Binding activity of Valeriana fauriei root extract on GABAA receptor flunitrazepam sites and distribution of its active ingredients in the brain of mice – A comparison with that of V. officinalis root

Article abstract of DOI:10.1016/j.jep.2021.114262

Ethnopharmacological relevance: Valeriana fauriei root (VF) is a crude drug registered in the Japanese Pharmacopeia 17th Edition and a known substitute for V. officinalis (VO). Although VO has been pharmacologically evaluated for its sedative effects and mechanism of action, data regarding VF remain scarce. Aim of the study: We compared the binding affinity of VF and VO extracts, as well as examined the active ingredients in the VF extract, on flunitrazepam sites of γ-aminobutyric acid receptor type A (GABA_A receptor). Furthermore, we confirmed whether these active ingredients were distributed in the brain of mice orally administered the VF extract. Materials and methods: We prepared the assay system to evaluate the binding activity of flunitrazepam sites of GABA_A receptor using a 96-well plate and assessed the activities of VF and VO extracts. We then analyzed their constituents using HPLC with principal component analysis (PCA) and evaluated active ingredients correlated with their activities. The distribution of active ingredients in the plasma and brain of mice orally administered the VF extract prepared with different emulsifiers were analyzed by LC-MS/MS. Results: The ethanol extract of VF exhibited significantly higher activity on flunitrazepam sites of GABA_A receptor than VO. For the VF extract, kessyl glycol diacetate (KGD) was markedly associated with the binding activities; however, active ingredients included KGD, kessyl glycol 8-acetate (KG8), α-kessyl acetate (α-KA), and coniferyl isovalerate (CI). For VO, valerenic acid and five other compounds were associated with the binding affinity on flunitrazepam sites of GABA_A receptor. On emulsifying the VF extract with a fat-soluble glycerin fatty acid ester, the plasma and brain distributions of KGD tended to be higher, those of KG8 were significantly more than 10-times higher, and those of α-KA was lower than those of the VF extract emulsified with water-soluble gum arabic, after oral administration in mice. Conclusions: Based on the binding activity on flunitrazepam sites of GABA_A receptor and brain distribution, KGD, KG8, and α-KA can be considered active ingredients of VF. The addition of a fat-soluble emulsifier promoted the absorption of KGD, the main active ingredient, and KGD was metabolized to KG8 in the body. The present results suggest a possible mechanism underlying the sedative effect for VF, and these three compounds can be used as marker compounds to evaluate the quality of VF products.

Full text of DOI:10.1016/j.jep.2021.114262

Journal of Ethnopharmacology 278 (2021) 114262
Contents lists available at ScienceDirect
Journal of Ethnopharmacology
journal homepage: www.elsevier.com/locate/jethpharm
Binding activity of Valeriana fauriei root extract on GABAA receptor
flunitrazepam sites and distribution of its active ingredients in the brain of
mice – A comparison with that of V. officinalis root
,
,
Misato Otaa b, Hao Nia, Yasuhito Makia b, Daiki Katob, Shohei Moriguchib, Shuto Nakayamab,
Yuki Oiwab, Kan’ichiro Ishiuchia, Toshiaki Makinoa
,
*
a Department of Pharmacognosy, Graduate School of Pharmaceutical Sciences, Nagoya City University, 3-1 Tanabe-Dori, Mizuho-ku, Nagoya, Aichi, 467-8603, Japan
b Kuki Sangyo Co., Ltd., 11 Onoe-cho, Yokkaichi-shi, Mie, 510-0059, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Ethnopharmacological relevance: Valeriana fauriei root (VF) is a crude drug registered in the Japanese Pharma-
copeia 17th Edition and a known substitute for V. officinalis (VO). Although VO has been pharmacologically
evaluated for its sedative effects and mechanism of action, data regarding VF remain scarce.
Aim of the study: We compared the binding affinity of VF and VO extracts, as well as examined the active in-
gredients in the VF extract, on flunitrazepam sites of γ-aminobutyric acid receptor type A (GABAA receptor).
Furthermore, we confirmed whether these active ingredients were distributed in the brain of mice orally
administered the VF extract.
Valeriana fauriei root
γ-aminobutyric acid type A receptor
Kessyl glycol diacetate
Chemical compounds studied in this article:
CI
Coniferyl isovalerate
α
α
-KA
Materials and methods: We prepared the assay system to evaluate the binding activity of flunitrazepam sites of
GABAA receptor using a 96-well plate and assessed the activities of VF and VO extracts. We then analyzed their
constituents using HPLC with principal component analysis (PCA) and evaluated active ingredients correlated
with their activities. The distribution of active ingredients in the plasma and brain of mice orally administered
the VF extract prepared with different emulsifiers were analyzed by LC-MS/MS.
-Kessyl acetate
KG
Kessyl glycol
KG2
Kessyl glycol 2-acetate
KG8
Results: The ethanol extract of VF exhibited significantly higher activity on flunitrazepam sites of GABAA receptor
than VO. For the VF extract, kessyl glycol diacetate (KGD) was markedly associated with the binding activities;
Kessyl glycol 8-acetate
KGD
however, active ingredients included KGD, kessyl glycol 8-acetate (KG8), α-kessyl acetate (α-KA), and coniferyl
Kessyl glycol diacetate
PHB
isovalerate (CI). For VO, valerenic acid and five other compounds were associated with the binding affinity on
flunitrazepam sites of GABAA receptor. On emulsifying the VF extract with a fat-soluble glycerin fatty acid ester,
the plasma and brain distributions of KGD tended to be higher, those of KG8 were significantly more than 10-
p-hydroxybenzoic acid butyl ester
SDS
Sodium lauryl sulfate
Umbelliferone
Valerenic acid
times higher, and those of α-KA was lower than those of the VF extract emulsified with water-soluble gum
arabic, after oral administration in mice.
Conclusions: Based on the binding activity on flunitrazepam sites of GABAA receptor and brain distribution, KGD,
KG8, and α-KA can be considered active ingredients of VF. The addition of a fat-soluble emulsifier promoted the
absorption of KGD, the main active ingredient, and KGD was metabolized to KG8 in the body. The present results
suggest a possible mechanism underlying the sedative effect for VF, and these three compounds can be used as
marker compounds to evaluate the quality of VF products.
Device Regulatory Science Society of Japan, 2016) and is used to treat
diseases in women. The root of European valerian V. officinalis L. (VO) is
a crude drug registered in the 10th Edition of the European Pharmaco-
peia (European Directorate for the Quality of Medicines & Health Care,
1. Introduction
Dried root of Valeriana fauriei Briq. (VF) is a crude drug registered in
2020), frequently used to treat depression and insomnia or to improve
the Japanese Pharmacopeia 17th Edition (Pharmaceutical and Medical
* Corresponding author.
E-mail addresses: prota-misato@phar.nagoya-cu.ac.jp (M. Ota), winniesakurai@163.com (H. Ni), y_maki@kuki-info.co.jp (Y. Maki), katou@kuki-info.co.jp
(D. Kato), s_moriguchi@kuki-info.co.jp (S. Moriguchi), shuto.r0925@gmail.com (S. Nakayama), y_ooiwa@kuki-info.co.jp (Y. Oiwa), ishiuchi@phar.nagoya-cu.ac.
jp (K. Ishiuchi), makino@phar.nagoya-cu.ac.jp (T. Makino).
https://doi.org/10.1016/j.jep.2021.114262
Received 4 March 2021; Received in revised form 21 May 2021; Accepted 27 May 2021
Available online 8 June 2021
0378-8741/© 2021 Elsevier B.V. All rights reserved.

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
A list of abbreviations
KG8
KGD
NSB
PBS
PC
kessyl glycol 8-acetate
kessyl glycol diacetate
nonspecific binding
ANOVA one-way analysis of variance
Cbrain
Cplasma
CI
brain concentration
phosphate buffered saline (0.15 M, pH 7.2)
principal component
plasma concentration
coniferyl isovalerate
PCA
PHB
SDS
TB
PC analysis
DPM
EC50
GA
disintegrations per minute
half maximal effective concentrations
gum arabic
p-hydroxybenzoic acid butyl ester
sodium lauryl sulfate
total binding
GABA
GFE
γ-aminobutyric acid receptor type A, GABAA receptor
glycerin fatty acid ester
TLC
T-PBS
VF
thin layer chromatography
0.05% Tween 20 in PBS
Valeriana fauriei Briq. root
Valeriana officinalis L. root
α
-KA
α-kessyl acetate
KG
kessyl glycol
VO
KG2
kessyl glycol 2-acetate
sleep conditions as a sedative and tranquilizing drug (Weiss et al., 2002).
In 1886, VO was registered in the Japanese Pharmacopeia 1st Edition;
however, it disappeared after the 4th edition published in 1920.
