| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cell Biology/Signaling |
From the Department of Medical Sciences, Clinical Chemistry, Uppsala University, Sweden.
Correspondence to Agneta Siegbahn or Teet Velling, Department of Medical Sciences, Clinical Chemistry, Academic Hospital, SE-751 85 Uppsala, Sweden. E-mail agneta.siegbahn{at}medsci.uu.se or teet.velling@medsci.uu.se
Abstract
Background— We have previously reported the potentiation of PDGF-BB–induced chemotaxis of fibroblasts, vascular smooth muscle cells, and endothelial cells by FVIIa. Here we studied the role of TF/FVIIa and the induced signaling pathways in regulation of chemotaxis of human monocytes, fibroblasts, and porcine aorta endothelial cells.
Methods and Results— Human monocytes were obtained by using Ficoll-Paque gradient and the MACS system (for highly purified population), fibroblasts and PAE cells have been characterized previously. Inhibitors of selected signaling intermediates were used, and the effect of TF/FVIIa on the migratory response of all cells to chemotactic agents was analyzed. The induced signaling was studied by immunoprecipitation and Western blotting. TF/FVIIa complex selectively enhanced PDGF-BB–induced chemotaxis in a Src-family, PLC, and PAR-2–dependent manner. Using PAE cells we identified c-Src and c-Yes as the Src-family members activated by TF/FVIIa. We report for the first time the PAR-2 and Src family-dependent transactivation of PDGFRβ by TF/FVIIa involving phosphorylation of a subset of PDGFRβ tyrosines.
Conclusions— The described transactivation is a likely mechanism of TF/FVIIa-mediated regulation of PDGF-BB–induced chemotaxis. Similar behavior of 3 principally different cell types in our experimental setup may reflect a general function of TF in regulation of cell migration.
TF/FVIIa was found to selectively potentiate the PDGF-BB–induced chemotaxis of 3 principally different cell types. We provide data concerning the signaling pathways involved in this potentiation, and describe the TF/FVIIa-induced, PAR-2, c-Src, and c-Yes–dependent transactivation of PDGFRβ as a likely mechanism of its regulation.
Key Words: TF/FVIIa PDGFRβ transactivation cell signaling chemotaxis
On binding of the coagulation factor VII (FVII) to its receptor tissue factor (TF), the proteolytically active TF/FVIIa complex forms. Besides initiating the coagulation process in vivo,1 this complex is involved in various nonhemostatic functions such as wound healing, local inflammation, atherosclerosis, angiogenesis, metastasis and tumor invasion,2–5 and regulation of gene expression.6–8 Directed cell migration, regulated by complex signaling reactions, is a critical component of these processes.9,10 The importance of TF cytoplasmic domain, and the role of serine residues in the TF cytodomain for cell migration have been recently established11,12 and phosphorylation of several signaling intermediates on the TF/FVIIa interaction has been reported.11,13–16 Nevertheless, the precise TF-dependent mechanism regulating cell migration, and the utilized signaling pathways have not been exhaustively explored.
A G protein–coupled receptor (GPCR), protease activated receptor-2 (PAR-2), activated on different cell types by the TF/FVIIa complex,6 is believed to mediate the nonhemostatic functions of TF/FVIIa, including transmission of cellular signals,4,5 although there is evidence for the PAR-2-independent reactions17,18 (and our unpublished results). Recently, transactivation of epidermal growth factor receptor (EGFR) by GPCR was reported,19 and the precise mechanism was characterized.20,21 Transactivation of the platelet-derived growth factor receptor β (PDGFRβ) by different GPCR has also been described,22,23 but the used mechanism is unknown.
We have previously shown that the TF/FVIIa complex positively regulates the platelet derived growth factor-BB (PDGF-BB)–induced chemotaxis.3,12 The aim of this study was to investigate the molecular mechanism by which TF/FVIIa enhances cell migration toward the PDGF-BB. We used cell lines with constitutive TF expression (porcine aorta endothelial cell line transfected with human TF and PDGFRβ, and human foreskin fibroblasts 1064Sk), and freshly isolated lipopolysaccharide (LPS)-stimulated human monocytes.3,12,24
We show that TF/FVIIa engages in a cross-talk with PDGFRβ by selectively potentiating the response of the cells to PDGF-BB differently from 5 other chemotactic agents. This potentiation is regulated by PAR-2, c-Src and c-Yes kinases, and PLC isoform(s) other than PLC
1 and PLCβ3, but not by MAP kinases. Most importantly we report, for the first time, the Src kinase–dependent transactivation of PDGFRβ by TF/FVIIa, and suggest this as the mechanism of the observed potentiation of PDGF-BB–induced cell migration by TF/FVIIa.
