Antimycobacterial eudesmanolides from Inula helenium and Rudbeckia subtomentosa

Article abstract of DOI:10.1055/s-1999-14001

In a bioassay guided search for antimycobacterial compounds from higher plants, the root extracts of Elecampane (Inula helenium L.; Asteraceae) and Sweet Coneflower (Rudbeckia subtomentosa Pursh.; Asteraceae) were chemically investigated for their active constituents. Chromatographic fractions of root extracts of I. helenium, which exhibited significant activity against Mycobacterium tuberculosis, provided the known eudesmanolides alantolactone, isoalantolactone, and 11αH,13-dihydroisoalantolactone. Peracid epoxidation of alantolactone and isoalantolactone provided 5α-epoxyalantolactone and 4(15)α-epoxyisoalantolactone, respectively and oxidation of alantolactone with OsO₄ gave 11,13-dihydroxyalantolactone. Active fractions from R. subtomentosa contained the known alloalantolactone and 3- oxoalloalantolactone. The structures of the above compounds were established by spectroscopic methods including 1D and 2D NMR techniques as well as spectral comparison with previously reported data. The molecular structure of 5α-epoxyalantolactone was determined by single crystal X-ray diffraction. Eleven natural and semisynthetic eudesmanolides were tested in a radiorespirometric bioassay for activity against M. tuberculosis. 5α- Epoxyalantolactone and encelin from Montanoa speciosa showed minimum inhibitory concentrations (MICs) of 8 and 16 μg ml^-1, respectively. Alantolactone, isoalantolactone and its 4α,15-epoxide, 1,2-dehydro-3-epi- isotelekin and alloalantolactone gave MICs of 32 μg ml^-1. All other compounds showed MIC values of 128 μg ml^-1 or higher.

Full text of DOI:10.1055/s-1999-14001

Original Paper
351
Antimycobacterial Eudesmanolides from Inula helenium and
Rudbeckia subtomentosa
Charles L. Cantrell1, Laura Abate1, Frank R. Fronczek1, Scott G. Franzblau2, Leovigildo Quijano3,
and Nikolaus H. Fischer1,*
1 Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana, U.S.A.
2 U.S. Department of Health and Human Services, Laboratory Research Branch, GWL Hansenꢀs Disease Center, Baton Rouge, Louisiana, U.S.A.
3 Instituto de Quimica, Universidad Autonoma de Mexico, Mexico D. F., Mexico
Received: August 21, 1998; Revision accepted: January 16, 1999
plant extracts with significant in vitro antimycobacterial
activity (3), followed by a bioassay-guided isolation of active
constituents (3±6). For instance, crude flower extracts of the
Sea Daisy (Borrichia frutescens) exhibited significant activity,
which from active chromatographic fractions provided anti-
mycobacterial cycloartane-type triterpenes with minimum
inhibitory concentrations (MICs) of 710 mg ml21 (6).
Abstract: In a bioassay guided search for antimycobacterial
compounds from higher plants, the root extracts of Elecampane
(Inula helenium L.; Asteraceae) and Sweet Coneflower (Rudbeck-
ia subtomentosa Pursh.; Asteraceae) were chemically investi-
gated for their active constituents. Chromatographic fractions
of root extracts of I. helenium, which exhibited significant activ-
ity against Mycobacterium tuberculosis, provided the known eu-
desmanolides alantolactone, isoalantolactone, and 11aH,13-di-
hydroisoalantolactone. Peracid epoxidation of alantolactone
and isoalantolactone provided 5a-epoxyalantolactone and
4(15)a-epoxyisoalantolactone, respectively and oxidation of
alantolactone with OsO4 gave 11,13-dihydroxyalantolactone. Ac-
tive fractions from R. subtomentosa contained the known allo-
alantolactone and 3-oxoalloalantolactone. The structures of the
above compounds were established by spectroscopic methods
including 1D and 2D NMR techniques as well as spectral com-
parison with previously reported data. The molecular structure
of 5a-epoxyalantolactone was determined by single crystal X-
ray diffraction. Eleven natural and semisynthetic eudesmano-
lides were tested in a radiorespirometric bioassay for activity
against M. tuberculosis. 5a-Epoxyalantolactone and encelin from
Montanoa speciosa showed minimum inhibitory concentrations
(MICs) of 8 and 16 mg ml21, respectively. Alantolactone, isoalan-
tolactone and its 4a,15-epoxide, 1,2-dehydro-3-epi-isotelekin
and alloalantolactone gave MICs of 32 mg ml21. All other com-
pounds showed MICvalues of 128 mg ml21 or higher.