Conversely, VF has been registered in the Japanese Pharmacopeia since
the 2nd edition, published in 1891 until the current edition (Yanagisawa
et al., 1993; Yanagisawa, 2013; Ota et al., in press); this indicates the
history of VF as a substitute for VO in Japan.
2-acetate (KG2) (Takamura et al., 1973, 1975a, 1975b; Hikino et al.,
1980; Sakamoto et al., 1992). Both KGD-rich and α-KA-rich VF extract
reportedly prolong the pentobarbital-induced sleep time in mice
(Yoshitomi et al., 2000). Furthermore, it has been reported that the VF
extract possesses antidepressant-like effects in mice following restraint
stress by inhibiting brain-derived neurotrophic factor; the active ingre-
dient was identified as α-kessyl alcohol (Oshima et al., 1995; Choi et al.,
For VF and VO, the chemical composition of essential oils differs
markedly, with kessyl alcohols and their acetates identified as charac-
teristic ingredients of VF; for example, kessyl glycol diacetate (KGD)
(Fig. 1) (Houghton, 1988; Suzuki et al., 1993). Reportedly, there are two
types of VF: one mainly contains KGD, and the other contains less KGD
2020). In addition, the VF extract revealed anxiolytic effects in mice by
evoking physiological stresses using an elevated plus-maze test (Kim
et al., 2015). However, the relationship between the mechanism un-
derlying its pharmacological effects and active ingredients remains
elusive.
but high amounts of
α
-kessyl acetate (
α-KA) in the essential oil (Suzuki
Reportedly, inhalation of VO essential oil significantly prolonged the
pentobarbital-induced sleeping time in rats (Komori et al., 2006). A VO
extract significantly shortened sleep latency without altering the
sleep-wakefulness cycle in rats (Tokunaga et al., 2007; Han et al., 2018).
Valerenic acid, a major component of VO, has exhibited anxiolytic ef-
fects by modulating γ-aminobutyric acid receptor type A (GABAA re-
ceptor), and those effect was antagonized by bicuculline (Yuan et al.,
2004). Moreover, valerenic acid had binding affinity of the fluni-
trazepam and flumazenil sites of GABAA receptor, and neurons
et al., 1993). In contrast, VO contains large amounts of bornyl acetate
and isovaleric acid, although its sedative activity can be attributed to the
valerenic acid composition (Houghton, 1988; Hendriks et al., 1981;
Safaralie et al., 2008). As only a few ingredients in VF and VO are
similar, their pharmacological activities should be distinct.
Previous animal studies have revealed that a VF extract prolonged
the hexobarbital-induced sleeping time in mice and revealed active in-
gredients such as KGD, kessyl glycol 8-acetate (KG8), and kessyl glycol
Fig. 1. Chemical structure of the ingredients in Valeriana fauriei root (VF).
2

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
expressing β3-containing GABAA receptor were a major cellular sub-
strate (Benke et al., 2009). GABAA receptors are ligand-gated ion
channels expressed in the brain. Several psychological disorders,
including epilepsy, anxiety, and sleep disorders, can be treated by
activating the GABAA receptor response (Le et al., 2013). However, it
remains unknown whether the essential oil components of VF modulate
GABAA receptors and whether these highly lipophilic compounds are
distributed in the brain. Although GABAA receptor has some subtypes
(Sieghart and Savic, 2018), since valerenic acid had high affinity for the
flunitrazepam binding sites (Benke et al., 2009), thus in this study also
focused on the sites.
was sonicated in 1 ml of ethanol and centrifuged at 1.5 × 104 rpm for 5
min. The supernatants were transferred to fresh tubes, evaporated under
N2 flow at room temperature, and finally lyophilized. The ratios of ex-
tracts yielded were 2.9%, 3.1%, 2.1%, 2.7%, and 2.7% for Vf1–Vf5, and
1.8%, 2.6%, 1.0%, and 1.2% for Vo1–Vo4, respectively. The extracts
were dissolved in ethanol at 100 mg/ml and maintained at ꢀ 20 ◦C until
use.
2.4. Isolation of active ingredients from VF extract
Vf2 (1.0 kg) was extracted in 4 l of hexane by sonication for 1 h. The
solution was filtered using a filter paper, with the residue extracted in a
similar manner. The filtrate was evaporated at 40 ◦C under reduced
pressure to obtain the extract (11 g). The extract was subjected to open
silica gel chromatography eluted with ethyl acetate/hexane (2:8 → 5:5),
and the fractions F1 (4.5 g), F2 (0.39 g), F3 (0.37 g), F4 (0.051 g), F5
(0.11 g), F6 (5.9 g), F7 (0.12 g), and F8 (0.14 g) were obtained by silica
gel thin-layer chromatography (TLC) developed with hexane/ethyl ac-
etate (8:2) or chloroform/methanol (98:2), and colored by
anisaldehyde-sulfate solution. Fraction F6 exhibited the largest single
spot with an Rf value of 0.5, developed by chloroform/methanol (98:2),
and compound 1 (1.9 g) was obtained by recrystallization with 2-prop-
anol. Fraction F2 (0.39 g) was further applied to open silica gel chro-
matography eluted with chloroform/methanol (99:1), and then,
compound 2 (84 mg) was isolated. Fraction F4 (51 mg) was further
applied to open silica gel chromatography eluted with chloroform/
methanol (99:1), and then, compound 3 (21 mg) was isolated. Fraction
F8 (0.14 g) was further applied to open silica gel chromatography eluted
with chloroform/methanol (95:5), and then, compounds 4 (28 mg) and
5 (18 mg) were isolated.
The objective of the present study was to elucidate the binding af-
finity on flunitrazepam sites of GABAA receptor of the VF extract in rats
in vitro, to compare the activities with a VO extract, and to investigate
the correlation between VF and VO extract components. We then
examined the active ingredients in the VF extract using an in vitro assay,
confirming their distribution in the brains of mice following oral
administration.
2. Materials and methods
2.1. Materials
We used six batches of VF (Vf1–6) and four batches of VO (Vo1–4)
samples, and their product information is presented in Table 1. Vf1–5
and Vo1–4 samples were used for in vitro and comparison studies, and
Vf6 was used for the in vivo study. These samples were identified by
anatomical studies with reference (Wang et al., 1998). Voucher speci-
mens of these samples were deposited at Kuki Sangyo Co., Ltd.
The chemical structures of compounds were analyzed using 1H and
13C-NMR and ESITOFMS. Compounds 1, 2, 4, and 5 were identified as
2.2. Animals
KGD, α-KA, KG8, and kessyl glycol (KG), respectively, according to a
Male Wistar/ST rats (7 weeks old) and male ddY mice (5 weeks old)
were obtained from Japan SLC (Shizuoka, Japan). The experimental
procedures were conducted according to Nagoya City University
Guidelines for the Care and Use of Laboratory Animals, and the study
protocol was approved by the local Animal Ethics Committee of Nagoya
City University.
previous study (Nishiya et al., 1992). Compound 3 was identified as
coniferyl isovalerate (CI), which was first isolated from the aerial part of
Eremantus incanus (Bohlmann et al., 1980) (Fig. 1).
2.5. Synthesis of KGD derivatives by deacetylation of KGD
2.3. Preparation of the extracts of VF
In brief, KGD (0.500 g) was dissolved in 100 ml of 2 M trifluoroacetic
acid, and the solution was incubated at 30 ◦C for 24 h. After neutrali-
zation with sodium hydrogen carbonate, 100 ml of ethyl acetate was
added, and the organic layer was collected. The same procedure was
repeated twice for the aqueous layer, and the total organic layer was
evaporated and dried using anhydrous sodium sulfate.
Vf2 (10 g) was extracted with 40 ml of ethanol or 70% ethanol at
room temperature by standing for 48 h. The solution was filtered using a
filter paper, and the residue was extracted in a similar manner. The
filtrates were merged and evaporated, and the ratios of the extracts
yielded were 3.4% and 18%, respectively. Vf6 (3.0 kg) was similarly
extracted with ethanol (7.5 l), and the ratio of the extract yielded was
3.9%.
The organic layer was subjected to open silica gel chromatography
and eluted with chloroform/isopropanol (95:5), KG8 (5.4 mg), KG 2-ac-
etate (KG2) (0.125 mg), and KG (65.9 mg), and the compounds were
identified by employing above listed methods.
For the comparison study between VO and VF, each sample (0.50 g)
Table 1
Sample list of the roots of Valeriana fauriei and V. officinalis.
No.
Original plant species
Production area
Distributor
Lot No.
Obtained year
Voucher No.
