Characterization of 5,6- and 8,9-Epoxyeicosatrienoic Acids (5,6- and 8,9-EET) as Potent in Vivo Angiogenic Lipids*

The cytochrome P450 arachidonic acid epoxygenase metabolites, the epoxyeicosatrienoic acids (EETs) are powerful, nonregioselective, stimulators of cell proliferation. In this study we compared the ability of the four EETs (5,6-, 8,9-, 11,12-, and 14,15-EETs) to regulate endothelial cell proliferation in vitro and angiogenesis in vivo and determined the molecular mechanism by which EETs control these events. Inhibition of the epoxygenase blocked serum-induced endothelial cell proliferation, and exogenously added EETs rescued cell proliferation from epoxygenase inhibition. Studies with selective ERK, p38 MAPK, or PI3K inhibitors revealed that whereas activation of p38 MAPK is required for the proliferative responses to 8,9- and 11,12-EET, activation of PI3K is necessary for the cell proliferation induced by 5,6- and 14,15-EET. Among the four EETs, only 5,6- and 8,9-EET consid- ered statistically significant. These results identify the release of arachidonic acid from stores, catalyzed by phospholipases A2 and epoxide hydrolase(s)-mediated hydration (31) as two key components of the cellular mechanisms controlling EET biosynthesis, disposition, intracellular concentrations, and, ulti-mately, biological activity 17). The biosynthesis of EETs by and freshly isolated endothelial cells, and the role of hydration in cellular EET metabolism have been reported

Interest in the biochemical mechanism of angiogenesis, the de novo formation of blood vessels from pre-existing vessels, stems from the critical roles played by this process in the pathophysiology of inflammation, cancer, and cardiovascular diseases. Neo-vascularization requires the coordinated contributions of endothelial cell proliferation and migration leading to the formation of nascent capillary structures and finally, to new functional vessels. The eicosanoids prostaglandin E 2 , prostacyclin (1, 2), HETEs (3)(4)(5)(6)(7), and the epoxyeicosatrienoic acids (EET) 1 (8 -11) have been characterized as pro-angiogenic molecules. In addition, the roles of cyclooxygenases 1 and 2, as well as cytochrome P450 (P450) 4A and 2C isoforms in these processes have been suggested (4 -11). However, the individual and/or combined contribution of these lipids to in vivo angiogenic responses remains to be determined.
The P450 arachidonic acid epoxygenases catalyze the metabolism of endogenous pools of arachidonic acid to 5,6-, 8,11,and 14,, a reaction catalyzed predominantly by members of the P450 CYP2C gene subfamily (12). Furthermore, CYP2C8, CYP2C23, and Cyp2c44 have been identified as predominant stereoselective epoxygenases in several human, rat, and mouse organ tissues (12,13,17,18). Since the original description of the mitogenic properties of 14,15-EET (12,19), several studies have characterized 11,12and 14,15-EET as powerful mitogens using cultured cells derived from kidney, brain, and endothelium (4,8 -12,18 -21), and a role of 14,15-EET in mediating the mitogenic responses of EGF and HB-EGF has been thoroughly documented in cultured LLCPk cells (20,21). More recently, 11,12-EET was identified as an angiogenic molecule, and in vitro roles for cyclooxygenase-2 and EGF receptor identified (10,11). Furthermore, 11,12-EET has also been shown to activate PI3K and tyrosine kinases (10,22). However, only a few studies have addressed whether activation of these downstream pathways are indeed required to mediate these EET-induced cell responses. Whereas the mitogenic activity of 11,12-and 14,15-EET, and the signaling pathways for 14,15-EET are well documented, less is known regarding the proliferative and/or angiogenic properties of 5,6-and 8,9-EET and their potential mechanisms of action. The biosynthesis of 8,9-EET from endogenous fatty acid pools is well documented (12), and that of the labile 5,6-EET has been inferred from the identification of its hydration product, 5,6-dihydroxyeicosatrienoic acid (5,6-DHET), in several biological samples (12,17). Furthermore, these EETs are known to circulate, associated with plasma lipoproteins, in rat and human blood (12).
