Structural analysis of 3-α-acetyl-20(29)-lupene-24-oic acid, a novel pentacyclic triterpene isolated from the gum resin of Boswellia serrata, by NMR spectroscopy

Article abstract of DOI:10.1002/mrc.1212

3α-Acetyl-20(29)-lupene-24-oic acid (1) was isolated from the gum resin of Boswellia serrata. Its presence evidently suggests, that the oxidosqualene triterpene pathway of Boswellia serrata closely resembles the biosynthetic route already found in other plants. Complete ¹H and ¹³C spectral assignments were derived from 1D and 2D NMR spectra. This is the first compound with the lupene backbone combining a 3α-hydroxy or 3α-acetyl group with the 24-carboxyl group, a configuration which is typical of the classical boswellic acids. Copyright

Full text of DOI:10.1002/mrc.1212

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2003; 41: 629–632
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1212
Spectral Assignments and Reference Data
29
Structural analysis of
3-a-acetyl-20(29)-lupene-24-oic acid,
a novel pentacyclic triterpene isolated
from the gum resin of Boswellia serrata,
by NMR spectroscopy
H
30
20
21
19
H
12
26
22
18
11
Klaus Belsner,1 Berthold Bu¨ chele,1 Udo Werz2
and Thomas Simmet1∗
13
17
25
H
28
1
14
16
9
1
Department of Pharmacology of Natural Products and Clinical Phar-
macology, University of Ulm, D-89081 Ulm, Germany
Department of Organic Chemistry, University of Ulm, D-89081 Ulm,
2
8
15
10
O
2
27
7
5
Germany
4
3
O
H3C
6
Received 16 January 2003; revised 10 April 2003; accepted 14 April 2003
24
23
O
3a-Acetyl-20(29)-lupene-24-oic acid (1) was isolated from
the gum resin of Boswellia serrata. Its presence evidently
suggests, that the oxidosqualene triterpene pathway of
Boswellia serrata closely resembles the biosynthetic
route already found in other plants. Complete 1H and
13C spectral assignments were derived from 1D and 2D
NMR spectra. This is the first compound with the lupene
backbone combining a 3a-hydroxy or 3a-acetyl group
with the 24-carboxyl group, a configuration which is
typical of the classical boswellic acids. Copyright  2003
John Wiley & Sons, Ltd.
OH
Figure 1. Structure and numbering scheme of 3˛-acetyl-20(29)-lupene
-24-oic acid (1).
3˛-Acetyl-20(29)-lupene-24-oic acid (1): white powder upon
°
precipitation with water from methanol; m.p. 234 C, with decompo-
°
sition starting at 224 C. UV, ꢀmax 202 nm, shoulder at 225 nm; EIMS,
m/z 499(26), 498(82) (MC calc. for C32H50O4: 498.75), 484(5), 483(18)
(MC ꢀ CH3), 438(11) (MC ꢀ AcOH), 423(7) (438C ꢀ CH3), 395(12),
394(31) (438C ꢀ CO2), 379(7) (394C ꢀ CH3), 327(10) (379C ꢀ C3H6),
283(7), 281(5), 280(15) (MC ꢀ cleavage of 9–11 and 8–14),11 279(6),
271(5), 259(7), 257(8), 255(11), 247(5), 246(13), 234(5), 233(7),
232(9), 231(15), 230(8), 229(21), 221(21), 220(46), 219(50), 218(67)
(MC ꢀ cleavage of 9–11 and 8–14),11 217(17), 216(9), 215(12), 213(6),
206(14), 205(35), 204(35) (218C ꢀ CH2),11 189(54) (204C ꢀ CH3),11
55(59) (CH3COC); IR, 3054.76 (w, alkene stretch), 1735.23 and 1695.72
(s, C O stretch), 1456.42 (m, CH3 and CH2 stretch), 1379.63 (m,
dimethyl), 1265.19 (ester C–O stretching).
