• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Thromb Haemost. Author manuscript; available in PMC Oct 26, 2005.
Published in final edited form as:
PMCID: PMC1266294

Factor VIIa binding and internalization in hepatocytes


The liver is believed to be the primary clearance organ for coagulation proteases, including factor VIIa. However, at present, clearance mechanisms for FVIIa in liver are unknown. To obtain information on the FVIIa clearance mechanism, we investigated the binding and internalization of FVIIa in liver cells using a human hepatoma cell line (HEPG2), and primary rat and human hepatocytes as cell models. 125I-FVIIa bound to HEPG2 cells in a time-and dose- dependent manner. Anti-tissue factor antibodies reduced the binding by about 25%, whereas 50-fold molar excess of unlabeled FVIIa had no effect. HEPG2 cells internalized FVIIa with a rate of 10 fmoles/105 cells/h. In contrast to HEPG2 cells, FVIIa binding to primary rat hepatocytes was completely independent of TF, and excess unlabeled FVIIa partly reduced the binding of 125I-FVIIa to rat hepatocytes. Further, compared to HEPG2 cells, 3 to 4- fold more FVIIa bound to rat primary hepatocytes, and the bound FVIIa was internalized at a faster rate. Similar FVIIa binding and internalization profiles were observed in primary human hepatocytes. Plasma inhibitors had no effect on FVIIa binding and internalization in hepatocytes. In contrast, annexin V, which binds to phosphatidylserine, blocked the binding and internalization. Consistent with this, binding of gla-domain-deleted FVIIa to hepatocytes was markedly diminished. In summary, the data presented herein reveal differences between HEPG2 cells and primary liver cells in FVIIa binding and internalization, and suggest that the rapid turnover of membrane and not a receptor-mediated endocytosis may be responsible for internalization of FVII/FVIIa in primary hepatocytes.

Keywords: factor VIIa, hepatocytes, liver, internalization


The liver has long been recognized as a clearance organ for several coagulation proteins. Compared to other clotting factors, which have long circulating half-lives (20 to 72 h) [13], FVII has a relatively short half-life, approximately 5 h [4,5]. Once activated, the clotting proteases in general are cleared rapidly from the circulation with half-lives of less than 3 min [6,7]. In contrast, FVIIa has a long circulating half-life of 2 to 4 h [8,9]. The marked differences in circulating half-lives between zymogens and enzyme forms of other vitamin K-dependent coagulation proteins vs. FVII and FVIIa suggest that liver cells may have a specific and distinctive mechanism for the clearance of FVII/FVIIa.

Tissue factor (TF) is the cellular receptor for FVII/FVIIa. In vitro studies showed that binding of FVIIa to TF facilitates the FVIIa internalization in fibroblasts and tissue factor pathway inhibitor (TFPI) enhances the TF-mediated internalization of FVIIa [10]. This scenario may be true for many extravascular cells that express TF, including smooth muscle cells, pericytes and epithelial cells, but not for cells that come in contact with blood since they do not express TF under normal conditions [11,12]. Other receptors or mechanisms may play a role for FVII/FVIIa clearance in these cells. Since the liver is believed to be the primary clearance organ for coagulation proteases, Chang and Kisiel used a human hepatoma cell line, HuH7, to investigate the binding and uptake of FVIIa [13]. These studies revealed that 60% of 125I-FVIIa binding to HuH7 is blocked by anti-TF IgG and excess unlabelled FVIIa blocked the remaining binding by 50%. These studies also showed that FVIIa bound to non-TF binding sites was internalized with a rate of 0.4 fmol/min/106 cells; and cross-linking studies suggested that FVIIa bound to a cellular receptor, which is yet to be identified, with an apparent mol.wt. of 130,000 Daltons.

Although the HuH7 cell line was believed to maintain liver-specific functions [14], whether this cell line truly represents normal liver cells is an open question. In the present study, we have reinvestigated the binding and internalization of FVIIa using another human hepatoma cell line, HEPG2, and primary rat hepatocytes as a model system for FVIIa clearance by liver. Limited studies were also performed with primary human hepatocytes to validate the data obtained with rat primary hepatocytes. The data presented herein document clear differences between HEPG2 cells and primary hepatocytes in their ability to support FVIIa binding and internalization, and suggest that the rapid turnover of the membrane and not a receptor-mediated endocytosis may be responsible for the internalization of FVIIa in liver cells.

