FVIIa:TF Induces Cell Survival via G12/G13-Dependent Jak/STAT Activation and BclXL Production
Tissue factor (TF), apart from activating the extrinsic pathway of the blood coagulation, is a principal regulator of embryonic and oncogenic angiogenesis, inflammation, leukocyte reverse transmigration, and tumor progression. It has become clear that these events are mediated by intracellular signal transduction elicited by TF/factor VIIa (FVIIa) interaction, but the details of this signaling remain largely obscure. In this study, we show that FVIIa/TF-interaction produces STAT5 phosphorylation, STAT5 nuclear translocation and transactivation of a STAT5 reporter construct. FVIIa-dependent STAT5 activation was dependent on FVIIa proteolytic activity but not on generation of the downstream coagulation factors Xa and thrombin, nor on the TF cytoplasmic domain. FVIIa-induced STAT5 phosphorylation was dependent on functional G12/G13 class G proteins and Jak2 activity, but not Jak1 or Tyk2. Finally, we show that FVIIa leads to cell survival through a Jak2/STAT5-dependent production of the antiapoptotic STAT5 target BclXL as well as Jak2-dependent activation of the antiapoptotic protein PKB. In conclusion, our results show that FVIIa induces cell survival through STAT5-dependent BclXL production and Jak2-dependent activation of PKB. Finally, we demonstrated for the first time that TF/FVIIa-signal transduction is dependent on G12/G13 class G proteins.
Tissue factor (TF) is the main initiator of the coagulation cascade. On vessel rupture, this transmembrane protein, normally not present on cells that are in contact with the bloodstream, binds with high affinity to the zymogen factor VII (FVII). The subsequent generation of factor Xa, thrombin, and fibrin deposition finally results in the formation of a blood clot.1 In addition, a large variety of coagulation-independent functions for FVIIa/TF have been described. TF is absolutely required for embryonic blood vessel formation, because TF knockout mice are not viable due to an underdeveloped vascular system.2 Furthermore, both TF and its ligand FVIIa have been shown to play a role in tumor metastasis, whereas TF has been shown to be an independent risk factor for hepatic metastasis in patients suffering from colon cancer.3–5 Finally, the TF:FVIIa complex plays a role in sepsis and inflammation, because inhibition of this complex inhibits sepsis-induced mortality.6,7
It is generally accepted that TF:FVIIa interactions provokes intracellular signal transduction in TF-expressing cells. Among the cellular signaling cascades targeted by this interaction are the MAP kinase pathway, the PI3 kinase pathway, Ca2+ signaling and activation of small GTPases, but such pathways are not a hallmark for inflammatory signaling per se.8–12 In 1990, Bazan noted the high homology between the extracellular part of TF and the proinflammatory interferon class II type receptors, including the interferon γ receptor.13 Indeed, recent in silico analysis has shown that all current members of the cytokine receptor class II family are derived from an ancient TF gene, as TF is the only member of the protein family found in teleosts.14 However, the intracellular part of TF does not resemble a cytokine receptor, and thus a cytokine receptor–like mechanism is unlikely. Rather, TF serves as a docking site for FVIIa, the latter cleaving a protease-activated receptor similar to the thrombin receptor thus triggering signal transduction.15 Nevertheless, we explored the possibility of TF:FVIIa-induced activation of signal transduction that is normally associated with cytokine receptors, the Jak/STAT pathway.
We report that FVIIa induces STAT5 activation, subsequent nuclear translocation and transactivation of a STAT5-dependent reporter construct, as well as production of the STAT5 target BclXL. This activation was dependent on FVIIa proteolytic activity, G12/G13 GTPase activity, and Jak2, but not on the TF cytoplasmic tail. Finally, TF/FVIIa induced Jak2-dependent cell survival via activation of both STAT5 and PKB. Thus, our results show that FVIIa:TF complex formation results in activation of the Jak/STAT pathway. However, this pathway rather plays a role in FVIIa-induced cell survival than in inflammation. Thus, the activation of STAT transcription factors provides an obvious link between TF:FVIIa and their role in (patho)physiology along with cell survival and cancerous processes.