Vf1
Vf2
Vf3
Vf4
Vf5
Vf6
Vo1
Vo2
Vo3
Vo4
Valeriana fauriei
Mie, Japan
Mie, Japan
Hokkaido, Japan
Hokkaido, Japan
Hokkaido, Japan
Mie, Japan
Poland
Kuki Sangyo
Kuki Sangyo
Uchida
KUK-00002
2017
2018
2017
2017
2017
2019
2018
2018
2018
2018
KUK-00002
A2018031901
KUK-00008
KUK-00009
KUK-00016
A2019031401
KUK-00036
KUK-00037
KUK-00038
KUK-00039
A2018031901
EAG0110
Tochimoto
Maechu
1816001
20-SN2016A
Kuki Sangyo
Koujien
A2019031401
V. officinalis
–
–
–
–
´
Poland
Cafe de Savon
United States
Germany
Sakurakougetu
Tree of Life
Voucher specimens were deposited at Kuki Sangyo Co., Ltd. Samples without lot numbers are written as "–". Kuki Sangyo, Mie, Japan; Uchida, Uchida Wakanyaku,
´
Tokyo, Japan; Tochimoto, Tochimoto Tenkaido, Osaka, Japan; Maechu, Nara, Japan; Koujien, Osaka, Japan; Cafe de Savon, Yamanashi, Japan; Sakurakougetu, Shiga,
Japan; Tree of Life, Tokyo, Japan.
3

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
2.6. Preparation of brain homogenate
Milford, MA, USA). The analytical column used was a Mightysil RP-18
GP II (4.6 i.d. × 250 mm, 5 μm, Kanto Chemical). For condition 1, the
The preparation of the brain homogenate was performed with
modification according to methods described in previous studies (Risa
et al., 2004; Choi et al., 2017). Rats were sacrificed by CO2 inhalation,
the heads were separated from the body, and the cerebral cortex was
collected. Cerebral cortex samples from eight rats were homogenized in
40 ml of Tris-citrate buffer (50 mM, pH 7.1), and the suspension was
centrifuged at 4 ◦C for 15 min at 2.7 × 104 × g. The sediments were
washed three times with Tris-citrate buffer. The washed sediment was
resuspended in 40 ml Tris-citrate buffer. The suspension was incubated
at 37 ◦C for 30 min and then centrifuged for 10 min at 2.7 × 104 × g. The
final sediment was resuspended in 60 ml of Tris-citrate buffer and stored
in aliquots at ꢀ 80 ◦C until use. The protein concentration of this sus-
pension was 7.94 mg/ml, as measured using a BCA™ Protein Assay Kit
(Thermo Scientific, Rockford, IL, USA).
mobile phase was passaged using a linear gradient elution system of
0.1% formic acid (A) and methanol (B), at a flow rate of 1.0 ml/min,
with the following gradient profile: increasing from 10% B to 20% (0–7
min), increasing from 20% to 25% (7–13 min), 25% (13–28 min),
increasing from 25% to 35% (28–29 min), increasing from 35% to 45%
(29–43 min), increasing from 45% to 85% (43–48 min), increasing from
85% to 100% (48–53 min), and 100% (53–63 min). For condition 2, the
mobile phase used an isocratic system with 70% B. The column tem-
perature was 40 ◦C for both conditions, and the detection wavelengths
for conditions 1 and 2 were 210–340 nm and 198 nm, respectively. This
method was based on previous studies assessing VO (Navarrete et al.,
´
2006; Lucio-Gutierrez et al., 2012). The peaks of valerenic acid and KGD
were identified by comparison with standard compounds (valerenic
acid, Extrasynthese, Lyon, France).
2.9. Quantitative analysis using LC-MS/MS for VF extracts
2.7. Binding assay for flunitrazepam sites of GABAA receptor
LC/MS/MS system (Quattro Premier XE, Waters) was used. Herein,
the mass spectrometer used an electrospray ionization source in positive
ion mode with multiple reaction monitoring. The analytical column used
The GABAA receptor binding assay was performed as described
previously (Benke et al., 2009; Choi et al., 2017; Risa et al., 2004), with
the modification using a 96-well plate instead of a glass filter. The wells
was an Inertsil ODS-4, 2.1 i.d. × 100 mm, 3 μm (GL Sciences Inc., Tokyo,
of a 96-well plate were coated with poly L-lysine solution (20 μg/ml,
Japan). The mobile phase was delivered using a linear gradient elution
system with 0.5% acetic acid (A) and acetonitrile containing 0.5% acetic
acid (B), at a flow rate of 0.25 ml/min, with the following gradient
profile: increasing from 25% B to 30% (0–0.8 min), increasing from 30%
to 60% (0.8–3.2 min), 60% (3.2–7.6 min), increasing from 60% to 80%
(7.6–7.7 min), and 80% (7.7–9.0 min). The column temperature was
maintained at room temperature. The transitions (precursor to
daughter) were monitored, with retention times detected as 339.4 to
201.2 m/z for KGD (6.2 min), 297.4 to 291.2 m/z for KG2 and KG8 (4.4
and 5.1 min, respectively), 255.4 to 219.2 m/z for KG (3.0 min), 281.4 to
Sigma Aldrich, St. Louis, MO, USA) for 1 h at room temperature. After
washing the wells with H2O, 50 l of rat brain homogenate (4 mg pro-
tein/ml) was added to each well and dried by standing on a hot plate
(40 ◦C) overnight. After washing the wells with 200
l of 0.05% Tween
20 in phosphate-buffered saline (T-PBS; 0.15 M, pH 7.2) three times,
200 l of Block Ace® solution (blocking buffer; 40 mg/ml in Tris-citrate
μ
μ
μ
buffer, pH 7.1, Snow Brand Milk Products, Tokyo, Japan) was added to
wells, and the plate was incubated at 37 ◦C for 1 h. The samples dis-
solved in ethanol was collected into the test tubes, and dried up under N2
gas flow, and then, dissolved in blocking buffer directly by sonication at
room temperature for more than 30 min. After discarding the blocking
203.2 m/z for α-KA (7.4 min), 163.1 to 130.9 m/z for CI (6.2 min), and
195.2 to 139.0 m/z for p-hydroxybenzoic acid butyl ester (PHB, Nacalai
buffer in 96-well plate, 50 μl of each sample or blank solution in blocking
Tesque, 5.7 min; used as an internal standard). Linear regressions over
buffer and 50
μ
l of [3H]-flunitrazepam (1 nM in blocking buffer,
the concentration range 7.81 ng/ml– 2.00 μg/ml for KGD, KG2, KG8,
American Radiolabeled Chemicals, St. Louis, MO, USA) were added to
the wells; the plate was incubated at 4 ◦C overnight. Diazepam (Fujifilm
Wako Pure Chemicals, Osaka, Japan) was used as the positive control.
For the nonspecific binding (NSB) group, another excessive concentra-
α
-KA, KG, and CI were examined using the peak-area ratios of the
compounds to their internal standards and the least-squares method (r2
> 0.98).
tion of clonazepam (20 μM, Fujifilm Wako) was selected based on the
2.10. Animal experiments
previous literature (Benke et al., 2009; Risa et al., 2004). In addition,
total binding (TB) was determined using a blank blocking buffer. After
The Vf6 ethanol extract was mixed with the same amount of dextrin
(VIANDEX-BH®, Showa Sangyo, Tokyo, Japan), and the following
samples were prepared: sample A, glycerin fatty acid ester (GFE; EXCEL
S-95®, Kao, Tokyo, Japan) and sesame oil (Kuki Sangyo, Mie, Japan)
mixed in a ratio of 0.5:14 (w/w). Then, 0.79 g of Vf6 extract/dextrin
mixture was mixed with 10 ml of GFE/sesame oil mixture; sample B,
0.79 g of Vf6 extract/dextrin mixture was mixed with 10 ml of 5% gum
arabic (GA; Kanto Chemical, Tokyo, Japan). The difference between
these two emulsifiers is that GFE uses oil as the solvent, and GA uses
water. The Vf6 extract/dextrin mixture (0.79 g) corresponded to 10 g of
the Vf6 sample.
washing the wells with 200 μl of T-PBS five times, 100 μl of 10% sodium
lauryl sulfate (SDS, Nacalai Tesque, Kyoto, Japan) solution was added to
the wells, and the plate was incubated at room temperature for 1 h. The
contents of each well were transferred into liquid scintillation vials
containing 3 ml Clear-sol® (Nacalai Tesque), and the radioactivity
(disintegration per minute, DPM) was measured using a liquid scintil-
lation counter.
The following equation was employed to determine the displacement
percentage of radioligand binding:
(
)
DPMsample – DPMNSB
DPMTB – DPMNSB
Binding displacement (%) = 1 ꢀ
× 100
Mice were fasted overnight and then orally administered 10 ml/kg of
sample A or B. Thirty minutes after administration, mice were sacrificed
by CO2 inhalation, and blood samples were collected from the abdom-
inal vein into heparinized tubes; brain tissues were then separated. The
blood samples were centrifuged at 4 ◦C for 7 min at 1.4 × 104 × g, and
the plasma samples were maintained at ꢀ 20 ◦C until analysis. The
preparation of brain samples was performed in accordance with previ-
ously described methods, with slight modifications (Hashiguchi et al.,
2001; Fong et al., 2017). Individual brain tissue was washed with
ice-cold PBS, gently wiped with paper to remove excess water, and then
the weight of brain tissue was recorded. The brain tissue was dipped into
2-fold weight of ethanol, and homogenized using Pestle tissue grinder.
