Introduction
Various stress conditions in daily life influence human health and may cause physical
disorders, such as insomnia [1] and anxiety [2]. Benzodiazepines are widely used as sedative drugs but are associated with problematic
side effects, including oversedation [3] or ataxia [4]. Therefore, as one of the complementary and alternative medicinal therapies, aromatherapy
has recently attracted much attention for its ability to induce relaxation and to
ameliorate stress. Spikenard, the dried underground part of Nardostachys chinensis Batalin, is a member of the family Valerianaceae found growing from Himalayan regions
to southwest China. In traditional Chinese medicine, spikenard has been used as a
gastrointestinal antispasmodic [5] and a traditional herbal tranquilizer [6]. Spikenard is also an essential ingredient in scented sachets used to improve insomnia.
Our research focused on the characteristic fragrance of spikenard, the expression
of the sedative effect by fragrance inhalation, and the identification of active ingredients.
In our previous study, we investigated the sedative effect of the spikenard hexane
extract by inhalation administration in mice. We identified four hydrocarbon sesquiterpenoids
from the spikenard hexane extract as active compounds [7], [8] and also revealed that its sedative effect due to administration by inhalation was
expressed via olfactory stimulation and pulmonary absorption [9]. While preparing the spikenard essential oil to identify other volatile compounds
by GC-MS analysis, we detected several high intensity peaks in addition to the peaks
for hydrocarbon sesquiterpenoids mentioned above ([Fig. 1]). The aim of the present study was to identify those compounds and to determine
whether they contribute to the sedative effect of spikenard. The sedative activities
of isolated compounds were evaluated in a caffeine-treated excitatory mouse model,
using an inhalation administration method and an open-field test. The locomotor activity
of these mice was more than doubled by caffeine administration [8]; therefore, the sedative activity of each compound could be easily detected. We
also investigated the mechanism of the sedative action of the highly active compounds.
To examine the involvement of the GABAergic system, we performed a pentobarbital sleep
test with a GABAA-benzodiazepine receptor antagonist, flumazenil. We describe the characteristic compounds
contained in spikenard and involved in its sedative effect.
Fig. 1 A GC chromatogram of the GC-FID analysis of the spikenard oil. The retention index
(RI) and percent of composition of the ten main peaks are indicated below. 1: β-maaliene, RI 1405, 3.1 %; 2: aristolene, RI 1413, 2.1 %; 3: calarene, RI 1428, 13.5 %;
4: valerena-4,7(11)-diene, RI 1459, 1.9 %; 5: β-ionone, RI 1485, 3.6 %; 6: maaliol, RI 1562, 6.4 %; 7: spathulenol, RI 1573, 5.6 %;
8: patchouli alcohol, RI 1650, 11.6 %; 9: aristolen-1(10)-en-9-ol, RI 1661, 13.6 %;
and 10: 1(10)-aristolen-2-one, RI 1754, 4.4 %.
Results
Six oxygenated sesquiterpenoids were isolated from the spikenard acetone extract in
this study. [Fig. 1] shows the GC-FID chromatogram of the spikenard oil, and the calculated retention
index and content ratio of each compound identified in the spikenard oil in the present
study are indicated. [Fig. 2] shows the chemical structures of these compounds, of which six compounds were aristolane-type
and two compounds were guaiane-type. Aristolen-1(10)-en-9-ol (13.6 %) and patchouli
alcohol (11.6 %) were the main components of the spikenard oil, in addition to the
volatile hydrocarbon calarene (13.5 %). The sedative effects of the six oxygenated
compounds were evaluated.
Fig. 2 Chemical structures of sesquiterpenoids as numbered in the GC chromatogram in [Fig. 1].
To investigate whether the isolated oxygenated sesquiterpenoids contribute to the
effect of spikenard, we assessed the total spontaneous locomotor activity of caffeine-treated
mice in an open-field test. [Fig. 3] shows the total motor activity of mice exposed to various concentrations of sample
solutions for 60 min. Diazepam, a positive control, inhibited the increase in locomotor
activity of caffeine-treated mice dose-dependently, and the inhibition ratios at doses
of 1 and 5 mg/kg were 60 and 80 %, respectively. Inhalation of the spikenard oil significantly
sedated the caffeine-treated mice at a dose of 300 µg/cage (inhibition ratio, 41 %).
Of the isolated compounds, maaliol and β-ionone tended to show a sedative effect but not significantly. In contrast, both
1(10)-aristolen-2-one and spathulenol showed significant sedative activity at a dose
of 300 µg/cage, with the inhibition ratios of 34 and 53 %, respectively. Patchouli
alcohol and aristolen-1(10)-en-9-ol had dose-dependent sedative effects, and the inhibition
ratios at a dose of 300 µg/cage were 69 and 63 %, respectively, and these effects
were comparable to 1 mg/kg diazepam.