We report here (i) the characterization of 5,6-and 8,9-EET as powerful mitogens, and selective chemotactic lipids for primary cultures of mouse pulmonary endothelial cells; (ii) the identification of the intracellular signaling pathways associated with their proliferative and chemotactic activities; and (iii) the demonstration of a role for these two EETs in promoting de novo angiogenesis in vivo. The observation that inhibition of enzymatic EET hydration potentiates the angiogenic activities of 5,6-and 8,9-EET suggests that inhibitors of epoxide hydrolase(s) could serve to promote angiogenesis in pathophysiological conditions in which de novo vascularization is compromised.

Synthesis of EETs by Endothelial
Cells-Pulmonary murine microvascular endothelial cells were isolated and cultured as previously described (23). Cells at early passages (1-3) were used for these studies because the expression of CYP2Cs in cultured endothelial cells decreases with increased cell passage (24). To study cellular EET synthase activity, semiconfluent endothelial cells (passage 3) were cultured in serum-free medium in the presence or absence of arachidonic acid (10 M). After 4 or 24 h, the cells and the medium were removed from the plates, and the suspension centrifuged to obtain a cell pellet and a medium-containing supernatant. The medium was mixed with 10 ng each of 1-14 C-labeled DHETs (55 Ci/mol; 95 atom % enrichment in 14 C) or 1-14 C-labeled EETs (55 Ci/mol) and extracted with ethyl acetate. The EETs and DHETs present in the culture medium were purified by reversed-phase HPLC (25), converted to the corresponding pentafluorobenzylesters (EETs) or pentafluorobenzylester-trimethylsilylether derivatives, and quantified by NICI/GC/MS exactly as described (26). The cell pellets were suspended in PBS, mixed with 10 ng each of 1-14 C-labeled EETs and DHETs (55 Ci/mol), and extracted with CHCl 3 /CH 3 OH (2:1). The organic extracts were submitted to alkaline hydrolysis, and EETs and DHETs purified and quantified by NICI/ GC/MS exactly as described (26). Prior to use, synthetic EETs and DHETs were purified by reversed-phase HPLC (25). The labile 5,6-EET was purified by normal phase HPLC on a Sorb silica column (250 ϫ 4.6 mm; 5), using hexane containing 0.5% HOAc and 2% 2-propyl alcohol as mobile phase at 2 ml/min.
In some experiments endothelial cells were plated as above and serum-starved for 24 h. The cells were then cultured in serum-free medium containing [ 3 H]thymidine (1 Ci/well) with or without 5,6-, 8,9-, 11,12-, and 14,15-EETs or their corresponding DHETs. Some cells were also treated with 10 M MS-PPOH, ketoconazole, or ACU added alone or in combination with each EET (1 M, each). Twenty-four hours later, cells were collected and proliferation determined as above.
To determine the pathways involved in the EET-induced cell proliferation, 24 h serum-starved cells were incubated in serum-free medium containing [ 3 H]thymidine (10 Ci/ml) with or without the individual EET regioisomers (1 M), and in the presence or absence of either a MEK1 inhibitor (PD98059) (10 M), a p38 MAPK inhibitor (10 M) (cat. number 506126) or a PI3K inhibitor (wortmannin) (0.1 M) (all from Calbiochem). After 24 h, the cells were collected, and [ 3 H]thymidine incorporation was determined as above. At least three independent experiments with quadruplicate samples were performed for each set of experiments described above. The [ 3 H]thymidine incorporation assay was also corroborated by manual cell counting (not shown).
Migration Assay-Cell migration was assayed in transwell plates fitted with 8-m membrane filters (Corning Ware). Lower wells were coated by an overnight incubation with collagen type I (10 g/ml) at 4°C and then incubated 1 h at 37°C with bovine serum albumin (1% in PBS) to inhibit nonspecific cell migration. Serum-free medium with or without 1 M EETs was then added to the lower wells and 24 h serum-starved endothelial cells (5 ϫ 10 4 cells in 300 l of serum-free medium containing 0.1% bovine serum albumin) to the upper wells. To determine the contribution of ERK, p38, and PI3K in EET-induced migration, serum-starved cells were allowed to migrate as indicated above with the difference that PD98059 (10 M), P38 MAPK inhibitor (10 M), or wortmannin (100 nM) were added to both upper and lower wells. After 12 h at 37°C, cells on the top of the filter were removed by wiping, and the filters were then fixed in 4% formaldehyde in PBS. Migrating cells were stained with 1% crystal violet, and five randomly chosen fields from triplicate wells were counted at ϫ400 magnification. Three independent experiments were performed in duplicate.
