Anti-inflammatory and cytotoxic neoflavonoids and benzofurans from Pterocarpus santalinus

Article abstract of DOI:10.1021/np100871g

Five new benzofurans, pterolinuses A-E (1-5), six new neoflavonoids, pterolinuses F-J (8-13), and five known compounds (6, 7, 14-16) were isolated from an extract of Pterocarpus santalinus heartwood. All new structures were elucidated by spectroscopic met

Full text of DOI:10.1021/np100871g

ARTICLE
pubs.acs.org/jnp
Anti-inflammatory and Cytotoxic Neoflavonoids and Benzofurans
from Pterocarpus santalinus
||
Shou-Fang Wu,† Fang-Rong Chang,*,†,‡ Sheng-Yang Wang,§ Tsong-Long Hwang,^ Chia-Lin Lee,†,
||,r
Shu-Li Chen,† Chin-Chung Wu,† and Yang-Chang Wu*,†,
†Graduate Institute of Natural Products, Kaohsiung Medical University, Kaohsiung 807, Taiwan, Republic of China
‡Cancer Center, Kaohsiung Medical University Hospital, Kaohsiung 807, Taiwan, Republic of China
§Department of Forestry, National Chung-Hsing University, Taichung 402, Taiwan, Republic of China
^Graduate Institute of Natural Products, Chang Gung University, Taoyuan 333, Taiwan, Republic of China
Natural Medicinal Products Research Center and Center for Molecular Medicine, China Medical University Hospital, Taichung 40402,
Taiwan, Republic of China
rGraduate Institute of Integrated Medicine, College of Chinese Medicine, China Medical University, Taichung 40402, Taiwan, Republic
of China
S Supporting Information
b
ABSTRACT: Five new benzofurans, pterolinuses AꢀE (1ꢀ5), six new
neoflavonoids, pterolinuses FꢀJ (8ꢀ13), and five known compounds
(6, 7, 14ꢀ16) were isolated from an extract of Pterocarpus santalinus
heartwood. All new structures were elucidated by spectroscopic methods,
and configurations were confirmed by CD spectral data and optical
rotation values. The isolates were evaluated for anti-inflammatory and
cytotoxic activities. Six compounds (1, 2, 4, 6, 7, and 15) showed
significant inhibition in at least one anti-inflammatory assay. Compound
2 showed the best selective effect against superoxide anion generation in
human neutrophils with, an IC50 value of 0.19 μg/mL, and was 6.2-fold more
potent than the positive control LY294002. Compound 14 showed the highest
cytotoxicity against Ca9-22 cancer cells, with an IC50 value of 0.46 μg/mL.
terocarpus santalinus L. (Fabaceae), also named “red sanders”
or “red sandalwood”, is a rare, commercial tree in the
and anti-inflammatory activities of the isolates were also investigated
in this study.
P
Legume family. This species is distributed exclusively in well-
defined forest tracts of Andhra Pradesh in Southern India. It is
valuable in the international market and is most notably exported
from India to Japan and other countries.1 The fragrant red
heartwood is valued for making furniture and also as a source
of coloring and dyeing materials. In Buddhism, it represents
holiness and is said to prevent evil; thus, it is used for carved
statues and as a component of incense. In addition, P. santalinus
has been used as a folk remedy for treatment of inflammation, such
as in chronic bronchitis and chronic cystitis, fever, headaches, mental
aberrations, ulcers, cancer, etc.2,3 In previous phytochemical inves-
tigations, six sesquiterpenes,4 one isoflavone,5 two lignans,2 and two
aurone glycosides6 were isolated from this species. In preliminary
studies, we found that a MeOH extract of the heartwood exhibited
potent cytotoxicity (IC50 < 20 μg/mL) against six cancer cell lines
(HepG2, Hep3B, Ca9-22, A549, MCF7, MDA-MB-231). Subse-
quently, we isolated seven benzofurans (1ꢀ7) and nine neoflavo-
noids (8ꢀ16) from the extract using bioactivity-guided
fractionation. Compounds 1ꢀ5 and 8ꢀ13 are new, and their
structures were determined by spectroscopic methods. Cytotoxic
’ RESULTS AND DISCUSSION
The CH2Cl2-soluble portion of a MeOH extract of P. santalinus
heartwood was subjected to column chromatography (CC)
on silica gel, C18 gel, Sephadex LH-20, and preparative TLC
to yield five new benzofurans (1ꢀ5), six new neoflavonoids
(8ꢀ13), and five known compounds (6, 7, 14ꢀ16). The known
compounds were identified as dehydromelanoxin (6), melanoxin
(7),7 melanoxoin (14), S-30-hydroxy-4,40-dimethoxydalbergione
(15), and melannein (16)8 by comparison of the NMR and MS
data with those in published literature.
Compounds 1 and 6, isolated separately, both had the same
molecular weight. The ion at m/z 300.0999 [M]þ in HREIMS
indicated a molecular formula of C17H16O5 and 10 degrees of
unsaturation. Compounds 1 and 6 had similar 1H and 13C NMR
spectra. The IR spectra of 1 and 6 were also similar, with
absorptions for OH (3443 cmꢀ1) and aromatic (1514 and
Received: November 30, 2010
Published: April 13, 2011
r
Copyright
2011 American Chemical Society and
989
dx.doi.org/10.1021/np100871g J. Nat. Prod. 2011, 74, 989–996
American Society of Pharmacognosy
|

Journal of Natural Products
ARTICLE
previous reference9 as corresponding to C-4 (δH 6.97, δC 105.2)
and C-7 (δH 7.12, δC 96.5). Moreover, these two aromatic
protons showed HMBC correlations with C-5, C-6, C-9 and C-5,
C-6, C-8, C-9, respectively. The HMBC spectrum also showed
correlations of the OCH3 protons at δH 3.91 and 3.90 with C-6
and C-40, which was confirmed by NOESY corrections with H-7
and H-50, respectively. Two OH protons (δH 7.30, 7.75) showed
HMBC correlations with C-4, C-5, C-6 and C-20, C-30, which
indicated that they were attached at C-5 and C-30. Consequently,
compound 1 has an isoparvifuran skeleton,9 and it was named
pterolinus A.
