Dual action of acertannins as potential regulators of intracellular ceramide levels

Article abstract of DOI:10.1016/j.tetasy.2016.09.006

Derived from the genus maple (Acer), acertannins are a group of gallotannins which have a characteristic 1,5-anhydro-D-glucitol (1,5-AG) occupying the central core position in the tannic acid structure whose hydroxyl groups have one or more galloyl residues. We have synthesized all ten naturally-occurring acertannins and seven new acertannin derivatives from 1,5-AG. Side-by-side comparisons revealed that 2,4,6-tri-O-galloyl-1,5-AG 21 (maplexin E) and maplexin F 22 (2,3,6-tri-O-galloyl-1,5-AG) were good inhibitors of ceramidase (CDase). In contrast, the core anhydrosugar 1,5-AG 12 itself and 3′,4′,5′-trimethoxy benzoyl derivatives 23–25 did not show CDase inhibition. Metabolic labelling experiments using NBD-hexanoic acid revealed that 50?μM of 6-O-galloyl-1,5-AG 16 (Ginnalin B), 2,6-di-O-galloyl-1,5-AG 19 (Ginnalin A), and 4,6-di-O-galloyl-1,5-AG 4 increased intracellular NBD-labeled ceramide, by 2.3, 2.2, and 2.1-fold, respectively. It is noteworthy that these acertannins 16, 19, and 4 promoted ceramide synthase 3 (CERS3) gene expression. Acertannins, therefore, represent a new class of potential intracellular ceramide regulators exhibiting both CDase inhibition and ceramide synthase promotion.

Full text of DOI:10.1016/j.tetasy.2016.09.006

Tetrahedron: Asymmetry 27 (2016) 1177–1185
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Dual action of acertannins as potential regulators of intracellular
ceramide levels
a
a
b
c
Akiko Kamori a, Atsushi Kato a, , Shota Miyawaki , Junna Koyama , Robert J. Nash , George W. J. Fleet ,
⇑
Daisuke Miura d, Fumihiro Ishikawa d, Isao Adachi a
a Department of Hospital Pharmacy, University of Toyama, Toyama 930-0194, Japan
b Institute of Biological, Environmental and Rural Sciences/Phytoquest Limited, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom
c Chemistry Research Laboratory, Department of Chemistry, University of Oxford, Oxford OX1 3TA, United Kingdom
d FUSHIMI Pharmaceutical Co. Ltd, 1676 Nakazu-cho, Marugame, Kagawa 763-8605, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Derived from the genus maple (Acer), acertannins are a group of gallotannins which have a characteristic
1,5-anhydro-D-glucitol (1,5-AG) occupying the central core position in the tannic acid structure whose
Received 15 September 2016
Accepted 27 September 2016
Available online 12 October 2016
hydroxyl groups have one or more galloyl residues. We have synthesized all ten naturally-occurring acer-
tannins and seven new acertannin derivatives from 1,5-AG. Side-by-side comparisons revealed that 2,4,6-
tri-O-galloyl-1,5-AG 21 (maplexin E) and maplexin F 22 (2,3,6-tri-O-galloyl-1,5-AG) were good inhibitors
of ceramidase (CDase). In contrast, the core anhydrosugar 1,5-AG 12 itself and 30,40,50-trimethoxy benzoyl
derivatives 23–25 did not show CDase inhibition. Metabolic labelling experiments using NBD-hexanoic
acid revealed that 50 lM of 6-O-galloyl-1,5-AG 16 (Ginnalin B), 2,6-di-O-galloyl-1,5-AG 19 (Ginnalin
A), and 4,6-di-O-galloyl-1,5-AG 4 increased intracellular NBD-labeled ceramide, by 2.3, 2.2, and 2.1-fold,
respectively. It is noteworthy that these acertannins 16, 19, and 4 promoted ceramide synthase 3 (CERS3)
gene expression. Acertannins, therefore, represent a new class of potential intracellular ceramide regula-
tors exhibiting both CDase inhibition and ceramide synthase promotion.
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
and salvage pathway.4 Ceramide synthases are responsible for the
formation of both dihydroceramide and ceramide through
The stratum corneum is the outermost layer of the skin (thick-
ness 100–300 m) and serves as a skin barrier and prevents
N-acetylation of dihydrosphingosine and sphingosine.5 The activity
of ceramide synthase significantly increases during epidermal
differentiation and is clearly higher in keratinocytes and murine
epidermis than in most other tissue.6
l
transepidermal water loss.1 Intercellular lipids in stratum corneum
principally consist of fatty acids, cholesterol, and ceramides; these
lipids form characteristic multi-lamellar sheets in intercellular
spaces.2 Previous studies suggested that the ceramide comprised
about 50% of the major lipid species of the stratum corneum. This
content by percentage is significantly higher than fatty acid
(10–20%), and cholesterol (25%).3 It is now well recognized that
intercellular ceramide level is affected by age, gender, skin type,
pigmentation, and life style such as smoking habits etc.
On the other hand, in ceramide degradation, ceramidase
(CDase; N-acylsphingosine amidohydrolase; EC 3.5.1.23) is a key
enzyme in the regulation of ceramide metabolism and catalyzes
the degradative N-deacylation of ceramide to produce fatty acid
and sphingosine. Thereafter, the sphingosine can be converted into
sphingosine-1-phosphate through phosphorylation by a sphin-
gosine kinase. Sphingosine-1-phosphate acts as a motile second
messenger; it regulates various processes such as differentiation,
proliferation, and programmed cell death, either directly or by
serving as a source of signaling metabolites.7 Many previous
reports suggested that there is a significant decrease in ceramide
levels in the stratum corneum of atopic dermatitis patients com-
pared to age-matched healthy controls.8 It is thought that one of
the reasons is that the endogenous activities of CDases are gradu-
ally increased in proportion by aging, and so the component
amount of skin ceramide tends to decrease. Furthermore, the most
With regard to ceramide biosynthesis, previous studies sug-
gested the existence of a plurality of pathways. Among them, a
de novo biosynthetic pathway is mediated by serine palmitoyl-
CoA transferase (EC 2.3.1.50) and subsequent ceramide synthase
(EC 2.3.1.24) and has essential roles in the control of ceramide
levels; ceramide is also supplied by the sphingomyelinase pathway
⇑
Corresponding author.
E-mail address: kato@med.u-toyama.ac.jp (A. Kato).
http://dx.doi.org/10.1016/j.tetasy.2016.09.006
0957-4166/Ó 2016 Elsevier Ltd. All rights reserved.

1178
A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
commonly expressed reason for this belief was the exogenous
ceramidase from colonizing bacteria such as Pseudomonas aerugi-
nosa and Staphylococcus aureus mediate the decrease of ceramide.
Thus, CDase inhibitors and/or ceramide synthase activators might
have enormous therapeutic potential in cosmetics and skin disease.
At present, some compounds mimicking the structures of cera-
mide or designed ‘drug-like’ compounds have been reported as
CDase inhibitors (Fig. 1). N-Oleoyl-ethanol amine, a typical cera-
mide-mimicking inhibitor, showed inhibition in keratinocytes.9
Some N-oleoyl-ethanol amine analogues were also synthesized
but it was not an effective strategy for improving CDase inhibi-
screening program, we found that extracts from the sugar maple
(Acer saccharum) and red maple (Acer rubrum L.) and other Acer
species had CDase inhibitory effects. The deciduous tree genus Acer
is widespread in eastern and central North America; its tree sap is
used for producing maple syrup. Recently, new gallotannin deriva-
tives ‘Acertannins’ have been identified from these plants and have
been given the trivial names maplexins and ginnalins.16 These
acertannins varied in the number and location of galloyl groups
linked to different positions of a 1,5-anhydro-D-glucitol (1,5-AG)
core (Fig. 2). They are considered to play a role in the protection
from predation, and perhaps also serve as pesticides, and in plant
growth regulation. Beginning with the discovery of acertannins,
in addition to consumption as premium natural sweeteners, the
extracts of the Acer genus plants have attracted significant research
attention as possible treatments of various disorders including
hyperglycemia and cancer.17,18 However, there is limited knowl-
edge of their effects on skin components and their metabolic turn-
over. Herein we report the synthesis and biological evaluation of a
series of acertannins, which includes all 10 naturally-occurring
acertannins (maplexins A–F, ginnalin A–C, 3,6-di-O-galloyl-1,5-
tion.10
D-Erythro-2-(N-Myristoylamino)-1-phenyl-1-propanol (D-e-
MAPP), a lipophilic aromatic ceramide analogue, is a much more
potent inhibitor than OAE.11 This interesting result led to the dis-
covery of the more water soluble N-acyl-phenylaminopropanol-
type ceramidase inhibitors, such as B13 and LCL-464 (lysosome-
and mitochondrium-directed analogues; LCL-analogues).12 How-
ever, these typical ceramide analogues with long-chain alkyl moi-
eties tend to have poor pharmacological properties due to binding
with albumin, resulting in poor bioavailability. Thus, there is a
need to identify a new class of small molecule ceramidase inhibi-
tors with more favorable ‘drug-like’ pharmacological properties.