The half-maximal effective concentrations (EC50) were calculated
from the least square regression line plotted from 3 points that crossed
50% of the control logarithmic concentration values.
2.8. Qualitative analysis using HPLC
For each powdered sample (Vf1–5 and Vo1–4), 200 mg was soni-
cated in 10 ml of ethanol for 30 min. After centrifugation at 3500 rpm
for 10 min, the supernatants were filtered through a 0.45 μm membrane
filter and analyzed using an HPLC system (Alliance HPLC, Waters,
4

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
The brain homogenates were centrifuged at 4 ◦C for 7 min at 1.4 × 104 ×
sites of GABAA receptor. Both extracts exhibited similar binding activ-
ities in a concentration-dependent manner, and the concentration
related to the crude drug weight (1.00 g/ml) exhibited approximately
100% binding activity. For ethanol and 70% ethanol extracts, EC50
values were 161 crude drug mg/ml and 143 crude drug mg/ml,
respectively (Fig. 2). These values were equivalent to 5.38 extract mg/
ml and 25.9 mg/ml, respectively.
g. Then, plasma and brain samples (15 μl each) were mixed with 37.5 μl
of ethanol containing 140 ng/ml of umbelliferone (Fujifilm Wako) used
as an internal standard. The samples were centrifuged at 4 ◦C for 7 min
at 1.4 × 104 × g, and supernatants were analyzed using HPLC (DIONEX
Ultimate 3000) with an MS/MS (TSQ QUANTUM ACCESS MAX) system
(Thermo Fisher Scientific, MA, USA).
The mass spectrometer used an electrospray ionization source in
positive ion mode with multiple reaction monitoring. A Mightysil RP-18
3.3. Comparison of binding activity between ethanol extracts of VF and
VO
GP II, 3.0 i.d. × 150 mm, 3 μm (Kanto Chemical) analytical column was
used. The mobile phase was delivered using a linear gradient elution
system with 0.1% formic acid (A) and acetonitrile (B) at a flow rate of
0.5 ml/min, with the following gradient profile: increasing from 25% B
to 30% (0–1 min), increasing from 30% to 60% (14. min), 60% (4–11.8
min), increasing from 60% to 90% (11.8–11.9 min), and 90%
(11.9–15.9 min). The column temperature was set at 40 ◦C. The tran-
sitions (precursor to daughter) were monitored, and retention times
were as follows: 339.27 to 219.22 m/z for KGD (7.7 min), 237.20 to
219.22 m/z for KG2 and KG8 (5.3 and 6.1 min, respectively), 255.23 to
We prepared ethanol extracts of five batches of VF (Vf1–5) and four
batches of VO (Vo1–4). The binding activities of VF extracts were
significantly higher (p = 0.002) than those of VO extracts (Fig. 3).
3.4. Comparison of VF and VO constituents
Fig. 4 shows the chromatographs of Vf2 and Vo4 samples as repre-
sentative data. Vf2 presented a smaller number of UV-detectable peaks
when compared with Vo4 in the chromatograph using condition 1
(Fig. 4a and b); however, in the chromatograph using condition 2, KGD
was detected in Vf2 (Fig. 4c). We compared the peak areas of VF and VO
samples and observed characteristic peaks. To consider the extraction
efficiencies of each sample, the value of the peak area divided by the
extraction efficiency converted from the maximum value to 1 was
adopted. In the VF samples, the peak areas of 5 peaks (peak nos. 7, 23,
35, 36, and 44) and 13 peaks (peak nos. 8, 12, 18, 21, 25, 29, 30, 32, 33,
34, 40, 41, and 43) were significantly higher and lower than those in the
VO samples, respectively (Table 2).
219.22 m/z for KG (3.7 min), 263.28 to 203.34 m/z for α-KA (11.2 min),
163.14 to 131.13 m/z for CI (7.6 min), and 204.18 to 163.17 m/z for
umbelliferone (3.4 min). The standard solution was diluted with plasma
or brain homogenates prepared from normal rats. In plasma samples, the
calibration curve was found to be linear over the concentration range of
7.81 ng/ml–2.00
μ
g/ml for KGD, KG2, and KG, 1.96 ng/ml–2.00
μ
g/ml
for KG8, and 31.3 ng/ml–2.00
μ
g/ml for
α-KA, respectively. As CI
demonstrates a high binding rate to albumin in plasma and could not be
detected, the calibration curve was not plotted. In brain samples, the
calibration curve was linear over the concentration range of 1.95 ng/
ml–2.00
μ
g/ml for KGD, KG8, and
α-KA, 0.488 ng/ml–2.00
μg/ml for CI,
Then, PCA was performed for the peak area of 44 peaks for VF and
VO samples, as described in Table 2. Fig. 5 shows the PCA score plots of
samples and loading plots of the peaks. The first two PCs accounted for
78% (PC1, 54%; PC2, 24%). The PC1–PC2 score plot clearly shows the
separation between the VF and VO samples. The loading plots of
PC1–PC2 score showed four peaks (peak nos. 44, 35, 23, and 36) and
seven peaks (peak nos. 27, 29, 26, 32, 28, 41, and 34), which mainly
contributed to VF and VO, respectively, and peaks 2 and 10 contributed
to both VF and VO.
and 7.81 ng/ml–2.00 μg/ml for KG and KG2, respectively. These were
examined using the peak-area ratios of compounds to their internal
standards and the least-squares method (r2 > 0.99).
2.11. Statistical analysis
Principal component analysis (PCA) was performed using the R
software (Version 3.5.1, https://www.r-project.org). Statistical analyses
of the data included were performed by one-way analysis of variance
(ANOVA) and Dunnett’s s multiple t-tests for comparison of multiple
data, with two-way ANOVA employed for comparing dose-dependent
activities between two groups. The analyses were conducted using
Rstudio software (Version 1.1.447). A P-value of less than 0.05 (p <
0.05) was considered significant. All data are expressed as mean ±
standard error (S.E.). In the animal study, samples with p < 0.05,
determined by the Smirnov-Grubbs test, were removed as abnormal
values.
3. Results
3.1. Confirmation for the binding assay system
To validate the modified binding assay system using a 96-well plate,
we evaluated the competitive binding effects of diazepam, the repre-
sentative benzodiazepine, and radiolabeled flunitrazepam (5 nM), a
benzodiazepine agonist with specific activity on the GABAA receptor
benzodiazepine-binding site, in the wells of 96-well plates coated with
the rat brain homogenate. Diazepam exhibited significant competition
with flunitrazepam in a concentration-dependent manner; 300 nM
diazepam demonstrated nearly 100% binding capacity (Suppl. Fig. 1).
3.2. Binding effect of ethanol or 70% ethanol extract of VF on
flunitrazepam sites of GABAA receptor
Fig. 2. Binding capacity of ethanol or 70% ethanol extract of Valeriana fauriei
root (VF) to GABAA receptor in rat brain homogenate. Binding capacity was
calculated as the ratio to the value of excessive clonazepam binding. Data were
mean ± S.E. (n = 4).**p < 0.01, ***p < 0.001 vs control group (0 nM) by
Dunnett’s multiple t-test.
We prepared two different extracts from VF samples using ethanol
and 70% ethanol and evaluated their binding activities on flunitrazepam
5

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
3.7. Calculation of the binding activity of each compound contained in
ethanol extract of VF on flunitrazepam sites of GABAA receptor
The content of KGD, KG2, KG8, α-KA, KG, and CI in each ethanol or
70% ethanol extract of VF was measured by LC-MS/MS. In both extracts,
KGD was the major constituent detected compounds analyzed; the
contents of other constituents were one-tenth or less than that of KGD. CI
was present only in the ethanol extract, whereas the amount of KG8 and
KG in the 70% ethanol extract was higher than those in the ethanol
extract (Table 4).
We then calculated the binding titers of these compounds on fluni-
trazepam sites of GABAA receptor when ethanol and 70% ethanol ex-
tracts of VF presented EC50 values. For the ethanol extract of VF, the
EC50 value (161 crude drug mg/ml) was equivalent to the concentra-
tions of KG (0.18 mM), KG2 (0.01 mM), KG8 (0.08 mM), KGD (4.44
mM),
α-KA (0.26 mM), and CI (0.37 mM), respectively. From the
activity-concentration curve shown in Fig. 6, KG (0.18 mM) and KG2
(0.01 mM) did not exhibit any binding activity, whereas binding activ-
ities of KG8 (0.08 mM), KGD (4.44 mM), α-KA (0.26 mM), and CI (0.37
mM) were estimated to be 5%, 38%, 10%, and 14%, respectively.