Materials and Methods


Dulbecco’s modified Eagle medium (DMEM), Medium 199, fetal bovine serum (FBS), trypsin-EDTA (ethylenediamine tetra acetic acid), glutamine, and penicillin-streptomycin were obtained from Gibco-BRL Life Technologies (Grand Island, NY). Humulin® was obtained from Eli Lilly (Indianapolis, IN). Dexamethasone sodium phosphate was from American Regent Laboratories Inc. (Shirley, NY). M6L5 and G6L5 (blockers of the mannose and galactose receptors, respectively) were synthesized by Novo Nordisk (Maaloev, Denmark). These compounds are composed of either six mannose (M6L5) or six galactose (G6L5) residues coupled to a backbone of five lysine residues. M6L5 was shown to exhibit a nanomolar affinity for the mannose receptor and inhibits ligand binding to the receptor [15]. Other chemicals, reagent grade or better were from Sigma Chemical Co (St Louis, MO).

Proteins and antibodies

Recombinant human FVIIa, des-gla-FVIIa, zymogen FVII and FVIIa blocked with active site inhibitor, phenylalanyl-phenylalanyl-arginyl chloromethyl ketone (FFR-VIIa) were obtained from Novo Nordisk A/S (Maaloev, Denmark). Asialo-FVIIa was prepared by mixing FVIIa (5 ml of 1.4 mg/ml) with neuraminidase-agarose beads (2 ml) (Sigma) overnight at room temperature. The removal of sialic acids was confirmed by IEF-gel electrophoresis. Asialo-FVIIa had the same specific peptidolytic/proteiolytic activity as FVIIa. FII and FIX were purified from plasma as described earlier [16]. FX was from Enzyme Research Laboratories (South Bend, IN). Human alpha2-macroglobulin (α2-MG) was from Sigma; and antithrombin (AT) was from Kabi Pharmacia (Franklin, OH). Polyclonal antibodies against human TF were described earlier [17]. Antibodies against mouse TF were obtained by immunizing a rabbit with recombinant mouse TF. IgG from the antiserum was purified by Affi-Gel-Blue (Bio-Rad) chromatography. Anti-mouse TF IgG was found to inhibit rat TF activity effectively as determined by its ability to block FVIIa-induced activation of factor X supported by rat thromboplastin (more than 75% inhibition at 100 μg/ml IgG).


The human hepatoma cell line, HEPG2, was purchased from American Type Culture Collection (Rockville, MD). The cells were cultured at 37°C and 5% CO2 in a humidified incubator in DMEM containing Glutamax and high glucose, and supplemented with 1% glutamine, 1% penicillin/streptomycin and 10% fetal bovine serum. For binding studies, the cells were seeded in 24-well tissue culture plates at a concentration of 1x105 cells/well and cultured for two days.

Rat hepatocytes were harvested by a two-step collagenase perfusion as described earlier [18] with slight modifications. Briefly, rats were anesthetized with ketamine HCl (100 mg/kg) and xylazine HCl (15 mg/kg), I.M. After opening the abdominal cavity, loose ligatures were placed around the vena porta and both the lower and upper vena cava. The canula (with flow on) was inserted into the vena porta and the ligature was tightened. At this point, the lower vena cava was cut to allow flow-through. A small hole was cut in the heart and a tube inserted into the upper vena cava (between the liver and the diaphragm). The tube was secured by tightening the ligature. Then, the distal end of the vena cava was clamped to establish a flow through the liver. The liver was perfused with approximately 40 ml/min of calcium-free buffer (25 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonicacid), 110 mM NaCl, 5.4 mM KCl, 0.34 mM Na2HPO4, 0.345 mM KH2PO4 and 22 mM glucose) for about 5 to 10 minutes, followed by 200 ml of calcium-containing collagenase buffer (the perfusion buffer supplemented with 2 mM CaCl2 and 0.6 mg/ml collagenase) at 20 ml/min to loosen the cells. Thereafter, the liver was perfused with calcium containing wash-buffer (the perfusion buffer supplemented with 5 mM CaCl2) (40 ml/min) to remove the collagenase solution. At the end of the perfusion, the liver cells were loosened and filtered. The cells were washed twice in Medium 199 with 10% fetal bovine serum, 1% glutamine and 1% penicillin/streptomycin. Cell survival, as determined by trypan blue dye exclusion method, was 85 to 98%.

The primary rat hepatocytes were seeded at a concentration of 1.65x105 cells/well in 24-well tissue culture plates coated with rat-tail collagen (10 μg/cm2) in Medium 199 with 4% fetal bovine serum, 100 nM dexamethasone, 1 nM insulin, 1% glutamine and 1% penicillin/streptomycin. After 4 h, the medium was replaced by medium 199 as described above, but without fetal bovine serum. The hepatocytes were maintained with daily change of medium. The cells were used within 2–3 days of the harvest.

Primary human hepatocytes were obtained from In Vitro Technologies (Baltimore, MD) and seeded in 24-well tissue culture plates coated with collagen (10 μg/cm2) and cultured in Medium 199 essentially as described for rat primary hepatocytes.


