INTRODUCTION

TOP ABSTRACT INTRODUCTION EET HOMOLOGS:... MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EPOXYEICOSATRIENOIC ACIDS (EETs) are endothelium-dependent hyperpolarizing factor (EDHF) candidates. Similar to EDHFs, EETs are synthesized in and released from vascular endothelial cells. Similar to EDHFs, EETs dilate arteries by hyperpolarizing the membrane potential and relaxing underlying smooth muscle cells, e.g., by opening large-conductance Ca²⁺-activated K⁺ (BK_Ca) channels. Because of a high density in cell membranes and large conductances, BK_Ca channels are considered to be the most important determinant of vascular tone regulation and the major target of EDHFs. Smooth muscle cells in resistance-generating microvessels have been shown to be more sensitive to EDHF-induced relaxations than the myocytes in large conduit arteries (28). Because EDHFs are formed independently of nitric oxide (NO) synthase or cyclooxygenase activity, those EDHFs that arise from arachidonic acid are believed to be produced by endothelial oxygenases other than cyclooxygenases. One exciting prospect involves cytochrome P-450 epoxygenases. Briefly, it is hypothesized that ACh and bradykinin stimulate the formation of EDHF by making more arachidonate available for cytochrome P-450 epoxygenases (2, 29). In turn, cytochrome P-450 epoxygenases oxidize arachidonic acid to four cis-EETs: 14,15-, 11,12-, 8,9-, and 5,6-EETs, with each regioisomer consisting of an RS and/or SR optical antipode. Because multiple epoxygenase isoforms may be present in coronary arteries and may produce various proportions of EET regio- and stereoisomers, it is unclear which EET isomers are prerequisites for microvessel dilation and BK_Ca channel activation.

Multiple studies suggest that the varying amounts of EET regioisomers and enantiomers in different tissues and species reflect differences in in situ epoxygenase activity (see Ref. 44 for review). In human whole heart preparations, the total EET concentration is 60 ng/g tissue, which is only 6% of that found in rat whole liver and kidney (44). The EETs endogenous to heart are fairly evenly divided among four regioisomers, with some enantiomer specificity occurring for 14(R),15(S)-EET. Generating a similar EET profile from exogenous arachidonic acid, one cytochrome P-450 epoxygenase (CYP 2J2) is localized to the endothelium in large and small human coronary arteries (24). Tissue culture studies lend support to an endothelial localization, because EETs are synthesized by vascular endothelial cells and not by smooth muscle cells (2). However, it is unclear whether newly formed EETs remain in endothelial cells esterified to phospholipids (38, 39) or are released and stored in smooth muscle (8) and/or cardiac myocyte phospholipids (20). Thus the concentrations of EET regioisomers and enantiomers in coronary endothelium are unknown. The distribution of EETs in cardiac tissue has physiological relevance, because if EETs are stored in coronary endothelial phospholipids, then bradykinin could also stimulate, along with arachidonic acid, the release of preformed EETs and, thus, potentiate bradykinin-induced dilations (43). Unlike EDHF, such dilations would not be blocked by "arachidonate epoxygenase" inhibitors.

In contrast to the human coronary CYP 2J2 epoxygenase, a porcine coronary epoxygenase (CYP 2C8/34) forms only the 11,12-EET regioisomer from exogenous arachidonic acid (9). Moreover, -naphthoflavone induces the endothelial expression of the same cytochrome P-450 epoxygenase, which increases 11,12-EET synthesis and, less consistently, 8,9-EET synthesis. The latter finding raises the possibility that -naphthoflavone may variably induce the formation of an additional coronary epoxygenase isoform that catalyzes 8,9-EET synthesis. In contrast, the cytochrome P-450 isozyme present in human coronary endothelium is not inducible by -naphthoflavone (46). Thus species differences in coronary epoxygenases may be indicated. However, because different arachidonate epoxygenase isozymes can be readily induced by environmental factors [medications (15), diets (33), and salt intake (3)] and because of the uncertainty of the extent of genetic variability [a single amino acid substitution in cytochrome P-450 epoxygenase dramatically alters EET regio- and stereospecificity (12)], it is unclear whether these pioneering studies accurately reflect EET biosynthetic capacities in human or porcine coronary arteries. The studies emphasize that it is critical to identify the EET regioisomers and enantiomers that may be released from coronary arteries.

