Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Docosahexaenoate and eicosapentaenoate are the major n-3 fatty acids in fish oils. For over 20 years, dietary supplements with fish oils have been considered as indirectly cardioprotective because of their antithrombotic, antiatherosclerotic, and antihypertensive properties. Recent clinical studies suggest that docosahexaenoate may be the active agent in dietary fish oils responsible for lowering systemic blood pressure. In 1989, the daily ingestion of fish oil (5 g of docosahexaenoate and 10 g of eicosapentaenoate) was shown to reduce the blood pressure of patients with essential hypertension (Knapp and FitzGerald, 1989). These antihypertensive effects, which follow eating large amounts of fish oil containing a low docosahexaenoate/eicosapentaenoate ratio, are now widely accepted (Appel et al., 1993; Morris et al., 1993). Recently, the daily ingestion of fish containing small amounts of n-3 fatty acids (3.65 g) with a high docosahexaenoate/eicosapentaenoate ratio was also found to be antihypertensive (Mori et al., 1999). Moreover, daily ingestions of 4.0 g of docosahexaenoate reduced blood pressure in normotensive subjects (Mori et al., 2000). Together, these clinical studies raised the possibility that dietary docosahexaenoate has hypotensive effects.

Epoxyeicosatrienoic acids (EETs) are cytochrome P-450 epoxygenase metabolites of the n-6 fatty acid arachidonate, and they potently dilate coronary arterioles (Oltman et al., 1998; Zhang et al., 2001). Because bradykinin and acetylcholine stimulate the endothelial synthesis and release of EETs (Nithipatikom et al., 2000) and because EETs hyperpolarize and relax myocytes in small arteries, EETs have been hypothesized to be endothelial-dependent hyperpolarizing factors (EDHFs) (Campbell et al., 1996; Fisslthaler et al., 1999). Moreover, like EDHF-induced dilations in human vessels (Urakami-Harasawa et al., 1997), EET-induced dilations are more prominent in resistance vessels than in conduit vessels (Oltman et al., 1998). Recently, dietary docosahexaenoate (but not eicosapentaenoate) was found to enhance the blood flow in human arterioles after acetylcholine infusions in the presence of an inhibitor of nitric-oxide synthesis. Thus, dietary docosahexaenoate may enhance the release of an EDHF, perhaps by providing an EDHF precursor (Mori et al., 2000). Yet, unlike eicosapentaenoate (Needleman et al., 1979; Zhang et al., 2001), no docosahexaenoate metabolite has been reported to be a potent dilator of resistance vessels.

Early animal studies suggested that EETs hyperpolarized vascular smooth muscle cells through activation of BK_Ca channels (Hu and Kim, 1993). Because of a high density in myocyte membranes and large conductance properties, BK_Ca channels are an important determinant of vascular tone. Moreover, primary and secondary alterations in BK_Ca channel activities are common in hypertensive rats and mice (Liu et al., 1998; Pluger et al., 2000; Lohn et al., 2001). Consistent with a generalized activation of vascular BK_Ca channels, an increased whole body synthesis of EETs retards hypertension development in several rat models (Capdevila and Falck, 2001). Similarly, dietary fish oil and docosahexaenoate have hypotensive effects in rat (Hashimoto et al., 1999) and retard the development of hypertension in multiple rat models (McLennan et al., 1996; Frenoux et al., 2001).

Crude preparations of hepatic and renal cytochrome P-450 epoxygenases react with docosahexaenoate, eicosapentaenoate, and arachidonate to yield multiple EDP, epoxyeicosaquatraenoic (EEQ), and EET regioisomers, respectively, with little apparent substrate specificity (VanRollins et al., 1984, 1988). Moreover, a cloned "arachidonate" P-450 epoxygenase binds arachidonate and eicosapentaenoate with equal affinity but generates twice as much EEQ as EET (Graham-Lorence et al., 1997). Thus, like prostaglandins, the epoxides present in tissues may reflect primarily which polyunsaturated fatty acids are available to act as substrate. Interestingly, dietary fish oils cause docosahexaenoate to preferentially accumulate in cardiac tissue (Charnock et al., 1986; Watkins et al., 2001) where polyunsaturated fatty acid cytochrome P-450 epoxygenases are known to reside (Fisslthaler et al., 1999; Tsao et al., 2001).

