Bioactive furanocoumarins from stems of Clausena lansium

Article abstract of DOI:10.1016/j.phytochem.2014.08.002

Clausena lansium Skeels, a member of the Rutaceae, is a shrub or small tree with grapelike fruits. Several parts of this plant have been used in folk medicine. A bioactive constituent investigation of the stems of C. lansium herein resulted in isolation o

Full text of DOI:10.1016/j.phytochem.2014.08.002

Phytochemistry 107 (2014) 141–147
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Bioactive furanocoumarins from stems of Clausena lansium
⇑
Hang Liu, Fei Li, Chuang-Jun Li, Jing-Zhi Yang, Li Li, Nai-Hong Chen, Dong-Ming Zhang
State Key Laboratory of Bioactive Substance and Function of Natural Medicines, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical
College, Beijing 100050, People’s Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 March 2013
Received in revised form 28 May 2014
Available online 26 August 2014
Clausena lansium Skeels, a member of the Rutaceae, is a shrub or small tree with grapelike fruits. Several
parts of this plant have been used in folk medicine. A bioactive constituent investigation of the stems of C.
lansium herein resulted in isolation of four furanocoumarins, claucoumarins A–D, and 13 known analogs.
Their structures were elucidated on the basis of spectroscopic analyses, including UV, IR, MS, and NMR
experiments, and their absolute configurations were determined by CD experiments. Using an in vitro
system, several of these compounds showed selective neuroprotective effects at a concentration of
Keywords:
Clausena lansium
Rutaceae
Claucoumarins A–D
CD experiment
Neuroprotective activity
10 lM.
Ó 2014 Elsevier Ltd. All rights reserved.
1
. Introduction
Clausena lansium Skeels (Wampee), a fruit tree belonging to the
2. Results and discussion
Successive chromatographic purification of the 95% ethanol
extract of the stems of C. lansium on a silica gel column and pre-
parative high performance liquid chromatography afforded four
new (1–4) and thirteen known furanocoumarins (5–17) (Fig. 1).
Rutaceae, is widely distributed in southern China, whose edible
tropical fruit (Chinese Wampee), is locally called ‘huang pi guo’.
Several parts of the plant have also been used in traditional medi-
cine. For example, its leaves have been used for treatment of
coughs, asthma, and gastro-intestinal diseases, and its seeds for
acute and chronic gastro-intestinal inflammation and ulcers; the
fruit also has stomachic and cooling effects and is used ethnomedi-
cally as a vermifuge and for digestive disorders (Adebajo et al.,
2
0
Claucoumarin A (1) was obtained as a pale yellow oil, ½
a
ꢀ
+2.1
D
(c 0.1, MeOH). Its molecular formula, C21 22NaO , was established
H
6
+
by HRESIMS (393.1318 [M + Na] , calcd for 393.1309), implying 11
degrees of unsaturation. The IR spectrum displayed characteristic
absorptions of hydroxyl (3396 cm ), lactone (1725 cm ), and
aromatic ring (1622 and 1586 cm ) groups, and its UV spectrum
showed absorbances at kmax 200, 218, 248, and 303 nm. Its 1
NMR spectrum had two distinct pairs of doublets at d 6.40, 8.10
(each 1H, J = 10.0 Hz) assignable to H-3, H-4 on a coumarin nucleus
ꢁ
1
1
ꢁ1
ꢁ
2
009). Coumarins, amides, and carbazole alkaloids are major con-
stituents of C. lansium (Maneerat et al., 2010; Yang et al., 1988;
Li et al., 1991) and some display interesting biological activities,
including anti-trichomonal, antidiabetic, anti-inflammatory, hepa-
toprotective, and antioxidant effects (Adebajo et al., 2009). In a
previous study of a 95% ethanol extract of the stems of C. lansium,
H
H
0
0
and d
H
8.08, 7.06 (each 1H, J = 1.5 Hz) assignable to H-2 , H-3 on a
7.64 (H-5). The above
furan ring, as well as a lone singlet at d
H
1
7 carbazole alkaloids were found (Liu et al., 2012). In a continuing
information coupled with biogenetic considerations and literature
references (Adams et al., 2006; Maneerat et al., 2010) indicated the
presence of a furanocoumarin skeleton. Analysis of the 13C NMR
spectra, including HSQC, suggested a furanocoumarin system with
an additional C10 moiety, which consisted of two vinyl methyl
groups, one oxymethine, one methylene, two trisubstituted double
bonds, and two oxymethylene groups, respectively. Correlations in
search for bioactive metabolites from the same extract, four new
furanocoumarins, claucoumarins A–D (1–4), and 13 known analogs
(5–17) were obtained. This report describes the isolation and struc-
tural elucidation of the new compounds, and the determination of
their absolute configurations. Although compounds 5–8 were
known compounds, their absolute configurations were first deter-
mined by CD experiments. The neuroprotective activities of 1–17
were also evaluated.
1
1
00
the H– H COSY spectra of 1 indicated the presence of CH
2
(4 )-
0
0
00
00
CHOH (5 )-CH (6 ) [d
1H, dd, J = 13.5, 5.5 Hz, H-4 b), 4.36 (1H, m, H-5 ), 5.06 (1H, d,
H
2.11 (1H, dd, J = 13.5, 7.5 Hz, H-4 a), 1.99
0
0
00
(
0
0
00
J = 8.5 Hz, H-6 ), and 4.41 (1H, d, J = 4.5 Hz, 5 -OH)] and CH
(
2
OH
0
0
00
8 ) [d
H
3.93 (1H, dd, J = 12.5, 5.0 Hz, H-8 a), 3.82 (1H, dd,
⇑
Corresponding author. Tel./fax: +86 10 63165227.