Besides being synthesized in situ, the EETs in vascular endothelium may also reflect unesterified EETs taken up from the circulation. All four EET regioisomers are rapidly taken up and differentially incorporated into various endothelial cell phospholipids (38, 39, 45). However, limited studies suggest that the majority of plasma EETs is primarily esterified to lipoprotein phospholipids (27 nM) and neutral lipids (2.1 nM); only 0.9 nM EET is present in the plasma in unesterified form (16). Although the esterified EETs possess regio- and stereoselectivity, the regioisomer and enantiomer composition of circulating unesterified EETs is unknown. Thus it remains to be defined how much of the EETs in coronary arteries reflects in situ synthesis or extraction from blood.

	EET HOMOLOGS: EPOXYEICOSAQUATRAENOIC ACIDS

TOP ABSTRACT INTRODUCTION EET HOMOLOGS:... MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Diets rich in fish and fish oils appear to reduce the risk of fatal heart attacks (6, 30). Eicosapentaenoic and docosahexaenoic acids are the major n-3 fatty acids in fish oils and act as substrates for cytochrome P-450 expoxygenases (36, 37). Eicosapentaenoic acid is identical to the n-6 arachidonic acid, except it has an additional double bond at C-17,C-18, counting from the COOH end. Recent studies indicate that a cloned arachidonate epoxygenase binds eicosapentaenoic and arachidonic acids with equal affinity and yet yields twice as much n-3 epoxide [17(S),18(R)-epoxyeicosaquatraenoic acid (EEQ)] as n-6 epoxide [14(S)15(R)-EET (12)]. Moreover, 17(S),18(R)-EEQ is the sole eicosapentaenoate product generated by a monkey seminal vesicle epoxygenase (26). Whether a similar arachidonate epoxygenase occurs in coronary endothelial cells has not been established; however, other tissues such as kidney and liver have the capacity to synthesize three additional EEQ regioisomers: 14,15-EEQ, 11,12-EEQ, and 8,9-EEQ (35). These EEQs are close EET homologs that simply possess an additional double bond at C-17,C-18 (Fig. 1). The diol products of these same three regioisomers are excreted in the urine of men on fish oil diets (18). In summary, arachidonate epoxygenases produce not only EET regioisomers and enantiomers, but EEQ regioisomers and enantiomers as well.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Structures of epoxyeicosatrienoic acid (EET), epoxyeicosaquatraenoic acid (EEQ), and epoxydocosatetraenoic acid (EDT) regioisomers.

Regarding vascular activity, EEQ regioisomers will first inhibit platelet aggregation at concentrations below those affecting thromboxane A₂ synthesis; however, at higher, micromolar concentrations, EEQs also inhibit thromboxane synthesis (34). Similar to EETs (48), some EEQ regioisomers are also readily converted to epoxyprostaglandin E₂ homologs (25). Because fish oil diets will reduce blood pressure (18) and because of the close structural homology of EEQs with EET vasodilators, it is important that EEQ effects on vessel diameter be tested.

Epoxydocosatetraenoic acids (EDTs) are EET homologs in which two additional carbons are inserted at the COOH end (Fig. 1). The EDTs are formed by the chain elongation of EETs taken up by microvessel endothelial cells (unpublished observations). Because of the structural similarities of EDTs and EETs, it is critical to determine whether EDTs can also potently dilate coronary microvessels.

Although the specific EETs that are formed, stored, and released from the arterial endothelium have yet to be determined, EET-induced dilations can be regioisomer and enantiomer specific. For example, 11(R),12(S)-EET, but not the two 14,15-EET enantiomers, dilates rat renal arcuate (360 µm) arteries (49). In contrast, no regioselective effects are evident in rat preglomerular arterioles (17), cat middle cerebral (250-300 µm) arteries (11), bovine small (300 µm) coronary arteries (1), and canine epicardial arteriolar (60-150 µm) coronary arteries (28). The first report on EET stereospecific dilations focused on the 8,9-EET regioisomer; 8(S),9(R)-EET but not 8(R),9(S)-EET-constricted preglomerular arterioles (17). Subsequently, stereospecific effects have also been found for the 11,12-EET regioisomer; 11(R),12(S)-EET but not 11(S),12(R)-EET dilated renal small arcuate arteries (49). Thus regioisomeric and stereospecific EET-induced dilations occur in some small arteries but not in others. Such observations raise the possibility that EETs may also stereospecifically dilate the resistance-generating microvessels (<120 um) in heart.

If EET-induced dilations are mediated by BK_Ca channel activation, then the same regio- and enantioselectivity found for dilations should occur for BK_Ca channel activation. A parallel regiospecificity in dilation and BK_Ca channel activation occurred in an arcuate small artery (49), while a parallel nonregiospecificity was present in the cat cerebral (11) and bovine coronary (1, 21) small arteries. Indicative of a high degree of structural specificity, parallelism in stereoselectivity for dilations and BK_Ca channel activation was also found in the renal arcuate small artery (49). Thus, as with EET biosynthesis, small artery responses to EETs appear to involve a wide range of regio- and enantioselectivity. Such findings raise the possibility that similar regio- and enantioselectivity may be present in microvessels.

