Heterologous Prime-Boost Combinations Highlight the Crucial Role of Adjuvant in Priming the Immune System

Abstract

The induction and modulation of the immune response to vaccination can be rationally designed by combining different vaccine formulations for priming and boosting. Here, we investigated the impact of heterologous prime-boost approaches on the vaccine-specific cellular and humoral responses specific for a mycobacterial vaccine antigen. C57BL/6 mice were primed with the chimeric vaccine antigen H56 administered alone or with the CAF01 adjuvant, and boosted with H56 alone, or combined with CAF01 or with the squalene-based oil-in-water emulsion adjuvant (o/w squalene). A strong secondary H56-specific CD4⁺ T cell response was recalled by all the booster vaccine formulations when mice had been primed with H56 and CAF01, but not with H56 alone. The polyfunctional nature of T helper cells was analyzed and visualized with the multidimensional flow cytometry FlowSOM software, implemented as a package of the R environment. A similar cytokine profile was detected in groups primed with H56 + CAF01 and boosted with or without adjuvant, except for some clusters of cells expressing high level of IL-17 together with TNF-α, IL-2, and IFN-γ, that were significantly upregulated only in groups boosted with the adjuvants. On the contrary, the comparison between groups primed with or without the adjuvant showed a completely different clusterization of cells, strengthening the impact of the formulation used for primary immunization on the profiling of responding cells. The presence of the CAF01 adjuvant in the priming formulation deeply affected also the secondary humoral response, especially in groups boosted with H56 alone or o/w squalene. In conclusion, the presence of CAF01 adjuvant in the primary immunization is crucial for promoting primary T and B cell responses that can be efficiently reactivated by booster immunization also performed with antigen alone.

Keywords: CAF01; T cells; adjuvants; computational flow cytometry; intracellular cytokines; prime-boost regimens; priming.

Figure 1

Study design and sample collection. Groups of C57BL/6 mice were subcutaneously primed with H56 alone (H56) or combined with CAF01, and boosted at day 28 with H56 + CAF01, or H56 + o/w squalene or H56 alone. Blood samples were collected at day 0, 12, 28, 31, and 38 following priming; draining lymph nodes (dLN) and spleens (Spl) were collected at day 7, 28, 31, and 38. The abbreviations of the different prime-boost combinations are here reported and used in all the figures.

Figure 2

Ag85B-specific CD4⁺ T cell response. C57BL/6 mice were subcutaneously immunized as summarized in Figure 1. Draining lymph nodes were collected 7 and 28 days after priming, and 3 and 10 days after booster immunization and analyzed for the frequency of Ag-specific CD4⁺ T cells, identified by staining with Ag85B-specific MHC class II tetramers (Tet-Ag85B). (A) Scatter plot of CD44 versus Tet-Ag85B, gated on live CD4⁺ lymphocytes, shown from a single animal representative of each immunization group, collected 10 days after boosting. Activated tetramer⁺ cells are gated and the frequency reported. (B) Time course analysis of the frequencies of tetramer⁺ CD4⁺ T cells, detected in each group, reported as mean ± SEM of five mice per group. Mann–Whitney test for multiple pairwise comparisons was used for assessing statistical difference between groups receiving the same booster and primed with H56 alone or H56 + CAF01 adjuvant. Kruskal–Wallis test, followed by Dunn’s post test for multiple comparison, was used to assess statistical difference among all groups (P ≤ 0.05).

Figure 3

Intracellular cytokines production. C57BL/6 mice were subcutaneously immunized as summarized in Figure 1. Spleens were collected 7 and 28 days after priming, and 3 and 10 days after boosting. Splenocytes were restimulated for 6 h with H56 protein. (A) Dot plots showing the production of TNF-α, IFN-γ, IL-17, IL-4/IL-13 versus IL-2 assessed on live CD4⁺ CD44⁺ lymphocytes in each group, collected 10 days after boosting. (B) Percentages of T cells positive for both IL-2 and the indicated cytokines, with respect to total CD4⁺ CD44⁺ cells, elicited by different vaccine formulations reported as mean ± SEM of five mice per group. Mann–Whitney test for multiple pairwise comparisons was used for assessing statistical difference between groups receiving the same booster and primed with H56 alone or H56 + CAF01 adjuvant. Kruskal–Wallis test, followed by Dunn’s post test for multiple comparison, was used to assess statistical difference among all groups (P ≤ 0.05).