E-mail address: zhangdm@imm.ac.cn (D.-M. Zhang).
0
0
00
J = 12.5, 5.0 Hz, H-8 b), and 4.53 (1H, t, J = 5.0 Hz, 8 -OH)] units.
http://dx.doi.org/10.1016/j.phytochem.2014.08.002
031-9422/Ó 2014 Elsevier Ltd. All rights reserved.
0

1
42
H. Liu et al. / Phytochemistry 107 (2014) 141–147
R
1
5
4
6
3
'
3
2
4
8
a
a
2
'
O 7
8
O
O
R
2
8
''
9''
OH CH OH
2
1
1
''
O
1
2
3
4
5
R =H R2=
9
R =H R2=
1
1
O
O
O
O
3''
5''
7''
1
0''
10''
9''
7''
''
H
8''
5
''
R
1
=H
R
2
=
10 R =H
R2=
3''
O
O
1
O
O
O
4''
1
''
OH
OH
R
1
=H
R
2
=
3
''
11 R =H R2=
1
5''
O
OGlu
9''
8''
7''
1
1''
1''
5''
OCH3
12 R1=
R =H R2=
3''
R =H
2
1
O
O
O
1
0''
10''
9
''
7''
8''
1''
H
5
''
R
1
=
R =H 13 R =OH R =
2
1
2
O
3''
O
O
1
4
R =H R =OGlu
1 2
9
''
8''
OH
1
''
7
''
6
7
R =H R2=
5''
15 R =H R2=
1
1
O
O
O
3''
O
O
1
0''
O
16
R
=H
R
2
=OH
9
''
8''
1
1
''
OH
OH
6
''
R =H R2=
3''
17 R =H R2=
1
1
1
0''
OH
OGlu
9
''
8''
1
''
6''
8
R =H R2=
3''
1
10''
OH
Fig. 1. Compounds 1–17 isolated from C. lansium.
0
0
00
00
In the HMBC experiment, a correlation of H-1 (d
1
H
4.91)/C-8 (d
C
spectrum of 2, correlations for H-5 (d
H
4.54)/C-7 (d
C
32.8) and
0
0
00
00
00
30.5) indicated the C-10 side-chain to be at C-8. Taking the chem-
ical shifts and the degree of unsaturation into consideration, a ser-
C-8 (d
C
179.4); H-6 (d
H
1.84 and 1.70)/C-5 (d
C
75.7), C-7 (d
C
0
0
00
00
32.8), and C-8 (d
C
179.4); and H-7 (d
H
2.64)/C-6 (d
C
33.6) and
0
0
00
00
00
00
ies of HMBC correlations from H-1 (d
H
4.91) to C-2 (d
1.65) to C-2 (d
C
121.4) and
C-8 (d
C
179.4) suggested that C-5 was connected to the C-8 ester
carbonyl to form a five-membered lactone ring. HMBC correlations
0
0
00
00
00
C-3 (d
C
139.5), from H-9 (d
H
C
121.4), C-3 (d
C
00
00
00
00
00
00
1
39.5) and C-4 (d
C
47.8), from H-10 (d
135.7), and from H-8 (d
H
1.63) to C-6 (d
C
130.3)
were observed: H-1 (d
H
4.95)/C-2 (d
122.5), C-3 (d
C
122.5) and C-3 (d
138.3), and C-4 (d
122.5), C-3 (d
C
138.3);
0
0
00
00
00
00
00
00
and C-7 (d
C
0
H
3.93 and 3.82) to C-6 (d
20.9) suggested that the C10
)–CH –CH(OH)–CH@CCH (CH2-
OH)]. In the NOESY experiment, correlations of H-1 /H-9 and H-
/H-10 confirmed a 2 E, 6 Z-configuration for the two double
bonds; a correlation between H-4 and H-5 also supported the C-
0 side-chain being at C-8. Thus, the structure of 1 was established.
On the basis of the bulkiness rule for secondary alcohols (Frelek
and Szczepek, 1999), a negative Cotton effect at 350 nm in the
2 3 4
(OCOCF ) -induced CD spectrum indicated a 5 R configuration
C
H-9 (d
H
00
1.63)/C-2 (d
C
C
C
43.8);
0
00
00
00
00
1
30.3), C-7 (d
C
135.7), and C-10 (d
C
and H-4 (d
(d 75.7), C-6 (d
the C10 side-chain of 2 was as shown.
In the NOESY experiment, the correlation of H-1 /H-9 indi-
H
2.34 and 2.22)/C-2 (d
C
C
138.3), C-5
0
0
00
2
side-chain is [–OCH –CH@C(CH
3
2
3
C
C C
33.6), and C-9 (d 16.3). This indicated that
0
0
00
0
0
00
00
00
00
00
6
0
0
cated a 2 E-configuration for the double bonds. A NOE correlation
0
0
00
00
00
1
between H-5 and H-10 established that H-5 and the C-7 methyl
group were on same side of the five-membered lactone ring. The
absolute configurations of 2 were established according to the
reported sector rule (Minkin et al., 1977). The positive Cotton effect
at 219 nm in the CD spectrum indicated that the absolute configu-
00
Rh
of 1 (Fig. 2). Thus, claucoumarin A was characterized as 1.
20
00
00
Claucoumarin B (2) was isolated as a colorless oil, ½
aꢀ
ꢁ11.2
rations were 5 S, 7 R (Fig. 3). Hence, structure 2 was established
D
+
(
c 0.06, MeOH). Its HRESIMS at m/z 369.1346 [M + H] established
a molecular formula of C21 . Its IR spectrum showed absorp-
tions for six-membered lactone (1730 cm ), a five-membered lac-
and named claucoumarin B.