In summary, arachidonic epoxygenases generate a great variety of EET isomers and homologs. To evaluate the structural specificity of EETs as EDHFs, we investigated the ability of EET regio- and stereoisomers to dilate canine and porcine coronary microvessels and compared their potencies with those of EEQ and EDT homologs. Initially, canine and porcine microvessels were studied, because we wanted to establish that the remarkable sensitivity found earlier for (±)EETs was not unique to the dog (28). Thereafter, porcine microvessels became the focus of the EEQ and EDT studies, because EDHFs appear to play a critical role in limiting atherosclerotic sequelae, and the pig is a well-established atherosclerosis model that responds beneficially to diets enriched in n-3 fatty acids (30). We also examined whether EET stereoisomers and EDTs activated BK_Ca channels in a rat coronary small artery. The latter vessel was selected simply because EETs appear to interact directly with the BK_Ca channel in this preparation (23). Our findings indicate a low structural specificity for EETs as EDHFs and, thereby, raise the possibility that EET homologs may also be EDHFs.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION EET HOMOLOGS:... MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Synthesis of fatty epoxides. Regioisomers of fatty epoxides were synthesized by reacting 1-¹⁴C-labeled unsaturated fatty acids with m-chloroperoxybenzoic acid, which converts cis double bonds to (±)cis-epoxides (5). Accordingly, mixtures of four EET racemic regioisomers were synthesized from arachidonic methyl ester, five EEQ racemic regioisomers from eicosapentaenoic methyl ester, and four EDT racemic regioisomers from adrenic methyl ester. The individual methylated regioisomers were isolated by normal-phase HPLC (35, 41). Because of ready -lactone formation at acidic pH, the 5,6-EET and 5,6-EEQ regioisomers were not processed further. The remaining regioisomers (Fig. 1) were saponified and isolated by normal-phase HPLC (41). When assayed by reverse-phase HPLC, each racemic regioisomer was found to be >99% free of diol hydrolysis products and other epoxides (35, 40). The molecular weights, epoxide positions, numbers of double bonds, and absence of conjugated dienes were established using ultraviolet spectroscopy and gas chromatography-mass spectrometry (35, 41). In addition, each regioisomer was subjected to acid-catalyzed hydrolysis, and the products were converted to bis(trimethylsilyl) ether pentafluorobenzyl esters. The position of the resulting vicinal diol was confirmed by electron impact, positive- and negative-ion chemical ionization mass spectrometry (38, 41).

Isolation and preparation of canine and porcine coronary microvessels. The animal protocols were approved by the University of Iowa Animal Care and Use Committee and conform with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Mongrel dogs (7 male and 11 female, 2.9-3.7 kg) were heparinized and euthanized, and their hearts were immediately harvested as described previously (28). Ventricular epicardial arterioles [60-130 µm ID, 100 ± 24 (SD) µm] were excised, and fat and excess adventitia were removed. Also, the hearts of 20 male and 29 female pigs (95.3-163.4 kg) were harvested at a local slaughterhouse and placed in Krebs solution (see below for composition) within 9 min of execution. To remove entrapped blood and developing clots, the left anterior descending coronary artery was catheterized and gently flushed with 100 ml of heparinized Krebs solution. An India ink-gelatin physiological saline solution was injected to facilitate visualization of ventricular subepicardial arterioles (19). The stained porcine coronary arterioles [50-136 µm ID, 87 ± 17 (SD) µm, ~1.5 mm long] were excised under a dissecting microscope and trimmed of fat and connective tissue. Using a desiccator and microbalance (model UMT2, Metler, Toledo, OH), we found the pig microvessels to have a dry weight of 14.7 ± 7.4 (SD) µg (n = 12).