Materials and Methods
The antibodies raised against p-STAT5 (Tyr694), p-PKB, p-MAPK, and p-Tyk2 (Tyr1054/1055) were purchased from Cell Signaling Technologies. Antibodies against p-Jak1 (Tyr 1022/1023) and p-Jak2 (Tyr 1007/1008) were from Biosource, and antibodies against total Jak1 and 2 were from Upstate Biotech Inc. Antibodies against Tyk2 and BclXL were from Transduction Laboratories. Anti-FLAG and MTT were from Sigma. Total STAT 5 (directed against both A/B) G12 and G13 antibodies were from Santa Cruz. Recombinant FVIIa was purchased from Novo Nordisk, factor Xa was from Enzyme Laboratories, and thrombin was from Sigma. TAP was a kind gift from Dr G. Vlasuk (Corvas International, San Diego, Calif).
Both of the DNA constructs encoding kinase-dead Jak1 and Jak2 were generous gifts from Dr David Levy (Department of Pathology, New York University School of Medicine, New York). The dominant-negative mutant of Jak1 contains a 3-amino acid change in the kinase domain (FWYAPE→LTYAPV) and impairs its catalytic function. The dominant-negative mutant of JAK2 contains a single amino acid change (Lys→Ala) at the ATP-binding site. The kinase-deleted Tyk2 construct was kindly provided by Dr Sandra Pellegrini (Unité de Biologie des Interactions Cellulaires, Institut Pasteur, Paris). This construct encodes a protein containing amino acids 1 to 895 of Tyk2. FLAG-tagged STAT5A and B were from Dr James Ihle (St Jude’s Children’s Hospital, Memphis, TN), G12 and G13 dominant-negative constructs (G12G228A) and (G13G225A) were from Dr Stefan Offermanns (Institute of Pharmacology, University of Heidelberg), and the STAT5-responsive NTCP-luciferase construct was provided by Mary Vore (Graduate Center of Toxicology, University of Kentucky). The dominant-negative pXM-STAT5A (truncated at residue 750) construct was obtained from Dr P. Coffer (Utrecht University), and this construct serves as an inhibitor to both STAT5A and STAT5B.16 The insert was excised HindIII/XhoI and inserted into pcDNA3.1 (Invitrogen).
Cell Culture and Transfection
Baby hamster kidney cells, either stably transfected with full-length TF or with a TF cytoplasmic domain-deleted mutant truncated at residue 247 (all a generous gift from Dr L. Petersen, Novo Nordisk), were maintained in Dulbecco’s Modified Eagle’s Medium, supplemented with 10% fetal calf serum (FCS, Gibco) and penicillin/streptomycin, at 37°C and 5% CO2 in a humidified environment.8,17
Transient transfections were performed as follows: for FLAG-tagged STAT5A, FLAG-tagged STAT5B or the luciferase expression constructs, cells were grown in 6-wells plates until 50% confluence and transfected with 0.4 μg construct using 1.6 μL Enhancer and 5 μL Effectene transfection reagent (Qiagen). For transient transfections using dominant-negative constructs, 2.5 μg construct, 21 μg of Enhancer reagent, and 133 μL of Effectene reagent were used. Using GFP-expression constructs, the transfection efficiency was determined to be at least 75%
Stable dominant-negative STAT5-expressing BHKTF cells were generated as follows: cells were grown in 6-wells plates until 50% confluence and transfected with 0.4 μg pcDNA3.1, using Effectene reagent. Cells were grown in DMEM containing 700 μg/mL G418, and positive clones were selected and screened for mutant STAT5 expression. Three clones (Nos. 1, 7, and 10) were selected for the experiments based on mutant STAT5 expression.