FVIIa and other clotting factors were labeled with 125I using IODO-GEN (Pierce, Rockford, IL)-coated tubes and Na125I according to the manufacturer’s technical bulletin and as described previously [19]. Here, as in earlier studies [19,20], the radiolabeled FVIIa was intact with no apparent degradation and retained 80% or more of functional activity of unlabeled FVIIa

Binding and internalization of 125I-FVIIa

Near confluent monolayers of HEPG2, primary human or rat hepatocytes were washed twice in buffer A (10 mM HEPES, 150 mM NaCl, 4 mM KCl, and 11 mM glucose, pH 7.45) containing 5 mM CaCl2. The cells were then incubated in buffer B (buffer A with 1mg/ml bovine serum albumin and 5 mM CaCl2) and chilled on ice before 125I-FVIIa and other compounds were added to the cells. Where cells were treated with anti-TF antibodies, the monolayers were incubated with anti-TF antibodies for 30 to 45 min before the radiolabeled ligand was added to the cells. After incubating the cells with 125I-FVIIa for a fixed interval or varying time periods, the unbound material was removed and the monolayers were washed three times in ice-cold buffer B. Cell bound 125I-FVIIa was eluted by treating the cells with 0.1 M glycine (pH 2.4) for 5 min at room temperature. For internalization studies, the cells were incubated at 37°C and the bound material was eluted as described above and then the cells were extracted with lysis buffer (0.1M NaOH, 1% SDS). The radioactivity associated with this fraction was taken as FVIIa that was internalized. To determine the extent of FVIIa degradation, hepatocytes were incubated with 125I-FVIIa (10 nM), and at varying times small aliquots were removed from the dish, and precipitated by adding an equal volume of ice-cold 10% trichloroacetic acid. The samples were subjected to centrifugation (15,000 x g for 10 min) and the supernatants were removed and counted for the radioactivity.


Binding and internalization of FVIIa in HEPG2 cells

To investigate the binding of FVIIa to human hepatoma cells, HEPG2 cells were incubated with 125I-FVIIa (10 nM) for various time periods at 4°C or 37°C. The binding followed a hyperbolic pattern and reached a maximum at 2 h (14 fmoles/105 cells at 4°C). The binding pattern was similar at 4°C and 37°C, but the amount of FVIIa associated with the cells at 37°C was 50% higher than that was observed at 4°C (Fig. 1A). The binding was calcium dependent since in the absence of calcium ions FVIIa binding to cells was reduced by more than 75% (data not shown). To determine the specificity of FVIIa binding to HEPG2 cells, the cells were incubated with 125I-FVIIa for 2 h at 4°C in the presence or absence of unlabeled FVIIa (500 nM) or anti-human TF IgG (100 μg/ml). The unlabeled FVIIa failed to reduce 125I-FVIIa binding to the cells, whereas anti-TF antibodies reduced the 125I-FVIIa binding by 25% (Fig. 1B). Although the reduction of 25% was not statistically significant (p=0.2838, n=6), we consistently observed a minor inhibition of 125I-FVIIa binding in the presence of anti-TF IgG. Similar data were obtained when the experiment was carried out at 37°C (data not shown). Addition of increasing concentrations of 125I-FVIIa (1 to 100 nM) to HEPG2 cells (both in the absence and presence of excess unlabeled FVIIa or anti-TF IgG) showed a near linear increase in 125I-FVIIa binding to HEPG2 cells with no apparent evidence for saturability (data not shown).

Fig. 1
FVIIa binding and internalization in HEPG2 cells. Monolayers of HEPG2 cells were incubated with 125I-FVIIa (10 nM) at 4°C or 37°C for varying time periods and the cell surface bound (panel A) or the internalized (panel C) 125I-FVIIa was ...

When HEPG2 cells were incubated with 125I- FVIIa (10 nM) for varying time periods at 37°C, FVIIa bound to the cells in a similar manner as at 4°C, but also internalized. Approximately, 15 fmoles of FVIIa was internalized over 2 h (Fig. 1C). Neither the serine-protease scavenger α2-macroglobulin (α2-MG) (p=0.9327, n=5) nor antithrombin III (AT) in the presence of heparin (p=0.2746, n=3) had significant effect on the internalization of FVIIa (data not shown). Antibodies against TF had a slight inhibitory effect (15% inhibition), which did not reach statistical significance (p=0.0856, n=11) (Fig. 1D).