In continuation of our search for new structural types of
antimycobacterial natural products, active root extracts of
Inula helenium and Rudbeckia subtomentosa were chemically
investigated. Native Americans (Iroquois, Cherokees, and
Mohegans) used infusions and decoctions of I. helenium roots
for the treatment of lung disorders and against tuberculosis
(7). Since at 100mg ml21, the crude root extracts of I. helenium
and R. subtomentosa exhibited 100 and 99 percent inhibition,
respectively against M. tuberculosis, the active fractions were
chemically investigated for their active constituents.
Materials and Methods
1H- and 13C-NMR spectra were recorded in CDCl3 on a Bruker
ARX 300 spectrometer at 300 MHz (1H-NMR) and 75 MHz
(
13C-NMR). Mass spectra were obtained on a Hewlett-Packard
5971A GC-MS instrument. IR spectra were run on a Perkin-
Elmer 1760X spectrometer as a film on KBr plates. Vacuum-
liquid chromatographic (VLC) separations (8) were carried out
on silica gel (MN Kieselgel).
Key words: Inula helenium, Rudbeckia subtomentosa, R. mollis,
Montanoa speciosa, Asteraceae, sesquiterpene lactones, eudes-
manolides, Mycobacterium tuberculosis, antituberculosis activ-
ity.
Plant material
Roots of I. helenium L. were collected on June 25, 1995 and
obtained from Mr. George Sturtz of Aromagen, 31787 Peoria
Road, Albany, Oregon 97321, U.S.A. A voucher (Sturtz-Fischer
No 570) is deposited at the Louisiana State University
Herbarium. The crude plant extract of R. subtomentosa Pursh.
from a previous chemical study (9) was obtained from our
repository. The voucher (Cox No 4923) is deposited at
Louisiana State University Herbarium, U.S.A.
Introduction
Worldwide, the number of tuberculosis cases is currently on
the rise and it is estimated that the number of new infections
by Mycobacterium tuberculosis is greater than 8 million
annually and more than 3 million people die of this disease
each year (1), creating a need for the discovery and develop-
ment of new and more effective antituberculosis drugs (2).
Our previous research efforts focused on a search for crude
Extraction and isolation
Small pieces of fresh roots (4.5 kg) of I. helenium were dried at
room temperature for two weeks and then soaked in hexane
(8.5 l) for 24 hours. The solvent was decanted from the plant
residue and evaporated in vacuo to yield 15.2 g of crude
Planta Medica 65 (1999) 351±355
ꢀ Georg Thieme Verlag Stuttgart New York
·

352
Planta Med. 65 (1999)
Charles L. Cantrell, Laura Abate, Frank R. Fronczek, Scott G. Franzblau, Leovigildo Quijano, and Nikolaus H. Fischer
indicated that all of 1 had reacted (1.5 hours). The solution
was washed with 10 ml of 10 % aqueous NaHCO3 followed by
10 ml of H2O. The organic phase was dried with anhydrous
MgSO4, evaporated and the residue was separated by prepara-
tive TLC (SiO2, 1 mm) using hexane-EtOAc (5 : 2) as the mobile
phase to yield 103 mg of 7. Lactone 2a was converted to
epoxide 6 by a previously described method (13).