The molecular formula (C16H14O4) of 2 was identified from
the HRESIMS ion at m/z 271.0972 [M þ H]þ, indicating 10
degrees of unsaturation. IR absorptions at 3317(OH) and 1510
(aromatic) cmꢀ1 and UV absorptions at 321 and 284 nm were
characteristic of a 2-methylbenzofuran.9,10 A 14-carbon skeleton
was also proposed on the basis of the 1H and 13C NMR spectra.
In contrast with 1, the proton and carbon chemical shifts for the
methyl group were observed at δH 2.36 s and δC 9.6. One OCH3
group (δH 3.92), two OH groups (δH 7.23 br/8.60 br), and an
AA0BB0 (δH 6.96, d, J = 8.0 Hz/δH 7.63, d, J = 8.0 Hz) system
appeared in the 1H NMR spectrum. The OCH3 was connected
to C-6 on the basis of a HMBC correlation and a NOE
correlation with H-7. Hence, the two OH groups were assigned
at C-5 and C-40. Compound 2 was identified as indicated and has
been named pterolinus B.
Compound 3 showed an [M þ H]þ ion at m/z 317.1387 in
the HRESIMS (molecular formula C18H20O5), and nine degrees
of unsaturation were calculated. The IR spectrum showed
absorptions at 3408 (OH) and 1495, 1592 (aromatic) cmꢀ1
,
and UV maximum absorption were found at 207, 232, and
298 nm,7 indicating a 2,3-dihydrobenzofuran skeleton. On the
basis of 1D NMR, one OH, one CH3, and three OCH3 groups
were found. Aromatic singlet protons, at δH 6.83 (H-4) and 6.51
(H-7), showed NOE correlations with δH 3.78 (OCH3)/1.34
(CH3) and δH 3.74 (OCH3), respectively. The third OCH3 was
located at C-40 on the basis of an HMBC correlation with C-40
and an NOE correlation with H-50. The OH group connected at
C-30 according to the HMBC correlations with C-20 (δC 114.4)
and C-30 (δC 148.2). The relative orientations of H-2 and H-3
in substituted 2,3-dihydrobenzofurans can be judged easily by
1H NMR chemical shifts, with the trans form at ca. δH 5.11 and
3.38 vs the cis form at ca. δH 5.73 and 3.62.10 The chemical shifts
of H-2 and H-3 in 3 were δH 5.02 and 3.29, which are consistent
with a trans orientation. This assignment was confirmed by the
absence of a NOE correlation. In addition, a negative Cotton
effect at 319 nm and a positive Cotton effect at 255 nm in the CD
spectrum, as well as a specific optical rotation value of ꢀ27.8,
were similar to those of a known compound ((2S,3S)-3-methyl-
2-phenyl-2,3-dihydrobenzb[b]furan, a negative Cotton effect at
280 nm and a positive Cotton effect at 220 nm, [R]D ꢀ4.8).10
The data suggested that the absolute configurations were 2S and
3S. Thus, compound 3 was identified as shown and was named
pterolinus C.
1512 cmꢀ1, respectively) groups. However, the UV spectra of
1 and 6 were significantly different. As indicated in the literature,9
3-methyl- and 2-methyl-benzofurans show different λmax absorp-
tions, at ca. 321 and 235 nm, respectively. The UV spectrum of
1 showed a maximum absorption at 242 nm, whereas that of 6
showed a λmax at 324 nm, which were consistent with the
reported data. The methyl groups of 1 and 6 showed significantly
different 13C NMR chemical shifts (δC 13.6 in 1 and δC 9.6 in 6).9
Therefore, the B rings in 1 and 6 are attached at C-3 and C-2,
respectively, of the arylbenzofuran skeleton. On the basis of the
1H and 13C NMR spectra, compound 1 has a 14-carbon skeleton
along with one methyl (δH 2.45/δC 13.6) and two methoxy
groups (δH 3.91/δC 57.4, and δH 3.90/δC 57.0). Five olefinic sp2
methine carbons (δC 96.5, 105.2, 113.5, 116.9, and 121.4) and
nine quaternary carbons [six oxygenated sp2 carbons (δC 145.1,
147.3, 148.1, 148.4, 149.4, 150.9)] were observed according to
Compound 4 showed an [M þ Na]þ ion at m/z 357.0952
(C17H18O7Na), corresponding to nine degrees of unsaturation.
The IR spectrum showed OH, CdO, and aromatic absorptions,
and a UV maximum absorption occurred at 230 nm. A 14-carbon
skeleton was evident from 1D NMR, including seven methine
carbons (four olefinic and three sp3 carbons), seven quaternary
carbons (three oxygenated carbons, an oxygenated sp3 carbon,
and two carbonyl carbons), a methyl, and two OCH3 groups. In
1
HSQC and DEPT spectra. In the H NMR spectrum, an ABX
system (δH 6.99, J = 2.0 Hz; 7.07, J = 8.0 Hz; 6.94, J = 8.0, 2.0 Hz)
and two singlet protons (δH 6.97 and 7.12) were found. The
latter were identified on the basis of the HMQC spectra and a
990
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
1
Table 1. H NMR Data of Compounds 1ꢀ7 (400 MHz in acetone-d6, J values in parentheses)
1
2a
3
4
5
6
7
1
4.28 d (4.0)
4.25 d (4.0)
2
5.02 d (8.0)
4.77 dd (10.0, 4.0)
1.81 dq (10.0, 7.2)
4.68 d (7.6, 4.0)
3.28 p (7.6)
4.98 d (8.2)
3
3.29 m (8.0, 6.8, 0.8)
6.83 d (0.8)
3.27 p (8.2, 6.8)
6.66 d (1.2)
4
6.97 s
7.12 s
6.96 s
6.96 s
7.13 s
5
3.93 s
6.62 s
6
7
7.13 s
6.51 s
6.49
8
5.91 s
5.99 s
9
10
20
30
6.99 d (2.0)
7.63 d (8.0)
6.96 d (8.0)
6.91 d (2.0)
6.88 d (2.0)
6.87 d (2.0)
7.28 d (2.4)
6.91 d (2.0)
40
50
60
7.07 d (8.0)
6.96 d (8.0)
7.63 d (8.0)
6.93 d (8.0)
6.86 dd (8.0, 2.0)
3.78 s
6.89 d (8.0)
6.86 d (8.0)
7.04 d (8.4)
6.49 d (8.0)
6.94 dd (8.0, 2.0)
6.80 dd (8.0, 2.0)
6.76 d (8.0, 2.0)
7.21 dd (8.4, 2.4)
6.86 dd (8.0, 2.0)
5-OCH3
6-OCH3
7-OCH3
40-OCH3
5-OH
30-OH
40-OH
CH3
3.91 s
3.92 s
3.74 s
3.91 s
3.84 s
3.83 s
3.83 s
3.82 s
3.83 s
3.90 s
7.30 s
7.75 s
3.84 s
3.89 s
7.34 s
7.86 s
3.81 s
6.95 s
7.65 s
7.23 br
7.73 s
7.54 s
7.50 s
8.60 br
2.36 s
2.45 s
1.34 d (6.8)
1.04 d (7.2)
0.97 d (7.6)
2.37 s
1.32 d (6.8)
a Recorded by 500 MHz NMR.