It is important to note that the inability of either anti-TF IgG or excess unlabeled FVIIa to block 125I-FVIIa binding to HEPG2 cells represents inherent low affinity of binding of FVIIa to these cells and not an artifact. Parallel experiments conducted with MDA-MB-231 carcinoma cells, which constitutively express TF, showed a pronounced 125I-FVIIa binding to these cells, and anti-TF IgG or excess unlabeled FVIIa blocked the 125I-FVIIa binding by more than 80% (data not shown). Since we neither detected a significant specific binding nor internalization of FVIIa with the human hepatoma cell line, we decided to use primary cultures of hepatocytes to study the role of liver cells in FVIIa metabolism.

Binding and internalization of FVIIa in primary rat and human hepatocytes

Similar to HEPG2 cells, FVIIa bound to freshly isolated rat and human hepatocytes in a time-dependent manner; however, the binding was 2–3 fold higher (30–50 fmoles/105 cells) than that observed with HEPG2 cells (Fig. 2A and 2C). Although the rate of 125I-FVIIa binding to primary human hepatocytes, compared to rat hepatocytes, was lower at 4°C, 125I-FVIIa bound to both rat and human hepatocytes at a similar rate at 37°C. 125I-FVIIa binding to both rat and human hepatocytes could be blocked by 25–30% in the presence of excess unlabeled FVIIa whereas anti-TF IgG had no effect on the binding (Fig. 2B and 2D). FVIIa bound to primary rat and human hepatocytes was internalized at a rate about 10-times faster (~120 to 150 fmoles/105 cells/h) than that observed in HEPG2 cells (compare the data of Fig. 2E and 2F with Fig. 1C). These data indicate that 125I-FVIIa interacts with primary rat hepatocytes and human hepatocytes in a similar fashion, whereas significant differences exist in FVIIa binding and internalization between the liver cell line and primary hepatocytes. Therefore, we used rat hepatocytes for further binding studies.

Fig. 2
FVIIa binding and internalization in primary hepatocytes. Primary rat hepatocytes (panels A, B and E) or human hepatocytes (panels C, D and F) were incubated with 125I-FVIIa (10 nM) at 4°C or 37°C for varying time periods (panels A, C, ...

To determine FVIIa binding parameters, rat hepatocytes were incubated with varying concentrations of 125I-FVIIa (1 to 85 nM) for 2 h at 4°C in the presence or absence of 1 μM unlabeled FVIIa. FVIIa bound to rat hepatocytes in a dose-dependent manner (Fig. 3A) and the presence of unlabeled FVIIa reduced 125I-FVIIa binding to the cells by approximately 20–25%. With the concentrations tested, the binding was not fully saturable, suggesting that FVIIa binds to hepatocytes with a low affinity. Analysis of the binding data with hyperbola curve fitting program (Prism, GraphPad) gave a calculated Kd value of 200 nM and an estimated Bmax of 350 fmoles/105 cells, which gives an estimated number of 2.2 x 106 binding sites/cell.

Fig. 3
FVIIa binding to primary rat hepatocytes. (A) Dose-dependent FVIIa binding to hepatocytes. Varying concentrations of 125I-FVIIa (1 to 85 nM) were added to primary rat hepatocytes in the presence ([filled triangle]) or absence ([filled square]) of 1 μM unlabeled ...

In an attempt to obtain clues on potential receptor(s) that are responsible for FVIIa binding, rat hepatocytes were incubated at 4°C for 2 h with 10 nM of 125I-FVIIa in the presence or absence of various competitors. These experiments revealed that unlabeled FVIIa inhibits the binding by 25 percent (p=0.02, n=11), whereas antibodies against mouse TF had no significant effect on the binding (p=0.77, n=4) (Fig. 3B). M6L5, G6L5, AT/heparin, α2-MG, and RAP had no effect on the binding (n=3–7), whereas FIX reduced the binding by 30 percent (p<0.02, n=3) and FX by 15 percent, the latter effect lacking statistical significance (p=0.06, n=3) (Fig. 3B). Similar data was obtained in a single experiment performed with primary human hepatocytes. To investigate the involvement of negatively charged phospholipids in the membrane on FVIIa binding to hepatocytes, the cells were incubated with 125I-FVIIa in the presence or absence of annexin V, which is known to bind to negatively charged phospholipids and block the binding of clotting factors to phospholipids [21]. Annexin V abrogated the binding of 125I-FVIIa by about 40% (Fig. 3B), indicating the involvement of membrane phospholipids in mediating FVIIa binding to rat liver cells. The combination of annexin V and unlabeled FVIIa reduced 125I-FVIIa binding to hepatocytes more than 60%, which was significantly higher than that was obtained with FVIIa or annexin V, individually.

