INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ENDOTHELIAL CELLS PLAY an integral role in the maintenance of vascular tone by producing vasoactive substances in response to receptor-mediated and physical stimuli (i.e., ischemia and shear stress). The most widely characterized endothelium-derived vasodilators are nitric oxide (NO), formed from L-arginine by NO synthase, and prostaglandin I₂ (PGI₂), a product of cyclooxygenase-mediated metabolism of arachidonic acid (AA) (9, 21).

In the coronary circulation of many species of animals, including humans, endothelium-dependent relaxation responses are resistant to inhibitors of NO synthase and cyclooxygenase, suggesting the contribution of factors other than NO and PGI₂ (2, 8, 13, 14, 16, 20, 27, 28, 49, 50). These NO synthase- and cyclooxygenase-independent factors have been shown to produce relaxation by hyperpolarizing smooth muscle membranes and are referred to as endothelium-derived hyperpolarizing factors (EDHFs) (5, 10, 20, 27, 28). EDHFs contribute prominently to relaxation of conduit blood vessels in pathological conditions, such as atherosclerosis, in which the biological activity of NO is diminished (5). In resistance arterioles, EDHFs contribute substantially to agonist-induced, endothelium-dependent vasodilation in nonpathological states, suggesting an important role for these factors in the physiological regulation of blood flow (10). While the chemical identity of these putative hyperpolarizing substances remains in question, a number of recent investigations suggest that they may be noncyclooxygenase metabolites of AA (2, 7, 8, 13, 20, 49, 50).

Endothelial cells can metabolize AA to vasoactive products through several noncyclooxygenase enzymatic pathways, most notably the lipoxygenase and cytochrome P-450 pathways (29). Three distinct lipoxygenases have been described in the vascular endothelium: 5-, 12-, and 15-lipoxygenase, corresponding to the carbon where the oxygenation reaction occurs. Each of these enzymes generates a stereo-specific hydroperoxyeicosatetraenoic acid (HPETE). These highly unstable compounds are rapidly reduced by cellular peroxidases to the corresponding hydroxyeicosatetraenoic acid (HETE). Cytochrome P-450 enzymes metabolize AA to four regioisomeric epoxyeicosatrienoic acids (EETs), which are rapidly hydrolyzed to the corresponding dihydroxyeicosatrienoic acids (DHETs). In addition, cytochrome P-450 enzymes form several nonstereo-specific allylic oxidation and hydroxylation products, including 19-HETE and 20-HETE.

Several studies (2, 38, 39) have examined the formation of noncyclooxygenase metabolites of AA in cell lines and blood vessels derived from the canine and bovine conduit coronary vasculature. However, this has not been done in the coronary microvasculature, where EDHFs have their largest influence on the coronary circulation.

In this study, we examined the formation and biological activity of noncyclooxygenase metabolites of AA in the coronary microvasculature of the pig, a species whose coronary circulation is structurally and functionally quite similar to that of humans (41). Moreover, pigs develop spontaneous and diet-induced atherosclerosis and are prone to heritable hyperlipidemias resembling those seen in humans (33). Thus our findings are potentially relevant to the physiological and pathophysiological regulation of the coronary microcirculation of humans.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

M199, minimal essential medium (MEM) nonessential amino acids, MEM vitamin solution, HEPES, Hank's balanced salt solution, and trypsin were obtained from GIBCO-BRL; fetal bovine serum was obtained from HyClone Laboratories; fatty acid-free bovine serum albumin was obtained from Miles Laboratories; and gentamicin was obtained from Schering. Dithiothreitol was purchased from Boehringer Mannheim, and collagenase was from Worthington Biochemical. [5,6,8,9,11,12,14,15-³H] AA was obtained from American Radiolabeled Chemicals and Amersham, and 12(S)-HETE, 12(S)-HPETE, 13(S)-hydroperoxyoctadecadienoic acid [13(S)-HPODE], and 13(S)-hydroxyoctadecadienoic acid [13(S)-HODE] were purchased from Cayman Chemical. Cinnamyl-3,4-dihydroxy--cyanocinnamate (CDC) and 17-octadecynoic acid (17-ODYA) were purchased from BioMol. Polyclonal anti-12-lipoxygenase antibody (platelet type) was obtained from Oxford Biomedical Research, and Texas Red-conjugated goat anti-rabbit IgG was purchased from Molecular Probes. All other chemicals and compounds were purchased from Sigma.