KEYWORDS: NMR; 1H NMR; 13C NMR; 2D NMR; triterpene;
3-˛-acetyl-20(29)-lupene-24-oic acid; structural assignment
INTRODUCTION
Natural compounds from the family of pentacyclic triterpenes may
be pharmacotherapeutically important. Pentacyclic triterpenic acids
isolated from extracts of the gum resin of Boswellia serrata have
shown promising antiinflammatory effects1–3 and also antitumor4–6
and antiviral activity.7 We have recently reported structural data on
acetylated and deacetylated ˛- and ˇ-boswellic acids, including the 3-
˛-acetyl-9,11-dehydro derivatives of ˛- and ˇ-boswellic acid.8 During
the isolation of these boswellic acids, we found a novel compound,
which we have now identified as 3˛-acetyl-20(29)-lupene-24-oic acid
(1), with the molecular structure shown in Fig. 1.
3˛-Hydroxy-20(29)-lupene-24-oic acid (2) [obtained by alkaline
saponification of (1)]: EIMS, m/z 457(30), 456 (MC) (100), 442(9),
441(31) (MC ꢀ CH3), 413(9) (457C ꢀ CO2), 410(9), 394(6) (442C ꢀ
CO2), 239(15), 238(87) (MC ꢀ cleavage of 9–11 and 8–14),11 219(45),
218(72) (MC ꢀ cleavage of 9–11 and 8–14),11 204(44) (218C ꢀ CH2),11
189(56) (204C ꢀ CH3).11
NMR spectra
EXPERIMENTAL
Standard conditions were used for the acquisition of the 1D 1H
and 13C NMR spectra. 1H and 13C NMR spectra were measured on
a Bruker AMX 500 spectrometer operating at 500 and 125 MHz
Samples
Frankincense gum resin of Boswellia serrata (Klenk, Schwebheim,
Germany) was extracted using the principal methods of Winterstein
and Stein.9 Briefly, pulverized boswellic gum resin was extracted
with diethyl ether in a Soxhlet apparatus and the extract was pre-
cipitated with Ba(OH)2. The collected solids were acetylated with
boiling acetic anhydride and after removal of barium acetate, the
mixed anhydrides were cleaved with boiling methanol to yield the
3-acetyl derivatives of the triterpenoic acids. These samples were
purified by reversed-phase high-performance liquid chromatogra-
phy (HPLC) to homogeneity, i.e. ½99% purity as estimated by HPLC
and high performance thin-layer chromatography (HPTLC).10 UV
spectra were recorded during purification with a UVD 340S HPLC
photodiode-array detector (Dionex, Idstein, Germany). Mass spec-
tra were recorded with an SSQ 7000 mass spectrometer (Thermo
Finnigan, San Jose, CA, USA) in the electron ionization (EI) mode,
ionization energy 70 eV, source current 400 µA, temperature pro-
at 300 K. About 15 mg of
1 were dissolved in CDCl3 with
TMS as internal standard and transferred to a 5 mm, 400 MHz
tube (Kontes, Vineland, USA). DQF-COSY was recorded on a
Bruker Avance 400 MHz spectrometer with an FID resolution of
1.724/0.4310 Hz (F1/F2), matrix size 512 ð 2048 (F1 ð F2). Other
2D spectra were recorded with 1H–1HCOSY45, HMBC, HMQC,
HSQC and HMQC-TOCSY standard pulse sequences. Conditions:
1H–1HCOSY45 FID resolution 3.409/0.562 Hz (F1/F2), matrix size
338 ð 2048 (F1 ð F2); HMBC FID resolution 20.35/0.735 Hz (F1/F2),
matrix size 512 ð 4096 (F1 ð F2); HMQC-TOCSY spin lock period
15 ms, 18.668/1.419 Hz (F1/F2), matrix size 512 ð 2048 (F1 ð F2);
ROESY spin lock period 300 ms, FID resolution 6.181/1.545 Hz
(F1/F2), matrix size 512 ð 2048 (F1 ð F2).
gramme 30–230 C at 25 C minꢀ1, acceleration voltage 1.7 kV.
