Cellular localization and trafficking of tissue factor

Abstract

Tissue factor (TF) is the cellular receptor for clotting factor VIIa (FVIIa). The formation of TF-FVIIa complexes on cell surfaces triggers the activation of coagulation cascade and cell signaling. In the present study, we characterized the subcellular distribution of TF and its transport in fibroblasts by dual immunofluorescence confocal microscopy and biochemical methods. Our data show that a majority of TF resides in various intracellular compartments, predominantly in the Golgi. Tissue factor at the cell surface is localized in cholesterol-rich lipid rafts and extensively colocalized with caveolin-1. FVIIa binding to TF induces the internalization of TF. Of interest, we found that TF-FVIIa complex formation at the cell surface leads to TF mobilization from the Golgi with a resultant increase in TF expression at the cell surface. This process is dependent on FVIIa protease activity. Overall, the present data suggest a novel mechanism for TF expression at the cell surface by FVIIa. This mechanism could play an important role in hemostasis in response to vascular injury by increasing TF activity where and when it is needed.

Figure 1.

Cellular distribution of tissue factor in fibroblasts. (A) Immunostaining of tissue factor. Nonpermeabilized and Triton X-100–permeabilized WI-38 lung fibroblasts were immunostained with polyclonal rabbit anti–human TF IgG (10 μg/mL), followed by Oregon Green–conjugated anti–rabbit IgG. Fluorescence was viewed with a Perkin Elmer UltraVIEW laser scanning confocal microscope. (B) Quantification of the plasma membrane and intracellular tissue factor. TF levels at the plasma membrane and in the intracellular pool were determined by measuring pixel density of the fluorescence of immunostained cells using 3-D reconstructed images of z slices. The difference in the pixel density between the nonpermeabilized and permeabilized cells was taken as an estimate of the TF in the intracellular compartment (n = 31 cells from 4 different experiments, mean ± SEM). (C) Functional activities of the cell surface and the intracellular tissue factor. Tissue factor activity at the cell surface was blocked by incubating the intact cells with anti-TF IgG (10 μg/mL) for 1 hour at 4°C. Unbound antibodies were removed and the cells were washed 3 times before they were permeabilized by Triton X-100 (0.01% for 10 minutes) or lysed by freeze-thawing. TF activity was measured by adding FVIIa (10 nM) and factor X (175 nM), and measuring the generation of FXa in a chromogenic assay. To demonstrate that there were no free/excess anti-TF antibodies in the experimental system, a small amount of purified relipidated TF (to permeabilized cells) or fibroblast cell lysate (to lysed cells) was added to cells whose surface TF activity was blocked by anti-TF antibodies or to a buffer (n = 3; mean ± SEM).

Figure 2.

Intracellular distribution of tissue factor in fibroblasts. Fibroblasts were fixed, permeabilized, and stained with anti–human TF and an organelle-specific antibody. Left column represents TF staining; middle column, organelle-specific staining; and the right column, the overlay image (colocalization) of TF and organelle-specific marker. Organelle-specific antibodies/stain used as follows: SYTOX Green for nucleus, anti-EEA1 and anti-rab5 for early endosomes, anti-LAMP1 for lysosomes, anticalreticulin for endoplasmic reticulum, and anti–golgin-97 for the Golgi. Note: In contrast to EEA1, whose localization is limited to early endosomes, rab5 is also known to present in the cytoplasmic side of the plasma membrane and clathrin-coated vesicles.

Figure 3.

Factor VIIa–induced internalization and recycling of tissue factor. (A) Fibroblasts were surface-labeled with sulfo NHS-SS-biotin at 4°C and then incubated at 37°C for varying times with a control buffer or the buffer containing FVIIa or FFR-FVIIa (10 nM). The cells were then treated with the reducing agent, lysed, and immunoprecipitated with anti-TF beads. The immunoprecipitated samples were analyzed for the biotin to identify the internalized TF. TF biotin signal detected in cells that were treated with the reducing agent immediately following the biotinylation was taken as 100%. To show the extent of cell-surface TF biotinylation, immunoprecipitates of cells that were not treated with the reducing agent were analyzed for the biotin label. *Values significantly differ from the TF internalized in the absence of FVIIa (P < .02). (B) To measure the recycling of the internalized cell-surface TF, biotin-labeled cells were first incubated with FVIIa at 37°C for 60 minutes to allow TF internalization. Cells were washed with PBS, and the internalized receptors were chased by reincubation at 37°C for various time periods in duplicate samples. After the incubation, only 1 of 2 samples was reduced again. The cells were lysed and subjected to TF immunoprecipitation, followed by immunoblotting. Differences between the chased cells that were treated or not treated with the reducing agent represent the amount of TF recycled back to the cell surface. NS denotes no statistically significant difference. The data shown in the figure represent mean ± SEM (n = 3 to 6 for A, n = 3 for B).

Figure 4.

Factor VIIa mobilizes tissue factor from the Golgi. (A) Fibroblasts were exposed to FVIIa (10 nM) for different time intervals (0 to 120 minutes). Cells were fixed, permeabilized, and immunostained with rabbit polyclonal anti–human TF and monoclonal anti–human golgin-97 antibodies, followed by Rhodamine Red–labeled anti–rabbit IgG and Oregon Green–labeled anti–mouse IgG as secondary reporter antibodies. (B) Factor VIIa's protease activity and the binding to cell-surface tissue factor are essential for the trafficking of tissue factor from the Golgi. WI-38 cells were exposed to FVIIa (10 nM) or FFR-FVIIa (10 nM) alone for 2 hours at 37°C or first incubated with 20-fold molar excess of FFR-FVIIa or anti-TF IgG (10 μg/mL) for 30 minutes before FVIIa (10 nM) was added to the cells. The cells were stained with anti-TF and anti–golgin-97 antibodies and analyzed as described in panel A. Left panel images represent TF staining, middle panel images represent the Golgi staining, and right panel images represent the overlay of TF and the Golgi staining (colocalization). Insets in panel A show a magnified view of TF localization in the Golgi.

Figure 5.

Increase in tissue factor concentration in endosomes in response to factor VIIa. Fibroblasts were exposed to control vehicle or FVIIa (10 nM) for 2 hours at 37°C. The cells were fixed, permeabilized, and immunostained with anti–human TF and antibodies against either EEA1 (A) or rab5 (B). Left panel represents TF staining, middle panel represents the staining for early endosomes (EEA1 or rab5), and the right panel represents the overlay of TF and EEA1 (or rab5) staining.

Figure 6.

Factor VIIa binding to fibroblasts increases tissue factor expression at the cell surface. Fibroblasts were incubated with a control buffer, FVIIa (10 nM), or FFR-FVIIa (10 nM) for 2 hours at 37°C. Thereafter, the cells were washed with 5 mM EDTA to remove the bound FVIIa/FFR-FVIIa, and cell-surface TF levels were determined by incubating the cells for 2 hours at 4°C with ¹²⁵I-TF mAB (TF9H10) or ¹²⁵I-FVIIa (10 nM) (± polyclonal anti–human TF) and then determining TF mAB binding (B) or TF-specific FVIIa binding (C) to the cells. To measure TF functional activity, fresh FVIIa (10 nM) and factor X (175 nM) were added to the cells and the rate of factor X activation was measured (A). Cell-surface TF antigen and activity were significantly higher statistically (P < .003 to .008) in FVIIa-treated cells compared with control vehicle–treated cells or cells treated with FFR-FVIIa. No significant differences were found between control vehicle– and FFR-FVIIa–treated cells (n = 3 to 6, mean ± SEM).