Treated cells were lysed in 0.5 mL lysis buffer (20 mmol/L Tris pH 7.5, 150 mmol/L NaCl, 1 mmol/L Na2EDTA, 1 mmol/L EGTA, 1% Triton X-100, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L beta-glycerophosphate, 1 mmol/L Na3VO4, and 1 μg/mL leupeptin) and were collected by scraping. Subsequently, the lysate was centrifuged at 14 000 rpm for 2 minutes, and the supernatant was precleared with 20 μL 50% protein A-Sepharose for 1 hour. The lysate was then incubated overnight with 1 μg antibody. The immunocomplex was precipitated with 20 μL 50% protein A-Sepharose for 1 hour. The beads were washed three times with lysis buffer and were subsequently resuspended in 40 μL denaturing sample buffer (125 mmol/L Tris/HCl, pH 6.8; 4% SDS; 2% β-mercaptoethanol; 20% glycerol, 1 mg bromophenol blue).
Cell lysates were prepared and separated into nuclear and cytosolic fractions, as described before,18 with some modifications; stimulated and washed cells (1 to 7×106) were scraped in 200 μL ice cold hypotonic lysis buffer (20 mmol/L HEPES, pH 7.9, 1.5 mmol/L MgCl2, 10 mmol/L KCl, 0.5 mmol/L dithiothreitol, 0.5 mmol/L pefabloc, and 0.5 μg/mL of leupeptin and aprotinin).19 The lysate was passed 10 times through a 27-gauche needle. After centrifugation at 600g, the resultant pellets (nuclear fraction) were washed again in lysis buffer and were resuspended in 300 μL 1× sample buffer. The supernatants (cytosolic fraction) were further diluted to 300 μL with 3× sample buffer.
After treatment, cells were rapidly harvested by adding 100 μL of heated (95°C) sample buffer and the lysates were collected by scraping. After 5 minutes of incubation at 95°C, 30 μL of the lysates were loaded onto SDS-PAGE and subsequently transferred to a PVDF membrane. The membranes were blocked with Tris-buffered saline supplemented with 0.1% Tween (wash buffer) and 2% low-fat milk powder, and incubated with primary antibody overnight at 4°C, diluted 1:1000 in wash buffer containing 2.5% BSA. Subsequently, the membranes were incubated with a horseradish peroxidase–conjugated secondary antibody in wash buffer, containing 2% low-fat milk powder. The bands were visualized, using Lumilight plus ECL substrate from Roche and a chemiluminescence detector with a cooled CCD camera from Syngene.
Cells were grown in 24-wells and transfected with either the STAT5-responsive construct NTCP-luciferase, or a CMV-driven luciferase construct for 16 hours. Transfections were performed as described above. After transfection, the cells were serum-starved for another 16 hours. Subsequently, cells were stimulated with 100 nmol/L FVIIa for 3 hours. Cells were lysed in 100 μL luciferase assay buffer (25 mmol/L glycylglycine pH 7.9, 15 mmol/L MgSO4, 4 mmol/L EGTA, 1% Triton X-100, 1 mmol/L dithiothreitol) of which 50 μL was used for the detection of luciferase activity on a Packard Topcount microplate scintillation counter.
Cell viability was assessed using an MTT reduction assay (a colorimetric assay in which a tetrazolium compound is bioreduced by cells into a colored formazan product in direct proportion to the number of living cells in culture) in 24-well plates. Shortly, cells were starved for 1 day, after which FVIIa was added. Viability was measured at the times indicated in the graphs. The control was set at 100%.
Statistical Analysis and Quantification of the Western Blots
In all graphs, means and standard errors of at least triplicates are shown. When Western blots were analyzed for phosphorylation of STAT5, MAP kinase, or PKB, equal loading was determined by blotting for total protein as well. Antibody bands were quantified using Genetools from Syngene.