Protocols testing fatty epoxide potency. To test arteriole viability, 50 mM (dog) or 75 mM (pig) high-KCl (isotonic) solution was applied to a vessel preequilibrated at the original in situ length for 30 min at 20 mmHg (dog) or 40 mmHg (pig) luminal pressure. After 5 min, fresh Krebs solution was added to the chamber, and the arteriole diameter was allowed to return to the original baseline value. To test epoxide dilatory potency, the arterioles were first constricted to 35-65% of the resting diameter with 48.9 ± 8.3 (SD) nM (dog) or 45.9 ± 9.6 nM (pig) endothelin-1 (Phoenix Pharmaceutical, San Francisco, CA). Individual fatty epoxides of increasing concentrations (10¹⁶-10⁶ M) were directly added to the organ chambers, and the dilation responses were determined every 3 min. In a few studies, Krebs vehicle alone was added to assess the contribution of spontaneous dilation with time. On completion of a concentration-response study, a single dose of 100 µM sodium nitroprusside or 100 µM papaverine (Sigma Chemical, St. Louis, MO) was applied to test residual dilating capacity. In a few cases in which the agonists produced full-scale dilations, the bath was filled with fresh medium, and a single, repeat dose of endothelin was administered to test whether the vessel was viable and retained the capacity to constrict.

Solution preparation and fatty epoxide dilutions. All solutions were prepared on the day of the experiment. The Krebs solution consisted of (in mM) 131.5 NaCl, 5 KCl, 2.5 CaCl₂, 1.2 MgSO₄, 23.5 NaHCO₃, 1.2 KH₂PO₄, 0.026 Na₂EDTA, and 11 glucose, pH 7.4, and was aerated at room temperature with 20% O₂-5% CO₂-75% N₂. Freshly synthesized 1-¹⁴C-labeled fatty epoxides were stored at 5 mM for up to 1 mo at 80°C in ethanol. Just before use, the fatty epoxides were diluted with ice-cold Krebs buffer and maintained over ice. Concentrations of the stock solutions and initial dilutions were checked daily by liquid scintillation counting techniques (41). The final concentration of ethanol was <0.01%.

Isolation of smooth muscle cells from rat small coronary arteries for patch-clamp studies. Single smooth muscle cells were prepared from the secondary and tertiary branches (150-300 µm ID) of rat septal coronary arteries (23). Briefly, the arteries were digested in 1.0 ml of a mixture of papain (1.3 mg/ml, 11.9 U/mg; Sigma Chemical), collagenase (0.4 mg/ml, 364 U/mg; CLS-2, Worthington Biochemical, Lakewood, NJ), 0.49 mM EGTA, 10 mM taurine, 4.16 mM 1,4-dithiothreitol (Boehringer Mannheim, Indianapolis, IN), and 0.2% (wt/vol) fatty acid-free BSA. After being gently shaken at 37°C for 40 min, the vessels were transferred to 1 ml of Krebs solution and gently triturated with a fire-polished glass pipette until completely dissociated.

Single BK_Ca channel recording. Unitary membrane currents in individual smooth muscle cells were recorded using an inside-out patch-clamp technique (13). A single BK_Ca channel was identified by its high conductance, Ca²⁺ sensitivity, and current inhibition with 50-100 nM iberiotoxin. The current from individual BK_Ca channels was recorded using pCLAMP 8 software, with an Axopatch 200B integrating amplifier (Axon Instruments, Foster City, CA). The output of the amplifier was filtered through an eight-pole low-pass Bessel filter unit at 5 kHz (model 902 LPF, Frequency Devices, Haverhill, MA) and digitized at 50 kHz (12-bit resolution; Digidata 1200, Axon Instruments). After being filled with a solution of 140 mM KCl, 1.0 mM CaCl₂, 1.0 mM MgCl₂, 10 mM HEPES, and 1.0 mM EGTA (adjusted to pH 7.4 with KOH), the resistance of the pipette (Corning 7056, Warner Instrument, Hamden, CT) was 2-5 M, while the seal resistance was typically >10 G. The (cytosolic) medium in the bath, consisting of 140 mM KCl, 81.4 µM CaCl₂, 1.0 mM MgCl₂, 1.0 mM EGTA, and 10.0 mM HEPES, with pH adjusted to 7.35 with KOH, contained 1.0 µM free Ca²⁺ on the basis of calculations using Chelator software (Theo J. M. Schoenmakers, Dept. of Animal Physiology, University of Nijmagen, Toernooiveld, The Netherlands). On excision and exposure to air, each patch was routinely perfused with 0.0, 0.2, and 1.0 µM free Ca²⁺. By finding rapid, graded increases in BK_Ca activity that were reversible and reproducible, we established that inside-out patches, and not cell-attached preparations, were being studied. The BK_Ca channel open probability (P_o) was determined using the pStat program as implemented in pCLAMP 8 software. All the channel experiments were performed at room temperature (22°C).