H
20
O
6
Claucoumarin C (3) had a molecular formula of C22
established by the HRESIMS ion at m/z 489.1377 [M + Na] (calcd
26
H O11, as
ꢁ1
+
ꢁ1
ꢁ1
1
13
tone (1768 cm ), and an aromatic ring (1622 and 1586 cm ). The
D NMR spectra of 2 displayed an 8-oxygenated linear furano-
coumarin skeleton with an additional C10 moiety. In the HMBC
for C22
data indicated that 3 was an 8-oxygenated linear furanocoumarin
skeleton with an additional C moiety and a glucose residue. In
H26NaO11, 489.1367). Its H NMR and C NMR spectroscopic
1
5

H. Liu et al. / Phytochemistry 107 (2014) 141–147
143
2 3 4 2 2
Fig. 2. The Rh (OCOCF ) induced CD spectrum of compound 1 in CH Cl (left), and the bulkiness rule for secondary alcohols applied (right).
Fig. 3. CD Spectrum of compound 2 in CH
3
OH (left) and projection of the sector rule for compound 2 (right).
0
0
00
the HMBC spectrum of 3, correlations for H-4 (d
H
00
1.25) and H-5
interference from the hydroxy groups of the glucopyranosyl moi-
ety, compound 3 was treated with 1 M HCl, and its aglycone 3a
was obtained from acid hydrolysis. As shown in Fig. 4, a positive
Cotton effect at 297 nm in the induced CD spectrum was an indica-
00
00
00
(
(
[
d
H
1.23)/C-2 (d
C
85.9) and C-3 (d
71.2), and C-8 (d
] for the C
C
71.2); H-1 (d
131.2) showed the structure
5
side-chain at C-8. The HMBC
H
4.65)/C-2
00
d
C
85.9), C-3 (d
C
C
–OCH
2
–CH–COH (CH
3
)
2
0
0
00
correlation of the glucose anomeric proton d
the sugar residue at C-2 . Thus, the structure of 3 was established.
Its overall structure was the same as that of dahurin B (17)
Zhao et al., 2007). However, the main different signals were
observed for the glucose anomeric proton and carbon in the 1D
NMR spectra (Supplementary, S36 and S37). The absolute configu-
H
4.43/C-2 placed
tion of a 2 S configuration for 3a by means of the empirical helicity
00
rule. The large coupling constant (7.5 Hz) of the anomeric proton
suggested that the glucose was in a b-configuration, and its D-con-
figuration was established by GC analysis. Therefore, claucoumarin
C was characterized as 3.
The elemental composition of claucoumarin D (4) was estab-
24 5
lished as C22H O by HRESIMS. Based on analysis of the IR, UV,
(
00
ration of 3 and 17 at C-2 may thus be opposite. The absolute
00
00
configuration of the 2 ,3 -diol unit in 3 was determined using
an in situ dimolybdenum CD method developed by Snatzke and
Frelek (Di Bari et al., 2001; Frelek et al., 1999). In order to avoid
and 1D NMR spectra, 4 was also an 8-oxygenated linear furano-
coumarin, but the side-chain was a C11 moiety. In its HMBC spec-
0
0
00
H C
trum, correlations were observed: H-1 (d 4.94)/C-2 (d 119.9)
induced CD spectrum of compound 3a in DMSO (left), and conformations of the Mo4
+
complexes of compound 3a (right).
Fig. 4. The Mo
2
(OAc)
4
2

1
44
H. Liu et al. / Phytochemistry 107 (2014) 141–147
0
0
00
00
00
and C-3 (d
C
141.4); H-9 (d
H
1.55)/C-2 (d
C
119.9), C-3 (d
C
141.4),
In the Rh
2
(OCOCF
3
)
4
-induced CD spectrum (Frelek and Szczepek,
0
0
0000
00
00
00
and C-4 (d
C
41.5); H-4 (d
H
2.65)/C-2 (d
C
00
119.9), C-3 (d
C
141.4),
1999), a positive Cotton effect at 350 nm suggested a 5 S configura-
tion for 6; a positive Cotton effect at 350 nm indicated a 6 S
configuration for 8. Thus, 6 and 8 were determined to be (S,E,E)-8-
[(5-hydroxy-3,7-dimethylocta-2,6-dien-1-yl)oxy]-psoralen
(S,E,E)-8-[(6-hydroxy-3,7-dimethylocta-2,7-dien-1-yl)oxy]-psora-
len, respectively. In the CD spectrum of in situ-formed complexes
0
0
00
00
00
C-5 (d
.28)/C-4 (d
1.09), and H-11 (d
their chemical shifts, verified the presence of the C11 side-chain
C
126.5), and C-6 (d
C
00
137.4); H-5 (d
H
5.28) and H-6 (d
1.09), H-10
74.0), in combination with
H
00
00
00
5
C
41.5) and C-7 (d
C
74.0); and H-8 (d
H
00
00
(d
H
H
2.92)/C-7 (d
C
and
[
–OCH
2
–CH@C(CH
3
)–CH
2
–CH@CH–COCH
3
(CH
3
)
2
]. In the NOESY
00
00
00
experiment, correlation of H-1 /H-9 suggested a 2 E-configura-
tion for the double bonds. Meanwhile, the 5 E-configuration was
determined from the coupling constant between H-5 and H-6
J = 16.0 Hz) in the H NMR spectrum of 4 (performed in CDCl
with Mo
2
(OAc)
4
(Di Bari et al., 2001; Frelek et al., 1999), a positive
00
00
Cotton effect at 314 nm was observed, indicating a 6 S configuration
for 7, i.e. (S,E,E)-8-[(6,7-dihydroxy-3,7-dimethyloct-2-en-1-
yl)oxy)]-psoralen.
