Vasorelaxation induced by methyl cinnamate, the major constituent of the essential oil ofOcimum micranthum, in rat isolated aorta

Summary: The aim of the present study was to investigate the vascular effects of the E‐isomer of methyl cinnamate (E‐MC) in rat isolated aortic rings and the putative mechanisms underlying these effects. At 1–3000 μmol/L, E‐MC concentration‐dependently relaxed endothelium‐intact aortic preparations that had been precontracted with phenylephrine (PHE; 1 μmol/L), with an IC50 value (geometric mean) of 877.6 μmol/L (95% confidence interval (CI) 784.1–982.2 μmol/L). These vasorelaxant effects of E‐MC remained unchanged after removal of the vascular endothelium (IC50 725.5 μmol/L; 95% CI 546.4–963.6 μmol/L) and pretreatment with 100 μmol/L NG‐nitro‐l‐arginine methyl ester (IC50 749.0 μmol/L; 95% CI 557.8–1005.7 μmol/L) or 10 μmol/L 1H‐[1,2,4]oxadiazolo[4,3‐a]quinoxalin‐1‐one (IC50 837.2 μmol/L; 95% CI 511.4–1370.5 μmol/L). Over the concentration range 1–3000 μmol/L, E‐MC relaxed K+‐induced contractions in mesenteric artery preparations (IC50 314.5 μmol/L; 95% CI 141.9–697.0 μmol/L) with greater potency than in aortic preparations (IC50 1144.7 μmol/L; 95% CI 823.2–1591.9 μmol/L). In the presence of a saturating contractile concentration of K+ (150 mmol/L) in Ca2+‐containing medium combined with 3 μmol/L PHE, 1000 μmol/L E‐MC only partially reversed the contractile response. In contrast, under similar conditions, E‐MC nearly fully relaxed PHE‐induced contractions in aortic rings in a Ba2+‐containing medium. In preparations that were maintained under Ca2+‐free conditions, 600 and 1000 μmol/L E‐MC significantly reduced the contractions induced by exogenous Ca2+ or Ba2+ in KCl‐precontracted preparations, but not in PHE‐precontracted preparations (in the presence of 1 μmol/L verapamil). In addition, E‐MC (1–3000 μmol/L) concentration‐dependently relaxed the contractions induced by 2 mmol/L sodium orthovanadate. Based on these observations, E‐MC‐induced endothelium‐independent vasorelaxant effects appear to be preferentially mediated by inhibition of plasmalemmal Ca2+ influx through voltage‐dependent Ca2+ channels. However, the involvement of a myogenic mechanism in the effects of E‐MC is also possible.

electromechanical coupling; vasorelaxant effects; voltage‐operated calcium channels

Methyl cinnamate (MC) is a phenylpropanoid derivative that is abundant in the essential oil of several plants, such as Ocimum micranthum Willd.,[1] Alpina malaccensis var. Nobilis,[2] Alpinia zerumbet[3] and Kaempferia galanga L. (Zingiberaceae).[4] It is also a pleasant and strongly aromatic constituent of fruits (e.g. strawberries) and culinary spices (e.g. basil) that is used in the flavour industry. Because of its in vitro antifungal activity,[5] MC vapours released from the edible film of strawberry puree extend the shelf‐life of strawberries.[6]

Pharmacological reports on MC are scarce, but recent studies from our group have found interesting pharmacological effects of the E‐ and Z‐isomers of MC on smooth muscle tissues in rats.[1] In isolated tracheal smooth muscle preparations, both the essential oil of O. micranthum (EOOM) and its main constituent E‐MC (Fig. [NaN] ) inhibited cholinergic‐induced contractions more potently than contractions elicited by KCl, with IC50 values for the relaxant effects against carbachol‐induced contractions being significantly lower than those against KCl‐induced contractions.[1] This suggests that, in rat airway tissues, E‐MC acts against contractile responses that preferentially recruit receptor‐operated Ca2+ channels (ROCCs), a finding that clearly differs from other volatile compounds that preferentially act on electromechanically induced contractions.[7] , [8] , [9] In strips from rat gastrointestinal smooth muscle, both EOOM and E‐MC exerted equipotent antispasmodic effects against KCl‐ and carbachol‐induced contractions, which likely resulted from a decrease in Ca2+ levels in smooth muscle cells.[10] These findings support the incorporation of O. micranthum in herbal remedy preparations in folk medicine.[11]