Matrigel-based Capillary Formation Assay-Capillary-like formation was analyzed as described (28). Briefly, 96-well plates were coated with 50 l of Matrigel and incubated 30 min at 37°C. Serum-starved endothelial cells (1 ϫ 10 4 ) were plated over solidified Matrigel in 200 l of serum-free medium with or without EETs (1 M). To determine the contribution of EET synthases in capillary-like formation, endothelial cells were plated on Matrigel in 200 l of complete medium in the presence or absence of different concentrations of MS-PPOH or ketoconazole. Capillary-like structures were recorded (3 images per gel per treatment) hourly for a period of 10 h, and representative images taken 3 h after plating are shown. To quantify capillary-like network formation, cellular nodes were defined as junctions linking at least three cells, and they were counted from digital images. Four independent experiments were performed with a total of 12 images analyzed per treatment.
Western Blot Analysis and Reverse Transcriptase PCR Analysis-To evaluate the effects of EETs on ERK, p38, and Akt phosphorylation, semiconfluent endothelial cells were serum-starved for 24 h and then treated with the EET (1 M each) for 0, 10, and 40 min. The cells were washed with PBS, collected, suspended in 50 mM HEPES, pH 7.5, 150 mM NaCl, 1% Triton X-100, and centrifuged for 10 min at 14,000 rpm. Cell lysates were resolved by SDS/PAGE (10% gels; 50 g of total protein/lane) and transferred to Immobilon-P membranes (Millipore, Billerica, MA). Membranes were incubated with a rabbit anti-phospho-ERK, anti-phospho-p38, or anti-phospho-Akt antibody (all from Cell Signaling Technology). Immunoreactive proteins were visualized using a peroxidase-conjugated goat anti-rabbit and an ECL kit (Pierce). Total ERK, p38, and Akt content were verified by stripping the membranes in 50 mM Tris-HCl, pH 6.5, containing 2% SDS and 0.4% ␤-mercaptoethanol for 1 h at 55°C, and re-probing with a rabbit anti-ERK, anti-p38, or anti-Akt antibody (Cell Signaling Technology). EET synthase expression in endothelial cell lysates (50 g of protein/lane) was analyzed as above using a rabbit anti-CYP2C23 or anti-CYP2C11 (14) antibody, cross-reactive against the murine Cyp2c44and Cyp2c38 EET synthase, respectively. Purified cyp2c44 or cyp2c38 (20 ng/lane) was used as positive controls.
In Vivo Angiogenesis-The subcutaneous sponge model was used to determine the effects of EETs on in vivo angiogenesis (29). Sterile polyvinyl acetal CF-50 round sponges (8 ϫ 3 mm, a gift from Dr. J. M. Davidson, Vanderbilt University) were implanted under the dorsal skin of 129 SvJ male mice (6 weeks of age, 20 g body weight, n ϭ 5/treatment). The sponges were then injected every second day for 14 days with 50 l of either vehicle (corn oil), EETs (50 M), ACU (250 M), or a mixture of EET and ACU (50 and 250 M, respectively). Ten minutes before sacrifice, mice were injected intravenously with 50 l of rhodamine-dextran (M r 65, 2% in PBS, Sigma) to label blood vessels (30), and the sponges were subsequently collected and analyzed under an epifluorescence microscope. Rhodamine-dextran-positive structures were imaged, the color images converted to black and white pictures using Photoshop (Adobe), and processed as described (23). Vascularity within sponges was expressed as a percentage of area occupied by rhodaminedextran-positive structures per microscopic field. Three images/sponge with a total of 15 images per treatment were used for analysis.