Fig. 3 Total motor activity of caffeine-treated mice exposed to the vapor of oxidized sesquiterpenoids
(0, 30, and 300 µg/cage). Results represent mean ± SD of values for each group (n = 8).
*P < 0.05, **p < 0.01 vs. vehicle group; one-way ANOVA followed by Tukeyʼs test.
The involvement of the GABAergic system in the sedative effects of patchouli alcohol
and aristolen-1(10)-en-9-ol was investigated using a pentobarbital sleep test ([Fig. 4]). Orally administered diazepam prolonged pentobarbital-induced sleeping times in
a dose-dependent manner. Compared with the vehicle alone, diazepam significantly prolonged
sleep duration at 1 and 5 mg/kg by 44 min (64 % prolongation) and 107 min (299 % prolongation),
respectively. No effect was observed in the patchouli alcohol inhalation group; however,
inhalation of aristolen-1(10)-en-9-ol (300 µg/cage) significantly prolonged the sleeping
time (47 min, 84 % prolongation), and this effect was comparable to 1 mg/kg diazepam.
To investigate the involvement of the GABAergic system, the mice were pretreated with
flumazenil, a specific GABAA-benzodiazepine antagonist ([Fig. 5]). Pretreatment with flumazenil alone did not affect pentobarbital-induced sleep
duration, but the hypnotic activity of 1 mg/kg diazepam was inhibited by pretreatment
with 3 mg/kg flumazenil. In addition, the hypnotic effect of aristolen-1(10)-en-9-ol
was completely antagonized by flumazenil.
Fig. 4 Effects of vaporized patchouli alcohol (PA) and aristolen-1(10)-en-9-ol (ARI-9-ol)
on sleep duration in a pentobarbital sleep test. Results represent mean ± SD of values
for each group (n = 8). **P < 0.01 vs. vehicle group; one-way ANOVA followed by Tukeyʼs
test.
Fig. 5 Effect of flumazenil (FLU) on sleep duration in mice treated with diazepam (po) and
aristolen-1(10)-en-9-ol (ARI-9-ol) (inhalation). Results represent mean ± SD of values
for each group (n = 8). **P < 0.01 vs. vehicle group (Studentʼs t-test), ##p < 0.01 for FLU treatment vs. treatment without FLU (Studentʼs t-test). NS, not significant.
The influence of drug administration on the motor function was investigated using
a rota-rod test for 5 min. [Fig. 6] shows the endurance time (time to fall) of mice treated with diazepam or aristolen-1(10)-en-9-ol.
In the diazepam group, a dose of 0.2 mg/kg did not disturb the motor coordination.
However, doses of 1 and 5 mg/kg disturbed the motor coordination in a dose-dependent
manner, with mean endurance times of 195 and 37 sec, respectively. Half (4/8) of the
mice in the 1 mg/kg diazepam group and all (8/8) of the mice in the 5 mg/kg group
fell from the rod. Aristolen-1(10)-en-9-ol inhalation for 1 h did not affect the motor
coordination in the rota-rod test.
Fig. 6 Effect of aristolen-1(10)-en-9-ol (ARI-9-ol) on the rotating rod performance of mice.
Each bar represents mean ± SD of the endurance time during the 5 min rota-rod test
(n = 8). **P < 0.01 vs. vehicle group; one-way ANOVA followed by Tukeyʼs test.
Discussion
Spikenard has been widely used as a traditional herbal tranquilizer and in scented
sachets to improve insomnia. Given that inhalation is a noninvasive administration
route, we investigated the sedative effect of volatile compounds to assess the effectiveness
of spikenard as aromatherapy. On GC analysis of the spikenard oil, we identified several
high-intensity peaks in addition to the peaks for the hydrocarbon sesquiterpenoids
identified previously. These compounds were identified as oxidized sesquiterpenoids,
including aristolane- and guaiane-type compounds. Open-field testing of caffeine-treated
mice exposed to aristolen-1(10)-en-9-ol and patchouli alcohol demonstrated that these
compounds have high sedative activity when administered by inhalation. In the pentobarbital-sleep
test, to investigate the involvement of the GABAergic system, aristolen-1(10)-en-9-ol
significantly prolonged the sleeping duration. Because we could not find an appropriate
positive control drug with a known pharmacological mechanism using the same inhalation
administration route, we used diazepam as a positive control. The administration of
5 mg/kg diazepam prolonged the sleeping duration by 299 % and markedly impaired the
motor coordination, which is a well-known undesirable side effect of diazepam. Therefore,
5 mg/kg diazepam was strongly sedative. Diazepam (1 mg/kg) also prolonged the sleeping
duration and mildly impaired the motor coordination. Therefore, 1 mg/kg diazepam was
mildly sedative. The effect of aristolen-1(10)-en-9-ol was comparable to that of 1 mg/kg
diazepam and was completely inhibited by 3 mg/kg flumazenil. Therefore, the sedative
effect of aristolen-1(10)-en-9-ol appeared to involve agonism of the GABAA-benzodiazepine receptor. Rota-rod testing demonstrated that aristolen-1(10)-en-9-ol
induced relaxation without the motor impairment or marked sedation. The sedative effect
of patchouli alcohol appears to be mediated by a system other than the GABAergic system.