In the present study, we tested whether cytochrome P-450 epoxygenase metabolites of docosahexaenoate (EDPs) dilated coronary arterioles and activated BK_Ca channels in myocytes from coronary arterioles and small arteries. Dilatory responses in porcine microvessels were examined, because in this species dietary fish oil is known to enhance the endothelial formation of an EDHF (Shimokawa and Vanhoutte, 1989; Nagao et al., 1995). To determine a mechanism for EDP-induced dilations, the effects of 13,14-EDP on BK_Ca channels in inside-out patches from myocytes in porcine coronary arterioles were examined. To compare potencies with EETs, the effects of 13,14-EDP on BK_Ca channels in inside-out patches from rat coronary myocytes were investigated (Lu et al., 2001). Our findings indicate that cytochrome P-450 metabolites of the n-3 fatty acid docosahexaenoate potently dilate coronary microvessels by activating BK_Ca channels in vascular smooth muscle cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Synthesis of EDPs and 13,14-Dihydroxydocosapentaenoic Acid (DHDP). A mixture of six EDP regioisomers was generated by reacting the methyl ester of [1-¹⁴C]docosahexaenoic acid with m-chloroperoxybenzoic acid that converts cis double bonds to (±)-cis-epoxides (Chung and Scott, 1974). Individual regioisomers (Fig. 1) were isolated by normal phase-HPLC and collected over ice (VanRollins et al., 1989). One regioisomer, 4,5-EDP, was not processed further because it is chemically unstable due to ready formation of -lactones. The other five regioisomers (Fig. 1) were saponified and further isolated by normal phase-HPLC (VanRollins et al., 1989). Based on reversed-phase HPLC (VanRollins et al., 1989), each isomer was >99% free of diol hydrolysis artifacts as well as other regioisomers. Molecular weights, epoxide positions, number of double bonds, and absence of conjugated dienes were established using gas chromatography-mass spectrometry and ultraviolet spectroscopy (VanRollins et al., 1989). In addition, 10 mg of synthetic 13,14-EDP was subjected to acid-catalyzed hydrolysis to yield 13,14-DHDP (Fig. 1) (VanRollins et al., 1989). A 3.0-µg aliquot of the 13,14-DHDP product was derivatized to form a bis(trimethylsilyl) ether, pentafluorobenzylester, and the position of the resulting vicinal diol was confirmed by electron impact, positive- and negative-ion chemical ionization mass spectrometry (VanRollins et al., 1996).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Structure of EDP and DHDP metabolites of docosahexaenoate.

Isolation and Preparation of 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. Details for isolating porcine microvessels have been reported (Zhang et al., 2001). In brief, hearts of seven male (118 ± 10 kg) and 12 female pigs (128 ± 4.4 kg) were harvested at a local slaughterhouse and cooled over ice. After being gently flushed with heparinized Krebs' solution to remove blood, subepicardial arterioles were perfused with an India ink solution for visualization purposes. Individual arterioles [65-135-µm range and 94 ± 17 (S.D.) µm i.d.; ~1.5 mm in length and ~16 µg of dry weight] were excised under a dissecting microscope and trimmed of fat and adventitia.

Protocol Testing the Potency of Docosahexaenoate, EDPs, and 13,14-DHDP in Dilating Coronary Arterioles. To test arteriole viability, 75 mM isotonic KCl was applied to a vessel preequilibrated at the original in situ length for 30 min at 60 mm Hg 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 dilatory potency, the arterioles were first constricted to 37 to 61% of the resting diameter using 6.6 ± 3.1 nM (S.D.) endothelin-1 (Phoenix Pharmaceutical, Inc., San Francisco, CA). Docosahexaenoate (10¹⁰-10⁴ M) or EDPs/13,14-DHDP (10¹⁶-10⁶ M) were added in increasing concentrations directly to the organ chambers, and 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. Upon completing each concentration-dilation study, a single dose of 100 µM sodium nitroprusside or 100 µM papaverine (Sigma-Aldrich, St. Louis, MO) was applied to test residual dilating capacity. In the rare circumstance where a compound produced full-scale dilation, 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.

Protocol Testing the Stability of Fatty Epoxides in Arteriole Baths. Porcine coronary arterioles (n = 2) were isolated, pressurized, and preconstricted with potassium and endothelin as described above. Each vessel was incubated 37 min (the average application time for 12 doses, each dose requiring a 3-min wait for measuring vasoactivity) with a mixture of 2.0 µM 13,14-EDP (unlabeled) and 1.0 µM [1-¹⁴C]11,12-EET (76,000 dpm); the latter tracer was synthesized as described previously (VanRollins et al., 1996) and added to assess hydrolysis specificity. After vessel responses to epoxides and sodium nitroprusside were tested, the 20-ml bath solution was transferred to 80 ml of ice-cold methanol containing ~5 mg of Na₂CO₃. Each of the vessels studied was considered representative in that they responded appropriately to each of the agonists tested.