To date, three classes of ‘drug-like’ CDase inhibitors have been
reported. The quinolinone-based compounds (cerenib-1 and cere-
nib-2) were screened from a ChemBridge DIVERset library consist-
ing of approximately 50,000 drug-like compounds.13 These two
compounds inhibit cellular ceramidase activity; in doing so, they
reduced cellular sphingosine and S1P. The 5-fluorouracil releasing
drug carmofur was found to inhibit intracellular acid CDase activ-
ity in vivo.14 This effect was independent of the ability to generate
5-fluorouracil; carmofur also showed synergistic antiproliferative
actions with other anticancer drugs.15
anhydro-D
-glucitol)19 and seven other acertannin derivatives
(Fig. 2). An additional aim was to demonstrate the effects of these
compounds both on the inhibition of ceramide degradation
through ceramidase and the promotion of ceramide production
through ceramide synthase.
2. Results and discussion
2.1. Preparation of maplexins, ginnalins and acertannins
Acertannins were isolated from members of the maple (Acer)
genus in concentrations of only 0.0005–0.005 wt.%.19c Reports on
previously reported natural acertannins are fragmented and
limited; they have been given various names such as maplexin or
In a search for a new class of ceramide level regulator, we
turned our attention to plant materials. As part of an ongoing
OH
OH
OH
HO
HO
HO
HN
12
12
12
NH
NH2
NH
14
5
O
O
N
N
Ceramide
Sphingosine
O
C6-NBD-ceramide
(6-((N-(7-nitrobenz-2-oxa-1,3-diazol-4-
yl)amino)hexanoyl)sphingosine)
ii) lipophilic aromatic ceramide analogues
i) endocannabinoid-related compound
O
O
O
O
CH3
O2N
O2N
N
HN
HN
C13H27
HN
C13H27
OH
HN
7
7
11
HO
OH
CH3
OH
D-e-MAPP
OH
B13
OH
LCL-464
N-oleoylethanolamine
(OEA)
iv) 2,4-dioxopyrimidine-1-carboxamide derivatives
iii) quinolinone-based compounds
H
O
N
O
O
F
N
N
H
R
O
N
R
O
O
Carmofur: R=H
25a
: R = COOMe
Ceranib-1: R = Me
Ceranib-2: R = H
Figure 1. Structure of known ceramidase inhibitors.

A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
1179
OR4
(Fig. 2), (Table 1). N-Oleoyl-ethanol amine and carmofur were used
as positive controls. The basic core anhydrosugar, 1,5-AG 12 itself
exhibited less than 50 % inhibition of ceramidase, even at concen-
trations as high as 10 mM, whereas 2,4,6-tri-O-galloyl-1,5-AG 21
(maplexin E) and 2,3,6-tri-O-galloyl-1,5-AG 22 (maplexin F) were
good inhibitors of ceramidase, with IC50 values of 0.8 and
0.7 mM, respectively. These activities were slightly weaker than
OAE and carmofur (IC50 = 0.11 and 0.18 mM, respectively). Thus,
we next focused on the galloyl group and whether the number
and position could influence the inhibition activities of acertan-
nins. From a comparison of the potency of mono-O-galloyl-1,5-
AG 13–16, the galloyl substituent at the C2-position seems most
effective because 2-O-galloyl-1,5-AG 13 (ginnalin C) had a 3-fold
greater IC50 than the other mono-O-galloyl-1,5-AGs. In contrast,
the number of gallate groups does not necessarily correspond with
inhibition potency, e.g. the newly synthesized 2,3,4,6-tetra-O-gal-
loyl-1,5-AG 11 was five times weaker as an inhibitor than 22. Fur-
thermore, 30,40,50-trimethoxy benzoyl-1,5-AGs 23–25 obtained in
the present study showed no significant inhibitory activity toward
ceramidase. These results suggested that the presence of OH sub-
stituents of galloyl groups linked to a 1,5-AG core is an important
factor for the inhibition.
6
4
5
R3O
R2O
O
1
2
3
OR1
: R1 = R2 = R3 = R4 = H
12
13: R1 = Galloyl, R2 = R3 = R4 = H
14: R2 = Galloyl, R1 = R3 = R4 = H
15: R3 = Galloyl, R1 = R2 = R4 = H
: R4 = Galloyl, R1 = R2 = R3 = H
: R1 = R2 = Galloyl, R3 = R4 = H
16
17
18: R1 = R3 = Galloyl, R2 = R4 = H
19: R1 = R4 = Galloyl, R2 = R3 = H
20: R2 = R4 = Galloyl, R1 = R3 = H
: R = R4 = Galloyl, R1 = R2 = H
3
4
21
: R1 = R3 = R4 = Galloyl, R2 = H
22: R1 = R2 = R4 = Galloyl, R3 = H
7: R1 = R2 = R3 = Galloyl, R4 = H
10: R2 = R3 = R4 = Galloyl, R1 = H
: R1 = R2 = R3 = R4 = galloyl
11
23
: R4 = 3',4',5'-trimethoxy benzoyl, R1 = R2 = R3 = H
24: R3 = R4 = 3',4',5'-trimethoxy benzoyl, R1 = R2 = H
25: R1 = R2 = R3 = R4 = 3',4',5'-trimethoxy benzoyl
Figure 2. Structure of naturally-occurring acertannins 12–22 and related synthetic
acertannin compounds 4, 7, 10, 11, and 23–25.
ginnalin. The synthesis of the central core anhydrosugar 1,5-
anhydro-D
-glucitol (1,5-AG) has been reported several times:20 i)
2.3. Inhibition mechanism of acertannins for ceramidase
reduction of alkyl glucosides by TESH/Lewis acid: ii) reduction of
peracetylated phenyl thioglucoside by Raney-Ni: iii) reduction of
On the basis of these findings, it is clear that the addition of gal-
loyl groups to the 1,5-AG core can lead to the inhibition of cerami-
dase. However, it has not been completely explained why these
acertannins inhibit ceramidase because they do not have a similar-
ity to the ceramide structure. Acertannins are not related to previ-
ously reported ceramidase inhibitors (Fig. 1). Previous reports of
mutational and structural analyses have suggested that the mech-
anism of the cleavage of the N-acyl linkage of ceramide by cerami-
dase was similar to that observed for the Zn2+ dependent
carboxylpeptidases.22 This reaction starts with the deprotonation
of the coordinated water molecule mediated by His99 and
Arg160; the resulting hydroxide ion is then subjected to nucle-
ophilic attack on the carbonyl carbon atom of ceramide, thereby
the amide bond is hydrated. These analytical reports suggested
that Zn2+ plays an important role in ceramidase. Furthermore,
Inoue et al., compared the three-dimensional structures of the
active site from bacteria to mammals and found they closely
resembled each other by using homology modeling.22 In view of
these reports, we envisaged that the inhibition mechanism of acer-
tannins might involve interactions with the zinc ion. To test this
hypothesis, we investigated the chelating ability of 1,5-AG 12,
2,3,6-tri-O-galloyl-1,5-AG 22 (maplexin F), 2,4,6-tri-O-galloyl-1,5-
AG 21 (maplexin E), 2,3,4,6-tetra-O-galloyl-1,5-AG 11, 2,3,4,6-
tetra-O-(30,40,50-trimethoxy benzoyl)-1,5-AG 25, and carmofur
toward Zn2+ by using eriochrome black T (EBT) as a chelate indica-
tor. Ethylenediaminetetraacetic acid (EDTA) was used as a positive
control. The reaction mixture (4 mL) contained 0.25 mM of ZnCl2
with or without 0.5 mM samples. A quick color change was
observed in the presence of both EDTA and three acertannins 22,
21, and 11 (Fig. 3). Thus it was confirmed that these acertannins
have a chelating ability for Zn2+. In sharp contrast, when 1,5-AG
12, 2,3,4,6-tetra-O-(30,40,50-trimethoxy benzoyl)-1,5-AG 25 or car-
mofur were tested under the same conditions, no change in color
was found. These results suggest that 1,5-AG 12 itself does not
have a chelating ability and the OH group on the galloyl ring might
be interacting with Zn2+. Furthermore, the ‘drug-like’ CDase inhibi-
tor carmofur did not change the color. In order to confirm that the
CDase inhibition mechanism was based on chelating with Zn2+, we
compared the inhibition activity of these compounds in the pres-
ence or absence of saturated ZnCl2. It was found that the activity
of ceramidase was clearly lost in the presence of 2 mM of EDTA,
peracetylated
a-bromoglucoside by Bu3SnH under radical condi-
tions. For the purpose of preparing a series of acertannins, we used
the following methods (Scheme 1). Hydrogenation of crude com-
pound 3, which can be prepared by a condensation reaction of
2,3-dibenzylated 1,5-AG 1 and 3,4,5-tris-benzyloxy benzoic acid
2, gave 4,6-di-O-galloyl-1,5-AG 4. Highly regioselective cleavage
of the 4,6-benzylidene part of 5 was achieved by treatment with
sodium cyanoborohydride under acidic conditions. The main pro-
duct 6b was converted into 3,4,6-tri-O-galloyl-1,5-AG 7 by the
same procedure described for the synthesis of 4 from 1. The reac-
tion of 5 with benzyl chloride and sodium hydride afforded the
corresponding benzylated compounds 8a and 8b in 17% and 41%
yield, respectively. This ratio may be rationalized by the steric hin-
drance of the benzylidene group at the 4,6-position. The ratio of 8a
and 8b was in contrast to a previous report21 in which the
monobenzoylation of the hydroxyl groups of 4,6-O-benzylidene-
1,5-AG 5 gave more of the 3-O-benzyl derivative 8a. Compound
8b was converted into 3,4,6-tri-O-galloyl-1,5-AG 10 by debenzyli-
denation, followed by the same esterification protocol as in the
preparation of 4 in 44% yield from 8b. 2,3,4,6-Tetra-O-galloyl-1,5-
AG 11 was prepared from 1,5-AG by the same esterification proce-
dure described for the synthesis of compound 4 in 78% yield. Other
known gallotannins; ginnalin A 19, B 16, C 13, maplexins A 14, B
15, C 17, D 18, E 21, F 22, and 3,6-di-O-galloyl-1,5-AG 20 were pre-
pared by protection, and galloyl esterification of desired hydroxy
groups of 1,5-anhydro-D-glucitol, and finally deprotection. 6-O-
(30,40,50-Trimethoxybenzoyl)-1,5-AG 23 was prepared by the ester-
ification of 1,5-AG with one equivalent of 3,4,5-trimethoxybenzoyl
chloride in pyridine in 58% yield. Compound 1 was converted into
4,6-di-O-(30,40,50-trimethoxybenzoyl)-1,5-AG 24 by esterification
with 3,4,5-trimethoxybenzoyl chloride, followed by hydrogenoly-
tic debenzylation in 63% yield. Esterification of 1,5-AG with excess
amounts of 3,4,5-trimethoxybenzoyl chloride gave 2,3,4,6-tetra-O-
(30,40,50-trimethoxybenzoyl)-1,5-AG 25 in 45% yield.
