IL-23-dependent IL-17 drives Th1-cell responses following Mycobacterium bovis BCG vaccination

Abstract

The generation of effective type 1 T helper (Th1)-cell responses is required for immunity against intracellular bacteria. However, some intracellular bacteria require interleukin (IL)-17 to drive Th1-cell immunity and subsequent protective host immunity. Here, in a model of Mycobacterium bovis Bacille Calmette-Guerin (BCG) vaccination in mice, we demonstrate that the dependence on IL-17 to drive Th1-cell responses is a host mechanism to overcome bacteria-induced IL-10 inhibitory effects. We show that BCG-induced prostaglandin-E2 (PGE2) promotes the production of IL-10 which limits Th1-cell responses, while simultaneously inducing IL-23 and Th17-cell differentiation. The ability of IL-17 to downregulate IL-10 and induce IL-12 production allows the generation of subsequent Th1-cell responses. Accordingly, BCG-induced Th17-cell responses precede the generation of Th1-cell responses in vivo, whereas the absence of the IL-23 pathway decreases BCG vaccine-induced Th17 and Th1-cell immunity and subsequent vaccine-induced protection upon M. tuberculosis challenge. Importantly, in the absence of IL-10, BCG-induced Th1-cell responses occur in an IL-17-independent manner. These novel data demonstrate a role for the IL-23/IL-17 pathway in driving Th1-cell responses, specifically to overcome IL-10-mediated inhibition and, furthermore, show that in the absence of IL-10, the generation of BCG-induced Th1-cell immunity is IL-17 independent.

Figure 1. Vaccine-induced IL-17 responses are required for Th1 responses following BCG vaccination

B6 and il17ra^−/− mice were subcutaneously vaccinated or left unvaccinated with 1×10⁶ BCG and the number of IFNγ-producing CD4⁺ T cells (A) or CD8⁺ T cells (B) in DLN cells detected by intracellular staining and flow cytometry on day 14. Ag85B-specific Th1 cell responses were detected in BCG-vaccinated or unvaccianted B6 and il17ra^−/− DLNs on day 14 post vaccination using antigen-driven ELISpot assay (C). mRNA levels of specific genes relative to GAPDH expression was determined in DLNs cells of B6 and il17ra^−/− using RT-PCR (D). The data points represent the mean (±SD) of values from 4–6 mice (A–D). _*, p ≤ 0.05. _**, p ≤0.05. One of two or more independent experiments shown.

Figure 2. IL-23-dependent IL-17 is required for BCG vaccine-induced Th1 immunity and protection following M.tuberculosis challenge

B6 and il23p19^−/− were vaccinated with BCG or left unvaccinated and the number of Ag85B-specific Th17 cells (A) and Ag85B-specific Th1 cells (B) in DLNs determined on day 14 by antigen-driven ELISpot assay. B6, il23p19^−/− mice were vaccinated and rested for 30 days, then challenged with ~100 CFU M. tuberculosis by the aerosol route. 30 days later lung CFU was determined in unvaccinated and vaccinated mice (C). The number of Ag85B-specific CD4⁺ Th1 and Th17 cells were determined in DLNs of BCG-vaccinated or unvaccinated B6 mice (D) or il17ra^−/− (E) by antigen-driven ELISpot assay. Log₁₀ fold induction of IL-17 and IFNγ mRNA in CD4⁺ sorted cells (F) in B6 BCG-vaccinated DLN cells compared to Unvaccinated B6 control cells determined by RT-PCR. The data points represent the mean (±SD) of values from 4–6 mice (A–F). _*, p ≤0.05. _**, p ≤0.005. _***, p ≤0.0005. One of two independent experiments shown.