Hypertension is a common and progressive disorder that constitutes a major risk factor for diabetes, stroke, cardiovascular disease and renal disease. Numerous studies have focused on bioactive compounds from natural resources as potential substances for the treatment of hypertension. Efforts to characterize such effects have been pursued by our research group for many years in an attempt to identify novel antihypertensive compounds with vasodilator activity, especially compounds that are derived from essential oils of aromatic plants from the North‐Northeast regions of Brazil.[12] , [13] , [14] , [15] , [16] In a recent study, Kang et al.[17] reported the vasodilator effects of cinnamic acid (Fig. [NaN] ) on rat aorta by demonstrating the ability of this compound to endothelium‐dependently relax contractions induced by phenylephrine (PHE). Because MC is the methyl ester of cinnamic acid, it likely shares vasodilator properties with cinnamic acid. The present study was performed in rat isolated aortic rings to investigate this issue and assess the putative mechanisms that underlie the vascular effects of E‐MC.

In rat endothelium‐intact isolated aortic preparations (Fig. [NaN] b), sustained contractions in response to PHE (1 μmol/L; 0.96 ± 0.12 g) tended to be slightly enhanced by the lower concentrations (30, 100 and 300 μmol/L) of E‐MC, but the effects only reached statistical significance with 100 μmol/L E‐MC (P < 0.05, Holm‐Sidak test). This phenomenon was not observed in endothelium‐denuded preparations (Fig. [NaN] a,c). In contrast, at higher concentrations (600–3000 μmol/L), E‐MC concentration‐dependently inhibited PHE‐induced contractions (P < 0.001, anova; Fig. [NaN] b).

The tendency of E‐MC to enhance PHE‐induced contractions was abrogated by the mechanical removal of the endothelium (Fig. [NaN] c; magnitude of PHE‐induced contractions in endothelium‐denuded preparations: 1.04 ± 0.06 g) or incubation of endothelium‐intact aortic preparations with either NG‐nitro‐l‐arginine methyl ester (l‐NAME; 100 μmol/L; Fig. [NaN] d; magnitude: 1.36 ± 0.14 g) or 1H‐[1,2,4]oxadiazolo[4,3‐a]quinoxalin‐1‐one (ODQ; 10 μmol/L, Fig. [NaN] e; magnitude: 1.40 ± 0.13 g). Under these conditions, two‐way anova revealed that the contraction values at several concentrations of E‐MC (30–300 μmol/L) decreased compared with the effects of the same concentrations in endothelium‐intact aortic ring preparations. However, the IC50 value for the E‐MC‐induced reduction of PHE‐induced contractions in preparations with an intact endothelium (877.6 μmol/L; 95% confidence interval (CI) 784.1–982.2 μmol/L; n = 6) was not significantly different from that in endothelium‐denuded preparations (725.5 μmol/L; 95% CI 546.4–963.6 μmol/L; n = 8; P > 0.05, Mann–Whitney U‐test) or from preparations with an intact endothelium incubated with l‐NAME (749.0 μmol/L; 95% CI 557.8–1005.7 μmol/L; n = 7; P > 0.05, Mann–Whitney U‐test). Similarly, the vasorelaxant effects of E‐MC remained unchanged (P > 0.05, two‐way anova; Fig. [NaN] e) in endothelium‐intact aortic ring preparations that were contracted with PHE and preincubated for 10 min with ODQ (IC50 = 837.2 μmol/L; 95% CI 511.4–1370.5 μmol/L; n = 7).

To determine whether E‐MC exerts relaxant effects with similar potency in conductance and resistance vessels, a separate set of experiments was conducted with aortic and mesenteric arterial rings that were subjected to sustained contractions induced by 60 mmol/L KCl (contractions corresponded to 1.14 ± 0.12 and 0.68 ± 0.19 g in aortic and mesenteric arterial rings, respectively). In endothelium‐intact aortic rings, E‐MC (1–3000 μmol/L; n = 6) fully and concentration‐dependently relaxed the sustained contractions (P < 0.001, one‐way anova; Fig. [NaN] ), with an IC50 value (1144.7 μmol/L; 95% CI 823.2–1591.9 μmol/L) that was in the same order of magnitude as that observed for the E‐MC‐induced relaxant effects on PHE‐induced contractions (P > 0.05, Mann–Whitney U‐test). The inhibitory effect of E‐MC became significant at 600 μmol/L (P < 0.05, one‐way anova followed by the Holm‐Sidak test; Fig. [NaN] ).