0
0
00
1
(
3
).
Thus, claucoumarin D was characterized as 4.
The other nine known furanocoumarins, imperatorin (9)
(Harkar et al., 1984), wampetin (10) (Khan et al., 1983), (E,E)-8-
Compound 5 was assigned the same molecular formula as 2 by
HRESIMS. Its 1D NMR spectra suggested it had the same C10 side-
(7-hydroxy-3,7-dimethylocta-2,5-dienyloxy)
psoralen
(11)
00
chain as 2. In its HMBC spectrum, correlation of H-1 /C-5 located
the C10 side-chain at C-5. Compound 5 was a 5-oxygenated linear
furanocoumarin.
(Adams et al., 2006), anisolactone (12) (Lakshmi et al., 1984), allo-
isoimperatorin (13) (Razdan et al., 1982), xanthotoxol-8-O-b-D-
glucopyranoside (14) (Jia et al., 2008), 8-isopentenoyloxypsoralen
(15) (Georgievskii and Rybachenko, 1985), 8-hydroxyfuranocoum-
arin (16) (Harkar et al., 1984), and dahurin B (17) (Zhao et al., 2007)
were also identified on the basis of their spectroscopic profiles
(NMR, UV, and MS) and comparison to the published data. Detailed
The overall structure was the same as lucidafuranocoumarin B,
the literature reference of which only reported its relative configu-
00
00
ration, and absolute configurations at C-5 and C-7 were pre-
sumed (Sripisut et al., 2012). In the NOE difference spectrum of
0
0
00
1
13
5
, irradiation of the 10 -CH
3
gave an enhancement of H-6 b, while
assignments of the H and C NMR spectroscopic data for com-
pounds 1–4 are shown in Tables 1 and 2.
0
0
00
H-6 b was also enhanced when H-5 was irradiated. This suggested
00
00
00
that the 10 -CH
membered lactone ring. According to the reported sector rule
Minkin et al., 1977) and the positive Cotton effect at 241 nm in
the CD spectrum (Supplementary, S32), this indicated that the
3
, H-5 , and H-6 b were on same side of the five-
Compounds 1–17 were evaluated for their neuroprotective
effect on neuron-like PC12 cells induced by serum deprivation,
Ab25–35, and sodium nitroprusside (SNP) (Pytlowany et al., 2008)
in vitro using the MTT method. In the aforementioned two models,
the neuron growth factor (NGF) was used as a positive control,
while edaravone, a reactive oxygen species scavenger, was used
as an anti-oxidant positive control in SNP induced model. At
(
00
00
00
00
absolute configurations at C-5 and C-7 were 5 S, 7 R. Hence,
compound 5 was named 5-{[(E)-3-methyl-4-((2S,4R)-4-methyl-5-
oxotetrahydrofuran-2-yl)but-2-en-1-yl]oxy}-psoralen.
Compounds 6, 7, and 8 were assigned as 8-oxygenated linear fur-
anocoumarin derivatives because they possessed characteristic UV
and IR spectroscopic data as mentioned for 1. The chemical and
physical properties (optical rotation, MS, and NMR) were in good
agreement with those of reported compounds clausenalansimin A
10 lM, 7 protected cells against serum deprivation induced PC12
cell damage; 4 and 8 increased the cell survival rate on Ab25–35
treated group, while 4, 8, 9, and 16 blocked the toxicity induced
by SNP; 15 exacerbated the cytotoxicity of the three models; the
other compounds were inactive (Table 3).
(
Maneerat et al., 2010), (E)-9-((6,7-dihydroxy-3,7-dimethyloct-2-
en-1-yl)oxy)-7H-furo[3,2-g]chromen-7-one, and lansiumarin-C
Adams et al., 2006). However, the absolute configurations of the
3
. Conclusions
(
side-chains were not determined in the previous report. CD experi-
ments were thus employed to establish the absolute configurations.
Seventeen furanocoumarins, including four new compounds,
were obtained from stems of Clausena lansium, increasing the
Table 1
1
a
H
N
M
R
s
p
e
c
t
r
o
s
c
o
p
i
c
d
a
t
a
o
f
c
o
m
p
o
u
n
d
s
1
–
4
.
Position
1
2
3
4
3
4
5
6.40 d (10.0)
8.10 d (10.0)
7.64 s
8.08 d (1.5)
7.06 d (1.5)
4.91 d (7.0)
5.50 t (7.0)
2.11 dd (13.5, 7.5), 1.99 dd (13.5, 5.5)
4.36 m
5.06 d (8.5)
6.42 d (9.6)
8.13 d (9.6)
7.68 s
8.11 d (2.0)
7.08 d (2.0)
4.95 m
5.58 t (7.2)
2.34 dd (14.0, 7.2), 2.22 dd (14.0, 7.2)
6.41 d (9.5)
8.12 d (9.5)
7.64 s
8.09 d (2.0)
7.08 d (2.0)
4.65 d (5.0)
3.77 t (5.0)
1.25 s
6.41 d (9.6)
8.12 d (9.6)
7.67 s
8.09 d (2.0)
7.07 d (2.0)
4.94 d (7.2)
5.49 m
2.65 d (4.0)
5.28 m
5.28 m
0
2
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
4
5
6
7
8
9
1
1
3
5
8
4.54 m
1.84 m, 1.70 m
2.64 m
1.23 s
3.93 dd (12.5, 5.0), 3.82 (12.5, 5.0)
1.65 s
1.63 s
1.09 s
1.55 s
1.09 s
2.92 s
1.63 s
1.07 d (7.2)
0
0
0
0
0
1
0
0
0
0
0
0
-OH
-OH
-OH
4.50 s
4.41 d (4.5)
4.53 t (5.0)
Glu-1
Glu-2
Glu-3
Glu-4
Glu-5
Glu-6
4.43 d (7.5)
3.00 m
3.06 m
3.14 m
3.14 m
3.33 overlapped
a
1H NMR data (d) were measured in DMSO-d
at 500 MHz for 1 and 3, and at 400 MHz for 2 and 4.