the 1H NMR spectrum, an ABX system (δH 6.88, J = 2.0 Hz; 6.89
J = 8.0 Hz; 6.80 J = 8.0, 2.0 Hz) was found, together with an
olefinic methine, an oxymethine, and two OH groups. A
CH3ꢀCHꢀCHꢀOH fragment (1.04, J = 7.2 Hz, CH3/1.81,
J = 10.0, 7.2 Hz, H-3/4.77, J = 10.0, 4.0 Hz, H-2/4.28, J = 4.0 Hz,
OH) was present. Proton signals at δH 6.88 and 6.80 showed a
HMBC correlation with C-2 (δC 76.3), indicating that ring B was
attached at C-2. Two unique carbons (δC 62.5, 65.1) of ring A
were observed in the 13C NMR spectrum, and a singlet proton at
δH 3.93 showed HMBC correlations with δC 47.1, 65.1, 160.3,
and 189.5. On the basis of this data, there was a C4ꢀC-5 epoxide.
The olefinic proton (δH 5.91, s) showed correlations with δC
65.1, 160.3, 189.5, and 193.5, providing evidence that ring A also
contained two carbonyl groups. The OCH3 groups in 4 showed
HMBC correlations with C-7 and C-40 and NOESY enhance-
ments with H-8 and H-50, which indicated that the OCH3
substituents were located at C-6 and C-40. The coupling constant
(J = 7.6 Hz) of H-2 and H-3 in compound 4 indicated that the
dihedral angle φ of these vicinal protons was near 0ꢀ or 150ꢀ.
However, a weak NOE correlation was observed between H-2
and H-3, which suggested that both dihedral angles (0ꢀ or 150ꢀ)
were possible. Therefore, we were not able to determine the
structure of 4 by means of NMR techniques. We attempted to
determine the absolute configuration in 4 by a Moshers’ reaction,
but were unable to distinguish the positive side and negative side
after calculating δS-MTPA ꢀ δR-MTPA. Thus, the structure of 4 was
identified as shown, and it was named pterolinus D.
4 were absent and instead were replaced by carbons resonating at
δC 153.7 and 132.2. Thus, a 1,4-benzoquinone was present. H-2
and H-3 of compound 5showed a large coupling constant (J=10Hz)
and no NOE correlation, indicating that these two protons had
an anti conformation. Because NOE correlations were found
between H-2 and the two vicinal groups, C3-Me and H-5, any of
four isomers were possible. Therefore, the stereochemistry was
not defined. Compound 5 was named pterolinus E.
Compound 8 had a molecular formula of C18H18O6, with nine
degrees of unsaturation. On the basis of comparison with
previously reported 13C NMR data,11 a neoflavonoid skeleton
1
was identified in 8. The H NMR spectrum exhibited an ABX
system (δH 6.66, J = 8.4, 2.4 Hz/6.86, J = 8.4 Hz/6.71, J = 2.4 Hz),
two aromatic singlet signals (δH 6.73/7.30), three OH
groups (δH 7.48, br/8.16, br/10.16, s), and two OCH3 groups
1
(δH 3.80, s/3.91, s). H NMR, COSY, and HSQC data estab-
lished the partial connectivity of the C7ꢀC8ꢀC9 segment and
identified the carbons as an aliphatic CH, an olefinic CH, and an
1
olefinic CH2. Two singlet aromatic protons in the H NMR
spectrum revealed the presence of a 1,2,4,5-tetrasubstituted
phenyl ring A. In the HMBC spectrum, both H-6 and H-60
showed correlations with C-7, suggesting that the two phenyl
groups were connected through C-7. In addition, one OCH3
group (δH 3.91) showed an HMBC correlation with the carbonyl
carbon at δC 171.7, which indicated the presence of a methyl
ester. The OH proton at δH 10.16 showed HMBC correlations
with C-4 (δc 111.5) and C-6 (δc 119.2), while H-3 showed
correlations with C-1, C-2, CO, and C-5. These data indicated
that ring A was para-hydroxy substituted. Furthermore, on the
basis of the HMBC and NOESY spectra, the OCH3 group (δH
3.80) was assigned at C-40. The absolute configuration was
Compound 5 had the molecular formula C17H18O6, one
oxygen atom less than 4. The IR spectrum showed absorptions
indicative of OH, CdO, and aromatic groups, and the UV had
maximum absorption at 230 and 273 nm. The epoxide carbons of
991
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
992
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
identified as 7S by comparing the [R]D value (þ20.6) with
showed NOE correlations with H-6b, H-3, and H-50, respec-
tively, and the structure was assigned as shown. Compound 12
was named pterolinus I.
previous references (latifolin, 7R, [R]20 ꢀ26.7; 2,4-dihydrox-
D
ydalbergiquinol, 7S, [R]22D þ34.7).11,12 Thus, the structure of 8
was elucidated, and the compound was named pterolinus F.