2.2. Inhibitory potency of acertannins for ceramidase (ceramide
N-deacylase)
We first compared the ability of acertannins 4, 7, 10, 11, and
13–25 to inhibit a ceramidase (ceramide N-deacylase) activity

1180
A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
OBn
OH
OBn
BnO
BnO
R1
=
BnO
BnO
CO2H
O
R1O
R1O
BnO
R2O
R2O
HO
HO
HO
O
O
O
2
Pd(OH)2/C, H2
THF, MeOH
BnO
OBn
EDC, ClCH2CH2Cl
OBn
OH
HO
HO
R2
=
4
1
3
76% from 1
O
2
1) EDC,
HO
R2O
R2O
HO
BnO
BnO
Ph
O
ClCH2CH2Cl
O
NaCNBH3, HCl-gas
THF
O
O
O
O
HO
+
HO
HO
HO
2) Pd(OH)2/C, H2
OR2
OH
OH
OH
6b
66%
5
6a
5%
7
6b
92% from
2
1) EDC,
cat.TsOH, MeOH
HO
HO
HO
R2O
R2O
R2O
Ph
O
Ph
O
BnCl, NaH
DMF, 0°C
ClCH2CH2Cl
O
O
O
O
O
O
5
+
BnO
HO
2) Pd(OH)2/C, H2
OH
OBn
OBn
OH
8b
10
44% from
8a
17 %
8b
9
41 %
Scheme 1.
Table 1
the presence of zinc ions. In contrast, when applied with 2 mM
2,3,6-tri-O-galloyl-1,5-AG 22 (maplexin F) and 2,4,6-tri-O-galloyl-
1,5-AG 21 (maplexin E), the ceramidase activities did not recover.
Tannic acids have been used as rust prevention treatment agents
for metals for a long time. They react with the metal ions to form
a metal tannate, thereby forming a coating film whose degree of
protection can be controlled to some extent by the method of
application. This knowledge and our results might suggest that
acertannins 22 and 21 form a chelate but it is different from EDTA
with that of acertannins being a more complex and irreversible
Concentration of red maple components giving and related synthetic compounds 50%
inhibition of ceramidase
IC50 (mM)
Compound
Ceramidase
aNI
Anhydrosugar
1,5-Anhydro-
D
-glucitol 12 (1,5-AG)
Mono-O-galloyl-1,5-AG
2-O-Galloyl-1,5-AG 13 (Ginnalin C)
3-O-Galloyl-1,5-AG 14 (Maplexin A)
4-O-Galloyl-1,5-AG 15 (Maplexin B)
6-O-Galloyl-1,5-AG 16 (Ginnalin B)
1.0 0.1
3.0 0.4
2.3 0.2
3.5 0.5
conjugate interaction with Zn2+
.
Di-O-galloyl-1,5-AG
2.4. Effect of acertannins on ceramide production through
ceramide synthase
2,3-Di-O-galloyl-1,5-AG 17 (Maplexin C)
2,4-Di-O-galloyl-1,5-AG 18 (Maplexin D)
2,6-Di-O-galloyl-1,5-AG 19 (Ginnalin A)
3,6-Di-O-galloyl-1,5-AG 20
1.2 0.1
1.0 0.2
1.6 0.1
1.3 0.1
2.0 0.04
Dry skin, senile xerosis, is one of the most prevalent dermato-
logical problems seen in the elderly with prevalence rate ranges
of 30–58%.23 It is caused by the disruption of the equilibrium
between ceramide synthesis and ceramide degradation cycles
and thereby failure of multilamellar sheets in intercellular spaces
of stratum corneum. Consequently, it causes fissures at the skin
surface and decrease of skin barrier function and allows greater
transepidermal water loss, which also triggers induction of other
skin disorders.24 A possible strategy for suppressing age-depen-
dent degradation of skin ceramide and preventing transepidermal
water loss may be through not only the inhibition of ceramidase,
but also promotion of the ceramide synthesis. Epidermal cera-
mides are generated by i) a de novo pathway involving serine-
palmitoyl CoA transferase; ii) ceramide synthase, which catalyzes
the acylation of sphingosine to form ceramide; iii) sphingomyelin
hydrolysis involving sphingomyelinase; and iv) hydrolysis from
glucosylsphingolipid involving b-glucocerebrosidase. We focused
on the ceramide synthase pathway. We evaluated acertannins as
promoters of ceramide production in vitro in a normal human
4,6-Di-O-galloyl-1,5-AG 4
Tri-O-galloyl-1,5-AG
2,4,6-Tri-O-galloyl-1,5-AG 21 (Maplexin E)
2,3,6-Tri-O-galloyl-1,5-AG 22 (Maplexin F)
2,3,4-Tri-O-galloyl-1,5-AG 7
0.8 0.1
0.7 0.03
1.3 0.1
1.3 0.2
3,4,6-Tri-O-galloyl-1,5-AG 10
Tetra-O-galloyl-1,5-AG
2,3,4,6-Tetra-O-galloyl-1,5-AG 11
3.7 0.3
Methoxy O-galloyl-1,5-AG
6-O-(30,40,50-trimethoxy benzoyl)-1,5-AG 23
4,6-Di-O-(30,40,50-trimethoxy benzoyl)-1,5-AG 24
2,3,4,6-Tetra-O-(30,40,50-trimethoxy benzoyl)-1,5-AG 25
NI
NI
NI
Positive control
N-Oleoyl-ethanol amine (OEA)
Carmofur
0.17 0.01
0.11 0.02
a
Less than 50% inhibition at 10 mM.
but the activity was restored with 1 mM of ZnCl2 (Table 2). These
results indicated that the ceramidase activity was controlled by

A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
1181
Figure 3. Chelating ability of acertannins 22 and 21 and carmofur against Zn2+. EDTA was used as positive control.
nins, therefore, represent a new class of potential intracellular cer-
amide regulators showing both inhibition of CDase and promotion
of ceramide synthase.
Table 2
Effects of ZnCl2 on the inhibitory activity of 2 mM EDTA and acertannins 22 and 21 on
ceramidase
Compound
Inhibition of activity (%)
Without With ZnCl2 (1 mM)
76.2 1.8 27.3 4.0
4. Experimental section
4.1. Chemistry
EDTA
2,3,6-Tri-O-galloyl-1,5-AG 22 (Maplexin F) 73.6 3.5 74.4 1.8
2,4,6-Tri-O-galloyl-1,5-AG 21 (Maplexin E) 62.3 2.7 67.2 1.7
Infrared (IR) spectra were recorded on a HORIBA FT-IR spec-
trometer FT-720. Mass spectra (MS) were recorded on a JEOL
JMN-DX 303/JMA-DA 5000 spectrometer. Optical rotations were
measured with a JASCO DIP-360, JASCO P-1020 and HORIBA
SEPA-300 digital polarimeter. Proton nuclear magnetic resonance
keratinocyte cell line HFb16d by using a metabolic label method.
