doi: 10.1111/j.1472-8206.2008.00571.

ORIGINAL Vasorelaxation induced by the essential oil

ARTICLE
of Croton nepetaefolius and its constituents
in rat aorta are partially mediated by the
endothelium
Pedro Jorge Caldas Magalhãesa*, Saad Lahloub, Davi Matthews Jucáa,
Lı́via Noronha Coelho-de-Souzaa, Pedro Thiago Tibúrcio da Frotaa,
Adriana Maria Gurgel da Costaa, José Henrique Leal-Cardosoc
a
Departamento de Fisiologia e Farmacologia, Faculdade de Medicina, Universidade Federal do Ceará, Fortaleza, CE, Brazil
b
Departamento de Fisiologia e Farmacologia, Universidade Federal de Pernambuco, Recife, PE, Brazil
c
Instituto Superior de Ciências Biomédicas, Universidade Estadual do Ceará, Fortaleza, CE, Brazil

Keywords ABSTRACT
Croton nepetaefolius,
essential oil, Previously, we reported that essential oil of Croton nepetaefolius (EOCN) decreases
isolated aorta, blood pressure in normotensive rats, an effect that seems resulting from its
nitric oxide, vasodilatory action directly upon vascular smooth muscle. In the present study, we
rat, aimed to study the role of endothelium–nitric oxide pathway in the mediation of
vascular mesenteric bed vasodilatory effects of EOCN and two of its constituents, methyleugenol and
a-terpineol, using rat isolated thoracic aorta and mesenteric vascular bed prepara-
Received 22 October 2007;
tions. EOCN (1–300 lg/mL), in a concentration-dependent manner, relaxed isolated
revised 28 December 2007; endothelium-intact aortic rings precontracted with KCl 60 mM, with an IC50 value of
accepted 21 January 2008 26.7 (14.7–48.2) lg/mL. Either pretreatment of the tissue with L-NAME, a nitric
oxide synthase inhibitor, or mechanical endothelium removal increased significantly
the IC50 value to 66.6 (52.7–84.1) or 105.6 (91.3–122.2) lg/mL, respectively. In
*Correspondence and reprints:
pjcmagal@ufc.br endothelium-intact aortic rings precontracted with norepinephrine, EOCN (10–
200 lg/mL) produced a vasorelaxant action which was decreased by the pretreat-
ment of the aortic rings with methylene blue, a guanylate cyclase inhibitor. In
mesenteric bed preparations perfused under constant pressure, EOCN reverted the
reduction of mesenteric flow caused by KCl (60 mM), an effect that was attenuated by
L-NAME. Vasodilator responses to EOCN in mesenteric bed preparations were
mimicked by methyleugenol and a-terpineol, and were also significantly reduced in
the presence of L-NAME. In conclusion, EOCN has vasorelaxant effects in both a
resistance vascular bed and in a conduit artery. They seem attributed, at least in part,
to the actions of its main constituents methyleugenol and a-terpineol and appear
partially dependent upon the integrity of a functional vascular endothelium.
Inhibition of other transduction pathways may be involved in the mediation of
these effects.

Recently, an elegant review performed by Salatino et al.

INTRODUCTION
[1] revealed that Croton species and their substances are
Croton is a large genus of Euphorbiaceae, comprising among the highly studied themes in chemistry of natural
around 1300 species of trees, shrubs and herbs distributed products, pharmacology and ethnopharmacology, and
in tropical and subtropical regions of both hemispheres. new data are rapidly accumulating. Croton nepetaefolius

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177 169
170 P.J.C. Magalhães et al.

