Protective effects of epoxyeicosatrienoic acids on human endothelial cells from the pulmonary and coronary vasculature

Anuradha Dhanasekaran, Rula Al-Saghir, Bernardo Lopez, Daling Zhu, David D. Gutterman, Elizabeth R. Jacobs, Meetha Medhora

Abstract

Epoxyeicosatrienoic acids (EETs) are cytochrome P-450 (CYP) metabolites synthesized from the essential fatty acid arachidonic acid to generate four regioisomers, 14,15-, 11,12-, 8,9-, and 5,6-EET. Cultured human coronary artery endothelial cells (HCAECs) contain endogenous EETs that are increased by stimulation with physiological agonists such as bradykinin. Because EETs are known to modulate a number of vascular functions, including angiogenesis, we tested each of the four regioisomers to characterize their effects on survival and apoptosis of HCAECs and cultured human lung microvascular endothelial cells (HLMVECs). A single application of physiologically relevant concentration of 14,15-, 11,12-, and 8,9-EET but not 5,6-EET (0.75–300 nM) promoted concentration-dependent increase in cell survival of HLMVECs and HCAECs after removal of serum. The lipids also protected the same cells from death via the intrinsic, as well as extrinsic, pathways of apoptosis. EETs did not increase intracellular calcium concentration ([Ca2+]i) or phosphorylate mitogen-activated protein kinase p44/42 when applied to these cells, and their protective action was attenuated by the phosphotidylinositol-3 kinase inhibitor wortmannin (10 μM) but not the cyclooxygenase inhibitor indomethacin (20 μM). Our results demonstrate for the first time the capacity of EETs to enhance human endothelial cell survival by inhibiting both the intrinsic, as well as extrinsic, pathways of apoptosis, an important underlying mechanism that may promote angiogenesis and endothelial survival during atherosclerosis and related cardiovascular ailments.

  • cytochrome P-450
  • eicosanoids
  • epoxygenase
  • apoptosis
  • calcium signaling

epoxyeicosatrienoic acids EETs are cytochrome P-450 (CYP) metabolites of the lipid arachidonic acid (AA). They mediate numerous biological functions (9, 32, 43, 47) and have been identified as endothelium-derived hyperpolarizing factor that dilate bovine (8, 17) and porcine coronary vessels (15). The vasoactive roles of EETs in the lung have been less clear, and EETs have been reported to dilate as well as constrict pulmonary vessels (24, 48, 50, 59). In human coronary arteries isolated from patients with coronary artery disease, CYP metabolites, and not nitric oxide, mediate flow-induced vasodilation, while nitric oxide is found to participate in the vasodilator response in patients without coronary disease (35, 36). EETs are synthesized in vascular tissue by CYP epoxygenase enzymes (9, 15, 17, 44, 57). One of the more recently explored functions of these enzymes is their ability to promote growth (11, 21, 23) of endothelial cells (16, 31), including those cells isolated from human umbilical vein and coronary arteries (16, 41). Recombinant human epoxygenase 2C9 stimulates growth and differentiation of human lung microvascular endothelial cells (HLMVECs) in culture when introduced by viral-mediated gene transfer (31). Products of epoxygenases induce growth of rat cerebral microvascular endothelial cells (58, 38), while 14,15-EET (the most abundant regioisomer in the lung; Refs. 31, 57) and 8,9-EET (58) are angiogenic when embedded in subcutaneous Matrigel plugs in rats in vivo. 11,12-EET enhances vessel growth and convergence when applied in a chick chorioallantoic membrane model of angiogenesis (34), and related eicosanoids have been demonstrated to be potent angiogens in the rabbit cornea (28). 8,9- and 5,6-EET have recently also been described as angiogenic in subcutaneous sponges implanted in mice, and 8,9- as well as 11,12-EET enhanced mitogenesis via activation of the p38 mitogen-activated protein kinase (MAPK) pathway, whereas activation of phosphotidylinositol-3 kinase (PI-3 kinase) was necessary for cell proliferation induced by 14,15- and 5,6-EET (42). In addition to growth, epoxygenase is protective against reperfusion after ischemia in the brain (2) and heart (20, 46). EETs have also been reported to inhibit apoptosis in proximal tubule-like epithelial cells derived from pig kidney (10). Overexpression of epoxygenase 2J2 promoted proliferation of a number of tumor cell lines in culture, and three EET regioisomers, 8,9-, 11,12- and 14,15-EET, protected human carcinoma cells from apoptosis, thereby stimulating tumor growth and development (25). 14,15-EETs have been reported to increase intracellular calcium concentrations ([Ca2+]i) in vascular smooth muscle cells (14). Calcium is a ubiquitous second messenger for a range of cellular processes (5), including cell proliferation/survival and regulation of endothelium-dependent vascular relaxation such as production of NO, so we were interested to see if EETs were able to alter [Ca2+]i in endothelial cells. Previous studies indicated that 5,6-EET was a second messenger for activation of endothelial calcium entry in bovine coronary artery endothelial cells (19), and it also elevated calcium entry into astrocytes (45).

Because the profile of EET regioisomers varies with tissue-dependent expression of CYP enzymes, we examined the growth effect of each regioisomer in vascular cells derived from human lungs, as well as heart, as a first step to resolve the functions of these compounds in the vasculature. Protective and angiogenic effects of EETs would be important in the event that they were to be considered for therapy, e.g., to promote growth of new vessels in pathological conditions such as ischemic heart disease and acute lung injury. The cardiovascular system needs to undergo dynamic adaptation, including growth, remodeling, survival, and apoptosis (56). Because the vasoactive effects of EETs, and indeed that of several other vasoactive agents, are opposite in the pulmonary (24, 59) vs. coronary vessels (8, 15, 17), we tested the effects of these reagents on both types of endothelial cells. We report for the first time that three EET regioisomers (14,15-, 11,12-, and 8,9-EET) attenuate two different pathways of apoptosis in human lung and coronary artery endothelial cells.

