A balance between TFPI and thrombin-mediated platelet activation is required for murine embryonic development

Abstract

Tissue factor pathway inhibitor (TFPI) is a critical anticoagulant protein present in endothelium and platelets. Mice lacking TFPI (Tfpi(-/-)) die in utero from disseminated intravascular coagulation. They are rescued by concomitant tissue factor (TF) deficiency, demonstrating that TFPI modulates TF function in vivo. Recent studies have found TFPI inhibits prothrombinase activity during the initiation of coagulation and limits platelet accumulation during thrombus formation, implicating TFPI in modulating platelet procoagulant activity. To examine whether altered platelet function would compensate for the lack of TFPI and rescue TFPI-null embryonic lethality, Tfpi(+/-) mice lacking the platelet thrombin receptor, protease activated receptor 4 (PAR4; Par4(-/-)), or its coreceptor, PAR3, were mated. PAR3 deficiency did not rescue Tfpi(-/-) embryos, but >40% of expected Tfpi(-/-):Par4(-/-) offspring survived to adulthood. Adult Tfpi(-/-):Par4(-/-) mice did not exhibit overt thrombosis. However, they had focal sterile inflammation with fibrin(ogen) deposition in the liver and elevated plasma thrombin-antithrombin complexes, indicating activation of coagulation at baseline. Tfpi(-/-):Par4(-/-) mice have platelet and fibrin accumulation similar to Par4(-/-) mice following venous electrolytic injury but were more susceptible than Par4(-/-) mice to TF-induced pulmonary embolism. In addition, ∼30% of the Tfpi(-/-):Par4(-/-) mice were born with short tails. Tfpi(-/-):Par4(-/-) mice are the first adult mice described that lack TFPI with unaltered TF. They demonstrate that TFPI physiologically modulates thrombin-dependent platelet activation in a manner that is required for successful embryonic development and identify a role for TFPI in dampening intravascular procoagulant stimuli that lead to thrombin generation, even in the absence of thrombin-mediated platelet activation.

Figure 1

Survival of Tfpi^+/+:Par4^−/−, Tfpi^+/−:Par4^−/−, and Tfpi^−/−:Par4^−/− pups from birth to wean. Litters were observed over the first 21 days of life. Dead pups were genotyped. Pups surviving through 21 days were weaned and genotyped. Shown is the percentage of survivors for each genotype over time.

Figure 2

Some Tfpi^−/−:Par4^−/− mice have short, kinked, and/or curly tails. (A) Representative photos of short tailed Tfpi^−/−:Par4^−/− mice. (B) The body length to tail length ratio of Tfpi^−/−:Par4^−/− (n = 38) and Tfpi^+/−:Par4^−/− (n = 35) littermate male mice was determined at 8 weeks of age. An increased ratio, because of decreased tail length, was observed in 11 (29%) Tfpi^−/−:Par4^−/− mice. Data points represent ratios of individual mice and black lines the average ratio.

Figure 3

Sterile inflammatory infiltrates are present in the liver of Tfpi^−/−:Par4^−/− mice. Livers were isolated from Tfpi^−/−:Par4^−/− mice (first and second columns) and Par4^−/− mice (third column) and processed for immunohistochemistry. Serial sections were stained for T lymphocytes (first row, CD3-positive cells), neutrophils (second row, Ly6B.2-positive cells), macrophages (third row; F4/80-positive cells), and fibrin(ogen) (fourth row). Two types of lesions were observed in Tfpi^−/−:Par4^−/− mice: smaller lesions containing predominantly CD3⁺ cells and macrophages that were associated with fibrin deposition (first column) and large, necrotic, fibrin-rich lesions containing neutrophils that were surrounded by CD3⁺ cells (second column). Staining of tissue from Par4^−/− mice (third column) served as a control. In each image, lesions are indicated with arrowheads, and the central vein is indicated with the letter V. Slides were examined at room temperature using Olympus 10X UPlanFI and ×20 Plan objectives lenses mounted to a Olympus BX50 microscope. Coverslips were mounted using xylene-compatible mounting medium (Dako). Images were captured with a Nikon DS-Fi1 digital microscope camera and Nikon NIS Elements software (version 2.1).

Figure 4

Tfpi^−/−:Par4^−/− mice have enhanced thrombin generation in vivo and in vitro. (A) TAT complex in plasma is elevated in Tfpi^−/−:Par4^−/− mice compared with Par4^−/− mice (13.4 ± 18.1 ng/mL, n = 20 vs 6.7 ± 13.0 ng/mL, n = 12; mean ± standard deviation; P = .014). (B) Real-time thrombin generation in PPP from wild-type, Par4^−/−, and Tfpi^−/−:Par4^−/− mice was assessed by calibrated automated thrombography. Lines represent the mean thrombin generated in pooled samples (n = 2 for each genotype) from 3 separate experiments using a total of 6 mice for each genotype.

Figure 5

Lack of TFPI does not affect the dynamics of clot formation in Par4^−/− mice. Thrombus formation in the femoral vein was induced by electrolytic injury. For each genotype, 2 thrombi were formed in each of 3 mice, 1 in each femoral vein, totaling 6 thrombi analyzed. Lines represent the mean and error bars the standard deviation of the relative intensities for accumulation of rhodamine 6G-labeled platelets and AlexaFluor-647–labeled anti-fibrin. Platelet accumulation is significantly decreased in both Par4^−/− (P < .01) and Tfpi^−/−:Par4^−/− (P < .001) mice compared with wild-type mice. It is not significantly different between Par4^−/− and Tfpi^−/−:Par4^−/− mice (P > .05). Fibrin formation is not different among the genotypes (P > .05 for all comparisons).

Figure 6

Lack of TFPI increases the susceptibility to thrombosis of Par4^−/− mice in a TF-induced pulmonary embolism model. The time to respiratory arrest after injection of TF was measured. Shown is the percentage of survivors over time. The fraction of surviving Par4^−/− mice is greater than surviving Tfpi^−/−:Par4^−/− mice (P = .0363), and the fraction of both Par4^−/− and Tfpi^−/−:Par4^−/− surviving mice is greater than wild-type mice (P < .0001 and P = .0319, respectively).