Endothelial cell cultures. Studies were performed in a line of porcine coronary microvessel endothelial cells (MVEC) isolated from porcine coronary resistance vessels ~100 µm in diameter, as described previously (52, 53). Porcine conduit artery endothelial cells were isolated and identified as reported previously (48, 50). All cells were grown in M199 media supplemented with MEM nonessential amino acids, MEM vitamin solution, 15 mmol/l HEPES, 2 mmol/l glutamate, 50 µmol/l gentamicin, and 10% fetal bovine serum. The cultures were maintained until confluence at 37°C in a humidified atmosphere containing 5% CO₂. Stocks were subcultured weekly by trypsinization and passaged into six-well plates. All experiments were performed with cells between passage 5-7 unless otherwise specified.

Incubation of endothelial cells with radiolabeled AA. Confluent cultures were washed and then incubated for the indicated times with AA and radiolabeled AA in 1 ml of serum-free M199 containing 0.1 µmol/l fatty acid-free bovine serum albumin and supplemented as described above. When employed, all inhibitor compounds or vehicles were applied continuously beginning 1 h before incubation with AA. All incubations were conducted at 37°C in a 5% CO₂ incubator. The final concentrations of vehicles in the medium in all experiments were <0.1%. In some experiments, after incubation with AA, the cells were washed with AA-free medium and then incubated for 30 min with the calcium ionophore A-23187 (2 µmol/l). In other experiments, cells were pretreated with various concentrations of 13(S)-HPODE or 13(S)-HODE for 10 min and then incubated for 30 min with AA and [³H]AA along with 2 µmol/l A-23187.

Analyses of medium-associated lipids. AA metabolites were analyzed by reverse-phase HPLC using a Gilson system equipped with an automatic sample injector and a 2.1 × 150-mm C₁₈ Alltech column, as described previously (6, 48). Normal-phase HPLC was performed using an isocratic gradient with a solvent system consisting of n-heptane-isopropanol-glacial acetic acid (100:1:0.05) at a flow rate of 4 ml/min over 60 min. Chiral-phase separation was performed using a 5-µm N-(3,5 dinitrobenzoyl)-phenylglycine column and a mobile-phase gradient with a solvent system of n-heptane-isopropanol (99.6:0.4) at a flow rate of 0.7 ml/min. The standards were detected by absorbance at 235 nm. In some experiments, before analysis by HPLC, aliquots of the medium-associated lipids were chemically derivatized by methylation with diazomethane followed, in some cases, by acetylation with pyridine and acetic anhydride, as described previously (11).

Immunohistochemical detection of 12-lipoxygenase in porcine coronary microvessels. Hearts were obtained from pigs weighing 75-150 kg within 10 min of exsanguination at local slaughterhouses. The left anterior descending, left circumflex, and right coronary arteries were selectively cannulated and vigorously perfused with ice-cold Hank's balanced salt solution supplemented with 10,000 units penicillin, 10 mg streptomycin, 10,000 units heparin, and 0.1 g HEPES/l (pH 7.35) until the effluent was clear. After transport to the laboratory, the hearts were injected with a 2% solution of india ink in porcine skin gelatin (0.36 g/10 ml) to assist in visualization of the arterioles. Approximately 6-8 ml of the india ink solution was selectively injected into the left anterior descending and circumflex coronary arteries, and 4-6 ml was injected into the right coronary artery depending on the weight and size of the organ.

Isolated microvessel preparation. An isolated pressurized microvascular preparation was used to study coronary microvessels, as described previously (34, 35). Briefly, porcine coronary microvessels were dissected as described above and transferred to an organ chamber, which was placed on the stage of an inverted microscope equipped with a video camera, monitor, and calibrated video caliper. The microvessels were cannulated and pressurized to 80 mmH₂O under no-flow conditions. Oxygenated (20% O₂-5% CO₂-balance N₂), warmed (37°C) Krebs solution [containing (in mmol/l) 131.5 NaCl, 5 KCl, 2.5 CaCl₂, 1.2 MgCl₂, 23.5 NaHCO₃, 1.2 KH₂PO₄, and 11 glucose; pH 7.4] was continuously circulated through the organ chamber. Microvessels were allowed to equilibrate for 30 min, after which they were submaximally preconstricted with graded doses of endothelin (4-10 nmol/l), isotonic KCl (35-50 mmol/l, prepared by substituting an equimolar amount of KCl for NaCl), or acetylcholine (0.1-10 µmol/l).