MATERIALS AND METHODS

Cell Culture

Human microvascular endothelial cells of the lung (HLMVEC) and human coronary artery endothelial cells (HCAECs) were obtained from Clonetics (Walkersville, MD) and maintained as described by the manufacturer. Passages 4–10 were used. HLMVEC and HCAEC were cultured using EGM-2MV media [containing hydrocortisone, human fibroblastic growth factor, vascular endothelial growth factor (VEGF), insulin-like growth factor, ascorbic acid, human epidermal growth factor, gentamycin, amphotericin, and serum; Clonetics CC-4147]. Medium was changed every 2–3 days. EETs were obtained from Biomol (Plymouth Meeting, PA) and stored at −70°C under nitrogen or were synthesized and kindly provided by the laboratory of Dr. John Falck (Univ. of Texas, Southwestern Medical Center, Dallas, TX).

Survival Assays

Cell count.

Subconfluent cells in T 25-cm2 flasks were washed, detached with trypsin-EDTA (T/E), and neutralized with the trypsin neutralization solution (TNS) as recommended by the manufacturer (Clonetics). The cells were suspended in EGM-2MV at a density of 3 × 104 cells/well in 24-well plates for 24–48 h, at which point they were 60–80% confluent. They were serum starved in basal medium [EBM (Clonetics), supplemented with fetal bovine serum (0.1%)] for 24 h to arrest cell growth. Cells were then treated with EETs (0.075–600 nM), ethanol (vehicle), or complete medium, in quadruplicates for 24 h. Medium was suctioned off, and cells were detached with 10× T/E and manually counted by using a hemocytometer. Each experimental data point represents an average count of four wells from two to four independent experiments with at least two separate batches of cells.

MTT Assay

The cells (HCAEC and HLMVEC) were cultured in 60-mm dishes (∼1 × 106 cells) to ∼80% confluency and then maintained in either complete medium (EGM-2MV) or switched to basal medium for the next 42–48 h. The samples were treated with 14,15-EET (300 nM for HLMVECs) or 8,9-EET (300 nM for HCAECs) or vehicle. At the end of the incubation period, the survival of cells was determined by the MTT assay as described in the manufacturer's protocol (V-13154, Molecular Probes, Eugene, OR). Briefly, the cells were incubated for 3 h in phenol red-free medium containing 0.5% of the yellow mitochondrial dye 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide (MTT). The amount of blue formazan dye formed from MTT was proportional to the number of live cells. The MTT reaction was terminated by adding DMSO to the medium followed by incubation for 10 min at 37°C. The absorbance was read at 540 nm in a spectrophotometer.

Hoechst Staining

Cells (HCAEC and HLMVEC) were cultured in four-well chamber slides to 70% confluency, serum deprived, and treated with EETs as described for the MTT assay. Cells were stained with 1 μl of Hoechst 33342 (5 mg/ml; V-13244, Molecular Probes) in 1 ml basal medium and incubated for 30 min. Stained cells were then washed twice with PBS (Sigma, St. Louis, MO) and imaged under a fluorescent microscope by using a 460-nm filter.

Propidium Iodide Staining of Floating Cells

Cells (HCAEC and HLMVEC) were cultured in 60-mm dishes to 70% confluency, serum-deprived, and treated with EETs as described for the MTT assay. The medium was collected and centrifuged at 5,000 g for 5 min, and the cells in the pellet were treated with 1 μl in 50 μl propidium iodide (stock 1.0 mg/ml, Molecular Probes) and incubated for 20 min. The cells were then washed twice with PBS and imaged using a 620-nm filter in a fluorescent microscope.

Calcium Measurements

These experiments examined the effect of EETs on [Ca2+]i in HLMVECs or HCAECs that were cultured by using EGM-2MV medium on glass coverslips in 35-mm dishes. The cells were then placed in basal medium for 24 h to arrest cell growth. They were loaded with 5 μM fura-2 AM (Molecular Probes, Invitrogen, Carlsbad, CA) in basal medium and 0.02% pluronic acid in the dark for 45 min at room temperature. After being loaded, the cells were transferred to a 1 ml perfusion chamber on an inverted microscope and superfused with physiological salt solution (PSS) containing (in mM) 119 sodium chloride, 4.7 potassium chloride, 1.6 calcium chloride, 1.17 magnesium sulfate, 12 sodium bicarbonate, 1.18 sodium dihydrogen phosphate, 10 HEPES (pH 7.4), and 10 glucose, in the dark for 15 min at 37°C. [Ca2+]i was measured using an InCyt Im2 imaging system (Intracellular Imaging, Cincinnati, OH) mounted on an inverted microscope (Nikon TS-100F). The cells were alternatively excited at wavelengths of 340 and 380 nm, and images were recorded at emission wavelength of 510 nm by using an ultraviolet fluorescence objective (40× magnification). The [Ca2+]i was calculated on the basis of the fluorescence intensity ratios obtained by using excitation and emission wavelengths of 340/380 and 510 nm, respectively, and a standard curve generated using solutions with known calcium concentrations. This method was used because treatment with calcium ionophore or digitonin at the end of the experiment to calibrate levels of calcium often resulted in decreased fluorescence due to leakage of fura-2 from the cells. Although this ex vivo calibration technique may in theory introduce small errors in absolute [Ca2+]i values, measurements of changes in [Ca2+]i with stimulation are very reliable and may be more accurate than in vivo calibrations. After baseline [Ca2+]i was measured, 14,15-, 11,12-, 8,9-, or 5,6-EET (15 μM) or vehicle (ethanol) was added to the bath, and [Ca2+]i was recorded during a 10-min experimental period. Calcium concentrations under baseline conditions and after addition of vehicle or EETs were averaged to obtain a single value for statistical comparison.

