Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Vascular endothelium modulates blood pressure and flow by producing and releasing factors, which regulate the tone of the underlying smooth muscle. The pivotal role of endothelium-derived relaxing factor, now known to be nitric oxide (NO) generated from the metabolism of L-arginine, is well established (Palmer et al., 1987). In addition, other factors including AA and its metabolites may also contribute significantly to endothelium-mediated relaxation (for reviews, see Quilley et al., 1997; Campbell and Harder, 1999; Feletou and Vanhoutte, 1999; Quilley and McGiff, 2000). However, the molecular identity of the phospholipase(s) responsible for AA release in endothelial cells involved in the endothelium-dependent relaxation of smooth muscle has not been established.

AA can be released in vascular tissue via multiple pathways. By analogy to the other cell types that have been more extensively investigated, the majority of agonist-induced AA release in endothelial cells is likely mediated by intracellular phospholipase A₂ (PLA₂). Within the PLA₂ superfamily of enzymes, there are two distinct families of intracellular PLA₂. Both a Ca²⁺-dependent PLA₂ (cPLA₂) and a Ca²⁺-independent PLA₂ (iPLA₂) are present in mammalian cells (Wolf and Gross, 1985, 1996; Gross et al., 1993; for review, see Gijon and Leslie, 1997). In circulating cells (e.g., platelets, macrophages), the majority of intracellular PLA₂ activity appears to be mediated by cPLA₂, whereas iPLA₂ appears to be the predominant activity present in other cell types (Wolf and Gross, 1985; Miyake and Gross, 1992; Wolf et al., 1995). Recent studies indicate that both cPLA₂ and iPLA₂ can participate in agonist-induced AA release (Lehman et al., 1993; Akiba et al., 1999; Murakami et al., 1999). McHowat et al. (2001) recently described an iPLA₂ activity in endothelial cells that was selective for plasmalogen substrate and inhibited by (E)-6-(bromomethylene)tetrahydro-3-(1-naphthalenyl)-2H-pyran-2-one (BEL).

In the present study, we investigated the role of iPLA₂, iPLA₂-derived AA, and AA metabolites in acetylcholine (ACh)-induced, endothelium-dependent relaxation of rat mesenteric resistance arteries. To determine the types of intracellular phospholipases involved in ACh-mediated AA release in endothelial cells, BEL, a selective mechanism-based inhibitor, which possesses a 1000-fold selectivity for the iPLA₂ versus the cPLA₂ families of enzymes (Hazen et al., 1991), was utilized. The role of AA itself or AA metabolites was assessed utilizing inhibitors of the cyclooxygenase (indomethacin), lipoxygenase (esculetin), or cytochrome P450 monooxygenase (17-octadecynoic acid, 17-ODYA) pathways. We now report that BEL inhibits ACh-mediated vascular relaxation, which can be restored by provision of a cell-permeant AA derivative, and that iPLA₂-derived AA itself, and not an eicosanoid metabolite, is essential for ACh-mediated relaxation in these vessels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Vessel Preparation. Vessel isolation and cannulation methods were similar to those previously described (Boyle and Maher, 1995). Mesenteric tissue was removed from halothane-anesthetized Sprague-Dawley rats (250-350g) and the fourth branch of the mesenteric artery (200- to 250-µm outer diameter) was dissected at 4°C in a HEPES buffer consisting of 135 mM NaCl, 2.6 mM NaHCO₃, 0.34 mM Na₂HPO₄, 0.44 mM KH₂PO₄, 5 mM KCl, 1.4 mM CaCl₂, 1.17 mM MgSO₄, 0.025 mM EDTA, 10 mM HEPES, and 5.5 mM glucose (pH 7.35-7.40). Arteries were cannulated with two glass cannulae, mounted in a bath filled with HEPES buffer, and pressurized to 40 mm Hg transmural pressure; temperature was maintained at 34 ± 0.5°C. After an equilibration period of 60 min, the integrity of vascular smooth muscle and endothelial function was assessed by contracting the vessel with phenylephrine (PE; 1-10 µM), and then adding ACh (1 µM). Vessels with a strong uniform response to PE and in which ACh reversed the 10 µM PE-induced constriction by more than 90% for 5 min were studied. Vessels that did not respond in this manner were considered damaged and were not studied further. A camera attached to the video port on the microscope was used to visualize the artery, and a computer-based image analysis system was used to continuously monitor and record the inner and the outer vessel diameters. All solutions and drugs were added to the bath.