5a-Epoxyalantolactone (7): C15H20O3 (Mr: 248) colorless crys-
tals (EtOAc); m.p. 164±166 ꢁC; [a]2D5 : 184.2ꢁ (c, 0.0057,
CHCl3); IR (KBr): nmax = 1741 (C=O) cm21; EI-MS (70 eV): m/z
(% rel. int.) = 248 (M1, 9), 233 (M1 2 Me, 7), 204 (14), 192 (6),
159 (6), 149, (28), 133 (11), 126 (30), 123 (34), 109 (86), 95
(44), 81 (100), 67 (74), 55 (73); 1H-NMR and 13C-NMR see
Table 3.
OsO4 oxidation of 1
To an ice-cooled solution of 135 mg of a 50 % aqueous solution
of 4-methylmorpholine N-oxide was added a solution of
0.33 mg of OsO4 in 0.33 ml t-BuOH (14). Lactone 1 (126 mg),
dissolved in 3 ml of acetone, was added and the reaction
mixture stirred for 18 hours. A solution of aqueous 10 %
NaHSO3 (3 ml) was added and all solvents evaporated on a
rotary evaporator. The crude mixture (136 mg) was absorbed
on silica gel (208 mg) and separated on a VLC column (2.3 cm
in diameter) packed with 5 g of silica gel. Elution of the
column with 50 ml fractions of hexane and hexane-EtOAc
mixtures of increasing polarity resulted in pure 8 (102 mg),
which eluted with hexane-EtOAc (94 : 6).
extract. The residual plant material was resoaked for 24 hours
in CH2Cl2 (7.8 l) and subsequently extracted for 24 hours in
MeOH (7.8 l), yielding 23.3 g and 85.2 g, respectively. The
CH2Cl2 extract (23.3 g) was adsorbed onto silica gel (15.2 g)
and separated by VLC (8) into 11 fractions using solvents of
increasing polarity (hexane, ethyl acetate, and MeOH) as
summarized in Table 2. Fractions 2±4 were further separated
by repeated VLC procedures on silica gel, as described
previously (10, 11), to yield pure compounds 1, 2a, and 3.
11,13-Dihydroxyalantolactone (8): C15H22O4 (Mr: 266) colorless
oil; [a]2D5: 227.9ꢁ (c, 0.0068, CHCl3); IR (KBr): nmax = 3388 (OH)
cm21; EI-MS (70 eV): m/z (% rel. int.) = 266 (M1, 16), 251 (M1
2 Me, 1), 235 (10), 221 (6), 215 (19), 207 (3), 191 (12), 162
(100), 147 (33), 133 (15), 119 (25), 105 (96), 91 (89), 77 (30), 67
(18), 55 (37); 1H-NMR and 13C-NMR see Table 3.
Compounds 2b, 2c, and 4
Isoalloalantolactone (4) had been previously obtained from
Rudbeckia mollis Ell. (15). Encelin (2b) and 1,2-dehydro-3-epi-
isotelekin (2c) were isolated from Montanoa speciosa DC.,
Asteraceae (16).
Previously obtained CH2Cl2 root extract (1.75 g) of R. subto-
mentosa (9) was adsorbed on 1.6 g silica gel and placed onto a
VLC column (2.3 cm in diameter) packed with 25 g of silica
gel. The extract was separated into 8 fractions of increasing
polarity using hexane, EtOAc, and MeOH and mixtures there-
of, as listed in Table 2. Fraction 2 (120 mg) was adsorbed on
200 mg of silica gel and placed onto a VLC column (1.3 cm in
diameter) containing 5 g of silica gel. Subsequent elution
using 10 ml fractions of hexane-EtOAc of increasing polarity
afforded 17 mg of alloalantolactone (5a), which eluted with
hexane-EtOAc (95 : 5). Fraction 3 (88 mg) was separated as
described above for fraction 2 to afford 7 mg of compound 5a.
Fraction 4 (167 mg) was adsorbed on 240 mg of silica gel and
placed onto a VLC column (1.3 cm in diameter) packed with
6 g of silica gel. Elution of the column with 10 ml fractions of
hexane-EtOAc of increasing polarity afforded 19 mg of pure
5b, which eluted with hexane-EtOAc (3 : 1).