6

H. Liu et al. / Phytochemistry 107 (2014) 141–147
145
Table 2
J-815 spectropolarimeter, and IR spectra were recorded on a Nico-
let 5700 FT-IR spectrometer. NMR measurements were performed
on INOVA-500, Mercury-plus-400 and Bruker AV500-III spectrom-
1
3C NMR spectroscopic data of compounds 1–4a.
Position
1
2
3
4
eters. Chemical shifts are recorded in parts per million (d) DMSO-d
(d 2.491 and d 39.5 ppm). GC data were obtained using an Agilent
7890A instrument. HRESIMS were acquired using an Agilent 1100
series LC/MSD Trap SL mass spectrometer. Preparative HPLC was
performed on a Shimadzu LC-6AD instrument with a SPD-20A
6
2
3
4
159.7
114.0
145.2
116.3
114.0
125.6
147.7
130.5
143.1
147.7
107.0
69.3
159.7
114.1
145.3
116.4
114.4
125.6
147.9
130.2
143.4
147.8
107.0
68.9
122.5
138.3
43.8
75.7
33.6
159.8
114.2
145.2
116.4
113.5
125.9
146.6
131.2
142.2
147.7
107.0
73.8
159.7
114.0
145.3
116.3
114.3
125.6
147.9
130.2
143.4
147.8
107.0
69.0
119.9
141.4
41.5
126.5
137.4
74.0
H
C
4
5
6
7
8
8
2
3
1
2
3
4
5
6
7
8
9
1
1
a
detector, using a YMC-Pack ODS-A column (250 ꢂ 20 mm, 5
Column chromatography (CC) was performed with silica gel
200–300 mesh, Qingdao Marine Chemical Inc., Qingdao, People’s
Republic of China) and ODS (50 m, YMC, Japan), respectively.
lm).
a
(
0
0
l
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
TLC was carried out with glass pre-coated silica gel GF254 plates.
Spots were visualized under UV light or by spraying with 10%
H SO in EtOH followed by heating.
2 4
121.4
139.5
47.8
85.9
71.2
25.8
25.9
64.8
4.2. Plant materials
130.3
135.7
59.7
16.7
20.9
32.8
179.4
16.3
25.6
16.1
25.6
49.4
Stems of C. lansium were collected in Liuzhou, Guangxi, China,
in December 2008, and identified by Engineer Guang-Ri Long, For-
estry of Liuzhou. A voucher specimen is deposited at the Herbar-
ium of Institute of Materia Medica, Chinese Academy of Medical
Sciences & Peking Union Medical College (ID-S-2320).
0
0
0
0
0
1
15.2
Glu-1
Glu-2
Glu-3
Glu-4
Glu-5
Glu-6
105.0
74.1
76.7
69.7
76.7
60.7
4.3. Extraction and isolation
Air-dried, powdered stems of C. lansium (6.4 kg) were macer-
ated for 3 h with EtOH-H O (95:5, v/v, 20 L) and then extracted
2
under conditions of reflux (i.e. using the same solvents and appa-
ratus, twice each for 4 h). The combined extracts were evaporated
under reduced pressure to yield a dark brown residue (640 g). The
latter was suspended in water (2 L) and then partitioned with
a
13C NMR data (d) were measured in DMSO-d
00 MHz for 2 and 4.
6
at 125 MHz for 1 and 3, and at
1
Table 3
Neuroprotective activity of compounds 4, 7–9, 15, and 16 on PC12 cell models
induced by serum deprivation, Ab25–35, and sodium nitroprusside (10 M).
CHCl
The CHCl
eluted with petroleum ether (PE, b.p. 60–90 °C)-acetone (3:1, v/v)
to afford ten fractions A –A10 on the basis of TLC analysis. Fraction
(7.30 g) was applied to a silica gel column (200–300 mesh)
eluted with petroleum ether-acetone gradients (6:1 ? 4:1, v/v)
to give five subfractions, A a–A e. Separation of fraction A
(3.12 g) by MPLC (MeOH–H O, 50:50 ? 100:0, v/v) and preparative
HPLC (detection at 210 nm, 7 mL/min), successively, yielded 4
10 mg) and 15 (4 mg). Fraction A (12.93 g) was subjected to silica
gel CC (200–300 mesh) eluted with petroleum ether-acetone gradi-
ents (5:1 ? 3:1, v/v) to give six subfractions, A a–A f. The couma-
rin-containing fraction A b (4.12 g) was purified by MPLC (MeOH–
O, 35:65 ? 75:25, v/v) followed by preparative HPLC (detection
3
(3 ꢂ 2L), EtOAc (3 ꢂ 2L), and n-BuOH (3 ꢂ 2L), successively.
l
-soluble portion (195 g) was subjected to silica gel CC
3
Groupa
Cell survival rate (%)
Serum
1
Ab25–35
Sodium nitroprusside
M, 48 h) (350
A
2
deprivation (48 h) (15
l
l
M, 24 h)
Control
Model
Positive control 89.4 ± 3.2
4
7
8
9
1
1
100
65.8 ± 3.9
100
100
#
##
**
*
###
###
2
2
2
d
57.2 ± 1.3
76.7 ± 3.1
67.3 ± 1.5
70.0 ± 3.9
69.4 ± 1.9
62.8 ± 2.5
34.9 ± 2.3
60.5 ± 3.6
61.0 ± 2.9
84.3 ± 3.7
80.2 ± 2.2
60.4 ± 7.1
77.7 ± 3.3
76.7 ± 4.2
36.2 ± 0.9
74.9 ± 3.2
*
*
***
*
2
*
*
*
***
64.5 ± 3.3
79.9 ± 2.3
68.4 ± 4.8
63.5 ± 6.6
27.7 ± 2.1
64.2 ± 7.9
*
(
7
**
***
*
**
*
**
**
7
7
5
6
*
*
7
H
2
#
##
p < 0.001 vs control.
p < 0.05.
p < 0.01.
p < 0.001 vs model.