Baill. is an aromatic plant with a rich essential oil content EOCN induces its vascular smooth muscle relaxation.
corresponding to 1–3% of plant dry weight. This plant is Therefore, the present investigation was performed in rat
native to northeastern Brazil where it is popularly named isolated thoracic aorta and mesenteric vascular bed
‘marmeleiro sabiá’ or ‘marmeleiro-vermelho’ [2]. preparations to assess the role of vascular endothelium
Infusates and teas made from barks and leaves of and L-arginine/nitric oxide (NO) pathway in the medi-
C. nepetaefolius are commonly used in folk medicine ation of vasodilatory effects of EOCN and two of its main
because of its stomachic, carminative and intestinal anti- constituents, methyleugenol and a-terpineol.
spasmodic properties [3,4].
The essential oil of C. nepetaefolius (EOCN) is comprised
MATERIALS AND METHODS
predominantly of terpenes such as methyleugenol and
a-terpineol, which may be involved in its previously Plant material
reported hypotensive, inhibitory and anti-spasmodic Aerial parts of C. nepetaefolius were collected in May
effects on several smooth muscle tissues [5–9]. The 2003, from the region of Viçosa, State of Ceará, Brazil.
inhibitory action of EOCN has an interesting pharmaco- Plant identification was confirmed by Dr A.A. Fernan-
logical profile, as it seems mediated by intracellular des (Department of Biology, Federal University of
mechanisms largely independent of alterations on trans- Ceará). A voucher specimen (No. 3185) is deposited
membrane potential [7]. Furthermore, its large efficacy in the herbarium of Prisco Viana, Federal Universtity of
as a myorelaxant agent associated with its low acute Ceará.
toxicity (its LD50 is >3 g/kg body weight, per os) makes it
an agent of therapeutic potential [5,10]. Orally admin- Extraction and chemical analysis
istered, the EOCN promoted a dose-dependent anti- The EOCN was kindly provided by the Department of
nociceptive effect unrelated to opioid mechanism [10]. Organic and Inorganic Chemistry of Federal University
Previous studies from our laboratory have shown that of Ceará. It was prepared from freshly chopped leaves
EOCN dose-dependently decreased mean aortic pressure by steam distillation and analyzed chemically as
(MAP) and heart rate (HR) in both pentobarbital- previously described [5–9]. EOCN had the following
anaesthetized and conscious normotensive rats [8]. chemical composition (in % of oil weight): 1,8-cineole
Whilst EOCN-induced bradycardia appeared to depend (25.4%), methyleugenol (14.9%), xanthoxilin (10.1%),
upon the presence of an intact and functional para- b-caryophyllene (9.66%), sabinene (5.2%), a-terpineol
sympathetic nerve drive to the heart, EOCN-induced (4.96%) and other minor constituents [8]. These
hypotension appeared independent of the presence of an compounds were identified using a mass spectral
operational sympathetic nervous system [8]. In deoxy- library and [13C]-nuclear magnetic resonance spectro-
corticosterone acetate (DOCA)-salt hypertensive and scopy [11].
uninephrectomized control, conscious rats, EOCN also
decreased MAP and HR in a dose-related manner [9]. In Solutions and drugs
DOCA-salt rats, the hypotensive effects of EOCN were The perfusion medium used was fresh modified Tyrode
significantly enhanced while bradycardia was not solution (pH 7.4) of the following composition (mM):
affected. In isolated thoracic aorta preparations from NaCl 136, KCl 5, MgCl2 0.98, CaCl2 2, NaH2PO4 0.36,
DOCA-salt hypertensive rats, EOCN induced a reduction NaHCO3 11.9 and glucose 5.5. EOCN, methyleugenol
of phenylephrine-induced contraction. Arteries from (Sigma, St Louis, MO, USA) and a-terpineol (Sigma) were
DOCA rats showed enhanced sensitivity to EOCN, as dissolved in Tyrode’s solution, and sonicated just before
compared to uninephrectomized controls. This enhance- use. Norepinephrine (NE) hydrochloride (Sigma), acetyl-
ment appeared to be related mainly to an increase in choline chloride (Sigma), methylene blue hydrate
EOCN-induced vascular smooth muscle relaxation rather (Sigma) and L-NAME (Sigma) were first dissolved in
than to enhance sympathetic nervous system activity in distilled water and were brought to volume with
this hypertensive model [9]. These data thus supported Tyrode’s solution.
the hypothesis that EOCN may be a direct vasorelaxant
agent by a mechanism, probably myogenic, that turns to Animals
be more active in hypertensive rats. Male Wistar rats (200–250 g, n = 45) were obtained
However, no information is available in the inter- from our local colonies maintained at the Department of
national literature regarding the mechanisms by which Physiology and Pharmacology, Federal University of

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
Vasorelaxant effects of the essential oil of Croton nepetaefolius 171