X-Ray crystallographic analysis of 5a-epoxyalantolactone (7)
A colorless fragment of dimensions 0.62 3 0.35 3 0.35 mm
was used for data collection at T = 100 K on an Enraf-Nonius
CAD4 diffractometer equipped with MoKa radiation (l =
0.71073 Š) and a graphite monochromator. The temperature
of the sample was maintained by an Oxford Cryostream
device. Crystal data are: C15H20O3, Mr = 248.3, monoclinic
space group P21, a = 7.968 (6), b = 6.181 (2), c = 13.132 (4) Š, b
= 105.58 (6)ꢁ, V = 623.0 (8) Š3, Z = 2, dc = 1.324 g cm23
.
Intensity data were measured by v-2q scans of variable rate.
Two quadrants of data were collected within the limits 2 7 q
7 30ꢁ. Data reduction included corrections for background,
Lorentz, and polarization effects. 3938 Intensities were aver-
aged to yield 1974 unique data (Rint = 0.019), of which 1932
had I 6 0 and were used in the refinement. The structure was
solved by direct methods using SIR92 (17) and refined by full-
matrix least squares, treating nonhydrogen atoms anisotropi-
Epoxidations of 1 and 2 a
Lactone 1 (101 mg) in 10 ml of CH2Cl2, was added to a
solution of 10 ml of CH2Cl2 containing 1.2 equivalents m-
chloroperbenzoic acid (m-CPBA), and stirred at 0 ꢁC until TLC

353
Antimycobacterial Eudesmanolides from Inula helenium and Rudbeckia subtomentosa
Planta Med. 65 (1999)
cally. Hydrogen atoms were located in difference maps and
refined isotropically. Convergence was achieved with R =
0.031, Rw = 0.036, and GOF = 1.687. The Enraf-Nonius MolEN
programs (18) were used for all computations. The structure
of 7 is illustrated in Figure 1. A full list of crystallographic data
and parameters including fractional coordinates is deposited
at the Cambridge Crystallographic Data Center, University
Chemical Laboratory, 12 Union Road, Cambridge, CB2, 1EZ, UK.
Table 1 Percent inhibition of crude plant extracts of Inula helenium
and Rudbeckia subtomentosa against Mycobacterium tuberculosis
(H37Rv) at 1000 and 100 mg ml21
.
Species
plant part extract. 1000 mg ml21 100 mg ml21
solv.*
I. helenium
roots
roots
roots
H
D
M
D
D
D
D
100
100
100
100
96
100
100
83
99
43
±
Radiorespirometric bioassays
R. subtomentosa roots
leaves
stems
flowers
Bioassays were performed essentially as described previously
(6, 12). Experiments for M. tuberculosis were usually com-
pleted within ten days; rifampin was used as a positive
control with an MIC of 0.25±0.125mg ml21. The MIC values
(mg ml21) for lactones 1±8 are 1: 32; 2a: 32; 2b: 16; 2c: 32;
3: 6128; 4: 128; 5a: 32; 5b: 128; 6: 32; 7: 8; 8: 6128.
70
95
30
* H = hexane; D = dichloromethane; M = methanol.
a minor constituent. The structures of 5a and 5b were
determined by spectral comparison with previously reported
data (11, 21). Both 5a and 5b had been previously isolated
from Eupatorium quadrangularae (21), but this is the first
report of their isolation from R. subtomentosa.
Results and Discussion
The results of radiorespiratory antituberculosis bioassays of
crude extracts of I. helenium and R. subtomentosa are pre-
sented in Table 1. The hexane and CH2Cl2 root extracts of
I. helenium demonstrated 100 percent inhibition at 100mg
ml21. VLC procedures of the CH2Cl2 root extract of I. helenium
and subsequent bioassay of the fractions against M. tuber-
culosis indicated that the nonpolar fractions 2±4 were the
most active, all three fractions giving 99 percent inhibitions at
33mg ml21 (Table 2). Chemical investigation of these fractions
resulted in the isolation of the known eudesmanolides
alantolactone (1) (19, 20), isoalantolactone (2a) (19, 20), and
11,13-dihydroisoalantolactone (3) (20), as shown by spectro-
scopic comparison with data previously reported for these
three sesquiterpene lactones.