Western Blot

Endothelial cells (HCAECs and HLMVECS) were cultured in 35-mm dishes to 70% confluency, washed, and incubated with basal medium for 24 h. Vehicle or EETs or complete medium were added, and after 5 min the cells were kept on ice and washed three times with cold PBS. Proteins were solubilized and extracted with 50 μl RIPA buffer [50 mM Tris pH 8.0, 150 mM NaCl, 0.5% SDS, 1% Nonidet P40, 0.5% sodium deoxycholate, 1 mM EDTA, 1× protease inhibitor cocktail (Pharmingen, San Diego, CA), 1× phosphatase inhibitors (Calbiochem, San Diego, CA)]. The lysates were used to estimate their protein content with the Bio-Rad DC Protein Assay Reagent (Bio-Rad, Hercules, CA). Equal amounts of protein (10 μg) from each sample were electrophoresed on a 10% SDS-polyacrylamide gel with running buffer as described (3, 30, 60). The proteins in the gel were transferred to nitrocellulose as described (3). The membrane was treated with primary antibody for pERK 1/2 (1:1,000 dilution, no. 9101S, Cell Signaling, Charlottesville, VA) for 18 h at 4°C and washed three times before incubating with matched secondary antibody (1:5,000) for 45 min. The protein bands were developed with chemiluminescence reagents. The membrane was stripped, rinsed, and redeveloped by using antibody against ERK1/2 (p44/42, 1:1,000, no. 9102, Cell Signaling).

Annexin V Binding

Annexin V binds to phosphatidyl serine (PS), which appears in the outer leaflet of the plasma membrane in early apoptotic cells. Cells (HCAEC and HLMVEC) were cultured in 60-mm dishes to 70% confluency, serum deprived, and treated with EETs as described for the MTT assay. For some experiments, the EETs were replaced with 14,15-dihydroxyeicosatetraenoic acids (14,15-diHETEs) (300 nM) or 8,9-diHETEs (300 nM) both purchased from Cayman Chemicals (Ann Arbor, MI). Cyclooxygenase inhibitor (indomethacin, 20 μM, Sigma Chemicals, St. Louis, MO) or PI-3 kinase inhibitor (wortmannin, 10 μM, Alexis, Lausen, Switzerland) were added along with the EETs in certain experiments where indicated in the text. For Fas-induced apoptosis, confluent cultures of HLMVECs or HCAECs were grown in 60-mm dishes. Apoptosis was induced by adding anti-Fas antibody (100 ng/ml, Upstate Cell Signaling Solutions, Lake Placid, NY) for 8 h in basal medium in presence or absence of vehicle or one of four EET regioisomers (300 nM each). The cells were washed with PBS and treated with FITC-labeled annexin V (0.2 μg/ml) for 20 min at room temperature. Labeling with FITC-coupled annexin V was performed according to the manufacturer's protocol (BD Biosciences, San Diego, CA) as previously described (4). The labeled cells (10,000/sample) were analyzed by measuring fluorescent intensity with the use of a FACScan flow cytometer in conjunction with CellQuest software (BD Biosciences).

Activity of Caspase-3

As described for annexin binding, cells were serum deprived for 42–48 h or treated with anti-Fas antibody (100 ng/ml) for 24 h in basal medium in the presence or absence of vehicle or one of the four EET regioisomers (300 nM). Control cells were maintained in EGM-2MV or were treated with vehicle instead of the anti-Fas antibody. The treated cells were collected by centrifugation, washed twice with ice-cold PBS, resuspended in lysis buffer [5 mM MgCl2, 1 mM EGTA, 0.1% Triton X-100, and 25 mM HEPES (pH 7.5)], and stored overnight at −80°C as previously described (12, 13). Release of free 7-amino-4-trifluoromethyl coumarin (AFC) from the synthetic substrate Ac-DEVD-AFC (50 μM) at 37°C was determined by fluorescence measurements in a SpectraFluor Plus plate reader after excitation at wavelength 390 nm. Emission was measured at 535 nm.

Statistical Comparisons

All values are expressed as means ± SE from at least three or more samples in each experiment. Cell numbers after serum deprivation (48 h) or presence of serum (48 h) were compared with initial counts using Student's t-test. Effects of vehicle or increasing concentration of an EET regioisomer on cell counts were compared by ANOVA with post hoc Newman Keuls (normal distribution) or Tukey (nonparametric) tests. Intracellular calcium was compared using a paired t-test. Annexin V fluorescence and caspase activities under baseline conditions and after addition of vehicle (control) or EETs were compared by ANOVA on ranks followed by a Tukey test.

RESULTS

Reduced Survival (Apoptosis) of Human Endothelial Cells After Withdrawal of Growth Factors and Serum

HLMVECs and HCAECs were cultured in EGM-2MV, which contains a number of growth factors (see materials and methods) as well as serum. When HLMVECs reached 50–70% confluency, the cells were washed once in PBS and maintained in basal medium (EBM with 0.1% FBS) for 24 h. The cells were then counted by using a hemocytometer. After serum starvation, there was a decrease in cell number (57% of the original in 24 h) that continued to fall to only 32% of cells remaining after 48 h of serum withdrawal (Fig. 1A). This value compares to an increase (93%) in cell number observed after continued incubation with complete growth medium, EGM-2MV.

Fig. 1.

Survival of human lung microvascular endothelial cells (HLMVECs) after withdrawal of growth factors and protection by epoxyeicosatrienoic acids (EETs) A: HLMVECs were cultured to >50% confluency and then either depleted of growth factors or maintained in complete medium for 48 h, with one set of samples being counted to determine cell numbers at the 0-h time point (marked on the x-axis). Cells were then counted in the remaining samples after 48 h. While the cells in the complete growth medium proliferated, those that were serum deprived showed significant decrease in cell numbers with time. All values are denoted as means ± SE. *P < 0.05 compared with the 0-h time point. B–E: cells were growth arrested for 24 h by serum withdrawal and stimulated with vehicle or increasing concentrations of EETs (shown on x-axis) for the next 18–24 h in serum-free medium. The cells were then manually counted in a hemocytometer 48 h after withdrawal of growth factors. Results shown are means ± SE from 2 or more independent batches of cells. *P < 0.05 for analyses compared with vehicle control. Bar graphs represent concentration-dependent cell survival of HLMVECs with 14,15-EET (B), 11,12-EET (C), 8,9-EET (D), or 5,6-EET (E). Note protection of cell numbers from serum starvation by 14,15-, 11,12-, and 8,9-EET but not by 5,6-EET.