Statistical analysis. Concentration-response curves were curve fitted using a nonlinear regression program (Prism version 3.0, GraphPad, San Diego, CA). The same software calculated for each curve the maximal dilation, concentration that produced 50% of maximal vasodilation (EC₅₀), and Hill slope. For each epoxide and where appropriate, the maximal dilation, EC₅₀, Hill slope, and BK_Ca channel P_o were compared using one-way ANOVA plus Tukey or Dunn corrections for multiple comparisons. Whether maximal dilation was less than that from sodium nitroprusside was assessed using a one-tailed, paired Student's t-test. In all studies, P < 0.05 was considered statistically significant. Values are means ± SE unless otherwise stated.

	RESULTS

TOP ABSTRACT INTRODUCTION EET HOMOLOGS:... MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

All six EET stereoisomers dilated canine epicardial arterioles in a concentration-dependent manner (Fig. 2, A-C). As arithmetic means and SE, the EC₅₀ values of the SR and RS pairs were 4 ± 2 and 14 ± 8 pM, 64 ± 37 and 10 ± 10 pM, and 121 ± 102 and 15 ± 10 pM for 14,15-, 11,12-, and 8,9-EET, respectively. As an index of minimal detectable dilations for curves with low Hill slopes, the EC₂₀ values of the SR and RS pairs were 0.07 ± 0.04 and 0.84 ± 0.79 pM, 0.25 ± 0.13 and 0.22 ± 0.12 pM, and 1.42 ± 1.31 and 0.79 ± 0.65 pM for 14,15-, 11,12-, and 8,9-EET, respectively. Thus dilations were detectable at EET concentrations ranging from 70 fM to 1.42 pM. Moreover, the dilations were 50% complete for enantiomer concentrations of 4-121 pM.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Concentration dependency of EET regio- and stereoisomers in dilating canine (A-C) and porcine (D-F) coronary arterioles. Coronary arterioles were constricted with endothelin to 35-60% of their resting diameters, and an EET isomer was applied in a cumulative concentration fashion. Dilation of the vessel was measured using a microscope interfaced to a video micrometer. The highest vehicle concentration produced a maximum dilation of 25 ± 2% (n = 3) in dog (28) and 16 ± 5% (n = 6) in pig. At the completion of each experiment, sodium nitroprusside (SNP) was added to the vessel medium to test for residual smooth muscle responsiveness. In D, the porcine vessel was denuded of functional endothelium before its response to 14(R),15(S)-EET was examined. Values (means ± SE) represent percentage of the diameter change induced by endothelin; n, number of animals in each group.

The two optical antipodes of each EET regioisomer were indistinguishable in potency, efficacy, or Hill slope for the canine vessels (Table 1). However, in one case, the enantiomers from two different regioisomers possessed different potencies: 14(S),15(R)-EET was >100-fold more potent than 11(S),12(R)-EET. Regarding efficacies, five of the six enantiomers dilated the vessels as much as 100 nM sodium nitroprusside; however, the 14(S),15(R)-EET enantiomer produced relaxations that represented only 90% of the maximal possible dilations. Interestingly, each of the six enantiomers, as well as each of the other epoxides investigated, required concentration changes of over seven orders of magnitude to reach the maximal dilation. Thus the Hill slope was always <0.5. Finally, the average EC₅₀ for each enantiomer pair was not significantly different from EC₅₀ values reported for racemic EET mixes: 0.2 pM for (±)14,15-EET, 16 pM for (±)11,12-EET, and 10 pM for (±)8,9-EET (28). Thus no regiospecificity was detectable in the present or previous study (28). In summary, EET enantiomers and regioisomers were equally potent and efficacious in dilating canine coronary arterioles.

All six EET stereoisomers also dilated porcine coronary subepicardial arterioles in a concentration-dependent manner (Fig. 2, D-F). The EC₅₀ values of the SR and RS pairs were 3 ± 1 and 7 ± 5 pM, 30 ± 8 and 6 ± 3 pM, and 21 ± 13 and 24 ± 12 pM for 14,15-, 11,12-, and 8,9-EET, respectively. The EC₂₀ values of the SR and RS pairs were 0.02 ± 0.01 and 0.02 ± 0.01 pM, 1.58 ± 0.93 and 0.14 ± 0.11 pM, and 0.31 ± 0.16 and 1.25 ± 1.02 pM for 14,15-, 11,12-, and 8,9-EET, respectively. Thus dilations were detectable at EET concentrations ranging from 20 fM to 1.58 pM. Moreover, the dilations were 50% complete for enantiomer concentrations of 3-30 pM.

As with canine vessels, porcine vessels showed no regio- or stereospecificity in EET potency, efficacy, or Hill slope (Table 1). However, the maximal dilation by three of the six enantiomers was 88% of the sodium nitroprusside response. Therefore, there was no difference in EET responsiveness between canine epicardial and porcine subepicardial arterioles, except for a slight (12%) reduction in dilation efficacies in porcine vessels (Table 1).