For the 70% ethanol extract of VF, EC50 values (143 crude drug mg/
ml) were equivalent to the concentrations of KG (0.77 mM), KG2 (0.01
Fig. 3. Binding capacity of the ethanol extracts of Valeriana fauriei root (VF)
and V. officinalis roots (VO) to GABAA receptor in rat brain homogenate.
Binding capacity was calculated as the ratio to the value of excessive clonaze-
pam binding. Each diamond symbol represents a data point of 1 batch of the
sample (n = 5 for VF and n = 4 for VO). Diazepam used 1 batch, and data are
shown as the mean of quadruple experiments. The p-value was calculated using
the Student’s t-test.
mM), KG8 (0.19 mM), KGD (3.93 mM), and α-KA (0.23 mM), respec-
tively. Similar to the above calculation, KG (0.77 mM) and KG2 (0.01
mM) did not exhibit any binding activity, whereas binding activities of
KG8 (0.19 mM), KGD (3.93 mM), and
be 5%, 38%, and 9%, respectively.
α-KA (0.23 mM) were estimated to
3.8. Distribution of active ingredients in the brain of mice orally
administered VF extract
3.5. Correlation between binding activities on flunitrazepam sites of
GABAA receptor and the compounds in VF and VO
We added two types of emulsifiers, GFE and GA, to the ethanol
extract of VF, and then examined differences in the constituent con-
centrations, both in the plasma and brain, 30 min after oral adminis-
tration in mice. The brain (Cbrain) and plasma (Cplasma) concentrations of
We evaluated the correlation between the binding activities on flu-
nitrazepam sites of GABAA receptor and peak areas of VF and VO, as
shown in Table 3. The peaks showing significant differences between VF
and VO are described in Table 2, and characteristic peaks for PCA
analysis shown in Fig. 5 were selected. In the PCA analysis, peaks 2 and
10 contributing to both VF and VO correlated with the activity of VO but
not with that of VF. Among characteristic VF peaks, peak No. 44,
identified as KGD, was highly correlated with activity (r = 0.91).
Meanwhile, among the 16 characteristic peaks of VO, 9 peaks (peak nos.
8, 12, 18, 27, 28, 32, 34, 40, and 41) correlated with the activities (r ≥
0.7). Although three peaks (peak nos. 12, 28, and 40) were also detected
for VF, these peaks did not correlate with the activities. Peak No. 32 was
identified as valerenic acid.
constituent compounds are shown in Table 5. KGD, KG8, KG2, and α-KA
were detected in the plasma and brain of mice treated with both prep-
arations. CI was only marginally detected in the brain. On using either
GFE or GA for emulsification, KG in the plasma and brain were below the
quantification limit of determination because of large impurities.
Although KGD was the main component detected in the VF extract,
the brain distribution of KGD was the lowest among those of other
compounds, except for CI, on using GFE as the emulsifier. The brain
distributions of KG2 and KG8 were more than 10 times higher than those
of the other compounds. On using GA for emulsification, the brain dis-
tribution of KGD was the lowest among those of other compounds,
except for CI. On comparing the two preparations, the brain distribu-
tions of KG2, KG8, and KGD were higher when GFE was used as an
emulsifier than when GA was employed. However, the brain distribution
3.6. Exploration of active ingredients in VF extract
To explore the binding activities of active ingredients on fluni-
trazepam sites of GABAA receptor in VF, we prepared a hexane extract of
VF and conducted activity-guided fractionation. First, we prepared
fractions F1–8 and observed that significant activity of KGD contained
F6, as described below. Among other fractions, we observed that frac-
tions F2, F4, F5, F7, and F8 had significant effects (Suppl. Fig. 2). On
performing 1H and 13C-NMR analysis, fractions F2 and F4 were found to
of α-KA tended to be higher when GA was used rather than when GFE
was employed for emulsification. Cbrain/Cplasma values were calculated
by dividing Cbrain by the corresponding Cplasma. The values of KG8 and
KGD on utilizing GFE were similar to those obtained with GA, whereas
the values of
α-KA with GA were higher than when GFE was used.
contain
α-KA and CI, respectively, and F8 contained KG8 and KG.
4. Discussion
Fractions F5 and F7 were composed of KGD and other unknown com-
pounds, respectively.
In the present study, we first identified the active ingredients of VF
demonstrating sedative effects using a binding assay on flunitrazepam
sites of GABAA receptor, which focuses on the pharmacological target
for insomnia treatment (Le et al., 2013). VO, which belongs to the same
genus as VF, is typically employed as a botanical sedative drug in Eu-
ropean countries (Weiss et al., 2002; European Directorate for the
Quality of Medicines & Health Care, 2020; U.S. Pharmacopeial United
States Pharmacopeial Convention, 2020) and had the binding activity on
GABAA receptor (Yuan et al., 2004; Benke et al., 2009).
We evaluated the binding activity of each constituent in the VF
extract. Among the six compounds, CI exhibited the highest binding
activity. KGD, KG8, and
α-KA exhibited binding activities in a
concentration-dependent manner, and the activity was nearly 30% at
concentrations ranging between 3 and10 mM. KG and KG2 showed
binding activities of approximately 15% at a concentration of 10 mM,
and the effects of KG and KG2 were lower than those of KGD, KG8, and
α
-KA (Fig. 6).
6

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
Fig. 4. Comparison of the HPLC chromatographs of Valeriana fauriei root (Vf2) and V. officinalis root (Vo4). (a) Maximum chromatograph plot of 210–340 nm
analyzed by condition 1. (b) Enlarged view of the dotted line in Fig. 4a. (c) Chromatograph detected at 198 nm analyzed by condition 2. HPLC conditions are
described in the Materials and Methods section. Peak numbers are listed in Table 2.
In a general receptor binding assay system, the radiolabeled ligand
and receptor are separated using glass microfiber filters. As VF contains
a large amount of essential oil components, nonspecific binding on the
glass fiber was observed, and the analysis could not be performed. Thus,
we prepared a new assay system to evaluate the GABAA receptor binding
activity using [3H]-flunitrazepam and a rat cerebral cortex homogenate
in a 96-well plate. In this system, the representative benzodiazepine,
diazepam, exhibited an EC50 value of approximately 10–30 nM. In a
previous study using a general receptor binding assay (Watanabe et al.,
1986), the EC50 values of diazepam in competition with [3H]-fluni-
trazepam binding, using rat brain cerebellum and spinal cord, were
approximately 10 nM, which is similar to that observed in the present
study. This indicates that the tested system can be applied not only for
VF but also for various plant materials rich in essential oil components.
By employing this system, we compared the binding activities on
flunitrazepam sites of GABAA receptor between VF and VO extracts; the
latter has been previously reported (Yuan et al., 2004; Benke et al.,
2009). We observed that the binding activities of VF extracts were
significantly higher than those of VO extracts. Reportedly, mice treated
with the VO extract show a longer hexobarbital-induced sleeping time
than mice treated with the VF extract (Takamura et al., 1973). However,
it remains debatable whether these two plant species could be accurately
compared using only one individual of each species. The present study is
the first to compare the activities of VO and VF using several individual
constituents, and the VF extract exhibited significantly higher binding
activity on flunitrazepam sites of GABAA receptor than the VO extract. It
was reported that the molecular target of valerenic acid, the main
compound of VO, on GABAA receptor was β3-subunit among its fluni-
trazepam sites (Benke et al., 2009). However, we did not elucidate the
interactions between the major substrates for VF and any allosteric sites
of GABAA receptor. This is the limit of the present study, and the further
investigation is needed.
In Japan, VF was used as a folk medicine in the middle of the Edo era
(1603–1868). It was then used for hysteria like VO in European coun-
tries after the late Edo era when European culture had been imported
from the Netherlands to Japan. As hysteria was associated with female
medical disorders during the Meiji era (1868–1912), VF began to be
used as a women’s home medicine in Japan (Ota et al., in press).
Furthermore, as VF contains a higher amount of essential oil than VO
(Houghton, 1988), VF has been exported from Japan to Germany at a
high price between the Taisho (1912–1925) and early Showa era
(1926–1988) (Yanagisawa et al., 1993; Yanagisawa, 2014). The present
results suggest that VF is superior to VO not only in essential oil content
and but also sedative and anxiolytic activities.
Furthermore, we observed that VF constituents significantly differed
from those in VO, and PCA analysis revealed that they were located in
different clusters. Among these compounds, the KGD content highly
correlated with the binding activity of the VF extract on flunitrazepam
sites of GABAA receptor. Meanwhile, valerenic acid (peak No. 32) and
the other five compounds (peak nos. 8, 18, 27, 34, and 41) correlated
with the VO extract binding activity. Valerenic acid was also slightly
associated with the VF extract activity; however, the other compounds
contained in both VF and VO were not correlated with the activity.