The other compounds were inactive.
*
at 210 nm, 7 mL/min) to yield 2 (11 mg), 5 (1 mg), 6 (7 mg), 8
(11 mg), 9 (27 mg) 10 (23 mg), 11 (15 mg), 12 (13 mg), 13
*
*
*
**
(17 mg), and 16 (15 mg), respectively. Fraction A
applied to a silica gel column (200–300 mesh) eluted with CHCl
MeOH gradients (20:1 ? 10:1, v/v) to give nine subfractions, A a–
i. Fraction A e (3.43 g) was subject to MPLC (MeOH–H O,
5:75 ? 60:40, v/v), successively using preparative HPLC (detec-
9
(29.83 g) was
a
3
–
9
A
9
9
2
known molecular diversity of metabolites from the Rutaceae. Their
structures were elucidated by MS spectrometric and IR, NMR, CD
spectroscopic data. Compounds 5–8 were first determined their
absolute configurations by CD experiments. Using an in vitro sys-
tem, compounds 4, 7–9, and 16 showed selective neuroprotective
effects at a concentration of 10
. Experimental
.1. General
Optical rotations were measured on a JASCO P-2000 automatic
2
tion at 210 nm, 7 mL/min) to yield 1 (13 mg) and 7 (9 mg). The
n-BuOH-soluble portion (61.3 g) was fractionated via silica gel CC
eluting with CHCl
tions B –B18 on the basis of TLC analysis. Fraction B
subject to MPLC (MeOH–H O, 5:95 ? 35:65, v/v), successively
3
–MeOH (9:1 ? 6:1, v/v) to afford eighteen frac-
1
7
(2.15 g) was
lM.
2
using preparative HPLC (detection at 210 nm, 7 mL/min) to yield
3 (2 mg), 14 (6 mg), and 17 (4 mg).
4
4
4.3.1. Claucoumarin A (1)
2
0
Pale yellow oil; ½
) 200 (4.43), 218 (4.31), 248 (3.99), 303 (2.51) nm; IR
2931, 1725, 1622, 1586, 1441, 1401, 1327, 1150, 1094, 1029,
a
ꢀ
+2.1 (c 0.1, MeOH); UV (MeOH) kmax (log
D
e
mmax 3396,
digital polarimeter, while UV spectra were acquired with a Hitachi
UV-240 spectrophotometer. CD spectra were measured on a JASCO
ꢁ1
1
13
1004 cm ; for H NMR (DMSO-d
6
, 500 MHz) and
C NMR

1
46
H. Liu et al. / Phytochemistry 107 (2014) 141–147
(
DMSO-d
6
, 125 MHz) spectroscopic data, see Tables 1 and 2; Rh
-induced CD(CH Cl ) 350 (
ꢁ0.035) nm; HRESIMS m/z
93.1318 [M + Na] (calcd for C21 22NaO , 393.1309).
2
(-
ole was added, followed by heating to dryness at 60 °C for 2 h. The
dried reactant was partitioned between n-hexane and H
(0.2 mL), and the n-hexane fraction was subjected to gas chroma-
tography (GC) (column: DM-5, 0.25 mm ꢂ 30 m ꢂ 25 M; detec-
tor: FID; temperature: 280 °C; injector temperature: 250 °C;
carrier: N gas). The sugar was identified by comparison of the
OCOCF
3
3
)
4
2
2
D
e
2
O
+
H
6
l
4
.3.2. Claucoumarin B (2)
2
0
Colorless oil; ½
aꢀ
ꢁ11.2 (c 0.06, MeOH); UV (MeOH) kmax (log
2
D
e
2
9
) 197 (4.39), 217 (4.29), 246 (3.94), 301 (2.44) nm; IR
m
max 2971,
retention time with the authentic standard [t
glucose].
R
: 18.99 min for D-
936, 1768, 1730, 1622, 1586, 1440, 1401, 1326, 1148, 1094,
96, 927, 874 cm ; for 1H NMR (DMSO-d
ꢁ1
13
C
6
, 400 MHz) and
NMR (DMSO-d
6
, 100 MHz) spectroscopic data, see Tables 1 and
+2.14), 253 ( +0.36),
4
.5. Determination of absolute configuration of the secondary alcohol
2; CD (MeOH) 219 (
D
e
D
e
ꢁ0.22), 294 (
De
+
units in 1, 6, and 8 (Frelek and Szczepek, 1999)
and 334 (
D
e
ꢁ0.11) nm; HRESIMS m/z 369.1346 [M + H] (calcd
for C21
H
21
O
6
, 369.1333).
Following the reported procedure, a 1:2 mixture of secondary
alcohol/Rh (OCOCF for 1, 6, and 8 was subjected to CD measure-
ments at concentration of 0.1 mg/mL in anhydrous CH Cl . The first
CD spectrum was recorded immediately after mixing, and its time
evolution was monitored until stationary (about 10 min after mix-
ing). The inherent CD was subtracted. The observed sign of the
band at around 350 nm in the induced CD spectrum was correlated
to the absolute configuration of the secondary alcohol.