Effects of PLA₂ Inhibitors. In these experiments, we studied the effect of the iPLA₂ inhibitor, BEL, on 1 µM ACh relaxation of 10 µM PE-constricted arteries. After the initial PE and ACh application, BEL was added to the bath (10 min) followed by washout and reapplication of PE and ACh. The dose response to BEL was determined by repeating this sequence using increasing concentrations of BEL. A similar approach was used to test the effect of the nonspecific PLA₂ inhibitor, arachidonyl trifluoromethyl ketone (AACOCF₃). Assuming that ACh may activate iPLA₂ and thereby increase the effectiveness of the pretreatment with the suicide substrate BEL, additional experiments were conducted in which ACh (1 µM) was included with BEL during the 10-min pretreatment period. Preliminary time-control experiments in the absence of the inhibitors demonstrated that the ACh (1 µM) relaxation remained constant over 4 h of repeated ACh applications.

AA Effects on ACh Relaxation after BEL. These experiments were conducted to determine whether application of the cell-permeant analog of the iPLA₂ product AA, AA-methyl ester (AA-Me), could restore the ACh-induced relaxation attenuated by BEL pretreatment. Since fatty acids are transported across cell membranes with concomitant thioesterification to facilitate their utilization in subsequent oxidation or lipid metabolic processes, we explored the effects of AA-Me, which we envisaged would have access to intracellular membrane compartments (without thioesterification) where it would be de-esterified to AA by intracellular esterases. After demonstrating that ACh relaxation was blocked by BEL, the vessel was incubated in buffer containing AA-Me (1 pM-1 µM) for 30 min and then rechallenged with ACh (1 µM). Control experiments with ACh rechallenge after BEL were done without the AA-Me incubation. Additionally, to control for potential effects of the methyl ester moiety, experiments were also conducted in which the AA-Me was replaced with the methyl ester analogs of oleic and linoleic acids. To determine the endothelium dependence of the effects of AA-Me, it was also tested in vessels in which the endothelium had been removed by perfusion of the vessel lumen with air. Two controls were done to ensure that the endothelium denudation procedure did not damage the smooth muscle. First, we examined the response to PE before and after the procedure and, if there was more than a 10% decrease in the measured response after denudation (usually the vasoconstricting response to PE was slightly increased), the smooth muscle cells were considered damaged, and the vessel was not further studied. Second, endothelium-independent vasodilation in response to a fixed concentration of an NO donor (sodium nitroprusside) was measured before and after denudation. Again, if there was more than a 10% decrease in the vasodilatory response to NO (there was usually a slight increase in the response), the vessel was considered damaged and not studied further.

Role of AA Metabolites and NO. The effects of inhibitors of the cyclooxygenase (indomethacin; 10 µM), the cytochrome P450 (17-ODYA; 10 µM), the lipoxygenase (esculetin, 10 µM), or the nitric-oxide synthase (N-nitro-L-arginine, L-NNA; 100 µM) pathways on responses to ACh and AA-Me were also studied. Inhibitors were added to the bath before and during application of ACh and/or AA-Me, and tested over a range of concentrations previously demonstrated to be effective in other systems.

Statistical Analysis. After verification that PE constrictions were not significantly different, the maximum relaxation was measured and normalized as the percentage of the PE constriction. Results are expressed as mean ± S.E.M. The effects of the PLA₂ inhibitors were compared with the control responses using Student's t test with Bonferroni correction. A p value <0.01 was taken to indicate a significant difference. The nonparametric Mann-Whitney test was used for the other comparisons, with p < 0.05 taken to indicate a significant effect; n indicates the number of experiments, each obtained from a different rat.

Solutions and Drugs. Acetylcholine, phenylephrine, arachidonic acid, arachidonic acid methyl ester, linoleic acid methyl ester, oleic acid methyl ester, N-nitro-L-arginine, and indomethacin were purchased from Sigma-Aldrich (St. Louis, MO). 17-Octadecynoic acid, esculetin, BEL, and AACOCF₃ were purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of PLA₂ Inhibitors. Pretreatment of the rat mesenteric arteries with BEL resulted in concentration-dependent inhibition of ACh relaxation in PE-constricted mesenteric arteries. The BEL-induced inhibition of ACh-mediated relaxation was enhanced when ACh was concomitantly administered with BEL (Fig. 1). Pretreatment of the rat mesenteric arteries with 1 µM BEL produced 53 ± 13% inhibition when administered alone (Fig. 1, A and C) and 85 ± 2% inhibition when applied with ACh (1 µM) (Fig. 1, A and D). AACOCF₃, an inhibitor of both the iPLA₂ and cPLA₂ family enzymes, also inhibited ACh relaxation (by 71 ± 12% at 1 µM) (data not shown). Neither BEL nor AACOCF₃ had any significant inhibitory effect on the PE-induced vascular constriction (data not shown).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   iPLA2 inhibition of ACh-mediated endothelium-dependent relaxation. A, dose-response relationships for the effect of the iPLA2 inhibitor, BEL, or BEL in presence of ACh, on 1 µM ACh relaxation in 10 µM PE-constricted rat mesenteric arteries. B, typical control responses to PE and ACh prior to preincubation with the iPLA2 inhibitor. PE and ACh responses following 10-min preincubation with 1 µM BEL alone (C) or following 10-min preincubation with 1 µM BEL plus 1 µM ACh (D). OD, outside diameter; *, p < 0.01 versus control, n = 8.

AA Effects on ACh Relaxation after BEL. After ACh relaxation had nearly been abolished by BEL, preincubation with 1 pM to 1 µM AA-Me resulted in near-complete recovery of the ACh relaxation (90 ± 2, 84 ± 3, and 86 ± 7% relaxation, respectively) (Fig. 2). The highest concentration of AA-Me tested (1 µM) also had a significant relaxing effect by itself (Fig. 2A). Interestingly, after BEL treatment, the higher dose of AA-Me (1 µM) also produced recovery of only a transient ACh relaxation (Fig. 2B), in contrast to recovery of the complete and sustained ACh-induced relaxation with the lower dose (Fig. 2C). As shown in Fig. 3A, the direct relaxing action of 1 µM AA-Me was not significantly affected by BEL and was absent in endothelium-denuded vessels (Fig. 3, A and B). Neither oleic acid nor linoleic acid methyl esters produced any significant relaxation (Fig. 3A), indicating that the specificity of the relaxing action of AA-Me was due to the arachidonyl moiety.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Recovery of ACh relaxation by arachidonic acid methyl ester following iPLA2 inhibition with BEL. A, comparison of ACh (1 µM) relaxation in PE-constricted rat mesenteric arteries before and after BEL, and following preincubation with 1 µM, 1 nM, or 1 pM AA-Me. B, typical recovery of only transient ACh relaxation following preincubation with 1 µM AA-Me. C, typical recovery of full and sustained ACh relaxation following preincubation with 1 pM AA-Me. OD, outside diameter; *, p < 0.05 versus control ACh relaxation, n = 5.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Relaxation of PE-constricted arteries with various fatty acids. A, relaxation percentage of PE-constricted rat mesenteric arteries produced by AA-Me before BEL, after BEL, in the absence of endothelial cells, and by different fatty acid methyl esters (all during 30 min of treatment at 1 µM). OA-Me, oleic acid methyl ester; LA-Me, linoleic acid methyl ester. B, following removal of endothelial cells, PE-induced contractions were slightly increased, and both ACh and AA-Me relaxations were eliminated. OD, outside diameter. *, p < 0.05 versus PE constriction, n = 5.

Role of NO and AA Metabolites. The effect of inhibitors of the cyclooxygenase (indomethacin; 10 µM), the cytochrome P450 (17-ODYA; 10 µM), the lipoxygenase (esculetin; 10 µM), or the nitric-oxide synthase (NOS) pathways (L-NNA; 100 µM) on ACh and AA-Me responses are shown in Fig. 4. As presented in the figure, the control ACh-mediated relaxation (Fig. 4A), the AA-Me-recovered ACh relaxation after BEL treatment (Fig. 4B), and the direct relaxing effect of AA-Me (Fig. 4C) all had similar pharmacological responses to the inhibitors. The control ACh relaxation (Fig. 4A), the BEL-inhibited ACh relaxation restored by AA-Me (Fig. 4B), as well as the direct relaxing effect of AA-Me (Fig. 4C), were each significantly inhibited by pretreatment with the NOS inhibitor, L-NNA. In contrast, no significant inhibition was produced by the three AA metabolic pathway inhibitors, even using prolonged incubations (up to 1 h) and higher concentrations than those previously reported to have effects in other systems. Interestingly, the cyclooxygenase pathway inhibitor (indomethacin; 10 µM) significantly increased the direct relaxing effect of 1 µM AA-Me (Fig. 4C) as well as the recovery of ACh relaxation by 1 µM AA-Me, resulting in sustained rather than transient recovery of ACh relaxation (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Role of NO and AA metabolites. Effect of preincubation with either the L-NNA (100 µM), indomethacin (Indo., 10 µM), 17-ODYA (10 µM), or esculetin (10 µM) on control ACh (1 µM) relaxation of 10 µM PE-constricted rat mesenteric arteries (A), 1 µM AA-Me-recovered ACh relaxation after BEL (B), or AA-Me (1 µM) relaxation (C). *, p < 0.05 versus control relaxation, n = 5.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Recent studies indicate that iPLA₂ is involved in stimulus-evoked AA release in a variety of cell types (Gross et al., 1993; Lehman et al., 1993; Wolf et al., 1997; Creer and McHowat, 1998; Akiba et al., 1999; Murakami et al., 1999). In this study, we have shown that ACh relaxation was nearly abolished by the iPLA₂-selective inhibitor BEL and that the ACh relaxation inhibited by BEL could be recovered by preincubation with a minute quantity of a cell-permeant analog of AA (AA-Me). In addition, BEL-mediated inhibition of ACh-induced relaxation was potentiated by coapplication of BEL with ACh, as would be anticipated for a mechanism-based inhibitor, and is consistent with the notion that ACh activates iPLA₂. Collectively, the results strongly suggest that iPLA₂-derived AA is an important component of ACh-mediated endothelium-dependent relaxation in rat mesenteric resistance arteries.

The role of phospholipases in ACh-induced relaxation in mesenteric arteries has previously been demonstrated utilizing AACOCF₃ (Adeagbo and Henzel, 1998). However, the lack of effect of BEL in this prior study suggested that cPLA₂, not iPLA₂, was involved. Although this finding appears to conflict with our results, the highest BEL concentration tested (0.3 µM) in that earlier study is only effective in utilizing purified enzyme and is not effective in cellular systems. Indeed, submicromolar concentrations of BEL were ineffective at blocking ACh relaxation in the absence of cellular activation. Since AACOCF₃ inhibits the iPLA₂ and the cPLA₂ enzyme families (Lio et al., 1996), our results using AACOCF₃ and BEL are both consistent with prior findings and strongly suggest the involvement of iPLA₂ in the generation of the AA required for ACh-induced endothelium-dependent relaxation in mesenteric vessels.

Remarkably, preincubation with a minute quantity (10¹² M) of the cell-permeant methyl ester analog of the iPLA₂ product AA (AA-Me) resulted in near-complete recovery of the ACh relaxation nearly abolished by BEL. Indeed, this finding provides strong support for the concept that iPLA₂-derived AA is involved in ACh relaxation. At the highest concentration tested, AA-Me also had a direct vascular-relaxing action in our model, and the relaxing action of AA-Me was both endothelium-dependent and AA-specific (i.e., it did not occur with oleic or linoleic acid methyl esters, and native AA had a similar effect). Fatty acids are transported across the plasma membrane by either spontaneous or enzyme-catalyzed transmembrane flip flop followed by rapid thioesterification with CoA to form acyl-CoA. In initial experiments, we sought to "rescue" BEL-mediated inhibition of ACh-induced relaxation with nonesterified AA. However, reconstitution of ACh-induced relaxation after BEL treatment with nonesterified AA was inconsistent. In contrast, utilization of AA-Me (which avoids the trapping of AA as its thioester) produced consistent restoration of ACh-mediated relaxation, presumably through delivery of AA-Me to appropriate intracellular membrane compartments with subsequent hydrolysis to nonesterified AA. The data further indicate that AA metabolites of either the cytochrome P450 or lipoxygenase pathways do not appear to be important for ACh relaxation in this model, or for the ACh-recovering or the direct relaxing effects of AA-Me. Although these results may appear surprising given the recent evidence suggesting that cytochrome P450 monooxygenase metabolites of AA (notably epoxyeicosatrienoic acids) may act as an endothelium-dependent hyperpolarizing factor (Campbell et al., 1996), other investigators have not been able to demonstrate the importance of cytochrome P450 metabolites (Vanheel and Van de Voorde, 1997), and considerable evidence suggests the differing relative importance of NO and endothelium-dependent hyperpolarizing factor in specific arterial systems (Ding and Triggle, 2000). Indomethacin potentiated both the ACh-recovering and -relaxing effect of AA-Me (1 µM), suggesting that the presence of a high amount of AA can result in the production of constrictor prostanoids, as previously described (Adeagbo and Malik, 1991).