In our previous study of cycloartane-type triterpenes from
Borrichia frutescens (6), epoxides were more active against
M. tuberculosis than their corresponding alkene analogs, sug-
gesting that epoxidation of the C-5 double bond of 1 or the C-
4 (15) double bond of 2a may result in derivatives of higher
activity.
Epoxidation of 1 gave epoxide 7 which was identified by MS,
1H-, 13C-NMR, including 90ꢁ and 135ꢁ DEPT methods as well
as comparison of spectral data with values previously
reported for structurally related lactones (11, 20). The molec-
ular structure of 7 was confirmed by single crystal X-ray
diffraction (Fig.1).
Bioassays of CH2Cl2 extracts of various plant parts of R. sub-
tomentosa indicated that root extracts were most active
against M. tuberculosis with a 99 percent inhibition at 100mg
ml21, which after VLC fractionations gave eight fractions, their
activities being summarized in Table 2. Further VLC separa-
tion of the most active fractions 2 and 3 provided the known
alloalantolactone (5a) and fraction 4 afforded the known 3-
oxoalloalantolactone (5b) as the major component and 5a as
Epoxidation of 2a provided epoxide 6, which had been
previously isolated (13) and synthesized (22). Its structure
was determined by spectral comparison, mainly of 1H-NMR
with previously reported values (13), and was shown to be
essentially identical.
Table 2 Percent inhibition of Inula helenium and Rudbeckia subtomentosa fractions against M. tuberculosis (H37Rv).
Fraction
solvent percentages
(hexane : EtOAc : MeOH)
I. heleniuma
I. helenium
R. subtomentosa
R. subtomentosab
100 mg ml21
33 mg ml21
100 mg ml21
33 mg ml21
1
2
100 : 0 : 0
95 : 5 : 0
75 : 25 : 0
85 : 15 : 0
80 : 20 : 0
70 : 30 : 0
50 : 50 : 0
0 : 100 : 0
100 : 0 : 0
95 : 5 : 0
100
100
99
99
99
92
90
220
214
2310
219
94
99
79
100
91
99
5326
26
28
70
±
43
97
390 : 10 : 0
4
5
6
7
8
9
10
11
100
99
50 : 50 : 0
25 : 75 : 0
0 : 100 : 0
0 : 50 : 50
0 : 0 : 100
±
99
96
84
80
29
210
67
24
23
44
±
0 : 50 : 50
0 : 50 : 50
0 : 0 : 100
±
±
±
±
±
212
±
a
Using a column (6.5 cm in diameter) with 200 ml of solvent for each fraction.
Using a column (2.3cm in diameter) with 100 ml of solvent for each fraction.
b

354
Planta Med. 65 (1999)
Charles L. Cantrell, Laura Abate, Frank R. Fronczek, Scott G. Franzblau, Leovigildo Quijano, and Nikolaus H. Fischer
(2 c) showed MICs of 16mg ml21 and 32mg ml21, respectively.
Lactones 1, 2a, and 5a gave values of 32mg ml21 while
isoalloalantolactone (4) from R. mollis (15) and compound 5b
exhibited MICs of 128mg ml21
.
Previous structure-activity studies within a series of natural
and semisynthetic germacranolides suggest that the a-meth-
ylene-g-lactone moiety is an essential, but not sufficient,
structural requirement for significant in vitro activity against
M. tuberculosis (23). The necessity of the presence of an a-
methylene-g-lactone group is supported by the activity of
compound 2a with a MIC of 32mg ml21, when compared with
the inactive 11aH,13-dihydroderivative 3 with a value of
6128 mg ml21. The presence of a second alkylating site, such
as an a,b-unsaturated carbonyl group or an epoxide function
together with a moderate to high lipophilicity (24), seems to
enhance the in vitro antimycobacterial activity of sesquiter-
pene lactones. For instance, encelin (2b) is more active than
2a and 5a-epoxyalantolactone (7), the most active lactone
within this eudesmanolide series with a MIC of 8mg ml21, is
significantly more active than its precursor 1 with an MIC of
32mg ml21. Both the 4a,15-epoxide 6 and its precursor 2a
gave MICs of 32mg ml21, which suggests that the distance
between the two alkylating sites and possibly the stereo-
chemical relationship of the methylene lactone moiety and
the epoxide, seems to influence the antimycobacterial activ-
ity. In contrast to 1 with an MIC of 32mg ml21, its more polar
11,13-diol derivative (8), showed no biological activity with a
MIC of 6128 mg ml21. This is most likely due to the increased
polarity and/or the loss of the lactonic alkylating site.