Methods
Cell Lines
PAE cells transfected with human PDGF β-receptors and TF (PAE/PDGFRβ,TF cells), or only PDGF β-receptors, and human foreskin fibroblasts 1064Sk and 1523 expressing endogenous TF, have been characterized previously.3,12,25
Isolation of Monocytes
The monocyte-containing fraction was isolated from whole blood of healthy donors by centrifugation in a Ficoll-Paque gradient. The cells were stimulated with 10 ng/mL bacterial LPS for 3 hours, washed twice with PBS to remove the platelets and lymphocytes, and the adherent monocytic cells were recovered. A highly purified monocyte population was obtained using the MACS system (Miltenyi Biotech Inc).
Reagents
Recombinant human FVIIa and catalytically inactive FVIIa (FFR-FVIIa) were from Novo Nordisk A/S, recombinant human PDGF-BB from Amgen Inc, PAR-2 agonist peptide SLIGKV-NH2, LPS, f-MLP, Zymosan A and pertussis toxin (PTX) from Sigma- Aldrich, and the polyclonal neutralizing antibody to PAR-2 was a gift from Dr Wolfram Ruf (La Jolla, Calif). C-reactive protein (CRP) was from Chemicon, MCP-1, and interleukin (IL)-8 from R&D systems, and Ficoll-Paque from Amersham Biosciences. The used inhibitors were: PD98059 (MEK) (Cell Signaling); SU6656 (Src family), U73122 (PLC), and the inactive compound U73343 (Calbiochem). The following antibodies were purchased: phospho-Src family (pY416), phospho-PLC
1 (pY783), phospho-PLCβ3 (pS537), PLC
1 and PLCβ3 pAb, and HRP-linked anti-rabbit and anti-mouse IgG (Cell Signaling); c-Src (H-12) and c-Fyn (15) mAb, and c-Yes (3) and PDGFR pAb, and HRP-anti-goat (Santa Cruz); anti-phosphotyrosine (PY20-HRPO) and anti-FAK (Transduction Laboratories). The PDGFRβ site-specific phosphoantibodies were: pY740 (sc-17173), pY857 (sc-12907) and pY716 (sc-16569) (Santa Cruz); pY579 (GeneTex); rabbit monoclonal pY740 (32A9), pY771 (76D6), pY1021 (6F10) and pY751 (88H8) (Cell Signaling); and pY579 and pY1009 (LabFrontier). The Human PDGF-AB and PDGF-BB ELISA kits were from R&D Systems.
Complement Activation
Zymosan-activated serum (ZAS) with its predominant component, complement factor C5a, was obtained as described previously,26 aliquoted and stored at –70°C.
Chemotaxis Assay
The effect of FVIIa, FFR-FVIIa, or SLIGKV, and the selected pharmacological inhibitors, on chemotactic migration of the LPS-stimulated (10 ng/mL) monocytes, and fibroblasts and endothelial cells, was analyzed in a modified Boyden chamber assay applying a leading front technique essentially as described previously.12,27,28 For each set of experiments, the migration of cells toward the assay media without chemotactic agents served as a control and was set as 100%. The results are expressed as the mean of at least 3 observations in different parts of the filter, analyzed in duplicates, and presented as % of control.
Immunoprecipitation and Western Blotting
Subconfluent PAE/PDGFRβ,TF cells were serum-starved for 24 hours and seeded to type I collagen (Vitrogen) before analysis of Src family kinases. A subset of cells was preincubated with SU6656, and FVIIa, SLIGKV, or PDGF-BB (a positive control) were added as agonists for the indicated times. Immunoprecipitation with c-Src, c-Yes, c-Fyn, or PDGFRβ antibodies and immunoblotting were carried out as described.29 The membranes were exposed to autoradiography films, bands from the representative experiments were scanned with the CanonScan 4200F scanner, and the relative extent of the protein phosphorylation was quantified densitometrically using the ImageJ software.
Statistical Analyses
Statistical analysis of the data were performed using Student unpaired t test in Statistica for Windows and Graph Pad Prism 5.0 (GraphPad software, Inc). The results are expressed as mean±SEM, and probability values (2-tailed) <0.05 are considered statistically significant.
