Adjuvanticity of a synthetic cord factor analogue for subunit Mycobacterium tuberculosis vaccination requires FcRgamma-Syk-Card9-dependent innate immune activation

Abstract

Novel vaccination strategies against Mycobacterium tuberculosis (MTB) are urgently needed. The use of recombinant MTB antigens as subunit vaccines is a promising approach, but requires adjuvants that activate antigen-presenting cells (APCs) for elicitation of protective immunity. The mycobacterial cord factor Trehalose-6,6-dimycolate (TDM) and its synthetic analogue Trehalose-6,6-dibehenate (TDB) are effective adjuvants in combination with MTB subunit vaccine candidates in mice. However, it is unknown which signaling pathways they engage in APCs and how these pathways are coupled to the adaptive immune response. Here, we demonstrate that these glycolipids activate macrophages and dendritic cells (DCs) via Syk-Card9-Bcl10-Malt1 signaling to induce a specific innate activation program distinct from the response to Toll-like receptor (TLR) ligands. APC activation by TDB and TDM was independent of the C-type lectin receptor Dectin-1, but required the immunoreceptor tyrosine-based activation motif-bearing adaptor protein Fc receptor gamma chain (FcRgamma). In vivo, TDB and TDM adjuvant activity induced robust combined T helper (Th)-1 and Th-17 T cell responses to a MTB subunit vaccine and partial protection against MTB challenge in a Card9-dependent manner. These data provide a molecular basis for the immunostimulatory activity of TDB and TDM and identify the Syk-Card9 pathway as a rational target for vaccine development against tuberculosis.

Figure 1.

Th1/Th17 induction by TDB correlates with a distinct APC activation program. (A) C57BL/6 mice were s.c. immunized twice with H1 antigen mixed with CpG (CpG) or DDA-TDB (TDB) or left untreated (control). 2 wk after the last injection, draining lymph node cells were restimulated with 10 μg/ml H1 antigen for 96 h, followed by detection of IFN-γ and IL-17. Mean and SD (n = 3–4 mice) from representative experiment out of three are shown. (B) BMMs were stimulated for 96 h in triplicate with 100 ng/ml LPS, 1 μM CpG ODN, and TDB, with or without prestimulation with 20 ng/ml IFN-γ. NO production measured as nitrites by Griess assay. (right) TDB and TDM concentration of 20 μg/ml. Mean and SD of triplicate wells. (C) Stimulation of IFN-γ–primed BMMs as indicated. (D) Gene expression changes induced by TDB and CpG in IFN-γ–primed BMMs. (top) Number of probe sets up-regulated by >10-fold. (bottom) Hierarchical cluster analysis of z-score–normalized expression of regulated probe sets (Max-Min>200, Max/Min>10). Yellow indicates high, blue indicates low relative expression. (E–G) Cytokine mRNA and ELISA protein data. Il12b averages of probe sets 1449497_at and 1419530_at; Il1b probe set 1449399_a_at; Il6 probe set 1450297_at. ELISA data are from triplicate wells. (H) Up-regulation of CD80, CD86, CD40, and MHC-II on DCs after stimulation with 30 μg/ml TDB (48 h; red line) or 100 ng/ml LPS (24 h; black line) compared with unstimulated cells (gray shade) as determined by FACS analysis. Representative of two experiments.

Figure 2.

TDB-/TDM-induced APC activation is independent of Myd88, yet requires Syk activity. (A and B) Wild-type, Myd88^−/−, and Syk^−/− BMMs primed with IFN-γ were stimulated for 60 h (A) and 96 h (B) before NO release was measured. TDB and TDM were used at 20 μg/ml in suspension (A and B) and coated onto tissue culture plastic in titrations ranging from 4 μg/ml to 15 ng/ml (A). Mean and SD of triplicate wells. Experiments were repeated three times with similar results. (C) TDB-induced macrophage activation is sensitive to Piceatannol. IFN-γ–primed BMMs were pretreated for 30 min with increasing doses of Piceatannol before stimulation with TDB (5 or 20 μg/ml) or LPS (100 ng/ml). Mean and SD of triplicate wells. Representative of two experiments.

Figure 3.

TDB/TDM activate APCs via Card9–Bcl10–Malt1. (A) BMM were stimulated with 1 μM CpG or 40 μg/ml TDB for 72 h. Expression of iNOS, IL-1β, and IL-6 was determined by qRT-PCR. Mean and SD of quadruplicate wells, representative of three independent experiments. (B) Expression profiling of wild-type and Card9^−/− IFN-γ–primed BMMs 48 h after stimulation with CpG, TDB, or Curdlan. Hierarchical clustering of 606 probe sets with regulated expression (Max/Min > 5, Max-Min > 100). Yellow indicates high expression, blue indicates low relative expression. The genotype of BMMs is indicated with + (wild type) and – (Card9^−/−). (C) Card9 dependence of TDB-induced production of IL-12p40, IL-23, and IL-12p70 from DCs. LPS and Curdlan were included as controls. Shown are mean and SD of triplicate cultures, harvested 72 h after stimulation. (D) Lack of TNF-α production from DCs deficient in Bcl10 or Malt1. Mean and SD, repeated twice. (E) IFN-γ–primed BMMs from Dectin-1^−/−, Card9^−/−, and C57BL/6 mice were stimulated in triplicates for 72 h, as described in Fig. 2 A, before measurement of nitrites using the Griess assay. Mean and SD of representative experiment from two performed. (F) Expression of Nos2 mRNA in IFN-γ–primed BMMs from C57BL/6, Dap12^−/−, and Fcer1g^−/− mice 48 h after stimulation as indicated. Fold change relative to medium control. Mean and SD of quadruplicate determinations, representative of at least two experiments. (G) NO production by C57BL/6 and Fcer1g^−/− IFN-γ–primed BMMs 72 h after stimulation as indicated. Mean and SD of triplicate wells. Representative of three experiments.

Figure 4.

Induction of protective immunity to tuberculosis by TDB/TDM adjuvant requires Card9. (A) Wild-type and Card9^−/− mice were immunized via hind footpad injection on day 0 and 20 (arrows) with H1 in DDA or DDA/TDB liposomes. Mean and SD of the increase in footpad thickness is shown (footpad number, n = 12). Significantly different for comparison of wild-type and Card9^−/− after DDA/TDB injection at all time points (P < 0.01). Representative of two independent experiments. (B) Number of cells in draining lymph nodes 2 wk after the 2 injections. Data are from two independent experiments (n = 3–5 for each data point). (C) Production of IFN-γ and IL-17 by draining lymph node cells after restimulation with H1 protein. Mean and SD (n = 4), two experiments with similar results. (D) IgG2a levels in sera of mice injected three times. Control, untreated mice; D, DDA; D/T: DDA/TDB; WT, wild type. (E) Mice injected three times, or naive mice, were infected by aerosol with 100 CFU MTB. After 6 wk, mice were killed and the bacterial load was determined in the lung. (left) Wild-type mice; (right) Card9^−/− mice. Shown are mean and SD from six mice per group, except for unvaccinated Card9^−/− mice, where two mice died before day 42. (F) Frequency analysis of CD4 T cells producing IL-17 (left) or IFN-γ (right) from mediastinal lymph nodes 6 wk after MTB infection. Asterisks indicate significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001).