These results suggest that the KGD content may primarily contribute to
7

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
Table 2
HPLC peak areas of Valeriana fauriei root (VF) and V. officinalis root (VO).
Peak
No.
tR
Peak areas of VF
converted to extract
concentration
Peak areas of VO
converted to extract
concentration
P
(min)
value
1
12.9
13.2
15.5
18.2
20.4
22.2
26.4
31.3
34.2
35.0
38.4
39.3
40.9
43.6
47.0
47.4
48.8
50.0
50.7
51.1
51.4
51.5
51.6
51.9
52.0
52.8
53.1
53.3
53.6
53.8
54.0
54.4
54.7
55.1
55.2
55.8
56.0
56.4
56.8
57.7
58.3
58.6
59.1
9.1
(5.50 ± 2.31) × 103
(9.70 ± 9.70) × 103
(2.13 ± 0.91) × 105
(2.06 ± 1.55) × 104
n.d.
2
(8.93 ± 1.75) × 104
3
(5.87 ± 5.87) × 103
4
(1.32 ± 1.32) × 102
5
(3.43 ± 3.43) × 103
n.d.
6
(2.09 ± 0.60) × 104
(1.94 ± 1.18) × 104
n.d.
7
(2.81 ± 0.31) × 104
***
***
8
n.d.
(6.04 ± 1.16) × 104
n.d.
9
(1.51 ± 1.51) × 103
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
(3.46 ± 0.73) × 104
(2.71 ± 1.82) × 105
(6.95 ± 5.94) × 104
(3.74 ± 0.80) × 104
(7.67 ± 5.75) × 104
(1.31 ± 0.84) × 104
(8.02 ± 4.64) × 103
(2.82 ± 2.82) × 103
(8.87 ± 5.48) × 103
(1.72 ± 0.21) × 104
(2.10 ± 1.26) × 104
(3.44 ± 2.04) × 103
(6.73 ± 3.03) × 104
(2.57 ± 1.89) × 104
(6.99 ± 4.36) × 103
(5.46 ± 3.19) × 103
(6.98 ± 2.62) × 104
(1.47 ± 0.76) × 104
(1.60 ± 0.89) × 106
(2.12 ± 1.07) × 105
(2.06 ± 0.51) × 105
(3.80 ± 1.33) × 104
(4.48 ± 3.00) × 103
(9.86 ± 2.10) × 104
(3.32 ± 1.21) × 104
(9.56 ± 2.80) × 104
(6.14 ± 4.51) × 103
(8.13 ± 5.85) × 103
(2.17 ± 2.17) × 103
(5.58 ± 2.94) × 104
(1.81 ± 1.81) × 103
(2.62 ± 0.68) × 104
(8.57 ± 2.50) × 104
(1.13 ± 0.68) × 104
(3.32 ± 1.02) × 104
n.d.
n.d.
(3.60 ± 3.60) × 103
**
n.d.
(4.60 ± 2.82) × 103
n.d.
Fig. 5. Principal component analysis (PCA) score plots for 5 batches of
Valeriana fauriei root (VF) and 4 batches of V. officinalis root (VO) and loading
plots of peaks. The PCA was performed for the peak area of 44 peaks described
in Table 2. The top and right axes indicate the score plots of Valeriana samples,
and the bottom and left axes indicate the loading plots of peaks.
n.d.
n.d.
n.d.
***
(2.32 ± 0.48) × 104
n.d.
n.d.
*
n.d.
(8.47 ± 1.37) × 104
**
*
Table 3
n.d.
Correlation between the GABAA receptor binding activity of Valeriana fauriei
root (VF) or V. officinalis root (VO) and its peak area converted to extract
concentration.
n.d.
n.d.
n.d.
(2.53 ± 2.53) × 103
Characteristic species
Correlation between the activity of
either VF or VO and its peak area
converted to extract concentration
(r)
(1.02 ± 1.02) × 103
**
*
(7.31 ± 2.56) × 103
(2.36 ± 1.45) × 103
(9.45 ± 0.57) × 103
**
*
Peak No.
t
R (min)
VF
VO
n.d.
n.d.
**
*
VF, VO
VF
2
13.2
35.0
26.4
51.6
55.2
55.8
9.1
– 0.21
0.90
(9.20 ± 2.24) × 104
10
7
– 0.20
0.81
(5.70 ± 1.31) × 104
*
– 0.56
unreckonable
0.54
n.d.
23
35
36
44
8
– 0.45
n.d.
0.06
0.52
n.d.
0.16
0.50
(2.02 ± 1.24) × 103
**
**
0.91
unreckonable
0.81
n.d.
VO
31.3
39.3
50.0
51.4
52.0
52.8
53.1
53.3
53.6
53.8
54.4
54.7
55.1
57.7
58.3
59.1
unreckonable
0.18
n.d.
12
18
21
25
26
27
28
29
30
32
33
34
40
41
43
0.93
n.d.
**
unreckonable
unreckonable
unreckonable
unreckonable
unreckonable
0.18
0.85
(4.66 ± 0.57) × 105
***
0.48
0.50
Data are expressed as the mean ± S.E. (n = 5 for VF and n = 4 for VO). The
conversion of the peak area was calculated by dividing by the extraction effi-
ciency with a maximum value of 1.
0.17
0.82
0.90
*p < 0.05, **p < 0.01, and ***p < 0.001 between VF and VO by Student’s t-test.
n.d., not detected. tR, retention time.
0.21
0.30
0.40
– 0.15
0.70
0.51
unreckonable
unreckonable
– 0.83
0.42
the binding activity of VF on flunitrazepam sites of GABAA receptors. In
contrast, valerenic acid might not be the only contributor to the activity
of VO. In present study, although we used the VF and VO samples
derived from different origins, we could only compare the total com-
ponents of those two species because the number of samples was not
enough to mention the difference of active ingredients due to the origin.
Further investigation revealed that ethanol and 70% ethanol extracts
of VF exhibited similar binding activities on flunitrazepam sites of
GABAA receptor converted as the dosage related to the original crude
drug weight, suggesting that the active ingredients in VF would barely
0.73
0.91
unreckonable
unreckonable
0.78
0.53
The peaks were selected significantly higher peaks of VF or VO, as described in
Table 2 and characteristic peaks for the principal component analysis (PCA)
shown in Fig. 5, are shown in Fig. 4. Peak nos. 32 and 44 were identified as
valerenic acid and kessyl glycol diacetate, respectively, by comparison with
standard compounds. The conversion of the peak area was calculated by
dividing by the extraction efficiency with a maximum value of 1. When the peaks
were not detected, r values were shown as “unreckonable”. tR, retention time.
dissolve in water. We isolated and identified KGD, KG8, KG2, α-KA, KG,
and CI as active ingredients of VF. Among the activity-concentration
curves of these compounds, CI had the highest activity, followed by
activity at the concentration of EC50 values in VF extracts, it can be
suggested that these two compounds did not contribute to the binding
activity. As for the 70% ethanol extract of VF, CI did not contribute to
the activity as it was absent in this extract. The sum of the activity values
of other compounds was 52%, i.e., nearly 50%; hence, it can be sug-
gested that KGD contributed to the majority of binding activity on
KGD, α-KA, KG8, KG2, and KG. Among these compounds, KGD had the
highest concentration, followed by KG,
α
-KA, KG8, and KG2 in the 70%
ethanol extract, and by CI,
extract.
α-KA, KG, KG8, and KG2 in the ethanol
In both VF extracts, as KG and KG2 did not exhibit any binding
8

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
activity. Therefore, the three compounds, KGD, KG8, and α-KA, would
contribute to the binding activity of the ethanol extract of VF, as well as
70% ethanol extract, on flunitrazepam sites of GABAA receptor.
We elucidated that the three compounds, KGD, KG8, and α-KA, were
distributed in the mouse brain following oral administration of the VF
extract. Although KGD was the main component of the VF extract, only
0.144 ± 0.107% of KGD was distributed in the brain when GA, a water-
soluble emulsifier, was used; however, brain distribution was approxi-
mately doubled with GFE, a fat-soluble emulsifier. Moreover, the brain
distributions of KG8 and KG2 in mice treated with the GFE-emulsified
formulation were more than 10-times higher than that of mice treated
with the GA-emulsified preparation. Based on these findings, it can be
postulated that incorporating a fat-soluble emulsifier could promote the
absorption of KGD and that KGD might undergo de-esterification to KG8
and KG2 in the body. Although the brain distributions of KG8 and KGD
were higher in mice treated with the GFE-emulsified preparation, the
C
brain/Cplasma values of these two compounds did not differ between the
two formulations. As KG8 and KGD distribution depends on their plasma
concentrations, the transport of these two compounds from the blood to
Fig. 6. Binding capacity of constituents in Valeriana fauriei root extract to
GABAA receptor in rat brain homogenate. Binding capacity was calculated as
the ratio to the value of excessive clonazepam binding. Data were mean ± S.E.