Compound 9 was obtained as a 2:5 mixture with 15, according
to 1H NMR data, even when different purification methods were
used. Compound 15 was isolated as a pure compound that
contains a 1,4-benzoquinone ring, as elucidated in an earlier
report,11,13 while 9 contains 1,4-hydroxy-substituted phenyl rings
(see OH signals in Table 2). To further suggest that 15 was the
oxidative product of 9, the latter compound showed an [M]þ ion
at m/z 302.1151 (C17H16O5) in the HRGCEIMS, while the
former compound showed an [M þ H]þ ion at m/z 300.97
(C17H17O5) in the ESIMS, 2 amu less than 9. In addition, excess
NaBH4 was used to reduce compound 15 to confirm that 9 was
the hydroquinone counterpart of 15 (data are shown in Support-
ing Information). We postulate that 9 is oxidized gradually to 15
and that both compounds should have an S configuration at C-7.
Compound 9 was named pterolinus G.
Compound 13 had the molecular formula C171H19O5, corre-
sponding to nine degrees of unsaturation. The H NMR and
HMQC spectra demonstrated that the terminal double bond (C-8,
C-9) found in 8ꢀ12 was absent and instead was replaced by
substituted methylene (δH 2.01, 2.17, m, δC 33.6) and oxyge-
nated methylene (δH 4.09, m, δC 65.4) groups. On the basis of
nine degrees of unsaturation, the presence of a heterocyclic ring
C was postulated. HMBC correlations of H-5 (δH 6.38, s) with
C-4 (δC 148.5) and C-6 (δC 149.7), as well as correlations of H-9
with C-6 and C-7 (δC 41.7), indicated that C-9 was linked to C-6
through an oxygen atom. The positions of the two OCH3 and
two OH groups at C-4, C-30, C-3, and C-40, respectively, were
confirmed from the NOESY and HMBC data. Thus, 13 was
identified and was named pterolinus J.
Compounds 1ꢀ8, 10, 11, and 13ꢀ16 were screened for
cytotoxicity against six human cancer cell lines, liver (HepG2,
Hep3B), gingival (Ca9-22), lung (A549), and breast (MCF7,
MDA-MB-231), and for anti-inflammatory activity based on
effects against superoxide anion generation and elastase release
by human neutrophils in response to fMLP/CB.
Compounds 10 and 11 were isolated by RI-recycle HPLC
(EtOAcꢀhexanes, 2:1, tR = 32.4 and 35.6 min). Both compounds
had the molecular weight C20H23O6. On the basis of similar
NMR data, 10 and 11 both had a neoflavonoid C7ꢀC8ꢀC9
fragment and a 3-hydroxy-4-methoxyphenyl ring. However, un-
like 15, compounds 10 and 11 had a ꢀCH2ꢀCOꢀCH3 moiety
(δH 2.95, 3.06, J = 14.5 Hz, δC 52.3/δC 205.6/δH 2.14, s, δC
31.9) and an OH (δH 5.10, s) rather than a second carbonyl
group in ring A, based on 1D NMR and HMBC data. In the
HMBC data of 10, H-7 (δH 4.68) showed correlations with C-1
(δC 139.4) and C-2 (δC 185.8), and the proton of an OH group
showed correlations with C-4, C-5, C-6, and ꢀCH2ꢀCOꢀCH3.
Therefore, the ꢀCH2ꢀCOꢀCH3 and OH substituents were
both attached to the C-5 of ring A. On the other hand, the
HMBC spectrum of 11 also showed correlation patterns similar
These compounds could be divided into those with a benzo-
furan skeleton and those with a neoflavonoid skeleton. Among
the neoflavonoids (8, 10, 11, 13ꢀ16), compound 14 showed the
highest cytotoxicity against Ca9-22, with an IC50 value of 0.46
μg/mL, 15 showed significant cytotoxicity against Hep3B and
MDA-MB-231 with IC50 values of 2.39 and 3.34 μg/mL, 10
exhibited cytotoxicity against HepG2 and MDA-MB-231 with
IC50 values of 3.65 and 2.85 μg/mL, and compound 11 showed
selective cytotoxicity against A549 with an IC50 value of 3.97 μg/mL.
In contrast, compounds 13 and 16, which contain a pyran or
lactone ring C formed by cyclization of a ꢀC7ꢀC8ꢀC9ꢀOꢀ
moiety onto ring B, were inactive (IC50 >17.10 μg/mL). There-
fore, the heterocyclic ring C abolished cytotoxic activity. Com-
pounds 8, 10, 11, 14, and 15 have identical phenyl B rings, but
differences in the A rings and linkage between the two rings.
Compound 8, with a para-dihydroxy-substituted phenyl ring A,
exhibited only weak cytotoxicity (IC50 11.10ꢀ16.10 μg/mL)
compared with 10, 11, and 15, which contain benzoquinone or
similar ring systems containing carbonyl groups. Compound 14,
which has a carbonyl (CdO) linkage between the phenyl A and
B rings rather than an allyl (CHꢀCHdCH2) group, exhibited
potent cytotoxcity against Ca9-22. Therefore, compound 14
could be a potential candidate for anticancer drug development.
Among the benzofurans (1ꢀ7), compound 2 showed the
highest cytotoxicity against A549 and MCF7, with IC50 values of
2.34 and 1.74 μg/mL, while compound 6, with a different
substitution pattern on ring B, and compound 1, with ring B
attached at C-2 rather than C-3, exhibited no cytotoxicity.
Compounds 3 and 7, with a trans-configured saturated C2ꢀC3
bond, were inactive (>19.42 μg/mL) or weakly active
(7.70ꢀ15.42 μg/mL). Compound 4, with an opened ring C
and benzoquinone ring A, showed significant cytotoxicity against
Hep3B and MCF7 (IC50 2.08 and 3.31 μg/mL) and moderate
cytotoxicity against HepG2 and MDA-MB-241 (IC50 4.33 and
4.89 μg/mL). Compound 5, which is structurally identical to 4,
except for lacking the epoxide group, exhibited reduced activity
(IC50 6.96ꢀ16.65 μg/mL).
to those of 10. Both 10 and 11 showed positive [R]25 values
D
([R]25 142.12/105.96) and a negative Cotton effect at ca.