°
°
Molecular modeling
A structure was modeled employing the InsightII/Discover software
(Accelrys, San Diego, CA, USA). The structure sketched according
to the NMR restrictions was optimized with the discover-force-field
(CVFF91, conjugated gradient) and then run through a molecular
dynamics calculation from 0 to 450 K in 10 fs and kept there for
ŁCorrespondence to: Thomas Simmet, Department of Pharmacology of
Natural Products and Clinical Pharmacology, University of Ulm, D-89081
Ulm, Germany. E-mail: thomas.simmet@medizin.uni-ulm.de
Copyright  2003 John Wiley & Sons, Ltd.

630
Spectral Assignments and Reference Data
Table 1. 1H and 13C NMR assignments (υ in ppm, J in Hz) for 3˛-acetyl-20(29)-lupene-24-oic acid (1)a
HMBC,
HMBC,
13C
1H
J D 2
J D 3
HMQC-TOCSY
ROESY
COSY
10
20
1
170,4
21.36
34.41
20
2.07
1.07 (d, t, J2˛ 14.7, J2ˇ 5.7)
25
2˛, 2ˇ, 3
1˛, 1ˇ, 3
2˛, 2ˇ
1ˇ, 2˛, 2ε, 25
1˛, 2˛, 2ˇ
1˛, 1ˇ, 2ˇ, 3ˇ
1˛, 1ˇ, 2˛
2˛, 2ˇ
1.48
2
23.72
2.10
1.58
3
4
5
6
73.48
46,70
50.48
19.53
5.25 s
23
1
23
1.37 m
1.84
6˛, 6ˇ, 7b
6ˇ, 7
6˛ 7
1.62
7
8
9
34.08
40,81
49.68
37.66
21.08
¾1.4b
26
27
6˛, 6ˇ
11˛, 11ˇ, 12˛, 12ˇ, 13
9, 12˛, 12ˇ, 13
15ˇ
6˛, 6ˇ
26
25
1.31 (d, d J11˛ 10.3, J11ˇ 4.5)
25, 26
27
11˛, 11ˇ
10
11
12
1.42 (d, t, J11ˇ 12.2)
1.21 (d, d, d, d)
11ˇ, 12˛, 12ˇ, 13˛
11˛, 12˛, 12ˇ
12ˇ, 11ˇ
25.19
1.06 (d, t, J12ˇ 12.6)
1.68 (d, d, J11˛ 8.0, J11ˇ 6.2)
1.64
9, 11˛, 11ˇ, 13, 18
11˛, 11ˇ, 12˛, 12ˇ, 18, 19
16˛, 16ˇ
27, 29(z) (w), 30
29(z)
12˛, 11˛
13
14
15
38.02
42.92
27.43
18, 19
27
26
27
27, 15˛
1.66
1.00 (J16˛ 4.5)
1.36
15ˇ, 16˛, 16ˇ
15˛, 16˛, 16ˇ
16ˇ, 15˛, 15ˇ
16˛, 15˛, 15ˇ
7ˇ
16
35.56
28
15˛, 15ˇ
1.46
17
18
19
20
21
43.03
48.25
47.95
150.92
29.85
22˛,28
19
1.36 (d, t, J19˛ 11.1)
28
12˛, 12ˇ, 19, 21˛, 21ˇ
18, 21˛, 21e, 22˛, 22ˇ
29z
19
2.36 (d, d, d, J21ˇ 5.8, J21˛ 11.1)
18
29e, 29z, 30
28, 29(z)
18, 21˛, 21ˇ
30,19
19
1.90 m (J21˛ 12)
19, 21˛, 21ˇ
19, 21˛, 21ˇ
28
30
21ˇ, 20, 22,˛, 22ˇ
21˛, 20, 22˛, 22ˇ
22ˇ, 21˛, 21ˇ
1.31 m (J22ˇ 5.0)
22
39.98
1.37
1.17
1.18
28
22˛, 21˛, 21ˇ, 28
23
24
25
26
27
28
29
23.67
181.26
13.41
15.98
14.64
18.01
109.33
c
0.76
1.04
1˛
0.97
29(e) (w)
19, 29(e) (w)
30 (s)
0.78 md
4, 55(e) me
4, 67(z) m
1.65 m
22˛, 22ˇ
22ˇ
29(z), 30
29(e), 30
29(e), 29(z)
30 (w)
30
19.30
29(e), 29(z)
19, 21ˇ
a C-8 and C14 are only very weakly distinguishable by the 15˛-correlation of C-14, which is supported by comparison with 20(29)-lupeol-3ˇ-
acetate14 and 20(29)-lupenone.18 Coupling constants in Hz were obtained from DQF-COSY 400 MHz spectra.