Preparation of Docosahexaenoate, EDP, and 13,14-DHDP Solutions. All solutions were prepared on the day of the experiment. The Krebs' solution consisted of 119.7 mM NaCl, 5 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgSO₄, 23.5 mM NaHCO₃, 1.2 mM KH₂PO₄, 0.026 mM Na₂EDTA, and 11 mM glucose, pH 7.4, and was aerated at room temperature with 20% O₂, 5% CO₂, and 75% N₂. Upon being synthesized, [1-¹⁴C]EDPs and [1-¹⁴C]13,14-DHDP were stored at 5 mM concentrations for up to 40 days at 80°C in ethanol. Just before use, EDPs and 13,14-DHDP were subjected to serial dilutions with ice-cold Krebs' buffer and maintained over ice. The concentrations of the stock solutions and initial dilutions were checked daily by liquid scintillation counting techniques (Zhang et al., 2001). The final concentration of ethanol was <0.01%. In contrast to EDPs and 13,14-DHDP, sodium docosahexaenoate (100 mg of neat material; NuChek Prep, Elysian, MN) was initially suspended in 143 ml of isotonic NaCl solution and briefly stirred at 40°C until dissolved. Serial dilutions were done using a Krebs' buffer maintained at 25°C. No concentration of docosahexaenoate higher than 100 µM could be tested for vasoactivity because the high concentrations produced a cloudy Krebs' solution.

Isolation of Smooth Muscle Cells from Porcine Coronary Arterioles and Rat Coronary Small Arteries for Patch-Clamp Studies. Upon being rapidly excised, five porcine hearts were washed and suspended in a 4°C solution containing 145.0 mM NaCl, 4.0 mM KCl, 0.05 mM CaCl₂, 1.0 mM MgCl₂, 10.0 mM HEPES, and 10.0 mM glucose, previously adjusted to pH 7.4 with NaOH. Subepicardial microvessels were isolated under a dissecting microscope, and their diameters (60-120 µm) were measured using a video microscope monitor and micrometer. Single smooth muscle cells were isolated as described previously (Ye et al., 2000) but with modified digestion conditions to improve cell viability. In brief, the microvessels were incubated at 37°C for 35 to 40 min in a Ca²⁺-free Tyrode's solution, pH 7.4, containing 0.18% protease (type XXIV; Sigma-Aldrich), 0.15% collagenase (type 1A; Sigma-Aldrich), 0.12% trypsin inhibitor (type II-S; Sigma-Aldrich), 138.0 mM NaCl, 4.5 mM KCl, 0.5 mM MgCl₂, 0.33 mM Na₂HPO₄, 10 mM HEPES, and 5.5 mM glucose. Digested microvessels were washed by three transfers to fresh Krebs' solutions, pH 7.4, composed of 70.0 mM KOH, 70.0 mM KCl, 50.0 mM L-glutamic acid, 20.0 mM taurine, 0.5 mM MgCl₂, 1.0 mM K₂HPO₄, 0.5 mM EGTA, 10.0 mM HEPES, 5.0 mM creatine, 5.0 mM pyruvic acid, and 5.0 mM Na₂ATP. Single smooth muscle cells were isolated by triturating with a fire-polished glass pipette, stored at 4°C in the Krebs' solution, and used within 10 h.

1

1

Single BK_Ca Channel Recording. Unitary membrane currents in individual smooth muscle cells were recorded using a patch-clamp technique in an inside-out configuration (Hamill et al., 1981). In brief, isolated vascular smooth muscle cells were placed in a 1.0-ml chamber on the stage of an inverted microscope (CK 40; Olympus America Inc., Mellville, NY) and were superfused at 1 to 2 ml min¹ using a direct current-powered pump (model 700; Instech Laboratories, Inc., Plymouth Meeting, PA). Under these conditions, the bath solution was entirely replaced within 30 to 60 s. Borosilicate glass capillaries (Corning 7056; Warner Instrument, Hamden, CT) were used to fabricate patch pipettes. When filled with 140.0 mM KCl, 1.0 mM CaCl₂, 1.0 mM MgCl₂, 10.0 mM HEPES, and 1.0 mM EGTA, and adjusted to pH 7.4 with KOH, each electrode had a tip resistance between 4 and 10 M and a typical seal resistance greater than 10 G. Single BK_Ca channel currents were recorded with an Axopatch 200B integrating amplifier (Axon Instruments, Union City, CA), and the output of the amplifier was filtered through an eight-pole low pass Bessel filter unit (900B/9L8L; Frequency Devices, Haverhill, MA) at 5 kHz and digitized at 20 (pig) or 40 (rat) kHz (12-bit resolution, Digidata 1200; Axon Instruments). Software (pClamp, version 6.05; Axon Instruments) was used to generate the voltage-clamp protocols, and the resulting current recordings were stored in a Pentium-based personal computer (Dimension XPS T450; Dell Computer Corp., Round Rock, TX) for further analysis. A pStat program, implemented in the pClamp software, was later used to calculate the BK_Ca channel open probability (P_o):
where P_o is the single channel open probability, T is the duration of recording, t_j is the time spent with j = 1,2,... ... N channel openings, and N is the maximal number of simultaneous channel openings observed when P_o was high. An N not more than five was used for all P_o studies. Each dose-response curve was fitted using a Hill equation of the following form: P_o/P_o,max = 1/{1 + (S/EC₅₀)^H}, where P_o,max represents the maximal channel activity observed, S represents concentration of the chemical, EC₅₀ represents the concentration at half-maximal effect, and H is the Hill coefficient.

Statistical Analysis. Data are presented as mean ± S.E.M. values unless otherwise stated. Concentration-dilation curves were subjected to a nonlinear regression program (Prism version 3.0; GraphPad Software, San Diego, CA). From each fitted curve, Hill slope and concentrations that produced maximal dilation and 50% of maximal vasodilation (EC₅₀) were determined and evaluated using one-way analysis of variance followed by a Fisher correction (least significant difference) for multiple comparisons. Whether the maximal dilation was less than that induced by sodium nitroprusside was assessed using a one-tailed paired t test. Changes in P_o were evaluated using one-way analysis of variance and the Student's t test. In all studies, a p value of <0.05 was considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Each of the five EDP regioisomers dilated porcine arterioles in a concentration-dependent manner (Fig. 2A). The EC₅₀ values were 211 ± 172, 4.1 ± 2.1, 2.0 ± 1.4, 20.3 ± 13.3, and 1.6 ± 0.7 pM for 19,20-, 16,17-, 13,14-, 10,11-, and 7,8-EDP, respectively (Table 1). The dilations were thus 50% complete for EDP concentrations ranging from 1.6 to 211 pM. No EDP regiospecificity was detectable in the induced dilations. In contrast to the EDPs, the EC₅₀ for the diol 13,14-DHDP was 30 ± 22 nM (n = 7). Thus, the conversion of 13,14-EDP to 13,14-DHDP caused a 15,000-fold loss in potency (p < 0.001). As an estimate of minimal detectable responses for curves characterized by low Hill slopes (vide infra), the EC₂₀ values were 1960 ± 1630, 501 ± 403, 150 ± 132, 552 ± 240, and 199 ± 147 fM for 19,20-, 16,17-, 13,14-, 10,11-, and 7,8-EDP, respectively. Dilated arterioles were detectable at EDP concentrations ranging from 150 fM to 1.96 pM. Thus, EDPs were very potent dilators, with some EDP regioisomers starting to dilate coronary arterioles at 199 fM, and all EDPs completing 50% of the induced dilations by 211 pM.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Concentration dependence of EDPs (A) and docosahexaenoate (B) in dilating porcine coronary arterioles. Arterioles were constricted with endothelin to 35 to 60% of their resting diameters, and EDPs (A) or docosahexaenoate (B) were applied to the medium in an accumulative concentration manner. Vessel dilation was measured using a microscope interfaced to a video micrometer. The highest vehicle concentration applied produced a maximum dilation of 16 ± 5% (n = 6). At the completion of each experiment, sodium nitroprusside (SNP) was added to the vessel medium to test for residual smooth muscle responsiveness. Data are the mean percentage ± S.E.M. of the diameter change induced by endothelin, and n is the number of animals studied.

Each of the five EDP regioisomers produced slightly less than the maximal dilation induced by 100 µM sodium nitroprusside (Table 1). 13,14-EDP reached 93% of the dilation induced by sodium nitroprusside; however, the other four EDPs produced only 84 ± 2% of the maximal possible dilation. Therefore, at nanomolar concentrations all EDPs induced dilations that were 7 to 16% less than the maximal response produced by sodium nitroprusside (Fig. 2A). Interestingly, the maximal dilations were reached only after increasing the EDP concentration over 6 orders of magnitude. The Hill slope was always 0.6 and showed no regiospecificity (Table 1). Nevertheless, together the above-mentioned data indicated that EDPs were both highly potent and efficacious dilators of coronary arterioles.