FVIIa/TF Interaction Induces a Proteolytic Activity-Dependent, TF Cytoplasmic Tail-Independent Phosphorylation of STAT5
There is substantial sequence homology between the extracellular domains of TF and the interferon receptors. The intracellular domains of TF and the interferon receptors, however, are not much alike but an intriguing question remains whether TF together with its ligand, FVIIa, activates Jak/STAT pathways. Therefore, we set out to investigate the possible involvement of STATs in TF-dependent signal transduction and whether the intracellular tail is involved in such signaling. To this end, BHK cells either expressing full-length TF (BHKTF) or a cytoplasmic domain deletion mutant (BHKTFΔcyto) were stimulated with recombinant FVIIa and STAT phosphorylation was determined using phosphospecific antibodies and Western blotting.8,17 We determined FVIIa-induced phosphorylation states of STAT1, STAT3, and STAT5. Although we did not observe FVIIa-induced phosphorylation of STAT1 and STAT3 (results not shown), BHKTF cells reacted to FVIIa with a transient increase in phospho-STAT5 immunoreactivity, which was already detectable after 5 minutes, reached a maximum effect 10 minutes after stimulation and returned to basal levels after approximately 60 minutes, whereas pan-STAT5 immunoreactivity was not affected (Figure 1A). Importantly, BHKTFΔcyto cells responded to FVIIa stimulation with transient STAT5 phosphorylation indistinguishable of that observed in BHK cells transfected with full-length TF, and therefore, we concluded that cytoplasmic domain of TF is not involved in this phosphorylation. Interestingly, this FVIIa-induced phosphorylation was much more transient than that observed after stimulation with insulin, an established inducer of STAT5 phosphorylation (Figure 1A).19 We also assessed the importance of FVIIa proteolytic activity using active site-inhibited FVIIa (FFR-VIIa or FVIIai). We observed that although a 10-minute stimulation with FVIIa potently induced STAT5 activation, FVIIai was ineffective in this respect (Figure 1B). We concluded that the FVIIa-dependent STAT5 phosphorylation is mediated by the extracellular domain of TF and FVIIa proteolytic activity. Dose-response analysis of FVIIa-induced STAT5 phosphorylation revealed that enhanced STAT5 phosphorylation in response to this protease was already detectable at concentrations as low as 1 nmol/L (Figure 1C), lying well within the physiological ranges of FVII in plasma (10 nmol/L). Finally, we assessed which isoform of STAT5, STAT5A, or STAT5B is/are activated on FVIIa stimulation. To this end, we transiently transfected BHKTF with expression constructs encoding either FLAG-tagged STAT5A or FLAG-tagged STAT5B.20 Subsequently, cells were stimulated with FVIIa for various times and FLAG-tagged STAT5 was immunoprecipitated using a FLAG antibody. Finally, phosphorylation of STAT5 was determined using a phosphospecific antibody for STAT5. We observed that, whereas total levels of FLAG-STAT5A and FLAG-STAT5B remained unchanged, the phosho-STAT5 immunoreactivity was transiently increased with respect to both STAT5A and STAT5B (Figure 1D), and hence, both STAT5 isoforms are a target for FVIIa-dependent signaling.
FVIIa-Induced STAT Phosphorylation Is not Dependent on Generation of Thrombin or FXa
After exposure of TF to the bloodstream, interaction of TF with FVII will take place, leading to FVII activation. The thus-generated FVIIa converts FX to FXa, and in turn, FXa activates thrombin. Hence, addition of FVIIa to cells may lead to FXa and thrombin generation. In order to investigate a possible role of FXa or thrombin in FVIIa-dependent STAT5 activation, we tested the effect of the FXa inhibitor TAP and the thrombin inhibitor hirudin on FVIIa-evoked STAT5 phosphorylation. As is obvious from Figure 2, both FXa and thrombin are capable of inducing STAT5 phosphorylation, and the effects of FXa and thrombin are sensitive to 200 nmol/L of TAP and 25 U/mL of hirudin, respectively. These inhibitors, however, did not influence FVIIa-induced STAT5 phosphorylation (Figure 2). Thus, although multiple factors of the coagulation cascade have the capacity to activate STAT5, the FVIIa-mediated STAT5 phosphorylation does not rely on either FXa or thrombin generation.