Ceará, Fortaleza, Brazil. They were kept under conditions of the guanylate cyclase enzyme inhibition on tissues
of constant temperature (22 ± 2 C) with a 12-h-light/ contracted with NE (0.3 lM).
12-h-dark cycle and free access to food and water. All
animals were cared for in compliance with the Guide for Measurement of the mesenteric bed flow
the Care and Use of Laboratory Animals, published by In some rats, after removal of the thoracic aortic rings as
the US National Institutes of Health (NIH Publication described above, the superior mesenteric artery was
85–23, revised 1996). All procedures described here identified and cannulated. The cannulated artery and its
were reviewed by and had prior approval from local vascular bed are dissected out and mounted in an organ
animal ethics committee. device developed in our laboratory. Thereafter, the entire
mesenteric vascular bed was perfused at a constant
Measurement of the aortic contractile activity hydrostatic pressure of 53 cm H2O with a modified
Rats were stunned and then exsanguinated. Thoracic Tyrode solution (37 C), continuously aerated. Changes
aortae were removed and immersed in perfusion in the vascular resistance are measured by comparing
medium at room temperature. After removing adhering perfusion flow before and after drug administration. As
fat and connective tissue, the aorta was cut transver- constant pressure was maintained by regulating the
sally into cylindrical ring-like segments (1 · 5 mm) height of the liquid column in order to achieve always a
attached to steel wire triangular pieces (0.3 mm dia- pressure correspondent to 53 cm H2O (equivalent to
meter), which were suspended in 5-mL organ baths 40 mmHg), changes in the vascular resistance of the
containing continuously aerated perfusion medium at perfused bed were inversely proportional to changes in
37 C. Endothelium-containing strips were stretched the perfusion flow.
with a passive tension of 0.5 g and tension was An equilibration period of 30 min was allowed to
recorded using an isometric force transducer (Grass stabilize the mesenteric vascular bed, baseline
Model FTO3, Quincy, MA, USA) connected to a PC- perfusion flow. In control experiments, this was
based Dataq acquisition system (PM-1000; CWE Inc., followed by a single concentration of KCl (60 mM)
Akron, OH, USA). After an equilibration period of at perfusion for 90 min. This was done because a
least 60 min, control contractions were induced by preconstriction is required to induce vasodilation
adding a submaximal concentration (60 mM) of potas- [12]. In solutions with elevated K+ concentrations,
sium chloride (KCl) to the bath. When two successive Na+ was simultaneously decreased to maintain osmo-
control contractions showed similar amplitudes, prepa- larity. In the experimental group, after 20 min of the
rations were considered to be equilibrated. To assess the beginning of the perfusion with KCl, responses to
vasorelaxant effects of EOCN, preparations were exposed increasing concentrations (1–300 lg/mL) of EOCN,
during 5-min period to increasing concentrations methyleugenol or a-terpineol alone (n = 5, 4 and 4,
(1–300 lg/mL, n = 8) of EOCN once a sustained respectively) or in association with L-NAME (50 lM)
contraction elicited by a 60 mM KCl was established. (n = 5, 4 and 4, respectively) were performed in a
EOCN was directly added to the buffer solution in a cumulative manner. A washout was then allowed
cumulative manner. until the baseline flow rates were achieved. Concen-
In order to examine whether the vascular responses tration–response curves to EOCN, methyleugenol
are dependent on endothelial integrity or NO synthase or a-terpineol were determined only once in each
(NOS) activity, the relaxant effects of EOCN were mesenteric vascular bed preparation and they were
determined in preparations without functional endothe- directly added to the physiological solution.
lium (n = 5) or in endothelium-intact aortic rings
(n = 5) incubated with L-NAME (100 lM). Endothelium Statistical analysis
was removed immediately after dissection by gentle All the results are expressed as mean ± SEM. The
rubbing of the aortic lumen with a stainless steel wire. IC50 value, defined as the EOCN concentration (lg/mL)
Each endothelium-containing or -denuded preparation required to produce half maximum reduction of KCl-
was challenged at the beginning of the experiment with induced contraction, was used to evaluate vascular
1 lM of acetylcholine. The lack of acetylcholine-induced sensitivity to EOCN. It was calculated by interpolation
vasorelaxant effects was taken as evidence that the from semi-logarithmic plots, and expressed as geometric
preparation was effectively denuded of endothelium. means (95% confidence interval). The significance
Methylene blue (10 lM) was used to study the influence (P < 0.05) of the results was assessed by means of

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
172 P.J.C. Magalhães et al.