D
340 nm and positive Cotton effect at ca. 300 nm. By comparisons
with previous data (shown in Supporting Information),13 we
suggest that C-7 has an S configuration. However, we could not
identify the absolute configuration at C-5. Thus, 10 and 11 were
identified as epimers and were named pterolinus Ha and
pterolinus Hb. However, these two epi-isomers could possibly
be artifacts due to reaction with acetone.
The molecular formula of 12 was C18H20O6, with nine
degrees of unsaturation. On the basis of the 13C NMR data, a
15-carbon skeleton and three OCH3 groups were found. The 1H
NMR and COSY data revealed a ꢀC7ꢀC8ꢀC9 moiety; how-
ever, the H-7 signal was shifted to δH 3.50 (d, J = 7.5 Hz),
indicating that one aromatic ring was absent. Two quaternary,
one methylene, and one carbonyl carbon were found in the 1D
NMR spectra. From the HMBC data, H-7 showed key correla-
tions with C-1, C-6, C-10, and C-60, revealing that this partial
structure was linked to C-7. In addition, H-6a and H-6b showed
HMBC correlation with C-1, C-2, C-4, C-5, and C-10, and H-6b
showed a long-range coupling (J = 1.2 Hz) with H-8 (δH 5.05).
Moreover, H-50 showed HMBC correlations with C-5, C-40, and
C-60. Thus, compound 12 was tricyclic and linked between C-5
and C-60. The relative configuration was assigned from the
NOESY spectrum, in which H2-6 showed no correlation with
H-7, suggesting that these two protons are on a different side
than H-7. The absolute configuration could not be identified in
the current study. The three OCH3 groups (δH 3.22, 3.72, 3.81)
Compounds 1ꢀ8, 10, 11, and 13ꢀ16 were evaluated for
inhibitory effects on superoxide anion generation and elastase
993
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
release by human neutrophils in response to fMLP/CB. The
benzofurans generally showed more potent inhibition than neoflavo-
noids (Table 5). Compounds 1, 2, 4, 6, and 7 showed potent
inhibition of superoxide anion generation with IC50 values of 0.33,
0.19, 0.29, 0.73, and 0.69 μg/mL, respectively. The inhibition was
1.6ꢀ6.2-fold greater than that of the positive control LY294002
(a phosphatidylinositol-3-kinase inhibitor).14,15 In addition, com-
pound 4 showed significant inhibition of human neutrophil elastase
release with an IC50 value of 1.06 μg/mL, which was 1.9-fold higher
than the positive control. Among the neoflavonoids, compound 15
exhibited potent inhibition in both assays with IC50 values of 0.47 and
1.44 μg/mL (Table 5). On the basis of the results, compounds 1, 2, 4,
6, 7, and 15 merit consideration as leads for anti-inflammatory agents.
Table 3. 13C NMR Data (δ) of Compounds 1ꢀ7 (100 MHz in
acetone-d6)
1
2a
3
4
5
6
7
1
2
150.9 151.1
117.8 110.0
94.0
47.2
76.3
47.1
65.1
62.5
78.1
41.2
151.3
125.4
104.9
93.8
47.1
3
4
105.2 104.2 111.0
145.1 148.5 151.8
153.7
132.2
183.5
160.0
109.2
188.4
138.4
115.0
111.6
5
147.9b 142.2
6
147.3 147.1 145.6 189.5
96.5 95.7 96.8 160.3
149.1
96.3
148.7b
7
96.1
8
149.4 144.4 155.1 110.5
122.8 124.8 123.6 193.5
127.5 124.3 135.9 138.5
116.9 128.6 114.4 115.0
148.4 116.5 148.2 148.1
148.1 158.0 149.0 148.6
113.5 116.5 112.9 112.5
121.4 128.6 118.9 119.8
57.0
145.0
111.2
125.4
114.4
153.8
124.3
135.9
114.4
9
10
20
30
40
50
60
’ EXPERIMENTAL SECTION
147.8b 148.2b 148.2b
148.3b 148.7b 148.9b
General Experimental Procedures. CD data were measured on
a JASCO J-815 instrument. Optical rotations were recorded on a JASCO
P-1020 polarimeter. IR spectra were measured on a Mattson Genesis II
FT-IR spectrophotometer. UV spectra were obtained on a JASCO
V-530 UV/vis spectrophotometer. NMR spectra were run on a Varian
Unity-plus 400 MHz FT-NMR or a Varian Mercury-plus 400 MHz FT-
NMR spectrometer. The chemical shift (δ) values are in ppm (parts per
million) with d6-acetone, d4-MeOH, or d5-pyridine as the internal
standard, and coupling constants (J) are in Hz. Low-resolution ESI-
mass spectra were obtained on a VG Biotech Quattro 5022 mass
spectrometer in a positive or negative mode, high-resolution MS spectra
were obtained on Bruker Daltonics APEX R30e and JEOL JMS-700
mass spectrometers. A JASCO PU986 pump, a JASCO UV-1575
detector, a JASCO 887-30 mix. Module, a JAI LC-918 RI-recycle HPLC,
and an Ascentis C18 column (250 mm ꢁ 10 mm, 5 μm) were employed
for HPLC separations. The Biotage SP1 was used for flash liquid
chromatography. Silica gel 60 (40ꢀ60 mesh, Merck), C-18 (40ꢀ63
mesh, Merck), Sephadex LH-20, Celite 545, and Diaion HP-20 were
used for column chromatography. Silica gel plates (Kieselgel 60, F254,
0.20 nm, Merck) were used for TLC.
Plant Material. Heartwood (4.75 kg) of P. santalinus was imported
from India, provided by Mr. Mike Y. C. Wei, and was identified by Prof. Dr.
Sheng-Yang Wang, Department of Forestry, National Chung-Hsing Uni-
versity, Taichung, Taiwan. A voucher specimen (Pterocarpus 002) was
deposited in the Graduate Institute of Natural Products, Kaohsiung, Taiwan.
Extraction and Isolation. The heartwood was powdered and
extracted with MeOH (5 ꢁ 20 L). After removal of solvent, the crude
extract (ca. 480 g) was partitioned between CH2Cl2 and 50% aqueous
MeOH. The CH2Cl2 layer was concentrated in vacuo, and the CH2Cl2
extract (ca. 300 g) was subjected to CC on Celite 545 and eluted with n-
hexane (8 L), CH2Cl2 (40 L), EtOAc (20 L), acetone (10 L), and
MeOH (8 L). Six fractions (PS-C1 to PS-C6) were obtained. PS-C3 (ca.