FVIIa-Mediated STAT5 Phosphorylation Is Dependent on G12/G13, but not on Gi Heterotrimeric G Proteins
FVIIa is believed to induce signaling through a protease-activated receptor (PAR), a heterotrimeric G protein–coupled receptor that has high homology with the thrombin receptor. However, until now, the nature of the G proteins involved is highly unclear. Therefore, we decided to investigate the involvement of two classes of G proteins that are frequently activated on thrombin stimulation; Gi and G12/G13. To investigate the role of Gi, BHKTF cells were preincubated for 6 hours with the Gi inhibitor pertussis toxin (PTX). Whereas this toxin potently inhibited MAP kinase phosphorylation induced by lysophosphatidic acid (LPA), a well-known Gi activator, PTX did not inhibit FVIIa-induced STAT5 phosphorylation (Figure 3A), ruling out a role for Gi in FVIIa-induced STAT5 phosphorylation. Next, we determined the influence of G12/G13 in this system, through transient transfection of BHKTF cells with inactive G12a or G13α expression vectors. The transfection efficiency in our system was at least 75% as determined with GFP-expression constructs. As is clear from Figure 3B, both inactive G12a or G13α inhibited FVIIa-induced STAT5 phosphorylation. However, the use of dominant-negative G12a or G13α did not inhibit, but rather enhanced insulin-induced STAT5 phosphorylation. This demonstrates that functional G12 and G13 proteins are absolutely required for FVIIa-induced STAT5 phosphorylation.
FVIIa Induces STAT Phosphorylation via Activation of Jak2
The upstream kinases activating STATs are known as Jaks (Janus kinases) of which four forms exist: Jak1, Jak2, Jak3, and Tyk2. Using phosphospecific antibodies, the phosphorylation states of Jak1, Jak2, and Tyk2 were determined. Figure 4A shows that Jak1, Jak2, and Tyk2 phosphorylation are transiently upregulated after FVIIa stimulation. Whereas Jak2 and Tyk2 phosphorylation return to basal levels within 15 minutes, Jak1 phosphorylation was more persistent; phosphorylation was still observed after 45 minutes (not shown). Furthermore, phosphorylation of these kinases was already observed at FVIIa concentrations as low as 1 nmol/L (Figure 4B).
To establish the nature of Janus kinases in the phosphorylation of STAT5, we explored the role of these kinases in this system by transiently transfecting BHKTF cells with DNA-constructs, encoding kinase-dead variants of Jak1, Jak2, and Tyk2. These proteins have previously been shown to potently inhibit endogenous Jak1, Jak2, and Tyk2 enzymatic activity.21,22 As can be seen in Figure 4C, transfection of cells with kinase-dead Jak2 abrogated FVIIa-induced STAT5 phosphorylation, whereas the use of a kinase-dead Jak1 or Tyk2 mutant did not have this effect. We also preincubated BHKTF cells with the specific Jak2 inhibitor AG490. As can be seen in Figure 4D, a 25 μmol/L preincubation with this inhibitor abrogates FVIIa-induced STAT5 phosphorylation. We concluded that, although FVIIa induces activation of Jak1, Jak2, and Tyk2, only Jak2 is essential in FVIIa-dependent STAT5 phosphorylation.
Jak2 Activity Is Essential for FVIIa-Induced PKB Phosphorylation
To check whether the effects of the Jak2 inhibitor AG490 on STAT5 phosphorylation were specific, the effect of AG490 incubation on other FVIIa targets such as MAP kinase and the anti-apoptotic protein PKB was also investigated. AG490 incubation showed no effect on MAP kinase phosphorylation (Figure 5A); however, it did inhibit PKB phosphorylation (Figure 5B). Transfection of BHKTF cells with dominant-negative Jak2 showed similar effects (Figure 5C), pointing out a role for Jak2 in FVIIa-induced PKB activation.