(n = 4). **p < 0.01, ***p < 0.001 vs control group (0 nM) by Dunnett’s
multiple t-test. KGD, kessyl 2 glycol diacetate; KG2, kessyl glycol 2-acetate,
the brain might be saturated. For α-KA, the Cbrain/Cplasma value was
higher in mice treated with the GA-emulsified preparation than the GFE-
added preparation; this indicated that the transfer of
blood to the brain might be affected by drug additives.
α-KA from the
KG8, kessyl glycol 8-acetate;
coniferyl isovalerate.
α-KA, α-kessyl acetate; KG, kessyl glycol; CI,
Takamura et al. (1975b) and Hikino et al. (1980) have reported that
the enhancing effect of KG8 emulsified with GA on hexobarbital-induced
sleeping time was greater than that of KGD. In the present study, the
distribution of KG8 in the plasma and brain of mice treated with the VF
extract emulsified using GFE was significantly higher than that emulsi-
fied using GA, and the distribution of KGD was much less than that of
KG8. On employing GFE as the emulsifier for KG8, the sleep time
induced by KG8 might be considerably longer than that induced by KGD.
Table 4
GABAA receptor binding activity of each compound when activities of ethanol
and 70% ethanol extract of Valeriana fauriei root show 50% effective concen-
tration (EC50) values.
Ethanol extracta)
70% ethanol extractb)
Furthermore, both KGD-rich and
α-KA-rich types of VF prolonged
Compound
Concentration
(mM)c)
Activity
(%)d)
Concentration
(mM) c)
Activity
(%)d)
pentobarbital-induced sleeping time in mice (Yoshitomi et al., 2000).
Reportedly, KGD prolongs hexobarbital-induced sleep time in mice
(Takamura et al., 1973, 1975a, 1975b). However, there are no reports
KG
0.18
0.01
0.08
4.44
0.26
0.37
0
0.77
0.01
0.19
3.93
0.23
n.d.
0
KG2
KG8
KGD
0
0
available regarding the effects of
both KGD and
α-KA. Our present study revealed that
5
5
38
10
14
38
9
α
-KA demonstrated the binding activities on fluni-
α-KA
trazepam sites of GABAA receptor in vitro. Following oral administration,
these compounds were distributed in the brains of mice. Although pre-
vious in vivo studies used excessively high dosages (approximately
20–40 g of VF/kg), our present results suggest a possible underlying
mechanism and active ingredients of VF as folk medicine.
CI
0
EC50 values were calculated from the least square regression line plotted from 3
points that crossed 50% of the control logarithmic concentration values (Fig. 2).
a) EC50 value of ethanol extract is 161 mg (original crude drug)/ml which is
related to 5.38 mg (extract)/ml.
b) EC50 value of 70% ethanol extract is 143 mg (original crude drug)/ml which is
related to 25.9 mg (extract)/ml.
5. Conclusion
c) Estimated concentration of each compound in the sample at a concentration
equal to the EC50 value.
The extract of Valeriana fauriei root (VF) exhibited the binding ac-
tivity on flunitrazepam sites of GABAA receptor, which was significantly
higher than that of Valeriana officinalis root (VO). Kessyl glycol diacetate
d) GABAA receptor binding activity of each compound at that concentration was
calculated from the activity-concentration curve (Fig. 6).
(KGD), kessyl glycol 8-acetate (KG8), and α-kessyl acetate (α-KA) are
KG, kessyl glycol; KG2, kessyl glycol 2-acetate; KG8, kessyl glycol 8-acetate;
KGD, kessyl glycol diacetate; α-KA, α-kessyl acetate; CI, coniferyl isovalerate.
active ingredients present in VF that exhibit binding activity on fluni-
trazepam sites of GABAA receptor, and these three compounds can be
distributed to the brain. For KGD, which is the main active ingredient,
the addition of a fat-soluble emulsifier promoted absorption, and it was
metabolized to KG8 in the body. Thus, these compounds can be used as
marker compounds to evaluate the quality of VF products in terms of
their sedative effects.
flunitrazepam sites of GABAA receptor, whereas KG8 and
α-KA demon-
strated minor contributions, in the 70% ethanol extract of VF. Moreover,
our findings indicate that KG8 and α-KA would interact additively and
that other compounds in the extract might not contribute to the total
effects.
For the ethanol extract of VF, the sum of these values was 67%;
hence, it can be suggested that KGD contributed to the majority of the
binding activity on flunitrazepam sites of GABAA receptor, whereas
A list of author’s contributions
M.O., H.N., Y.M., D.K., S.M., and S.N. conducted the experiments. M.
O., T.M, and Y.O. taught and advised the experiments to H.N., Y.M., D.
K., S.M., and S.N. Y.O. performed statistical analysis. K.I. performed
NMR analysis. M.O. wrote the draft article, T.M. and M.O. are presiding
over the study.
KG8, α-KA, and CI presented a minor contribution, and these compounds
or other unknown compounds in the ethanol extract interact counter-
actively among the compounds. CI exhibited the highest in vitro binding
activity on flunitrazepam sites of GABAA receptor. However, CI was
scarcely distributed in the plasma and brain owing to the high binding
rate to albumin, revealing that CI did not contribute to the actual
9

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
Table 5
The brain (Cbrain) and plasma (Cplasma) concentrations of the compounds 30 min after the oral administration of two VF extract preparations, emulsified with glycerin
fatty acid ester (GFE) and dissolved with oil, or emulsified with gum arabic (GA) and dissolved with water, at the crude drug dosage of 10 g/kg to mice.
Compound
c)
The kinds of emulsifier
Dose in VF extract (mg/kg)a)
0.179
Cbrain (ng/g)
Cplasma (ng/ml)
Distribution in brain (%)b)
Cbrain/C
plasma
KG2
GFE
GA
137 ± 26 **
12.2 ± 7.5
n.d.
76.5 ± 14.3 **
6.80 ± 4.21
unreckonable
1.78 ± 0.00
1.26 ± 0.11
1.17 ± 0.22
1.31 ± 0.34
1.43 ± 0.56
0.822 ± 0.619
4.70 ± 2.58
unreckonable
unreckonable
2.19 ± 2.19
378 ± 74.0 **
33.1 ± 5.8
219 ± 81.6
60.3 ± 17
114 ± 47
188 ± 46
n.d.
KG8
KGD
GFE
GA
1.55
93.4
4.51
6.09
506 ± 130 *
36.0 ± 8.2
32.6 ± 8.41 *
2.32 ± 0.53
GFE
GA
206 ± 48
0.221 ± 0.051
0.144 ± 0.107
1.15 ± 0.53
135 ± 100
α-KA
GFE
GA
51.7 ± 23.7
290 ± 161
6.43 ± 3.57
CI
GFE
GA
0.409 ± 0.379
0.973 ± 0.522
0.00672 ± 0.00622
0.0160 ± 0.0086
n.d.
Data were expressed as mean ± S.E. (n = 6–7). *p < 0.05 and **p < 0.01 vs treated with GA-added preparation group by Student’s t-test.
a) The dose in the VF extract was calculated based on the individual compound content in the VF extract.
b) Distribution in the brain (%) was calculated by dividing the measured Cbrain of the compound by its dose in the VF extract.
c) Cbrain/Cplasma was calculated by dividing the measured Cbrain of the compound by its corresponding Cplasma
When the compounds were not detected, Cbrain/Cplasma values shown as “unreckonable”.
.
KG, kessyl glycol; KG2, kessyl glycol 2-acetate; KG8, kessyl glycol 8-acetate; KGD, kessyl glycol diacetate;
α
-KA,
α
-kessyl acetate; CI, coniferyl isovalerate; trace, less
than 1 μg/g for KG in Cbrain and 1.5 μg/ml for KG in Cplasma; n.d., not detected.
Navarrete, A., Avula, B., Choi, Y.W., Khan, I.A., 2006. Chemical fingerprinting of
valeriana species: simultaneous determination of valerenic acids, flavonoids, and
phenylpropanoids using liquid chromatography with ultraviolet detection. J. AOAC
Int. 89, 8–15.
Declaration of competing interest
T.M. received grant support from Tsumura & Co., Kracie Pharma-
ceuticals, and JPS Pharmaceuticals. M.O., Y.M., D.K., S.M., S.N., and Y.
O. are employee of Kuki Sangyo.
Nishiya, K., Kimura, T., Takeya, K., Itokawa, H., 1992. Sesquiterpenoids and iridoid
glycosides from Valeriana fauriei. Phytochemistry 10, 3511–3514.