Cells were cultured in the presence of 50
lM acertannins 4, 7,
(1H NMR) spectra were recorded on
a JEOL JNM-AL 300
10, 11, and 13–25 and 10 M NBD-hexanoic acid (Matreya LLC.,
l
PA, USA) for 24 h, and the fluorescence of generated intracellular
NBD-ceramide were detected with LAS4000 (Fujifilm, Tokyo,
Japan) (Fig. 4). The results showed the ratio to non-treated cells.
6-O-Galloyl-1,5-AG 16 (Ginnalin B), 2,6-di-O-galloyl-1,5-AG 19
(Ginnalin A), and 4,6-di-O-galloyl-1,5-AG 4 clearly increased intra-
cellular NBD-labeled ceramide in normal human keratinocyte, with
increases of 2.3, 2.2, and 2.1-fold, respectively. These three acertan-
nins have a C6-gallate in common, indicating that the gallate group
at the C6-position may be a factor in promoting potency against
ceramide synthase. Furthermore the ceramide production
increased dose-dependently (Fig. 5). To confirm this promotion
effect, we furthermore performed quantitative RT-PCR against cer-
amide synthase 3 (CERS3), which is highly expressed in ker-
(300 MHz) spectrometer, using tetramethylsilane as an internal
standard. The following abbreviations are used: s = singlet,
d = doublet, t = triplet, q = quartet, m = multiplet, br = broad. Col-
umn chromatography was carried out on a FUJI SILYSIA CHEMICAL
Silica gel PSQ100B (100 mesh) for flash chromatography. Reactions
were monitored by TLC (Merck TLC silica gel 60 F254) and visual-
ized by charring with Hanessian’s stain.
4.1.1. 4,6-Di-O-galloyl-1,5-anhydro-D-glucitol 4
To compound 2 (4.0 g, 8.8 mmol) and 1-ethyl-3-(3-dimethy-
laminopropyl) carbodiimide (EDC) (1.5 g, 9.7 mmol) in 1,2-dichlor-
oethane (80 mL), compound 120d (1.4 g, 4.1 mmol) was added
under Ar. The reaction mixture was allowed to warm to 75 °C for
12 h. The reaction mixture was poured into 0.1 M HCl (200 mL),
and diluted with CHCl3 (200 mL). The organic phase was washed
with H2O (200 mL) and dried (MgSO4). Filtration and
concentration under reduced pressure gave the crude compound
3 (4.0 g) as a yellow oil. A mixture of unpurified compound 3,
20% Pd(OH)2 on carbon (100 mg), and 50% Methanol/THF (80 mL)
was treated with hydrogen (0.7 MPa, 40 °C) for 8 h. Filtration
through Celite and purification by silica gel column
chromatography (10:1 CHCl3/MeOH) gave 1.45 g (76% from 1) of
atinocytes.25 The relative amount of CERS3 gene was calculated
DDCt
using the 2ꢀ
method with glyceraldehyde 3-phosphate dehy-
drogenase (GAPDH) as an internal control. 6-O-galloyl-1,5-AG 16
(Ginnalin B), 2,6-di-O-galloyl-1,5-AG 19 (Ginnalin A), and 4,6-di-
O-galloyl-1,5-AG 4 clearly increased the expression of CERS3, by
1.5, 1.7, and 1.4-fold, respectively (Fig. 6). These results suggested
that the increase of internal ceramide production is due to promo-
tion of the expression of ceramide synthase at gene level.
3. Conclusion
compound 4 (4,6-di-O-galloyl-1,5-anhydro-
D-glucitol) as white
solid. [
a]
25 = +63.4 (c 1.0, MeOH); 1H NMR (D2O, 300 MHz), d
D
For the purpose of clarifying the difference in inhibition
requirements between the acertannins, we synthesized all 10 nat-
urally-occurring acertannins and synthesized seven new acertan-
nin derivatives and performed side-by-side comparisons. The
main features are as follows: (a) 2,4,6-tri-O-galloyl-1,5-AG 21
(maplexin E) and 2,3,6-tri-O-galloyl-1,5-AG 22 (maplexin F) were
good inhibitors of ceramidase, with IC50 values of 0.8 and
0.7 mM, respectively; (b) the presence of galloyl groups linked to
hydroxyls of the 1,5-AG core is an important factor for the inhibi-
tion; (c) acertannins have a chelating ability with Zn2+, which is
assumed to be one of the reasons for the CDase inhibition; (d) 6-
O-galloyl-1,5-AG 16 (Ginnalin B), 2,6-di-O-galloyl-1,5-AG 19
(Ginnalin A), and 4,6-di-O-galloyl-1,5-AG 4 raised intracellular
ceramide levels and promoted the CERS3 expression level. Acertan-
3.62–3.76 (m, 4H), 3.96 (d, 1H, 9.9), 4.08–4.24 (m, 2H), 4.94 (t,
1H, 6.6, 7.3), 6.78 (s, 2H), 6.91 (s, 2H); 13C NMR (D2O, 75 MHz), d
64.5, 69.8, 69.9, 73.3, 75.5, 76.0, 110.4, 110.6, 120.5, 138.7, 139.0,
144.9, 145.1, 167.8, 168.1. IR mmax (KBr) 526, 565, 607, 661, 763,
811, 867, 997, 1037, 1099, 1240, 1348, 1452, 1540, 1616, 1700,
3396 cmꢀ1
;
ESI-TOFMS m/z: found 491.07917 [M+Na+];
C20H20NaO+13 requires for 491.08016.
4.1.2. 4-O-Benzyl-1,5-anhydro-D-glucitol 6a and 6-O-Benzyl-1,5-
anhydro- -glucitol 6b
D
Compound 5 (2.52 g, 10 mmol) and NaCNBH3 (5.02 g, 80 mmol)
were dissolved in THF (50 mL) after which HCl-gas/1,4-dioxane
solution (ca4 mol/L) was added until the pH of the solution
became 2–3 at 0 °C under Ar. The mixture was stirred at the same

1182
A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
Figure 4. Effect of acertannins 4, 7, 10, 11, and 13–25 on ceramide synthesis. Ceramide synthesis was evaluated by using metabolic label methods with NBD–hexanoic acid.
3.5
6-O-Galloyl-1,5-AG (16)
3.0
2.5
2.0
1.5
1.0
0.5
0.0
2,6-di-O-Galloyl-1,5-AG (19)
4,6-di-O-Galloyl-1,5-AG (4)
(µM)
10
50
100
compound dose
Figure 6. Promotion effect of acertannins 16, 19, and 4 on expression of ceramide
synthase 3 (CERS3) gene.
Figure 5. Dose-dependent effect of acertannins 16, 19, and
4
on ceramide
synthesis.
compound 6a (0.14 g, 5.4%) and 6b (1.69 g, 65.6%) as a colorless
25
oil, respectively. 4-O-Benzyl-1,5-anhydro-
D
-glucitol 6a: [
a
]
D
=
+39.7 (c 1.1, MeOH); 1H NMR (CD3OD, 300 MHz), d 3.14 (t, 1H,
11.0), 3.21 (ddd, 1H, 2.2, 5.1), 3.27–3.31 (overlap, 1H), 3.44–3.51
(m, 2H), 3.60 (dd, 1H, 11.8, 5.1), 3.78 (dd, 1H, 11.8, 2.2), 3.89 (dd,
1H, 11.0, 4.9), 4.55–4.96 (m, 2H), 7.26–7.42 (m, 5H); 13C NMR (CD3-
OD, 75 MHz), d 62.7, 70.9, 71.8, 75.8, 79.5, 80.4, 81.7, 128.6, 128.9,
temperature for 18 h. The reaction was quenched by addition of
saturated NaHCO3 aq (50 mL). The organic phase was washed with
water (100 mL) and dried (MgSO4). Filtration and condensation
under reduced pressure gave a light yellow oil, which was purified
by silica gel column chromatography (1:1 Hexane/AcOEt) to give

A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
1183
129.1, 129.3, 140.0; IR
m
max (KBr) 513, 609, 660, 698, 740, 844, 899,
reduced pressure. The residue was purified by silica gel column
chromatography (10:1 CHCl3/MeOH) to give compound
(650 mg, 97%) as a white solid. [
25 = +7.2 (c 1.0, MeOH); 1H
995, 1064, 1114, 1211, 1246, 1330, 1365, 1454, 1496, 2862, 2916,
9
3035 cmꢀ1; HRMS (FAB) m/z: found 255.1231 [M+H+]; C13H19O5+
a]
D
requires for 255.1232. 6-O-Benzyl-1,5-anhydro-
[a]
D
1H, 10.6), 3.27–3.33 (m, 1H), 3.40–3.42 (m, 2H), 3.53–3.62 (m,
D
-glucitol 6b:
NMR (CD3OD, 300 MHz), d 3.09–3.17 (m, 2H), 3.22–3.47 (m, 3H),
3.59 (dd, 1H, 11.7, 5.9), 3.82 (dd, 1H, 11.7, 2.2), 3.97 (dd, 1H,
11.3, 5.1), 7.24–7.39 (m, 5H); 13C NMR (CD3OD, 75 MHz), d 63.1,
68.9, 71.9, 74.1, 79.2, 79.3, 82.3, 128.7, 129.0, 129.3, 140.3; IR mmax
(KBr) 439, 489, 551, 590, 651, 694, 748, 833, 883, 925, 995, 1041,
1106, 1157, 1211, 1265, 1334, 1376, 1450, 1496, 1542, 1650,
2869, 2911, 2977, 3031, 3062, 3436 cmꢀ1; MS(EI): m/z: found
254.1155 [M+]; C13H18O+5 requires for 254.1154.
