Caspase-2-dependent dendritic cell death, maturation, and priming of T cells in response to Brucella abortus infection

Abstract

Smooth virulent Brucella abortus strain 2308 (S2308) causes zoonotic brucellosis in cattle and humans. Rough B. abortus strain RB51, derived from S2308, is a live attenuated cattle vaccine strain licensed in the USA and many other countries. Our previous report indicated that RB51, but not S2308, induces a caspase-2-dependent apoptotic and necrotic macrophage cell death. Dendritic cells (DCs) are professional antigen presenting cells critical for bridging innate and adaptive immune responses. In contrast to Brucella-infected macrophages, here we report that S2308 induced higher levels of apoptotic and necrotic cell death in wild type bone marrow-derived DCs (WT BMDCs) than RB51. The RB51 and S2308-induced BMDC cell death was regulated by caspase-2, indicated by the minimal cell death in RB51 and S2308-infected BMDCs isolated from caspase-2 knockout mice (Casp2KO BMDCs). More S2308 bacteria were taken up by Casp2KO BMDCs than wild type BMDCs. Higher levels of S2308 and RB51 cells were found in infected Casp2KO BMDCs compared to infected WT BMDCs at different time points. RB51-infected wild type BMDCs were mature and activated as shown by significantly up-regulated expression of CD40, CD80, CD86, MHC-I, and MHC-II. RB51 induced the production of cytokines TNF-α, IL-6, IFN-γ and IL12/IL23p40 in infected BMDCs. RB51-infected WT BMDCs also stimulated the proliferation of CD4(+) and CD8(+) T cells compared to uninfected WT BMDCs. However, the maturation, activation, and cytokine secretion are significantly impaired in Casp2KO BMDCs infected with RB51 or Salmonella (control). S2308-infected WT and Casp2KO BMDCs were not activated and could not induce cytokine production. These results demonstrated that virulent smooth strain S2308 induced more apoptotic and necrotic dendritic cell death than live attenuated rough vaccine strain RB51; however, RB51, but not its parent strain S2308, induced caspase-2-mediated DC maturation, cytokine production, antigen presentation, and T cell priming.

Figure 1. Cell death of BMDCs infected with B. abortus strain RB51 and its parent wild type strain S2308.

(A) PI and Annexin V staining of Brucella-infected BMDCs at 24 h post infection. (B) The LDH release from RB51- or S2308-infected BMDCs derived from WT and Casp2KO mice. The MOI used was 5. The data represent the means ± standard deviations (SDs) from three independent experiments. The asterisk sign (**) represent the significant differences P-value <0.01, respectively, of a LDH release level from infected Casp2KO BMDCs compared to that from WT BMDCs.

Figure 2. Caspase-2 enzyme activity in infected WT BMDCs induced by RB51 and S2308.

The results were quantified by densitometry and normalized to the β-actin content. The data represent the means ± standard deviations from three independent experiments. The sign (*) represents a statistically significant difference (P-value <0.05) of caspase-2 enzyme activity in Brucella-infected BMDCs compared to uninfected control. A statistically significant difference of caspase-2 enzyme activities in RB51 or S2308-infected WT BMDCs was labeled with (**) (P-value <0.05).

Figure 3. Growth kinetics of RB51 and S2308 inside infected WT and Casp2KO

BMDCs. WT and Casp2KO BMDCs were infected with RB51 or S2308 at a MOI of 5∶1. The numbers of live intracellular bacteria at different time points post infection were evaluated by CFU examination. Data are means ± standard deviations from three independent experiments. The sign (*) represents a statistically significant difference (P-value <0.05) of S2308 survival inside infected WT and Casp2KO BMDCs. A statistically significant difference of RB51 survival inside infected WT and Casp2KO BMDCs was labeled with (**) (P-value <0.05).

Figure 4. BMDC maturation in response to infection with rough Brucella strain RB51 or its parent smooth strain S2308.

WT or Casp2KO BMDCs were infected with S2308, RB51, or S. typhimurium strain SL1344 (MOI: 5). At 24 h post infection, the expression of three costimulatory cell surface markers CD40 (A), CD80 (B), and CD86 (C) was measured by flow cytometry. For each surface molecule studied, a cytometry analysis histogram from one representative experiment is presented on the left, and the complication histogram on the right includes summarized results (means ± standard deviations of the means) from four independent experiments. The asterisk sign (*) denotes statistically significant difference (P-value <0.05) between an infection group and the medium control group. The sign (**) represents statistically significant difference (P-value <0.05) between WT and Casp2KO BMDCs infected with the same bacteria.

Figure 5. Surface expression of MHC class I and II molecules of BMDCs in response to infection with RB51 or S2308.

WT or Casp2KO BMDCs were infected with Brucella strains S2308 or RB51, or S. typhimurium strain SL1344 (MOI: 5). At 24 h post infection, the surface expression of two MHC class molecules H-2Kb (A) and I-Ab (B) was measured by flow cytometry. For each surface molecule studied, a cytometry analysis histogram from one representative experiment is presented on the left, and the complication histogram on the right includes results (means ± standard deviations of the means) from four independent experiments. Data present means ± SDs of three independent experiments. The asterisk sign (*) denotes statistically significant difference (P-value <0.05) between an infection group and the medium control group. The sign (**) represents statistically significant difference (P-value <0.05) between WT and Casp2KO BMDCs infected with the same bacteria.

Figure 6. The effects of caspase-2 on production of cytokines in Brucella-infected BMDCs.

To assess the effect of caspase-2 on DC function, the protein levels of TNF-α (A), IL-6 (B), IL12/IL23p40 (C), and IFN-γ (D) from the culture supernatants of S2308- or RB51-infected WT or Casp2KO BMDCs were collected and analyzed using ELISA. Supernatants were obtained at 24 h post infection. The results represent cytokine secretion (means ± standard deviations of the means of three experiments). The asterisk sign (*) denotes statistically significant difference (P-value <0.05) between an infection group and the medium control group. The sign (**) represents statistically significant difference (P-value <0.05) between WT and Casp2KO BMDCs infected with the same bacteria.

Figure 7. Effect of caspase-2 on priming of allogeneic T-cells through Brucella-infected BMDCs.

At 24 h post infection, WT and Casp2KO BMDCs that were infected or not infected (NI) with RB51 were co-cultured with CFSE-labeled naïve CD4+ T cells (A) or CD8+ T cells (B) from BALB/c mice for 5 days at different BMDC/T-cell ratios. The proliferation of naïve T-lymphocytes was analyzed by flow cytometry. The data are means ± standard deviations of the means of five independent experiments. Statistical differences for comparisons with uninfected DCs are indicated by the sign (**) (P-value <0.01).