RESULTS
Vasorelaxant effects of EOCN in rat isolated aortic
rings precontracted with K+ and the influence of
the endothelium removal or L-NAME
The cumulative addition of the EOCN (1–300 lg/mL)
produced a concentration-related vasorelaxant response
(P < 0.001, one-way ANOVA) in rings of rat aorta with
intact endothelium precontracted with 60 mM KCl
(Figure 1a, upper trace), with an EC50 value correspond-
ing to 26.7 (14.7–48.2) lg/mL (Figure 1b). The inhi-
bitory effect of EOCN became significant at a
concentration of 3 lg/mL (P < 0.05, Holm–Sidak test;
Figure 1b). In endothelium-denuded rings (Figure 1a,
lower trace), the vasorelaxant action of the EOCN was
significantly decreased, as the IC50 value was signifi-
cantly increased to 105.6 (91.3–122.2) lg/mL
(P < 0.05, unpaired Student’s t-test). The EC50 for the
vasorelaxant effect of EOCN was also significantly
augmented by L-NAME (10)4 M) to 66.6 (52.7–
84.1) lg/mL (P < 0.05, unpaired Student’s t-test;
Figure 1b). In presence of L-NAME or in denuded tissues,
the inhibitory effect of EOCN became significant at a
concentration of 3 or 30 lg/mL, respectively (P < 0.05,
Holm–Sidak test; Figure 1b). Contractions induced by
60 mM KCl in endothelium-denuded rings (0.80 ±
0.14 g) or in rings with intact endothelium maintained
in presence of L-NAME (1.01 ± 0.25 g) were signifi-
Figure 1 Vasorelaxant effects of the essential oil of Croton nep-
cantly higher than that observed in endothelium intact
etaefolius (EOCN) in rat isolated aortic rings. (a) Typical traces
showing the vasorelaxant effects of the EOCN. Tissues were preparations (0.33 ± 0,07 g) (P < 0.05, Holm–Sidak
contracted with a submaximal concentration (60 mM) of KCl (K60) test).
and, after a stable plateau has been achieved, EOCN (1 to 300 lg/
mL) was cumulatively added to produce vascular relaxation in Vasorelaxant effects of EOCN in rat isolated aortic
intact (upper trace, +endothelium) or denuded (lower trace, rings precontracted with norepinephrine and the
)endothelium) aortic rings. Dashed lines indicate baseline to each influence of methylene blue
experiment. (b) Graph showing mean values of the vasorelaxant
In endothelium-intact rat aortic rings precontracted with
effects of EOCN on rat isolated aortic rings with intact endothelium
(control, n = 8), without endothelium ()endothelium, n = 5) or
NE (0.3 lM, Figure 2a), the cumulative addition of EOCN
endothelium-containing rings pretreated with L-NAME 100 lM (10, 100 and 200 lg/mL) produced a vasorelaxant
(L-NAME 100, n = 5). Responses are expressed as % of K+ (60 mM)- action correspondent, respectively, to 33.8 ± 9.6,
induced contraction before EOCN addition. Vertical bars indicate 71.4 ± 9.7 and 82.8 ± 7.1% of the contraction obtained
SEM. The vasorelaxant effect of EOCN was significantly (P < 0.001, in the absence of EOCN (Figure 2b). In the presence of
two-way ANOVA) reduced by either mechanical removal of the methylene blue (10)5 M), vasorelaxant effects of EOCN
vascular endothelium or L-NAME pretreatment. *P < 0.05, value
(10, 100 and 200 lg/mL) were about 13.3 ± 4.7,
significantly different when compared with control responses
46.7 ± 4.1 and 67.4 ± 4.4% of the control NE-induced
obtained in absence of EOCN at each experimental protocol (Holm–
Sidak test). contraction, respectively. These responses were signifi-
cantly reduced only at 10 and 100 lg/mL of EOCN
(P < 0.05, unpaired Student’s t-test). Contractions in-
paired or unpaired Student’s t-tests, and one- or two-way duced by NE (0.3 lM) in rings with intact endothelium
analysis of variance (ANOVA), followed by Holm–Sidak’s maintained in presence of methylene blue (1.09 ±
multiple comparison test. 0.14 g) were significantly higher than that observed in

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
Vasorelaxant effects of the essential oil of Croton nepetaefolius 173

(a)

(b)

Figure 2 Effects of methylene blue on the relaxation induced by

EOCN in rat isolated aortic rings. (a) Typical traces showing the
vasorelaxant effects of the EOCN (10, 100 and 200 lg/mL) on
endothelium-intact aortic rings contracted with norepinephrine
(0.3 lM; NE 0.3) in absence (left panel) or in presence of methylene
blue (10 lM, Met Blue 10) (right panel). Dashed lines indicate
(c)
baseline to each experiment. (b) Graph showing mean values of the
vasorelaxant effects of EOCN on rat isolated aortic rings with intact
endothelium in absence (control) or in presence of Met Blue.
Ordinate values express relaxation which was expressed as % of the
NE-induced contraction before EOCN addition. Vertical bars indicate
SEM (n = 5 per group). The vasorelaxant effects of EOCN were
significantly (*P < 0.05, paired Student’s t-test) reduced in the
presence of Met blue.

control preparations (0.69 ± 0,10 g) (P < 0.05, paired

Student’s t-test).

Inhibitory effects of EOCN in rat isolated

Figure 3 Vasorelaxant effects of EOCN in rat mesenteric vascular
mesenteric vascular bed preparations
bed preparations. (a) Reversible mesenteric flow responses of
Under constant pressure (53 cm H2O), rat mesenteric increasing concentrations of EOCN (10, 100 and 300 lg/mL) in rat
bed perfused with a solution containing a normal K+ mesenteric vascular bed preparations initially perfused by a high
concentration (5 mM) showed a control flow of potassium (60 mM, K60) solution for @60 min. Ordinate values are
2.9 ± 0.4 mL/min. Subsequent perfusion of the mesen- expressed as percentage of the control flow obtained during
teric bed in vitro with a K+-enriched (60 mM) solution mesenteric perfusion with normal solution ([K+]e 5 mM) and
produced a vasoconstrictor response, as occurred a vertical bars indicate SEM (n = 7). (b) and (c) show that EOCN
elicited vasodilator responses were fully abolished by L-NAME
significant reduction (P < 0.05, paired Student’s t test)
(n = 5). *P < 0.05 by Holm–Sidak test, compared with corre-
of the control mesenteric flow to a value of 1.8 ±
sponding baseline values obtained in the absence of EOCN.
0.3 mL/min, which is correspondent to 60.5 ± 5.9% of
the control flow (Figure 3a). After mesenteric flow each concentration) produced an enhancement of the
reached a plateau under high K+ perfusion, cumulative mesenteric flow towards control flow values, in a
addition of EOCN (10, 100 and 300 lg/mL, 5 min for reversible manner, although this effect was significant