17 = +80.7 (c 0.3, MeOH); 1H NMR (CDCl3, 300 MHz), d 3.14 (t,
1H), 3.64–3.70 (m, 2H), 3.91 (dd, 1H, 11.4, 5.5), 4.53 (d, 2H),
13
7.27–7.34 (m, 5H);
C NMR (CDCl3, 75 MHz), d 69.6, 69.9, 70.1,
71.3, 73.8, 78.8, 127.9, 127.9, 128.5, 137.6; IR (KBr) 412, 489,
536, 609, 667, 698, 752, 867, 906, 1010, 1076, 1141, 1211, 1241,
1322, 1369, 1454, 1469, 1650, 1789, 2861, 3351 cmꢀ1; HRMS
(FAB) m/z: found 254.1155 [M+]; C13H18O5+ requires for 254.1154.
4.1.6. 3,4,6-Tri-O-galloyl-1,5-anhydro-D-glucitol 10
4.1.3. 2,3,4-Tri-O-galloyl-1,5-anhydro-
D
-glucitol 7
Compound 10 was synthesized from compound 9 according to
Compound 7 was synthesized from compound 6b according to
the same procedure described for the synthesis of compound 4.
the same procedure described previously for the synthesis of com-
pound 4. White solid; [a]
25 = +46.4 (c 0.2, MeOH); 1H NMR (CD3OD,
D
[
a]
D
17 = ꢀ3.2 (c 0.3, MeOH); 1H NMR (CD3OD, 300 MHz), d 3.49–
300 MHz), d 3.45 (t, 1H, 11.7, 10.3), 3.88–3.96 (m, 2H), 4.09 (dd, 1H,
10.3, 5.5), 4.23 (dd, 1H, 12.4, 4.8), 4.45 (dd, 1H, 12.4, 1.8), 5.29–5.41
(m, 2H), 6.95 (s, 2H), 7.00 (s, 2H), 7.08 (s, 2H); 13C NMR (CD3OD,
75 MHz), d 62.5, 68.3, 69.1, 69.7, 76.6, 77.0, 108.9, 109.0, 119.1,
119.8, 138.4, 138.5, 138.8, 144.9, 145.0, 145.1, 165.9, 166.6,
166.7; IR mmax (KBr) 433, 764, 806, 868, 1003, 1033, 1095, 1222,
1319, 1454, 1539, 1616, 1701, 3394 cmꢀ1; MS (FAB): m/z 621;
HRMS (FAB) m/z: found 621.1097 [M+H+]; C27H25O1+7 requires for
621.1092.
3.57 (m, 2H), 3.62–3.70 (m, 2H), 4.24 (dd, 1H, 11.0, 5.5), 5.14–
5.22 (m, 1H), 5.27 (t, 1H, 9.9), 5.66 (t, 1H, 9.9), 6.87 (s, 2H), 6.93
(s, 2H), 6.93 (s, 2H); 13C NMR (CD3OD, 75 MHz), d 62.2, 67.9,
70.4, 71.2, 75.4, 80.7, 110.3, 120.4, 120.6, 140.2, 146.3, 146.4,
167.1, 167.2, 167.7; IR mmax (KBr) 764, 868, 995, 1030, 1095,
1207, 1322, 1454, 1538, 1616, 1701, 3421 cmꢀ1; ESI-TOFMS m/z:
found 621.1093 [M+H+]; C27H25O1+7 requires for 621.1092.
4.1.4. 3-O-Benzyl-4,6-O-benzylidene-1,5-anhydro-
D-glucitol 8a
and 2-O-benzyl-4,6-O-benzylidene-1,5-anhydro- -glucitol 8b
D
4.1.7. 2,3,4,6-Tetra-O-galloyl-1,5-anhydro-
Compound 11 was synthesized from 1,5-anhydro-
according to the same procedure described previously for com-
pound 4 synthesis. White solid; [
17 = +71.6 (c 0.2, MeOH); 1H
NMR (CD3OD, 300 MHz), d 3.66 (t, 1H, 11.0), 4.05 (ddd, 1H, 9.9,
2.2), 4.30–4.36 (m, 2H), 4,48 (dd, 1H, 12.8, 2.2), 5.29 (m, 1H),
5.51 (t, 1H, 9.9), 5.77 (t, 1H, 9.9), 6.96 (s, 2H), 7.02 (s, 2H), 7.03
(s, 2H), 7.15 (s, 2H); 13C NMR (CD3OD, 75 MHz), d 63.7, 68.0,
70.3, 71.0, 75.3, 78.0, 110.4, 120.4, 120.4, 120.6, 121.1, 139.9,
140.0, 140.2, 146.2, 146.4, 167.0, 167.2, 167.6, 168.0; IR mmax
(KBr) 765, 868, 995, 1030, 1091, 1207, 1323, 1338, 1454, 1539,
1616, 1704, 3386 cmꢀ1. MS (FAB): m/z 773; HRMS (FAB) m/z:
found 773.1208 [M+H+]; C34H29O2+1 requires for 773.1201.
D-glucitol 11
Compound 520d (2.52 g, 10 mmol) was dissolved in DMF
(25 mL) and 60% NaH (0.44 g, 11 mmol) was added at 0 °C under
Ar. The mixture was cooled to ꢀ5 °C and BnCl (1.26 g, 10 mmol)
was added dropwise over 1 h. The mixture was stirred at the
same temperature for 19 h. The reaction was quenched by
addition of MeOH (10 mL), and diluted with 0.1 N-HCl (200 mL).
The products were extracted with AcOEt (200 mL); the organic
phase was washed with water (200 mL), dried (MgSO4), and
concentrated under reduced pressure. The residue was purified
by silica gel column chromatography (4:1 Hexane/AcOEt) to give
compound 8a (1.14 g, 16.8%) and 8b (2.76 g, 40.6%). 3-O-Benzyl-
D-glucitol
a]
D
4,6-O-benzylidene-1,5-anhydro-
D-glucitol 8a as a light yellow
solid, respectively. [
a
]
D
25 = +3.9 (c 1.0, CHCl3); 1H NMR (CDCl3,
300 MHz), d 2.36 (s, 1H), 3.29–3.44 (m, 2H), 3.53–3.80 (m, 4H),
4.04 (dd, 1H), 4.32 (dd, 1H), 4.88 (dd, 2H), 5.56 (s, 1H), 7.31–7.49
(m,10H); 13C NMR (CDCl3, 75 MHz), d 68.9, 69.9, 70.0, 71.6, 74.7,
82.2, 82.7, 101.3, 126.0, 128.0, 128.1, 128.3, 128.6, 129.0, 137.4,
138.4; IR mmax (KBr) 440, 505, 559, 624, 648, 694, 744, 903, 1003,
1068, 1103, 1138, 1227, 1281, 1365, 1412, 1454, 1496, 1542,
1650, 1809, 2858, 2927, 2978, 3028, 3062 cmꢀ1. MS (FAB): m/z
343; HRMS (FAB) m/z: found 343.1542 [M+H+]; C20H23O5+
4.1.8. 2,6-Di-O-galloyl-1,5-anhydro- -glucitol 19 (Ginnalin A)
[a]
20 = +22.5 (c 0.6, MeOH). [lit.19a 20 = +22 (c 0.6, acetone)];
D
D
[
a]
D
1H NMR (CD3OD, 300 MHz), d 3.26–3.30 (m, 2H, overlap), 3.47–
3.55 (m, 2H), 3.66 (t, 1H, 8.1, 7.9), 4.05 (dd, 1H, 11.0, 5.0), 4.33
(m, 1H), 4.50 (d, 1H, 11.0), 7.03 (s, 2H), 7.04 (s, 2H); 13C NMR
(CD3OD, 75 MHz), d 64.9, 67.9, 71.9, 73.1, 76.9, 80.1, 110.1, 110.2,
121.1, 121.3, 139.8, 139.9, 146.4, 146.5, 167.8, 168.3.