Statistical Analysis-The Student's t test for comparisons between two groups, and analysis of variance using Sigma-Stat software for statistical differences between multiple groups. p Յ 0.05 was considered statistically significant.

Microvascular Endothelial Cells Express CYP2C Isoforms
and Biosynthesize EETs-To establish the presence of a functional epoxygenase in mouse lung endothelial cells, we quantified cellular EETs and DHETs by the isotope ratio GC/MS method (26). Lung endothelial cells generate epoxygenase metabolites from endogenous precursors. Whereas EETs were found only in the cell pellet (ϳ0.37 ng/mg total proteins), their hydrated products, the DHETs, were mostly secreted into the culture medium (ϳ0.65 ng/mg). Incubation of the cells with exogenous arachidonic acid for 4 or 24 h significantly increased the yield of epoxygenase products. Interestingly, the amount of total cell epoxygenase products (EETs ϩ DHETs) was higher at 4 h than at 24 h (36 versus 24 ng/mg, respectively), suggesting that depletion of the precursor arachidonic acid and further metabolism (i.e. ␤-oxidization) could account for these changes (17). Furthermore, within the first 4 h, ϳ80% of the EETs (ϳ3.6 ng/mg), and Ͼ95% of the DHETs (ϳ30 ng/mg) formed were secreted into the medium. In contrast, after 24 h the EETs were found exclusively in the cells (ϳ1.3 ng/mg), whereas most of the DHETs formed (95% of the total; ϳ20 ng/mg) were secreted into the culture medium.
These results identify the release of arachidonic acid from glycerophospholipid stores, catalyzed by phospholipases A2 and epoxide hydrolase(s)-mediated hydration (31) as two key components of the cellular mechanisms controlling EET biosynthesis, disposition, intracellular concentrations, and, ultimately, biological activity (12,17). The biosynthesis of EETs by cultured and freshly isolated endothelial cells, and the role of hydration in cellular EET metabolism have been reported (17).
Several members of the CYP2C P450 gene subfamily have been identified as the predominant arachidonic acid epoxygenases in several organ tissues (12), and CY2C23, 2C11, 2C8, and 2C9 are characterized as renal and vascular epoxygenases in rat and human tissues (12-14, 17, 18, 32, 33). The murine homologues of rat CYP2C23 and 2C11, Cyp2c44 and 2c38, are cloned, and the recombinant proteins were shown to catalyze arachidonic acid epoxidation (15,16). Amplification by reverse transcriptase-PCR of pulmonary endothelial cell RNAs using Cyp2c38-and 2c44-specific PCR primers, showed the presence of Cyp2c44 transcripts, but the absence of detectable Cp2c38 mRNAs (Fig. 1A). Western blots using anti-CYP2C23 or -2C11 antibodies, cross-reactive toward Cyp2c44 and 2c38, respectively, demonstrated the presence of Cyp2c44 (Fig. 1B), but not Cyp2c38 (not shown), in cell lysates. We conclude that Cyp2c44 is the predominant Cyp2c mouse EET synthase present in the cultured lung endothelial cells, and the enzyme responsible for most of cellular EET biosynthesis.
EETs Are Mitogenic for Microvascular Endothelial Cells-To determine EET mitogenic activity, we incubated primary cultures of pulmonary murine endothelial cells with different amount of each EET regioisomer, and estimated their effects on cell proliferation by measurements of [ 3 H]thymidine incorporation. All four EETs induced significant increases in cell proliferation whether the cells were incubated with the EETs in the absence ( Fig. 2A) or presence (Fig. 2B) of serum. In the absence of serum, a significant increased proliferation over control cells (cells incubated with vehicle, ethanol, only) was observed when EETs were used at 0.5 M. Maximum induction was observed when cells were incubated with 1 M EETs, and this was independent from the cell culture conditions. The most effective regioisomer, 5,6-EET, induced 2-or 2.5-fold increases in proliferation in cells cultured in the presence or absence of serum, respectively (Fig. 2, A and B).