FVIIa Induces STAT5 Translocation and Transactivation
STAT5 translocation to the nucleus is essential for STAT5 function, and thus if FVIIa-dependent STAT5 phosphorylation is relevant, it should result in such a translocation. Hence, BHKTF were subjected to FVIIa for various times, and subsequently, cell fractionation was used to produce nucleus-enriched and cytosol-enriched fractions. FVIIa treatment produced a marked increase in STAT5 phosphorylation in both the cytosolic and the nuclear fraction (Figure 6A) that was observed at concentrations as low as 1 nmol/L (Figure 6B). Strikingly, STAT5 phosphorylation in the cytosol was enhanced compared with STAT5 phosphorylation in the nucleus. The latter might be due to the presence of nuclear-localized STAT phosphatases, such as TC-PTP, which have been shown to dephosphorylate STAT5 in the nucleus.23 To further establish the relevance of FVIIa-induced STAT5 phosphorylation, we made use of a STAT5-responsive luciferase reporter construct. Cells were transfected with either this construct or a constitutive luciferase-expressing construct. After starvation, cells were stimulated with 100 nmol/L FVIIa for 3 hours, and luciferase activity was determined. Figure 6C shows that a 3-hour treatment with FVIIa strongly enhances luciferase activity. Thus, FVIIa provokes Jak2-dependent STAT5 phosphorylation, nuclear translocation, and transactivation.
FVIIa Induces Jak2/STAT5-Dependent Expression of BclXL
BclXL is an antiapoptotic protein that has been shown to be upregulated as a result of STAT5 activation.24,25 Because FVIIa in our systems activates STAT5, we decided to investigate FVIIa-induced BclXL expression. Therefore, we starved BHKTF cells for 2 hours and stimulated them with 100 nmol/L of FVIIa. As can be seen in Figure 7A, stimulation of these cells results in an increase in BclXL in time that showed to be concentration-dependent (Figure 7B). This expression could not be reversed with 10 μmol/L of AG490, but BclXL expression could be inhibited with the same concentration of Jak2 inhibitor (25 μmol/L) that also potently inhibited STAT5 phosphorylation (Figure 7C).
To investigate the role of STAT5 in FVIIa-induced BclXL expression, we constructed BHKTF cells stably expressing dominant-negative STAT5 (Figure 7D). Whereas wild-type cells considerably upregulate BclXL after FVIIa stimulation, dominant-negative STAT5-expressing cells show only a moderate upregulation of BclXL (Figure 7E), whereas activation of FVIIa targets such as MAP kinase, PKB, and Jak2 were not affected (Figure 7F). Therefore, we concluded that FVIIa-induced BclXL expression is mediated, at least partially, via STAT5.
Jak2/STAT5 Are Involved in FVIIa-Induced Cell Survival
Because Jak2 and STAT5 were shown to be involved in FVIIa-driven BclXL expression and Jak2 activity was shown to be essential for PKB phosphorylation in BHKTF cells, we decided to investigate the role of Jak2 and STAT5 in FVIIa-induced cell survival, a recently described phenomenon.26,27 Therefore, cells were starved and preincubated with 10 μmol/L and 25 μmol/L AG490 before stimulation with FVIIa. FVIIa (100 nmol/L) efficiently promoted cell survival in serum-starved cells (Figure 8A) and 10 μmol/L of AG490 did not inhibit this effect. However, 25 μmol/L of AG490, the same concentration that efficiently inhibited STAT5 phosphorylation and BclXL expression, also inhibited FVIIa-induced survival. Finally, to address the role of STAT5, wild-type BHKTF cells and STAT5 mutant cells were starved and incubated with FVIIa. Whereas wild-type cells show a considerable survival after 1 and 2 days of FVIIa stimulation, STAT5 mutant cells show a greatly diminished cell survival (Figure 8B). Therefore, we conclude that FVIIa induces cell survival via Jak2 and STAT5 activation.
It is now generally recognized that TF, apart from its role in coagulation, is an important mediator in inflammation and cancerous processes.3–7 The molecular basis, however, of this action remains unclear. In the present report, we provide evidence that FVIIa provokes Jak2-dependent phosphorylation of STAT5 via G12/G13, followed by nuclear translocation and transactivation by this transcription factor. Furthermore, we show that Jak2 and STAT5 are involved in FVIIa-driven BclXL expression and cell survival.