Membrane potential recordings. Changes in porcine coronary microvessel smooth muscle potentials were measured using a similar experimental protocol to that used for microvessel reactivity. Porcine microvessels were obtained as described above, transected longitudinally, and immobilized luminal side up in a temperature-regulated chamber (series 20, Warner Instruments) on the stage of an inverted microscope. Membrane potentials were recorded using conventional microelectrode techniques, as previously described (40). The vessels were continuously superfused with warmed (37°C), oxygenated Krebs solution and impaled with glass capillary micropipettes with tip resistances of 75-110 M when filled with 3 M KCl. The microelectrodes were connected to a high-input impedance preamplifier (Axoclamp-2A, Axon Instruments), and the recording chamber was grounded with an Ag/AgCl pellet. Transmembrane potentials were displayed on an oscilloscope, acquired, and stored on an 80486-based computer using pCLAMP 5.5 (Axon Instruments) software with data filtered at 1 kHz. The criteria for successful impalement were the following: 1) intracellular recordings between 40 and 60 mV (30), 2) membrane potential returned to zero on removal of the pipette from the tissue, 3) >15 mV depolarization on application of 50 mmol/l KCl before beginning the experiment and a second equal depolarization to 50 mmol/l KCl at the end of the experiment (to ensure tissue viability and integrity), and 4) the same impalement was maintained throughout the experiment. The membrane potential was stable for at least 10 min before beginning the experiments.

Single K+ channel recordings. Single coronary artery smooth muscle cells were prepared from small porcine coronary arteries using previously described methods (3). Briefly, septal coronary arteries with their secondary branches (150-300 µm in intraluminal diameter) were carefully dissected from the left ventricle and isolated free of the surrounding myocardium and connective tissue under a dissection microscope, as previously described (19). Microvessels were placed into 1 ml of ice-cold cell dissociation solution [consisting of (in mM) 110.0 NaCl, 5.0 KCl, 0.16 CaCl₂, 2.0 MgCl₂, 10.0 HEPES, 10.0 NaHCO₃, 0.5 NaH₂PO₄, 0.5 KH₂PO₄, 10.0 glucose, 0.49 EGTA, and 10.0 taurine; pH 6.9 (buffered with NaOH)] to which 1.3 mg/ml (11.9 U/mg) papain, 0.64 mg/ml dithiothreitol, 0.4 mg/ml collagenase, and 0.2% (wt/vol) bovine serum albumin were added. After 30 min of gentle shaking at 37°C, the vessels were transferred into 1 ml of Kraftbrühe solution [containing (in mM) 70.0 KOH, 40.0 KCl, 50.0 L-glutamic acid, 20.0 taurine, 0.5 MgCl₂, 1.0 K₂HPO₄, 0.5 EGTA, 10.0 HEPES, 5.0 creatine, 5.0 pyruvic acid, and 5.0 Na₂ATP; pH 7.38 (buffered with KOH)] and then gently triturated with a fire-polished glass pipette until completely dissociated. The resulting smooth muscle cell suspension was stored at room temperature and used within 8 h of isolation. All solutions were vigorously oxygenated for 30 min before starting the experiment.

(1)