A previous study found that the essential oil of Valeriana wallichi, which contains patchouli alcohol (40.2 %), has an antidepressant effect [10]. Furthermore, administration of V. wallichi essential oil increased cerebral serotonin and norepinephrine levels to a similar
extent as imipramine. Imipramine also has a sedative effect; therefore, the mechanism
of action of patchouli alcohol could involve the cerebral monoamine system. However,
further research is needed to confirm this possible mechanism. In the present study,
we identified active components of spikenard and provided evidence supporting the
traditional sedative use of this compound. Although more studies are required to further
elucidate the properties of aristolen-1(10)-en-9-ol, our research suggests that aristolen-1(10)-en-9-ol
may be an effective aromatherapy, providing mild sedation.
Materials and Methods
General
Triethyl citrate (purity > 99 %, GC; Wako Pure Chemical Industries, Ltd.), an odorless
solvent, was used to dissolve the isolated compounds or the spikenard oil. Caffeine
(purity > 98.5 %, TLC) and diazepam (purity > 98 %, Ti) were purchased from Wako Pure
Chemical Industries, Ltd., flumazenil (purity > 99 %, HPLC) was obtained from Sigma-Aldrich,
Inc., and pentobarbital sodium salt (purity > 98 %, N) was obtained from Tokyo Chemical
Industry Co., Ltd. Caffeine and pentobarbital were dissolved in physiological saline,
and diazepam and flumazenil were suspended in a solution of dimethyl sulfoxide : Tween-80:physiological
saline = 1 : 1 : 8. Column chromatography was performed using 60 N silica gel (40–50 µm,
Kanto Chemical Co., Ltd.) and Cosmosil® 75C18-PREP (Nacalai Tesque, Inc.). Preparative HPLC was performed using Cosmosil® 5C18-AR-2 (150 mm × 20 mm ID), and the compounds were detected using UV (SPD-10A; Shimadzu
Corporation) and a refractive index (RI-72, Showa Denko K. K.) detector. All chemicals
and reagents were of the highest grade.
Plant material
Spikenard (lot number: 0080) was purchased from Mitsuboshi Pharmaceutical Co., Ltd.
and identified by the authors. A voucher specimen of the plant material (No. 3735)
was deposited in the Herbarium of Medicinal Plants Garden, Tokyo Metropolitan Institute
of Public Health.
Preparation of the essential oil
Spikenard oil was isolated from the dried plant material by hydrodistillation for
3 h using a Clevenger-type apparatus according to previously reported methods [11].
GC-FID and GC-MS analysis
Quantitative analysis of the volatile component: GC-FID analyses were performed on a GC-2014 (Shimadzu Corporation) with a flame
ionization detector, and operating conditions were as follows: Column, Rtx®-5MS (Restek
Corporation), 30 m × 0.25 mm, 0.25 µm film thickness; column temperature, 60 °C (1 min
hold) = 180 °C (Δ2 °C/min) = 280 °C (Δ10 °C/min, 15 min hold); carrier gas, helium
(39.2 cm/s); injector, 230 °C; detector, 300 °C; injection volume, 1 µL (splitless).
Qualitative analysis of the volatile component: GC-MS analyses were performed on a GC-2010/GCMS-QP2010 Plus (Shimadzu Corporation),
with operating conditions as follows: column and column temperature as above; carrier
gas, helium (38.3 cm/s); injector, 230 °C; interface, 250 °C; ion source, 200 °C;
injection volume, 1 µL (splitless). MS was operated in the EI mode at an ionization
voltage of 70 eV over an m/z range from 30 to 400 amu. Retention indices were determined relative to the retention
times of a series of n-alkane standards (10–20 carbons; GL Sciences, Inc.), measured under the chromatographic
conditions described above.
Extraction and isolation
Spikenard (150 g) was subjected to acetone extraction three times at room temperature.