Results
TF/FVIIa Enhances the Migration of Human Monocytes, Fibroblasts, and PAE/PDGFRβ,TF Cells to PDGF-BB, but not to Five Other Potent Chemoattractants, in a PAR-2, Src-Family, and PLC-Dependent Manner
The cell surface expression of TF was comparable in LPS-stimulated monocytes, fibroblasts, and PAE/PDGFRβ,TF cells agreeing with our previously published data.3,12,30 Concentration gradients of MCP-1, f-MLP, CRP, complement factor ZAS/C5a, and IL-8 all induced migration of human monocytes without any significant effect of the added FVIIa or FFR-FVIIa (Figure 1A through 1E). However, migration of all 3 cell types toward a concentration gradient of PDGF-BB was strongly enhanced by both 10 nmol/L and 100 nmol/L FVIIa or 50 µmol/L SLIGKV, and it was abolished by blocking the function of PAR-2 (Figures 1F and 2
B). To investigate the role of Src-family kinases in the effect of FVIIa, the cells were treated with 2 µmol/L of the Src family kinase inhibitor SU6656. SU6656 abolished the hyperchemotactic response induced by FVIIa as well as by SLIGKV in all cell types tested (Figures 1F and 2
B). Incubation of the cells with 0.5 µg/mL PTX, which interferes with G
0 and G
i-type G protein
-subunit–mediated signals, abolished chemotaxis toward 10 nmol/L f-MLP but did not reverse the effect of FVIIa on PDGF-BB–induced migration (Figure 2B). Similarly, the MEK-inhibitor PD98059 had no effect on any of the cell types in the same assay (Figure 2B). We have previously shown that the TF/FVIIa complex formation leads to the activation of PLC.3 The widely used PLC inhibitor U73122 eliminated the enhanced chemotactic response (Figure 2B). To exclude that the inhibitors affect the basal cell migration the cells pretreated with the increasing concentrations of SU6656 and U73122, and the same concentrations of pertussis toxin and PD98059 as used in Figure 2, were subjected to migration assays. No change in cell migration was noted (see supplemental Figures I and II, available online at http://atvb.ahajournals.org), and our results thus confirm the selective inhibition of the effect of FVIIa by the used inhibitors.
|
|
PLC
1, which was recently suggested to regulate the early migratory response mediated by adhesion and growth factor receptors,31 was not activated by TF/FVIIa or by the SLIGKV peptide in PAE/PDGFRβ,TF cells. Likewise, the phosphorylation of PLCβ3 active site Ser 537 was not affected by TF/FVIIa (data not shown) leaving the involved U73122-sensitive PLC isoform to be identified.
c-Src and c-Yes, but not c-Fyn, Are Activated by TF/FVIIa
Analysis of Src family kinases in monocytes was hampered by a strong activation of these enzymes during the purification process, and therefore the PAE/PDGFRβ,TF cells were chosen for this purpose. Prominent phosphorylation of the active site Y416 of Src-family kinases was detected on the addition of FVIIa or SLIGKV to the cells adhering to type I collagen, and it was reduced to the background level by SU6656 (Figure 3A; collagen was used as an adhesive substrate to ensure comparable conditions to the migration assay). To discriminate between c-Src, c-Yes, and c-Fyn kinases, the common Src-family members in the 3 cell types used, these enzymes were immunoprecipitated from the PAE/PDGFRβ,TF cells treated as above, and analyzed in Western blotting with the pY416 Src-family antibody. Figure 3B depicts the result of this experiment whereby c-Src and c-Yes, but not c-Fyn, are activated on the formation of the TF/FVIIa complex. Activation of these enzymes by integrin type collagen receptors was not detected under these conditions (Figure 3B, "Col").
|
TF/FVIIa Complex Transactivates PDGFRβ
To test the possibility that the potentiation of chemotaxis toward PDGF-BB by TF/FVIIa is attributable to a direct effect of TF/FVIIa on PDGFRβ, the PAE/PDGFRβ,TF cells were serum starved overnight and incubated with the suggested physiological concentration of FVIIa (10 nM). PDGFRβ was immunoprecipitated from cell lysates followed by Western blotting with the pan-phosphotyrosine antibody. In repeated experiments both a time and concentration-dependent increase in PDGFRβ phosphorylation was detected (P ranging from 0.003 to 0.04, and from 0.032 to 0.036; Figure 4A and 4B, respectively) indicating that the engagement of TF by FVIIa triggers activation of PDGFRβ. The phosphorylation of PDGFRβ induced by low concentrations of PDGF-BB (0.1 ng/mL) was enhanced by 10 nmol/L FVIIa (data not shown) thus correlating with the results of cell migration assays performed under similar conditions (Figure 1F). Reduction of this phosphorylation by a Src family inhibitor SU6656 suggested a role of the TF/FVIIa-activated Src kinase(s) in this process (Figure 4B). To exclude the possibility that the phosphorylation of PDGFRβ was induced by a nonspecific binding of FVIIa to the cell surface protein other than TF, the PAE cells transfected with PDGFRβ (but not with TF), were incubated with 100 nmol/L of FVIIa, but no change in PDGFRβ phosphorylation was detected (data not shown). Increasing concentrations of SLIGKV led to phosphorylation of PDGFRβ that was significantly (P=0.012) reduced by SU6656 (Figure 4C) supporting the possibility that at least one of the ways to activate PDGFRβ by TF/FVIIa may be mediated by PAR-2. Thrombin binding to PAR-1 has been shown to induce secretion of PDGF and activation PDGFR.32 To test whether similar mechanism was used by FVIIa/PAR-2 we analyzed PDGF-AB and PDGF-BB in conditioned media of the FVIIa-treated cells. As no secreted PDGF was detected this mechanism of PDGFRβ activation could be excluded (data not shown). These results show for the first time that the TF/FVIIa complex transactivates PDGFRβ.
