Tissue factor expressed by circulating cancer cell-derived microparticles drastically increases the incidence of deep vein thrombosis in mice

Abstract

Background: The risk of thrombotic complications such as deep vein thrombosis (DVT) during tumor development is well known. Tumors release into the circulation procoagulant microparticles (MPs) that can participate in thrombus formation following vessel injury. The importance of this MP tissue factor (TF) in the initiation of cancer-associated DVT remains uncertain.

Objective: To investigate how pancreatic cancer MPs promote DVT in vivo.

Methods: We combined a DVT mouse model in which thrombosis is induced by flow restriction in the inferior vena cava with one of subcutaneous pancreatic cancer in C57BL/6J mice. We infused high-TF and low-TF tumor MPs to determine the importance of TF in experimental cancer-associated DVT.

Results: Both tumor-bearing mice and mice infused with tumor MPs subjected to 3 h of partial flow restriction developed an occlusive thrombus; fewer than one-third of the control mice did. We observed that MPs adhered to neutrophil extracellular traps (NETs), which are functionally important players during DVT, whereas neither P-selectin nor glycoprotein Ib were required for MP recruitment in DVT. The thrombotic phenotype induced by MP infusion was suppressed by hirudin, suggesting the importance of thrombin generation. TF carried by tumor MPs was essential to promote DVT, as mice infused with low-TF tumor MPs had less thrombosis than mice infused with high-TF tumor MPs.

Conclusions: TF expressed on tumor MPs contributes to the increased incidence of cancer-associated venous thrombosis in mice in vivo. These MPs may adhere to NETs formed at the site of thrombosis.

Keywords: cancer; cell-derived microparticles; neutrophils; tissue factor; venous thrombosis.

Fig. 1

(A–C) Three-hour IVC stenosis in control WT mice (WT control, n=7), tumor-bearing mice (Cancer, n=8) and mice infused with cancer cell-derived MPs (MPs, n=10). (A) Percentage of mice with a thrombus. (B and C) Values for weight and length of each thrombus. Horizontal bars in dot plots represent median. (D) Representative images of fluorescence depicting accumulation of calcein-AM-labeled Panc02-derived MPs within the thrombosis area (n=3 mice). The pictures show the IVC part immediately below stenosis.

Fig. 2

(A–C) Three-hour IVC stenosis in control P-selectin deficient mice (PselKO + PBS, n=6) and P-selectin deficient mice infused with cancer cell-derived MPs (PselKO + MPs, n=7). (A) Percentage of mice with a thrombus. (B and C) Values for weight and length of each thrombus. Horizontal bars in dot plots represent median. (D–F) One-hour IVC stenosis in GPG290 pretreated mice (GPG290, n=9) and in GPG290-pretreated mice infused with cancer cell-derived MPs (GPG290 + MPs, n=7). (D) Percentage of mice with a thrombus. (E and F) Values for weight and length of each thrombus. (G and H) Representative images depicting Panc02-derived MPs interacting with NETs but not with unstimulated neutrophils. MPs were labeled with DiD (red) and NET DNA was stained with Hoechst 33342 (blue) after incubation of MPs with washed mouse neutrophils pre-incubated in absence (no stimulation) or in presence of 50 μM PAF to induce NET formation. (G) Experiment performed under static conditions. The photo on the right is at a higher magnification as indicated by bar. (H) Experiment performed in a flow chamber, at a shear rate of 100s⁻¹. The flow direction was from left to right. The photo on the right shows the absence of MP binding after treatment with DNase-1.

Fig. 3

(A–C) One-hour IVC stenosis in control WT mice (WT control, n=9), mice infused intravenously with cancer cell-derived MPs (MPs, n=13) and in mice infused with cancer cell-derived MPs and pretreated with hirudin (MPs + Hir, hirudin 8U/g, n=6). (A) Percentage of mice with a thrombus. (B and C) Values for weight and length of each thrombus. Horizontal bars in dot plots represent median. (D–F) One-hour IVC stenosis in WT mice infused with unmodified Panc02-derived MPs (Native MPs, n=19) and in mice infused with LowTF MPs (LowTF MPs, n=19), (D) Percentage of mice with a thrombus. (E and F) Values for weight and length of each thrombus. Horizontal bars in dot plots represent median.