Denudation of the endothelium in porcine microvessels reduced neither the potency nor the efficacy of 14(R),15(S)-EET (Table 1). If anything, endothelial denudation may have slightly raised the vessel responsiveness to 14(R),15(S)-EET, so that it matched the dilations induced by sodium nitroprusside. Thus, similar to canine epicardial arterioles (28), porcine subepicardial arterioles did not require an endothelium to mediate the dilations (Fig. 2D); i.e., the EETs acted directly on microvessel smooth muscle cells (see below). Thus, independent of the endothelium, EET enantiomers and regioisomers equally and potently dilated canine and porcine coronary arterioles.

The four EEQ regioisomers also dilated porcine arterioles in a concentration-dependent manner (Fig. 3A). The EC₅₀ values were 0.7 ± 0.4, 2 ± 2, 0.4 ± 0.2, and 0.9 ± 0.2 pM for 17,18-, 14,15-, 11,12-, and 8,9-EEQ, respectively. The EC₂₀ values were 0.02 ± 0.01, 0.47 ± 0.32, 0.02 ± 0.01, and 0.04 ± 0.02 pM for 17,18-, 14,15-, 11,12-, and 8,9-EEQ, respectively. Thus dilations were detectable at EEQ concentrations of 20-470 fM. Moreover, the dilations were 50% complete as EEQ regioisomer concentrations reached 0.4-2 pM. Interestingly, the average EC₅₀ of all four EEQ regioisomers was 0.9 ± 0.5 pM, which was slightly less than the average EC₅₀ of the six EET enantiomers tested in pig (3.7 ± 1.4 pM, P < 0.05) and in dog (7.2 ± 2.9 pM, P < 0.05), as well as of the four EDT regioisomers (5.7 ± 2.5 pM, P < 0.05). Three of the four EEQ regioisomers produced maximal dilations that represented only 90% of that achieved by 100 nM sodium nitroprusside. Thus, similar to three EET enantiomers dilating porcine subepicardial vessels, three of the EEQ regioisomers did not quite match the efficacy of sodium nitroprusside. In any case, the EETs and their EEQ homologs were potent and efficacious dilators of porcine coronary microvessels.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Concentration dependency of EEQ regioisomers in dilating porcine coronary arterioles. Coronary arterioles were constricted with endothelin. Dilations were induced by EEQ isomers (A) and 11,12-EEQ (B) in vessels with or without prior endothelial denudation and measured as described in Fig. 2 legend. Values (means ± SE) represent percentage of the diameter change induced by endothelin.

As with EETs, no regiospecificity was detectable in EEQ potency, efficacy, or Hill slope (Table 2). Moreover, the potency, efficacy, and Hill slope were indistinguishable for EETs and those EEQ homologs that differed only by an extra double bond at C-17,C-18, i.e., 14,15-EET vs. 14,15-EEQ, 11,12-EET vs. 11,12-EEQ, and 8,9-EET vs. 8,9-EEQ (Tables 1 and 2). Thus adding one extra double bond at the saturated tail end (Fig. 1) did not alter EET vasoactivity. Finally, 11,12-EEQ did not require the presence of an endothelium to dilate porcine arterioles (Fig. 3B); the potency, efficacy, and Hill slope of (±)11,12-EEQ were unchanged by prior removal of the endothelium (Table 2). Thus, as with EETs, EEQs dilated microvessels by directly relaxing vascular smooth muscle cells.

The four EDT regioisomers also dilated porcine arterioles in a concentration-dependent manner (Fig. 4). The EC₅₀ values were 75 ± 55, 12 ± 10, 53 ± 53, and 915 ± 899 pM for 16,17-, 13,14-, 10,11-, and 7,8-EDT, respectively. The EC₂₀ values were 2.10 ± 1.36, 0.12 ± 0.08, 0.03 ± 0.03, and 1.84 ± 1.79 pM for 16,17-, 13,14-, 10,11-, and 7,8-EDT, respectively. Thus dilations were detectable at EDT concentrations ranging from 30 fM to 2.1 pM. Moreover, the dilations were 50% complete at EDT concentrations of 12-915 pM. In contrast to three EEQ regioisomers, only one EDT regioisomer (16,17-EDT) produced a maximal dilation that was 15% less than that induced by sodium nitroprusside. Thus, similar to the parent EETs and their EEQ homologs, EDTs were potent and efficacious dilators of porcine coronary microvessels.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Concentration dependency of EDT regioisomers in dilating porcine coronary arterioles. Coronary arterioles were constricted with endothelin, and dilations to EDT isomers were determined as described in Fig. 2 legend. Values (means ± SE) represent percentage of the diameter change induced by endothelin.