Statistics. Results are reported as means ± SE. Data from two groups were analyzed by unpaired or Student's t-tests as appropriate. Comparison of multiple treatment groups was performed by repeated measures analysis of variance followed by a Newman-Keuls post hoc analysis. Probability values of 0.05 or less were considered to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Metabolism of AA by MVEC. We investigated AA metabolism by MVEC using several different incubation protocols. In all experiments, the most abundant product identified in the medium was unmodified AA. Several products were also detected which comigrated with authentic prostaglandins and their metabolites, the production of which was blocked by pretreatment with indomethacin (10 µmol/l) (Fig. 1). In addition, a major metabolite, hereafter referred to as "unknown," was observed, which comigrated with authentic 12-HETE. The metabolite was formed in a time-dependent fashion, with detectable amounts of the product observed as soon as 15 min after exposure to AA (Table 1). Pretreatment with clotrimazole (10 µmol/l) or 17-ODYA (10 µmol/l), inhibitors of cytochrome P-450, did not inhibit formation of the unknown metabolite (data not shown). However, the lipoxygenase inhibitor CDC (10 µmol/l), a hydroxy-cinnamic acid derivative, eliminated its production (Fig. 2). CDC also appeared to decrease the production of prostaglandins in some experiments; however, this was inconsistent and not quantified. As has been reported for noncyclooxygenase metabolites in other cell lines, detection of the unknown metabolite in the incubation medium was observed to be diminished as the passage number of the cells increased, with minimal amounts of the product detected in incubations with cells beyond passage 7. In contrast with MVEC, conduit porcine coronary artery endothelial cells at all passage numbers formed very little of this unknown product under identical incubation conditions (data not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Metabolism of arachidonic acid (AA) by microvessel endothelial cells (MVEC): effects of the cyclooxygenase inhibitor indomethacin. Cells were pretreated with vehicle (top) or 10 µmol/l indomethacin (bottom) for 1 h and then incubated with 5 µmol/l [3H]AA for 16 h. After the medium was removed, the cells were washed and incubated with 2 µmol/l A-23187 for 30 min. The lipids contained in the medium were extracted, separated by reverse-phase HPLC, and then assayed for radioactivity. Similar chromatograms were obtained from two other identically treated sets of cultured cells in both groups. The main radiolabeled products in the medium were AA, prostaglandins (PGs), and an unknown metabolite. The peak, which eluted from the column just before AA (bottom), was inconsistently observed and not formed in sufficient quantities to permit identification. Retention times for AA and 12-hydroxyeicosatetraenoic acid (HETE) standards were 38.3 and 25.3 min, respectively. dpm, Disintegrations per minute.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Metabolism of AA by MVEC: effects of the lipoxygenase inhibitor cinnamyl-3,4-dihyroxy--cyanocinnamate (CDC). Cells were pretreated for 1 h with vehicle (top) or 10 µmol/l CDC (bottom) and then incubated with 5 µmol/l [3H]AA and 2 µmol/l A-23187 for 30 min. The lipids contained in the medium were then extracted and separated by reverse-phase HPLC using a different column and gradient than those employed in Fig. 1. The chromatograms are representative of those obtained from 3 experiments performed under identical conditions. Retention times for AA and 12-HETE standards were 52.1 and 37.4 min, respectively.

Characterization of unknown AA metabolite as 12(S)-HETE, a lipoxygenase product. To investigate whether the unknown metabolite produced by MVEC might be a -oxidation product of AA (11), experiments were performed as described above except that cells were incubated with [1-¹⁴C]AA rather than [³H]AA, which has ³H present at carbons 5, 6, 8, 9, 11, 12, 14, and 15. Formation of the metabolite was observed with [1-¹⁴C]AA, indicating that it retained the carboxyl carbon of AA (data not shown). This observation indicated that the unknown metabolite was not a chain-shortened product formed through -oxidation.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Comigration of the unknown metabolite with 12-HETE by normal-phase HPLC. After initial separation by reverse-phase HPLC, the unknown metabolite was fraction collected and subjected to normal-phase HPLC using an isocratic gradient with a solvent system consisting of n-heptane-isopropanol-glacial acetic acid (100:1:0.05) at a flow rate of 4 ml/min over 60 min. Top: separation of authentic radiolabeled 12-HETE and 15-HETE standards; bottom: representative chromatogram of the unknown metabolite. The product was resolved into a major peak, which comigrated with authentic 12-HETE, and a minor peak, which eluted from the column just before 12-HETE. This peak was not present in sufficient quantities to permit identification. The other radiolabeled peak is the solvent front (SF).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Chiral analysis of 12-HETE produced by MVEC. Chiral separation of 12(S)-HETE and 12(R)-HETE standards was obtained with a 5-µm N-(3,5-dinitrobenzoyl)-phenylglycine column and an isocratic mobile phase. Absorbance was measured at 235 nm (top). Inset: separation of both enantiomers when mixed together before injecting onto the column. Bottom: authentic radiolabeled 12(S)-HETE, the unknown 12-HETE metabolite produced by MVEC (unknown), and equivalent amounts of both radiolabeled compounds (Mix) were subjected to the same chiral separation as described in top except that radioactivity was detected with an on-line flow scintillation counter.

View larger version (87K): [in this window] [in a new window]   Fig. 5.   Immunohistochemical localization of 12-lipoxygenase protein in porcine coronary microvessels. Microvessels were dissected, fixed, sectioned, and incubated with a polyclonal anti-12-lipoxygenase antibody (A) or stained with Verhoeff-van Gieson's stain (B). The tissue sections were then examined by confocal laser scanning microscopy (A) or phase-contrast microscopy (B) at ×60 (original magnification). Pictures shown are representative of results obtained with multiple sections of microvessels taken from 3 hearts.