In endothelium‐intact mesenteric arterial rings (Fig. [NaN] ), E‐MC (1–3000 μmol/L) relaxed sustained contractions induced by KCl (60 mmol/L), with an IC50 value of 314.5 μmol/L (95% CI 141.9–697.0 μmol/L; (n = 6), which was significantly lower (P < 0.05, Mann–Whitney U‐test) than the IC50 obtained for the relaxant effect of E‐MC on KCl‐induced contractions in aortic rings. The inhibitory effect of E‐MC became significant at 300 μmol/L (P < 0.05, one‐way anova followed by the Holm‐Sidak test; Fig. [NaN] ).

In aortic preparations with an intact endothelium that were incubated in Ca2+‐free medium containing 400 μmol/L EGTA in the presence of high KCl (60 mmol/L), increasing concentrations of CaCl2 (0.1–20 mmol/L; n = 12) concentration‐dependently evoked contractions (P < 0.001, one‐way anova), an effect that became significant at 0.5 mmol/L (P < 0.05, one‐way anova followed by the Holm‐Sidak test) and reached a maximal at 20 mmol/L (Fig. [NaN] a). Contractions induced by the addition of Ca2+ were significantly reduced (P < 0.01, two‐way anova) by 600 and 1000 μmol/L E‐MC (n = 6–7; P < 0.01, two‐way anova; Fig. [NaN] a). The maximum response to 20 mmol/L CaCl2 (138.1%) was significantly reduced to 85.5% and 50.7% in the presence of 600 and 1000 μmol/L E‐MC, respectively (P < 0.05, two‐way anova followed by the Holm‐Sidak test; Fig. [NaN] a).

In preparations that were maintained in Ca2+‐free medium containing 400 μmol/L EGTA in the presence of high KCl (60 mmol/L), increasing concentrations of Ba2+ (0.1–20 mmol/L; n = 15) evoked the expected concentration‐dependent increase in the force of contractions (P < 0.001, one‐way anova), an effect that became significant at a concentration of 0.5 mmol/L (P < 0.05; Fig. [NaN] b). The maximum effect (105.3%) was significantly reduced (P < 0.05, two‐way anova followed by the Holm‐Sidak test; Fig. [NaN] b) to 66.4% and 43.5% following 5 min pretreatment of the preparations with 600 and 1000 μmol/L E‐MC (n = 7–8), respectively.

Endothelium‐intact aortic preparations that were incubated in Ca2+‐free medium in the presence of PHE (1 μmol/L) and verapamil (1 μmol/L) exhibited concentration‐dependent contractions to increasing concentrations of CaCl2 (0.1–20 mmol/L; n = 13; P < 0.001, one‐way anova). The effect became significant at 1 mmol/L E‐MC (P < 0.05, one‐way anova followed by the Holm‐Sidak test) and reached a maximum at 20 mmol/L (Fig. [NaN] c). Contractions induced by the addition of Ca2+ remained unaffected by 600 and 1000 μmol/L E‐MC (n = 6–7; P > 0.05, two‐way anova; Fig. [NaN] c).

After preparations had been bathed in 2 mmol/L Ca2+‐containing modified Krebs'–Henseleit solution (MKHS), a contractile response was evoked with 150 mmol/L K+ and the subsequent addition of PHE (3 μmol/L) to achieve maximum functional activation (Fig. [NaN] a). Under these conditions, the addition of E‐MC (1000 μmol/L; n = 6) during the steady state of sustained contractions resulted in significant relaxation (P < 0.05, Holm‐Sidak test; Fig. [NaN] a), an effect that corresponded to 55.2 ± 5.6% of the contractions induced by 100 mmol/L K+ (0.64 ± 0.19 g; Fig. [NaN] c). In contrast, in preparations that were maximally stimulated with 150 mmol/L K+ in the presence of 2 mmol/L Ba2+, the same concentration of E‐MC (1000 μmol/L) induced a more pronounced relaxant effect (P < 0.05, unpaired Student's t‐test) compared with preparations maintained in Ca2+‐containing medium, almost fully relaxing the combined contraction (approximately 110.8 ± 6.1% of the contraction induced by 100 mmol/L K+, corresponding to 1.28 ± 0.05 g; n = 6; Fig. [NaN] b,c).