Because the DHETs are formed and secreted by endothelial cells (see above), and because they have been reported to modulate endothelial cell functions (34), we determined their ability to stimulate endothelial cells proliferation. As shown in Fig.  2, none of the DHETs stimulated endothelial cell proliferation independently of their dose or culture conditions (Fig. 2, C  and D).
To determine the contribution of the endogenous EETs to endothelial cell proliferation, cells were cultured in complete medium containing different concentrations of the EET synthase inhibitors MS-PPOH or ketoconazole. Compared with controls (cells incubated with ethanol only), the epoxygenase inhibitors decreased endothelial cell proliferation in a dose-dependent manner (Fig. 3A). Maximal inhibition was obtained at 20 M MS-PPOH or ketoconazole, concentrations shown to block by nearly 90% microsomal epoxygenase activity (12,27). To test the capacity of exogenously added EETs to rescue cell proliferation from EET synthase inhibition, endothelial cells were treated with MS-PPOH or ketoconazole in the presence or the absence of each EET (1 M). The EETs were here used at 1 M because, within a range of 0 to 5 M, this was the concentration inducing maximal proliferation (Fig. 2). As shown in Fig. 3B, both inhibitors (10 M each) significantly decreased endothelial cell proliferation, and co-incubation with EETs rescued their proliferation to levels comparable or slightly higher to those observed in cells treated with vehicle (ethanol) only.
The EETs are enzymatically hydrated to DHETs by epoxide hydrolases, particularly, cytosolic epoxide hydrolase (31). As mentioned above, hydration does contribute to cellular EET disposition, and the administration of inhibitors of this enzyme to rats has been associated with reduction in the urinary excretion of DHETs (35). To determine whether the inhibition of cytosolic epoxide hydrolase alters cell proliferation, endothelial cells were cultured as indicated above in the absence or presence of different concentrations of the hydrolase inhibitor ACU (17). As shown in Fig. 3A, ACU increased cell proliferation in a dose-dependent manner, with a maximal stimulation obtained between 10 and 40 M (Fig. 3A). These data suggest that preventing endogenous EET hydration enhances endothelial cell proliferation, in agreement with the finding that only exogenously added EETs, but not DHETs, stimulate endothelial cells proliferation (Fig. 2). Interestingly, co-administration of ACU (10 M) and EETs (1 M) further enhanced endothelial cell proliferation compared with EETs alone (Fig. 3B). We concluded from these experiments that experimental manipulation of cellular EET levels by either inhibiting their biosynthesis or their degradation has opposing anti-and proproliferative effects. This is consistent with a role for them as endogenous pro-mitogenic lipids. In addition to establishing a functional link between EET biosynthesis and cell prolifera-tion, these studies expand the list of lipid-derived mitogens to include the 5,6-and 8,9-EETs.

5,6-and 8,9-EET Promote Microvascular Endothelial Cell Migration and Tubule-like Structure Formation-Endothelial
cell migration is an important step in angiogenesis as it confers these cells with the capacity to form capillary-like structures both in vivo and in vitro. We therefore evaluated the ability of the EETs to induce endothelial cell migration using transwell plates. As shown in Fig. 4A, 5,6-and 8,9-EET caused a significant increases in cell migration (ϳ2-fold over cells treated with vehicle only), while 11,12-and 14,15-EET showed no detectable chemotactic activity. The fact that 11,12-EET has been previously characterized as a chemotactic EET in cultures of cerebral and umbilical endothelial cells (8,10,11), albeit at higher concentration, raises the interesting possibility of species and/or vascular bed EET-specific effects on cell migration.
We next determined whether the combined mitogenic and pro-migratory activities of 5,6-and 8,9-EET led to the formation of capillary-like structures. Endothelial cells were plated in serum-free medium on solidified Matrigel, and the ability of the EETs (1 M each) to promote the formation of capillary-like structures was compared with that of cells treated with vehicle (ethanol) only. Within the first 3 h of treatment, 5,6-and 8,9-EET induced capillary-like structures, as determined by the ability of endothelial cells to sprout, branch, and form ring-like structures (Fig. 4, B and C). In contrast, no significant changes were observed with 11,12-or 14,15-EET. Interestingly, at 0.5 M, 5,6-was the only EET able to induce capillary formation (not shown).