Docosahexaenoate also dilated porcine subepicardial arterioles in a concentration-dependent manner (Fig. 2B). However, because of solubility problems at greater than 100 µM, the maximal response achieved for docosahexaenoate was only 80 ± 9% and was less than the 102 ± 2% obtained with sodium nitroprusside (p = 0.023). Perhaps more importantly, the apparent EC₅₀ of docosahexaenoate was 8.4 ± 2.6 µM and no dilation was detectable at less than 100 nM. Thus, the parent docosahexaenoate was 100,000 times less potent in dilating coronary arterioles than the EDP metabolites that had an average EC₅₀ of 48 pM.

Upon being added to the arteriole bath, 13,14-EDP underwent little hydrolysis over a 37-min period. Based on integration of absorbance at 194 nm, only 1.6 and 2.8% (n = 2) of 2.0 µM 13,14-EDP were converted to 13,14-DHDP, which eluted at 12.97 min (Fig. 3A, bottom). Based on integration of radioactivity, only 1.1 and 2.8% of [1-¹⁴C]11,12-EET were converted to [1-¹⁴C]11,12-DHET, which eluted at 13.40 min (Fig. 3A, top). Because the starting concentration for 11,12-EET (1.0 µM) was one-half that of 13,14-EDP (2.0 µM) and because EETs and DHETS have much lower absorbances than EDPs and DHDPs (Fig. 3B, bottom), the small amount of 11,12-DHET generated was not readily detectable at 194 nm (Fig. 3A, bottom). Interestingly, the total radioactivity recovered was only 40 and 44% of that applied but was similar to the 46 and 50% recovered from baths without arterioles. Perhaps more importantly, the ultraviolet absorbing compounds that eluted at 38.4 and 46.2 min were not 13,14-EDP metabolites but contaminants of the extracting solvents. Together, the above-mentioned data indicate that both 13,14-EDP and 11,12-EET were stable for 37 min during incubations in the arteriole baths.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   A, HPLC chromatogram of lipids in bath of coronary arteriole incubated with 13,14-EDP and 11,12-EET. In brief, 2 µM 13,14-EDP (unlabeled) and 1 µM [1-14C]11,12-EET were incubated 37 min with an arteriole. The bath solution was collected, and lipids extracted and resolved by reversed-phase HPLC. Effluent radioactivity (top) and UV absorption at 194 nm (bottom) were monitored with an on-line flow scintillation and photodiode array detector, respectively. B, HPLC chromatograms of substrates prior to addition to bath. Note that compared with 11,12-EET, which has three double bonds, 13,14-EDP with five double bonds has greater absorption at 194 nm (bottom). Unlabeled arrows depict retention times of [1-14C]11,12-DHET (top) and 13,14-DHDP (bottom) standards. Each panel contains data from a single incubation, but identical chromatograms were obtained from duplicate incubations.

One accepted mechanism mediating dilations induced by EETs-cytochrome P-450 epoxygenase metabolites of the n-6 fatty acid arachidonate involves activations of BK_Ca channels and hyperpolarization of coronary smooth muscle cells (for review, see Zhang et al., 2001). Like EETs (Lu et al., 2001), 13,14-EDP also activated BK_Ca channels in myocytes from porcine coronary arterioles in a dose-dependent manner. Increasing the concentration of 13,14-EDP recruited more and more BK_Ca channels (Fig. 4A). Under these experimental conditions, the ethanol vehicle had a maximum concentration of 0.0002% and did not activate BK_Ca channels. Channel recruitment by 13,14-EDP was associated with an increased open probability for BK_Ca channels (Fig. 4B). The open probability increased from 0.062 ± 0.022 (no EDP) to a maximum of 0.34 ± 0.09 (100 pM 13,14-EDP). Open probability climbed 6.04 ± 1.03 fold (p = 0.02) when calculated for each experiment, and 5.5-fold when means were compared. Perhaps more germane for mechanism considerations, the EC₅₀ of BK_Ca channel activation was 6.6 ± 0.6 pM (n = 5), and although statistically different (p = 0.017), was comparable with the 2 pM EC₅₀ observed for coronary arteriolar dilation. The Hill coefficient for BK_Ca channel activation was 1.1. In contrast, the Hill coefficient for arteriolar dilation was only 0.429 (Table 1). Thus, in the same arteriolar vessel, BK_Ca channel activation and dilations had the same EC₅₀ but different Hill coefficients. To our knowledge, this is the first report on eicosanoid potencies and BK_Ca channel activity in porcine coronary arterioles.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   13,14-EDP activated BKCa channels in myocytes from porcine coronary arterioles. Single BKCa channel currents were recorded on inside-out membrane patches, maintained at +60 mV membrane potential and in the presence of 200 nM cytosolic free Ca2+. A, raw tracings of BKCa channels with 15 pA currents that were exposed to increasing concentrations of 13,14-EDP (0, 0.1 pM, 10 pM, and 1 nM); c, current level observed for closed channels. B, after normalizing for different channel numbers in each patch, the Po of BKCa channels was plotted as a function of increasing concentrations of 13,14-EDP. Data represent mean Po/Po,max values ± S.E.M. (n = 5). In addition, 13,14-EDT data were fitted to a Hill equation; the EC50 value represents the EDP concentration at half-maximal effect, and H represents the Hill coefficient.

