Selective suicide of cross-presenting CD8+ dendritic cells by cytochrome c injection shows functional heterogeneity within this subset

Abstract

Cross-presentation as a fundamental pathway of activating CD8(+) T cells has been well established. So far the application of this concept in vivo is limited, and the mechanisms that specialize CD8(+) dendritic cells (DCs) for this task are not fully understood. Here we take advantage of the specific cytosolic export feature of cross-presenting DCs together with the property of cytosolic cytochrome c (cyt c) in initiating Apaf-1-dependent apoptosis selectively in cross-presenting DCs. A single i.v. injection of cyt c in B6 mice produced a 2- to 3-fold reduction in splenic CD8(+) DCs but not in Apaf-1-deficient mice. Functional studies both in vivo and in vitro showed that cyt c profoundly abrogated OVA-specific CD8(+) T cell proliferation through its apoptosis-inducing effect on cross-presenting DCs. More importantly, in vivo injection of cyt c abolished the induction of cytotoxic T lymphocytes to exogenous antigen and reduced subsequent immunity to tumor challenge. In addition, only a proportion of CD8(+) DCs that express abundant IL-12 and Toll-like receptor 3 were efficient cross-presenters. Our data support the hypothesis that cross-presentation in vivo requires cytosolic diversion of endocytosed proteins, conferring cross-presentation specialization to a proportion of CD8(+) DCs. We propose that DCs incapable of such transfer, even within the CD8(+) DC subset, are unable to cross-present. Our model opens an avenue to specifically target cross-presenting DCs in vivo for manipulating cytotoxic T lymphocyte responses toward infections, tumors, and transplants.

Fig. 1.

Horse cyt c-mediated apoptosis of cross-presenting DCs. (a) Flow cytometry of splenic DCs 24 h after treatment with 5 mg of either horse or yeast cyt c. Numbers indicate the relative percentage of CD11c⁺ cells. Results are typical of >10 experiments with three to five mice per group. (b) Absolute DC counts per spleen after horse or yeast cyt c treatment. Means ± SEM of three separate experiments with four to five mice per group are indicated. (c) Total cell numbers after horse or yeast cyt c treatment. (d) DC subset counts. AutoMACS-purified DCs from B6 or Apaf-1^−/− mice were cultured for 24 h at the indicated doses of cyt c. DCs were gated on CD11c^hiCD8⁺ (Left) or CD11c^hiCD8⁻ (Right) expression. Means of triplicate wells ± SEM are shown. ***, P ≤ 0.001; n.s., not significant. (e) Apoptosis assay. Purified CD11c⁺ cells were cultured in triplicate for 6 h at the indicated doses of cyt c and stained with CD8, annexin V, and propidium iodide for analysis by flow cytometry. Means ± SEM (percentage of annexin V⁺ propidium iodide⁺ cells) are shown. One representative experiment of three is depicted.

Fig. 2.

Cyt c treatment preferentially inhibits cross-presentation in vivo. B6 mice were left untreated or given three daily doses of i.v. cyt c starting the day before injection of OCS (a) or PK15-OVA (b). CFSE-labeled OT-I cells were adoptively transferred into all mice (a and b Left). Some mice were coinjected with equal numbers of CFSE-labeled OT-II (b Right). The FACS profiles of OT-II proliferation were similar for both types of cellular Ag, and for the sake of clarity only the profiles for PK15-OVA are shown. Results are representative of multiple experiments with four to five mice per group.

Fig. 3.

Effects of cyt c are mediated through an Apaf-1-dependent mechanism. CFSE-labeled OT-I cells were adoptively transferred into WT- or Apaf-1^−/−-reconstituted bm1 mice on day 2. Mice were untreated or given three daily doses of cyt c starting the day before receiving OCS. Proliferation was assessed 60 h later with representative flow cytometry plots (a) and quantitation of proliferating cells (b) shown. Bars represent the average number of proliferating cells relative to values obtained in untreated mice ± SEM. Results are representative of more than six experiments with four to five mice per group. n.s., not significant.

Fig. 4.

Cyt c inhibits endogenous anti-OVA CD8⁺ responses. Cyt c-treated or untreated B6 mice were primed in vivo with OCS. (a) Endogenous in vivo CTL responses were determined after 7 d. Means ± SEM are shown. (b) Endogenous CD8⁺ T cell proliferation. Spleens were harvested 7 d after in vivo priming, CFSE-labeled, and cultured in duplicate for 4 d with 100, 10, or 0 μg/ml OVA. Dot plots are gated on live CD8⁺ cells. Numbers indicate the percentage of proliferated CD8⁺ cells. (c) Total numbers of CFSE^dimCD8⁺ cells from cyt c-treated immunized, untreated immunized, or unimmunized mice (mean ± SEM). (d) Endogenous CD4⁺ T cell proliferation. Dot plots are gated on live CD4⁺ cells. Numbers indicate the percentage of proliferated CD4⁺ cells. Data are representative of three independent experiments with three to four mice per group.

Fig. 5.

Cyt c impairs subsequent immunity to B cell lymphoma. B6 mice were untreated or given three daily injections of cyt c starting the day before in vivo priming with OCS. After 7 d recipient mice were inoculated with Eμ-myc-GFP or Eμ-myc-OVA cells. After another 7 d spleens were analyzed for GFP⁺ B220⁺ tumor cells. Total tumor numbers per spleen (a and c) and representative FACS plots (b) are shown (representative of three experiments with three to five mice per group). n.s., not significant. Numbers indicate the percentage of tumor cells of all live cells.

Fig. 6.

Cyt c-resistant DCs are incapable of cross-presentation but retain the capacity for MHC class II presentation. Residual splenic CD11c⁺ DCs from pooled cyt c-treated (after 24 h) or untreated mice were sorted into CD8⁺ or CD8⁻ subsets. Each subset was cultured continuously in duplicate for 60 h with the indicated doses of OVA together with CFSE-labeled OT-I (a) or OT-II (b) cells. Numbers of proliferating cells are shown (mean ± SEM). Data are representative of three individual experiments.

Fig. 7.

Cyt c treatment preferentially reduces IL-12 production and TLR3 expression by CD8⁺ DCs. DC subsets were prepared as for Fig. 6 and cultured in duplicate for 24 h with a mixture of CpG, IFN-γ, and GM-CSF or media alone. (a) IL-12p70 production as measured by ELISA of culture supernatants (mean ± SEM). Data are representative of three individual experiments. **, undetectable. (b) TLR3 expression as measured by quantitative real-time PCR. Values are expressed relative to GAPDH levels.