To explore the role of cellular EET synthase(s) in the formation of capillary-like structures, endothelial cells were plated on solidified Matrigel and cultured in serum-containing medium to which we added MS-PPOH or ketoconazole. Between 2.5 and 5.0 M, both epoxygenase inhibitors markedly reduced the formation of capillary-like structures (Fig. 5, A and B), indicating that blocking endogenous EET biosynthesis impairs the capacity of endothelial cells to form these structures. Significant inhibition of tube formation was also observed in human endothelial cells treated with MS-PPOH or miconazole, although these effects were reached only at high concentration of inhibitors (9). In contrast to our results, 11,12-and 14,15-EET have been reported to promote the formation of capillary-like structures and angiogenesis by using either synthetic 11,12-EET (10), or as a consequence of overexpressed CYP2C9-directed 14,15-EET biosynthesis (9). However, in the CYP2C9 overexpression experiments, the spectrum of EET metabolites generated by CYP2C9 was not characterized. Furthermore, capillary-like structures were examined only after 18 h (9) or 14 days (10) from plating. Nevertheless, the effects described for these EETs give further support to the idea of EET isomers and/or vascular bed-specific effects for the epoxygenase metabolites.
EETs Activate Distinct Signaling Pathways in Microvascular Endothelial Cells-11,12-and 14,15-EET have been shown to participate as second messengers in several signaling pathways, including growth factor-dependent mitogenic cascades (10, 20, 36 -38). Among the various mitogenic signaling pathways, the activation of ERK, p38 MAPK, and PI3K/Akt play important roles in endothelial cell function. To determine whether EETs could activate these pathways in murine lung endothelial cells, serum-starved endothelial cells were treated with each EET (1 M) for 0, 10, and 40 min and ERK, p38 MAPK, and Akt activation analyzed by Western blot. Within the first 10 min of addition, all four EETs caused a significant activation of Akt (Fig. 6A); however, only for 8,9-and 11,12-EET was the response long lasting and persisting for at least 30 min after EET addition (Fig. 6A). Whereas all four EETs caused marked and transient increases in ERK activation (Fig.  6A), 8,9-was the most potent EET in activating p38 MAPK (Fig. 6A). These results show that these eicosanoids are capable of activating mitogenic signaling cascades known to regulate endothelial cell functions, such as proliferation and migration.
Most of the intracellular pathways activated in endothelial cells have been analyzed following treatment with 11,12and/or 14,15-EET, or by transfecting human endothelial cells with Cyp2c isoforms. In agreement with our data, treatment of human endothelial cells with 11,12-EET leads to transient Akt activation (38), and this event seems to be related to the ability of this eicosanoid to induce EGF receptor phosphorylation (10). Moreover, treatment of bovine artery endothelial cells with 14,15-EET activates ERK1/2, which then stimulates the phosphorylation and activation of endothelial nitric-oxide synthase (39). However, in most of the studies with endothelial cells, only one or a few EET signaling pathways have been analyzed, and the downstream signaling molecule(s) activated by the 5,6and 8,9-EETs are yet to be identified. We now provide evidence that the three major pathways involved in endothelial cells function, namely PI3K/Akt, p38 MAPK, and ERK1/2, can be activated by the four EETs, and that the 8,9-EETs is the most potent activator of p38 MAPK. Based on this finding we next determined whether activation of one or more of these pathways was important to modulated endothelial cell proliferation and migration.