Fig.1 Molecular structure of 7 at 100 K.
In an attempt to synthesize the 5,6-diol derivative of lactone
1, it was reacted with OsO4 (14). The 1H-NMR spectrum of
product 8 indicated that the olefinic H-6 doublet of 1 at d 5.11
was nearly unchanged (d 5.03) in 8. Instead, the two
diagnostic exocyclic methylene lactone signals (H-13) were
missing in 8, suggesting that glycol formation had occurred
exclusively at the 11,13-rather than at the 5,6-position. 1H-
NMR values of 8 are summarized in Table 3 with peak
assignments based on spectral comparison with data of
structurally related analogs (10, 11, 19, 20). Attempts to
determine the relative configuration of C-11 in 8 by NMR
experiments were inconclusive.
The eudesmanolides isolated from I. helenium, R. subtomento-
sa, lactone
4 from R. mollis (15), and lactones 2 b, 2c
previously obtained from M. speciosa (16), as well as their
semisynthetic derivatives, 6, 7, and 8, were tested for their
biological activities against M. tuberculosis. The M. speciosa
constituents, encelin (2b) and 1,2-dehydro-3-epi-isotelekin
Acknowledgements
Financial support from the National Institute of Allergy and
Infectious Diseases, National Institutes of Health under intra-
Table 3 1H-NMR (300 MHz) and 13C-NMR (75.4 MHz) spectral data of compounds 7 and 8 (CDCl3).a
Atom
7
8
1H-NMR
13C-NMR
1H-NMR
13C-NMR
1
2
3a
3b
4
5
6b
7a
8a
9a
9b
10
11
±
±
±
±
37.7 t*
16.5 t
29.5 t
±
37.1 d**
67.5 s
61.2 d
37.4 d**
75.2 d
39.6 t*
±
32.6 s
136.7 s
169.7 s
123.8 t
±
24.0 q
18.1 q
±
±
±
42.3 t*
16.9 t
32.9 t
±
38.6 d
152.9 s
112.5 d
44.9 d
78.0 d
42.6 t*
±
33.0 s
77.4 s
177.9 s
64.2 t
±
28.8 q
23.1 q
±
1.54 m
1.54 m
2.47 m
±
5.03d (3.5)
2.98 dd (3.5, 5.6)
5.12 ddd (3.0, 3.0, 5.6)
1.54 dd (3.0, 15.0)
2.14 dd (3.0, 15.0)
±
1.34 m
±
2.89 d (2.6)
3.66 dddd (2.5, 2.6, 2.9, 8.8)
4.66 ddd (1.9, 4.5, 8.8)
1.55 dd (1.9, 15.0)
1.87 dd (4.5, 15.0)
±
±
±
5.77 d (2.5)
6.39 d (2.9)
1.10 s
1.03d (2.7)
±
±
±
12
13a
13b
14
15
11-OH
13-OH
3.72 br d (12.0)
3.93 d (12.0)
1.22 s
1.12 d (7.6)
3.77 br s
±
±
2.93 br d (9.9)
±
a
Expressed as d values in ppm, with J values in Hz in parentheses.
* Assignments in the same column are interchangeable.

355
Planta Med. 65 (1999)
Antimycobacterial Eudesmanolides from Inula helenium and Rudbeckia subtomentosa
agency agreement Y1-AI-5016 is gratefully acknowledged. We
thank Anita N. Biswas and Ken E. White for technical
assistance and George Sturtz for generously providing I. helen-
ium roots.