We repeated the experiment with HCAECs starting with a lower level of confluency (40–50%) to make sure the cells had adequate opportunity to proliferate because EETs have been reported to enhance proliferation of these cells. Once again, we observed only 38% of the original cell number after withdrawal of serum for 48 h, while the cell density increased by 230% if serum was not removed (Fig. 2A). Thus the serum deprivation not only resulted in growth arrest of the human endothelial cells but significantly decreased cell numbers (Figs. 1A and 2A).

Fig. 2.

Survival of human coronary artery endothelial cells (HCAECs) after withdrawal of growth factors and protection by EETs. A: HCAECs were cultured to 40–50% confluency and then either depleted of growth factors or maintained in complete medium for 48 h, while one set of samples were counted for the 0 h time point. Cells were then counted in the remaining samples after 48 h. While the cells in the complete growth medium proliferated, those that were serum deprived showed significant decrease in cell numbers with time. All values are denoted as means ± SE. *P < 0.05 compared with the 0-h time point. B–E: cells were growth arrested for 24 h as described in materials and methods and stimulated with vehicle or increasing concentrations of EETs (shown on x-axis) for the next 24 h under serum-depleted conditions. The cells were then manually counted 48 h after growth factor withdrawal. Results shown are means ± SE from 2 or more independent batches of cells. *P < 0.05 for analyses compared with vehicle control. Bar graphs represent concentration-dependent growth of HCAECs with 14,15-EET (B), 11,12-EET (C), 8,9-EET (D), or 5,6-EET (E). Note protection of cell numbers from serum withdrawal by 14,15-, 11,12-, and 8,9-EET but not by 5,6-EET, which were added 24 h after growth arrest.

Protective Effect of EETs on Endothelial Cell Survival After Serum Deprivation

We tested the effects of all four EET regioisomers on human endothelial cell survival. The cells were incubated in basal medium for 24 h before treating them with each of four EET regioisomers or vehicle with continued incubation in basal medium for another 18–24 h. The cells were therefore maintained under serum-free conditions for 42–48 h but incubated in the presence of 14,15-, 11,12-, 8,9-, 5,6-EETs, or vehicle (single application) for the last 18–24 h. Three of the lipids increased the number of surviving cultured pulmonary and coronary endothelial cells compared with the vehicle-treated controls. 14,15-EET increased survival of both HLMVEC (Fig. 1B) and HCAEC (Fig. 2B). The maximal increase was 60 ± 13.5% and 46 ± 10.8% for each cell type, respectively. The responses were concentration dependent with effective range between 0.75 and 600 nM and with peak responses seen at 600 nM for HLMVEC and 30 nM for HCAEC. Other EET regioisomers were also tested (Figs. 1, C and D, and 2, C and D). 11,12-EET and 8,9-EET but not 5,6-EET increased numbers of adherent HLMVECs and HCAECs. 11,12-EET induced an 87% increase in HLMVEC at 300 nM and 45% increase in cell count of HCAEC at a concentration of 75 nM. Cells treated with 8,9-EET showed a maximal increase in counts of slightly more than 70% for both HLMVEC and HCAECs at concentrations of 300 and 30 nM, respectively. 5,6-EET did not increase or decrease HLMVECs or HCAECs compared with vehicle-treated cells, and this was observed after application of at least two independently obtained batches of the lipid that were freshly prepared before treatment. Interestingly, the same batches of 5,6-EET increased proliferation and survival of human coronary artery smooth muscle cells in culture (results not shown).

We tested the viability of the endothelial cells that remained adherent after serum deprivation compared with cells maintained in EGM-2MV (control) or after treatment with an inducer of the extrinsic pathway of apoptosis, anti-Fas antibody, by the colorimetric conversion of MTT to formazan. Serum deprivation decreased cell viability of HLMVECs, while 14,15-, 11,12-and 8,9-EETs were able to protect these cells to 78 ± 1.7%, 88 ± 0.9%, and 85 ± 1.7% of the control value, respectively (Fig. 3A). Results with HCAECs were similar, and 14,15-, 11,12-, and 8,9-EETs were able to protect these cells to 86 ± 13.7%, 80 ± 2.0%, and 85 ± 4.2% of the control value, respectively (Fig. 4A). Viability of cells after Fas-induced apoptosis also decreased and was rescued in both cell types by addition of three isomers of EETs (Figs. 3B and 4B). To demonstrate that the floating cells were dead as opposed to only detached, these cells were harvested and stained with propidium iodide (Figs. 3C and 4C). Serum and growth factor deprivation considerably increased the population of floating cells that stained positive (red) with propidium iodide but were protected by treatment with either 14,15-EET (300 nM in HLMVECs, Fig. 3C) or 8,9-EET (300 nM in HCAECs, Fig. 4C), respectively. Finally we looked for condensation of chromatin after serum deprivation by staining with Hoechst 33342 (Figs. 3D and 4D). Once again, removal of growth factors resulted in condensed chromatin (marked with arrows), which was protected by either 14,15-EET (300 nM in HLMVECs, Fig. 3D) or 8,9-EET (300 nM in HCAECs Fig. 4D), respectively. Therefore, human endothelial cells started dying by 24 h incubation in serum-free medium, and this effect continued to increase with time.

Fig. 3.