In aortic ring preparations incubated in Ca2+‐free medium with 1 mmol/L EGTA, the protein tyrosine phosphatase inhibitor sodium orthovanadate (2 mmol/L) induced sustained contractions corresponding to 1.11 ± 0.13 g (n = 6). When contractions reached a steady state, the cumulative addition of E‐MC (1–3000 μmol/L) significantly and concentration‐dependently reduced orthovanadate‐induced contractions (P < 0.001, one‐way anova; Fig. [NaN] ). The effect of E‐MC became significant at 30 μmol/L (P < 0.01, one‐way anova followed by the Holm‐Sidak test), with an IC50 value of 611.1 μmol/L (95% CI 317.4–1176.5 μmol/L), which was in the same order of magnitude (P > 0.05, Mann–Whitney U‐test) as the one evoked by E‐MC against contractions elicited by PHE.

The present study demonstrated that E‐MC at higher concentrations had vasorelaxant properties that occurred independent of endothelium integrity. The E‐isomer of methyl cinnamate inhibited, with the same potency, both electromechanical and pharmacomechanical coupling and the contractions induced by sodium orthovanadate under Ca2+‐free conditions. In contrast, E‐MC inhibited contractions induced by the cumulative addition of Ca2+ when aortic preparations were previously maintained in Ca2+‐free medium that contained high K+, but not with a mixture of PHE and verapamil, suggesting that E‐MC preferentially interferes with contractile responses evoked by Ca2+ influx through voltage‐operated calcium channels (VOCCs). However, a myogenic mechanism may also be involved.

At concentrations > 600 μmol/L, E‐MC significantly and concentration‐dependently reduced PHE‐induced contractions. One possibility is that this vasorelaxant effect is mediated by an endothelial l‐arginine/nitric oxide (NO) pathway through peripheral muscarinic receptor activation. However, such a possibility seems unlikely because the vasorelaxant effects of E‐MC were not reduced after blockade of endothelial function by removing the endothelium or pretreatment with l‐NAME. In addition, the vasorelaxant effects of E‐MC remained unaffected after pretreating aortic rings with the muscarinic receptor antagonist atropine (AA Vasconcelos‐Silva, unpubl. data, 2013). The vasorelaxant effects of E‐MC are also not attributable to the putative ability of this substance to activate soluble guanylate cyclase (sGC) because the vasorelaxant effect remained unchanged after preincubation of aortic preparations with the sGC inhibitor ODQ.

Importantly, such findings differ from the vascular effects reported for the E‐MC analogues cinnamic acid and ethyl cinnamate, which induced endothelium‐dependent vasodilation in rat aorta.[17] , [18] Pretreatment of aortic tissues with indomethacin and methylene blue blunted the vasodilator effect of ethyl cinnamate,[18] whereas ODQ and l‐NAME reduced the cinnamic acid‐induced relaxant effect.[18] Thus, a reasonable consideration is that the carboxylate moiety substituent (Fig. [NaN] ) is important for conferring the ability of these compounds to interfere with the modulatory role of the endothelial layer in vascular smooth muscle cells. The inhibitory effects of cinnamic acid and ethyl cinnamate may involve the release of NO and prostacyclin from endothelial cells. At low concentrations, E‐MC appeared to induce endothelium‐dependent potentiating actions on PHE‐induced contractions. A reasonable hypothesis to explain such potentiation is the inhibition of endothelial NO production or release because the potentiating effect was abolished both in endothelium‐denuded rings and in the presence of l‐NAME. Notwithstanding, the potentiating effect of E‐MC in the presence of PHE was small in magnitude, which clearly did not interfere with the more evident vasorelaxant action induced by E‐MC. Notably, E‐MC itself did not change the resting tone of aortic preparations with intact endothelia (data not shown), which precludes any possible contractile effect of E‐MC that could be related to its ability to release Ca2+ from the sarcoplasmic reticulum.

With regard to excitation–contraction coupling in vascular smooth muscle, high KCl is well known to induce membrane depolarization, which, in turn, opens VOCCs, promotes Ca2+ influx, increases [Ca2+]i and elicits sustained contractions following myosin light chain phosphorylation.[19] Phenylephrine contracts vascular smooth muscles as a result of α1‐adrenoceptor activation.[20] The present study showed that E‐MC inhibited both KCl‐ and PHE‐induced contractions in aortic preparations with intact endothelia. Generally, α1‐adrenoceptor activation may cause contraction in certain large conduit vessels without the involvement of depolarizing effects (i.e. pharmacomechanical coupling), but such a phenomenon depends on the vessel type and agonist concentration.[21] Once activated, α1‐adrenoceptors recruit a biphasic response that is characterized by an initial phasic event that derives from intracellular Ca2+ release from the sarcoplasmic reticulum and a subsequent tonic stimulus that results from Ca2+ influx from the extracellular compartment, especially through ROCCs.[22]