We showed that STAT5 phosphorylation is independent of the TF cytoplasmic tail, but relies on the proteolytic activity of FVIIa. These data are most consistent with the often-proposed platform hypothesis, in which TF serves as a docking site for FVIIa, which in turn proteolytically cleaves and thereby activates a protease-activated G protein–coupled receptor. One of the proteins that has frequently been implicated in FVIIa:TF-dependent signaling is protease-activated receptor 2 (PAR-2),15 although involvement of any of the known PARs in BHKTF has been excluded.28
A transient phosphorylation of STAT5 coinciding with activation of Jak1, Jak2, and Tyk2 was observed. Although only functional Jak2 was required for STAT5 responses, other Janus kinases may have downstream targets such as other members of the STAT family of signal transducers, but obviously further work is needed to substantiate this notion.
We were able to show that FVIIa upregulated the STAT5 target BclXL, suggesting a role for FVIIa-induced STAT5 phosphorylation in the prevention from apoptosis. Such a role has indeed been shown by ourselves and others, and could form an explanation for the role of TF and FVIIa in tumor growth and metastasis.3–5,26,27 Indeed, also in this article, we show that FVIIa induces survival in cells that are stimulated to go into apoptosis, but this survival was inhibited by the Jak2 inhibitor AG490 or expression of dominant-negative STAT5. Moreover, Jak2 activity in our system proved essential for activation of the antiapoptotic kinase PKB, the latter being a key player in FVIIa-induced cell survival.26,27 Indeed, Jak2 has been found to be an upstream activator of PKB before.29 Therefore, we hypothesize that Jak2 exerts its antiapoptotic effects through activation of both STAT5 and PKB, thus positively influencing tumor growth and possibly metastasis. Currently, it is unknown whether Jak2 influences PKB activity directly, or via PKB’s upstream kinase PI-3 kinase, but previous work suggests that PI-3 kinase may be an intermediate in Jak2-dependent PKB activation.10,11
Induction of STAT5, Jak1, and Jak2 phosphorylation was already observed at concentrations as low as 1 nmol/L. Because plasma levels of FVII are normally maintained at 10 nmol/L, this concentration of FVIIa appears to be physiologically relevant. FVIIa-induced activation of proteins such as c-Akt/PKB occurs at slightly higher concentrations (10 nmol/L), suggesting that FVIIa-induced STAT5 phosphorylation is a relatively efficient process. It is possible that this difference might be due to activation of different G proteins coupling to the FVIIa proteolytic target. We indeed show for the first time that FVIIa-induced Jak2/STAT5 activation occurs via the heterotrimeric G proteins G12/G13, whereas Gi played no role. Both the G12/G13 and Gi classes of G proteins were investigated because these signal transducers are also frequently activated by thrombin. A third class of G proteins, also activated by thrombin is Gq. Activation of the latter G protein leads to transient calcium release and subsequent calcium-dependent signal transduction. However, in contrast to thrombin, FVIIa does not elicit calcium signals in BHKTF cells, ruling out Gq activation.28
In conclusion, FVIIa induces activation of the Jak/STAT pathway via G proteins, leading to enhanced cell survival.
This research was sponsored by the Dutch Cancer Fund (UVA 98-1855).
Original received November 6, 2003; revision received February 26, 2004; accepted March 2, 2004.
Mueller BM, Reisfeld RA, Edgington TS, Ruf W. Expression of tissue factor by melanoma cells promotes efficient hematogenous metastasis. Proc Natl Acad Sci USA. 1992; 89: 11832–11836.
Taylor FB Jr, Chang AC, Peer G, Li A, Ezban M, Hedner U. Active site inhibited Factor VIIa (DEGR VIIa) attenuates the coagulant and interleukin-6 and-8, but not tumor necrosis factor responses of the baboon to LD100. Blood. 1998; 91: 1609–1615.
Poulsen LK, Jacobsen N, Sørensen BB, Bergenhem NC, Kelly JD, Foster DC, Thastrup O, Ezban M, Petersen LC. Signal transduction via the mitogen activated protein kinase pathway induced by binding of coagulation factor VIIa to tissue factor. J Biol Chem. 1998; 273: 6228–6232.