No regiospecificity was detectable in EDT potency, efficacy, or Hill slope (Table 3). Perhaps more importantly, the three parameters differed little for EETs and the EDT homologs that represented two carbon "frame shifts," i.e., 14,15-EET vs. 16,17-EDT, 11,12-EET vs. 13,14-EDT, and 8,9-EET vs. 10,11-EDT (Tables 1 and 3). Thus adding two carbons to the COOH end (Fig. 1) did not alter EET vasoactivity.

11,12-EET enantiomers and (±)13,14-EDT also activated BK_Ca channels in inside-out patches from smooth muscle cells isolated from rat coronary small arteries [see (±)13,14-EDT in Fig. 5, A and B]. In this preparation, 50 nM (±)11,12-EET maximally increased P_o of BK_Ca channels (23). At 50 nM, 11(R),12(S)-EET, 11(S),12(R)-EET, and (±)13,14-EDT increased the P_o by 39 ± 11% (n = 6), 59 ± 20% (n = 6), and 64 ± 18% (n = 6), respectively, compared with the vehicle control (Fig. 5C). Thus 11,12-EET enantiomers and a two-carbon-longer homolog had equal efficacy in increasing P_o. It is interesting that the increased P_o was rapidly reversed when the patch was perfused with medium lacking test compound but containing fatty acid-free albumin (Fig. 5A).

View larger version (39K): [in this window] [in a new window]   Fig. 5.   Effects of (±)13,14-EDT and two 11,12-EET enantiomers on the open probabilities (Po) of large-conductance Ca2+-dependent K+ (BKCa) channels in smooth muscle cells from rat small coronary arteries. Single BKCa channel currents were recorded on inside-out membrane patches and identified as described in MATERIALS AND METHODS. Po of individual channels was determined at a membrane potential of +60 mV and with 1 µM free Ca2+ in the (cytosolic) perfusate. A: time course of Po as 50 nM 13,14-EDT was added to and removed from the perfusate. B: expanded section from A demonstrating that cytosolic 50 nM 13,14-EDT had increased the total current through a typical BKCa channel of 15 pA; c, current level when the channel was closed. C: effects of 13,14-EDT, 11(R),12(S)-EET, 11(S),12(R)-EET, and ethanol vehicle (Ctrl) on Po. Values (means ± SE) represent change above baseline Po values. *P < 0.05 compared with ethanol.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EET HOMOLOGS:... MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrated that femto- to picomolar concentrations of EETs dilated 50- to 140-µm-ID canine and porcine coronary arterioles. The EET-induced dilations were not regio- or enantioselective. Moreover, femto- to picomolar concentrations of the EEQ and EDTs were as effective as EETs in dilating coronary microvessels. On average, the EEQs were slightly more potent than EETs or EDTs in dilating porcine coronary microvessels. Such modest increases in EEQ potency may simply reflect the presence of an additional cis double bond (7). Thus EEQs and EDTs were at least as potent as EETs in dilating coronary microvessels. Moreover, EETs and their EEQ homologs dilated the arterioles by directly relaxing smooth muscle cells.

To determine whether the same structural specificity observed in microvessel dilation occurred in BK_Ca channel activations, we tested BK_Ca channels in inside-out patches from smooth muscle cells freshly isolated from rat coronary small arteries. In contrast to experiments with cell-attached preparations, our inside-out patch studies focused on interactions with intact epoxides, because EET metabolism and conversions to second messengers were eliminated by removal of the cytosol. Other inside-out patches require prior 0.5 mM GTP fortifications for EETs to activate membrane-delimited G proteins (21). However, (±)11,12-EET activated our BK_Ca channel preparation to the same extent (2.5 nM EC₅₀ with a 1.2 Hill slope factor) whether GTP was absent or present (23). Using this preparation, we found that 50 nM 11(R),12(S)-EET activated the BK_Ca channel as much as 11(S),12(R)-EET. Similar to the 11,12-EET-induced dilations in coronary microvessels, BK_Ca channel activation was not stereospecific. Moreover, the 11,12-EET homolog 13,14-EDT produced comparable BK_Ca channel activation. Thus the structural properties observed in coronary microvessel dilations were paralleled in the BK_Ca channel activations.