requires 343.1546. 2-O-Benzyl-4,6-O-benzylidene-1,5-anhydro-
D
-
4.1.9. 6-O-Galloyl-1,5-anhydro-
21 = +21.6 (c 1.0, MeOH). [lit.19b
[a]
D
D
-glucitol 16 (Ginnalin B)
glucitol 8b:
[a
]
D
25 = ꢀ28.5 (c 1.1, CHCl3); 1H NMR (CDCl3,
[
a]
19 = +14.8 (c 2.5,
D
300 MHz), d 2.68 (d, 1H), 3.25–3.37 (m, 2H), 3.44 (t, 1H, 9.16),
3.52–3.67 (m, 2H), 3.83 (ddd, 1H, 9.16, 2.2), 3.99 (dd, 1H, 11.0,
5.9), 4.28 (dd, 1H, 11.0), 4.71 (dd, 2H), 5.49 (s, 1H), 7.23–7.50 (m,
5H); 13C NMR (CDCl3, 75 MHz), d 68.5, 68.8, 71.0, 73.5, 74.9, 77.8,
81.2, 101.9, 126.3, 127.9, 128.0, 128.3, 128.6, 129.2, 137.2, 138.1;
IR mmax (KBr) 440, 490, 551, 590, 651, 694, 748, 833, 883, 925,
995, 1041, 1106, 1211, 1265, 1334, 1376, 1450, 1496, 1542,
1589, 1820, 2870, 2912, 2966, 3031, 3062 cmꢀ1; MS (FAB): m/z
343; HRMS (FAB) m/z: found 343.1547 [M+H+]; C20H23O5+
requires 343.1546.
acetone)]; 1H NMR (CD3OD, 300 MHz), d 3.16 (t, 1H, 10.3, 11.0),
3.25–3.46 (m, 4H), 3.86 (dd, 1H, 11.0, 5.5), 4.27 (dd, 1H, 12.1,
5.5), 4.47 (dd, 1H, 12.1, 1.8), 7.02 (s, 2H); 13C NMR (CD3OD,
75 MHz), d 62.7, 71.0, 71.6, 72.8, 77.7, 80.9, 110.3, 121.2, 126.0,
140.0, 146.5, 167.8.
4.1.10. 2-O-Galloyl-1,5-anhydro-
19 = +63.6 (c 0.28, MeOH). [lit.19b
[a]
D
D
-glucitol 13 (Ginnalin C)
[
a]
20 = +64.8 (c 2.7, DMSO)];
D
1H NMR (CD3OD, 300 MHz), d 3.22–3.36 (overlap, m, 4H), 3.57–
3.64 (m, 2H), 3.84 (d, 1H, 10.6), 4.04 (m, 1H), 7.04 (s, 2H); 13C
NMR (CD3OD, 75 MHz), d 62.9, 67.8, 72.0, 73.2, 77.0, 82.5, 110.2,
121.1, 139.8, 146.3, 167.7.
4.1.5. 2-O-Benzyl-1,5-anhydro-D-glucitol 9
Compound 8b (900 mg, 2.63 mmol) was dissolved in MeOH
(30 mL), and treated with a catalytic amount of TsOH and the reac-
tion mixture heated to reflux for 12 h. The reaction mixture was
neutralized with anion exchange resin, and concentrated under
4.1.11. 3-O-Galloyl-1,5-anhydro-
20 = +25 (c 0.4, MeOH). [lit.19c
[a]
D
NMR (D2O, 300 MHz), d 3.34–3.44 (m, 2H), 3.62–3.88 (m, 4H), 4.02
D
-glucitol 14 (Maplexin A)
[
a]
25 = +25 (c 0.28, MeOH)]; 1H
D

1184
A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
(m, 1H,) 5.04 (t, 1H, 10.3, 8.8), 7.20 (s, 2H); 13C NMR (D2O, 75 MHz),
d 61.7, 68.5, 68.7, 69.4, 80.5, 80.9, 110.9, 121.2, 139.1, 145.3, 168.8.
+30.8 (c 0.25, MeOH); 1H NMR (CDCl3, 300 MHz), d 3.21 (t, 1H,
11.0, 9.9), 3.35 (dd, 1H, 10.3, 8.8), 3.46–3.54 (m, 2H), 3.58–3.63
(m, 1H), 3.86 (s, 6H), 3.87 (s, 3H), 3.96 (dd, 1H, 11.3, 4.8), 4.48 (dd,
1H, 11.7, 5.5), 4.57 (dd, 1H, 11.7, 1.1), 7.27 (s, 2H); 13C NMR (CDCl3,
75 MHz), d 56.4, 60.9, 64.9, 69.6, 70.0, 70.7, 78.5, 78.6, 107.4, 124.4,
4.1.12. 4-O-Galloyl-1,5-anhydro-
20 = +13.3 (c 0.06, MeOH). [lit.19c
[a]
D
D
-glucitol 15 (Maplexin B)
[
a]
20 = +15 (c 0.06, MeOH)];
D
1H NMR (D2O, 300 MHz), d 3.47–3.70 (m, 6H), 3.98 (dd, 1H, 10.6,
5.1), 4.87 (t, 1H, 9.9, 7.7), 7.13 (s, 2H); 13C NMR (D2O, 75 MHz), d
61.2, 69.6, 70.1, 72.1, 76.2, 79.0, 110.8, 120.8, 139.3, 145.3, 167.9.
142.9, 153.0, 153.0, 167.1; IR mmax (KBr) 451, 471, 520, 601, 644, 759,
860, 902, 995, 1052, 1122, 1176, 1230, 1338, 1419, 1461, 1504,
1589, 1712, 2842, 2942, 3405 cmꢀ1; HRMS (FAB) m/z: found
359.13514 [M+H+]; C16H23O+9 requires for 359.13418.
4.1.13. 2,3-Di-O-galloyl-1,5-anhydro-
D
-glucitol 17 (Maplexin C)
[
a]
20 = +11.3 (c 0.12, MeOH). [lit.19c
[a
]
D
20 = +13 (c 0.12, MeOH)];
4.1.19. 4,6-Di-O-(30,40,50-trimethoxy benzoyl)-1,5-anhydro-
D-
glucitol 24
D
1H NMR (CD3OD, 300 MHz), d 3.31–3.35 (m, 1H), 3.40 (t, 1H, 10.6),
3.62–3.70 (m, 2H), 3.85 (dd, 1H, 11.8, 2.6), 4.15 (dd, 1H, 10.6, 5.8),
5.02 (m, 1H), 5.35 (t, 1H, 9.5, 9.2), 6.91 (s, 2H), 7.00 (s, 2H); 13C
NMR (CD3OD, 75 MHz), d 62.6, 67.8, 69.8, 71.4, 77.9, 82.7, 110.3,
110.3, 120.5, 121.3, 139.9, 140.1, 146.3, 146.4, 167.4, 168.2.
Compound 1 (0.5 g, 1.45 mmol) was dissolved in pyridine
(30 mL) and 3,4,5-trimethoxybenzoyl chloride (1.0 g, 4.35 mmol)
was added under Ar. The reaction mixture was stirred at 40 °C
for 19 h, the mixture was poured into water (250 mL), and diluted
with CHCl3 (250 mL). The organic phase was washed with 0.5 N
HCl (250 mL) and water (250 mL), and dried (MgSO4). Filtration
and concentration under reduced pressure gave a light yellow oil
which was hydrogenated with 5% Pd on carbon (50 mg) and hydro-
gen (0.3 MPa, 40 °C) in methanol (50 mL) for 18 h. Filtration
through Celite and purification by silica gel column chromatogra-
phy (1:1 Hexane/AcOEt) gave compound 24 (0.51 g, 63% from 1)
4.1.14. 2,4-Di-O-galloyl-1,5-anhydro- -glucitol 18 (Maplexin D)
[a]
22 = +8.8 (c 0.17, MeOH). [lit.19c 20 = +6 (c 0.17, MeOH)]; 1H
D
D
[
a]
D
NMR (CD3OD, 300 MHz), d 3.39 (t, 1H, 10.6), 3.49–3.66 (m, 3H),
3.87 (m, 1H), 4.06 (dd, 1H, 10.6, 5.5), 5.14 (t, 1H, 9.2, 9.9), 5.33 (t,
1H, 9.2, 9.5), 6.94 (s, 2H), 6.99 (S, 2H); 13C NMR (CD3OD,
75 MHz), d 62.4, 69.8, 70.6, 71.0, 78.5, 80.6, 110.3, 120.5, 121.2,
139.8, 140.1, 146.3, 146.4, 167.3, 168.1.
as a white solid. mp: 182-184 °C; [a]
18 = +47.5 (c 0.25, CHCl3); 1H
D
NMR (CDCl3, 300 MHz), d 3.28–3.38 (m, 2H), 3.72–3.89 (m, 20H),
4.09 (dd, 1H, 11.4, 5.1), 4.27 (dd, 1H, 12.1, 5.5), 4.66 (dd, 1H,
12.1, 2.6), 5.13 (t, 10.6, 9.2), 7.26 (s, 4H); 13C NMR (CDCl3,
75 MHz), d 56.3, 60.9, 64.0, 69.4, 70.8, 72.6, 76.6, 107.2, 107.4,
124.0, 124.6, 142.6, 143.0, 153.0, 166.0, 166.3; IR mmax (KBr) 528,
647, 759, 856, 937, 995, 1037, 1095, 1130, 1180, 1226, 1338,
1415, 1461, 1508, 1592, 1708, 2838, 2884, 2954, 3004, 3220,
3301, 3482 cmꢀ1; HRMS (FAB) m/z: found 553.1920 [M+H+];
4.1.15. 2,4,6-Tri-O-galloyl-1,5-anhydro-
E)
D
-glucitol 21 (Maplexin
[a] [a]
21 = +12.5 (c 0.12, MeOH). [lit.19c 20 = +10 (c 0.13, MeOH)];
D D
1H NMR (CD3OD, 300 MHz), d 3.41 (t, 1H, 12.0, 10.6), 3.82 (m, 1H),
3.99 (t, 1H, 9.5, 8.4), 4.11–4.19 (m, 2H), 4.37 (dd, 1H, 11.0, 2.6), 5.00
(m, 1H), 5.18 (t, 1H, 9.5), 7.04 (s, 4H), 7.06 (s, 2H); 13C NMR
(CD3OD, 75 MHz), d 64.1, 68.1, 72.6, 73.2, 74.9, 78.3, 110.3, 110.4,
110.5, 121.1, 121.3, 139.9, 140.1, 140.1, 146.5, 146.5, 146.5,
167.5, 167.7, 168.1.