Next, the extent of FVIIa internalization and degradation in liver cells was determined by incubating 125I-FVIIa (10 nM) with rat hepatocytes at 37°C for increasing time periods. Internalization of FVIIa at 37°C was extensive with 110 fmoles of FVIIa internalized/105 cells within the first 1 h (Fig. 4A). FVIIa degradation products were similarly increased over the time with proportionally more degradation products accumulating at the later time points. Approximately 100 fmoles of FVIIa was degraded and excreted at the end of a 2 h incubation period. The presence of 1 μM of unlabeled FVIIa reduced the internalization of radiolabeled FVIIa by 20% (Fig. 4B) whereas cold competition with other gla-domain containing clotting factors had no significant effect on FVIIa internalization (data not shown). Neither α2-MG nor AT/heparin affected the internalization (Fig. 4B). As expected, TFPI was also without an effect on FVIIa internalization. Preincubation of FVIIa overnight with α2-MG or AT to form FVIIa-inhibitor complexes (Fig. 4C) did not enhance the extent of FVIIa internalization (data not shown). Annexin V, which inhibits FVIIa binding to hepatocytes, inhibited its internalization by 35% (Fig. 4B). The combination of annexin V and unlabeled FVIIa further reduced the internalization of radiolabeled FVIIa.

Fig. 4
Internalization and degradation of FVIIa in primary hepatocytes. (A) 125I-FVIIa (10 nM) was added to primary rat hepatocytes and 125I-FVIIa internalized and degraded at varying time periods at 37°C was determined as described in the methods (mean ...

Binding and internalization of various variants of FVIIa

Since asialo glycoproteins were shown to be internalized rapidly in liver cells via receptor-mediated endocytosis [22], we next investigated the internalization of asialo-FVIIa in primary rat hepatocytes. All experiments were carried out in parallel with FVIIa in order to allow direct comparison. Binding of asialo-FVIIa to primary hepatocytes was very similar to that of FVIIa (50 fmoles asialo-FVIIa bound to 105 cells when 10 nM of 125I-asialo-FVIIa was incubated with hepatocytes for 2 h at 4°C; Fig. 5A). Asialo-FVIIa was internalized at a slightly lower rate as compared to FVIIa (~80 fmoles/105/h) (Fig. 5B). As with FVIIa, neither AT/heparin nor α2-MG affected the internalization of asialo FVIIa (data not shown). Thus, these data revealed no significant differences between asialo-FVIIa and FVIIa in their binding and internalization in rat hepatocytes. In contrast, gla-domain deleted FVIIa (des-gla-FVIIa) bound poorly to hepatocytes (14 fmoles of des-gla-FVIIa vs. 50 fmoles of FVIIa/well) (Fig. 5A). Consistent with the binding, the internalization of des-gla-FVIIa was also decreased as compared to FVIIa (Fig. 5B). These data indicate that FVIIa binds to hepatocytes via a mechanism that involves its gla-domain.

Fig. 5
Binding and internalization of asialo- and des-gla FVIIa in primary hepatocytes. Primary rat hepatocytes were incubated at 37°C with 10 nM of either 125I-FVIIa (closed squares), 125I-asialo-FVIIa (open triangles), or 125I-des-gla-FVIIa (inverted ...

Binding and internalization of other gla-domain clotting factors in rat hepatocytes

Next, we investigated whether other gla-domain containing clotting factors, like FIX and FX as well as their activated forms bind and internalize in a similar fashion to that observed for FVIIa. Primary rat hepatocytes were exposed to 125I-labeled FVII, FIX, FX, or their activated forms, and the binding and internalization of these factors at 37°C was determined. No marked differences were observed in FVII and FVIIa binding and internalization. As shown in Fig. 6A and 6B, FIX and FX bound and internalized to a much lesser extent than that of FVII. Compared to FVII, approximately 5–10 times less FIX and FX were bound and internalized in hepatocytes. Binding of the activated forms of these clotting factors was also found to be very low compared to FVIIa (Fig. 6C). Although FIXa and FXa were internalized at a higher rate, compared to their unactivated forms, still it was much lower compared to FVIIa internalization (Fig. 6D). These data indicate that the FVIIa binding to hepatocytes may represent a specific interaction of FVIIa with the hepatocytes rather than a simple interaction of the gla domain with cell surface phospholipids.

Fig. 6
Binding and internalization of FVIIa and other gla-domain clotting factors in primary hepatocytes. Primary rat hepatocytes were incubated at 37°C with 10 nM of either 125I-FVII, 125I-FIX or 125I-FX (panels A and B), or their activated forms (panels ...