The acetone extract was concentrated under reduced pressure and silica gel column
chromatography (25 × 5 cm) was performed using the concentrate (12 g). The column
was first eluted with n-hexane to yield fractions 1–5, and then eluted with n-hexane-EtOAc mixtures with increasing polarity (40 : 1 to 2 : 1) to yield fractions
6–12. Fraction 7 (159.3 mg) was purified by preparative HPLC [solvent, CH3CN : H2O (70 : 30); flow rate, 6.0 mL/min], which yielded β-ionone (5.66 mg). Fraction 8 (268.5 mg) was applied to an ODS column (14 × 2.6 cm)
and eluted with CH3CN : H2O (6 : 4 to 10 : 0) to yield fractions 1–5, while further preparative HPLC of the
fraction 8–3 (177.2 mg) [solvent, CH3CN : H2O (82 : 18); flow rate, 6.0 mL/min] yielded patchouli alcohol (154.3 mg). Fraction
10 (978.3 mg) was chromatographed over ODS using the same conditions described above.
Seven fractions were obtained, and further preparative HPLC of fraction 10–2 [734.7 mg;
solvent, CH3CN : H2O (70 : 30); flow rate, 6.0 mL/min] yielded 1(10)-aristolen-2-one (62.8 mg) as fraction
10–2–1. Fraction 10–2–2 was further purified by preparative HPLC [solvent, CH3CN : H2O (73 : 27); flow rate, 6.0 mL/min] to yield spathulenol (60.5 mg), aristolen-1(10)-en-9-ol
(244.3 mg), and maaliol (78.2 mg). The purities of isolated compounds were over 98 %
(GC-FID).
Animals
Four-week-old male ddY mice were purchased from Japan SLC, Inc. Prior to experimentation,
the mice were acclimatized for one week to a temperature of 25 °C ± 2 °C, humidity
of 50 % ± 10 %, and a 12-h light/12-h dark cycle. All behavioral observations were
recorded between 10 : 00 and 15 : 00 h. Experiments were performed in accordance with
the Kitasato University guidelines for animal care, handling, and termination, which
are in line with the international and Japanese guidelines for animal care and welfare
(approval number: FR07–2, date: May 1, 2013).
Inhalation administration
The fragrance components were dissolved in triethyl citrate (30 µL total) for the
experiments. A thick paper disk (2.5 × 3.0 cm, 1 mm in thickness) was permeated with
sample solution and placed on a hotplate diffuser (approximately 70 °C) in the upper
portion of the cage. The cylindrical cage was made of transparent polycarbonate (22 cm
in height, 25 cm in diameter) and covered with vinyl film so the vapor from the sample
solution could pervade the cage. A mouse treated with caffeine or pentobarbital was
placed in the center of the cage 30 min after charging the solution. The fragrance
inhalation dose was set to 30 or 300 µg/cage, on the basis of previous studies [7].
Caffeine administration test
Caffeine was orally administered to mice at a dose of 20 mg/kg. Each mouse was placed
in a cage filled with vapor from the sample solution 30 min after administration,
and the locomotor activity was measured for another 60 min using an open-field test.
As a positive control, diazepam was administered at doses of 0.2, 1, and 5 mg/kg (po)
30 min before caffeine peroral administration. The test was performed according to
the method described by Kobayashi et al. [12] in the same cylindrical cages equipped with a passive infrared sensor (PYS-001;
Muromachi Kikai Co., Ltd.). Spontaneous motor activity was recorded using a passive
infrared sensor detection system (Supermex; Muromachi Kikai Co., Ltd.) and analyzed
using CompACT AMS software (Muromachi Kikai Co., Ltd.).
Pentobarbital sleep test
Mice were administered pentobarbital (ip) at a dose of 30 mg/kg and immediately placed
in a cage filled with vapor of the sample solution (300 µg/cage). Sleep duration was
defined as the time difference between loss and recovery of the righting reflex. As
a positive control, diazepam at a dose of 0.2, 1, and 5 mg/kg (po) was administered
30 min before pentobarbital injection. Flumazenil, a specific GABAA-benzodiazepine receptor antagonist, was administered (3 mg/kg, ip) 60 min prior to
pentobarbital injection.
Rota-rod test
A rota-rod treadmill (MK-600; Muromachi Kikai Co., Ltd) was used in this study. To
adapt to the rota-rod, one day before the test, mice were placed on the rod rotating
at 28 rpm for 5 min each hour three times. On the test day, one more episode of training
was performed, and the mice that failed to stay on the rod were excluded from the
experiment. The duration, up to 5 min, that mice could remain on the rota-rod was
recorded after 60 min of sample inhalation or 30 min after diazepam administration.
Statistical analysis
All results are expressed as the mean ± standard deviation (SD). The data were analyzed
using one-way analysis of variance (ANOVA) followed by Tukeyʼs multiple comparison
test. For the flumazenil treatment experiment ([Fig. 5]), Studentʼs t-test was used to compare the values from the two groups. All statistical
analyses were performed using Prism 5 (GraphPad Software, Inc.), and p values < 0.05
were considered to be significant.
Supporting information
Spectral data for the isolated compounds are available as Supporting Information.