Effects of 13(S)-HPODE on production of 12-HETE by MVEC. In primary cultured airway epithelial cells and cultured neuronal HT22 cells, 12-lipoxygenase activity is modulated by the oxidation status of the cells (18, 42). Shornick and Holtzman (42) demonstrated that a brief exposure of airway epithelial cells to a lipid hydroperoxide, 13(S)-HPODE, dramatically increased 12-lipoxygenase activity. To investigate whether 12-HETE release by MVEC might be accentuated by treatment with 13(S)-HPODE, MVEC at passage 9 were exposed to graded concentrations (0.1, 1, and 10 µmol/l) of the compound for 10 min. The cells were then incubated with 5 µmol/l [³H]AA for 15 min, after which lipids contained in the medium were extracted and separated by reverse-phase HPLC. In vehicle-treated cells, very little radiolabeled 12-HETE was detected in the medium, a finding that was consistently observed when cells were beyond passage 7 (Fig. 6). Treatment with 13(S)-HPODE resulted in concentration-dependent increases in the release of 12-HETE to levels equal to or greater than those produced by early passage cells. As was observed by Shornick and Holtzman (42) in airway epithelial cells, formation of prostaglandins was diminished by 13(S)-HPODE. When 13(S)-HPODE (10 µmol/l) was incubated with [³H]AA in the absence of MVEC, no 12-HETE was produced, excluding a nonspecific oxidation process (data not shown). Incubation of MVEC with 1-10 µmol/l 13(S)-HODE, which is structurally identical to 13(S)-HPODE except that the hydroperoxy group is reduced to a hydroxyl group, did not alter release of 12-HETE or prostaglandins (data not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Concentration-dependent increases in 12-HETE production by MVEC after treatment with 13(S)-hydroperoxyoctadecadienoic acid (HPODE). Cells at passage 9 were exposed to vehicle or 13(S)-HPODE (0.1-10 µmol/l) for 10 min and incubated with 5 µmol/l [3H]AA for 15 min, and the lipids contained in the medium were then extracted and separated by reverse-phase HPLC. Note that in vehicle-treated cells (top), minimal 12-HETE production was detected, a finding that was consistently observed when cells were beyond passage 7. The chromatograms shown are representative of those obtained from 3 experiments in each group.

Effects of 12(S)-HETE and 12(S)-HPETE on microvascular reactivity. The effects of 12(S)-HETE (n = 7) and 12(S)-HPETE (n = 4) on porcine coronary microvascular reactivity were examined. Isolated microvessels were pressurized to 80 mmH₂O of distention pressure and submaximally preconstricted with endothelin followed by extraluminal application of 12(S)-HETE or 12(S)-HPETE. Both compounds produced extremely potent concentration-dependent dilation [EC₅₀ values (expressed as log [M]) were 13.2 ± 0.4 and 13.1 ± 0.6, respectively] (Fig. 7). When microvessels were preconstricted with acetylcholine instead of endothelin, responses to 12(S)-HETE were virtually identical (data not shown). Administration of 12(S)-HETE to nonpreconstricted microvessels, however, resulted in no significant change in diameter (Fig. 7). In some experiments, the effects of 13(S)-HPODE on microvessels preconstricted with endothelin were examined. The compound produced concentration-dependent vasodilation that was less potent than that observed with 12(S)-HETE or 12(S)-HPETE [35 ± 6 and 56 ± 5% dilation at 10 and 1 µmol/l 13(S)-HPODE, respectively (n = 7)].

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Vasodilatory effects of 12(S)-HETE and 12(S)-hydroperoxyeicosatetraenoic acid (HPETE) in porcine coronary microvessels preconstricted with endothelin or KCl. Isolated porcine coronary microvessels (distention pressure 80 mmHg, no flow) were preconstricted with endothelin (4-10 nmol/l) to 41 ± 4% of resting internal diameter (120 ± 26 µm) followed by the application of 12(S)-HETE (n = 7) or 12(S)-HPETE (n = 4). In separate experiments, microvessels were preconstricted with graded doses of isotonic KCl (n = 3, 35-50 mmol/l) to 46 ± 6% of resting diameter before application of 12(S)-HETE. In other experiments, 12(S)-HETE was administered to vessels that were not preconstricted (n = 4). Values are expressed as means ± SE.