b No separation between ˛- and ˇ-peaks.
c No cross-peak derived from the carboxyl carbon.
d Very weakly resolved multiplet.
e Weakly resolved multiplet both for H-29(e) and (z).
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 629–632

631
Spectral Assignments and Reference Data
100 fs. In addition to the molecular mechanics calculation, the
structure was finally minimized with MOPAC/AM1,12,13 BFGS-
optimizer (Broyden–Fletcher–Goldfarb–Shanno pseudo–Newton
optimization algorithm), maximum gradient 0.1. The generated
PDB files were visualized and printed with Swiss-PdbViewer.14
Distances were calculated from the model. In the case of swivelling
bonds, dihedrals were modified to obtain minimal distances without
application of an additional restrained optimization, because the
results were only used to illustrate possible ROEs.
RESULTS AND DISCUSSION
Individual carbons were assigned on the basis of the 13C spectrum
and with the J-resolved 13C spectrum. The two- and three- bond
13C–1H cross peaks of the methyl 1H singlets served as the starting
point.15 The 3˛-hydroxy orientation appears to correspond to that of
the known boswellic acids.8,16 The pronounced H-1˛/H-25 and H-
9/C-25 together with the large J9 –11˛ D 10.3 Hz coupling indicates
a relaxed chair–chair conformation of both rings A and B. The
carboxyl C-24 did not show any cross peaks, but was identified
at υ D 181.26 ppm after comparison with our recently published
results.8
Figure 3. Spatial model in ball and stick representation calculated
according to the NMR restraints using the semi-empirical MOPAC/AM1
method. The isopropenyl substituent is directed to the background. All
rings are arranged in a nearly perfect plane by trans-connections, in
contrast to the typical boswellic acids, where rings C and D are
cis connected.
Starting at H-9, the spin system spanning to H-22 via H-11,
H-12, H-13, H-18, H-19, H-21 was identified by HMQC-TOCSY.
The conformation of the cyclopentyl neighborhood was deduced
from the H-19 DQF-COSY cross peaks with H-13, H-18 and H-
21. The principal ddd pattern of H-19 is degenerated to a dt with
J21˛ ³ J18 ³ 11.1 Hz, intensity distribution 1 : 1 : 2 : 2 : 1 : 1. The large
coupling constants afford a planar orientation of these hydrogens,
which is only possible with a transaxial conformation of H-13 and
H-18. Left alone with DQF-COSY the dihedral angle between H-18
˚
were slightly below 2 A, while the minimal distance of H-27/H-29(e)
˚
was 2.04 A. ROE signals corresponding to these distance calculations
were found as listed in Table 1.
CONCLUSION
The discovery of 3˛-acetyl-20(29)-lupene-24-oic acid (1) adds a new
member to the range of pentacyclic triterpenic acids found in the
gum resin of Boswellia serrata. The relative content of 1 in boswellic
resin is low, probably because it is a byproduct in the biosynthetic
formation of boswellic acids. The respective enzymatic pathways
from oxidosqualene to similar pentacyclic triterpenes involve an
intermediate lupene cation,17 which rearranges to the final amyrin,
ursol or lupenyl structure preserving the original hydroxyl group
orientation. In all known plant triterpenes, except of those isolated
from the gum resin of Boswellia species, ˇ-orientation of the hydroxy
group has been found. The combination of the ˛-hydroxy and the
24-carboxyl group of this acid and the ˛- and ˇ-amyrin backbone
of the usual lead triterpenes of Boswellia serrata may simply indicate
low selectivity of the enzymes catalyzing the oxygenation of C-24 for
alterations at ring E. This could be verified if the respective 3˛-lupeol
can be found in the neutral fraction of the resin gum extract.