2
3 4
)
4
.3.3. Claucoumarin C (3)
Yellow amorphous powder; ½
MeOH) kmax (log ) 217 (4.28), 246 (3.91), 306 (2.45) nm; IR
398, 2920, 1716, 1622, 1588, 1442, 1403, 1157, 1084, 1033,
20
aꢀ
ꢁ31.2 (c 0.1, MeOH); UV
2
2
D
(
3
8
e
m
max
ꢁ1
1
13
77 cm ; for H NMR (DMSO-d
6
, 500 MHz) and
C NMR
(
DMSO-d
SIMS m/z 489.1377 [M + Na] (calcd for C22
Compound 3a: Mo (OAc) -induced CD (DMSO) 297 (De
6
, 125 MHz) spectroscopic data, see Tables 1 and 2; HRE-
+
H26NaO11, 489.1367).
2
4
+0.47) nm.
4
7
.6. Determination of absolute configuration of the diol unit in 3a and
by Snatzke’s method (Di Bari et al., 2001; Frelek et al., 1999)
4
.3.4. Claucoumarin D (4)
Yellow oil; UV (MeOH) kmax (log
3.97), 303 (2.31) nm; IR max 3134, 2977, 1713, 1623, 1585,
e
) 202 (4.38), 220 (4.31), 248
(
1
m
According to the published procedure, a 1:1.2 mixture of diol/
438, 1400, 1330, 1145, 1083, 989, 941, 877 cm ; for 1H NMR
ꢁ1
2 4
Mo (OAc) for 3a and 7 was subjected to CD measurements at con-
(
DMSO-d
scopic data, see Tables 1 and 2; HRESIMS m/z 391.1553 [M + Na]
calcd for C22 24NaO , 391.1516).
6
, 400 MHz) and 13C NMR (DMSO-d
, 100 MHz) spectro-
6
centration of 0.1 mg/mL in anhydrous DMSO. The first CD spec-
trum was recorded immediately after mixing, and its time
evolution was monitored until stationary (about 10 min after mix-
ing). The inherent CD was subtracted. The observed sign of the
diagnostic band at around 300 nm in the induced CD spectrum
+
(
H
5
4.3.5. 5-{[(E)-3-methyl-4-((2S,4R)-4-methyl-5-oxotetrahydrofuran-2-
yl)but-2-en-1-yl]oxy}-psoralen (5)
0
0
2
0
was correlated to the absolute configuration of the 2 , 3 -diol unit.
Colorless oil; ½
aꢀ
ꢁ28.3 (c 0.05, MeOH); UV (MeOH) kmax (log
D
e) 200 (4.27), 218 (4.35), 303 (2.11) nm; IR
m
max 2976, 2937, 1766,
1
8
731, 1622, 1580, 1456, 1354, 1198 1126, 1073, 1000, 927,
4
.7. Neuroprotection bioassays
ꢁ1
26 cm ; CD (MeOH) 241 (
D
e
+1.48), 325 (
D
e
ꢁ1.21) nm; HRE-
+
SIMS m/z 369.1330 [M + H] (calcd for C21
H
21
O
6
, 369.1333).
Pheochromocytoma (PC12) cells were incubated in DMEM sup-
plied with 5% fetal bovine serum and 5% equine serum as basic
medium. PC12 cells in logarithmic phase were cultured at a density
of 5000 cells per well in a 96-well microtiter plate. After 24 h incu-
bation, the medium of model group was changed to DMEM or basic
4
.3.6. (S,E,E)-8-[(5-hydroxy-3,7-dimethylocta-2,6-dien-1-yl)oxy]-
psoralen (6)
Yellow oil; ½
CD(CH Cl ) 350 (
2
0
a
ꢀ
+30.8 (c 0.08, MeOH);Rh
+0.11) nm; ESIMS m/z 355.1 [M + H] .
2
(OCOCF
3 4
) -induced
D
+
2
2
D
e
medium with 15
50 M SNP for 24 h. Tested compounds dissolved in dimethyl
l
M Ab25–35 for 48 h, or basic medium with
3
l
4
.3.7. (S,E,E)-8-[(6,7-dihydroxy-3,7-dimethyloct-2-en-1-yl)oxy]-
psoralen (7)
Colorless oil; ½
CD (DMSO) 314 (
sulfoxide (DMSO) were added to each well for >1000 fold dilution
in the model medium at the same time. Each sample was tested in
2
0
a
D
ꢀ
ꢁ18.9 (c 0.08, MeOH); Mo
2
(OAc)
4
-induced
D
triplicate. After the incubation at 37 °C in 5% CO for 24 h, 10
l
l of
MTT (5 mg/ml) was added to each well and incubated for another
h, then liquid in the wells was removed. DMSO (100 l) was
2
+
e +0.37) nm; ESIMS m/z 395.1 [M + Na] .
4
l
4
.3.8. (S,E,E)-8-[(6-hydroxy-3,7-dimethylocta-2,7-dien-1-yl)oxy]-
psoralen (8)
Yellow oil; ½
CD(CH Cl ) 350 (
added to each well. The absorbance was recorded on a microplate
reader (Bio-Rad model 550) at a wavelength of 570 nm. Analysis of
variance (ANOVA) followed by Newman–Keuls post hoc test were
performed to assess the differences between the relevant control
and each experimental group. P-values of <0.05, <0.01, and
2
0
a
ꢀ
+14.2 (c 0.07, MeOH); Rh
+0.13) nm; ESIMS m/z 377.1 [M + Na] .
2
(OCOCF
3 4
) -induced
D
+
2
2
D
e
4.4. Acid hydrolysis of compound 3
<0.001 were regarded as statistically significant. Data were
expressed as mean ± SEM as indicated.