Cell death after induction of apoptosis and protection by EETs (HLMVEC). Cells were serum deprived (SD) or treated with vehicle, anti-Fas antibody, or EGM-2MV in presence of vehicle or EET. Controls for serum-deprived cells were treated with EGM-2MV for 42–48 h, while controls for Fas-induced apoptosis were exposed to vehicle instead of anti-Fas antibody. A and B: cell viability was determined using the MTT assay (see materials and methods). *P < 0.05 for analyses compared with serum-deprived or anti-Fas antibody treated samples. C: floating cells were harvested and stained with propidium iodide. Note increase in number and intensity of stained cells after serum deprivation, which is considerably restored by preincubation with 14,15-EET (300 nM). D: HLMVECs were stained with Hoechst reagent, and nuclei were imaged. There is visible condensation of chromatin in serum-deprived cells (marked with arrows) that is absent in control cells cultured in EGM-2MV and reduced in cells pretreated with 14,15-EET (300 nM).

Fig. 4.

Cell death after induction of apoptosis and protection by EETs (HCAEC). Cells were serum deprived or treated with vehicle, anti-Fas antibody, or EGM-2MV in presence of vehicle or EET. Controls for serum-deprived cells were treated with EGM-2MV for 42–48 h, while controls for Fas-induced apoptosis were exposed to vehicle instead of anti-Fas antibody. A and B: cell viability was determined by using the MTT assay (see materials and methods). *P < 0.05 for analyses compared with serum-deprived or anti-Fas antibody-treated samples. C: floating cells were harvested and stained with propidium iodide. Note increase in number and intensity of stained cells after serum deprivation, which is considerably restored by preincubation with 8,9-EET (300 nM). D: HCAECs were stained with Hoechst reagent, and nuclei were imaged. There is visible condensation of chromatin in serum-deprived cells (marked with arrows) that is absent in control cells cultured in EGM-2MV and reduced in cells pretreated with 8,9-EET (300 nM).

Calcium Measurements

Addition of 3 nM to 1 μM EET to primary cultures of HLMVEC and HCAEC did not evoke any increase in [Ca2+]i (data not shown). However, addition of at least 15 μM of 14,15-, 11,12-, or 8,9-EET to HLMVECs showed a significant increase in [Ca2+]i that peaked ∼4 min after application of these three regioisomers and lasted for at least 5 min (Fig. 5, A and B). The increase in intracellular calcium peaked at ∼120 s after addition of 15 μM of 14,15- and 8,9-EETs to HCAECs while 15 μM of 11,12-EETs demonstrated a slower rise with sustained increase over 500 s (>8 min) after application (Fig. 6, A and B). Neither 5,6-EET (15 μM) nor vehicle (ethanol) altered [Ca2+]i when added to HLMVECs or HCAECs. The baseline for intracellular calcium ranged from 20 nM (HCAECs) or 30 nM (HLMVECs) and significantly increased to 40 nM (HCAECs) or 50 nM (HLMVECs) after treatment with EET regioisomers.

Fig. 5.

Effect of EETs on intracellular calcium concentration ([Ca2+]i) in HLMVECs. A: tracings from single cells under baseline conditions after treatment with vehicle (ethanol) and 14,15-, 11,12-, 8,9-, and 5,6-EET. B: summary of effect of vehicle and EETs on [Ca2+]i in all cells studied. *Significant difference (P < 0.05) from the baseline value within the same group for which number of cells analyzed are denoted in the graph legends.

Fig. 6.

Effect of EETs on [Ca2+]i in HCAECs. A: tracings from single cells under baseline conditions after treatment with vehicle (ethanol) and 14,15-, 11,12-, 8,9-, and 5,6-EET. B: summary of effect of vehicle and EETs on [Ca2+]i in all cells studied. *Significant difference (P < 0.05) from the baseline value within the same group for which number of cells analyzed are denoted in the graph legends.

EETs Protect Human Endothelial Cells from Apoptosis Induced by Growth Factor Removal or Treatment with anti-Fas Antibody

Serum deprivation is known to trigger the intrinsic pathway of apoptosis (1). EETs were seen to increase survival of human endothelial cells under this condition. We did not observe stimulation of the proliferative ERK pathway by application of EETs to endothelial cells after 24 h of growth factor withdrawal. Figure 7 depicts HCAECs as well as HLMVECs stimulated for 5 min with 75 nM of 14,15-EET. While EGM-2MV induced a rapid phosphorylation of ERK1/2 (Fig. 7, lanes 3 and 6), the enzymes were not phosphorylated by stimulation with EETs (Fig. 7, lanes 2 and 5). Similar results were observed with the other regioisomers (11,12-, 8,9-, and 5,6-EETs), as well as after stimulation for 15 min, 30 min, 1 h, or 4 h of treatment with EETs at concentrations up to 300 nM (results not included). We also did not observe a significant increase in incorporation of [3H]thymidine in serum-starved HLMVECs treated with 300 nM of 14,15-EET or HCAECs in the presence of 8,9-EET (300 nM) (results not shown). We therefore tested the role of all four EET regioisomers on protection of human endothelial cells from apoptosis induced by serum deprivation (intrinsic pathway) and also by the anti-Fas antibody (extrinsic pathway) to determine if these lipids inhibited both forms of programmed cell death. Using two independent assays, annexin V binding (early marker) and activation of intracellular caspase-3, we monitored whether EETs were able to protect cells from apoptosis. Typical results of a single experiment obtained by flow cytometry are depicted in the top panels, while average values from at least three independent experiments are graphed below these in Figs. 811. Application of 14,15-EET (300 nM) or 8,9-EET (300 nM) significantly inhibited apoptosis by serum deprivation in HLMVECs or HCAECs, respectively, by using binding of annexin V as an indicator of programmed cell death (Figs. 8, A, B, and F, and 10, A, B, and F). The corresponding diHETEs administered at similar concentrations (300 nM) were not as effective (Figs. 8C and 10C). 14,15-diHETE decreased annexin binding (9.6% lower) compared with serum-deprived, vehicle-treated cells, but at the same time it also demonstrated a significant increase in apoptosis (∼160%) compared with EGM-2MV controls (Fig. 8, A and F). To investigate a signaling pathway that may relay the effect of 14,15- or 8,9-EET on HLMVECs or HCAECs, respectively, we repeated the experiment in the presence of two different pharmacological inhibitors: the cyclooxygenase inhibitor indomethacin (20 μM) and the PI-3 kinase inhibitor wortmannin (10 μM) (Figs. 8, D–F, and 10, D–F). The wortmannin significantly attenuated the protection by EETs.