Earlier pharmacokinetic studies in rats showed that FVIIa accumulates in highly perfused organs including the stomach, small intestine, kidney and liver [23,24]. Although the liver is thought to be the main clearance organ for glycoproteins and coagulation proteins [2527], little is known about how various coagulation proteins, particularly FVII/FVIIa, are cleared by the liver. We are aware of only a single study where attempts were made to investigate mechanisms by which FVIIa binds and internalizes in liver cells (HuH7 cell line) [13]. This study revealed that the majority of FVIIa bound to HuH7 cells via TF, but a small portion of FVIIa also binds and internalizes independent of TF via an as yet unidentified receptor. Our present data, obtained with HEPG2 cells, significantly differs from the data obtained with HuH7 cells. In the present study, we observed no statistical significant reduction in 125I-FVIIa binding to HEPG2 cells in the presence of excess unlabeled FVIIa, suggesting that these cells do not possess a high affinity specific cellular receptor for FVIIa. A small amount of FVIIa specifically binds to HEPG2 cells via TF; however, this binding was not statistically significant. Consistent with the binding data, HEPG2 cells have a relatively low capacity to internalize FVIIa (0.17 fmol/min/105 cells). It is likely that trace amounts of TF present in these cells might have facilitated this internalization. It is interesting to note that the rate of FVIIa internalization via a specific receptor, other than TF, in HuH7 cells was also very slow (0.04 fmol/min/105 cells). These data suggest that hepatoma cell lines do not support an active internalization of FVIIa.

Compared to HEPG2 cells, two to three-fold more FVIIa bound to primary rat hepatocytes. A small part of this binding represents FVIIa binding to a specific cellular component (which was not TF) since excess unlabeled FVIIa but not anti-mouse TF IgG inhibited the 125I-FVIIa binding to rat primary hepatocytes. It is relevant to point out here that the anti-mouse TF antibodies used in this study effectively blocks FVIIa binding to rat TF as the antibodies block rat TF-FVIIa-induced activation of FX by more than 75%. FVIIa internalization in rat primary hepatocytes strikingly differs from that observed in human hepatoma cell lines, whereas it closely resembles the internalization found in primary human hepatocytes. In contrast to a slow internalization of FVIIa in HEPG2 cells, FVIIa was internalized rapidly in both primary rat and human hepatocytes. The internalization rate of FVIIa in primary rat hepatocytes was about 15-fold higher than that was observed in HEPG2 cells.

Plasma proteinase inhibitors α2-MG and antithrombin have been shown earlier to bind to FIXa and FXa and play a role in clearance of these proteases [6,7]. In the present study, addition of these inhibitors had no effect on either FVIIa binding or its internalization in HEPG2 cells or primary hepatocytes, suggesting that FVIIa is a poor substrate to these proteinase inhibitors. Earlier studies from our laboratory showed that antithrombin, in the absence of TF, is a poor inhibitor of FVIIa [28]. In the present study, we found that only minimal amounts of FVIIa- α2-MG complexes are formed after 2 h when FVIIa was incubated with 50-fold molar excess of α2-MG . As expected, FXa and α2-MG rapidly formed complexes under similar conditions (data not shown). Although prolonged incubation of FVIIa with α2-MG and antithrombin resulted in significant amounts of FVIIa-inhibitor complexes, it did not increase the extent of FVIIa binding and internalization, indicating that the plasma inhibitors may not influence FVIIa binding and internalization in liver cells. Similar to these inhibitors, TFPI had no effect on FVIIa binding and internalization in primary hepatocytes. This is an expected finding since TFPI, at physiological relevant concentrations, does not bind to FVIIa in the absence of TF [29].

Although our present data show that FVIIa was rapidly internalized by primary hepatocytes, it is unclear whether it is due to a receptor-mediated endocytosis or simply the result of a high membrane turnover in hepatocytes. Rat hepatocytes have been reported to have a very high rate of membrane turnover [30]. Our inability to block the majority of 125I-FVIIa binding in hepatocytes with the excess of unlabelled FVIIa or a specific competing substrate suggests that FVIIa associates with hepatocytes with a low affinity binding, probably with phospholipids in the cell membrane. The ability of annexin V to partly block the FVIIa binding and internalization supports this hypothesis. The observation that the combination of annexin V and unlabeled FVIIa block the bulk of radiolabeled FVIIa binding to hepatocytes and a relatively poor binding of gla-domain less FVIIa to hepatocytes suggest that phospholipids and a potential epitope within the gla domain accounts for most of the FVIIa binding affinity to hepatocytes. If this was true, then the affinity of a coagulation protein for the hepatocyte membrane dictates its clearance rate when it is in its zymogen form. Once activated, the protease interactions with plasma inhibitors regulate their clearance. Since FVIIa reacts poorly with plasma inhibitors, it may explain why the clearance rate for FVIIa is similar to that of its zymogen form. However, we also found no evidence that FVIIa interaction with plasma inhibitors actually enhanced its internalization. Although these data are somewhat surprising to us, we were unable to find any evidence in the literature that binding of FVIIa, or other clotting proteases (such as FIXa and FXa), to α2-MG or antithrombin enhanced the internalization of these clotting proteases in cultured hepatocytes.