Effects of 12(S)-HETE on vascular smooth muscle membrane potential The observation that the potent 12(S)-HETE-induced microvascular dilation was blocked by preconstriction with KCl suggested a K⁺ channel-dependent mechanism of action. Therefore, we examined the effects of 12(S)-HETE at similar concentrations to those producing vasodilation on microvessel smooth muscle membrane potential. Microvessels were dissected, transected longitudinally, immobilized endothelium side up in a recording chamber, and impaled with micropipettes. The resting membrane potential in microvascular smooth muscle (90-190 µm in diameter, n = 3) was 51 ± 3.5 mV, similar to values previously reported by Miura and Gutterman (20). Application of KCl (50 mmol/l) resulted in depolarization to 32.7 ± 2.3 mV, which was rapidly reversed on washout (Fig. 8). Application of 7 nmol/l endothelin induced significant depolarization of membrane potential to 38 ± 2.4 mV. In the continuous presence of 7 nmol/l endothelin, 1 pmol/l 12(S)-HETE produced significant hyperpolarization, restoring the membrane potential to the resting value (51.7 ± 2.3 mV), whereas 1 nmol/l 12(S)-HETE induced further hyperpolarization (71 ± 3.2 mV). In contrast with these findings in coronary microvessels, the same concentrations of 12(S)-HETE did not significantly hyperpolarize conduit porcine coronary arteries (n = 3; data not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Concentration-dependent hyperpolarization produced by 12(S)-HETE in porcine coronary microvessels. Isolated porcine coronary microvessels (n = 3, 120-190 µm) were transected longitudinally, immobilized, and impaled from the luminal side with a micropipette. Membrane potentials (in mV) were continuously measured during the treatment protocol. Values are expressed as means ± SE; *P < 0.05 vs. previous baseline; #P < 0.05 vs. endothelin-treated baseline; P < 0.05 vs. previous 12(S)-HETE (1012 M) value.

Effects of 12-HETE on BK_Ca channel activity in coronary smooth muscle cells We then examined the effects of 12-HETE on BK_Ca channel activity in isolated myocytes from small coronary arteries (150-300 µm) obtained from pig hearts. Membrane patches from freshly dissociated arteries were richly endowed with BK_Ca channels, which were activated by cytoplasmic Ca²⁺. At a membrane potential of +60 mV, no channel activity was observed in the absence of Ca²⁺. However, BK_Ca channel open probability increased approximately eightfold when free Ca²⁺ was raised from 200 nmol/l to 1 µmol/l (data not shown). The BK_Ca channel openings were completely inhibited by application of 100 nmol/l iberiotoxin to the pipette solution or superfusate (data not shown).

View larger version (31K): [in this window] [in a new window]   Fig. 9.   Effects of 12-HETE on large-conductance Ca2+-activated K+ (BKCa) channel activity in membrane patches obtained from small porcine coronary arteries. Top: sequential raw current tracings of BKCa channel activities are shown at baseline, after introduction of 3 nmol/l 12(S)-HETE, and after washout of 12(S)-HETE from the perfusion chamber. The bottom tracing shows that iberiotoxin (100 nmol/l) inhibits 12(S)-HETE-induced BKCa channel activation. Currents were recorded from inside-out patches at a membrane potential of +60 mV with the bath solution containing 1.0 µmol/l free [Ca2+]. Individual channel open probabilities (Po) values are listed on each tracing. Bottom: summary effects of 5 nmol/l 12-HETE on Po of the BKCa channel. Data are presented as means ± SE (n = 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

There are four major findings of the present study. First, porcine coronary MVEC actively metabolize AA through a lipoxygenase pathway to form 12(S)-HETE. Second, release of 12-HETE by MVEC into the extracellular fluid is increased by a brief exposure to 13(S)-HPODE. Third, both 12(S)-HETE and its precursor, 12(S)-HPETE, produce potent dilation of porcine coronary microvessels in vitro. Finally, the 12(S)-HETE-induced dilation is endothelium independent and blocked by depolarizing the microvessels with KCl. Similar concentrations of 12(S)-HETE produce hyperpolarization of porcine coronary microvascular smooth muscle cells and activate BK_Ca channels in isolated myocytes from coronary microvessels, indicating that 12(S)-HETE can function as a hyperpolarizing factor in the coronary microcirculation. Together, these findings suggest that the 12-lipoxygenase pathway could contribute to endothelium-dependent, K⁺ channel-mediated dilation of the coronary microvasculature, particularly in the setting of oxidative stress.