Camerer E, Røttingen JA, Gjernes E, Larsen K, Skartlien AH, Iversen JG, Prydz H. Coagulation factors VIIa and Xa induce cell signaling leading to the upregulation of the egr-1 gene. J Biol Chem. 1999; 274: 32225–32233.
Versteeg HH, Hoedemaeker I, Diks SH, Stam JC, Spaargaren M, van Bergen en Henegouwen PM, van Deventer SJ, Peppelenbosch MP. Factor VIIa/tissue factor-induced signaling via activation of Src-like kinases, Phosphatidylinositol 3-kinase, and Rac. J Biol Chem. 2000; 275: 28750–28756.
Versteeg HH, Sørensen BB, Slofstra SH, Van den Brande JH, Stam JC, van Bergen en Henegouwen PM, Richel DJ, Petersen LC, Peppelenbosch MP. VIIa/tissue factor interaction results in a tissue factor cytoplasmic domain-independent activation of protein synthesis, p70, and p90 S6 kinase phosphorylation. J Biol Chem. 2002; 277: 27065–27072.
Røttingen JA, Enden T, Camerer E, Iversen JG, Prydz H. Binding of human factor VIIa to tissue factor induces cytosolic Ca2+ signals in J82 cells, transfected Cos-1 cells, Madin Darby canine kidney cells and in human endothelial cells, induced to synthesize tissue factor. J Biol Chem. 1995; 270: 4650–4660.
Bazan JF. Structural design and molecular evolution of a cytokine receptor superfamily. Proc Natl Acad Sci USA. 1990; 87: 6934–6938.
Camerer E, Huang W, Coughlin SR. Tissue factor and factor X-dependent activation of protease-activated receptor 2 by factor VIIa. Proc Natl Acad Sci USA. 2000; 97: 5255–5260.
Buitenhuis M, Baltus B, Lammers JW, Coffer PJ, Koenderman L. Signal transducer and activator of transcription 5a (STAT5a) is required for eosinophil differentiation of human cord blood-derived CD34+ cells. Blood. 2003; 101: 134–142.
Sørensen BB, Freskgård PO, Nielsen LS, Rao LV, Ezban M, Petersen LC. Factor VIIa-induced p42/44 mitogen-activated protein kinase activation requires the proteolytic activity of factor VIIa and is independent of the tissue factor cytoplasmic domain. J Biol Chem. 1999; 274: 21349–21354.
Mao PL, Beauchemin M, Bedard PA. Quiescence-dependent activation of the p20K promoter in growth-arrested chicken embryo fibroblasts. J Biol Chem. 1993; 268: 8131–8139.
Chen J, Sadowski HB, Kohanski RH, Wang L. STAT5 is a physiological substrate of the insulin receptor. Proc Natl Acad Sci USA. 1997; 94: 2295–2300.
Wang D, Stravopodis D, Teglund S, Kitazawa J, Ihle JN. Naturally occurring dominant negative variants of Stat5. Mol Cell Biol. 1996; 16: 6141–6148.
Zong CS, Chan J, Levy DE, Horvath C, Sadowski HB, Wang LH. Mechanism of STAT3 activation by insulin-like growth factor I receptor. J Biol Chem. 2000; 275: 15099–15105.
Velazquez L, Mogensen KE, Barbieri G, Fellous M, Uze G, Pellegrini S. Distinct domains of the protein tyrosine kinase tyk2 required for binding of interferon-α/β and for signal transduction. J Biol Chem. 1995; 270: 3327–3334.
Aoki N, Matsuda T. A nuclear protein tyrosine phosphatase or-PCP is a potential negative regulator of the PRL-mediated signaling pathway: dephosphorylation and deactivation of signal transducer and activator of transcription 5A and 5B by TC-PTP in nucleus. Mol Endocrinol. 2002; 16: 58–69.
Gesbert F, Griffin JD. Bcr/Abl activates transcription of the Bcl-X gene through STAT5. Blood. 2000; 96: 2269–2276.
Sørensen BB, Rao LV, Tornehave D, Gammeltoft S, Petersen LC. Anti-apoptotic effect of coagulation factor VIIa. Blood. 2003; 102: 1708–1715.