EETs Promote Microvascular Endothelial Cell Proliferation by Activating EET Regioisomer-specific Signaling Pathways-To determine the signaling pathway/s involved in EETinduced proliferation, serum-starved endothelial cells (to minimize endogenous Akt, ERK, and p38 MAPK activity, see also Fig. 6A) were incubated with each EET (1 M) in the presence or absence of either PD98059 (a MEK1/ERK inhibitor), wortmannin (a PI3K inhibitor), or a p38 MAPK inhibitor. After 24 h, cell proliferation was evaluated by measurements of [ 3 H]thymidine incorporation. As seen in Fig. 6B, all four EETs rescued cell proliferation from PD98059 inhibition, suggesting that ERK is not involved in EET-mediated mitogenesis. In contrast, whereas only 5,6-and 14,15-EET rescued cell proliferation from p38 MAPK inhibition (Fig. 6B), 8,9-and 11,12-EET restored cell proliferation in PI3K/Akt-inhibited cells. These data indicate that the proliferative response of endothelial cells following treatment with either 8,9-or 11,12-EET involves p38 MAPK activation, whereas the mitogenic activity of 5,6-and 14,15-EET are mediated by the PI3/Akt kinase. Our results suggest that these eicosanoids are capable of stimulating the same cell function by activating quite different cellular pathways. In this regard, it has been shown that the ability of 11,12-EET to induce endothelial cell proliferation results from PI3K/Akt-mediated down-regulation of the cyclin-dependent kinase inhibitor p27 kip1 (38). Moreover, studies with CYP2C9transfected endothelial cells suggested that the mitogenic activity of 11,12-EET, one of the EETs formed by the enzyme (18), results from activation of the EGF receptor and subsequent PI3K/Akt-mediated activation (10). In contrast, our studies show that pulmonary endothelial cell proliferation is primarily mediated by 11,12-EET via activation of p38 MAPK. As mentioned above, this apparent discrepancy could reflect EET isomer specific and/or vascular bed specific effects. Finally, Zhang and Harder (8) have shown that cerebral capillary endothelial cell proliferation can be stimulated by all four EETs via activation of a tyrosine kinase receptor, with 8,9-EET being the most potent and 5,6-EET being the less active. On the other hand, our results identified 5,6-and 8,9-EETs as the most potent mitogens among the four EETs and show that their mitogenic activity is dependent on PI3K/Akt or p38 MAPK activation, respectively. Again, these differences are not surprising considering that (i) endothelial cells derived from dif-ferent organs were used for the study, and (ii) it is well established that mature endothelial cells have distinctive functional properties, depending on the tissue of origin (40). Finally, an important issue when comparing results is that of the chemical lability of 5,6-EET. All our studies were done with 5,6-EET stocks that were purified and characterized by HPLC immediately prior to use.
ERK and PI3K/Akt Activation Is Required for 5,6-and 8,9-EET-induced Chemotaxis-As discussed, 5,6-and 8,9-EET induce significant endothelial cell migration (Fig. 4A), a cellular function also known to involve several protein kinase signaling cascades. To determine the pathways involved in EET-induced endothelial cell migration, we evaluated the chemotactic activity of these EETs (1 M each) in cells that were cultured in the absence or the presence of PD98059, wortmannin, or a p38 MAPK inhibitor. As shown in Fig. 6C, neither 5,6-nor 8,9-EET rescued cell migration from ERK or PI3K/Akt inhibition. On the other hand, 5,6-EET, but not 8,9-EET, rescued endothelial cell migration from p38 kinase inhibition (Fig. 6C), suggesting that this kinase is not involved in 5,6-EET-dependent chemotaxis. To the best of our knowledge, this is the first study showing not only that 5,6-and 8,9-EETs enhance endothelial cell migration, but also that they achieve this effect by activating distinct intracellular pathways.