AA is rapidly metabolized by porcine coronary MVEC to prostaglandins and a product that we chemically identified as 12-HETE. Chiral HPLC analysis indicated that the product consisted entirely of 12(S)-HETE, which is formed through lipoxygenase metabolism (29, 42). Histological studies suggested the presence of immunoreactive 12-lipoxygenase protein localized to the endothelium of intact coronary microvessels. Earlier studies (37-39) demonstrated that 12-HETE is also synthesized by human and canine epicardial coronary arteries and bovine coronary artery endothelial cells. The enzymatic origin of the 12-HETE formed by human coronary arteries was not determined, but its production was blocked by treatment with nordihydroguaiaretic acid, a nonselective lipoxygenase inhibitor (37). While no prior studies have examined the metabolism of radiolabeled AA in the coronary microvasculature, reports indicate that 12(S)-HETE is a major product of AA metabolism in murine cerebral microvessels, where it is presumably formed by lipoxygenase metabolism within the endothelium (22-24). Bacalein and nordihydroguaiaretic acid also blocked production of 12-HETE in those studies (IC₅₀ values for both inhibitor compounds were <10 µmol/l). More recently, platelet type 12-lipoxygenase expression was demonstrated in microvascular endothelial cell lines derived from the rat brain and human foreskin microvessels using reverse transcription-polymerase chain reaction and immunoblot analyses (31). Exposure of the endothelial cells to vascular endothelial growth factor resulted in a fivefold increase in the synthesis of 12(S)-HETE concomitant with increases in cell migration and proliferation. Inhibition of 12-lipoxygenase blocked 12(S)-HETE production, cell proliferation, and cell migration, whereas application of 12(S)-HETE potently and selectively stimulated cell migration. Together, these findings suggest that the 12-lipoxygenase pathway may play an important role in regulation of angiogenesis.

Increased formation of lipoxygenase products in the setting of oxidative stress has been demonstrated in several models, tissues, and species. Ischemia-reperfusion in isolated rat hearts increased 12-HETE formation (4, 25), and oxidant stress in bovine coronary artery endothelial cells stimulated the production of 15-HETE (1). Shornick and Holtzman (42) demonstrated that brief exposure to 13(S)-HPODE dramatically increased 12-lipoxygenase activity in primary cultured rat airway epithelial cells, and, furthermore, depletion or derivatization of intracellular glutathione potentiated the effects of 13(S)-HPODE. The authors concluded that 12-lipoxygenase activity in these cells is tonically regulated by a balance between lipid hydroperoxides (that stimulate) and glutathione (that inhibits) enzymatic activity.

Similar to their findings in airway epithelial cells, the formation of prostaglandins in MVEC was suppressed by pretreatment with 13(S)-HPODE, suggesting that the stimulatory effects of 13(S)-HPODE are specific for 12-HETE. It is unlikely that the 13(S)-HPODE-induced increase in 12-HETE release was simply related to enhanced substrate availability due to inhibition of prostaglandin synthesis, because indomethacin eliminated prostaglandin production without significantly increasing 12-HETE formation (Fig. 1).

To examine the possibility that nonoxidative effects of 13(S)-HPODE might mediate the increase in 12(S)-HETE release, we treated MVEC with 13(S)-HODE. 13(S)-HODE is structurally identical to 13(S)-HPODE except that it has a hydroxyl rather than a hydroperoxy group, is not a substrate for peroxidases, and does not generate free electrons. Unlike 13(S)-HPODE, 13(S)-HODE did not affect 12-HETE or prostaglandin release. The stimulatory effects of 13(S)-HPODE on 12-HETE release were observed after only a 10-min exposure. Therefore, when considered together, these results suggest that 13(S)-HPODE most likely acts through an oxidation-dependent posttranslational mechanism. Whereas hydroperoxide-induced activation of 12-lipoxygenase enzymatic activity may be responsible, as suggested by Shornick and Holtzman, other mechanisms of action are possible. For example, 12-HETE is avidly taken up by endothelial cells and incorporated into phospholipids (47). It is therefore conceivable that 13(S)-HPODE could enhance release of 12-HETE by blocking incorporation of the compound into cell lipids or by stimulating phospholipase activity. Additional studies are required to definitively establish the mechanisms by which 13(S)-HPODE stimulates 12-HETE release from MVEC and the mechanisms by which 13(S)-HPODE produces dilation of coronary microvessels.