Fig. 7.

ERK is not activated in human endothelial cells treated with EET. HCAECs and HLMVECs were grown to 70% confluency and serum starved for 24 h. They were then stimulated with vehicle (lanes 1 and 4), 14,15-EET (75 nM, lanes 2 and 5) for 5 min, and EGM-2MV (lanes 3 and 6). The cells were lysed in the presence of phosphatase inhibitors, and the lysates were developed by Western analysis with antibody to phosphorylated ERK (p-ERK, top) to determine if the EET activated this enzyme. Bottom: same gel treated with antibody for ERK to demonstrate protein loading. Samples treated with vehicle or 14,15-EET did not show increase in phosphorylation of ERK as did cells treated with growth factors. Similar results were obtained for all 4 regioisomers of EETs at concentrations ranging from 10 to 600 nM and 3 other time points (15 min, 30 min, 1 h, and 4 h; results not shown).

Fig. 8.

Protection of serum-deprived HLMVECs by 14,15-EET (annexin binding). HLMVECs were cultured to 80% confluency, and controls were maintained in EGM-2MV. Other samples were serum deprived and either pretreated with vehicle or 14,15-EET (300 nM) or 14,15-dihydroeicosatetraenoic acid (14,15-diHETE) (300 nM) or inhibitors in the presence of 14,15-EET. Cells were incubated with binding buffer containing annexin V as described in materials and methods. A–E: typical histograms comparing fluorescent intensity of control cells with those deprived of serum and treated as labeled. F: bar graph showing means + SE of percent increase of annexin binding over control (100%). WM, wortmannin (10 μM). IM, indomethacin (20 μM). *P < 0.05 compared with serum-deprived cells; **P < 0.05 compared with serum-deprived cells that were rescued with 14,15-EET (3rd bar from left). Note attenuation of protection by wortmannin but not indomethacin, indicating that action of 14,15-EET is mediated by the phosphotidylinositol-3 kinase (PI-3 kinase) pathway via inhibition of caspase-3 activity. FL1-H, arbitrary fluorescent units.

We also tested all three EET regioisomers, 14,15-, 11,12-, or 8,9-EET (300 nM each) and demonstrated significant protection of HLMVECs (Fig. 9, A–F) and HCAECs (Fig. 11, A–F) from increased binding of annexin V after induction of apoptosis by engaging the Fas receptor to stimulate the extrinsic pathway of apoptosis. Controls in these experiments (Figs. 9F and 11F) were treated with vehicle instead of the anti-Fas antibody.

Fig. 9.

Protection of Fas-induced apoptosis in HLMVECs by EET (annexin binding). HLMVECs were cultured to 80% confluency and were pretreated with one of the four EET followed by anti-Fas antibody (100 ng/ml). Cells were assayed after incubation with annexin V (see materials and methods). A–E: typical histograms of cells treated with anti-Fas antibody with or without the pretreatment with EET regioisomers as labeled. F: bar graph depicting the means + SE of percent increase of annexin binding compared with controls not treated with anti-Fas antibody. *P < 0.05 compared with apoptotic cells induced with anti-Fas antibody. Three EET regioisomers, 14,15-EET, 11,12-EET, and 8,9-EET, but not 5,6-EET, protected HLMVECs from Fas-induced apoptosis.

Fig. 10.

Protection of serum-deprived HCAECs by 8,9-EET (annexin binding). HCAECs were cultured to 80% confluency, and controls were maintained in EGM-2MV. Other samples were serum deprived and either pretreated with vehicle or 8,9-EET (300 nM) or diHETE (300 nM) or inhibitors in the presence of 8,9-EET. Cells were incubated with binding buffer containing annexin V as described in materials and methods. A–E: fluorescent intensity comparisons of control cells with those deprived of serum and treated as labeled in each panel. F: means + SE of percent increase of annexin binding over control (100%). *P < 0.05 compared with serum-deprived cells; **P < 0.05 compared with serum-deprived cells that were rescued with 8,9-EET (third bar). SD, serum deprived; IM, indomethacin (20 μM); WM, wortmannin (10 μM). Note attenuation of protection by wortmannin but not indomethacin, indicating that action of 14,15-EET is mediated by the PI3 kinase pathway, via inhibition of caspase-3 activity.

Fig. 11.

Protection of Fas-induced apoptosis in HCAECs by EET (annexin binding). HCAECs were cultured to 80% confluency and were pretreated with one of the four EET followed by anti-Fas antibody (100 ng/ml). Cells were assayed after incubation with annexin V (see materials and methods). A–E: cells treated with anti-Fas antibody with or without pretreatment with EET regioisomers as labeled. F: means ± SE of percent increase of annexin binding compared with controls not treated with anti-Fas antibody. *P < 0.05 compared with apoptotic cells induced with anti-Fas antibody. Three EET regioisomers 14,15-EET, 11,12-EET, 8,9-EET but not 5,6-EET protected HLMVECs from Fas-induced apoptosis.

We further tested 14,15-EET (300 nM) and 8,9-EET in HLMVECs and HCAECs (300 nM) for protection against apoptosis by serum deprivation by using a second assay for apoptosis, caspase-3 activity (Figs. 12, A and C). Once again the EETs protected the cells, including in the presence of indomethacin, but this action was blocked by wortmannin. There was also a significant decrease in caspase-3 activity after treatment with anti-Fas antibody in both HLMVECs (Fig. 12B) and HCAECs (Fig. 12D) that were pretreated with the three EET regioisomers, while 5,6-EETs did not alter caspase-3 activity. Some of these experiments were carried out with an esterified derivative of 5,6-EET, which is more stable than the free acid while retaining biological activity. The esterified derivative of 5,6-EET was no more effective in protecting cells against caspase-3 activity than the native lipid.

Fig. 12.