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
174 P.J.C. Magalhães et al.

only at 300 lg/mL (P < 0.05, Holm–Sidak test) (n = 7). produce any significant vasorelaxant effect (P > 0.05,
This effect is indicative of vasodilation. At 300 lg/mL, ANOVA; Figure 3c).
the mesenteric flow was increased to a value corre-
sponding to 85.3 ± 5.2% of the control flow (Figure 3b). Inhibitory effects of methyleugenol and a-terpineol
When mesenteric bed was previously perfused in in rat isolated mesenteric bed preparations
control period with a solution containing both K+ After K+-reduced mesenteric flow reached a plateau,
(5 mM) and L-NAME (10)4 M) (n = 5), its flow was methyleugenol (10, 100 and 300 lg/mL) caused a
significantly reduced (P < 0.05, paired Student’s t-test) recovery of perfusion flow, in a concentration-dependent
by the subsequent perfusion with high K+ solution (from manner (P < 0.001, one-way ANOVA, Figure 4a). This
2.8 ± 0.3 to 0.9 ± 0.1 mL/min, which is correspondent effect was significant at 100 and 300 lg/mL (P < 0.05,
to 33.9 ± 4.8% of the control flow) (Figure 3). Under Holm–Sidak test), which reverted the flow to 73.0 ± 2.3
these conditions, cumulative addition of EOCN (10, 100 and 97.9 ± 1.6% (n = 4) of the control flow, respec-
and 300 lg/mL, 5 min for each concentration) did not tively. In the presence of L-NAME (50 lM), the methyl-
eugenol vasorelaxant effect was reduced as it was
significant only at 300 lg/mL (P < 0.05, Holm–Sidak
(a)
test) (n = 4).
Another constituent of EOCN, a-terpineol (10, 100
and 300 lg/mL) also caused a vasorelaxation, although
its effect was significant only at 300 lg/mL (P < 0.05,
Holm–Sidak test, Figure 4b), which reverted the mesen-
teric flow to 93.8 ± 3.4% of the control flow (n = 7).
The vasorelaxant effect of a-terpineol (300 lg/mL) was
abolished by the previous perfusion of the mesenteric bed
with L-NAME (50 lM, n = 7).

DISCUSSION
(b)
Our previous hypothesis that EOCN-induced hypotension
results mainly from a vasodilatory action of EOCN
directly upon vascular smooth muscle [9] is further
corroborated by the present in vitro findings, that were
obtained using either a conduit artery or a resistance
vascular bed preparations [13,14]. Our results showed
that EOCN induces concentration-dependent vasodilator
effects in potassium (60 mM)-precontracted rat aortic
ring or mesenteric vascular bed preparations with intact
endothelium from normotensive rats. This effect seems
partly because of release of an endothelium-derived
relaxing factor, notably NO, as it was partly reduced by
Figure 4 Vasorelaxant effects of methyleugenol and a-terpineol in either endothelium removal or pretreatment with
rat mesenteric vascular bed preparations. Vasodilator responses to L-NAME.
of increasing concentrations (10–300 lg/mL) of methyleugenol (a) Vasodilator responses to EOCN in mesenteric bed
or a-terpineol (b) on potassium (60 mM)-precontracted rat mesen- preparations are reported for the first time, and were
teric vascular bed preparations in the absence (left panel) or in the mimicked by methyleugenol and a-terpineol, two con-
presence (right panel) of L-NAME (50 lM). Ordinate values are stituents of the EOCN. It might have suggested that
expressed as percentage of the control flow obtained during
EOCN-induced vasorelaxant effects have been related to
mesenteric perfusion with normal solution ([K+]e 5 mM) and
vertical bars indicate SEM (n = 4–7 per group). Vasodilator
a putative toxic effect of this essential oil. However, two
responses to methyleugenol and a-terpineol were also significantly lines of evidences do not support this hypothesis. First, in
reduced by L-NAME. *P < 0.05, by Holm–Sidak test, compared with our experimental conditions, all vasodilator responses to
corresponding baseline values obtained in the absence of EOCN. EOCN and its constituents were reversible. Second, oral