The present results indicate that nanomolar concentrations of EETs can dilate microvessels by interacting with the BK_Ca channel directly or via perturbations in the lipid environment as a result of solubilization of EETs in cell plasma membrane. In our inside-out patch preparation, EETs are 1,000 times more potent than the parent arachidonate in activating the BK_Ca (23). Thus the presence of an oxirane ring is clearly important for BK_Ca channel activation. The equal efficacy of EET regioisomers in dilating microvessels further suggests that the oxirane ring is adequately positioned when present between C-8, and C-18. The stereochemistry of the epoxide did not affect BK_Ca channel activation. The addition of a double bond at C-17,C-18 did not reduce EET vasoactivity; therefore, a saturated aliphatic "tail" is probably not required for EET activation of the BK_Ca channel. Finally, inserting two carbons at the COOH end did not reduce BK_Ca channel activity. Thus the length of the COOH "head" portion is not critical for EET activations of BK_Ca channels. In summary, increasing the number of cis double bonds is known to enhance fatty acid activation of BK_Ca channels (7). Moreover, the addition of a cis-oxirane ring to a polyunsaturated fatty acid increases BK_Ca channel potency 1,000-fold (23). In the present study, the combined vasodilation and BK_Ca channel data suggest that the localization and stereochemistry of the cis-oxirane ring are not critical for activation of BK_Ca channels in coronary myocytes. A similar lack of regio- and stereoselectivity was recently reported for EET inhibitions of L-type Ca²⁺ channels (4).

The mechanisms by which EETs, EDTs, and EEQs interact with rat coronary BK_Ca channels are unknown. In some coronary artery preparations, EETs stimulate ADP-ribosylation of the G protein G_s with subsequent membrane-delimited activation of -units in BK_Ca channels (10). However, in our rat coronary preparation, EETs activate BK_Ca channels in the absence of GTP (23). For this reason, we believe that fatty epoxides may attach by hydrophobic bonds on or close to the BK_Ca channels and activate BK_Ea channels by conformational changes due to hydrogen bonding between the oxirane ring and the pore-forming - and the regulatory -subunits. There are multiple hydrophobic regions in both subunits by which fatty epoxides could affect voltage-sensor and Ca²⁺-sensitive components (32). Alternatively, selective binding to -units (31) or -units at the cytosolic side of the K⁺ pore (22) could explain our observed concomitant increases in voltage and Ca²⁺ sensitivities (23). Future experiments involving cloned channels and site-directed mutagenesis will clarify the critical binding sites.

The present studies also indicate that the mechanism by which fatty epoxides induce microvessel dilations may involve more than simple direct activation of BK_Ca channels. EETs in the femto- to picomolar range dilate canine and porcine microvessels, whereas >1,000-fold higher concentrations are required to activate the rat BK_Ca channels. One possible explanation for this disparity in potencies is that the BK_Ca channels examined in the present study were isolated from coronary vessels, which are significantly larger than the arterioles used for dilation studies. BK_Ca channel activations in myocytes from small arteries of different species routinely require nanomolar or higher EET concentrations (11, 14, 21, 49). Thus arteriolar BK_Ca channels may be more sensitive to EET activation than those of small arteries. Yet EETs are more potent as dilators than as BK_Ca channel activators, even when small arteries of the same size are studied (49). Thus differences in vessel size or organ source do not appear to explain the disparity in potencies. In the present study, the Hill slopes for EETs and their homologs ranged from 0.242 to 0.411. In contrast, the Hill slope for BK_Ca channel activation by EETs in the rat small coronary arteries is 1.2 (23). One possible explanation for the different Hill slopes is that EETs dilate microvessels by binding at multiple K⁺ channels with different affinities. At femto- to picomolar concentrations, EETs may preferentially bind to K⁺ channels that are not BK_Ca channels, e.g., voltage-gated, ATP-sensitive, or small or intermediate Ca²⁺-activated K⁺ channels (20, 42). Another possible explanation for the low Hill slope characterizing microvessel dilations is that EETs interact with G proteins of limited availability and thereby alter subsequent EET binding (21). Finally, EETs may also produce microvascular second messengers, which, in turn, increase K⁺ channel activity. Clearly, exciting possibilities exist for determining the mechanisms of actions for EETs and their homologs.

In conclusion, EETs and their homologs are potent dilators of coronary microvessels and less potent activators of BK_Ca channels. Because of a low structural specificity for EETs, the EDHF target sites may also bind EEQs and EDTs. Interestingly, the EEQs are formed by the same arachidonate epoxygenases that generate EETs. Large amounts of eicosapentaenoate substrate are readily provided by fish oil diets. We speculate that, in addition to dilating coronary microvessels, EEQs may also potently reduce peripheral resistances and contribute to the hypotensive effects of fish oil diets. Thus it is important that the levels of EETs, EEQs, and EDTs present in arteries and released into the circulation be determined.