-term depleted group. For both panels the geometric mean titers (GMT) with 95% confidence interval (CI) are shown. Representative data of one out of two mouse studies are shown. ns: not statistically significant. Four days post challenge mice were sacrificed and the viral load in lung was measured by (C) plaque assay. Mean pfu per half lung ± SEM is shown. (D) Viral load detected by RT-qPCR. Mean L-gene copies ± SEM is shown; the black line indicates the lower limit of detection (LLOD = 15 gcs). Plaque assay data showed significant differences between TBS-treated control mice and all other groups (***p < 0.0005, **p < 0.005 Dunn’s method). RT-qPCR data showed significant differences between TBS-treated control mice and vaccinated control mice, CD4 depl. mice, and CD8 depl. mice (****p < 0.00005, ***p < 0.0005 Dunn’s method). Pooled data of two separately performed mouse studies are shown (n = 9 or 10).

Figure 2

Efficacy in CD8 T cell (β2M -/-)- or antibody (T11µMT)-deficient C57BL/6 mice. C57BL/6 WT mice (n = 5), T11µMT (n = 5), or β2M -/- (n = 4) were vaccinated (IN) twice 3 weeks apart with 1 × 10⁸ TCID₅₀ MVA-RSV. WT mice treated with TBS served as controls. RSV challenge (IN, 10⁶ pfu) was performed 2 weeks after last vaccination. Four days post challenge mice were sacrificed, and the viral load in lung was measured by (A) plaque assay. Mean pfu per half lung ± SEM is shown. (B) Viral load detected by RT-qPCR. Mean L-gene copies ± SEM is shown; the black line indicates the lower limit of detection (LLOD = 15 gcs). Plaque assay data (pfu) showed significant differences between TBS-treated control mice and vaccinated control mice (*p < 0.05, Dunn’s method). RT-qPCR data (L-gene copies) showed significant differences between TBS-treated control mice and all other groups (***p < 0.0005, **p < 0.005, Tukey test). ns: not statistically significant.

Figure 3

Efficacy of passively transferred RSV-specific antibodies. Either BALB/c mice (n = 5) were vaccinated (IN) twice 3 weeks apart with 1 × 10⁸ TCID₅₀ MVA-RSV or non-vaccinated BALB/c mice received 1 ml of serum (IP) from MVA-RSV-vaccinated (positive serum) or mock-vaccinated control mice (mock serum) 1 day before challenge. (A) RSV-specific IgG in serum was measured on Day 39 (4 days post serum transfer) by ELISA. Geometric mean titers (GMT) with 95% confidence interval (CI) are shown. Titers were comparable between MVA-RSV-vaccinated mice and mice receiving positive serum (p = 0.4127, Dunn`s method). (B) RSV-specific mucosal IgA and IgG in BAL fluids were measured on Day 39 (4 days post serum transfer) by ELISA. GMT with 95% CI is shown. RSV-specific mucosal IgG titers were comparable between vaccinated mice and mice receiving RSV-specific serum antibodies (p = 0.2313, Dunn’s method). For Figures 3B and 1B , identical results for RSV-specific mucosal IgA and IgG for MVA-RSV-vaccinated control animals were shown as this experiment was run in parallel to reduce the number of animals required. (C) Viral load in lung was measured 4 days post RSV challenge (IN, 10⁶ pfu) by plaque assay. Mean pfu per half lung ± SEM is shown. Viral load (pfu) determined by plaque assay was significantly (****p < 0.00005;Mann–Whitney U-test) reduced by the passive transfer of RSV-specific antibodies compared to the viral load of mice receiving the mock serum. (D) Viral load detected by RT-qPCR. Mean L-gene copies ± SEM is shown; the black line indicates the lower limit of detection (LLOD = 15 gcs). There was no statistically significant difference (p = 0.066, Mann–Whitney U-test) in RSV-specific L-gene copies between mice receiving RSV-specific antibodies and mice receiving the mock serum. Pooled data of two separately performed mouse studies are shown (n = 9 or 10). ns: not statistically significant.

Figure 4

Role of IgA for RSV protection. C57BL/6 WT (n = 5 to 10) or IgA-deficient (IgA-/-, n = 5 to 10) mice were vaccinated (IN) twice 3 weeks apart with 1 × 10⁸ TCID₅₀ MVA-RSV. C57BL/6 WT mice treated with TBS served as controls. (A) RSV-specific IgG in serum was measured on Day 34 (13 days post last vaccination) by ELISA. Geometric mean titers (GMT) with 95% confidence interval (CI) are shown. Titers were similar between vaccinated WT mice and IgA -/- mice (p = 0.054, Mann–Whitney U-test). (B) RSV-specific mucosal IgA and IgG in BAL fluids were measured on Day 39 (4 days post challenge) by ELISA. GMT with 95% CI is shown. Mucosal IgG titers were comparable between vaccinated WT mice and IgA -/- mice (p = 0.4206, Mann–Whitney U-test). Mucosal IgA was only detected in vaccinated WT mice but not in IgA -/- mice, as expected. (C) Viral load in lung was measured 4 days post RSV challenge (IN, 10⁶ pfu) by plaque assay. Mean pfu per half lung ± SEM is shown. The lungs of vaccinated IgA -/- mice were completely free from infectious virus analogous to the vaccinated WT control group (*p < 0.05; Dunn’s method). (D) Viral load detected by RT-qPCR. Mean L-gene copies ± SEM is shown; the black line indicates the lower limit of detection (LLOD = 15 gcs). L-gene copies were significantly reduced in IgA -/- mice compared to TBS-treated WT mice (**p < 0.005, ***p < 0.0005, Tukey test). ns: not statistically significant. For Figures 4C, D and 2A, B , identical results for plaque assay (pfu) and RT-qPCR (L-gene copies) for TBS-treated and vaccine-treated control animals were shown as this experiment was run in parallel to reduce the number of animals required.