Tumor-derived exosomes confer antigen-specific immunosuppression in a murine delayed-type hypersensitivity model

Abstract

Exosomes are endosome-derived small membrane vesicles that are secreted by most cell types including tumor cells. Tumor-derived exosomes usually contain tumor antigens and have been used as a source of tumor antigens to stimulate anti-tumor immune responses. However, many reports also suggest that tumor-derived exosomes can facilitate tumor immune evasion through different mechanisms, most of which are antigen-independent. In the present study we used a mouse model of delayed-type hypersensitivity (DTH) and demonstrated that local administration of tumor-derived exosomes carrying the model antigen chicken ovalbumin (OVA) resulted in the suppression of DTH response in an antigen-specific manner. Analysis of exosome trafficking demonstrated that following local injection, tumor-derived exosomes were internalized by CD11c+ cells and transported to the draining LN. Exosome-mediated DTH suppression is associated with increased mRNA levels of TGF-β1 and IL-4 in the draining LN. The tumor-derived exosomes examined were also found to inhibit DC maturation. Taken together, our results suggest a role for tumor-derived exosomes in inducing tumor antigen-specific immunosuppression, possibly by modulating the function of APCs.

Figure 1. Characterization of tumor exosomes.

(A) EM micrographs of exosomes isolated from EL4, EG7, B16 and MO5 cell culture supernatants. (B) Western blot analysis of exosomes and cell lysates. 10 µg of proteins were loaded per lane. (C) IP detection of OVA protein (40∼45 kD) in both cell lysates and exosomes. (D) FACS analysis of MHC class I, MHC class II and CD81 expression on cells and exosomes.

Figure 2. Suppression of OVA-specific DTH response by local administration of EG7 exosomes.

(A) Mice pre-sensitized with OVA were injected with 10 µg of EL4 exosomes plus 30 µg of OVA, 10 µg of EG7 exosomes plus 30 µg of OVA, 30 µg of OVA alone, 10 ug of EL4 exosomes alone or 10 ug of EG7 exosomes alone in 50 µl of PBS in their right hind paws. The left hind paws were all challenged with 30 µg of OVA in 50 µl of PBS. Paw swellings of both treated (right) and contralateral (left) paws were measured 24 h and 48 h post-challenge as the increase in footpad thickness (×0.01 mm). The results shown are from one representative experiment and are the means ± SD with n = 5. (B) The mean increase of footpad thickness of the treated paws in PBS group (OVA alone) at each time point was set to 1, and the increases of footpad thickness in EL4 exosomes plus OVA group and EG7 exosomes plus OVA group were normalized as fold increase. Figures show the pooled results of three independent experiments and are the means ± SD with n = 15. Significance at **: P<0.01; *: P<0.05; NS: not significant.

Figure 3. Suppression of OVA-specific DTH response by local administration of MO5 exosomes.

Mice pre-sensitized with OVA were injected with 10 µg of B16 exosomes, 10 µg of MO5 exosomes or PBS alone in their right hind paws and were challenged with OVA at both hind paws. Paw swellings were measured 24 h and 48 h post-challenge. (A) Representative results showing the increase in footpad thickness (×0.01 mm) of treated and contralateral paws. n = 5. (B) Pooled results of two independent experiments showing the fold increase in footpad thickness as compared to the treated paws in PBS group. n = 10. **: P<0.01; *: P<0.05; NS: not significant.

Figure 4. OVA-containing tumor exosomes were not effective in suppressing KLH-specific DTH response.

Mice pre-sensitized with KLH were treated with 10 µg of exosomes or PBS in their right hind paws and were challenged with KLH antigen at both hind paws. Paw swellings were measured 24 h and 48 h later. The increases in footpad thickness (×0.01 mm) of one representative experiment (n = 5) (A) and the normalized fold increases in footpad thickness of two independent experiments (n = 10) (B) are shown. NS: not significant.

Figure 5. Exosome in vivo trafficking in DTH model.

PKH67-labeled exosomes were injected in the right footpad of OVA-sensitized mice as in the DTH experiment. Footpads and the popliteal LNs were harvested, cryo-sectioned and examined by immunofluorescence. Similar observations were made with different tumor exosomes and data show the representative figures of MO5 exosomes. (A) 24 h post-injection, exosomes (green) were captured by dermal CD11c+ cells (red) in footpads and transported to the treated-side LN. (B) 48 h post-injection, large numbers of exosome-internalized CD11c+ cells (red, upper left panel) appear in the treated-side LN. Exosomes (or exosome-containing cells) were also physically adjacent to CD3+ T cells (red, lower left panel). Only very few exosomes were observed in the contralateral LN. (C) TUNEL staining for apoptotic cells (red) in both side LNs 48 h post-injection. Magnification: 20×.

Figure 6. qRT-PCR analysis of cytokines and FoxP3 mRNA levels in the draining popliteal LN associated with DTH suppression.

Panels show the relative mRNA levels of TGF-β1 (A), IL-4 (B), IL-10 (C), IFN-γ (D) and FoxP3 (E) normalized to β-actin mRNA level in the treated-side popliteal LNs 48 h after EL4 exosomes, EG7 exosomes or PBS treatment at the time of OVA challenge. n = 5. **: P<0.01; *: P<0.05; NS: not significant.

Figure 7. Tumor exosomes inhibit BMDC maturation and induce TGF-β1 production.

(A) Day 8 BMDCs (purity >90%) were treated with 10 µg/ml of tumor exosomes or cultured untreated for 3 days. The expression of I-A^b and CD86 were analyzed by FACS. LPS treatment (1 µg/ml) for 24 h was used as a DC maturation control. (B) TGF-β1 protein levels (pg/ml) in DC culture supernatants after exosome treatment. Data show the mean values of two independent experiments ± SD. (C) TGF-β1 contents in exosome preparations (pg/10 µg of exosomes). For each exosome sample, the data shown represent the mean value of three preparations ± SD.