Our data on FIX and FX binding and internalization by hepatocytes show that these zymogens have a poor affinity for the hepatocyte membrane, which fits well with their much prolonged half-life in circulation. These data also suggest that although FVIIa binding to hepatocytes involves the gla-domain of the protease and negatively charged phospholipids on the cell surface, other determinants in FVIIa may also play a role in the binding. Otherwise, one would expect a similar or higher binding and internalization of FIX and FX in hepatocytes since FIX and FX were shown to bind to phospholipid membranes with a much higher affinity (10 to 100-fold higher) than FVIIa [31]. A higher internalization rate of activated FIX and FX, compared to their zymogen counter parts, could be the result of enhanced receptor-mediated internalization of these proteases alone or alternatively in complex with some locally produced inhibitors.

The data presented in the manuscript raises a valid question whether there is a specific receptor-mediated mechanism in liver cells for the clearance of FVIIa or other clotting proteases. The data in the literature on the binding and internalization of clotting factors in liver cells is scant. Earlier studies on tissue distribution of 125I-FVIIa injected into rats showed that FVIIa was distributed into many organs, including liver, kidneys, spleen, lungs, adrenals and pituitary gland [23,24]. When the amount of FVIIa distributed into various tissues was expressed as ng/gm tissue weight, there is little difference in the relative amount of FVIIa accumulated in various organs. These data indicate that the accumulation is not organ specific but rather related to the amount of blood the organ receives, together with the size of the organ, that determines how much FVIIa is cleared by the particular organ. The liver, being a very large organ and heavily perfused, plays a major role in the clearance.

Overall our data show that hepatocytes internalize FVIIa probably by a mechanism involving the membrane turnover rather than receptor-mediated endocytosis. More importantly our data highlights the fact that hepatoma cell lines, which originate from hepatocellular carcinomas, do not mimic primary hepatocytes and thus may not be a valid cell model system for clearance studies. Application of human primary hepatocytes is therefore preferable, however, it is impractical to obtain in large quantities of human primary hepatocytes on a continued basis. Rat primary hepatocytes appear to closely resemble human hepatocytes in their ability to support binding and internalization of FVIIa and other clotting proteins. Therefore, rat primary hepatocytes can be adopted as a model cell system to investigate potential clearance mechanisms of clotting factors.


The work is partly supported by NIH grant HL58869.