Figure 3. BCG-induced IL-10 inhibits generation of Th1 responses

B6 DCs were either untreated (Un), stimulated with BCG (MOI-5) for 24 hours and supernatants assayed for IL-23 (A), IL-12 (B) and IL-10 (C) production. B6 DCs were untreated (Un), stimulated with BCG and isotype antibody control (BCG+Iso) or BCG and IL-10 neutralizing antibody (BCG+α-IL-10) and the levels of IL-12 determined (D). In some experiments, B6 DCs were left untreated (Un), stimulated with BCG and isotype antibody control (BCG+Iso) or BCG and IL-10 neutralizing antibody (BCG+α-IL-10) and cultured with naive OT-II TCR-Tg CD4⁺ T cells and OVA_323–339. On day 5, IFNγ protein levels determined in culture supernatants (E). Samples were treated in triplicates (A–E). B6 and il-10^−/− mice were vaccinated with BCG as described in Figure 1 and Ag85B-specific Th1 responses detected in DLNs on day 14 by ELISpot assays (F). B6, il10^−/− BCG vaccinated mice were rested for 30 days, following which they were challenged with ~100 CFU M. tuberculosis by the aerosol route. 30 days later lung CFU was determined in unvaccinated and vaccinated mice (G). Data points represent mean (±SD) from 4–5 mice (F–G). nd-not detectable, ns-not significant, _*, p ≤0.05. _**, p ≤0.005. _***, p ≤0.0005. One of two independent experiments shown.

Figure 4. BCG-induced PGE2 drives IL-23 and Ag85B-specific Th17 responses

DCs were untreated (Un), stimulated with BCG (MOI-5), or BCG and COX2 inhibitor (BCG+Celecoxib; 10µM). Since Celecoxib was dissolved in DMSO, a similar volume of DMSO was added into BCG treated wells (BCG+DMSO). Supernatants assayed for levels of PGE2 (A), IL-10 (B) and IL-12 (C). B6 DCs were stimulated with BCG+DMSO or BCG and COX2 inhibitor (BCG+Celecoxib) and cultured with naive OT-II TCR-Tg CD4⁺ T cells and OVA_323–339 and protein levels of IFNγ (D) determined in supernatants on day 5. Untreated samples shown in Figure 3E and 4D are the same. B6 DCs were either left untreated (Un), stimulated with BCG (MOI-5) or BCG and COX2 inhibitor (BCG+Celecoxib; 10µM) and levels of IL-23 determined (E). B6 DCs were stimulated with BCG+DMSO or BCG and COX2 inhibitor (BCG+Celecoxib) and cultured with naive OT-II TCR-Tg CD4⁺ T cells and OVA_323–339 and protein levels of IL-17 (F) determined in day 5 supernatants. Samples were treated in triplicates (A–F). B6 mice were BCG vaccinated or BCG vaccinated and treated with COX2 inhibitor (NS-398) and the Ag85B-specific Th17 responses (G) or Th1 responses (H) were determined in DLNs on day 14 by ELISpot assay. B6 BCG vaccinated mice were treated with DMSO or with COX2 inhibitor between days 1–15 post vaccination following which they were rested for 30 days and challenged with ~100 CFU M. tuberculosis by the aerosol route. 30 days later lung CFU was determined in unvaccinated and vaccinated mice (I). Data points represent mean (±SD) from 4–5 mice (G). nd-not detectable, ns-not significant, _*, p ≤0.05. _**, p ≤0.005. _***, p ≤0.0005. One of two independent experiments shown.

Figure 5. IL-17 is not required to drive Th1 responses in the absence of IL-10

DCs from B6 mice were left untreated (Un), stimulated with BCG (BCG), or BCG and IL-17 (100 ng/ml; BCG+IL-17) and supernatants assayed for levels of IL-12 (A) and IL-10 (B). Samples were treated in triplicates (A–B). B6 mice were BCG vaccinated and the serum was analyzed for PGE2 protein levels post vaccination (C), while DLNs were cultured in media for 24 hours and supernatants assayed for IFNγ protein (D). il10^−/− BCG vaccinated mice were treated with isotype control antibody or IL-17 neutralizing antibody and the Ag85B-specific Th1 responses were determined in DLNs on day 14 (E). B6 BCG-vaccinated mice treated with IL-17-neutralizing antibody or control isotype antibody were also included (E). The data points represent the mean (+/− SD) of values for 4–5 mice. nd-not detectable. *p< 0.05. **p<0.005, _***, p ≤0.0005, ns-not significant. One of two experiments shown.