A series of experiments was performed by contracting aortic vessels with PHE in Ca2+‐free medium that contained verapamil to prevent VOCCs influencing the resulting contractions. Under these conditions, E‐MC did not cause any inhibitory effect on contractions elicited by the exogenous addition of Ca2+, even when E‐MC was applied at concentrations that induced pronounced vasodilatory effects (> 50% efficacy) on PHE‐induced contractions in the absence of the VOCC blocker. Based on these findings, a possible conclusion is that E‐MC was inert against contractions mediated by phenylalkylamine‐insensitive Ca2+ influx through ROCCs. In addition, PHE likely recruited pathways that are distinct from ROCCs to induce contractions under normal conditions in rat aorta, and the inhibitory actions of E‐MC against VOCC‐elicited mechanisms appear to be sufficient to significantly relax sustained contractions elicited by PHE when the extracellular concentration of Ca2+ is maintained at physiological levels.

Evidence suggests that E‐MC acts against contractile responses that preferentially recruit VOCCs rather than ROCCs. First, in aortic preparations that are depolarized with high KCl maintained under Ca2+‐free conditions, the contractions induced by CaCl2, which are attributable to an increase in Ca2+ influx through VOCCs, were significantly reduced by E‐MC. Second, under the same conditions, E‐MC was effective against contractile responses to increasing concentrations of Ba2+, an ion that is preferentially permeable through VOCCs rather than ROCCs.[23] , [24] Such a hypothesis is further supported by the relaxant effect of E‐MC against contractions induced in aortic preparations by the addition of saturating depolarizing concentrations of K+, whereas PHE concomitantly elicited the recruitment of ROCCs. The relaxant effect of E‐MC was more evident in Ba2+‐containing medium than in Ca2+‐containing medium, thus indicating the preferential recruitment of VOCCs. These conclusions are consistent with the postulation of Othman et al.[18] regarding the ability of ethyl cinnamate to inhibit Ca2+ influx into vascular smooth muscle cells of the rat aorta.

Such a mode of action of E‐MC in vascular tissue differs from the one that was reported previously in isolated tracheal smooth muscles. In fact, in the latter preparations, both EOOM and its main constituent E‐MC more potently inhibited cholinergic‐induced contractions than KCl‐induced contractions.[1] This suggests that E‐MC acts against contractile responses that preferentially recruit pharmacomechanical coupling in rat airway tissues, a finding that is clearly different from other essential oils and volatile compounds that preferentially act on electromechanically induced contractions.[7] , [8] , [9]

Notable, however, is the finding that E‐MC relaxed contractions evoked pharmacomechanically or electromechanically in aortic rings with the same pharmacological potency. Although a preferential action on VOCC‐mediated contractile phenomena appears to be consistent, a putative myogenic mechanism should also be considered. In fact, E‐MC was able to relax contractions induced by sodium orthovanadate in endothelium‐intact aortic preparations that were maintained in Ca2+‐free medium. The phosphorylation of tyrosine by tyrosine kinases is able to activate many intracellular signalling pathways, resulting in various cellular events, including the contraction of vascular smooth muscles.[25] The level of phosphorylated residues of tyrosine is regulated by a balance between the actions of tyrosine kinases and tyrosine phosphatases. Sodium orthovanadate, a potent protein tyrosine phosphatase inhibitor, induces smooth muscle contractions through the reduction of dephosphorylated tyrosine residues, thus indirectly increasing the amount of phosphorylated tyrosine.[26] This suggests that the vasorelaxant effects of E‐MC could partially depend on the amount of phosphorylated tyrosine residues. We hypothesize that more direct experimental methods that are able to measure tyrosine phosphorylation may unequivocally demonstrate the occurrence of such a phenomenon.

Interestingly, E‐MC more potently relaxed mesenteric arterial rings than aortic rings. Although we are unable to definitively explain the higher potency of E‐MC in mesenteric arterial vessels, such a feature suggests a preferential action of this compound in tissues that are more involved in haemodynamic control and ultimately determine peripheral vascular resistance.[27] Further experiments are necessary to assess the cardiovascular effects of E‐MC in normotensive rats. One may suggest that the E‐MC‐induced vasorelaxant effects are related to possible toxic effects. However, two lines of evidence refute this possibility. First, under our experimental conditions, all the vasodilator responses to this isomer were reversible (data not shown). Second, the LD50 of EOOM and E‐MC has been reported to be 5[28] and 2.1 g/kg body weight (BG Pinheiro & ASB Silva, unpubl. data, 2008), respectively, when administered in rats. For an LD50 that ranges from 2.00 to 4.99 g/kg, the essential oil is classified as having low toxicity. Because of its safety and current use as a food additive, E‐MC may be considered a flavouring substance that possesses the advantage of having vasodilator properties.