1. Hasselback R, Hjort PF. Effect of heparin on in vivo turnover of clotting factor. J Appl Physiol. 1960;15:945–48. [PubMed]
2. O’Reilly RA, Aggeler PM, Leong LS. Studies on the coumarin anticoagulation drugs: The pharmacodynamics of warfarin in man. J Clin Invest. 1963;42:1542–51. [PMC free article] [PubMed]
3. Bjorkman S, Berntorp E. Pharmacokinetics of coagulation factors: clinical relevance for patients with haemophilia. Clin Pharmacokinet. 2001;40:815–32. [PubMed]
4. Loeliger EA, van der Esch B, ter Haar Romney-Wachter CC, Booij HL. Factor VII; its turnover rate and its possible role in thrombogenesis. Thromb Diath Haemorrh. 1960;4:196–200. [PubMed]
5. Fair DS. Quantitation of Factor VII in the Plasma of Normal and Warfarin-Treated Individuals by Radioimmunoassay. Blood. 1983;62:784–91. [PubMed]
6. Fuchs HE, Pizzo SV. Regulation of Factor Xa In Vitro in Human and Mouse Plasma and In Vivo in Mouse. J Clin Invest N Y. 1983;72:2041–49. [PMC free article] [PubMed]
7. Fuchs HE, Trapp HG, Griffith MJ, Roberts HR, Pizzo SV. Regulation of factor IXa in vitro in human and mouse plasma and in vivo in the mouse. Role of the endothelium and the plasma proteinase inhibitors. J Clin Invest. 1984;73:1696–703. [PMC free article] [PubMed]
8. Seligsohn U, Kasper CK, Osterud B, Rapaport SI. Activated factor VII: presence in factor IX concentrates and persistence in the circulation after infusion. Blood. 1979;53:828–37. [PubMed]
9. Erhardtsen E. Pharmacokinetics of recombinant activated factor VII (rFVIIa) Semin Thromb Hemost. 2000;26:385–91. [PubMed]
10. Iakhiaev A, Pendurthi UR, Voigt J, Ezban M, Rao LVM. Catabolism of factor VIIa bound to tissue factor in fibroblasts in the presence and absence of tissue factor pathway inhibitor. J Biol Chem. 1999;274:36995–7003. [PubMed]
11. Fleck RA, Rao LVM, Rapaport SI, Varki N. Localization of human tissue factor antigen by immunostaining with monospecific, polyclonal anti-human tissue factor antibody. Thromb Res. 1990;59:421–37. [PubMed]
12. Drake TA, Morrissey JH, Edgington TS. Selective cellular expression of tissue factor in human tissues: Implications for disorders of hemostasis and thrombosis. Am J Pathol. 1989;134:1087–97. [PMC free article] [PubMed]
13. Chang GTG, Kisiel W. Internalization and degradation of recombinant human coagulation factor VIIa in the human hepatoma cell line HuH7. Thromb Haemost. 1995;73:231–38. [PubMed]
14. Nakabayashi H, Taketa K, Miyano K, Yamane T, Sato J. Growth of human hepatoma cells lines with differentiated functions in chemically defined medium. Cancer Res. 1982;42:3858–63. [PubMed]
15. Biessen EA, Noorman F, van Teijlingen ME, et al. Lysine-based cluster mannosides that inhibit ligand binding to the human mannose receptor at nanomolar concentration. J Biol Chem. 1996;271:28024–30. [PubMed]
16. Bajaj SP, Rapaport SI, Prodanos C. A simplified procedure for purification of human prothrombin, factor IX and factor X. Prep Biochem. 1981;11:397–412. [PubMed]
17. Rao LVM. Characterization of anti-tissue factor antibody and its use in immunoaffinity purification of human tissue factor. Thromb Res. 1988;51:373–84. [PubMed]
18. Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol. 1976;13:29–83. [PubMed]
19. Le DT, Rapaport SI, Rao LVM. Relations between factor VIIa binding and expression of factor VIIa/tissue factor catalytic activity on cell surfaces. J Biol Chem. 1992;267:15447–54. [PubMed]
20. Iakhiaev A, Pendurthi UR, Rao LVM. Active site blockade of factor VIIa alters its intracellular distribution. J Biol Chem. 2001;276:45895–901. [PubMed]
21. Tait JF, Gibson D, Fujikawa K. Phospholipid binding properties of human placental anticoagulant protein-I, a member of the lipocortin family. J Biol Chem. 1989;264:7944–49. [PubMed]
22. Schwartz AL. The hepatic asialoglycoprotein receptor. CRC Crit Rev Biochem. 1984;16:207–33. [PubMed]
23. Beeby TL, Chasseaud LF, Taylor T, Thomsen MK. Distribution of the recombinant coagulation factor 125I-rFVIIa in rats. Thromb Haemost. 1993;70:465–68. [PubMed]
24. Thomsen MK, Diness V, Nilsson P, Rasmussen SN, Taylor T, Hedner U. Pharmacokinetics of recombinant factor VIIa in the rat--a comparison of bio-, immuno- and isotope assays. Thromb Haemost. 1993;70:458–64. [PubMed]
25. Deykin D. The role of the liver in serum-induced hypercoagulability. J Clin Invest. 1966;45:256–63. [PMC free article] [PubMed]
26. Meijer DK, Scholtens HB, Hardonk MJ. The role of the liver in clearance of glycoproteins from the general circulation, with special reference to intestinal alkaline phosphatase. Pharm Weekbl Sci. 1982;4:57–70. [PubMed]
27. Kelly DA, Summerfield JA. Hemostasis in liver disease. Semin Liver Dis. 1987;7:182–91. [PubMed]
28. Rao LVM, Rapaport SI, Hoang AD. Binding of factor VIIa to tissue factor permits rapid antithrombin III/heparin inhibition of factor VIIa. Blood. 1993;81:2600–2607. [PubMed]
29. Broze GJ, Jr, Girard TJ, Novotny WF. Regulation of coagulation by a multivalent Kunitz-type inhibitor. Biochem. 1990;29:7539–45. [PubMed]
30. Scharschmidt BF, Lake JR, Renner EL, Licko V, Van Dyke RW. Fluid phase endocytosis by cultured rat hepatocytes and perfused rat liver: implications for plasma membrane turnover and vesicular trafficking of fluid phase markers. Proc Natl Acad Sci U S A. 1986;83:9488–92. [PMC free article] [PubMed]
31. McDonald JF, Shah AM, Schwalbe RA, Kisiel W, Dahlback B, Nelsestuen GL. Comparison of naturally occurring vitamin K-dependent proteins: correlation of amino acid sequences and membrane binding properties suggests a membrane contact site. Biochem. 1997;36:5120–5127. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...