Tumor-Derived Exosomes Induce the Formation of Neutrophil Extracellular Traps: Implications For The Establishment of Cancer-Associated Thrombosis

Abstract

Cancer patients are at an increased risk of developing thromboembolic complications. Several mechanisms have been proposed to explain cancer-associated thrombosis including the release of tumor-derived extracellular vesicles and the activation of host vascular cells. It was proposed that neutrophil extracellular traps (NETs) contribute to the prothrombotic phenotype in cancer. In this study, we evaluated the possible cooperation between tumor-derived exosomes and NETs in cancer-associated thrombosis. Female BALB/c mice were orthotopically injected with 4T1 breast cancer cells. The tumor-bearing animals exhibited increased levels of plasma DNA and myeloperoxidase in addition to significantly increased numbers of circulating neutrophils. Mice were subjected to either Rose Bengal/laser-induced venous thrombosis or ferric chloride-induced arterial thrombosis models. The tumor-bearing mice exhibited accelerated thrombus formation in both models compared to tumor-free animals. Treatment with recombinant human DNase 1 reversed the prothrombotic phenotype of tumor-bearing mice in both models. Remarkably, 4T1-derived exosomes induced NET formation in neutrophils from mice treated with granulocyte colony-stimulating factor (G-CSF). In addition, tumor-derived exosomes interacted with NETs under static conditions. Accordingly, the intravenous administration of 4T1-derived exosomes into G-CSF-treated mice significantly accelerated venous thrombosis in vivo. Taken together, our observations suggest that tumor-derived exosomes and neutrophils may act cooperatively in the establishment of cancer-associated thrombosis.

Figure 1

Tumor-bearing mice display systemic signs of neutrophil activation. 4T1 tumor cells (5 × 10⁴) were orthotopically injected into the mammary fat pads of female BALB/c mice. Blood was collected for hemogram analysis 21 days after tumor cell inoculation as described in the Methods section. The following parameters were analyzed: (A) hematocrit; (B) platelet count; (C) monocyte count; (D) lymphocyte count and (E) neutrophil count in control (open bars) and 4T1 tumor-bearing mice (black bars). The plasma levels of (F) DNA and (G) myeloperoxidase (MPO) were evaluated in control (open bars) and 4T1-bearing mice (black bars) as described in the Methods section. *P < 0.05; **P < 0.01; unpaired, two-tailed Student’s t-test. N = 5 mice per group.

Figure 2

4T1 tumor-bearing mice exhibit a NET-dependent prothrombotic state. (A) Venous thrombosis was evaluated in the Rose Bengal/laser-induced injury in control (●, n = 9) and 4T1 tumor-bearing (■, n = 15) mice as described in the Methods section. Alternatively, control (▼, n = 4) and 4T1-bearing (▲, n = 8) mice were treated with DNase 1 (10 μg, i.v.) 15 min before vessel injury. The data represent the occlusion time after photochemical injury in the jugular vein. (B) Arterial thrombosis was evaluated in the ferric chloride model in (●, n = 9) control and (■, n = 9) 4T1 tumor-bearing mice as described in the Materials and Methods section. Alternatively, control (▼, n = 6) and 4T1-bearing (▲, n = 9) mice were treated with DNase 1 (10 μg, i.v.) 15 min before vessel injury. The data represent the occlusion time after ferric chloride injury in the carotid artery. Each data point represents one individual mouse. **P < 0.01; ***P < 0.001; analysis of variance (ANOVA) with Tukey’s posttest. (C) Fluorescence microscopy analysis of longitudinal cryosections of arterial thrombi from control (left) and 4T1 tumor-bearing mice (right). (D) Higher magnification of the thrombus from a 4T1 tumor-bearing mouse. The sections were stained with anti-Ly6G (green) and Hoechst (blue). Extracellular DNA fibers are indicated by arrows. Bars = 50 μm.