In conclusion, the findings of the present study indicate that E‐MC concentration‐dependently induces vasorelaxation that occurs independently of the integrity of the vascular endothelium or NO synthase (NOS) or sGC activation. The effect of E‐MC appears to be preferentially mediated through the inhibition of plasmalemmal Ca2+ influx through voltage‐dependent Ca2+ channels. However, the possibility that this vasorelaxation may partially occur intracellularly, likely through the inhibition of contractions that occur independent of Ca2+ influx from the extracellular milieu, cannot be discarded.

Adult male Wistar rats (200–250 g) were maintained under conditions of constant temperature (22 ± 2°C) and a 12 h light–dark cycle with free access to food and water. All animal care was in compliance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH publication no. 85–23, revised 1996; http://www.nap.edu/catalog.php?record%5fid=5140, accessed 5 August 2014). All the procedures were reviewed by and had prior approval from the local animal ethics committee (11/2012).

The E‐isomer of methyl cinnamate was dissolved in 0.5% Tween 80 and sonicated just before use. The final concentrations were achieved with the addition of physiological solution. The highest vehicle concentration used in the bath chamber was 0.075% (v/v; a concentration that is devoid of significant effects; data not shown). The physiological solution was fresh MKHS buffer (pH 7.4) with the following composition (in mmol/L): NaCl 118; KCl 4.7; NaHCO3 25; CaCl2·2H2O 2.5; KH2PO4 1.2; MgSO4·7H2O 1.2; glucose 10. Phenylephrine hydrochloride, acetylcholine chloride, EGTA, ODQ, sodium orthovanadate, l‐NAME hydrochloride and verapamil were all purchased from Sigma (St Louis, MO, USA) and dissolved in distilled water; final concentrations in the bath chamber were achieved by the addition of MKHS, with the exception of EGTA, which was added directly to Ca2+‐free MKHS.

Rats acutely sedated with tribromoethanol (250 mg/kg, i.p.) were killed by cervical dislocation immediately followed by exsanguination. The thoracic aorta and second branch of the superior mesenteric artery were removed and cut into cylindrical ring‐like segments (1 × 5 mm and 0.2 × 2 mm, respectively). Contractions were recorded with aortic rings mounted in a 5 mL organ bath. The rings were attached with triangular steel wire (0.3 mm diameter) to isometric force transducers (ML870B60/C‐V; ADInstruments, Bella Vista, NSW, Australia) and a data‐acquisition system (PowerLab 8/30; ADInstruments). Rings of the mesenteric artery were mounted in a 610M‐DMT Wire Myograph System (DMT, Aarhus, Denmark) with two tungsten wires (40 μm) that passed through the lumen of each ring, one fixed to a micrometer for length adjustments and the other connected to a force transducer. The organ bath contained warm (37°C) MKHS that was bubbled continuously with 5% CO2 in O2. Aortic and mesenteric arterial rings were stretched with a passive tension of 1 g and 5 mN, respectively. For endothelium denudation, the intimal surface of the aortic rings was rubbed gently with a stainless steel wire (0.3 mm diameter). Lack of endothelial functionality was confirmed pharmacologically by the absence of relaxation after the addition of acetylcholine (1 μmol/L) on the plateau of a PHE (0.1 μmol/L)‐induced contraction. The vessels were initially exposed twice to 60 mmol/L KCl to verify their functional integrity. After 30 min, the rings were contracted with a concentration of PHE (0.1 μmol/L) that induces 50%–70% of the contraction induced by KCl. Acetylcholine (1 μmol/L) was then added to assess endothelium integrity. The endothelium was considered intact when acetylcholine‐induced relaxation that was > 70% of the contraction induced by PHE. Sixty minutes later, each ring was subjected once to one of the following six series of experiments.

In this series of experiments, the effects of cumulative concentrations of E‐MC (1–3000 μmol/L) on sustained contractile responses to PHE (1 μmol/L) were studied in either endothelium‐intact or ‐denuded aortic ring preparations that were maintained in MKHS. To assess the role of sGC and NOS in mediating the vascular effects of E‐MC, experiments were performed in PHE (1 μmol/L)‐contracted aortic ring preparations with intact endothelia that were incubated for 5–10 min with ODQ (10 μmol/L) or l‐NAME (100 μmol/L), respectively.

In this series of experiments, the effects of cumulative concentrations of E‐MC (1–3000 μmol/L) on sustained contractile responses to KCl (60 mmol/L) were studied in endothelium‐intact aortic or mesenteric arterial rings maintained in MKHS.

This series of experiments was performed in aortic ring preparations that were depolarized by 60 mmol/L KCl in Ca2+‐free medium to assess the inhibitory effects of E‐MC on contractions induced by the exogenous addition of Ca2+ or Ba2+. The Ba2+ ions flow through VOCCs but are poorly permeable through ROCCs. The Ca2+‐free solutions were prepared by omitting CaCl2 from normal MKHS. After verifying tissue responsiveness in Ca2+‐containing medium, the preparation was maintained in Ca2+‐free MKHS in the presence of 60 mmol/L K+ and 400 μmol/L EGTA to promote VOCC activation. Thereafter, cumulative concentration–response curves were constructed to Ca2+ (0.1–20 mmol/L) and Ba2+ (0.1–20 mmol/L). Subsequently, the preparation was washed again by changing the bath chamber solution to remove Ca2+ or Ba2+ from the medium and E‐MC (600 and 1000 μmol/L) was then added to the preparation for 5 min and cumulative concentration–response curves to CaCl2 (0.1–20 mmol/L) or Ba2+ (0.1–20 mmol/L) were again constructed. A maximal response to exogenous Ca2+ or Ba2+ was presumed when further increases in the Ca2+ or Ba2+ concentration did not induce significant additional contractions. The contractile response obtained with the first concentration–response curve for CaCl2 or BaCl2 was used as a control and the contractions were calculated as a function of the value observed for initial contractions in response to 60 mmol/L KCl.

The same protocol as Series 3 (construction of cumulative concentration–response curves for 0.1–20 mmol/L Ca2+) was used in aortic ring preparations that were maintained under Ca2+‐free conditions but stimulated with PHE (1 μmol/L) in the presence of verapamil (1 μmol/L) to obtain contractile responses that are preferentially mediated by ROCCs.

To further determine whether E‐MC acts on contractile responses induced through voltage‐operated Ca2+ entry, aortic rings were maximally stimulated with 100 mmol/L KCl in the presence of either Ca2+ (2 mmol/L) or Ba2+ (2 mmol/L). After contractions developed, the K+ concentration in the bath solution was increased further in steps of 25 mmol/L to 150 mmol/L to confirm that maximal voltage‐operated contractions were achieved while the contractile force was recorded continuously. After the preparation reached a plateau, PHE (3 μmol/L) was added in the presence of the highest K+ concentration to also activate the receptor‐operated Ca2+ entry pathway. Then, E‐MC (1000 μmol/L) was added to the bath solution and relaxation was recorded until an asymptotic value could be computed.

To assess whether vasorelaxation elicited by E‐MC occurs at least partially through a reduction of Ca2+ sensitivity in the intracellular contractile pathway, the inhibitory effects of increasing concentrations of E‐MC (1–3000 μmol/L) on contractions elicited by sodium orthovanadate (2 mmol/L) were studied in aortic rings with intact endothelia that were incubated in Ca2+‐free medium containing 1 mmol/L EGTA.

Results are expressed as the mean ± SEM. Peak deflections were used to measure the magnitude of the concentration–response curves, which are expressed as a percentage of a given contractile agent in the absence of E‐MC. The IC50 value, defined as the E‐MC concentration (in μmol/L) that was required to produce a half‐maximum reduction of a given contractile stimulus, was used to evaluate the vascular sensitivity to E‐MC. The IC50 was calculated by interpolation from semilogarithmic plots and is expressed as the geometric mean, with 95% CI. The significance (P < 0.05) of the results was assessed using unpaired Student's t‐test, the Mann–Whitney U‐test and one‐ or two‐way anova, followed by the Holm‐Sidak multiple‐comparison test when appropriate.

This study was financed by a grant from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) through the Instituto Nacional de Biomedicina do Semiárido Brasileiro (INCT‐IBISAB‐CNPq), Edital MCT/CNPq N8014/2010 and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

The authors declare no conflicts of interest.

Graph: Chemical structure of the E ‐isomer of methyl cinnamate and its analogues ethyl cinnamate and cinnamic acid.