We did not observe any MVEC products that comigrated with EETs, cytochrome P-450-derived metabolites of AA, or their diol metabolites, the DHETs. This have may been due to downregulation of cytochrome P-450 enzymatic activity, which has been reported in cultured cells (12, 24). EETs have been shown to be formed by, and/or to produce relaxation of, coronary blood vessels in several species (2, 6, 12, 13, 20, 35). The proposed mechanism of EET-induced vasorelaxation is activation of K⁺ channels, and cytochrome P-450 2C was recently identified as an EDHF synthase in conduit porcine coronary artery (2, 7, 13, 17, 20). Inhibitors of cytochrome P-450 monooxygenases have been shown to block endothelium-dependent vasodilatory responses in the coronary microcirculation of the rat and dog (8, 51). However, in some studies (46, 49), cytochrome P-450 inhibitors did not block endothelium-dependent responses. The effects of pharmacological inhibitors of lipoxygenase and cytochrome P-450 on endothelium-dependent vasodilatory responses must be interpreted cautiously. For example, metabolism of AA through either of these enzymatic pathways may simultaneously yield dilator and constrictor compounds. Furthermore, endothelial cells actively incorporate EETs and HETEs into membrane phospholipids (47, 48). Incorporation of EETs into porcine coronary arteries potentiated bradykinin-induced vasorelaxation (48), suggesting that accumulation of EETs in cell lipids may influence vascular function. Such preformed stores of EETs would not be eliminated by inhibitors of cytochrome P-450 enzymes. Thus the inability of enzymatic inhibitors to significantly alter vascular reactivity does not exclude the possibility that metabolites of these pathways could importantly influence vascular function.

12(S)-HETE and 12(S)-HPETE produced potent vasodilation and hyperpolarization of preconstricted porcine coronary microvessels, with EC₅₀ values of <1 pmol/l. Moreover, 12(S)-HETE potently activated BK_Ca channels in membrane patches from small porcine coronary arteries. These findings were surprising for several reasons. First, in rat models of hypertension, blood pressure was reduced by inhibitors of 12-lipoxygenase (32, 43), suggesting a vasoconstrictor role for this pathway. In angiotensin II-induced hypertension, however, the hypotensive actions of 12-lipoxygenase inhibitors were attributed to increases in prostaglandin I₂ production (44). Interactions between the lipoxyenase and cyclooxygenase pathways would make it difficult to interpret how a single metabolite, such as 12(S)-HETE, might contribute to the overall regulation of vascular resistance. In addition, several studies have examined the effects of 12-lipoxygenase metabolites on vascular reactivity in conduit blood vessels. For example, 12-HPETE was reported to induce dose-dependent constriction of cat coronary arteries (45), whereas 12-HETE and 12-HPETE produced little effect on norepinephrine-preconstricted rabbit aortas or U-46619-constricted canine epicardial coronary arteries (36, 39). We also found that concentrations of 12(S)-HETE that induced marked relaxation and hyperpolarization of coronary microvessels did not significantly relax or hyperpolarize epicardial porcine coronary arteries. The mechanisms responsible for the disparate effects of 12(S)-HETE on small versus large coronary arteries are unclear. However, compounds that activate K⁺ channels to produce vasorelaxation have been shown to be more potent in small compared with large arteries derived from the same species (26, 28, 35, 46). Indeed, similar findings were observed with EETs and DHETs in canine coronary vessels (35). Like 12(S)-HETE, EETs and DHETs appear to produce potent coronary microvascular dilation by activating K⁺ channels. That these compounds can each potently and selectively dilate coronary microvessels suggests that AA metabolites derived from distinct enzymatic pathways may play redundant roles in regulation of the coronary microcirculation.

In summary, our results raise the possibility that products of lipoxygenase metabolism could contribute to dilatory responses in the porcine coronary microcirculation. When considered together with our biochemical studies with 13(S)-HPODE, these results suggest that oxidative stress may preferentially stimulate 12-lipoxygenase activity in coronary microvascular endothelium, which in turn generates products that potently dilate coronary microvessels. Such a mechanism could contribute importantly to pathophysiological processes such as ischemia-induced coronary microvascular dilation. Moreover, oxidative upregulation of the 12-lipoxygenase pathway in the coronary microcirculation could serve as a mechanism to maintain endothelium-dependent coronary blood flow in conditions such as atherosclerosis, in which the biological activity of NO is diminished (15), presumably by excess free radical production.