Figure 3

Exosome levels are increased in tumor-bearing mice. (A) Quantification of exosomes isolated from the plasma of control (●, n = 5) and 4T1 tumor-bearing (■, n = 10) mice. (B) Representative histogram indicating the size distribution of exosomes isolated from control (left panel) and 4T1 bearing-mice (right panel). Exosomes were isolated from the plasma as described in the Methods section. (C) CD63 was quantified in the plasma of control (open bar) and 4T1-bearing mice (black bar) as described in the Methods section. *P < 0.05; **P < 0.01; unpaired, two-tailed Student’s t-test.

Figure 4

Tumor-derived exosomes induce NET formation. (A) Histogram indicating the size distribution of exosomes isolated from the culture supernatants of 4T1 cells. Exosomes were isolated and quantified as described in the Methods section. Bone marrow cells isolated from G-CSF-treated mice were incubated for 3 h at 37 °C in the absence (B) or presence (C) of 0.1 µg of 4T1-derived exosomes (Exo). Immunofluorescence staining for DNA (blue), Ly6G (green) and citrullinated histone (red) was performed as described in the Methods section. Scale bars = 20 μm. (D) The quantification of NETs was performed as described in the Methods section. **P < 0.01; unpaired, two-tailed Student’s t-test. Experiments were run in triplicate (n = 3).

Figure 5

NETs interact with tumor-derived procoagulant exosomes. (A) Flow-cytometric analysis of TF expression in 4T1 cells. Black region represents labeling with a rabbit polyclonal anti-murine TF antibody and a phycoerythrin-conjugated secondary antibody. Gray regions represent cells labeled with IgG isotype control and the same phycoerythrin-conjugated secondary antibody. (B) Procoagulant activity of 4T1-derived exosomes. Exosomes were isolated and quantified from culture supernatants and further assayed for procoagulant activity, as described in the Methods section. Control bar represents the coagulation time of murine plasma alone. The asterisks indicate P < 0.001 relative to control plasma (Student’s t-test). Experiments were performed in triplicate. (C) Representative image showing 4T1-derived exosomes interacting with NETs. Exosomes were labeled with DilC18 (red), and NET DNA was stained with Hoechst 33342 (blue). Cells were isolated and stimulated with PMA for 3 hs to induce NET formation, before incubation with 4T1 exosomes. Scale bar = 20 μm.

Figure 6

G-CSF and tumor-derived exosomes accelerate venous thrombosis in tumor-free mice. (A) Neutrophil counts in the peripheral blood of control (n = 5) and G-CSF-treated (n = 5) mice. (B) The venous thrombosis model was applied to control mice (●, n = 12), mice treated with G-CSF (■, n = 6), control mice infused with 100 µg of 4T1-derived exosomes (▲, n = 3), or mice treated with G-CSF infused with 100 µg of 4T1-derived exosomes (▼, n = 5). The data represent the mean occlusion time after photochemical-induced vascular injury, with each data point representing one individual mouse. **P < 0.01; ***P < 0.001; analysis of variance (ANOVA) with Tukey’s posttest. (C) Fluorescence microscopy analysis of cryosections of venous thrombi from control mice treated with exosomes (upper) and mice treated with G-CSF + exosomes (bottom). (D) Higher magnification of the venous thrombi from control mice treated with exosomes (upper) and mice treated with G-CSF + exosomes (bottom). The sections were stained with Ly6G (green) and Hoechst (blue). Extracellular DNA fibers are indicated by arrows. Bars = 50 μm.

Figure 7

Schematic representation of the cooperation between tumor-derived EVs and neutrophils in tumor progression. The development of a thrombotic state in cancer patients is dependent on tumor-induced systemic alterations, including neutrophilia and increased levels of EVs (microvesicles and/or exosomes). Tumor-derived G-CSF acts in an endocrine fashion, stimulating the bone marrow to produce and export neutrophils to the bloodstream. The tumor also releases a large number of EVs into the circulation. The interaction between tumor-derived exosomes and G-CSF-primed neutrophils favors the release of NETs. Increased NET formation may favor tumor progression in various ways: 1) through the establishment of cancer-associated thrombosis; 2) by facilitating tumor cell arrest and encouraging metastasis; or 3) by promoting vascular damage and organ dysfunction.