Oshima, Y., Matsuoka, S., Ohizumi, Y., 1995. Antidepressant principles of Valeriana
fauriei roots. Chem. Pharm. Bull. 43, 169–170.
Appendix A. Supplementary data
Ota, M., Utaka, I., Makino, T. Historical study on the usage and the name of crude drug
"Valerian" in Japan. Kampo Med. (in press).
Pharmaceutical and Medical Device Regulatory Science Society of Japan, 2016. The
Japanese Pharmacopoeia, seventeenth ed. Jiho, Tokyo, Japan.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jep.2021.114262.
¨
Risa, J., Risa, A., Adsersen, A., Gauguin, B., Stafford, G.I., van Staden, J., Jager, A.K.,
2004. Screening of plants used in southern Africa for epilepsy and convulsions in the
GABAA-benzodiazepine receptor assay. J. Ethnopharmacol. 93, 177–182.
Safaralie, A., Fatemi, S., Sefidkon, F., 2008. Essential oil composition of Valeriana
officinalis L. roots cultivated in Iran. Comparative analysis between supercritical CO2
extraction and hydrodistillation. J. Chromatogr. A 1180, 159–164.
References
Benke, D., Barberis, A., Kopp, S., Altmann, K.H., Schubiger, M., Vogt, K.E., Rudolph, U.,
¨
Mohler, H., 2009. GABAA receptors as in vivo substrate for the anxiolytic action of
Sakamoto, T., Mitani, Y., Nakajima, K., 1992. Psychotropic effects of Japanese valerian
root extract. Chem. Pharm. Bull. 40, 758–761.
valerenic acid, a major constituent of valerian root extracts. Neuropharmacology 56,
174–181.
Sieghart, W., Savic, M.M., 2018. International union of basic and clinical pharmacology.
CVI: GABAA receptor subtype-and function-selective ligands: key issues in
translation to humans. Pharmacol. Rev. 70, 836–878.
Bohlmann, F., Zdero, C., King, R.M., Robinson, H., 1980. Sesquiterpene lactones from
Eremanthus species. Phytochemistry 19, 2663–2668.
Choi, H.S., Ko, B.S., Kim, H.D., Hong, K.B., Suh, H.J., 2017. Effect of valerian/hop
mixture on sleep-related behaviors in Drosophila melanogaster. Biol. Pharm. Bull. 40,
1101–1110.
Suzuki, H., Zhang, B.C., Harada, M., Iida, O., Satake, M., 1993. Quantitative studies on
terpenes of Japanese and European valerians. Jap. J. Pharmacog. 47, 305–310.
Takamura, K., Kakimoto, M., Kawaguchi, M., Iwasaki, T., 1973. Pharmacological studies
on the constituents of crude drugs and plants. I. Pharmacological actions of Valeriana
officinalis LINNE var. latifolia MIQUEL. Yakugaku Zasshi 93, 599–606.
Takamura, K., Kawaguchi, M., Nabata, H., 1975a. The preparation and pharmacological
screening of kessoglycol derivative. Yakugaku Zasshi 95, 1198–1204.
Takamura, K., Nabata, H., Kawaguchi, M., 1975b. The pharmacological action on the
kessoglycol 8-monoacetate. Yakugaku Zasshi 95, 1205–1209.
Choi, J.H., Lee, M.J., Chang, Y., Lee, S., Kim, H.J., Lee, S.W., Kim, Y.O., Cho, I.H., 2020.
Valeriana fauriei exerts antidepressant-like effects through anti-inflammatory and
antioxidant activities by inhibiting brain-derived neurotrophic factor associated with
chronic restraint stress. Rejuvenation Res. 23, 245–255.
European Directorate for the Quality of Medicines & Health Care, 2020. European
Pharmacopoeia, tenth ed. Council of Europe, Strasbourg, France, p. 1660.
Fong, S.Y.K., Li, C., Ho, Y.C., Li, R., Wang, Q., Wong, Y.C., Xue, H., Zuo, Z., 2017. Brain
uptake of bioactive flavones in Scutellariae Radix and its relationship to anxiolytic
effect in mice. Mol. Pharm. 14, 2908–2916.
Tokunaga, S., Takeda, Y., Niimoto, T., Nishida, N., Kubo, T., Ohno, T., Matsuura, Y.,
Kawahara, Y., Shinomiya, K., Kamei, C., 2007. Effect of valerian extract preparation
(BIM) on the sleep-wake cycle in rats. Biol. Pharm. Bull. 30, 363–366.
United States Pharmacopeial Convention, 2020. The United States Pharmacopoeia 43rd
Edition – the National Formulary, 38th Edition. United States Pharmacopeial
Convention, North Bethesda, Maryland, United States.
Han, S.H., Kim, H.D., Ko, K., Shin, J.C., Suh, H.J., 2018. Preparation of valerian
(Valeriana officinalis) extract powder and its sleep potentiating activity. J. Korean
Soc. Food Sci. Nutr. 47, 131–135.
Hashiguchi, W., Nagatomo, I., Akasaki, Y., Uchida, M., Tominaga, M., Takigawa, M.,
2001. Influences of caffeine to nitric oxide production and zonisamide concentration
in the brain of seizure-susceptible EL mice. Psychiatr. Clin. Neurosci. 55, 319–324.
Hendriks, H., Bos, R., Allersma, D.P., Malingre, T.M., Koster, A.S., 1981. Pharmacological
screening of valerenal and some other components of essential oil of Valeriana
officinalis. Planta Med. 42, 62–68.
Wang, B., Zheng, T.K., Lu, C.L., Xu, X.J., Yin, H.B., 1998. Identification of three kind of
Valeriana species by anatomical characteristics. J. Chin. Med. Mater. 21, 177–179.
Watanabe, Y., Shibuya, T., Khatami, S., Salafsky, B., 1986. Comparison of typical and
atypical benzodiazepines on the central and peripheral benzodiazepine receptors.
Jpn. J. Pharmacol. 42, 189–197.
Weiss, R.F., Fintelmann, V., 2002. Lehrbuch der Phytotherapie. Hippokrates-Verlag,
Stuttgart, Germany.
Hikino, H., Hikino, Y., Kobinata, H., Aizawa, A., Konno, C., Oh’izumi, Y., 1980. Sedative
principles of valeriana roots. Jap. J. Pharmacog. 34, 19–24.
Yanagisawa, K., Yamada, M., Matsumoto, Y., 1993. The transition of psychotropic drugs
in Japanese Pharmacopoeia (JP) (Part 6) the knowledge and judgment about the sort
as the medicine on Valerianae Radix. J. Jpn. Hist. Pharm. 28, 12–19.
Yanagisawa, K., 2013. Transitions of psychotropic drugs in the Japanese pharmacopoeia
(JP) (Part 13) Transitions of the standards and test methods of Valerianae Radix in
the JP 1 (1886) and JP X VI (2011) and comparison between the USP and BP.
Yakushigaku Zasshi 48, 63–74.
Houghton, P.J., 1988. The biological activity of Valerian and related plants.
J. Ethnopharmacol. 22, 121–142.
Kim, J.S., Ahn, J.D., Cho, S.I., 2015. Effects of Valerianae Radix et Rhizoma extract on
psychological stress in mice. Phcog. Mag. 11, 381–388.
Komori, T., Matsumoto, T., Motomura, E., Shiroyama, T., 2006. The sleep-enhancing
effect of valerian inhalation and sleep-shortening effect of lemon inhalation. Chem.
Senses 31, 731–737.
Yanagisawa, K., 2014. Transitions of psychotropic drugs in the Japanese pharmacopoeia
(JP) (Part 14) Transitions in the standards and test methods of Valerian root in the
Deutsches Arzneibuch (DAB) and comparison between the USP, BP, EP and JP, and
Le, M.C., Monyer, H., 2013. GABAergic interneurons shape the functional maturation of
the cortex. Neuron 77, 388–405.
´
Lucio-Gutierrez, J.R., Coello, J., Maspoch, S., 2012. Enhanced chromatographic
fingerprinting of herb materials by multi-wavelength selection and chemometrics.
Anal. Chim. Acta 710, 40–49.
10

M. Ota et al.
Journal of Ethnopharmacology 278 (2021) 114262
Oleum Valerianae (Japan kesso oil) listed in DAB6 (1926). Yakushigaku Zasshi 49,
Yuan, C.S., Mehendale, S., Xiao, Y., Aung, H.H., Xie, J.T., Ang-Lee, M., 2004. The
γ-aminobutyric acidergic effects of Valerian and valerenic acid on rat brainstem
neuronal activity. Anesth Analg 98, 353–358.
176–189.
Yoshitomi, S., Watanabe, M., Kawanishi, F., Satake, M., 2000. Quality and their sedative
effects of Japanese valerian roots. Jap. J. Pharmacog. 54, 55–60.
11