°
°
and H-19 could be around 180 or 0 . The ROEs illustrated in Fig. 2
resolved this question as part of the ROE-led reconstruction above
and below the ring plane.
Figure 3 shows a picture of the optimized structure. Calculated
dihedral angles from this model structure were H-12˛/H-13 D
°
°
°
°
172.48 , H-13/H-18 D 171.71 , H-18/H-19 D 163.57 , H-19/H-21˛ D
°
149.06 , H-19/H-21ˇ D 14.72 , respectively, which is in accordance
with the observed coupling constants. Clearly, the last two dihedrals
between the cyclopentyl hydrogens would explain a large coupling
constant; however only one has been found. This may be caused
by the conformational uncertainty of cyclopentyl rings, that was not
resolved with these settings of the BFGS-optimizer. The calculated
˚
minimal distance from H-29(e) to H-12˛ is 0.96 A, which explains the
relatively strong ROE. Distances between H-25/H-26, H-27/H-12˛
H
REFERENCES
H
1. Ammon H, Mack T, Singh G, Safayhi H. Planta Med. 1991; 57:
203.
H
2. Sander O, Herborn G, Rau R. Z. Rheumatol. 1998; 57: 11.
3. Gerhardt H, Seifert F, Buvari P, Vogelsang H, Repges R. Z.
Gastroenterol. 2001; 39: 11.
H
H
4. Hoernlein RF, Orlikowsky Th, Zehrer C, Niethammer D,
Sailer ER, Simmet Th, Dannecker GE, Ammon HPT. J. Pharmacol.
Exp. Ther. 1999; 288: 613.
H
5. Heldt MR, Syrovets T, Winking M, Sailer ER, Safayhi H,
Ammon HPT, Simmet Th. Naunyn-Schmiedeberg’s Arch. Pharma-
col. 1997; 355(Suppl.): 30.
O
6. Belsner K, Bu¨chele B, Syrovets T, Simmet Th. Naunyn-
Schmiedeberg’s Arch. Pharmacol. 2001; 363(Suppl.): 44.
7. Hussein G, Miyashiro H, Nakamura N, Hattori M, Kakiuchi N,
Shimotohno K. Phytother. Res. 2000; 14: 510.
8. Belsner K, Bu¨chele B, Werz U, Syrovets T, Simmet Th. Magn.
Reson. Chem. 2003; 41: 115.
O
H3C
H
O
OH
Figure 2. Illustration of the ROEs observed with
9. Winterstein A, Stein G. Hoppe Seyler’s Z. Physiol. Chem. 1932; 208:
9.
10. Bu¨chele B, Zugmaier W, Simmet Th. J. Chromatogr. B 2003; 791:
21.
3˛-acetyl-20(29)-lupene-24-oic acid (1), which were important to
resolve the conformation. Additional interactions between H-12ˇ/H-30
and H-29(z)/H-18 are also important, but were excluded for clarity.
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 629–632

632
Spectral Assignments and Reference Data
11. Shiojima K, Suzuki H, Kodera N, Kubota K, Tsushima S,
Ageta H, Chang HC, Chen YP. Chem. Pharm. Bull. 1994; 42: 2193.
12. Dewar MJS, Zoebisch EG, Healy EF, Stewart JJP. J. Am. Chem.
Soc. 1985; 107: 3902.
15. Reynolds WF, McLean S, Burke SJ, Jacobs H. Magn. Reson. Chem.
2001; 39: 757.
16. Schweizer S, Eichele K, Ammon HPT, Safayhi H. Planta Med.
2000; 66: 781.
13. Stewart JJP. J Comput.-Aided Mol. Des. 1990; 4: 1.
14. Guex N, Diemand A, Peitsch MC. Trends Biochem. Sci. 1999; 24:
364.
17. Poralla K. In Isoprenoids including Carotenoids and Steroids, vol. 2,
Cane DE (ed). Elsevier: Oxford, 1999; 300.
18. Kim E-M, Jung H-R, Min T-J. Bull. Korean Chem. Soc. 2001; 22: 59.
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 629–632