Caspase-3 activity. HLMVECs (A and B) and HCAECs (C and D) were serum starved or treated with anti-Fas antibody (100 ng/ml) after pretreatment with vehicle or the 4 regioisomers of EETs or in the presence of inhibitors as labeled. Caspase-3 activity was measured as described in materials and methods. *Significant difference (P < 0.05) compared with serum-deprived cells or after anti-Fas antibody treatment. Controls for serum-deprived cells were cultured in EGM-2MV, while those for Fas-induced apoptosis were not treated with anti-Fas antibody. HLMVECs and HCAECs showed increased caspase-3 activity after serum deprivation or engagement of the Fas receptor, which was attenuated by 14,15-, 11,12-, or 8,9-EET, but not 5,6-EET. This action was not seen in presence of the PI-3 kinase inhibitor WM but was not altered by the cyclooxygenase inhibitor IM.

DISCUSSION

Our results demonstrate that 14,15-, 11,12-, and 8,9-EETs significantly improve survival of human lung as well as coronary endothelial cells in culture in a concentration-dependent manner. The maximal protection we detected (45%-87%) was observed in both types of human endothelial cells compared with vehicle-treated controls or cells treated with the 5,6-regioisomer. Cell numbers decreased by 24 h after removal of growth factors and serum due to cell death, which continued to progress during the next 24 h. Growth factor deprivation is known to trigger the intrinsic pathway of apoptosis. We observed significant increase in two indicators of apoptosis, annexin V binding and caspase-3 activation, which were reduced by regioisomers of EETs that rescued cell number, indicating the role of EET in blocking the intrinsic pathway of apoptosis to enhance cell viability. Our results also provide preliminary evidence that the protection by EETs was mediated by the PI-3 kinase pathway but not by detectable activation of MAPK or detectable increase in [Ca2+]i in these two types of human endothelial cells. The intrinsic pathway of apoptosis could not be as effectively rescued by diHETEs (300 nM), nor was protection by 14,15-EET in HLMVECs or 8,9-EETs in HCAECs blocked by inhibition of cyclooxygenase. Caspase-3, an enzyme activated late in apoptosis, can be involved in both the intrinsic as well as extrinsic pathways.

We were therefore interested to determine if EETs also attenuated apoptosis induced by the extrinsic pathway and tested action of all four regioisomers on HLMVECs and HCAECs treated with anti-Fas antibodies. Once again, 14,15-, 11,12-, and 8,9-EETs but not 5,6-EET partially blocked apoptosis in both cell types. The lack of response by 5,6-EET prompted us to consider that the lipid may have been hydrolyzed before it reacted with the cells, so we repeated experiments by using two fresh batches of reagent, but observed similar results. We also applied a more stable methyl ester of this lipid synthesized in the laboratory of Dr. Falck, and once again we failed to demonstrate significant protection. In unrelated experiments, the same batches of 5,6-EET did increase proliferation of human coronary artery smooth muscle cells and was the only EET regioisomer from the four tested to do this (results not shown). It is still, however, possible that endothelial cells hydrolyze this regioisomer, very rapidly rendering it inactive.

Recombinant human epoxygenases CYP2C8 (16) and CYP 2C9 (31, 41) and EETs (15, 34, 42, 53, 58) induce proliferation and differentiation of endothelial cells, promoting angiogenesis. Survival of endothelial cells, as well as regulation of apoptosis, plays an important part in angioadaptation of the vascular system (56) under steady-state conditions, as well as during vascular development and remodeling. EETs are found in (15, 31) as well as synthesized by endothelial cells (44), and free/esterified 14,15-, 11,12-, and 8,9-EETs have been detected in human plasma (52). We hypothesized that one of the roles of EETs at physiologically relevant concentrations may be to maintain survival of endothelial cells as well as promote angiogenesis by protecting these cells during tissue dissolution and differentiation, which are important steps in the formation of new vessels from existing ones.

As with other metabolites of AA (e.g., thromboxane and prostaglandins), the actions of the CYP metabolites could be mediated via receptors. Although no specific EET receptor has been purified, evidence for a high-affinity binding site on mononuclear cell membranes specific to 14(R),15(S)-EET that is attenuated by cholera toxin and stimulates intracellular cAMP has been reported (54, 55). More recently, 11,12-EET stimulated ADP-ribosylation in bovine vascular smooth muscle cells, but not endothelial cells, in a GTP-dependent and G protein-coupled receptor Gsα-dependent manner, to activate K+ channels (26, 27). Transcription of tissue plasminogen activator (t-PA) by EETs is also mediated by Gs (40), and interestingly, release of t-PA in human endothelium is stimulated by bradykinin (7), which is known to release EETs in HCAECs (33). 8,9-EET (100 nM) induced DNA synthesis in rat cerebral capillary endothelial cells to the same extent as VEGF (1 nM), and this action was abolished by the tyrosine kinase inhibitor genistein. Further, 14,15- (31) and 8,9-EET (42, 58) were able to induce angiogenesis when embedded in matrix plugs in rats and mice in vivo. Despite these biological functions of EETs, purification of receptors has been challenging and remains a necessary step toward deciphering the molecular pathways regulated by these lipids. It remains to be tested if the antiapoptotic effects of EETs we observed in endothelial cells is receptor mediated or the EETs act as intracellular second messengers.

11,12-EET (100 nM) has been shown to stimulate a number of signaling pathways in human coronary endothelial cells. It activates tyrosine kinases and phosphatases, ERK1/2, and p38 MAPK, induces expression of cyclin D1, and inhibits c-jun NH2-terminal kinase activity in human endothelial cells (16, 41). However, it is not clear how long the cells used in some of these studies were deprived of serum (16). EETs at a concentration of 0.5 μM and higher were recently observed to stimulate DNA synthesis in mouse pulmonary artery endothelial cells after prolonged (>48 h) serum starvation (42). We did not test the lipids for activation or ERK or incorporation of thymidine at this higher concentration in human endothelial cells. Interestingly 8,9- and 11,12-EET stimulated p38 MAPK, whereas 5,6- and 14,15-EET induced survival via the PI-3 kinase pathway. We did not observe activation of the ERK pathways by treating 24-h serum/growth factor-starved HLMVECs or HCAECs with 14,15-EETs but demonstrated phosphorylation of ERK by treating the cells with growth factors and serum present in complete medium (Fig. 7). It is possible that the cells in our experiments were already programmed to die (as seen by a decrease in cell count, Fig. 1), and cultured human cells may be more susceptible to apoptosis by serum deprivation than mouse cells that could still proliferate after growth arrest (42).