Relaxation produced by ACh before BEL treatment, ACh relaxation recovered by AA-Me after BEL, as well as the direct relaxing effect of AA-Me, were all nearly completely abolished by preincubation with the NOS inhibitor, L-NNA. The findings that ACh and AA relaxations are blocked by NOS inhibitors are similar to those reported previously (Koller et al., 1993). Thus, it seems likely that both ACh and AA produce relaxation that is dependent on activation of endothelial NOS (eNOS). In addition, both the relaxing effect of AA-Me and the AA-Me-recovered ACh relaxation were not affected by BEL but were blocked by the NOS inhibitor. Thus, BEL-mediated inhibition of ACh relaxation could not be due to a direct effect of BEL on NOS activity. Rather, the finding that preincubation with only picomolar concentrations of AA-Me (which had no direct relaxing effect) resulted in full recovery of the NOS-dependent ACh relaxation after BEL suggests that iPLA₂-generated AA plays an essential permissive role in the series of biochemical events leading to eNOS activation after ACh stimulation.

ACh muscarinic receptors in endothelial cells are coupled via G proteins to the activation of phospholipase C, which results in the generation of inositol triphosphate and Ca²⁺ release from intracellular stores. Previously, we have provided evidence to propose the calcium-depletion hypothesis of iPLA₂ activation. The essential elements of this hypothesis include the calcium-dependent association of calmodulin with iPLA₂, leading to inhibition of iPLA₂ enzyme activity (Wolf and Gross, 1996), and the activation of iPLA₂ by dissociation of iPLA₂ from calmodulin after internal store calcium depletion (Wolf et al., 1997). The present results are consistent with the notion that ACh-mediated internal store calcium depletion results in de-inhibition of the calmodulin iPLA₂ complex, leading to the release of AA, thereby facilitating ACh-induced vascular relaxation.

Concomitant activation of intermediate conductance ACh- and Ca²⁺-activated K⁺ channels results in membrane hyperpolarization and Ca²⁺ influx through nonselective membrane cation channels (Himmel et al., 1993; Nilius et al., 1997). The increase in [Ca²⁺]_i is important for the activation of eNOS, a Ca²⁺-calmodulin-dependent enzyme (Presta et al., 1997). Although the role of AA in this process is unclear, iPLA₂-derived AA is capable of directly activating transfected Kv1.1 channels in Sf9 cells (Gubitosi-Klug et al., 1995). Furthermore, some had proposed that Ca²⁺ entry mechanisms other than depletion-activated channels may be important in agonist-evoked Ca²⁺ influx (for review, see Elliott, 2001). Finally, recent studies demonstrated that AA mediates activation of a novel noncapacitive Ca²⁺ entry pathway following activation of transfected muscarinic receptors in human embryonic kidney 293 cells (Shuttleworth and Thompson, 1998; Mignen and Shuttleworth, 2000). Although the effect of AA on ACh-induced [Ca²⁺]_i increases in endothelial cells is not known, these recent studies suggest that the potential importance of AA in ACh-mediated endothelium-dependent relaxation may be related to an essential role of AA in ACh-mediated increases in [Ca²⁺]_i. Such an effect may be mediated either by a direct effect of AA on K⁺ channels or by AA-induced activation of noncapacitive Ca²⁺ entry.