C
26H33O+13 requires for 553.1921.
4.1.20. 2,3,4,6-Tetra-O-(30,40,50-trimethoxy benzoyl)-1,5-anhy-
dro- -glucitol 25
1,5-Anhydro- -glucitol 3,4,5-
4.1.16. 2,3,6-Tri-O-galloyl-1,5-anhydro-
D
-glucitol 22 (Maplexin
20 = +68 (c 0.1, MeOH)]; 1H
D
D
F)
D
(0.5 g,
3.05 mmol)
and
[
a]
20 = +65.8 (c 0.2, MeOH). [lit.19d
[
a]
trimethoxybenzoyl chloride (5.6 g, 24.4 mmol) were dissolved in
pyridine (50 mL), and the mixture was stirred for 18 h at 75 °C.
The reaction was quenched by addition of water (250 mL), and
diluted with AcOEt (300 mL). The aqueous phase was extracted
by AcOEt (300 mL). The organic phase was combined and washed
with 0.5 N-HCl (200 mL) and water (200 mL), and dried (MgSO4).
Filtration and concentration under reduced pressure afforded a
residue which was purified by silica gel column chromatography
(10:1 dichloromethane/AcOEt) to give compound 25 (1.30 g,
D
NMR (CD3OD, 300 MHz), d 3.40 (t, 1H, 10.6), 3.51–3.66 (m, 3H),
3.98 (t, 1H, 9.2), 4.18 (dd, 1H, 11.0, 5.5), 4.95–5.04 (m, 2H), 6.90
(s, 1H), 6.95 (s, 1H), 7.03 (s, 2H), 7.05 (s, 2H); 13C NMR (CD3OD,
75 MHz), d 62.7, 68.0, 72.9, 73.4, 75.0, 81.0, 110.4, 110.4, 121.1,
121.2, 140.0, 140.1, 146.5, 146.5, 167.7, 167.7.
4.1.17. 3,6-Di-O-galloyl-1,5-anhydro-
D-glucitol 20
[a]
D
20 = +67 (c 0.2, MeOH); 1H NMR (CD3OD, 300 MHz), d 3.44
(dd, 2H), 3.54–3.77 (m, 3H), 3.95 (dd, 1H, 11.8, 5.9), 4.34 (dd, 1H,
11.8, 6.9), 4.52 (dd, 1H, 11.8, 1.8), 5.04 (t, 1H, 8.8), 7.03 (s, 2H),
7.09 (s, 2H); 13C NMR (CD3OD, 75 MHz), d 64.9, 69.8, 70.1, 71.1,
80.1, 80.9, 110.1, 110.3, 121.3, 121.8, 139.7, 139.9, 146.4, 146.5,
168.3, 168.4; IR mmax (KBr) 443, 482, 528, 601, 648, 764, 868,
1038, 1091, 1223, 1315, 1338, 1454, 1539, 1608, 1697, 3629,
45.4%) as a white solid. [a]
18 = +5.1 (c 0.35, CHCl3); 1H NMR (CDCl3,
D
300 MHz), d 3.61 (t, 1H, 11.0, 9.9), 3.83–3.92 (m, 36H), 4.07 (ddd,
1H, 9.5, 5.1, 2.6), 4.37 (dd, 1H, 12.1, 5.1), 4.50 (dd, 1H, 11.0, 5.5),
4.77 (dd, 1H, 12.1, 2.6), 5.35 (ddd, 1H, 9.9, 5.5), 5.60 (t, 1H, 9.5),
5.86 (t, 1H, 9.9), 7.13 (s, 2H), 7.15 (s, 2H), 7.27 (s, 2H), 7.34 (s,
2H); 13C NMR (CDCl3, 75 MHz), d 56.2, 56.2, 56.3, 60.9, 60.9, 63.5,
67.3, 69.8, 70.5, 74.4, 77.0, 107.1, 107.2, 107.2, 107.3, 123.8,
123.9, 124.6, 142.7, 142.8, 142.8, 142.9, 152.9, 153.0, 153.0,
153.0, 165.2, 165.3, 165.8; IR mmax (KBr) 412, 443, 532, 678, 759,
863, 933, 1002, 1126, 1176, 1218, 1338, 1415, 1461, 1504, 1589,
1727, 2838, 2942, 2996, 3108 cmꢀ1; HRMS (FAB) m/z: found
941.3076 [M+H+]; C46H53O2+1 requires for 941.3079.
3672, 3749 cmꢀ1
.
4.1.18. 6-O-(30,40,50-Trimethoxy benzoyl)-1,5-anhydro-
D-glucitol
23
1,5-Anhydro-D-glucitol (1.0 g, 6.1 mmol) and 3,4,5-trimethoxy-
benzoyl chloride (1.4 g, 6.1 mmol) were dissolved in pyridine
(30 mL), and the mixture was stirred for 3 h at 0 °C. The reaction
was quenched by addition of water (150 mL), and diluted with
AcOEt (200 mL). The aqueous phase was extracted by AcOEt
(200 mL). The organic phase was combined and washed with 0.5 N
HCl (100 mL) and water (200 mL), and dried (MgSO4). Filtration
and concentration under reduced pressure gave a residue which
4.2. Ceramidase inhibition assay
Ceramidase was isolated and purified from rat brain as
described.26 Briefly, male Wistar rat brains (16 week), obtained
from Japan SLC, Inc. (Hamamatsu, Japan) were homogenized in
buffer (20 mM phosphate buffer containing 0.25 M sucrose,
1 mM EDTA, and 1 mM AEBSF, pH 7.4) using a Dounce homoge-
was purified by silica gel column chromatography (1:1 Hexane/
AcOEt) to afford compound 23 (1.27 g, 58%), colorless oil [
18
a]
=
D

A. Kamori et al. / Tetrahedron: Asymmetry 27 (2016) 1177–1185
1185
nizer. The homogenate was centrifuged at 1000ꢁg for 10 min, and
Acknowledgments
the supernatant fractions were thereafter centrifuged at 25,000ꢁg
for 10 min. The supernatant fractions were furthermore cen-
trifuged at 100,000ꢁg for 60 min and then the pellet was used as
a source for ceramidase. All the steps were carried out at 4 °C.
For ceramidase activities, the reaction mixture contained 5 mg/
mL rat brain ceramidase in lysis buffer (25 mM Tris/HCl, pH 8.0,
We wish to thank Dr. Ryuta Miyatake (University of Toyama)
for technical support in instrumental analysis.
References
1. (a) Coderch, L.; Lopez, O.; de la Maza, A.; Parra, J. L. Am. J. Clin. Dermatol. 2003, 4,
107–129; (b) Mutanu Jungersted, J.; Hellgren, L. I.; Hogh, J. K.; Drachmann, T.;
Jemec, G. B.; Agner, T. Acta Derm. Venereol. 2010, 90, 350–353; (c) Meckfessel,
M. H.; Brandt, S. J. Am. Acad. Dermatol. 2014, 71, 177–184.
2. (a) Kuempel, D.; Swartzendruber, D. C.; Squier, C. A.; Wertz, P. W. Biochim.
Biophys. Acta 1998, 1372, 135–140; (b) Wertz, P. W. Acta Derm. Venereol. Suppl.
2000, 208, 7–11; (c) Feingold, K. R.; Elias, P. M. Biochim. Biophys. Acta 2014,
1841, 280–294; (d) Sahle, F. F.; Gebre-Mariam, T.; Dobner, B.; Wohlrab, J.;
Neubert, R. H. Skin Pharmacol. Physiol. 2015, 28, 42–55.
3. (a) Elias, P. M. J. Control. Release 1991, 15, 199–208; (b) Wertz, P. W.; van den
Bergh, B. Chem. Phys. Lipids 1998, 91, 85–96; (c) Wertz, P. W. In Skin Barrier;
Elias, P., Feingold, K., Eds.; Taylor & Francis: New York, 2006; p 33.
4. Zheng, W.; Kollmeyer, J.; Symolon, H.; Momin, A.; Munter, E.; Wang, E.; Kelly,
S.; Allegood, J. C.; Liu, Y.; Peng, Q.; Ramaraju, H.; Sullards, M. C.; Cabot, M.;
Merrill, A. H., Jr. Biochim. Biophys. Acta 2006, 1758, 1864–1884.