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
Vasorelaxant effects of the essential oil of Croton nepetaefolius 175

acute toxicity LD50 value for EOCN [10] has been found As is well known, NO is synthesized from guanidine
greater than 3000 mg/kg, and that for methyleugenol groups of L-arginine [20] which has been reported to
[15] and a-terpineol [16] was about 1200 and stimulate the production of cGMP leading to a vaso-
2900 mg/kg, respectively. These values are indicative relaxation in smooth muscle cells [21]. As in NE-
of low toxicity of the EOCN and its constituents. precontracted aortic rings pretreated with methylene
In the present study, putative participation of the blue, an inhibitor of guanylate cyclase [22], the effect of
vascular endothelium in mediation of EOCN-induced EOCN was significantly reduced, we hypothesized that
relaxation has been investigated. Our results show that EOCN may relax vascular tissue through an interference
the vasorelaxant effects of EOCN in aortic ring prepara- with the NO-cGMP cellular pathway. However, whether
tions were partially, but significantly reduced by EOCN acts by increasing guanylate cyclase activity or
mechanical removal of the endothelium, as evidenced decreasing the phosphodiesterase enzymatic actions
by the significant increase in the IC50 of EOCN-induced remains to be further investigated.
reduction of the potassium-induced contractions. This It has been described that the vasorelaxation induced
suggests that the vasorelaxation induced by EOCN is by NO is dependent on a reduction in intracellular
partly mediated by NO or prostacyclin release. Given the calcium in smooth muscle cells, not only due to the
fact that vasorelaxant responses to EOCN were also activation of soluble guanylate cyclase, but also due to
partially, but significantly reduced following NOS inhi- increased K+ efflux following the activation of K+ chan-
bition by L-NAME [17] supports the hypothesis that nels, which may to lead away to hyperpolarization of
EOCN may latterly stimulate endothelial NO production smooth muscle cells [23–25]. A putative EOCN-induced
and/or release. Whether EOCN is directly interacting opening of a K+ channel seems unlikely, because EOCN
with endothelial NOS (e-NOS) or with other factors, completely relaxed the contraction induced by 60 mM K+.
which may increase the e-NOS activity remains to be In fact, in sustained contractions induced by high K+
further investigated. A similar mechanism of action of concentrations, a K+ channel opener cannot hyperpolar-
EOCN can be advanced in mesenteric bed preparations ize smooth muscle, as the transmembrane potential come
because EOCN-induced vasorelaxant effects (at 300 lg/ close to K+ equilibrium potential [26]. On the other hand,
mL) were also partially reduced in the presence of endothelium-dependent vasorelaxants, such as acetyl-
L-NAME, but remained unaffected by the inactive isomer choline, could also release an endothelium-derived
D-NAME. hyperpolarizing factor (EDHF), which induces vasorelax-
As the contractions obtained after endothelium re- ation by membrane hyperpolarization [27]. However, the
moval, NO synthase or guanilate cyclase inhibition were release of such EDHF as primary mechanism of action
higher than that observed in endothelium intact aortic EOCN appears improbable because preliminary experi-
rings, a possible functional antagonism could be present. ments showed that a concentration of EOCN (200 lg/
However, two lines of evidence argue against the possi- mL), which produced almost maximal relaxation of both
bility of this non-specific decrease in the EOCN-induced rat and guinea-pig aortae, caused no significant changes
relaxation. Firstly, significant decreases in the vasorelax- in intracellularly measured transmembrane potential of
ant effects of EOCN in denuded preparations were observed guinea-pig maintained in normal or in depolarizing
only at low concentrations, whereas at higher concentra- 60 mM K+ solution (unpublished data).
tions the endothelium removal appears to be ineffective The present work demonstrates that the EOCN-
upon EOCN-elicited effects. This profile of action was induced vasorelaxant effects were only partially de-
observed in either KCl- or NE-contracted preparations. creased after either endothelium removal or L-NAME
Thus, if a decrease in the intensity of relaxation occurs by a pretreatment, indicating that other mechanisms may be
functional antagonism, then it would have been observed involved in the inhibitory effects of EOCN. Recently, we
at all concentrations used. Secondly, under similar con- reported that EOCN reduces basal tone of the isolated
ditions described in the present study, the vasodilator guinea-pig ileum [7]. The myorelaxant and anti-spas-
effects of the essential oil of Croton zehntneri [18] or Labd-8 modic activity of EOCN is myogenic and independent of
(17)-en-15-oic acid (Labd-8) [19] observed experimen- neural activity and possibly reflects an ability of EOCN to
tally in aortic rings, were not influenced by the endothelial decrease the sensitivity of the contractile proteins to
abrasion or pharmacological impairment, despite the calcium. This pharmacological profile also appears to be
contractions recorded also showed higher amplitudes important in vascular tissue. In fact, experiments
when compared to control conditions. performed in isolated guinea-pig aorta preparations

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
176 P.J.C. Magalhães et al.

