Impact of ChAdOx1 or DNA Prime Vaccination on Magnitude, Breadth, and Focus of MVA-Boosted Immunogen-Specific T Cell Responses

Abstract

The efficacy of anti-viral T-cell vaccines may greatly depend on their ability to generate high-magnitude responses targeting a broad range of different epitopes. Recently, we created the HIV T-cell immunogen HTI, designed to generate T-cell responses to protein fragments more frequently targeted by HIV controllers. In the present study, we aim to maximize the breadth and magnitude of the T-cell responses generated by HTI by combining different vaccine vectors expressing HTI. We evaluated the ability to induce strong and broad T-cell responses to the HTI immunogen through prime vaccination with DNA plasmid (D) or Chimpanzee Adenovirus Ox1 (ChAdOx1; C) vectors, followed by a Modified Virus Ankara (MVA; M) vaccine boost (DDD, DDDM, C, and CM). HTI-specific T-cell responses after vaccination were measured by IFN-γ-ELISpot assays in two inbred mice strains (C57BL/6 and BALB/c). CM was the schedule triggering the highest magnitude of the response in both mice strains. However, this effect was not reflected in an increase in the breadth of the response but rather in an increase in the magnitude of the response to specific immunodominant epitopes. Immunodominance profiles in the two mouse strains were different, with a clear dominance of T-cell responses to a Pol-derived peptide pool after CM vaccination in C57BL/6. Responses to CM vaccination were also maintained at higher magnitudes over time (13 weeks) compared to other vaccination regimens. Thus, while a ChAdOx1 prime combined with MVA booster vaccination generated stronger and more sustained T-cell responses compared to three DNA vaccinations, the ChAdOx1 primed responses were more narrowly targeted. In conclusion, our findings suggest that the choice of vaccine vectors and prime-boost regimens plays a crucial role in determining the strength, duration, breadth, and focus of T-cell responses, providing further guidance for selecting vaccination strategies.

Keywords: HIV; T-cell vaccine; immune memory; immunodominance.

Figure 1

Vaccination schedule. The number of animals used, the week of each immunization, and the week of post-immunization (wpi) sampling are indicated for each vaccination regimen. C57BL/6 mice are represented by black and BALB/c mice by white. Created using Biorender.com.

Figure 2

Magnitude and breadth of the T-cell response to HTI. INFγ SFC/10⁶ spleen cells and the number of reactive peptide pools, after stimulation with 17 pools containing OLPs covering the HTI sequence, in C57BL/6 and BALB/c mice vaccinated with four different prime-boost regimens (DDD, DDDM, C, and CM). Statistically significant differences (p < 0.05) between treatment regimens were tested with the Kruskal–Wallis test with Dunn’s correction for multiple comparisons. Significant adjusted Dunn’s p-values are shown as ** < 0.01. Trends (p < 0.1) are indicated by grey numbers.

Figure 3

Distribution of the T-cell responses along the different HIV protein fragments that form HTI. The mean percentage of the total magnitude of the IFN-γ T-cell response to each of the four HIV proteins from which HTI fragments are derived is compared between C57BL/6 (left) and BALB/c (middle) mice and related to their proportions in the HTI sequence (right). Statistically significant differences were tested with a two-way ANOVA and multiple comparisons were performed with Tukey’s test. Sources of variation and the p-values of the different two-way ANOVA factors are indicated in the table below each graph. Statistically significant Tukey’s multiple comparisons between regimens for the different HTI protein fragments are indicated by colors (purple for Gag, Blue for Pol, Green for Vif, and yellow for Nef). Adjusted Tukey’s p-values are shown as * p < 0.05, by ** <0.01, by *** <0.001, and **** <0.0001.

Figure 4

Focus of the response to HTI. The pool-by-pool magnitude of the response is shown for C57BL/6 mice (up graph) and BALB/c mice (down graph). The graph shows significant Kruskal–Wallis and adjusted Dunn’s p-values comparing the four vaccination regimens (DDD, DDDM, C, and CM) responses for each pool. Statistical significance for Dunn’s p-values is shown as * if p < 0.05 by, ** if <0.01, and by *** if <0.001. At the bottom of each graph, a heat map indicates the % of animals responding to each pool.

Figure 5

OLP response mapping. (A) Responses above the threshold to individual OLPs in C57BL/6 and BALB/c mice vaccinated with the DDDM regimen (B) Responses to individual OLPs comparing DDDM with the CM regimen in C57BL/6. Responses between vaccination regimens were compared using multiple t-tests with false discovery rate correction. (C) IC50 of individual peptides comparing DDDM with the CM regimen in C57Bl/6. IC50s between OLPs were compared using the Kruskal–Wallis test with Dunn’s correction. Statistical significance is shown by p and q values.

Figure 6

Maintenance of the IFN-γ T-cell response over time. Magnitude (upper graph) and breadth (lower graph) are shown. Statistically significant differences were tested with a two-way ANOVA and multiple comparisons performed with Tukey’s test. Sources of variation and the p-values of the different two-way ANOVA factors are indicated in the table below each graph. Adjusted Tukey’s p-values are shown as * p < 0.05, by ** <0.01, and by *** <0.001.

Figure 7

Maintenance of the focus of the response. The responses of the individual pool responses at 3 and 13 wpi were compared using the Mann–Whitney test, with FDR correction for multiple comparisons. Only significant p-values (p < 0.05) with q < 0.01 are shown.