100 g) was subjected to silica gel CC using a hexaneꢀEtOAcꢀMeOH
gradient solvent system (3:1:0, 1:1:0, 0:1:0, 0:40:1, 0:20:1, 0:10:1,
0:6:1), to give 11 fractions (PS-C3.1 to PS-C3.11).
112.4
119.5
113.3
119.4
112.9
118.9
5-OCH3
6-OCH3
7-OCH3
40-OCH3
CH3
57.4
56.8
58.1
56.9
56.9
57.7b
56.9b
14.7
57.2b
56.9b
17.8
57.0
13.6
56.9
19.4
57.4
10.3
57.2
19.3
9.6
a Recorded by 125 MHz NMR. b The chemical shifts may be
interchanged.
PS-C3.5.6.7 (478 mg) was separated by RP-MPLC with a MeOH and
H2O gradient system and divided into five subfractions, while com-
pound 13 (7.6 mg) was isolated by RP-HPLC (Acentis, 250 ꢁ 10 mm,
57% MeOH(aq)) from subfraction 3. PS-C3.5.6.8 (0.27 g) was sepa-
rated by RP-MPLC with 60% MeOH(aq), and eight subfractions were
obtained, while compounds 1 and 8 were isolated by PTLC with pure
CH2Cl2 from subfraction 7 (70.7 mg). PS-C3.5.6.9 (533 mg) was
separated by RP-MPLC with a MeOH and H2O gradient to give nine
fractions, while 2 (19.2 mg) and 9 (9.7 mg) were isolated by PTLC with
pure CH2Cl2 from subfraction 3 (117.6 mg).
Pterolinus A (1): brown gum; UV λMeOHmax nm (log ε) 242 (4.13),
297 (3.97), 309 (3.84), 350 (3.33); IR (neat) νmax 3443, 2930, 1621,
1514 cm ;
ꢀ1 13C and 1H NMR data, see Table 1; HREIMS m/z 300.0999
[M]þ (calcd 300.0999 for C17H16O5).
Pterolinus B (2): amorphous, pale brown powder; UV λMeOHmax nm
(log ε) 226 (4.00), 284 (4.05), 321 (4.28), 337 (4.08); IR (neat) νmax
1
3317, 2921, 1677, 1510 cmꢀ1
;
13C and H NMR data, see Table 2;
ESIMS m/z 270.99 [M þ H]þ; HRESIMS m/z 271.0972 [M þ H]þ
(calcd 271.0970 for C16H15O4).
Pterolinus C (3): amorphous, pale brown powder; [R]25 ꢀ27.8
D
(c 0.095, MeOH); UV λMeOHmax nm (log ε) 207 (4.44), 232 (3.93), 289
(3.72), 298(3.69);CD ε315 ꢀ11.08, Δε258 29.26 (MeOH; c=0.3mg/mL);
Compounds 14 (3.6 g) and 16 (ca. 8 g) were obtained as precipitated
solids from PS-C3.6 and PS-C3.7, and 6 (144 mg) was obtained from
PS-C3.5. PS-C3.5 (31.2 g) was subjected to repeated silica gel CC with a
CH2Cl2 and MeOH gradient solvent system, and 11 fractions (PS-
C3.5.1 to PS-C3.5.11) were obtained. PS-C3.5.6 (20.70 g) was subjected
to CC on Sephadex LH-20 and eluted with acetone and CH2Cl2 (1:1) to
give 10 fractions (PS-C3.5.6.1 to PS-C3.5.6.10). PS-C3.5.6.3 (6.0 g) was
repeatedly chromatographed on RP-MPLC, Sephadex LH-20, NP-RI-
HPLC, silica gel, and PTLC to give 3 (9.8 mg), 10 (25 mg), 11 (15.1
mg), 12 (0.6 mg), and 15 (42 mg). PS-C3.5.6.5 and PS-C3.5.6.6 were
mixed (total ca. 6.8 g) and subjected to CC with Sephadex LH-20, RP-
MPLC, silica gel, and PTLC to give 4 (16.5 mg), 5 (12 mg), and 7 (1.6 g).
IR (neat) νmax 3408, 2928, 1592, 1495 cmꢀ1 13C and 1H NMR data,
;
see Table 3; ESIMS m/z 316.93 [M þ H]þ; HRESIMS m/z 317.1387
[M þ H]þ (calcd 317.1389 for C18H21O5).
Pterolinus D (4): amorphous, brown powder; [R]25D ꢀ116.0 (c 0.10,
MeOH); UV λMeOHmax nm (log ε) 230 (4.04), 264 (3.86), 278 (3.86);
CD Δε329 ꢀ93.23, Δε293 þ26.64 (MeOH; c = 0.3 mg/mL); IR (neat)
νmax 3433, 2922, 1715, 1667, 1609, 1511 cmꢀ1 13C and 1H NMR data,
;
see Table 3; ESIMS m/z 357 [M þ Na]þ; HRESIMS m/z 357.0952
[M þ Na]þ (calcd 357.0950 for C17H18O7Na).
Pterolinus E (5): amorphous, brown powder; [R]25D ꢀ10.4 (c 0.10,
MeOH); UV λMeOHmax nm (log ε) 230 (4.16), 273 (4.04), 313 (3.82);
CD Δε384 þ9.67, Δε313 ꢀ7.51 (MeOH; c = 0.3 mg/mL); IR (neat) νmax
994
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
Table 4. 13C NMR Data (δ) of Compounds 8ꢀ15 (100 MHz
in acetone-d6)
Table 5. Inhibitory Effects of Compounds 1ꢀ8, 10, 11, and
13ꢀ16 on Superoxide Anion Generation and Elastase Release
by Human Neutrophils in Response to FMLP/CB
8
9
10a 11a 12b
13
14c
15
16b
superoxide anion
elastase release
1
148.5 123.3 139.4 139.5 82.2 118.2 112.6 152.0 112.4
142.0 148.7 185.8 185.9 185.0 117.2 159.8 187.6 112.2
115.6 102.1 101.8 101.8 101.3 141.7 101.1 109.1 114.9
111.5 147.6d 174.8 174.8 198.9 148.5 156.3 160.3 152.5
156.5 141.0 69.9 69.9 74.1 101.8 140.3 187.3 100.6
119.2 117.1 143.4 143.3 40.2 149.7 118.7 132.5 155.8
48.9 47.9 46.8 47.2 51.8 41.7 199.3 48.0 149.7
2
compound
IC50a (μg/mL) or (Inh %)b
IC50 (μg/mL) or (Inh %)
3
1
0.33 ( 0.38
0.19 ( 0.03
2.28 ( 0.79
0.29 ( 0.07
NT
2.54 ( 0.57
(47.83 ( 4.63)f
(9.73 ( 2.95)d
1.06 ( 0.24
4
2
5
3
6
4
7
5
2.13 ( 0.19
8
141.5 142.8 140.9 140.9 141.4 33.6
117.0 115.9 116.2 116.2 117.3 65.4
139.8 111.8
118.2 161.6
6
0.73 ( 0.20
0.69 ( 0.19
NT
3.69 ( 0.19
9
10
20
30
40
50
60
7
(33.33 ( 7.33)e
3.22 ( 1.01
136.7 138.2 135.3 135.5 133.7 139.0 132.1 134.3 129.2
117.0 117.0 116.4 116.5 118.5 113.4 117.3 116.9 120.0
147.6 147.6d 146.9 147.2 147.9 149.0 148.0 148.0d 148.3
148.0 147.3d 147.2 146.9 148.0 146.7 151.8 148.2d 149.4
113.0 112.8 112.2 112.2 109.1 116.2 111.1 113.1 112.2
120.9 120.8 120.5 120.3 128.1 122.6 122.0 121.0 116.7
52.0
8
10
(29.44 ( 7.71)d
(34.15 ( 7.76)d
2.10 ( 0.29
NT
(19.53 ( 7.60)
(2.84 ( 5.74)
(16.59 ( 6.99)
3.62 ( 0.71
11
13
14
15
0.47 ( 0.04
3.18 ( 0.48
1.18 ( 0.24
1.44 ( 0.11
1-OCH3
2-OCH3
4-OCH3
5-OCH3
30-OCH3
16
(ꢀ1.53 ( 2.90)
1.99 ( 0.35
57.9
LY294002c
56.9 56.3 56.4
56.8
56.9
57.3 55.9
a Concentration necessary for 50% inhibition (IC50). b Percentage of
inhibition (Inh %) at 10 μg/mL concentration. Results are presented as
mean ( SEM (n = 3ꢀ5). c LY294002, a phosphatidylinositol-3-kinase
inhibitior, was used as a positive control for superoxide anion generation
and elastase release. d p < 0.05 compared with the control value. e p < 0.01
compared with the control value. f p < 0.001 compared with the control
value. “NT” means not tested.
56.0
40-OCH3 56.9 57.0 56.4 56.3 56.9
55.8 56.9 56.1
COOCH3 53.4
COOCH3 171.7
CH2
52.3 52.1
31.9 31.9
205.6 205.7
COCH3
COCH3
Pterolinus I (12): colorless oil; [R]25 ꢀ30.7 (c 0.10, EtOH); UV
D
λMeOHmax nm (log ε) 250(3.86);CDε315 ꢀ132.37 (EtOH; c=0.3mg/mL);
a Recorded by 125 MHz NMR in acetone-d6. b Recorded by 150 MHz
NMR in acetone- d6. c Measured in pyridine-d5. d The chemical shifts are
interchangeable.
IR (neat) νmax 3405, 2925, 1713, 16467 1510 cmꢀ1 13C and 1H NMR
;
data, see Table 3; ESIMS m/z 333.02 [M þ H]þ; HRESIMS m/z
333.1339 [M þ H]þ (calcd 333.1338 for C18H21O6).
Pterolinus J (13): amorphous, white powder; [R]25D þ11.6 (c 0.10,
EtOH); UV λMeOHmax nm (log ε) 227 (4.10), 289 (3.83), 300 (3.76);
CD ε320 ꢀ5.75, ε309ꢀ6.17, Δε257 ꢀ23.11 (EtOH; c = 0.3 mg/mL); IR
3419, 2924, 1681, 1606, 1509 cm ;
ꢀ1 13C and 1H NMR data, see Table 3;
ESIMS m/z 341 [M þ Na]þ; HRESIMS m/z 341.0998 [M þ Na]þ
(calcd 341.1001 for C17H18O6Na).
(neat) νmax 3433, 2921, 1708, 1609, 1507 cmꢀ1 13C and 1H NMR data,
;
Pterolinus F (8): yellow, colorless oil; [R]25D þ20.6 (c 0.11, CH2Cl2);
UV λMeOHmax nm (log ε) 220 (4.23), 253 (3.88), 285 (3.62), 341 (3.68);
CD Δε363 ꢀ4.15, Δε307 ꢀ12.12, Δε266 þ2.51 (EtOH; c = 0.3 mg/mL);
ε329 ꢀ2.57, Δε266 þ2.13, Δε255 ꢀ2.12 (MeOH; c = 0.5 mg/mL); IR
see Table 3; ESIMS m/z 302.98 [M þ H]þ; HRESIMS m/z 303.1233
[M þ H]þ (calcd 303.1232 for C17H19O5).
In Vitro Cytotoxicity Assay. Fractions and isolates were tested
against lung (A549), gingival (Ca9-22), breast (MEA-MB-231 and
MCF7), and liver (HepG2 and Hep3B) cancer cell lines using estab-
lished colorimetric MTT assay protocols.16 Doxorubicin was used as a
positive control.
(neat) νmax 3419, 2955, 1677, 1623, 1510 cmꢀ1 13C and 1H NMR data,
;
see Table 3; ESIMS m/z 331.10 [M þ H]þ; HRESIMS m/z 331.1185
[M þ H]þ (calcd 331.1182 for C18H19O6).
Pterolinus G (9): brown gum; UV λMeOHmax nm (log ε) 261 (3.92);
In Vitro Anti-inflammatory Assay. The assay procedure was
IR (neat) νmax 3420, 2928, 1670, 1600, 1514 cmꢀ1 13C and 1H NMR
;
performed as previously reported.17ꢀ19
data, see Table 3; HREIMS m/z 300.1000 [M]þ (calcd 300.0998 for
C17H16O5), 302.1151 [M]þ (calcd 302.1151 for C17H18O5).
’ ASSOCIATED CONTENT
Pterolinus Ha (10): brown gum; [R]25D þ142.12 (c 0.12, CH2Cl2); UV
λMeOHmax nm (log ε) 232 (4.43), 280 (4.12), 399 (3.37); CD ε343 ꢀ23.87,
Δε325 ꢀ19.51, Δε295 þ4.33 (EtOH; c = 0.3 mg/mL); IR (neat) νmax
Supporting Information. 1H, 13C NMR and HRMS data
of all new compounds along with the stereochemical comparison
of compounds 8, 10, 11, and 15 are available free of charge via the
Internet at http://pubs.acs.org.
S
b
3392, 2922, 1704, 1664, 1609, 1507 cmꢀ1 13C and 1H NMR data, see
;
Table 3; ESIMS m/z 359.14 [M þ H]þ; HRESIMS m/z 359.1494 [M þ
H]þ (calcd 359.1495 for C20H23O6).
Pterolinus Hb (11): brown gum; [R]25D þ106.0 (c 0.094, CH2Cl2); UV
λMeOHmax nm (log ε) 232 (4.66), 282 (4.36), 399 (3.50); CD ε349 ꢀ20.13,
’ AUTHOR INFORMATION
Corresponding Author
ε305 þ24.37 (EtOH; c = 0.3 mg/mL); IR (neat) νmax 3418, 2922, 1705,
1606, 1510 cmꢀ1
;
13C and H NMR data, see Table 3; ESIMS m/z
1
*Tel: þ886-7-3121101, ext. 2162. Fax: þ886-7-3114773. E-mail:
aaronfrc@kmu.edu.tw (F.-R.C.). Tel: þ886-4-22057153. Fax:
þ886-4-22060248. E-mail: yachwu@mail.cmu.edu.tw (Y.-C.W.).
359.14 [M þ H]þ; HRESIMS m/z 359.1494 [M þ H]þ (calcd
359.1495 for C20H23O6).
995
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996

Journal of Natural Products
ARTICLE
’ ACKNOWLEDGMENT
This work was supported by grants from the National Science
Council, Taiwan (NSC100-2917-I-037-002), and Department of
Health, Executive Yuan, Taiwan (DOH99-TD-C-111-002). The
authors deeply appreciate Dr. K.-H. Lee for the comments and
suggestions, as well as the English revision by Dr. S. L. Morris-
Natschke (Natural Products Research Laboratories, UNC).
’ REFERENCES
(1) Parkash, E.; Sha Valli Khan, P. S.; Sreenivasa Rao, T. J. V.; Meru,
E. S. J. For. Res. 2006, 11, 329–335.
(2) Cho, J. Y.; Park, J.; Kim, P. S.; Yoo, E. S.; Baik, K. U.; Park, M. H.
Biol. Pharm. Bull. 2001, 24, 167–171.
(3) Kwon, H. J.; Hong, Y. K.; Kim, K. H.; Han, C. H.; Cho, S. H.;
Choi, J. S.; Kim, B. W. J. Ethnopharmacol. 2006, 105, 229–234.
(4) Kumar, N.; Ravindranath, B.; Seshadri, T. R. Phytochemistry
1974, 13, 633–636.
(5) Krishnaveni, K. S.; Rao, J. V. Phytochemistry 2000, 53, 605–605.
(6) Kesari, A. N.; Gupta, R. K.; Watal, G. Phytochemistry 2004,
65, 3125–3129.
(7) Donnelly, B. J.; Donnelly, D. M. X.; O’Sullivan, A. M.; Prendergast,
J. P. Tetrahedron 1969, 25, 4409–4414.
(8) Donnelly, D. M. X.; O’Reilly, J.; Whalley, W. B. Phytochemistry
1975, 13, 2287–2290.
(9) Muangnoicharoen, N.; Frahm, A. W. Phytochemistry 1981,
20, 291–293.
(10) Juhꢀasz, L.; Szilꢀagyi, L.; Antus, S.; Visy, J.; Zsila, F.; Simonyi, M.
Tetrahedron 2002, 58, 4261–4265.
(11) Muangnoicharoen, N.; Frahm, A. W. Phytochemistry 1982,
21, 767–727.
(12) Awale, S.; Shrestha, S. P.; Tezuka, Y.; Ueda, J. Y.; Matsushige,
K.; Kadota, S. J. Nat. Prod. 2005, 68, 858–864.
(13) Eyton, W. B.; Ollis, W. D.; Sutherland, I. O. Tetrahedron 1966,
21, 2683–2696.
(14) Vlahos, C. J.; Matter, W. F.; Hui, K. Y.; Brown, R. F. J. Biol.
Chem. 1994, 269, 5241–5248.
(15) Yu, H. P.; Hsieh, P. W.; Chang, Y. J.; Chung, P. J.; Kuo, L. M.;
Hwang, T. L. Biochem. Pharmacol. 2009, 78, 983–992.
(16) Mosmann, T. J. Immunol. Methods 1983, 65, 55–63.
(17) Yen, C. T.; Hwang, T. L.; Wu, Y. C.; Hsieh, P. W. Eur. J. Med.
Chem. 2009, 44, 1933–1940.
(18) Chang, H. L.; Chang, F. R.; Chen, J. S.; Wang, H. P.; Wu, Y. H.;
Wang, C. C.; Wu, Y. C.; Hwang, T. L. Eur. J. Pharmacol. 2008,
586, 332–339.
(19) Hwang, T. L.; Yen, S. H.; Leu, Y. L.; Chern, C. Y.; Hsu, H. C.
Br. J. Pharmacol. 2006, 48, 78–87.
996
dx.doi.org/10.1021/np100871g |J. Nat. Prod. 2011, 74, 989–996