|
TF/FVIIa Induces Phosphorylation of a Subset of PDGFRβ Cytoplasmic Tyrosines
To get an insight into the signaling pathways potentially activated by TF/FVIIa through PDGFRβ transactivation we analyzed the phosphorylation of individual PDGFRβ tyrosines, which in their phosphorylated state are known to regulate major cellular functions by binding the SH-2 domain-containing enzymes and adapter proteins.33 Of the 8 PDGFRβ autophosphorylation sites analyzed only 4 (tyrosines 579, 771, 1009, and 1021) were phosphorylated on the TF/FVIIa complex formation (Figure 5 and Table). Four other sites, Y716, 740, 751, and 857, were phosphorylated on PDGF-BB stimulation but no significant increase of the signal in FVIIa-stimulated cells was detected (Figure 5).
|
|
Discussion
We have previously shown that TF/FVIIa complex potentiates PDGF-BB–induced chemotaxis in fibroblasts and vascular smooth muscle cells3 and endothelial cells.12 In this study, using the freshly isolated LPS-stimulated human monocytes together with fibroblasts and endothelial cells, we studied the molecular mechanism by which TF/FVIIa influences PDGF-BB–induced cell migration. Whereas monocyte migration induced by 5 well-characterized chemotactic factors was unaffected by TF/FVIIa, PDGF-BB–induced migration of all 3 cell types was strongly enhanced by TF/FVIIa indicating a cross-talk between TF/FVIIa and PDGFRβ signaling. Here we show that TF, after ligation of FVIIa, induces the PAR-2 and Src family-dependent phosphorylation of PDGFRβ. We thus report, for the first time, the phenomenon of transactivation of PDGFRβ by TF/FVIIa and suggest a role for this event in regulation of cell migration.
Incubation of cells with FVIIa-induced PDGFRβ tyrosine phosphorylation in time and concentration-dependent manner (Figure 4). Involvement of PAR-2 and the TF/FVIIa-activated Src kinases in this reaction was supported by a similar induction of pY PDGFRβ by SLIGKV and the reduction of the effect of both FVIIa and SLIGKV by a Src family inhibitor SU6656. Closer mapping of specific tyrosines in PDGFRβ with the site-specific phosphoantibodies revealed a distinct TF/FVIIa-induced pattern of phosphorylation compared with that induced by PDGF-BB (Figure 5). Phosphorylation of the Src binding site (Y579) supports the role of Src kinases in the transactivation. The results summarized in Table 1 indicate that the TF/FVIIa-induced PDGFRβ transactivation is likely to amplify some of the TF/FVIIa-induced signals in parallel with those triggered by TF/FVIIa via separate mechanisms.
Transactivation of receptor tyrosine kinases (RTK) by activated GPCR has been recognized as one of the means to mediate a variety of cellular responses34 (and references therein). E.g. PAR-1–induced signals have been shown to trigger activation of different growth factor receptors in this manner,21,35,36 and the involvement of EGFR transactivation in PAR-2–dependent signaling has been described.37–39 One of the possible ways of PDGFRβ transactivation in our system could be stimulation of PDGF secretion by the TF/FVIIa/PAR-2 complex and the subsequent activation of PDGFRβ by its native ligand. However, we could not detect PDGF-AB or BB in conditioned media of the FVIIa-treated cells, which excludes such possibility. Involvement of the recently identified selective PDGFRβ ligand PDGF-DD40 is unlikely as it would trigger phosphorylation of all tyrosines in the receptor similarly to PDGF-BB. Although our data support the dependence of TF/FVIIa-induced transactivation of PDGFRβ, and hyperchemotaxis, on the activation of PAR-2 (see Figures 1F, 2A, and 4![]()
C), parallel PAR-2–independent pathway(s) may exist that are regulated by TF cytoplasmic domain or the lateral interactions with other cellular receptors11 (and our unpublished results).