The Cyp2c44 Epoxygenase Is Expressed in Angiogenic Blood Vessels-The expression of CYP2C isoforms and the formation of epoxygenase metabolites by cultured endothelial cells has been reported (reviewed in Ref. 24 and the present study). Furthermore, isolated rat vascular tissues catalyze EET formation, and express a CYP2C23 epoxygenase (33). However, other than the demonstration of the EETs as endogenous constituents of human and rat plasma (12), there is scant evidence of vascular CYP2C expression in vivo (12,15). To determine whether the murine Cyp2c44 epoxygenase is expressed in angiogenic blood vessels, we performed immunofluorescence analysis on sections from tumors originated by human cancer cells grown into nude mice using: (a) anti-mouse CD31 antibody to visualize vessels within the tumor, and (b) anti-rat CYP2C23 antibody, cross-reactive toward mouse Cyp2c44. A human tu-mor cell line was selected for these studies because it lacks proteins that are immunoreactive with CYP2C23 antibodies, thus facilitating the identification of mouse Cyp2c44-positive structures. Fig. 7A illustrates the expression and co-localization of the Cyp2c44 epoxygenase with the CD31 endothelium marker, indicating that the mouse-derived tumor vessels express this epoxygenase, and that it could play a role in angiogenesis in vivo. 5,6-and 8,9-EET Are Potent Angiogenic Factors in Vivo-Endothelial cell proliferation, migration, and the development of capillary-like morphology, combined responses elicited only after treatment with 5,6-and 8,9-EET (Figs. 2 and 4), are essential components of angiogenesis. To test the angiogenic activity of 5,6-and 8,9-EET in vivo, we utilize a subcutaneous sponge model (29). Inert sponges, implanted subcutaneously in the back of adult mice, were injected every other day with either vehicle (oil), 5,6-EET, or 8,9-EET (50 M each). After 14 days, we retrieved the sponges and compared their vessel density by direct observation and by analysis of vessel-associated rhodamine-dextran fluorescence. Sponges injected with 5,6-or 8,9-EET show significantly increased vessel density compared with sponges injected with vehicle only, demonstrating clearly in vivo de novo vascularization (Fig. 7, B-D).
To determine whether the potentiating effects of ACU on the EET-stimulated endothelial cell proliferation in vitro (Fig. 3) have functional significance to in vivo angiogenesis, the epoxide hydrolase inhibitor was injected either alone or a combination with either 5,6-or 8,9-EET into implanted sponges, and its effects on angiogenic activity determined as above. As shown in Fig. 7, the pro-angiogenic activity of 5,6-and 8,9-EET was significantly enhanced when they were co-injected with ACU (Fig. 7, B-D). Furthermore, the administration of ACU alone also caused significant increases in vascularization when compared with vehicle only, suggesting that inhibition of exog-FIG. 6. EET-induced cell proliferation and migration requires activation of specific downstream signaling molecules. A, serumstarved endothelial cells were either kept untreated or treated for 10 and 40 min with the different EETs indicated (1 M each). Levels of phosphorylated as well as total ERK, Akt, and p38 were determined on total cell lysates (50 g/lane) by Western blot analysis. B, endothelial cells were cultured as described in the legend to Fig. 2A with or without 1 M EETs alone or in combination with the kinase inhibitors indicated and proliferation measured 24 h after. Values are changes relative to cells grown in the absence of EETs and kinase inhibitors (ethanol only was added) and are averages calculated from three experiments. C, endothelial cells were serum-starved for 24 h and then plated in transwell dishes coated with 10 g/ml collagen I. Cells were then allowed to migrate for 12 h at 37C toward serum-free medium with or without 1 M EETs alone or in combination with the kinase inhibitors indicated. Values are changes relative to cells migrating in the absence of EETs and kinase inhibitors (ethanol only added) and are averages calculated from three experiments. Differences between untreated and EET-treated cells (*), or cells treated with EETs alone or in combination with the kinase inhibitors indicated (**) were significant with p Ͻ 0.05. enous EET hydration increases angiogenic potential. Finally, analysis of EET and DHETs present in the sponges at 14 days indicted that ACU prevented by 40% EET hydration (not shown).
Several groups, using different models of angiogenesis, have documented the pro-angiogenic activities of synthetic 11,12and 14,15-EET (8 -11,17). However, this is the first study that characterizes the 5,6-and 8,9-EET as potent pro-angiogenic molecules, and that addresses the mechanisms responsible for their mitogenic and chemotactic activity. In addition, our in vivo angiogenesis assay not only shows the ability of 5,6-and 8,9-EETs to induce de novo vascularization, but also to promote the formation of functionally intact vessels, as determined by vessel-associated rhodamine-dextran fluorescence. Finally, we demonstrate that preventing 5,6-and 8,9-EET enzymatic hydration could be used to enhance the angiogenic activity of these endogenous lipids.