Prof. Dr. Nikolaus H. Fischer
Department of Chemistry
Louisiana State University
Baton Rouge
Louisiana 70803
U.S.A.
References
E-mail: fischer@chem.lsu.edu
Fax: 11 225-388-2695
1
Sepkowitz, K. A., Raffalli, J., Riley, L., Kiehn, T. E., Armstrong, D.
(1995) Clinical Microbiology Reviews 8, 180±199.
Cohn, D. L., Bustreo, F., Raviglione, M. C. (1997) Clinical Infectious
2
Diseases 24 (Suppl. 1), S121±130.
Cantrell, C. L., Fischer, N. H., Urbatsch, L., Franzblau, S. G. (1998)
3
Phytomedicine 5, 139±147.
Rajab, M. S., Cantrell, C. L., Franzblau, S. G., Fischer, N. H. (1998)
4
Planta Med. 64, 2±4.
Lu, T., Cantrell, C. L., Robbs, S. L., Franzblau, S. G., Fischer, N. H.
5
(1998) Planta Med. 64, 665±667.
Cantrell, C. L., Lu, T., Fronczek, F. R., Fischer, N. H., Adams, L. B.,
6
Franzblau, S. G. (1996) J. Nat. Prod. 59, 1131±1136.
Moerman, D. E. (1986) Medicinal Plants of Native America, Vol. II,
7
p. 642, Ann Arbor, Michigan, USA.
Coll, J. C., Bowden, B. F. (1986) J. Nat. Prod. 49, 934±936.
Vasquez, M., Quijano, L., Fronczek, F. R., Macias, F. A., Urbatsch, L.
8
9
E., Cox, P. B., Fischer, N. H. (1990) Phytochemistry 29, 561±565.
Kashman, Y., Lavie, D., Glotter, E. (1967) Israel J. Chem. 5, 23±27.
Bohlmann, F., Mahanta, P. K., Jakupovic, J., Rastogi, R. C., Natu, A.
10
11
A. (1978) Phytochemistry 17, 1165±1172.
Collins, L. A., Franzblau, S. G. (1997) Antimicrob. Agents Chemo-
12
ther. 41, 1004±1009.
Jakupovic, J., Jaensch, M., Bohlmann, F., Dillon, M. O. (1988)
13
Phytochemistry 27, 3551±3556.
Renoud-Grappin, M., Vanucci, C., Lhommet, G. (1994) J. Org.
14
Chem. 59, 3902±3905.
Vasquez, M., Quijano, L., Urbatsch, L. E., Fischer, N. H. (1992)
15
Phytochemistry 31, 2051±2054.
Quijano, L., Gomez-Garibay, F., Trejo-B., R. I., Rios, T. (1991)
16
Phytochemistry 30, 3293±3295.
Altomare, A., Cascarano, G., Giacovazzo, C., Guagliardi, A., Burla,
17
M. C., Polidori, G., Camalli, M. (1994) J. Appl. Cryst. 27, 435±436.
Fair, C. K. (1990) MolEN. An Interactive System for Crystal
18
Structure Analysis; Enraf-Nonius, Delft, The Netherlands.
Milman, A. (1990) Khim. Prir. Soedin. 3, 307±320.
Marshal, J. A., Cohen, N. (1964) J. Org. Chem. 29, 3727±3729.
Okunade, A. L., Wiemer, D. F. (1985) Phytochemistry 24, 1199±
19
20
21
1201.
22
Chen, C., Yang, L., Lee, T. T., Shen, Y., Zhang, D., Pan, D., McPhail, A.
T., Liu, S., Li, D., Cheng, Y., Lee, K. (1994) Bioorg. and Med. Chem. 2,
137±145.
Fischer, N. H., Lu, T. S., Cantrell, C. L., Castaæeda-Acosta, J., Quijano,
23
L., Franzblau, S. G. (1998) Phytochemistry 49, 559±564.
Connell, N. D., Nikaido, H. (1994) in: Tuberculosis, Pathogenesis,
24
Protection and Control, (Bloom, B. R., ed.), pp. 333, ASM Press,
Washington DC.