Compound 3 (2 mg) was dissolved in 1 M HCl (aq) (5 mL) and
heated at 60 °C for 8 h under constant stirring. The reaction mix-
ture was diluted with H
The EtOAc layers were combined and evaporated to dryness and
then subjected to preparative HPLC using the mobile phase CH3-
2
O and extracted with EtOAc (3 ꢂ 5 mL).
Acknowledgements
We are grateful to the Department of Instrumental Analysis,
Institute of Materia Medica, Chinese Academy of Medical Sciences
and Peking Union Medical College for the measurement of the UV,
IR, CD, NMR, MS, and HRESIMS spectra. This research program is
supported by the National Natural Science Foundation of China
(No. 21272278).
2
CN/H O (21:79) to give the aglycone 3a (0.3 mg). The aqueous lay-
ers were combined, evaporated and cryodesiccated. The residue
was dissolved in anhydrous pyridine (1 mL), and then L-cysteine
methyl ester hydrochloride (2 mg) was added. The mixture was
stirred at 60 °C for 2 h, and then 0.2 mL of N-trimethylsilylimidaz-

H. Liu et al. / Phytochemistry 107 (2014) 141–147
147
Harkar, S., Razdan, T.K., Waight, E.S., 1984. Steroids, chromone, and coumarins from
Angelica officinalis. Phytochemistry 23, 419–426.
Appendix A. Supplementary data
Jia, X.D., Zhao, X.Z., Feng, X., Wang, M., Sun, H., Dong, Y.F., 2008. Coumarins of
Angelica dahurica var. formosana (II). Chin. Tradit. Herb. Drugs 39, 1768–1771.
Khan, N.U., Naqvi, S.W.I., Ishratullah, K., 1983. Wampetin, a furocoumarin from
Clausena wampi. Phytochemistry 22, 2624–2625.
Lakshmi, V., Prakash, D., Raj, K., Kapil, R.S., Popli, S.P., 1984. Monoterpenoid
furanocoumarin lactones from Clausena anisata. Phytochemistry 23, 2629–
2631.
Li, W.S., McChesney, J.D., El-Feraly, F.S., 1991. Carbazole alkaloids from Clausena
lansium. Phytochemistry 30, 343–346.
Liu, H., Li, C.J., Yang, J.Z., Ning, N., Si, Y.K., Li, L., Chen, N.H., Zhao, Q., Zhang, D.M.,
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.2014.
0
8.002.
References
Adams, M., Ettl, S., Kunert, O., Wube, A.A., Haslinger, E., Bucar, F., Bauer, R., 2006.
Antimycobacterial activity of geranylated furocoumarins from Tetradium
daniellii. Planta Med. 72, 1132–1135.
2012. Carbazole alkaloids from the stems of Clausena lansium. J. Nat. Prod. 75,
677–682.
Adebajo, A.C., Iwalewa, E.O., Obuotor, E.M., Ibikunle, G.F., Omisore, N.O., Adewunmi,
C.O., Obaparusi, O.O., Klaes, M., Adetogun, G.E., Schmidt, T.J., Verspohl, E.J., 2009.
Pharmacological properties of the extract and some isolated compounds of
Clausena lansium stem bark: anti-trichomonal, antidiabetic, anti-inflammatory,
hepatoprotective and antioxidant effects. J. Ethnopharmacol. 122, 10–19.
Di Bari, L., Pescitelli, G., Pratelli, C., Pini, D., Salvadori, P., 2001. Determination of
absolute configuration of acyclic 1,2-diols with Mo2(OAc)4.1, Snatzke’s method
revisited. J. Org. Chem. 66, 4819–4825.
Maneerat, W., Prawat, U., Saewan, N., Laphookhieo, S., 2010. New coumarins from
Clausena lansium twigs. J. Braz. Chem. Soc. 21, 665–668.
Minkin, V.I., Legrand, M., Rougier, M.J., 1977. Determination of Configurations by
Dipole Moments, CD or ORD, vol. 2. Georg Thieme, Stuttgart, pp. 135–136.
Pytlowany, M., Strosznajder, J.B., Jesko, H., Cakala, M., Strosznajder, R.P., 2008.
Molecular mechanism of PC12 cell death evoked by sodium nitroprusside, a
nitric oxide donor. Acta Biochim. Pol. 55, 339–347.
Razdan, T.K., Kachroo, V., Harkar, S., Koul, G.L., 1982. Furanocoumarins from
Heracleum canescens. Phytochemistry 21, 923–927.
Frelek, J., Szczepek, W.J., 1999. Rh
chiroptical studies on steroidal alcohols. Tetrahedron Asymmetry 10, 1507–
520.
2 3 4
(OCOCF ) as an auxiliary chromophore in
Sripisut, S., Phakhodee, W., Ritthiwigrom, T., Cheenpracha, S., Laphookhieo, S., 2012.
Lucidafuranocoumarins B and C from the twigs of Feroniella lucida: absolute
configurations of lucidafuranocoumarin C. Phytochem. Lett. 5, 309–312.
Yang, M.H., Chen, Y.Y., Huang, L., 1988. Three novel cyclic amides from Clausena
lansium. Phytochemistry 27, 445–450.
1
Frelek, J., Geiger, M., Voelter, W., 1999. Transition metal complexes as auxiliary
chromophores in chiroptical studies on carbohydrates. Curr. Org. Chem. 3, 117–
1
46.
Georgievskii, V.P., Rybachenko, A.I., 1985. Spectral-luminescence properties of
natural coumarin derivatives and their use for group identification. Chem. Nat.
Compd. 6, 766–769.
Zhao, X., Feng, X., Jia, X., Wang, M., Shan, Y., Dong, Y., 2007. New coumarin glucoside
from Angelica dahurica. Chem. Nat. Compd. 43, 399–401.