showed that the concentration-dependent contractions REFERENCES

elicited by phorbol ester in calcium-free medium with
1 Salatino A., Salatino M.L.F., Negri G. Traditional uses, chem-
EGTA (2 mM) were significantly reduced by EOCN istry and pharmacology of Croton species (Euphorbiaceae).
(200 lg/mL), an finding suggestive of the ability of J. Braz. Chem. Soc. (2007) 18 11–33.
EOCN in inhibiting protein kinase C (Magalhães P.J.C., 2 Craveiro A.A., Fernandes A.G., Andrade C.H.S., Matos F.J.A.,
personal communication). Alencar J.W., Machado M.I.L. Óleos essenciais de plantas do
The vasorelaxant activity of EOCN in mesenteric nordeste. Edições UFC, Fortaleza, CE, Brazil, 1981.
vascular bed preparations appear partly attributed to 3 Craveiro A.A., Andrade C., Matos F.J.A., Alencar J.W., Dantas
T.N.C. Fixed and volatile constituents of Croton aff. Nepetaefo-
the actions of two of its constituents, methyleugenol and
lius. J. Nat. Prod. (1980) 43 756–757.
a-terpineol. These constituents were also able to induce a 4 Leal-Cardoso J.H., Fonteles M.C. Pharmacological effects of
reversible and concentration-dependent vasorelaxation essential oils of plants of the northeast of Brazil. An. Acad. Bras.
that was reduced or abolished under blockade of NOS by Cienc. (1999) 71 207–213.
L-NAME. Methyleugenol has been reported to induce 5 Magalhães P.J.C., Criddle D.N., Tavares R.A., Melo E.M., Mota
vasorelaxant effects on isolated potassium-precontracted T.L., Leal-Cardoso J.H. Intestinal myorelaxant and anti-spas-
thoracic aortic preparations from normotensive rats with modic effects of the essential oil of Croton nepetaefolius, and its
constituents cineole, methyl-eugenol and terpineol. Phytother.
an IC50 (32.2 lg/mL) [28] very close to that reported
Res. (1998) 12 172–177.
herein with the EOCN (26.6 lg/mL). As only 14.9% of 6 Magalhães P.J.C., Lahlou S., Vasconcelos dos Santos M.A.,
the EOCN was methyleugenol, this observation indicates Pradines T.L., Leal-Cardoso J.H. Myorelaxant effects of the
that methyleugenol is not the only constituent that essential oil of Croton nepetaefolius on the contractile activity of
mediates the vasorelaxant activity of EOCN. To the best the guinea-pig tracheal smooth muscle. Planta Med. (2003) 69
of our knowledge, this is also the first demonstration of 874–877.
the a-terpineol-induced relaxant properties on vascular 7 Magalhães P.J.C., Lahlou S., Leal-Cardoso J.H. Antispasmodic
effects of the essential oil of Croton nepetaefolius on guinea-pig
smooth muscle. Other constituents, such a 1,8-cineole
ileum: a myogenic activity. Fundam. Clin. Pharmacol. (2004)
(the main constituent with 25.4% of total weight of the
18 539–546.
oil), may also contribute to the vasorelaxant activity of 8 Lahlou S., Leal-Cardoso J.H., Magalhães P.J.C., Coelho-de-Souza
EOCN as it was reported to inhibit significantly the KCl- A.N., Duarte G.P. Cardiovascular effects of the essential oil of
induced contractions in aortic rings from normotensive Croton nepetaefolius in rats: role of the autonomic nervous
rats [29]. system. Planta Med. (1999) 65 553–557.
9 Lahlou S., Leal-Cardoso J.H., Magalhães P.J.C. Essential oil of
Croton nepetaefolius decreases blood pressure through an action
CONCLUSION upon vascular smooth muscle: studies in DOCA-salt hyperten-
sive rats. Planta Med. (2000) 66 138–143.
In conclusion, the vasorelaxant effects of EOCN are 10 Abdon A.P., Leal-Cardoso J.H., Coelho-de-Souza A.N., Morais
attributed, at least in part, to the actions of its main S.M., Santos C.F. Antinociceptive effects of the essential oil of
constituents methyleugenol and a-terpineol and seem Croton nepetaefolius on mice. Braz. J. Med. Biol. Res. (2002) 35
partially dependent upon the integrity of functional 1215–1219.
endothelium. Beside the involvement of endothelial 11 Alencar J.W., Craveiro A.A., Matos F.J.A. Kovats indices as a
L-arginine/NO pathways, inhibition of other transduction pre-selection routine in mass spectra library search of volatiles.
J. Nat. Prod. (1984) 47 890–892.
pathways may be involved in the mediation of EOCN-
12 Boric M.P., Figueroa X.F., Donoso M.V., Paredes A., Poblete I.,
induced vasodilatory effects in vascular smooth muscle. Huidobro-Toro J.P. Rise in endothelium-derived NO after
stimulation of rat perivascular sympathetic mesenteric nerves.
Am J. Physiol. Heart Circ. Physiol. (1999) 277 H1027–H1035.
ACKNOWLEDGEMENTS
13 Tanaka Y., Tang G., Takizawa K. et al. Kv channels contribute
This work was funded by CNPq and FUNCAP grants. The to nitric oxide- and atrial natriuretic peptide-induced relaxation
authors are grateful to Prof Nilberto Robson Falcão of a rat conduit artery. J. Pharmacol. Exp. Ther. (2006) 317
341–354.
do Nascimento for its assistance in vascular mesenteric
14 O’Sullivan S.E., Kendall D.A., Randall M.D. Heterogeneity in the
bed preparation experiments. L.N. Coelho-de-Souza,
mechanisms of vasorelaxation to anandamide in resistance and
P.T.T. da Frota and A.M.G. da Costa are integrants of conduit rat mesenteric arteries. Br. J. Pharmacol. (2004) 142
the PIBIC/CNPq Fellowship Program. D.M. Jucá is a 435–442.
postgraduate student in Pharmacology (MSc, CNPq 15 NTP (National Toxicology Program) Toxicology and carcino-
Fellow). genesis of methyleugenol (CAS No. 93-15-12) in F344/N rats