In Figure 2 we have used the published optimal concentrations of the inhibitors; eg, in the case of SU6656 the selectivity for Src (at the 2 µmol/L concentration) relative to a number of other kinases was shown to be over 6 times higher.41 Under these conditions the used inhibitors that were found not to influence the basal chemotaxis induced by PDGF-BB (supplemental Figures I and II) strongly inhibited the TF/FVIIa-induced enhanced cell migration (Figure 2B). This finding is supported by a previous report on stimulation of Src-kinases by TF16 and suggests a similar functional role for these enzymes in all 3 cell types used. Our results further show that the c-Src and c-Yes kinases are selectively activated by TF/FVIIa and participate in transactivation of PDGFRβ. Other specific functions for these Src family enzymes in a given context remain to be identified.
Recently, a role for c-Src–mediated activating phosphorylation of PLC
1 on Y783 for cell migration was demonstrated.31 We have shown that TF/FVIIa complex triggered PLC activity,3 and U73122 eliminated the enhanced cell migration in our assays (Figure 2). However, neither PLC
1 nor PLCβ3 that both are targets of U73122, were activated by TF/FVIIa (data not shown). c-Src and c-Yes are therefore likely to act via other U73122-sensitive PLC isoform(s), or function in parallel with a putative Src-family–independent PLC subtype. Pertussis toxin, which perturbs the signaling through G
0-type nucleotide-binding proteins42 that are engaged in PAR-2–activated pathways,43 had no effect on cell migration (Figure 2B). Our results rather implicate the pertussis toxin-insensitive G
q subfamily members, which also connect to PAR-2, in TF/FVIIa signaling.
TF/FVIIa signaling in monocyte-derived macrophages has been described earlier,44,45 but our study is so far the first one to use freshly isolated human monocytes in this context. A recent paper by Muth et al45 reported a consistent upregulation of IL-8 gene in response to FVIIa in activated monocyte-derived macrophages, and migration of breast cancer cells with constitutive TF-expression toward FVIIa was entirely dependent on TF/FVIIa-induced IL-8 production.7 In our experimental system a weak effect of IL-8 on chemotaxis of the short-term cultured human monocytes was not altered by addition of FVIIa, which rules out the possibility that the FVIIa-dependent potentiation of chemotaxis is regulated by IL-8 via an autocrine loop.
Our findings with 3 principally different cell types indicate that the TF/FVIIa-induced elevated migratory response to PDGF-BB could be of broader functional importance. However, in a different cellular context FVIIa does not have an additive effect on PDGF-BB–stimulated events that involve cell migration.4 Nevertheless, a synergy between TF and PDGFRβ has been suggested in support of our results.46
In conclusion, we report the functional cooperation between TF/FVIIa and PDGF β-receptor in chemotaxis of 3 principally different cell types. This cross talk involves PAR-2, Src-family kinases c-Src and c-Yes, and isoform(s) of phospholipase C other than
1 or β3. We show that TF/FVIIa induces the phosphorylation of a defined subset of PDGFRβ cytoplasmic tyrosines and thus report, for the first time, the phenomenon of transactivation of PDGFRβ by TF/FVIIa that is likely to play a role in the regulation of PDGF-BB–induced cell migration. Characterization of the signaling intermediates involved in this process, the activated signaling pathways, and their role in TF/FVIIa-dependent biological functions are a subject of ongoing studies.
Acknowledgments
The authors thank Dr Wolfram Ruf for the gift of the anti–PAR-2 antibody and acknowledge Birgitta Fahlström for skillful technical assistance.
Sources of Funding
This work was supported by the Swedish Research Council, the Swedish Heart and Lung Foundation, and King Gustav V and Queen Victoria Foundation.
Disclosures
None.
Footnotes
Original received July 2, 2007; final version accepted October 28, 2007.
References
This article has been cited by other articles:
![]() |
S.K. Payeli, R. Latini, C. Gebhard, A. Patrignani, U. Wagner, T.F. Luscher, and F.C. Tanner Prothrombotic Gene Expression Profile in Vascular Smooth Muscle Cells of Human Saphenous Vein, but Not Internal Mammary Artery Arterioscler. Thromb. Vasc. Biol., April 1, 2008; 28(4): 705 - 710. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2008 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |