Cytochrome P450 epoxygenases 2C8 and 2C9 are implicated in hypoxia-induced endothelial cell migration and angiogenesis

Cytochrome P450 (CYP) epoxygenases of the 2B, 2C and 2J subfamilies are expressed in endothelial cells and metabolize arachidonic acid to four regio-isomeric epoxides (5,6-, 8,9-, 11,12- and 14,15-epoxyeicosatrienoic acid or EETs). CYP 2J enzymes seem to be constitutively expressed, whereas the expression of CYP 2C isoforms can be induced by a number of different pharmacological and mechanical stimuli (for a review, see Fleming, 2001). Interest in the vascular actions of EETs was initially linked to their identification as endothelium-derived hyperpolarizing factors (EDHFs) (Campbell et al., 1996; Fisslthaler et al., 1999), but it is now generally appreciated that these compounds mediate a number of membrane-potential-independent effects. The consequences of an increased intracellular production of EETs range from changes in the intracellular concentration of Ca²⁺ (Graier et al., 1995), and the activation of protein kinase A (Imig et al., 1999) and tyrosine kinases and phosphatases (Hoebel and Graier, 1998; Fleming et al., 2001), to the activation of MAP kinase phosphatases and inhibition of the c-Jun N-terminal kinase (Potente et al., 2002). Although some of these responses may result from the direct interaction of an EET with the target molecule; for example the binding of EETs to the TrpV4 channel may account for effects on Ca²⁺ signaling (Watanabe et al., 2003), some of the consequences of increased EET production can be attributed to the transactivation of the epidermal growth factor (EGF) receptor (Chen et al., 2002; Michaelis et al., 2003). Indeed, the release of heparin-binding EGF-like growth factor from the endothelial cell surface and the subsequent activation of EGF-receptor-dependent signaling have been suggested to underlie the proliferative and angiogenic effects of CYP-derived EETs (Michaelis et al., 2003).

Of the CYP 2C epoxygenase family, CYP 2C8 and 2C9 are expressed in native human endothelial cells (Lin et al., 1996; Hillig et al., 2003). However, the expression of these enzymes (mRNA and protein) decreases rapidly following cell isolation, which means that to assess the consequences of CYP 2C activation in cultured cells, it is necessary to either induce CYP expression pharmacologically or to transfect the cells with an expression vector encoding the isoform of interest. To achieve high transfection efficiencies, adenoviral vectors are generally used but although this approach facilitates the elucidation of EET-dependent signaling, it does not accurately reflect a physiological/pathophysiological response. The aim of the present investigation was to determine the effects of a potent angiogenic stimulus (hypoxia) on the expression of CYP 2C protein in endothelial cells and to further elucidate the mechanisms by which CYP 2C-derived EETs promote the degradation of the extracellular matrix and endothelial cell migration: two essential steps in the process of angiogenesis.

Matrigel was from BD Biosciences, 11,12-EET was purchased from Cayman Chemicals (Massy, France), KT 5720 and AG 1478 were from Calbiochem (Darmstadt, Germany) and thrombin was from Haemochrom Diagnostica GmbH (Essen, Germany). MS-PPOH and 14,15-epoxyeicosa-5(Z)-enoic acid (14,15-EEZE) were synthesized as described (Gauthier et al., 2002). Sulfaphenazole, myelin basic protein, the antibody recognizing β-actin and all other chemicals were from Sigma.

Human umbilical vein endothelial cells (HUVEC) were either isolated as described (Busse and Lamontagne, 1991) or purchased from Cell Systems/Clonetics (Solingen, Germany). Cells were cultured in MCDB 131 (Gibco Life Technology, Karlsruhe, Germany), supplemented with 8% fetal calf serum (FCS), L-glutamine (10 mmol/l), basic fibroblast growth factor (1 ng/ml), epidermal growth factor (0.1 ng/ml), endothelial cell growth-stimulating factor from bovine brain (ECGS/H, 0.4%), penicillin (50 U/ml) and streptomycin (50 μg/ml). First or second passage endothelial cells were used throughout. Porcine coronary artery endothelial cells were isolated and cultured as described previously (Popp et al., 1996). Where appropriate, hypoxic conditions were achieved by incubating cells in an air-tight Heraeus incubator (Hanau, Germany) with 5% CO₂ and 1% O₂ balanced with N₂.

Endothelial cells (80-90% confluent) were serum-starved for 24 hours prior to infection with adenoviral vectors. Cells were incubated with recombinant adenoviruses (10 pfU/cell) expressing CYP 2C9 sense or antisense cDNA in medium without antibiotics for 4 hours at 37°C followed by recovery in the presence of 2% FCS. The infection efficiency was between 90 and 100% and, as reported previously (Michaelis et al., 2005), endothelial cells infected with the CYP 2C9 sense adenovirus generated approximately twofold more 11,12- and 14,15-EET under basal conditions than cells infected with the control (CYP 2C9 antisense) virus and the increase in EET production was sensitive to the CYP 2C9 inhibitor, sulfaphenazole.

In some experiments, an antisense oligonucleotide approach was used to prevent the hypoxia-induced expression of CYP 2C8/9; HUVEC were treated with the oligonucleotides (2 μmol/l) using GeneTrans II according to the manufacturer's protocol (MobiTec, Göttingen, Germany) with CYP 2C sense and antisense oligonucleotides (antisense, 5′-TCC ATT GAA GCC TTC TCT TCT T-3′; sense, 5′-AAG AAG AGA AGG CTT CAA TGG A-3′, in both cases the three 5′ nucleotides were modified with phosphothioate; MWG-Biotech, Ebersberg, Germany). The sequence of these oligonucleotides spans the ATG and is 100% identical with CYP 2C8 and contains one mismatch to the other three human CYP 2C isoforms. As a consequence of the high homology between the 2C enzymes the oligonucleotides used were not able to differentiate between isoforms.

Following exposure to hypoxia for 16 hours, endothelial cells were cultured for a further 2 hours in normoxic conditions in the absence or presence of MS-PPOH. Thereafter, the cells were harvested by scraping and the cell pellets (approximately 8×10⁶ cells) were suspended in 100 μl of 0.2 M KCl and 100 μl methanol. The cells were hydrolyzed for 2 hours using NaOH (0.5 N), then neutralized with HCl (2M) and deuterated internal standards (5-HETE-d₈, 12-HETE-d₈, 15-HETE-d₈, 20-HETE-d₆, 8,9-EET-d₈, 11,12-EET-d₈ and 14,15-EET-d₈) were added. Liquid-liquid extraction was performed twice using 1 ml hexane/ethyl acetate (1:1). After evaporation of the solvent under a gentle stream of nitrogen in a vacuum block, residues were reconstituted with 50 μl methanol/water (1:1, v/v) and determined with a Sciex API4000 mass spectrometer operating in multiple reaction monitoring (MRM) mode. Chromatographic separation was performed on a Gemini C18 column (150×2 mm I.D., 5 μm particle size; Phenomenex, Aschaffenburg, Germany).

Endothelial cells were transiently co-transfected with the non-coding 5′ region (-2088 to +21, kindly provided by P. Maurel, Montpellier, France) of CYP 2C9 subcloned into pGL3basic (Promega) and pcDNA3.1myc-His/LacZ (Invitrogen, Karlsruhe, Germany). After 2 hours, the cells were exposed to either normoxia or hypoxia for 60 minutes. Thereafter, the cells were lysed, and luciferase and galactosidase activities were assayed according to the manufacturer's protocols (Promega, Mannheim, Germany and Applied Biosystems, Darmstadt, Germany). Promoter activity was determined as luciferase activity relative to that of galactosidase. Preliminary experiments using an empty vector demonstrated that luciferase production from the vector construct was not influenced by hypoxia per se.

Total RNA was isolated from cultured endothelial cells using phenol and guanidine isothiocyanate (Tri®Reagenz, Sigma, Germany). Random hexanucleotide primers were used for reverse transcription of equal amounts of RNA. The cDNA was used for real-time PCR using Taqman probes for the detection of the specific amplification products. The oligonucleotides used were derived from the human CYP 2C8 sequence (2C8 forward, 5′-GGACTTTATCGATTGCTTCCTG-3′; reverse, 5′-CCATATCTCAGAGTGGTGCTTG-3′; FAM and dabcyl-labeled Taqman probe, 5′-TTGGCACTGTAGCTGATCTATTTGTTGCTGGA-3′). To ensure equal amounts of cDNA were used, the 18S RNA was amplified (Assay on Demand, Applied Biosystems) and the amount of cDNA in the samples was calculated on the basis of the amplification of a serial dilution of a plasmid (CYP 2C8) or the serial dilution of the cDNA (18S RNA). The CYP 2C8 levels were normalized to that of 18S. At least two RT reactions were performed using each RNA preparation and at least two PCR reactions were performed with each cDNA sample.

Cells were treated as described in the Results section, fixed in phosphate-buffered formaldehyde solution (4%), permeabilized using Triton X-100 (0.2%) and blocked with bovine serum albumin solution (3%) and horse serum (5%) in phosphate-buffered saline. CYP 2C was detected using a specific polyclonal CYP 2C antibody (kindly provided by E. Morgan, Atlanta, GA) and HIF 1α using a monoclonal antibody (BD Transduction Laboratories). The cells were then washed and incubated with fluorescent secondary antibodies (Alexa), mounted and images were acquired by laser-scanning microscopy (LSM 510 meta, Carl Zeiss, Jena, Germany).

Endothelial cell migration was investigated using a modified Transwell chamber system. After pre-treatment of the cells as described in the Results section, endothelial cells were counted and seeded on membrane inserts (FluoroBlok, 3 μm pore size, BD Bioscience, Heidelberg, Germany) in the presence of MCDB 131 supplemented with 0.1% BSA. The lower chamber contained MCDB 131 supplemented with 4% FCS, L-glutamine (10 mmol/l), basic fibroblast growth factor (0.5 ng/ml), epidermal growth factor (0.05 ng/ml) and endothelial cell growth-stimulating factor from bovine brain (ECGS/H, 0.2%). After 20 hours, the cells on the upper surface of the filter were removed mechanically and cells that had migrated into the lower compartment were fixed (4% paraformaldehyde in PBS), stained with DAPI and counted (10 images per well, ×30 magnification). Invasion assays were performed in the same manner but using Matrigel-coated membrane inserts with a pore size of 8 μm (BD Bioscience, Heidelberg, Germany).

Endothelial cells were treated as described in the Results section and Rac activity was determined using a Rac-Pak pull-down assay according to the manufacturer's instructions (Upstate Biotechnology, Lake Placid, NY).

Endothelial cells were infected with CYP 2C9 sense or antisense adenoviruses and cultured with or without sulfaphenazole (30 μmol/l) for 24 hours. Crude membrane fractions were prepared by repeated cycles of freezing and thawing and MT-MMP activity assessed using a commercially available kit (Chemicon International, Temecula, CA).

Zymography was performed as described (Bouloumié et al., 2001) using myelin basic protein (MBP) polymerized in an SDS gel as the substrate. Cell supernatants were separated by SDS-PAGE, and following renaturation of the proteins with 2.5% Triton-X100, the gels were incubated for 22 hours in a buffer containing Tris-HCl (50 mmol/l), CaCl₂ (5 mmol/l) and NaN₃ (0.02%). Gels were then stained with Coomassie Blue and MBP degradation assessed densitometrically.

Fibrin gels were prepared using thrombin (0.5 U/μl)-polymerized fibrinogen (1.5 mg/ml in MCDB 131). Endothelial cells were infected with CYP 2C9 sense or antisense adenoviruses 24 hours after seeding onto the fibrin gels and cultured as described above, in the absence or presence of 14,15-EEZE (10 μmol/l). Total tube length was determined after 14 days.

To investigate the effect of hypoxia-induced CYP 2C expression on tube formation, cells were cultured in normoxic or hypoxic conditions for 24 hours in the absence and presence of 14,15-EEZE (10 μmol/l). Thereafter, the cells were seeded onto Matrigel (3×10⁴ cells/cm²) and cultured under normoxic or hypoxic conditions, in the absence and presence of 14,15-EEZE. Tube formation was assessed after 14 hours and quantified by counting the number of branch points.

All experiments with chick embryos were performed in ovo. A window (7-10 mm in diameter) was cut into the eggshell of 3-day-old embryos, resealed with transparent film, and incubated for a further 6 days. For the chorioallantoic membrane (CAM) assay, solvent or substances described in the Results section were mixed with a 1% methylcellulose solution. Aliquots (10 μl) of the resulting 0.5% methylcellulose solution were applied to bacteriological-grade Petri dishes, air dried for 1 hour and then placed onto the 9-day-old CAM. Two discs were placed on each CAM ∼10-20 mm apart. Eggs were then incubated under normoxic or hypoxic (14% O₂) conditions for 3 days. To better visualize the vascular system of the CAM, Intralipid was injected underneath the membrane and 20% Luconyl Black (BASF, Ludwigshafen, Germany) in PBS was injected into a vitelline vein. Photographs were taken using a Nikon SMZ1000 stereomicroscope. The angiogenic response was quantified as described (Brunner et al., 2000).

Data are expressed as the mean±s.e.m. Statistical analyses were performed by one-way ANOVA followed by Bonferroni's multiple comparison test. Values of P<0.05 were considered statistically significant.

To assess the effects of hypoxia on CYP 2C8/9 expression we measured the activity of a reporter construct as well as levels of endogenous CYP 2C RNA and protein. Exposure of HUVEC, transfected with a CYP 2C9 promoter-luciferase construct, to hypoxia (1% O₂) for 60 minutes elicited a significant increase (approximately fivefold) in the activity of the reporter gene construct (Fig. 1A).

In non-transfected cells, the endogenous expression of CYP 2C8 RNA (Fig. 1B) and CYP 2C protein (Fig. 1C) was time-dependently increased following exposure to hypoxia. The increase in CYP 2C RNA was rapid and reached an elevated steady-state level after approximately 4 hours. Increased protein levels were evident after approximately 8 hours and remained elevated over the entire observation period (up to 24 hours). Hypoxia was also associated with a marked increase in the expression of HIF 1α. The increase in CYP 2C expression was associated with an elevation in EET production. The hypoxia (16 hours)-induced increase in intracellular levels of 11,12-EET and 11,12-dihydroxyeicosatrienoic acid (11,12-DHET) was inhibited in the presence of the CYP epoxygenase inhibitor MS-PPOH (Fig. 1D). No effect of hypoxia on the formation of the other EET regioisomers (5,6-, 8,9- or 14,15-EET) was detected (data not shown).

To determine the consequences of an increase in endogenous CYP 2C8/9 expression and activity on endothelial cell migration, experiments were performed using endothelial cells pre-exposed to hypoxia (24 hours) to increase CYP 2C expression as well as with CYP 2C antisense oligonucleotides. The hypoxia-induced expression of CYP 2C was unaffected by sense oligonucleotides but was significantly attenuated in cells treated with antisense oligonucleotides (Fig. 2A). As reported previously (Fisslthaler et al., 1999), scrambled oligonucleotides had no effect on CYP expression (data not shown). Endothelial cell migration under normoxic conditions tended to be slightly enhanced following incubation with the antisense oligonucleotides although this effect did not achieve statistical significance. However, treatment with the antisense oligonucleotides significantly attenuated endothelial cell migration in cells pre-exposed to hypoxia (Fig. 2B). Corresponding to a role for a CYP-dependent potentiation of endothelial cell migration under hypoxia, the migration of cells maintained under normoxic conditions was not significantly affected by either the epoxygenase inhibitor MS-PPOH or the EET antagonist 14,15-EEZE whereas the migration of endothelial cells pretreated with hypoxia was significantly attenuated by both substances (Fig. 2C).

To exclude additional effects of hypoxia on endothelial cells and to elucidate the specific role of CYP 2C8/9 on endothelial cell migration, extracellular matrix degradation and angiogenesis, we used endothelial cells infected with an adenovirus coding for CYP 2C9. Although CYP 2C8 was endogenously expressed in human endothelial cells we chose to overexpress CYP 2C9 (the EET profile generated by CYP 2C8 and 2C9 are similar), as its activity can be inhibited by sulfaphenazole (Mancy et al., 1996). Human endothelial cells were infected with adenoviruses expressing CYP 2C9 mRNA in the sense or antisense orientation and seeded on Transwell filters. More of the CYP-2C9-overexpressing cells had migrated to the lower portion of the filter after 20 hours than cells treated with the control (CYP 2C9 antisense) adenovirus. The CYP-2C9-induced increase in endothelial cell migration was inhibited by sulfaphenazole (Fig. 3A,B).

As cell migration is generally accompanied by an increase in Rac activity, we investigated the effect of CYP 2C9 on Rac using a Rac-PAK pull-down assay. Overexpression of CYP 2C9 markedly increased the association of Rac with PAK-1, an effect that was sensitive to treatment with sulfaphenazole (Fig. 4A). The exogenous application of the CYP 2C8/9 product 11,12-EET, increased Rac activity to a similar extent as vascular endothelial growth factor (VEGF) (Fig. 4B).

We next determined the effect of CYP 2C9 on MMP activity using a commercially available MT-MMP assay. MMP activity was markedly increased in membrane fractions isolated from CYP-2C9-overexpressing endothelial cells and was inhibited by sulfaphenazole (Fig. 5A). MMP activity was also assessed by zymography using MBP as a substrate. Enhanced degradation of MBP by a protein with the apparent molecular mass of ∼150 kDa was readily apparent in the lanes containing the supernatant from CYP-2C9-overexpressing cells (Fig. 5B). The time-dependent increase in MMP activity was also evident in cells treated with 11,12-EET (1 μmol/l) for 4-8 hours (Fig. 5C).

To determine the consequence of CYP 2C9 overexpression on endothelial cell invasion, HUVEC were seeded on Matrigel-coated Transwell filters and migration was monitored after 24 hours. Significantly more CYP-2C9-overexpressing cells were detected in the lower chamber than cells treated with the control virus. This CYP-2C9-dependent induction of cell invasion was sensitive to both sulfaphenazole, and the EET-antagonist 14,15-EEZE (Fig. 6A). The non-selective MMP inhibitor GM 6001 completely abolished CYP-2C9-induced cell invasion (Fig. 6B).

To determine the effect of 14,15-EEZE on CYP-2C9-induced angiogenesis, we monitored tube formation in a fibrin gel. After 14 days, the formation of capillary-like structures was apparent in cultures containing CYP-2C9-overexpressing cells, but not in those cells infected with the control (antisense) virus or CYP-2C9-overexpressing endothelial cells cultured in the presence of 14,15-EEZE (Fig. 6C).

To further assess the role of CYP epoxygenases in hypoxia-induced angiogenesis, we determined the effects of 14,15-EEZE on endothelial cell tube formation under normoxic and hypoxic conditions. Human endothelial cells were cultured in the absence or presence of 14,15-EEZE (24 hours) before being seeded on Matrigel and cultured for an additional 14 hours either under normoxic or hypoxic conditions. Although tube-like structures were evident after 14 hours in CYP-expressing endothelial cells treated with solvent (0.01% DMSO), 14,15-EEZE completely prevented tube formation under hypoxic conditions. The EET antagonist, which did not promote endothelial cell apoptosis or compromise cell viability (data not shown), did not influence tube formation in cells maintained under normoxic conditions (Fig. 7).

To demonstrate that the hypoxia-induced increase in CYP 2C expression and CYP-2C-dependent formation of endothelial cell tubes was not a phenomenon restricted to human umbilical vein endothelial cells, additional experiments were performed using porcine coronary artery endothelial cells. Hypoxia also induced a time-dependent increase in CYP 2C protein in porcine endothelial cells (Fig. 8A), which was associated with an increase in endothelial cell migration through Transwell filters (Fig. 8B). The latter effect was attributed to an increase in CYP activity as the effect was prevented in the presence of 14,15-EEZE and MS-PPOH. Hypoxia also potentiated the formation of tube-like structures by porcine endothelial cells and again this effect was abrogated by the EET antagonist and the epoxygenase inhibitor (Fig. 8C).

The CAM assay was used to assess the effect of 14,15-EEZE on hypoxia-induced angiogenesis in vivo. CAMs were treated with discs containing either solvent or 14,15-EEZE and incubated under normoxic or hypoxic (14% O₂) conditions. Hypoxia increased CAM vessel density compared with that recorded in eggs maintained in normoxia and was sensitive to the EET-antagonist 14,15-EEZE (Fig. 9).

The results of the present investigation demonstrate that in two different types of endothelial cells CYP 2C8/9 expression is upregulated by hypoxia and that the subsequent endothelial cell migration and tube formation are inhibited by an EET antagonist and a CYP 2C inhibitor as well as by downregulating CYP 2C8/9 expression using antisense oligonucleotides. The EET antagonist also abolished hypoxia-induced angiogenesis in the chick chorioallantoic membrane in vivo. As such, these data provide the first indication that CYP 2C8/9 is involved in the process of hypoxia-induced angiogenesis. Indeed, CYP-2C9-derived EETs induce endothelial cell migration, activate metalloproteases and initiate the degradation of the extracellular matrix.

CYP 2C8/9 epoxygenases are expressed in native endothelial cells but RNA and protein levels decrease rapidly following cell isolation. Although the expression of CYP 2C8/9 in cultured cells can be restored by different stimuli, the mechanisms involved in its regulation are not well understood (for a review, see Fleming, 2004). The expression of several CYP enzymes is modulated by changes in oxygen tension, for example, hypoxia downregulates CYP 2J2 (Marden et al., 2003) whereas transient cerebral ischemia elicits the upregulation of CYP 2C11 in rats (Alkayed et al., 2002). Furthermore, the exposure of rats to hypoxia for 48 hours induces the expression of a sulfaphenazole-sensitive CYP epoxygenase, most probably CYP 2C11 (Earley et al., 2003). As the promoter regions of several CYP 2C genes contain multiple HIF 1α sites, we analyzed the effect of hypoxia on CYP 2C8 and 2C9 expression in cultured endothelial cells. We found that hypoxia increased the activity of the CYP 2C9 promoter and enhanced the expression of CYP 2C8/9 mRNA and protein in human and porcine endothelial cells. Not only was CYP 2C8/9 expression increased by hypoxia but an EET antagonist prevented tube formation in endothelial cells maintained under hypoxic conditions, while leaving cells cultured in a normoxic environment, which express negligible amounts of CYP 2C8/9, completely unaffected.

CYP-derived EETs are now recognized as intracellular signal transduction molecules exerting different effects on a variety of cell types including the activation of kinase cascades as well as the modulation of gene expression (Fleming, 2001). In endothelial cells, CYP 2C9 overexpression and 11,12-EET application induce proliferation and angiogenesis (Michaelis et al., 2003) and an angiogenic effect of EETs derived from astrocytes has also been described (Munzenmaier and Harder, 2000; Zhang and Harder, 2002). As the process of angiogenesis involves endothelial cell migration and the degradation of the extracellular matrix in addition to cell proliferation, we assessed the effects of CYP 2C upregulation on these processes. Endothelial cell migration under control conditions (i.e. uninfected and cultured under normoxic conditions) was insensitive to interference with EET generator or effector pathways. However, the migration of hypoxia-treated cells was attenuated by an epoxygenase inhibitor and an EET antagonist. Moreover, treatment of these cells with CYP 2C antisense oligonucleotides to prevent the hypoxia-induced upregulation of CYP 2C8/9 expression also markedly attenuated endothelial cell migration, indicating that CYP 2C-derived EETs are able to modulate endothelial cell migration. To exclude other effects elicited by hypoxia, we repeated the same experiments using endothelial cells infected with an adenoviral CYP 2C9 expression vector and found that the overexpression of CYP 2C9 was associated with a marked increase in endothelial cell migration. Our results contrast with the effects of EETs on smooth muscle cell migration because 11,12-EET has been reported to inhibit the PDGF-induced migration of rat aortic smooth muscle cells (Sun et al., 2002). However as EETs as well as an inhibitor of the soluble epoxide hydrolase (which increases EET levels) effectively prevented the PDGF-stimulated proliferation and expression of cyclin D1 in human vascular smooth muscle cells (Davis et al., 2002), it seems that EETs exert opposing effects on endothelial cell and vascular smooth muscle cell proliferation and migration.

