Author for correspondence: Márcio Roberto V. Santos, Department of Physiology, Federal University of Sergipe, São Cristóvão-SE, 49100 000, Brazil (fax +55 79 2105 6474, e-mail email@example.com).
Abstract: Citronellol is an essential oil constituent from the medicinal plants Cymbopogon citratus, Cymbopogon winterianus and Lippia alba which are thought to possess antihypertensive properties. Citronellol-induced cardiovascular effects were evaluated in this study. In rats, citronellol (1–20 mg/kg, i.v.) induced hypotension, which was not affected by pre-treatment with atropine, hexamethonium, Nω-nitro-l-arginine methyl ester hydrochloride or indomethacin, and tachycardia, which was only attenuated by pre-treatment with atropine and hexamethonium. These responses were less than those obtained for nifedipine, a reference drug. In intact rings of rat mesenteric artery pre-contracted with 10 μM phenylephrine, citronellol induced relaxations (pD2 = 0.71 ± 0.11; Emax = 102 ± 5%; n = 6) that were not affected by endothelium removal, after tetraethylamonium in rings without endothelium pre-contracted with KCl 80 mM. Citronellol strongly antagonized (maximal inhibition = 97 ± 4%; n = 6) the contractions induced by CaCl2 (10−6 to 3 × 10−3 M) and did not induce additional effects on the maximal response of nifedipine (10 μM). Finally, citronellol inhibited the contractions induced by 10 μM phenylephrine or 20 mM caffeine. The present results suggest that citronellol lowers blood pressure by a direct effect on the vascular smooth muscle leading to vasodilation.
The essential oils are natural, complex, multi-component systems composed mainly of terpenes in addition to some other non-terpene components. These volatile substances are commonly found in aromatic plants and their therapeutic potential has been evaluated lately [1–3]. Studies in animals have demonstrated beneficial properties of essential oils in the cardiovascular system such as antithrombotic, antiplatelet, endothelial protective, vasorelaxant and hypotensive activities [2–5]. Recent reports have shown that the cardiovascular effects of essential oils are also observed in humans such as improvement in coronary flow , hypotensive and bradycardic effects .
The use of medicinal plants as an alternative to conventional medicine in the treatment of cardiovascular diseases has increased considerably worldwide. Because of these trends, many reports have evaluated the effects of several medicinal plants and their constituents on the cardiovascular system, aiming to provide a scientific basis for the therapeutic applications. In this context, essential oils extracted from medicinal plants have been largely studied and their therapeutic potential has been demonstrated in animals .
Citronellol is one of the essential oil constituents from some medicinal plants used in folk medicine as an antihypertensive drug, such as Cymbopogon citratus , Cymbopogon winterianus  and Lippia alba . It belongs to a family of natural products derived from C5 isoprene units and which has known cardioprotective effects . Some pharmacological effects such as antibacterial, antifungal, antispasmodic and anticonvulsant activity were described for citronellol . Due to the high use of medicinal plants containing citronellol and the absence of information about effects of this essential oil on the cardiovascular system, the objective of this work was to evaluate the cardiovascular effects of citronellol by using in vitro and in vivo approaches.
Material and Methods
Drugs. The drugs used were: (RS)-(±)-citronellol: [α]27D = 0 (c 0.017, CHCl3), 95.0% (GC) of purity and FW = 156.27 (fig. 1) (from Dierberger, Brazil), Nω-nitro-l-arginine methyl ester hydrochloride, atropine sulphate, indomethacin, l-phenylephrine chloride, acetylcholine chloride, hexamethonium, tetraethylamonium, caffeine, ethyleneglycol bis (β-aminoethylether)-N,N,N′,N′-tetraacetic acid (all from SIGMA, St. Louis, MO, USA), sodium thiopental (CRISTÁLIA , São Paulo, Brazil), heparin (ARISTON, São Paulo, Brazil) and nifedipine (RBI, Natick, MA, USA). In the preparation of the stock solutions, citronellol was diluted in Tyrode/cremophor (0.15% v/v) solution, for in vitro experiments, or saline/cremophor (0.15% v/v) solution, for in vivo. Other drugs were diluted in saline (in vivo) or Tyrode (in vitro) solutions only. All stock solutions were maintained at 0°C and diluted to the desired concentration with distilled water or saline, when necessary. Cremophor in the concentrations used showed no effect in control experiments (fig. 4A).
Solutions. The composition of the normal Tyrode’s solution was (mM): NaCl 158.3, KCl 4.0, CaCl2 2.0, NaHCO3 10.0, C6H12O6 5.6, MgCl2 1.05 and NaH2PO4 0.42. The K+-depolarizing solutions (KCl 60 and 80 mM) were prepared by replacing 60 or 80 mM KCl in Tyrode’s solution with equimolar NaCl respectively. In nominally without Ca2+ solution, CaCl2 was omitted, and in Ca2+-free solution, CaCl2 was omitted and 1 mM of ethyleneglycol bis (β-aminoethylether)-N,N,N′,N′-tetraacetic acid, a calcium chelator, was added.
Animals. Male Wistar normotensive rats (200–300 g) were obtained from colonies maintained in the Department of Physiology, Federal University of Sergipe, Sergipe, Brazil. They were maintained in a large cage under controlled conditions of temperature and lighting (lights on: 06:00–18:00 hr), fed rodent diet and tap water ad libitum. All procedures were approved by the Animal Research Ethics Committee of the Federal University of Sergipe, Brazil (Protocol number 72/2006) and were in compliance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication 85-23, revised 1996).
Effect of citronellol on haemodynamic parameters in non-anaesthetized rats. The measurements of the mean arterial pressure and heart rate were performed as described by Lahlou et al. . After the stabilization period, mean arterial pressure and heart rate were recorded in conscious and freely moving rats before (baseline values) and after i.v. in bolus administration of nifedipine (0.1, 0.5, 1 and 5 mg/kg), an L-type voltage-operated Ca2+ channel blocker  or citronellol (1, 5, 10 and 20 mg/kg) for obtainment of dose–response curves. Similar records with citronellol were obtained after treatment with atropine, (2 mg/kg; i.v.; 30 min.), a non-selective muscarinic receptor antagonist , Nω-nitro-l-arginine methyl ester hydrochloride (20 mg/kg; i.v.; 30 min.), an inhibitor of the nitric oxide synthase , indomethacin (5 mg/kg; i.v.; 30 min.), a potent non-selective COX inhibitor , or hexamethonium (20 mg/kg; i.v.; 30 min.), a nicotinic receptor antagonist and ganglionic blocker , separately. After protocols, all animals were killed by exsanguination under thiopental anaesthesia.
Tissue preparation. The tissue preparation was performed as described in Menezes et al. . Rats were killed by exsanguination under ether anaesthesia and the superior mesenteric artery was removed, cleaned from connective and fat tissues and sectioned in rings (1–2 mm). These rings were suspended in organ baths containing 10 ml of Tyrode’s solution, gassed with carbogen and maintained at 37°C under a resting tension of 0.75 g for 60 min. (stabilization period). The isometric tension was recorded by a force transducer (Model TRI210; Letica, Barcelona, Italy) coupled to an amplifier-recorder. When necessary, the endothelium was removed and its functionality was assessed by the ability of acetylcholine chloride (10 μM) to induce more than 70% relaxation of phenylephrine (10 μM) tonus. The absence of the relaxation for acetylcholine chloride was taken as evidence that the rings were functionally denuded of endothelium.
Effect of citronellol on phenylephrine (10 μM) or KCl 80 mM tonus in rings with and without endothelium or after tetraethylamonium incubation. Contractions of the vessels were induced with 10 μM phenylephrine or 80 mM KCl in rings with or without the endothelium. During the tonic phase of the contraction, citronellol (6.4 × 10−4, 1.9 × 10−3, 6.4 × 10−3, 1.9 × 10−2, 6.4 × 10−2, 1.9 × 10−1, 6.4 × 10−1 and 1.9 M, cumulatively) was added to the organ bath. The extent of relaxation was expressed as the percentage of phenylephrine- or KCl-induced contraction. Furthermore, curves for citronellol were obtained before and after incubation with 1 mM of tetraethylamonium, a non-selective K+ channel blocker , in rings without the endothelium.
Effect of citronellol on CaCl2 contractions in endothelium-denuded rings. The effect of citronellol on CaCl2 contractions in endothelium-denuded rings was assessed by using a protocol described by Santos et al. . Cumulative concentration–response curves for CaCl2 (3 × 10−6 to 3 × 10−3 M) were obtained in rings without the endothelium exposed to nominally without Ca2+ solution with KCl 60 mM before and after pre-incubation separately with citronellol (1.9 × 10−1 M, 6.4 × 10−1 M and 1.9 M) for 15 min. The results were expressed as percentages of the maximal response for CaCl2 alone and the curves were statistically compared.
Effect of citronellol on phenylephrine- and caffeine-induced contractions in Ca2+-free solution. The effect of citronellol on phenylephrine- or caffeine-sensitive calcium intracellular stores was assessed by using a protocol described by Sakata and Karaki  and Adaramoye et al. . The transient contractions (n = 6) were obtained in endothelium-denuded rings by 10 μM phe or 20 mM caffeine in Ca2+-free solution before and after incubation with citronellol (6.4 × 10−3 to 1.9 M) for 20 min. The results were expressed as percentages of the response induced by phenylephrine or caffeine alone.
Statistical analysis. Values were expressed as the mean ± S.E.M. The results were analysed with one- or two-way repeated-measures anova followed by the Bonferroni post-test or paired or unpaired Student’s t-test. The pD2 values of in vitro experiments were obtained by nonlinear regression. All procedures were performed by using Graph Pad Prism 3.02™. (GraphPad Software, Inc., San Diego, CA, USA)
Effect of citronellol on haemodynamic parameters in non-anaesthetized rats.
In non-anaesthetized rats, baseline mean arterial pressure and heart rate values were 117 ± 5 mmHg and 310 ± 15 bpm respectively. In these animals, the intravenous bolus injections of citronellol (1, 5, 10 and 20 mg/kg) or nifedipine (0.1, 0.5, 1 and 5 mg/kg) induced hypotension associated with tachycardia that was not dose-dependent (one-way anova followed by the Bonferroni post-test) (figs 2 and 3A). The citronellol responses were less than those obtained for nifedipine, a reference drug. The duration of citronellol effect on mean arterial pressure and heart rate was approximately 40 sec. Pre-treatment with atropine rose only the heart rate baseline value from 310 ± 15 to 406 ± 11 bpm, and Nω-nitro-l-arginine methyl ester hydrochloride rose only the blood pressure baseline value from 117 ± 5 to 154 ± 2 mmHg. However, the pre-treatment with hexamethonium only decreased the blood pressure baseline value from 115 ± 8 to 89 ± 3 mmHg. In the pre-treatment with indomethacin, any parameter was modified. These data were analysed with the paired Student’s t-test (p < 0.01 versus Control). The pre-treatment with these drugs did not change the hypotension induced by citronellol. However, atropine and hexamethonium attenuated significantly the tachycardia (0.5 ± 1, 2 ± 1, 5 ± 2 and 3 ± 3%; and 4 ± 2, 3 ± 3, 5 ± 1 and 6 ± 4% respectively) (two-way anova followed by the Bonferroni post-test; p < 0.01) (fig. 3B).
Effect of citronellol on phenylephrine (10 μM) or KCl 80 mM tonus in rings with and without the endothelium or after tetraethylamonium.
In rings of rat mesenteric artery with functional endothelium (Control) pre-contracted with phenylephrine, citronellol induced relaxations (pD2 = 0.71 ± 0.11; Emax = 102 ± 5%; n = 6) that were not changed after removal of the endothelium (pD2 = 1.01 ± 0.09; Emax = 105 ± 4%; n = 6) (fig. 4A) or pre-contraction with KCl 80 mM (pD2 = 1.23 ± 0.14; Emax = 121 ± 6%; n = 6) (fig. 4B). In endothelium-denuded rings pre-contracted with phenylephrine (10 μM), tetraethylamonium (1 mM) was not able to change citronellol relaxations (pD2 = 0.71 ± 0.08; Emax = 107 ± 4%; n = 6) (fig. 4B).
Effect of citronellol on CaCl2 contractions in endothelium-denuded rings.
In endothelium-denuded rings incubating with depolarizing and nominally without Ca2+ solution, CaCl2 (3 × 10−6 to 3 × 10−3 M) was able to induce contractions that were strongly inhibited by citronellol in doses of 1.9 × 10−1 M, 6.4 × 10−1 M and 1.9 M (maximal inhibition = 38 ± 13; 76 ± 2*** and 97 ± 4%***; n = 6 respectively; two-way anova followed by the Bonferroni post-test. ***p < 0.01) (fig. 5).
Effect of citronellol on phenylephrine- and caffeine-induced contractions in Ca2+-free solution.
In mesenteric rings under a Ca2+-free solution, citronellol inhibited transient contractions induced by 10 μM phenylephrine (maximal inhibition = 95.3%) or by 20 mM caffeine (maximal inhibition = 98%) (fig. 6).
In Brazil, many hypertensive patients with associated cardiovascular diseases daily drink tea of medicinal plants containing citronellol and this study demonstrated the possible benefits of this essential oil on the cardiovascular system. Our results demonstrated that citronellol appears to have a calcium-blocking property as many drugs used in the treatment of hypertension such as amilodipine, nifedipine and verapamil . The vasorelaxant activity could be used as a potential substance for antihypertensive treatment.
Our results have demonstrated that, in conscious and freely moving rats, the intravenous administration of citronellol induced intense hypotension associated with tachycardia. Nifedipine, a voltage operated calcium channel (VOCC) L-type (dihydropyridine sensitive) blocker , similar to citronellol, was also able to induce hypotension associated with tachycardia. However, the effects of nifedipine were significantly more than those obtained for citronellol at similar doses. The cardiovascular effects for nifedipine were observed with doses ten times less (0.1–5 mg/kg) than those for citronellol (1–20 mg/kg).
As established in the literature, the vascular tone of the arterial bed underlies the maintenance of peripheral resistance in the circulation and it is the major contributor to the control of blood pressure . Furthermore, in most vascular beds, the activation of muscarinic receptors in the endothelial cells induces vasorelaxation by the release of endothelium-derived relaxant factors (EDRFs), including nitric oxide and cyclooxygenase metabolites, such as prostaciclin (PGI2) . In order to verify the role of muscarinic receptors in hypotensive and tachycardic responses induced by citronellol, we performed experiments in animals pre-treated with atropine, a non-selective antagonist of these receptors. Under this condition, the hypotensive response did not significantly change, suggesting that citronellol does not act via muscarinic receptor activation. However, the tachycardic response was attenuated. The absence of a tachycardic response, after administration of citronellol, appears to be due to the intense positive inotropic response of atropine treatment. Since the heart rate was strongly rose, the tachycardic response induced by the indirect action of citronellol, the response to baroreflex was possibly attenuated. In order to confirm this hypothesis, experiments were performed with hexamethonium, a ganglionic blocker. In this condition, the tachycardic response induced by citronellol was also attenuated without a change in hypotension, suggesting that tachycardia appears to be of reflex origin.
To determine whether the hypotensive effect could involve the release of nitric oxide or PGI2 by an independent way of muscarinic activation, we performed experiments with Nω-nitro-l-arginine methyl ester hydrochloride, an inhibitor of nitric oxide synthase , and indomethacin, a potent non-selective COX inhibitor , separately. As with atropine, neither Nω-nitro-l-arginine methyl ester hydrochloride nor indomethacin was able to alter hypotension, suggesting that nitric oxide or PGI2appears not to be participating in this effect.
In mesenteric artery intact rings, citronellol induced vasorelaxation in a concentration-dependent manner of phenylephrine-induced tonus. In endothelium-denuded rings, or after incubation with tetraethylamonium, a non-selective K+ channel blocker , the relaxant effect induced by citronellol was not modified, suggesting that the presence of the endothelium and the K+ channels is not essential for relaxant response expression. Then, an endothelium- and K+ channel-independent pathway probably are implicated in this effect.
These initial results are in agreement with other reports that have demonstrated that several essential oils present potent hypotensive effect through a direct vasorelaxant action [19,24,25]. In addition, the in vitro findings are in agreement with those obtained in the in vivo experiments, which demonstrated that hypotension induced by citronellol does not appear to be mediated by the endothelial muscarinic receptors or EDRFs.
Interestingly, as blood volume in the rat is approximately 10 ml to each 100 g of body weight , the administration of 1 mg/kg in the in vivo experiments is the same as the incubation with 10 μg/ml (6.4 × 10−2 M) of citronellol in the in vitro experiments. Both doses were able to induce approximately the same percentage of response (31.0 ± 1.9% of hypotension and 41.6 ± 4.6% of vasorelaxation).
Calcium is a primary regulator of tension in the vascular smooth muscle . The maintenance of smooth muscle contraction depends on Ca2+ influx from the extracellular space through voltage- and/or receptor-operated calcium channels (VOCCs and/or receptor operated calcium channels respectively) .
It is also suggested that the increase in external K+ concentration (KCl 80 mM) induces smooth muscle contraction through VOCC activation. This contraction is inhibited by Ca2+ channel blockers or by removal of external Ca2+ and is, therefore, entirely dependent on Ca2+ influx . Based on this assumption, we evaluated the effect of citronellol on endothelium-denuded rings pre-contracted with K+-depolarizing solutions (KCl 80 mM). This experiment revealed that citronellol induced vasorelaxation similar to those observed in rings pre-contracted with phenylephrine. This result suggests that citronellol could inhibit Ca2+ influx through VOCCs.
In order to strengthen the above hypothesis, we constructed concentration–response curves for CaCl2 in a calcium-free medium before and after incubation with citronellol. In this condition, citronellol significantly attenuated the CaCl2-induced contractions. As reported by Chan et al. , nifedipine also inhibited the concentration–response curve for CaCl2, which strongly supports that citronellol could possibly act as a calcium channel blocker.
As citronellol appears to act through the decreasing Ca2+ influx, other pathways cannot be discarded, as for example inhibition of calcium release from calcium intracellular stores. The activation of phosphoinositide turnover in response to receptor activation, as α-adrenoceptors by phenylephrine, is crucial to the cytoplasmatic calcium increase through calcium release of the intracellular stores and consequently contraction. Phenylephrine induces a rapid and transient increase in inositol triphosphate (IP3) formation in vascular smooth muscle, which releases calcium of the IP3-sensitive calcium intracellular stores [30,31]. Furthermore, caffeine/ryanodine-sensitive calcium intracellular store could be activated by caffeine, which activates the ryanodine receptor and also leads to intracellular Ca2+ release .
To investigate the involvement of these stores in the vasorelaxant response induced by citronellol, we performed experiments using rings contracted by phenylephrine or caffeine, in Ca2+-free solution, in the absence and presence of citronellol. Thus, citronellol inhibited transient contractions induced by phenylephrine and caffeine, suggesting that citronellol also interferes in the calcium mobilization of both IP3- and caffeine-sensitive calcium intracellular stores.
The literature has demonstrated that other monoterpenes presented an effect similar to citronellol. Lahlou et al.  showed that terpinen-4-ol, a main monoterpene constituent of Alpinia zerumbet essential oil, promoted hypotension that can be attributed to a direct vasorelaxant action. Aydin et al.  showed that carvacrol, a monoterpene constituent of many essential oils, presents hypotensive and vasorelaxant action possibly due to block of vascular L-type calcium channels. Lahlou et al.  and Guedes et al.  demonstrated that rotundifolone, a monoterpene isolated from Mentha x villosa essential oil, presented hypotensive action through decrease in the peripheral vascular resistance caused by vasorelaxation. This effect was caused by inhibiting voltage-dependent Ca2+ channels and intracellular Ca2+ release .
In summary, our results demonstrate that citronellol lowers blood pressure by a direct effect on the vascular smooth muscle leading to vasodilation.
This work was supported by CNPq and FAPITEC/SE, Brazil.