Because of the surprisingly high potency of 13,14-EDP in activating BK_Ca channels in myocytes from porcine arterioles, the effects of 13,14-EDP on BK_Ca channels in myocytes from rat small coronary arteries were also studied. Although less sensitive to calcium, the BK channels in rat coronary arteries and their responses to EETs/DHETs have been well characterized (Lu et al., 2001). As with porcine arterioles, the BK_Ca channels in rat coronary arteries were activated by 13,14-EDP in a dose-dependent manner. Moreover, increasing the concentration of 13,14-EDP recruited more and more BK_Ca channels (Fig. 5A, top). The channel recruitment by 13,14-EDP was associated with an increased probability of opening the BK_Ca channels (Fig. 5A, bottom), which climbed from 0.058 ± 0.026 (no EDP) to a maximum of 0.23 ± 0.05 (100 pM 13,14-EDP). Open probability increased 3.6 ± 0.4-fold (p = 0.02) when calculated for each experiment and 4.0-fold when means were compared. Perhaps more importantly, the EC₅₀ of BK_Ca channel activation was 2.2 ± 0.6 pM (n = 7) and similar (but statistically different, p = 0.001) to the 6.6 ± 0.6 pM (n = 5) found for porcine arterioles. Because the 1.5 Hill coefficient found for rat myocytes was similar to the 1.1 (p = 0.15) observed for porcine myocytes, there were only minor differences in BK_Ca channel responsiveness to 13,14-EDP by myocytes from rat small arteries and porcine arterioles. More importantly, 13,14-EDP was a highly potent channel activator in both preparations.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   13,14-EDP (A) but not docosahexaenoate (B) activated BKCa channels in myocytes from rat small coronary arteries. Single BKCa channel currents were recorded as described in Fig. 4, but in the presence of 1 µM cytosolic free Ca2+. Top, raw tracings of BKCa channels with 15-pA currents that were exposed to increasing concentrations of 13,14-EDP (A) and docosahexaenoate (B). Bottom, normalized Po of BKCa channels was plotted as a function of increasing concentrations of 13,14-EDP (A) or docosahexaenoate (B). Data represent mean Po values ± S.E.M. (n = 7).

Compared with 13,14-EDP, docosahexaenoate weakly activated BK_Ca channels (Fig. 5B, top). In fact, up to 10 µM docosahexaenoate (Fig. 5B, bottom) completely failed to activate BK_Ca channels (n = 7; p = 0.350). In contrast, vessel dilations were apparent at 1 µM docosahexaenoate (Fig. 2B). Thus, docosahexaenoate weakly dilated coronary arterioles at concentrations that had no effect on the BK_Ca channels in myocytes isolated from small coronary arteries.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present report, femtomolar-to-picomolar concentrations of EDPs dilated 65 to 135-µm-i.d. porcine coronary arterioles. The EC₅₀ values ranged from 1.6 to 211 pM. As found for coronary arteriolar dilations induced by epoxides of arachidonate (EETs) and eicosapentaenoate (EEQs), EDP-induced dilations were not regioselective. Moreover, the EC₅₀ value for individual EDPs was very similar to that of individual regioisomers derived from eicosapentaenoate and arachidonate (Zhang et al., 2001). Like 11,12-EET, 13,14-EDP underwent little hydrolysis during the time that a dilation was induced (0.25%/3 min). Thus, compared with endothelial cells isolated from large arteries (VanRollins et al., 1996), whole arterioles and endothelial cells from microvessels (M. VanRollins, unpublished data) may have little epoxide hydrolase activity. Interestingly, unlike 11,12-EET (Oltman et al., 1998), 13,14-EDP hydrolysis dramatically lowered the potency for arteriolar dilation. Together, these findings indicate that EDPs are some of the most potent dilators of coronary microvessels known and suggest that the induced dilations do not require prior conversions to diols.