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177
Vasorelaxant effects of the essential oil of Croton nepetaefolius 177

and B6C3F1 mice (gavage studies). U.S. Department of Health 23 Tare M., Parkington H.C., Coleman H.A., Neild T.O., Dusting
and Human Services, Washington DC, 1998, DRAFT NTP-TR- G.J. Hyperpolarization and relaxation of arterial smooth muscle
491; NIH Publication No. 98-3950. caused by nitric oxide derived from endothelium. Nature (1990)
16 MSDS (Material Safety Data Sheets) section 11, Sigma-Aldrich 346 69–71.
Company, available at: http://www.sigmaaldrich.com (accessed 24 Lang R.J., Harvey J.R., Mcphee G.J., Klemm M.F.. Nitric oxide
on May 16, 2007). and thiol reagent modulation of Ca2+ activated (BKCa) channels
17 Moncada S., Palmer R.M.J., Higgs E.A. Nitric oxide: physiology, myocytes of the guinea-pig taenia caeci. J. Physiol. (2000) 525
pathophysiology and pharmacology. Pharmacol. Rev. (1991) 363–376.
43 109–142. 25 Robertson B.E., Schubert R., Hescheler J., Nelson M.T. CGMP-
18 Siqueira R.J., Magalhães P.J.C., Leal-Cardoso J.H., Duarte G.P., dependent protein kinase activates Ca-activated K channels in
Lahlou S. Cardiovascular effects of the essential oil of Croton cerebral artery smooth muscle cells. Am. J. Physiol. (1993) 265
zehntneri leaves and its main constituents, anethole and C299–C303.
estragole, in normotensive conscious rats. Life Sci. (2006) 78 26 Adeagbo A.S., Triggle C.R. Varying extracellular [K+]: a
2365–2372. functional approach to separating EDHF- and EDNO-related
19 Lahlou S., Barros-Correia C.A., Santos M.V. et al. Mechanisms mechanisms in perfused rat mesenteric arterial bed. J. Cardio-
underlying the cardiovascular effects of a labdenic diterpene vasc. Pharmacol. (1993) 21 423–429.
isolated from Moldenhawera nutans in normotensive rats. 27 Feletou M., Vanhoutte P.M. Endothelium-dependent hyperpo-
Vascul. Pharmacol. (2007) 46 60–66. larisation of canine coronary smooth muscle. Br. J. Pharmacol.
20 Palmer R.M., Ferrige A.G., Moncada S. Nitric oxide release (1988) 93 515–524.
accounts for the biological activity of endothelium-derived 28 Lahlou S., Figuereido A.F., Magalhães P.J.C., Leal-Cardoso J.H.,
relaxing factor. Nature (1987) 327 524–526. Duarte G.P. Cardiovascular effects of methyleugenol, a natural
21 Arnold W.P., Mittal C.K., Katsuki S., Murad F. Nitric oxide constituent of many plant essential oils, in normotensive rats.
activates guanylate cyclase and increases guanosine 3¢:5¢-cyclic Life Sci. (2004) 74 2401–2412.
monophosphate levels in various tissue preparations. Proc. 29 Lahlou S., Figuereido A.F., Magalhães P.J.C., Leal-Cardoso J.H.
Natl. Acad. Sci. U.S.A. (1977) 74 3203–3207. Cardiovascular effects of 1,8-cineole, a terpenoid oxide present
22 Gruetter C.A., Kadowitz P.J., Ignarro L.J. Methylene blue in many plant essential oils, in normotensive rats. Can. J.
inhibits coronary arterial relaxation and guanylate cyclase Physiol. Pharmacol. (2002) 80 1125–1130.
activation by nitroglycerin, sodium nitrite, and amyl nitrite.
Can. J. Physiol. Pharmacol. (1981) 59 150–156.

ª 2008 The Authors Journal compilation ª 2008 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 22 (2008) 169–177