Signaling, the fate of a cell to survive, proliferate, or differentiate, is controlled by transient biochemical cascades that may ultimately alter enzymatic activity and gene expression, responses that are often regulated by calcium. An array of homeostatic and sensory mechanisms exist in most cells to shape calcium signals in space and time and channel them to carry out specific cellular responses (6, 29). For example, the lipid derivative sphingosine-1-phosphate has been reported to act as an intracellular lipid messenger, regulating calcium mobilization, cell growth, and survival (51). Endothelial nitric oxide synthase that protects vascular endothelial function is also regulated by [Ca2+]i.

Others have reported that EETs increase [Ca2+]i in endothelial cells (19, 37). Induction of epoxygenase in endothelial cells (19, 22) causes capacitative calcium entry and membrane hyperpolarization. 11,12-EET has a similar effect in cultured porcine aortic endothelial cells and the human endothelial cell line EA.hy926 (37). Recovery of depleted [Ca2+]i by 11,12- and 8,9- but not 14,15- and 5,6-EET has been reported in smooth muscle cells from hamster vas deferens and implicated to regulate transition from a stationary to proliferative growth state (18). We observed a sustained rise in [Ca2+]i in HLMVECs and HCAECs after application of 15 μM 14,15-, 11,12-, and 8,9- but not 5,6-EETs. Although we initially hypothesized that EETs would increase calcium when applied to the endothelial cells at concentrations up to 300 nM, we did not observe this rise in [Ca2+]i until we reached a concentration of at least 15 μM. This was much higher than the amount of EETs needed to increase survival of human endothelial cells from serum deprivation or apoptosis induced by anti-Fas antibody. Thus we do not believe that the ability of EETs to increase [Ca2+]i plays a role in mediating its protective effects on endothelial cell survival, but we cannot exclude that this effect may be caused by localized increase in [Ca2+]i that may initiate intracellular signals before being incorporated into the endoplasmic [Ca2+]i stores and therefore not detected in our assay.

Although 14,15-EET has been reported to inhibit apoptosis of cultured proximal tubule-like epithelial cells LLCPKc14 (10), this effect was observed at a concentration of 10 μM, which is above the physiological concentration that is present in HCAECs (33). Also 14,15-EET was the only regioisomer tested and described to have a protective role in these cells against apoptosis induced by serum withdrawal, etoposide, hydrogen peroxide, or excess free AA. The effect of 14,15-EET was determined by assaying DNA laddering, Hoechst staining, and annexin V binding and was mediated via the PI-3 kinase-Akt signaling pathway. We observed similar findings: protection of human endothelial cells from apoptosis by EETs that was mediated by the PI-3K pathway. Future studies analyzing the effect of EETs on each step in this pathway at a molecular level, which is beyond the scope of this study, are needed to define the exact mechanism of action of EETs in survival of endothelial cells. The only other example of antiapoptotic activity of EETs (100 nM of 14,15-, 11,12-, or 8,9-EET, while 5,6-EET was not tested) has recently been reported in human carcinoma cells that are programmed to promote tumor formation and resist cell death (25). Apoptosis plays an important role in development and normal physiology, as well as pathology, so that the role of EETs in suppressing apoptosis contributes to the existing list of biological functions of these lipids. This is a first step in characterizing a prosurvival and antiapoptotic activity of EET regioisomers in endothelial cells.

Many enzymes that catalyze formation of EETs generate more than one regioisomer, resulting in a characteristic profile of EETs in each tissue bed, the net effects of which will be ultimately responsible for mediation of biological activity. Determining whether the synergistic effects of EETs on vascular protection may be greater or lesser than of each regioisomer remains to be investigated by using combinations of EETs.

We conclude that in addition to angiogenesis and anti-inflammatory functions such as induction of t-PA expression (40), inhibition of NF-κB and VCAM expression (39), inhibition of smooth muscle migration (49), and protection against ischemic injury (2, 20, 46), physiological concentrations of EETs also protect endothelial cells from programmed cell death. This report is the first analyses of all four regioisomers of EETs as protective agents against apoptosis of human endothelial cells and inhibition of caspase-3 activity in any cell type. Impressively, application of a physiological concentration of one of three regioisomers, 14,15-, 11,12-, or 8,9-EET, can rescue HLMVECs as well as HCAECs from apoptosis induced by either the intrinsic or extrinsic pathways. Preliminary investigations into the mechanism for this protection suggests that EETs may function via the PI3-kinase-Akt survival signaling pathway. Our findings have potentially important implications for therapeutic application of these naturally occurring lipids to promote maintenance of endothelial function as well as cell survival during acute lung injury or coronary disease.

GRANTS

Financial support was provided by National Heart, Lung, and Blood Institute Grants HL-069996 (M Medhora), HL-49294 (E. R. Jacobs), and HL-68627 (E. R. Jacobs) and a Veterans Affairs Merit Review Award (D. D. Gutterman). B. Lopez was supported as a Fulbright Scholar awarded by the Fulbright Commission/Spanish Ministry of Education, Culture and Sports (MECD Award FU2003–0973).

Acknowledgments

Excellent technical help from Stephanie Gruenloh for culturing the cells and carrying out the survival studies is gratefully acknowledged. We thank Dr. John Falck and his laboratory at the University of Texas, Southwestern Medical Center, Dallas, TX, for synthesizing and providing EETs to confirm our results.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract