Cytotoxic and antimicrobial constituents of the bark of Diospyros maritima collected in two geographical locations in Indonesia

Article abstract of DOI:10.1021/np040027m

Bioactivity-directed fractionation of extracts of two Diospyros maritima bark samples from Indonesia, one collected at sea level in a beach forest in Java and the other collected at a slight elevation away from the sea shore on the island of Lombok, yielded a diverse set of secondary metabolites. The naphthoquinone plumbagin (1), although found in extracts of both specimens, constituted a much larger percentage of the former sample, which also yielded a series of plumbagin dimers, maritinone (2), chitranone (3), and zeylanone (4). The latter sample yielded a new naphthoquinone derivative, (4S)-shinanolone (5), and a new natural product coumarin, 7,8-dimethoxy-6-hydroxycoumarin (6), along with three other analogues of plumbagin, 2-methoxy-7-methyljuglone (7), 3-methoxy-7-methyljuglone (8), and 7-methyljuglone (9). The structures of compounds 5 and 6 were elaborated by physical, spectral, and chemical methods. All of the isolates were evaluated in both cytotoxicity and antimicrobial assays, and structure-activity relationships of these naphthoquinones are proposed. Plumbagin (1) and maritinone (2) were evaluated also for in vivo antitumor activity in the hollow fiber assay, but both were found to be inactive.

Full text of DOI:10.1021/np040027m

1
156
J . Nat. Prod. 2004, 67, 1156-1161
Cytotoxic a n d An tim icr obia l Con stitu en ts of th e Ba r k of Diospyr os m a r itim a
Collected in Tw o Geogr a p h ica l Loca tion s in In d on esia
†
‡
†,‡
†
†
§
J ian-Qiao Gu, Tyler N. Graf, Dongho Lee, Hee-Byung Chai, Qiuwen Mi, Leonardus B. S. Kardono,
⊥
⊥
⊥
†
†
Fransisca M. Setyowati, Rachman Ismail, Soedarsono Riswan, Norman R. Farnsworth, Geoffrey A. Cordell,
J ohn M. Pezzuto,†,| Steven M. Swanson, David J . Kroll, J oseph O. Falkinham, III, Monroe E. Wall,
†
‡
∇
‡,#
‡
†,O
,‡
Mansukh C. Wani, A. Douglas Kinghorn, and Nicholas H. Oberlies*
Program for Collaborative Research in the Pharmaceutical Sciences and Department of Medicinal Chemistry
and Pharmacognosy, College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street,
Chicago, Illinois 60612, Natural Products Laboratory, Research Triangle Institute, P.O. Box 12194,
Research Triangle Park, North Carolina 27709-2194, Research Center for Chemistry, Indonesian Institute of Science,
Serpong, 15310 Tangerang, Indonesia, Herbarium Bogoriense, Research and Development Center for Biology, Indonesian
Institute of Science, 16122 Bogor, Indonesia, and Department of Biology, Virginia Polytechnic Institute and State University,
Blacksburg, Virginia 24061-0406
Received February 5, 2004
Bioactivity-directed fractionation of extracts of two Diospyros maritima bark samples from Indonesia,
one collected at sea level in a beach forest in J ava and the other collected at a slight elevation away from
the sea shore on the island of Lombok, yielded a diverse set of secondary metabolites. The naphthoquinone
plumbagin (1), although found in extracts of both specimens, constituted a much larger percentage of
the former sample, which also yielded a series of plumbagin dimers, maritinone (2), chitranone (3), and
zeylanone (4). The latter sample yielded a new naphthoquinone derivative, (4S)-shinanolone (5), and a
new natural product coumarin, 7,8-dimethoxy-6-hydroxycoumarin (6), along with three other analogues
of plumbagin, 2-methoxy-7-methyljuglone (7), 3-methoxy-7-methyljuglone (8), and 7-methyljuglone (9).
The structures of compounds 5 and 6 were elaborated by physical, spectral, and chemical methods. All
of the isolates were evaluated in both cytotoxicity and antimicrobial assays, and structure-activity
relationships of these naphthoquinones are proposed. Plumbagin (1) and maritinone (2) were evaluated
also for in vivo antitumor activity in the hollow fiber assay, but both were found to be inactive.
Diospyros maritima Blume (Ebenaceae) is a shrub
graphically distinct samples displayed promising antican-
cer activity, and thus, each was investigated individually
via bioactivity-directed fractionation, using KB cells (A3919)
and Lu1 cells (A5248) to monitor the chromatographic
purifications.
1
indigenous to Southeast Asia. The sap of this species has
2
been used as a vesicant in Indonesia, and the stems are
used traditionally in Taiwan as a treatment for rheumatic
3
diseases. Crude extracts of D. maritima have displayed a
4
range of biological effects, such as ichthyotoxicity and
The cytotoxic compound plumbagin (1) along with the
inactive compounds lupeol and scopoletin were found to be
common to both accessions. Several active compounds that
can be considered dimers of 1, maritinone (2), chitranone
(3), and zeylanone (4), were isolated from the A3919
sample. Two new compounds were isolated from the A5248
sample: a naphthoquinone derivative, (4S)-shinanolone (5),
and a new natural coumarin, 7,8-dimethoxy-6-hydroxycou-
marin (6); these two compounds were not active in either
the anticancer or antimicrobial assays. Further, three
known active naphthoquinones, all of which are analogues
of 1, but not dimers, 2-methoxy-7-methyljuglone (7), 3-meth-
oxy-7-methyljuglone (8), and 7-methyljuglone (9), were
isolated from A5248, along with the inactive compounds
betulin and methyl R-orcinolcarboxylate. The identification
of compounds 1-4, the structure elucidation data of 5 and
6, and the biological evaluation of all the isolates performed
in anticancer and antimicrobial assays are described
herein.
antibacterial activity.5 Previous phytochemical studies on
,6
D. maritima have resulted in the isolation of structurally
7
diverse secondary metabolites, including alkanol ester,
8
9
2,9,10
aliphatic lactone, benzenoid, coumarin,
kaurane diter-
phenylpropanoid,17 sterol,10,15
9
2,5,8-16
pene, naphthoquinone,
and triterpenoid types.3
,5,7,9,10,18-20
Recently, several puri-
fied naphthoquinones from this plant were found to exhibit
activity against bacteria, fungi,9 guppy fish, human
5
1,9
1
5
tumor cells, and seed germination of Lactuca sativa L.
var. Great Lakes1 and to effect platelet aggregation.
,9
21
As a part of our collaborative program to search for new
anticancer compounds from plants,22 two samples of the
bark from D. maritima were collected from different
geographical locations in Indonesia. One of these was
gathered in J uly 1996 near sea level in a beach forest on
the western side of J ava (A3919), while the other was
gathered in October 2001 at elevation, farther from the
beach, on Lombok Island (A5248). Extracts of these geo-
*
To whom correspondence should be addressed. Tel: (919) 541-6958.
Resu lts a n d Discu ssion
Fax: (919) 541-6499. E-mail: oberlies@rti.org.
†
University of Illinois at Chicago.
Research Triangle Institute.
Research Center for Chemistry, Indonesian Institute of Science.
Research and Development Center for Biology, Indonesian Institute
The naphthoquinone plumbagin (1) was the most abun-
dent compound isolated from the D. maritima bark sample
A3919, accounting for more than 0.25% of the weight of
the dried plant material, and this confirms earlier reports
‡
§
⊥
of Science.
|
Present address: Heine Pharmacy Building, Purdue University, West
9
of its high concentration in this taxon. Even though the
Lafayette, IN 47907.
∇
#
13
Virginia Polytechnic Institute and State University.
Deceased J uly 6, 2002.
Present address: College of Pharmacy, The Ohio State University,
C NMR data exactly matched the reported literature
values,9 complete 1H and C NMR data are shown in
13
O
Columbus, OH 43210.
Supplemental Table 1 (see Supporting Information) for two
1
0.1021/np040027m CCC: $27.50
© 2004 American Chemical Society and American Society of Pharmacognosy
Published on Web 07/09/2004

Cytotoxic Constituents of Diospyros maritima
J ournal of Natural Products, 2004, Vol. 67, No. 7 1157
Another plumbagin dimer, chitranone (3), was isolated,
and its 13C NMR data were reported recently. However,
the assignments for positions C-7 and C-7′ could be
distinguished by using a combination of the HSQC data,
the observed HMBC correlations from H-7′ to C-3, and the
ortho-coupled, doublet splitting patterns of 7.5 Hz in the
1
1
H NMR data for both H-7′ and H-8′, as opposed to the
more complex splitting for H-6, H-7, and H-8. Moreover, a
combination of the HSQC data and the allylic splitting
1
pattern observed for H-11′ in the H NMR data led to the
1
reassignment of positions C-11 and C-11′. As with 1, the
1
23
H NMR data for 3 are included to update the literature,
especially for positions H-6, H-7, and H-8 versus H-7′ and
H-8′ (Supplemental Table 2, see Supporting Information).
Zeylanone (4) is a dimer of 1 where the C-11′ methyl
group of one unit has been condensed at the C-3 position
of a second unit. Together with a bond between positions
C-2 and C-3′, this results in the formation of a five-
membered ring.24 Sankaram et al. presented a structure
elucidation of 4 using 1H NMR data and chemical shift
24
1
3
reagents, but, to the best of our knowledge, the C NMR
data for 4 have not been reported previously. Thus, a
1
combination of the HSQC, HMBC, and H NMR data was
used to assign the 13C NMR data (Supplemental Table 2,
see Supporting Information). Each half of the molecule
could be distinguished by a series of key HMBC correla-
tions (Supplemental Figure 1, see Supporting Information).
For instance, correlations from the H-11 methyl protons
to the C-1 quinone, along with correlations from H-8 to C-1,
the C-5-hydroxyl proton to C-6 and C-10, and cross cor-
relations between positions H-6 and H-8 were used to
assign one-half of the molecule, while similar HMBC
correlations from H-8′ to the C-1′ quinone, from the C-5′-
hydroxyl proton to C-6′, and cross correlations between H-6′
and H-8′ led to the assignments for the other half of the
molecule. Other key observations are the HMBC correla-
tions from both H-11′ and H-3 to C-4, whereas there were
no HMBC correlations to C-4′, as would be predicted due
to being located four bonds from the nearest proton. The
stereochemistry at positions C-2 and C-3 was not assigned
due to the paucity of compound isolated.
From sample A5248, a new compound (5), [R]25
+14° (c
D
0
.1, MeOH), was obtained as a pale yellowish amorphous
powder, and the molecular formula was deduced as C11
H
12
O
3
+
by HREIMS (obsd m/z 192.0778 for [M] ). Compound 5
exhibited a yellowish fluorescence on silica gel TLC plates
under UV light at 365 nm, and it showed UV absorption
maxima at 217, 267, and 329 nm, suggesting the presence
of an acetophenone chromophore.25 In the IR spectrum,
-
1
absorption bands at 3526-3109 cm (hydrogen-bonded
-
1
25
1
OH) and 1641 cm (carbonyl) were apparent. The
NMR and COSY spectra showed three isolated spin sys-
tems, inclusive of a hydrogen-bonded hydroxyl at δ 12.44
1H, s, OH-8), a methyl group at δ 2.38 (3H, br s, CH -6)
coupled with two aromatic protons at δ 6.76 (1H, br s,
H-7) and 6.87 (1H, br s, H-5), and a methine at δ 4.89
(1H, dd, J ) 7.4 and 3.8 Hz, H-4) coupled with a methylene
at δ 2.20 and 2.36 (each 1H, m, H -3), which were further
coupled with another methylene unit at δ 2.67 and 2.97
(each 1H, m, H -2). We were not able to assign the
respective R and â orientations of the methylene signals
-2 and H -3; neither of the J values at H-4 were large
enough to indicate a â relationship to either of the H -3
H
H
(
H
3
H
1
reasons. These H NMR data are superior to the earlier
H
literature,10 largely due to advances in instrumentation in
1
3
the past 30 years. Moreover, the C NMR data for the
structurally related plumbagin dimers, maritinone (2) and
zeylanone (4), have not been published, and their identi-
fication was facilitated greatly by comparison with 1.
A comparison of the NMR and mass spectral data
between compounds 1 and 2 suggested that 2 was a dimer
H
2
H
2
H
2
2
2
of 1 with C
2
symmetry (Supplemental Table 1).10 The
protons, the NOE difference spectrum of H-4 showed
correlations to both H-3a and H-3b, and the results of
splitting pattern for H-6 and H-7 was simplified due to the
absence of H-8, which established the dimer linkage as C-8
to C-8′.
chemical shift difference experiments in CDCl
3 5 5
and C D N
were inconclusive. One hydroxyl group was placed at C-4

1
158 J ournal of Natural Products, 2004, Vol. 67, No. 7
Gu et al.
Ta ble 1. 1H, 13C, DEPT-135, and HMBC NMR Data for Compounds 5, 5s, and 5r
4S)-shinanolone (5)
(
5s
5r
position
δC (CDCl3)
DEPT
δH (CDCl3)a
HMBC (HfC)
δH (C5D5N)a
δH (C5D5N)a
δH (C5D5N)a
∆δH((S-R)
1
203.9
34.8
C
CH2
2
2
a
2.97, ddd
68.1, 203.9
2.98, ddd
(18, 6.6, 4.6)
2.63, ddd
(18, 9.7, 4.9)
2.30, m
2.22, m
5.02, dd
(8.4, 3.8)
7.17, s
2.59, m
2.71, m
2.69, m
2.84, m
-0.10
-0.13
(
18, 8.1, 4.8)
2.67, ddd
18, 8.3, 4.8)
b
68.1, 203.9
(
3
3
4
a
b
31.7
68.1
CH2
CH
2.36, m
2.20, m
4.89, dd
2.28, m
2.28, m
6.41, dd
(4.3, 4.0)
6.86, s
2.35, m
2.28, m
6.42, dd
(6.7, 3.8)
6.57, s
-0.07
68.1, 145.9, 203.9
0
b
S
(7.4, 3.8)
5
6
7
8
9
1
118.9
149.1
118.1
163.2
113.8
145.9
22.6
CH
C
CH
C
C
C
6.87, s
22.6, 68.1, 113.8, 118.1
22.6, 113.8, 118.9
+0.29
+0.07
6.76, s
6.81, s
6.86, s
6.79, s
0
CH3-6
OH-8
CH3
2.38, s
12.44, s
118.1, 118.9, 149.1
113.8, 118.1, 163.2
2.18, s
12.98, s
2.15, s
2.09, s
+0.06
a
δH, multiplicity (J in Hz). b Absolute configuration.
Ta ble 2. Cytotoxic Activity of Constituents from the Bark of
D. maritima
as a result of correlations between H-5 (δ
H
6.87) and C-4
a-d
(δ
C
68.1) in the HMBC spectrum (Table 1) and the spatial
compound
KB
Lu1
LNCaP
HUVEC
correlation between H-4 and H-5 observed in the NOE
difference spectra. The absolute configuration was deter-
1
2
3
4
7
8
9
0.3
0.3
0.3
2.1
1.9
2.4
4.1
0.3
1.4
1.1
14.5
0.2
1.8
0.4
2.2
2.2
3.3
3.4
4.1
3.7
0.4
2.2
2.1
3.3
2.0
5.5
5.7
mined by a convenient Mosher ester procedure carried out
in NMR tubes,26,27 in which 5 was treated with (R)-(-)- and
(S)-(+)-R-methoxy-R-(trifluoromethyl)phenylacetyl (MTPA)
chloride to obtain the (S)- (5s) and (R)-MTPA (5r ) esters,
respectively. Analysis of the ∆δΗ(S-R) data between these
compounds (Table 1) showed a positive difference in
13.2
a
Results are expressed as EC50 values (µg/mL). b Key to cell
lines used: KB ) human oral epidermoid carcinoma; Lu1 )
human lung cancer; LNCaP ) hormone-dependent human pros-
tate cancer; HUVEC ) human umbilical vein endothelial cells.
3
chemical shift for the protons at C-5 and CH -6 and a
negative effect for the protons at C-2 and C-3, and this
indicated that the absolute configuration at C-4 was S.
Accordingly, the structure of 5 was assigned as 4S,8-
dihydroxy-6-methyl-1-tetralone, to which the trivial name
c
Compounds 5 and 6, betulin, lupeol, methyl R-orcinolcarboxylate,
and scopoletin were regarded as inactive (ED50 values >5 µg/mL
in all cell lines). d Typical average EC50 value for camptothecin
against KB cells is 0.01 µg/mL.
(
4S)-shinanolone has been accorded. The structure of
2
0
shinanolone, with [R]
D
-22.8°, was reported by Kuro-
yanagi et al. in 1971,25 and its absolute configuration at
C-4 was proposed as R through its dibenzoate by utilizing
an aromatic chirality method.
natural product, was assigned as 7,8-dimethoxy-6-hydroxy-
coumarin. This compound was reported previously to be a
synthetic substance, but the structure was elucidated only
1
29
Compound 6, mp 145-146 °C, was obtained as pale
yellowish needles, and the molecular formula was deter-
by H NMR data, which are somewhat inconsistent with
the values reported herein.
mined as C11
H
10
O
5
by HREIMS (obsd m/z 222.0535). This
Eight other compounds of known structure were isolated
from D. maritima sample A5248, as described in the
Experimental Section. A much smaller quantity of 1 was
obtained relative to sample A3919. Also, whereas analogues
of 1 that represented dimers of the parent compound were
found in the A3919 specimen, A5248 yielded monomeric
analogues that were methylated in an alternate manner,
compound exhibited UV absorption maxima at 209, 330,
and 383 nm, suggesting the presence of considerable
conjugation in the molecule. In the IR spectrum, absorption
-1
bands at 1690 (R,â-unsaturated ester), 1641 and 1546 cm
aromatic ring) were apparent. The 1H NMR spectrum
showed four isolated spin systems, inclusive of two methoxy
groups at δ 3.94 and 4.00 (each 3H, s), one pair of coupled
vinylic protons at δ 6.24 (1H, d, J ) 9.6 Hz, H-3) and 7.93
(
3
0
H
such as 2-methoxy-7-methyljuglone (7), 3-methoxy-7-
methyljuglone (8),30 and 7-methyljuglone (9). Lupeol and
scopoletin were common to both samples, and betulin and
methyl R-orcinolcarboxylate31 were identified only in A5248
by comparison of their physical and spectral data with
reported values. Among these isolates, 7, 8, and methyl
R-orcinolcarboxylate were obtained from this species for the
first time.
9
H
(
1H, dd, J ) 9.6 and 0.4 Hz, H-4), and one aromatic proton
at δ 6.71 (1H, d, J ) 0.4 Hz, H-5). A coumarin skeleton
was determined for 6 on the basis of the characteristic AB
H
1
spin system for H-3 and H-4 in the H NMR spectrum and
seven degrees of unsaturation. The two methoxy groups
in 6 were placed at C-7 and C-8 due to their appearance
at ca. δ
O-substituted groups in an aromatic ring, instead of at ca.
56, as found for an isolated methoxy or a methoxy group
neighboring only one O-substituted functionality. Fur-
thermore, the above assignments were supported by the
spatial correlations between H-4 and H-3 and H-5, along
C
61, typical for methoxy moieties neighboring two
All of the isolates were examined with a panel of
cytotoxicity and antimicrobial assays (Tables 2 and 3). With
respect to the former, 1 was potently cytotoxic against all
four of the cancer cell lines tested, and a preliminary
structure-activity relationship (SAR) may be inferred. On
the basis of data obtained with KB cells, for example,
compounds 1-3, which all have an allylic methyl group at
the 2 (and 2′) position(s), were more cytotoxic by ap-
proximately an order of magnitude than those naphtho-
quinones lacking such a moiety (4 and 7-9). Against the
δ
C
2
8
3 3
with OCH -7 and OCH -8 observed in NMR NOE difference
experiments, and 6,7,8-trimethoxycoumarin was generated
after compound 6 was treated directly with diazomethane
in diethyl ether solution.28 Accordingly, compound 6, a new

Cytotoxic Constituents of Diospyros maritima
J ournal of Natural Products, 2004, Vol. 67, No. 7 1159
Ta ble 3. Antimicrobial Activity of Constituents from the Bark of D. maritimaa-c
antibacterial activity (µg/mL)
antifungal activity (µg/mL)
Gram-positive
M. luteus S. aureus
mycobacteria
M. smegmatis M. avium
yeast
filamentous
C. albicans
C. neoformans
S. cerevisiae
A. niger
compd
C
I
C
I
C
I
C
I
C
I
C
I
C
I
C
I
1
2
3
4
7
8
9
600 80
>1000 200
600
300
100 >1000
300 >1000 600 >1000
20 >1000 600 40
90
50 >800 200
50 400 50
9
40
10
3
100
40
200
10
100
200
20
20
3
10
50
40
80
10
500
200
10
9
3
10
50
400
20
40
6
30
100
9
90
80
20
>1000
>1000
>800
>1000 10 >1000
>800 50 >800
6
5 >1000
6
200
400
800
3
800 50
400
800
>1000
>1000 700 >1000 >1000 >1000 >1000 >1000 700 >1000 >1000
1000 20 >1000
5000 200 5000
>1000 >1000
>1000 >1000
>1000 >1000
200
600
40
300
5
20
0.3
20
1
300
0.3
80
1
0.3
300
300
5
b
nitidine
300 >5000 1000 >5000 3000
200
3000
>5000 3000
a
C ) cidal ) 99.9% inhibition of growth; I ) inhibitory ) 50% inhibition of growth. b Used as a positive control against some yeast
c
species. Compounds 5 and 6, lupeol, methyl R-orcinolcarboxylate, and scopoletin were regarded as inactive (values >1000 µg/mL against
all organisms).
utilized for the DEPT-135 and 13C NMR spectra using a Bruker
5 mm QNP probe. Low-resolution ESIMS were determined on
a Finnigan LCQ instrument with an electrospray interface.
Column chromatography was carried out on silica gel 60 (70-
other three cell lines, 1 displayed consistently strong
cytotoxicity, although the proposed SAR did not necessarily
hold true. The new compound, 5, a naphthoquinone ana-
logue in which one of the quinone carbonyls was reduced
to an alcohol, was completely inactive, suggesting that an
2
30 mesh, Merck, Darmstadt, Germany), and fractions were
monitored via TLC (silica gel 60 F254 plates, 0.25 mm
thickness) visualized with vanillin-sulfuric acid in EtOH. A
Varian ProStar HPLC system was utilized for preparative-
scale HPLC.
intact quinone may be necessary for potency. Recent
studies have reported the chemopreventive effects of 1.32
Thus, the compound was evaluated for chemotherapeutic
efficacy via an in vivo antitumor hollow fiber assay.33,34 No
significant growth inhibition of tumor cells was observed
at either the intraperitoneal or subcutaneous sites, with
KB, LNCaP, or Lu1 cancer cell lines, when the mice were
treated with doses of 1 that ranged from 1.25 to 10 mg/kg.
Doses of 7.5 and 10 mg/kg were lethal to one and two out
of six mice, respectively, and a dose of 5 mg/kg caused
significant loss of body weight. In a similar fashion, 2 was
tested in the hollow fiber assay and also found to be
inactive. Interestingly, the mice could tolerate doses of 2
as high as 25 mg/kg, although death was observed after
two injections when the concentration of 2 was increased
to 50 mg/kg.
For sample A5248 (see below), melting points were deter-
mined on a Fisher-J ohns melting point apparatus and are
uncorrected. Optical rotations, UV spectra, CD spectra, and
IR spectra were measured with a Perkin-Elmer 241 polarim-
eter, a Beckman DU-7 spectrometer, a J ASCO-710 spectropo-
larimeter, and an ATI Mattson FT-IR spectrometer, respec-
tively. NMR spectra were recorded on a Bruker DPX-360 NMR
spectrometer. HREIMS and LREIMS were recorded on a
Finnigan MAT 95 (70 eV) and J EOL GC-mate II mass
spectrometer. Compounds were visualized by dipping TLC
plates into a phosphomolybdic acid reagent (Aldrich, Milwau-
kee, WI) followed by charring at 110 °C for 5-10 min.
P la n t Ma ter ia l. Two collections of the bark of D. maritima
were obtained in Indonesia, and voucher specimens have been
deposited at the Field Museum of Natural History, Chicago,
IL (accession no. A3919 and A5248). Sample A3919 was
collected at 5-10 m above sea level in J uly 1996 at Sukabumi,
West J ava, while A5248 was obtained at ∼350 m above sea
level in October 2001 at West Lombok County, Lombok Island.
Extr a ction a n d Isola tion . For sample A3919, the dried
bark (892 g) was percolated with hot MeOH overnight (2 × 3
L). This concentrated extract was partitioned between 10%
Compound 1 is well known to have strong antimicrobial
activity.9 Mahoney et al. found that 1 inhibited the
germination of spores of Aspergillus flavus more effectively
than juglone, an analogous naphthoquinone that lacks an
allylic methyl group at the 2 position. Against a related
fungus, A. niger, a clear SAR was not observed among the
series of structurally related naphthoquinones isolated in
this study, although 1 was one of the most potent com-
pounds tested against this fungus (Table 3). Against the
three yeast species, 9 was the most potent, while in the
antibacterial assays, 3 was the most potent. In general,
the isolated compounds displayed broad antimicrobial
activity. Of note was the strong antimycobacterial activity
observed against both slowly growing (Mycobacterium
avium) and rapidly growing (M. smegmatis) species. Fur-
ther, strong antimicrobial activity was detected against
pathogenic yeasts, Candida albicans and Cryptococcus
neoformans, and against the filamentous fungus A. niger.
,35
36
(
aqueous) MeOH and hexane, and then the aqueous MeOH
fraction was partitioned further between CHCl
-MeOH (4:1)
and H O. The organic layer was washed with 1% saline to
3
2
3
7
remove tannins. This organic fraction (51.2 g) was separated
further on a flash silica gel column using a gradient of 100%
3
hexane to 100% CHCl to 20% MeOH. The fractions were
combined into 13 pools based on TLC properties and examined
against the KB assay. Pools 2-4 were highly cytotoxic (0%
survival tested at 2 µg/mL), and no further purification of pool
2 was required to isolate compound 1 (2.27 g, yield 0.25%, 95%
pure by HPLC). A large percentage of pools 3 and 4 was
determined to be 1 as well. A small quantity of 1 was found
also in pools 5 and 6, contributing to their bioactivity, and the
inactive compound lupeol (41.5 mg, yield 0.047%) was isolated
from these pools using HPLC with an isocratic solvent system
of 100% MeOH on an 8 µm Inertsil ODS-3 column (10 mm i.d.
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. For sample A3919
(see below), melting points were determined on a Bristoline
hot-stage apparatus and are uncorrected. UV spectra were
measured on a Varian ProStar 330 photodiode array detector.
All NMR experiments were performed in CDCl with TMS as
3
×
50 mm guard and 20 mm i.d. × 250 mm column; Metachem
Technologies, Torrance, CA). Pool 7, having strong cytotoxicity
(
∼480 mg), was separated on a second flash silica gel column
an internal standard; gs-COSY, ROESY, gs-HSQC, gs-HMBC,
using a gradient of 100% hexane to 100% EtOAc to 20%
MeOH. Fractions were combined into 12 pools (7-1 through
7-12) based on TLC properties and tested versus the KB assay.
Pools 7-4 through 7-7 (∼150 mg) were combined and purified
1
and H NMR spectra were run on a Bruker AMX-500 instru-
ment using a Bruker 5 mm broad-band inverse probe with
z-gradient or a Nalorac 3 mm microinverse broad-band probe
with z-gradient, while a Bruker DPX-300 instrument was
2
via RP-HPLC using a gradient of 80:20 MeOH-H O up to

1
160 J ournal of Natural Products, 2004, Vol. 67, No. 7
Gu et al.
1
00% MeOH over 50 min using the aforementioned Metachem
(MeOH) ∆ꢀ216 +7.87, ∆ꢀ234 -1.55, ∆ꢀ271 -3.03, ∆ꢀ325 +0.44; UV
(MeOH) λmax (log ꢀ) 217 (4.19), 267 (3.95), 329 (3.51) nm; IR
column to isolate 3 (23.3 mg isolated here along with 23.5 mg
isolated from pools 8-2-5 and 8-2-6, mentioned below, total
yield 0.0052%). Another fraction from this HPLC experiment
-
1
1
(dried film) νmax 3526-3109, 1641, 1362, 1048 cm ; H and
1
3
+
C NMR, see Table 1; EIMS m/z 192 [M] (71), 164 (20), 149
+
(
∼26 mg) was purified further using a step MeOH-CHCl
3
(14), 135 (100), 107 (19); HREIMS m/z 192.0778 [M] (calcd
gradient that started at 0:100 up to 10:90 over 30 min,
remained isocratic for 30 min, then increased to 20:80 over 10
min and remained isocratic thereafter, using a 5 µm YMC-
Pack DIOL-120 NP column (20 mm i.d. × 250 mm column;
YMC Inc., Milford, MA) to give compounds 4 (3.4 mg, yield
12 3
for C11H O , 192.0790).
P r ep a r a tion of th e (R)- a n d (S)-MTP A Ester Der iva -
tives of 5. Compound 5 (1.5 mg) and 4-(dimethylamino)-
pyridine (0.2 mg) were transferred into a clean NMR tube, and
this mixture was dried under vacuum. Deuterated pyridine
(0.5 mL) and (S)-(+)-R-methoxy-R-(trifluoromethyl)phenyl-
acetyl chloride (6 µL) were added into the NMR tube im-
mediately under a N2 gas stream, then the NMR tube was
sealed and shaken to mix the sample and MTPA chloride
evenly. The reaction NMR tube was permitted to stand at room
temperature and monitored every 20 min by 1H NMR. (R)-
MTPA monoester derivative 5r was afforded completely after
3 h, and then the reaction was stopped by shaking the
unsealed NMR tube under air. Treatment of 5 (1.5 mg) with
(R)-(-)-R-methoxy-R-(trifluoromethyl)phenylacetyl chloride, as
0
.00038%) and 2 (16.4 mg isolated here along with 38.5 mg
isolated from pools 8-2-5 and 8-2-6, mentioned below, total
yield 0.0062%). Pool 8 from the initial column (∼15 g) was
separated on a second flash silica gel column using a gradient
of 100% hexane to 100% EtOAc to 60% MeOH to give 10 pools
(8-1 through 8-10). Scopoletin (1.2 g, yield 0.13%) was isolated
by filtering a MeOH suspension of pool 8-7 (∼3.7 g) through a
fritted funnel. Bioactive pool 8-2 (∼1.3 g) was separated on a
flash diol column using a gradient of 100% hexane to 100%
CHCl
1
3
3
to 100% MeOH to yield 13 pools (8-2-1 through 8-2-
3). Pool 8-2-5 (∼50 mg) was purified using a gradient of 70:
0 MeOH-H O up to 100% MeOH over 90 min on a 5 µm YMC
1
described above, yielded the (S)-MTPA ester derivative 5s. H
2
NMR data for 5s and 5r are presented in Table 1 (data were
ODS-A column (20 mm i.d. × 50 mm guard and 20 mm i.d. ×
obtained from the reaction NMR tube directly and were
1
1
2
50 mm column) to give compounds 2 and 3; the same
assigned on the basis of the correlations of the H- H COSY
spectrum).26
purification system was used on pool 8-2-6 (∼50 mg) to give 2
and 3 also.
7
,8-Dim eth oxy-6-h yd r oxycou m a r in (6): pale yellowish
For A5248, the plant material (500 g) was extracted by
maceration with MeOH (1 L × 3). The resultant extract was
needles; mp 145-146 °C; UV (MeOH) λmax (log ꢀ) 209 (4.50),
3
1
30 (4.04), 383 (3.45) nm; IR (dried film) νmax 3348, 1690, 1641,
diluted with H
2
O to afford an aqueous MeOH solution (90%)
-1
1
546 cm ; H NMR (CDCl
3
3
) δ 3.94 (3H, s, OCH -7), 4.00 (3H,
and then partitioned with n-hexane (0.8 L × 3). The lower
s, OCH
Hz, H-5), 7.93 (1H, dd, J ) 0.4, 9.6 Hz, H-4); C NMR (CDCl
-7), 61.9 (q, OCH -8), 99.2 (d, C-5), 107.5 (s,
3
-8), 6.24 (1H, d, J ) 9.6 Hz, H-3), 6.71 (1H, d, J ) 0.4
layer was concentrated and partitioned between 5% MeOH in
13
3
)
H
2
O (1 L) and CHCl
3
(1 L × 3). The CHCl
3
-soluble extract (10
δ 61.6 (q, OCH
3
3
g, ED50 4.4 µg/mL against the Lu1 cell line) was subjected to
silica gel column chromatography by elution with increasing
concentrations of MeOH in CHCl
Fractions 1, 2, and 3 were active when tested against the Lu1
cell line (ED50 0.19, 3.2, and 5.9 µg/mL, respectively). 7-Me-
thyljuglone (9, 6 mg) and 1 (4 mg) were purified from a portion
C-10), 112.6 (d, C-3), 136.8 (s, C-7), 139.2 (d, C-4), 148.9 (s,
C-8), 151.8 (s, C-9), 154.0 (s, C-6), 162.0 (s, C-2); EIMS m/z
to give six pooled fractions.
+
3
2
22 [M ] (100), 207 (99), 179 (41), 151 (19), 136 (20), 69 (59);
+
HREIMS m/z 222.0535 [M] (calcd for C11
10 5
H O , 222.0514).
P r ep a r a tion of 6,7,8-Tr im eth oxycou m a r in . 6,7,8-Tri-
methoxycoumarin (5 mg) was generated in 100% yield by
treating 6 directly for 10 min with diazomethane in diethyl
ether solution, which was prepared using Diazald KIT diazo-
methane generator (Aldrich-Sigma). Pale yellow powder; mp
(
12 mg) of fraction 1 (400 mg, CHCl
preparative TLC developed twice by n-hexane-EtOAc (20:1).
Further purification of fraction 2 (1.5 g, CHCl -MeOH, 45:1)
3
-MeOH, 50:1) using
3
over silica gel by elution with gradient mixtures of n-hexane-
EtOAc (stepwise, 40:1 to 8:1) gave three subfractions (i-iii).
Methyl R-orcinolcarboxylate (5 mg) was obtained from sub-
fraction i (27 mg) using preparative TLC developed twice by
n-hexane-CHCl -MeOH (6:1:0.4). Similarly, 2-methoxy-7-
methyljuglone (7, 10 mg), 3-methoxy-7-methyljuglone (8, 10
mg), and lupeol (25 mg) were purified from subfraction ii (65
mg) by preparative TLC developed twice by n-hexane-CHCl
EtOAc (3:1:0.4). In turn, fraction 3 (5.2 g, CHCl
28
1
13
1
00-102 °C (lit. 102-104 °C). The UV, H NMR, C NMR,
28,41
and EIMS data are consistent with literature values.
2
-Meth oxy-7-m eth ylju glon e (7): brown amorphous pow-
1
13
der; mp 171-173 °C (dec); UV, H NMR, C NMR, and EIMS
3
data, consistent with literature values.30
3
-Meth oxy-7-m eth ylju glon e (8): orange needles; mp 209-
42 1 13
2
11 (lit. 209-210 °C); UV, H NMR, C NMR, and EIMS
data, consistent with literature values.
3
-
30
3
-MeOH, 40:
, stepwise, n-hexane-acetone,
:1 to 1:1) to afford betulin (2.1 g), scopoletin (1.5 g), and a
7
-Meth ylju glon e (9): brown amorphous powder; mp 114-
1
9
) was further fractionated (SiO
2
43,44 1 13
1
16 °C (lit.
data, consistent with literature values.
116 °C); UV, H NMR, C NMR, and EIMS
9
,44
subfraction iv (240 mg, n-hexane-acetone, 6:1), which was
purified by preparative TLC developed twice by n-hexane-
acetone (3:2) to afford compounds 5 (15 mg) and 6 (36 mg).
Cytotoxic P oten tia l. All of the compounds were evaluated
for cytotoxicity against the KB (human oral epidermoid
carcinoma), Lu1 (human lung cancer), LNCaP (hormone-
dependent human prostate cancer), and HUVEC (human
umbilical vein endothelieal cells) cell lines using established
procedures.4
9
P lu m ba gin (1): orange needles; mp 75-76 °C (lit. 75-76
C); UV spectrum consistent with literature values.3 H, C,
8 1
13
°
and DEPT-135 NMR data are given in Supplemental Table 1
5,46
(see Supporting Information) and are consistent with literature
9
,10
values.
In Vivo Hollow F iber Assa y. Compounds 1 and 2 were
Ma r itin on e (2): orange-red amorphous powder; mp 197-
tested against this assay according to established protocols
9
33,34
1
98 °C (lit. 199-200 °C); UV spectrum consistent with
with some modification.
The compounds were regarded as
1
0 1
13
literature values. H, C, and DEPT-135 NMR data are given
in Supplemental Table 1 (see Supporting Information).
Ch itr a n on e (3): orange amorphous powder; mp 112-115
inactive if <50% cell growth inhibition was observed when
compared with PBS control.
Mea su r em en t of An tim icr obia l Activities. Antimicrobial
activity of pure compounds was measured by a zone-of-
3
9
°
C (lit. 116-118 °C); UV spectrum consistent with literature
23 1 13
47
values.
H, C, and DEPT-135 NMR data are given in
inhibition assay against strains of Micrococcus luteus, Sta-
Supplemental Table 2 (see Supporting Information).
Zeyla n on e (4): orange amorphous powder; mp 208 °C (lit.
phylococcus aureus, Mycobacterium avium, M. smegmatis,
Saccharomyces cerevisiae, Cryptococcus neoformans, Candida
albicans, and Aspergillus niger. The minimal inhibitory con-
centration (MIC) was defined as the lowest concentration
resulting in inhibition of growth; the term “cidal” was used
for 99.9% inhibition, which was evidenced by clearing, and the
term “inhibitory” was used for 50% inhibition, which was
evidenced by a hazy spot.
4
0
2
08 °C); UV spectrum consistent with literature values.24 1H,
1
3
C, and DEPT-135 NMR data are given in Supplemental
Table 2 (see Supporting Information).
S,8-Dih yd r oxy-6-m et h yl-1-t et r a lon e [(4S)-sh in a n ol-
4
2
5
25
on e] (5): pale yellowish powder; [R] 589 +14°, [R] 578 +17°,
2
5
25
25
[R] 546 +21°, [R] 436 +34°, [R] 365 +29° (c 0.1, MeOH); CD nm

Cytotoxic Constituents of Diospyros maritima
J ournal of Natural Products, 2004, Vol. 67, No. 7 1161
Ack n ow led gm en t. This investigation was supported in
part by grant U19-CA-52956, funded by the National Cancer
Institute, NIH, Bethesda, MD. D.L. and N.H.O. gratefully
acknowledge partial support via a Research Scholar Grant to
N.H.O. from the American Cancer Society (RSG-02-024-01-
CDD). We thank Dr. K. Fagerquist, Mass Spectrometry
Facility, Department of Chemistry, University of Minnesota,
Minneapolis, MN, and Dr. A. Anderson, Research Resources
Center, University of Illinois at Chicago, for obtaining the mass
spectral data. We are grateful to the Nuclear Magnetic
Resonance Laboratory and Protein Research Facility of the
Research Resources Center, University of Illinois at Chicago,
for the provision of certain NMR and CD spectral facilities used
in this investigation.
(21) Shen, Z.; Dong, Z.; Cheng, P.; Li, L.; Chen, Z.; Liu, J . Planta Med.
2
003, 69, 605-609.
(
22) Kinghorn, A. D.; Farnsworth, N. R.; Soejarto, D. D.; Cordell, G. A.;
Pezzuto, J . M.; Udeani, G. O.; Wani, M. C.; Wall, M. E.; Navarro, H.
A.; Kramer, R. A.; Menendez, A. T.; Fairchild, C. R.; Lane, K. E.;
Forenza, S.; Vyas, D. M.; Lam, K. S.; Shu, Y. Z. Pure Appl. Chem.
1
999, 71, 1611-1618.
(
(
(
(
23) Sankaram, A. V. B.; Srinivasarao, A.; Sidhu, G. S. Phytochemistry
1976, 15, 237-238.
24) Sankaram, A. V. B.; Rao, A. S.; Shoolery, J . N. Tetrahedron 1979,
3
5, 1777-1782.
25) Kuroyanagi, M.; Yoshihira, K.; Natori, S. Chem. Pharm. Bull. 1971,
19, 2314-2317.
26) Su, B. N.; Park, E. J .; Mbwambo, Z. H.; Santarsiero, B. D.; Mesecar,
A. D.; Fong, H. H. S.; Pezzuto, J . M.; Kinghorn, A. D. J . Nat. Prod.
2
002, 65, 1278-1282.
(27) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J . Am. Chem.
Soc. 1991, 113, 4092-4095.
(28) Tsukamoto, H.; Hisada, S.; Nishibe, S. Chem. Pharm. Bull. 1985, 33,
4
069-4073.
Su p p or tin g In for m a tion Ava ila ble: Tables of NMR data for
compounds 1-4 and a figure that graphically represents the key
HMBC correlations for compound 4 are available free of charge via
the Internet at http://pubs.acs.org.
(
29) Kang, K. S.; Ahn, B. Z. Arch. Pharm. Res. 1986, 9, 115-117.
(30) Budzianowski, J . Phytochemistry 1995, 40, 1145-1148.
(31) Lee, S. S.; Chang, S. M.; Chen, C. H. J . Nat. Prod. 2001, 64, 1548-
1551.
(
32) Sugie, S.; Okamoto, K.; Rahman, K. M.; Tanaka, T.; Kawai, K.;
Yamahara, J .; Mori, H. Cancer Lett. 1998, 15, 177-183.
33) Mi, Q.; Lantvit, D.; Reyes-Lim, E.; Chai, H.; Zhao, W.; Lee, I. S.;
Peraza-Sanchez, S.; Ngassapa, O.; Kardono, L. B.; Riswan, S.;
Hollingshead, M. G.; Mayo, J . G.; Farnsworth, N. R.; Cordell, G. A.;
Kinghorn, A. D.; Pezzuto, J . M. J . Nat. Prod. 2002, 65, 842-850.
(34) Hollingshead, M. G.; Alley, M. C.; Camalier, R. F.; Abbott, B. J .; Mayo,
J . G.; Malspeis, L.; Grever, M. R. Life Sci. 1995, 57, 131-141.
(35) Tripathi, R. D.; Tripathi, S. C.; Dixit, S. N. Agric. Biol. Chem. 1980,
44, 2483-2485.
(36) Mahoney, N.; Molyneux, R. J .; Campbell, B. C. J . Agric. Food Chem.
2000, 48, 4418-4421.
(37) Wall, M. E.; Wani, M. C.; Brown, D. M.; Fullas, F.; Oswald, J . B.;
J osephson, F. F.; Thornton, N. M.; Pezzuto, J . M.; Beecher, C. W.
W.; Farnsworth, N. R.; Cordell, G. A.; Kinghorn, A. D. Phytomedicine
1996, 3, 281-285.
(38) Boisvert, L.; Brassard, P. J . Org. Chem. 1988, 53, 4052-4059.
(39) Gunaherath, G. M. K. B.; Gunatilaka, A. A. L.; Thomson, R. H. J .
Chem. Soc., Perkin Trans. 1 1988, 407-410.
(40) Dinda, B.; Das, S. K.; Hajra, A. K. Indian J . Chem. Sect. B 1995, 34,
525-528.
Refer en ces a n d Notes
(
(
1) Higa, M.; Noha, N.; Yokaryo, H.; Ogihara, K.; Yogi, S. Chem. Pharm.
Bull. 2002, 50, 590-593.
(
(
(
(
(
(
(
2) Meyer, T. M. Rec. Trav. Chim. 1947, 66, 193-198.
3) Chang, C. I.; Kuo, Y. H. Chem. Pharm. Bull. 1998, 46, 1627-1629.
4) Spies, J . R. J . Econ. Entomol. 1933, 26, 285-288.
5) Khan, M. R.; Timi, D. Fitoterapia 1999, 70, 194-196.
6) Khan, M. R. Fitoterapia 1998, 64, 471-473.
7) Kuo, Y. H.; Chang, C. I. Chem. Pharm. Bull. 2000, 48, 1211-1214.
8) Kuo, Y. H.; Chang, C. I.; Kuo, Y. H.; Huang, S. L. J . Chin. Chem.
Soc. 1998, 45, 111-114.
(9) Higa, M.; Ogihara, K.; Yogi, S. Chem. Pharm. Bull. 1998, 46, 1189-
1
193.
(
(
(
(
10) Tezuka, M.; Takahashi, C.; Kuroyanagi, M.; Satake, M.; Yoshihira,
K.; Natori, S. Phytochemistry 1973, 12, 175-183.
11) Yoshihira, K.; Tezuka, M.; Takahashi, C.; Natori, S. Chem. Pharm.
Bull. 1971, 19, 851-854.
12) Tikkanen, L.; Matsushima, T.; Natori, S.; Yoshihira, K. Mutat. Res.
1
983, 124, 25-34.
13) Higa, M.; Himeno, K.; Yogi, S.; Hokama, K. Chem. Pharm. Bull. 1987,
(41) Greger, H.; Hofer, O.; Robien, W. J . Nat. Prod. 1983, 46, 510.
(42) Giles, R. G. F.; Roos, G. H. P. J . Chem. Soc., Perkin Trans. 1 1976,
2057-2060.
3
5, 4366-4367.
(
14) Higa, M. Chem. Pharm. Bull. 1988, 36, 3234-3236.
15) Kuo, Y. H.; Chang, C. I.; Li, S. Y.; Chou, C. J .; Chen, C. F.; Kuo, Y.
H.; Lee, K. H. Planta Med. 1997, 63, 363-365.
(
(43) Luckner, B. Pharmazie 1970, 25, 261-263.
(44) Krohn, K.; Broser, E. J . Org. Chem. 1984, 49, 3766-3771.
(45) Likhitwitayawuid, K.; Angerhofer, C. K.; Cordell, G. A.; Pezzuto, J .
M.; Ruangrungsi, N. J . Nat. Prod. 1993, 56, 30-38.
(46) Oberlies, N. H.; Burgess, J . P.; Navarro, H. A.; Pinos, R. E.; Soejarto,
D. D.; Farnsworth, N. R.; Kinghorn, A. D.; Wani, M. C.; Wall, M. E.
J . Nat. Prod. 2001, 64, 497-501.
(
16) Kuo, Y. H.; Chang, C. I.; Kuo, Y. H. J . Chin. Chem. Soc. 1996, 43,
5
11-514.
17) Kuo, Y. H.; Huang, S. L.; Chang, C. I. Phytochemistry 1998, 49, 2505-
507.
(
2
(
(
18) Chang, C. I.; Kuo, Y. H. J . Nat. Prod. 1999, 62, 309-310.
19) Kuo, Y. H.; Chang, C. I.; Kuo, Y. H. Chem. Pharm. Bull. 1997, 45,
(47) Cain, C. C.; Henry, A. T.; Waldo, R. H., III; Casida, L. J ., J r.;
Falkinham, J . O., III. Appl. Environ. Microbiol. 2000, 66, 4139-4141.
1
221-1222.
20) Kuo, Y. H.; Chang, C. I.; Kuo, Y. H. Phytochemistry 1997, 46, 1135-
137.
(
1
NP040027M