Figure 4

Computational analysis of polyfunctional profiles of T cells. C57BL/6 mice were subcutaneously immunized as summarized in Figure 1. CD4⁺ and CD44⁺ cells from restimulated splenocytes were manually gated with FlowJov10, concatenated within the same immunization group, randomly downsampled to 15,000 cells and exported as uncompensated cells. Data were then compensated, logically transformed, and scaled with FlowSOM. A minimal spanning tree was obtained from the flowSet including FCS files of the groups primed with H56 + CAF01 and boosted with H56, H56 + CAF01, and H56 + o/w squalene. Size of the nodes is relative to the percentage of cells present in each cluster. Each node is represented by a star chart indicating the relative mean fluorescence intensity values of each cytokine; the height of each sector indicates the intensity, if the part reaches the border of the circle, the cells have maximal expression for that cytokine. A relative increase (blue nodes) or decrease (pink nodes) of at least twofold of the cell percentage in nodes of the CAF/CAF (A) or CAF/SQL (B) groups compared with the CAF/H56 group is shown. A magnification or colored nodes, indicated by numbers, is reported on the right.

Figure 5

Comparison of FlowSOM clustering in different prime-boost combinations. Scatter plots showing the frequency of cells within the 49 nodes obtained with the FlowSOM analysis. Groups with the same priming and different boost are shown in (A,B) (CAF/CAF versus CAF/H56, CAF/SQL versus CAF/H56, respectively); groups with the same boost and different priming are shown in (C,D) (CAF/CAF versus H56/CAF and CAF/SQL versus H56/SQL, respectively). Each dot represents a node, and the X and Y axis report the percentage of cells clustered into this node calculated in the indicated prime-boost combination. Within the dashed lines are reported nodes (black) with a similar frequency of cells assessed in the two immunization combinations; nodes with a frequency of cells >2-fold with respect to the other immunization group are reported in blue or red. The threshold value was chosen according to default threshold in FlowSOM package. The number of nodes is reported in each sector.

Figure 6

Antigen-specific IgG response. C57BL/6 mice were subcutaneously immunized as summarized in Figure 1. H56-specific IgG serum titers were analyzed on day 0, 12, 28, 31, and 38 following priming, by enzyme-linked immunosorbent assay. Antibody titers are expressed as the reciprocal of the highest serum dilution with an OD value ≥0.2 after subtraction of background value (diluent buffer). Values are reported as geometric mean titers (GMT) ± 95% CI of 8–10 mice per group from two independent experiments. Mann-Whitney test for multiple pairwise comparisons was used for assessing statistical difference between groups receiving the same booster and primed with H56 alone or H56 + CAF01 adjuvant, ^#P ≤ 0.05. Kruskal–Wallis test, followed by Dunn’s post test for multiple comparison, was used to assess statistical difference between all groups. *P ≤ 0.05.

Figure 7

Surface plasmon resonance analysis of H56-specific sera. A Biacore assay was performed on sera collected 10 days after booster immunization. Sera were diluted 200-fold and injected over H56 previously immobilized on CM4 sensor chip. (A) Sensorgrams of H56-specific antibodies binding (RU) from sera of single animals primed with the CAF01 adjuvant (blue lines) or with H6 alone (dashed lines). (B) Mean RU values ± SEM of five sera for group; filled bars indicate groups primed with CAF01 adjuvant, open bars with antigen alone. Mann–Whitney test for multiple pairwise comparisons was used for assessing statistical difference between groups receiving the same booster and primed with H56 alone or H56 + CAF01 adjuvant, ^#P ≤ 0.05. Kruskal–Wallis test, followed by Dunn’s post test for multiple comparison, was used to assess statistical difference between all groups. **P ≤ 0.01.