Endothelial cell migration is a complex process during which the cells need to dissolve old contacts with the subcellular matrix at the same time as forming new contacts at the leading edge of the cell. GTPases of the Rho family play an important role in the regulation of the restructuring of the actin cytoskeleton and in the process of migration (Ridley, 2001). The results of the present investigation show that CYP-2C9-induced endothelial cell migration is accompanied by the activation of Rac and its association with the downstream effector kinase PAK-1. The latter serine/threonine kinase has previously been associated with disassembly of focal adhesions and actin stress fibers (Frost et al., 1998). Other simultaneously activated processes are also implicated in endothelial cell migration. We previously reported that CYP-2C9-induced endothelial cell proliferation is accompanied by the attenuated expression of p27^{Kip 1} (Potente et al., 2003). However, in addition to its role in cell cycle regulation, this cyclin-dependent kinase inhibitor possesses cell-cycle-independent functions and inhibits cellular changes that normally occur during cell locomotion (e.g. lamellipodia formation and reorganization of actin filaments and focal adhesions) most probably by binding to RhoA, thereby inhibiting its activation (Diez-Juan and Andres, 2003; Besson et al., 2004). It is therefore possible that the downregulation of p27^{Kip 1} also contributes to the migratory effect of CYP 2C9 reported here.

To migrate, cells must also be able to degrade the extracellular matrix, a process that involves a number of different metalloproteases (Chang and Werb, 2001). In the present study, we have shown that metalloprotease activity is enhanced in CYP-2C9-overexpressing cells as well as in cells treated with the CYP 2C9 product, 11,12-EET. Moreover, CYP-2C9-overexpressing cells migrated through a Matrigel-coated filter faster than CYP-deficient cells. Although we did not identify the metalloprotease activated by 11,12-EET, it is probably a member of the ADAM (a disintegrin and metalloprotease) family, because different ADAMs have been reported to degrade the MBP used as a substrate in the zymography experiments (Moss et al., 1997) and elicit the release of HB-EGF from endothelial cells (Asakura et al., 2002; Hinkle et al., 2004), which is another consequence of enhanced CYP 2C9 activity (Michaelis et al., 2003).

Since the first proposal that CYP-derived metabolites of arachidonic acid may be linked to vascular smooth muscle hyperpolarisation and relaxation, it has become clear that the arachidonic acid epoxides are important intracellular signaling molecules that modulate much more than membrane potential. Indeed it is likely that the importance of EETs in vascular homeostasis has been largely underestimated because of the labile nature of the EET-forming enzymes in cell culture and the lack of appropriate experimental models and tools. This also means that the contribution of CYP-derived products in the vast majority of the experimental models based on cell culture systems to address topics related to vascular signaling/homeostasis and angiogenesis has been overlooked. In addition to the data obtained showing that CYP-derived EETs enhance endothelial cell migration and MMP activity in an overexpression system, our results showing that hypoxia increases CYP 2C8/9 expression in cultured endothelial cells and that hypoxia-induced endothelial tube formation in vitro and angiogenesis in vivo can be prevented by an EET antagonist highlight the importance of EETs as signaling molecules in the angiogenic process.

The authors are indebted to Isabel Winter for expert technical assistance. This work was partially supported by Philip Morris, the Deutsche Forschungsgemeinschaft (FI 830/2-2), the National Institute of Health (NIH GM31278, to J.R.F.), by a Young Investigators Grant (to U.R.M.) from the Medical Faculty of the Johann-Wolfgang-Goethe-Universität and the Marie Christine Held und Erika Hecker Stiftung.