Graph: image%5ft/cep12289-fig-0001-t.gif

Graph: Typical trace recordings of the vasorelaxant effect of E ‐methyl cinnamate (E‐ MC ; 1–3000  μ mol/L) on sustained contractions induced by phenylephrine ( PHE ; 1  μ mol/L) in (a) endothelium‐denuded (E−) and (b) endothelium‐intact (E+) isolated aortic preparations from normotensive rats. (c–e) Concentration–effect curves for the relaxant effects of E‐ MC (1–3000  μ mol/L) in (c) E+ (○) and E− (●) aortic rings precontracted with PHE (1  μ mol/L) and in E+ preparations pretreated with (d) 100  μ mol/L NG ‐nitro‐ l ‐arginine methyl ester (▲) or (e) 10  μ mol/L 1H‐[1,2,4]oxadiazolo[4,3‐a]quinoxalin‐1‐one (■). Data are the mean ±  SEM ( n  = 6–8 per group). * P  < 0.05 compared with contractions in the absence of E‐ MC in E+ preparations (one‐way anova followed by the Holm‐Sidak test); †P  < 0.05 compared with the same concentration in E+ preparations (two‐way anova followed by the Holm‐Sidak test).

Graph: image%5ft/cep12289-fig-0002-t.gif

Graph: Concentration–effect curves for the relaxant effects of E ‐methyl cinnamate (E‐ MC ; 1–3000  μ mol/L) in endothelium‐intact aortic (▽) and mesenteric arterial (▼) rings. In both preparations, sustained contractions were induced by KC l (60 mmol/L) and are expressed as a percentage of the contraction immediately before the addition of E‐ MC. Data are the mean ±  SEM ( n  = 6–8 per group). * P  < 0.05, first significant effect compared with values without E‐ MC (one‐way anova followed by the Holm‐Sidak test); †P  < 0.001 compared with the relaxing effects of E‐ MC in aortic rings (two‐way anova ).

Graph: image%5ft/cep12289-fig-0003-t.gif

Graph: Inhibitory effects of E‐methyl cinnamate (E‐ MC ; 600 and 1000  μ mol/L) on the cumulative concentration–effect curve with the addition of extracellular (a) Ca 2+ or (b) Ba 2+ (0.1–20 mmol/L) in endothelium‐intact, KC l‐stimulated aortic ring preparations incubated in Ca 2+ ‐free medium. (♦), control; (♢), 600  μ mol/L E‐ MC ; (■), 1000  μ mol/L E‐ MC. (c) Inhibitory effects of E‐ MC on the cumulative concentration–effect curve with the addition of extracellular Ca 2+ (0.1–20 mmol/L) in endothelium‐intact, phenylephrine‐stimulated aortic ring preparations incubated in Ca 2+ ‐free medium. (●), control; (○), 600  μ mol/L E‐ MC ; (□), 1000  μ mol/L E‐ MC. Data are the mean ±  SEM ( n  = 6–15 per group). * P  < 0.05 for control curve or curves obtained in the presence of E‐ MC (one‐way anova followed by the Holm‐Sidak test); †P  < 0.05 for E‐ MC compared with concentration–reponse curves of Ca 2+ or Ba 2+ alone as a control (two‐way anova ).

Graph: image%5ft/cep12289-fig-0004-t.gif

Graph: (a) Effects of E ‐methyl cinnamate (E‐ MC ) on rat aortic preparations precontracted with both K + and phenylephrine ( PHE ) in Ca 2+ ‐containing medium (upper trace) or Ba 2+ ‐containing medium (lower trace). Typical traces show that aortic preparations were maximally stimulated with supramaximal concentrations of K + (100–150 mmol/L) and PHE (3  μ mol/L). Once the steady state of this contraction was reached, E‐ MC (1000  μ mol/L) was added and tissue relaxation (arrow) was measured. (b) Mean ±  SEM values of the relaxant effect. Results are expressed as a percentage of contractions evoked by 100 mmol/L KC l ( n  = 6 per group). * P  < 0.05 compared with preparations maintained in Ca 2+ ‐containing medium (unpaired Student's t ‐test).

Graph: image%5ft/cep12289-fig-0005-t.gif

Graph: Effects of cumulative and increasing concentrations of E ‐methyl cinnamate (E‐ MC ; 1–3000  μ mol/L) on sustained contractions induced by the tyrosine phosphatase inhibitor sodium orthovanadate (3 mmol/L) in rat aortic ring preparations under Ca 2+ ‐free conditions. Results are the mean ±  SEM ( n  = 6). * P  < 0.05, first significant effect compared with the values before E‐ MC (one‐way anova followed by the Holm‐Sidak test).

Graph: image%5ft/cep12289-fig-0006-t.gif