5. Mullen, T. D.; Hannun, Y. A.; Obeid, L. M. Biochem. J. 2012, 441, 789–802.
6. Holleran, W. M.; Williams, M. L.; Gao, W. N.; Elias, P. M. J. Lipid Res. 1990, 31,
1655–1661.
0.5% Triton X-100) and 30 lM NBD-ceramide (Sigma–Aldrich Co;
St. Louis, Mo, USA), and the incubations were performed for 15 h
at 37 °C with or without acertannins. The reaction was quenched
by addition of methanol. The resulting reaction mixture was dried
under nitrogen and re-dissolved in 20 lL of methanol. The samples
were separated on a TLC silica gel-60 plate (Merck Millipore) using
chloroform:methanol:25% ammonia solution (90:20:0.5) and
their fluorescence were detected with LAS4000 (Fujifilm, Tokyo,
Japan). Ceramidase activity was determined by fluorescence of
NBD-hexanoic acid from NBD-ceramide.
4.3. Chelating ability test
The reaction mixture (4 mL) contained 0.25 mM ZnCl2 in NH4Cl-
7. (a) Spiegel, S.; Milstien, S. J. Biol. Chem. 2002, 277, 25851–25854; (b) Mao, C.;
Obeid, L. M. Biochim. Biophys. Acta 2008, 1781, 424–434; (c) Maceyka, M.;
Harikumar, K. B.; Milstien, S.; Spiegel, S. Trends Cell Biol. 2012, 22, 50–60.
8. Imokawa, G. J. Dermatol. Sci. 2009, 55, 1–9.
9. (a) Sugita, M.; Willians, M.; Dulaney, J. T.; Moser, H. W. Biochim. Biophys. Acta
1975, 398, 125–131; (b) Houben, E.; Uchida, Y.; Nieuwenhuizen, W. F.; De
Paepe, K.; Vanhaecke, T.; Holleran, W. M.; Rogiers, V. Skin Pharmacol. Physiol.
2007, 20, 187–194.
NH3 buffer and Eriochrome black T (EBT) indicator (10
allowed to stand at room temperature with or without 10 mM
acertannins (200 L). The reversible effect of saturated ZnCl2 on
the CDase inhibition was studied by adding 1 mM ZnCl2 with
2 mM EDTA or acertannins to the standard CDase assay method.
lL), and
l
10. Grijalvo, S.; Bedia, C.; Triola, G.; Casas, J.; Llebaria, A.; Teixidó, J.; Rabal, O.;
Levade, T.; Delgado, A.; Fabriàs, G. Chem. Phys. Lipids 2006, 144, 69–84.
11. Bielawska, A.; Greenberg, M. S.; Perry, D.; Jayadev, S.; Shayman, J. A.; McKay, C.;
Hannun, Y. A. J. Biol. Chem. 1996, 271, 12646–12654.
12. (a) Bhabak, K. P.; Kleuser, B.; Huwiler, A.; Arenz, C. Bioorg. Med. Chem. 2013, 21,
874–882; (b) Saied, E. M.; Arenz, C. Cell Physiol. Biochem. 2014, 34, 197–212.
13. Draper, J. M.; Xia, Z.; Smith, R. A.; Zhuang, Y.; Wang, W.; Smith, C. D. Mol. Cancer
Ther. 2011, 10, 2052–2061.
14. Realini, N.; Solorzano, C.; Pagliuca, C.; Pizzirani, D.; Armirotti, A.; Luciani, R.;
Costi, M. P.; Bandiera, T.; Piomelli, D. Sci. Rep. 2013, 3, 1035.
15. Pizzirani, D.; Pagliuca, C.; Realini, N.; Branduardi, D.; Bottegoni, G.; Mor, M.;
Bertozzi, F.; Scarpelli, R.; Piomelli, D.; Bandiera, T. J. Med. Chem. 2013, 56, 3518–
3530.
16. Bi, W.; Gao, Y.; Shen, J.; He, C.; Liu, H.; Peng, Y.; Zhang, C.; Xiao, P. J.
Ethnopharmacol. 2016, 189, 31–60.
17. Honma, A.; Koyama, T.; Yazawa, K. J. Enzyme Inhib. Med. Chem. 2011, 26, 176–
180.
18. Gonzalez-Sarrias, A.; Yuan, T.; Seeram, N. P. Food Chem. Toxicol. 2012, 50, 1369–
1376.
19. (a) Bock, K.; laCour, N. F.; Jensen, S. R.; Nielsen, B. J. Phytochemistry 1980, 19,
2033; (b) Song, C.-Q.; Zhang, N.; Xu, R.-S.; Song, G.-Q.; Sheng, Y. Acta Chim.
Sinica 1982, 40, 1142–1147; (c) Wan, C.; Yuan, T.; Li, L.; Kandhi, V.; Cech, N. B.;
Xie, M.; Seeram, N. P. Bioorg. Med. Chem. Lett. 2012, 22, 597–600; (d) Yuan, T.;
Wan, C.; Liu, K.; Seeram, N. P. Tetrahedron 2012, 68, 959–964; (e) Kutani, N.
Chem. Pharm. Bull. 1960, 8, 72–76.
20. (a) Bennek, J. A.; Gray, G. R. J. Org. Chem. 1987, 52, 892–897; (b) Yuan, C.;
Hollingsworth, R. I. Tetrahedron Lett. 2011, 52, 5421–5423; (c) Ruttens, B.;
Blom, P.; van Hoof, S.; Hubrecht, I.; van der Eycken, J.; Sas, B.; van hemel, J.;
Vandenkerckhove, J. J. Org. Chem. 2007, 72, 5514–5522; (d) Nie, X.; Wang, G. J.
Org. Chem. 2005, 70, 8687–8692.
4.4. Ceramidase synthase assay
A human keratinocyte cell line PHK16-0b were cultured with
keratinocyte growth medium (Lonza, Basel, Switzerland) in the
presence or absence of 50 lM acertannins 4, 7, 10, 11, and 13–
25, and then labelled with NBD-hexanoic acid (Sigma–Aldrich Co,
St. Louis, Mo, USA) for 24 h in a water-jacket incubator at 37 °C
under 5% CO2. The cells were harvested, washed three times with
phosphate-buffered saline and then homogenized for 10 min
under sonication. The protein concentration was determined using
a BCA Protein Assay kit (Thermo Fisher Scientific K.K., Kanagawa,
Japan) with bovine serum albumin as the standard. NBD-labeled
ceramide was extracted by the Bligh & Dyer extraction method.27
The pooled extracts were dried under nitrogen and re-dissolved
in 20 lL of methanol. The samples were separated on a TLC silica
gel-60 plate (Merck Millipore) using chloroform:methanol:25%
ammonia solution (90:20:0.5) and their fluorescence was detected
with LAS4000 (Fujifilm, Tokyo, Japan).
4.5. Quantitative PCR analysis
Total RNAs were isolated from PHK16-0b cells using RNeasy
plus mini kit (Qiagen; Hilden, Germany) according to the manufac-
turer’s instructions. Quantitative RT-PCR with Mx3000P (Agilent;
Santa Clara, California, USA) was performed by using One step
SYBR PrimeScript PLUS RT-PCR kit (Takara-Bio; Shiga, Japan). The
21. Kondo, Y. Agric. Biol. Chem. 1977, 41, 2481–2482.
22. Inoue, T.; Okino, N.; Kakuta, Y.; Hijikata, A.; Okano, H.; Goda, H. M.; Tani, M.;
Sueyoshi, N.; Kambayashi, K.; Matsumura, H.; Kai, Y.; Ito, M. J. Biol. Chem. 2009,
284, 9566–9577.
23. (a) White-Chu, E. F.; Reddy, M. Clin. Dermatol. 2011, 29, 37–42; (b) Hayama, K.;
Takano, Y.; Tamura, J.; Tagami, H.; Terui, T. Australas. J. Dermatol. 2015, 56, 36–
39.
24. Takenouchi, M.; Suzuki, H.; Tagami, H. J. Invest. Dermatol. 1986, 87, 574–576.
25. Levy, M.; Futerman, A. H. IUBMB Life 2010, 62, 347–356.
26. (a) El Bawab, S.; Bielawska, A.; Hannun, Y. A. J. Biol. Chem. 1999, 274, 27948–
27955; (b) Thayyullathil, F.; Chathoth, S.; Hago, A.; Patel, M.; Szulc, Z. M.;
Hannun, Y.; Galadari, S. Biochim. Biophys. Acta 1811, 2011, 242–252.
27. Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37, 911–917.
relative amount of target gene was calculated using the 2ꢀ
DDCt
method with glyceraldehyde 3-phosphate dehydrogenase
(GAPDH) as an internal control.
Primers used in PCR reaction are as follows:
CERS3
50-TCTTGCAGGTCCTTCACCTT-30
50-CCTCGTCATCACTCCTCACA-30
(sense)
(anti-sense)
GAPDH
50-CATGAGAAGTATGACAACAGCCT-30
50-AGTCCTTCCACGATACCAAAGT-30
(sense)
(anti-sense)