In contrast to EDPs, the parent docosahexaenoate was only a weak dilator of coronary microvessels. The EC₅₀ value for docosahexaenoate-induced dilations was 8.4 ± 2.6 µM and was greater than that of arachidonate (0.48 ± 0.19 µM; n = 14) prepared and administered under identical conditions (p = 0.0003). Accordingly, docosahexaenoate was 16- and 160,000-fold less potent than arachidonate and EDPs, respectively, in dilating coronary arterioles. One explanation for the low dilator potency is that the abluminally presented docosahexaenoate must first penetrate the smooth muscle barrier, enter endothelial cells for epoxygenation, and be released as vasodilatory EDPs. If EDP formation mediates the docosahexaenoate-induced dilations, the marked differences in potencies between EDPs and docosahexaenoate suggest that only a small amount of newly formed EDP is released to relax underlying smooth muscle cells and dilate whole arterioles.

13,14-EDP potently activated BK_Ca channels in coronary myocytes from porcine arterioles and rat small arteries. Unlike other preparations (Li and Campbell, 1997), the porcine and rat inside-out patch preparations are similar in that they require no GTP to demonstrate BK_Ca channel activation by fatty epoxides and diols (Lu et al., 2001). Moreover, the channels responded similarly to 13,14-EDP, generating comparable EC₅₀ values and maximum channel open probabilities. Yet, the patch preparations differed in that a similar state of activation occurred in the arterioles at 5-fold lower cytosolic calcium concentrations than used for small arteries. Such a difference in calcium sensitivity suggests the presence of BK_Ca channel isoforms, which may be either vessel- or species-specific. Unfortunately, characterizing and identifying the different isoforms require the analysis of large numbers of channels (Toro et al., 1998). Moreover, quantitating the different proportions of isoforms in coronary conduits and arterioles awaits the advent of specific antibodies. Both goals are beyond the scope of the present investigation. Perhaps more important regarding whether EDPs act as EDHFs, the measured EC₅₀ for 13,14-EDP activation of BK_Ca channels was 1000-fold less than reported for epoxides and diols derived from the n-6 fatty acid arachidonate when analyzed under identical conditions (Lu et al., 2001; Zhang et al., 2001). Thus, EDPs are the most potent fatty acid epoxides known to activate BK_Ca channels (Lu et al., 2001; Zhang et al., 2001). In fact, we are unaware of any BK_Ca channel activator as potent as 13,14-EDP.

In concentrations up to 10 µM, docosahexaenoate did not activate BK_Ca channels in the smooth muscle cells from coronary small arteries. Moreover, when prepared and administered under identical conditions, arachidonate does not activate BK_Ca channels (Lu et al., 2001). In light of the comparable EDP potencies for activating the BK_Ca channels in myocytes from porcine arteriolar and rat small arteries, docosahexaenoate was probably a poor activator of arteriolar BK_Ca channels. Assuming a comparable deficiency in BK_Ca channel activation in porcine arterioles as found in rat small arteries, there would not be significant BK_Ca channel activation at the docosahexaenoate doses (EC₅₀ value of 8.4 µM) that induce arteriolar dilations. As mentioned above, the weak dilations in whole arterioles may be readily explained by cytochrome P-450 epoxygenases in endothelial cells converting a small amount of docosahexaenoate to EDPs. 13,14-EDP is a highly potent activator of BK_Ca channels in porcine arterioles, and at least 10,000,000-fold more potent than docosahexaenoate in activating the BK_Ca channels in small arteries. The inability of either docosahexaenoate or arachidonate to activate BK_Ca channels may result from the plasmalemmal patches lacking the biosynthetic machinery (cytochrome P-450 epoxygenases) for converting polyunsaturated fatty acids to epoxides. In any case, it is clear that cytochrome P-450 epoxygenases transform docosahexanoate to very potent coronary vasodilators, BK_Ca channel activators, and inhibitors of platelet aggregation (VanRollins, 1995). All three actions may contribute to the antithrombotic effects of fish and fish oil diets.

Regarding the mechanism of action for EDP-induced dilations, the EC₅₀ value of 13,14-EDP for activating BK_Ca channels matched that of arteriolar dilation. Moreover, a tendency to activate channels in the femtomolar range was indicated (Fig. 4B). Both findings suggested that BK_Ca channel activation is the predominant mechanism for EDP-induced dilations. In contrast, the EC₅₀ values for BK_Ca channel activation by arachidonate epoxides/diols are orders of magnitude greater than those for dilations (Lu et al., 2001; Zhang et al., 2001), even when the same vessel from the same species is examined (Zou et al., 1996). Such a disparity in EC₅₀ values suggests that activation of potassium channels other than BK_Ca channels may mediate the arteriolar dilations induced by epoxides and diols derived from arachidonate (Zhang et al., 2001). Thus, unlike EET vasoactivity, EDP-induced dilations may be mediated exclusively by BK_Ca channel activation.

However, the molecular mechanisms responsible for EDP-induced dilations are more complex than suggested by simply matching dilation EC₅₀ values with those of BK_Ca channel activation. Cytoplasmic elements are absent from patches in the inside-out configuration; in such simple plasmalemmal preparations, the EDPs seem to act by directly binding to the BK_Ca channels or their immediate environs. The open probability responses of the coronary channels had Hill slopes of 1.1 (pig) and 1.5 (rat) and indicate that BK channels may possess one or two EDP binding sites. Thus, in the rat small arteries, the EDPs may bind on or close to the -pores or -regulatory units of BK_Ca channels and, by hydrogen bonding with the epoxide ring, induce conformational changes favoring EDP binding to a second site (Lu et al., 2001; Zhang et al., 2001). Future binding studies with cloned BK_Ca channels and site-directed mutagenesis are needed to clarify the precise binding sites.

Surprisingly, arteriolar dilations by fatty epoxides are characterized by Hill slopes less than 1.0. The Hill slope determined for dilations by 13,14-EDP was 0.429, and the average Hill slope for the five EDP regioisomers was 0.52 ± 0.04 (n = 5). In comparison, the dilation curves of arachidonate- and eicosapentaenoate-derived epoxides had Hill slopes of 0.32 ± 0.02 (n = 6 enantiomers) and 0.39 ± 0.02 (n = 4), respectively (Zhang et al., 2001). Thus, EDP-dilation curves had Hill slopes that are 33% (p < 0.05) and 62% (p < 0.001) greater than that of arachidonate and eicosapentaenoate epoxides, respectively. Notwithstanding, the Hill slope for arteriolar dilation by 13,14-EDP was only one-third the 1.1 value measured for BK_Ca channel activation. One possible explanation for the consistently low Hill slopes characterizing arteriolar dilations is that fatty epoxides interact with G proteins of limited availability in intact vessels, and thus inhibit subsequent EDP binding (Li and Campbell, 1997). Alternatively, the threshold activities of fatty epoxides may result from interactions with several BK_Ca channel isoforms of varying affinities or with other (non-BK_Ca) potassium channels. In any case, the fatty epoxide interactions with BK_Ca channels in whole arterioles seem to involve more than a simple, direct interaction with BK_Ca channels.

In conclusion, EDPs are cytochrome P-450 epoxygenase metabolites of docosahexaenoate, one of the major n-3 fatty acids in fish oils. In the present study, 13,14-EDP activated BK_Ca channels with an EC₅₀ value of 2.2 to 6.6 pM, which is over 1000 times more potent than observed for epoxides (EETs)/diols (DHETs) derived from arachidonate. To our knowledge, no cyclooxygenase or lipoxygenase metabolite of n-3 fatty acids has been reported to be more active than the corresponding n-6 fatty acid product. As a result of the potent BK_Ca channel activation by 13,14-EDP, hyperpolarization and relaxation of vascular smooth muscle cells should follow. Consistent with this expectation, 13,14-EDP dilated coronary arterioles with an EC₅₀ value of 2 pM. EDPs are the first docosahexaenoate metabolites shown to have potent dilator actions on resistance vessels. Moreover, the EDHF-like actions make it an important goal to determine whether EDPs are synthesized and released from vascular endothelial cells in response to bradykinin and acetylcholine. An endothelial release of EDPs could explain the enhanced endothelial-dependent relaxations and hypotensive effects that accompany dietary supplements of docosahexaenoate. In addition, we speculate that an enhanced release of EDPs and EEQs may mediate reductions in myocardial infarctions (Tavani et al., 2001) and ischemia-induced arrhythmias associated with dietary fish oils (Leaf, 2001).